Studies in Sound Symbolism
Studies in Sound Symbolism
Studies in Sound Symbolism
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
GOTHENBURG MONOGRAPHS IN LINGUISTICS 17<br />
<strong>Studies</strong> <strong>in</strong><br />
<strong>Sound</strong> <strong>Symbolism</strong><br />
Åsa Abel<strong>in</strong><br />
Doctoral Dissertation<br />
publicly defended <strong>in</strong> Stora Hörsalen,<br />
Humanisten, Göteborg University,<br />
on May 14, 1999, at 10.00<br />
for the degree of Doctor of Philosophy<br />
Department of L<strong>in</strong>guistics, Göteborg University, Sweden, 1999
ABSTRACT<br />
This thesis <strong>in</strong>vestigates how the Swedish lexicon is structured with<br />
respect to sound symbolism, the productivity of phonesthemes and cross<br />
language similarities <strong>in</strong> certa<strong>in</strong> areas of sound symbolism.<br />
The Swedish lexicon has been analyzed with emphasis on the sound<br />
symbolic properties of <strong>in</strong>itial and f<strong>in</strong>al consonant clusters, and to a<br />
certa<strong>in</strong> extent of vowels. Approximately 1, 000 lexemes were judged to<br />
be sound symbolic and the outcome of the analysis are tentative<br />
phonesthemes, i.e. motivated connections between mean<strong>in</strong>gs and<br />
consonant clusters.<br />
Almost all Swedish <strong>in</strong>itial consonant clusters and many of the f<strong>in</strong>al<br />
consonant clusters can carry sound symbolic mean<strong>in</strong>gs. Lexically<br />
<strong>in</strong>frequent clusters are utilized to a larger extent than lexically frequent<br />
clusters. No two consonant clusters have exactly the same semantic<br />
profile. Phonesthemes have different sound symbolic strength, i.e. some<br />
are clearly sound symbolic (i.e. a high percentage of the words<br />
beg<strong>in</strong>n<strong>in</strong>g with a certa<strong>in</strong> cluster are sound symbolic), and carry either<br />
one mean<strong>in</strong>g or several mean<strong>in</strong>gs. Other (candidates for) phonesthemes<br />
are weaker and not so clearly sound symbolic.<br />
The mean<strong>in</strong>gs of most phonesthemes are relatable to the senses:<br />
hear<strong>in</strong>g, vision or tactile sensation, or they are metaphorically or<br />
metonymically connected to the senses. The most common semantic<br />
features occurr<strong>in</strong>g are often relatable to synaesthesia.<br />
The productivity of phonesthemes was tested <strong>in</strong> experiments of<br />
production and understand<strong>in</strong>g. The experiments show, for<br />
<strong>in</strong>terpretation, that no constructed word is <strong>in</strong>terpreted as expected by<br />
all subjects, but that all of the constructed words are <strong>in</strong>terpreted<br />
correctly by some subjects. The most common semantic features found<br />
<strong>in</strong> the lexical analysis are often also the most successfully <strong>in</strong>terpreted by<br />
subjects. For production, the experiments <strong>in</strong>dicate that subjects tend to<br />
encode the semantic features <strong>in</strong> <strong>in</strong>itial clusters rather than <strong>in</strong> f<strong>in</strong>al<br />
clusters. F<strong>in</strong>al consonant clusters seem to be of less importance than the<br />
<strong>in</strong>itial clusters <strong>in</strong> new sound symbolic words <strong>in</strong> Swedish.<br />
For the contrastive studies, the general results are that there are both<br />
similarities and dissimilarities between the expressions <strong>in</strong> the different<br />
languages. The variation is greater for some semantic fields than for<br />
others.<br />
KEY WORDS: sound symbolism, lexical structure, synaesthesia,<br />
productivity, universals
Acknowledgements<br />
The first idea to write this thesis emerged when I was work<strong>in</strong>g <strong>in</strong> the<br />
Lexical Database and Svensk Ordbok project at Språkdata <strong>in</strong> the 1980's,<br />
where I alphabetically plowed through large parts of the Swedish<br />
vocabulary.<br />
My greatest thanks goes to my supervisor Jens Allwood, for always<br />
support<strong>in</strong>g me and for be<strong>in</strong>g so consistent <strong>in</strong> always misunderstand<strong>in</strong>g<br />
what is not perfectly clear. I am also grateful to Elisabeth Ahlsén who has<br />
given me many valuable comments and especially encouraged my<br />
experiment<strong>in</strong>g.<br />
I also want to thank all my colleagues and former colleagues at the<br />
l<strong>in</strong>guistics department, especially Sally Boyd, Beatrice Dorriots, Johan<br />
Hagman, Jerker Järborg, Per L<strong>in</strong>dblad, Lars Malmsten, Kerst<strong>in</strong> Nelfelt,<br />
Shirley Nicholson, Joakim Nivre, Sören Sjöström, Sven Strömqvist,<br />
Nicole Takolander, Hans Vappula, Åsa Wengel<strong>in</strong>, Ulla Veres, and all<br />
others who have helped me <strong>in</strong> various ways over the years. A special<br />
thanks goes to Cather<strong>in</strong>e Paterson who did the cover design.<br />
I also want to express my gratitude to my <strong>in</strong>formants from many parts of<br />
the world.<br />
The persons who I want to thank especially are my children Tove and<br />
Ellen who are for sure very tired of the word "avhandl<strong>in</strong>g". They have<br />
shown great patience with my <strong>in</strong>termittent absentm<strong>in</strong>dedness, but they<br />
have also shown great <strong>in</strong>terest <strong>in</strong> the subject of this thesis and given me<br />
<strong>in</strong>terest<strong>in</strong>g comments from their po<strong>in</strong>t of view. F<strong>in</strong>ally I want to thank my<br />
family, my mother and father, my sisters Ulla and Susanne and all my<br />
other friends.<br />
Göteborg <strong>in</strong> April, 2000<br />
Åsa Abel<strong>in</strong>
Contents<br />
Chapter 0 Introduction 1<br />
Chapter 1 Background 3<br />
Chapter 2 Theoretical framework 49<br />
Chapter 3 Method 71<br />
Chapter 4 Analysis of <strong>in</strong>itial consonant clusters 77<br />
Chapter 5 Analysis of f<strong>in</strong>al consonant clusters, vowels and<br />
comb<strong>in</strong>ations. 143<br />
Chapter 6 Some contrastive studies <strong>in</strong> sound symbolism 189<br />
Chapter 7 Experiments with words constructed from phonesthemes 219<br />
Chapter 8 Summary and discussion 245<br />
Bibliography 273<br />
Appendix 1: <strong>Sound</strong> symbolic roots of <strong>in</strong>itial clusters<br />
Appendix 2: Interjections of cross language study<br />
Appendix 3: Test sheets of chapter 7
<strong>Studies</strong> <strong>in</strong> <strong>Sound</strong> symbolism<br />
0 Introduction 1<br />
1 Background 3<br />
1.1 Purpose of the chapter 3<br />
1.2 Term<strong>in</strong>ology 3<br />
1.2.1 Onomatopoeia 3<br />
1.2.2 <strong>Sound</strong> symbolism 3<br />
1.2.3 Phonestheme 4<br />
1.2.4 Ideophone 4<br />
1.2.5 Morpheme 5<br />
1.2.6 Conclusion 8<br />
1.3 Is sound symbolism the rule or the exception <strong>in</strong><br />
language? 9<br />
1.3.1 <strong>Sound</strong> symbolism is an exception 9<br />
1.3.2 <strong>Symbolism</strong> is fundamental to language 9<br />
1.3.3 <strong>Sound</strong> symbolism is both <strong>in</strong>side and outside of<br />
language 11<br />
1.3.4 Evaluation of discussion of sound symbolism 12<br />
1.4 Is sound symbolism productive or not? 12<br />
1.4.1 <strong>Sound</strong> symbolism is not productive 12<br />
1.4.2 <strong>Sound</strong> symbolism is productive 13<br />
1.4.3 Greater or lesser degree of productivity 13<br />
1.4.4 Evaluation of the discussion of productivity 13<br />
1.5 The question of etymology 14<br />
1.6 The phylogenesis of language 18<br />
1.7 Universality versus language specificity 20<br />
1.7.1 <strong>Sound</strong> symbolism is universal 20<br />
1.7.2 <strong>Symbolism</strong> is not universal 21<br />
1.7.3 Evaluation of the discussion of universality<br />
versus language specificity <strong>in</strong> sound symbolism 22<br />
1.8 Context 23<br />
1.9 A framework for models of sound symbolism 26<br />
1.10 Expression and content of sound symbolism 28<br />
1.10.1 Expression 29
1.10.2 Content 29<br />
1.10.3 Expression and content <strong>in</strong> different analyses 29<br />
1.10.4 Results - data from different authors 31<br />
1.10.5 Experimental results 37<br />
1.11 Possible explanations of sound symbolism 41<br />
1.11.1 Miscellaneous explanations and proprioception 41<br />
1.11.2 Synaesthesia 42<br />
1.11.3 Other neurological and biological explanations 44<br />
1.11.4 Non- biological explanations 46<br />
2 Theoretical framework 49<br />
2.1 General considerations 49<br />
2.2 Static-dynamic, conventionality and<br />
arbitrar<strong>in</strong>ess 49<br />
2.3 Semantic analysis 51<br />
2.3.1 Conceptions of mean<strong>in</strong>g 51<br />
2.3.2 Semantic features and semantic fields 52<br />
2.4 Basic relations between expression and content 54<br />
2.5 The nature of phonesthemes 57<br />
2.6 Considerations for a model 59<br />
2.6.1 Relations between the categories 63<br />
2.6.2 An explanatory model for sound symbolism 66<br />
3 Method 71<br />
3.1 Stage 1: Collection of lexical material 71<br />
3.2 Stage 2: Cross-language comparisons 74<br />
3.2.1 Cross-language thesaurus studies 74<br />
3.2.2 Cross-language <strong>in</strong>formant studies 74<br />
3.3 Stage 3: Experiments 76<br />
3.3.1 Experiments with neologisms 76<br />
3.4 Further method 76<br />
4 Analysis of the <strong>in</strong>itial consonant clusters 77<br />
4.1 Data analysis, a short overview 77<br />
4.2 Results 80<br />
4.2.1 More and less sound symbolic clusters 80<br />
4.2.2 Proportions of motivated root morphemes. Summary 86<br />
4.2.3 Types of mean<strong>in</strong>g 88
4.3 Frequent semantic features 93<br />
4.3.1 Pejorative 93<br />
4.3.2 <strong>Sound</strong> 100<br />
4.3.3 Long th<strong>in</strong> form 105<br />
4.3.4 Quick or strong movement 107<br />
4.3.5 Wetness 111<br />
4.3.6 The most sound symbolic clusters 115<br />
4.4 Frequent clusters 116<br />
4.4.1 The cluster sl- 118<br />
4.4.2 The cluster sn- 119<br />
4.4.3 The cluster kn- 121<br />
4.4.4 The cluster kr- 124<br />
4.4.5 The cluster fn- 126<br />
4.4.6 The cluster kn- 127<br />
4.4.7 The cluster gn- 127<br />
4.4.8 The cluster spr- 129<br />
4.4.9 The cluster pj- 130<br />
4.5 Typical or unique mean<strong>in</strong>gs 131<br />
4.6 Patterns of semantic features 134<br />
4.7 Discussion and conclusions 140<br />
5 Analysis of f<strong>in</strong>al consonant clusters, vowels and<br />
comb<strong>in</strong>ations 143<br />
5.1 F<strong>in</strong>al clusters 144<br />
5.2 Summary of the analysis of semantic features for<br />
f<strong>in</strong>al clusters 150<br />
5.3 Properties of consonant clusters of Nusvensk Frekvensordbok 152<br />
5.3.1 Summary of the analysis of properties of different<br />
f<strong>in</strong>al clusters 164<br />
5.4 Vowels 164<br />
5.4.1 Vowel pairs and triplets 164<br />
5.4.2 Vowels <strong>in</strong> light/gaze-words 166<br />
5.4.3 The vowel [P] 167<br />
5.4.4 Summary of vowels 167<br />
5.5 Comparisons of f<strong>in</strong>al clusters of roots from different sources<br />
167<br />
5.5.1 NFO4 and Sigurd (1965) 167
5.5.2 The most frequent f<strong>in</strong>al clusters <strong>in</strong> NFO4 171<br />
5.5.3 Discussion of f<strong>in</strong>al clusters <strong>in</strong> Svensk<br />
Baklängesordbok and of Nusvensk Frekvensordbok 172<br />
5.6 Comb<strong>in</strong>ations of <strong>in</strong>itial and f<strong>in</strong>al consonant clusters 174<br />
5.6.1 Initial cluster + no f<strong>in</strong>al cluster 175<br />
5.6.2 No <strong>in</strong>itial cluster + f<strong>in</strong>al cluster 179<br />
5.6.3 Initial cluster + f<strong>in</strong>al cluster 182<br />
5.6.4 Summary of comb<strong>in</strong>ations 185<br />
5.7 Summary and discussion of <strong>in</strong>itial and f<strong>in</strong>al clusters,<br />
and vowels 186<br />
6 Some contrastive studies <strong>in</strong> sound symbolism 189<br />
6.1 Introduction 189<br />
6.2 The Thesaurus study 189<br />
6.2.1 Method 189<br />
6.2.2 Results 190<br />
6.2.2.1Words for 'stupidity' <strong>in</strong> English 190<br />
6.2.2.2 Words for 'stupidity' <strong>in</strong> Swedish 190<br />
6.2.2.3 Words for 'surface structure' <strong>in</strong> English 191<br />
6.2.2.4 Words for 'surface structure' <strong>in</strong> Swedish 192<br />
6.2.3 Conclusions of the Thesaurus study 193<br />
6.3 Some <strong>in</strong>terjections <strong>in</strong> different languages 194<br />
6.3.1 Swedish expressive <strong>in</strong>terjections 195<br />
6.3.2 Discussion of Swedish expressive <strong>in</strong>terjections,<br />
commands and greet<strong>in</strong>gs 196<br />
6.3.3 Phonological and phonetic similarities and<br />
dissimilarities between <strong>in</strong>terjections of different languages 197<br />
6.4 Imitations of animal calls 202<br />
6.4.1 Expressions for animal calls <strong>in</strong> Swedish and other<br />
languages 202<br />
6.4.2 A test of expressions for animal calls of different<br />
languages 205<br />
6.4.3 Results from a test of expressions for animal calls<br />
<strong>in</strong> different languages 206<br />
6.4.4 Discussion of the test on identify<strong>in</strong>g animal calls 208<br />
6.5 Conclusions of studies of expressive <strong>in</strong>terjections<br />
and expressions of animal calls 208
6.6 Test of cross language <strong>in</strong>terpretation of Swedish<br />
onomatopoeic and other sound symbolic words 210<br />
6.6.1 Method 210<br />
6.6.2 Results of <strong>in</strong>terpretation of cross language Swedish<br />
onomatopoeic and other sound symbolic words 211<br />
6.6.3 Conclusions from the test on cross language<br />
<strong>in</strong>terpretation of Swedish onomatopoeic and other<br />
sound symbolic words 215<br />
6. 7 General conclusions and discussion of the cross<br />
language studies 217<br />
7 Experiments with words constructed from<br />
phonesthemes 219<br />
7.1 Production and understand<strong>in</strong>g 215<br />
7.1.1 Forced choice for production - from mean<strong>in</strong>g to<br />
phonological (graphic) form 223<br />
7.1.2 Forced choice for understand<strong>in</strong>g - from phonological<br />
form to mean<strong>in</strong>g 226<br />
7.1.3 Free production test from constructed words to<br />
mean<strong>in</strong>gs 229<br />
7.1.4 Free production from mean<strong>in</strong>g to constructed word 232<br />
7.1.5 Match<strong>in</strong>g test of nonsense words and mean<strong>in</strong>gs 238<br />
7.1.6 Summary of results of tests on <strong>in</strong>terpret<strong>in</strong>g mean<strong>in</strong>gs<br />
and sounds 239<br />
8 Summary and discussion 245<br />
8.1 The research questions were as follows: 245<br />
8.1.1 Question 1 246<br />
8.1.2 Question 2 252<br />
8.1.3 Question 3 255<br />
8.1.4 Question 4 257<br />
8.2 Comparison of the studies 258<br />
8.3 Possible explanations of onomatopoeia and sound symbolism 263<br />
8.3.1 Pejoratives 265<br />
8.3.2 Summary 266<br />
8.4 Predictions for sound symbolism <strong>in</strong> Swedish 268<br />
8.5 Ma<strong>in</strong> objectives and further research 271
Introduction<br />
The purpose of this thesis is to study some aspects of the wide field of<br />
onomatopoeia and other sound symbolism with special reference to<br />
Swedish. A large part of the study is devoted to a description, semantic and<br />
phonological, of onomatopoeia and other sound symbolism 1 of <strong>in</strong>itial and<br />
f<strong>in</strong>al consonant clusters, and also of vowels.<br />
One ma<strong>in</strong> <strong>in</strong>terest is the issue of productivity, which is studied ma<strong>in</strong>ly with<br />
the aid of experiments.<br />
Furthermore, the issue of universals is addressed through partial<br />
comparisons with other languages and also through tests of perception of<br />
Swedish onomatopoeia and sound symbolism.<br />
The purpose is f<strong>in</strong>ally to construct a model for sound symbolism as a<br />
central part of language, and to construct an explanatory model for the<br />
semantic aspect of sound symbolism <strong>in</strong> Swedish.<br />
The research questions of the thesis are the follow<strong>in</strong>g:<br />
1. What are the properties of sound symbolic sequences <strong>in</strong> Swedish? More<br />
specifically the questions are:<br />
Which <strong>in</strong>itial and f<strong>in</strong>al consonant clusters are used <strong>in</strong> sound<br />
symbolism?<br />
Which mean<strong>in</strong>gs are used <strong>in</strong> sound symbolism?<br />
How do these comb<strong>in</strong>e <strong>in</strong> phonesthemes?<br />
What are the sound symbolic characteristics of some vowels?<br />
How do <strong>in</strong>itial and f<strong>in</strong>al clusters and vowels comb<strong>in</strong>e <strong>in</strong> words?<br />
2. Are phonesthemes productive <strong>in</strong> Swedish? And, if so, are some<br />
phonesthemes more productive than others? Are neologisms created or<br />
<strong>in</strong>terpreted <strong>in</strong> accordance with the semantic model of chapter 2 and the<br />
analysis of chapters 4 and 5?<br />
3. Are there similarities or dissimilarities, <strong>in</strong> some specific aspects of sound<br />
symbolism, between different languages?<br />
4. Do non-Swedish speakers <strong>in</strong>terpret Swedish phonesthemes <strong>in</strong> accordance<br />
with the semantic model of chapter 2 and the analysis of chapters 4 and 5?<br />
1 cf. p. 4<br />
1
1 Background<br />
1.1 Purpose of the chapter<br />
The area of onomatopoeia and other sound symbolism has not been<br />
central <strong>in</strong> l<strong>in</strong>guistics. Nevertheless, most l<strong>in</strong>guists have had someth<strong>in</strong>g to<br />
say on the subject, <strong>in</strong> many cases us<strong>in</strong>g different term<strong>in</strong>ologies. In this<br />
chapter there will be a thematic overview of some of the most <strong>in</strong>terest<strong>in</strong>g<br />
contributions. The themes that are especially important are the questions<br />
of <strong>in</strong>tegration <strong>in</strong>to language (grammar and lexicon), of productivity, of<br />
universality and of explanation. The chapter will end with a summary of<br />
forms and mean<strong>in</strong>gs of onomatopoeic and sound symbolic expressions.<br />
1.2 Term<strong>in</strong>ology<br />
There are two problems with the term<strong>in</strong>ology <strong>in</strong> this area. First, there are<br />
a lot of terms. Second, the terms are not always used <strong>in</strong> the same way, and<br />
they are seldom def<strong>in</strong>ed. The greatest confusion, if one doesn't know the<br />
author's purpose, is perhaps that the term symbolic is often used to mean<br />
sound symbolic (i. e. words or other expressions hav<strong>in</strong>g a sound structure<br />
that is not <strong>in</strong>dependent of their mean<strong>in</strong>g), and not as a contrast to<br />
<strong>in</strong>dexical and iconic (cf. Peirce (1955) and Allwood and Andersson<br />
(1976)). In this thesis, the term symbolic is therefore not used as a<br />
synonym to sound symbolic. If the term is used, it is used <strong>in</strong> the<br />
traditional way, for conventional non-motivated signs.<br />
1.2.1 Onomatopoeia<br />
Nordberg (1986) gives a def<strong>in</strong>ition. ‘‘Onomatopoeia <strong>in</strong> a restricted sense<br />
refers to imitation of natural sounds, e.g. of animals’’. I do not believe it<br />
is necessary to restrict the term onomatopoeia to natural sounds even<br />
though, of course, this is a special k<strong>in</strong>d of onomatopoeia. I will use the<br />
term onomatopoeia for all k<strong>in</strong>ds of sound imitation.<br />
1.2.2 <strong>Sound</strong> symbolism<br />
Nordberg (1986) writes ‘‘<strong>Sound</strong> symbol or phonestheme ... is the<br />
synesthetic comb<strong>in</strong>ation of a certa<strong>in</strong> sound or sound sequence with a<br />
particular notion or a particular connotative content.’’ Jakobson and<br />
Waugh (1979) def<strong>in</strong>e sound symbolism as ‘‘an <strong>in</strong>most, natural association<br />
between sound and mean<strong>in</strong>g’’. Malkiel (1994) uses the term<br />
phonosymbolism to mean the same, he writes, as sound symbolism.<br />
3
I will use the term sound symbolism as a general term for an iconic or<br />
<strong>in</strong>dexical relationship between sound and mean<strong>in</strong>g, and also between<br />
sound and sound (which is onomatopoeia). Onomatopoeia is then a special<br />
case of sound symbolism. Onomatopoeia and other sound symbolism will<br />
also be termed motivated expressions.<br />
1.2.3 Phonestheme<br />
Phonesthemes are sometimes called word aff<strong>in</strong>ities or verbal aff<strong>in</strong>ities and<br />
are described as associated with ‘‘a marg<strong>in</strong>al set of vocables which are<br />
semantically fluid, more expressive than cognitive’’ (Jakobson and<br />
Waugh, 1979) or as ‘‘the group<strong>in</strong>g of similar mean<strong>in</strong>gs about similar<br />
sound’’ (Bol<strong>in</strong>ger, 1965). (The relation between phonesthemes and<br />
morphemes is discussed <strong>in</strong> 1.2.5). Householder's (1946) def<strong>in</strong>ition of a<br />
phonestheme is: ‘‘a phoneme or cluster of phonemes shared by a group of<br />
words which also have <strong>in</strong> common some element of mean<strong>in</strong>g or function,<br />
though the words may be etymologically unrelated’’ The def<strong>in</strong>ition above<br />
could be improved by chang<strong>in</strong>g ‘‘words’’ <strong>in</strong>to ‘‘morphemes’’ (for a<br />
discussion of the morpheme, see 1.2.5).<br />
Another term is psychomorphs (Markell and Hemp, 1960), which is used<br />
<strong>in</strong> approximately the same sense as phonesthemes. Bol<strong>in</strong>ger (1950) also<br />
uses the term submorphemic differentials or submorphs which are<br />
described as ‘collocations of phonemes common to a set of words and<br />
suggestive of a stronger or vaguer semantic <strong>in</strong>terconnection. He uses the<br />
term affective morphemes as synonymous to phonesthemes. Nordberg<br />
(1986) co<strong>in</strong>s the term sound words (onomatopoeic elements) as the jo<strong>in</strong>t<br />
designation for ideophones, sound symbols, phonesthemes and<br />
onomatopoeia s<strong>in</strong>ce he claims that the boundaries are fuzzy. Nordberg<br />
also writes that onomatopoeic elements are not such vaguely imitative<br />
words with normal phonotactics as e.g. Swedish susa 'sigh', klucka<br />
'cluck', fladdra 'flutter', etc. but purely sound-illustrative sequences,<br />
rem<strong>in</strong>iscent of the sound balloons of comic strips.<br />
1.2.4. Ideophone<br />
Childs (1994) discusses the problem of def<strong>in</strong><strong>in</strong>g ideophones and claims<br />
that often one or more of the follow<strong>in</strong>g criteria are met: ideophones often<br />
have unusual phonological characteristics, and they often display very<br />
little morphology. Syntactically they are often set apart from the rest of<br />
4
an utterance. In some languages they constitute a separate syntactic<br />
category, <strong>in</strong> others not. Semantically they often underscore the mean<strong>in</strong>g<br />
of a verb, and often, but not always, they are sound symbolic. In many<br />
cases ideophones show a close connection to gestures.<br />
This k<strong>in</strong>d of phenomenon seems to correspond best to the ‘‘sound words’’<br />
of adolescent language, described by Nordberg (1986), (e.g. krch, ppff,<br />
då<strong>in</strong>g, ‘‘har hon håret så här, så här: tsscchht ’’ (‘‘she wears her hair like<br />
this, like this: tsscchh’’, etc.)) but not to phonesthemes.<br />
The term ideophone will not be used <strong>in</strong> my own analysis of onomatopoeia<br />
and sound symbolism <strong>in</strong> Swedish, partly because different authors use the<br />
term <strong>in</strong> a non-uniform way. Even if I do not treat ideophones <strong>in</strong> my own<br />
analysis I will use the term as it is described above, <strong>in</strong> discuss<strong>in</strong>g some of<br />
the authors below.<br />
1.2.5 Morpheme<br />
What exactly is the relation between phonesthemes and morphemes? Are<br />
phonesthemes really morphemes or are they someth<strong>in</strong>g different? The<br />
most common def<strong>in</strong>ition of a morpheme is ‘‘a m<strong>in</strong>imal mean<strong>in</strong>g carry<strong>in</strong>g<br />
unit’’, but that def<strong>in</strong>ition <strong>in</strong> this context seems to apply to phonesthemes<br />
rather than to what traditionally has been labeled morphemes: The lexeme<br />
bjäfs (gewgaws) can be analyzed as one morpheme. But it can, at the same<br />
time, be analyzed as consist<strong>in</strong>g of the m<strong>in</strong>imal mean<strong>in</strong>g units bj 1-<br />
'pejorative', -E- 'pejorative, and -fs 'pejorative'. It seems that these are<br />
really the m<strong>in</strong>imal mean<strong>in</strong>g carry<strong>in</strong>g units.<br />
However, these m<strong>in</strong>imal mean<strong>in</strong>g carry<strong>in</strong>g units do not build up the whole<br />
mean<strong>in</strong>g of the word bjäfs. The mean<strong>in</strong>g of a sound symbolic word is<br />
more than the mean<strong>in</strong>g of its parts, but this also goes for many lexicalized<br />
compounds, e.g. blackbird means more than 'black bird'. In the case of<br />
other sound symbolic words, however, it seems that the mean<strong>in</strong>g of a<br />
word is almost always more than the mean<strong>in</strong>g of its parts. As a contrast,<br />
the mean<strong>in</strong>g of onomatopoeic words, like plask, is often not more than the<br />
mean<strong>in</strong>g of its parts: pl- 'wetness' and -sk 'wetness'.<br />
1Phonemes will be marked with bold type and are not put between slashes. The symbol P<br />
will be used for the short rounded close-mid vowel, e.g. the first vowel <strong>in</strong> muttra<br />
(mutter).<br />
5
Also, <strong>in</strong> a productive perspective, neologisms are sometimes created or<br />
understood out of (one or more) m<strong>in</strong>imal mean<strong>in</strong>g units (phonesthemes)<br />
and these can thus be seen as build<strong>in</strong>g blocks for word mean<strong>in</strong>g.<br />
Distributionally, phonesthemes are not whole words (lexemes) but parts<br />
of words and could thus not qualify as free morphemes. They could<br />
perhaps be described as bound morphemes, i. e. as affixes. However,<br />
there is a fact that contradicts this <strong>in</strong>terpretation, on the expression side:<br />
bound morphemes <strong>in</strong> Swedish can <strong>in</strong>clude a vowel or consist of a s<strong>in</strong>gle<br />
consonant or of a consonant cluster. The doma<strong>in</strong> of the phonestheme is<br />
often a consonant cluster. This shows that the expression side of<br />
phonesthemes is only partially different from other bound morphemes, <strong>in</strong><br />
Swedish. But there are also other more problematic facts on the<br />
expression side <strong>in</strong> traditional morphology (for a discussion see e.g.<br />
Källström (1988), e.g. suppletion (be, am are, is, was, were) and<br />
<strong>in</strong>flect<strong>in</strong>g and fusioned decl<strong>in</strong>ation (e.g. f<strong>in</strong>na, fann, funnen), so it is not<br />
clear that this would dist<strong>in</strong>guish phonesthemes from morphemes.<br />
Another problem can be illustrated by the word flämta (pant), where flis<br />
a phonestheme imitat<strong>in</strong>g movement. The rest of the word -ämta would<br />
have to be called a restmorph. Restmorphs comprise a problematic, but<br />
not unusual, category <strong>in</strong> morphology. A way to avoid too many<br />
restmorphs is to say that <strong>in</strong> sound symbolic words there are mean<strong>in</strong>g units<br />
<strong>in</strong> <strong>in</strong>dependent dimensions to morphemes, e. g. <strong>in</strong> the morphemes bjäfs<br />
and flämta; bjäfs and flämta are mean<strong>in</strong>g units on one level which at the<br />
same time conta<strong>in</strong> bj- (<strong>in</strong> bjäfs ) and fl- (<strong>in</strong> flämta), with pejorative<br />
mean<strong>in</strong>gs on another level. Examples such as these show that it is a<br />
problem to determ<strong>in</strong>e how far the morphological analysis should go (see<br />
Källström, 1988).<br />
Phonesthemes can be placed <strong>in</strong> a hierarchy between phonemes and<br />
morphemes, where morphemes and phonesthemes both are mean<strong>in</strong>g<br />
bear<strong>in</strong>g units while phonemes are not (figure 1.1).<br />
6
morphemes<br />
phonesthemes<br />
phonemes<br />
Figure 1.1. The hierarchy between phonemes and morphemes.<br />
Phonesthemes are built up from phonemes, morphemes can (partly) be<br />
built up from phonesthemes, but phonesthemes are never built up of<br />
morphemes.<br />
The morpheme is a connection between expression and mean<strong>in</strong>g. When<br />
the expression varies, e.g. dog-s [z] - cat-s [s] they are called allomorphs.<br />
A parallel dist<strong>in</strong>ction could be made between e.g. the pejoratives pj-, bj-,<br />
fj- which can be called ‘‘allophonests 2’’ to a phonestheme (with pejorative<br />
mean<strong>in</strong>g), <strong>in</strong> an item-and-arrangement analysis. (Alternatively, pj-, bjand<br />
fj- can be analyzed as labial obstruent + j.) These examples show that<br />
these types of analyses can be done with phonesthemes, as well as with<br />
regular morphemes.<br />
A certa<strong>in</strong> consonant cluster can sometimes be the expression of a<br />
phonestheme, but sometimes not, (e.g. kl- kladdig (sound symbolic),<br />
klöver (not sound symbolic). The phonestheme appears <strong>in</strong> <strong>in</strong>teraction with<br />
the word mean<strong>in</strong>g, when it fits the expression 3. That means that we have<br />
homonymy/polysemy <strong>in</strong> phonesthemes, e.g. kl- can mean 'adhesion' as <strong>in</strong><br />
kladdig (sticky), klibbig (sticky), kletig (smeary), klick (dollop), 'wetness'<br />
as <strong>in</strong> klafsa (squelch), it can be 'onomatopoeic' as <strong>in</strong> klang (clang),<br />
klappra (clatter), klatsch (slap), klicka (click), kl<strong>in</strong>g (t<strong>in</strong>kle), klirra<br />
(j<strong>in</strong>gle), klämta (toll), klucka (lap) and it can mean a 'certa<strong>in</strong> form' as <strong>in</strong><br />
klimp (lump), kloss (block), klot (ball), klubba (club), klump (lump) and<br />
it is 'pejorative' as <strong>in</strong> kludda (daub), klotter (doodle), etc, cf. chapter 4.<br />
2This term will, however, not be used further.<br />
3There can also be <strong>in</strong>teraction between different sounds <strong>in</strong> the word, cf. analysis <strong>in</strong> 5.5.<br />
7
We also have a variation between some phonesthemic mean<strong>in</strong>g and none,<br />
e.g. we have no phonesthemic mean<strong>in</strong>g <strong>in</strong> klöver (clover). Certa<strong>in</strong><br />
consonant clusters are almost always phonesthemic, e.g. pj-, while others<br />
are less phonesthemic, e.g. tr-. This means that there is a variation <strong>in</strong> the<br />
strength of the connection between content and expression <strong>in</strong><br />
phonesthemes, a quantitative dimension of the morphology.<br />
Consequently, there are arguments for and aga<strong>in</strong>st whether phonesthemes<br />
are morphemes or not. The mean<strong>in</strong>g of the concept morpheme is vague,<br />
but I suggest that phonesthemes belong to morphology. It is one type of<br />
morpheme, a type which is special on the semiotic and semantic side: the<br />
relation to what is denoted is often iconic or <strong>in</strong>dexical and what is denoted<br />
are often sounds, experiences of the senses, emotions, etc.<br />
Phonesthemes are also special <strong>in</strong> that they have a low degree of autonomy.<br />
They are bound morphemes which often can be analytically dist<strong>in</strong>guished<br />
<strong>in</strong> a fashion similar to free morphemes. They will be referred to as<br />
phonesthemes and regarded as a type of morpheme which, because of<br />
their dependent and motivated nature, perhaps could rather be called<br />
submorphemes.<br />
1.2.6 Conclusion<br />
Above, there has been a survey of different terms and usages.<br />
Subsequently, the term onomatopoeia will be used to mean all k<strong>in</strong>ds of<br />
sound imitation, phonestheme will be used to mean bound submorphemic<br />
(cf. above) str<strong>in</strong>gs (e.g. consonant clusters) which have <strong>in</strong> common a<br />
certa<strong>in</strong> element of mean<strong>in</strong>g or function. The relation between sound and<br />
mean<strong>in</strong>g is often iconic or <strong>in</strong>dexical, as well as symbolic. The term sound<br />
symbolism will be used for the general phenomenon of motivated<br />
relations between sound and mean<strong>in</strong>g, <strong>in</strong>clud<strong>in</strong>g onomatopoeia. Thus<br />
sound symbolism, <strong>in</strong> a sense, is used almost oppositely to the sense of<br />
symbolism that was suggested by Peirce, i. e. it focuses on what he called<br />
icons and <strong>in</strong>dexes but not symbols. Most of the words discussed also<br />
conta<strong>in</strong> a conventional arbitrary element which means we are deal<strong>in</strong>g with<br />
what could be called iconic and <strong>in</strong>dexical symbols. (For further<br />
discussion, see chapter 2.)<br />
There also seems to be a further term<strong>in</strong>ological issue here, <strong>in</strong> that sound<br />
symbolism tends to be reserved for universal phenomena and<br />
8
phonesthemes for language specific phenomena. I would, however, prefer<br />
to use sound symbolism as a more general term and use the term universal<br />
sound symbolism when this is the issue.<br />
The relationship between the most important terms discussed here can be<br />
illustrated <strong>in</strong> the follow<strong>in</strong>g way (figure 1.2).<br />
sound symbolism or<br />
motivated expressions<br />
onomatopoeia other sound symbolism<br />
phonesthemes<br />
free morphemes<br />
phonesthemes<br />
9<br />
free morphemes<br />
Figure 1.2. The relationship between some of the most important terms<br />
discussed <strong>in</strong> 1.2.<br />
1.3. Is sound symbolism the rule or the exception<br />
<strong>in</strong> language?<br />
1.3.1 <strong>Sound</strong> symbolism is an exception<br />
Saussure (1916) who held the view that l<strong>in</strong>guistic signs are arbitrary<br />
wrote that ‘‘onomatopoeic words are never organic elements of a<br />
l<strong>in</strong>guistic system’’. Bühler (1933/1969) said that onomatopoeia is a<br />
reversion, s<strong>in</strong>ce language has evolved beyond primitive needs and means<br />
of self-expression. The genesis of language is a measure of its success at<br />
arbitrary symbolic (i. e. not what I call sound symbolic) representation.<br />
1.3.2 <strong>Sound</strong> symbolism is fundamental to language<br />
Already von der Gabelentz (1891) wrote about the ‘‘sound symbolic<br />
feel<strong>in</strong>g’’, the experience that sound and mean<strong>in</strong>g are <strong>in</strong>alienably<br />
<strong>in</strong>terconnected for the naive members of the speech community. For
example, naive Germans would say ‘‘that Frenchmen are silly when they<br />
name e<strong>in</strong> Pferd Schewall’’ (Jakobson and Waugh, p. 182).<br />
An opposite view was held by Saussure's contemporary Jespersen (1922<br />
a), who went as far as to claim that ‘‘languages <strong>in</strong> the course of time grow<br />
richer and richer <strong>in</strong> symbolic’’ (that is sound symbolic) ‘‘words’’ and<br />
‘‘develop towards a greater number of easy and adequate expressions –<br />
expressions <strong>in</strong> which sound and sense are united <strong>in</strong> a marriage-union<br />
closer than was ever known to our remote ancestors’’.<br />
To the extent that African so called ideophones are discussed (by e.g.<br />
Samar<strong>in</strong>, 1978 4 ) these are usually considered part of language proper.<br />
One might question if this is because ideophones play a much more central<br />
role <strong>in</strong> certa<strong>in</strong> African languages than <strong>in</strong> European languages, or whether<br />
there is another, freer tradition <strong>in</strong> describ<strong>in</strong>g non-European and primarily<br />
spoken languages. Diffloth (1976) writes that <strong>in</strong> describ<strong>in</strong>g languages with<br />
a structure remote from ones own language, it is often self-evident to<br />
segment roots <strong>in</strong>to smaller, significant units with their own sound<br />
symbolic value, thereby touch<strong>in</strong>g on the subject of how grammatical<br />
tradition may obscure <strong>in</strong>terest<strong>in</strong>g phenomena <strong>in</strong> a language.<br />
Lakoff and Johnson (1980), show<strong>in</strong>g many examples, claim that <strong>in</strong> all of<br />
language, <strong>in</strong>clud<strong>in</strong>g syntax, there are many cases of non-arbitrary<br />
relations between form and mean<strong>in</strong>g. This seems to po<strong>in</strong>t to a view of<br />
non-arbitrar<strong>in</strong>ess be<strong>in</strong>g <strong>in</strong>tegrated <strong>in</strong> language and not <strong>in</strong> a separate<br />
category.<br />
A phenomenon <strong>in</strong> Swedish which is similar to ideophones (cf. 1.2.4) are<br />
the ‘‘sound words’’ of adolescent language, described by Nordberg<br />
(1986), e.g. krch, ppff, då<strong>in</strong>g. These are usually not considered to belong<br />
to the core of Swedish, if described at all. However, they show <strong>in</strong>terest<strong>in</strong>g<br />
similarities to e.g. the ideophones <strong>in</strong> Gbeya (a dialect of Gbaya, spoken <strong>in</strong><br />
Central Africa), (Samar<strong>in</strong>, 1978); they show reduplication, lengthen<strong>in</strong>g,<br />
phonotactic ‘‘divergencies’’ and sentence f<strong>in</strong>al position. The greatest<br />
4<br />
Samar<strong>in</strong> def<strong>in</strong>es ideophony as ‘‘the foreground<strong>in</strong>g of phonological elements <strong>in</strong> word (or<br />
lexeme) composition <strong>in</strong> both spontaneous creations and fully <strong>in</strong>stitutionalized lexicons,<br />
usually associated with semantic categories of an attributive nature, commonly affective,<br />
and sometimes also <strong>in</strong> true onomatopoeia’’<br />
10
difference to Gbeyan seems to be on the semantic side; Gbeyan words<br />
seem to have more specific mean<strong>in</strong>gs, that can be assigned dictionary-type<br />
def<strong>in</strong>itions, e.g. ndadak ndadak ‘‘(wood that) doesn’t split well <strong>in</strong><br />
chopp<strong>in</strong>g’’.<br />
1.3.3 <strong>Sound</strong> symbolism is both <strong>in</strong>side and outside<br />
of language<br />
Householder (1946), writes that the vocabulary of English falls <strong>in</strong>to three<br />
parts with regard to arbitrar<strong>in</strong>ess of structure: 1) those clearly and<br />
completely arbitrary, their mean<strong>in</strong>g unaffected by the sound, 2) those<br />
made up, <strong>in</strong> whole or part, of phonesthemes, 3) those belong<strong>in</strong>g primarily<br />
to the first group but with their mean<strong>in</strong>g colored or altered <strong>in</strong> vary<strong>in</strong>g<br />
degree by secondary association with phonesthemes.<br />
In the Oxford Dictionary of English Etymology (1967) the consonant wr -<br />
is a separate entry mean<strong>in</strong>g ‘‘consonant comb<strong>in</strong>ation occurr<strong>in</strong>g <strong>in</strong>itially <strong>in</strong><br />
many words imply<strong>in</strong>g twist<strong>in</strong>g or distortion’’. The dictionary further<br />
states that the mean<strong>in</strong>g of twist<strong>in</strong>g has many correspondences <strong>in</strong> other<br />
Germanic languages5. Wescott (1975) uses the term ‘‘microlanguage’’ for the core of language,<br />
which is subject to well-known grammatical rules – i.e. conventional<br />
language. For other doma<strong>in</strong>s of speech, e.g. baby talk, exclamations,<br />
verbal art – ‘‘language that is alienated from conventionally structured<br />
speech’’, he co<strong>in</strong>s the cover term ‘‘allolanguage’’. He says, ‘‘one of the<br />
characteristics of allolanguage is a closer relation between sound and sense<br />
than obta<strong>in</strong>s <strong>in</strong> microlanguage’’, ‘‘a retention ... of an older and simpler<br />
manner of self expression alongside one that is more recent and<br />
complex.’’ <strong>Sound</strong> symbolism, be<strong>in</strong>g one aspect of allolanguage, is thus<br />
part of language, but on the periphery, accord<strong>in</strong>g to Wescott.<br />
<strong>Sound</strong>s can enter a language by means of sound symbolic words (e.g.<br />
H<strong>in</strong>ton, 1986). Also, sound changes often do not affect sound symbolic<br />
words, so that phonemes that should have disappeared <strong>in</strong> a language or<br />
have become restricted to certa<strong>in</strong> environments are still to be found <strong>in</strong><br />
sound symbolic vocabulary (as po<strong>in</strong>ted out by H<strong>in</strong>ton, Nichols, Ohala<br />
5Cf. the Swedish clusters kr- and sn-, <strong>in</strong> chapter 4.<br />
11
1994). E. g. <strong>in</strong>itial r- is rare <strong>in</strong> the non sound-symbolic vocabulary of<br />
F<strong>in</strong>nish (Austerlitz, 1994).<br />
There are also suggestions of tendencies to use a more reduced phonemic<br />
<strong>in</strong>ventory <strong>in</strong> sound symbolism (e.g. Oswalt 1994).<br />
1.3.4 Evaluation of discussion of sound symbolism<br />
The view of the difference between sound symbolic words and other<br />
words assumed <strong>in</strong> this thesis is the follow<strong>in</strong>g: For most words, the<br />
ord<strong>in</strong>ary speaker will, on reflection, agree that there is no motivation for<br />
them (e.g. for horse) but he/she will say that there is a motivation for<br />
sound symbolic words. Of course there will be a border area where<br />
different speakers will disagree or f<strong>in</strong>d it hard to judge.<br />
Bol<strong>in</strong>ger (1950) takes a radical methodological approach to morphology<br />
<strong>in</strong> general, with phonesthemes (which are also called affective morphemes<br />
<strong>in</strong> contrast to neutral morphemes, <strong>in</strong> spite of the impossibility of stat<strong>in</strong>g<br />
where the neutral ends and the affective beg<strong>in</strong>s) as one type of possible<br />
output from a strictly synchronic analysis. Accord<strong>in</strong>g to Bol<strong>in</strong>ger, the<br />
alternative approach of tak<strong>in</strong>g etymology <strong>in</strong>to account is not available<br />
s<strong>in</strong>ce ord<strong>in</strong>ary, naive speakers' judgements are important for discover<strong>in</strong>g<br />
morphemes. Bol<strong>in</strong>ger suggests the existence of numerous phonesthemes<br />
but also claims that they are too fluid to be penned with limits. One<br />
possible way to evaluate morphemes, accord<strong>in</strong>g to Bol<strong>in</strong>ger, would be<br />
through lexical and textual frequencies (cf. chapter 4). Bol<strong>in</strong>ger claims<br />
that roughly half of the words, <strong>in</strong> English, beg<strong>in</strong>n<strong>in</strong>g with gl- have the<br />
implication ‘‘visual’’. Bol<strong>in</strong>ger claims that as percentages go this is better<br />
than some of the paradigmatic suffixes (with regard to a constant<br />
association of mean<strong>in</strong>g and form), though of course gl- is never more<br />
than sporadically productive.<br />
1.4 Is sound symbolism productive or not?<br />
1.4.1 <strong>Sound</strong> symbolism is not productive<br />
Samar<strong>in</strong> (1978), <strong>in</strong> his studies of Gbeyan, has not found support for the<br />
creation of new ideophones (cf. 1.2.4). In traditional etymology the<br />
explanation of new co<strong>in</strong>ages is often just by ‘‘analogy’’ with one other<br />
word (which implies non-productivity).<br />
12
1.4.2 <strong>Sound</strong> symbolism is productive<br />
Most of the above mentioned l<strong>in</strong>guists who are specifically <strong>in</strong>terested <strong>in</strong><br />
the phenomenon and who view it as an <strong>in</strong>tegral part of language, also<br />
regard it as productive. For example, Nordberg (1986) who studied sound<br />
words, deals almost exclusively with new co<strong>in</strong>ages.<br />
1.4.3 Greater or lesser degree of productivity<br />
Rhodes (1994) discusses onomatopoeia, aural images (mapp<strong>in</strong>g sound onto<br />
sound) and forms based on aural images. He dist<strong>in</strong>guishes between ‘‘wild’’<br />
and ‘‘tame’’ words, these be<strong>in</strong>g the ends of a scale. ‘‘At the extreme wild<br />
end the possibilities of the human vocal tract are utilized to their fullest to<br />
imitate sounds of other than human orig<strong>in</strong>. At the tame end the imitated<br />
sound is simply approximated by an acoustically close phoneme or<br />
phoneme comb<strong>in</strong>ation.’’<br />
Bol<strong>in</strong>ger does an assonance-rime analysis of English monosyllables (cf.<br />
Bol<strong>in</strong>ger 1950) where the <strong>in</strong>itial consonant(s) constitute the assonance and<br />
the rema<strong>in</strong>der of the syllable is the rime. He argues that assonance-rime<br />
analysis (of tame words) is morphology because assonances and rimes do<br />
not comb<strong>in</strong>e productively. That, however, does not mean that a<br />
construction is frozen. He <strong>in</strong>troduces the term ‘‘active’’ for constructions<br />
that produce monosyllables cont<strong>in</strong>uously, at a slow rate.<br />
1.4.4 Evaluation of the discussion of productivity<br />
The hypothesis that will be tested <strong>in</strong> the experiments is that phonesthemes<br />
are productive to a greater or lesser degree, i.e. that some phonesthemes<br />
are more productive than others. The <strong>in</strong>termittent occurrence of new<br />
forms <strong>in</strong> speech, prose and fiction, (which fit <strong>in</strong>to a pattern), especially <strong>in</strong><br />
child literature, constitutes an argument for this claim. The opposite view<br />
would mean that new co<strong>in</strong>ages would be phonetically and semantically<br />
haphazard. However, with that view, the fairly wide-spread and easy<br />
comprehension of new forms would be difficult to account for. The<br />
concept of phonestheme <strong>in</strong>volves stronger or weaker productivity.<br />
The dist<strong>in</strong>ction between understand<strong>in</strong>g and production might also be<br />
fruitful. When be<strong>in</strong>g presented with deliberately constructed nonsense<br />
words <strong>in</strong> the experiments of this thesis, listeners have no objections to or<br />
difficulties <strong>in</strong> assign<strong>in</strong>g some <strong>in</strong>terpretation to them (cf. 7.1.3).<br />
13
Another problem, <strong>in</strong> this area of research is to decide where the<br />
borderl<strong>in</strong>e goes between lexicalized and more temporary, newly created,<br />
forms. In other words, what I experience as a neologism can be an<br />
established word <strong>in</strong> a subgroup or an (ext<strong>in</strong>ct) dialectal word. 6<br />
1.5 The question of etymology<br />
In an etymological perspective this part of the vocabulary is less static <strong>in</strong><br />
one aspect, more static <strong>in</strong> another; onomatopoeic words are constantly<br />
recreated, but this also makes them keep much the same form throughout<br />
the ages, as they don't always undergo general changes of sound and<br />
mean<strong>in</strong>g Jespersen (1922 a) gives the example of cuckoo which has not<br />
changed its vowel from [u] to [U], as <strong>in</strong> cut, but is pronounced [kuku]. An<br />
onomatopoeic word is constra<strong>in</strong>ed by the sound it imitates.<br />
In account<strong>in</strong>g for onomatopoeic and other sound symbolic expressions, a<br />
synchronic explanation will be given, irrespective of whether the<br />
explanation is historically ‘‘true’’ or ‘‘false’’. Some examples will be given<br />
of what some critics of classical etymology have stated. First, however, it<br />
should be po<strong>in</strong>ted out that there seems to be a general agreement that<br />
onomatopoeic (and also sound symbolic) words usually do not undergo<br />
the same phonological changes as other words, e.g. accord<strong>in</strong>g to Grimm's<br />
law, as long as they still have a sound imitative mean<strong>in</strong>g. They constantly<br />
get renewed and sometimes reshaped by fresh imitation. Wescott (1975)<br />
writes: ‘‘Thus, the reconstructed proto-<strong>in</strong>doeuropean forms *pap- ‘‘teat’’,<br />
*tut- ‘‘to hoot’’ and *kuku- ‘‘cuckoo’’, appear <strong>in</strong> English as pap, toot, and<br />
cuckoo rather than, as comparative philologists would normally predict,<br />
*faf, *thuth and *houhg.’’<br />
6<br />
Exist<strong>in</strong>g forms, like place-names, can also be given a (new) mean<strong>in</strong>g, befitt<strong>in</strong>g their<br />
form, as <strong>in</strong> the book ‘‘The mean<strong>in</strong>g of Liff’’ (Adams & LLoyd, 1983), which states that<br />
‘‘In life, there are many hundreds of common experiences, feel<strong>in</strong>gs, situations and even<br />
objects which we all know and recognize, but for which no words exist. On the other<br />
hand, the world is littered with thousands of spare words, which spend their time do<strong>in</strong>g<br />
noth<strong>in</strong>g but loaf<strong>in</strong>g on signposts po<strong>in</strong>t<strong>in</strong>g at places...’’ These names and mean<strong>in</strong>gs are<br />
then paired like e.g., Cranleigh: a mood of irrational irritation with everyone and<br />
everyth<strong>in</strong>g, Burbage: The sound made by a lift full of people all try<strong>in</strong>g to breathe politely<br />
through their noses, or Plymouth: to relate an amus<strong>in</strong>g story to someone without<br />
remember<strong>in</strong>g that it was they who told it to you <strong>in</strong> the first place.<br />
14
On the other hand, Jespersen (1922 a) also remarks that ‘‘words that have<br />
been symbolically expressive may cease to be so <strong>in</strong> consequence of<br />
historical development, either phonetic, semantic or both.’’ An example,<br />
accord<strong>in</strong>g to Jespersen, is the word crow which is not now so good an<br />
imitation of the sound made by the bird as OE crawe was. And wh<strong>in</strong>e,<br />
pipe were better imitations when the vowel was still i as <strong>in</strong> Danish hv<strong>in</strong>e,<br />
pibe. But the sound made by a small bird is still pronounced with an i <strong>in</strong><br />
peep.<br />
Furthermore, writes Jespersen, some words have <strong>in</strong> the course of time<br />
become more expressive than they were at first. This phenomenon he calls<br />
secondary echoism or secondary symbolism. Patter ‘‘to talk rapidly or<br />
glibly’’ is to all <strong>in</strong>tents a truly symbolical word, even though it comes<br />
from pater (paternoster) and at first meant to repeat that prayer.<br />
An <strong>in</strong>terest<strong>in</strong>g example <strong>in</strong> Swedish is (atjo) prosit as a polite reaction to a<br />
sneeze, of which the latter word is probably apprehended as<br />
onomatopoeic, at least by children. (The word atjo is of course motivated.<br />
It is a conventionalization of a sound from a bodily reaction and the<br />
relation between the sound and the mean<strong>in</strong>g is <strong>in</strong>dexical (cf. the crossl<strong>in</strong>guistic<br />
comparison <strong>in</strong> chapter 5). The Lat<strong>in</strong> orig<strong>in</strong>al mean<strong>in</strong>g of prosit<br />
(which is ‘‘måtte det gagna’’ (may it be of use)) is probably quite dead <strong>in</strong><br />
the mental lexicons of most speakers. On the other hand, accord<strong>in</strong>g to<br />
Jespersen (1922 a), there are words which we feel <strong>in</strong>st<strong>in</strong>ctively to be<br />
adequate to express the ideas they stand for and others, the sound of which<br />
are felt to be more or less <strong>in</strong>congruous with their signification. These<br />
feel<strong>in</strong>gs of adequacy or <strong>in</strong>congruity are both examples of etymological<br />
creativity. A Swedish example of <strong>in</strong>congruity is perhaps munter (joyful) 7.<br />
Jespersen (1922 a, b) talks about ‘‘symbolism at work’’ where both sound<br />
and sense fit. Also, he writes, through changes <strong>in</strong> mean<strong>in</strong>g, too, some<br />
words have become more expressive than they were formerly. His<br />
example is m<strong>in</strong>iature, which, because of the i, has come to mean ‘‘a small<br />
picture’’ <strong>in</strong>stead of ‘‘image pa<strong>in</strong>ted with a m<strong>in</strong>imum of vermilion’’.<br />
Jespersen claims that ‘‘sound symbolism makes some words more fit to<br />
survive’’. The word roll, <strong>in</strong> French rouler, etc. derived from Lat<strong>in</strong> rota +<br />
7Because of the [P], cf. 5.4.3.<br />
15
dim<strong>in</strong>utive end<strong>in</strong>g -ul- ga<strong>in</strong>ed its popularity <strong>in</strong> English, Dutch, German,<br />
and Scand<strong>in</strong>avian languages, because the sound is suggestive of the sense.<br />
He also talks of sound symbolism be<strong>in</strong>g <strong>in</strong> action when borrow<strong>in</strong>g words<br />
from other languages and when co<strong>in</strong><strong>in</strong>g nouns, verbs, etc. from (place)<br />
names. To sum up then, accord<strong>in</strong>g to Jespersen, onomatopoeia and other<br />
sound symbolism do not always date back to the earliest times, and it is<br />
mostly un<strong>in</strong>terest<strong>in</strong>g to lay any aspects of ‘‘correct etymology’’ on words<br />
affected by onomatopoeia and sound symbolism.<br />
Wescott (1978) would make the same claim for most words, but especially<br />
for slang and proper names. He takes a polygenetic view of word orig<strong>in</strong>s<br />
and assumes that ‘‘lexical ancestry is relative rather than absolute <strong>in</strong><br />
nature’’ and that because of this derivational relativity ‘‘there is a gradual<br />
‘fade-out’ <strong>in</strong> the etymological antecedence of any lexeme and that this<br />
fade-out effect, <strong>in</strong> turn, leads <strong>in</strong>escapably to subjectivity <strong>in</strong> the assessment<br />
of degrees of lexical ancestry’’. Of the various processes that contribute to<br />
the development of polygenetic lexemes are, e.g. sound repetition,<br />
<strong>in</strong>different varieties, elision or ‘‘chopp<strong>in</strong>g’’, the conversion of spoken<br />
language <strong>in</strong>to written language and the consequent feedback effect of<br />
writ<strong>in</strong>g on speech. 8<br />
Von der Gabelentz (1891) observes historically ‘‘false’’ but synchronically<br />
‘‘true’’ etymologies based on collective agreement with<strong>in</strong> a given speech<br />
community. He writes that words l<strong>in</strong>ked together by both sound and<br />
mean<strong>in</strong>g manifest ‘‘elective aff<strong>in</strong>ities’’ (wahlverwandtschaften) able to<br />
modify both the shape and the content of the words <strong>in</strong>volved. Here, it is<br />
natural to refer to Householder's (1946) def<strong>in</strong>ition of a phonestheme: ‘‘a<br />
phoneme or cluster of phonemes shared by a group of words which also<br />
have <strong>in</strong> common some element of mean<strong>in</strong>g or function, though the words<br />
may be etymologically unrelated.’’ (Traditionally, words are said to be<br />
etymologically related if they can be traced back to the same word, but<br />
not if they conta<strong>in</strong> the same phonestheme, as a part of words. It would<br />
probably be fruitful for etymological study to explore the concept of<br />
8<br />
There is another k<strong>in</strong>d of etymology, namely folk etymology, which I believe is an<br />
expression for the same mental process as <strong>in</strong> productive sound symbolism. There seems<br />
to exist a human <strong>in</strong>cl<strong>in</strong>ation for motivated signs. Instead of us<strong>in</strong>g sounds that fit the<br />
mean<strong>in</strong>g, exist<strong>in</strong>g words are reshaped through us<strong>in</strong>g other but similar-sound<strong>in</strong>g<br />
morphemes to make the result<strong>in</strong>g word more morphologically transparent.<br />
16
phonestheme, i.e. sound symbolic l<strong>in</strong>ks could be just as valid as other<br />
morphological l<strong>in</strong>ks.)<br />
Bol<strong>in</strong>ger (1968) also holds the view that traditional etymology is not<br />
relevant for expla<strong>in</strong><strong>in</strong>g phonesthemes (for methodological reasons, cf.<br />
above). He also describes phonesthemes like this: ‘‘Given a particular<br />
word for a particular th<strong>in</strong>g, if other words for similar th<strong>in</strong>gs come to<br />
resemble that word <strong>in</strong> sound, then, no matter how arbitrary the<br />
relationship between sound and sense was to beg<strong>in</strong> with, the sense is now<br />
obviously tied to the sound. The relationship between sound and sense is<br />
still arbitrary, as far as the outside world is concerned (and would appear<br />
that way absolutely to a foreigner), but with<strong>in</strong> the system it is no longer<br />
so’’ (p. 242). Bol<strong>in</strong>ger (1950) also argues that when two expressions are<br />
encountered <strong>in</strong> the same area of greater or lesser specificity of mean<strong>in</strong>g,<br />
and are also similar <strong>in</strong> form, they are likely to exercise a k<strong>in</strong>d of magnetic<br />
attraction one upon the other. The attraction may be extremely remote.<br />
Most speakers of English, accord<strong>in</strong>g to Bol<strong>in</strong>ger (1950), when they hear<br />
ambush, are likely to th<strong>in</strong>k of someone hid<strong>in</strong>g <strong>in</strong> the bushes. Likewise<br />
with hierarchy: one tends to hear the element higher. (This phenomenon<br />
could be called '<strong>in</strong>terpretive folk etymology'.) Bol<strong>in</strong>ger (1950) is very<br />
consistent <strong>in</strong> his synchronic (and spoken language) approach, which<br />
sometimes leads to absurdities. His ma<strong>in</strong> arguments for the irrelevance of<br />
etymology is that it doesn't match speakers' judgements and that it isn't<br />
compatible with morpheme convergence and divergence.<br />
Malkiel (1994) discusses the role of phonosymbolic (i.e. sound symbolic)<br />
<strong>in</strong>terference <strong>in</strong> the sound development of words. He especially discusses<br />
the example of the older French word for close developed from Lat<strong>in</strong><br />
claudere be<strong>in</strong>g replaced (<strong>in</strong> Late Antiquity and the Middle Ages) by the<br />
word fermer under the <strong>in</strong>fluence of the word for iron (ferrum), a then<br />
newly <strong>in</strong>troduced, highly prestigious metal.<br />
It is sometimes po<strong>in</strong>ted out that every word has its own history, which is<br />
an anomalistic po<strong>in</strong>t of view (<strong>in</strong> the sense of antique Greek-Roman<br />
dist<strong>in</strong>ction of ‘‘anomaly vs. analogy’’). But one could also talk about<br />
different rule systems conflict<strong>in</strong>g. Malkiel, for example, refers to<br />
Meillet's (1931) discussion on phonetic m<strong>in</strong>i-systems expla<strong>in</strong><strong>in</strong>g<br />
exceptional lexical units handed down from one generation to another.<br />
17
These m<strong>in</strong>i-systems could well have existed side by side with the basic<br />
phonological structure.<br />
Rhodes (1994) writes that the abundant etymological dictionary comments<br />
like ‘‘orig<strong>in</strong> uncerta<strong>in</strong>’’ or ‘‘prob. ak<strong>in</strong> to’’ for English monosyllables<br />
should conv<strong>in</strong>ce one that these are <strong>in</strong>novated cont<strong>in</strong>uously, but at a slow<br />
rate (active comb<strong>in</strong>ation, <strong>in</strong> contrast to productive and frozen). There is<br />
also, he writes, an ongo<strong>in</strong>g re<strong>in</strong>terpretation of forms of various historical<br />
sources as new <strong>in</strong>stances of assonances (<strong>in</strong>itial consonants). An example is<br />
the largely Germanic sl- liquid (slop, slush) classifier (a type of<br />
phonestheme) which also, semantically, <strong>in</strong>cludes the sl- <strong>in</strong> Lat<strong>in</strong>ate sluice<br />
(< Vulgar Lat<strong>in</strong> *exclusa), (Swe: sluss), <strong>in</strong> spite of different etymological<br />
orig<strong>in</strong>s.<br />
1.6 The phylogenesis of language<br />
From the question of etymology it is natural to go on to a related question<br />
<strong>in</strong> which onomatopoeia and sound symbolism have figured, namely the<br />
question of the phylogenesis of language.<br />
Plato, <strong>in</strong> the dialogue Cratylos, treats the contest between the two basic<br />
l<strong>in</strong>guistic forces, convention and nature. The argument concerned the<br />
nature of names; Cratylos meant that names were given by nature,<br />
Socrates that they were conventional.<br />
Most theories of the orig<strong>in</strong> of language have built on onomatopoeia, the<br />
‘‘bow-wow theory’’ (e.g. Herder, 1772) (cf. Müller, 1861) for the names<br />
‘‘bow-wow’’, etc) or on gestures, (e.g. Herman Paul, accord<strong>in</strong>g to Fano,<br />
1962). Other variants are ‘‘div<strong>in</strong>e orig<strong>in</strong>’’, (e.g. Süssmilch, 1767) or ‘‘the<br />
pooh-pooh theory’’ (e.g. Rousseau, 1822), which says that the orig<strong>in</strong> of<br />
language lies <strong>in</strong> <strong>in</strong>terjectional, emotional cries triggered by strong feel<strong>in</strong>gs<br />
of fear, pa<strong>in</strong>, lust, etc. The ‘‘yo-he-ho theory’’ says that <strong>in</strong> collective<br />
labor, rhythmically produced spoken sounds help coord<strong>in</strong>ate the actions of<br />
many <strong>in</strong>dividuals. ‘‘The s<strong>in</strong>g-song theory’’ (e.g. Jespersen, 1922 a) means<br />
that the orig<strong>in</strong> of language lies <strong>in</strong> dance, song and related expressive<br />
vocaliz<strong>in</strong>g.<br />
Max Müller (1861), (who has co<strong>in</strong>ed humorous expressions) is a<br />
representative for the ‘‘d<strong>in</strong>g-dong theory’’, which might appear to give a<br />
sound symbolic explanation. It is, however, based on his reconstruction of<br />
18
400-500 Indo-European roots. He writes that man had an <strong>in</strong>nate<br />
<strong>in</strong>cl<strong>in</strong>ation to associate certa<strong>in</strong> types of sound with certa<strong>in</strong> types of objects<br />
and actions which echoed <strong>in</strong> him <strong>in</strong> a way which is analogous to an<br />
object's resonance when struck. The reconstructed Indo-European roots<br />
are an outgrowth of the d<strong>in</strong>g-dong effect. Müller would not, as would e.g.<br />
Jespersen, say that sound symbolism is productive <strong>in</strong> language, but that<br />
this <strong>in</strong>st<strong>in</strong>ct to give articulate expression for rational concepts <strong>in</strong> the<br />
human m<strong>in</strong>d has disappeared s<strong>in</strong>ce there is no longer a need for it, once<br />
language is established.<br />
One theory of how spoken language has developed from gestural language<br />
is ‘‘the mouth-gesture theory’’ (Paget, 1930 and Jóhannesson, 1949). It<br />
says, briefly, that the organs of speech tend to move <strong>in</strong> unison with hand<br />
and arm movements when these are used <strong>in</strong> sign language or when us<strong>in</strong>g<br />
tools. If such movements of the speech organs are accompanied by<br />
vocalizations, then the result<strong>in</strong>g sounds (that are similar to sounds <strong>in</strong><br />
articulated speech) eventually get the same mean<strong>in</strong>g as the gestures. The<br />
gestures of the organs of articulation are recognized by the hearer because<br />
the hearer unconsciously reproduced <strong>in</strong> his m<strong>in</strong>d the actual gesture which<br />
had produced the sound.<br />
Jóhannesson (1949) presents material from both Indo-European and<br />
Semitic sources. In Indo-European languages, he claims that about 5% of<br />
the words could be attributed to <strong>in</strong>terjection, probably of emotion, and<br />
10% to onomatopoeia, leav<strong>in</strong>g 85% expla<strong>in</strong>able as direct or <strong>in</strong>direct<br />
derivations from mouth-gesture.<br />
Other theories of the orig<strong>in</strong> of language have focused more on why and<br />
less on how language appeared, e.g. because of social need (Révész, 1946)<br />
or cerebral development or as a consequence of early child language<br />
development. An example of the later is ‘‘the babble-luck theory’’<br />
(Thorndike, 1943), which is to be understood as a theory of the selective<br />
re<strong>in</strong>forcement of <strong>in</strong>itial, random babbl<strong>in</strong>g, which later became words.<br />
Thorndike has been criticized for not expla<strong>in</strong><strong>in</strong>g how the parents of the<br />
earliest babblers were able to shape <strong>in</strong>fant vocalizations.<br />
Some l<strong>in</strong>guists (explicitly) believe <strong>in</strong> monogenesis, others <strong>in</strong> polygenesis<br />
of human language. For a much more extensive review on the subject, see<br />
Hewes (1977). I will not take a stand on these theories of the orig<strong>in</strong> of<br />
19
language, but suspect that the solution probably lies <strong>in</strong> a comb<strong>in</strong>ation of<br />
different theories, one of them be<strong>in</strong>g the bow-wow variant.<br />
The forego<strong>in</strong>g section leads up to the question of the role of onomatopoeia<br />
and sound symbolism <strong>in</strong> the ontogenesis of language. One of the more<br />
recent <strong>in</strong>vestigations, Williams (1991), concerns the phenomenon of [da]<br />
universally hav<strong>in</strong>g a deictic function <strong>in</strong> early language acquisition. She<br />
f<strong>in</strong>ds support for the mechanism of cross-modal transfer exist<strong>in</strong>g as a<br />
bridge to the acquisition of language and she holds that this can shed some<br />
light on sound symbolism <strong>in</strong> adult language. The cross-modal transfer<br />
doesn't disappear entirely once it has fulfilled its function, even though<br />
this aspect may later be regarded as superfluous. Without any other<br />
comparison this is similar to Müller's (1861) views on the genesis of<br />
language.<br />
1.7 Universality versus language specificity<br />
One important question, which is not always clearly accounted for, is that<br />
of the universality or language specificity of sound symbolism.<br />
Universality can be a feature of the semantic side, the expressive side or<br />
more abstractly.<br />
1.7.1 <strong>Sound</strong> symbolism is universal<br />
Samar<strong>in</strong> (e.g. 1978) writes that it is a universal fact that all languages use<br />
onomatopoeia and sound symbolism, but that the expressions differ. He<br />
does not make any claims for the semantic side, other than that all<br />
languages have an expressive function.<br />
Bol<strong>in</strong>ger (1965) writes that the tendency of forms to mold themselves on<br />
other forms with like mean<strong>in</strong>gs and of mean<strong>in</strong>gs to mold themselves on<br />
other mean<strong>in</strong>gs, conveyed by like words, is universal. Allot (1973), us<strong>in</strong>g<br />
data from many languages, claims that there is a tendency for words that<br />
sound similar to have similar mean<strong>in</strong>gs <strong>in</strong> languages not known to be<br />
historically related.<br />
The mean<strong>in</strong>g of [i] has been well studied, and claims have been made for<br />
its universality <strong>in</strong> sound symbolism (Ultan, 1978; Jespersen, 1933).<br />
There seems to be a general agreement that phonesthemes are language<br />
specific, a conclusion which is not necessary; besides, phonesthemes are<br />
20
often similar <strong>in</strong> related languages. A reasonable claim as concerns<br />
universality is that ‘‘the <strong>in</strong>tral<strong>in</strong>guistic variation recapitulates the<br />
<strong>in</strong>terl<strong>in</strong>guistic variation’’ (Allwood, 1985). Wescott (1975) handles the<br />
same phenomena by claim<strong>in</strong>g that what is not <strong>in</strong> the microlanguage<br />
(conventionally structured speech) is to be found <strong>in</strong> the allolanguage<br />
(speech which is alienated from conventionally structured speech e.g.<br />
<strong>in</strong>fantile babbl<strong>in</strong>g, non-grammatical <strong>in</strong>terjection, poetry and song). He<br />
claims universality <strong>in</strong> processes like reduplication and lengthen<strong>in</strong>g. Some<br />
experimental l<strong>in</strong>guists have explicitly tackled the question of universality<br />
by test<strong>in</strong>g people speak<strong>in</strong>g different languages, e.g. Osgood (1962) (see<br />
1.10.5).<br />
Allwood (1983) discusses language <strong>in</strong> general, about the relationship<br />
between language and thought and its connection with universal and<br />
relativistic standpo<strong>in</strong>ts. The conclusions relevant to onomatopoeia and<br />
other sound symbolism are, <strong>in</strong> brief, that universal traits <strong>in</strong> language<br />
mostly concern the semantic, not the expression side and that semantic<br />
universals are most likely <strong>in</strong> areas that are biologically grounded, e.g.<br />
motoric and perceptual activities. Examples are spatial relations and color<br />
terms.<br />
1.7.2 <strong>Sound</strong> symbolism is not universal<br />
Some l<strong>in</strong>guists seem to have taken the fact that there are differences<br />
between languages concern<strong>in</strong>g (presumptive) onomatopoeia and other<br />
sound symbolism as a proof of the non-existence of sound symbolism, on<br />
the grounds that if sound symbolism did exist, it should be universal <strong>in</strong><br />
form and content. This conclusion rests on some unarticulated<br />
assumptions concern<strong>in</strong>g the way <strong>in</strong> which sound symbolism ought to be<br />
accounted for, e. g. as <strong>in</strong>nate, cf. Bol<strong>in</strong>ger (1968), ‘‘If there were a real<br />
connection between the sound of a word and its mean<strong>in</strong>g, a person who<br />
did not know the language would be able to guess the word if he knew the<br />
mean<strong>in</strong>g and guess the mean<strong>in</strong>g if he heard the word. This almost never<br />
happens, even with words that imitate sounds’’.<br />
Austerlitz (1994) has studied vowels <strong>in</strong> e.g. F<strong>in</strong>nish and suggests that there<br />
is a language-specific correlation between recentness and exploitability <strong>in</strong><br />
sound symbolism. In F<strong>in</strong>nish the vowels (and the consonants) are<br />
unequally rooted <strong>in</strong> the system. The vowel ö /ø/ is the most recent arrival<br />
21
(and is secondary, and marked, as all front rounded vowels are) <strong>in</strong> the<br />
sound system and it is the vowel which is the most sound symbolic.<br />
1.7.3 Evaluation of the discussion of universality<br />
versus language specificity <strong>in</strong> sound symbolism<br />
An important question is whether phonesthemes are language specific, as<br />
Bol<strong>in</strong>ger (1950) claims. If so, then the orig<strong>in</strong> of each phonestheme could<br />
be attributed solely to chance, a ‘‘cluster<strong>in</strong>g effect’’ (which would be<br />
<strong>in</strong>terest<strong>in</strong>g <strong>in</strong> itself). Another possibility is that phonesthemes show<br />
universal tendencies, even if ‘‘only’’ <strong>in</strong> cases concern<strong>in</strong>g semantic<br />
features. If they show universal tendencies, the explanation could partly<br />
have to do with <strong>in</strong>nateness, and partly <strong>in</strong>volve a basic relationship between<br />
environment and <strong>in</strong>dividual.<br />
But it may be that, <strong>in</strong>stead of the dichotomy between certa<strong>in</strong> sounds which<br />
are universal and certa<strong>in</strong> phonesthemes which are language specific, one<br />
should dist<strong>in</strong>guish between (a) universal sound symbolism perta<strong>in</strong><strong>in</strong>g to<br />
certa<strong>in</strong> sounds and sound clusters and (b) the phenomenon of<br />
‘‘phonesthemicity’’, i. e. the tendency of certa<strong>in</strong> sound-mean<strong>in</strong>g<br />
comb<strong>in</strong>ations to mold themselves on other comb<strong>in</strong>ations, perta<strong>in</strong><strong>in</strong>g to all<br />
sounds and clusters, etc. In the latter case the result could well be<br />
language specific while the thrust for analogy itself is universal. The next<br />
question is if the cluster<strong>in</strong>g effect is enhanced <strong>in</strong> certa<strong>in</strong> types of mean<strong>in</strong>g<br />
or sound. There might also be a frequency effect emanat<strong>in</strong>g from<br />
syntagmatic context.<br />
If the semantic-phonetic relationships of motivated words could be<br />
analytically treated one by one, my assumption is that the existence of<br />
universality <strong>in</strong> phonesthemes on the phonetic side (i. e. that e.g. imitation<br />
of ‘‘wet sounds’’ is done with the same speech sounds <strong>in</strong> different<br />
languages) is most likely at a level of (comb<strong>in</strong>ations of) dist<strong>in</strong>ctive<br />
features, e.g. voiceless, fricative, etc.<br />
On the semantic/functional side phonesthemes are probably partly<br />
universal and partly language specific. The result here probably depends,<br />
to a great extent, on how abstract the semantic categorization is. Also,<br />
some semantic fields are more likely to conta<strong>in</strong> sound symbolism, e.g.<br />
[DIMINUTIVE].<br />
22
Accord<strong>in</strong>g to Allwood, as referred to above (1983), the universals <strong>in</strong><br />
general to be expected, viz. universals of content are likely to be related<br />
to perception and perhaps to motoric behavior. This proposal does not,<br />
however, expla<strong>in</strong> all semantic categories <strong>in</strong>volved <strong>in</strong> sound symbolism.<br />
Whorf (1956) has an argument for universals of content: ‘‘<strong>in</strong> the<br />
psychological experiments, human subjects seem to associate the<br />
experiences of bright, cold, sharp, hard, high, light (<strong>in</strong> weight), quick,<br />
high-pitched, narrow, and so on <strong>in</strong> a long series, with each other; and<br />
conversely, the experiences of dark, warm, yield<strong>in</strong>g, soft, blunt, low,<br />
heavy, slow, low-pitched, wide, etc. <strong>in</strong> another long series. This occurs<br />
whether the WORDS for such associated experiences resemble them or<br />
not, but the ord<strong>in</strong>ary person is likely to NOTICE a relation to words only<br />
when it is a relation of likenesses to such a series <strong>in</strong> the vowels and<br />
consonants of words’’ (p. 267 f).<br />
It is not always clear if the authors mentioned above have discussed<br />
understand<strong>in</strong>g, production or both. It is most clear <strong>in</strong> the experiments<br />
conducted, cf. 1.10.5).<br />
1.8 Context<br />
Words (and parts of them) are always perceived <strong>in</strong> a context, which<br />
<strong>in</strong>fluences their <strong>in</strong>terpretation. The types of context that are <strong>in</strong>terest<strong>in</strong>g<br />
for sound symbolism are:<br />
i) phonetic/phonological and semantic<br />
ii) only semantic<br />
iii) situational<br />
The <strong>in</strong>terpretation of the mean<strong>in</strong>g connected with e.g. a consonant cluster<br />
can thus be <strong>in</strong>fluenced by phonetic/phonological and semantic<br />
(phonesthemic) context e.g. other consonants, vowels, <strong>in</strong>tonation, etc.<br />
connected with a certa<strong>in</strong> mean<strong>in</strong>g.<br />
The <strong>in</strong>terpretation can also be <strong>in</strong>fluenced (e.g. disambiguated) by semantic<br />
context only. For example, a consonant cluster can have as a weak<br />
phonestheme mean<strong>in</strong>g 'wetness'; the cluster pl-, which is otherwise<br />
ma<strong>in</strong>ly 'pejorative', can also mean 'wetness'. In the context of other words<br />
23
hav<strong>in</strong>g to do with wetness (and e.g. a word like r<strong>in</strong>na (flow)) a neologism<br />
beg<strong>in</strong>n<strong>in</strong>g with pl- could adopt a mean<strong>in</strong>g of wetness.<br />
The disambiguation of a consonant cluster can also depend on the speech<br />
situation <strong>in</strong> which e.g. a neologism or an ambiguous word is uttered, as<br />
for all words. An example of an ambiguous word is klabb (‘‘wet snow’’<br />
or ‘‘chunk of wood’’). The cluster kl- can mean 'wetness' (wet snow) or<br />
'short wide form' (chunk of wood).<br />
It is probable that at least some motivated words, e.g. the sound words of<br />
teenagers (like tssccht <strong>in</strong> ‘‘she wears her hair like this : tssccht’’), are<br />
more dependent on situational context than more arbitrary lexical<br />
morphemes; a person, who was not present <strong>in</strong> the situation were the word<br />
‘‘tssccht’’ was uttered, cannot understand exactly how the girl's hair<br />
looked.<br />
Grammont (1933), who was ma<strong>in</strong>ly <strong>in</strong>terested <strong>in</strong> the evocative value of<br />
vowels, especially <strong>in</strong> word forms reduplicated with a vowel change, <strong>in</strong><br />
different languages (e.g. ritsch-ratsch, piff-paff-puff) claimed that the<br />
mean<strong>in</strong>g of a vowel manifests itself when it is prompted by the mean<strong>in</strong>g<br />
of the text or when it at least does not stand <strong>in</strong> contradiction to it. The<br />
degree depends on the subjectivity of speakers and listeners as well as on<br />
situations, e.g. affective speech and poetry be<strong>in</strong>g favorable.<br />
As will be seen <strong>in</strong> chapter 7, the experiments <strong>in</strong> the present work are<br />
ma<strong>in</strong>ly done without variation of context. The reason for this is the desire<br />
to know how much (if any) of the sound symbolic mean<strong>in</strong>g is conveyed by<br />
certa<strong>in</strong> consonant clusters themselves, without context consist<strong>in</strong>g of<br />
vowels, other consonants, other words, <strong>in</strong>tonation, gestures, etc.<br />
However, <strong>in</strong> these experiments situational context is present from another<br />
aspect. As is shown <strong>in</strong> chapter 4, some consonant clusters have a higher<br />
frequency of motivated words with a certa<strong>in</strong> mean<strong>in</strong>g while other<br />
consonant clusters are less dom<strong>in</strong>ated by motivated words. This is often<br />
reflected <strong>in</strong> the results of the tests described <strong>in</strong> chapter 7. It is reasonable<br />
to believe that neologisms beg<strong>in</strong>n<strong>in</strong>g with not so clearly profiled clusters<br />
are more dependent on the l<strong>in</strong>guistic or extral<strong>in</strong>guistic context for their<br />
<strong>in</strong>terpretation.<br />
24
An example of phonetic/phonological and semantic (phonesthemic)<br />
context could be an <strong>in</strong>vented name like Pjäfser which isn't very attractive<br />
because of the pejorative pj- and the pejorative -fs (like <strong>in</strong> hafs, slafs,<br />
tjafs, krafs, rufs, bjäfs). These two clusters make the weakly pejorative -<br />
E- come to life and add to the pejorative impression. (In addition to this<br />
the suffix -er also has a pejorative nuance.)<br />
The cluster fl-, which ma<strong>in</strong>ly means 'quick movement' also has the<br />
phonestheme mean<strong>in</strong>g 'pejorative' (and few others). In a neologism like<br />
flafs it is likely that the whole word will be <strong>in</strong>terpreted as a pejorative<br />
because of the pejorative end<strong>in</strong>g -fs. The less common pejorative mean<strong>in</strong>g<br />
of fl- is activated because of the phonological/semantic context of -fs.<br />
From this we can see that words with sound symbolism are neither more<br />
context dependent nor less context dependent than other words. Certa<strong>in</strong><br />
clusters are more context dependent, while other are less context<br />
dependent, depend<strong>in</strong>g on the lexical sound symbolic strength of the cluster<br />
(cf. chapter 4).<br />
While the cluster pj- is mostly pejorative, another clearly sound symbolic<br />
cluster kl- carries several mean<strong>in</strong>gs, i.e. 'sound', 'talk<strong>in</strong>g', 'pejorative',<br />
'wetness', 'adhesion' and 'short wide form'. What determ<strong>in</strong>es the mean<strong>in</strong>g<br />
of a neologism beg<strong>in</strong>n<strong>in</strong>g with kl- must be either phonesthemic, other<br />
semantic or situational context. Look<strong>in</strong>g at the examples <strong>in</strong> Appendix 1<br />
there seems to be a tendency that kl- words mean<strong>in</strong>g 'short wide form'<br />
are monosyllabic and end<strong>in</strong>g with a gem<strong>in</strong>ate consonant. They do not end<br />
with fricative clusters; these appear here to be reserved for 'sound' and<br />
'wetness' mean<strong>in</strong>gs. Phonesthemic context seems to be important but it is<br />
not possible, at this stage, to give rules for this. (Cf. the discussion <strong>in</strong><br />
Allwood (1982) about mean<strong>in</strong>g potential and contextual conditions for<br />
different mean<strong>in</strong>gs. Cf. also 5.6.)<br />
The <strong>in</strong>fluence of context expla<strong>in</strong>s why the same sound (sequence) mostly<br />
unambiguously can appear <strong>in</strong> both sound symbolic words (as<br />
phonesthemes) and <strong>in</strong> non-sound symbolic words (as just a phoneme<br />
sequence).<br />
Nerman (1954) makes an <strong>in</strong>terest<strong>in</strong>g analysis of the sound symbolism of<br />
vowels and consonants <strong>in</strong> Swedish poetry. He writes that they always stand<br />
25
<strong>in</strong> a context, and really the whole poem, or at least the stanza, ought to be<br />
quoted for every example and that all means conspire: content, rhythm<br />
and r<strong>in</strong>g. However, his analysis is not preceded by a lexical<br />
(paradigmatic) analysis.<br />
1.9 A framework for models of sound symbolism<br />
Models of onomatopoeia and sound symbolism must take such factors <strong>in</strong>to<br />
account as discussed <strong>in</strong> 1.3-1.8. One factor is the position of onomatopoeia<br />
and other sound symbolism <strong>in</strong> language, (i. e. <strong>in</strong> grammar and lexicon).<br />
Are they primarily central or somewhere on the periphery of language? I<br />
claim that onomatopoeia and sound symbolism are central and <strong>in</strong>side<br />
language and a part of morphology.<br />
However, some newly created motivated expressions (like the sound<br />
words of teenagers, e.g. krch, ppff, då<strong>in</strong>g) seem to be more context<br />
dependent than most established adjectives, nouns and verbs and therefore<br />
on the periphery of language - words are created <strong>in</strong> subgroups and are<br />
used for a limited time (cf. Kots<strong>in</strong>as, 1994). The study of sound<br />
symbolism naturally leads to the study of processes of language<br />
development: How do words and sounds enter language or disappear from<br />
language?<br />
Phonesthemes vary <strong>in</strong> regard to productivity. The question of<br />
productivity also <strong>in</strong>volves the issue of production of new expressions vs.<br />
understand<strong>in</strong>g of such expressions. <strong>Sound</strong> symbolic neologisms are not<br />
created all the time, but are created now and then. But when they are<br />
created, they are easily understood by the listener; nonsense words can be<br />
<strong>in</strong>terpreted almost immediately.<br />
Onomatopoeia and other sound symbolism are probably driv<strong>in</strong>g forces <strong>in</strong><br />
etymological development (and some claim they were so <strong>in</strong> the genesis of<br />
language).<br />
The general framework has to <strong>in</strong>clude a view of language as a dynamic<br />
phenomenon with components of different dignity for different functions.<br />
<strong>Sound</strong> symbolism is always possibly present, but new expressions (words<br />
or phonesthemes) can be seen as float<strong>in</strong>g <strong>in</strong> and out of language. Whether<br />
or not they are actually ‘‘<strong>in</strong> language’’ depends on situation, stylistic<br />
context, time span under study and geographical area studied.<br />
26
The description and explanation of sound symbolic phenomena is related<br />
to the question of universality. If there are universals <strong>in</strong> (different aspects<br />
of) sound symbolism, this is compatible with <strong>in</strong>nateness of sound<br />
symbolism.<br />
Many of the above mentioned po<strong>in</strong>ts are relevant for most aspects of<br />
language, but perhaps more typically so for onomatopoeia and other<br />
sound symbolism, which are often more productive and universal. This<br />
leads back to the first question of their status <strong>in</strong> language. Other sound<br />
symbolism and onomatopoeia are different from other expressions,<br />
consider<strong>in</strong>g some of the po<strong>in</strong>ts mentioned above (productivity and<br />
universality). But, above all, they are different because of the nonarbitrary<br />
dimension.<br />
The dimensions of the framework are:<br />
1) universality - language specificity<br />
2) <strong>in</strong>nateness<br />
3) degree of conventionalization<br />
4) productivity<br />
5) centrality <strong>in</strong> language (historically, genetically, frequentially)<br />
6) types of context determ<strong>in</strong>ation<br />
and they can be related 9 <strong>in</strong> the follow<strong>in</strong>g way:<br />
<strong>in</strong>nate<br />
universal<br />
not <strong>in</strong>nate language specific<br />
conventional<br />
conventionalization<br />
centrality productivity type of context<br />
determ<strong>in</strong>ation<br />
Figure 1.3 A framework model for sound symbolism.<br />
27<br />
actual sound<br />
symbolic mean<strong>in</strong>g<br />
9The arrows <strong>in</strong> the model <strong>in</strong> figure 1.3 stand for different types of relationships, which<br />
are spelled out <strong>in</strong> the text below.
The model <strong>in</strong> figure 1.3 shows the follow<strong>in</strong>g: Assum<strong>in</strong>g that sound<br />
symbolism is central <strong>in</strong> language, this fact is compatible with both<br />
<strong>in</strong>nateness and non-<strong>in</strong>nateness of sound symbolism. If it is <strong>in</strong>nate it must<br />
be universal and if it is not <strong>in</strong>nate it is language specific (or universal due<br />
to pure chance). In both cases new expressions can be produced, based on<br />
<strong>in</strong>nate capacity and phonesthemes which are pre-exist<strong>in</strong>g because of<br />
<strong>in</strong>nateness or because of convention. Naturally there are no <strong>in</strong>nate forms<br />
that are unaffected by convention, and therefore universals, like i<br />
connected to smallness, are not absolute. Smallness does not implicate i<br />
and i does not implicate smallness. Language specific expressions can be<br />
said to be created by convention while universal expressions are affected<br />
by convention.<br />
Context affects the mean<strong>in</strong>g of all expressions. The mean<strong>in</strong>g potentials of<br />
both the language specific conventional clusters like kl- ('sound',<br />
'talk<strong>in</strong>g', 'wetness', 'adhesion', 'shortwide form' and 'pejorative') and<br />
more universal phonesthemes like i (smallness, high pitch, light) are<br />
disambiguated by context (phonesthemic, situational, etc.).<br />
A lexical description of sound symbolism <strong>in</strong> Swedish is presented <strong>in</strong><br />
chapters 4 and 5. The lexical description, which shows the mean<strong>in</strong>g<br />
potentials for consonant clusters, treats sound symbolism as a part of<br />
language and not as an exception. The description is a basis for<br />
predictions of sound symbolism (chapter 8). It does not claim<br />
universality. The explanatory model of chapter 2, which draws on the<br />
above discussion and describes <strong>in</strong> detail the nature of the motivated<br />
relation between expression and mean<strong>in</strong>g of sound symbolic words, is,<br />
however, easily relatable to many languages.<br />
1.10 Expression and content of sound symbolism<br />
Before I cont<strong>in</strong>ue with the presentation of more concrete results obta<strong>in</strong>ed<br />
by various l<strong>in</strong>guists, I will make some comments about the expressions<br />
and contents of onomatopoeia and other sound symbolism, i.e. phonemes,<br />
sounds, phonological features, semantic fields, and level of semantic<br />
categories.<br />
28
1.10.1 Expression<br />
The expressive side of onomatopoeia and other sound symbolism is<br />
usually described <strong>in</strong> terms of phonemes, i.e. as belong<strong>in</strong>g to the language<br />
<strong>in</strong> question, or as sounds (phones) not belong<strong>in</strong>g to the phonemic<br />
<strong>in</strong>ventory of the language <strong>in</strong> question. More seldom it is described <strong>in</strong><br />
terms of some k<strong>in</strong>d of phonological or phonetic features.<br />
In this presentation the phonemes are written as <strong>in</strong> the literature, but one<br />
has to keep <strong>in</strong> m<strong>in</strong>d that phonemes are relative to the phoneme system of<br />
the language under consideration.<br />
1.10.2 Content<br />
In terms of content, especially if one is <strong>in</strong>terested <strong>in</strong> similarities between<br />
languages, it is convenient to work with semantic fields (Trier, 1934,<br />
Lehrer, 1974) and semantic features. When deal<strong>in</strong>g with certa<strong>in</strong> k<strong>in</strong>ds of<br />
onomatopoeia, this is not so important; dogs sound about the same <strong>in</strong><br />
different countries, so different expressions can be fairly<br />
straightforwardly compared. Interjections, on the other hand, often<br />
express different aspects of human emotions and are therefore often more<br />
difficult to compare.<br />
While an expression such as aj (ouch) can probably be translated fairly<br />
easily <strong>in</strong>to other languages, many other exclamations are probably better<br />
described with semantic components <strong>in</strong> relation to a semantic field, rather<br />
than just translated to the nearest synonym. E.g. ohoj (ahoy) has an<br />
expression with a similar content (EwE) <strong>in</strong> Ososo, but this is not used to<br />
adults. Schas (shoo) <strong>in</strong> Icelandic has a special expression when directed to<br />
sheep (hau hau <strong>in</strong>stead of sch). Hoppsan (whoops) <strong>in</strong> Slovenian is<br />
translated differently depend<strong>in</strong>g on whether the surprise is experienced <strong>in</strong><br />
a negative or positive way (Oho: and Oi).<br />
1.10.3 Expression and content <strong>in</strong> different analyses<br />
It is often difficult to compare the results obta<strong>in</strong>ed by different l<strong>in</strong>guists<br />
because of different levels of analytic semantic categories. The categories<br />
can be very broad, e.g. th<strong>in</strong>gs and appearances (Jespersen, 1922 a) or<br />
more narrow, e.g. color (Samar<strong>in</strong>, 1978). Also, the categories differ<br />
from one l<strong>in</strong>guist to another. In the context of onomatopoeia and other<br />
sound symbolism it is difficult but perhaps desirable to strive towards a<br />
29
ase level à la Brown (1958). Bol<strong>in</strong>ger (1950) is of the op<strong>in</strong>ion that the<br />
level of specificity determ<strong>in</strong>es which morpheme analysis is made. ‘‘The<br />
lower the specificity of mean<strong>in</strong>g, the larger is the number of forms that<br />
may be subsumed under one morpheme.’’ As an example of this, he gives<br />
Nida's characterization of the 'suffix' /-´r/ i hammer, ladder, spider,<br />
otter, badger, and water as hav<strong>in</strong>g some sort of 'grammatical mean<strong>in</strong>g'.<br />
In spite of these difficulties I will try to compare the semantic<br />
categorizations of different l<strong>in</strong>guists, especially those who have done a<br />
more extensive analysis.<br />
The most elaborated taxonomy for onomatopoeia and other sound<br />
symbolism is Jespersen's (1922 a). He describes it as ‘‘a prelim<strong>in</strong>ary<br />
enumeration of the most obvious classes, with a small fraction of the<br />
examples ... collected.’’ The classes are:<br />
1. direct imitation, e.g. splash, klonk (onomatopoeia)<br />
2. orig<strong>in</strong>ator of the sound, e.g. cuckoo, or nicknames of nations from<br />
recurr<strong>in</strong>g words.<br />
3. movement (<strong>in</strong>separable from sound), e.g. flicker, snatch, slide<br />
4. th<strong>in</strong>gs and appearances (this seems to be form and light), e. g. ‘‘ a<br />
thick stick, a knot of wood, a bit of food, a protuberance, a small<br />
hill’’; gloom, light, dunkel<br />
5. states of m<strong>in</strong>d, e.g. grunt, sulky, also pejoratives, e.g. bosh<br />
6. size and distance, e.g. teeny<br />
7. length and strength of words and sounds; this gives an emotional<br />
effect, e.g. Danish langsommelig or Hungarian short words for commands<br />
and long words for entreaty; the category also mirrors perfective, plural,<br />
distance <strong>in</strong> time and space.<br />
A few po<strong>in</strong>ts of criticism of Jespersen's taxonomy are the follow<strong>in</strong>g:<br />
1) The basis of division is not homogeneous. The basic pr<strong>in</strong>ciple is<br />
semantic, but how this is applied is also open to discussion. The ma<strong>in</strong><br />
critique is, however: a) category 4 is a very broad semantic category, it<br />
covers almost everyth<strong>in</strong>g. I believe Jespersen wants to emphasize the<br />
<strong>in</strong>terpretation of visual perception. If so, then auditory perception ought<br />
to be treated <strong>in</strong> the same way, as th<strong>in</strong>gs and appearances; b) class 7 is the<br />
result of a categorization based on form and not on content.<br />
30
2) The categories are not mutually exclusive, either empirically (many<br />
words can be placed under both 1 and 3) or, analytically. Category 7 has<br />
several types of content which characterize other categories, e.g.<br />
'distance' and 'emotional effect').<br />
3) The categories are not exhaustive. It is, for example, evident that there<br />
is sound symbolism <strong>in</strong> the semantic spheres of 'wetness' and 'dryness'.<br />
The categories suggested <strong>in</strong> this thesis (see chapter 2) agree on some<br />
po<strong>in</strong>ts with those of Jespersen though they are a bit more detailed. The<br />
category 'th<strong>in</strong>gs and appearances' takes up examples which would be<br />
classified as 'form' and 'light' by me. The class of 'pejoratives' of the<br />
present thesis falls under Jespersen's 'states of m<strong>in</strong>d'. The category<br />
'orig<strong>in</strong>ator of the sound' is just a normal extension of mean<strong>in</strong>g from Act<br />
to Actor, (e. g. sökande (pres. part). – sökande (noun), (see Malmgren,<br />
1988), which also applies to words which are not sound symbolic <strong>in</strong> any<br />
way. Jespersen's last category does not have a correspondence <strong>in</strong> this<br />
thesis but is <strong>in</strong>terest<strong>in</strong>g and deals with, e.g. ‘‘the emotional value of some<br />
'mouth-fill<strong>in</strong>g' words’’, e.g. evig – ev<strong>in</strong>derlig (eternal), vex – aggravate<br />
and slang words like splendidious for splendid.<br />
None of the authors below has tried to relate the categories which they<br />
have found to each other, <strong>in</strong> order to expla<strong>in</strong>, other than partially, the<br />
phenomena of sound symbolism. In contrast, one important aim of this<br />
thesis is to give an explanatory model for the semantic categories that<br />
reoccur <strong>in</strong> sound symbolism <strong>in</strong> Swedish.<br />
1.10.4 Results - data from different authors<br />
The results presented <strong>in</strong> this section are primarily second-hand data, a<br />
large part of them from the authors discussed <strong>in</strong> 1.1-1.8. These data<br />
arrived at by earlier l<strong>in</strong>guists will be summarized, without any evaluation.<br />
Therefore, this section is a guide to what sounds and mean<strong>in</strong>gs different<br />
l<strong>in</strong>guists have studied. Comparisons between the authors are difficult for<br />
several reasons; e.g. Bol<strong>in</strong>ger claims there is an unclear boundary<br />
between neutral and affective morphemes, Wescott <strong>in</strong>cludes the whole of<br />
allolanguage, others have studied contrast, etc.<br />
None of the authors has claimed to give a complete <strong>in</strong>ventory of sound<br />
symbolism, usually they are e.g. simply illustrat<strong>in</strong>g an argument.<br />
31
Sometimes they just mention sound symbolic contents without giv<strong>in</strong>g<br />
examples of the relevant sounds or sound comb<strong>in</strong>ations.<br />
Table 1 Phonesthemic sounds and mean<strong>in</strong>gs of different authors.<br />
Examples are given <strong>in</strong> regular spell<strong>in</strong>g except for some IPA-symbols. The<br />
language described is English if noth<strong>in</strong>g else is written.<br />
Bol<strong>in</strong>ger (1950):<br />
sound content<br />
gl- phenomena<br />
light<br />
of<br />
fl- phenomena<br />
movement<br />
of<br />
itr <strong>in</strong>termittent<br />
ow steady<br />
Er <strong>in</strong>tense<br />
kr- bent<br />
-amble locomotion<br />
-ust surface formation<br />
-usty old<br />
-lessness <strong>in</strong>difference<br />
i dim<strong>in</strong>utive<br />
-utter discont<strong>in</strong>uity<br />
-ash hit, fragments<br />
tw- twist<strong>in</strong>g motion<br />
st- arrest<br />
sp-t rush of liquid<br />
str-p l<strong>in</strong>e hav<strong>in</strong>g<br />
st-nt<br />
breadth<br />
piece of<br />
performance<br />
s k swift movement<br />
-ump awkward, heavy<br />
Bloomfield (1933):<br />
sound content<br />
32<br />
fl- mov<strong>in</strong>g light<br />
fl- movement <strong>in</strong> air<br />
gl- unmov<strong>in</strong>g light<br />
sl- smoothly wet<br />
k r - noisy impact<br />
skr- grat<strong>in</strong>g impact or<br />
sound<br />
sn- breath-noise<br />
s n - quick separation<br />
snor<br />
movement<br />
creep<br />
dJ- up-and-downmovement<br />
b - dull impact<br />
-E violent movement<br />
-E´ big light or noise<br />
-awns quick movement<br />
-im (´ ) small light or<br />
noise<br />
-Um clumsy<br />
-Et(´ ) particled<br />
movement<br />
Humboldt (1836/1907, German)<br />
sound content<br />
st- firmness<br />
n - sharp cutt<strong>in</strong>g<br />
w - random movement<br />
u hollow and dark
Rhodes (1994):<br />
Classifiers:<br />
sound content<br />
st- 1 dimensional (stick, staff, stem)<br />
str- 1 dimensional<br />
flexible<br />
33<br />
(str<strong>in</strong>g, strand, strip)<br />
fl- 2 dimensional (flap, flat, floor)<br />
S/sk - 2 dimensional, flexible (sheet, scarf)<br />
n - 3 dimensional (knob, knot, node, nut)<br />
sp- cyl<strong>in</strong>drical (spool, sp<strong>in</strong>e, spike)<br />
dr-/tr- liquid (dr<strong>in</strong>k, dra<strong>in</strong>, trickle, trough)<br />
Paths:<br />
sound content<br />
tr-/dr- simple (track, trip, drive, drag)<br />
p-/b- ‘‘anchored’’ (push, pop, bump, bounce)<br />
j-/tS - short (jerk, jiggle, jagged, chop)<br />
w - back and forth (wag, wiggle, wobble)<br />
p - abrupt onset (pop, p<strong>in</strong>g, peep)<br />
b - abrupt, loud onset (boom, bang, beep)<br />
b l - loud, air-<strong>in</strong>duced sound (blat, blast, blab)<br />
kl- abrupt onset (clank, click, clip, clop)<br />
r - irregular onset (rip, roar, roll)<br />
y loud, vocal tract noise (yell, yap, yak)<br />
Q- low pitch, slow onset (thump, thud)<br />
p l - abrupt onset (pl<strong>in</strong>k, plop,<br />
plunk)<br />
k r - abrupt onset (creak, crack, crunch)<br />
tS- irregular onset (chirp, cheep, chatter)<br />
w- poorly resolvable onset (whiz, whack, wham)<br />
z- poorly resolvable onset (zip, z<strong>in</strong>g, zap, zoom)<br />
dr- liquid (drip, dra<strong>in</strong>, drop, drizzle)<br />
sl- liquid (slop, slush)<br />
fl- liquid (flow, flush, flood)<br />
m - liquid (mud, mush, mire, marsh)
Plato:<br />
(sounds and mean<strong>in</strong>gs discussed<br />
<strong>in</strong> Cratylos, on Greek))<br />
form content<br />
r movement<br />
i all f<strong>in</strong>eness (it<br />
can penetrate<br />
ph, ps, s, z<br />
everyth<strong>in</strong>g)<br />
everyth<strong>in</strong>g<br />
similar to<br />
airstreams<br />
d, t b<strong>in</strong>d<strong>in</strong>g and<br />
stand<strong>in</strong>g still<br />
l glid<strong>in</strong>g<br />
movement<br />
gl someth<strong>in</strong>g<br />
sticky<br />
n - <strong>in</strong>side (it is<br />
pronounced<br />
<strong>in</strong>side the<br />
mouth)<br />
a size (largeness?)<br />
e (eta) length (long?)<br />
o roundness<br />
Sapir (1921):<br />
sound process content<br />
reduplication distribution,<br />
plurality,<br />
repetition,<br />
customary<br />
activity,<br />
<strong>in</strong>crease of size,<br />
added <strong>in</strong>tensity,<br />
cont<strong>in</strong>uance<br />
34<br />
Malkiel (1978):<br />
sound process content<br />
reduplication<br />
+vowel change<br />
Jespersen (1918):<br />
disorder,<br />
confusion,<br />
rubbish, thrash<br />
sound content<br />
m roundness<br />
Jespersen (1922 a):<br />
no sounds content<br />
direct imitation<br />
orig<strong>in</strong>ator of<br />
the sound<br />
movement<br />
th<strong>in</strong>gs and<br />
appearances<br />
states of m<strong>in</strong>d<br />
size and<br />
distance<br />
length and<br />
strength of<br />
words and<br />
sounds<br />
Wescott (1975): (content is often<br />
not mentioned)<br />
sound content<br />
u w emotive<br />
z ‘‘an unusual<br />
semantic<br />
function’’<br />
‘‘sound<br />
repetition’’
‘‘sound<br />
alternation’’<br />
‘‘allol<strong>in</strong>guistic<br />
prefix’’<br />
‘‘pentestheme’’<br />
Householder (1946)<br />
no sounds content<br />
dislike<br />
destruction<br />
projection<br />
protuberance<br />
(short and<br />
roundish)<br />
collectives: heap<br />
or pile,<br />
cluster or knot,<br />
large<br />
shapeless piece,<br />
<strong>in</strong>def<strong>in</strong>ite<br />
number or<br />
amount<br />
35<br />
thick, coarse,<br />
soft substance<br />
dull, loud,<br />
<strong>in</strong>dist<strong>in</strong>ct noise<br />
coward, failure<br />
or deceiv<strong>in</strong>g<br />
past tense, past<br />
participle<br />
deficient <strong>in</strong><br />
some desirable<br />
quality<br />
Sigurd, B. (1965) (Swedish):<br />
sound mean<strong>in</strong>g<br />
fj- pejorative<br />
fn- pejorative<br />
pj- pejorative<br />
-ms pejorative<br />
-mp pejorative<br />
-sk pejorative<br />
-b(e)l pejorative<br />
-m(e)l pejorative<br />
Some generalizations<br />
Some phonesthemes are the same <strong>in</strong> English and Swedish, e.g. fl-<br />
‘‘phenomena of movement’’, e.g. flicker, flutter; fladdra, flaxa, while<br />
others are different <strong>in</strong> English and Swedish, e.g. English: fl-‘‘mov<strong>in</strong>g<br />
light’’ e.g. flicker, gl- ‘‘unmov<strong>in</strong>g light’’, e.g. gleam, and Swedish bl-,<br />
gn-, e.g. blänka, gnistra ‘‘light’’ (not to mention the discrepancies<br />
between different analysts of English). Therefore, there is not complete<br />
universality of expression on the phoneme level, s<strong>in</strong>ce for example<br />
English does not use the cluster gn- for light phenomena (it is not even a<br />
consonant cluster <strong>in</strong> English).
There are sound symbolic contents <strong>in</strong> some languages which do not occur<br />
<strong>in</strong> Swedish, e.g. <strong>in</strong> the field of color. This <strong>in</strong>dicates that all sound<br />
symbolic contents are not the same <strong>in</strong> all languages. On the other hand,<br />
size (dim<strong>in</strong>utive) stands out for itself s<strong>in</strong>ce it seems to occur <strong>in</strong> almost all<br />
languages – and <strong>in</strong> a similar phonetic form (Ultan, 1978).<br />
The categories on the content side of the lists above can be summarized as<br />
belong<strong>in</strong>g to the follow<strong>in</strong>g semantic fields. This is one possible<br />
classification, ma<strong>in</strong>ly with a departure <strong>in</strong> the senses:<br />
Hear<strong>in</strong>g<br />
sound, noise (dull, loud, <strong>in</strong>dist<strong>in</strong>ct, big, small, rush<strong>in</strong>g liquid)<br />
Vision<br />
light (small, big, mov<strong>in</strong>g, dark)<br />
Touch<br />
surface structure or substance (thick, coarse, soft)<br />
Movements<br />
(random, twist<strong>in</strong>g, swift, locomotion, up and down, quick, violent; or<br />
stop)<br />
Forms<br />
(bent, l<strong>in</strong>e hav<strong>in</strong>g breadth, projection or protuberance, round, hollow)<br />
M<strong>in</strong>d<br />
attitude, emotive (<strong>in</strong>difference, dislike)<br />
Pejorative<br />
(old, awkward, heavy, coward, failure, deceiv<strong>in</strong>g, deficient, clumsy,<br />
disorder, confusion, rubbish, thrash)<br />
Size<br />
dim<strong>in</strong>utive<br />
augmentative (<strong>in</strong>crease <strong>in</strong> size, added <strong>in</strong>tensity)<br />
Number<br />
collectives (heap or pile, cluster or knot, <strong>in</strong>def<strong>in</strong>ite number)<br />
<strong>in</strong>termittent (distribution <strong>in</strong> space, plurality, repetition, discont<strong>in</strong>uity)<br />
Various<br />
liquid<br />
steady, firmness<br />
destruction (hit, fragments, sharp cutt<strong>in</strong>g)<br />
<strong>in</strong>side<br />
tense (past tense, past participle)<br />
piece of performance<br />
unusual semantic function<br />
36
Three of the features are perceptive: 'hear<strong>in</strong>g', 'vision' and 'touch'. Taste<br />
and smell, however, do not occur. Perception of 'movement' is often cooccurr<strong>in</strong>g<br />
with perception of sound, a contiguity relationship (cf.<br />
discussion of H<strong>in</strong>ton, Nichols and Ohala (1994) <strong>in</strong> 1.11). 'Form' is<br />
likewise, perceptually, closely connected with 'vision' and secondarily also<br />
with 'touch' (cf. Brown, 1958).<br />
The semantic category 'm<strong>in</strong>d' (attitudes and emotions) is a category that<br />
probably comes naturally and can be seen as <strong>in</strong>dexically related to<br />
expressions for dislike, etc. This may also be the connection to at least<br />
some pejorative expressions. The size categories can be seen as iconically<br />
related to speech sound (<strong>in</strong> accordance with Ohala, 1994).<br />
It seems, from this overview of the literature, that the most common<br />
semantic categories are related to the three senses hear<strong>in</strong>g, vision, touch<br />
(situated <strong>in</strong> the cortex of the human bra<strong>in</strong>, <strong>in</strong> contrast with smell and<br />
taste) or they are metaphorically, metonymically, <strong>in</strong>dexically or iconically<br />
related to the senses.<br />
1.10.5 Experimental results<br />
Sapir (1929), (who communicated with Jespersen on sound symbolism and<br />
who wrote his master's thesis on J.G. Herder's (1772) "Essay on the<br />
orig<strong>in</strong> of speech") wrote about ‘‘latent expressive symbolism’’, a type of<br />
relationship, e.g. <strong>in</strong> words like teeny and t<strong>in</strong>y, which is ‘‘directly<br />
expressive of the difference <strong>in</strong> mean<strong>in</strong>g’’. He conducted some<br />
experiments. In one of them more than 500 subjects were asked to decide<br />
which of the nonsense words mil and mal meant a large table and which<br />
meant a small table. 80% agreed that mal is better suited to the large<br />
table. An <strong>in</strong>vestigation by Bentley and Varon (1933) <strong>in</strong>dicated that [a]<br />
sounds are felt to be larger than [i] sounds <strong>in</strong> the proportion 4:1, and also<br />
that [a] is rounder and [i] is more angular <strong>in</strong> the proportion 3:1 and [a] is<br />
softer and [i] is harder <strong>in</strong> the proportion 2:1.<br />
Sapir's experiments were further developed by Newman (1933). He tested<br />
both vowels and consonants with respect to the small-large and the brightdark<br />
dimensions. His results were the follow<strong>in</strong>g: vowels agree with<br />
articulatory position (front-back:small-large); also consonants agree with<br />
37
articulatory position (labial-dental: small-large); accentuation of vowels is<br />
heavily <strong>in</strong> favor of largeness and darkness.<br />
Newman (1933) also made a study of a Thesaurus with respect to the<br />
categories of 'greatness', 'smallness', 'size' and 'littleness'. He did not,<br />
however, f<strong>in</strong>d great support for sound symbolism here.<br />
The subject of correspondence between mean<strong>in</strong>g of speech sound<br />
sequences and abstract graphic figures was <strong>in</strong>vestigated by Usnadze (1924)<br />
and Köhler (1930). Two nonsense l<strong>in</strong>e draw<strong>in</strong>gs and two nonsense words<br />
maluma and takete were presented to subjects who were asked to decide<br />
which sound matched which draw<strong>in</strong>g. The overwhelm<strong>in</strong>g majority<br />
assigned maluma to the rounded figure and takete to the angular one. This<br />
result has been shown for several other languages (Holland and<br />
Wertheimer, 1964; Davis, 1961).<br />
The longest series of experiments concerns the question of whether, and<br />
to what degree, lexical oppositions <strong>in</strong> mean<strong>in</strong>g bear any consistent lawful<br />
relationship to the symbolic properties of sounds. Tsuru and Fries (1933)<br />
<strong>in</strong>itiated match<strong>in</strong>g experiments, where verbal data <strong>in</strong> the manner of<br />
Köhler's maluma study were used. Lists of pairs of opposites <strong>in</strong> two<br />
different languages were prepared, e.g. big/small - gross/kle<strong>in</strong>, and<br />
presented orally. Subjects who knew only one of these languages were to<br />
match the correspond<strong>in</strong>g words. They succeeded with a certa<strong>in</strong>ty<br />
exceed<strong>in</strong>g chance. Other l<strong>in</strong>guists have followed try<strong>in</strong>g to elim<strong>in</strong>ate the<br />
semantic cues, e. g. by match<strong>in</strong>g two languages that are both unknown to<br />
the speakers. The results were less successful <strong>in</strong> these cases as no semantic<br />
dimension could be available to the subject.<br />
A recent cont<strong>in</strong>uation of these experiments was done by Lapolla (1994).<br />
S<strong>in</strong>ce many of the earlier results have had methodological weaknesses he<br />
started a new set of experiments, on tonal morphology <strong>in</strong> Mandar<strong>in</strong><br />
Ch<strong>in</strong>ese. The results show that there is a cross-l<strong>in</strong>guistic tendency toward<br />
associat<strong>in</strong>g acoustically acute 10 segments with 'small' category words, and<br />
acoustically grave 11 segments with 'big' category words. These results are<br />
expla<strong>in</strong>ed with reference to the ‘‘frequency code’’ theory developed by<br />
10high frequency energy<br />
11low frequency energy<br />
38
Ohala (1994). This theory implies that sound symbolism is a manifestation<br />
of a larger ethological phenomenon that is also seen <strong>in</strong> the vocal<br />
communication and certa<strong>in</strong> facial expressions 12 of other species. In<br />
<strong>in</strong>tonational communication of affect and <strong>in</strong> sound symbolic vocabulary<br />
there are sound-mean<strong>in</strong>g correlations where high F0 signifies smallness,<br />
non-threaten<strong>in</strong>g attitude, desire for goodwill of the receiver, etc, whereas<br />
low F0 conveys largeness, threat, self-confidence, and self-sufficiency.<br />
The common connections between segments (consonants and vowels) with<br />
low frequency energy 13 and largeness and segments with high frequency<br />
energy and smallness are also <strong>in</strong> accordance with the frequency code.<br />
Wilde (1958), Müller (1960) and others have studied another area of<br />
motivated expressions, namely the expressions of emotions. The present<br />
thesis will not <strong>in</strong>vestigate this area, see however Abel<strong>in</strong> and Allwood<br />
(1984). Wisseman (1954) studied the creation of onomatopoeic words<br />
from noises. Various noises were presented to the subjects, who were not<br />
able to observe how they were produced. They were then <strong>in</strong>structed to<br />
<strong>in</strong>vent or select names for the noises. The research design has been<br />
criticized, e.g. Hörmann (1979), but it seems that vowels represented the<br />
pitch and qualitative color feature of the noise: the i sound imitates a<br />
light, spiky noise, the u sound a low dark noise. The number of syllables<br />
<strong>in</strong> the <strong>in</strong>vented words was not proportional to the length of the noise; it<br />
reflected sections of the noise sequence. A noise with an abrupt beg<strong>in</strong>n<strong>in</strong>g<br />
was represented by a word beg<strong>in</strong>n<strong>in</strong>g with a voiceless plosive p, t, k. A<br />
gradually start<strong>in</strong>g noise was described by a word beg<strong>in</strong>n<strong>in</strong>g with s or ts.<br />
Also Osgood (1962) studied (universal) phonetic symbolism. Osgood's<br />
semantic differential (Osgood et al, 1957) test<strong>in</strong>g the basic dimensions of<br />
‘‘value’’, ‘‘strength’’ and ‘‘potency’’ has been widely used <strong>in</strong> test<strong>in</strong>g<br />
emotional content <strong>in</strong> words, also for phonetic symbolism. Nonsense<br />
syllables were rated on the semantic differential. The results for<br />
American and Japanese speakers were almost identical. For both groups<br />
12The smile is connected with high F2 (and higher formant frequencies) s<strong>in</strong>ce retract<strong>in</strong>g<br />
the corners of the mouth shortens the vocal tract and thus raises those frequencies.<br />
13In consonants, voiceless obstruents have higher frequency than voiced, and dental,<br />
alveolar, palatal and front velars have higher frequencies (of bursts and frication noise)<br />
than labials and back velars. Dentals have higher frequencies of formant transitions than<br />
labials. High front vowels have higher F2 and high back vowels the lowest F2.<br />
39
front consonants (e. g. p) were more pleasant than back consonants (e.g.<br />
g); high frequency sounds were associated with smallness and impotence.<br />
However, it seems that his semantic categories are too restricted and the<br />
task of plac<strong>in</strong>g words/concepts on a 1–7 scale is not really feasible.<br />
<strong>Symbolism</strong> of French vowels was studied by Chasta<strong>in</strong>g (1958). i is treated<br />
<strong>in</strong> association with acuteness, smallness, lightness, rapidity, and closeness.<br />
Also consonantal oppositions were studied by Chasta<strong>in</strong>g (1965, 1966).<br />
Stops are hard, cont<strong>in</strong>uants soft, r is rough, strong, hard, etc. <strong>in</strong> contrast<br />
to l which is smooth, weak, light-weighted, etc.<br />
Fonagy (1963) compared i and u <strong>in</strong> Hungarian. In his <strong>in</strong>vestigations of<br />
children and adults i was quicker than u for 94%, smaller for 88%,<br />
prettier for 83%, friendlier for 82%, harder for 71 %, whereas u was<br />
thicker for 98%, hollower and darker for 97%, sadder and blunter for<br />
92%, more bitter for 86% and stronger for 80% . The responses to r<br />
were that r was wild, pugnacious, manly, roll<strong>in</strong>g, harder for the<br />
overwhelm<strong>in</strong>g majority.<br />
A recent study by Sereno (1994) concerned lexical organization.<br />
Depart<strong>in</strong>g from the results of a lexical study concern<strong>in</strong>g which vowels are<br />
the most common <strong>in</strong> English verbs and nouns, a reaction time experiment<br />
was performed. In this experiment the subjects categorized nouns and<br />
verbs. The results showed that there was a connection between syntactic<br />
class and phonological form <strong>in</strong> English. Verbs, with front vowels (which<br />
are lexically most frequent) were recognized faster than verbs with back<br />
vowels, while nouns with back vowels (which are lexically more frequent)<br />
were recognized more quickly than nouns with front vowels. This<br />
connection is <strong>in</strong>dependent of the frequency of the words. It is <strong>in</strong>terest<strong>in</strong>g<br />
to note this dist<strong>in</strong>ction between front and back vowels. The dist<strong>in</strong>ction<br />
between front and back vowels is perceptually salient and often also<br />
occurr<strong>in</strong>g <strong>in</strong> size-sound symbolism.<br />
Sereno (1994) thus proposes that noun/verb-categories and front/back<br />
classification of vowels (i. e. acoustical-perceptual classification) are<br />
organization pr<strong>in</strong>ciples of the lexicon and explicitly coded.<br />
As we have seen, there has been quite a large amount of different k<strong>in</strong>ds of<br />
experimentation <strong>in</strong> this area. The tests, which have been commented upon<br />
40
above, have mostly been concerned with isolated consonants (i. e. not with<br />
consonant clusters) and vowels and, semantically, with contrast<strong>in</strong>g ends on<br />
a scale (semantic oppositions). The experiments of this thesis treat<br />
Swedish (mostly) and more specifically the semantic value (not necessarily<br />
<strong>in</strong> opposition to some other category) of <strong>in</strong>itial consonant clusters.<br />
1.11 Possible explanations of sound symbolism<br />
The different explanations of sound symbolism are of vary<strong>in</strong>g types but<br />
have often focussed on synaesthesia and proprioception and, generally, on<br />
the question of a biological or non-biological base.<br />
1.11.1 Miscellaneous explanations and<br />
proprioception<br />
Let us now further <strong>in</strong>vestigate some of the ideas concern<strong>in</strong>g the nature of<br />
the relations between mean<strong>in</strong>g and expression <strong>in</strong> sound symbolism (cf. fig.<br />
2.3). This discussion will then serve as a basis for a suggestion of an<br />
explanatory model of sound symbolism to be presented <strong>in</strong> chapter 2.<br />
Different explanations have been put forward by different l<strong>in</strong>guists. Some<br />
l<strong>in</strong>guists like Publius Nigidius Figulus, 98–45 BC (accord<strong>in</strong>g to Jespersen,<br />
1922 a, p. 396), Fónagy (1963) or Peterfalvi (1965) have speculated on<br />
the connection between articulatory movements and mean<strong>in</strong>g. Nigidius<br />
Figulus claimed that <strong>in</strong> pronounc<strong>in</strong>g vos one puts forward one's lips and<br />
sends out breath <strong>in</strong> the direction of the other person, while this is not the<br />
case with nos 14 (i.e. an <strong>in</strong>dexical relation).<br />
Peterfalvi (1965) claims that vowels articulated towards the exterior of<br />
the body are judged to be ‘‘light’’ whereas those articulated towards the<br />
<strong>in</strong>terior of the body are judged to be ‘‘dark’’ because ‘‘the further you<br />
penetrate the body, the darker it is there.’’ It seems that he is <strong>in</strong>fluenced<br />
by sense analogy, i.e. the word dark can be used for different pe r ceptual<br />
phenomena. Fonagy (1963) claims that various movements of the tongue<br />
and the jaw bear likeness to different emotions.<br />
These and similar theories connect articulation with mean<strong>in</strong>g directly and<br />
disregard the acoustic (or visual) l<strong>in</strong>k <strong>in</strong> a communicative situation. As<br />
for the area of non-verbal communication, Fonagy's and Peterfalvi's<br />
14This is of course not correct.<br />
41
‘‘gestures’’ cannot even have an observer, if one excepts that front<br />
articulations can be seen. However, there might be proprioception<br />
<strong>in</strong>volved.<br />
Bol<strong>in</strong>ger (1968) suggests the follow<strong>in</strong>g metaphor: ‘‘the digital island floats<br />
on an analog sea’’. Digital stands for consonants and vowels, which are<br />
arbitrary, analogue stands for phenomena like prosody and gestures,<br />
which are not completely arbitrary. ‘‘... now and then a bit of the<br />
analogue sea washed over the digital island’’, e.g. when i stands for<br />
smallness and o for largeness. ‘‘The size of the mouth cavity ... is matched<br />
with the mean<strong>in</strong>g’’. If there is someth<strong>in</strong>g <strong>in</strong> this explanation, a number of<br />
factors are left out, e.g. the acoustic l<strong>in</strong>k. The explanation is a bit better if<br />
one compares the frequency of acoustic energy <strong>in</strong> i with that <strong>in</strong> o, or a,<br />
which is connected with the size of the mouth cavity and other vocal tract<br />
cavities. (see e.g. Jakobson, Fant, Halle, 1957).<br />
Darw<strong>in</strong> (1872) proposes a possible explanation for sound symbolic words<br />
related to emotions, e.g. disgust (related to 'pejorative'), and for<br />
<strong>in</strong>terjections. The explanation is based on the <strong>in</strong>st<strong>in</strong>ctive contractions of<br />
facial muscles connected with a certa<strong>in</strong> emotion. This type of explanation<br />
would mean that pejoratives can be classified as <strong>in</strong>dexical (cf. discussion<br />
<strong>in</strong> chapter 2).<br />
A recent <strong>in</strong>vestigation on the topic of universal sound symbolism <strong>in</strong><br />
deictic words was done by Traunmüller (1996). He found that pairs of<br />
demonstratives <strong>in</strong> which there is a vocalic opposition have an advantage <strong>in</strong><br />
the struggle for existence <strong>in</strong> languages when F2' is higher <strong>in</strong> the proximal<br />
than <strong>in</strong> the distal form; he also found that nasals are preferred <strong>in</strong> first<br />
person pronouns while stops and other obstruents are preferred <strong>in</strong> second<br />
person pronouns. He also offers explanations that <strong>in</strong>volve aff<strong>in</strong>ities with<br />
the association of pitch with size, the proprioceptive qualities of speech<br />
sounds and oral po<strong>in</strong>t<strong>in</strong>g gestures.<br />
1.11.2 Synaesthesia<br />
Sometimes it is stated <strong>in</strong> pass<strong>in</strong>g that sound symbolism can be expla<strong>in</strong>ed by<br />
synesthesia, i.e. neurological connections between the sound side and the<br />
semantic side of a word, morpheme or phonestheme, when the mean<strong>in</strong>g is<br />
<strong>in</strong> some way connected to one of the other senses, e.g. 'sight', 'touch' or<br />
to categories perceived with several of the senses, e.g. 'form', 'surface<br />
42
structure', 'movement'. (This resembles Müller's (1861) ‘‘d<strong>in</strong>g dong<br />
theory’’.)<br />
Luria's (1977) patient, the mnemonist S. is a well known synesthete.<br />
When he heard a tone vibrat<strong>in</strong>g with 50 cycles/second at 100 dB he saw a<br />
brown stripe aga<strong>in</strong>st a darker background with red tongues at the edges.<br />
At the same time he experienced the taste of sweet-sour borsjtj. When he<br />
heard [r] he always saw a ragged l<strong>in</strong>e. Persons with synaesthesia are<br />
unusual, and the experiences vary from person to person (cf. e.g.<br />
Cytowic, 1989).<br />
Aristotle claimed that the senses were clearly separated from each other.<br />
Newton, on the other hand, tried to f<strong>in</strong>d a numerical correlation between<br />
e.g. the wave length of green light and a certa<strong>in</strong> frequency of sound.<br />
However, he could not f<strong>in</strong>d this.<br />
The problems with an explanation <strong>in</strong> terms of synaesthesia are several.<br />
Accord<strong>in</strong>g to Cytowic (1989) synaesthesia is an idiosyncratic<br />
phenomenon, which varies between the persons who experience it.<br />
Cytowic argues that ‘‘rather than be<strong>in</strong>g merely a more <strong>in</strong>tense form of<br />
metaphoric speech, one can look at cross-modal metaphor as an abstract,<br />
l<strong>in</strong>guistic derivative of the stuff of synaesthesia’’. Jakobson and Waugh<br />
(1979), however, mention different f<strong>in</strong>d<strong>in</strong>gs of tendencies of the<br />
correlation of speech sounds with colors: redness of a, yellowness and<br />
whitishness of e and i and darkness of o and u.<br />
Accord<strong>in</strong>g to Marks (1990) the synaesthetic connections do not vary too<br />
much between the experiences of different persons. For example, the<br />
frequency of the second formant of vowels can be related to black (dark)<br />
and white (light) so that [u] has a low frequency and dark color while [i]<br />
has a high frequency and light color.<br />
In order to give a more conv<strong>in</strong>c<strong>in</strong>g explanation, one would need to do it<br />
<strong>in</strong> terms of neurophysiology. For example Freud (1891) suggested that<br />
language could be represented <strong>in</strong> a ‘‘field’’ <strong>in</strong> the border area between the<br />
temporal, parietal and occipital lobes, where all properties of the<br />
perception of a th<strong>in</strong>g were connected <strong>in</strong> a network (smell, taste, visual<br />
appearance, sound, etc.)<br />
43
Consider<strong>in</strong>g the apparent similarities between the semantic categories <strong>in</strong><br />
s y n a e s t h e s i a – color, form, motion, texture, lum<strong>in</strong>escence,<br />
dynamics (but also numbers, days of the week and months of the year,<br />
moral judgements) – s e n s e a n a l o g i e s (sense analogies are<br />
l<strong>in</strong>guistic metaphors, e.g. the expressions dark tones, warm colors, cf.<br />
Abel<strong>in</strong>, 1988) and s o u n d s y m b o l i s m , i.e. sense related categories,<br />
one might hypothesize a common ground for synaesthesia on the one hand<br />
and language phenomena like sense analogies and sound symbolism on the<br />
other.<br />
This is <strong>in</strong> agreement with Geschw<strong>in</strong>d (1965, accord<strong>in</strong>g to Cytowic, 1989):<br />
language depends on stable <strong>in</strong>termodal associations, especially visualauditory<br />
and tactile-auditory. These are the most common modes of<br />
synaesthesia. The relation between sound and mean<strong>in</strong>g <strong>in</strong> sound<br />
symbolism should then stand for a neurological connection; this can be<br />
<strong>in</strong>terpreted as an <strong>in</strong>dexical relation s<strong>in</strong>ce it is not iconic or conventional.<br />
Accord<strong>in</strong>g to Cytowic (1989), there is beh<strong>in</strong>d synaesthesia a purely<br />
neurological process which is connected to certa<strong>in</strong> cell clusters. He studied<br />
the blood flow to cortical association centra for vision-hear<strong>in</strong>g-touch<br />
dur<strong>in</strong>g subjects synaesthetic experiences. He expected an <strong>in</strong>crease <strong>in</strong> these<br />
association centra but did not f<strong>in</strong>d such an <strong>in</strong>crease. Instead, the<br />
connections seemed to occur <strong>in</strong> the limbic system, which was studied by<br />
<strong>in</strong>troduc<strong>in</strong>g electric stimulation dur<strong>in</strong>g operations. Cytowic argues that, <strong>in</strong><br />
some <strong>in</strong>dividuals, the developmentally earlier, ‘‘suppressed’’ limbic<br />
system sometimes overrides the cortex and that the boundaries between<br />
the senses then disappear.<br />
Related to the issue of synaesthesia and language universals is the study of<br />
verbs of perception <strong>in</strong> 50 languages by Viberg (1984). This study showed<br />
that there are implicational universals <strong>in</strong> the order:<br />
SEE>HEAR>FEEL>TASTE>SMELL. That is, if a language has only one<br />
verb of perception, the basic mean<strong>in</strong>g is 'see'. If it has two, the basic<br />
mean<strong>in</strong>gs are 'see' and 'hear' etc.<br />
1.11.3 Other neurological and biological<br />
explanations<br />
A specific, neurological explanation for why languages often represent<br />
movement with the same sort of sound symbolic forms that they use for<br />
44
the representation of non-l<strong>in</strong>guistic sounds is found <strong>in</strong> H<strong>in</strong>ton, Nichols and<br />
Ohala (1994): ‘‘ Certa<strong>in</strong> rhythmic movements often directly produce<br />
sound. But beyond that, the rhythms of sound and the rhythms of<br />
movement are so closely l<strong>in</strong>ked <strong>in</strong> the human neural system that they are<br />
virtually <strong>in</strong>separable. This is illustrated <strong>in</strong> the very natural human<br />
physical response to rhythmic music, <strong>in</strong> the forms of hand clapp<strong>in</strong>g, foot<br />
tapp<strong>in</strong>g, danc<strong>in</strong>g, rhythmical physical labor, etc. ... humans ... are also<br />
capable of the reverse : translat<strong>in</strong>g rhythmic movements <strong>in</strong>to sounds,<br />
<strong>in</strong>clud<strong>in</strong>g sound-symbolic language forms.’’ (pp. 3-4) In other words, at<br />
least part of ‘‘the sound symbolic feel<strong>in</strong>g’’ is not someth<strong>in</strong>g that one has<br />
learned.<br />
Related to the area of synaesthesia and sense analogy is the research area<br />
of bimodal perception e.g. <strong>in</strong> the form of the McGurk effect (e.g.<br />
Massaro et al 1993). Various experiments show how e.g. auditory<br />
perception is <strong>in</strong>fluenced by e.g. visual or tactile perception, and po<strong>in</strong>ts to<br />
an <strong>in</strong>terconnection of the senses. Which phonemes (or sound types) that<br />
are perceived do not, then, depend purely on acoustic <strong>in</strong>formation but also<br />
on <strong>in</strong>formation from other senses.<br />
Ohala (e.g.1994) discusses the frequency code and offers a solution for<br />
sound symbolism of dim<strong>in</strong>utives and augmentatives. In contrast with<br />
Brown (1958) his theory po<strong>in</strong>ts to <strong>in</strong>nateness through an ethologically<br />
based, phonetically plausible theory for why sound symbolism exists <strong>in</strong><br />
language. He identifies a l<strong>in</strong>k between sound symbolism <strong>in</strong> vowels,<br />
consonants, tones and <strong>in</strong>tonation. The common factor is high-low F0 (of<br />
tones and <strong>in</strong>tonation), noise frequency (of obstruents) or F2 (of vowels<br />
and sonorant consonants). High frequency is connected to smallness, low<br />
frequency to largeness. However, <strong>in</strong> Swedish there are several examples<br />
of <strong>in</strong>itial [p] usually denot<strong>in</strong>g smallness <strong>in</strong> spite of be<strong>in</strong>g grave ([p] has<br />
noise at low frequencies), e.g. pipa, pil, pilla, pilt, p<strong>in</strong>gla, p<strong>in</strong>ne, pippi,<br />
(however followed by [i], which has F2 at a high frequency). Another<br />
possible explanation of why [p] can denote smallness is the fact that [p] has<br />
weak noise.<br />
Clark and Clark (1977) discuss different k<strong>in</strong>ds of categories, perceptual,<br />
e.g. color, shape, spatiality, cognitive categories, e.g. number, negation,<br />
evaluation, cause and effect, time, and social categories, e.g. k<strong>in</strong>ship<br />
terms, personal pronouns.<br />
45
1.11.4 Non-biological explanations<br />
A non-biological explanation of size-sound symbolism is offered by<br />
Brown (1958), who claims that associations between, e.g., sound and size<br />
are simply learned through experience. Large objects usually produce<br />
dark (low frequency) sounds when pushed or moved <strong>in</strong> certa<strong>in</strong> ways,<br />
whereas small objects produce bright (high frequency) sounds, (i. e. the<br />
relation is <strong>in</strong>dexical). Thus, universality does not have to imply<br />
<strong>in</strong>nateness. Brown (1958) also anticipates the studies on multimodal<br />
perception and writes that perception does not have to be connected to a<br />
particular receptor but is a matter of the whole body.<br />
Lakoff and Johnsson (1980), expla<strong>in</strong> part of sound symbolism with<strong>in</strong> the<br />
framework of their theory for metaphors. Underly<strong>in</strong>g sound symbolism is<br />
the ‘‘conduit metaphor’’ def<strong>in</strong><strong>in</strong>g a spatial relationship between form and<br />
content: ‘‘l<strong>in</strong>guistic expressions are conta<strong>in</strong>ers’’, and their mean<strong>in</strong>gs are<br />
the content of those conta<strong>in</strong>ers. We expect small conta<strong>in</strong>ers to have small<br />
contents, large conta<strong>in</strong>ers to have large contents. Therefore: more form is<br />
more content. Examples of this is reduplication: He ran and ran and ran,<br />
He is very, very tall or lengthen<strong>in</strong>g: He is bi-i-i-i-i-ig.<br />
Diffloth (1994) goes aga<strong>in</strong>st ma<strong>in</strong>stream work on sound symbolism with<br />
unusual data and unusual explanations. He shows data from Bahnar, a<br />
language of Vietnam, <strong>in</strong> which i expresses largeness and a expresses<br />
smallness. He claims that there is nevertheless an iconic basis for this – <strong>in</strong><br />
proprioceptive sensation. He concludes that iconicity can be both<br />
physiologically motivated and culturally relative at the same time. From<br />
this follows, he proposes, that iconic patterns, be<strong>in</strong>g language specific,<br />
must be described <strong>in</strong> the grammars of language. He criticizes current<br />
l<strong>in</strong>guistic models for not be<strong>in</strong>g able to describe a direct relation between<br />
phonetics and semantics. Also, phonetics is unable to represent the<br />
phonetic parameters needed for Bahnar, e.g. perceived size of the tongue<br />
<strong>in</strong> the oral cavity. He proposes an aesthetic component of the grammar.<br />
His discussions on a proprioceptive explanation are very <strong>in</strong>terest<strong>in</strong>g but<br />
seem isolated <strong>in</strong> comparison with the elaborate Frequency Code theory of<br />
Ohala (see 1.10.5). Nevertheless, I th<strong>in</strong>k, it is preferable to give an<br />
explanation of Bahnar’s size-sound symbolism rather than discard<strong>in</strong>g it as<br />
an exception. A multimodal model for sound symbolism should<br />
preferably <strong>in</strong>corporate the facts of Bahnar, i. e. proprioception.<br />
46
Hamano (1994) has studied palatalization <strong>in</strong> Japanese sound symbolism.<br />
Palatalization of alveolar stops and fricatives is associated with<br />
‘‘childishness’’ and ‘‘immaturity’’. He connects this to studies on language<br />
acquisition report<strong>in</strong>g palatalization as one of the universal characteristics<br />
of early stages of children's language acquisition. Palatalization is also<br />
reported as one of the commonest features of baby-talk, i.e. child directed<br />
speech (Snow and Ferguson, 1977). A possible explanation of the Swedish<br />
pejorative clusters pj-, bj-, fj- which do not fit <strong>in</strong>to a sense related<br />
pattern (see chapters 2 and 4) could be that they have another orig<strong>in</strong>, <strong>in</strong><br />
child directed speech. Consider<strong>in</strong>g Hamanos f<strong>in</strong>d<strong>in</strong>gs these clusters can be<br />
<strong>in</strong>terpreted as marked (they are unusual), rather than hav<strong>in</strong>g their orig<strong>in</strong><br />
<strong>in</strong> <strong>in</strong>terjections. Hamanos explanation is basically <strong>in</strong>dexical (causal).<br />
A study of the frequency of Swedish consonant clusters <strong>in</strong> runn<strong>in</strong>g text<br />
(Stenström, 1984) shows that the seven most <strong>in</strong>frequent <strong>in</strong>itial consonant<br />
clusters are fj-, bj-, pj-, mj-, spl-, vr-, nj- (where nj- is the most<br />
unusual). In other words, bj-, pj- and fj- are among the most <strong>in</strong>frequent<br />
clusters, also <strong>in</strong> runn<strong>in</strong>g text.<br />
A pragmatic explanation for a part of sound symbolism, namely<br />
vocatives, is given by Jacobsen (1994). In Nootka languages there seem to<br />
be two ma<strong>in</strong> sound-symbolic tendencies that shape vocative forms:<br />
saliency and brevity. Saliency means that the word conta<strong>in</strong>s a prom<strong>in</strong>ent<br />
syllable that will attract the attention of the addressee. For example, nonhigh<br />
vowels are more salient because they are <strong>in</strong>tr<strong>in</strong>sically more <strong>in</strong>tense;<br />
fall<strong>in</strong>g pitch is argued to be more salient because it <strong>in</strong>volves a rapid<br />
change <strong>in</strong> pitch. The notion of saliency can, however, be criticized. Even<br />
if non-high vowels are acoustically more <strong>in</strong>tense, they are perceived as<br />
hav<strong>in</strong>g the same strength as high vowels. Also, a rais<strong>in</strong>g F0 can be as<br />
quick as a fall<strong>in</strong>g.<br />
The discussion of different explanations of sound symbolism is cont<strong>in</strong>ued<br />
<strong>in</strong> 8.3 where the arguments are related to the f<strong>in</strong>d<strong>in</strong>gs of the lexical and<br />
experimental studies of this thesis.<br />
47
2 Theoretical framework<br />
2.1 General considerations<br />
The preced<strong>in</strong>g chapter was an overview of term<strong>in</strong>ological issues and<br />
general questions concern<strong>in</strong>g onomatopoeia and other sound symbolism.<br />
This chapter will penetrate further <strong>in</strong>to the relation between mean<strong>in</strong>g and<br />
expression <strong>in</strong> sound symbolism. There will be an attempt at an explanation<br />
of the relation between expression and mean<strong>in</strong>g, with the aid of the<br />
concepts 'icon' and '<strong>in</strong>dex'.<br />
In this thesis the standpo<strong>in</strong>t is that language <strong>in</strong> important ways is a<br />
psychological phenomenon. The semantics of lexical analysis is therefore<br />
not concerned with referents, but with reference: correspondence between<br />
l<strong>in</strong>guistic form and external data. For a discussion on this subject, see<br />
Allwood and Andersson (1976).<br />
The validity of the model presented, for example concern<strong>in</strong>g the<br />
explanations or the number and types of semantic categories provided, will<br />
be supported <strong>in</strong> terms of coherence, e.g. recurrent relations between<br />
expression and content <strong>in</strong> the analyses, and by consensus, between subjects<br />
<strong>in</strong> the experiments.<br />
2.2 Static-dynamic, conventionality and arbitrar<strong>in</strong>ess<br />
The vocabulary of language is both static and dynamic (<strong>in</strong> the sense of<br />
undergo<strong>in</strong>g change). Some parts are less static than other parts, e.g. sound<br />
symbolic words can be less static <strong>in</strong> both a diachronic and a microgenetic<br />
perspective (i.e., dur<strong>in</strong>g the development of language <strong>in</strong> a situation) than<br />
more arbitrary words are, s<strong>in</strong>ce sound symbolic words can easily be<br />
created. The phonesthemes, however, are stable and sound symbolic<br />
words are created with the assistance of phonesthemes.<br />
Thus, onomatopoeic or other sound symbolic neologisms come and go, but<br />
the phonesthemes, out of which neologisms can be created, are stable over<br />
49
longer periods of time. An example (concern<strong>in</strong>g the pejorative fj-) from<br />
SAOB (Ordbok över svenska språket, utgiven av Svenska Akademien) is<br />
the word fjåka (now, if exist<strong>in</strong>g at all, only <strong>in</strong> some dialects) which meant<br />
'våp, narr, tok' (silly or crazy person). The first written <strong>in</strong>stance is from<br />
1732; there are similar words with similar mean<strong>in</strong>gs <strong>in</strong> Norwegian dialects.<br />
The orig<strong>in</strong> of the word is described as "unclear". In other words, fjåka is no<br />
longer exist<strong>in</strong>g <strong>in</strong> standard Swedish, but the phonestheme fj- 'pejorative'<br />
seems to have been the same <strong>in</strong> 1732 as today; the phonesthemes are more<br />
stable than the words. Another more recent example is the word slobb. It<br />
was created for a newspaper article (around 1985) and describes the sound<br />
of porridge land<strong>in</strong>g <strong>in</strong> a bowl. The same expression may be created aga<strong>in</strong> <strong>in</strong><br />
another context. One phonestheme used is the phonestheme sl- 'wetness'.<br />
In words, the connection between expression and content can be described<br />
as arbitrary or motivated. The motivated words can be + or - conventional;<br />
neologisms are -conventional. In reality a word cannot be completely nonconventional<br />
because the phonesthemes of sound symbolic neologisms are<br />
not. As discussed <strong>in</strong> connection with the model shown <strong>in</strong> figure 1.3,<br />
language specific expressions are created by convention while expressions<br />
emanat<strong>in</strong>g from <strong>in</strong>nate (universal) processes are affected by convention.<br />
arbitrary door<br />
+conventional -conventional<br />
motivated shriek iiiiik<br />
The field of <strong>in</strong>terest <strong>in</strong> this thesis consists of the motivated expressions,<br />
both + and - conventional. Arbitrar<strong>in</strong>ess is a precondition for<br />
conventionalization. Arbitrary words are thus always conventional while<br />
motivated words can be conventionalized or not. In language there can not<br />
be non-conventional arbitrary words. Somebody can <strong>in</strong>vent words that fit<br />
50
<strong>in</strong>to this square but nobody would understand them. The present claim is<br />
that the non-conventional can be understood if it is motivated. In some<br />
cases non-conventional words can be said to be constructed out of<br />
conventional phonesthemes, e.g. the expression fnölp is constructed for<br />
someth<strong>in</strong>g that is 'silly' (i.e. 'pejorative').<br />
The perspective on words <strong>in</strong> this thesis is synchronic, not diachronic. This<br />
is also the case as concerns semantic relations and extension of mean<strong>in</strong>g.<br />
Language is not static, but <strong>in</strong> the microperspective static enough.<br />
2.3 Semantic analysis<br />
2.3.1 Conceptions of mean<strong>in</strong>g<br />
Mean<strong>in</strong>g is more or less context dependent, especially so <strong>in</strong> the reference of<br />
a lexeme. The mean<strong>in</strong>gs of concrete nouns like horse or spoon are less<br />
context dependent than pronouns like he or it.<br />
Lexemes can be claimed to have a central, <strong>in</strong>tensional mean<strong>in</strong>g. It is<br />
possible to describe this mean<strong>in</strong>g as prototypes, mean<strong>in</strong>g components or<br />
dist<strong>in</strong>ctive features. One can also make a dist<strong>in</strong>ction between core mean<strong>in</strong>g<br />
and the vaguer emotive and associative mean<strong>in</strong>gs and sometimes between<br />
denotation and connotation. Another conception, proposed by Allwood<br />
(1989) is 'mean<strong>in</strong>g potential'. This means, briefly, that e.g. a word form has<br />
a union of potential mean<strong>in</strong>gs, one of which is decided on by different<br />
contextual conditions. Homonymy only has to be resorted to <strong>in</strong> word<br />
classification where the different mean<strong>in</strong>gs are not relatable via semantic<br />
relations like metonymy or metaphor.<br />
The concept of mean<strong>in</strong>g potential is useful also <strong>in</strong> the analysis of<br />
phonesthemes, and will be used <strong>in</strong> this thesis. A certa<strong>in</strong> consonant cluster<br />
can have one or several mean<strong>in</strong>gs. Where there are several mean<strong>in</strong>gs many<br />
of these are usually relatable to each other, e.g. light – form lett<strong>in</strong>g through<br />
light – look<strong>in</strong>g – gaze – reflect<strong>in</strong>g smooth surface – movement on such a<br />
surface (gl-), which all can be characterized as metonymical (<strong>in</strong>dexical).<br />
Other examples are sound – wetness – adhesion – pejorative (kl-) and<br />
51
wetness – smooth surface – quick movement – long th<strong>in</strong> form (sl-). (Some<br />
of these are described <strong>in</strong> more detail <strong>in</strong> chapter 4.6.)<br />
The mean<strong>in</strong>gs of onomatopoeic and other sound symbolic neologisms are<br />
very context sensitive. What does, for example, the word slasli (which is<br />
newly created) mean? Well, someth<strong>in</strong>g hav<strong>in</strong>g to do with wetness. If we<br />
get more context, like <strong>in</strong> the test match<strong>in</strong>g columns of neologisms, where<br />
each column has a special mean<strong>in</strong>g, (cf. table 7.8), subjects correctly pair<br />
word columns with mean<strong>in</strong>gs. The results depend on the possibility to<br />
compare the words <strong>in</strong> the different columns and compare these with the<br />
different suggested mean<strong>in</strong>gs. We get an effect from phonological and<br />
semantic context.<br />
Emotive and associative mean<strong>in</strong>gs are often not considered part of the core<br />
mean<strong>in</strong>g of words. I will not adopt this standpo<strong>in</strong>t but I claim that<br />
associative/emotive mean<strong>in</strong>gs belong to the mean<strong>in</strong>g potential of words and<br />
phonesthemes. I claim that, apart from the lexicalized emotive/associative<br />
mean<strong>in</strong>gs which <strong>in</strong>dividual lexemes may have, there are also phonesthemes<br />
as parts of <strong>in</strong>dividual lexemes, e. g. sl- mean<strong>in</strong>g wetness <strong>in</strong> e. g. slaska<br />
(splash), slafsa (squelch), slabba (splash), slem (slime).<br />
2.3.2 Semantic features and semantic fields<br />
The status of different types of semantic units, i.e. how concepts are best<br />
described, has been the subject of a long debate (see Allwood and<br />
Andersson (1976), Allwood (1989), Kukkonen (1989), Lehrer (1974),<br />
Miller and Johnsson Laird (1976) and Lyons (1977)). The most important<br />
units proposed are essences (def<strong>in</strong>ed by necessary and sufficient<br />
conditions) discussed e.g. by Aristotle and semantic features or<br />
components, discussed by e.g. Katz and Postal (1964), Leech (1969). The<br />
semantic features can be of different k<strong>in</strong>ds, e.g. <strong>in</strong> the form of a restricted<br />
number of primitive (universal) dist<strong>in</strong>ctive features with +, – values (i. e.<br />
mean<strong>in</strong>g components) or more generally: features found empirically <strong>in</strong><br />
semantic analyses and which are possible to decompose further (i. e.<br />
semantic features or mean<strong>in</strong>g postulates). Mean<strong>in</strong>g components and<br />
semantic features have also proven to be a useful tool <strong>in</strong> systematic<br />
52
lexicology. In this thesis are used semantic features, i.e. features found<br />
empirically <strong>in</strong> the semantic analysis, but not mean<strong>in</strong>g components, <strong>in</strong> the<br />
sense of a restricted number of apriori primitive features.<br />
Semantic fields can be characterized, us<strong>in</strong>g semantic features, as consist<strong>in</strong>g<br />
of words (lexemes) with related mean<strong>in</strong>gs which have at least one common<br />
semantic feature <strong>in</strong> common and are analyzed through mean<strong>in</strong>g<br />
components or semantic features. Fields can be used to show paradigmatic<br />
relations between the words (Lehrer, 1974) or syntagmatic relations<br />
(Porzig, 1934).<br />
Examples of semantic fields relevant for this study are 'pejoratives' and<br />
'form'. The aim was not to do a semantic field analysis. Instead fields and<br />
properties of fields will be used <strong>in</strong> the discussions, e.g. sometimes the field<br />
of 'pejoratives' is discussed, sometimes the feature [PEJORATIVE] and<br />
sometimes words like fjantig (fussy) or fjollig (foolish). In this way<br />
different authors and different <strong>in</strong>formants who have given <strong>in</strong>formation<br />
related to semantic fields at different levels of abstraction could be <strong>in</strong>cluded<br />
<strong>in</strong> the discussion .<br />
One aim of this thesis is to give a psychologically valid description and the<br />
pr<strong>in</strong>ciple assumed here is that the appropriate level is the one that works<br />
best <strong>in</strong> the analysis. Dur<strong>in</strong>g data collection the root morpheme level was<br />
preferred. The mean<strong>in</strong>gs of roots were later, <strong>in</strong> the analysis, decomposed<br />
<strong>in</strong>to semantic features.<br />
S<strong>in</strong>ce some of the studies concern a number of words with similar (and<br />
vague) content, with<strong>in</strong> a restricted number of semantic categories, which<br />
are translated <strong>in</strong>to different languages, it is helpful to work with semantic<br />
fields, and semantic features. An example: when translat<strong>in</strong>g to an unknown<br />
language, and with <strong>in</strong>formants of vary<strong>in</strong>g proficiency <strong>in</strong> Swedish (or<br />
English) it is difficult to ask for word to word translations of e.g. words like<br />
oh, ah, whoops (to take some <strong>in</strong>terjections) but easier to ask for words with<br />
mean<strong>in</strong>g of positive surprise or negative mood etc.<br />
53
It can also be the case that an expression, for example for pa<strong>in</strong>, can be<br />
translated differently <strong>in</strong> different contexts. For example there is, <strong>in</strong> Ososo, a<br />
special expression of pa<strong>in</strong> for when somebody has died.<br />
Other <strong>in</strong>terjections, like atjo (for a sneeze) are easily translatable word for<br />
word, e.g. [haptsI] <strong>in</strong> Hungarian and [Itçími] <strong>in</strong> Nigerian Ososo.<br />
2.4 Basic relations between expression and content<br />
The field of <strong>in</strong>terest which was analyzed here can be def<strong>in</strong>ed as<br />
"synchronously motivated verbal signs". By this def<strong>in</strong>ition non-verbal signs<br />
such as gestures, sign-posts etc are excluded. For practical reasons the<br />
greater part of prosody such as sentence <strong>in</strong>tonation, phrase <strong>in</strong>tonation etc<br />
was also excluded.<br />
<strong>Sound</strong> symbolism can be divided <strong>in</strong>to three types:<br />
Free morphemes: 1) onomatopoeic<br />
2) expressive <strong>in</strong>terjections<br />
Non-free morphemes: 3) sound symbolic phonesthemes<br />
(these can be part of 1) and 2))<br />
The basic dist<strong>in</strong>ctions between these three groups are:<br />
1) onomatopoeia is ma<strong>in</strong>ly iconic, e.g. mjau, plask<br />
2) expressive <strong>in</strong>terjections are ma<strong>in</strong>ly <strong>in</strong>dexical, e.g. aj, atjo<br />
3) sound symbolic phonesthemes are motivated, but not so clearly iconic<br />
or <strong>in</strong>dexical, e.g. glänsa, fjanta<br />
The three groups can be characterized <strong>in</strong> more detail:<br />
1) Onomatopoeia is realized <strong>in</strong> complete free morphemes e.g. mjau, voff.<br />
Onomatopoeia is a verbal imitation of any sound. The relation between<br />
expression and content has an element of arbitrar<strong>in</strong>ess. The mean<strong>in</strong>g is a<br />
54
(sort of) sound. The expression side is more or less conventionalized, that<br />
is, adjusted to the phonology of the language <strong>in</strong> question.<br />
The expression has universal traits with a language specific superstructure<br />
which is affected by the phonology of the <strong>in</strong>dividual language. The relation<br />
between expression and content is basically iconic. (The sound of the<br />
rooster can serve as an illustration; it is (with <strong>in</strong>formants' transcriptions) <strong>in</strong><br />
Swedish kuckeliku, <strong>in</strong> German kikeriki, <strong>in</strong> Serbocroatian kukuricu,<br />
Macedonian kukurica, Italian chicchiricchi, Syrian kuckylyku, Urdu<br />
kokelongkong). Figure 2.1 illustrates onomatopoeia:<br />
sound<br />
step1: iconic relation<br />
l<strong>in</strong>guistic form<br />
(step2:<br />
l<strong>in</strong>guistic expression)<br />
conventionalization<br />
Figure 2.1 Schematic illustration 1 of onomatopoeia.<br />
2) Expressive <strong>in</strong>terjections are also realized <strong>in</strong> complete free morphemes<br />
e.g. atjo, aj, oj. The relation between expression and content is nonarbitrary.<br />
The expression can be described as stand<strong>in</strong>g <strong>in</strong> an <strong>in</strong>direct causal<br />
relation to the content which is a bodily or mental reaction. The expression<br />
is more or less conventionalized and there are universal traits <strong>in</strong> the<br />
expressions, naturally with a language specific superstructure. Its base is<br />
<strong>in</strong>dexical. Figure 2.2 can illustrate:<br />
1 L<strong>in</strong>guistic form is a phonetic form associated with a mean<strong>in</strong>g.<br />
55
emotion, bodily reaction<br />
step 1: <strong>in</strong>dexical relation<br />
expression (cry, i.e.<br />
sound)<br />
step 2: iconic relation<br />
l<strong>in</strong>guistic form<br />
step 3: conventionalization<br />
l<strong>in</strong>guistic<br />
expression<br />
Figure 2.2 Schematic illustration of expressive <strong>in</strong>terjections.<br />
That is, a bodily reaction produces a sound, e.g a sneeze or a scream. That<br />
scream etc. is then verbally imitated (like <strong>in</strong> onomatopoeia) and then<br />
conventionalized <strong>in</strong>to a word of a language.<br />
The more vivid the connection between e.g. an emotion and its verbal<br />
expression is, the stronger the causal component is. But the form of the<br />
l<strong>in</strong>guistic sign is not an icon of the emotion etc.! The causal relation is<br />
<strong>in</strong>terest<strong>in</strong>g <strong>in</strong>sofar as the expression could not be just any k<strong>in</strong>d of<br />
expression.<br />
Probably many signs are a comb<strong>in</strong>ation of iconicity, <strong>in</strong>dexicality and<br />
symbol. Atjo is a good example of a word were the <strong>in</strong>dexical relation (a<br />
natural relation) is clear. The sound of the sneeze is imitated and then<br />
conventionalized - the symbol for a sneeze differs <strong>in</strong> different languages.<br />
The word aj is not so clearly <strong>in</strong>dexical; the connection with pa<strong>in</strong> is not as<br />
obvious as the connection between the sneeze and the sound of a sneeze,<br />
even though the sound for pa<strong>in</strong> can be seen as dependent on the shape of<br />
the mouth cavity while, e.g. one is scream<strong>in</strong>g from pa<strong>in</strong>. The mouth is<br />
opened wide, caus<strong>in</strong>g an [a]-like sound <strong>in</strong> all languages (see Darw<strong>in</strong>,<br />
1872).<br />
56
3) Phonesthemes are realised systematically with<strong>in</strong> complete (traditional)<br />
morphemes, e.g. fl- <strong>in</strong> flicker, flutter. In sound symbolic phonesthemes the<br />
relation between expression and content is of a type which is experienced<br />
as motivated by the typical speaker of a language. (In a more compell<strong>in</strong>g<br />
degree than the feel<strong>in</strong>g, e.g. by English speakers that horse is a better<br />
expression than häst, for that four-legged animal.) The expression side is<br />
language specific, and conventional, but the content side has universal<br />
traits.<br />
The mean<strong>in</strong>g of a phonestheme is experienced by the speaker/listener as<br />
somehow correspond<strong>in</strong>g to the expression. The ma<strong>in</strong> problem is to describe<br />
the type of relation that exists between, e.g. fl- and 'quick or strong<br />
movement'. What the relation might be, <strong>in</strong> detail and <strong>in</strong> relation to iconicity<br />
and <strong>in</strong>dexicality, is discussed <strong>in</strong> sections 2.5 and 2.6.<br />
2.5 The nature of phonesthemes<br />
First there is the problem with onomatopoeic phonesthemes like fr- <strong>in</strong> e.g.<br />
frusta or gn- <strong>in</strong> e.g. gnägga, gnälla. In the k<strong>in</strong>d of analysis undertaken <strong>in</strong><br />
this chapter these words ought to belong to category 1, onomatopoeic<br />
words, s<strong>in</strong>ce they are iconic (even though <strong>in</strong> a word like gnägga it is<br />
conventionalized almost beyond recognition). However, I have preferred to<br />
put the onomatopoeic phonesthemes <strong>in</strong> category 3, even though they are<br />
not problematic semiotically, because they are often metaphorically<br />
connected with other phonesthemes. Gn- which is 'sound', like <strong>in</strong> (gnissla,<br />
gnälla) is metaphorically connected with gn- : 'way of talk<strong>in</strong>g' (like <strong>in</strong><br />
gnata, gnola, gnälla). Here is aga<strong>in</strong> the problem of dist<strong>in</strong>guish<strong>in</strong>g mean<strong>in</strong>gs<br />
(which also holds for other morphology) s<strong>in</strong>ce 'way of talk<strong>in</strong>g' really<br />
belongs to the same mean<strong>in</strong>g potential as 'sound' does. In this analysis,<br />
however, 'way of talk<strong>in</strong>g' is isolated s<strong>in</strong>ce it is both frequent and<br />
<strong>in</strong>terest<strong>in</strong>g.<br />
The follow<strong>in</strong>g can be stated and should be considered <strong>in</strong> a model for<br />
phonesthemes:<br />
57
1. Some sounds/sound comb<strong>in</strong>ations are (judged to be) better suited for<br />
some (types of) mean<strong>in</strong>gs, with<strong>in</strong> a given language or for many languages.<br />
2. Some mean<strong>in</strong>gs are better suited for be<strong>in</strong>g expressed with some of these<br />
sounds/sound comb<strong>in</strong>ations.<br />
Why? It isn´t plausible that the only explanation is that a number of words<br />
happen to have similar phonetic structures l<strong>in</strong>ked to mean<strong>in</strong>gs and that this<br />
"tie" lately has become productive, which would suggest that<br />
diachronically the phonesthemes would be arbitrary. In that case the<br />
mean<strong>in</strong>g dimensions (see chapter 4) would not be so few and they ought to<br />
be more difficult to relate to each other. The relatable mean<strong>in</strong>g dimensions<br />
can be structure <strong>in</strong>ternal proofs (coherence).<br />
The expression has universal traits (especially if the mean<strong>in</strong>g of the<br />
phonestheme can be related to onomatopoeia.) The content of the<br />
phonesthemes is partly universal and partly language specific because of an<br />
<strong>in</strong>teraction between e.g. <strong>in</strong>nateness on the one hand and environment on the<br />
other (see discussion on explanations of sound symbolism <strong>in</strong> 1.11)<br />
The phonesthemes that appear from the lexical analysis (chapter 4) are<br />
prelim<strong>in</strong>ary until either there have appeared neologisms that support them,<br />
the more the better, or until they have found support <strong>in</strong> tests.<br />
Figure 2.3 can illustrate the problem of expla<strong>in</strong><strong>in</strong>g other sound symbolism,<br />
i.e., phonesthemes which are neither iconic or <strong>in</strong>dexical at first sight.<br />
mean<strong>in</strong>g<br />
l<strong>in</strong>guistic expression<br />
?-relation<br />
Figure 2.3 Schematic illustration of other sound symbolism.<br />
58
There is a motivated relation between mean<strong>in</strong>g and form, but the question<br />
is: How should the relation illustrated by the arrow best be described?<br />
The analogous representation for a conventional, arbitrary sign would be:<br />
mean<strong>in</strong>g<br />
l<strong>in</strong>guistic expression<br />
purely conventional relation<br />
Figure 2.4 Schematic illustration of a conventional, arbitrary sign.<br />
without an arrow, show<strong>in</strong>g that the form is not motivated by the mean<strong>in</strong>g<br />
(see figure 2.4). What exactly does then the arrow represent <strong>in</strong><br />
phonesthemes? It can be stated that it represents an ord<strong>in</strong>ary speakers'<br />
<strong>in</strong>tuition that the form is motivated by the content. In section 1.11 it was<br />
shown how this <strong>in</strong>tuition can be expla<strong>in</strong>ed <strong>in</strong> different ways, and it was<br />
suggested that the relation is <strong>in</strong> fact <strong>in</strong>dexical or iconic. It can be concluded<br />
that most of the explanations discussed <strong>in</strong> 1.11 support the idea that the<br />
relation is <strong>in</strong>dexical, iconic or a comb<strong>in</strong>ation of both.<br />
2.6. Considerations for a model<br />
As concerns more specified explanations an eclectic approach to<br />
explanation of sound symbolism would be to say that most of the<br />
explanations discussed <strong>in</strong> 1.11 could be valid, either simultaneously or for<br />
different types of sound symbolism.<br />
However, some explanations seem more plausible than others. One basic<br />
dist<strong>in</strong>ction is that between <strong>in</strong>nateness and learn<strong>in</strong>g. Innateness can be of at<br />
least three different k<strong>in</strong>ds. It can mean: 1) that there are <strong>in</strong>nate abilities by<br />
nature of biological endowment, 2) that there are <strong>in</strong>nate specific<br />
59
predispositions for the ability, which develops <strong>in</strong> a certa<strong>in</strong> way over time,<br />
<strong>in</strong> <strong>in</strong>teraction with the environment, and 3) that there are <strong>in</strong>nate nonspecific<br />
predispositions, the development of which heavily relies on the<br />
environment. Learn<strong>in</strong>g is compatible with the last two types of <strong>in</strong>nateness.<br />
It seems that there are recurr<strong>in</strong>g semantic features, which also are easy to<br />
relate to each other. This suggests that a model of <strong>in</strong>nate predispositions for<br />
learn<strong>in</strong>g certa<strong>in</strong> connections between expression and content is the most<br />
plausible one. If there were no such predispositions, the semantic<br />
categories <strong>in</strong>volved <strong>in</strong> sound symbolism (if there were sound symbolism at<br />
all) would most probably be haphazard. As we will see <strong>in</strong> chapter 4 the<br />
semantic categories <strong>in</strong> sound symbolism are not unlimited but restricted to<br />
a number of types. In the analysis of sound symbolic words (the method of<br />
excerption of words is described <strong>in</strong> chapter 3) the follow<strong>in</strong>g categories,<br />
which are differentiated by their mean<strong>in</strong>g, were found:<br />
<strong>Sound</strong><br />
Movement<br />
Light<br />
Surface structure<br />
Consistency (plasticity)<br />
Wetness<br />
Dryness<br />
Attitude<br />
Slang<br />
Jocular<br />
Pejorative<br />
Mental feel<strong>in</strong>g<br />
Bodily feel<strong>in</strong>g<br />
Separation<br />
Putt<strong>in</strong>g together (convergence)<br />
Dim<strong>in</strong>utive<br />
Augmentative<br />
Form<br />
Iterative<br />
60
It is obvious that some of the categories seem to be of similar types, and<br />
they can be structured <strong>in</strong> the follow<strong>in</strong>g way (figure 2.5):<br />
61
cognition spatiality repetition<br />
sense attitude mental bodily structure size form iterative<br />
impr. feel<strong>in</strong>g feel<strong>in</strong>g<br />
sound attitude pejorative separation dim<strong>in</strong>utive augmentative<br />
movement slang putt<strong>in</strong>g together<br />
light jocular (convergence)<br />
surface<br />
structure<br />
consistency<br />
(plasticity)<br />
wetness<br />
dryness<br />
Figure 2.5 Relations between the recurr<strong>in</strong>g semantic categories <strong>in</strong> Swedish sound symbolism.
Most of the categories can also be described as related to sense impressions<br />
<strong>in</strong> a more or less direct way. The categories are either related to one sense<br />
or to several senses. The simpler cases are 'sound', related to hear<strong>in</strong>g, 'light'<br />
and 'movement' related to see<strong>in</strong>g and 'consistency' related to touch. The<br />
experience of 'surface structure', 'wetness' or 'dryness' is a comb<strong>in</strong>ation of<br />
tactile, auditive and visual experience. 'Form' <strong>in</strong>volves the senses see<strong>in</strong>g<br />
and touch<strong>in</strong>g. 'Separation' and 'putt<strong>in</strong>g together' have a more abstract<br />
relation to the expression. 'Dim<strong>in</strong>utive' and 'augmentative' are also more<br />
abstract (but can be expla<strong>in</strong>ed with reference to the frequency code, cf.<br />
discussion <strong>in</strong> 1.10.5.) as well as 'iterative' (but it is clearly iconic). F<strong>in</strong>ally<br />
there are the expressive categories of mental and bodily emotion and<br />
attitude which are cognitive categories.<br />
The categories are not mutually exclusive. E.g. 'cognition' and 'spatiality'<br />
could be comb<strong>in</strong>ed e.g. <strong>in</strong> prepositions like <strong>in</strong> and other basic relations.<br />
'Mental feel<strong>in</strong>g' and 'bodily feel<strong>in</strong>g' can comb<strong>in</strong>e. However, for the analysis<br />
of significant properties of phonesthemes the above categorisation is<br />
useful.<br />
The categories are identical to or comparable with the semantic categories<br />
listed <strong>in</strong> the end of 1.10.4, which were based on the studies of different<br />
l<strong>in</strong>guists. The only really new classes are 'separation' (e.g. spl- spl<strong>in</strong>ta,<br />
split, splits, splitter, spr- spreta, sprida) and 'putt<strong>in</strong>g together' (e.g. knknipa,<br />
knyta, knyckla, tv- tv<strong>in</strong>g, tv<strong>in</strong>na).<br />
2.6.1 Relations between the categories<br />
The onomatopoeic free morphemes or phonesthemes are <strong>in</strong> a special class<br />
s<strong>in</strong>ce the iconicity concerns language sounds imitat<strong>in</strong>g sounds. Almost all<br />
the <strong>in</strong>itial consonant clusters correspond to an onomatopoeic phonestheme.<br />
This means that this category is co-occurr<strong>in</strong>g with many of the other<br />
categories, e.g. many consonant clusters can be both 'onomatopoeic' and<br />
'pejorative', however not necessarily <strong>in</strong> the same root morpheme (see<br />
further discussion <strong>in</strong> chapter 4).<br />
63
One semantic category, 'light', can be classified as a metaphorical extension<br />
from sound phenomena. The basis for extension is sense analogy - an<br />
iconic phenomenon based on some sort of similarity, e.g. the word warm<br />
can be used not only for temperature but also for colour ('light') and tones<br />
('sound') i. e. us<strong>in</strong>g phonesthemes of one sensory modality for words of<br />
another sensory modality. (cf Abel<strong>in</strong> 1988). (This is not the same as<br />
synesthesia - a neurological, <strong>in</strong>dexical phenomenon, e.g. the experience of<br />
hear<strong>in</strong>g a certa<strong>in</strong> tone because of tast<strong>in</strong>g someth<strong>in</strong>g that is salty, cf. 1.11.2).<br />
Appeal<strong>in</strong>g to sense analogy we can expla<strong>in</strong> the category of 'light'. Sense<br />
analogy concerns conventional relations between words. These relations<br />
are there because we see some sort of similarity between a particular type<br />
of sound and a particular type of light. Underly<strong>in</strong>g sense analogy (as well<br />
as synaesthesia and sound symbolism) could be "stable <strong>in</strong>termodal<br />
connections". An example of sense analogy with sound as a po<strong>in</strong>t of<br />
departure is gn-, as <strong>in</strong> gnälla (squeak), which can be seen as extended to<br />
gn- <strong>in</strong> gnistra (glimmer) <strong>in</strong> a fashion similar to the mean<strong>in</strong>g extension of<br />
e.g. the words skrikig (loud), gäll (shrill), dov (dull) which orig<strong>in</strong>ally<br />
described sounds but later also colours.<br />
Many of the rema<strong>in</strong><strong>in</strong>g categories can be expla<strong>in</strong>ed satisfactorily with the<br />
metaphoric and metonymical model of Lakoff and Johnson (1980) and<br />
Lakoff (1987). <strong>Sound</strong> symbolism can be seen as depend<strong>in</strong>g on<br />
1) <strong>in</strong>nate capacities for metaphor (similarity, icon), especially<br />
personification, e.g. gr- 'negative mood' and for metonymy (contiguity,<br />
<strong>in</strong>dex) i. e. closeness <strong>in</strong> time and space, e.g. gl- denotes both 'light' and<br />
'form lett<strong>in</strong>g through light').<br />
and on<br />
2) the result of learn<strong>in</strong>g from one's senses through use of the <strong>in</strong>nate<br />
capacities, (e. g. how someth<strong>in</strong>g sounds when mov<strong>in</strong>g <strong>in</strong> a certa<strong>in</strong> way)<br />
where the <strong>in</strong>nate capacity for metaphor is put to use.<br />
This model seems to <strong>in</strong>corporate the ideas of both Cytowic and Brown, as<br />
described above.<br />
64
Lakoff (1987, p. 8) writes "human categorization is essentially a matter of<br />
both human experience and imag<strong>in</strong>ation - of perception, motor activity and<br />
culture on the one hand, and of metaphor, metonymy and mental imagery<br />
on the other." This would also expla<strong>in</strong> why sound symbolism is partly<br />
universal, partly language specific. Universality from the semantic<br />
perspective depends among other th<strong>in</strong>gs on human capacity for metaphor<br />
and metonymy; experience is shaped by metaphor and metonymy.<br />
Lakoff and Johnsson (1980) write: "the most obvious ontological<br />
metaphors are those where the physical object is further specified as be<strong>in</strong>g<br />
a person." How does this apply to sound symbolism? <strong>Sound</strong>s of language<br />
naturally cannot be seen as persons, but they can be seen (heard) as<br />
belong<strong>in</strong>g to persons of a certa<strong>in</strong> type, mood etc. Do<strong>in</strong>g this is a k<strong>in</strong>d of<br />
personification, which is a basic conceptual strategy. The material this<br />
personification is work<strong>in</strong>g on is based on our experience of how people<br />
sound when of a certa<strong>in</strong> type, <strong>in</strong> a certa<strong>in</strong> mood etc. This personification<br />
could expla<strong>in</strong> the 'negative mood' apply<strong>in</strong>g to persons.<br />
The categories Movement, Form, Consistency, Surface structure, Wetness,<br />
Dryness and Dim<strong>in</strong>utive are simply expla<strong>in</strong>ed through metonymy<br />
(closeness <strong>in</strong> time and space) between sound and manipulation of objects<br />
dur<strong>in</strong>g the language learn<strong>in</strong>g process. Already Brown (1958) said that what<br />
was mysteriously called physiognomic (by Werner, 1953) is a result and<br />
consequence of a simple act of learn<strong>in</strong>g. In everyday experience there is a<br />
correlation between certa<strong>in</strong> physical attributes of objects and the noises<br />
produced by these objects. A large object is more likely to produce a deep<br />
sound than a small object when it is pushed or when it falls. The<br />
relationship between size and deep sounds can therefore be learned; equally<br />
a connection between sharpness and high frequency and between roundness<br />
or bluntness and a low frequency, cf. however Ohala (1994).<br />
Werner (1953) <strong>in</strong>troduced the concept of physiognomy. Accord<strong>in</strong>g to him,<br />
<strong>in</strong> the primitive perception of animals, children etc. th<strong>in</strong>gs are not treated<br />
objectively, but physiognomically, i. e. as if they expressed an <strong>in</strong>ner life<br />
and had a m<strong>in</strong>d; a landscape is cheerful, a cup on its side tired etc. Werner,<br />
<strong>in</strong> contrast to Lakoff and Johnsson (1980) treated this as an exception.<br />
65
Werner and Kaplan (1963) developed a genetic theory for symbol<br />
formation accord<strong>in</strong>g to which objects are expressive, e. g. two headlights<br />
look like eyes, music can be threaten<strong>in</strong>g. This transcendence of expressive<br />
qualities does not only form the basis for analogies, metaphors and similes,<br />
it is also the basis for see<strong>in</strong>g similarities <strong>in</strong> such unrelated th<strong>in</strong>gs as rounded<br />
shapes <strong>in</strong> a draw<strong>in</strong>g and the word "maluma".<br />
Lakoff and Johnsson's theory is <strong>in</strong> agreement with both Brown, Werner and<br />
Kaplan, and Cytowic (i. e. sound symbolism is formed by experience of the<br />
world (learn<strong>in</strong>g) (cf. Brown, 1958) but it is also formed by predispositions<br />
of perception (Werner and Kaplan, 1963, Cytowic, 1989). This po<strong>in</strong>ts to<br />
the conclusion that sound symbolism works <strong>in</strong> much the same way as other<br />
concepts except that much more is dependent on iconicity and <strong>in</strong>dexicality.<br />
2.6.2 An explanatory model for sound symbolism<br />
From the considerations discussed above a number of dimensions for a<br />
model of sound symbolism can be extracted. The general background to the<br />
model is the framework presented <strong>in</strong> 1.9, e.g. sound symbolism is central to<br />
language, it can be <strong>in</strong>nate or conventional, it is productive and context<br />
sensitive. The present model tries to expla<strong>in</strong> the nature of the relation<br />
between mean<strong>in</strong>g and expression, "the arrow with a question-mark". Some<br />
dimensions from the model <strong>in</strong> 1.9 then appear aga<strong>in</strong>, e.g. <strong>in</strong>nateness vs.<br />
learn<strong>in</strong>g (the latter connected with conventionality). The dimensions are the<br />
follow<strong>in</strong>g:<br />
1. <strong>in</strong>nateness<br />
2. learn<strong>in</strong>g<br />
3. icon (metaphor)<br />
4. <strong>in</strong>dex (metonymy)<br />
5. conventionalization<br />
6. the sound symbolic categories:<br />
<strong>Sound</strong><br />
Movement<br />
Light<br />
Surface structure<br />
66
Consistency (plasticity)<br />
Wetness<br />
Dryness<br />
Attitude<br />
Slang<br />
Jocular<br />
Pejorative<br />
Mental feel<strong>in</strong>g<br />
Bodily feel<strong>in</strong>g<br />
Separation<br />
Putt<strong>in</strong>g together (convergence)<br />
Dim<strong>in</strong>utive<br />
Augmentative<br />
Form<br />
Iterative<br />
The sound symbolic categories concern Swedish s<strong>in</strong>ce they build on the<br />
analysis of the Swedish lexical material. In a description of another<br />
language they might be slightly different but they would be expla<strong>in</strong>able <strong>in</strong><br />
terms of icon and <strong>in</strong>dex and mostly relatable to sense impressions or<br />
emotions.<br />
<strong>in</strong>nateness<br />
(icon,<strong>in</strong>dex)<br />
learn<strong>in</strong>g<br />
sound<br />
symbolism<br />
Figure 2.6 An explanatory model for sound symbolism, which shows the<br />
cause of the motivated relation between expression and mean<strong>in</strong>g.<br />
67<br />
conventionalization<br />
phonesthemes<br />
for<br />
onomatopoeia<br />
pejoratives,<br />
surface<br />
structure,<br />
movement,<br />
form, etc.
The model <strong>in</strong> figure 2.6 tries to capture the relations between the different<br />
concepts that are fundamental to sound symbolism. It says that the relation<br />
between expression and content <strong>in</strong> sound symbolism can be caused by<br />
either <strong>in</strong>nateness or learn<strong>in</strong>g. What is predisposed can be affected by<br />
learn<strong>in</strong>g, and what is learned is depend<strong>in</strong>g on <strong>in</strong>nate categories for th<strong>in</strong>k<strong>in</strong>g,<br />
such as <strong>in</strong>dex (metonymy) or icon (metaphor). <strong>Sound</strong> symbolism is the<br />
result of both <strong>in</strong>nate capacities and learn<strong>in</strong>g. Some types of sound<br />
symbolism depend more on an <strong>in</strong>dexical relation (connected with learn<strong>in</strong>g)<br />
e.g. phonesthemes for 'movement', 'form' etc. while other types of sound<br />
symbolism depend more on an iconic relation, e.g. phonesthemes for<br />
'onomatopoeia' and perhaps 'pejoratives', but these have nevertheless been<br />
put <strong>in</strong> the same box. The model also shows that sound symbolic expression,<br />
and maybe also conent, is always conventionalized, to a greater or lesser<br />
degree. This expla<strong>in</strong>s the fact that e.g. words for a certa<strong>in</strong> animal call can<br />
differ between languages at the same times as it is motivated. The model<br />
can be seen as a static model where the processes to the left are necessary<br />
prerequisites for the existence of phonesthemes. It could also be seen as a<br />
basis for how sound symbolic expressions have emerged <strong>in</strong> <strong>in</strong>dividuals <strong>in</strong><br />
<strong>in</strong>teraction with<strong>in</strong> a language system, with the result of sound symbolic<br />
expressions <strong>in</strong> that particular language.<br />
The concepts iconicity and <strong>in</strong>dexicality can be used for describ<strong>in</strong>g the<br />
nature of the different explanations of the relation between expression and<br />
form <strong>in</strong> sound symbolism described above. The concepts iconicity and<br />
<strong>in</strong>dexicality (superord<strong>in</strong>ate to metaphor and metonymy) can also be used<br />
for describ<strong>in</strong>g the semantic relations between the semantic features<br />
reoccurr<strong>in</strong>g <strong>in</strong> sound symbolism (cf. 2.6) The relations between some of the<br />
semantic categories will be further specified <strong>in</strong> 4.6. Figure 2.6 An<br />
explanatory model for sound symbolism, which shows the cause of the<br />
motivated relation between expression and mean<strong>in</strong>g.<br />
The model <strong>in</strong> figure 2.6 tries to capture the relations between the different<br />
concepts that are fundamental to sound symbolism. It says that the relation<br />
between expression and content <strong>in</strong> sound symbolism can be caused by<br />
either <strong>in</strong>nateness or learn<strong>in</strong>g. What is predisposed can be affected by<br />
68
learn<strong>in</strong>g, and what is learned is depend<strong>in</strong>g on <strong>in</strong>nate categories for th<strong>in</strong>k<strong>in</strong>g,<br />
such as <strong>in</strong>dex (metonymy) or icon (metaphor). <strong>Sound</strong> symbolism is the<br />
result of both <strong>in</strong>nate capacities and learn<strong>in</strong>g. Some types of sound<br />
symbolism depend more on an <strong>in</strong>dexical relation (connected with learn<strong>in</strong>g)<br />
e.g. phonesthemes for 'movement', 'form' etc. while other types of sound<br />
symbolism depend more on an iconic relation, e.g. phonesthemes for<br />
'onomatopoeia' and perhaps 'pejoratives', but these have nevertheless been<br />
put <strong>in</strong> the same box. The model also shows that sound symbolic expression,<br />
and maybe also conent, is always conventionalized, to a greater or lesser<br />
degree. This expla<strong>in</strong>s the fact that e.g. words for a certa<strong>in</strong> animal call can<br />
differ between languages at the same times as it is motivated. The model<br />
can be seen as a static model where the processes to the left are necessary<br />
prerequisites for the existence of phonesthemes. It could also be seen as a<br />
basis for how sound symbolic expressions have emerged <strong>in</strong> <strong>in</strong>dividuals <strong>in</strong><br />
<strong>in</strong>teraction with<strong>in</strong> a language system, with the result of sound symbolic<br />
expressions <strong>in</strong> that particular language.<br />
The concepts iconicity and <strong>in</strong>dexicality can be used for describ<strong>in</strong>g the<br />
nature of the different explanations of the relation between expression and<br />
form <strong>in</strong> sound symbolism described above. The concepts iconicity and<br />
<strong>in</strong>dexicality (superord<strong>in</strong>ate to metaphor and metonymy) can also be used<br />
for describ<strong>in</strong>g the semantic relations between the semantic features<br />
reoccurr<strong>in</strong>g <strong>in</strong> sound symbolism (cf. 2.6) The relations between some of the<br />
semantic categories will be further specified <strong>in</strong> 4.6.<br />
69
3 Method<br />
The method used <strong>in</strong> this <strong>in</strong>vestigation is a comb<strong>in</strong>ation of <strong>in</strong>tuition and<br />
empirical studies (experiments). Us<strong>in</strong>g <strong>in</strong>tuition about language is an<br />
<strong>in</strong>dispensable part of all l<strong>in</strong>guistic research, as the researcher cannot avoid<br />
hav<strong>in</strong>g some knowledge of the subject. However, <strong>in</strong>tuition is not<br />
sufficient, and will therefore be supported by empirical studies.<br />
This <strong>in</strong>vestigation of sound symbolic words <strong>in</strong> language is concentrated on<br />
<strong>in</strong>itial and f<strong>in</strong>al consonant clusters <strong>in</strong> Swedish. It consists of three ma<strong>in</strong><br />
stages.<br />
1. A number of Swedish dictionaries were excerpted with the purpose of<br />
establish<strong>in</strong>g tentative phonesthemes for Swedish. The material collected<br />
was stored <strong>in</strong> spreadsheats, which could be manually searched for roots,<br />
key words and semantic features.<br />
2. Stage 2, which is a smaller part of the study, consists of partial<br />
excerptions from other languages. The excerptions were done us<strong>in</strong>g<br />
mean<strong>in</strong>gs, or, <strong>in</strong> some cases, phonesthemes as the po<strong>in</strong>t of departure.<br />
3. A number of tests were performed <strong>in</strong> order to study the tentative<br />
Swedish phonesthemes found <strong>in</strong> stage 1.<br />
Below follows an elaboration of the three stages. In each chapter there is a<br />
more detailed description of the method.<br />
3.1 Stage 1: Collection of lexical material<br />
Initial consonant clusters<br />
S<strong>in</strong>ce the study has its po<strong>in</strong>t of departure <strong>in</strong> the Swedish language it was<br />
<strong>in</strong>itiated by an excerption of sound symbolic words <strong>in</strong> Swedish. The<br />
lexical material consists of excerptions from Svenska Akademiens Ordlista<br />
(SAOL 10, The Wordlist of the Swedish Academy), a word list with some<br />
semantic <strong>in</strong>formation, and Svensk ordbok (SOB, Dictionary of Swedish),<br />
a dictionary with def<strong>in</strong>itions, exemplifications etc.<br />
71
The excerption of words from SOB was done <strong>in</strong> the follow<strong>in</strong>g way. To<br />
qualify as a candidate for be<strong>in</strong>g a possibly <strong>in</strong>terest<strong>in</strong>g word, one or two of<br />
the follow<strong>in</strong>g criteria should be met:<br />
(i) The word is clearly onomatopoeic.<br />
(ii) At least two 1 word roots with the same consonant sequence and<br />
similar mean<strong>in</strong>g can be found <strong>in</strong> the lexical material, where the likeness<br />
does not come from trivial morphological relatedness like derivation of<br />
e.g. nouns or adjectives. As an example, fjant, fjanta, fjantig are counted<br />
as forms of the same root. Kladd, kladda och kladdig are another<br />
example. In other words: compounds, derivations or words belong<strong>in</strong>g to<br />
different parts of speech are seen as different <strong>in</strong>stances of one root<br />
morpheme (conta<strong>in</strong><strong>in</strong>g a phonestheme).<br />
Through the method described above there is no absolute guarantee that<br />
all possible phonesthemes for Swedish are found. Those that are found are<br />
chosen with approximately the same criteria. The emphasis has been on an<br />
analysis of all words, rather than a detailed analysis of a few words. The<br />
purpose of this is to get an overview of the ma<strong>in</strong> traits <strong>in</strong> the Swedish<br />
lexical material.<br />
The Swedish lexicon (represented by SOB) was manually excerpted for<br />
root morphemes with onomatopoeic or other sound symbolic mean<strong>in</strong>gs,<br />
(motivated words.) Only words beg<strong>in</strong>n<strong>in</strong>g with consonant clusters were<br />
considered <strong>in</strong> this excerption.<br />
"Key words" from the SOB were also registered <strong>in</strong> the excerption. A key<br />
word is a formal representative of the sense connected to a root<br />
morpheme and usually denotes e.g. 'sound', 'shape', 'texture' (sense<br />
related categories), 'pejorative' etc.<br />
A key word is either a word <strong>in</strong> the def<strong>in</strong>ition of the actual word, or a<br />
word <strong>in</strong> the def<strong>in</strong>ition of a word which is <strong>in</strong> the def<strong>in</strong>ition of the actual<br />
word, or another word <strong>in</strong> the paraphrase of a more peripheral men<strong>in</strong>g of<br />
the actual word, or a synonym. Ideally the key word should have been <strong>in</strong><br />
the def<strong>in</strong>ition of the actual word, but unfortunately SOB is not completely<br />
1 For lexically <strong>in</strong>frequent clusters, with less than 13 roots, 1 root sufficed if it had one of<br />
the recurr<strong>in</strong>g semantic features.<br />
72
consistent <strong>in</strong> its way of giv<strong>in</strong>g def<strong>in</strong>itions. Sometimes a word is only given<br />
a very short def<strong>in</strong>ition, e.g. a synonym.<br />
The key word is always a word which is written <strong>in</strong>to the def<strong>in</strong>ition of<br />
the lexicon entry of the word under consideration and, <strong>in</strong> a most cases, a<br />
word <strong>in</strong> the def<strong>in</strong>ition of the actual word.<br />
This function of key words is to ensure a certa<strong>in</strong> amount of objectivity,<br />
through consensus, <strong>in</strong> the selection procedure. It depends then on earlier<br />
semantic analyses by several persons. Key words are sound symbolic<br />
categories or hyponyms to these. They can be found <strong>in</strong> Appendix 1.<br />
Examples of key words from SOB are: ljud (sound), rörelse (movement),<br />
spetsig (po<strong>in</strong>ted), ljus (light), äcklig (nasty), vatten (water), tjock (thick),<br />
klibbig, (sticky), ohyfsad (rude), löjlig (ridiculous). Naturally there were<br />
expectations as to what semantic categories would be <strong>in</strong>terest<strong>in</strong>g; the key<br />
words are related to the senses, to 'form', 'mental feel<strong>in</strong>g' etc. Words<br />
with def<strong>in</strong>itions conta<strong>in</strong><strong>in</strong>g appropriate key words were, however, not<br />
considered if a connection between expression and mean<strong>in</strong>g was not felt,<br />
us<strong>in</strong>g l<strong>in</strong>guistic <strong>in</strong>tuition, as e.g. for skam. No part of the expression side<br />
of this word seems to mirror the words oförmåga or förkastlig of the<br />
def<strong>in</strong>ition. The opposite is also the case; a connection seemed to be present<br />
but an appropriate key word was not to be found <strong>in</strong> the def<strong>in</strong>ition. 2<br />
Some words which can be found <strong>in</strong> SAOL 10 are described as colloquial,<br />
while they <strong>in</strong> fact are out of date. Nevertheless they are not difficult to get<br />
a vague understand<strong>in</strong>g of. Examples are words like pjask, pjasker and<br />
pjalt 3 that have a clearly pejorative r<strong>in</strong>g. These and the words of the<br />
previous paragraph could probably easily be (mis)<strong>in</strong>terpreted <strong>in</strong> an<br />
appropriate context!<br />
With an attitude of accept<strong>in</strong>g doubtful words rather than discard<strong>in</strong>g them,<br />
about 1,000 words – <strong>in</strong> the sense of root morphemes – were registered<br />
for further analysis. Accord<strong>in</strong>g to Nusvensk frekvensordbok 4 (Allén et<br />
2 There are also words that, from the po<strong>in</strong>t of view of sound, would seem to be<br />
appropriate for mirror<strong>in</strong>g a certa<strong>in</strong> content but do not. Examples are fläns (part of a<br />
mach<strong>in</strong>e), skralna (to beg<strong>in</strong> to head, of w<strong>in</strong>d; sail<strong>in</strong>g term), and slubb (sort of sp<strong>in</strong>n<strong>in</strong>g<br />
mach<strong>in</strong>e).<br />
3 In SAOL 12 these words are no longer <strong>in</strong>cluded.<br />
73
al, 1980) the number of lexical morphemes <strong>in</strong> Swedish amount to around<br />
8,300 (suffixes add another 562 morphemes).<br />
It is important to keep <strong>in</strong> m<strong>in</strong>d, both what concerns the data collection and<br />
the data analysis, that the borders between the different mean<strong>in</strong>gs are<br />
fuzzy and partly subjective, and the result<strong>in</strong>g diagrams and tables are to<br />
be seen as po<strong>in</strong>ts of departure for different <strong>in</strong>vestigations, experiments<br />
and discussions. The primary aim of the data collection is to get an<br />
overview of the structure of the clusters and to show patterns of forms<br />
and contents.<br />
F<strong>in</strong>al clusters<br />
F<strong>in</strong>al consonant comb<strong>in</strong>ations with recurr<strong>in</strong>g mean<strong>in</strong>gs were also<br />
excerpted from Svensk baklängesordbok (1981) (Reverse Order<br />
Dictionary of Swedish) and Nusvensk frekvensordbok (Allén et al, 1980)<br />
(Frequency Dictionary of Present-Day Swedish). They were analyzed <strong>in</strong> a<br />
fashion similar to the one described above.<br />
3.2 Stage 2: Cross-language comparisons<br />
With a focus on both the expression and the content side of phonesthemes,<br />
some comparisons with other languages were done. The emphasis was on<br />
certa<strong>in</strong> easily def<strong>in</strong>ed mean<strong>in</strong>gs. Also, some contrastive experiments of<br />
<strong>in</strong>terpretation, test<strong>in</strong>g Swedish phonesthemes on native speakers of other<br />
languages, were made.<br />
3.2.1 Cross-language thesaurus studies<br />
The lexical study consists of two studies, one <strong>in</strong> Swedish and one <strong>in</strong><br />
English. It is concentrated on the semantic fields of 'stupidity' and<br />
'surface structure' – 'rough' and 'smooth'. These fields were chosen<br />
because they were different <strong>in</strong> type, they were quite common ('stupidity'<br />
is a subclass of 'pejorative') and also quite uncomplicated semantically.<br />
There were three <strong>in</strong>formants from each language.<br />
3.2.2 Cross-language <strong>in</strong>formant studies<br />
Self-imitative <strong>in</strong>terjections<br />
The Swedish primary <strong>in</strong>terjections (see Ideforss, 1928) have been<br />
translated to 8 different languages with the help of <strong>in</strong>formants. The<br />
languages are Icelandic, English, Polish, Hungarian, F<strong>in</strong>nish, Ososo,<br />
74
Malagasi and Slovenian. The material is presented <strong>in</strong> Appendix 2 and it is<br />
grouped accord<strong>in</strong>g to type of <strong>in</strong>terjection. In some cases it was difficult to<br />
translate word for word. The semantic contents of an <strong>in</strong>terjection <strong>in</strong>stead<br />
had to be described (with semantic features) after which the <strong>in</strong>formant<br />
gave the closest correspondence <strong>in</strong> her own language.<br />
Interjections imitat<strong>in</strong>g animal sounds<br />
Swedish <strong>in</strong>terjections imitat<strong>in</strong>g animal calls were translated, with the help<br />
of <strong>in</strong>formants, <strong>in</strong>to 17 different languages: Icelandic, English, Polish,<br />
Hungarian, F<strong>in</strong>nish, Ososo, Malagasi, Slovenian, Korean, Japanese,<br />
Ch<strong>in</strong>ese, Estonian, Urdu, Persian, Kurdish, Arabic, and Spanish.<br />
Interpretation of animal sounds<br />
Sixteen of the <strong>in</strong>terjections imitat<strong>in</strong>g animal calls from different languages<br />
were chosen to test how speakers of other languages would <strong>in</strong>terpret<br />
them. The ones chosen were those whose expression sides were most<br />
different between the languages studied, <strong>in</strong> order to make the task as<br />
difficult as possible. The subjects <strong>in</strong> this test were 15 persons, with the<br />
follow<strong>in</strong>g first languages: Swedish (8 subjects) French (2 subjects),<br />
English, Hungarian, Czech, Slovenian, Ososo. All were tested on the same<br />
occasion. They listened to pronunciations of the words for the animal calls<br />
and they saw them transcribed <strong>in</strong> IPA. They were told to guess which<br />
animal had gotten its call conventionalized <strong>in</strong> each way, and to write down<br />
their answers.<br />
Cross-language experiment<br />
Furthermore, one person represent<strong>in</strong>g each of the languages: Arabic,<br />
Spanish, German, Dutch, Ibo and English, took part <strong>in</strong> an experimental<br />
study of persons who did not have Swedish as a first language. The study<br />
was conducted <strong>in</strong> much the same way as one of the experiments described<br />
below and <strong>in</strong> 7.1, (test 2.a) – a free choice test from expression to<br />
mean<strong>in</strong>g. The ma<strong>in</strong> difference was that the subjects were confronted<br />
mostly with real words. There were, however, also a few neologisms<br />
based on phonesthemes.<br />
75
3.3 Stage 3: Experiments<br />
3.3.1 Experiments with neologisms<br />
To further penetrate the status and productivity of the tentative Swedish<br />
phonesthemes of language users of today, a number of tests were<br />
performed (see Appendix 3).<br />
The tests <strong>in</strong>vestigated both production and understand<strong>in</strong>g of written<br />
neologisms, which were, <strong>in</strong> three of the tests, modelled on the tentative<br />
phonesthemes. The tests consisted of free and forced choice tasks. There<br />
was was an additional match<strong>in</strong>g test where two neologisms were to be<br />
matched with two mean<strong>in</strong>gs. The subjects were 14–15 native Swedish<br />
speakers. In one test they were <strong>in</strong>structed to freely <strong>in</strong>terpret neologisms<br />
and <strong>in</strong> another they were supposed to produce neologisms out of different<br />
concepts written on the test sheet. In the next set of tests the subjects<br />
should, <strong>in</strong> <strong>in</strong>terpret<strong>in</strong>g neologisms, chose from three alternative mean<strong>in</strong>gs<br />
written on the test sheet, or, <strong>in</strong> produc<strong>in</strong>g neologisms, choose among three<br />
alternative neologisms, also written on the test sheet. The design of the<br />
experiments is described <strong>in</strong> detail <strong>in</strong> 7.1.<br />
3.4 Further method<br />
Explanatory models for sound symbolism are also constructed, see 1.9,<br />
2.6.2. These models aim at expla<strong>in</strong><strong>in</strong>g the nature of the motivated<br />
connection between expression and mean<strong>in</strong>g <strong>in</strong> sound symbolism.<br />
Comparisons between the results of the lexical studies, the experiments –<br />
also the cross-language experiments – some of the cross language<br />
comparisons 4 ,and the models are made <strong>in</strong> chapter 8.<br />
4 Interjections and animal sounds of the cross language comparisons belong to the classes<br />
expressive <strong>in</strong>terjections and onomatopoeia which are not <strong>in</strong>cluded <strong>in</strong> the models of 1.9<br />
and 2.6.2.<br />
76
4 Analysis of the <strong>in</strong>itial consonant clusters<br />
4.1. Data analysis, a short overview<br />
All the entries of the lexicon Svensk Ordbok (1986) were exam<strong>in</strong>ed<br />
(around 65, 000 lexemes, which corresponds to approximately 8, 300 1<br />
morphemes, which is the number of morphemes <strong>in</strong> Swedish accord<strong>in</strong>g to<br />
Allén et al, 1980). Around 1,250 words (root morphemes, cf. 3.1) were<br />
judged to be possibly motivated.<br />
Of all the 37 <strong>in</strong>itial consonant clusters of Swedish all but one are used for<br />
sound symbolism. The 37 clusters are: bj-, bl-, br-, dr-, dv-, fj-, fl-,<br />
fn-, fr-, gl-, gn-, gr-, kl-, kn-, kr-, kv-, mj-, nj-, pj-, pl-, pr-,<br />
sk-, skr-, skv-, sl-, sm-, sn-, sp-, spj-, spl-, spr-, st-, str-, sv-,<br />
tr-, tv-, vr-. The unused one (dv- ) has a lexical frequency of 3<br />
morphemes. Thus, almost all clusters seem to be used, but they are used<br />
to vary<strong>in</strong>g extents and for different purposes.<br />
(Now and then one hears authentic examples of e.g. onomatopoeic<br />
expressions which use non-standard clusters (cf. Garlén (1988) like<br />
sklofsa ("walk <strong>in</strong> mud"). It is also the case that a cluster like spl-, which<br />
normally is not onomatopoeic <strong>in</strong> Swedish, is often used for imitat<strong>in</strong>g<br />
sounds, for example <strong>in</strong> comic strips – perhaps <strong>in</strong>fluenced by English. In<br />
comic strips expressions like splofs and splafs are common.)<br />
All <strong>in</strong>itial consonant clusters were analyzed closely, both those that appear<br />
to have a greater and those that seem to have a smaller amount of<br />
motivated root morphemes (cf. 3.1).<br />
For each cluster, each motivated root morpheme was classified accord<strong>in</strong>g<br />
to its motivated semantic features. (cf. 2.6). A root morpheme can have<br />
one or more motivated semantic features, e.g. skvalpa (lap, ripple, splash,<br />
spill) has 'sound', 'wetness'. Even though part of the mean<strong>in</strong>g might also<br />
be due to -alpa and not only to skv-, such possibilities were not<br />
1This number of morphemes is the result of an <strong>in</strong>vestigation of a one million word<br />
newspaper corpus (NFO 4). It is most probable that newspaper language conta<strong>in</strong>s fewer<br />
roots than spoken language, but these figures are not known at present. The estimation of<br />
65, 000 lexemes is exclud<strong>in</strong>g transparent compounds.<br />
77
considered at this stage. 2 The semantic features 'wetness' and 'sound' of<br />
skvalpa were, <strong>in</strong> this analysis, attributed to the cluster skv- because there<br />
are two or more root morphemes beg<strong>in</strong>n<strong>in</strong>g with skv- that have the same<br />
semantic features. The SOB-def<strong>in</strong>ition of skvalpa is: "vara i rörelse och<br />
därvid ge ifrån sig ett kluckande och plaskande ljud – om vatten o.d."(my<br />
italics) ("to be mov<strong>in</strong>g and thereby emit a gurgl<strong>in</strong>g and splash<strong>in</strong>g sound –<br />
of water, etc")<br />
The SOB-def<strong>in</strong>itions of other skv- lexemes are:<br />
skvimpa (splash to and fro) "skvalpa med små rörelser – ofta så att vätska<br />
spills ut" (The semantic features are the same as for skvalpa (lap with<br />
small movements – often so that liquid is spilled out), with the addition of<br />
'dim<strong>in</strong>utive' – which is probably due to the i, cf. Ultan, 1978.)<br />
skvätta (splash, squirt) "fara i väg i skvättar" (go off <strong>in</strong> squirts), skvätt:<br />
"liten mängd vätska" (small amount of liquid) The semantic features are<br />
'movement', 'dim<strong>in</strong>utive', 'wetness'.<br />
skvala (pour, gush, rush) "r<strong>in</strong>na rikligt och ljudligt (flow abundantly and<br />
noisily). R<strong>in</strong>na (flow) is def<strong>in</strong>ed as "förflytta sig nedåt (längs viss yta) i<br />
sammanhängande formlös mängd – om vätska e.d." (move downwards<br />
(along a certa<strong>in</strong> surface) <strong>in</strong> a cont<strong>in</strong>uous shapeless quantity – about liquid,<br />
etc) and riklig (abundant) is def<strong>in</strong>ed as "förekommande i stor mängd eller<br />
omfattn<strong>in</strong>g" (occurr<strong>in</strong>g <strong>in</strong> large quantity or range) and so the semantic<br />
features of skvala are 'movement', 'wetness', 'sound' and 'augmentative'.<br />
As we see the features 'movement', 'wetness' and 'sound' are all present<br />
<strong>in</strong> three of these four words. There are around eight more words<br />
(consider<strong>in</strong>g the root morphemes) beg<strong>in</strong>n<strong>in</strong>g with skv-.<br />
(The cluster -Vlpa ends the follow<strong>in</strong>g words (consider<strong>in</strong>g the root<br />
morphemes): valpa (whelp), skvalpa (lap), stolpa (walk with long paces),<br />
pulpa (pulp), hjälpa (help), stjälpa (tip over), skölpa (hollow out with a<br />
special tool) – no other root morphemes with the features 'wetness' or<br />
'sound', but 4 root morphemes with the feature 'movement'.)<br />
2 For an analysis of comb<strong>in</strong>ed effects of <strong>in</strong>itial cluster, f<strong>in</strong>al clusters and, to some extent,<br />
of vowels, see 5.6.<br />
78
Diagram 4.1 shows the number of motivated root morphemes for every<br />
cluster. The number of all root morphemes and of the motivated root<br />
morphemes for each cluster can be studied <strong>in</strong> diagrams 4.2 and 4.3. Also<br />
the ratios between the number of sound symbolic root morphemes on the<br />
one hand (irrespective of exact mean<strong>in</strong>g) and total number of root<br />
morphemes beg<strong>in</strong>n<strong>in</strong>g with a certa<strong>in</strong> cluster was calculated, see diagram<br />
4.4.<br />
In addition to this, the various semantic features (cf. 4.2.3) for every<br />
cluster and the ratios for the occurrence of different semantic features per<br />
total number of root morphemes (for every cluster) were put <strong>in</strong>to the<br />
chart. These figures are used for the follow<strong>in</strong>g diagrams. 3<br />
Diagram 4.5 shows the absolute numbers of all <strong>in</strong>itial consonant clusters<br />
for all semantic features.<br />
In diagrams 4.6 to 4.15 the distributions of separate semantic features<br />
over different clusters can be seen.<br />
Diagrams 4.16 to 4.23 show the sound symbolic profiles of different<br />
clusters, i.e. how the semantic features are differently distributed.<br />
In addition to this, the relations between the different types of mean<strong>in</strong>g<br />
with<strong>in</strong> a cluster were analyzed synchronically.<br />
4.2 Results<br />
In this section some of the results from the lexical analysis are presented.<br />
The same material will be studied from different angles and <strong>in</strong> greater or<br />
lesser detail. The focus will be either on the different consonant clusters,<br />
or on the different mean<strong>in</strong>gs. There will be general patterns and patterns<br />
for certa<strong>in</strong> clusters or mean<strong>in</strong>gs.<br />
3 If all dialectal, archaic and slang words had been counted the ratios of motivated root<br />
morphemes would have been higher for most consonant clusters. (This reflects the<br />
phenomenon of slang words float<strong>in</strong>g <strong>in</strong> and out of language at another rate, cf. 1.3 - 1.4).<br />
As an example the follow<strong>in</strong>g root morphemes, found <strong>in</strong> Svenska Akademiens ordlista<br />
över svenska språket (SAOL 10), are not <strong>in</strong>cluded: skvattra "snattra"(quack, gabble,<br />
chatter), skrålla (coll.) "löjlig damhatt" (silly lady's hat), skryp (dial.) "slösaktig"<br />
(wasteful). The ma<strong>in</strong> reason for not count<strong>in</strong>g these is that they are not covered <strong>in</strong> SOB;<br />
special studies would have to be made of dialectal and slang dictionaries. There seem to<br />
be more dialectal and archaic words <strong>in</strong> SAOL than <strong>in</strong> SOB.<br />
79
4.2.1 More and less sound symbolic clusters<br />
The 36 <strong>in</strong>itial consonant clusters are very different with respect to how<br />
many root morphemes they conta<strong>in</strong>. They also differ considerably <strong>in</strong> how<br />
many sound symbolic root morphemes there are for each cluster.<br />
Diagram 4.1 shows the number of motivated root morphemes for all 36<br />
clusters.<br />
Diagram 4.1 shows e.g. that sl- has the highest number of sound symbolic<br />
clusters, namely 83, while nj- has the fewest, namely only 1. In absolute<br />
numbers, the 6 clusters that have most motivated root morphemes are (<strong>in</strong><br />
order of descend<strong>in</strong>g frequency): sl-, sn-, kn-, kr-, kl-, sp-. Of these 6<br />
clusters 3 beg<strong>in</strong> with an s and 3 beg<strong>in</strong> with a k. There is reason to wonder<br />
if nj- should be counted as a sound symbolic cluster at all, but nj-, as well<br />
as all other clusters, must be seen <strong>in</strong> the light of how many root<br />
morphemes there are <strong>in</strong> the total vocabulary for each cluster. This<br />
comparison is done <strong>in</strong> diagram 4.2.<br />
80
sl<br />
sn<br />
kn<br />
kr<br />
kl<br />
sp<br />
st<br />
gl<br />
tr<br />
fl<br />
skr<br />
pl<br />
bl<br />
sm<br />
str<br />
gr<br />
sk<br />
sv<br />
spr<br />
pr<br />
gn<br />
kv<br />
dr<br />
vr<br />
br<br />
fr<br />
fn<br />
skv<br />
mj<br />
fj<br />
pj<br />
spl<br />
bj<br />
spj<br />
tv<br />
nj<br />
0 20 40 60 80 100<br />
Diagram. 4.1. More and less sound symbolic <strong>in</strong>itial consonant clusters. Number of<br />
motivated root morphemes per cluster.<br />
81
st<br />
pr<br />
tr<br />
kr<br />
sl<br />
sp<br />
br<br />
sn<br />
gr<br />
sk<br />
kl<br />
pl<br />
fl<br />
fr<br />
str<br />
kn<br />
dr<br />
gl<br />
sv<br />
bl<br />
skr<br />
kv<br />
sm<br />
spr<br />
fj<br />
gn<br />
tv<br />
vr<br />
bj<br />
mj<br />
skv<br />
fn<br />
spl<br />
pj<br />
spj<br />
nj<br />
0 50 100 150 200 250 300<br />
82<br />
no of motiv roots<br />
all roots<br />
Diagram. 4.2 More and less sound symbolic consonant clusters.<br />
Number of root morphemes and number of motivated root<br />
morphemes per cluster.
Diagram 4.2 shows the proportions between total number of root<br />
morphemes and the number of motivated root morphemes. The diagram<br />
shows for example that the cluster pj- conta<strong>in</strong>s both few root morphemes<br />
and few motivated root morphemes. So there is a high proportion of<br />
motivated ones. It also shows that tr- has many root morphemes, quite a<br />
few motivated root morphemes, but a lower proportion of motivated root<br />
morphemes. For percentages of the proportion motivated root<br />
morphemes/all root morphemes, see diagram 4.4 (It should be noted that<br />
one motivated root morpheme may conta<strong>in</strong> one or more motivated<br />
semantic feature.)<br />
Diagram. 4.3 shows the same <strong>in</strong>formation as diagram 4.2. but here the<br />
clusters are ranked from the most sound symbolic to the least sound<br />
symbolic, as <strong>in</strong> 4.1. The diagram shows that lexically frequent clusters do<br />
not necessarily conta<strong>in</strong> a large proportion of sound symbolic morphemes,<br />
e.g. pr-. (A large proportion of the non-motivated morphemes beg<strong>in</strong>n<strong>in</strong>g<br />
with pr- are due to loan words 4 and it is also quite difficult to count the<br />
morphemes.)<br />
Table 4.1 (related to diagrams 4.1, 4.2 and 4.3) shows the number of root<br />
morphemes and the number of motivated root morphemes for the 36<br />
clusters. The table is ordered alphabetically. The clusters are ranked from<br />
the most sound symbolic (<strong>in</strong> absolute numbers) to the least sound<br />
symbolic.<br />
4 It would be <strong>in</strong>terest<strong>in</strong>g, <strong>in</strong> further research, to study the <strong>in</strong>fluence of loan words on the<br />
frequencies of sound symbolic root morphemes per cluster. Words older than a 100<br />
years, etc. could easily be found with the aid of the mach<strong>in</strong>e readable version of SAOB<br />
(Ordbok över svenska språket, utgiven av Svenska Akademien). It could be hypothesized<br />
that there would be a higher frequency of sound symbolic root morphemes among the<br />
older root morphemes. On the other hand, root morphemes that confirm with the native<br />
pattern could be preferred <strong>in</strong> loans, cf. discussion <strong>in</strong> 1.5.)<br />
83
sl<br />
sn<br />
kn<br />
kr<br />
kl<br />
sp<br />
gl<br />
st<br />
tr<br />
fl<br />
skr<br />
pl<br />
bl<br />
sm<br />
str<br />
gr<br />
sk<br />
spr<br />
sv<br />
pr<br />
gn<br />
kv<br />
dr<br />
vr<br />
br<br />
fn<br />
fr<br />
skv<br />
mj<br />
fj<br />
pj<br />
spl<br />
bj<br />
spj<br />
tv<br />
nj<br />
0 50 100 150 200 250 300<br />
84<br />
no of motiv roots<br />
all roots<br />
Diagram 4.3. Number of root morphemes and number of motivated root<br />
morphemes per cluster, which are ordered from highest number of sound<br />
symbolic root morphemes to lowest number of sound symbolic root<br />
morphemes.
Table 4.1 Number of root morphemes and motivated root morphemes.<br />
total number of motivated root Rank<br />
root morphemes morphemes<br />
bj- 13 5 32<br />
bl- 64 33 13<br />
br- 130 11 25<br />
dr- 74 16 23<br />
fj- 24 7 29<br />
fl- 96 36 10<br />
fn- 10 10 26<br />
fr- 87 10 26<br />
gl- 69 45 7<br />
gn- 22 20 21<br />
gr- 126 26 16<br />
kl- 115 51 5<br />
kn- 77 59 3<br />
kr- 169 53 4<br />
kv- 56 17 22<br />
mj- 11 7 29<br />
nj- 3 1 36<br />
pj- 7 6 31<br />
pl- 100 34 12<br />
pr- 200 21 20<br />
sk- 116 25 17<br />
skr- 63 35 11<br />
skv- 10 9 28<br />
sl- 150 83 1<br />
sm- 55 31 14<br />
sn- 126 62 2<br />
sp- 137 47 6<br />
spj- 6 4 34<br />
spl- 9 5 32<br />
spr- 34 23 18<br />
st- 279 45 7<br />
str- 77 30 15<br />
sv- 65 23 18<br />
tr- 169 44 9<br />
tv- 20 3 35<br />
vr- 19 12 23<br />
85
Table 4.1 shows, for example, that 5 out of 13 root morphemes beg<strong>in</strong>n<strong>in</strong>g<br />
with bj- have a sound symbolic mean<strong>in</strong>g. The rema<strong>in</strong><strong>in</strong>g 8 (13-5)<br />
unmotivated root morphemes beg<strong>in</strong>n<strong>in</strong>g with bj- are:<br />
bjuda (<strong>in</strong>vite, offer, etc)<br />
bjugg (barley)<br />
bjussa (slang for bjuda)<br />
bjälke (balk)<br />
björk (birch)<br />
björkna (a fish)<br />
björn (bear)<br />
björna (demand repayment)<br />
and the 5 motivated root morphemes are:<br />
Key word<br />
bjäbba uppnosigt prat (cheeky talk)<br />
bjäfs överdriven (exaggerated)<br />
bjällra kl<strong>in</strong>gande (chim<strong>in</strong>g bell)<br />
bjärt lysande (sh<strong>in</strong><strong>in</strong>g)<br />
bjässe mycket stor (very big)<br />
(Only key words, not the complete def<strong>in</strong>itions, are given above.)<br />
4.2.2 Proportions of motivated root morphemes.<br />
Summary<br />
To sum up part of the forego<strong>in</strong>g comparisons, diagram 4.4 shows the<br />
proportions between number of sound symbolic morphemes and total<br />
number of morphemes <strong>in</strong> percentages.<br />
86
fn<br />
gn<br />
skv<br />
pj<br />
kn<br />
spr<br />
spj<br />
gl<br />
mj<br />
vr<br />
spl<br />
sm<br />
skr<br />
sl<br />
bl<br />
sn<br />
kl<br />
str<br />
fl<br />
bj<br />
sv<br />
sp<br />
pl<br />
nj<br />
kr<br />
kv<br />
fj<br />
tr<br />
sk<br />
dr<br />
gr<br />
st<br />
tv<br />
pr<br />
fr<br />
br<br />
0 20 40 60 80 100<br />
Diagram 4.4. Percentage of motivated root morphemes per cluster.<br />
87<br />
%
Diagram 4.4. shows the proportion of motivated root morphemes <strong>in</strong><br />
percentage of the total number of root morphemes per <strong>in</strong>dividual cluster.<br />
The root morphemes are the same as <strong>in</strong> diagrams 4.1–4.3. We can see that<br />
the clusters fn-, gn-, skv-, pj-, kn-, spr-, spj-, gl- mj- and vr- have a<br />
greater ratio of motivated root morphemes, well over 60%. The cluster<br />
fn- is the most sound symbolic, at 100 %. These clusters are all lexically<br />
<strong>in</strong>frequent clusters with the exception for kn- and gl- which are<br />
comparatively larger, (cf. diagrams 4.2 and 4.3). It is clear that lexically<br />
<strong>in</strong>frequent clusters are exploited sound symbolically to a higher degree<br />
than lexically frequent clusters are (cf. the discussion <strong>in</strong> 8.3). The clusters<br />
br-, pr-, and fr- have the lowest proportion of sound symbolic root<br />
morphemes. Of these clusters pr- is very frequent lexically and br- is<br />
quite frequent too (cf. diagrams 4.2, 4.3). This phenomenon is however<br />
not symmetrical; there are lexically very frequent clusters like sl- which<br />
are also sound symbolic to quite a high degree (55%). An <strong>in</strong>terest<strong>in</strong>g<br />
result is that the two most sound symbolic clusters, fn- and gn-, end with<br />
n while the three least sound symbolic clusters, pr-, fr- and br-, all end<br />
with an r.<br />
4.2.3 Types of mean<strong>in</strong>g<br />
After calculat<strong>in</strong>g the ratios of motivated root morphemes per cluster, an<br />
analysis was made of the semantic features appear<strong>in</strong>g <strong>in</strong> each cluster. The<br />
analysis can be done <strong>in</strong> detail or more abstractly. The analysis presented<br />
<strong>in</strong> this chapter is quite detailed, with more specific semantic features. For<br />
example there are the categories 'narrow form', round form', 'th<strong>in</strong><br />
form', hollow form', 'short wide form', 'crooked form', 'long th<strong>in</strong><br />
form', 'w<strong>in</strong>d<strong>in</strong>g form' and 'small end form'. On a more abstract level<br />
these can of course be classified as 'form'. The same goes for e.g. 'surface<br />
structure' which is further analyzed <strong>in</strong>to 'rough surface structure',<br />
'smooth surface structure', 'hardness'. Some categories are broader than<br />
their names imply. This should be noted for 'pejorative', which <strong>in</strong>cludes<br />
someth<strong>in</strong>g more generally negative and 'talk<strong>in</strong>g' which <strong>in</strong>cludes all sounds<br />
made by humans.<br />
The categories of diagram. 4.5 are related to those presented <strong>in</strong> chapter 2<br />
<strong>in</strong> the manner listed below. They are not mutually exclusive, s<strong>in</strong>ce it was<br />
judged to be <strong>in</strong>terest<strong>in</strong>g to count certa<strong>in</strong> frequent special cases of<br />
mean<strong>in</strong>g. Therefore e.g. 'mental feel<strong>in</strong>g' is subdivided <strong>in</strong>to 'bad mood'<br />
and 'other mental feel<strong>in</strong>g'. However, <strong>in</strong> the follow<strong>in</strong>g discussions and<br />
88
diagrams 'other mental feel<strong>in</strong>g', etc is simply called 'mental feel<strong>in</strong>g' 5. The<br />
feature 'slang' is a stylistic feature and <strong>in</strong> fact adds another dimension.<br />
<strong>Sound</strong>: <strong>Sound</strong><br />
Talk<strong>in</strong>g<br />
Beat<br />
Movement: Movement<br />
Quick or strong movement<br />
Walk<strong>in</strong>g<br />
Fall<strong>in</strong>g<br />
Potential movement<br />
Quickness<br />
Light: Light<br />
Gaze<br />
Surface structure: Rough surface structure<br />
Smooth surface structure<br />
Consistency: Soft consistency<br />
(Plasticity) Hardness<br />
Slackness<br />
Stiffness<br />
Wetness: Wetness<br />
Adhesion<br />
Dryness: Dryness<br />
Attitude: Attitude<br />
Secrecy<br />
Slang: Slang<br />
Jocular: Jocular<br />
Pejorative: Pejorative<br />
Destruction<br />
Mental feel<strong>in</strong>g: Mental feel<strong>in</strong>g<br />
Bad mood<br />
Bodily feel<strong>in</strong>g: Bodily feel<strong>in</strong>g<br />
Suffocation<br />
Separation: Separation<br />
5 Other such cases are 'other movement' which is called 'movement', 'light emission'<br />
which is called 'light', 'other wetness' which is called 'wetness', 'other attitude' which is<br />
called 'attitude', 'other mental feel<strong>in</strong>g' which is called 'mental feel<strong>in</strong>g', 'other bodily<br />
feel<strong>in</strong>g' which is called 'bodily feel<strong>in</strong>g', 'other form' which is called 'form', and 'other<br />
iterative' which is called 'iterative'.<br />
89
Putt<strong>in</strong>g together Putt<strong>in</strong>g together<br />
Dim<strong>in</strong>utive: Dim<strong>in</strong>utive<br />
Augmentative: Augmentative<br />
Form: Form<br />
Round form<br />
Short-wide form<br />
Th<strong>in</strong> form<br />
Hollow form<br />
W<strong>in</strong>d<strong>in</strong>g form<br />
Long th<strong>in</strong> form<br />
Narrow form<br />
Small end form<br />
Bent form<br />
Iterative: Iterative<br />
F<strong>in</strong>e gra<strong>in</strong><br />
The most frequent mean<strong>in</strong>gs (shown <strong>in</strong> table 4.2) are, <strong>in</strong> descend<strong>in</strong>g<br />
order:<br />
Table 4.2 The most frequent semantic features, <strong>in</strong> descend<strong>in</strong>g order.<br />
semantic freq semantic freq semantic freq<br />
feature<br />
feature<br />
feature<br />
'pejorative' 163 'gaze' 15 'attitude' 5<br />
'sound' 107 'th<strong>in</strong> form' 14 'jocular' 5<br />
'long th<strong>in</strong> form' 97 'smooth surface' 13 'fall<strong>in</strong>g' 5<br />
'quick or strong 67 'slang' 12 'soft<br />
4<br />
movement'<br />
consistency'<br />
'wetness' 63 'beat' 12 'small end<br />
form'<br />
4<br />
'talk<strong>in</strong>g' 55 'slackness' 11 'secrecy' 4<br />
'light' 32 'rough surface<br />
structure'<br />
10 'iterative' 3<br />
'dim<strong>in</strong>utive' 31 'separation' 9 'bent form' 3<br />
'round form' 23 'putt<strong>in</strong>g<br />
together'<br />
9 'narrow form' 3<br />
'way of walk<strong>in</strong>g' 22 'hardness' 8 'stiffness' 2<br />
90
'destruction' 21 'bodily feel<strong>in</strong>g' 8 'potential<br />
movement'<br />
2<br />
'w<strong>in</strong>d<strong>in</strong>g form' 20 'f<strong>in</strong>e gra<strong>in</strong>' 8 'light tactile<br />
sensation'<br />
2<br />
'short-wide<br />
form'<br />
18 'hollow form' 8 'augmentative' 1<br />
'bad mood' 18 'quickness' 6 'dryness' 1<br />
'form' 16 'adhesion' 6<br />
91
pejorative<br />
sound<br />
long th<strong>in</strong> form<br />
quick or strong movem<br />
wetness<br />
talk<strong>in</strong>g<br />
light<br />
dim<strong>in</strong>utive<br />
round form<br />
way of walk<strong>in</strong>g<br />
destruction<br />
w<strong>in</strong>d<strong>in</strong>g form<br />
bad mood<br />
shortwide form<br />
form<br />
gaze<br />
th<strong>in</strong> form<br />
smooth surface<br />
beat<br />
slang<br />
slackness<br />
rough surf. str<br />
separation<br />
putt<strong>in</strong>g together<br />
hardness<br />
bodily feel<strong>in</strong>g<br />
f<strong>in</strong>e gra<strong>in</strong><br />
hollow form<br />
quickness<br />
adhesion<br />
attitude<br />
jocular<br />
fall<strong>in</strong>g<br />
soft cons<br />
small end form<br />
secrecy<br />
bent form<br />
narrow form<br />
stiffness<br />
pot. movem.<br />
light tactile sens.<br />
augmentative<br />
dryness<br />
0 20 40 60 80 100 120 140 160 180<br />
Diagram, 4.5 The diagram shows the extent to which different semantic<br />
features are exploited by all <strong>in</strong>itial consonant clusters, <strong>in</strong> absolute<br />
numbers. It shows e.g. that 'pejorative' is the most frequent semantic<br />
feature.<br />
92
Comb<strong>in</strong>ations of features<br />
The high frequency of some of the features is probably due to their<br />
ability to comb<strong>in</strong>e with other features. 'Pejorative' and 'sound' often<br />
comb<strong>in</strong>e with other features <strong>in</strong> a root morpheme, e.g. 'walk<strong>in</strong>g', 'talk<strong>in</strong>g'<br />
('way of walk<strong>in</strong>g' and 'talk<strong>in</strong>g' are very often comb<strong>in</strong>ed with<br />
'pejorative'). A feature like 'w<strong>in</strong>d<strong>in</strong>g form', on the other hand, is very<br />
low frequent (only found <strong>in</strong> two clusters, kr- and sn-) and this low<br />
frequency might be partly a consequence of lower ability to comb<strong>in</strong>e with<br />
other features. Likewise 'bad mood' is conf<strong>in</strong>ed to the clusters gr-, vrand<br />
tr-. The frequencies of different semantic features for each cluster<br />
and how different features comb<strong>in</strong>e can be studied <strong>in</strong> Appendix 1.<br />
4.3 Frequent semantic features<br />
A detailed account for how semantic features exploit the five most<br />
frequent clusters (cf. diagram. 4.5) is shown below. These features are<br />
'pejorative', 'wetness', 'sound', 'long th<strong>in</strong> form' and 'quick or strong<br />
movement'.<br />
Diagram 4.6 shows the feature 'pejorative' and how it is distributed <strong>in</strong><br />
terms of percentage over 28 clusters. (The percentage is calculated on<br />
number of features per total number of root morphemes, for every<br />
cluster.)<br />
4.3.1 Pejorative<br />
Diagram 4.6 shows the percentage of morphemes with a pejorative<br />
feature. For example, 71 % of all morphemes beg<strong>in</strong>n<strong>in</strong>g with pj- and<br />
44% of all morphemes beg<strong>in</strong>n<strong>in</strong>g with fn- have a pejorative mean<strong>in</strong>g.<br />
Almost all clusters have morphemes with pejorative features. As can be<br />
seen, pj- is the cluster with the highest percentage of pejorative root<br />
morphemes. fn- comes second and then two more cluster end<strong>in</strong>g <strong>in</strong> j: nj-<br />
(which is however to small to be considered <strong>in</strong>terest<strong>in</strong>g) and fj-.<br />
93
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
sp<br />
sk<br />
pr<br />
str<br />
tr<br />
sv<br />
sm<br />
kl<br />
kr<br />
gr<br />
pl<br />
fl<br />
gl<br />
mj<br />
skv<br />
sn<br />
dr<br />
bl<br />
kn<br />
skr<br />
bj<br />
vr<br />
sl<br />
gn<br />
fj<br />
nj<br />
fn<br />
pj<br />
Diagram 4.6. Percent root morphemes with the feature 'pejorative' for different clusters.
It is also <strong>in</strong>terest<strong>in</strong>g to study the absolute numbers of morphemes with<br />
pejorative features, s<strong>in</strong>ce some clusters are lexically very frequent and<br />
others are very <strong>in</strong>frequent. The morphemes with a pejorative feature for<br />
the clusters are shown <strong>in</strong> table 4.3 (the absolute number of root<br />
morphemes for each cluster is placed after every cluster). The clusters<br />
are presented <strong>in</strong> order of descend<strong>in</strong>g relative frequency (cf. diagram 4.6).<br />
Table 4.3 The morphemes with the feature pejorative, the <strong>in</strong>itial<br />
consonant clusters <strong>in</strong>volved, and their absolute frequencies.<br />
pj-: pjatt (squirt), pjoller (babble,<br />
twaddle), pjosk (coddl<strong>in</strong>g), pjunk (coddl<strong>in</strong>g;<br />
pul<strong>in</strong>g), pjåkig (mawkish, bad) (SAOL also<br />
lists pjalt (stackare), pjunk ("gnäll, pjosk"),<br />
pjåk ("pjunk")) 5<br />
fn-: fnask (prostitute), fnatta (run about),<br />
fnoskig (dotty), fnurra (cf. grumpy) 4<br />
nj-: njugg (niggardly) 1<br />
fj-: fjant (busybody), fjollig (foolish), fjompig<br />
(foolish), fjuttig (<strong>in</strong>significant), fjäsa<br />
(show off), fjäska (fawn on) (SAOL also lists<br />
fjoskig (fnoskig) (dotty) 6<br />
gn-: gnat (nagg<strong>in</strong>g), gneta (st<strong>in</strong>t), gnidare<br />
(miser) 3<br />
sl-: slabba (splash), slabbertacka<br />
(gossipmonger), sladdra (gossip), slafs<br />
(slopp<strong>in</strong>ess), slampa (slut), slams<br />
(slovenl<strong>in</strong>ess), slarvig (slipshod), slas<br />
(slugish and slipshod person), slasa<br />
(walk lazily), slask (slush), slatt (heeltap),<br />
slattrig (gossip<strong>in</strong>g), slidder (gossip), sl<strong>in</strong>ka (wench),<br />
slisk (sweet stuff), slok (bloke),<br />
sludder (slurred speech), slum (slum),<br />
slusk (shabby-look<strong>in</strong>g fellow), slyna (bitch),<br />
slyngel (young rascal), slö (<strong>in</strong>dolent),<br />
slödder (riff-raff), slösa (squander) 24<br />
vr-: vräkig (flashy), vränga<br />
(turn <strong>in</strong>side out), vrövel (silly talk) 3<br />
bj-: bjäbba (squabble, bicker), bjäfs<br />
95
(gewgaws) 2<br />
skr-: skral (poor, bad), skrodera (brag),<br />
skrubb (cubby-hole), skrutt (rubbish),<br />
skrymt (hypocrisy), skräppa (brag),<br />
skrävla (brag), skröna (tall story) 8<br />
kn-: knackig (quite bad), knal<br />
(economically weak), knasig (crazy),<br />
kneg (job), kneken (vara på kneken:<br />
be down one's luck), knodd (bounder),<br />
knorva (make creased<br />
and wr<strong>in</strong>kled), knutte (person who is<br />
one-sidedly focused on a certa<strong>in</strong> activity),<br />
knöl (bastard), knös (rich fellow) 10<br />
bl-: bladdra (babble), blaffa (big ugly<br />
colour patch), blaj (silly talk), blarr (silly<br />
talk), blaska (splash), bliga (stare stupidly),<br />
bluddra (talk nonsense), blunder (<br />
blunder) (SAOL also lists (blaha<br />
(nonsense talk)) 8<br />
dr-: drasut (lanky fellow), dratta (fall),<br />
dravel (twaddle), dregel (dribble), drulle<br />
(clumsy fool), drummel (lout), drägg<br />
(dregs), drälla (spill), dröse (unstructured<br />
amount) 9<br />
sn-: snafs (dirt), snaskig (smutty), snatta<br />
(p<strong>in</strong>ch), snicksnack (chit-chat), sniken<br />
(greedy), snobb (dandy), snorkig (snooty),<br />
snusk (smutt<strong>in</strong>ess), snylta (sponge),<br />
snål (st<strong>in</strong>gy), snärta (young th<strong>in</strong>g),<br />
snöd (sordid), snöplig (<strong>in</strong>glorious) 13<br />
skv-: skvallra (gossip) 1<br />
mj-: mjäkig (namby-pamby) 1<br />
gl-: glop (whipper-snapper), glufsa<br />
(scoff), glunkas (det glunkas: there is a<br />
rumour), glupande (ravenous), glupsk<br />
(greedy), glåpord (taunt) 6<br />
fl-: flacka (rove), flamsig (silly, giggly),<br />
96
fl<strong>in</strong>a (sneer, cackle), flitter (?), floskel 6<br />
(empty phrase), flabba (cackle), flopp<br />
(ignom<strong>in</strong>ious failure), flummig (<strong>in</strong>tellectually<br />
unclear). SAOL also lists flepig ("mjäkig",<br />
namby-pamby) 8<br />
pl-: pladuska (irregular spot), plottra<br />
(potter about), plufsig (bloated), plump<br />
(rude), plump (blot), plundra (plunder),<br />
plussig (bloated) 7<br />
gr-: grumlig (muddy), grummel<br />
(dregs), grums (dregs), grumsa<br />
(grumble), gräll (glar<strong>in</strong>g), gräma (grieve),<br />
gräslig (terrible) 7<br />
kr-: krafs (trash), krake (wretch), kram<br />
(trash), krams (trash), kratta (funk), kreta<br />
(badly whittle), krimskrams (knick-knacks),<br />
kruserlig (ceremonious), krångel (fuss), kräk<br />
(wretch) 10<br />
tr-: tradig (bor<strong>in</strong>g), traggla (plod through),<br />
trams (rubbish, drivel), traska (trot,<br />
jog), trassel (tangle, muddle), troll<br />
(troll), truls (unordered collection) 7<br />
kl-: klanka (grouse, grumble), klantig<br />
(clumsily stupid), klotter (scribble, doodle),<br />
kludda (daub, smudge), klyscha 7 (cliché;<br />
lump of spittle), klåpare (bungler, botcher) 6<br />
sm-: smicker (blarney), smolk<br />
(particle of dirt), smuts (filth) 3<br />
sv-: svamla (drivel), svassa (strut,<br />
swagger), svulstig (bombastic) 3<br />
str-: strul (fuss), strunt (rubbish),<br />
(stuck-up blighter) 2<br />
pr-: pracka (fob), pryl (awl), prångla (utter -<br />
counterfeit co<strong>in</strong>) 3<br />
6 The word floskel comes from Lat<strong>in</strong> flosculus which means "little flower; showy<br />
decoration <strong>in</strong> speech" but it also fits <strong>in</strong>to the Swedish pattern (cf. discussion <strong>in</strong> 1.5)<br />
7 Another <strong>in</strong>terest<strong>in</strong>g word, which can be said to come from cliché, but which fits <strong>in</strong>to the<br />
pejorative group (and also with all the root morphemes denot<strong>in</strong>g way of talk<strong>in</strong>g, etc.)<br />
especially <strong>in</strong> connection with the dialectal mean<strong>in</strong>g of klyscha (lump of spittle).<br />
97
sk-: skavank (flaw), skorv (old tub) 2<br />
sp-: spill (waste), spol<strong>in</strong>g (whipper-snapper) 2<br />
The follow<strong>in</strong>g diagram (4.7) shows the absolute number of features for<br />
the clusters. 8 This diagram shows that sl- has 24 root morphemes with a<br />
pejorative feature, sn- has 13, etc. It is clear that a large number of all<br />
pejorative morphemes beg<strong>in</strong> with an s. Sl-, sn- and skr- beg<strong>in</strong> a large<br />
number of morphemes that have a pejorative component (but these<br />
clusters are not the most dom<strong>in</strong>ated by sound symbolic root morphemes,<br />
cf. diagram 4.4).<br />
Note aga<strong>in</strong> that one root morpheme can have more than one motivated<br />
mean<strong>in</strong>g component. For example 'sound' and 'pejorative' are often<br />
comb<strong>in</strong>ed as <strong>in</strong> bluddra, or 'sound', 'wetness' and 'movement' as <strong>in</strong><br />
skvimpa. 'Walk<strong>in</strong>g' is often comb<strong>in</strong>ed with an additional slightly<br />
pejorative mean<strong>in</strong>g component (e. g. svassa has 'pretentious manner').<br />
'Talk<strong>in</strong>g' is also often comb<strong>in</strong>ed with an additional pejorative mean<strong>in</strong>g<br />
component (e.g. pladdra implies 'nonsense').<br />
8 The reason why there are 30 clusters here <strong>in</strong>stead of 28 as <strong>in</strong> diagram 4.6 is that the<br />
frequency of 1 morpheme is below 1% <strong>in</strong> a large cluster like st-.<br />
98
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
fl<br />
spr<br />
nj<br />
mj<br />
st<br />
sp<br />
sk<br />
str<br />
bj<br />
sv<br />
sm<br />
vr<br />
pr<br />
gn<br />
fn<br />
pj<br />
gl<br />
kl<br />
fj<br />
pl<br />
gr<br />
tr<br />
bl<br />
s kr<br />
dr<br />
kr<br />
kn<br />
sn<br />
sl<br />
k<br />
Diagram 4.7 Number of root morphemes with the 'pejorative' feature.
Comments on the feature 'pejorative'<br />
As stated above, pj- is the cluster with the highest percentage of<br />
pejorative root morphemes. All two-consonant clusters conta<strong>in</strong><strong>in</strong>g a jphoneme<br />
are pejorative and to quite a high degree.<br />
Look<strong>in</strong>g at absolute figures, clusters beg<strong>in</strong>n<strong>in</strong>g with s – especially sl- snand<br />
skr- – are dom<strong>in</strong>at<strong>in</strong>g among the pejorative morphemes. Sl- is by far<br />
connected with the greatest number of pejorative morphemes.<br />
4.3.2 <strong>Sound</strong><br />
The semantic feature 'sound' ranks as number 2 <strong>in</strong> sound symbolic<br />
frequency (cf. diagram 4.5). Diagram 4.8 shows that fn- is the cluster<br />
with the highest proportion of morphemes hav<strong>in</strong>g the feature 'sound'. Fnis<br />
a lexically <strong>in</strong>frequent cluster and thus the tendency from 'pejorative'<br />
and is repeated: lexically <strong>in</strong>frequent clusters tend to be highly sound<br />
symbolic. Fn- is also the most sound symbolic cluster of all (100%)<br />
count<strong>in</strong>g the different semantic features.<br />
100
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
st<br />
pr<br />
tr<br />
fl<br />
fr<br />
pl<br />
gl<br />
fj<br />
sk<br />
kr<br />
br<br />
skr<br />
sm<br />
sn<br />
mj<br />
bj<br />
kl<br />
kv<br />
gn<br />
kn<br />
skv<br />
fn<br />
Diagram 4.8 Percent root morphemes with the feature 'sound' for different clusters.
The morphemes with the feature 'sound' for the clusters are shown <strong>in</strong><br />
table 4.4 (the absolute number of root morphemes for each cluster is<br />
placed after every cluster). The clusters are presented after descend<strong>in</strong>g<br />
relative frequency (cf. diagram 4.8):<br />
Table 4.4 The morphemes with the feature 'sound', the <strong>in</strong>itial consonant<br />
clusters <strong>in</strong>volved and their absolute frequencies.<br />
fn-: fnissa (giggle), fnittra (giggle), fnysa<br />
(snort) 3<br />
skv-: skval (pour<strong>in</strong>g), skvalp (lapp<strong>in</strong>g) 2<br />
kn-: knacka (knock), knaka (creak), knall<br />
(bang), knapra (nibble), knarr (creak<strong>in</strong>g),<br />
knastra (crackle), knatter (rattle), knirka<br />
(creak), knirra (creak), knistra (crackle),<br />
knittra (crackle), knorr (grumbl<strong>in</strong>g), knot<br />
(murmur<strong>in</strong>g), knyst (the least sound),<br />
knäppa (click) 15<br />
gn-: gnissla (squeak), gnod<strong>in</strong>g (a fish with a<br />
grumbl<strong>in</strong>g sound), gnägga (wh<strong>in</strong>ny), gnälla<br />
(squeak) 4<br />
kv-: kvacka (quack), kvarka (strangles),<br />
kvida (whimper), kvillra (twitter), kvirra<br />
(grumble), kvitter (chirp), kväda (s<strong>in</strong>g),<br />
kväka (croak) 8<br />
kl-: klafs (squelch), klamp (tramp), klang<br />
(clang), klappa (knock), klappra (klatter),<br />
klatsch (crack), klick (lump), klifs (squelch),<br />
kl<strong>in</strong>g (r<strong>in</strong>g<strong>in</strong>g), kl<strong>in</strong>k (t<strong>in</strong>kl<strong>in</strong>g), klirra (j<strong>in</strong>gle),<br />
klocka (bell), klucka (cluck), klämta (toll),<br />
kläpp (clapper) 15<br />
bj-: bjällra (bell) 1<br />
mj-: mjau (meow) 1<br />
sn-: snarka (snore), snarpa (creak), snarra<br />
(burr), snattra (quack), snusa (sniff),<br />
snyfta (sob), snyta (blow one's nose),<br />
snäppa (click), snärp (duck's sound),<br />
snörvla (snuffle) 10<br />
102
sm-: smack (smack), smaska (guzzle),<br />
smatter clatter), smätta (flick) 4<br />
skr-: skramla (rattle), skrap (scrap<strong>in</strong>g),<br />
skrälla (blare) 3<br />
br-: braka (crash), brassa (fire away),<br />
brum (grumble), brus (buzz), bräka (bleat),<br />
bröl (growl) 6<br />
kr-: krafsa (scratch), krakel (row), kras<br />
(crack), krasch (crash), kraxa (croak),<br />
krysta (bear down), kråka (crow), krämta<br />
(hawk) 8<br />
sk-: skall (bark<strong>in</strong>g), skalla (resound), skallra<br />
(rattle), skorr (burr), skott (shot), skångra<br />
(vibrate) 6<br />
fj-: fjärta (fart) 1<br />
gl-: glam (laugh<strong>in</strong>g and talk<strong>in</strong>g), glissando,<br />
gläfs (yelp) 3<br />
pl-: pladask (flop), plask (splash), pl<strong>in</strong>g<br />
(t<strong>in</strong>g-a-l<strong>in</strong>g), plums (plop) 4<br />
fr-: frasa (rustle), frusta (snort), fräsa (hiss) 3<br />
fl-: flabb (cackle), flöjt (flute) 2<br />
tr-: trumma (drum), trumpet<br />
(trumpet) 2<br />
pr-: prassel (rustle), prutt (fart) 2<br />
st-: stampa (stamp), stepp (tap-dance),<br />
stön (groan) 3<br />
The list (as well as diagrams 4.8 and 4.9) shows that kn- is frequent both<br />
percentally and absolutely.<br />
Comments on the feature 'sound'<br />
Diagram 4.9 shows how many root morphemes have the feature 'sound',<br />
<strong>in</strong> absolute figures. Four of the five largest ones beg<strong>in</strong> with k: kl-, kn-,<br />
kv- and kr-. Kl-, kn- and kv- are also percentally (cf. diagram 4.8)<br />
among the clusters more dom<strong>in</strong>at<strong>in</strong>g for the feature 'sound'.<br />
103
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
mj<br />
fj<br />
bj<br />
skv<br />
fl<br />
pr<br />
tr<br />
st<br />
gl<br />
fr<br />
fn<br />
skr<br />
sm<br />
pl<br />
gn<br />
sk<br />
br<br />
kr<br />
kv<br />
sn<br />
kn<br />
kl<br />
Diagram 4.9 Number of root morphemes with the feature 'sound' for different clusters.
4.3.3 Long th<strong>in</strong> form<br />
The feature 'long th<strong>in</strong> form' ranks as number 3 <strong>in</strong> sound symbolic<br />
frequency accord<strong>in</strong>g to diagram 4.5. This is one of the many 'form'features.<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
spj<br />
s tr<br />
spr<br />
sp<br />
sl<br />
sn<br />
Diagram 4.10 Percent of the feature 'long th<strong>in</strong> form' for different<br />
clusters.<br />
Diagram 4.10 shows that the semantic feature 'long th<strong>in</strong> form' is<br />
dom<strong>in</strong>ated by three three-consonant clusters: spj-, str-, and spr-. All<br />
these beg<strong>in</strong> with s (a fact of all Swedish <strong>in</strong>itial three consonant clusters)<br />
just as all but two of the other clusters, namely sp-, sl-, sn-, st-, sk- and<br />
sv-. The only ones that do not beg<strong>in</strong> with s are tr- and tv-. ( However, t<br />
is a dental like s.) Aga<strong>in</strong> the highest ranked cluster spj- is a lexically very<br />
<strong>in</strong>frequent cluster.<br />
The morphemes with the feature 'long th<strong>in</strong> form' for the clusters are<br />
shown <strong>in</strong> table 4.5 (the absolute number of root morphemes for each<br />
cluster are placed after every cluster). The clusters are presented after<br />
descend<strong>in</strong>g relative frequency (cf. diagram 4.10).<br />
tr<br />
105<br />
tv<br />
st<br />
sk<br />
sv
Table 4.5 The morphemes with the feature 'long th<strong>in</strong> form', the <strong>in</strong>itial<br />
consonant clusters <strong>in</strong>volved and their absolute frequencies.<br />
spj-: spjut (spear), spjäla (lath) 2<br />
str-: streamer (streamer), streck (stroke),<br />
stretch (stretch), strigel (strop), stril<br />
(nozzle), strimla (strip), strimma (streak),<br />
stripa (wisp), strut (cone), strå (straw),<br />
stråk (band), stråle (ray), stråt (way),<br />
sträcka (stretch), sträng (str<strong>in</strong>g), ström<br />
(stream), strössel (spr<strong>in</strong>kles) 17<br />
spr-: spricka (crack), spr<strong>in</strong>ga (narrow<br />
open<strong>in</strong>g), spr<strong>in</strong>t (p<strong>in</strong>), sprits (squirt),<br />
sprund (slit), spröjs (w<strong>in</strong>dow-bar),<br />
spröt (rib) 7<br />
sp-: spaljé (trellis), spalt (column),<br />
spant (rib), sparre (rafter), sparris<br />
(asparagus), spatel (spatula), spene (dug),<br />
spenslig (slender), spetig (sk<strong>in</strong>ny), spets<br />
(po<strong>in</strong>t), spett (spit), spigg (stickleback), spik<br />
(nail), spila (spike l<strong>in</strong>g), sp<strong>in</strong>del (spider),<br />
sp<strong>in</strong>kig (th<strong>in</strong>), sp<strong>in</strong>na (sp<strong>in</strong>), spira (spire),<br />
spole (bobb<strong>in</strong>), spont (tongue), spång<br />
(foot-bridge), spänta (split wood), spö (twig) 23<br />
sl-: sladd (cord), slamsa (rag), slana (scaffold<br />
pole), slang (tube), slank (slender), slant<br />
(a fish<strong>in</strong>g rod), slejf (strap), slimmad (fitted),<br />
sl<strong>in</strong>ga (coil), slips (tie), slits (slit), släde (sleigh) 12<br />
sn-: snabel (trunk), snigel (snail), snilja<br />
(a thread), snitsel (paperstrip), sno (tw<strong>in</strong>e),<br />
snodd (cord), snok (grass snake), snorkel<br />
(snorkel), snugga (cutty), snöre (str<strong>in</strong>g) 10<br />
tr-: tratt (funnel), tross (hawser), truta<br />
(pout), tryne (snout), tråd (thread), tråg<br />
(trough), trål (trawl), träns (braid) 8<br />
tv-: tv<strong>in</strong>na (tw<strong>in</strong>e) 1<br />
st-: stake (stake), stav (staff), sticka<br />
(spl<strong>in</strong>ter), stift (p<strong>in</strong>), stig (path), stilett<br />
(stiletto), stock (log), stolpe (pole), stylta<br />
106
(stilt), stång (pole), stör (stake), stötta<br />
(prop) 12<br />
sk-: skakel (shaft), skalm (shaft), skalpell<br />
(scalpel) 3<br />
sv-: svabba (swab), svans (tail) 2<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
sp<br />
s tr<br />
sl<br />
st<br />
sn<br />
tr<br />
Diagram 4.11 Number of morphemes with the feature 'long th<strong>in</strong> form'<br />
for different clusters.<br />
Comments on the feature 'long th<strong>in</strong> form'.<br />
Diagram 4.11 shows that sp- is the largest cluster, <strong>in</strong> absolute figures, for<br />
the semantic feature 'long th<strong>in</strong> form'.<br />
4.3.4 Quick or strong movement 9<br />
Diagram 4.12 shows that fl- has the highest percent of morphemes hav<strong>in</strong>g<br />
the feature 'quick or strong movement'. Fl- is a quite large cluster. In<br />
second place comes spr- which is a lexically <strong>in</strong>frequent cluster even if it<br />
is not extremely small. Vr- <strong>in</strong> third place is also lexically <strong>in</strong>frequent.<br />
The morphemes with the feature 'quick or strong movement' for the<br />
clusters are shown <strong>in</strong> table 4.6 (the absolute number of root morphemes<br />
9 The reason for collaps<strong>in</strong>g quick movement and strong movement is that <strong>in</strong> some root<br />
morphemes both these mean<strong>in</strong>gs are present (although <strong>in</strong> others only one is present).<br />
However, it seems cumbersome to make three categories <strong>in</strong>stead of one.<br />
spr<br />
107<br />
sk<br />
spj<br />
sv<br />
tv
for each cluster are placed after every cluster). The clusters are presented<br />
accord<strong>in</strong>g to descend<strong>in</strong>g relative frequency (cf. diagram 4.12).<br />
20<br />
18<br />
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
fl<br />
spr<br />
vr<br />
sv<br />
gn<br />
fn<br />
skv<br />
Diagram 4.12 Percentage 'quick or strong movement' for different<br />
clusters.<br />
Table 4.6 The morphemes with the feature 'quick or strong movement',<br />
the <strong>in</strong>itial consonant clusters <strong>in</strong>volved, and their absolute frequencies.<br />
fl-: flacka (rove), fladdra (flutter), flagga<br />
(flag) flamma (flame), flanera (be out for a<br />
stroll), flaxa (flutter), flimra (flicker), fluga<br />
(fly), fluktuation (fluctuation), flyga (fly), fly<br />
(flee), flyta (float), flytta (move), flåsa<br />
(puff) fläkta (fan), flämta (pant), flänga<br />
(be dash<strong>in</strong>g about), flöda (flow) 18<br />
spr-: sprallig (frolicsome), sprattla<br />
(flounder), spritta (jump), sprudla (bubble),<br />
spruta (spurt), sprutt (speed) 6<br />
vr-: vricka (scull), vrida (twist), vräka<br />
(heave) 3<br />
sv-: svaja (sw<strong>in</strong>g), svalla (surge), svepa<br />
(sweep), sv<strong>in</strong>g (sw<strong>in</strong>g), svirvel (swivel),<br />
svämma (overflow), sväng (sweep),<br />
svärm (swarm), sväva (float) 9<br />
sl<br />
108<br />
kn<br />
sk<br />
tr<br />
fr<br />
kr<br />
sp<br />
st
gn-: gnida, (rub) gno (rub), gnugga (rub) 3<br />
fn-: fnatta (run about)) 1<br />
skv-: skvimpa (splash to and fro) 1<br />
sl-: sladda (skid), sl<strong>in</strong>ka (slip), sl<strong>in</strong>ta (slide),<br />
slipprig (slippery), slira (skid), slita (pull),<br />
slugga (beat), slunga (fl<strong>in</strong>g), slänga (throw) 9<br />
kn-: knixa (bob), knyck (jerk) 2<br />
sk-: skaka (shake), skalv (quake), skippa<br />
(skip) 3<br />
tr-: tromb (tornado), tromla (rotat<strong>in</strong>g<br />
cyl<strong>in</strong>drical sieve), trumla (treat<br />
someth<strong>in</strong>g by enclos<strong>in</strong>g it <strong>in</strong> a rotat<strong>in</strong>g<br />
drum) 3<br />
fr-: frossa (the shivers), frusa (gush) 2<br />
kr-: kratsa (scrape), kränga (heave over) 2<br />
sp-: spasm (spasm), spurt (spurt) 2<br />
st-: studsa (bounce), stöppla (give pa<strong>in</strong>ted<br />
surface a certa<strong>in</strong> look by strik<strong>in</strong>g a brush<br />
aga<strong>in</strong>st it), stöt (thrust) 3<br />
18<br />
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
fl<br />
sv<br />
sl<br />
spr<br />
vr<br />
tr<br />
st<br />
gn<br />
Diagram 4.13 Number of morphemes with the feature 'quick and strong<br />
movement' for different clusters.<br />
109<br />
sk<br />
fr<br />
kn<br />
kr<br />
sp<br />
skv<br />
fn
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
kr<br />
str<br />
dr<br />
sn<br />
pl<br />
sp<br />
sv<br />
bl<br />
sl<br />
kl<br />
sm<br />
spr<br />
skv<br />
Diagram 4.14 Percent root morphemes with the feature 'wetness' for different clusters.
Comments on the feature 'quick or strong<br />
movement'.<br />
Diagram 4.13 shows that fl- has by far the greatest number of<br />
morphemes with the feature 'quick and strong movement'. Fl- is then<br />
most frequent both absolutely and percentally.<br />
4.3.5 Wetness<br />
The last semantic feature to be exam<strong>in</strong>ed is 'wetness', which ranks as<br />
number 5 <strong>in</strong> sound symbolic frequency accord<strong>in</strong>g to diagram 4.5.<br />
Diagram 4.14 shows that skv- is the cluster with the highest proportion of<br />
morphemes hav<strong>in</strong>g the feature wetness. (Almost half of the morphemes<br />
beg<strong>in</strong>n<strong>in</strong>g with skv- have the feature 'wetness', cf. table 4.16) The next to<br />
largest cluster, sm-, has only 13% morphemes with this feature. The<br />
cluster skv- is small and conta<strong>in</strong>s a total of ten root morphemes. The<br />
morphemes with the 'wetness' feature for the clusters are the follow<strong>in</strong>g<br />
(the absolute number of root morphemes for each cluster are placed after<br />
every cluster). The clusters are presented accord<strong>in</strong>g to descend<strong>in</strong>g relative<br />
frequency (cf. diagram 4.14).<br />
Table 4.7 The morphemes with the feature 'wetness', the <strong>in</strong>itial consonant<br />
clusters <strong>in</strong>volved, and their absolute frequencies.<br />
skv-: skval (gush), skvalta (ripple), skvimpa<br />
(splash), skvätt (drop) 4<br />
spr-: sprej (spray), spr<strong>in</strong>kler (spr<strong>in</strong>kler),<br />
sprudla (bubbla), spruta (spurt) 4<br />
sm-: smegma (secretion), smet (sludge),<br />
smetana (a sour thick cream), sm<strong>in</strong>k<br />
(make-up), smälta (melt), smörja (grease) 6<br />
kl-: klabb (stick), kladd (daub), klafs<br />
(squelch), klena (daub), kleta (daub), klibba<br />
(stick), klick (daub), klifs (squelch), klucka<br />
(gurgle), klunk (gulp) 10<br />
sl-: slabba (splash), slafsa (squelch),<br />
slam (ooze), slask (slush), slatt (drop), slem<br />
(slime), slicka (lick), slipprig (slippery), slisk<br />
(sweet stuff), slur<strong>in</strong>g (soup), slurk (swig),<br />
111
sluss (sluice) 12<br />
bl-: blaska (splash), blod (blood), bläck (<strong>in</strong>k),<br />
blöt (wet) 4<br />
sv-: svabba (swab), svett (sweat), svämma<br />
(overflow) 3<br />
sp-: spackel (putty), spad (liquid), sperma<br />
(sperm), spilla (spill), spola (flush), spott<br />
(spittle), sputum (phlegm) 7<br />
pl-: plask (splash), plums (plop), plurr<br />
(water) 3<br />
sn-: snaskig (smutty), snigel (snail), snor<br />
(snot), snuva (cold) 4<br />
dr-: dregel (dribble), droppa (drip) 2<br />
str-: strila (spr<strong>in</strong>kle), ström (stream) 2<br />
kr-: kram (cloggy), kräm (cream) 2<br />
Comments on the feature 'wetness'<br />
Diagram. 4.15. shows <strong>in</strong> absolute numbers how many morphemes have<br />
the feature 'wetness' for each cluster. The largest cluster is sl- (twelve<br />
<strong>in</strong>stances) followed by kl- (ten <strong>in</strong>stances). 8 of the 13 'wetness' clusters<br />
beg<strong>in</strong> with s. Three beg<strong>in</strong> with an unvoiced stop and two beg<strong>in</strong> with a<br />
voiced stop.<br />
Summary of and discussion of the five most<br />
frequent features<br />
As stated above, skv- is the cluster with the highest percentage of<br />
'wetness' root morphemes. This cluster only conta<strong>in</strong>s ten morphemes so –<br />
as well as for the pejorative root morphemes – small clusters are<br />
proportionally more utilized for sound symbolism. The clusters sl-, kl-,<br />
sp-, sm-, skv-, spr-, sn-, bl-, pl-, and sv- are most utilized for the<br />
mean<strong>in</strong>g 'wetness', i.e. s or <strong>in</strong>itial unvoiced plosives are preferred. Sland<br />
kl- (both end<strong>in</strong>g with l) have the highest number of root morphemes<br />
with the feature 'wetness'.<br />
112
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
sl kl sp sm skv sn bl spr sv pl kr dr str<br />
Diagram 4.15 Number of root morphemes with the feature 'wetness' for different clusters.
The cluster sl- is the most frequent sound symbolic cluster for both<br />
'pejorative' and 'wetness'.<br />
Look<strong>in</strong>g at absolute figures, clusters beg<strong>in</strong>n<strong>in</strong>g with s, especially sl- sn-,<br />
and skr-, are dom<strong>in</strong>at<strong>in</strong>g among the 'pejorative' morphemes. Sl- has by<br />
far the greatest number of pejorative morphemes.<br />
Similar results for cluster types were obta<strong>in</strong>ed for other clusters : 'light'<br />
is dom<strong>in</strong>ated by voiced plosives + l or n: bl-, gl-, gn- (dl-, bn-, dn- are<br />
not allowed <strong>in</strong> Swedish) 10, 'bad (negative) mood' is constra<strong>in</strong>ed to some<br />
plosives + r: gr-, tr-, vr- 11 and 'wetness' is dom<strong>in</strong>ated by: sl-, kl-, sp-,<br />
sm-, skv-, spr-, sn-, bl-, pl-, sv-, i.e. s and <strong>in</strong>itial unvoiced plosives are<br />
preferred (the only exception is bl-). The phoneme l is also quite<br />
common.<br />
10 There are root morphemes beg<strong>in</strong>n<strong>in</strong>g with other clusters that have been classified as<br />
‘light’ but these ma<strong>in</strong>ly concern quality of colour: bjärt (gaudy), gräll (glar<strong>in</strong>g),<br />
prunkande (dazzl<strong>in</strong>g), prålig (garish).<br />
11 There is only one exception: knarrig (creaky).<br />
114
4.3.6 The most sound symbolic clusters<br />
Table 4.8 shows which the most frequent clusters are, for every mean<strong>in</strong>g.<br />
Table 4.8. The most sound symbolic clusters <strong>in</strong> absolute numbers and<br />
proportionally, for the five most frequent mean<strong>in</strong>gs.<br />
freq % examples <strong>in</strong> % % freq examples<br />
mean<strong>in</strong>g <strong>in</strong><br />
absolutenumbers<br />
'pejorative' sl- 24 16 sladder<br />
(gossip), slok<br />
(bloke),<br />
slödder (riffraff)<br />
'sound' kl-,<br />
kn-<br />
'long th<strong>in</strong><br />
form'<br />
'quick or<br />
strong<br />
movement'<br />
15 13,<br />
19<br />
klang (clang),<br />
klirra (j<strong>in</strong>gle)<br />
knarra<br />
(creak),<br />
knittra<br />
(crackle)<br />
sp- 23 17 spant (rib),<br />
spett (spit)<br />
spö (twig)<br />
fl- 18 19 fladdra<br />
(flutter),<br />
flamma<br />
(flame),<br />
flimra<br />
(flicker).<br />
fläkta (fan)<br />
'wetness' sl- 12 8 slask (slush)<br />
slem (slime)<br />
slipprig<br />
(slippery)<br />
115<br />
pj- 71 5 pjoller<br />
(babble),<br />
pjosk<br />
(coddl<strong>in</strong>g),<br />
pjunk<br />
fn- 33 3<br />
(coddl<strong>in</strong>g)<br />
fnissa<br />
(giggle),<br />
fnysa (snort)<br />
spj- 33 2 spjut (spear),<br />
spjäla (lath)<br />
fl- 19 18 fladdra<br />
(flutter),<br />
flamma<br />
(flame),<br />
flimra<br />
(flicker)<br />
fläkta (fan)<br />
skv- 40 4 skval (gush),<br />
skvalta<br />
(ripple),<br />
skvimpa<br />
(splash)
The <strong>in</strong>formation given <strong>in</strong> table 4.8, for the five most frequent mean<strong>in</strong>gs,<br />
will now be expla<strong>in</strong>ed. For the feature 'pejorative' (which is the most<br />
frequent mean<strong>in</strong>g, see diagram 4.5) the most high frequent cluster (<strong>in</strong> the<br />
sense of the greatest number of roots) is sl-, and it has a frequency of 24<br />
root morphemes. Sl- can therefore be said to be a very pejorative cluster.<br />
However, many words beg<strong>in</strong> with sl- and only 16% of these are <strong>in</strong> fact<br />
pejorative. It is also <strong>in</strong>terest<strong>in</strong>g to look at which cluster is dom<strong>in</strong>ated to<br />
the highest degree by the feature 'pejorative'. This turns out to be pjwhere<br />
5 root morphemes (71% of the root morphemes beg<strong>in</strong>n<strong>in</strong>g with<br />
pj-) are pejorative. So, <strong>in</strong> the sense of be<strong>in</strong>g dom<strong>in</strong>ated by a certa<strong>in</strong><br />
mean<strong>in</strong>g, pj- can be said to be the most pejorative cluster. The same<br />
comparisons are made for the rema<strong>in</strong><strong>in</strong>g 4 semantic features.<br />
The clusters that are most frequent, <strong>in</strong> absolute numbers, show a strong<br />
tendency to end with l. The clusters that are most frequent proportionally<br />
show a weak tendency to end with j. In both groups the clusters beg<strong>in</strong><br />
with a voiceless obstruent.<br />
4.4 Frequent clusters<br />
So far we have studied how clusters are distributed over different<br />
mean<strong>in</strong>gs. We will now look at what mean<strong>in</strong>gs different clusters hold.<br />
The follow<strong>in</strong>g section shows how different mean<strong>in</strong>gs are distributed over<br />
different clusters. The clusters exam<strong>in</strong>ed are the most frequent <strong>in</strong> absolute<br />
numbers: sl-, sn-, kn-, kr- (cf. diagram 4.3) and some of the most<br />
frequent percentally: fn-, kn-, gn-, spr-, pj- (cf. diagram 4.4).<br />
Also those clusters that are almost unique for a certa<strong>in</strong> mean<strong>in</strong>g will be<br />
discussed, e.g. 'pejorative' has the unique cluster fj-, i.e., fj- is almost<br />
only 'pejorative'. Fl- is to a high degree 'quick or strong movement', fn-<br />
'pejorative', gl- 'light', bl- 'light' or 'gaze', kr- 'w<strong>in</strong>d<strong>in</strong>g form', kn-<br />
'round form', skr- 'pejorative' and 'destruction', sl- 'pejorative', str-<br />
'long th<strong>in</strong> form', br- 'sound' and 'destruction', gr- 'bad mood' or 'hollow<br />
form', (vr- 'bad mood', tr- 'bad mood'), kl- is 'wetness', 'short-wide<br />
form' or 'adhesion', kv- is 'suffocation', mj- 'softness', sk- 'hardness',<br />
skv- 'wetness', sl- 'slackness' and sp-, st- and str- are 'long th<strong>in</strong> form'<br />
(cf. table 4.16).<br />
116
25<br />
20<br />
15<br />
% m.comp/all<br />
10<br />
5<br />
0<br />
walk<strong>in</strong>g<br />
pot. movem.<br />
smooth surf.<br />
beat<br />
talk<strong>in</strong>g<br />
strong movem<br />
quick or<br />
slackness<br />
long th<strong>in</strong><br />
wetness<br />
pejorative<br />
Diagram 4.16 Semantic features of the cluster sl-. Percentages of semantic features for all root morphemes.
4.4.1 The cluster sl-<br />
The first cluster to be studied is sl-, the largest cluster <strong>in</strong> absolute<br />
numbers. Diagram 4.16 shows the semantic features of the consonant<br />
cluster sl-. The words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g<br />
features are shown <strong>in</strong> table 4.8. The absolute frequencies (not the<br />
percentages) of the semantic features are shown to the right.<br />
Table 4.8 The sound symbolic morphemes of the sl- cluster, the semantic<br />
features <strong>in</strong>volved, and their absolute frequencies.<br />
'pejorative': slabba (splash),<br />
slabbertacka (gossipmonger),<br />
sladdra (chatter), slafs (slopp<strong>in</strong>ess)<br />
slampa (slut), slams (slovenl<strong>in</strong>ess),<br />
slarvig (slipshod), slas (lazy and<br />
careless person), slasa (walk<br />
heavily and shuffl<strong>in</strong>g), slask (slush),<br />
slatt (drop), slattrig (slack), slidder<br />
(gossip), sl<strong>in</strong>ka (wench), slisk<br />
(sweet stuff), sloka (droop), sludder<br />
(slurred speech), slum (slum), slusk<br />
(shabby look<strong>in</strong>g fellow), slyna<br />
(bitch), slyngel (young rascal), slö<br />
(sluggish), slödder (riff-raff),<br />
slösa (waste) 24<br />
'wetness': slabba (splash), slafs<br />
(slopp<strong>in</strong>ess), slam (ooze), slask<br />
(slush), slatt (drop), slem (slime),<br />
slicka (lick), slipprig (slippery), slisk<br />
(sweet stuff), slur<strong>in</strong>g (a soup), slurk<br />
(swig), sluss (sluice)<br />
(cf. discussion <strong>in</strong> 1.5) 12<br />
'long th<strong>in</strong> form': sladd (cord),<br />
slamsa (rag), slana (scaffold pole),<br />
slang (tube), slank (slender), slant<br />
(a fish<strong>in</strong>g rod), slejf (strap), slimmad<br />
(fitted), sl<strong>in</strong>ga (coil), slips (tie), slits (slit),<br />
släde (sleigh). 12<br />
118
'slackness': slack (slack), sladdrig<br />
(flabby), slak (slack), slamsa (rag),<br />
slana (scaffold pole), slang (tube),<br />
slankig (flaccid), slapp (slack), sliddrig<br />
(flabby), sl<strong>in</strong>kig (slabby), sloka (droop) 11<br />
'quick or strong movement': sladda<br />
(skid), sl<strong>in</strong>ka (slip), sl<strong>in</strong>ta (slip),<br />
slipprig (slippery), slira (skid), slita<br />
(tear), slugga (slug), slunga (fl<strong>in</strong>g), slänga<br />
(fl<strong>in</strong>g) 9<br />
'talk<strong>in</strong>g': slabbertacka (gossipmonger),<br />
sladdra (chatter), slattrig (gossip<strong>in</strong>g),<br />
slidder (gossip), sludder (slurred<br />
speech) 5<br />
'beat': slag (beat), slå (beat), slägga<br />
(sledge hammer) 3<br />
'smooth surface': slipa (polish), slät<br />
(smooth), slätt (pla<strong>in</strong>) 3<br />
'potential movement': slutta (slant),<br />
slänt (slope) 2<br />
'walk<strong>in</strong>g': slasa (walk heavily and<br />
shuffl<strong>in</strong>g), släntra (saunter) 2<br />
The semantic feature 'slackness' is unique for this cluster.<br />
The relations between some of the different mean<strong>in</strong>gs of sl- are discussed<br />
below <strong>in</strong> 4.6.<br />
Sl- has a different mean<strong>in</strong>g profile than sn- (cf. diagram 4.17). However,<br />
they are both dom<strong>in</strong>ated by the feature 'pejorative'.<br />
4.4.2 The cluster sn-<br />
The second most sound symbolic cluster, <strong>in</strong> absolute numbers, is sn-.<br />
Diagram 4.17 shows the semantic features of this consonant cluster.<br />
119
10<br />
9<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
0<br />
pejorative<br />
sound<br />
long th<strong>in</strong> form<br />
quickness<br />
wetness<br />
Diagram 4.17 Semantic features of the cluster sn-. Percentages of<br />
semantic features for all root morphemes.<br />
In table 4.9 the words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g<br />
features are shown. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 4.9 The sound symbolic morphemes of the sn- cluster, the semantic<br />
features <strong>in</strong>volved, and their absolute frequencies.<br />
'pejorative': snafs (dirt), snaskig (smutty),<br />
snatta (p<strong>in</strong>ch), snicksnack (chit-chat),<br />
sniken (greedy), snobb (dandy), snorkig<br />
(snooty), snusk (smutt<strong>in</strong>ess), snylta<br />
(sponge), snål (greedy), snärta<br />
(young th<strong>in</strong>g), snöd (sordid), snöplig<br />
(<strong>in</strong>glorious) 13<br />
'sound': snarka (snore), snarpa (creak),<br />
snarra (burr), snattra (quack), snusa<br />
120<br />
small end form<br />
talk<strong>in</strong>g<br />
way of walk<strong>in</strong>g<br />
w<strong>in</strong>d<strong>in</strong>g form<br />
dim<strong>in</strong>utive<br />
slang
(sniff), snyfta (sob), snyta (blow nose),<br />
snäppa (click) , snärp (duck´s sound),<br />
snörvla (snuffle) 10<br />
'long th<strong>in</strong> form': snabel (trunk), snigel<br />
(snail), snilja (a thread) , snitsel (paper<br />
strip), sno (tw<strong>in</strong>e), snodd (cord), snok (grass<br />
snake), snorkel (snorkel), snultra (a long<br />
th<strong>in</strong> fish), snöre (str<strong>in</strong>g) 10<br />
'quickness': snabb (quick), snappa (snatch),<br />
snar (who reacts without delay), sno<br />
(tw<strong>in</strong>e), snudig (quick), snärj (hurry) 6<br />
'talk<strong>in</strong>g': snacka (chat), snicksnack<br />
(chit-chat), snubba (rebuke), snäsa (snub) 4<br />
'wetness': snaskig (smutty), snigel (snail),<br />
snor (snot), snuva (head cold) 4<br />
'small end form': snagga (clip short), snibb<br />
(tip), snip (lip), snopp (tip) 4<br />
'w<strong>in</strong>d<strong>in</strong>g form': snirkel (scroll), snurra<br />
(twist), snäcka (shell) 3<br />
'dim<strong>in</strong>utive': snitt (cut), snugga (cutty),<br />
snutt (small piece) 3<br />
'slang': snofsa upp (make elegant), snubbe<br />
(person), snut (cop) 3<br />
'walk<strong>in</strong>g': snava (trip), snubbla (trip) 2<br />
4.4.3 The cluster kn-<br />
The follow<strong>in</strong>g diagram (4.18) shows the semantic features of the cluster<br />
kn-, the third largest cluster <strong>in</strong> absolute numbers.<br />
The words (represent<strong>in</strong>g root morphemes) of the semantic features of knare<br />
shown <strong>in</strong> table 4.10. The absolute frequency (not the percentages) of<br />
the semantic features are shown to the right.<br />
121
20<br />
18<br />
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
round form<br />
sound<br />
pejorative<br />
together<br />
Diagram 4.18 Semantic features of the cluster kn-. Percentages of<br />
semantic features for all root morphemes.<br />
Table 4.10 The sound symbolic morphemes of the kn- cluster, the<br />
semantic features <strong>in</strong>volved and their absolute frequencies.<br />
'sound': knacka (knock), knaka (creak)<br />
, knall (bang), knapra (nibble), knarr<br />
(creak<strong>in</strong>g) , knastra (crackle), knatter (rattle)<br />
, knirka (creak), knirra (creak), knistra<br />
(crackle), knittra (crackle), knorr<br />
(grumbl<strong>in</strong>g), knot (murmur<strong>in</strong>g), knyst<br />
(the least sound), knäppa (click) 15<br />
'round form': knagg (wooden stick), knalle<br />
(hillock), knapp (button), knast (twig <strong>in</strong><br />
122<br />
way of walk<br />
dim<strong>in</strong>utive<br />
beat<br />
quick or<br />
strong movem
wood), knick (sharp curve on pipe), knoge<br />
(fist), knollrig (frizzy), knop (knot), knopp<br />
(bud), knorr (curl), knota (bone), knottra<br />
(goose-pimples), knubbig (chubby), knödel<br />
(cooked bun of patatoes), knöl (bump) 15<br />
'pejorative': knackig (quite bad), knal<br />
(economically weak), knasig (crazy), kneg<br />
(job), kneken (vara på kneken: be down<br />
one's luck), knodd (bounder), knorva (make<br />
creased and wr<strong>in</strong>kled), knutte (person who<br />
is onesidedly focused on a certa<strong>in</strong> activity),<br />
knöl (bastard), knös (rich fellow) 10<br />
'putt<strong>in</strong>g together': knipa (p<strong>in</strong>ch), knippa<br />
(bunch), knipsa (staple), knut (knot),<br />
knyckla (crumple up), knyppla (make lace),<br />
knyta (tie) 7<br />
'walk<strong>in</strong>g': knaggla (move forward slowly<br />
and with difficulty), knalla (trot), knata<br />
(run), knoga (walk laboriously) 4<br />
'dim<strong>in</strong>utive': knatte (little fellow), knott<br />
(gnat), knåp (f<strong>in</strong>icky job) 3<br />
'beat': knocka (knock out), knuff (push),<br />
knäck (crack) 3<br />
'quick or strong movement': knixa (bob),<br />
knyck (jerk) 2<br />
Kn- conta<strong>in</strong>s many words with the mean<strong>in</strong>g 'round form', but the cluster<br />
is not unique for this mean<strong>in</strong>g.<br />
Kn- also belongs to the clusters <strong>in</strong> the next section, s<strong>in</strong>ce it is also one of<br />
the 4 most sound symbolic cluster percentally.<br />
123
10<br />
9<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
0<br />
w<strong>in</strong>d<strong>in</strong>g form<br />
pejorative<br />
sound<br />
way of walk<strong>in</strong>g<br />
124<br />
th<strong>in</strong> form<br />
rough surf<br />
destruction<br />
quick or<br />
strong movem<br />
Diagram 4.19 Semantic features of the cluster kr-. Percentages of<br />
semantic features for all root morphemes.<br />
4.4.4 The cluster kr-<br />
The last of the four largest clusters (<strong>in</strong> absolute numbers) is kr-, shown <strong>in</strong><br />
diagram 4.19. In table 4.11 the words (represent<strong>in</strong>g root morphemes) of<br />
the semantic features of kr- are shown. The absolute frequency (not the<br />
percentages) of the semantic features are shown to the right.<br />
Table 4.11 The sound symbolic morphemes of the kr- cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'w<strong>in</strong>d<strong>in</strong>g form': krans (wreath),<br />
kr<strong>in</strong>gelikrokar (lots of bends <strong>in</strong> different<br />
wetness
directions), kr<strong>in</strong>gla (pretzel), krok (hook),<br />
krokan (croquembouche), krulla (curl), krum<br />
(crooked), krumbukt (w<strong>in</strong>d<strong>in</strong>gs), krumelur<br />
(flourish), krumpen (crooked and bent from<br />
age or disease), krusa (curl), krusiduller<br />
(ornaments), krycka (crutch), kråma (prance<br />
about), kräkla (crosier), kräla (crawl),<br />
krök (bend) 17<br />
'pejorative': krafs (trash), krake (wretch),<br />
kram (trash), krams (trash), kratta (funk),<br />
kreta (badly whittle), krimskrams (knick-knacks),<br />
kruserlig (ceremonious), krångel (fuss), kräk<br />
(wretch) 10<br />
'sound': krafsa (scratch), krakel (row), kras<br />
(crack), krasch (crash), kraxa (croak), krysta<br />
(bear down), kråka (crow), krämta (hawk) 8<br />
'th<strong>in</strong> form': krakmandel (soft-shell<br />
almond), krokett (croquette), krusta (crust),<br />
krustad (croustade) 4<br />
'walk<strong>in</strong>g': kravla (crawl), krylla (crawl with),<br />
krypa (crawl), kräla (crawl) 4<br />
'rough surface structure': kratta (rake),<br />
krås (ruffle), kräpp (crepe) 3<br />
'destruction': kracka (decompose molecules<br />
through heat<strong>in</strong>g), krackelera (crackled), krossa<br />
(crush) 3<br />
'wetness': kram (cloggy), kräm (cream) 2<br />
'quick or strong movement': kratsa<br />
(scrape), kränga (heave over) 2<br />
Kr- is dom<strong>in</strong>ated by the unusual semantic feature 'w<strong>in</strong>d<strong>in</strong>g form' and the<br />
cluster and mean<strong>in</strong>g are almost uniquely connected. (However 'w<strong>in</strong>d<strong>in</strong>g<br />
form' also occurs <strong>in</strong> some root morpheme with sn-.)<br />
The clusters described above are the four largest ones. Taken together<br />
they are dom<strong>in</strong>ated by the semantic features 'pejorative' and 'sound'.<br />
125
The largest clusters, percentally<br />
The next group of clusters that are to be studied more closely are the ones<br />
that are the percentally most sound symbolic, cf. diagram 4.4.<br />
4.4.5 The cluster fn-<br />
Fn- is the <strong>in</strong>itial consonant cluster with the largest percentage of sound<br />
symbolic root morphemes. The percentages of the semantic features are<br />
shown <strong>in</strong> diagram 4.20.<br />
45<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
pejorative<br />
sound<br />
quick or strong<br />
movem<br />
Diagram 4.20 Semantic features of the cluster fn-. Percentages of<br />
semantic features for all root morphemes.<br />
The words (represent<strong>in</strong>g root morphemes) of the semantic features of fnare<br />
presented <strong>in</strong> table 4.12. The absolute number of words (not the<br />
percentages) of the semantic features are shown to the right. For all<br />
lexically <strong>in</strong>frequent clusters, i.e. clusters conta<strong>in</strong><strong>in</strong>g less than 13 root<br />
126<br />
th<strong>in</strong> form<br />
dryness
morphemes, also root morphemes with the frequency 1 are counted (cf.<br />
chapter 3).<br />
Table 4.12 The sound symbolic morphemes of the fn- cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'pejorative': fnask (prostitute), fnatta (run<br />
about), fnoskig (dotty), fnurra (cf. grumpy) 4<br />
'sound': fnissa (giggle), fnittra (giggle),<br />
fnysa (snort) 3<br />
'quick or strong movement': fnatta (run about) 1<br />
'th<strong>in</strong> form': fnasig (chapped) 1<br />
'dryness': fnöske (t<strong>in</strong>der) 1<br />
The cluster fn- is dom<strong>in</strong>ated by 'pejorative'. All fn- root morphemes are<br />
sound symbolic. (However, as shown <strong>in</strong> diagrams 4.6 and 4.7, the feature<br />
'pejorative' is spread over many clusters.)<br />
4.4.6 The cluster kn-<br />
Kn- is the <strong>in</strong>itial consonant cluster with the fourth largest percentage of<br />
sound symbolic root morphemes. It is also the third largest cluster <strong>in</strong><br />
absolute numbers. The diagrams and the list of root morphemes can be<br />
studied above <strong>in</strong> 4.4.3.<br />
4.4.7 The cluster gn-<br />
Gn- is the <strong>in</strong>itial consonant cluster with the second largest percentage of<br />
sound symbolic root morphemes. The percentages of the semantic features<br />
are shown <strong>in</strong> diagram 4.21.<br />
The words (represent<strong>in</strong>g root morphemes) of the semantic features of gnare<br />
presented <strong>in</strong> table 4.13. The absolute number of words (not the<br />
percentages) of the semantic features are shown to the right.<br />
127
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
talk<strong>in</strong>g<br />
sound<br />
pejorative<br />
quick or<br />
strong movem<br />
Diagram 4.21 Semantic features of the cluster gn-. Percentages of<br />
semantic features for all root morphemes.<br />
Table 4.13 The sound symbolic morphemes of the gn- cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'talk<strong>in</strong>g': gnabb (bicker<strong>in</strong>g), gnat (nagg<strong>in</strong>g),<br />
gnola (hum), gny (d<strong>in</strong>), gnöla (klaga) 5<br />
'sound': gnissla (squeak), gnod<strong>in</strong>g (a fish<br />
with a grumbl<strong>in</strong>g sound), gnägga (wh<strong>in</strong>ny),<br />
gnälla (squeak) 4<br />
'pejorative': gnat (nagg<strong>in</strong>g), gneta (st<strong>in</strong>t),<br />
gnidare (miser) 3<br />
'smooth surface structure': gnida (rub),<br />
gno (rub), gnugga (rub) 3<br />
'quick or strong movement': gnida,(rub),<br />
gno (rub), gnugga (rub) 3<br />
'light': gnista (spark), gnistra (sparkle) 2<br />
128<br />
smooth surf<br />
light
The semantic features of gn- are primarily 'talk<strong>in</strong>g' and 'sound', which<br />
semantically are very close. An analysis of relations between some of the<br />
semantic features of gn- are presented <strong>in</strong> 4.6.<br />
4.4.8 The cluster spr-<br />
The <strong>in</strong>itial consonant cluster with the fifth largest percentage of sound<br />
symbolic root morphemes is spr-. The percentages of the semantic<br />
features are shown <strong>in</strong> diagram 4.22:<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
long th<strong>in</strong> form<br />
quick or strong<br />
movement<br />
Diagram 4.22 Semantic features of the cluster spr-. Percentages of<br />
semantic features for all root morphemes.<br />
129<br />
separation<br />
wetness
Table 4.14 shows the words (represent<strong>in</strong>g root morphemes) of the<br />
semantic features of spr-. The absolute number of words (not the<br />
percentages) of the semantic features are shown to the right.<br />
Table 4.14 The sound symbolic morphemes of the spr- cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'long th<strong>in</strong> form': spricka (crack), spr<strong>in</strong>ga<br />
(narrow open<strong>in</strong>g), spr<strong>in</strong>t (p<strong>in</strong>), sprits (squirt),<br />
sprund (slit), spröjs (w<strong>in</strong>dow-bar), spröt (rib) 7<br />
'quick or strong movement': sprallig<br />
(frolicsome), sprattla (flounder), spritta<br />
(jump), sprudla (bubble), spruta (spurt),<br />
sprutt (speed) 6<br />
'separation': sprej (spray), spreta (sprawl),<br />
sprida (spread), spr<strong>in</strong>kler (spr<strong>in</strong>kler),<br />
spränga (burst), sprätta (flick) 6<br />
'wetness': sprej (spray), spr<strong>in</strong>kler (spr<strong>in</strong>kler),<br />
sprudla (bubbla), spruta (spurt) 4<br />
Spr- is quite unique for the mean<strong>in</strong>g 'separation'. (The mean<strong>in</strong>g does,<br />
however, occur <strong>in</strong> sp- and spl-).<br />
4.4.9 The cluster pj-<br />
Pj- is the <strong>in</strong>itial consonant cluster with the third largest percentage of<br />
sound symbolic root morphemes. The percentages of the semantic<br />
features are shown <strong>in</strong> the follow<strong>in</strong>g diagram (4.23):<br />
The words (represent<strong>in</strong>g root morphemes) of the semantic features of pjare<br />
presented <strong>in</strong> table 4.15. The absolute number of words (not the<br />
percentages) of the semantic features are shown to the right.<br />
130
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
pejorative way of<br />
talk<strong>in</strong>g<br />
Diagram 4.23 Semantic features of the cluster pj-. Percentages of<br />
semantic features for both root morphemes.<br />
Table 4.15 The sound symbolic morphemes of the pj- cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'pejorative': pjatt (squirt), pjoller (babble,<br />
twaddle), pjosk (coddl<strong>in</strong>g), pjunk (coddl<strong>in</strong>g;<br />
pul<strong>in</strong>g), pjåkig (mawkish, bad) 5<br />
'sound': pjoller (babble, twaddle) 1<br />
The cluster is clearly dom<strong>in</strong>ated by the feature 'pejorative'.<br />
Summary for percentages<br />
Also for the four percentally most sound symbolic clusters the semantic<br />
features 'pejorative' and 'sound' are dom<strong>in</strong>at<strong>in</strong>g, together with 'talk<strong>in</strong>g'.<br />
Only the cluster spr- is deviat<strong>in</strong>g from this tendency. The cluster kn- is<br />
among the most sound symbolic clusters, both <strong>in</strong> absolute numbers and<br />
percentally.<br />
4.5 Typical or unique mean<strong>in</strong>gs<br />
In Table 4.16 all the clusters with typical or unique mean<strong>in</strong>gs are<br />
presented. Clusters that are dom<strong>in</strong>ated by a certa<strong>in</strong> mean<strong>in</strong>g (e.g. 'quick<br />
or strong movement' <strong>in</strong> fl-), or have a mean<strong>in</strong>g which is almost unique<br />
for that cluster (e.g. 'w<strong>in</strong>d<strong>in</strong>g form': kr-), have those mean<strong>in</strong>gs <strong>in</strong> italics.<br />
For small clusters, i.e. those with less than 13 root morphemes, also<br />
mean<strong>in</strong>gs that have the frequency 1 are counted.<br />
131
Table 4.16 Typical<br />
or unique mean<strong>in</strong>gs<br />
of <strong>in</strong>itial clusters.<br />
bl- 33<br />
pejorative 8<br />
light 9<br />
gaze 5<br />
talk<strong>in</strong>g 4<br />
wetness 4<br />
round form 3<br />
br- 11<br />
sound 6<br />
destruction 5<br />
fj- 7<br />
pejorative 6<br />
sound 1<br />
fl- 36<br />
quick or strong<br />
movement<br />
18<br />
pejorative 8<br />
th<strong>in</strong> form 6<br />
sound 2<br />
light 2<br />
fn- 10<br />
pejorative 4<br />
sound 3<br />
th<strong>in</strong> form 1<br />
quick or strong<br />
movement<br />
1<br />
dryness 1<br />
gl- 45<br />
light 14<br />
smooth surface<br />
structure<br />
7<br />
pejorative 6<br />
gaze 6<br />
dim<strong>in</strong>utive 5<br />
form 4<br />
sound 3<br />
gn- 20<br />
talk<strong>in</strong>g 5<br />
sound 4<br />
pejorative 3<br />
smooth surface<br />
structure<br />
3<br />
quick or strong<br />
movement<br />
3<br />
light 2<br />
gr- 21<br />
hollow form 8<br />
bad mood 6<br />
pejorative 7<br />
talk<strong>in</strong>g 5<br />
kl- 51<br />
sound 15<br />
short wide form 11<br />
wetness 10<br />
adhesion 6<br />
pejorative 6<br />
talk<strong>in</strong>g 3<br />
kn- 59<br />
round form 15<br />
sound 15<br />
pejorative 10<br />
putt<strong>in</strong>g together 7<br />
walk<strong>in</strong>g 4<br />
dim<strong>in</strong>utive 3<br />
beat 3<br />
quick or strong<br />
movement<br />
2<br />
kr- 53<br />
w<strong>in</strong>d<strong>in</strong>g form 17<br />
132<br />
pejorative 10<br />
sound 8<br />
walk<strong>in</strong>g 4<br />
th<strong>in</strong> form 4<br />
rough surface<br />
structure<br />
3<br />
destruction 3<br />
wetness 2<br />
quick or strong<br />
movement<br />
2<br />
kv- 17<br />
sound 8<br />
suffocation 5<br />
destruction 2<br />
dim<strong>in</strong>utive 2<br />
mj- 7<br />
f<strong>in</strong>e gra<strong>in</strong> 3<br />
soft consistency 2<br />
sound 1<br />
pejorative 1<br />
pj- 6<br />
pejorative 5<br />
talk<strong>in</strong>g 1<br />
sk- 25<br />
hardness 8<br />
sound 6<br />
long th<strong>in</strong> form 4<br />
quick or strong<br />
movement<br />
3<br />
pejorative 2<br />
round form 2<br />
skr- 35<br />
talk<strong>in</strong>g 12<br />
pejorative 8<br />
destruction 7<br />
rough surface<br />
structure<br />
5
sound 3<br />
skv- 9<br />
wetness 4<br />
sound 2<br />
talk<strong>in</strong>g 1<br />
pejorative 1<br />
movement 1<br />
sl- 83<br />
pejorative 24<br />
wetness 12<br />
long th<strong>in</strong> form 12<br />
slackness 11<br />
quick or strong<br />
movement<br />
9<br />
talk<strong>in</strong>g 5<br />
beat 3<br />
smooth surface str 3<br />
walk<strong>in</strong>g 2<br />
potential<br />
movement<br />
2<br />
sn- 62<br />
pejorative 13<br />
long th<strong>in</strong> form 10<br />
sound 10<br />
quickness 6<br />
talk<strong>in</strong>g 4<br />
small end form 4<br />
wetness 4<br />
w<strong>in</strong>d<strong>in</strong>g form 3<br />
dim<strong>in</strong>utive 3<br />
slang 3<br />
walk<strong>in</strong>g 2<br />
sp- 47<br />
long th<strong>in</strong> form 23<br />
wetness 7<br />
jocular 4<br />
th<strong>in</strong> form 3<br />
gaze 2<br />
quick or strong<br />
movement<br />
2<br />
attitude 2<br />
pejorative 2<br />
dim<strong>in</strong>utive 2<br />
spr- 23<br />
long th<strong>in</strong> form 7<br />
separation 6<br />
quick or strong<br />
movement<br />
6<br />
wetness 4<br />
str- 30<br />
long th<strong>in</strong> form 17<br />
walk<strong>in</strong>g 3<br />
stiffness 2<br />
wetness 2<br />
light 2<br />
pejorative 2<br />
f<strong>in</strong>e gra<strong>in</strong> 2<br />
tr- 44<br />
bad mood 8<br />
long th<strong>in</strong> form 8<br />
pejorative 7<br />
walk<strong>in</strong>g 5<br />
dim<strong>in</strong>utive 4<br />
round form 3<br />
quick or strong<br />
movement<br />
3<br />
talk<strong>in</strong>g 2<br />
short wide form 2<br />
sound 2<br />
133
4.6 Patterns of semantic features<br />
The different semantic features of a certa<strong>in</strong> consonant cluster often do not<br />
only constitute a list of haphazard features, but <strong>in</strong>stead they are related.<br />
These k<strong>in</strong>ds of relations between phonesthemes (e.g. sl-:'smooth surface<br />
structure' (slät) -> 'quick or strong movement (slira, sl<strong>in</strong>gra)) can also be<br />
found with<strong>in</strong> the same phonestheme (e.g. bl-: 'light' (blixt) –>'reflect<strong>in</strong>g<br />
surface structure potentially giv<strong>in</strong>g light' (blank), at a more detailed level<br />
of analysis. In other words, there are many clusters which conta<strong>in</strong> root<br />
morphemes that have similar, but not exactly the same, semantic features.<br />
These features are related <strong>in</strong>dexically or iconically and can be seen as<br />
belong<strong>in</strong>g to the same mean<strong>in</strong>g potential and activated <strong>in</strong> different<br />
contexts. This is especially strik<strong>in</strong>g <strong>in</strong> the case of root morphemes<br />
connected to 'light', beg<strong>in</strong>n<strong>in</strong>g with the clusters gl-, bl- and gn-. Most of<br />
these words have been classified as 'light' or 'gaze' <strong>in</strong> the analysis above,<br />
but as shown here they could be further analyzed. Some root morphemes<br />
can denote light, others perception of light, yet others potentiality for<br />
perception of light, etc. The different relations to 'light' are paraphrased<br />
below. Almost all relations between phonesthemic mean<strong>in</strong>gs can be<br />
described as some k<strong>in</strong>d of potentiality. Potentiality is not a semantic<br />
feature but it describes the relation between different mean<strong>in</strong>gs. By<br />
potentiality I mean a relation between mean<strong>in</strong>gs where one mean<strong>in</strong>g<br />
describes a possible prerequisite for another mean<strong>in</strong>g, e.g. form can be<br />
seen from the perspective of be<strong>in</strong>g able to let through a bit of light, or a<br />
reflect<strong>in</strong>g (glossy) surface can send out light, cf. the paraphrases below.<br />
The emergence of these relations probably come from simultaneity <strong>in</strong><br />
perception, i.e. a k<strong>in</strong>d of <strong>in</strong>dexical relation. In some cases the relation is<br />
iconic.<br />
The structures are similar but not exactly parallel, so the clusters will be<br />
presented one at a time:<br />
gl-<br />
'form through which light<br />
potentially can be perceived'<br />
glipa (narrow space), glugg<br />
(aperture), glänt (slightly open),<br />
glänta (glade)<br />
134
'perception of light'<br />
glana (stare), glo (stare), glutta<br />
(take a glance), glimt (glimpse)<br />
'production or source of light'<br />
glans (lustre), glimma (gleam), glimra (gleam), gl<strong>in</strong>dra<br />
(gleam), glisa (sh<strong>in</strong>e), glittra (glitter), glänsa (sh<strong>in</strong>e), glöd<br />
(glow), gloria (halo), glåmig (pale)<br />
'smooth surface that potentially<br />
reflects light'<br />
glas (glass), glasyr (glaz<strong>in</strong>g), glatt<br />
(smooth), glimmer (gleam<strong>in</strong>g),<br />
gl<strong>in</strong>der (a sh<strong>in</strong>y troll<strong>in</strong>g-spoon),<br />
glaciär (glacier)<br />
'movement on such a surface'<br />
glida (glide)<br />
135
l-<br />
gn-<br />
'perception of light'<br />
blick (gaze), bliga (stare), blänga<br />
(glare), blända (bl<strong>in</strong>d), bl<strong>in</strong>d (bl<strong>in</strong>d),<br />
bl<strong>in</strong>ka (bl<strong>in</strong>k)<br />
'light'<br />
blixt (lightn<strong>in</strong>g), bl<strong>in</strong>ka (tw<strong>in</strong>kle), blänka (sh<strong>in</strong>e), blossa<br />
(flare), blek (pale)<br />
'reflect<strong>in</strong>g surface potentially giv<strong>in</strong>g light'<br />
blank (sh<strong>in</strong>y)<br />
136<br />
(iconic)<br />
'light surface'<br />
bläs (blaze), black (drab)<br />
'momentary light'<br />
gnista (spark), gnistra (sparkle)<br />
'action which produces smooth surface'<br />
gnida (rub), gno (rub), gnugga (rub)
'sound aris<strong>in</strong>g from, and simultaneous<br />
with, such an action'<br />
gnida (rub), gno (rub), gnugga (rub)<br />
'similar sound aris<strong>in</strong>g from similar action'<br />
gnissla (squeak)<br />
'similar sounds aris<strong>in</strong>g from humans or<br />
creatures'<br />
gnabb (bicker<strong>in</strong>g), gnat (nagg<strong>in</strong>g),<br />
gnod<strong>in</strong>g (a murmur<strong>in</strong>g fish), gnola<br />
(hum), gny (d<strong>in</strong>), gnägga (neigh), gnälla<br />
(wh<strong>in</strong>e), gnöl (grumble)<br />
In the last cluster we can see <strong>in</strong>dexical and iconic relations between such<br />
basic categories as 'sound', 'movement' and 'light', e.g. the relation<br />
between the sound and the action is often simultaneous (<strong>in</strong>dexical) or the<br />
relation between this sound and another sound (like <strong>in</strong> gnissla) is likeness<br />
(iconic). Gn- can, <strong>in</strong> other words, be used for other sounds than those<br />
aris<strong>in</strong>g simultaneously with a polish<strong>in</strong>g movement. These are synchronic<br />
relations but one can imag<strong>in</strong>e the historical changes through mean<strong>in</strong>g<br />
extension. (The etymological relations, as described <strong>in</strong> e.g. Hellquist<br />
(1966), are not clear.)<br />
Common for the three clusters is the mean<strong>in</strong>g 'light', i. e. the process<br />
which is the source for visual perception. It is of course central. Different<br />
extensions of mean<strong>in</strong>g from 'light' are possible. As can be seen, there are<br />
"empty slots" <strong>in</strong> the mean<strong>in</strong>g patterns for the clusters. For example, gndoes<br />
not have adjectives hav<strong>in</strong>g to do with 'reflect<strong>in</strong>g surface', only verbs<br />
mean<strong>in</strong>g 'produc<strong>in</strong>g reflect<strong>in</strong>g surface'. And it does not have root<br />
morphemes for 'perception of light'. These can be accidental gaps. A<br />
speculative explanation of why gn- does not have words for 'perception<br />
137
of light' is that these words often are verbs and that the verbs beg<strong>in</strong>n<strong>in</strong>g<br />
with gn- are used for the feature 'sound'.<br />
Similar relations can be found for other semantic fields and for other<br />
clusters, sometimes as different root morphemes beg<strong>in</strong>n<strong>in</strong>g with a certa<strong>in</strong><br />
cluster, sometimes as semantic features of a root morpheme. Mean<strong>in</strong>gs<br />
often comb<strong>in</strong>ed are e.g. 'sound' - 'movement' - 'destruction' (e.g. braka),<br />
'destruction' - 'pejorative' (e.g bråte), 'wetness' - 'pejorative' (e.g.<br />
slemmig), 'sound' - 'surface structure' (e.g skrovlig), etc, and the relation<br />
between the mean<strong>in</strong>gs (except for 'wetness' - 'pejorative') is nearness <strong>in</strong><br />
space and time (<strong>in</strong>dexical).<br />
The semantic pattern of sl- is similar:<br />
pejorative (slabba (splash), slafs<br />
(slopp<strong>in</strong>ess), slampa (slut), slarvig<br />
(slipshod), slasa (walk lazily),<br />
sladdrig (flabby))<br />
wetness (slafsa (slop), slipprig (slipprig), slem (phlegm))<br />
smooth surface structure (slipa<br />
(polish), slät (smooth))<br />
quick or strong movement (sl<strong>in</strong>ka<br />
(slide), sl<strong>in</strong>ta (glide), slira (skid),<br />
sl<strong>in</strong>gra (w<strong>in</strong>d))<br />
138<br />
potential<br />
movement
139<br />
(slutta (slope),<br />
slänt (slope))<br />
long th<strong>in</strong> form (slang (tube),<br />
slimmad (slimmed), slana (scaffold<br />
pole), slejf (strap), sl<strong>in</strong>ga (coil), slips<br />
(tie))<br />
slackness (slamsa (rag), slankig<br />
(limp), slana (scaffold pole), slejf<br />
(strap), sl<strong>in</strong>ga (wreath), slips (tie))<br />
The relations are the follow<strong>in</strong>g:<br />
Wetness causes smooth surface structure (<strong>in</strong>dex).<br />
Smooth surface causes quick or strong movement (<strong>in</strong>dex).<br />
That which moves quickly often has a long th<strong>in</strong> form (<strong>in</strong>dex).<br />
That which has same form as long th<strong>in</strong> form is often slack (<strong>in</strong>dex).<br />
Mean<strong>in</strong>g hierarchies?<br />
One aim of the analysis was to f<strong>in</strong>d mean<strong>in</strong>g hierarchies with<strong>in</strong> clusters,<br />
e.g. that sound symbolism presupposes sound imitation, that movement<br />
presupposes sound imitation, that wetness always implies pejorative, but<br />
such strong claims can not be made. However, there are strong<br />
tendencies:<br />
For all 36 clusters 'pejorative' is the largest category, followed by<br />
'sound'. As mentioned earlier, clusters are often both 'sound' (or<br />
'talk<strong>in</strong>g') and 'pejorative'. Exceptions are br-, fr-, kv- which are<br />
'sound' but not 'pejorative'. Spj-, spl-, spr- are neither 'sound' nor<br />
'pejorative'.<br />
Clusters with 'quick or strong movement' also have the semantic feature<br />
'sound' (th<strong>in</strong>gs that move quickly often sound) except for one cluster:<br />
spr-. 'Wetness' is almost always co-occurr<strong>in</strong>g with 'sound imitation', also<br />
probably due to the fact that wet th<strong>in</strong>gs often sound.
Some clusters are used for a variety of mean<strong>in</strong>gs (sl-, kl-) while others<br />
are more specialized on one mean<strong>in</strong>g (fj-, pj-).<br />
4.7 Discussion and conclusions<br />
This chapter has shown that certa<strong>in</strong> semantic features are connected with<br />
certa<strong>in</strong> <strong>in</strong>itial consonant clusters, <strong>in</strong> different comb<strong>in</strong>ations and to<br />
different extents. From the analysis of sound symbolism of the <strong>in</strong>itial<br />
consonant clusters of Swedish some general conclusions can be drawn:<br />
* There is a difference between clusters <strong>in</strong> the number of root<br />
morphemes that are motivated. This is partly due to how lexically<br />
frequent the cluster is. There is also a difference between clusters <strong>in</strong> how<br />
big a proportion of the root morphemes are motivated. Some clusters are<br />
simply more used for motivated root morphemes. The variation is<br />
between 8% (br-) and 100% (fn-). It seems clear that lexically <strong>in</strong>frequent<br />
clusters are exploited for sound symbolism to a higher degree.<br />
(One might ponder about the importance of total or percental number for<br />
the language user's feel<strong>in</strong>g of degree of motivation of a cluster.<br />
Nevertheless this analysis has mostly dealt with percentages. Another<br />
related issue is the issue of textual frequency of motivated root<br />
morphemes, which has not been dealt with here.)<br />
* The mean<strong>in</strong>g profiles for most clusters differ. Different clusters (and<br />
different cluster types) seem to be fitt<strong>in</strong>g for different (types of)<br />
mean<strong>in</strong>gs.<br />
* There are often, on a more detailed level of analysis, <strong>in</strong>terest<strong>in</strong>g<br />
relations between the different root morphemes classified as belong<strong>in</strong>g to<br />
the same phonestheme, and sometimes there is a connection between the<br />
different mean<strong>in</strong>gs of a cluster profile, see the analysis of gl-, bl- gnand<br />
sl- above. The mean<strong>in</strong>gs are related <strong>in</strong>dexically or iconically and can<br />
be seen as belong<strong>in</strong>g to the same mean<strong>in</strong>g potential and activated <strong>in</strong><br />
different contexts. It is a question of detail of analysis how many<br />
mean<strong>in</strong>gs are described.<br />
140
* There is also a frequency difference between the mean<strong>in</strong>gs. Some<br />
mean<strong>in</strong>gs are quite frequent, e.g. 'sound' and 'pejorative', other mean<strong>in</strong>gs<br />
are less frequent, e.g. 'adhesion' and 'round form'.<br />
* Some mean<strong>in</strong>gs seem to comb<strong>in</strong>e more often with other mean<strong>in</strong>gs, e.g.<br />
pejorative + walk<strong>in</strong>g or talk<strong>in</strong>g.<br />
The next chapter will present an analysis of sound symbolism of f<strong>in</strong>al<br />
consonant clusters and of vowels where phonesthemes connected with the<br />
semantic features found <strong>in</strong> the analysis of <strong>in</strong>itial consonant clusters will be<br />
searched for. The results of the analysis of <strong>in</strong>itial consonant clusters (as<br />
well as that of f<strong>in</strong>al clusters) have been the base for the experiments with<br />
neologisms presented <strong>in</strong> chapter 7. It was also the base for part of the<br />
cross l<strong>in</strong>guistic studies presented <strong>in</strong> chapter 6.<br />
141
5. Analysis of f<strong>in</strong>al consonant clusters and<br />
vowels and of comb<strong>in</strong>ations<br />
It seems clear that a greater or lesser part of the sound symbolic<br />
mean<strong>in</strong>g of root morphemes can be attributed not only to an <strong>in</strong>itial<br />
consonant sequence but also to the f<strong>in</strong>al consonants, sometimes <strong>in</strong><br />
comb<strong>in</strong>ation with vowels (<strong>in</strong> different positions). It could be possible<br />
that the semantic feature 'quick or strong movement' of e. g. fladdra<br />
(flutter) is attributable not only to fl- (cf. 4.3.4) but also to -dr-. For all<br />
words end<strong>in</strong>g <strong>in</strong> -dr- the dom<strong>in</strong>at<strong>in</strong>g semantic feature is 'talk<strong>in</strong>g'<br />
(bladdra (talk nonsense), pladdra (babble), sladdra (chatter), bluddra<br />
(talk nonsense), sluddra (slur one's words) but also 'quick or strong<br />
movement' (fladdra and bläddra (turn over the pages). So, <strong>in</strong> the case<br />
of fladdra, fl- is most strongly tied to 'quick and strong movement', but<br />
-dr- also adds to the mean<strong>in</strong>g of the word. In addition it is possible that<br />
the vowels add to the sound symbolic flavor of a word, and that a<br />
neologism like fliddra would depict a yet quicker movement.<br />
To facilitate understand<strong>in</strong>g of the study, the reader is rem<strong>in</strong>ded of the<br />
procedure of decid<strong>in</strong>g which f<strong>in</strong>al clusters should be counted as tentative<br />
phonesthemes (cf. 3.1). When there are at least two 1 root morphemes<br />
end<strong>in</strong>g with the same cluster (which is sometimes followed by an<br />
obligatory vowel, i. e. strictly speak<strong>in</strong>g they are semif<strong>in</strong>al) and hav<strong>in</strong>g<br />
similar mean<strong>in</strong>g (ascerta<strong>in</strong>ed by key words), one phonestheme is<br />
established. If there are roots (normally at least two of each) with<br />
different, sound symbolic, mean<strong>in</strong>gs, different phonesthemes are<br />
established, e.g. -mla 'talk<strong>in</strong>g' (svamla, mumla) and 'quick or strong<br />
movement' (famla, vimla, tromla, fumla, rumla, drumla, tumla). It can<br />
also be the case that words have clearly different mean<strong>in</strong>gs, e.g. blaska.<br />
There is one mean<strong>in</strong>g 'splash' and one pejorative mean<strong>in</strong>g for<br />
'newspaper'. This word is then counted as two roots, and the analysis<br />
proceeds as above. In many cases a word has more than one mean<strong>in</strong>g,<br />
but only one of them is sound symbolic.<br />
1 Or, for lexically <strong>in</strong>frequent clusters, 1 root<br />
143
5.1 F<strong>in</strong>al clusters<br />
It is not as easy to enumerate the f<strong>in</strong>al sequences as it is to list the <strong>in</strong>itial<br />
sequences. In f<strong>in</strong>al sequences, morphological structure has to be taken<br />
<strong>in</strong>to account s<strong>in</strong>ce some sequences only occur <strong>in</strong> <strong>in</strong>flected or derived<br />
forms, e.g. -ndsk <strong>in</strong> bondsk. Many forms are difficult to evaluate,<br />
accord<strong>in</strong>g to the discussion <strong>in</strong> Sigurd (1965, pp. 67–69), who adopts a<br />
set of mechanical rules which exclude certa<strong>in</strong> forms (secondary forms)<br />
which can be assumed to break the natural phonotactic pattern.<br />
Thus, for practical reasons, f<strong>in</strong>al clusters were studied <strong>in</strong> the follow<strong>in</strong>g<br />
ways. Primarily they were studied with the aid of 'Svensk<br />
Baklängesordbok' (1981), which is a list of most Swedish lemmas 2<br />
(ordered alphabetically after the end<strong>in</strong>gs of the words only). The f<strong>in</strong>al<br />
clusters that are followed by a have been analyzed, partly s<strong>in</strong>ce the<br />
clusters were easy to f<strong>in</strong>d <strong>in</strong> that way, but ma<strong>in</strong>ly because some clusters<br />
cannot have word f<strong>in</strong>al position (e.g. C+r- and C+l-sequences). These<br />
are ma<strong>in</strong>ly verbs. (In addition, roots from Nusvensk Frekvensordbok 4<br />
(NFO4) were analyzed, see 5.3 and 5.5). The roots that were excerpted<br />
were those that have one or more of the semantic features discussed<br />
earlier. For this reason also s<strong>in</strong>gle occurrences of f<strong>in</strong>al clusters with<br />
these semantic features were registered. The aim has been to use<br />
different materials <strong>in</strong> order to study as many roots as possible. Not all<br />
f<strong>in</strong>al clusters are sound symbolic 3, as almost all <strong>in</strong>itial clusters are.<br />
The result<strong>in</strong>g 27 sound symbolic f<strong>in</strong>al clusters before a (semi f<strong>in</strong>al) are<br />
shown <strong>in</strong> table 5.1 below. There, gem<strong>in</strong>ate consonants, and not only<br />
consonant clusters, are <strong>in</strong>cluded. The semantic features recurr<strong>in</strong>g more<br />
than once for each f<strong>in</strong>al cluster are with<strong>in</strong> brackets.<br />
Table 5.1 The 27 sound symbolic f<strong>in</strong>al clusters (before a and preceded<br />
by a short vowel), from Svensk Baklängesordbok. 100% <strong>in</strong>dicates that<br />
there is only one root. This root is <strong>in</strong> all <strong>in</strong>dicated cases sound symbolic.<br />
All other clusters have at least two roots with similar mean<strong>in</strong>g (or, for<br />
lexically <strong>in</strong>frequent clusters, 1 root) accord<strong>in</strong>g to the method described<br />
<strong>in</strong> 3.1.<br />
2Svensk Baklängesordbok conta<strong>in</strong>s lemmas from SAOL and from newspaper articles,<br />
and does not exactly match Svensk Ordbok.<br />
3 Approximately 22% of the f<strong>in</strong>al clusters from NFO4 are sound symbolic, cf. table<br />
5.17.<br />
144
%<br />
/tSa/ (sound, quick or strong movement)<br />
/f:a/ (slang4, quick or strong movement)<br />
/ska/ (wetness, pejorative)<br />
/bla/ (talk<strong>in</strong>g)<br />
/fla/ (pejorative, talk<strong>in</strong>g)<br />
/Nla/ (quick or strong movement)<br />
/l:a/ (pejorative, round form, walk<strong>in</strong>g)<br />
/mla/ (quick or strong movement, talk<strong>in</strong>g, sound)<br />
/pla/ (quick or strong movement) 1/15 100%<br />
/rpla/ (sound) 1/1 100%<br />
/spla/ (quick or strong movement) 1/1 100%<br />
/rla/ (sound)<br />
/sla/ (sound, talk<strong>in</strong>g)<br />
/ampa/ (walk<strong>in</strong>g)<br />
/mpa/ (walk<strong>in</strong>g, pejorative, quick or strong<br />
movement, short-wide form)<br />
/p:a/ (quick or strong movement)<br />
/bra/ (talk<strong>in</strong>g) 1/1 100%<br />
/dra/ (talk<strong>in</strong>g, quick or strong movement)<br />
/<strong>in</strong>dra/ (light)<br />
/lra/ (quick or strong movement, sound, talk<strong>in</strong>g)<br />
/imra/ (light)<br />
/r:a/ (quick or strong movement, sound, talk<strong>in</strong>g)<br />
/tra/ (quick or strong movement, sound, talk<strong>in</strong>g,<br />
pejorative)<br />
/fsa/ (pejorative, quick or strong movement,<br />
walk<strong>in</strong>g, sound)<br />
/msa/ (long th<strong>in</strong> form, pejorative, talk<strong>in</strong>g, quick or<br />
strong movement)<br />
/lta/ (walk<strong>in</strong>g, quick or strong movement)<br />
4 'Slang' is a stylistic feature and thus belongs to another dimension, cf. 4.2.3.<br />
5 1/1 means that there is one root morpheme which is sound symbolic.<br />
145
sta/ (talk<strong>in</strong>g, quick or strong movement)<br />
Examples of sound symbolic root morphemes end<strong>in</strong>g with these f<strong>in</strong>al<br />
clusters from Svensk Baklängesordbok are found <strong>in</strong> table 5.2 below.<br />
Some f<strong>in</strong>al clusters are rare and occur <strong>in</strong> only one or a few words. These<br />
are (translated <strong>in</strong> the table below): klatscha, ratscha, rutscha (three out of<br />
four root morphemes), knaggla, raggla, traggla (three out of four root<br />
morphemes), porla, sorla (two out of four root morphemes) and stöppla,<br />
sörpla, haspla, slabbra (there is only one root morpheme for each of these<br />
last four f<strong>in</strong>al clusters).<br />
146
Table 5.2 <strong>Sound</strong> symbolic roots,<br />
translation and categorization.<br />
klatscha crack sound<br />
ratscha rip sound<br />
rutscha slide quick or strong<br />
movement<br />
haffa nab slang<br />
blaffa large (ugly) color slang<br />
patch<br />
klaffa tally slang<br />
fiffa smarten (up) slang<br />
sniffa sniff slang<br />
piffa smarten (up) slang<br />
skoffa shovel slang<br />
roffa rob slang<br />
buffa nudge slang<br />
skuffa push slang<br />
luffa tramp slang<br />
fluffa fluff (up) slang<br />
knuffa push quick or strong<br />
movement<br />
puffa push quick or strong<br />
movement<br />
ruffa foul slang<br />
gruffa squabble slang<br />
tuffa puff slang<br />
töffa puff slang<br />
blaska splash wetness<br />
plaska splash wetness<br />
sjaska soil pejorative<br />
slaska splash wetness<br />
smaska slurp wetness<br />
snaska munch wetness<br />
vaska wash wetness<br />
babbla babble talk<strong>in</strong>g<br />
rabbla rattle talk<strong>in</strong>g<br />
gaffla gabble talk<strong>in</strong>g<br />
taffla muck th<strong>in</strong>gs up pejorative<br />
fiffla fiddle pejorative<br />
ruffla fiddle pejorative<br />
knaggla plod along movement<br />
traggla go on about talk<strong>in</strong>g<br />
dangla dangle quick or strong<br />
movem.<br />
rangla be lanky quick or strong<br />
movem.<br />
skrangla be rickety quick or strong<br />
movem.<br />
d<strong>in</strong>gla dangle quick or strong<br />
movem.<br />
147<br />
p<strong>in</strong>gla t<strong>in</strong>kle sound<br />
r<strong>in</strong>gla coil quick or strong<br />
movem.<br />
kr<strong>in</strong>gla pretzel w<strong>in</strong>d<strong>in</strong>g form<br />
v<strong>in</strong>gla stagger quick or strong<br />
movem.<br />
fjolla foolish woman pejorative<br />
lolla scatter-bra<strong>in</strong>ed<br />
woman<br />
pejorative<br />
rolla pa<strong>in</strong>t with a round form<br />
roller<br />
skrolla scroll round form<br />
stolla fool of a woman pejorative<br />
bulla bun round form<br />
lulla reel walk<strong>in</strong>g<br />
rulla roll round<br />
drulla blunder walk<strong>in</strong>g<br />
krulla curl round form<br />
famla grope quick or strong<br />
movem.<br />
ramla tumble fall<strong>in</strong>g<br />
skramla rattle sound<br />
svamla drivel talk<strong>in</strong>g<br />
strimla strip long th<strong>in</strong> form<br />
vimla swarm quick or strong<br />
movem.<br />
tromla rotat<strong>in</strong>g quick or strong<br />
cyl<strong>in</strong>drical sieve movement<br />
fumla fumble quick or strong<br />
movem.<br />
humla bumble-bee sound<br />
mumla mumble talk<strong>in</strong>g<br />
rumla be on the spree quick or strong<br />
movement<br />
drumla blunder quick or strong<br />
movem.<br />
tumla tumble quick or strong<br />
movem.<br />
stöppla give pa<strong>in</strong>ted quick or strong<br />
surface a certa<strong>in</strong><br />
look by strik<strong>in</strong>g a<br />
brush aga<strong>in</strong>st it<br />
movement<br />
sörpla dr<strong>in</strong>k noisily sound<br />
haspla reel quick or strong<br />
movem.<br />
porla ripple sound<br />
sorla ripple sound<br />
rassla rattle sound<br />
prassla rustle sound<br />
tassla whisper talk<strong>in</strong>g<br />
gnissla squeak sound<br />
tissla whisper talk<strong>in</strong>g<br />
vissla whistle sound<br />
rossla wheeze talk<strong>in</strong>g<br />
klampa tramp walk<strong>in</strong>g
slampa slut pejorative<br />
trampa tramp walk<strong>in</strong>g<br />
stampa stamp walk<strong>in</strong>g<br />
dimpa tumble quick or strong<br />
movem.<br />
fimpa stub short-wide form<br />
limpa loaf short-wide form<br />
klimpa get lumpy short-wide form<br />
skvimpa splash to and fro quick or strong<br />
movement<br />
fjompa be silly pejorative<br />
dumpa dump slang<br />
gumpa jog quick or strong<br />
jumpa jump from one<br />
piece of float<strong>in</strong>g<br />
ice to another<br />
movem.<br />
walk<strong>in</strong>g<br />
skumpa jog quick or strong<br />
movem.<br />
klumpa form lumps short-wide form<br />
plumpa make blots form<br />
pumpa pump quick or strong<br />
movem.<br />
rumpa rump slang<br />
sumpa blow a th<strong>in</strong>g slang<br />
stumpa t<strong>in</strong>y tot dim<strong>in</strong>utive<br />
klappa pat quick or strong<br />
movement<br />
snappa snatch quick or strong<br />
movem.<br />
rappa strike quick or strong<br />
movem.<br />
tappa drop quick or strong<br />
movem.<br />
greppa grip quick or strong<br />
movem.<br />
steppa tap-dance quick or strong<br />
movem.<br />
kippa flop about quick or strong<br />
movem.<br />
skippa skip quick or strong<br />
movem.<br />
trippa trip along walk<strong>in</strong>g<br />
tippa tip over quick or strong<br />
movem.<br />
vippa sw<strong>in</strong>g up and quick or strong<br />
down<br />
movem.<br />
hoppa jump quick or strong<br />
movem.<br />
loppa flea dim<strong>in</strong>utive<br />
glopp sleet wetness<br />
moppa mop quick or strong<br />
movem.<br />
noppa pluck quick or strong<br />
movem.<br />
snoppa top and tail quick or strong<br />
movem.<br />
148<br />
poppa pop up quick or strong<br />
movem.<br />
droppa drip quick or strong<br />
movem.<br />
proppa cram quick or strong<br />
movem.<br />
stropp sl<strong>in</strong>g round form<br />
guppa jolt quick or strong<br />
movem.<br />
knäppa flick quick or strong<br />
movem.<br />
snäppa snap one's f<strong>in</strong>gers quick or strong<br />
movem.<br />
slabbra chatter talk<strong>in</strong>g<br />
bladdra talk nonsense talk<strong>in</strong>g<br />
fladdra flutter quick or strong<br />
movem<br />
pladdra babble talk<strong>in</strong>g<br />
sladdra chatter talk<strong>in</strong>g<br />
bluddra talk nonsense talk<strong>in</strong>g<br />
sluddra slur one's words talk<strong>in</strong>g<br />
bläddra turn over the quick or strong<br />
pages<br />
movem.<br />
gl<strong>in</strong>dra gleam light<br />
t<strong>in</strong>dra tw<strong>in</strong>kle light<br />
dallra wobble quick or strong<br />
movement<br />
skallra rattle sound<br />
pillra potter at quick or strong<br />
movem.<br />
tillra trickle quick or strong<br />
movem.<br />
kvillra ripple, twitter sound<br />
jollra babble talk<strong>in</strong>g<br />
pjollra babble talk<strong>in</strong>g<br />
knollra curl round form<br />
bullra rumble sound<br />
mullra rumble sound<br />
myllra swarm quick or strong<br />
movem.<br />
bjällra bell sound<br />
skimra shimmer light<br />
flimra flicker quick or strong<br />
movem., light<br />
glimra gleam light<br />
darra tremble quick or strong<br />
movem.<br />
blarra talk nonsense talk<strong>in</strong>g<br />
knarra creak sound<br />
snarra burr talk<strong>in</strong>g<br />
irra wander about quick or strong<br />
movem.<br />
dirra tremble quick or strong<br />
movem.<br />
klirra j<strong>in</strong>gle sound
knirra creak sound<br />
pirra t<strong>in</strong>gle bodily feel<strong>in</strong>g<br />
stirra stare gaze<br />
virra wander about quick or strong<br />
movem.<br />
kvirra make a fuss talk<strong>in</strong>g<br />
svirra whirl quick or strong<br />
movem.<br />
skorra burr sound<br />
morra growl talk<strong>in</strong>g<br />
knorra grouse talk<strong>in</strong>g<br />
burra ruffle up quick or strong<br />
movem.<br />
hurra hurrah talk<strong>in</strong>g<br />
kurra rumble sound<br />
plurra fall <strong>in</strong>to the water quick or strong<br />
movement<br />
murra growl talk<strong>in</strong>g<br />
snurra sp<strong>in</strong> quick or strong<br />
movem.<br />
surra hum sound<br />
tjattra jabber talk<strong>in</strong>g<br />
klattra be awkward pejorative<br />
smattra clatter sound<br />
knattra rattle sound<br />
snattra quack sound<br />
skvattra quack sound<br />
glittra glitter quick or strong<br />
movement<br />
splittra spl<strong>in</strong>ter quick or strong<br />
movement<br />
fnittra giggle talk<strong>in</strong>g<br />
knittra sound high<br />
pitched and<br />
sound<br />
iterated<br />
kvittra chirp sound<br />
klottra scrawl quick or strong<br />
movem.<br />
plottra potter about pejorative<br />
knottra get goose- rough surface<br />
pimples structure<br />
huttra shiver quick or strong<br />
movement<br />
kuttra coo sound<br />
muttra mutter talk<strong>in</strong>g<br />
puttra chug sound<br />
hafsa scamp a th<strong>in</strong>g pej. quick or<br />
strong movement<br />
sjafsa shuffle quick or strong<br />
movement<br />
tjafsa talk drivel pej., talk<strong>in</strong>g<br />
lafsa shuffle pej., walk<strong>in</strong>g<br />
klafsa squelch sound, wetness,<br />
walk<strong>in</strong>g<br />
slafsa slop sound, pejorative<br />
149<br />
nafsa snap quick or strong<br />
movem.<br />
snafsa snap quick or strong<br />
movem.<br />
rafsa rummage quick or strong<br />
movem.<br />
krafsa scratch quick or strong<br />
movem.<br />
tafsa fiddle about with quick or strong<br />
a th<strong>in</strong>g movem.<br />
lufsa lumber walk<strong>in</strong>g<br />
glufsa gobble down pejorative<br />
plufsa plop quick or strong<br />
movem.<br />
rufsa ruffle quick or strong<br />
movem.<br />
tufsa tousle quick or strong<br />
movem.<br />
gläfsa yelp sound<br />
räfsa rake quick or strong<br />
movem.<br />
jamsa talk nonsense talk<strong>in</strong>g, pej.<br />
flamsa fool about pejorative<br />
slamsa rag slackness, long<br />
th<strong>in</strong> form<br />
ramsa str<strong>in</strong>g long (th<strong>in</strong>) form<br />
tramsa talk rubbish talk<strong>in</strong>g, pej.<br />
remsa strip long th<strong>in</strong> form<br />
slimsa rag long th<strong>in</strong> form<br />
plumsa plop quick or strong<br />
movem.<br />
mumsa munch quick or strong<br />
movem.<br />
grumsa grumble talk<strong>in</strong>g<br />
dalta coddle pejorative<br />
halta limp walk<strong>in</strong>g<br />
palta muffle up quick or strong<br />
movem.<br />
skralta be rickety quick or strong<br />
movem.<br />
skvalta ripple wetness, quick or<br />
strong movem.<br />
bulta beat beat<br />
rulta waddle walk<strong>in</strong>g<br />
tulta toddle about walk<strong>in</strong>g<br />
stulta toddle walk<strong>in</strong>g<br />
bylta muffle up quick or strong<br />
movem<br />
stylta walk stiff-legged walk<strong>in</strong>g<br />
gasta yell talk<strong>in</strong>g<br />
hasta hasten quick or strong<br />
movem.<br />
kasta throw quick or strong<br />
movem.<br />
rista cut quick or strong<br />
movem.
ista burst destruction<br />
hosta cough talk<strong>in</strong>g<br />
pusta pant talk<strong>in</strong>g<br />
frusta snort talk<strong>in</strong>g<br />
hysta throw quick or strong<br />
movem.<br />
150<br />
nysta w<strong>in</strong>d quick or strong<br />
movem.<br />
knysta breath<strong>in</strong>g a word talk<strong>in</strong>g<br />
krysta bear down talk<strong>in</strong>g<br />
5.2. Summary of the analysis of semantic features<br />
for f<strong>in</strong>al clusters<br />
The most common semantic feature of the f<strong>in</strong>al clusters is 'quick or<br />
strong movement', which is present <strong>in</strong> 15 f<strong>in</strong>al clusters. Table 5.3<br />
compresses all the features and f<strong>in</strong>al clusters of table 5.2. The table can<br />
also be compared with table 5.1, where <strong>in</strong>stead the semantic features for<br />
each f<strong>in</strong>al cluster are shown. The semantic features are shown, <strong>in</strong><br />
descend<strong>in</strong>g order, <strong>in</strong> table 5.3.<br />
Table 5.3 The semantic features of roots from Svensk Baklängesordbok.<br />
Listed are also the clusters correspond<strong>in</strong>g to each feature and the total<br />
frequencies of clusters.<br />
Semantic feature Frequency of<br />
f<strong>in</strong>al clusters<br />
F<strong>in</strong>al clusters6 quick or strong movement 85 ffa, Nla, pla, spla,<br />
mpa, mla, ppa, dra,<br />
lra, rra, tra, fsa, msa,<br />
lta, sta<br />
talk<strong>in</strong>g 37 bla, fla, mla, sla, bra,<br />
lra, rra, msa, dra,<br />
tra, sta, ska<br />
sound 34 tSa, mla, rpla, rla,<br />
sla, lra, rra, tra, fsa<br />
pejorative 20 fla, lla, mpa, tra, fsa,<br />
msa<br />
slang 19 ffa, mpa<br />
walk<strong>in</strong>g 15 lla, ampa, fsa, lta<br />
wetness 9 ska<br />
long th<strong>in</strong> form 5 msa<br />
round form 7 lla<br />
light 5 <strong>in</strong>dra, imra<br />
short-wide form 4 mpa<br />
6 Double consonant grapheme stands for a phonologically long consonant.
In table 5.4 the rank<strong>in</strong>g of table 5.3 is compared with diagram 4.5, which<br />
shows the most frequent mean<strong>in</strong>gs for i n i t i a l consonant clusters.<br />
Table 5.4 The most frequent semantic features for <strong>in</strong>itial and f<strong>in</strong>al<br />
clusters.<br />
rank <strong>in</strong>itial clusters f<strong>in</strong>al clusters<br />
1 pejorative quick or strong<br />
movement<br />
2 sound talk<strong>in</strong>g<br />
3 long th<strong>in</strong> form sound<br />
4 quick or strong pejorative<br />
movement<br />
5 wetness slang<br />
6 talk<strong>in</strong>g walk<strong>in</strong>g<br />
7 light wetness<br />
8 dim<strong>in</strong>utive long th<strong>in</strong> form<br />
9 round form round form<br />
10 walk<strong>in</strong>g light<br />
11 destruction short-wide form<br />
The semantic features <strong>in</strong> bold type are those n<strong>in</strong>e that are among the<br />
eleven most common features for both groups. '<strong>Sound</strong>' is ranked second<br />
for <strong>in</strong>itial clusters and third for f<strong>in</strong>al clusters. 'Quick or strong<br />
movement' is ranked first for f<strong>in</strong>al clusters and fourth for <strong>in</strong>itial clusters.<br />
'Pejorative' is ranked fourth for f<strong>in</strong>al clusters but first for <strong>in</strong>itial clusters.<br />
The semantic features of the six most frequent f<strong>in</strong>al clusters are among<br />
the semantic features of the ten most frequent <strong>in</strong>itial clusters. In other<br />
words, many of the most common semantic features are the same for<br />
<strong>in</strong>itial clusters and f<strong>in</strong>al clusters. However, 'pejorative' is not as common<br />
<strong>in</strong> f<strong>in</strong>al clusters as <strong>in</strong> <strong>in</strong>itial.<br />
Some clusters can occur both <strong>in</strong>itially and f<strong>in</strong>ally (before a), namely sk,<br />
bl, fl, spl, sl, br, dr, tr and st. Four of these can have the same<br />
semantic feature. These are bl (talk<strong>in</strong>g), fl (pejorative), sl (talk<strong>in</strong>g), tr<br />
(quick or strong movement, sound, talk<strong>in</strong>g, pejorative). However, there<br />
are no conventional words that both beg<strong>in</strong> and end with these<br />
comb<strong>in</strong>ations.<br />
151
5.3 Properties of consonant clusters of Nusvensk<br />
Frekvensordbok<br />
Us<strong>in</strong>g the root analysis with paraphrases of Nusvensk Frekvensordbok 4<br />
(NFO4), an analysis of f<strong>in</strong>al clusters that resembles the analysis of <strong>in</strong>itial<br />
clusters was made, i.e. sound symbolic roots were excerpted from a<br />
larger set of roots with the aid of key words (cf. Appendix 1). NFO4 is a<br />
frequency dictionary of Swedish morphemes, and the material consists of<br />
1 million words from newspaper articles. List 4.3 of NFO4 conta<strong>in</strong>s the<br />
roots ordered <strong>in</strong> reverse. The reason not to analyze only the NFO4<br />
material is that there are many more words <strong>in</strong> Svensk Baklängesordbok.<br />
The roots from the NFO4 material are chosen <strong>in</strong> the follow<strong>in</strong>g way: they<br />
shall conta<strong>in</strong> at least one root end<strong>in</strong>g of a cluster, if this root belongs to<br />
one of the semantic categories listed <strong>in</strong> chapter 4.<br />
The most common clusters, absolutely and percentally, are shown <strong>in</strong><br />
diagrams 5.1 and 5.2, and <strong>in</strong> the follow<strong>in</strong>g diagrams (5.3 to 5.11) the<br />
semantic profiles of some of the most common clusters are shown.<br />
The first diagram (5.1) shows how the clusters rank when the total<br />
number of sound symbolic roots are counted. -Nk and -sk are the clusters<br />
with the largest number of sound symbolic roots. Thereafter follow - fs, -<br />
nd and -tr. -Nk, -sk, -fs and -tr are described <strong>in</strong> detail below. Diagram<br />
5.1 corresponds to diagram 4.1 of <strong>in</strong>itial clusters.<br />
The next diagram (5.2) shows the clusters that have the highest degree of<br />
sound symbolic roots. The first five have a percentage of 100% because<br />
there is only one – sound symbolic – root, for every cluster. These are<br />
not studied <strong>in</strong> further detail. The follow<strong>in</strong>g ones with quite a high<br />
percentage are -fs, -dr and -lr, followed by -ml and -Nl, These are<br />
described <strong>in</strong> detail below. Also -bl is described. Diagram 5.2 corresponds<br />
to diagram 4.4 of <strong>in</strong>itial clusters.<br />
152
20<br />
18<br />
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
ksn<br />
ltn<br />
spl<br />
mj<br />
rS<br />
jf<br />
lb<br />
mt<br />
lt<br />
ps<br />
rl<br />
Sk<br />
rj<br />
ld<br />
ls<br />
lm<br />
rv<br />
nt<br />
ns<br />
ms<br />
sl<br />
Nl<br />
bl<br />
lk<br />
lr<br />
dr<br />
st<br />
ml<br />
mp<br />
tr<br />
nd<br />
fs<br />
sk<br />
Nk<br />
Diagram 5.1 7 More and less sound symbolic f<strong>in</strong>al clusters. Number of motivated root morphemes per cluster.<br />
7 In this and the follow<strong>in</strong>g diagrams "N" stands for [N] and "S" stands for [S].
100<br />
90<br />
80<br />
70<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
nt<br />
lt<br />
st<br />
ns<br />
rj<br />
ld<br />
nd<br />
rv<br />
ps<br />
rl<br />
mt<br />
ls<br />
lk<br />
lm<br />
mp<br />
sk<br />
bl<br />
ms<br />
sl<br />
tr<br />
Nk<br />
Sk<br />
Nl<br />
ml<br />
lr<br />
dr<br />
ksn<br />
fs<br />
ltn<br />
spl<br />
mj<br />
rS<br />
jf<br />
lb<br />
Diagram 5.2 Percent of motivated root morphemes per cluster, for f<strong>in</strong>al consonant clusters.
The follow<strong>in</strong>g diagrams, 5.3 to 5.11 show the semantic profiles for<br />
different f<strong>in</strong>al clusters. These diagrams can be compared with the<br />
diagrams 4.16 – 4.23, which show the semantic profiles for different<br />
<strong>in</strong>itial clusters. Diagram 5.3 shows the semantic profile of -fs.<br />
60<br />
50<br />
40<br />
30<br />
20<br />
10<br />
0<br />
pejorative sound quickness<br />
Diagram 5.3 Semantic features of the cluster -fs. Percentages concern the<br />
number of semantic features <strong>in</strong> proportion to all roots of the cluster.<br />
Diagram 5.3 shows the semantic features of the consonant cluster -fs. The<br />
words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g features are<br />
shown <strong>in</strong> table 5.5. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 5.5 The sound symbolic morphemes of the -fs cluster, the semantic<br />
features <strong>in</strong>volved, and their absolute frequencies.<br />
'pejorative': hafs (slovenl<strong>in</strong>ess), tjafs<br />
(drivel), slafs (slopp<strong>in</strong>ess), rafs (trash),<br />
krafs (knick-knacks), lufsa (shamble),<br />
kalufs (forelock), rufsa (ruffle),<br />
tufsa (tousle), bjäfs (gewgaws) 10<br />
'sound': klafs (squelch), gläfs (yelp) 2<br />
'quickness': nafs (snap), rafs (trash) 2<br />
155<br />
%
10<br />
9<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
0<br />
long th<strong>in</strong> form<br />
quick or strong<br />
movement<br />
sound<br />
Diagram 5.4 The semantic profile of -Nk. Percentages concern the<br />
number of semantic features <strong>in</strong> proportion to all roots of the cluster.<br />
Diagram 5.4 shows the semantic features of the consonant cluster -Nk. In<br />
table 5.6. the words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g<br />
features are shown. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 5.6 The sound symbolic morphemes of the -Nk cluster, the semantic<br />
features <strong>in</strong>volved, and their absolute frequencies.<br />
'long th<strong>in</strong> form': bank (bar), dank<br />
(th<strong>in</strong> candle), hank (a band), planka<br />
(plank), rank (tall and slender), sp<strong>in</strong>kig<br />
(slender), rynka (furrow) 7<br />
'quick or strong movement': fl<strong>in</strong>k<br />
(quick), sl<strong>in</strong>k (slip), v<strong>in</strong>k (wave) 3<br />
beat<br />
156<br />
slang<br />
wetness<br />
%
'sound': banka (knock), dunk (thump<strong>in</strong>g),<br />
stånka (puff and blow) 3<br />
'beat': banka (knock), dunka (thump) 2<br />
'slang': pank (broke), grunka (gadget) 2<br />
'wetness': klunk (gulp), stänk (splash) 2<br />
The follow<strong>in</strong>g diagram (5.5) shows the semantic profile of -sk.<br />
8<br />
7<br />
6<br />
5<br />
4<br />
3<br />
2<br />
1<br />
0<br />
wetness pejorative beat sound<br />
Diagram 5.5 The semantic profile of -sk. Percentages concern the number<br />
of semantic features <strong>in</strong> proportion to all roots of the cluster.<br />
Diagram 5.5 shows the semantic features of the consonant cluster -sk. The<br />
words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g features are<br />
shown <strong>in</strong> table 5.7. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 5.7 The sound symbolic morphemes of the -sk cluster, the semantic<br />
features <strong>in</strong>volved, and their absolute frequencies.<br />
'wetness': blask (slops), plask (splash),<br />
slask (slush), smaska (guzzle), vaska<br />
(wash), loska (spit), mäsk (mash),<br />
träsk (swamp) 8<br />
157<br />
%
'pejorative': sjask (cad), fnask<br />
(prostitute), snaska (be messy), fnoskig<br />
(dotty), fjäsk (fawn<strong>in</strong>g) 5<br />
'beat': daska (slap), plask (splash), piska (whip) 3<br />
'sound': pladask (flop), smaska (guzzle) 2<br />
The follow<strong>in</strong>g diagram (5.6) shows the semantic profile of -tr.<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
talk<strong>in</strong>g sound pejorative<br />
Diagram 5.6 The semantic profile of -tr. Percentages concern the number<br />
of semantic features <strong>in</strong> proportion to all roots of the cluster.<br />
The diagram 5.6 shows the semantic features of the consonant cluster -tr.<br />
In table 5.8 the words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g<br />
features can be studied. The absolute frequencies (not the percentages) of<br />
the semantic features are shown to the right.<br />
Table 5.8 The sound symbolic morphemes of the -tr cluster, the semantic<br />
features <strong>in</strong>volved, and their absolute frequencies.<br />
'talk<strong>in</strong>g': tjattra (jabber), snattra (quack),<br />
fnittra (giggle), muttra (mutter),<br />
yttra (utter) 5<br />
158<br />
%
'sound': smattra (clatter), knattra (rattle),<br />
kvittra (chirp), puttra (simmer) 4<br />
'pejorative': tjattra (jabber), plottra (fritter) 2<br />
The follow<strong>in</strong>g diagram (5.7) shows the semantic profile of -Nl.<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
quick or strong movement<br />
Diagram 5.7 The semantic profile of -Nl. The percentage is of the<br />
semantic feature, for all roots of the cluster.<br />
Diagram 5.7 shows the semantic feature of the consonant cluster -NNl. The<br />
words (represent<strong>in</strong>g root morphemes) of the only feature is shown <strong>in</strong><br />
table 5.9. The absolute frequency (not the percentage) of the semantic<br />
feature is shown to the right.<br />
159<br />
%
Table 5.9 The sound symbolic morphemes of the -Nl cluster, the semantic<br />
feature <strong>in</strong>volved, and its absolute frequency.<br />
'quick or strong movement': d<strong>in</strong>gla<br />
(dangle), s<strong>in</strong>gla (dance), v<strong>in</strong>gla (wobble),<br />
jonglera (juggle) 4<br />
The follow<strong>in</strong>g diagram (5.8) shows the semantic profile of -bl. The<br />
percentage is of the semantic feature for all roots of the cluster.<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
talk<strong>in</strong>g<br />
Diagram 5.8 The semantic profile of -bl. The percentage is of semantic<br />
features for all roots of the cluster.<br />
Diagram 5.8 shows the semantic feature of the consonant cluster -bl. The<br />
words (represent<strong>in</strong>g root morphemes) of the only feature is shown <strong>in</strong><br />
table 5.10. The absolute frequency (not the percentage) of the semantic<br />
feature is shown to the right.<br />
Table 5.10 The sound symbolic morphemes of the -bl cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'talk<strong>in</strong>g': babbla (babble), rabbla (rattle),<br />
käbbla (bicker), jubla (shout with joy) 4<br />
160<br />
%
The follow<strong>in</strong>g diagram (5.9) shows the semantic profile of -dr.<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
quick or<br />
strong<br />
movement<br />
talk<strong>in</strong>g pejorative<br />
Diagram 5.9 The semantic profile of -dr. Percentages concern the<br />
number of semantic features <strong>in</strong> proportion to all roots of the cluster.<br />
This diagram shows the semantic features of the consonant cluster -dr. In<br />
table 5.11 the words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g<br />
features are shown. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 5.11 The sound symbolic morphemes of the -dr cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'quick or strong movement': fladdra (flutter), 2<br />
bläddra (turn over the pages)<br />
'talk<strong>in</strong>g': pladdra (babble), sluddra (slur) 2<br />
'pejorative': pladdra (babble), sluddra (slur) 2<br />
The next diagram (5.10) shows the semantic profile of -lr. Percentages<br />
are of semantic features for all roots of the cluster.<br />
161<br />
%
45<br />
40<br />
35<br />
30<br />
25<br />
20<br />
15<br />
10<br />
5<br />
0<br />
sound quick or<br />
strong<br />
movement<br />
Diagram 5.10 The semantic profile of -lr. Percentages are of semantic<br />
features for all roots of the cluster.<br />
Diagram 5.10 shows the semantic features of the consonant cluster -lr.<br />
The words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g features are<br />
shown <strong>in</strong> table 5.12. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 5.12 The sound symbolic morphemes of the -lr cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'sound': skallra (rattle), bullra (make a noise),<br />
mullra (rumble), bjällra (j<strong>in</strong>gle) 4<br />
'quick or strong movement': dallra (tremble),<br />
myllra (swarm) 2<br />
The last diagram (5.11) shows the semantic profile of -ml.<br />
162<br />
%
20<br />
18<br />
16<br />
14<br />
12<br />
10<br />
8<br />
6<br />
4<br />
2<br />
0<br />
quick or<br />
strong<br />
movement<br />
movement sound<br />
Diagram 5.11 The semantic profile of -ml. Percentages concern the<br />
number of semantic features <strong>in</strong> proportion to all roots of the cluster.<br />
Diagram 5.11 shows the semantic features of the consonant cluster -ml.<br />
The words (represent<strong>in</strong>g root morphemes) of the dom<strong>in</strong>at<strong>in</strong>g features are<br />
shown <strong>in</strong> table 5.13. The absolute frequencies (not the percentages) of the<br />
semantic features are shown to the right.<br />
Table 5.13 The sound symbolic morphemes of the -ml cluster, the<br />
semantic features <strong>in</strong>volved, and their absolute frequencies.<br />
'quick or strong movement': vimla (swarm), 3<br />
fumla (fumble), tumla (tumble)<br />
'movement': famla (grope), ramla (fall down) 2<br />
'sound': skramla (rattle), humla (bumble-bee) 2<br />
163<br />
%
5.3.1 Summary of the analysis of properties of<br />
different f<strong>in</strong>al clusters<br />
As well as for the <strong>in</strong>itial clusters, the mean<strong>in</strong>g profiles for the sound<br />
symbolic f<strong>in</strong>al clusters vary. Different clusters are connected with<br />
different mean<strong>in</strong>gs. Also the f<strong>in</strong>al clusters have different semantic<br />
profiles. Some f<strong>in</strong>al clusters have several semantic features, others have<br />
only one.<br />
Of these absolutely and percentally most common clusters (-Nk, -sk,-fs, -<br />
nd, -tr, -dr, -lr, -ml, -Nl), 5 out of 9 end with a liquid and are ma<strong>in</strong>ly<br />
connected with 'quick or strong movement', 'talk<strong>in</strong>g', 'sound', and<br />
'pejorative'. -sk is dom<strong>in</strong>ated by 'wetness' and -fs is clearly 'pejorative'<br />
5.4. Vowels<br />
As mentioned earlier, vowels may also have sound symbolic mean<strong>in</strong>g. To<br />
delimit the analysis of vowels, the contrastive effect of different vowels<br />
was studied.<br />
5.4.1 Vowel pairs and triplets<br />
In order to <strong>in</strong>vestigate the semantic effect of different vowels, word pairs<br />
were searched for. However, there are not so many cases of word pairs<br />
with contrast<strong>in</strong>g vowels. Those that were found contrast the vowels [i]-[a],<br />
[i]-[a]-[u], [i]-[o], [i]-[P]. Most of them imitate sound (8 out of 19). The<br />
examples (taken from Svensk Baklängesordbok) are: klibba-klabba<br />
(adhesion 8), slibba-slabba (wetness), slidder-sladder 9 (talk<strong>in</strong>g, pejorative),<br />
klifsa-klafsa (wetness, sound), klitsch-klatsch (sound), ritsch-ratsch<br />
(sound), snick-snack (talk<strong>in</strong>g), r<strong>in</strong>gla-rangla (movement), tissla-tassla<br />
(talk<strong>in</strong>g), rispa-raspa (rough surface structure), dirr and darr<br />
(movement), knirra-knarra (sound), knistra-knastra (sound), knittraknattra<br />
(sound), slimsa-slamsa (form) snipp-snapp-snorum (<strong>in</strong> nonsense<br />
verse), dripp-dropp (sound), tick-tock (sound), klimp-klump (form).<br />
From the def<strong>in</strong>itions <strong>in</strong> Svensk Ordbok it is not evident that there is a<br />
mean<strong>in</strong>g dist<strong>in</strong>ction (size or pitch) between [i] and [a] except for some<br />
cases, but clearly [i] never stands for anyth<strong>in</strong>g large or low-pitched. (For<br />
an <strong>in</strong>vestigation of English reduplicatives see Thun, 1963.)<br />
8Shown with<strong>in</strong> parenthesis is the semantic feature for both words <strong>in</strong> the pair. Some<br />
words share two semantic features.<br />
9 In fact, many of these pairs occur as semi-lexicalized compounds, e.g. <strong>in</strong> poetry and<br />
verse.<br />
164
However, there are a number of words found <strong>in</strong> Svensk Baklängesordbok<br />
which show that root morphemes with end<strong>in</strong>gs conta<strong>in</strong><strong>in</strong>g [i] often have<br />
the mean<strong>in</strong>gs 'smallness', 'quickness' or 'high pitch', while [P] has the<br />
mean<strong>in</strong>gs of darker sounds or bigger movements – while [a] seems quite<br />
neutral. Compar<strong>in</strong>g the end<strong>in</strong>gs [<strong>in</strong>ka] and [Pnka] <strong>in</strong> table 5.14, there is, <strong>in</strong><br />
bl<strong>in</strong>ka (tw<strong>in</strong>kle) and lunka (jog along), a contrast <strong>in</strong> small, quick<br />
movement versus big, slow movement. In kl<strong>in</strong>ka (t<strong>in</strong>kle) and dunka<br />
(thump) there is a contrast between high pitched and low pitched sound.<br />
Table 5.14 The vowels i, P, a before the end<strong>in</strong>g -NNka<br />
k<strong>in</strong>ka wh<strong>in</strong>e<br />
l<strong>in</strong>ka limp<br />
bl<strong>in</strong>ka tw<strong>in</strong>kle<br />
kl<strong>in</strong>ka t<strong>in</strong>kle<br />
sl<strong>in</strong>ka slip<br />
sp<strong>in</strong>ka split (wood)<br />
v<strong>in</strong>ka wave (hand)<br />
dunka thump<br />
pjunka coddle<br />
lunka jog along<br />
klunka gulp down<br />
runka shake<br />
banka bang<br />
klanka grouse<br />
planka plank<br />
slanka make slim<br />
ranka creeper<br />
vanka saunter<br />
svanka be sway-backed<br />
The vowels with the end<strong>in</strong>g -lra can also be compared. Most of the -lra<br />
words imitate sounds or movements where i clearly stands for smallness<br />
or high pitch and [P] for low pitch, e.g. tillra (trickle) and mullra<br />
(rumble).<br />
165
Table 5.15 The vowels a, i, O, P before the end<strong>in</strong>g -lra.<br />
dallra quiver<br />
skallra rattle<br />
pillra potter at<br />
tillra trickle<br />
kvillra ripple, twitter<br />
jollra prattle<br />
pjollra twaddle<br />
knollra frizz<br />
bullra rumble<br />
mullra rumble<br />
The i and [P] vowels <strong>in</strong> words end<strong>in</strong>g with -ttra show similar tendencies,<br />
cf. e.g. kvittra (twitter) and muttra (mutter).<br />
Table 5.16 The vowels a, i, O, P before the end<strong>in</strong>g -tra.<br />
tjattra chatter<br />
klattra make a fuss<br />
smattra clatter<br />
knattra rattle<br />
snattra quack<br />
skvattra quack<br />
glittra glitter<br />
splittra spl<strong>in</strong>ter<br />
fnittra giggle<br />
knittra sound high<br />
pitched and<br />
iterated<br />
kvittra twitter<br />
knottra get goosepimples<br />
huttra shiver<br />
kuttra coo<br />
muttra mutter<br />
puttra chug<br />
5.4.2 Vowels <strong>in</strong> light/gaze-words<br />
The follow<strong>in</strong>g list conta<strong>in</strong>s most light or gaze root morphemes:<br />
blek (pale), blick (gaze), bliga (stare), bl<strong>in</strong>d (bl<strong>in</strong>d), bl<strong>in</strong>ka (bl<strong>in</strong>k), blixt<br />
(lightn<strong>in</strong>g), blossa (flare), blända (bl<strong>in</strong>d), blänga (glare), blänka (sh<strong>in</strong>e),<br />
flimra (flicker), glana (stare), glans (luster), glatt (glossy), glimma<br />
(gleam), glimra (gleam), glimta (gleam), gl<strong>in</strong>dra (gleam), glisa (sh<strong>in</strong>e),<br />
glittra (glitter), glo (stare), gloria (halo), glutta (take a glance), glåmig<br />
166
(pale), glänsa (sh<strong>in</strong>e), glöd (glow), gnista (spark), gnistra (sparkle), t<strong>in</strong>dra<br />
(tw<strong>in</strong>kle), skimra (shimmer).<br />
The stressed vowels of these 30 root morphemes are <strong>in</strong> 16 cases i, <strong>in</strong> 1<br />
case e, <strong>in</strong> 4 cases E, <strong>in</strong> 1 case Ø, <strong>in</strong> 3 cases a (2 [a] and 1 [A:]), <strong>in</strong> 2 cases o,<br />
<strong>in</strong> 2 cases u, and <strong>in</strong> 1 case [P]. The dom<strong>in</strong>at<strong>in</strong>g vowel is clearly i.<br />
Furthermore, 24 of the vowels are front vowels; only 5 are back vowels<br />
and 1 is central. It seems thus that the feature 'light' is connected with<br />
front vowels (high F2).<br />
5.4.3 The vowel [P]<br />
S<strong>in</strong>ce there seemed to be a tendency for words with [P] (a short, medial,<br />
half closed, rounded vowel) to have a pejorative mean<strong>in</strong>g (see table 5.2),<br />
this vowel was studied more closely. In the whole material (Appendix 1)<br />
the number of pejorative features is 163. Of these features 33 belong to<br />
roots with [P], i.e. 20%. The number of non-pejorative features is 789 and<br />
70 of these belong to roots with a [P], i.e. 8%. A Chi square test showed<br />
that there is a significant correlation between 'pejorative' and [P] (p. <<br />
0.0001).<br />
5.4.4 Summary of vowels<br />
From the examples <strong>in</strong> this material it seems quite clear that stressed<br />
vowels are connected with the semantic dimensions 'size' and 'high-low<br />
pitch' (which are connected through the frequency code); i tends to have<br />
the mean<strong>in</strong>gs 'smallness', 'quickness', 'high pitch', while [P] seems to have<br />
the mean<strong>in</strong>gs of 'low pitch' and 'largeness'. a seems quite neutral. The<br />
special study of [P] showed that this sound is over-represented among<br />
roots with the pejorative feature.<br />
5.5. Comparisons of f<strong>in</strong>al clusters of roots from<br />
different sources<br />
5.5.1 NFO4 and Sigurd (1965)<br />
S<strong>in</strong>ce the question of which f<strong>in</strong>al comb<strong>in</strong>ations should be taken <strong>in</strong>to<br />
account is somewhat open, a comparison is made between NFO4 and<br />
Sigurd (1965). Table 5.17 shows the f<strong>in</strong>al clusters which are found <strong>in</strong> the<br />
two different sources. The comparison with Sigurd (1965) is <strong>in</strong>cluded to<br />
rem<strong>in</strong>d the reader that the f<strong>in</strong>al clusters of NFO4 are not exactly the same<br />
167
as those of other analyses and to show which clusters are found <strong>in</strong> both<br />
sources.<br />
168
Table 5.17 The f<strong>in</strong>al clusters of<br />
NFO4, those of NFO4 that are<br />
sound symbolic, and, as a<br />
comparison, the primary f<strong>in</strong>al<br />
clusters accord<strong>in</strong>g to Sigurd<br />
(1965).<br />
Clusters<br />
for all<br />
reverse<br />
order<br />
sorted<br />
roots of<br />
NFO4<br />
Clusters<br />
for sound<br />
symbolic<br />
roots<br />
from<br />
NFO4<br />
lb + +<br />
mb +<br />
rb<br />
gd +<br />
jd +<br />
ld + +<br />
md +<br />
nd + +<br />
vd +<br />
jf + +<br />
lf +<br />
mf +<br />
rf +<br />
sf<br />
rg +<br />
dg<br />
rj + +<br />
ntg<br />
nS +<br />
rk<br />
nSS<br />
rS + +<br />
tS +<br />
dj<br />
lj +<br />
mj + +<br />
nj +<br />
rj +<br />
lC<br />
tj<br />
vj<br />
jk +<br />
lk + +<br />
mk<br />
Nk + +<br />
rk +<br />
sk + +<br />
lsk +<br />
msk +<br />
Primary<br />
clusters<br />
accord<strong>in</strong>g<br />
to Sigurd<br />
169<br />
nsk +<br />
psk +<br />
Sk + +<br />
bl +<br />
mbl<br />
ndl<br />
fl<br />
gl<br />
Nl +<br />
rjl<br />
gl<br />
jl<br />
kl<br />
NNkl<br />
rkl<br />
ml +<br />
pl<br />
mpl<br />
spl +<br />
rl +<br />
Nsl<br />
jsl<br />
nsl<br />
psl<br />
sl +<br />
ntl<br />
stl<br />
vl<br />
rvl<br />
ksl<br />
gm<br />
hm<br />
km<br />
lm + +<br />
rm +<br />
sm<br />
stm<br />
tm<br />
Nn +<br />
jn<br />
kn<br />
ln +<br />
mn +<br />
pn<br />
rn<br />
sn<br />
tn<br />
ltn +<br />
ksn + +<br />
jp<br />
lp<br />
mp + +<br />
rp +
sp +<br />
br<br />
rbr<br />
dr +<br />
ldr<br />
ndr<br />
dr +<br />
rdr<br />
fr<br />
gr<br />
kr<br />
Nkr<br />
lr +<br />
mr<br />
nr<br />
pr<br />
tr +<br />
rtr<br />
str<br />
kstr<br />
vr<br />
lvr<br />
bs<br />
ds<br />
fs + +<br />
js +<br />
ks +<br />
ls + +<br />
ms + +<br />
ns + +<br />
Ns +<br />
ps + +<br />
mps<br />
rps<br />
ts +<br />
ft +<br />
rft +<br />
kt<br />
Nt<br />
jt +<br />
kt +<br />
Nkt<br />
rkt<br />
lt + +<br />
mt +<br />
nt + +<br />
pt +<br />
st + +<br />
fst +<br />
Nst +<br />
jst +<br />
kst +<br />
mst +<br />
170<br />
nst +<br />
St +<br />
kv<br />
lv +<br />
rv + +<br />
nts
Table 5.17 shows the follow<strong>in</strong>g: There are, <strong>in</strong> the NFO-material, 151<br />
different end<strong>in</strong>gs of roots. Of these, 33 are used systematically for sound<br />
symbolism.<br />
The rightmost column shows which of these clusters are primary f<strong>in</strong>al<br />
comb<strong>in</strong>ations accord<strong>in</strong>g to Sigurd (1965). Quite a few root f<strong>in</strong>al<br />
comb<strong>in</strong>ations <strong>in</strong> NFO4 are not the same as Sigurd’s word f<strong>in</strong>al<br />
comb<strong>in</strong>ations. Also some of Sigurd’s f<strong>in</strong>al comb<strong>in</strong>ations do not occur <strong>in</strong><br />
the NFO-material: tsk, pst, lft, lst, NNd, NNt.<br />
5.5.2 The most frequent f<strong>in</strong>al clusters <strong>in</strong> NFO4<br />
The follow<strong>in</strong>g tables (5.18 and 5.19) show which f<strong>in</strong>al clusters are the<br />
most frequent for different semantic features, both absolutely and<br />
percentally. They show that fs is the most sound symbolic f<strong>in</strong>al cluster<br />
(pejorative) both <strong>in</strong> absolute numbers and percentally.<br />
Table 5.18 The most frequent f<strong>in</strong>al clusters <strong>in</strong> absolute numbers. Their<br />
semantic features are shown to the right.<br />
Cluster Cluster<br />
frequency<br />
Semantic feature<br />
fs 10 pejorative<br />
sk 8 wetness<br />
Nk<br />
sk<br />
tr<br />
st<br />
7<br />
5<br />
5<br />
5<br />
long th<strong>in</strong> form<br />
pejorative<br />
talk<strong>in</strong>g<br />
sound<br />
lk<br />
bl<br />
Nl<br />
sl<br />
4<br />
4<br />
4<br />
4<br />
long th<strong>in</strong> form<br />
talk<strong>in</strong>g<br />
quick or strong<br />
movement<br />
sound<br />
lr 4 sound<br />
tr<br />
ms<br />
4<br />
4<br />
sound<br />
pejorative<br />
171
Table 5.19 The most frequent f<strong>in</strong>al clusters <strong>in</strong> percent 10 of the total<br />
number of roots for clusters. Their semantic features are shown to the<br />
right.<br />
Cluster % Semantic feature<br />
fs 59 pejorative<br />
lr 44 sound<br />
Nl 40 quick or strong<br />
movement<br />
Sk 29 attitude<br />
sl 27 sound<br />
ms 27 pejorative<br />
dr 25 quick or strong<br />
movement<br />
dr 25 talk<strong>in</strong>g<br />
dr 25 pejorative<br />
bl 24 talk<strong>in</strong>g<br />
lr 22 quick or strong<br />
movement<br />
The accumulated frequencies of table 5.18 show that the most common<br />
semantic feature is 'pejorative' with 19 roots (distributed over 3 clusters)<br />
followed by 'sound' with 17 roots (distributed over 4 clusters), 'long th<strong>in</strong><br />
form' with 11 roots distributed over 2 clusters, and 'talk<strong>in</strong>g' with 9 roots<br />
distributed over 2 clusters.<br />
5.5.3 Discussion of f<strong>in</strong>al clusters <strong>in</strong> Svensk<br />
Baklängesordbok and <strong>in</strong> Nusvensk Frekvensordbok<br />
As po<strong>in</strong>ted out <strong>in</strong> the beg<strong>in</strong>n<strong>in</strong>g of this chapter, it is not self-evident which<br />
are the f<strong>in</strong>al clusters <strong>in</strong> Swedish. The semif<strong>in</strong>al clusters (before a) of<br />
Svensk Baklängesordbok (the Reverse Order Dictionary), the f<strong>in</strong>al<br />
clusters of Nusvensk Frekvensordbok 4, and of Sigurd (1965) are<br />
partially overlapp<strong>in</strong>g sets. For practical reasons, percentages are<br />
calculated only on Nusvensk Frekvensordbok 4.<br />
Likewise, the roots of the two dictionaries also constitute partially<br />
overlapp<strong>in</strong>g sets. The lemmas (from which roots were extracted) of<br />
Svensk Baklängesordbok come from the Wordlist of the Swedish<br />
Academy (Svenska Akademiens ordlista över svenska språket, 1974) and<br />
10 The follow<strong>in</strong>g clusters have a 100%-frequency because they have only one root, which<br />
is sound symbolic: lb (round form), jf (long th<strong>in</strong> form), rS (walk<strong>in</strong>g), mj (bodily<br />
feel<strong>in</strong>g), spl (quick or strong movement), ltn (destruction) and ksn (mental feel<strong>in</strong>g).<br />
172
from the same newspaper material as the material of Nusvensk<br />
Frekvensordbok. Consequently Svensk Baklängesordbok conta<strong>in</strong>s more<br />
roots. The similarities and the differences <strong>in</strong> the analyses of the two<br />
dictionaries are as follows (cf. also tables 5.4 and 5.24).<br />
The most common semantic features<br />
The most common semantic features are, <strong>in</strong> order:<br />
Svensk Baklängesordbok: 'quick or strong movement', 'talk<strong>in</strong>g', 'sound',<br />
'pejorative', 'slang', 'walk<strong>in</strong>g', 'wetness', 'long th<strong>in</strong> form', 'round form',<br />
'light', 'short-wide form'.<br />
Nusvensk Frekvensordbok 4: 'pejorative', 'sound', 'long th<strong>in</strong> form',<br />
'talk<strong>in</strong>g', 'wetness', 'quick or strong movement'.<br />
For both dictionaries the follow<strong>in</strong>g features are the same: 'quick or strong<br />
movement', 'talk<strong>in</strong>g', 'sound', 'pejorative', 'wetness' and 'long th<strong>in</strong> form'.<br />
(These six features are also the six most common features of <strong>in</strong>itial<br />
clusters, however <strong>in</strong> a different order, cf. table 5.4).<br />
The most common f<strong>in</strong>al consonant clusters (<strong>in</strong>clud<strong>in</strong>g all sound symbolic<br />
semantic features) are, <strong>in</strong> descend<strong>in</strong>g order (tables 5.20 and 5.21).<br />
Table 5.20 The most common f<strong>in</strong>al consonant clusters (for all sound<br />
symbolic semantic features) <strong>in</strong> Svensk Baklängesordbok. Cf. table 5.2.<br />
mp 20<br />
fs 18<br />
tr 18<br />
ml 13<br />
st 12<br />
lr 12<br />
lt 11<br />
ms 10<br />
Nl 8<br />
sl 7<br />
dr 7<br />
sk 6<br />
fl 4<br />
tS 3<br />
173
In Nusvensk Frekvensordbok 4 the most common f<strong>in</strong>al consonant clusters,<br />
<strong>in</strong>clud<strong>in</strong>g all sound symbolic semantic features, are (cf. diagram 5.1):<br />
Table 5.21 The most common f<strong>in</strong>al consonant clusters (for all sound<br />
symbolic semantic features) <strong>in</strong> Nusvensk Frekvensordbok 4.<br />
Nk 19<br />
sk 18<br />
fs 14<br />
nd 12<br />
tr 11<br />
mp 8<br />
ml 7<br />
st 7<br />
dr 6<br />
lr 6<br />
As can be seen <strong>in</strong> the two lists above, many, but not all, of the most<br />
common clusters are the same, namely mp, fs, tr, ml, st, lr, dr, and sk,<br />
<strong>in</strong> slightly different order.<br />
The follow<strong>in</strong>g clusters used sound symbolically are found <strong>in</strong> Svensk<br />
Baklängesordbok but not <strong>in</strong> Nusvensk Frekvensordbok 4: fl, tS, gl, pl,<br />
rpl and br.<br />
The differences between the two dictionaries seem to be greater for the<br />
consonant clusters than for the semantic features, but the most common<br />
consonant clusters are to a great extent the same. Study<strong>in</strong>g not only the<br />
most common clusters, there are more sound symbolic clusters <strong>in</strong> Svensk<br />
Baklängesordbok. This might be due to the fact, described above, that<br />
Svensk Baklängesordbok is based on a larger selection.<br />
5.6 Comb<strong>in</strong>ations of <strong>in</strong>itial and f<strong>in</strong>al consonant<br />
clusters<br />
Initial and f<strong>in</strong>al clusters can comb<strong>in</strong>e to create mean<strong>in</strong>g, e.g. 'wetness' as<br />
<strong>in</strong> blaska, plaska, slaska, smaska, snaska. The f<strong>in</strong>al comb<strong>in</strong>aton -ska is<br />
connected with 'wetness' and the <strong>in</strong>itial clusters bl-, pl-, sl-, sm- and sncan<br />
also stand for 'wetness' (among other th<strong>in</strong>gs).<br />
174
The f<strong>in</strong>al clusters are comb<strong>in</strong>ed <strong>in</strong> different ways with the <strong>in</strong>itial clusters,<br />
apart from with s<strong>in</strong>gle consonants. There are four groups:<br />
1. <strong>in</strong>itial cluster + no f<strong>in</strong>al cluster (e.g. skval 11)<br />
2. no <strong>in</strong>itial cluster + f<strong>in</strong>al cluster (e.g. babbla)<br />
3 a. <strong>in</strong>itial cluster + f<strong>in</strong>al cluster with the same mean<strong>in</strong>g<br />
(e.g. blaska, wetness+wetness)<br />
3 b. <strong>in</strong>itial cluster + f<strong>in</strong>al cluster with a different mean<strong>in</strong>g<br />
(e.g. drumla, pejorative + quick or strong movement)<br />
The analysis of group 1 is based on the analysis of chapter 4 (of Svensk<br />
Ordbok). The follow<strong>in</strong>g analysis of groups 2 and 3 is based on the f<strong>in</strong>al<br />
clusters of Svensk Baklängesordbok. As the analysis <strong>in</strong> 5.5. shows, the<br />
most common semantic features are the same for both <strong>in</strong>itial clusters and<br />
f<strong>in</strong>al clusters.<br />
5.6.1 Initial cluster + no f<strong>in</strong>al cluster<br />
In order to study sound symbolism of the <strong>in</strong>itial clusters isolated from<br />
f<strong>in</strong>al clusters (cf. discussion <strong>in</strong> 4.1), the roots with a sound symbolic<br />
<strong>in</strong>itial cluster n o t end<strong>in</strong>g <strong>in</strong> a f<strong>in</strong>al cluster were studied more closely.<br />
Table 5.22 shows the frequencies of semantic features when roots end<strong>in</strong>g<br />
with clusters or gem<strong>in</strong>ates are not taken <strong>in</strong>to account. The table shows the<br />
semantic features, the frequencies and root examples, and the clusters are<br />
ranked <strong>in</strong> frequency order. The words are extracted from Appendix 1.<br />
The roots of Table 5.22 are a subset of the roots <strong>in</strong> Appendix 1. Table<br />
5.22 thus only shows the roots which do not end with a f<strong>in</strong>al cluster or<br />
gem<strong>in</strong>ate, whereas Appendix 1 conta<strong>in</strong>s roots both with and without f<strong>in</strong>al<br />
clusters or gem<strong>in</strong>ates.<br />
11 It is possible that there is sound symbolism also <strong>in</strong> s<strong>in</strong>gle consonants, <strong>in</strong>itially as well<br />
as f<strong>in</strong>ally. For example, [S] could have the semantic feature 'separation' as <strong>in</strong> skiva,<br />
skilja, sk<strong>in</strong>na. Also, s<strong>in</strong>gular sounds with special characteristics could attract sound<br />
symbolic mean<strong>in</strong>gs, e.g. [r] is a sound which is different from other sounds and [C] has<br />
low frequency <strong>in</strong> Swedish. This question is not <strong>in</strong>vestigated <strong>in</strong> this thesis, but is left for<br />
further research.<br />
175
Table 5.22 Comb<strong>in</strong>ations of <strong>in</strong>itial cluster + no f<strong>in</strong>al cluster. The<br />
recurr<strong>in</strong>g semantic features of <strong>in</strong>itial clusters <strong>in</strong> frequency order. Listed<br />
to the right are the roots.<br />
features freq. the words<br />
pejorative 45 bliga, drasut, dröse, fjäsa,<br />
fl<strong>in</strong>a, glop, glupa, glåpord,<br />
gnat, gneta, gnidare, gräma,<br />
gräslig, klåpa, knal, knasig,<br />
kneg, kneken, knöl, knös,<br />
krake, kreta, kruserlig,<br />
kräk, mjäkig, pjåkig, pryl,<br />
skral, slas, slasa, slok, slyna,<br />
slö, slösa, sniken, snyta,<br />
snål, snöd, snöpa, spol<strong>in</strong>g,<br />
strul, tradig, trasa, vräkig,<br />
vrövel<br />
long th<strong>in</strong> form 36 skåra, slana, släde, snabel,<br />
sno, snok, snöre, spene,<br />
speta, spik, spila, spira,<br />
spole, spö, spjut, spjäla<br />
spröt, stake, stav, stig,<br />
stilett, stör, streamer,<br />
strigel, stril, stripa, strut,<br />
strå, stråk, stråle, stråt,<br />
truta, tryne, tråd, tråg, trål<br />
talk<strong>in</strong>g 23 gnat, gnola, gny, gnöla,<br />
gråta, gräla, klaga, knota,<br />
kverulans, prat, pruta,<br />
skri, skrodera, skryta,<br />
skrål, skräna, skröna,<br />
smäda, snäsa, svada, svära,<br />
vrål, vrövel<br />
sound 21 braka, brus, bräka, bröl,<br />
fnysa, frasa, fräsa, gnod<strong>in</strong>g,<br />
knaka, knot, krakel, kras,<br />
kråka, kvida, kväda, kväka,<br />
skval, snusa, snyta, stön,<br />
stril<br />
176
quick or strong<br />
movement<br />
18 flanera, fluga, flyga, flyta,<br />
flåsa, flöda, frusa, gnida,<br />
gno, skaka, slira, slita,<br />
spruta, stöt, svepa, sväva,<br />
vrida, vräka<br />
wetness 15 blod, blöt, klena, kleta,<br />
kram, kräm, skval, smet,<br />
snor, snuva, spad, spola,<br />
sputum, spruta, stril<br />
th<strong>in</strong> form 13 flaga, flak, flarn, flisa, fnasig,<br />
krakmandel, krokett,<br />
krusta, krustad, spade,<br />
smooth surface<br />
structure<br />
spat, spatel, spån<br />
11 glací, glaciär, glas, glatt,<br />
glida, glissando, gnida, gno,<br />
slipa, slät, spegel<br />
destruction 11 bryta, bråte, bräck,<br />
bränn<strong>in</strong>g, frat, fräta, kvarn,<br />
skrot, skräp, trasa, vrak<br />
light 10 blek, bläs, glacé, glisa,<br />
gloria, glåmig, glöd, prål,<br />
spat, spraka<br />
gaze 10 bliga, glana, glaukom, glo,<br />
glosögd, glosa, plira, snegla,<br />
spana, speja<br />
dim<strong>in</strong>utive 8 gli, glipa, knåp, krabat,<br />
smula, små, späd, vret<br />
walk<strong>in</strong>g 8 knata, krypa, kräla, snava,<br />
strosa, ströva, trava, träda<br />
round form 7 blåsa, knop, knota, knödel,<br />
knöl, kruka, skopa<br />
w<strong>in</strong>d<strong>in</strong>g form 7 krok, krokan, krusa, kråma,<br />
kräla, krök, spiral<br />
bad mood 6 gräla, trulig, tråkig, träta,<br />
vrede, vresig<br />
hollow form 5 grav, grop, gryt, gräva,<br />
gröpa<br />
form 4 glipa, glob, plös, pryl<br />
bodily feel<strong>in</strong>g 4 kval, kvav, kväva, svida<br />
soft consistency 4 mjuk, plym, plymå, plysa<br />
f<strong>in</strong>e gra<strong>in</strong> 4 mjäla, mjöl, stöv, strö<br />
rough surface structure 4 fryna, krås, skråma, sträv<br />
separation 4 spagat, split, spreta, sprida<br />
fall<strong>in</strong>g 3 dråsa, drälla, drösa<br />
putt<strong>in</strong>g together 3 knipa, knut, knyta<br />
177
hardness 3 skal, skare, skava<br />
slackness 3 slak, slana, sloka<br />
quickness 3 snar, sno, snudig<br />
stiffness 3 styv, strak, stram<br />
slang 2 plit, snut<br />
beat 2 slag, slå<br />
mental feel<strong>in</strong>g 2 snopen, tråna<br />
jocular 2 spe, spjuver<br />
f<strong>in</strong>e gra<strong>in</strong> 2 stöv, strö<br />
In order to make a general comparison with the earlier results of chapter<br />
4 concern<strong>in</strong>g the semantic features found <strong>in</strong> the analysis of <strong>in</strong>itial<br />
consonant clusters, part of the results of diagram 4.5 are presented as a<br />
table below (table 5.24).<br />
The 11 most frequent semantic features for the comb<strong>in</strong>ation <strong>in</strong>itial cluster<br />
+ no f<strong>in</strong>al cluster (and no gem<strong>in</strong>ate consonant) are shown <strong>in</strong> table 5.23.<br />
This rank<strong>in</strong>g can be compared with diagram 4.5, which shows the<br />
frequencies of all the semantic features, both with and without f<strong>in</strong>al<br />
clusters.<br />
Table 5.23 The 11 most frequent<br />
semantic features for the comb<strong>in</strong>ation<br />
<strong>in</strong>itial cluster + no f<strong>in</strong>al cluster.<br />
178<br />
Table 5.24 The 11 most frequent<br />
semantic features of diagram 4.5<br />
show<strong>in</strong>g <strong>in</strong>itial clusters irrespective of<br />
f<strong>in</strong>al cluster.<br />
pejorative 45 pejorative 163<br />
long th<strong>in</strong> form 36 sound 107<br />
talk<strong>in</strong>g 23 long th<strong>in</strong> form 97<br />
sound 21 quick or strong<br />
movement<br />
67<br />
quick or strong<br />
movement<br />
18 wetness 63<br />
wetness 15 talk<strong>in</strong>g 55<br />
th<strong>in</strong> form 13 light 32<br />
smooth surface<br />
11 dim<strong>in</strong>utive 31<br />
structure<br />
destruction 11 round form 23<br />
light 10 way of walk<strong>in</strong>g 22<br />
gaze 10 destruction 21
Table 5.23 shows the 11 most frequent semantic features of table 5.22<br />
(<strong>in</strong>itial clusters + no f<strong>in</strong>al clusters or gem<strong>in</strong>ates), whereas table 5.24<br />
shows the 11 most frequent features of diagram 4.5 (or Appendix 1),<br />
which conta<strong>in</strong>s roots with <strong>in</strong>itial clusters either with or without f<strong>in</strong>al<br />
clusters or gem<strong>in</strong>ates.<br />
Naturally, there are fewer words <strong>in</strong> the table show<strong>in</strong>g <strong>in</strong>itial cluster + no<br />
f<strong>in</strong>al cluster. In both tables, i.e. for both <strong>in</strong>itial cluster + no f<strong>in</strong>al cluster<br />
and <strong>in</strong>itial cluster + f<strong>in</strong>al cluster, 'pejorative' is the most frequent feature.<br />
The features 'long th<strong>in</strong> form', 'sound', 'quick or strong movement',<br />
'wetness' and 'talk<strong>in</strong>g' are among the six most frequent features <strong>in</strong> both<br />
tables. In other words the same six features are among the six most<br />
frequent features <strong>in</strong> both cases, only <strong>in</strong> partially different order. This<br />
means that the semantic features are the same, and also approximately as<br />
frequent, irrespective of whether one analyzes roots with only <strong>in</strong>itial<br />
consonant clusters or roots with both <strong>in</strong>itial and f<strong>in</strong>al consonant clusters.<br />
This result strengthens one outcome of the analysis of chapter 4: the<br />
analysis of semantic features of <strong>in</strong>itial clusters actually showed the<br />
semantic features of the <strong>in</strong>itial consonant clusters and not of the f<strong>in</strong>al<br />
clusters.<br />
5.6.2 No <strong>in</strong>itial cluster + f<strong>in</strong>al cluster<br />
Table 5.25 shows the comb<strong>in</strong>ations of no <strong>in</strong>itial cluster + f<strong>in</strong>al cluster, <strong>in</strong><br />
frequency order.<br />
179
Table 5.25 Comb<strong>in</strong>ations of no <strong>in</strong>itial cluster + f<strong>in</strong>al cluster. The<br />
recurr<strong>in</strong>g semantic features of f<strong>in</strong>al clusters <strong>in</strong> frequency order. Listed to<br />
the right are the roots.<br />
semantic feature freq. rank the words<br />
quick or strong<br />
movement<br />
<strong>in</strong> 5.22<br />
51 5 rutscha, puffa, dangla,<br />
rangla, d<strong>in</strong>gla, r<strong>in</strong>gla,<br />
v<strong>in</strong>gla, famla, vimla,<br />
fumla, rumla, tumla,<br />
haspla, dimpa, gumpa,<br />
pumpa, rappa, tappa,<br />
kippa, tippa, vippa, hoppa,<br />
moppa, poppa, guppa,<br />
dallra, pillra, tillra,<br />
myllra, darra, irra,<br />
dirra, virra, burra,<br />
huttra, hafsa, sjafsa,<br />
nafsa, rafsa, tafsa, rufsa,<br />
tufsa, räfsa, mumsa,<br />
palta, bylta, hasta, kasta,<br />
rista, hysta, nysta<br />
talk<strong>in</strong>g 18 3 babbla, rabbla, gaffla,<br />
mumla, tassla, tissla,<br />
rossla, jollra, morra,<br />
hurra, murra, tjattra,<br />
muttra, jamsa, gasta,<br />
hosta, pusta, tjafsa<br />
sound 14 4 ratscha, p<strong>in</strong>gla, humla,<br />
sörpla, porla, sorla, rassla,<br />
vissla, bullra, mullra,<br />
kurra, surra, kuttra,<br />
puttra<br />
slang 12 - haffa, fiffa, piffa, roffa,<br />
buffa, luffa, ruffa, tuffa,<br />
töffa, dumpa, rumpa,<br />
sumpa<br />
pejorative 9 1 taffla, fiffla, ruffla, lolla,<br />
hafsa, tjafsa, lafsa, jamsa,<br />
dalta<br />
walk<strong>in</strong>g 7 13 lulla, jumpa, lafsa, lufsa,<br />
halta, rulta, tulta<br />
round form 3 14 rolla, bulla, rulla<br />
short-wide form 2 - fimpa, limpa<br />
light 2 10 t<strong>in</strong>dra, skimra<br />
long th<strong>in</strong> form 2 2 ramsa, remsa<br />
180
The rank<strong>in</strong>g of the most frequent words can be compared with the<br />
rank<strong>in</strong>g of table 5.22. This comparison shows that 'pejorative' is not as<br />
frequently coded <strong>in</strong> the f<strong>in</strong>al clusters as <strong>in</strong> the <strong>in</strong>itial clusters.<br />
On the other hand, 'slang' (ma<strong>in</strong>ly coded <strong>in</strong> -ffa), 'walk<strong>in</strong>g' and 'quick or<br />
strong movement' are more frequent <strong>in</strong> words with only f<strong>in</strong>al clusters,<br />
compared with words with only <strong>in</strong>itial clusters. This might be due to the<br />
verbal nature of the clusters end<strong>in</strong>g <strong>in</strong> a vowel. In the same way as table<br />
5.23 is compared with table 5.24, table 5.25 can be compared with table<br />
5.26, which conta<strong>in</strong>s all the roots (and not just the ones without <strong>in</strong>itial<br />
clusters) taken from table 5.2.<br />
Table 5.26 All roots from table 5.2, ordered by semantic features, <strong>in</strong><br />
frequency order.<br />
semantic feature freq. the words<br />
quick or strong<br />
movement<br />
85 rutscha, knuffa, puffa, dangla, rangla, skrangla,<br />
d<strong>in</strong>gla, r<strong>in</strong>gla, v<strong>in</strong>gla, famla, vimla, tromla,<br />
fumla, rumla, drumla, tumla, stöppla, haspla,<br />
rumla, dimpa, skvimpa, gumpa, skumpa, pumpa,<br />
klappa, snappa, rappa, tappa, greppa, steppa,<br />
kippa, skippa, tippa, vippa, hoppa, moppa,<br />
noppa, snoppa, poppa, droppa, proppa, guppa,<br />
knäppa, snäppa, fladdra, bläddra, dallra, pillra,<br />
tillra, myllra, flimra, darra, irra, dirra, virra,<br />
svirra, burra, plurra, snurra, glittra, splittra,<br />
klottra, huttra, hafsa, sjafsa, nafsa, snafsa, rafsa,<br />
krafsa, tafsa, plufsa, rufsa, tufsa, räfsa, plumsa,<br />
mumsa, palta, skralta, skvalta, bylta, hasta, kasta,<br />
rista, hysta, nysta<br />
talk<strong>in</strong>g 37 babbla, rabbla, gaffla, traggla, svamla, mumla,<br />
tassla, tissla, rossla, slabbra, bladdra, pladdra,<br />
sladdra, bluddra, sluddra, jollra, pjollra, blarra,<br />
snarra, kvirra, morra, knorra, hurra, murra,<br />
tjattra, fnittra, muttra, tjafsa, jamsa, grumsa,<br />
tramsa, gasta, hosta, pusta, frusta, knysta, krysta<br />
sound 34 klatscha, ratscha, p<strong>in</strong>gla, skramla, humla, sörpla,<br />
porla, sorla, rassla, prassla, gnissla, vissla,<br />
kvillra, skallra, bullra, mullra, bjällra, knarra,<br />
klirra, knirra, skorra, kurra, surra, smattra,<br />
knattra, snattra, skvattra, knittra, kvittra, kuttra,<br />
puttra, gläfsa, klafsa, slafsa<br />
181
slang 19 haffa, blaffa, klaffa, fiffa, sniffa, piffa, skoffa,<br />
roffa, buffa, skuffa, luffa, fluffa, ruffa, gruffa,<br />
tuffa, töffa, dumpa, rumpa, sumpa<br />
pejorative 19 taffla, fiffla, ruffla, fjolla, lolla, stolla, slampa,<br />
fjompa, klattra, plottra, lafsa, glufsa, flamsa,<br />
dalta, slafsa, hafsa, tjafsa, tramsa, jamsa<br />
walk<strong>in</strong>g 15 lulla, drulla, klampa, trampa, stampa, jumpa,<br />
trippa, lufsa, halta, rulta, tulta, stulta, stylta,<br />
klafsa, lafsa<br />
wetness 9 blaska, plaska, slaska, smaska, snaska, vaska,<br />
glopp, klafsa, skvalta<br />
round form 7 rolla, skrolla, bulla, krulla, stropp, rulla, knollra<br />
long th<strong>in</strong> form 5 strimla, slamsa, ramsa, remsa, slimsa<br />
light 5 gl<strong>in</strong>dra, t<strong>in</strong>dra, skimra, glimra, flimra<br />
short-wide form 4 fimpa, limpa, klimpa, klumpa<br />
dim<strong>in</strong>utive 2 stumpa, loppa<br />
fall<strong>in</strong>g 1 ramla<br />
bodily feel<strong>in</strong>g 1 pirra<br />
gaze 1 stirra<br />
rough surface structure 1 knottra<br />
beat 1 bulta<br />
destruction 1 brista<br />
slackness 1 slimsa<br />
movement 1 knaggla<br />
w<strong>in</strong>d<strong>in</strong>g form 1 kr<strong>in</strong>gla<br />
form 1 plumpa<br />
The analysis shows that the six most frequent features are the same and<br />
that they are so <strong>in</strong> exactly the same order. This result strengthens the<br />
analysis of f<strong>in</strong>al clusters <strong>in</strong> table 5.25.<br />
5.6.3 Initial cluster + f<strong>in</strong>al cluster<br />
The tables 5.27 and 5.28 show comb<strong>in</strong>ations of f<strong>in</strong>al clusters and <strong>in</strong>itial<br />
clusters. The examples are extracted from table 5.2. Table 5.27 shows<br />
comb<strong>in</strong>ations where the semantic features are the same <strong>in</strong>itially and<br />
f<strong>in</strong>ally, and table 5.28 shows comb<strong>in</strong>ations where the features are<br />
different <strong>in</strong>itially and f<strong>in</strong>ally. The follow<strong>in</strong>g conclusions can be drawn:<br />
1. Fifty words have the same semantic feature <strong>in</strong>itially and f<strong>in</strong>ally,<br />
whereas 34 words have different features. However, of these 34 words, 8<br />
have <strong>in</strong>itial and f<strong>in</strong>al features that are similar to each other: 'sound' +<br />
'talk<strong>in</strong>g' and 'w<strong>in</strong>d<strong>in</strong>g form' + 'round form'. It is more common, then,<br />
182
for semantic features to be the same, <strong>in</strong>itially and f<strong>in</strong>ally, than to be<br />
different.<br />
2. The semantic features of a root can be the same <strong>in</strong>itially and f<strong>in</strong>ally for<br />
each root with double features 12. These are the follow<strong>in</strong>g, listed <strong>in</strong> order<br />
of frequency. The mean<strong>in</strong>gs of <strong>in</strong>itial clusters come from the analysis <strong>in</strong><br />
Appendix 1, and the mean<strong>in</strong>gs of f<strong>in</strong>al clusters come from the analysis <strong>in</strong><br />
table 5.2.<br />
Table 5.27 Semantic features (with absolute frequencies of the roots) that<br />
are the same <strong>in</strong>itially and f<strong>in</strong>ally <strong>in</strong> one and the same root.<br />
doubl<strong>in</strong>g<br />
freq. the words represent<strong>in</strong>g<br />
features<br />
root morphemes<br />
sound 18 klatscha, skramla,<br />
prassla, gnissla, skallra,<br />
kvillra, bjällra, knarra,<br />
klirra, knirra, skorra,<br />
smattra, knattra, snattra,<br />
knittra, kvittra, klafsa, gläfsa<br />
talk<strong>in</strong>g 9 svamla, slabbra, bladdra,<br />
sladdra, bluddra, sluddra,<br />
blarra, tramsa, grumsa<br />
pejorative 9 fjolla, stolla, slampa, fjompa,<br />
plottra, slafsa, glufsa, flamsa,<br />
tramsa<br />
wetness 4 blaska, plaska, slaska, snaska<br />
quick or strong 3 tromla, skippa, fladdra<br />
movement<br />
light 3 gl<strong>in</strong>dra, flimra, glimra<br />
walk<strong>in</strong>g 2 trampa, stulta<br />
short-wide form 2 klimpa, klumpa<br />
long th<strong>in</strong> form 1 slamsa<br />
Table 5.28 shows which different semantic features are comb<strong>in</strong>ed <strong>in</strong>itially<br />
and f<strong>in</strong>ally. A word represent<strong>in</strong>g a root morpheme may occur <strong>in</strong> two<br />
places if it conta<strong>in</strong>s two features <strong>in</strong>itially or two features f<strong>in</strong>ally, e.g.<br />
plumsa: pl- means both 'sound' and 'wetness' <strong>in</strong> this root morpheme.<br />
12 There are 50 words but 51 features, as can be seen <strong>in</strong> the table, because one word<br />
(tramsa) conta<strong>in</strong>s two features <strong>in</strong>itially and two features f<strong>in</strong>ally.<br />
183
Table 5.28 Semantic features that are different <strong>in</strong>itially and f<strong>in</strong>ally <strong>in</strong> one and<br />
the same root.<br />
<strong>in</strong>itial:<br />
f<strong>in</strong>al:<br />
semantic<br />
feature of<br />
f<strong>in</strong>al<br />
cluster:<br />
slang quick or<br />
strong<br />
movement<br />
semantic<br />
feature of<br />
<strong>in</strong>itial<br />
cluster:<br />
sound klappa<br />
steppa<br />
knäppa<br />
snäppa<br />
krafsa<br />
plumsa<br />
pejorative blaffa drumla<br />
klottra<br />
plufsa<br />
beat knuffa<br />
bad mood gruffa<br />
talk<strong>in</strong>g gruffa<br />
destruction skrangla<br />
splittra<br />
skralta<br />
184<br />
wetness walk<strong>in</strong>g round<br />
form<br />
smaska klampa<br />
stampa<br />
talk<strong>in</strong>g<br />
snarra<br />
kvirra<br />
knorra<br />
fnittra<br />
frusta<br />
knysta<br />
krysta<br />
drulla pjollra<br />
w<strong>in</strong>d<strong>in</strong>g<br />
form<br />
snurra krulla<br />
iterative stampa<br />
quickness snappa<br />
fall<strong>in</strong>g droppa<br />
wetness droppa<br />
plurra<br />
plumsa<br />
skvalta<br />
light glittra<br />
dim<strong>in</strong>utive splittra<br />
The follow<strong>in</strong>g conclusions can be drawn from table 5.28.<br />
1. The semantic feature which is most often comb<strong>in</strong>ed with other features<br />
is f<strong>in</strong>al 'quick or strong movement' – but this could partly be an effect of<br />
the choice to study the semif<strong>in</strong>al "a- end<strong>in</strong>gs".<br />
2. The most common feature comb<strong>in</strong>ations are:
'sound' + 'talk<strong>in</strong>g' (7 words)<br />
'sound' + 'quick or strong movement' (6 words)<br />
'wetness' + 'quick or strong movement' (4 words)<br />
In the comb<strong>in</strong>ation 'sound' + 'talk<strong>in</strong>g' the two features are closely related<br />
and they strengthen each other.<br />
3. '<strong>Sound</strong>' is the most common <strong>in</strong>itial feature <strong>in</strong> these comb<strong>in</strong>ations.<br />
4. The otherwise common semantic features 'sound' and 'pejorative' do not<br />
have f<strong>in</strong>al position <strong>in</strong> these comb<strong>in</strong>ations. 'Talk<strong>in</strong>g' and 'walk<strong>in</strong>g' are much<br />
more common <strong>in</strong> f<strong>in</strong>al position than <strong>in</strong> <strong>in</strong>itial.<br />
5.6.4 Summary of comb<strong>in</strong>ations<br />
The general results of the study of comb<strong>in</strong>ations of <strong>in</strong>itial and f<strong>in</strong>al<br />
clusters are the follow<strong>in</strong>g: Sometimes the <strong>in</strong>itial cluster carries the sound<br />
symbolic mean<strong>in</strong>g because the word ends <strong>in</strong> a s<strong>in</strong>gle consonant (Group 1).<br />
The most common semantic feature of Group 1 is 'pejorative', which is<br />
also the case for all the sound symbolic roots <strong>in</strong> Appendix 1. The same six<br />
features are among the six most frequent ones <strong>in</strong> both groups and this<br />
<strong>in</strong>dicates that the analysis of chapter 4 showed what it was <strong>in</strong>tended to<br />
show: the mean<strong>in</strong>g of <strong>in</strong>itial clusters.<br />
When there are both <strong>in</strong>itial and f<strong>in</strong>al consonant clusters <strong>in</strong> a root, these<br />
can comb<strong>in</strong>e by hav<strong>in</strong>g the same mean<strong>in</strong>g (Group 3a), e.g. bl- and -aska<br />
('wetness') – cf. table 5.27. Alternatively, different mean<strong>in</strong>gs can comb<strong>in</strong>e<br />
<strong>in</strong> a word (Group 3b), e.g. glittra – cf. table 5.28. The most common<br />
comb<strong>in</strong>ations are those <strong>in</strong> which the mean<strong>in</strong>g is the same <strong>in</strong>itially and<br />
f<strong>in</strong>ally. The most common mean<strong>in</strong>g for group 3a is 'sound' + 'sound',<br />
followed by 'talk<strong>in</strong>g' + 'talk<strong>in</strong>g' and 'pejorative' + 'pejorative'. The most<br />
common mean<strong>in</strong>g comb<strong>in</strong>ations for Group 3b is 'sound' + 'talk<strong>in</strong>g',<br />
followed by 'sound' + 'quick or strong movement'.<br />
F<strong>in</strong>ally, sometimes only the f<strong>in</strong>al cluster (or gem<strong>in</strong>ate) carries the sound<br />
symbolic mean<strong>in</strong>g, because the word beg<strong>in</strong>s with a s<strong>in</strong>gle consonant<br />
(Group 2). The most common mean<strong>in</strong>g is 'quick or strong movement'<br />
followed by 'talk<strong>in</strong>g' and 'sound'.<br />
185
5.7 Summary and discussion of <strong>in</strong>itial and f<strong>in</strong>al<br />
clusters, and vowels<br />
This and the preced<strong>in</strong>g chapter have shown that certa<strong>in</strong> semantic features<br />
are connected with certa<strong>in</strong> <strong>in</strong>itial and f<strong>in</strong>al consonant clusters and with<br />
vowels – <strong>in</strong> different constellations and to different extents. A study of<br />
comb<strong>in</strong>ations of <strong>in</strong>itial and f<strong>in</strong>al clusters shows that the sound symbolic<br />
mean<strong>in</strong>g of a root morpheme can also be dependent on the comb<strong>in</strong>ations<br />
of phonemes of the whole word, e.g. sl- 'wetness' + a + -sk(a) 'wetness'<br />
(=slaska (splatter)). (However, both 'slisk' and 'slusk' have the feature<br />
'pejorative' and not 'wetness' because of the semantic profile of sl-, cf.<br />
4.4.1).<br />
Partly the same semantic features are used <strong>in</strong> both <strong>in</strong>itial and f<strong>in</strong>al<br />
clusters, while the vowels sometimes add a dimension of 'size' (smalllittle,<br />
high pitch-low pitch). The 'light/gaze' words, where the vowel is<br />
almost always a front vowel, could be <strong>in</strong>cluded <strong>in</strong> the 'size' group, s<strong>in</strong>ce<br />
'light' can be seen as synaesthetically connected with high pitch. The<br />
vowels, when they are used sound symbolically, thus often seem to have<br />
other semantic features than the consonant clusters do, and they add a<br />
special mean<strong>in</strong>g to roots – which otherwise get their sound symbolic<br />
mean<strong>in</strong>g from <strong>in</strong>itial and/or f<strong>in</strong>al clusters – e.g. by show<strong>in</strong>g if a sound is<br />
high pitched or low pitched, or if a movement is quick or slow.<br />
For typically sound-symbolic (where all or almost all root morphemes are<br />
sound symbolic) or high frequent <strong>in</strong>itial clusters, vowels, and f<strong>in</strong>al<br />
clusters, the mean<strong>in</strong>g of a word, especially of a neologism, should be<br />
predictable.<br />
A nonsense expression like pjaffla : pj- ('pejorative') + a + -fla<br />
('pejorative') ought to be <strong>in</strong>terpreted as a pejorative word. A neologism<br />
like brullra : br- ('sound') + P ('low pitch') + -lra ('sound' or 'quick or<br />
strong movement') ought to be <strong>in</strong>terpreted as imitat<strong>in</strong>g sound (a low<br />
pitched one) while a nonsense expression like flillra: fl- ('quick or strong<br />
movement') + i ('smallness', 'quickness' or 'high pitch') + -lra ('sound'<br />
186
or 'quick or strong movement') ought to be <strong>in</strong>terpreted as someth<strong>in</strong>g<br />
hav<strong>in</strong>g to do with quick movement 13.<br />
Look<strong>in</strong>g at real words, there can be a comb<strong>in</strong>ation effect like <strong>in</strong> skrangla<br />
(skranglig-rickety) "which has an unsteady construction, or has become<br />
unsteady because of long usage". This def<strong>in</strong>ition goes well with the<br />
phonesthemes skr- 'destruction' and -ngla 'quick or strong movement'.<br />
Another example is klatscha (crack): kl- 'sound' and -tscha [tSa] 'sound'.<br />
A more unconventional word klitscha (which is not <strong>in</strong> the dictionary)<br />
would stand for a more high pitched sound 14. For real words there can<br />
always be an effect of conventionalization which makes them less<br />
predictable.<br />
Conventionalization can always take over, at least partly, either <strong>in</strong> the<br />
arbitrary direction or <strong>in</strong> choos<strong>in</strong>g phonesthemic mean<strong>in</strong>gs. The<br />
phonestheme level is an underly<strong>in</strong>g level which can be activated by<br />
different types of context.<br />
The next chapter will present contrastive studies of some semantic fields<br />
of sound symbolism. In chapter 7 f<strong>in</strong>al clusters, as well as <strong>in</strong>itial clusters,<br />
will be discussed aga<strong>in</strong> <strong>in</strong> connection with neologisms.<br />
13 In the experiments with nonsense words <strong>in</strong> chapter 6 only the effect of the <strong>in</strong>itial<br />
clusters were tested. In further research, however, the <strong>in</strong>tention is to also test the<br />
comb<strong>in</strong>ed effects.<br />
14Two roommates were asked about the mean<strong>in</strong>g of this expression. They answered that<br />
it was not a real word, and then one of them gave a possible mean<strong>in</strong>g 'the sound of a<br />
bicycle spoke that breaks'. The other person said it meant 'the sound that comes when<br />
you drive on a th<strong>in</strong> layer of ice'. Both these def<strong>in</strong>itions have to do with a high pitched<br />
sound.<br />
187
6 Some contrastive studies <strong>in</strong> sound symbolism<br />
6.1 Introduction<br />
The phenomena of onomatopoeia and other sound symbolism have been<br />
described <strong>in</strong> part for different languages. Some earlier contrastive<br />
f<strong>in</strong>d<strong>in</strong>gs are described <strong>in</strong> the overview of 1.10.4.<br />
The aim of this chapter is to study some of the semantic features and<br />
consonant clusters discussed above <strong>in</strong> order to analyze what differences<br />
and similarities there are between some languages.<br />
In the first section I will present a contrastive lexical study made on some<br />
selected mean<strong>in</strong>gs <strong>in</strong> a Thesaurus. The languages compared are Swedish<br />
and English. I will then present a contrastive study of some expressive<br />
<strong>in</strong>terjections, commands and greet<strong>in</strong>gs and a study of words for animal<br />
sounds. Furthermore I will present a contrastive experiment of<br />
<strong>in</strong>terpretation of Swedish sound symbolic words.<br />
6.2 The Thesaurus study<br />
6.2.1 Method<br />
In this experiment <strong>in</strong>formants chose the best words for 'stupidity',<br />
'roughness' and 'smoothness' <strong>in</strong> English and Swedish out of a Thesaurus.<br />
From the Swedish Thesaurus (Br<strong>in</strong>g, 1930) a couple of mean<strong>in</strong>gs were<br />
chosen, namely: 'dumhuvud' (501), 'oförstånd' (499), and 'glatthet' (255),<br />
'skrovlighet' (256). In the English Roget´s Thesaurus (1977) the<br />
follow<strong>in</strong>g mean<strong>in</strong>gs were chosen: 'fool' (501), 'imbecility' (499),<br />
'<strong>in</strong>sanity' (503), 'smoothness' (255), and 'roughness' (256)<br />
Lists of all words under these head<strong>in</strong>gs were then <strong>in</strong>spected by native<br />
<strong>in</strong>formants, three for the Swedish material and three for the English<br />
material. The <strong>in</strong>formants were given the <strong>in</strong>struction "to mark words, the<br />
expressive forms of which are felt to be especially adequate for their<br />
content". 1<br />
1 S<strong>in</strong>ce I am not an active speaker of English and s<strong>in</strong>ce I have deepened my <strong>in</strong>tuitions of<br />
Swedish phonesthemes from the analysis of chapter 4, I used subjects for this task.<br />
189
6.2.2 Results<br />
6.2.2.1 Words for 'stupidity' <strong>in</strong> English<br />
S<strong>in</strong>ce English does not have the clusters typically pejorative for Swedish:<br />
fj-, fn-, pj-, there could be no similarities between Swedish and English<br />
here. (Cf. below and diagram 4.6.)<br />
There were 4 words where all three English subjects agreed:<br />
n<strong>in</strong>compoop, blockhead, dunce and dull.<br />
Two of the three subjects agreed on 9 words: simpleton, dolt, booby, oaf,<br />
clod, silly, muddleheaded, addleheaded, fool.<br />
Isolated contributions from either of the subjects were: tomfool, ass,<br />
noodle, gawk, mooncalf, dotard, driveler, old fogy, Simple Simon, goose,<br />
lout, idiot, dotage, idiocy, fatuity, gidd<strong>in</strong>ess, drivel, dote, stultify, bov<strong>in</strong>e,<br />
feeblem<strong>in</strong>ded, obtuse, stolid, fatuous, drivel<strong>in</strong>g, bewildered, maudl<strong>in</strong>,<br />
stupidity, foolishness, rashness, bra<strong>in</strong>less, childlike, vacant.<br />
Many of these words are not what we would normally th<strong>in</strong>k of as sound<br />
symbolic. Nevertheless there seems to be a slight tendency for the<br />
preferred ones to conta<strong>in</strong> a long [u] or an [a] and perhaps [d] <strong>in</strong>itially or<br />
medially. The value of the <strong>in</strong>itial consonant clusters is difficult to judge<br />
from this material.<br />
6.2.2.2 Words for 'stupidity' <strong>in</strong> Swedish<br />
All three Swedish subjects agreed on the follow<strong>in</strong>g words: pjosker 2 ,<br />
pjasker, fjoller, fnasker, fjanter, flep, flepig, pjåk, fjolleri, fjant, pjoller,<br />
pjosk, fjollig, fjoskig, fjantig, pjoskig, pjollrig, pjåkig, pjaskig.<br />
The frequencies of the <strong>in</strong>itial consonant clusters are the follow<strong>in</strong>g:<br />
pj- : 9 words<br />
fj- : 7 words<br />
fl- : 2 words<br />
fn-: 1 word<br />
2 The Swedish words listed under the different categories of the Thesaurus are not<br />
translated s<strong>in</strong>ce the Thesaurus head<strong>in</strong>g, e.g. "stupidity" shows the semantic category.<br />
What is <strong>in</strong>terest<strong>in</strong>g is not the exact mean<strong>in</strong>g of all these words, but which sounds are the<br />
most frequent for each Thesaurus head<strong>in</strong>g.<br />
190
Except for fl- the result above mirrors the results of the percentally most<br />
frequent clusters well (diagram 4.6).<br />
Two of the subjects agreed on the follow<strong>in</strong>g words:<br />
mähä, tafser, tossa, tafsig, fjoskighet, fjanta, fjollas, pjollra, pjoska, pjåka,<br />
pjask.<br />
The most common clusters here are:<br />
pj-: 4 words<br />
fj-: 3 words<br />
There is a clear preference for the fj- and pj- clusters among the Swedish<br />
<strong>in</strong>formants.<br />
Isolated contributions were: våp, tåp, jöns, fån, sjasker, drummel, tos<strong>in</strong>g,<br />
stolle, drönig, tåpig, -snut, pund-, mes, schajas, flack, flärdfull, larvig,<br />
korkad, tafs, sjask, jolt, fåna, jolta, drumlig, slapp, slö, våpig, fånig,<br />
sjaskig, taskig, tölpig, show<strong>in</strong>g an additional preference for the vowels<br />
[o:], [O] and [a]. The consonant cluster which is the most common is dr-<br />
(3 roots). The cluster dr- is pejorative accord<strong>in</strong>g to the analysis <strong>in</strong><br />
chapter 4.<br />
6.2.2.3 Words for 'surface structure' <strong>in</strong> English<br />
For the categories of 'roughness' and 'smoothness' there where two<br />
English subjects who made an assessment.<br />
'Roughness'<br />
Both subjects agreed on the follow<strong>in</strong>g words: crest, ruffle, crumple,<br />
rumple, rugged, jagged, gnarly, scraggly, scraggy, craggy, cragged,<br />
prickly, bristly, bushy.<br />
The clusters cr- and scr- and the consonant r are favored.<br />
Isolated contributions were: asperity, corrugation, shag, cross-gra<strong>in</strong>ed,<br />
hirsute, shaggy, nappy, thatch, whiskers, feather, rough, cr<strong>in</strong>kle. Among<br />
these words there are two <strong>in</strong>stances of cr- and five <strong>in</strong>stances of s<strong>in</strong>gle r.<br />
'Smoothness'<br />
Both subjects agreed on the words sleek and glossy.<br />
191
Isolated contributions were: plane, level, polish, velvety, glassy, gloss,<br />
roller, roll, oily, silken, silky.<br />
The preferred consonant is l.<br />
6.2.2.4 Words for 'surface structure' <strong>in</strong> Swedish<br />
'Roughness'<br />
All three subjects agreed on the follow<strong>in</strong>g words: knotter and knottrig.<br />
Two out of three agreed on the follow<strong>in</strong>g: knotighet, knut, knyla, knotig,<br />
knagglig, klippig, knollrig, krullig, krusig, stripig, skrapa, skrynka,<br />
skrynkla, knottra, skrovlig, skrynkig, skrynklig 3<br />
These words show a preference for r and for the clusters kn- and skrand<br />
kr-. The kr- and skr- words both have the semantic feature 'rough<br />
surface structure' (<strong>in</strong> chapter 4) but kn- words are analyzed as 'round<br />
form'. The cluster with the highest number of root morphemes with the<br />
feature 'rough surface' structure (<strong>in</strong> the lexical analysis) is skr-.<br />
Isolated contributions were: raggighet, knöl, kornighet, rufsighet, burr,<br />
stripa, ludenhet, lurv, tovig, purrighet, ludd, borst, ragg, lugg, rugg,<br />
stubb, test, ull, kvist, ris, tistel, tova, tuva, kvast, visp, buckla, kugge,<br />
krusn<strong>in</strong>g, rysch, frans, dun, plysch, schagg, sträv, fårad, uddig, tandad,<br />
taggig, törnig, risig, buskig, tuvig, lummig, sträv, tofsig, noppig.<br />
These contributions, from s<strong>in</strong>gular subjects, show a preference for the<br />
vowel [P].<br />
'Smoothness'<br />
All three subjects agreed on glätta, glatta, glanska and glansa show<strong>in</strong>g<br />
preference for gl-.<br />
Two out of three agreed on: glättn<strong>in</strong>g, glans, slipprighet, slirn<strong>in</strong>g,<br />
glidn<strong>in</strong>g, blanka, glasera, glida, slira, glatt, glansig, glansk and slipprig.<br />
3 All words excerpted by the subjects were counted, and as a consequence the list also<br />
conta<strong>in</strong>s words that represent the same root.<br />
192
The recurr<strong>in</strong>g clusters are gl-, sl-, bl-, which all conta<strong>in</strong> the liquid l.<br />
These are, except for sl-, related to 'light', but as was shown <strong>in</strong> 4.6 there<br />
is an <strong>in</strong>dexical relation between 'light', 'reflect<strong>in</strong>g surface' and 'movement<br />
on reflect<strong>in</strong>g surface'. Slipprig and slira were analyzed as 'quick or<br />
strong' movement <strong>in</strong> the lexical analysis, (while slipa, slät, slätt were<br />
analyzed as 'smooth surface structure'). Sl- is also to a high degree<br />
connected with 'wetness'.<br />
Isolated contributions were: glänsa, glänsande, blank, glas, smörjn<strong>in</strong>g,<br />
smörja, smärgel, kristall, siden, silke, sammet, lackera, polera, slipa,<br />
kana, stryka, mangla, valsa, oljig, smidig, mjäll.<br />
Also <strong>in</strong> these words the l is very frequent, however not part of an <strong>in</strong>itial<br />
cluster.<br />
6.2.3 Conclusions of the Thesaurus study<br />
There was greater agreement between Swedish and English among the<br />
preferred sounds for 'surface structure', both 'roughness' (r, skr) and<br />
'smoothness' (sl, gl) than for 'pejorative'. An explanation to this might<br />
be that 'surface structure' is closer to a potentially common phenomenon,<br />
namely sound imitation, s<strong>in</strong>ce strok<strong>in</strong>g different surfaces gives different<br />
sound effects.<br />
Sometimes the isolated contributions perhaps mirror the fact that the task<br />
at hand for the test subjects can be difficult to keep <strong>in</strong> m<strong>in</strong>d or that the<br />
language feel<strong>in</strong>g sometimes runs amock among words with similar<br />
mean<strong>in</strong>gs. Suddenly all words can feel motivated. Nevertheless, this<br />
method is a possible way of compar<strong>in</strong>g different languages, through the<br />
<strong>in</strong>tuitions of <strong>in</strong>formants. A problem with us<strong>in</strong>g the Thesaurus is that the<br />
word corpus listed is very heterogeneous. Examples of this are several<br />
words (however not chosen by the Swedish subjects) under the category<br />
'roughness': ojämnhet (unevenness), knävelborr (big and bushy<br />
moustache), brottyta (fracture), kartnagel (deformed nail), vårta (wart)<br />
and tuppkam (cock's crest).<br />
For cross l<strong>in</strong>guistic comparisons the Thesaurus method used above is an<br />
alternative to free production with<strong>in</strong> different semantic fields, (for those<br />
languages that have a Thesaurus). Naturally, native <strong>in</strong>formants have to be<br />
used.<br />
193
6.3 Some <strong>in</strong>terjections <strong>in</strong> different languages<br />
Expressive <strong>in</strong>terjections is one of the sound symbolic types described <strong>in</strong><br />
2.4. The relation between expression and mean<strong>in</strong>g is ma<strong>in</strong>ly <strong>in</strong>dexical; the<br />
expression is caused by a bodily or mental reaction. It is an <strong>in</strong>terest<strong>in</strong>g<br />
task to analyze how similar the expressive <strong>in</strong>terjections are <strong>in</strong> different<br />
languages (the question of universal traits of the expression). It is also<br />
<strong>in</strong>terest<strong>in</strong>g to see, for the expressive <strong>in</strong>terjections of Swedish, to what<br />
extent the expressive <strong>in</strong>terjections are unsystematical, i. e. not relatable to<br />
phonesthemes.<br />
Expressive <strong>in</strong>terjections, commands and greet<strong>in</strong>gs (cf. Ideforss, 1928,<br />
categorization) are exemplified mostly with a fragment of the examples <strong>in</strong><br />
Ideforss. These were translated <strong>in</strong>to 8 different languages with the help of<br />
<strong>in</strong>formants. The languages are Icelandic, English, Polish, Hungarian,<br />
F<strong>in</strong>nish, Ososo, Malagasi and Slovenian. The results are presented <strong>in</strong><br />
Appendix 2 and grouped accord<strong>in</strong>g to type of <strong>in</strong>terjection. The expressive<br />
<strong>in</strong>terjections, commands, and greet<strong>in</strong>gs are listed <strong>in</strong> table 6.2 and <strong>in</strong><br />
Appendix 2. They are written <strong>in</strong> normal spell<strong>in</strong>g <strong>in</strong> table 6.2.<br />
In some cases it was difficult to ask subjects to translate word for word.<br />
The semantic contents of the <strong>in</strong>tended <strong>in</strong>terjection <strong>in</strong>stead had to be<br />
described (with<strong>in</strong> the semantic categories of table 6.2), after which the<br />
<strong>in</strong>formant gave the closest correspondence <strong>in</strong> his own language. The way<br />
the semantic contents was described to the <strong>in</strong>formant depended on his or<br />
her competence <strong>in</strong> Swedish or English, e.g. "Give me a word or<br />
expression for when you want to be depreciatory" or "Give me a word or<br />
expression for when you th<strong>in</strong>k somebody or someth<strong>in</strong>g is bad or ugly,<br />
etc".<br />
The semantic features used <strong>in</strong> earlier chapters, which correspond to the<br />
features of these <strong>in</strong>terjections, are shown <strong>in</strong> table 6.1.<br />
194
Table 6.1 The semantic features of <strong>in</strong>terjections and their<br />
correspondences <strong>in</strong> earlier chapters.<br />
Features of<br />
Correspond<strong>in</strong>g features<br />
<strong>in</strong>terjections<br />
of earlier chapters<br />
pejorative pejorative<br />
positive attitude<br />
surprise mental feel<strong>in</strong>g<br />
song talk<strong>in</strong>g<br />
other bodily or mental bodily feel<strong>in</strong>g or mental<br />
feel<strong>in</strong>g<br />
feel<strong>in</strong>g<br />
commands talk<strong>in</strong>g<br />
greet<strong>in</strong>gs talk<strong>in</strong>g<br />
6.3.1 Swedish expressive <strong>in</strong>terjections<br />
In table 6.2, a sample of the Swedish expressive <strong>in</strong>terjections is presented.<br />
Table 6.2 Examples (<strong>in</strong> ord<strong>in</strong>ary spell<strong>in</strong>g, mostly from Ideforss, 1928) of<br />
Swedish expressive <strong>in</strong>terjections, commands and greet<strong>in</strong>gs.<br />
EXPRESSIVE<br />
hihi(hi)<br />
mums<br />
pejorative<br />
mm<br />
namnam<br />
bu<br />
iih<br />
usch<br />
grr<br />
hu<br />
surprise<br />
COMMANDS<br />
"t" [|]<br />
oj<br />
to animals<br />
blä<br />
oh<br />
schas<br />
ha<br />
åh<br />
ptroo<br />
håhå(jaja)<br />
ä<br />
"t" [|]<br />
tvi<br />
öh<br />
åhå<br />
"p" [Ö]<br />
ä<br />
song<br />
to persons<br />
äh<br />
lala(la)<br />
jaja<br />
bah<br />
trala(la)<br />
aja(baja)<br />
asch<br />
sch<br />
äsch<br />
other bodily or<br />
vyss<br />
isch<br />
mental feel<strong>in</strong>g<br />
lull lull<br />
uh<br />
urrk<br />
aha<br />
aj<br />
pst<br />
bu<br />
fy<br />
brr<br />
tss<br />
positive<br />
hm<br />
åhej(åhå)<br />
GREETINGS<br />
ohoj<br />
tjo(hej)<br />
ah<br />
puh<br />
pust<br />
hoho<br />
oh<br />
ojojoj<br />
åh<br />
vojvoj<br />
haha(ha)<br />
håhå(jaja)<br />
atjo<br />
195
6.3.2 Discussion of Swedish expressive<br />
<strong>in</strong>terjections, commands and greet<strong>in</strong>gs<br />
Some short comments are made here on the phonetic/phonological<br />
structure of the <strong>in</strong>terjections <strong>in</strong> table 6.2 <strong>in</strong> order to relate them to the<br />
analyses <strong>in</strong> chapters 2, 4 and 5. They often consist of only a vowel,<br />
sometimes extra long, or of only a consonant, often extra long. They can<br />
beg<strong>in</strong> with a vowel or a consonant. The more open vowels are preferred,<br />
especially [A] and [a] but also [O]. The most common phonemes are j and<br />
h.<br />
Of the <strong>in</strong>itial clusters, bl-, tv-, ts, tr-, br, gr-, ptr and pst, that occur<br />
among these <strong>in</strong>terjections, bl-, tr- and gr- conform with the analysis <strong>in</strong><br />
chapter 4, i.e. bl- can be 'pejorative', tr- can be 'talk<strong>in</strong>g' and gr- can be<br />
'bad mood'. The clusters ts, ptr and pst are unconventional. The f<strong>in</strong>al<br />
clusters are very few: -rk, -st and -ms. None of these conform to the<br />
mean<strong>in</strong>gs of f<strong>in</strong>al clusters of chapter 5.<br />
This result is <strong>in</strong> accordance with the analysis presented <strong>in</strong> 2.4, which said<br />
that the sound side of expressive <strong>in</strong>terjections are realized more<br />
unsystematically, i.e. they are not (partly) built up by phonesthemes as<br />
sound symbolic roots are. But, as can be seen, the borders are not<br />
absolute between expressive <strong>in</strong>terjections and sound symbolic<br />
phonesthemes. Three <strong>in</strong>terjections – blä (bl- 'pejorative'), tralala (tr-<br />
'talk<strong>in</strong>g') and grr (gr- 'bad mood') – conform with the phonesthemic<br />
analysis.<br />
Typical for the <strong>in</strong>terjections are the special sounds or sound comb<strong>in</strong>ations<br />
that occur, e.g. click sounds and non-standard phonotactic comb<strong>in</strong>ations,<br />
e.g. [ptro:], [ts:], [pst] and [hm], and isolated consonants like [S]. CVstructure<br />
and reduplicated CV-structure are also common, as well as the<br />
lengthen<strong>in</strong>g of vowels or consonants. Consonant frames with a shift of<br />
vowel, e.g. asch, äsch, usch, isch also occur. Here it is quite obvious that<br />
the vowel quality imparts different mean<strong>in</strong>gs (cf. 5.4). The same goes for<br />
ah, äh, eh, i(h), ö(h). Prosody is probably important but has not been<br />
studied here. None of the phonological or phonetic characteristics is<br />
clearly connected with a certa<strong>in</strong> semantic feature.<br />
196
To summarize, few of the <strong>in</strong>terjections have (conventional) <strong>in</strong>itial<br />
consonant clusters. Of those 5 conventional clusters that occur, 3 conform<br />
with the analysis of chapter 4: bl- (pejorative), tr- (talk<strong>in</strong>g) and gr- (bad<br />
mood). The f<strong>in</strong>al clusters of <strong>in</strong>terjections are very few, and none of them<br />
conforms with the analysis <strong>in</strong> chapter 5.<br />
6.3.3 Phonological and phonetic similarities and<br />
dissimilarities between <strong>in</strong>terjections of different<br />
languages<br />
One <strong>in</strong>terest<strong>in</strong>g question now is: Are there phonological and phonetic<br />
similarities between the <strong>in</strong>terjections of different languages? Expressive<br />
terms, which mirror bodily and mental states, could be similar <strong>in</strong><br />
different languages, with a larger or smaller language specific<br />
superstructure. (cf. Abel<strong>in</strong> and Allwood, 1984). An analysis of my<br />
material (presented <strong>in</strong> Appendix 2) gives the follow<strong>in</strong>g result:<br />
Each semantic category shows its sound pattern tendencies for the<br />
different languages. These tendencies are enhanced by big differences<br />
between the categories. The pejorative <strong>in</strong>terjections, for example, often<br />
conta<strong>in</strong> an [u:] or an [O], the positive <strong>in</strong>terjections an [i] or an [a], surprise<br />
often an [O] or an [a]. 'Pa<strong>in</strong>' (cf. Swedish aj) has an [a] and a [j] or a<br />
diphtongized open vowel. 'Sneeze' (cf. Swedish atjo 4 ) has a voiceless<br />
fricative or an affricate <strong>in</strong> all the languages be<strong>in</strong>g studied, and the<br />
<strong>in</strong>terjection for go away (to an animal) (cf. Swedish schas) has a voiceless<br />
fricative. 'Good taste' has a nasal <strong>in</strong> these 8 languages, and 'scar<strong>in</strong>g<br />
somebody' has a voiced stop (mostly b) <strong>in</strong> all these languages except for<br />
two.<br />
When count<strong>in</strong>g all <strong>in</strong>stances of vowels and consonants <strong>in</strong> the <strong>in</strong>terjections<br />
<strong>in</strong> Appendix 2 (exclud<strong>in</strong>g the Swedish ones, s<strong>in</strong>ce they are greater <strong>in</strong><br />
number) the follow<strong>in</strong>g tendencies are found, shown <strong>in</strong> table 6.3. (Which<br />
languages contributed to which sounds can be seen <strong>in</strong> Appendix 2 and<br />
table 6.3). Many of the categories of table 6.3 are more specific than<br />
those <strong>in</strong> table 6.2, depend<strong>in</strong>g on what answers the <strong>in</strong>formants gave, cf.<br />
4 Different ways of conventionaliz<strong>in</strong>g this physical outburst (which of course can vary for<br />
different <strong>in</strong>dividuals and from time to time) is, <strong>in</strong> Swedish, mirrored <strong>in</strong> older spell<strong>in</strong>gs of<br />
this word: atschi, atsji, aptschäh, apschohoj, aaah-tschah-katsch katsch, kaa-kah, tschah,<br />
tjas, hlutt, schtschi, tjihitt, tjihihitt, tjitjisit (Ideforss, 1928, p. 25).<br />
197
Appendix 2. The categories 'song' and 'greet<strong>in</strong>gs' were not elicited from<br />
the <strong>in</strong>formants.<br />
Table 6.3 The most common vowels and consonants for different<br />
categories of <strong>in</strong>terjections. The languages are Icelandic, English, Polish,<br />
Hungarian, F<strong>in</strong>nish, Ososo, Malagasi and Slovenian.<br />
vowels consonants<br />
pejorative O (9), u (9), i (5)<br />
O:<br />
Icelandic: 2<br />
English: 1<br />
Polish: 2<br />
Hungarian: 1<br />
Ososo: 2<br />
Slovenian: 1<br />
u:<br />
Icelandic: 1<br />
English: 3<br />
Hungarian: 3<br />
F<strong>in</strong>nish: 2<br />
i:<br />
Hungarian: 1<br />
F<strong>in</strong>nish: 2<br />
Slovenian: 2<br />
positive i (6), a (5)<br />
i:<br />
Icelandic: 2<br />
English: 2<br />
F<strong>in</strong>nish: 2<br />
a:<br />
Icelandic: 1<br />
English: 2<br />
F<strong>in</strong>nish: 1<br />
Ososo: 1<br />
198<br />
j (10), f (6), h (6)<br />
j:<br />
Icelandic: 3<br />
English: 1<br />
Polish: 1<br />
Hungarian: 4<br />
Ososo: 1<br />
f:<br />
English: 2<br />
Hungarian: 1<br />
Ososo: 1<br />
Slovenian: 2<br />
h:<br />
F<strong>in</strong>nish: 5<br />
Ososo: 1<br />
j (3), p (3)<br />
j:<br />
Icelandic: 1<br />
English: 1<br />
F<strong>in</strong>nish: 1<br />
p:<br />
Icelandic: 1<br />
English: 1<br />
F<strong>in</strong>nish: 1
surprise O (7), a (4)<br />
O:<br />
English: 1<br />
Polish: 1<br />
Ososo: 1<br />
Slovenian: 4<br />
a:<br />
Icelandic: 1<br />
English: 1<br />
Ososo: 1<br />
Malagasi: 1<br />
pa<strong>in</strong> a (9), u (6)<br />
a:<br />
Icelandic: 2<br />
English: 2<br />
Hungarian: 2<br />
F<strong>in</strong>nish: 1<br />
Malagasi: 1<br />
u:<br />
Icelandic: 2<br />
Hungarian: 2<br />
Ososo: 2<br />
freez<strong>in</strong>g O (2)<br />
O:<br />
Polish: 2<br />
199<br />
j (3), h (3)<br />
j:<br />
Hungarian: 1<br />
Ososo: 1<br />
Slovenian: 1<br />
h:<br />
Icelandic: 1<br />
Hungarian: 1<br />
Slovenian: 1<br />
j (7)<br />
j:<br />
Icelandic: 1<br />
Hungarian: 4<br />
F<strong>in</strong>nish: 1<br />
Malagasi: 1<br />
b (3), r (3)<br />
b:<br />
English: 1<br />
Polish: 1<br />
Hungarian: 1<br />
r:<br />
English: 1<br />
Polish: 1<br />
Hungarian: 1<br />
thoughtfulness - h (5), m (6)<br />
h:<br />
Icelandic: 1<br />
English: 2<br />
Polish: 1<br />
F<strong>in</strong>nish: 1<br />
m:<br />
Icelandic: 2<br />
English: 2<br />
Polish: 1<br />
F<strong>in</strong>nish: 1
exhaustion u (4)<br />
u:<br />
English: 1<br />
Polish: 1<br />
F<strong>in</strong>nish: 2<br />
sudden <strong>in</strong>sight a (10)<br />
a:<br />
English: 4<br />
Hungarian: 2<br />
Ososo: 2<br />
Slovenian: 2<br />
sneeze i (8), a (6)<br />
i:<br />
English: 1<br />
Polish: 1<br />
Hungarian: 1<br />
Ososo: 3<br />
Slovenian: 2<br />
a:<br />
Icelandic: 1<br />
English: 1<br />
Polish: 1<br />
Hungarian: 1<br />
Slovenian: 2<br />
good taste a (6)<br />
a:<br />
Icelandic: 3<br />
English: 1<br />
Polish: 2<br />
commands to<br />
animals<br />
i (3), u (5)<br />
i:<br />
Icelandic: 2<br />
Hungarian: 1<br />
u:<br />
Icelandic: 2<br />
English: 1<br />
Ososo: 1<br />
Malagasi: 1<br />
200<br />
h (4)<br />
h:<br />
Hungarian: 1<br />
F<strong>in</strong>nish: 3<br />
h (7)<br />
h:<br />
English: 4<br />
Hungarian: 1<br />
Ososo: 1<br />
Slovenian: 1<br />
t (6), C (5)<br />
t:<br />
Icelandic: 1<br />
English: 1<br />
Hungarian: 1<br />
Ososo: 1<br />
Slovenian: 2<br />
C :<br />
Icelandic: 1<br />
Polish: 1<br />
Ososo: 1<br />
Slovenian: 2<br />
m (11), n (7)<br />
m:<br />
Icelandic: 4<br />
English: 1<br />
Polish: 2<br />
F<strong>in</strong>nish: 1<br />
Slovenian: 3<br />
n:<br />
Icelandic: 3<br />
Polish: 2<br />
Slovenian: 2<br />
S (7)<br />
S:<br />
Icelandic: 1<br />
English: 1<br />
Polish: 1<br />
Hungarian: 2<br />
F<strong>in</strong>nish: 1<br />
Malagasi: 1
mild warn<strong>in</strong>g to<br />
children<br />
- s (3)<br />
s:<br />
Icelandic: 3<br />
be quiet - S (4), t (3)<br />
S:<br />
English: 1<br />
Hungarian: 1<br />
F<strong>in</strong>nish: 1<br />
Malagasi: 1<br />
t :<br />
Polish: 1<br />
scar<strong>in</strong>g somebody u (5)<br />
u:<br />
English: 1<br />
Polish: 2<br />
Hungarian: 2<br />
201<br />
Hungarian: 2<br />
b (6)<br />
b:<br />
Icelandic: 1<br />
English: 1<br />
Polish: 2<br />
Hungarian: 1<br />
F<strong>in</strong>nish: 1<br />
The table shows that there are many vowels and consonants that are<br />
similar <strong>in</strong> the languages Icelandic, English, Polish, Hungarian, F<strong>in</strong>nish,<br />
Ososo, Malagasi and Slovenian for the categories 'pejorative', 'pa<strong>in</strong>',<br />
'sneeze', 'good taste', 'be quiet', 'thoughtfulness' respectively. These<br />
vowels and consonants of all the functions of table 6.3 are (approximately)<br />
[i], [a], [O], [u] and [p], [b], [m], [f], [t], [s], [n], [r], [j], [C], [S], [h]. Most of<br />
these consonants are labial or dental. There are no consonants produced<br />
beh<strong>in</strong>d the hard palate, except for [h]. The vowels are few and mostly [i],<br />
[u], and [a].<br />
It also seems that a certa<strong>in</strong> sound (or sound comb<strong>in</strong>ation) is preferred <strong>in</strong> a<br />
certa<strong>in</strong> language and is used for a variety of mean<strong>in</strong>gs. The Polish<br />
<strong>in</strong>formant, for example, prefers the sound comb<strong>in</strong>ations [ux], [Ox] and [O],<br />
while the Hungarian <strong>in</strong>formant prefers [jOj], [jaj], [juj].<br />
There are also different degrees of conventionalization <strong>in</strong> the language,<br />
both depend<strong>in</strong>g on speakers and on the situation (wild and tame forms <strong>in</strong><br />
Rhodes' (1995) term<strong>in</strong>ology). Thus, <strong>in</strong> Swedish there are both expiration
and the <strong>in</strong>terjection uh for 'tiredness', and an imitation of the horse's<br />
neigh or the <strong>in</strong>terjection gnägg. 5<br />
6.4 Imitations of animal calls<br />
A subgroup to the onomatopoeic <strong>in</strong>terjections (cf. 2.4) are those that<br />
imitate animal sounds.<br />
sound imitative<br />
Animal call imitative Imitative of physical and bodily reactions Other<br />
Figure 6.1. Relations between different types of onomatopoeic<br />
<strong>in</strong>terjections.<br />
Many animals have laryngeal and supraglottal organs and cavities similar<br />
to man, but with different resonance properties. Most of them produce<br />
sounds <strong>in</strong> the same way as humans (cf. L<strong>in</strong>dblad, 1992). These sounds<br />
seem to be sufficiently similar to allow for imitation <strong>in</strong> human languages.<br />
6.4.1 Expressions for animal calls <strong>in</strong> Swedish and<br />
other languages<br />
The human expressions that imitate animal calls are easy to categorize<br />
semantically. (I presume that e.g. a Swedish cat sounds the same as a cat<br />
from geographically distant countries.) Interjections imitat<strong>in</strong>g animal calls<br />
have been translated, with the help of <strong>in</strong>formants, <strong>in</strong>to 16 different<br />
languages: Icelandic, English, Polish, Hungarian, F<strong>in</strong>nish, Ososo,<br />
Malagasi, Korean, Japanese, Ch<strong>in</strong>ese, Estonian, Urdu, Persian, Kurdish,<br />
Arabic, and Spanish (see table 6.4).<br />
5There is reason to believe that comic strips have contributed to a conventionalization of<br />
both writ<strong>in</strong>g and speech (via writ<strong>in</strong>g).<br />
202
Table 6.4.a Interjections imitat<strong>in</strong>g 7 animal calls, from 16 different languages.<br />
mouse cat goat/<br />
sheep<br />
Korean tSikTSik jaON<br />
njaON<br />
Japanese tSiju: nija:<br />
nijaN<br />
Ch<strong>in</strong>ese tSitSitSi mimio<br />
mijao<br />
dog pig horse cow<br />
mœœ: mUNmUN kulkul hihi:hiN Im:E:<br />
mœhœ: kENkEN<br />
mE:mE: waNwaN bu:bu: hihiN mo:<br />
mimi:<br />
mimie:<br />
w´w´w´ IhIhIh ahahaha Na:<br />
F<strong>in</strong>nish pyppyp miau mœ:œ: hauhau<br />
vuhvuh<br />
rØhrØh ihaha: am:u:<br />
Estonian pi:p miau mØk:<br />
mœ:<br />
auxaux ØhØh ihaha mu:<br />
Urdu tSitSi miaow mE:<br />
bE:<br />
bONbON – - m:<br />
Persian - miau bE:bE: howhow korokoro hihihi:<br />
daNtSo<br />
mE:<br />
Kurdish DikDikDik miau be: wOwwOw – - ma:<br />
me:<br />
bOrabOr<br />
Arabic - miau me:<br />
ba:<br />
habhab – hEehEe bu:<br />
Polish pipi miau me: hauhau – hihi mu:<br />
English skwi:k<br />
i:k<br />
miau ba: bauwau<br />
wOfwOf<br />
japjap<br />
Swedish pi:p pi:p miau bœ:<br />
mœ:<br />
vuv:uv<br />
vufvuf<br />
203<br />
OiNkOiNk neigh mu:<br />
nØf:nØf: gnEg:nEg m¨<br />
Spanish - miau be:be. guau OiNkOiNk - mu:<br />
Icelandic pi:p pi:p miau mœ: vufvuf ¯ njihaha møù<br />
Hungarian - miau bœ:œœ<br />
mœkmœk<br />
vauvau rØf:rØf: - mu:<br />
Ososo - miau bœœ bOubOu imitation - mu:<br />
Table 6.4.b Interjections imitat<strong>in</strong>g another 8 animal calls, from 16 different languag<br />
KoreanJapaneseCh<strong>in</strong>eseF<strong>in</strong>nishEstonian<br />
crow cuckoo owl duck hen rooster frog chicken<br />
ka:kka:k PUkukPuUNkwakkokotœpUkukpuUNkwakkoko ka:ka: kak:o: ho:kho:k ga:ga: kokoko kokekoko - -<br />
kokjo: - -<br />
tSijatSija pukupuku ahaha ka:kak kukuku kIkIkI - -<br />
va:kva:k kuk:u:kuk:u:<br />
huhu: kvak-kvak kO:tkotkot kuk:ukie:ku<br />
kva:k -<br />
- kuku uhu: prœ:k kaka kukeleku: - -<br />
Urdu kajkaj kokokoko - kvakva kONkONkON<br />
kOk´lONkON<br />
- -
PersianKurdishArabicPolishEnglishSwedish<br />
SpanishIcelandic<br />
kr¨k-<br />
--kr¨k<br />
HungarianOsoso<br />
204<br />
- -<br />
VarVar hoho huo: kwak- VodVod kukulikwakkoko<br />
- - ba:bu:kwak- kOkOkokO qOqOqOqO - -<br />
ba:bu:kwak - - hu:hu: waq ququ qoqoqo - -<br />
krakra kuku uhu: kwakwa koko kukuriku - -<br />
kOwkOw kuku: - kwak- kl´kkl´k kOk´kwakdud´ldu:krakskraks<br />
kuku: huhu: kvakkvak kakaka<br />
- - - kwakkwak<br />
- -<br />
kvak:vak: pi:p pi:p<br />
kPkElik¨:<br />
klPkklPk<br />
- kikiriki: - -<br />
- ¨hu: bra:bra: kakaka kPkElikP<br />
- pi:ppi:p<br />
ka:rka:r - - haphap kOtkOt kukuriku - tCiptCip<br />
- - u:u: kuOk klOkklOk kukuru:kuOkOkOrOkO<br />
The follow<strong>in</strong>g conclusions can be drawn from comparisons between<br />
expressions for animal sounds:<br />
1. There is imitation of animal calls <strong>in</strong> all the languages <strong>in</strong> the study.<br />
2. No animal call imitation is exactly the same <strong>in</strong> all languages.<br />
krukkruk tCiatCia<br />
3. Some animal call imitations are similar <strong>in</strong> the different languages,<br />
while others vary more. For example, the cat's sound is conventionalized<br />
as [miau] <strong>in</strong> all languages except <strong>in</strong> Korean where it is [njaON], Japanese<br />
where it is [nijaN] and Ch<strong>in</strong>ese where it can be [mimio] (but also [mijao]).<br />
The expression for the dog's sound, however, varies greatly between the<br />
languages. One possible explanation for this is that the calls of the<br />
different animals differ <strong>in</strong> complexity so that for the complicated animal<br />
calls, different languages attend to different acoustic characteristics with a<br />
po<strong>in</strong>t of departure <strong>in</strong> the phonology of the language. Another possibility<br />
is that some animals simply have a bigger repertoire of calls.<br />
4. There are mostly similarities between expressions for animal calls on<br />
the level of phonological features. For example, all imitations of the cats<br />
meow <strong>in</strong>clude a nasal, and the imitation of the rooster always conta<strong>in</strong>s a
voiceless plosive, velar or uvular. Sometimes the similarity is <strong>in</strong> the form<br />
of reduplication. The dog says [vuv:uv] or [vuf:vuf] <strong>in</strong> Swedish, the same<br />
<strong>in</strong> Icelandic, and [vau vau] <strong>in</strong> Hungarian, so <strong>in</strong> these languages the <strong>in</strong>itial<br />
consonant is the same. In Ososo the dog says [bOubOu]. The likeness with<br />
the previous examples lies <strong>in</strong> the <strong>in</strong>itial consonant which is voiced and<br />
labial, and that there is an [u] vowel.<br />
5. Prosody is probably important but not analyzed here.<br />
6. With<strong>in</strong> each language here, as well as for the <strong>in</strong>terjections, there are<br />
likenesses which can be supposed to depend on, or reflect, the<br />
phonological structure of the language <strong>in</strong> question. For example, Arabic<br />
has [q] where e.g. Swedish has [k], Swedish has [¨] <strong>in</strong>stead of [u], Korean<br />
and Japanese often have a f<strong>in</strong>al [N].<br />
6.4.2 A test of expressions for animal calls of<br />
different languages<br />
Sixteen of the expressions for animal calls from different languages were<br />
chosen to test how speakers from other languages would <strong>in</strong>terpret them.<br />
Those chosen were the ones where the sound shape of the expression was<br />
the most deviant <strong>in</strong> the corpus <strong>in</strong> tables 6.4.a and 6.4.b, <strong>in</strong> order to make<br />
the test as difficult as possible. The expressions of animal calls are listed<br />
<strong>in</strong> table 6.5.<br />
The subjects <strong>in</strong> this test were 15 people, with the follow<strong>in</strong>g mother<br />
tongues: Swedish (8 subjects), French (2 subjects), English, Hungarian,<br />
Czech, Slovenian, Ososo. All were tested on the same occasion. They<br />
listened to pronunciations of the words for the animal calls and they saw<br />
them transcribed as <strong>in</strong> table 6.5. They were told to guess which animals<br />
had gotten their calls conventionalized <strong>in</strong> this way, and to write down<br />
their answers.<br />
205
Table 6.5 The 16 expressions of animal calls chosen for the <strong>in</strong>terpretation<br />
experiment.<br />
CALL ANIMAL LANGUAGE<br />
1. vau vau dog Hungarian<br />
mπ´ 2. : cow Icelandic<br />
3. njihaha horse Hungarian<br />
4. hap hap duck Hungarian<br />
5. rØf rØf pig Hungarian<br />
6. bra: bra: duck Icelandic<br />
7. kOt kOt hen Hungarian, F<strong>in</strong>nish<br />
8. kOkEkO‘ko: rooster Japanese<br />
9. prœ:k duck Estonian<br />
10. pakUk: pakUk: cuckoo Korean<br />
11. njaON cat Korean<br />
12. bu: cow, pig Arabic, Czech,<br />
Japanese<br />
13. kOk´lONkON rooster Urdu<br />
14. tSi tSi mouse Urdu<br />
15. maN maN dog Korean<br />
16. hab: hab: dog Arabic<br />
6.4.3 Results from a test of expressions of animal<br />
calls <strong>in</strong> different languages<br />
The results were as listed <strong>in</strong> table 6.6. Sometimes a subject had written<br />
down more than one answer. All answers have been counted except for<br />
when subjects gave a judgement on a word <strong>in</strong> his/her first language.<br />
Table 6.6 Assignment of animal calls to animals, by 8 speakers of Swedish<br />
and 7 speakers of other languages: French (2 subjects), English,<br />
Hungarian, Czech, Slovenian, Ososo.<br />
8 Swedish 7 other languages<br />
expected answer other answer expected answer other answer<br />
1.dog 8 6<br />
2. cow 8 7<br />
3. horse 5 donkey 5 3 donkey 2<br />
jackal 1 cat 1<br />
4. duck 0 ? 5 0 ? 4<br />
goose 1 frog 2<br />
hen 1<br />
hippopotamus 1<br />
5. pig 7 dog 1 2 dog 3<br />
? 1<br />
206
6. duck 0 sheep 5 1 donkey 1<br />
crow 2 raven 1<br />
lion 1 crow 1<br />
wolf 1 sheep 1<br />
? 1 ? 2<br />
7. hen 2 goose 1 4 ? 3<br />
squirrel 1<br />
? 4<br />
8. rooster 8 5 hen 1<br />
pigeon 1<br />
9. duck 0 turkey 1 2 frog 1<br />
sheep 1 ? 4<br />
magpie 1<br />
? 5<br />
10. cuckoo 0 hen 4 1 turkey 2<br />
rooster 1 hen 1<br />
? 3 rooster 1<br />
? 2<br />
11. katt 7 ? 1 7<br />
12. cow, pig<br />
(cow)<br />
2 sheep 2 5 (cow) frog 1<br />
owl 2 ? 1<br />
? 2<br />
13. rooster 7 hen 1 5 ? 2<br />
14. mouse 0 squirrel 1 1 bird 2<br />
great titmouse 1 sparrow 1<br />
small bird 1 chicken 1<br />
snake 1 ? 2<br />
? 3<br />
15. dog (ch<strong>in</strong>.) 1 ch<strong>in</strong>. peacock 1 0 Ch<strong>in</strong>. peacock 1<br />
pek<strong>in</strong>gese dog ? 6<br />
rabbit 1<br />
? 5<br />
16. dog 1 duck 1 1 frog 1<br />
? 6 duck 1<br />
? 4<br />
207
6.4.4 Discussion of the test on identify<strong>in</strong>g animal<br />
calls<br />
The animals that were correctly identified by all listeners (who gave an<br />
answer) were: dog (<strong>in</strong> Hungarian, [vau vau]), cow (<strong>in</strong> [mπ´<br />
Icelandic, :])<br />
and cat (<strong>in</strong> Korean, [njaON]). The Arabic word for the dog’s bark, [hab:<br />
hab:] and the Korean [maN maN] were not as accurately identified. The<br />
three <strong>in</strong>stances of the duck's sound (from Hungarian, Icelandic and<br />
Estonian) all gave a variety of answers. Only one animal sound was not<br />
identified by anybody, [hap hap] (Hungarian duck). The animal sounds<br />
that were identified by only 1 person were [bra: bra:] (Icelandic duck),<br />
[pakUk: pakUk:] (Korean cuckoo, however, was <strong>in</strong>terpreted as other<br />
birds), [tSi tSi] (mouse <strong>in</strong> Urdu, however, <strong>in</strong>terpreted as various small<br />
animals), [maN maN] (Korean dog).<br />
In general it can be said that certa<strong>in</strong> animal calls were more difficult to<br />
identify, others were easier. The words for the duck sounds were difficult<br />
while the word for the cat's meow was easy, <strong>in</strong> spite of the fact that the<br />
word chosen for a meow for the test was the one that was most deviant.<br />
There was no clear difference between the larger group of Swedish<br />
speakers and speakers of the other languages.<br />
Several times the <strong>in</strong>formants guessed on the correct k<strong>in</strong>d of animal, even<br />
though they did not give exactly the expected answer, e.g. 'bird' (see 6, 8,<br />
10) or 'small animal' (see 14). This means that they identified the size of<br />
the animal <strong>in</strong> question. Some subjects were more unsure than others. No<br />
subject guessed throughout on certa<strong>in</strong> animals. A few subjects were more<br />
imag<strong>in</strong>ative and more specific. 6<br />
6.5 Conclusions of studies of expressive<br />
<strong>in</strong>terjections and expressions of animal calls<br />
Interjections and expressions of animal calls exist <strong>in</strong> all the languages <strong>in</strong><br />
the study. There are similar sounds and mean<strong>in</strong>gs <strong>in</strong> all the languages, but<br />
there are also categories that seem to be specific to a certa<strong>in</strong> language,<br />
e.g. Icelandic has a special command 'go away' directed to sheep. The<br />
expressions of the <strong>in</strong>terjections are not the same <strong>in</strong> all the languages, but<br />
on the other hand they are not totally different; for some categories they<br />
6 In question 15 there might be an error; somebody must have whispered "Ch<strong>in</strong>ese" aloud.<br />
It is still <strong>in</strong>terest<strong>in</strong>g that several subjects judged "Ch<strong>in</strong>ese" to be fitt<strong>in</strong>g.<br />
208
are very similar, e.g. 'sneeze', 'good taste' cf. table 6.3. These expressions<br />
have an onomatopoeic basis which can be conventionalized <strong>in</strong> different<br />
ways <strong>in</strong> the different languages (the expressive <strong>in</strong>terjections also have an<br />
<strong>in</strong>dexical basis, i.e. a bodily reaction produces a sound, e.g. a sneeze or a<br />
scream, cf. 2.4). The phonological system <strong>in</strong>fluences the perception of the<br />
sounds and the choice of an adequate expression for imitation, e.g. [D] <strong>in</strong><br />
Icelandic. There are, however, tendencies for the whole material: most of<br />
these consonants are labial or dental. The vowels are few and mostly [i],<br />
[a], [u].<br />
The study of expressions of animal calls shows similar phonological<br />
tendencies (the semantic categories <strong>in</strong>vestigated were fixed, as specific<br />
animal calls were asked for): The imitations of the animal calls are not the<br />
same <strong>in</strong> all the languages, but on the other hand they are not totally<br />
different. Also, some expressions of animal calls are more alike <strong>in</strong> the<br />
different languages, e.g. Persian cat [miau] and F<strong>in</strong>nish cat [miau] while<br />
others vary more, e.g. Korean pig [kulkul] and Swedish pig [nØf:nØf:] cf.<br />
table 6.4. The imitations of animal calls can be conventionalized <strong>in</strong><br />
different ways <strong>in</strong> the different languages, e.g. Swedish has [P] or [¨]<br />
where other languages have [u], for example Swedish [kPkelik¨:] and<br />
Hungarian [kukuriku]. The most common vowels are [i], [a], [u], i.e.,<br />
closed or open vowels.<br />
In the <strong>in</strong>terpretation test of expressions of animal calls (cf. 6.4.2–6.4.4),<br />
some animals were identified by all listeners (dog, cow and cat), some<br />
less correctly identified and one animal was not identified at all (duck,<br />
Hungarian). The <strong>in</strong>terpretation test shows that, given the context of<br />
animal calls, it was quite easy for listeners to <strong>in</strong>terpret animal calls from<br />
languages other than their own.<br />
Prosody, voice quality and gestures are most probably important for both<br />
<strong>in</strong>terjections and imitation of animal calls. These forms for expression<br />
need to be studied further.<br />
209
6.6 Test of cross language <strong>in</strong>terpretation of<br />
Swedish onomatopoeic and other sound symbolic<br />
words<br />
6.6.1 Method<br />
In order to see how non-Swedish speakers <strong>in</strong>terpret Swedish<br />
onomatopoeic and other sound symbolic expressions, an experiment<br />
(similar to that <strong>in</strong> 7.1.3) was performed.<br />
The subjects were six persons who did not have Swedish as a first<br />
language, one each from of the languages Arabic, Spanish, German,<br />
Dutch, Ibo and English.<br />
A list of 15 written words was presented and read aloud to each subject.<br />
One of the words was repeated. Three of the words were not real ones<br />
but were <strong>in</strong>stead constructed out of phonesthemes. The subjects were told<br />
to write down their answers <strong>in</strong> a language of their own choice. The words<br />
are:<br />
fjompig<br />
skvalpa<br />
skrälle<br />
pjaltig (constructed)<br />
vresig<br />
glansig<br />
slabbig<br />
bjaltig (constructed)<br />
blankig (constructed end<strong>in</strong>g)<br />
pladdig (constructed)<br />
trumpen<br />
kladdig<br />
slabbig<br />
fladdrig<br />
grubbel<br />
stripig<br />
The subjects were <strong>in</strong>structed to try to <strong>in</strong>terpret each word and suggest an<br />
appropriate mean<strong>in</strong>g. What is here called the conventional <strong>in</strong>terpretation<br />
of a constructed word is one where the answer has a semantic feature that<br />
210
can be connected with a certa<strong>in</strong> consonant cluster <strong>in</strong> accordance with the<br />
analysis of chapter 4. These <strong>in</strong>itial clusters: pj-, bj-, pl- and bl- were<br />
successful <strong>in</strong> different parts of the test <strong>in</strong> chapter 7 (especially pj) and<br />
could thus be assumed to represent a Swedish norm.<br />
Subjects who knew some Swedish were told to mark the words that they<br />
knew already. After the test the subjects were free to orally elaborate on<br />
their answers.<br />
6.6.2 Results of <strong>in</strong>terpretation of cross language<br />
Swedish onomatopoeic and other sound symbolic<br />
words<br />
The subjects seldom <strong>in</strong>terpreted the words <strong>in</strong> the conventional way.<br />
However, nearly all of the answers given belong to semantic features of<br />
the model <strong>in</strong> 2.6.2 and the semantic features of chapter 4. It thus seems<br />
that certa<strong>in</strong> semantic categories are preferred to be expressed sound<br />
symbolically, but that the expression can vary.<br />
The subjects will first be accounted for one by one:<br />
Arabic speaker : This subject had very little knowledge of Swedish and<br />
knew none of the words. He answered <strong>in</strong> English. He guessed accord<strong>in</strong>g to<br />
convention on one word. The mean<strong>in</strong>g suggestions belonged to the<br />
follow<strong>in</strong>g categories: The ones <strong>in</strong>terpreted <strong>in</strong> an acceptable way, as<br />
concerns phonesthemes, were:<br />
fjompig - big mouth ('pejorative')<br />
pjaltig - speaks wrong ('pejorative', 'talk<strong>in</strong>g')<br />
slabbig - slap him/her on face ('beat')<br />
bjaltig - speaks wrong ('pejorative', 'talk<strong>in</strong>g')<br />
trumpen - drum’s sound ('sound')<br />
fladdrig - eruption ('quick or strong movement')<br />
For some of the mis<strong>in</strong>terpreted words (mis<strong>in</strong>terpreted <strong>in</strong> terms of<br />
phonesthemes), the semantic categories of the given answers were<br />
nevertheless <strong>in</strong> accordance with the sound symbolic categories of the<br />
model <strong>in</strong> 2.6.2 (These mis<strong>in</strong>terpreted words are labeled "unconventional<br />
but possible category" <strong>in</strong> table 6.7).<br />
211
'wetness'<br />
'light'<br />
'quick or strong movement', 'wetness'<br />
'quick or strong movement'<br />
The responses that did not fit <strong>in</strong>to any category of the semantic model (or<br />
were difficult to <strong>in</strong>terpret) are: shut up, close the door, k<strong>in</strong>d of penalty,<br />
<strong>in</strong>tuitive, my country.<br />
Spanish speaker : This subject had a good knowledge of Swedish, both<br />
passively and actively. Her answers were given <strong>in</strong> Spanish, after which<br />
they were translated. The subject knew 4 of the words and guessed<br />
accord<strong>in</strong>g to convention, <strong>in</strong> terms of phonesthemes, on 3. The ones<br />
<strong>in</strong>terpreted <strong>in</strong> an acceptable way, as concerns phonesthemes, are:<br />
skrälle - chim<strong>in</strong>g ('sound') 7<br />
blankig - opaque ('light') 8<br />
As for the Arabic speaker, for some of the mis<strong>in</strong>terpreted words (<strong>in</strong><br />
terms of phonesthemes), the semantic categories of the given answers<br />
were <strong>in</strong> accordance with sound symbolic categories of the model <strong>in</strong> 2.6.2.<br />
'adhesion'<br />
'beat'<br />
'long th<strong>in</strong> form' (2)<br />
'rough surface structure'<br />
'destruction'<br />
German speaker: This subject had a good passive knowledge of Swedish,<br />
but he did not know any of the words on the list. The answers were given<br />
<strong>in</strong> English. He guessed accord<strong>in</strong>g to convention on 3 words. The words<br />
that were given an acceptable <strong>in</strong>terpretation, <strong>in</strong> terms of phonesthemes,<br />
were:<br />
fjompig - condescend<strong>in</strong>g ('pejorative')<br />
vresig - angry ('mental feel<strong>in</strong>g')<br />
glansig - sh<strong>in</strong>y ('light')<br />
7 Skr- does not imitate a chim<strong>in</strong>g sound at all, but <strong>in</strong> the previous analysis no dist<strong>in</strong>ction<br />
has been made between different k<strong>in</strong>ds of sound.<br />
8 This is of course the opposite of 'light'.<br />
212
lankig - sh<strong>in</strong>y ('light')<br />
pladdig - looks like a blot of e.g. yogurt ('wetness')<br />
fladdrig - someth<strong>in</strong>g which is decomposed, th<strong>in</strong> ('th<strong>in</strong> form')<br />
The rema<strong>in</strong><strong>in</strong>g mean<strong>in</strong>g suggestions can be classified <strong>in</strong> the follow<strong>in</strong>g<br />
semantic fields:<br />
'small size'<br />
'rough surface structure' (wr<strong>in</strong>kled sk<strong>in</strong> of tomato)<br />
'form' (distorted)<br />
'pejorative' (distorted; heavy, fat, uncontrolled; unordered)<br />
'mood' (aggressive - of females)<br />
More difficult to classify <strong>in</strong> the previously discussed categories are the<br />
answers: "powerful"; "no more energy".<br />
Dutch speaker : This speaker had studied Swedish for a year and was very<br />
fluent. She had lived <strong>in</strong> Sweden for a very short period. She gave her<br />
answers <strong>in</strong> Swedish or English. She knew 1 of the words and guessed<br />
accord<strong>in</strong>g to convention on 1 word. The words that were given a<br />
plausible <strong>in</strong>terpretation, <strong>in</strong> terms of phonesthemes, were:<br />
skvalpa - call names at someone ('talk<strong>in</strong>g', 'pejorative')<br />
skrälle - say someth<strong>in</strong>g with a lot of sound ('sound', 'talk<strong>in</strong>g')<br />
pjaltig - snobbish ('pejorative')<br />
slabbig - unorganized ('pejorative')<br />
trumpen - garbage ('pejorative')<br />
kladdig - to write <strong>in</strong> an ugly way ('pejorative')<br />
stripig - someth<strong>in</strong>g with many stripes ('long th<strong>in</strong> form')<br />
For the mis<strong>in</strong>terpreted words (<strong>in</strong> terms of phonesthemes), the semantic<br />
categories given were, however, <strong>in</strong> accordance with the model:<br />
'form'<br />
'mental feel<strong>in</strong>g'<br />
'round form'<br />
'quick or strong movement' (2).<br />
One answer could not be classified accord<strong>in</strong>g to the model: 'void'.<br />
213
Ibo (and English) speaker: This speaker knew none of the words and did<br />
not guess accord<strong>in</strong>g to convention <strong>in</strong> any case. The words that were given<br />
a plausible <strong>in</strong>terpretation (acceptable category), <strong>in</strong> terms of phonesthemes,<br />
were:<br />
slabbig - lazy ('pejorative')<br />
pladdig - dirty ('pejorative')<br />
trumpen - loud ('sound')<br />
grubbel - grumble ('talk<strong>in</strong>g')<br />
The semantic categories given were <strong>in</strong> accordance with the model<br />
(unconventional but possible category) for some of the mis<strong>in</strong>terpreted<br />
words (<strong>in</strong> terms of phonesthemes):<br />
'putt<strong>in</strong>g together'<br />
'bodily feel<strong>in</strong>g'<br />
Seven of the answers could not be classified accord<strong>in</strong>g to the model<br />
(unconventional category): deep, slowness, positive, closed, tight, empty,<br />
wide.<br />
English speaker This subject had good knowledge of Swedish, Norwegian<br />
and German and answered <strong>in</strong> English. She knew 4 of the test words and<br />
guessed accord<strong>in</strong>g to convention on 2 of the words. The words that were<br />
given an acceptable <strong>in</strong>terpretation, <strong>in</strong> terms of phonesthemes, were:<br />
skvalpa - to gossip ('talk<strong>in</strong>g')<br />
skrälle - to compla<strong>in</strong>, wh<strong>in</strong>e ('talk<strong>in</strong>g')<br />
vresig - twist<strong>in</strong>g ('quick or strong movement')<br />
fladdrig - flimsy ('th<strong>in</strong> form')<br />
grubbel - annoy<strong>in</strong>g, irritat<strong>in</strong>g stuff ('bad mood')<br />
However, the semantic categories given were <strong>in</strong> accordance with the<br />
model (unconventional but possible category), for some of the<br />
mis<strong>in</strong>terpreted words (<strong>in</strong> terms of phonesthemes):<br />
'round form'<br />
'adhesion'<br />
'sound'<br />
214
One of the answers could not be classified accord<strong>in</strong>g to the model<br />
(unconventional category): tricked<br />
6.6.3 Conclusions from the test on cross language<br />
<strong>in</strong>terpretation of Swedish onomatopoeic and sound<br />
symbolic words<br />
The follow<strong>in</strong>g table, 6.7, shows the numbers of answers with different<br />
degrees of accuracy for each subject.<br />
Table 6.7 Summary of the numbers of answers with different degrees of<br />
accuracy <strong>in</strong> the cross l<strong>in</strong>guistic word <strong>in</strong>terpretation study.<br />
subject’s<br />
language<br />
known<br />
words<br />
guess <strong>in</strong><br />
accordance<br />
with<br />
conven-<br />
tion<br />
accept.<br />
category<br />
215<br />
unconventional<br />
but<br />
possible<br />
category<br />
unconventional<br />
category<br />
Arabic 0 1 6 4 5 0<br />
Spanish 4 3 2 6 0 0<br />
German 0 3 6 5 2 0<br />
Dutch 1 1 7 5 1 0<br />
Ibo 0 0 4 2 7 2<br />
English 4 2 5 3 1 0<br />
sum 9 10 30 25 15 2<br />
Few of the words were known to the subjects, not even to those who had<br />
good knowledge of Swedish. There were also quite few correct guesses.<br />
Their performance is better when their answers are classified <strong>in</strong>to<br />
semantic categories which are related to phonesthemes of the <strong>in</strong>itial<br />
clusters (column "acceptable category").<br />
The frequencies of words that were judged accord<strong>in</strong>g to convention (<strong>in</strong><br />
terms of phonesthemes) are shown <strong>in</strong> table 6.8.<br />
All 15 test words are represented <strong>in</strong> the results and there is no<br />
great preference for any word or words.<br />
no<br />
answer
Table 6.8 Frequencies of words that were most often successfully<br />
<strong>in</strong>terpreted accord<strong>in</strong>g to the Swedish norm, by the six speakers.<br />
Test words Number of speakers<br />
fladdrig 3<br />
skrälle 3<br />
trumpen 3<br />
blankig 2<br />
fjompig 2<br />
grubbel 2<br />
pjaltig 2<br />
pladdig 2<br />
skvalpa 2<br />
slabbig 2<br />
vresig 2<br />
bjaltig 1<br />
glansig 1<br />
kladdig 1<br />
stripig 1<br />
Table 6.9 shows which semantic categories were used most often, when<br />
the subjects guessed <strong>in</strong> an unconventional way (i.e. there is no possibility,<br />
accord<strong>in</strong>g to the lexical analysis of chapter 4, that the consonant cluster <strong>in</strong><br />
question can have this mean<strong>in</strong>g).<br />
This result means that even when subjects guess <strong>in</strong> an unconventional way,<br />
they still guess with<strong>in</strong> the semantic categories of the model. The answers<br />
that can not be classified with<strong>in</strong> the semantic categories of the model (the<br />
column “unconventional category” of table 6.7) are fewer. Most of them<br />
were produced by the Arabic and Ibo speaker, which suggests an<br />
<strong>in</strong>fluence of cultural (or l<strong>in</strong>guistic) differences, i.e. European vs. non-<br />
European. 9<br />
9 These results are very prelim<strong>in</strong>ary and clearly a larger <strong>in</strong>vestigation is needed.<br />
216
Table 6.9 Most commonly preferred mean<strong>in</strong>gs for unconventional but<br />
possible mean<strong>in</strong>gs.<br />
semantic features number of<br />
words<br />
quick or strong movement 4<br />
wetness 2<br />
adhesion 2<br />
long th<strong>in</strong> form 2<br />
rough surface structure 2<br />
round form 2<br />
form 2<br />
pejorative 1<br />
putt<strong>in</strong>g together 1<br />
bodily feel<strong>in</strong>g 1<br />
mood 1<br />
mental feel<strong>in</strong>g 1<br />
small size 1<br />
destruction 1<br />
beat 1<br />
sound 1<br />
light 1<br />
6.7 General conclusions and discussion of the<br />
cross language studies<br />
The follow<strong>in</strong>g conclusions can be drawn from the different contrastive<br />
studies reported <strong>in</strong> this chapter: The Thesaurus study has shown that the<br />
expression of phonesthemes can differ between languages as closely<br />
related as English and Swedish. In some phonesthemes, however, the<br />
expression (and mean<strong>in</strong>g) is the same.<br />
The study of <strong>in</strong>terjections shows that there are similarities of expression<br />
as well as of mean<strong>in</strong>g between unrelated languages. There is variation of<br />
expression with<strong>in</strong> semantic categories, but certa<strong>in</strong> sounds are still<br />
preferred.<br />
The study of expressions for animal calls <strong>in</strong> different languages shows<br />
that there are similarities as well as differences. The understand<strong>in</strong>g test<br />
217
showed that it is quite easy for subjects to guess correctly on expressions<br />
for animal calls <strong>in</strong> other languages. Some animals were more difficult to<br />
identify. Often the type of animal was identified, e.g. bird or small<br />
animal.<br />
The last study, on <strong>in</strong>terpretation of Swedish sound symbolic words and<br />
non-words (constructed out of phonesthemes), showed that it was quite<br />
difficult for subjects to <strong>in</strong>terpret the mean<strong>in</strong>gs <strong>in</strong> accordance with the<br />
Swedish norm, but that the mis<strong>in</strong>terpretations most often were with<strong>in</strong> the<br />
semantic categories of the model <strong>in</strong> 2.6.2.<br />
The general conclusion from these studies is that there are greater<br />
differences between languages for expression than for mean<strong>in</strong>g, <strong>in</strong> sound<br />
symbolism. The more onomatopoeic expressions are easier to <strong>in</strong>terpret<br />
than other sound symbolism for speakers of other languages.<br />
Interpretation of other sound symbolism often goes wrong (because<br />
expressions differ <strong>in</strong> different languages), but the semantic categories<br />
guessed on are most often with<strong>in</strong> the semantic model of this thesis.<br />
218
7 Experiments with constructed words conta<strong>in</strong><strong>in</strong>g<br />
phonesthemes<br />
7.1 Production and understand<strong>in</strong>g<br />
In section 1.4 the hypothesis was formulated that phonesthemes are<br />
productive, to a greater or lesser degree. In order to test this hypothesis I<br />
have carried out several experiments. The purpose of the experiments<br />
that will be described <strong>in</strong> this chapter is to test this hypothesis, also <strong>in</strong><br />
more detail for some of the phonesthemes. In chapter 4 the results of the<br />
lexical analysis made by one person were presented. The results of this<br />
analysis are to a certa<strong>in</strong> extent dependent on the idiosyncrasies of the<br />
<strong>in</strong>dividual subject, the material and other circumstances and are therefore<br />
prelim<strong>in</strong>ary. Therefore the results of the lexical analysis were used as a<br />
basis for test<strong>in</strong>g the sound symbolic value of certa<strong>in</strong> consonant clusters on<br />
a great number of <strong>in</strong>dividuals.<br />
Method and materials<br />
The tests (see Appendix 3) are constructed as free choice, multiple-choice<br />
and match<strong>in</strong>g tests.<br />
Test 1.a. is a free choice test, which goes from mean<strong>in</strong>g to expression, to<br />
test the production of sound symbolism, e.g. "Invent a short word for<br />
somebody who is stupid".<br />
Test 1.b. is a forced choice test which also goes from mean<strong>in</strong>g to<br />
expression. An example of a task is: "Which of the follow<strong>in</strong>g words fits<br />
best for a person who is silly: a) pjotig b) brotig c) splotig?" Only one<br />
answer should be possible accord<strong>in</strong>g to the previous analysis. For some of<br />
the questions, one of the answers is supposed to be clearly contradictory,<br />
i. e. the expected answer<strong>in</strong>g score is zero (where there are contrast<strong>in</strong>g<br />
mean<strong>in</strong>gs, e. g. 'dry'-'wet'). In a few questions a word which sounds<br />
similar to the test word, but with a non-expected cluster is <strong>in</strong>cluded to test<br />
which is more important, word analogy or sound symbolic clusters, e.g.<br />
"Which of the follow<strong>in</strong>g words best describes a broken (trasig) object". a)<br />
bjatig b) skratig c) tratig?" Skr- is the expected cluster but tratig is very<br />
similar to trasig.<br />
219
Test 2.a is a free choice test which goes from expression to mean<strong>in</strong>g <strong>in</strong><br />
order to test the understand<strong>in</strong>g of presumptive sound symbolic clusters,<br />
e.g. "What would be a good mean<strong>in</strong>g for the word fnotig?"<br />
In test 2.b, a forced choice test (which also goes from expression to<br />
mean<strong>in</strong>g), each nonsense word shall be matched with a certa<strong>in</strong> number of<br />
recurr<strong>in</strong>g mean<strong>in</strong>gs of a certa<strong>in</strong> abstractness, e.g. "someth<strong>in</strong>g which is<br />
soft". More specifically, for each nonsense word tested there will be 3<br />
alternative mean<strong>in</strong>gs, only 1 of which is possible accord<strong>in</strong>g to the<br />
previous analysis. An example of this is "What do you th<strong>in</strong>k slatig means:<br />
a) somebody who is silly b) someth<strong>in</strong>g that is dry c) someth<strong>in</strong>g which is<br />
unpleasant?"<br />
Test 3 is a match<strong>in</strong>g test where the subject has to choose between two<br />
different mean<strong>in</strong>gs and two different expressions. An example of this is:<br />
Which word best describes a th<strong>in</strong>g that is wet and which word best<br />
describes a th<strong>in</strong>g that is dry: fnottig or skvottig?"<br />
For the tests 2a and b, nonsense words were constructed from <strong>in</strong>itial<br />
consonant clusters that are commonly used for onomatopoeia and other<br />
sound symbolism, accord<strong>in</strong>g to the lexical analysis <strong>in</strong> chapter 4. To avoid<br />
motivated mean<strong>in</strong>g appear<strong>in</strong>g <strong>in</strong> the non-<strong>in</strong>itial part of the neologism the<br />
words constructed are short. Also the end<strong>in</strong>gs (-t or -t:) of the neologisms<br />
were checked <strong>in</strong> Svensk baklängesordbok (1981) <strong>in</strong> order to a v o i d<br />
motivated mean<strong>in</strong>gs <strong>in</strong> the end<strong>in</strong>gs.<br />
A complication with us<strong>in</strong>g monosyllables is that this syllable structure,<br />
when repeated <strong>in</strong> a list, gives an impression of 'beat' or 'sound imitation',<br />
especially when it ends <strong>in</strong> a long consonant. For this reason the test words<br />
do not have a long consonant (except <strong>in</strong> a few test cases). Furthermore,<br />
the test words have been added with the semantically neutral end<strong>in</strong>g -ig.<br />
A problem with this might be, however, that <strong>in</strong> us<strong>in</strong>g longer words, more<br />
consonants and vowels will be <strong>in</strong>volved, and it is more difficult to keep<br />
control over entire test words.<br />
220
Aside from construct<strong>in</strong>g monosyllabic words (with an added end<strong>in</strong>g -ig),<br />
another simplification is that <strong>in</strong> natural sound symbolic words, the<br />
semantic feature, e. g. 'pejorative', of an <strong>in</strong>itial consonant cluster is often<br />
repeated <strong>in</strong> the f<strong>in</strong>al consonant cluster (cf. 5.6).<br />
As was shown <strong>in</strong> diagram 4.5, the 13 most common semantic features of<br />
Swedish <strong>in</strong>itial clusters are, <strong>in</strong> descend<strong>in</strong>g order: 'pejorative', 'sound',<br />
'long th<strong>in</strong> form', 'quick or strong movement', 'wetness', 'talk<strong>in</strong>g', 'light',<br />
'dim<strong>in</strong>utive', 'round form', 'walk<strong>in</strong>g', 'destruction', 'w<strong>in</strong>d<strong>in</strong>g form', 'bad<br />
mood'. These semantic features were used <strong>in</strong> the tests except for 'sound',<br />
'dim<strong>in</strong>utive', 'round form' and 'way of walk<strong>in</strong>g' 1. Also some more<br />
<strong>in</strong>frequent features, shown <strong>in</strong> the list below, were used. The formulations<br />
on the test sheets correspond to the semantic features above <strong>in</strong> the way<br />
shown <strong>in</strong> table 7.1.<br />
Table 7.1 The semantic features correspond<strong>in</strong>g to the formulations on the<br />
test sheets.<br />
formulations semantic features<br />
löjlig (silly) pejorative<br />
fånig (silly) pejorative<br />
obehaglig (unpleasant) pejorative<br />
dum och klumpig (stupid and<br />
clumsy)<br />
pejorative<br />
pratar dumheter (talks nonsense) pejorative, talk<strong>in</strong>g<br />
rak (straight) long th<strong>in</strong> form<br />
smal form (narrow form) long th<strong>in</strong> form<br />
rörelse fram och tillbaka (movement<br />
back and forth)<br />
quick or strong movement<br />
plötslig rörelse (sudden movement) quick or strong movement<br />
(mycket) blöt ((very) wet) wetness<br />
1 '<strong>Sound</strong>' was not tested s<strong>in</strong>ce it is usually easier for speakers to produce sound imitative<br />
words. 'Dim<strong>in</strong>utive', 'round form', and 'walk<strong>in</strong>g' were not <strong>in</strong>cluded because the f<strong>in</strong>al lexical<br />
analysis was not completed at the time the test was constructed, and therefore these categories<br />
were not <strong>in</strong>cluded. The category of 'form' is, however, well represented <strong>in</strong> the study.<br />
221
hur en människa kan låta (how a<br />
person can sound)<br />
talk<strong>in</strong>g<br />
starkt lysande (sh<strong>in</strong><strong>in</strong>g brightly) light<br />
trasig (broken) destruction<br />
sl<strong>in</strong>grig form (w<strong>in</strong>d<strong>in</strong>g form) w<strong>in</strong>d<strong>in</strong>g form<br />
arg (angry) bad mood<br />
på dåligt humör (bad mood) bad mood<br />
ihålig (hollow) hollow form<br />
pekar åt olika håll (po<strong>in</strong>t<strong>in</strong>g <strong>in</strong><br />
different directions)<br />
separation<br />
går ihop (comes together) putt<strong>in</strong>g together<br />
(mycket) torr ((very) dry) dryness<br />
tydlig hård ytstruktur (pronounced<br />
hard surface structure)<br />
rough surface structure<br />
följsam ytstruktur (adaptable surface<br />
structure)<br />
soft consistency<br />
sitter fast (sticks) adhesion<br />
vidhäftande (stick<strong>in</strong>g) adhesion<br />
hårt slag (hard beat) beat<br />
The most common consonant clusters, <strong>in</strong> absolute numbers and<br />
percentally, are shown <strong>in</strong> diagrams 4.1 and 4.4. In this test all of these<br />
clusters except nj- are used. The tests <strong>in</strong>cludes the clusters: kl-, gr-, vr-<br />
, spj-, str-, sp-, mj-, kv-, skv-, skr-, kr-, bl-, sm-, gr-, st-, tr-<br />
, bl-, br-, spl-, gn-, spr-, fr-, sk-, kn-, gl-, sv-, dr-, pl-, sl-,<br />
sn, fl-, bj-, pj-, fn- and fj-.<br />
The test words are grotig, gratig, vrotig, vratig, spjotig, spjatig, stratig,<br />
spatig, mjatig, kvatig, skvatig, skratig, bjetig, bjatig, kretig, kratig, bletig,<br />
blatig, smatig, statig, tratig, platig, fnatig, krotig, strotig, trotig, pjotig,<br />
brotig, splotig, snitig, gnitig, gritig, klatig, spratig, smitig, gritig, spitig,<br />
platig, snatig, smatig, dratig, blattig, mjattig, flattig, snattig, smattig,<br />
drattig, etc (see Appendix 3).<br />
222
Test procedure<br />
The test questions were written four to a page and the pages were stapled<br />
together. The test questions were mixed <strong>in</strong> order to keep the test words<br />
from <strong>in</strong>fluenc<strong>in</strong>g each other (e.g. no two questions about 'light' were put<br />
on the same page). The subjects were told not to go back and check and<br />
were not given time for that either. Half of the subjects were tested <strong>in</strong> the<br />
order: 1 (a)(b) -2(a)(b) -3 and the other half <strong>in</strong> the order 2 (a)(b) -1(a)(b)<br />
-3 (see Appendix 3).<br />
Test (a) which is free production, is the most difficult, test (b) which is<br />
multiple-choice, is less difficult, and test 3, which is match<strong>in</strong>g, is the<br />
easiest. The largest number of presumed answers are expected for test 3<br />
and the least number of presumed (or consistent) answers are expected for<br />
1a and 2a.<br />
The subjects were l<strong>in</strong>guistics students at the beg<strong>in</strong>ner's level, with<br />
Swedish as a first language, males and females, ages 21–57. There were<br />
15 <strong>in</strong>formants. 14 of them took the tests from mean<strong>in</strong>g to expression and<br />
15 <strong>in</strong>formants took the tests from expression to mean<strong>in</strong>g. Thirteen of<br />
them took the match<strong>in</strong>g test.<br />
7.1.1 Forced choice for production – from<br />
mean<strong>in</strong>g to phonological (graphic) form<br />
Of the 39 test questions where a mean<strong>in</strong>g was presented along with three<br />
alternatives for phonetic (graphic) form, one of which was possible<br />
accord<strong>in</strong>g to the model, 28 test questions showed a majority for expected<br />
answers, 4 had a shared majority for the expected and an unexpected<br />
answer and 7 showed a majority for an unexpected answer. Sometimes<br />
the majority was overwhelm<strong>in</strong>g, sometimes not, see table 7.2.<br />
The follow<strong>in</strong>g 19 (out of 39) test questions received the best results: no.<br />
37, 38, 39, 35, 30, 31, 32, 29, 21, 22, 23, 24, 13, 16, 9, 10, 5, 7 and 8<br />
(cf. Appendix 3). These questions concerned the features and words<br />
shown <strong>in</strong> table 7.2.<br />
223
Table 7.2 The semantic features of the 19 test questions that received the best<br />
results <strong>in</strong> the forced choice test of choos<strong>in</strong>g phonological form from<br />
different semantic features. Words are presented <strong>in</strong> order from best match<br />
with expected to worst. 14 <strong>in</strong>formants.<br />
semantic feature word ratio of expected<br />
answers<br />
'pejorative' fjotig 14/14<br />
'dryness' fnotig 13/14<br />
'long th<strong>in</strong> form' stratig 13/14<br />
'wetness' spatig 13/14<br />
'bad mood' vratig 13/14<br />
'pejorative' pjotig 13/14<br />
'wetness' platig 12/14<br />
'pejorative' bjatig 12/14<br />
'quick or strong<br />
movement'<br />
flattig 12/14<br />
'adhesion' klatig 12/14<br />
'bad mood' tratig 12/14<br />
'separation' spratig 11/14<br />
'hollow form' gratig 11/14<br />
'wetness' svotig 11/14<br />
'quick or strong<br />
movement'<br />
skottig 11/14<br />
'pejorative' drotig 11/14<br />
'hollow form' grotig 10/14<br />
'w<strong>in</strong>d<strong>in</strong>g form' krotig 9/14<br />
'talk<strong>in</strong>g' snatig 9/14<br />
Seen <strong>in</strong> relation to the number of <strong>in</strong>stances of each feature there were <strong>in</strong><br />
the test, the semantic features shown <strong>in</strong> table 7.3 are the most successfully<br />
<strong>in</strong>terpreted:<br />
224
Table 7.3 The most successfully <strong>in</strong>terpreted semantic features of table 7.2.,<br />
14 <strong>in</strong>formants.<br />
'wetness:' 3/3 2 (100%)<br />
'hollow form': 2/2 (100%)<br />
'separation': 1/1(100%)<br />
'dryness': 1/1(100%)<br />
'pejorative': 4/6 (67%)<br />
'bad mood': 2/3 (67%)<br />
'quick or strong movement': 2/4 (50%)<br />
'w<strong>in</strong>d<strong>in</strong>g form': 1/2 (50%)<br />
'talk<strong>in</strong>g': 1/2 (50%)<br />
'adhesion': 1/2 (50%)<br />
'long th<strong>in</strong> form': 1/3 (33%)<br />
The questions that did not give the expected answers perta<strong>in</strong>ed to the<br />
features and words <strong>in</strong> table 7.4 (for a majority of subjects).<br />
Table 7.4 Semantic features for neologisms that did not give the expected<br />
answers. The substitutions of words (with frequencies) are shown <strong>in</strong> the<br />
right column. 14 <strong>in</strong>formants.<br />
'w<strong>in</strong>d<strong>in</strong>g form' stratig (6) <strong>in</strong>stead of kratig (4)<br />
'light' kratig (7) <strong>in</strong>stead of blatig (3)<br />
'light' snitig (8) <strong>in</strong>stead of gnitig (2)<br />
'quick or strong movement' mjatig (6) <strong>in</strong>stead of flatig (3)<br />
'light' snotig (9) <strong>in</strong>stead of gnotig (2)<br />
'light' kretig (6) <strong>in</strong>stead of gletig (4)<br />
It is clear that the 'light' feature gets least expected answers. The clusters<br />
tested are the well known bl-, gl- and gn- clusters, but obviously<br />
someth<strong>in</strong>g more is required to get a light-associative effect. The 'light'<br />
words might also belong to a closed lexical class which would imply that<br />
gl- , gn-, bl- are not productive, unlike phonesthemes such as kl- or<br />
2 i.e. 3 of the 3 wetness words were identified by most subjects.<br />
225
pj-. Another possible explanation, <strong>in</strong> light of the results <strong>in</strong> 5.4, is that the<br />
vowels have to be taken <strong>in</strong>to account. The lexical analysis showed that<br />
front vowels, and especially i, are much more common <strong>in</strong> root<br />
morphemes with the features 'light' or 'gaze'. Two of the constructed<br />
words <strong>in</strong> the test, gnitig and gletig, had front vowels (i and e) but were<br />
nevertheless not <strong>in</strong>terpreted accord<strong>in</strong>g to expectations.<br />
The feature most successfully coded <strong>in</strong> phonesthemes is 'wetness', for<br />
which 3 out of 3 words where identified by most subjects, 'hollow form'<br />
(2 out of 2), 'separation' (1 out of 1) and 'dryness' (1 out of 1), cf. table<br />
7.3. The feature that is the least successfully <strong>in</strong>terpreted is 'light'; none of<br />
the 4 test words were <strong>in</strong>terpreted accord<strong>in</strong>g to expectations, cf. table 7.4.<br />
S<strong>in</strong>ce some of the vowels <strong>in</strong> the constructed words were <strong>in</strong> accordance<br />
with expectations, the non-expected results of the 'light' words can not be<br />
expla<strong>in</strong>ed with reference to the importance of vowels. It seems rather that<br />
the 'light' phonesthemes are not productive. One explanation for why<br />
they are not productive could have to do with the fact that 'light' is the<br />
only category which is metaphorically connected with 'sound', i .e. by<br />
sense analogy. Many other categories are connected metonymically with<br />
sound, i.e. they occur simultaneously, e.g. 'movement', 'surface<br />
structure', 'wetness', 'form', or are more abstract (cf. 2.5, 2.6).<br />
7.1.2 Forced choice for understand<strong>in</strong>g - from<br />
phonological form to mean<strong>in</strong>g<br />
Out of the 38 test questions where a constructed word was presented<br />
along with three mean<strong>in</strong>g alternatives, one of which was possible (or most<br />
possible 3) accord<strong>in</strong>g to the model of 2.6.2 and the analysis of chapter 4,<br />
29 showed a majority for expected answers, 1 had a shared majority for<br />
expected and unexpected answers and 8 had a majority for an unexpected<br />
answer. Sometimes the majority was overwhelm<strong>in</strong>g, sometimes not. The<br />
test questions that had the best results were the follow<strong>in</strong>g: 24, 32, 26, 29,<br />
21, 23, 25, 18, 10, 12, 5, 3, 4, 14. These questions concerned the words<br />
and features presented <strong>in</strong> table 7.5.<br />
3 The only exception is fnotig (question 23) where 'rough surface structure' is slightly more<br />
expectable than 'dryness'. However, as can be seen <strong>in</strong> table 7.5, fnotig is <strong>in</strong>terpreted as<br />
'dryness' by 15 out of 15 subjects.<br />
226
Table 7.5 The words <strong>in</strong> the test questions that received the best results <strong>in</strong> the<br />
forced choice test of choos<strong>in</strong>g mean<strong>in</strong>gs for different constructed words. 15<br />
<strong>in</strong>formants.<br />
word semantic feature ratio of expected<br />
answers<br />
grotig 'hollow form' 15/15<br />
fnotig 'dryness' 15/15<br />
snattig 'talk<strong>in</strong>g' 15/15<br />
pjotig 'pejorative' 15/15<br />
skottig 'quick or strong<br />
movement'<br />
14/15<br />
spratig 'separation' 14/15<br />
grotig 'bad mood' 11/15<br />
fjotig 'pejorative' 13/15<br />
vratig 'bad mood' 13/15<br />
blatig 'pejorative' 12/15<br />
spitig 'long th<strong>in</strong> form' 12/15<br />
snatig 'talk<strong>in</strong>g' 11/15<br />
skratig 'rough surface<br />
structure'<br />
10/15<br />
flattig 'quick or strong<br />
movement'<br />
10/15<br />
The semantic features which were the most successful <strong>in</strong> the forced choice<br />
test of choos<strong>in</strong>g mean<strong>in</strong>gs for different constructed words were 'hollow<br />
form' (for gr-), 'dryness' (for fn-), 'talk<strong>in</strong>g' (for sn-), and 'pejorative'<br />
(for pj-)<br />
The questions that did not give the expected results concerned the forms<br />
and mean<strong>in</strong>gs <strong>in</strong> table 7.6 (for a majority of speakers).<br />
227
Table 7.6 Words that did not give the expected answers. The substitutions of<br />
features. 15 <strong>in</strong>formants.<br />
word substitutions of<br />
features<br />
gnotig 'talk<strong>in</strong>g' <strong>in</strong>stead of<br />
'light'<br />
flatig 'long th<strong>in</strong> form' <strong>in</strong>stead<br />
of 'quick or strong<br />
movement'<br />
gletig 'separation' <strong>in</strong>stead of<br />
'light'<br />
gnitig 'stick<strong>in</strong>g' <strong>in</strong>stead of<br />
'light'<br />
glatig 'pejorative' <strong>in</strong>stead of<br />
'light'<br />
smatig 'pejorative' <strong>in</strong>stead of<br />
'beat'<br />
blatig 'hollow form' <strong>in</strong>stead<br />
of 'light'<br />
228<br />
comment<br />
This is <strong>in</strong> fact more<br />
expectable accord<strong>in</strong>g to<br />
the analysis <strong>in</strong> 4.5<br />
This is also expectable.<br />
There might be an<br />
<strong>in</strong>fluence from the<br />
word flat (flat).<br />
In this case a similar<br />
word gnatig has no<br />
<strong>in</strong>fluence. The third,<br />
but not chosen,<br />
alternative was actually<br />
'talk<strong>in</strong>g' which is also<br />
possible accord<strong>in</strong>g to<br />
the model<br />
'Pejorative' is <strong>in</strong> fact<br />
also expectable<br />
accord<strong>in</strong>g to the model<br />
'Pejorative' is <strong>in</strong> fact<br />
also expectable<br />
accord<strong>in</strong>g to the model<br />
The semantic feature which is the least successful is 'light' (for gn-, gland<br />
bl-). Even where the vowel was front, as <strong>in</strong> gletig and gnitig, the<br />
result was not accord<strong>in</strong>g to expectations.<br />
Summary of forced choice tests<br />
In the first test, 28 out of 39 test questions showed a majority for<br />
expected answers accord<strong>in</strong>g to the model, and 19 of these received good<br />
results, i.e. at least 64% (9/14) expected answers (table 7.2). In the second<br />
test, 29 out of 38 test questions showed a majority for expected answers,
and 13 of these received good results, i.e. at least 67% (10/15) expected<br />
answers.<br />
In both tests the semantic feature which was the least successfully<br />
<strong>in</strong>terpreted or produced was 'light'. Even when the vowel was front,<br />
accord<strong>in</strong>g to expectations, the constructed words were not <strong>in</strong>terpreted or<br />
produced accord<strong>in</strong>g to expectations. It thus seems as if the phonesthemes<br />
connected with 'light' are not productive. In the test of mean<strong>in</strong>g to<br />
phonological form, 28 of the 39 test questions showed a majority for<br />
expected answers. The six phonesthemes that were the most successfully<br />
coded were 'pejorative': fj-, 'dryness': fn-, long th<strong>in</strong> form': str-,<br />
'wetness': sp-, 'bad mood': vr-, 'pejorative': pj-. (The features most<br />
successfully coded totally were 'wetness', 'hollow form', 'separation' and<br />
'dryness'.) In the test from phonological form to mean<strong>in</strong>g, 29 of the 38<br />
test questions showed a majority for expected answers. The six<br />
phonesthemes most successfully <strong>in</strong>terpreted were 'hollow form': gr-,<br />
'dryness': fn-, 'talk<strong>in</strong>g': sn-, 'pejorative': pj-, 'quick or strong<br />
movement': sk- and 'separation': spr-. For both the <strong>in</strong>terpretation and<br />
production tests, the most successful cases were 'dryness': fn- and<br />
'pejorative': pj-. Both of these are lexically low frequency clusters that<br />
are sound symbolic to a very high degree.<br />
7.1.3 Free production test from constructed words<br />
to mean<strong>in</strong>gs<br />
The free production test from constructed forms to mean<strong>in</strong>gs asked the<br />
question "What do you th<strong>in</strong>k would be a good mean<strong>in</strong>g for ...?" This gave<br />
the results shown <strong>in</strong> table 7.7.<br />
229
Table 7.7 Free <strong>in</strong>terpretations by 15 subjects of 6 constructed words.<br />
skvatig fnotig vratig krötig pjotig skratig<br />
härsken tråkig vrensk stöddig tokig skarp<br />
(rancid) (bor<strong>in</strong>g) (refractory) (stuck up) (mad) (sharp)<br />
blöt obeslutsam tjatig trög pjoskig glad<br />
(wet) (irresolute) (nagg<strong>in</strong>g) (sluggish) (mawkish) (happy)<br />
tröttsam<br />
(tir<strong>in</strong>g)<br />
- sned<br />
(crooked)<br />
ihopklumpad<br />
(lumped<br />
together)<br />
barnslig<br />
(childish)<br />
bristfällig<br />
(defective)<br />
blöt<br />
(wet)<br />
torr och<br />
fladdrig<br />
(dry and<br />
flapp<strong>in</strong>g)<br />
vriden<br />
(twisted)<br />
trött och lite<br />
sjuk<br />
(tired and a<br />
little ill)<br />
fånig<br />
(silly)<br />
snurrig<br />
(giddy,<br />
crazy)<br />
blöt fnasig grov tillgjord petig dålig<br />
(wet) (chapped) (coarse) (affected) (f<strong>in</strong>ical) (bad)<br />
kaxig krokig tjatig knotig dum glad<br />
(cocky) (crooked) (nagg<strong>in</strong>g) (bony) (stupid) (happy)<br />
schlager- pustande skrytsam sluddrande gnetig trög<br />
musik<br />
(popular<br />
song)<br />
(pant<strong>in</strong>g) (boastful) (slurr<strong>in</strong>g) (crabbed) (sluggish)<br />
flamsig löjlig pratig rörig pjåskig kantig<br />
(silly) (ridiculous) (chatty) (messy) (mawkish) (angular)<br />
galen gammal arg mosig ojämn knyckig<br />
(crazy) (old) (angry) (fuddled) (uneven) (jerky)<br />
blöt knäpp vrängd krokig plaskig smal<br />
(wet) (stupid) (turned<br />
<strong>in</strong>side out)<br />
(crooked) (splashy) (narrow)<br />
något kantigt<br />
och blött<br />
(sth angular<br />
and wet)<br />
något svårt<br />
(sth difficult)<br />
pratsam<br />
tråkig<br />
(talkative<br />
bor<strong>in</strong>g)<br />
<strong>in</strong>krökt<br />
(focused on<br />
one´s own<br />
problems)<br />
besvärlig<br />
(troublesome)<br />
vissen<br />
(withered)<br />
skvalande skrumpen tjatig urgröpt fånig full med<br />
(pour<strong>in</strong>g) (shrunk) (nagg<strong>in</strong>g) (hollowed (ridiculous) revor<br />
out)<br />
(full of rips)<br />
kantig eller<br />
hård<br />
(angular or<br />
hard)<br />
knölig<br />
(knobbly)<br />
arg<br />
(angry)<br />
berusad<br />
(drunk)<br />
liten<br />
rar<br />
(small<br />
sweet)<br />
trasig<br />
(ragged)<br />
230
skratta<br />
(laugh)<br />
halvdålig* 4<br />
(half bad)<br />
frysa<br />
(freeze)<br />
knäpp*<br />
tokrolig*<br />
(crazy<br />
funny)<br />
vrida<br />
(wr<strong>in</strong>g)<br />
halvnerriven<br />
vägreklam*<br />
(half torn<br />
down<br />
advertizement<br />
by the<br />
road side)<br />
krånglig<br />
(trouble-<br />
231<br />
some)<br />
försupen*<br />
(sottish)<br />
passande<br />
(suitable)<br />
löjlig*<br />
liten*<br />
(ridiculous<br />
small)<br />
skrattande<br />
(laugh<strong>in</strong>g)<br />
en härjad<br />
smal kv<strong>in</strong>na*<br />
(a worn and<br />
haggard th<strong>in</strong><br />
woman)<br />
1. skvatig gives 6/15 expected answers, i.e., words with a semantic<br />
feature 'wetness'. Of the other mean<strong>in</strong>gs only one gives an answer with a<br />
feature 'wetness' – pjotig<br />
2. fnotig gets the expected 'pejorative' <strong>in</strong> 5/14 (perhaps more depend<strong>in</strong>g<br />
on which answers should be classified as 'pejorative'). It also gets the<br />
expected 'dry' <strong>in</strong> 2/14. (cf. diagram 4.20).<br />
3. vratig gives 3/15 clear answers conta<strong>in</strong><strong>in</strong>g the feature 'bad mood' and<br />
some other answers which are less clear.<br />
4. krötig does not give words with the expected semantic feature 'w<strong>in</strong>d<strong>in</strong>g<br />
form' but rather po<strong>in</strong>t to some sort of 'pejorative', which is the second<br />
expected feature for kr- (cf. diagram 4.19).<br />
5. pjotig, where the expected semantic feature 'pejorative' gives 10/15.<br />
6. skratig, where expected semantic feature 'destruction' gives 7/15.<br />
'Pejorative' has a higher percentage of skr- than 'destruction', but<br />
'destruction' is typical for skr-, cf. table 4.16.<br />
The nonsense word that is most successful is pjotig (pejorative). Pj- is<br />
percentally the fourth most sound symbolic cluster (cf. diagram 4.4) and<br />
is ma<strong>in</strong>ly 'pejorative' (cf. diagram 4.23). The second most successful<br />
word is skratig, which only belongs to the eleventh most sound symbolic<br />
cluster (percentally), skr-. However, it is a typical cluster for<br />
'destruction'. In third place comes skvatig (wetness).<br />
In other words, no constructed word is <strong>in</strong>terpreted as expected by all<br />
subjects, but all of the constructed words, except krötig (w<strong>in</strong>d<strong>in</strong>g form),<br />
are <strong>in</strong>terpreted correctly by some subjects. Krötig is, however, given a<br />
4 The * marks the answers from the subject who deviated most <strong>in</strong> the "match<strong>in</strong>g test of<br />
nonsense words and mean<strong>in</strong>gs" (see 7.1.5).
second best <strong>in</strong>terpretation, 'pejorative'. The success rank<strong>in</strong>g of these six<br />
phonesthemes are thus, <strong>in</strong> order from best match to worst: 'pejorative'<br />
(for pj-), 'destruction' (for skr-), 'wetness' (for skv-), 'pejorative' (for<br />
fn-), 'bad mood' (for vr-), 'w<strong>in</strong>d<strong>in</strong>g form' (for kr-).<br />
There are examples both of where the associations have gone to a word<br />
with a similar phonetic form, and examples of where the phonestheme has<br />
an <strong>in</strong>fluence <strong>in</strong> spite of a similar sound<strong>in</strong>g word 5. An example of the first<br />
case is krötig where the associations often seem to go to kröka (dr<strong>in</strong>k<br />
alcohol): slang word): sluddrande, berusad, försupen. Skratig sometimes<br />
gives associations to skratt (laughter): glad, skrattande. There are,<br />
however, many answers under skratig which conta<strong>in</strong> the mean<strong>in</strong>g<br />
component of 'destruction': bristfällig, lite sjuk, dålig, vissen, full med<br />
revor, trasigt, en härjad smal kv<strong>in</strong>na.<br />
The conclusion is that these nonsense words are <strong>in</strong>terpreted <strong>in</strong> accordance<br />
with expectations by some subjects, but not all. The nonsense words have<br />
a phonesthemic mean<strong>in</strong>g potential that is used by some subjects <strong>in</strong> the test<br />
situation. The phonestheme most easily <strong>in</strong>terpreted was 'pejorative': pj-.<br />
Explanations for this might have to do with the fact that pj- is a lexically<br />
low frequent highly sound symbolic cluster. 'Pejorative' is also the most<br />
frequent semantic feature accord<strong>in</strong>g to the lexical analysis. The cluster<br />
most difficult to <strong>in</strong>terpret was 'w<strong>in</strong>d<strong>in</strong>g form': kr-. Kr- is lexically high<br />
frequent but also sound symbolic to quite a high degree, however not<br />
only with the feature 'w<strong>in</strong>d<strong>in</strong>g form'. It seems as lexically low frequent<br />
highly sound symbolic clusters are easy to access.<br />
7.1.4 Free production from mean<strong>in</strong>g to constructed<br />
word.<br />
The question "Make up a short word for somebody or someth<strong>in</strong>g which is<br />
(has) ...?", where subjects were supposed to <strong>in</strong>vent an expression for one<br />
of the mean<strong>in</strong>gs 'silly', 'w<strong>in</strong>d<strong>in</strong>g', 'angry', 'dry', 'wet', and 'rough<br />
surface', gave the results shown <strong>in</strong> table 7.8.<br />
5 There are, as is well known, many types of relations <strong>in</strong> the lexicon, and thus relations<br />
between words, as well as between phonesthemes, cf. e.g. Garman (1990).<br />
232
Table 7.8 Words produced by 14 subjects for 6 different mean<strong>in</strong>gs. The<br />
words are non-words. Most of them follow the Swedish phonotactic rules.<br />
'silly' 'w<strong>in</strong>d<strong>in</strong>g' 'angry' 'dry' 'wet' 'rough<br />
surface'<br />
smurk slirv gurp lirv slish pritt<br />
spjal islig kral spri subl flarb<br />
flong plyr orn krasp svåsk donk<br />
fjän ril vrag fnat plat raster<br />
koos pis faaby kirl trippp hitt<br />
flutt sjl<strong>in</strong>g grol fnus svurp gilb<br />
floppig siloln börr kritto slasli teppig<br />
pjöl krel vram fnöl svomm knupp<br />
fjutt kril burr krasp mollo skrak<br />
krumpig vr<strong>in</strong>lig trossk prusskig schjaflig dank<br />
fnölp tirori rark fnuskig splass n<strong>in</strong><br />
- slio furn kln splurr klik<br />
loup* 6 evans* hram* srrats* paupe* kovo*<br />
knork vrom dramm frok blu tlak<br />
When it comes to the production of forms, we are not restricted to <strong>in</strong>itial<br />
clusters; the subjects have also used vowel qualities, reduplication and<br />
f<strong>in</strong>al clusters to express the different mean<strong>in</strong>gs. However, <strong>in</strong>itial clusters<br />
are much more common <strong>in</strong> these neologisms. An analysis of the<br />
neologisms produced follows below.<br />
Semantic feature 'pejorative'<br />
For the mean<strong>in</strong>g 'silly' (semantic feature: 'pejorative') the produced<br />
<strong>in</strong>itial clusters <strong>in</strong> the test are, <strong>in</strong> frequency order: fl- (3), fj- (2), and then<br />
one <strong>in</strong>stance each of pj-, fn-, kn-, kr-, sm- and spj-. These results can<br />
be compared with diagrams 4.6 and 4.7 where we see that pj- fn- and fjare<br />
very frequent percentally, while kn- and kr- are very frequent <strong>in</strong><br />
absolute numbers. Thus both percental and absolute frequency of<br />
phonesthemes <strong>in</strong> the lexical analysis of chapter 4 correlate with the<br />
6 The * marks the answers from the subject who deviated most <strong>in</strong> the "match<strong>in</strong>g test of<br />
nonsense words and mean<strong>in</strong>gs" (see 7.1.5).<br />
233
clusters that the subjects use <strong>in</strong> free production of sound symbolic words.<br />
The most frequently produced cluster fl- is quite frequent <strong>in</strong> absolute<br />
numbers but not so frequent percentally. Only one of the produced words<br />
conta<strong>in</strong>s a cluster which is not pejorative at all, namely spj-.<br />
The f<strong>in</strong>al consonant clusters produced are, <strong>in</strong> frequency order: -rk (2),<br />
and one <strong>in</strong>stance each of -mp and -lp. These clusters can be compared<br />
with tables 5.2 and 5.3 and diagrams 5.1 and 5.2. The tables show that -<br />
mp can be pejorative (but is usually not), and that -rk and -lp are not<br />
sound symbolic at all. The conclusion here must be that the f<strong>in</strong>al<br />
consonant clusters are of less importance than the <strong>in</strong>itial ones <strong>in</strong><br />
produc<strong>in</strong>g new pejorative words.<br />
Semantic feature 'w<strong>in</strong>d<strong>in</strong>g form'<br />
For the semantic feature 'w<strong>in</strong>d<strong>in</strong>g form', the produced <strong>in</strong>itial clusters <strong>in</strong><br />
the test are, with two <strong>in</strong>stances each: kr-, vr-, sl- pl- and sjl-. These can<br />
be compared with table 4.16. The clusters that have the feature 'w<strong>in</strong>d<strong>in</strong>g<br />
form' are: kr- (typical) and sn- (possible). In other words, kr- is<br />
expected. The cluster sjl- breaks the phonotactic pattern.<br />
The f<strong>in</strong>al consonant clusters produced are, with one <strong>in</strong>stance each: -rv, -<br />
ln and -ns. Of these three, -rv and -ns can be sound symbolic accord<strong>in</strong>g<br />
to the analysis <strong>in</strong> chapter 5 (cf. tables 5.2 and 5.3). There are however, <strong>in</strong><br />
this analysis, no f<strong>in</strong>al consonant clusters with the feature 'w<strong>in</strong>d<strong>in</strong>g form'<br />
(only one word kr<strong>in</strong>gla).<br />
Instead, there seems to be an iconicity <strong>in</strong> any position <strong>in</strong> the word for the<br />
mean<strong>in</strong>g 'w<strong>in</strong>d<strong>in</strong>g form' reflected e.g. <strong>in</strong> a contrast between s and l, <strong>in</strong><br />
the words slirv, islig, sjl<strong>in</strong>g, siloln, slio. As seen earlier (cf. for<br />
example diagram 4.1), sl- is also the most sound symbolic cluster <strong>in</strong><br />
absolute numbers. This issue is left for further research.<br />
The vowels are ma<strong>in</strong>ly i, and there is a dom<strong>in</strong>ance for front, closed or<br />
half closed vowels (with high F2).<br />
234
Semantic feature 'bad mood'<br />
For the mean<strong>in</strong>g 'angry' (semantic feature: 'bad mood') the <strong>in</strong>itial clusters<br />
produced <strong>in</strong> the test are, <strong>in</strong> order of frequency: vr- (2) and one <strong>in</strong>stance<br />
of each of gr-, kr-, tr-, dr- and hr-. The most frequent ones, vr-, as trand<br />
gr- (and only these clusters) also have the semantic feature 'bad<br />
mood' accord<strong>in</strong>g to the earlier analysis. All the produced clusters except<br />
one have an r, which is what the conventional ones have <strong>in</strong> common. The<br />
last one, hr- breaks the phonotactic pattern. The phonestheme 'bad mood'<br />
seems to have the expression obstruent + r.<br />
The f<strong>in</strong>al consonant clusters produced are: -rn (2) (it is unclear if the<br />
subjects pronounced this as a cluster or as a retroflex) and one <strong>in</strong>stance<br />
each of -rp, -rk, and -ssk. Obviously r dom<strong>in</strong>ates also <strong>in</strong> the f<strong>in</strong>al<br />
clusters, the only exception be<strong>in</strong>g -ssk. Accord<strong>in</strong>g to the analysis <strong>in</strong><br />
chapter 5, there are no conventional f<strong>in</strong>al clusters with the semantic<br />
feature 'bad mood', so the <strong>in</strong>itial cluster phonemes seem to have been<br />
transferred to the f<strong>in</strong>al position.<br />
There are no front, closed or half closed vowels <strong>in</strong> these words but there<br />
is a preference for vowels with lower F2.<br />
Semantic feature 'dryness'<br />
For the semantic feature 'dryness' the <strong>in</strong>itial clusters produced <strong>in</strong> the test<br />
are, <strong>in</strong> order of frequency: fn- (4), kr- (3) and one <strong>in</strong>stance each of fr-,<br />
pr-, spr-, srr-, kln (a whole word). The most frequent one fn- has the<br />
semantic feature 'dryness' accord<strong>in</strong>g to the earlier analysis, and fn- is the<br />
only cluster with the feature 'dryness'. The rema<strong>in</strong><strong>in</strong>g ones consequently<br />
do not have the feature 'dryness'; srr- and kln break the phonotactic<br />
pattern.<br />
The f<strong>in</strong>al consonant clusters produced are: -sp, -sk, -rv, -ts and -dt.<br />
'Dryness' is not a feature of f<strong>in</strong>al clusters accord<strong>in</strong>g to the analysis of<br />
chapter 5. The cluster -dt breaks the phonotactics. Aga<strong>in</strong>, the subjects<br />
seem to encode the semantic feature <strong>in</strong> the <strong>in</strong>itial cluster.<br />
235
For 'dryness' there are more clusters that break the phonotactic pattern<br />
than for the other semantic features. Perhaps this has to do with the fact<br />
that 'dryness' is a very <strong>in</strong>frequent sound symbolic feature, e.g. <strong>in</strong><br />
comparison with wetness.<br />
Semantic feature 'wetness'<br />
For the semantic feature 'wetness', the <strong>in</strong>itial clusters produced <strong>in</strong> the test<br />
are, <strong>in</strong> order of frequency: sl- (2), sv- (2), spl- (2) and one <strong>in</strong>stance each<br />
of pl-, tr-, schj- [Sj], bl- and bw-. The clusters with the feature<br />
'wetness' are two of the most common ones, sl- and sv- (but not spl-).<br />
The rema<strong>in</strong><strong>in</strong>g ones, pl- and bl-, also have the feature 'wetness', but not<br />
tr-. Schj- [Sj] breaks the phonotactic pattern and bw- has the sound [w]<br />
which is not even a phoneme <strong>in</strong> Swedish. So, for this feature, four<br />
expected clusters are used, two unexpected ones and two unconventional.<br />
The f<strong>in</strong>al consonant clusters produced are: -bl, -sk, -rp, -sl(i) and -fl.<br />
Of these, the only f<strong>in</strong>al clusters for 'wetness', accord<strong>in</strong>g to the analysis of<br />
chapter 5, is -sk. -bl is not phonotactically possible <strong>in</strong> the absolute f<strong>in</strong>al<br />
position.<br />
In both the <strong>in</strong>itial and f<strong>in</strong>al clusters produced for 'wetness' the most<br />
common phonemes are the follow<strong>in</strong>g: l (6 <strong>in</strong>stances), s (5 <strong>in</strong>stances) and<br />
p (4 <strong>in</strong>stances), as could be expected, cf. diagrams 4.14 and 4.15.<br />
Semantic feature 'rough surface structure'<br />
For the semantic feature 'rough surface structure', the <strong>in</strong>itial clusters<br />
produced <strong>in</strong> the test are: fl- (2) and one <strong>in</strong>stance each of pr-, skr-, kn-,<br />
kl-, tl-. Of these only skr- has the feature 'rough surface structure'. Tlbreaks<br />
the phonotactic pattern. There are other cluster with this feature,<br />
fr- and kr-, accord<strong>in</strong>g to the analysis <strong>in</strong> chapter 4, but they were not used<br />
by the subjects. Perhaps the formulation <strong>in</strong> Swedish : 'tydlig, hård<br />
ytstruktur' was not clear enough.<br />
The f<strong>in</strong>al consonant clusters produced are: -Nk (2), -rb and -lb. In the<br />
analysis of chapter 5, the only roots classified as 'rough surface structure'<br />
236
are knottra (ttr) and rispa-raspa (sp). Rough surface structure is hardly a<br />
feature of f<strong>in</strong>al consonant clusters.<br />
Discussion of free production tests from mean<strong>in</strong>g<br />
to constructed word<br />
Subjects seem to encode semantic features <strong>in</strong> the <strong>in</strong>itial clusters rather<br />
than <strong>in</strong> the f<strong>in</strong>al ones when produc<strong>in</strong>g new sound symbolic words. This is<br />
mirrored <strong>in</strong> the results of chapter 5: many of the semantic features are<br />
not conventionally encoded <strong>in</strong> specific f<strong>in</strong>al consonant clusters. However,<br />
subjects do produce f<strong>in</strong>al clusters <strong>in</strong> these words, but these f<strong>in</strong>al clusters<br />
seem to be either mirrored <strong>in</strong> the <strong>in</strong>itial clusters or are haphazard. There<br />
is of course also the possibility that semantic features other than those <strong>in</strong><br />
the study are relevant for f<strong>in</strong>al clusters.<br />
Subjects do produce clusters that break phonotactics, both <strong>in</strong>itially and<br />
f<strong>in</strong>ally. For sound symbolically low frequent features (like 'dryness'),<br />
with few possible clusters, there are more unconventional forms<br />
produced.<br />
The semantic features that to the greatest extent were expressed accord<strong>in</strong>g<br />
to the model were 'pejorative' (163 7), 'bad mood' (18) and 'wetness'<br />
(63). Less successfully expressed were 'w<strong>in</strong>d<strong>in</strong>g form' (20), 'dryness' (1)<br />
and 'rough surface' (10). There is a tendency for frequent features to be<br />
more successfully coded. The <strong>in</strong>itial consonant clusters of 'pejorative',<br />
with the exception of one, are accord<strong>in</strong>g to expectations. The most<br />
commonly produced <strong>in</strong>itial consonant clusters for 'bad mood' are <strong>in</strong><br />
accordance with the phonesthemes vr-, tr-, gr-. Of the rest of the words<br />
produced all except one have an r. For the <strong>in</strong>itial consonant clusters of<br />
'wetness' the most common phonemes are: l, s and p. The feature<br />
'w<strong>in</strong>d<strong>in</strong>g form' stands out <strong>in</strong> that it uses the non-expected phonemes s and<br />
l <strong>in</strong> different positions of the words. This might be an effect of the<br />
trigger word 'sl<strong>in</strong>grig form' ('w<strong>in</strong>d<strong>in</strong>g form').<br />
7 The absolute frequencies of these features, for the <strong>in</strong>itial clusters, are shown <strong>in</strong> table 4.2.<br />
237
A match<strong>in</strong>g test with context<br />
Yet another study was done <strong>in</strong> order to shed further light on the results<br />
presented <strong>in</strong> table 7.8. A match<strong>in</strong>g test between the mean<strong>in</strong>gs and all<br />
columns of words of table 7.8 8 was performed on a group of 9 l<strong>in</strong>guistics<br />
students with Swedish as their first language. The subjects were told to<br />
match the six columns of expressions with a random list of the six<br />
mean<strong>in</strong>gs.<br />
This test gave 100 % correct results. The results depend partly on the<br />
possibility of compar<strong>in</strong>g the words <strong>in</strong> the different columns. We get an<br />
effect from context. An isolated read<strong>in</strong>g of the columns might be<br />
somewhat more difficult, and might not yield the same degree of correct<br />
results. Even more difficult would be to read (or hear) one word at a<br />
time and suggest a mean<strong>in</strong>g. Another <strong>in</strong>terest<strong>in</strong>g test would be to ask<br />
people to pick out the word most suitable for every mean<strong>in</strong>g of each<br />
labeled column.<br />
7.1.5 Match<strong>in</strong>g test of nonsense words and<br />
mean<strong>in</strong>gs<br />
In this test (Test 3), pairs of nonsense words, e.g. fnottig-skvottig, were<br />
to be matched with pairs of contrast<strong>in</strong>g mean<strong>in</strong>gs, e.g. 'wet' - 'dry'. The<br />
subjects were 13 of those <strong>in</strong> tests 1 and 2. This test gave the follow<strong>in</strong>g<br />
results, for each question:<br />
Question 1 (skv- 'wetness', fn- 'dryness'): 92 % expected answers<br />
Question 2 (str- 'long th<strong>in</strong> form', kr- 'w<strong>in</strong>d<strong>in</strong>g form'): 84% expected<br />
answers<br />
Question 3 (spr- 'separation', kn- 'putt<strong>in</strong>g together': 84% expected<br />
answers<br />
Question 4 (skr- 'rough surface structure', mj- 'soft consistency'): 100%<br />
expected answers<br />
Question 5 (fj- 'silly' (pejorative), vr- 'arg' (bad mood): 100% expected<br />
answers.<br />
8 One column of words is e.g. slirv, islig, plyr, ril, pis, sjl<strong>in</strong>g, siloln, krel, kril, vr<strong>in</strong>lig,<br />
tirori, slio, evans, vrom.<br />
238
In question 1, one person (out of 13) deviated, on questions 2 and 3 two<br />
persons deviated. The same person deviated on questions 1, 2 and 3. In<br />
other words, one person stands for 3 out of 5 unexpected answers. (This<br />
is the person who's answers are marked with an * <strong>in</strong> tables 7.8. and 7.7.)<br />
7.1.6 Summary of results of tests on <strong>in</strong>terpret<strong>in</strong>g<br />
mean<strong>in</strong>gs and sounds<br />
The forced choice tests<br />
The forced choice tests (1b and 2b) had very good results for 19 out of<br />
39 test questions <strong>in</strong> the test from mean<strong>in</strong>g to constructed words, and 28<br />
test questions showed a majority for expected answers. In the test from<br />
constructed words to mean<strong>in</strong>g (1a and 2a) 14 out of 38 test questions<br />
showed very good results, and 29 test questions showed a majority for<br />
expected answers. This is an overall good result for the <strong>in</strong>terpretation of<br />
the hypothesized phonesthemes. The phonesthemes that were most<br />
successfully <strong>in</strong>terpreted <strong>in</strong> the test from mean<strong>in</strong>g to constructed words<br />
were fj- 'pejorative', fn- 'dryness', str- 'long th<strong>in</strong> form', sp- 'wetness',<br />
vr- 'bad mood', and pj- 'pejorative'. The phonesthemes that were most<br />
successfully <strong>in</strong>terpreted <strong>in</strong> the test from constructed words to mean<strong>in</strong>g<br />
were gr- hollow form', fn- 'dryness', sn- 'talk<strong>in</strong>g' and pj- 'pejorative'.<br />
Comb<strong>in</strong><strong>in</strong>g the results of both tests, pj- 'pejorative' and fn- 'dryness'<br />
were the most successful.<br />
The question arises whether the <strong>in</strong>terpretation of semantic features <strong>in</strong><br />
general correlate with the most frequent features accord<strong>in</strong>g to the lexical<br />
analysis. The features 'pejorative': pj- and 'dryness': fn- (both lexically<br />
small clusters with a high degree of sound symbolism) showed the highest<br />
expected correlations between mean<strong>in</strong>g and sound among the six most<br />
successful features for both forced choice mean<strong>in</strong>g to constructed word<br />
and constructed word to mean<strong>in</strong>g tests (tables 7.2 and 7.5). It seems then<br />
that it is generally easier to <strong>in</strong>terpret and code phonesthemes of lexically<br />
low frequent, highly sound symbolic clusters. The mean<strong>in</strong>g that gave the<br />
smallest number of expected answers was 'light', (also for <strong>in</strong>terpret<strong>in</strong>g<br />
sounds). There were five test words with the expected feature 'light' and<br />
the clusters tested were bl-, gl- and gn-. Even when the vowel was <strong>in</strong><br />
accordance with the expectations of the lexical analysis, i.e. hav<strong>in</strong>g an i or<br />
239
e vowel like <strong>in</strong> gletig and gnitig, the words were not <strong>in</strong>terpreted as 'light'<br />
by a majority of speakers, and the category 'light' was not coded with bl-<br />
, gl- or gn- by a majority of speakers. , To quite a great extent, gn- is a<br />
lexically low frequent, highly sound symbolic cluster, as is the case with<br />
bl- and gl. However, this does not seem to be enough for productivity.<br />
As mentioned above, the 'light' phonesthemes (which are not productive 9)<br />
have the only semantic feature which is metaphorically connected with<br />
sound.<br />
The free production tests<br />
The free production of mean<strong>in</strong>gs or words tests (1a and 2a) gave a<br />
variety of mean<strong>in</strong>gs and new sound symbolic words, many of which were<br />
predicted but some were not. In the free production test from form to<br />
mean<strong>in</strong>g no constructed word was <strong>in</strong>terpreted as expected by all subjects,<br />
but all of the constructed words (except one which is given a second best<br />
<strong>in</strong>terpretation) are <strong>in</strong>terpreted correctly by some of the subjects. The<br />
success rank<strong>in</strong>g of the six phonesthemes are, <strong>in</strong> order from best match to<br />
worst: 'pejorative' (for pj-), 'destruction' (for skr-), 'wetness' (for skv-<br />
), 'pejorative' (for fn-), 'bad mood' (for vr-), 'w<strong>in</strong>d<strong>in</strong>g form' (for kr-).<br />
The features (of the six tested) that were most successful were<br />
'pejorative', 'bad mood', and 'wetness'. Also the fact that all the test<br />
subjects will<strong>in</strong>gly produced nonsense words, and mean<strong>in</strong>gs from nonsense<br />
words, is an <strong>in</strong>terest<strong>in</strong>g result. Thus, these nonsense words are <strong>in</strong>terpreted<br />
<strong>in</strong> accordance with expectations by some subjects, but not all. The<br />
nonsense words have a phonesthemic mean<strong>in</strong>g potential that is used by<br />
some subjects <strong>in</strong> the test situation.<br />
The free production test from mean<strong>in</strong>g to form shows that phonesthemes<br />
are used. This test also shows that for all semantic features both percental<br />
and absolute frequencies of the lexical analysis correlate with the clusters<br />
that the subjects use <strong>in</strong> free production of sound symbolic words. Subjects<br />
seem to encode semantic features <strong>in</strong> the <strong>in</strong>itial clusters rather than <strong>in</strong> the<br />
f<strong>in</strong>al ones when produc<strong>in</strong>g new sound symbolic words. They also produce<br />
clusters that break phonotactic rules, both <strong>in</strong>itially and f<strong>in</strong>ally. The<br />
9 There may well be productive 'light' phonesthemes, but it might be the case that the<br />
expression side of these phonesthemes has to <strong>in</strong>clude a vowel, e.g. i.<br />
240
semantic features that to the greatest extent were expressed accord<strong>in</strong>g to<br />
the model, were 'pejorative', 'bad mood' and 'wetness'. Less successfully<br />
identified were 'w<strong>in</strong>d<strong>in</strong>g form', 'dryness' and 'rough surface'. There is a<br />
tendency for frequent features to be more successfully coded.<br />
For both free production tests, the most successful features (out of six<br />
possible) were 'pejorative', 'bad mood' and 'wetness'.<br />
Thus <strong>in</strong> both the forced choice tests and free production tests the most<br />
successful semantic feature is 'pejorative'.<br />
The match<strong>in</strong>g test<br />
The match<strong>in</strong>g test (c), where two words are to be matched with two<br />
semantic features, produced the highest percentages of expected answers.<br />
The results partly depend on the possibility to compare the word pairs.<br />
We get an effect from context.<br />
Summary and discussion<br />
Most of the phonesthemes are more or less successfully <strong>in</strong>terpreted or<br />
coded, while some are more clearly not <strong>in</strong>terpreted or coded. In table 7.9<br />
below, the phonesthemes that were the most successfully <strong>in</strong>terpreted and<br />
coded <strong>in</strong> the experiments are presented, <strong>in</strong> order to show which are<br />
recurr<strong>in</strong>g <strong>in</strong> the different experiments.<br />
241
Table 7.9 The phonesthemes most successfully <strong>in</strong>terpreted or coded <strong>in</strong><br />
experiments with neologisms.<br />
Test:<br />
Phonestheme:<br />
most<br />
successfulphonesthemes<br />
leastsuccessfulphonesthemes<br />
forced<br />
choice<br />
from<br />
mean<strong>in</strong>g to<br />
phonol.<br />
form<br />
pejorative:<br />
fjdryness:fnlong<br />
th<strong>in</strong><br />
form:<br />
strwetness:spbad<br />
mood:<br />
vrpejorative:pj-<br />
light:<br />
gn-<br />
forced<br />
choice from<br />
phonol.<br />
form to<br />
mean<strong>in</strong>g<br />
hollow form:<br />
grdryness:fntalk<strong>in</strong>g:snpejorative:pjquick<br />
or<br />
strong<br />
movement:<br />
skseparation:sprlight:gnlight:gl-<br />
242<br />
free<br />
production<br />
from<br />
mean<strong>in</strong>g to<br />
phonol.<br />
form<br />
pejorative<br />
(several<br />
clusters)<br />
bad mood<br />
(several<br />
clusters)<br />
wetness<br />
(several<br />
clusters)<br />
w<strong>in</strong>d<strong>in</strong>g<br />
form,<br />
dryness,<br />
rough<br />
surface<br />
free<br />
production<br />
from<br />
phonol.<br />
form to<br />
mean<strong>in</strong>g<br />
pejorative:<br />
pj-<br />
w<strong>in</strong>d<strong>in</strong>g<br />
form:<br />
kr-<br />
As can be seen from table 7.9 the most successful phonestheme is<br />
'pejorative': pj- and the least successful is 'light': gn- and 'w<strong>in</strong>d<strong>in</strong>g form'<br />
(kr-). Pj- and gn- are lexically low frequent, highly sound symbolic<br />
clusters.<br />
The clusters of most of the most successful phonesthemes <strong>in</strong> table 7.9<br />
above – fj-, fn-, vr-, pj-, spr-, skr- and mj- – are lexically low<br />
match<strong>in</strong>g test<br />
rough<br />
surface<br />
structure vs.<br />
soft<br />
consistency:<br />
skr- vs mjpejorative<br />
vs.<br />
bad mood:<br />
fj- vs. vr-<br />
-
frequent, very sound symbolic clusters. Str-, sp-, gr-, sn- and sk- are<br />
not. The most successful semantic feature of table 7.8 is 'pejorative'. The<br />
general success of the feature 'pejorative' is discussed <strong>in</strong> chapter 8. Of the<br />
least successful phonesthemes, gn- is lexically low frequent, highly sound<br />
symbolic, but not gl- and kr-. The degree of success <strong>in</strong> these tests is not<br />
then exclusively restricted to the lexically low frequent, highly sound<br />
symbolic clusters.<br />
243
8 Summary and discussion<br />
The purpose of this thesis has been to study different aspects of sound<br />
symbolism – with special reference to Swedish. The largest part of the<br />
descriptive study has been devoted to Swedish phonesthemes. Initial and<br />
f<strong>in</strong>al consonant clusters were primarily studied, but vowels were also<br />
<strong>in</strong>cluded.<br />
Other important issues have been: study<strong>in</strong>g productivity of new sound<br />
symbolic words and similarities between languages (connected with the<br />
issue of universals), and f<strong>in</strong>d<strong>in</strong>g a reasonable explanatory model for (a<br />
part of) sound symbolism <strong>in</strong> Swedish.<br />
The role of sound symbolism <strong>in</strong> language was discussed and a model for<br />
the position of sound symbolism <strong>in</strong> language was constructed, tak<strong>in</strong>g a<br />
number of factors <strong>in</strong>to account, such as <strong>in</strong>nateness, learn<strong>in</strong>g, productivity,<br />
context and conventionalization. (cf. 1.9).<br />
The sound symbolic properties of consonant clusters and vowels were<br />
described – the expressions and the mean<strong>in</strong>gs. The mean<strong>in</strong>gs of<br />
phonesthemes were found to be relatable to each other and different<br />
explanations for the relationship between sound and mean<strong>in</strong>g <strong>in</strong> sound<br />
symbolism were also discussed, especially <strong>in</strong> relation to <strong>in</strong>dexicality and<br />
iconicity. A semantic model for expla<strong>in</strong><strong>in</strong>g the different semantic features<br />
of sound symbolism was constructed.<br />
With a po<strong>in</strong>t of departure <strong>in</strong> the description of Swedish phonesthemes a<br />
number of studies were made <strong>in</strong> order to <strong>in</strong>vestigate universality and,<br />
above all, the role of productivity.<br />
8.1 The research questions were as follows:<br />
1. What are the properties of sound symbolic sounds and sound sequences<br />
<strong>in</strong> Swedish? More specifically the questions are:<br />
Which <strong>in</strong>itial and f<strong>in</strong>al consonant clusters are used <strong>in</strong> sound<br />
symbolism?<br />
245
Which mean<strong>in</strong>gs are used <strong>in</strong> sound symbolism?<br />
How do these comb<strong>in</strong>e <strong>in</strong> phonesthemes?<br />
What are the sound symbolic characteristics of some vowels?<br />
How do <strong>in</strong>itial and f<strong>in</strong>al clusters and vowels comb<strong>in</strong>e <strong>in</strong> words?<br />
2. Are phonesthemes productive <strong>in</strong> Swedish? And, if so, are some<br />
phonesthemes more productive than others? Are neologisms created or<br />
<strong>in</strong>terpreted <strong>in</strong> accordance with the semantic model of chapter 2 and the<br />
analysis of chapters 4 and 5?<br />
3. Are there similarities or dissimilarities between different languages <strong>in</strong><br />
some aspects of sound symbolism?<br />
4. Do non-Swedish speakers <strong>in</strong>terpret Swedish phonesthemes <strong>in</strong><br />
accordance with the semantic model of chapter 2 and the analysis of<br />
chapters 4 and 5?<br />
The results perta<strong>in</strong><strong>in</strong>g to these questions are summarized and discussed<br />
below:<br />
8.1.1 Question 1<br />
What are the properties of sound symbolic sequences <strong>in</strong> Swedish?<br />
The lexical study of <strong>in</strong>itial consonant clusters h<strong>in</strong>ts at about 1,000 roots<br />
with sound symbolic beg<strong>in</strong>n<strong>in</strong>gs of the 65,000 lexemes (or 8,300 root<br />
morphemes 1) <strong>in</strong> the Swedish vocabulary. The <strong>in</strong>itial consonant clusters<br />
vary <strong>in</strong> the degree to which they are sound symbolic and the same is true<br />
for the f<strong>in</strong>al clusters and the vowels. Almost all <strong>in</strong>itial consonant clusters<br />
and about 22% of the f<strong>in</strong>al clusters are used for sound symbolism. (From<br />
the experiments <strong>in</strong> chapter 7 it seems that the f<strong>in</strong>al consonant clusters are<br />
of less importance than the <strong>in</strong>itial ones <strong>in</strong> produc<strong>in</strong>g new words, see<br />
below.)<br />
1 This number of morphemes is the result of an <strong>in</strong>vestigation of a one million word<br />
newspaper corpus (NFO 4). It is most probable that newspaper language conta<strong>in</strong>s<br />
more roots than spoken language, but these figures are not known at present. The<br />
estimate of 65, 000 lexemes excludes transparent compounds.<br />
246
A restricted number of mean<strong>in</strong>gs that are semantically relatable to each<br />
other are used <strong>in</strong> Swedish phonesthemes. These are <strong>in</strong> most cases<br />
connected with perception or cognition (cf. the models <strong>in</strong> 1.9 and 2.6.2).<br />
A consonant cluster usually has more than one possible sound symbolic<br />
mean<strong>in</strong>g, and the semantic profiles vary for different clusters.<br />
The most frequent semantic features for <strong>in</strong>itial and f<strong>in</strong>al clusters are<br />
partly the same and partly different. For the <strong>in</strong>itial clusters, the 10 most<br />
common features are, <strong>in</strong> descend<strong>in</strong>g order:<br />
Table 8.1 The 10 most common<br />
semantic features of <strong>in</strong>itial clusters.<br />
'pejorative'<br />
'sound'<br />
'long th<strong>in</strong> form'<br />
'quick or strong movement'<br />
'wetness'<br />
'talk<strong>in</strong>g'<br />
'light'<br />
'dim<strong>in</strong>utive'<br />
'round form'<br />
'walk<strong>in</strong>g'<br />
For the f<strong>in</strong>al clusters, the 10 most common semantic features of Svensk<br />
Baklängesordbok and the 6 most common features of Nusvensk<br />
Frekvensordbok 4 (NFO 4) are shown <strong>in</strong> table 8.2.<br />
When the most common semantic features of f<strong>in</strong>al clusters of Svensk<br />
Baklängesordbok and Nusvensk Frekvensordbok 4 are comb<strong>in</strong>ed, the<br />
follow<strong>in</strong>g features are most frequent, <strong>in</strong> descend<strong>in</strong>g order: 'sound' and<br />
'pejorative', 'talk<strong>in</strong>g', 'quick or strong movement', 'long th<strong>in</strong><br />
form', 'wetness' and 'slang' ,'walk<strong>in</strong>g', 'round form', 'light'.<br />
247
The first six features are also the six most common features of <strong>in</strong>itial<br />
clusters, cf. table 8.1 (cf. also table 5.4).<br />
Table 8.2 The 10 most common semantic features of f<strong>in</strong>al clusters of<br />
Svensk Baklängesordbok and the six most common features of the<br />
Nusvensk Frekvensordbok 4.<br />
Svensk Baklängesordbok Nusvensk Frekvensordbok 4 2<br />
'quick or strong movement' 'pejorative'<br />
'talk<strong>in</strong>g' 'sound'<br />
'sound' 'long th<strong>in</strong> form'<br />
'pejorative' 'talk<strong>in</strong>g'<br />
'slang' 'wetness'<br />
'walk<strong>in</strong>g' 'quick or strong movement'<br />
'wetness'<br />
'long th<strong>in</strong> form'<br />
'round form'<br />
'light'<br />
This can be compared with the features that were the most successfully<br />
coded and <strong>in</strong>terpreted <strong>in</strong> the experiments of chapter 7. In the experiments<br />
of free production tests from mean<strong>in</strong>g to constructed word, 'pejorative',<br />
'bad mood', and 'wetness' were the most successful.<br />
Of the six features listed above 'pejorative and 'wetness' were the most<br />
successful <strong>in</strong> the experiments of free production tests from mean<strong>in</strong>g to<br />
constructed word.<br />
In the forced choice tests the most successful semantic features were<br />
'pejorative' and 'dryness' ('sound', 'dim<strong>in</strong>utive', 'round form' and<br />
'walk<strong>in</strong>g' were not tested <strong>in</strong> the forced choice experiments, see 7.1).<br />
2 There are only six semantic features <strong>in</strong> this column s<strong>in</strong>ce they are the result of the<br />
analysis presented <strong>in</strong> table 5.18, which showed the most common clusters and their<br />
semantic features; other semantic features are therefore not very frequent <strong>in</strong> Nusvensk<br />
Frekvensordbok 4.<br />
248
Of the six features listed above, 'pejorative' was the most successful <strong>in</strong> the<br />
forced choice tests.<br />
The phonological characteristics of the most common sound symbolic<br />
clusters, <strong>in</strong>itially and f<strong>in</strong>ally and <strong>in</strong> particular for certa<strong>in</strong> mean<strong>in</strong>gs, were<br />
as follows:<br />
Initial clusters<br />
For <strong>in</strong>itial clusters, the most common ones (cf. table 4.8) are sl-<br />
(absolutely) and pj- (percentally). For the five most frequent mean<strong>in</strong>gs,<br />
<strong>in</strong> absolute numbers, there is a strong tendency for the consonant clusters<br />
to end with l. For the five most frequent mean<strong>in</strong>gs, percentally, there is a<br />
slight tendency that the consonant clusters end with j. The clusters end<strong>in</strong>g<br />
with j are lexically low frequent ones and are thus marked. This can<br />
make them more useful for sound symbolic functions, cf. 8.3.<br />
Consider<strong>in</strong>g specific semantic features, the results for the <strong>in</strong>itial clusters<br />
are: all two-consonant clusters conta<strong>in</strong><strong>in</strong>g a j-phoneme are 'pejorative'<br />
and they are so to quite a great extent. Look<strong>in</strong>g at absolute figures,<br />
clusters beg<strong>in</strong>n<strong>in</strong>g with s, especially sl-, sn-, skr- are dom<strong>in</strong>ant among<br />
the pejorative roots. The semantic feature 'bad mood' is ma<strong>in</strong>ly coded<br />
with the clusters gr-, vr- and tr- (but these clusters can have many other<br />
mean<strong>in</strong>gs). Furthermore 'light' is dom<strong>in</strong>ated by voiced plosives + l or n –<br />
bl-, gl-, gn- (dl-, bn-, dn- are not allowed <strong>in</strong> Swedish) – and 'wetness' is<br />
dom<strong>in</strong>ated by sl-, kl-, sp-, sm-, skv-, spr-, sn-, bl-, pl-, sv-, i.e. s or<br />
<strong>in</strong>itial unvoiced plosives are preferred (the only exception is bl-). The<br />
feature 'long th<strong>in</strong> form' is dom<strong>in</strong>ated by three three-consonant clusters:<br />
spj-, str-, spr-. These all beg<strong>in</strong> with s (a fact of all Swedish <strong>in</strong>itial threeconsonant<br />
clusters). In addition, with the exception of two clusters<br />
connected with the feature 'long th<strong>in</strong> form', the above mentioned clusters<br />
plus sp-, sl-, sn-, st-, sk- and sv-, also beg<strong>in</strong> with an s. The only ones<br />
that do not beg<strong>in</strong> with s are tr- and tv-. These are however, voiceless<br />
dentals like s.<br />
249
Most clusters that are percentally most frequent are lexically very<br />
<strong>in</strong>frequent ones, cf. discussion <strong>in</strong> 8.3.<br />
The results of the lexical analysis of chapter 4 concern<strong>in</strong>g the pj-<br />
'pejorative' cluster, the gr-, tr-, vr- 'bad mood', the skr- 'destruction',<br />
the skv- 'wetness' and the fn- 'dryness' clusters correlate with the results<br />
of the free production or free <strong>in</strong>terpretation experiments of chapter 7 (cf.<br />
question 2 below).<br />
F<strong>in</strong>al clusters<br />
As discussed <strong>in</strong> the beg<strong>in</strong>n<strong>in</strong>g of chapter 5, it is not self-evident which are<br />
the f<strong>in</strong>al clusters <strong>in</strong> Swedish. Morphological structure has to be taken <strong>in</strong>to<br />
account s<strong>in</strong>ce some sequences only occur <strong>in</strong> <strong>in</strong>flected or derived forms<br />
(e.g. -ndsk <strong>in</strong> bondsk) and some sequences, which are obviously<br />
<strong>in</strong>terest<strong>in</strong>g, cannot occur <strong>in</strong> the absolutely f<strong>in</strong>al position (e. g. -dr- as <strong>in</strong><br />
fladdra. The mean<strong>in</strong>g profiles for the sound symbolic f<strong>in</strong>al clusters differ,<br />
as is also the case for <strong>in</strong>itial clusters, cf. diagrams 5.3–5.11.<br />
For the f<strong>in</strong>al clusters, the ma<strong>in</strong> results are as follows: The most common<br />
clusters with sound symbolic mean<strong>in</strong>g, <strong>in</strong> order of frequency –<br />
<strong>in</strong>dependent of mean<strong>in</strong>g - are the follow<strong>in</strong>g (from NFO4): -Nk, -sk, -fs,<br />
-nd, -tr (absolutely) or -fs, -dr, -lr, -ml, -Nl (percentally) 3. Of these, 5<br />
of 9 end with a liquid. Of the 11 percentally most sound symbolic f<strong>in</strong>al<br />
consonant clusters, 8 end with a liquid and 3 with a voiceless obstruent (s<br />
or k). Of the 13 most common f<strong>in</strong>al consonant clusters <strong>in</strong> absolute<br />
numbers, 6 end with a liquid and 7 with a voiceless obstruent (s, t or k).<br />
The high frequency of liquids and voiceless obstruents among the f<strong>in</strong>al<br />
clusters could be due to lexical dom<strong>in</strong>ance of these clusters. (Swedish f<strong>in</strong>al<br />
clusters can also end <strong>in</strong> voiced obstruents and nasals.) It is beyond the<br />
scope of this thesis to <strong>in</strong>vestigate this issue. The high sound symbolic<br />
frequency of liquids and voiceless obstruents could also have to do with<br />
the fact that they are especially suited for certa<strong>in</strong> mean<strong>in</strong>gs. As seen <strong>in</strong><br />
table 5.3, however, none of the two groups are especially tied to any<br />
3 The five most frequent f<strong>in</strong>al clusters have a percentage of 100% because there is only<br />
one – sound symbolic – root, for every cluster. These are not considered here, cf.<br />
however, diagram 5.2.<br />
250
semantic feature. In the NFO 4 material, -fs (pejorative) is the most<br />
common phonestheme both absolutely and percentally (cf. tables 5.18 and<br />
5.19).<br />
It is not the case, for f<strong>in</strong>al clusters, that the high frequent semantic<br />
features seem to prefer certa<strong>in</strong> sounds or sound comb<strong>in</strong>ations, cf. table<br />
5.3. The more low frequent semantic features, however, use certa<strong>in</strong><br />
consonant clusters: 'wetness' -ska; 'long th<strong>in</strong> form' -msa; 'round form' -<br />
lla; 'light' -<strong>in</strong>dra, -imra; 'shortwide form' -mpa. These clusters, except<br />
for -<strong>in</strong>dra, -imra ('light') are, as can be seen <strong>in</strong> table 5.3, used by other<br />
semantic features as well.<br />
Vowels<br />
Vowels, a selection of which were studied <strong>in</strong> m<strong>in</strong>imal pairs, seem to have<br />
other semantic characteristics than consonant clusters, often modify<strong>in</strong>g the<br />
mean<strong>in</strong>g of consonant clusters, e.g. be<strong>in</strong>g dim<strong>in</strong>utive or augmentative. [i]<br />
tends to have the mean<strong>in</strong>gs 'smallness', 'quickness', 'high pitch', while [P]<br />
seems to have the mean<strong>in</strong>gs of 'low pitch' and 'largeness'. [a] seems<br />
neutral. The vowel [P] is also connected with 'pejorative'. The semantic<br />
feature 'light' is connected with front vowels, especially with [i].<br />
Comb<strong>in</strong>ations<br />
The study of comb<strong>in</strong>ations of <strong>in</strong>itial and f<strong>in</strong>al consonant clusters showed<br />
that it is more common for semantic features to be the same <strong>in</strong>itially and<br />
f<strong>in</strong>ally, than to be different. If new semantic features had been searched<br />
for <strong>in</strong> the <strong>in</strong>vestigation of f<strong>in</strong>al clusters, there is a possibility that this<br />
result would have to be modified, because of the appearance of new roots.<br />
The semantic features under consideration are however, mostly the same<br />
<strong>in</strong>itially and f<strong>in</strong>ally.<br />
It seems then, that <strong>in</strong>itial and f<strong>in</strong>al clusters strengthen each other rather<br />
than contribut<strong>in</strong>g to different sound symbolic mean<strong>in</strong>gs <strong>in</strong> the same root.<br />
However, for those consonant clusters that can occur both <strong>in</strong>itially and<br />
f<strong>in</strong>ally it is not the case that they are comb<strong>in</strong>ed <strong>in</strong> the same word, e.g. bland<br />
-bl, but neologisms as blabbla are conceivable.<br />
251
8.1.2 Question 2<br />
Are phonesthemes productive <strong>in</strong> Swedish? And, if so, are some<br />
phonesthemes more productive than others? Are neologisms created and<br />
<strong>in</strong>terpreted <strong>in</strong> accordance with the semantic model of chapters 1 and 2 and<br />
the analysis of chapters 4 and 5? The model of 1.9 says, among other<br />
th<strong>in</strong>gs, that sound symbolism is productive. Furthermore, the model of<br />
2.6.2 claims that semantic features of sound symbolism are, due to the<br />
<strong>in</strong>nateness of categories of th<strong>in</strong>k<strong>in</strong>g - such as predispositions for see<strong>in</strong>g<br />
contiguity and similarity, and due to learn<strong>in</strong>g - relatable to sense<br />
impressions and emotions, restricted to certa<strong>in</strong> types.<br />
The characteristics of the consonant clusters (cf. above, question 1) are<br />
mirrored <strong>in</strong> the experiments of chapter 7 <strong>in</strong> the follow<strong>in</strong>g way:<br />
Forced choice tests<br />
The phonesthemes, <strong>in</strong> the constructed words, that were the most<br />
successfully coded, i.e. given a phonological form based on a presented<br />
semantic feature, were 'pejorative': fj-, 'dryness': fn-, long th<strong>in</strong> form':<br />
str-, 'wetness': sp-, 'bad mood': vr-, 'pejorative': pj-. The features most<br />
successfully coded totally were 'wetness', 'hollow form', 'separation' and<br />
'dryness'. The six phonesthemes most successfully <strong>in</strong>terpreted were<br />
'hollow form': gr-, 'dryness': fn-, 'talk<strong>in</strong>g': sn-, 'pejorative': pj-, 'quick<br />
or strong movement': sk- and 'separation': spr-. For both the<br />
<strong>in</strong>terpretation and production tests, the most successful phonesthemes<br />
were 'dryness': fn- and 'pejorative': pj-. Both of these are (lexically low<br />
frequent) clusters that are sound symbolic to a very high degree. A<br />
possible explanation for the success of these phonesthemes is that they can<br />
be more easily accessed. Phonesthemes could be stored <strong>in</strong> a way different<br />
from other morphemes s<strong>in</strong>ce there is a motivated connection between<br />
sound and mean<strong>in</strong>g. This could make them more reliable and more<br />
frequent <strong>in</strong> production and <strong>in</strong>terpretation.<br />
The 'light' words stand out because of many <strong>in</strong>stances of bad results, both<br />
<strong>in</strong> the word-to-mean<strong>in</strong>g and mean<strong>in</strong>g-to-word tests. It is clear that the<br />
'light' category gets the least number of expected answers. The 'light'<br />
words might belong to a closed lexical class which would imply that gl-,<br />
252
gn-, bl-, are not productive, unlike phonesthemes such as kl- or pj-.<br />
Even when the 'light' words had a vowel with high F2, as many of the<br />
'light' words have accord<strong>in</strong>g to 5.4, the results of the experiment were no<br />
better. The semantic feature 'light' is positioned quite high <strong>in</strong> the analysis<br />
of the most common features (cf. diagram 4.5).<br />
Free production tests<br />
Phonesthemes<br />
In the free production test from expression to mean<strong>in</strong>g, the percentally<br />
most common phonestheme pj- (pejorative) was the most accurately<br />
identified. In the free production test from mean<strong>in</strong>g to expression the jclusters,<br />
among others, for 'pejorative' were produced as expected. The<br />
r-clusters for 'bad mood' were also produced as expected, plus some<br />
additional r-clusters, both <strong>in</strong>itial and f<strong>in</strong>al. For the feature 'wetness',<br />
there was a majority of s-, l- and p-clusters (as expected), both <strong>in</strong>itially<br />
and f<strong>in</strong>ally.<br />
Explanations of these results, for the j-clusters, could have to do with the<br />
facts that these are (lexically low frequent) clusters that are sound<br />
symbolic to a very high degree; they could thus be more easily accessed<br />
and this could, <strong>in</strong> turn, make them more reliable – and frequent – <strong>in</strong><br />
production and <strong>in</strong>terpretation. As concerns the r-clusters for 'bad mood'<br />
they can have their base <strong>in</strong> expressions which are spontaneous<br />
vocalizations <strong>in</strong> connection with an angry feel<strong>in</strong>g, and the s-, l- and pclusters<br />
for 'wetness' have their base <strong>in</strong> sound imitation. The question of<br />
whether one of these bases gives better results than the other <strong>in</strong><br />
<strong>in</strong>terpretation and production cannot be answered from these experiments<br />
alone; further experimentation is needed.<br />
The phonestheme most difficult to <strong>in</strong>terpret was 'w<strong>in</strong>d<strong>in</strong>g form': kr-.<br />
Semantic features<br />
The semantic features of the free production mean<strong>in</strong>g to word<br />
experiments that were most successfully expressed (<strong>in</strong> accordance with the<br />
lexical analysis) were 'pejorative' (163), 'bad mood' (18) and 'wetness'<br />
(63). Less successfully expressed were 'w<strong>in</strong>d<strong>in</strong>g form' (20), 'dryness' (1)<br />
253
and 'rough surface' (10) (The absolute frequencies of these semantic<br />
features, for the <strong>in</strong>itial clusters, are shown with<strong>in</strong> the parenthesis, cf. table<br />
4.2.) These results can be compared with the most frequent semantic<br />
features accord<strong>in</strong>g to the lexical analysis summarized <strong>in</strong> tables 8.1 and<br />
8.2. Of the three most successful features – 'pejorative', 'bad mood' and<br />
'wetness' – of the free production experiments, 'pejorative' and 'wetness'<br />
are among the six most common features accord<strong>in</strong>g to the lexical analysis.<br />
A probable explanation for these similarities is that phonesthemes of the<br />
most frequent semantic features are stored <strong>in</strong> such a way that they are<br />
more accurately accessed by the language user. The way they are stored is<br />
dependent on stable <strong>in</strong>termodal connections, cf. the discussion <strong>in</strong> 1.11.2.<br />
Of the three less successful features of the free production experiments -<br />
'w<strong>in</strong>d<strong>in</strong>g form', 'dryness' and 'rough surface' - none belonged to the 10<br />
most common semantic features of the lexical analysis. A probable<br />
explanation for this is that the phonesthemes of the less frequent semantic<br />
features are stored <strong>in</strong> such a way that they are less accurately accessed.<br />
The conclusion is that phonesthemes are – more or less – productive, both<br />
<strong>in</strong> production of new forms and understand<strong>in</strong>g of neologisms. There is a<br />
tendency for the most common semantic features to be more successfully<br />
coded, <strong>in</strong> accordance with the ma<strong>in</strong> results of the analysis of<br />
phonesthemes <strong>in</strong> chapter 4; these might be more accessible. It could also<br />
be the case that categories related to negative emotions (e.g. 'pejorative',<br />
'bad mood') are more important to humans (at least Swedes) than more<br />
abstract categories like 'form'. It also seems that subjects tend to encode<br />
semantic features <strong>in</strong> the <strong>in</strong>itial clusters rather than <strong>in</strong> the f<strong>in</strong>al ones <strong>in</strong> free<br />
production, and it seems that for low frequent features (like 'dryness'),<br />
with few possible clusters, there are a larger number of unconventional<br />
forms produced. These tendencies need to be <strong>in</strong>vestigated further.<br />
An <strong>in</strong>terest<strong>in</strong>g result from the free choice test from expression to<br />
mean<strong>in</strong>g is that the mean<strong>in</strong>gs produced all belong to the classes found <strong>in</strong><br />
the lexical study. Even the <strong>in</strong>formants who started the test by freely<br />
suggest<strong>in</strong>g mean<strong>in</strong>gs based on constructed neologisms (and therefore had<br />
no expectancies as to what the answers ought to be) produced mean<strong>in</strong>gs<br />
254
with<strong>in</strong> these classes (although not always with<strong>in</strong> the expected class for a<br />
certa<strong>in</strong> nonsense word - the reason for this might be <strong>in</strong>dividual contextual<br />
<strong>in</strong>fluences at the moment of the test). These classes are: 'pejorative' (often<br />
'destruction'), 'mental feel<strong>in</strong>g' (often irritated), 'surface structure',<br />
'wetness', 'form', 'consistency', 'movement', 'dim<strong>in</strong>utive', 'sound'. These<br />
results can also be compared with the most frequent semantic features<br />
accord<strong>in</strong>g to the lexical analysis summarized <strong>in</strong> table 8.1. Six of the n<strong>in</strong>e<br />
general semantic features result<strong>in</strong>g from the free choice test from<br />
expression to mean<strong>in</strong>g ('pejorative', 'wetness', 'form', 'movement',<br />
'dim<strong>in</strong>utive' and 'sound') are among (or superord<strong>in</strong>ate to) the n<strong>in</strong>e most<br />
common semantic features of table 8.1. This supports the model of 2.6.2<br />
where the phonesthemes for these categories are seen as a result of<br />
<strong>in</strong>nateness, learn<strong>in</strong>g and conventionalization; the semantic categories of<br />
phonesthemes are predictable rather than haphazard. This model shows<br />
that <strong>in</strong> many cases the semantic features of phonesthemes are potentially<br />
relatable to neurological connections between the senses, cf. 8.3.2. It<br />
could also imply that phonesthemes that concern the most frequent<br />
semantic features are stored <strong>in</strong> such a way that they are most readily and<br />
accurately accessed 4.<br />
8.1.3 Question 3<br />
Are there similarities or dissimilarities between different languages <strong>in</strong><br />
various aspects of sound symbolism,?<br />
In the Thesaurus study of the concepts 'stupidity', 'rough surface<br />
structure', and 'smooth surface structure', for English and Swedish, the<br />
follow<strong>in</strong>g was found: The phonological agreement between words<br />
belong<strong>in</strong>g to these semantic fields <strong>in</strong> English and Swedish was greater<br />
among the words for 'surface structure' than for the words for<br />
'pejoratives'. One obvious reason is that some of the clusters used <strong>in</strong><br />
Swedish (fj-, fn-, pj-) are not allowed <strong>in</strong> English. Another reason could<br />
be that 'surface structure' is closer to a potentially common phenomenon,<br />
namely sound imitation, s<strong>in</strong>ce strok<strong>in</strong>g different surfaces give different<br />
4 This is a question for further experiments, e.g. lexical decision experiments, and<br />
development of on l<strong>in</strong>e models.<br />
255
sound effects. Of course, Swedish and English have many cognates but not<br />
<strong>in</strong> these results (cf. 6.2), except for some cases.<br />
In the tests with <strong>in</strong>formants concern<strong>in</strong>g different <strong>in</strong>terjections <strong>in</strong> 8<br />
languages the follow<strong>in</strong>g similarities and dissimilarities were found<br />
concern<strong>in</strong>g expressions: The pejorative <strong>in</strong>terjections often conta<strong>in</strong> an u or<br />
an O, the positive <strong>in</strong>terjections an a, 'surprise' often an a or an O. 'Pa<strong>in</strong>'<br />
(e.g. Swedish aj) has a diphtongized open vowel, 'sneeze' (e.g. Swedish<br />
atjo) has an affricate and the <strong>in</strong>terjection for go away (to an animal) (e.g.<br />
Swedish schas) has a fricative <strong>in</strong> all the 8 languages.<br />
In the tests with <strong>in</strong>formants of 16 different languages concern<strong>in</strong>g different<br />
expressions that are imitative of animal noises, the ma<strong>in</strong> results are the<br />
follow<strong>in</strong>g: No animal call imitation is exactly the same <strong>in</strong> all the 16<br />
languages. Some animal call imitations are more alike, e.g. the cat's<br />
meow, while others vary more, e.g. the dog's bark<strong>in</strong>g. One reason for<br />
this could be that some animal cries are more complicated than others.<br />
There are, however, always similarities on the level of phonetic features<br />
between the expressions for the same animal call <strong>in</strong> the different<br />
languages. With<strong>in</strong> each language there are, as well as for the expressive<br />
<strong>in</strong>terjections, similarities which can be assumed to depend on the<br />
phonological structure of the language <strong>in</strong> question.<br />
In the identification test, expressions for different animal sounds by<br />
speakers of 9 different languages were tested on 15 persons with 6<br />
different first languages. The ma<strong>in</strong> results are that some animals were<br />
identified by all listeners (e.g. dog, cow and cat), some less correctly<br />
identified and one animal not identified at all (e.g. duck, Hungarian<br />
expression). The <strong>in</strong>terpretation test shows that, given the context of<br />
animal calls, it was quite easy for listeners to <strong>in</strong>terpret animal calls from<br />
languages other than their own (cf. table 6.6).<br />
Thus, for these contrastive studies, the follow<strong>in</strong>g general conclusion can<br />
be drawn. There are both similarities and dissimilarities between the<br />
256
expressions <strong>in</strong> the different languages 5. This is true for most phenomena<br />
<strong>in</strong> language, and sound symbolism is no exception here. The variation is<br />
greater for some semantic fields than for others; expressions for rough or<br />
smooth surface structure are more alike than expressions for pejoratives.<br />
A possible explanation of this is that 'surface structure' is closer to a<br />
potentially common phenomenon, namely sound imitation, either because<br />
strok<strong>in</strong>g different surfaces give different sound effects or because of the<br />
auditory-tactile neural connection. Certa<strong>in</strong> animals are imitated more<br />
consistently than others, possibly because certa<strong>in</strong> animal cries are shorter<br />
and less complicated than others, cf. e.g. the cries of the cat and the<br />
rooster.<br />
8.1.4 Question 4<br />
Do non-Swedish speakers <strong>in</strong>terpret Swedish phonesthemes <strong>in</strong> accordance<br />
with the semantic model of chapter 2 and the analysis of chapters 4 and 5?<br />
In the test of understand<strong>in</strong>g Swedish sound symbolic words (and some<br />
neologisms constructed from phonesthemes) the general results are as<br />
follows: The subjects seldom <strong>in</strong>terpreted the words correctly. However,<br />
the answers given almost all belong to categories with<strong>in</strong> the semantic<br />
model of 2.6.2 (cf. table 6.7). It therefore seems that for certa<strong>in</strong><br />
categories, often related to sense impressions or emotions, there is a<br />
preference for sound symbolic expressions; it is more natural for<br />
speakers to imag<strong>in</strong>e that an unknown word, where the expression is<br />
supposed to reflect the mean<strong>in</strong>g, means e.g. 'quick or strong movement'<br />
than that it means e.g. 'my country'.<br />
The conclusion from the studies under questions 3 and 4 is that there are<br />
greater differences between these languages 6 for expression than for<br />
mean<strong>in</strong>g. The expressions imitat<strong>in</strong>g animal sounds are easier to <strong>in</strong>terpret<br />
than other sound symbolism for speakers of different languages.<br />
5 For crossl<strong>in</strong>guistic studies <strong>in</strong> other semantic fields, see e.g. Viberg (1999).<br />
6 Swedish, English, Icelandic, Polish, Hungarian, F<strong>in</strong>nish, Ososo, Malagasi,<br />
Slovenian, Korean, Japanese, Ch<strong>in</strong>ese, Estonian, Urdu, Persian, Kurdish, Arabic and<br />
Spanish. Persons who <strong>in</strong>terpreted Swedish sound symbolic words spoke Arabic,<br />
Spanish, German, Dutch, Ibo, English.<br />
257
Interpretation of other sound symbolism often goes wrong (because<br />
expressions probably differ <strong>in</strong> different languages), but the semantic<br />
categories guessed at are most often with<strong>in</strong> the semantic models of this<br />
thesis.<br />
8.2. Comparison of the studies<br />
The next table (8.3) is a comparison between the most frequent<br />
phonesthemes of the lexical studies with the most accurately <strong>in</strong>terpreted<br />
and coded phonesthemes of the experimental studies. The most common<br />
clusters and the most common mean<strong>in</strong>gs are also shown. The data are<br />
from different tables and diagrams as <strong>in</strong>dicated <strong>in</strong> table 8.3.<br />
Table 8.3 The most frequent 7 phonesthemes or the most successful 8<br />
phonesthemes (their consonant clusters and mean<strong>in</strong>gs), whichever is most<br />
appropriate, from the different studies. The examples are presented <strong>in</strong><br />
frequency order. In some cases, however, consecutive examples might<br />
have the same frequency. 9<br />
7 Accord<strong>in</strong>g to the lexical analysis.<br />
8 In the case of the experiments where subjects produced or <strong>in</strong>terpreted neologisms.<br />
9 Squares are empty for the experiments where there are no relevant data.<br />
258
lexically <strong>in</strong>itial<br />
consonant<br />
clusters<br />
experimentally<br />
f<strong>in</strong>al consonant<br />
clusters<br />
forced choice 14<br />
mean<strong>in</strong>g to<br />
expression<br />
phonesthemes consonant<br />
clusters<br />
pejorative:<br />
sl-, pjsound:<br />
kl-, kn-, fnlong<br />
th<strong>in</strong> form:<br />
sp-, spjquick<br />
or strong<br />
movement:<br />
flwetness:<br />
sl-, skv-10 quick or strong<br />
movement: -NNl,<br />
talk<strong>in</strong>g: -tr,<br />
sound: -st,<br />
pejorative: -fs,<br />
wetness: -sk,<br />
long th<strong>in</strong> form<br />
-Nk11 fj-: pejorative<br />
fn-: dryness<br />
str-: long th<strong>in</strong><br />
form<br />
sp-: wetness<br />
vr-:bad mood<br />
pj-<br />
:pejorative15 259<br />
<strong>in</strong> absolute<br />
numbers:<br />
sl-, sn-, kn-,<br />
kr-, kl-, sp-,<br />
gl-, st-, tr-<br />
(cf. diagram<br />
4.1)<br />
<strong>in</strong> %:<br />
fn-, gn-, skv-<br />
, pj-, kn-,<br />
spr-, spj-, gl-<br />
, mj-, vr- (cf.<br />
diagram 4.4)<br />
-Nk, -sk,-fs, -<br />
nd, -tr (cf.<br />
diagram 5.1<br />
and table 5.21)<br />
-dr, -lr, -ml,<br />
-NNl 12<br />
10From table 4.8.<br />
11 From tables 5.3 and 5.18.<br />
12 From NFO4, diagram 5.2 (except for the 1 root clusters).<br />
13 From NFO4 and Svensk Baklängesordbok, section 5.5.3.<br />
mean<strong>in</strong>gs<br />
pejorative,<br />
sound, long<br />
th<strong>in</strong> form,<br />
quick or strong<br />
movement,<br />
wetness,<br />
talk<strong>in</strong>g, light,<br />
dim<strong>in</strong>utive,<br />
round form,<br />
walk<strong>in</strong>g (cf.<br />
diagram 4.5)<br />
sound,<br />
pejorative,<br />
talk<strong>in</strong>g,<br />
quick or strong<br />
movement,<br />
long th<strong>in</strong> form,<br />
wetness13 wetness,<br />
hollow form<br />
separation<br />
dryness,<br />
pejorative (cf.<br />
table 7.3)
cross<br />
cultural<br />
<strong>in</strong>terpretation<br />
forced choice<br />
expression to<br />
mean<strong>in</strong>g16 free production<br />
expression to<br />
mean<strong>in</strong>g<br />
free production<br />
mean<strong>in</strong>g to<br />
expression<br />
gr-: hollow<br />
form<br />
fn-: dryness<br />
sn-: talk<strong>in</strong>g<br />
pj-: pejorative<br />
sk-: quick or<br />
strong<br />
movement<br />
spr-:<br />
separation17 pj-: pejorative<br />
skr-:<br />
destruction<br />
skv-:<br />
wetness18 pejorative: fl-,<br />
fjbad<br />
mood: rclusters<br />
wetness: sl-,<br />
sv-19 fl20-: quick or<br />
strong<br />
movement<br />
skr- :<br />
destruction<br />
tr-: bad mood<br />
260<br />
quick or strong<br />
movement21 14 The semantic feature 'sound' was not tested.<br />
15 From table 7.2.<br />
16 For both the production and <strong>in</strong>terpretation experiments added together, the most<br />
successful phonesthemes were fn-: 'dryness' and pj-: 'pejorative'<br />
17 From table 7.5.<br />
18 From table 7.7.<br />
19 From table 7.8.<br />
20Phonesthemes of words that were most successfully <strong>in</strong>terpreted accord<strong>in</strong>g to the<br />
Swedish norm. Cf. table 6.8.<br />
21 Most preferred mean<strong>in</strong>g for unconventional but possible mean<strong>in</strong>gs. 'Pejorative' and<br />
'sound' were NOT among the more preferred mean<strong>in</strong>gs. Cf. table 6.9.
Table 8.3 shows, among other th<strong>in</strong>gs, the follow<strong>in</strong>g relations between<br />
different studies: The most successful <strong>in</strong>itial phonesthemes of the<br />
experimental studies that correspond to <strong>in</strong>itial phonesthemes of the most<br />
frequent semantic features are: pj-: pejorative (<strong>in</strong>itial consonant clusters,<br />
forced choice mean<strong>in</strong>g to expression, forced choice expression to<br />
mean<strong>in</strong>g, free production expression to mean<strong>in</strong>g); skv-:wetness (<strong>in</strong>itials,<br />
free production expression to mean<strong>in</strong>g); sl-:wetness (<strong>in</strong>itials, free<br />
production mean<strong>in</strong>g to expression) <strong>in</strong> this most condensed version of the<br />
results.<br />
Sl- is the percentally largest cluster for 'wetness' (and THE most sound<br />
symbolic cluster – for all semantic features – <strong>in</strong> absolute figures), pj- is<br />
the percentally largest 'pejorative' cluster and skv- is the percentally fifth<br />
largest wetness cluster 22. Pj- and skv- are both small but percentally<br />
highly sound symbolic clusters.<br />
An explanation for these results is that phonesthemes that are<br />
proportionally large (i. e. where a certa<strong>in</strong> mean<strong>in</strong>g has a proportionally<br />
large part of the (sound symbolic) roots of a cluster) are stored <strong>in</strong> such a<br />
way that they are more accurately accessed by the language user.<br />
However, all phonesthemes of the free production experimental results<br />
above match the lexical analysis. A possible explanation for most of these<br />
similarities could be that the phonesthemes of semantic features are<br />
relatable to neurological connections between the senses (<strong>in</strong> accordance<br />
with the model of 2.6.2).<br />
The most successful phonesthemes (which were <strong>in</strong> fact not very<br />
successful, cf. table 6.8) of the cross cultural <strong>in</strong>terpretation experiment do<br />
not reflect the frequent Swedish semantic feature 'pejorative'. An<br />
explanation for the lack of success of this feature could be that the<br />
negative emotion that 'pejorative' is based on is a cultural trait of<br />
Swedish, but not of the other languages <strong>in</strong>vestigated here. However, of the<br />
phonesthemes that did succeed, one of the clusters, fl-, is the<br />
22 Cf. also table 4.8.<br />
261
proportionally most sound symbolic cluster for 'quick or strong<br />
movement', and the other two clusters skr- and tr- are the proportionally<br />
the most sound symbolic clusters for the semantic features 'destruction'<br />
and 'bad mood' respectively. These results are then <strong>in</strong> accordance with<br />
Swedish competence. One obvious explanation for the overall bad results<br />
is that non-native Swedes do not connect accurate expressions with the<br />
semantic features as a natural consequence of different phonological<br />
learn<strong>in</strong>g environments. 23<br />
The vowels were not taken <strong>in</strong>to consideration <strong>in</strong> this table, s<strong>in</strong>ce they<br />
were studied <strong>in</strong> a different way, which <strong>in</strong>volved contrast<strong>in</strong>g some<br />
mean<strong>in</strong>gs and sounds <strong>in</strong> m<strong>in</strong>imal pairs.<br />
The cross-l<strong>in</strong>guistic comparisons concern<strong>in</strong>g <strong>in</strong>terjections and the<br />
expressions for animal cries are not comparable to the data <strong>in</strong> table 8.3,<br />
because they concern semantic fields other than the Swedish lexical study<br />
and the experiments. The Thesaurus study, compar<strong>in</strong>g English and<br />
Swedish, showed that, for the concepts 'stupidity', 'rough surface<br />
structure' and 'smooth surface structure', similar phonemes were used for<br />
the last two concepts, but not for the first. A possible explanation for this<br />
is that 'surface structure' is closer to a potentially simultaneous<br />
phenomenon, namely sound imitation, s<strong>in</strong>ce strok<strong>in</strong>g different surfaces<br />
give different sound effects. The comparison of different languages <strong>in</strong> the<br />
studies of <strong>in</strong>terjections showed certa<strong>in</strong> agreements <strong>in</strong> vowels. Most of the<br />
<strong>in</strong>terjections consist of vowels and s<strong>in</strong>gle consonants and are therefore not<br />
comparable with sound symbolism of <strong>in</strong>itial or f<strong>in</strong>al consonant clusters.<br />
For the animal sounds, there were great expressive similarities between<br />
the different languages. It was quite easy for listeners of different first<br />
languages to correctly identify animals from the way <strong>in</strong> which their cries<br />
were represented <strong>in</strong> different languages, cf. summary <strong>in</strong> 8.1.3.<br />
Compar<strong>in</strong>g these last two studies with the Thesaurus study (only <strong>in</strong>volv<strong>in</strong>g<br />
the related languages Swedish and English but show<strong>in</strong>g great differences),<br />
23 However, as can be seen <strong>in</strong> table 6.9, subjects mostly guessed with<strong>in</strong> the expected<br />
semantic categories of the model. A potential explanation for these preferences has to<br />
do with neurological connections between the senses, i. e. the parts of the model <strong>in</strong><br />
2.6.2 connected with <strong>in</strong>nate predispositions.<br />
262
the conclusion is that the expressions where there is a more direct (ma<strong>in</strong>ly<br />
<strong>in</strong>dexical or iconic) connection between expression and mean<strong>in</strong>g are more<br />
alike <strong>in</strong> different languages.<br />
8.3 Possible explanations of onomatopoeia and<br />
sound symbolism<br />
Why is it that certa<strong>in</strong> consonant clusters are connected with certa<strong>in</strong><br />
mean<strong>in</strong>gs? And why is it that we want certa<strong>in</strong> mean<strong>in</strong>gs to be expressed<br />
sound symbolically, as well as with conventional morphemes; what is the<br />
function of sound symbolic morphemes as compared with full<br />
morphemes?<br />
The first question is easy to answer when it concerns sound imitation. The<br />
expressive side of words are articulated sounds and articulation can<br />
imitate sounds, <strong>in</strong> a more or less conventionalized way. One possible<br />
explanation to other sound symbolism is that phonesthemes were<br />
orig<strong>in</strong>ally onomatopoeic and later developed <strong>in</strong>to metaphorically (e.g.<br />
gn-: 'sound' and later also 'light') or metonymically (e.g. skv-: 'sound'<br />
and later also 'wetness') related mean<strong>in</strong>gs (cf. 2.6.1). Underly<strong>in</strong>g this are<br />
the <strong>in</strong>nate capacities for metaphor and metonymy. This is <strong>in</strong> accordance<br />
with the theories of e.g. Herder (cf. 1.6.) about the orig<strong>in</strong> of language.<br />
This is not an impossible view s<strong>in</strong>ce almost all <strong>in</strong>itial consonant clusters<br />
have the feature 'sound' or 'talk<strong>in</strong>g' (which later feature is a sub-category<br />
of 'sound', cf. 4.2.3) to a greater or lesser extent. The only exceptions are<br />
pj-, spj-, spl- and spr-.<br />
There are some l<strong>in</strong>guistic facts about Swedish and analyses of several<br />
languages by various l<strong>in</strong>guists that show along what l<strong>in</strong>es other sound<br />
symbolism <strong>in</strong> Swedish can partly be expla<strong>in</strong>ed. A tendency is that lexically<br />
<strong>in</strong>frequent clusters (i.e. marked clusters) are exploited sound symbolically<br />
to a higher degree than lexically frequent clusters. It also seems as if a<br />
large proportion of the three-consonant s-clusters are used for sound<br />
symbolism <strong>in</strong> Swedish, cf. diagram 4.4. This is <strong>in</strong> accordance with the<br />
discussion of H<strong>in</strong>ton, Nichols, Ohala (1994) about marked sounds be<strong>in</strong>g<br />
used for sound symbolism. Also, sounds and sound comb<strong>in</strong>ations<br />
263
otherwise non-existent <strong>in</strong> the language (wild forms) occur <strong>in</strong><br />
onomatopoeic and sound symbolic neologisms.<br />
Another aspect of markedness is that sounds that are new <strong>in</strong> a language<br />
are often used sound symbolically (cf. Austerlitz, 1994, on ø <strong>in</strong> F<strong>in</strong>nish).<br />
In Swedish, the latest great consonant change was the collapse of several<br />
consonant clusters (e. g. skj-, stj-, sj-, kj-, tj-) <strong>in</strong>to the fricatives S and<br />
ç. A possible explanation is that the clusters conta<strong>in</strong><strong>in</strong>g j that were left<br />
over – spj-, bj-, fj-, pj- – became more unusual and marked <strong>in</strong> Swedish<br />
and therefore useful for sound symbolism.<br />
Classify<strong>in</strong>g the different mean<strong>in</strong>gs of sound symbolism, it is clear that<br />
most of them fall under the perceptual category, but also under cognitive<br />
factors, viz. evaluation, but not under a social category. This h<strong>in</strong>ts at<br />
sound symbolism be<strong>in</strong>g <strong>in</strong> some way biologically grounded, rather than<br />
learned.<br />
Most of the mean<strong>in</strong>gs of the phonesthemes are relatable to the senses<br />
(probably so because of metaphorical or metonymical extensions from<br />
sound and/or neurological connections between the senses), apart from the<br />
'pejoratives' and 'mental feel<strong>in</strong>g'. Thus the phonesthemes are related to<br />
hear<strong>in</strong>g, see<strong>in</strong>g, touch<strong>in</strong>g, (but not to smell<strong>in</strong>g, tast<strong>in</strong>g) 24. The semantic<br />
features occurr<strong>in</strong>g over and over aga<strong>in</strong> are relatable to stable <strong>in</strong>termodal<br />
connections, i.e. one sense is connected to another. There are, however,<br />
many problems with an explanation <strong>in</strong> terms of synaesthesia (see 1.11.2).<br />
Neurological aspects relevant for a model (see chapter 2) are neurological<br />
connections between the senses, or a common ground for the senses.<br />
Modalities that have especially strong connections (accord<strong>in</strong>g to<br />
Geschw<strong>in</strong>d, 1965, accord<strong>in</strong>g to Cytowic, 1989) are visual-auditory and<br />
tactile-auditory.<br />
Semantic features hav<strong>in</strong>g to do with movements can ontogenetically be<br />
expla<strong>in</strong>ed <strong>in</strong> terms of metonymy, i.e., simultaneity <strong>in</strong> time and space<br />
between sound and movement, and metaphor. In some cases<br />
24 There also exist, on a lexical level, mean<strong>in</strong>g extensions from sensory modalities to<br />
mental phenomena (see Abel<strong>in</strong>, 1988).<br />
264
proprioception might be <strong>in</strong>volved, as <strong>in</strong> an explanation of pejoratives<br />
where a feel<strong>in</strong>g of disgust is experienced simultaneously with contractions<br />
of certa<strong>in</strong> facial muscles, which contractions have an effect on articulation<br />
(cf. Darw<strong>in</strong>, "1965").<br />
Ohala's (e.g. 1994) frequency code offers a solution for sound symbolism<br />
of dim<strong>in</strong>utives and augmentatives. The common factor is high-low<br />
frequency for F0 (of <strong>in</strong>tonation), noise or F2. High frequency is<br />
connected to smallness, low frequency to largeness.<br />
The second question – Why is it that we want certa<strong>in</strong> mean<strong>in</strong>gs to be<br />
expressed sound symbolically, as well as with conventional morphemes? –<br />
can be answered with appeal to a k<strong>in</strong>d of redundancy <strong>in</strong> the l<strong>in</strong>guistic<br />
sign. In language, there is partly a "triple articulation". The "third<br />
articulation" is mean<strong>in</strong>g bear<strong>in</strong>g and motivated (cf. 1.2.5), and this makes<br />
sound symbolic words very effective. There are, <strong>in</strong> sound symbolic<br />
words, other ties between expression and mean<strong>in</strong>g than merely the<br />
arbitrary conventional ones, namely those that are motivated but still to<br />
some extent conventional. We can still wonder if sound symbolism is a<br />
remnant from earlier stages, where it could have had a high survival<br />
value because of its connection with the senses and with th<strong>in</strong>gs present at<br />
the moment someth<strong>in</strong>g is spoken. An argument for this is, for example,<br />
Ohala's frequency code. Also, the connection between emotions and facial<br />
contractions possibly result<strong>in</strong>g <strong>in</strong> certa<strong>in</strong> sounds for pejoratives could also<br />
be a remnant from earlier stages. On the other hand, it is possible, as<br />
Jespersen claimed, that languages grow richer and richer <strong>in</strong> sound<br />
symbolic words.<br />
8.3.1 Pejoratives<br />
The pejorative phonesthemes resist a simple explanation. Is there an<br />
ugl<strong>in</strong>ess code? An ugl<strong>in</strong>ess metaphor? Or do they have to do with a basic<br />
dist<strong>in</strong>ction between approval and disapproval? One of the 6 basic<br />
emotions, accord<strong>in</strong>g to e.g. Ekman (1973) is disgust, which is an emotion<br />
underly<strong>in</strong>g pejoratives.<br />
265
Darw<strong>in</strong> (1872) gives a possible explanation for sound symbolic words<br />
related to emotions, e.g. disgust (related to pejoratives) and <strong>in</strong>terjections.<br />
The explanation builds on the <strong>in</strong>st<strong>in</strong>ctive contractions of facial muscles<br />
connected with a certa<strong>in</strong> emotion. Pejoratives could be of a more<br />
<strong>in</strong>dexical nature, the result of <strong>in</strong>terjections which <strong>in</strong> turn could be<br />
conditioned by <strong>in</strong>st<strong>in</strong>ctive facial contractions. A similar type of word is<br />
one that could be termed truly iconic. These are words like pluta, truta<br />
and pussa, perhaps gr<strong>in</strong>a and a few more, where the pronunciation of the<br />
vowel <strong>in</strong> particular can be seen as be<strong>in</strong>g dependent on the shape of the<br />
face and sometimes connected to emotions.<br />
Hamano (1994) proposes an explanation <strong>in</strong> child directed speech. He<br />
studied palatalization <strong>in</strong> Japanese sound symbolism. Palatalization of<br />
alveolar stops and fricatives is associated with "childishness" and<br />
"immaturity". He connects this to studies on language acquisition<br />
report<strong>in</strong>g palatalization as one of the universal characteristics of early<br />
stages of children’s language acquisition. Palatalization is also reported as<br />
one of the commonest devices of baby-talk, i.e., child directed speech<br />
(Snow and Ferguson, 1977). Perhaps this is a way of expla<strong>in</strong><strong>in</strong>g the<br />
Swedish pejorative clusters bj-, fj-, pj-, which do not fit <strong>in</strong>to a<br />
synaesthetic and metaphoric pattern.<br />
The high frequency of pejoratives <strong>in</strong> Swedish phonesthemes may well be a<br />
cultural peculiarity, s<strong>in</strong>ce it was not mirrored <strong>in</strong> the cross cultural<br />
<strong>in</strong>terpretation experiment. The pejorative could be based on a general<br />
negative emotion, which could be favored <strong>in</strong> a certa<strong>in</strong> social environment.<br />
Naturally, further cross l<strong>in</strong>guistic research is needed before any<br />
conclusions can be drawn with respect to this.<br />
8.3.2 Summary<br />
To summarize:<br />
1) The unique feature of phonesthemes <strong>in</strong> relation to ord<strong>in</strong>ary morphemes<br />
is that they are bound morphemes that are both mean<strong>in</strong>g bear<strong>in</strong>g and<br />
motivated (cf. 1.2.5). Words with phonesthemes are very effective s<strong>in</strong>ce<br />
there are ties between expression and mean<strong>in</strong>g other than merely the<br />
266
arbitrary conventional, namely the motivated and to some extent the<br />
conventional.<br />
2a) Certa<strong>in</strong> mean<strong>in</strong>gs are suitable to be conveyed symbolically <strong>in</strong> sound<br />
e.g. mean<strong>in</strong>gs related to the senses because of the <strong>in</strong>teraction between, on<br />
the one hand, <strong>in</strong>nate capacities based on neurological connections between<br />
the senses and, on the other hand, learn<strong>in</strong>g <strong>in</strong> the present. The<br />
neurological connections which (accord<strong>in</strong>g to Geschw<strong>in</strong>d, 1965,<br />
accord<strong>in</strong>g to Cytowic, 1989) are especially strong are the connections<br />
visually-auditorily and tactile-auditorily. They make it possible for sound<br />
imitative expressions and phonesthemes to lead to other sound symbolic<br />
connections between these sound sequences and e.g. forms and movements<br />
(which have to do with visual and/or tactile experience), and surface<br />
structures and consistency (which have to do with tactile experience). The<br />
learn<strong>in</strong>g process, of which a part is perceiv<strong>in</strong>g sounds and mimick<strong>in</strong>g<br />
them, <strong>in</strong> comb<strong>in</strong>ation with conventionalization, causes the sound symbolic<br />
expressions to vary (slightly) <strong>in</strong> different languages.<br />
2b) The connection between the expression and mean<strong>in</strong>g of pejoratives is<br />
more difficult to relate to neurological connections between the senses and<br />
can be given various explanations, connected with e. g. markedness, child<br />
language or proprioception. The emotion of disgust might, however, be<br />
beh<strong>in</strong>d the semantic feature 'pejorative'.<br />
3) The result that f<strong>in</strong>al clusters predom<strong>in</strong>antly end with a liquid or a<br />
voiceless obstruent could perhaps be expla<strong>in</strong>ed by the auditory salience <strong>in</strong><br />
the case of several voiceless obstruents; what is heard well is more useable<br />
<strong>in</strong> general. It could be that such sounds are chosen for sound symbolism<br />
s<strong>in</strong>ce they could then fulfill functions other than merely the dist<strong>in</strong>ctive. It<br />
is more difficult to give an explanation of the predom<strong>in</strong>ance of liquids<br />
s<strong>in</strong>ce this is a group based on both acoustic similarities and common<br />
phonological pattern<strong>in</strong>g.<br />
The usefulness of sound symbolism <strong>in</strong> general is connected with the<br />
strong tie between mean<strong>in</strong>g and expression, discussed <strong>in</strong> 3.3. It is<br />
imag<strong>in</strong>able that peoples' reactions are quicker, stronger and more<br />
267
accurate to sound symbolic expressions, and therefore they can be e.g.<br />
more effective, more emotionally arous<strong>in</strong>g, and more poetic.<br />
8.4 Predictions for sound symbolism <strong>in</strong> Swedish<br />
When a new non-arbitrary word is created or understood it is most likely<br />
to have the follow<strong>in</strong>g characteristics:<br />
1. Its semantic content belongs to those described <strong>in</strong> chapter 4 (except for<br />
'light') and 5 25.<br />
<strong>Sound</strong>: <strong>Sound</strong><br />
Talk<strong>in</strong>g<br />
Beat<br />
Movement: Movement<br />
Quick or strong movement<br />
Walk<strong>in</strong>g<br />
Fall<strong>in</strong>g<br />
Potential movement<br />
Quickness<br />
Light: Light<br />
Gaze<br />
Surface structure: Rough surface structure<br />
Smooth surface structure<br />
Consistency: Soft consistency<br />
(Plasticity) Hardness<br />
Slackness<br />
Stiffness<br />
Wetness: Wetness<br />
Adhesion<br />
25 The order of presentation follows the order <strong>in</strong> 2.6 and 4.2.3, which is based on the<br />
prelim<strong>in</strong>ary lexical frequencies of this study. After the lexical study of <strong>in</strong>itial clusters<br />
was completed the order of the semantic features was somewhat changed, but the<br />
orig<strong>in</strong>al order was kept <strong>in</strong> some lists. It therefore does not represent an analytical order.<br />
For the order of frequency of the completed lexical study of <strong>in</strong>itial clusters, see diagram<br />
4.5.<br />
268
Dryness: Dryness<br />
Attitude: Attitude<br />
Secrecy<br />
Slang: Slang<br />
Jocular: Jocular<br />
Pejorative: Pejorative<br />
Destruction<br />
Mental feel<strong>in</strong>g: Mental feel<strong>in</strong>g<br />
Bad mood<br />
Bodily feel<strong>in</strong>g: Bodily feel<strong>in</strong>g<br />
Suffocation<br />
Separation: Separation<br />
Putt<strong>in</strong>g together: Putt<strong>in</strong>g together<br />
Dim<strong>in</strong>utive: Dim<strong>in</strong>utive<br />
Augmentative: Augmentative<br />
Form: Form<br />
Round form<br />
Short-wide form<br />
Th<strong>in</strong> form<br />
Hollow form<br />
W<strong>in</strong>d<strong>in</strong>g form<br />
Long th<strong>in</strong> form<br />
Narrow form<br />
Small end form<br />
Bent form<br />
Iterative: Iterative<br />
F<strong>in</strong>e gra<strong>in</strong><br />
2. Its <strong>in</strong>itial and f<strong>in</strong>al consonant cluster is one that is appropriate for its<br />
semantic content, see chapters 4 (except for 'light') and 5. The most sound<br />
symbolic <strong>in</strong>itial clusters <strong>in</strong> absolute numbers and percentally, for the five<br />
most frequent mean<strong>in</strong>gs, are shown <strong>in</strong> table 8.4.<br />
Table 8.4 The most sound symbolic <strong>in</strong>itial clusters <strong>in</strong> absolute numbers<br />
and percentally, for the five most frequent mean<strong>in</strong>gs.<br />
269
mean<strong>in</strong>g <strong>in</strong> freq. % <strong>in</strong> % % freq.<br />
freq.<br />
'pejorativ sl- 24 16 pj- 71 5<br />
e'<br />
'sound' kl-, 15 13, fn- 33 3<br />
kn-<br />
19<br />
'long th<strong>in</strong> sp- 23 17 spj- 33 2<br />
form'<br />
'quick or fl- 18 19 fl- 19 18<br />
strong<br />
movement<br />
'<br />
'wetness' sl- 12 8 skv- 40 4<br />
The most frequent semantic features for f<strong>in</strong>al clusters accord<strong>in</strong>g to NFO 4<br />
are shown <strong>in</strong> table 8.5. Clusters and cluster frequencies are shown to the<br />
right (cf. tables 5.4, 5.18 and 5.19).<br />
Table 8.5 The most frequent semantic features for f<strong>in</strong>al clusters accord<strong>in</strong>g<br />
to NFO 4.<br />
mean<strong>in</strong>g <strong>in</strong><br />
freq.<br />
freq. <strong>in</strong> % %<br />
'quick or<br />
Nl 4 Nl, 40,<br />
strong<br />
movement'<br />
dr,<br />
lr<br />
25,<br />
22<br />
'talk<strong>in</strong>g' tr, bl 5, 4 dr, 25,<br />
bl 24<br />
'sound' st, sl, 5 ,4, lr, sl 44,<br />
lr, tr 4, 4<br />
27<br />
'pejorative' fs, 10, fs, 59,<br />
sk, 5, 4 ms, 27,<br />
ms<br />
dr 25<br />
'wetness' sk 8 - -<br />
'long th<strong>in</strong><br />
Nk, 7, 4 - -<br />
form'<br />
lk<br />
270
8.5 Ma<strong>in</strong> objective and further research<br />
The ma<strong>in</strong> objective of this thesis is the description of Swedish<br />
phonesthemes, which can constitute a basis for further research <strong>in</strong> this<br />
area.<br />
Another important result is the <strong>in</strong>sight that sound symbolism is not an<br />
issue of whether or not phonesthemes exist. Instead, sound symbolism is<br />
present <strong>in</strong> consonant clusters <strong>in</strong> vary<strong>in</strong>g degrees. Some clusters are more<br />
sound symbolic than others while some mean<strong>in</strong>gs are more frequent than<br />
others. Some phonesthemes are more easily <strong>in</strong>terpretable than others.<br />
Some phonesthemes are more productive than others and <strong>in</strong> slightly<br />
vary<strong>in</strong>g ways for different speakers. It is obvious that sound symbolic<br />
categories <strong>in</strong> Swedish show a conceptual organization that is compatible<br />
with both <strong>in</strong>nate categories of th<strong>in</strong>k<strong>in</strong>g, such as similarity and contiguity,<br />
and with learn<strong>in</strong>g, <strong>in</strong> connection with sense impressions and emotion.<br />
An <strong>in</strong>terest<strong>in</strong>g cont<strong>in</strong>uation of this <strong>in</strong>vestigation is to study the effect of<br />
context on the <strong>in</strong>terpretation of phonesthemes, e.g. <strong>in</strong> experiments with<br />
neologisms <strong>in</strong> different sentence contexts.<br />
Special spoken corpora would also be <strong>in</strong>terest<strong>in</strong>g to study, <strong>in</strong> search of<br />
neologisms (<strong>in</strong> context), e.g. corpora of child and adolescent language.<br />
It would also be <strong>in</strong>terest<strong>in</strong>g to study the stability of (the most common)<br />
phonesthemes over time, e.g. by analysis of "dead" words <strong>in</strong> the SAOB<br />
(Svenska Akademiens ordlista över svenska språket).<br />
Yet another <strong>in</strong>terest<strong>in</strong>g cont<strong>in</strong>uation would be to study the dispersion of<br />
(the most common) phonesthemes <strong>in</strong> dialect lexica.<br />
And, of course, it would be very <strong>in</strong>terest<strong>in</strong>g to expand the cross-l<strong>in</strong>guistic<br />
comparisons to more semantic fields and to more languages.<br />
It is also very tempt<strong>in</strong>g to refresh e.g. the experiments of chapter 6 and<br />
the lexical decision experiment (Abel<strong>in</strong>, 1996) with auditive stimuli, with<br />
the aid of reaction time programs which were not available at the time<br />
271
when I <strong>in</strong>itiated the experimentation. Naturally, the context effect should<br />
also be studied <strong>in</strong> this way. Connected with this is the aim to construct a<br />
psychol<strong>in</strong>guistic model for onl<strong>in</strong>e process<strong>in</strong>g for the perception and<br />
production of sound symbolic words.<br />
272
Bibliography<br />
Abel<strong>in</strong>, Å., Allwood, J. (1984) Tolkn<strong>in</strong>g av känsloprosodi - en kontrastiv studie,<br />
unpubl. paper, Göteborg University, Department of L<strong>in</strong>guistics.<br />
Abel<strong>in</strong>, Å. (1988) 'Patterns of synaesthesia <strong>in</strong> the Swedish vocabulary,' <strong>in</strong> <strong>Studies</strong> <strong>in</strong><br />
Computer-Aided lexicology, Almqvist & Wiksell Int. Stockholm<br />
Abel<strong>in</strong>, Å. (1996) 'A lexical decision experiment with onomatopoeic, sound symbolic<br />
and arbitrary words', <strong>in</strong> TMH-QPSR 2/1996. Stockholm<br />
Adams, D., Lloyd, J. (1983) The mean<strong>in</strong>g of Liff, Pan Books and Faber & Faber.<br />
London<br />
Allén, S., Eeg-Olofson, M., Gavare, R., Sjögreen, C. (1981) Svensk baklängesordbok,<br />
Esselte Studium. Stockholm<br />
Allén, S., Berg, S., Järborg, J., Löfström, J., Ralph, B., Sjögreen, C. (1980)<br />
Nusvensk frekvensordbok 4, Almqvist & Wiksell International. Stockholm<br />
Allott, R. (1973) The physical foundation of language, ELB Pr<strong>in</strong>ters. Seaford Sussex<br />
Allwood, J. (1983) Om dist<strong>in</strong>ktionerna mellan semantik och pragmatik, Gothenburg<br />
papers <strong>in</strong> theoretical l<strong>in</strong>guistics S3, Göteborg University, Department of L<strong>in</strong>guistics<br />
Allwood, J. (1983) 'Kan man tänka oberoende av språk' <strong>in</strong> U. Teleman (ed.) Tal och<br />
tanke, Liber förlag. Lund<br />
Allwood, J. (1985) 'Tvärkulturell kommunikation', <strong>in</strong> Papers <strong>in</strong> anthropological<br />
l<strong>in</strong>guistics 12, Göteborg University, Department of L<strong>in</strong>guistics, pp. 9-61<br />
Allwood, J. (1989) Om begrepp – deras bestämn<strong>in</strong>g, analys och konstruktion (ms).<br />
Allwood, J., Andersson, L-G.(1976) Semantik. (Semantics). GULING 1. Göteborg<br />
University, Department of L<strong>in</strong>guistics<br />
Austerlitz, R. (1994) 'F<strong>in</strong>nish and Gilyak sound symbolism – an <strong>in</strong>terplay between<br />
system and history'. <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds) <strong>Sound</strong> symbolism,<br />
Cambridge University Press. Cambridge, pp. 249-260<br />
Bentley, M., Varon, E. (1933) An accessory study of phonetic symbolism Amer. J.<br />
Psychol. 45, 76-86<br />
Bloomfield, L. (1933) Language, Henry Holt. New York<br />
Bol<strong>in</strong>ger, D. (1968) Aspects of language, Harcourt, Brace & World, Inc. New York<br />
Bol<strong>in</strong>ger, D. (1950), Rime, assonance and morpheme analysis, Word 6: 117-136<br />
Bol<strong>in</strong>ger, D. (1965) Forms of English: Accent, morpheme, order, Hokuou Publish<strong>in</strong>g<br />
Company. Tokyo<br />
Br<strong>in</strong>g, S. C. (1930) Svenskt ordförråd, Natur och Kultur. Stockholm<br />
Brown, R. W. (1958) Words and th<strong>in</strong>gs, The Free Press. New York<br />
273
Bühler, K. (1969) "L'onomatopée et la fonction représentative du langage", <strong>in</strong> Essais sur<br />
le langage. J.-P. Pariente (ed.) Les Editions de M<strong>in</strong>uit. Paris (Partial translation of work<br />
orig<strong>in</strong>ally published 1933 as Sprachtheorie ) Fisher. Jena<br />
Chasta<strong>in</strong>g, M. (1958) "Le symbolisme de voyelles: significations des "i" I& II", Journal<br />
de Psychologie 55, 403-423 & 461-481<br />
Chasta<strong>in</strong>g, M. (1965) "Pop - fop - pof - fof", Vie et Langage 159, 311-317<br />
Chasta<strong>in</strong>g, M. (1966) "Si les r étaient des l", (Part 1) Vie et langage 173, 468-472; (Part<br />
2) 174, 502-507<br />
Childs, G.T. (1994) 'African ideophones'. <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds.)<br />
<strong>Sound</strong> symbolism, Cambridge University Press. Cambridge, pp. 178-204<br />
Clark, H., Clark, E. (1977) Psychology and language, An <strong>in</strong>troduction to<br />
psychol<strong>in</strong>guistics, Harcourt, Brace, Jovanovich. New York<br />
Cytowic, (1989) Synaesthesia - a union of the senses, Spr<strong>in</strong>ger Verlag. New York<br />
Darw<strong>in</strong>, C. (1872) The expression of emotion <strong>in</strong> man and animals, John Murray. London<br />
Davis, R. (1961) The fitness of names to draw<strong>in</strong>gs: A cross-cultural study <strong>in</strong><br />
Tanganyika, Brit. J. Psychol. 52, 259-268<br />
Diffloth, G. (1976) 'Expressives <strong>in</strong> Semai', Oceanic l<strong>in</strong>guistics special publication no.<br />
13: Austroasiatic <strong>Studies</strong> I, Honolulu<br />
Diffloth, G., (1994) 'i:big, a:small', <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds) <strong>Sound</strong><br />
symbolism, Cambridge University Press. Cambridge<br />
Ekman, P. (1973) 'Cross-cultural studies of facial expressions' <strong>in</strong> P. Ekman (ed.)<br />
Darw<strong>in</strong> and facial expression , Academic Press. New York (pp. 169-222)<br />
Fano, G. (1962) Saggio sulle orig<strong>in</strong>e del l<strong>in</strong>guaggio. Con una storia critica della dottr<strong>in</strong>e<br />
glottogoniche, Guil<strong>in</strong> E<strong>in</strong>audi Editore. Tor<strong>in</strong>o<br />
Fónagy, I. (1963) Die metaphern <strong>in</strong> der phonetik. The Hague<br />
Freud, S. (1891) On aphasia, a critical study, International Univ. Press. New York<br />
Gabelentz, G. v. d. (1891) Die Sprachwissenschaft, ihre Aufgaben, Methoden und<br />
bisherigen Ergebnisse. Leibzig<br />
Garlén, C. (1988) Svenskans fonologi: i kontrastiv och typologisk belysn<strong>in</strong>g,<br />
Studentlitteratur. Lund<br />
Garman, M. (1990) Psychol<strong>in</strong>guistics, Cambridge University Press. Cambridge<br />
Grammont, M. (1933) Traité de phonétique. Paris<br />
Hamano, S. (1994) 'Palatalization <strong>in</strong> Japanese sound symbolism', <strong>in</strong> L. H<strong>in</strong>ton, J.<br />
Nichols, J.J. Ohala (eds) <strong>Sound</strong> symbolism, Cambridge University Press. Cambridge,<br />
pp. 148-157<br />
Hellquist, E. (1966) Svensk etymologisk ordbok I-II. Gleerup. Lund<br />
274
Herder, J. G. (1772/1969) Essay on the orig<strong>in</strong> of speech <strong>in</strong> P.H. Salus (ed.) On<br />
language: Plato to von Humboldt, Holt, R<strong>in</strong>ehart and W<strong>in</strong>ston. New York, pp.147-166<br />
Hewes, G. W. (1977) 'Language orig<strong>in</strong> Theories' <strong>in</strong> D. M. Rumbaugh (ed) Language<br />
Learn<strong>in</strong>g by a Chimpanzee, New York Academic Press. New York<br />
H<strong>in</strong>ton, L. (1986) 'Musical diffusion and l<strong>in</strong>guistic diffusion'. <strong>in</strong> C. Frisbee (ed)<br />
Anthropology and Music: Essays <strong>in</strong> Honor of David P. McAllester. Detroit Monographs<br />
<strong>in</strong> Musicology 9. Detroit. Information Coord<strong>in</strong>ators<br />
H<strong>in</strong>ton, L., Nichols, J., Ohala, J.J., (1994) 'Introduction: <strong>Sound</strong> symbolic processes',<br />
<strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds) <strong>Sound</strong> symbolism, Cambridge University<br />
Press. Cambridge, pp. 1-12<br />
Holland, M. K., Wertheimer, M. (1964) 'Some physiognomic aspects of nam<strong>in</strong>g, or,<br />
maluma and takete revisited', Percept. Mot. Skills 19, 111-117<br />
Householder, F.W. (1946) On the problem of sound and mean<strong>in</strong>g, An English<br />
phonestheme (discussion <strong>in</strong> Language 23 feb 1946)<br />
Humboldt W. v. (1836/1907) Über die Verschiedenheit des menschlichen Sprachbaues<br />
und ihren E<strong>in</strong>fluss auf die geistige Entwicklung des Menschengeschlechts. Gesammelte<br />
Schriften, Vol. VII, Part 1. Königlich--Preussische Akademie der Wissenschaften. Berl<strong>in</strong><br />
Hörmann, H. (1979) Psychol<strong>in</strong>guistics: an <strong>in</strong>troduction to research and theory, Spr<strong>in</strong>ger-<br />
Verlag. New York<br />
Ideforss, H. (1928) De primära <strong>in</strong>terjektionerna i nysvenskan, 1, Primära impulsioner<br />
och imperationer, Carl Bloms boktryckeri. Lund<br />
Jacobsen Jr., W. H. (1994) 'Nootkan vocative vocalism and its implications' <strong>in</strong> L.<br />
H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds) <strong>Sound</strong> symbolism, Cambridge University Press.<br />
Cambridge, pp. 23-39<br />
Jakobson, R., Fant, G., Halle, M. (1957) Prelim<strong>in</strong>aries to speech analysis, the<br />
dist<strong>in</strong>ctive features and their acoustic correlates, M.I.T. Press. Cambridge, Mass.<br />
Jakobson, R., Waugh, L. (1979) The sound shape of language, Harvester Press, Sussex<br />
Jespersen, O. (1918) "Nogle men-ord", studier tillegnade Esaias Tegner, Lund<br />
Jespersen, O. (1922 a) Language - its nature, development and orig<strong>in</strong>, Allen and Unw<strong>in</strong>.<br />
London<br />
Jespersen, O. (1922 b) "Lydsymbolik", Nordisk tidskrift för vetenskap, konst och<br />
<strong>in</strong>dustri (II), Stockholm<br />
Jesperssen, O. (1933) "Symbolic value of the vowel i" [1922] <strong>in</strong> L<strong>in</strong>guistica, College<br />
Park. Maryland<br />
Jo'hannesson, A. (1949) Orig<strong>in</strong> of language: four essays, Reykjavík, H.F. Leiftur,<br />
Blackwell. Oxford<br />
Katz, J.J., Postal, F. (1964) An <strong>in</strong>tegrated theory of l<strong>in</strong>guistic descriptions, M.I.T.<br />
Press. Cambridge, Mass.<br />
Kots<strong>in</strong>as, U. (1994) Ungdomsspråk, Hallgren & Fallgren. Uppsala<br />
275
Kukkonen, P. (1989) Från konst till vetenskap, PhD Thesis, Hels<strong>in</strong>gfors university<br />
Källström, R. (1988) Morfologi, GULING 9, Göteborg University, Department of<br />
L<strong>in</strong>guistics.<br />
Köhler W. (1930) Gestalt psychology, Bell and Sons Ltd. London<br />
Lakoff, G., Johnsson, M. (1980) Metaphors we live by, The University of Chicago<br />
Press. Chicago and London<br />
Lakoff, G. (1987) Women, fire, and dangerous th<strong>in</strong>gs: what categories reveal about the<br />
m<strong>in</strong>d, University of Chicago Press. Chicago<br />
Lapolla, R.J, (1994) 'An experimental <strong>in</strong>vestigation <strong>in</strong>to phonetic symbolism as it relates<br />
to Mandar<strong>in</strong> Ch<strong>in</strong>ese' <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds.) <strong>Sound</strong> symbolism,<br />
Cambridge University Press. Cambridge<br />
Leech, G. N. (1969) Towards a semantic description of English, Longman. London<br />
Lehrer, A. (1974) Semantic fields and lexical structure, North-Holland. Amsterdam,<br />
London<br />
L<strong>in</strong>dblad, P. (1992) Rösten, Studentlitteratur. Lund<br />
Luria, A. (1977) Mannen med det obegränsade m<strong>in</strong>net, Wahlström & Widstrand.<br />
Stockholm<br />
Lyons, J. (1977) Semantics I-II, Cambridge University Press. Cambridge<br />
Malkiel, Y. (1978) 'From phonosymbolism to morphosymbolism', Fourth LACUS<br />
Forum, Columbia, S.C.<br />
Malkiel, Y. (1994) 'Regular sound development, phonosymbolic orchestration,<br />
disambiguation of homonyms' <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds) <strong>Sound</strong><br />
symbolism, Cambridge University Press. Cambridge<br />
Malmgren, S-G, (1988) 'On regular polysemy <strong>in</strong> Swedish' <strong>in</strong> <strong>Studies</strong> <strong>in</strong> Computer-Aided<br />
Lexicology, Almqvist & Wiksell International. Stockholm<br />
Markell, N.N. & Hamp, E.P. (1960) "Connotative mean<strong>in</strong>gs of certa<strong>in</strong> phoneme<br />
Sequences", <strong>Studies</strong> <strong>in</strong> l<strong>in</strong>guistics 15, pp. 47-61<br />
Marks L.E. (1990) Synaesthesia: perception and metaphor <strong>in</strong> F. Burwick & W. Pape<br />
(eds) Aesthetic Illusion: theoretical and historical approaches, Berl<strong>in</strong> New York, W. de<br />
Gruyter<br />
Massaro, D. W., Tsuzaki, M., Cohen, M. M:, Gesi, A., & Heredia, R. (1993),<br />
'Bimodal speech perception : An exam<strong>in</strong>ation across languages'. <strong>in</strong> Journal of Phonetics,<br />
21<br />
Meillet, A. (1931) Review of Wahlgren, 1930 Bullet<strong>in</strong> de la Société de l<strong>in</strong>guistique de<br />
Paris 31 (2): 113-114<br />
Miller, G. A., Johnsson-Laird, P.N. (1976) Language and perception, Harward<br />
University Press. Cambridge, Mass.<br />
Müller, A. L., (1960) Experimentelle Untersuchungen zur stimmlichen Darstellung von<br />
Gefühlen, Dissertation. Gött<strong>in</strong>gen<br />
276
Müller, M.F. (1861/1961) Three lectures on the science of language and its place <strong>in</strong><br />
general education, repr. Indological Book House. Benares<br />
Nerman, T. (1954) Troll i ord, <strong>in</strong>rim och uddrim i svensk vers, Tidens förlag. Stockholm<br />
Newman, S. (1933) Further experiments <strong>in</strong> phonetic symbolism, American Journal of<br />
Psychology 45, 53-75<br />
Nordberg, B. (1986) 'The use of onomatopoeia <strong>in</strong> the conversational style of<br />
adolescents', FUMS rapport nr 132, Uppsala University. FUMS. Uppsala<br />
Ohala, J. J., (1994) The frequency code underlies the sound-symbolic use of voice pitch,<br />
<strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds.) <strong>Sound</strong> symbolism, Cambridge University<br />
Press. Cambridge<br />
Osgood, C. E. (1962 a) Psychol<strong>in</strong>guistic relativity and universality, Proc. 16th Int.<br />
Congr. Psychol. (Bonn 1960) North-Holland. Amsterdam<br />
Osgood, C. E. (1962 b) <strong>Studies</strong> on the generality of affective mean<strong>in</strong>g systems, Amer.<br />
Psychologist 17, 10-28<br />
Osgood, C. E., G. J. Suci, and P. H. Tannenbaum (1957) The measurment of mean<strong>in</strong>g,<br />
University of Ill<strong>in</strong>ois Press. Urbana, Ill<strong>in</strong>ois<br />
Oswalt, R. L. (1994) 'Inanimate imitatives <strong>in</strong> English', <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J.<br />
Ohala (eds.) <strong>Sound</strong> symbolism, Cambridge University Press. Cambridge, pp. 293-306<br />
The Oxford dictionary of English etymology (1967), ed. by C. T. Onions, University<br />
press. Oxford<br />
Paget, R. (1930) Human Speech, Harcourt, Brace. New York<br />
Peirce, C.S. (1955) 'Logic as semiotic: The theory of signs', <strong>in</strong> Buchler, J. (ed)<br />
Philosophical writ<strong>in</strong>gs of Peirce Dover. New York<br />
Peterfalvi, J.-M. (1965) 'Les recherches expérimentales sur le symbolisme phonétique',<br />
American Journal of Psychology 65, 439-473<br />
Plato (c. 428-348 B.C.) Cratylos dialogue (1962) Loeb Classical Library. Harward<br />
University Press. Cambridge<br />
Porzig, W. (1934/1973) 'Wesenhafte Bedeutungsbeziehungen'. Wortfeldsforschung<br />
1973. 78-103<br />
Révész, G. (1946) Ursprung und Vorgeschichte der Sprache, A. Francke. Bern<br />
Rhodes, R. (1994) 'Aural images', <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds.) <strong>Sound</strong><br />
symbolism, Cambridge University Press. Cambridge, pp. 276-292<br />
Roget's thesaurus of English words and phrases (1977) Pengu<strong>in</strong> Books.<br />
Harmondsworth. Middlesex<br />
Rousseau, J.-J. (1822/1969) Essay on the orig<strong>in</strong> of languages <strong>in</strong> P.H. Salus (ed.) On<br />
Language: Plato to von Humboldt, Holt, R<strong>in</strong>ehart and W<strong>in</strong>ston. New York, pp. 138-146<br />
Samar<strong>in</strong>, W. (1978) 'L<strong>in</strong>guistic adaptation to speech function', <strong>in</strong> W. Mc Cormack, S.A.<br />
Wurm (eds) Approaches to language: Anthropological issues, Mouton. The Hague<br />
277
Sapir, E. (1921) Language, Harcourt Brace. New York<br />
Sapir, E. (1929) 'A study <strong>in</strong> phonetic symbolism', <strong>in</strong> D. Mandelbaum (ed.) Selected<br />
writ<strong>in</strong>gs, Berkeley. California<br />
Saussure, F. de (1916/1983) Course <strong>in</strong> general l<strong>in</strong>guistics, Duckworth. London<br />
Sereno J. A. (1994) 'Phonosyntactics', <strong>in</strong> L. H<strong>in</strong>ton, J. Nichols, J.J. Ohala (eds) <strong>Sound</strong><br />
symbolism, Cambridge University Press. Cambridge<br />
Sigurd, B. (1965) Phonotactic sructures <strong>in</strong> Swedish, Berl<strong>in</strong>gska boktryckeriet. Lund<br />
Sigurd, B. (1993) 'Att tillverka varumärken', Språkvård 3, Svenska språknämnden.<br />
Stockholm, pp. 9-14<br />
Snow, C.E., Ferguson, C.A., (eds), (1977) Talk<strong>in</strong>g to Children: Language Input and<br />
Acquisition, Cambridge University Press. New York<br />
Stenström, C. (1984) Svenskans <strong>in</strong>itiala konsonantkluster - en frekvensstudie, Unpubl.<br />
paper, Dept. of l<strong>in</strong>guistics. Göteborg University, Department of L<strong>in</strong>guistics.<br />
Süssmilch, J.P. (1767), Versuch e<strong>in</strong>es Beweises, dass die erste Sprache ihren Ursprung<br />
nicht vom Menschen, sondern alle<strong>in</strong> vom Schöpfer erhalten habe, Akad. Vorl. Berl<strong>in</strong><br />
Svensk Ordbok (1986) Esselte Studium. Stockholm<br />
Svenska Akademiens ordlista över svenska språket (1974) 10 uppl., P.A. Norstedt &<br />
Söner förlag. Stockholm<br />
Svenska Akademiens ordlista över svenska språket (1998) 12 uppl., Norstedts.<br />
Stockholm<br />
Ordbok över svenska språket, utgiven av Svenska Akademien (SAOB) (1926),<br />
L<strong>in</strong>dstedts universitetsbokhandel. Lund<br />
Thorndike, E. L. (1943) Man and his works, Harvard University Press. Cambridge,<br />
Mass.<br />
Thun, N. (1963) Reduplicative words <strong>in</strong> English: a study of formations of the types ticktock,<br />
hurly-burly, and shilly-shally. PhD Thesis, Stockholm University<br />
Traunmüller, H. (1996) '<strong>Sound</strong> symbolism <strong>in</strong> deictic words', TMH-QPSR 2/1996.<br />
Stockholm<br />
Trier, J. (1934) Das sprachliche Feld, Neue Jahrbucher fur Wissenschaft und<br />
Jugendbildung 10, 428-449<br />
Tsuru, S., Fries, H. S. (1933) 'A problem <strong>in</strong> mean<strong>in</strong>g', Journal of Gen. Psychology 8,<br />
281-284<br />
Ullman, S. (1972) Semantics An <strong>in</strong>troduction to the science of mean<strong>in</strong>g, Blackwell.<br />
Oxford<br />
Ultan, R. (1978) 'Size-sound symbolism', <strong>in</strong> J. Greenberg (ed) Universals of human<br />
language 2,<br />
278
Usnadze, D. (1924) 'E<strong>in</strong> experimenteller Beitrag zum problem der psychologischen<br />
Grundlagen der Namengebung', Psychologische Forschung 5, 24-43.<br />
Werner, H. (1953/1957) Comparative psychology of mental development, International<br />
Universities Press. New York<br />
Werner, H., Kaplan, B. (1963) Symbol formation, Wiley. New York<br />
Wescott, R. (1975) 'Allol<strong>in</strong>guistics: explor<strong>in</strong>g the peripheries of speech', <strong>in</strong> P. Reich<br />
(ed),The Second LACUS Forum, The Hornbeam Press, Columbia, South Carol<strong>in</strong>a, 497-<br />
513<br />
Wescott, R. (1978) 'Lexical polygenesis: Words as resultants of multiple l<strong>in</strong>guistic<br />
pressures' <strong>in</strong> Semantics and Lexicology<br />
Whorf, B. L. (1956/1967) Language, thought and reality: selected writ<strong>in</strong>gs, ed J.B.<br />
Carroll (ed), The MIT paperback series 5, MIT Press, Cambridge, Mass.<br />
Viberg, Å. (1984) 'The verbs of perception: a typological study', L<strong>in</strong>guistics, 21, 1.<br />
Viberg, Å. (1993) 'Crossl<strong>in</strong>guistic perspectives on lexical organization and lexical<br />
progression' <strong>in</strong> K. Hyltenstam & Å. Viberg (eds), Progression and Regression <strong>in</strong><br />
language, Cambridge University Press. Cambridge, pp. 340-385<br />
Wilde, K. (1958) 'Naive und kunstlerische Formen des graphischen Ausdrucks', Ber.<br />
21. Kongr. Dtsch. Ges. Psychol., Gött<strong>in</strong>gen: Verlag fur Psychologie, Bonn, pp. 157-<br />
159<br />
Wisseman, H. (1954) Untersuchungen zur Onomatopoie, Part I: Die<br />
sprachpsychologischen Versuche, W<strong>in</strong>ter. Heidelberg<br />
279
From SOB<br />
Word: Key words,SOB Category<br />
bjäbba uppnosigt prat pejorative, talk<strong>in</strong>g<br />
bjäfs överdriven pejorative<br />
bjällra kl<strong>in</strong>gande sound<br />
bjärt lysande light<br />
bjässe mycket stor augmentative<br />
black ljus, färg light<br />
bladdra prata strunt pejorative, talk<strong>in</strong>g<br />
blaffa större fläck m<strong>in</strong>dr tilltalan pejorative<br />
blaj men<strong>in</strong>gslöst prat pejorative, talk<strong>in</strong>g<br />
blank glansig yta light<br />
blarr struntprat pejorative, talk<strong>in</strong>g<br />
blaska röra i vatten plaska dålig pejorative, wetness<br />
blek färg light<br />
blemma blåsa round form<br />
blick seende gaze<br />
bliga titta envetet, dumt, fånstirra pejorative, gaze<br />
bl<strong>in</strong>d se gaze<br />
bl<strong>in</strong>ka ögonen gaze<br />
blixt ljus light<br />
block massivt stycke shortwide form<br />
blod vätska wetness<br />
blond ljus färg light<br />
blossa br<strong>in</strong>na, låga light<br />
bluddra struntprat pejorative, talk<strong>in</strong>g<br />
blunda ögonen gaze<br />
blunder misstag pejorative<br />
blåsa rundad round form<br />
bläck vätska färgad wetness<br />
bläddra blemma, blåsa round form<br />
blända ljus light<br />
blänga betrakta gaze<br />
blänka ljus light<br />
bläs ljus strimma light<br />
blöt vatten wetness<br />
braka ljud sound<br />
brassa skjuta , stora krafter sound<br />
brista sönder destruction<br />
brum ljud sound<br />
brus ljud sound<br />
bryta kraftig, spricka destruction<br />
bråte förbrukat, oanvändbart destruction<br />
bräck krosskada destruction<br />
bräka ljud sound<br />
bränn<strong>in</strong>g splitras destruction<br />
bröl ljud sound<br />
Appendix 1 Sida 1
drasut nedsättande pejorative<br />
dratta slarv, fumlig fall<strong>in</strong>g, pejorative<br />
dravel struntprat talk<strong>in</strong>g, pejorative<br />
dregel saliv, oavsiktligt wetness, pejorative<br />
dribbla rörelser movement<br />
drill ton sound<br />
droppa falla fall<strong>in</strong>g, wetness<br />
drulle ohyfsat, vårdslöst pejorative<br />
drummel klumpigt, vårdslöst pejorative<br />
dråsa falla klumpigt fall<strong>in</strong>g<br />
drägg botten, slödder pejorative<br />
drälla planlöst, vårdslöst, förflytta fall<strong>in</strong>g, pejorative<br />
drämma slå hårt beat<br />
dröse osorterad, ostrukturerad mängd pejorative<br />
drösa falla, tätt, okontrollerat fall<strong>in</strong>g<br />
fjant åtlöje pejorative<br />
fjolla saknar stadga i karaktär pejorative<br />
fjompig löjlig pejorative<br />
fjuttig futtig, torftig pejorative<br />
fjärta ljud, ofrivilligt sound<br />
fjäsa göra sig till pejorative<br />
fjäsk överdriven pejorative<br />
flabb skratt sound, pejorative<br />
flacka planlöst fara q or s movement, pejorative<br />
fladdra rörelse q or s movement<br />
flaga tunt stycke th<strong>in</strong> form<br />
flagga vifta q or s movement<br />
flak flat form, större stycke th<strong>in</strong> form<br />
flamma stark eld, orolig, plötslig light, q or s movement<br />
flamsa bullersam, tröttsamt, vårdslöst, stojigt pejorative<br />
flanera vandra utan bestämt mål q or s movement<br />
flarn bräckligt stycke poröst th<strong>in</strong> form<br />
flaxa slå häftigt, flygande q or s movement<br />
flimra dallrande ljus q or s movement, light<br />
fl<strong>in</strong>a m<strong>in</strong>dre vackert, försmädligt pejorative<br />
fl<strong>in</strong>ga litet lätt tunt stycke th<strong>in</strong> form<br />
fl<strong>in</strong>k snabbt quickness<br />
flisa litet tunt vasst stycke th<strong>in</strong> form<br />
flitter oäkta, värdelös pejorative<br />
flopp snöplig, misslyckande pejorative<br />
floskel högtravande <strong>in</strong>tetsäg. löjlig pejorative<br />
fluffa lätt luftig ruska airy consistency<br />
fluga flygande q or s movement<br />
flukta kasta snabb blick gaze<br />
fluktuation ständigt stigande och fallande q or s movement<br />
flummig <strong>in</strong>tellektuellt oredig pejorative<br />
flyga förflytta sig q or s movement<br />
fly hastigt lämna q or s movement<br />
Appendix 1 Sida 2
flyta transporteras q or s movement<br />
flytta till annan plats q or s movement<br />
flåsa tungt snabbt q or s movement<br />
fläkta blåsa q or s movement<br />
flämta snabbt häftigt q or s movement<br />
flänga rusa framochtillbaka hetsigt q or s movement<br />
flöda r<strong>in</strong>na stor mängd q or s movement<br />
flöjel v<strong>in</strong>driktn<strong>in</strong>g lätt skiva th<strong>in</strong> form<br />
flöjt ton luftström skarp kant sound<br />
fnasig fjälla th<strong>in</strong> form<br />
fnask nedsättande pejorative<br />
fnatta spr<strong>in</strong>ga planlöst q or s movement, pejorative<br />
fnissa skratta sound<br />
fnittra skratta sound<br />
fnoskig tokig pejorative<br />
fnurra osams pejorative<br />
fnysa häftigt ljudligt förakt sound<br />
fnöske porös dryness<br />
fradga skummande wetness<br />
frasa ljud sound<br />
frat söndergnagt skräp destruction<br />
frossa skakn<strong>in</strong>g köldkänslor q or s movement<br />
frotté poröst rough surface structure<br />
frusa häftigt välla fram q or s movement<br />
frusta ljud sound<br />
fryna rynka rough surface structure<br />
fräsa ljud sound<br />
fräta upplösa destruction<br />
glacé glänsande light<br />
glací jämn yta smooth surface str<br />
glaciär is smooth surface str<br />
glam högljudd sound<br />
glana stirra gaze<br />
glans jämnt sken blank yta light<br />
glas glänsande smooth surface str<br />
glatt slät blank hal smooth surface str<br />
glaubersalt glas smooth surface str<br />
glaukom syn gaze<br />
gli litet dim<strong>in</strong>utive<br />
glida jämnt smooth surface str<br />
glimma lysa svagt skiftande light<br />
glimmer glänsande light<br />
glimra lysa svagt skiftande light<br />
glimt blänk ljus light<br />
gl<strong>in</strong>der glänsande light<br />
gl<strong>in</strong>dra glimra light<br />
glipa smalt form, dim<strong>in</strong>utive<br />
Appendix 1 Sida 3
glisa lysa light<br />
glissando glida ton sound, smooth surface<br />
glittra lysa starkt snabbt light<br />
glo stirra dumt gaze<br />
glob klotformig form<br />
glop slyngel pejorative<br />
glopp snö regn wetness<br />
gloria ljus light<br />
glosögd ögon gaze<br />
glosa glo gaze<br />
glufsa mycket ovårdat pejorative<br />
glugg liten form, dim<strong>in</strong>utive<br />
glunkas sägas skvaller pejorative<br />
glupa glupskt pejorative<br />
glupsk nedsätt pejorative<br />
glutta titta gaze<br />
glytt barnunge dim<strong>in</strong>utive<br />
glåmig blek light<br />
glåpord förolämpande yttrande pejorative<br />
gläfs svagt hundskall sound<br />
glänsa sken blank light<br />
glänta m<strong>in</strong>dre öppen form, dim<strong>in</strong>utive<br />
glöd sken light<br />
gnabb smågräl talk<strong>in</strong>g<br />
gnat småaktig upprepad talk<strong>in</strong>g, pejorative<br />
gneta småsaker, oväsentlig, snål pejorative<br />
gnida upprepade rörelser smooth surface str, q or s mo<br />
gnidare snål pejorative<br />
gnissla ljud sound<br />
gnista br<strong>in</strong>nande glödande light<br />
gnistra korta starka ljusglimtar light<br />
gno små snabba rörelser smooth surface str, q or s mo<br />
gnod<strong>in</strong>g knorrhane sound<br />
gnola sjunga talk<strong>in</strong>g<br />
gnugga små korta rörelser smooth surface str, q or s mo<br />
gnutta mycket liten dim<strong>in</strong>utive<br />
gny ljud talk<strong>in</strong>g<br />
gnägga kraftigt läte sound<br />
gnälla läte sound<br />
gnöla klaga talk<strong>in</strong>g<br />
gramse arg irriterad bad mood<br />
grav grävd hålighet i marken hollow form<br />
grift grav hollow form<br />
groll ovänskap bad mood<br />
grop djup hålighet hollow form<br />
grotta håltum hollow form<br />
grubba grop hollow form<br />
grubbla dyster bad mood<br />
Appendix 1 Sida 4
gruff gräl bad mood, talk<strong>in</strong>g<br />
grumlig ogenomsk<strong>in</strong>lig oönskad pejorative<br />
grummel grums pejorative<br />
grums oönskad pejorative<br />
grumsa klagomål pejorative, talk<strong>in</strong>g<br />
grunka föremål vard slang<br />
grym samvetslöst tillfogar lidande bad mood<br />
grymta ljud talk<strong>in</strong>g<br />
gryt jordhåla hollow form<br />
gråta röst talk<strong>in</strong>g<br />
gräla högljudd respektlös diskussion bad mood, talk<strong>in</strong>g<br />
gräll syn alltför starkt light, pejorative<br />
gräma oönskad pejorative<br />
gräslig stark olust pejorative<br />
gräva djupare grop hollow form<br />
gröpa gräva m<strong>in</strong>dre hål hollow form<br />
klabb kort tjockt stycke shortwide form<br />
klabb blöt wetness<br />
klack utskjutande, del shortwide form<br />
kladd blöt vidhäftande wetness<br />
klafs ljud våt klibbig sound, wetness<br />
klaga framföra missnöje talk<strong>in</strong>g<br />
klamp grovt stycke shortwide form<br />
klamp ljud sound<br />
klamra gripa adhesion<br />
klandra ogillande yttra talk<strong>in</strong>g<br />
klang toner sound<br />
klanka upprepat förebrå småaktigt talk<strong>in</strong>g, pejorative<br />
klanta dum klumpig pejorative<br />
klappa upprepat slå sound<br />
klappra ljud sound<br />
klatsch ljud sound<br />
klena kladdig wetness<br />
kleta blöt vidhäftande wetness, adhesion<br />
klibba kladdig wetness, adhesion<br />
klick klump smetig wetness, shortwide form<br />
klick ljud sound<br />
klifs ljud blött sound, wetness<br />
klimp m<strong>in</strong>dre stycke shortwide form<br />
kl<strong>in</strong>g ljud sound<br />
kl<strong>in</strong>k spel sound<br />
klirra ljud sound<br />
klister klibbig adhesion<br />
klocka klang sound<br />
kloss grovt stycke shortwide form<br />
klot runt shortwide form<br />
klotter slarvig pejorative<br />
klubba kloss shortwide form<br />
klucka ljud sound, wetness<br />
Appendix 1 Sida 5
kludda dålig slarvig pejorative<br />
klump oformligt större stycke shortwide form<br />
klunk vätska wetness<br />
kluns klumpig pejorative<br />
klåpa dålig slarvig pejorative<br />
klämta r<strong>in</strong>ga sound<br />
klänga hålla kvar adhesion<br />
kläpp ljud, klump sound, shortwide form<br />
klätt liten topp shortwide form<br />
klättra fästpunkt adhesion<br />
knacka upprepade slag lystr<strong>in</strong>g sound<br />
knackig dålig pejorative<br />
knagg p<strong>in</strong>ne round form<br />
knaggla röra sig walk<strong>in</strong>g<br />
knaka ljud sound<br />
knal svag knapp pejorative<br />
knall ljud sound<br />
knalla gå långsamt vard walk<strong>in</strong>g<br />
knalle liten rund höjd round form<br />
knapp skiva, kula round form<br />
knapra knastrande sound<br />
knarr ljud sound<br />
knarrig vresig bad mood<br />
knasig tokigt dum pejorative<br />
knast utväxt round form<br />
knastra ljud sound<br />
knata spr<strong>in</strong>ga vard walk<strong>in</strong>g<br />
knatte liten dim<strong>in</strong>utive<br />
knatter ljud sound<br />
kneg nedsättande pejorative<br />
kneken förfallet pejorative<br />
knick tvär böj round form<br />
knipa ihop putt<strong>in</strong>g together<br />
knippa hophållna putt<strong>in</strong>g together<br />
knipsa nypa putt<strong>in</strong>g together<br />
knirka ljud sound<br />
knirra ljud sound<br />
knistra ljud sound<br />
knittra ljud sound<br />
knixa liten hastig q or s movement<br />
knocka slå beat<br />
knodd nedsättande pejorative<br />
knoga mödosamt gå walk<strong>in</strong>g<br />
knoge upphöjn<strong>in</strong>g round form<br />
knollra smålockig round form<br />
knop knut round form<br />
knopp liten kula round form<br />
knorr liten vriden form round form<br />
knorr muttrande kurrande sound<br />
Appendix 1 Sida 6
knorva veckig rynkig pejorative<br />
knot knorrhane sound<br />
knota muttrande klaga talk<strong>in</strong>g<br />
knota förtjockad round form<br />
knott liten mygga dim<strong>in</strong>utive<br />
knottra liten upphöjn<strong>in</strong>g round form<br />
knubbig kort tjock round form<br />
knuff stöt beat<br />
knut sammanfogn<strong>in</strong>g putt<strong>in</strong>g together<br />
knutte nedsätt pejorative<br />
knyck plötslig rörelse q or s movement<br />
knyckla ihop putt<strong>in</strong>g together<br />
knyppla flätas putt<strong>in</strong>g together<br />
knyst ljud sound<br />
knyta samman putt<strong>in</strong>g together<br />
knåp smått obetydligt dim<strong>in</strong>utive<br />
knäck brott slag beat<br />
knäppa ljud sound<br />
knödel bulle round form<br />
knöl m<strong>in</strong>dre rundad round form<br />
knöl otrevlig pejorative<br />
knös nedsätt pejorative<br />
krabat litet dim<strong>in</strong>utive<br />
kracka sönderdela destruction<br />
krackelera f<strong>in</strong>a sprickor destruction<br />
krafs värdelösa småsaker pejorative<br />
krafsa ljud sound<br />
krake beklagansvärd pejorative<br />
krakel högljutt sound<br />
krakmandel tunt bräckligt skal th<strong>in</strong> form<br />
kram värdelösa pejorative<br />
kram våt wetness<br />
krams värdelösa småsaker pejorative<br />
krans r<strong>in</strong>g sammanflätad w<strong>in</strong>d<strong>in</strong>g form<br />
kras ljud sound<br />
krasch ljud sound<br />
kratsa rörelse q or s movement<br />
kratta fåra rough surface structure<br />
kratta dålig pejorative<br />
kravla mödosam förflytta walk<strong>in</strong>g<br />
kraxa ljud sound<br />
kreta karva dålig pejorative<br />
krimskrams värdelösa småsaker pejorative<br />
kr<strong>in</strong>gelikrokar mängd krökar w<strong>in</strong>d<strong>in</strong>g form<br />
kr<strong>in</strong>gla form av en båge w<strong>in</strong>d<strong>in</strong>g form<br />
krock häftig sammanstötn<strong>in</strong>g beat<br />
krok långsmalt halvcirkel w<strong>in</strong>d<strong>in</strong>g form<br />
krokan krusiduller, snirklig w<strong>in</strong>d<strong>in</strong>g form<br />
krokett frasig th<strong>in</strong> form<br />
Appendix 1 Sida 7
krossa sönder destruction<br />
kruka cyl<strong>in</strong>drisk bukig round form<br />
krulla små lockar w<strong>in</strong>d<strong>in</strong>g form<br />
krum böjd w<strong>in</strong>d<strong>in</strong>g form<br />
krumbukt kraftfull sl<strong>in</strong>grande rörelse w<strong>in</strong>d<strong>in</strong>g form<br />
krumelur krokig l<strong>in</strong>je w<strong>in</strong>d<strong>in</strong>g form<br />
krumpen böjd w<strong>in</strong>d<strong>in</strong>g form<br />
krusa full av små vågor w<strong>in</strong>d<strong>in</strong>g form<br />
kruserlig tillgjord pejorative<br />
krusiduller form sl<strong>in</strong>grande l<strong>in</strong>jer w<strong>in</strong>d<strong>in</strong>g form<br />
krusta skal th<strong>in</strong> form<br />
krustad bakverk i form av bägare th<strong>in</strong> form<br />
krycka böjt w<strong>in</strong>d<strong>in</strong>g form<br />
krylla rörlig walk<strong>in</strong>g<br />
krypa förflytta sig walk<strong>in</strong>g<br />
krysta samtal, formuler<strong>in</strong>g sound<br />
kråka fågel sound<br />
kråma stolt vrida w<strong>in</strong>d<strong>in</strong>g form<br />
krångel besvärligt pejorative<br />
krås veckad remsa rough surface str<br />
kräk klandervärd pejorative<br />
kräkla <strong>in</strong>rullat krön w<strong>in</strong>d<strong>in</strong>g form<br />
kräla förflytta sig sl<strong>in</strong>grande walk<strong>in</strong>g, w<strong>in</strong>d<strong>in</strong>g form<br />
kräm tjockflytande wetness<br />
krämta harkla sound<br />
kränga vickande rörelser q or s movement<br />
kräpp tunt krusig yta rough surface str<br />
krök kraftig sväng w<strong>in</strong>d<strong>in</strong>g form<br />
kvacka läte sound<br />
kvadda förstöra destruction<br />
kval lidande bodily feel<strong>in</strong>g<br />
kvalm osund äckel bodily feel<strong>in</strong>g<br />
kvarka sound<br />
kvarn nedbrytande destruction<br />
kvav kvalmig kväva bodily feel<strong>in</strong>g<br />
kverulans klaga talk<strong>in</strong>g<br />
kvick små snabba rörelser q or s movement<br />
kvida ljud sound<br />
kvillra ljud sound<br />
kvirra kverulera sound<br />
kvissla liten blåsa dim<strong>in</strong>utive<br />
kvist liten smal dim<strong>in</strong>utive<br />
kvitter läte sound<br />
kväda sjunga sound<br />
kväka läte sound<br />
kvälja lukt äckla bodily feel<strong>in</strong>g<br />
kväva svårighet att andas bodily feel<strong>in</strong>g<br />
mjau läte sound<br />
Appendix 1 Sida 8
mjuk angenäm känsel jämn slät soft consistency<br />
mjäkig alltför vek pejorative<br />
mjäla mjölf<strong>in</strong> f<strong>in</strong>e gra<strong>in</strong><br />
mjäll f<strong>in</strong> konsistens soft consistency<br />
mjäll fjäll f<strong>in</strong>e gra<strong>in</strong><br />
mjöl f<strong>in</strong>mald f<strong>in</strong>e gra<strong>in</strong><br />
njugg onödigt snål pejorative<br />
pjatt obetydlig hålln<strong>in</strong>gslös pejorative<br />
pjoller men<strong>in</strong>gslöst prat pejorative, talk<strong>in</strong>g<br />
pjosk överdriven pejorative<br />
pjunk gnäll veklighet pejorative<br />
pjåkig dålig pejorative<br />
pladask klumpigt ljudligt fall sound<br />
pladder oavbrutet <strong>in</strong>nehållslöst prat talk<strong>in</strong>g<br />
pladuska stor utbredn<strong>in</strong>g oönskad pejorative<br />
planka fyrsidigt stycke form<br />
plask häftigt vatten ljud sound, wetness<br />
platt ytor form<br />
pligg kort spetsigt stift form, dim<strong>in</strong>utive<br />
pl<strong>in</strong>g ljud sound<br />
pl<strong>in</strong>t lådformigt avsmalnande form<br />
plira titta halvt tillslutna gaze<br />
plissé regelbunden f<strong>in</strong> veckn<strong>in</strong>g rough surface structure<br />
plit starkt vard. slang<br />
ploj skämtsamt spratt slang<br />
plomb bly, fylln<strong>in</strong>g form<br />
plottra spridda oväsentliga smådetaljer pejorative<br />
plufsig slappt fet pejorative<br />
plugg litet cyl<strong>in</strong>driskt form, dim<strong>in</strong>utive<br />
plump fläck pejorative<br />
plump grov ohyfsad pejorative<br />
plums ljud sound, wetness<br />
plundra hänsynslöst pejorative<br />
plunta vard. slang<br />
plur<strong>in</strong>g starkt vard slang<br />
plurr vatten wetness<br />
plussig uppsvälld pejorative<br />
plutt liten dim<strong>in</strong>utive<br />
plym fjäder soft consistency<br />
plymå dyna soft consistency<br />
plysa luckra upp ull soft consistency<br />
plysch sammetsliknande långhårigt soft consistency<br />
plätt litet runt form, dim<strong>in</strong>utive<br />
plös kilformigt form<br />
pracka lura pejorative<br />
prassel ljud sound<br />
Appendix 1 Sida 9
prat tal talk<strong>in</strong>g<br />
prick mycket liten rundad platt form, dim<strong>in</strong>utive<br />
prilla portion slang<br />
prillig småtokig slang<br />
propp föremål täppa hål öppn<strong>in</strong>g form<br />
propsa envist kräva talk<strong>in</strong>g<br />
prunka lysande färger light<br />
pruta övertala talk<strong>in</strong>g<br />
prutt fjärt sound<br />
pryl grovt nålformigt form<br />
pryl onyttig överflödig pejorative<br />
pryttel slang<br />
prål strålande light<br />
prång liten trång form, dim<strong>in</strong>utive<br />
prångla tvivelaktigt knep pejorative<br />
pröjs vard. slang<br />
pröjsare vard slang<br />
skaka rörelse q or s movement<br />
skakel stång long th<strong>in</strong> form<br />
skal hårt tunt hölje hardness<br />
skalk hård kant hardness<br />
skall läte sound<br />
skalla ljuda starkt sound<br />
skalle hårdhet hardness<br />
skallra ljud sound<br />
skalm avlång rörlig long th<strong>in</strong> form<br />
skalpell rak long th<strong>in</strong> form<br />
skalv darra q or s movement<br />
skare hårt ytskikt hardness<br />
skarp spetsig kant hardness<br />
skava hårt tryck föras upprepat hardness<br />
skavank fel pejorative<br />
skippa hoppa q or s movement<br />
skopa m<strong>in</strong>dre halvklotformigt round form<br />
skorpa hårdnad yta hardness<br />
skorr ljud sound<br />
skorvig yta hård ojämn hardness<br />
skorv gammal förfallen pejorative<br />
skott ljud sound<br />
skovel stor bred rundad round form<br />
skångra skaka starkt ljud sound<br />
skåra långsträckt fördjupn<strong>in</strong>g long th<strong>in</strong> form<br />
skrabba gammal utsliten destruction<br />
skraj vard. slang<br />
skral dålig pejorative<br />
skraltig dålig svag destruction<br />
skramla ljud sound<br />
skranglig ostadig destruction<br />
skrap ljud sound<br />
Appendix 1 Sida 10
skratta ljud talk<strong>in</strong>g<br />
skri läte starkt talk<strong>in</strong>g<br />
skrocka läte talk<strong>in</strong>g<br />
skrodera skryta talk<strong>in</strong>g, pejorative<br />
skrot skräp värdelöst destruction<br />
skrovlig grov ojämn yta rough surface str<br />
skrubb förvar<strong>in</strong>gsutrymme pejorative<br />
skrubba gnida hårt grov ojämn rough surface str<br />
skrubbor bannor talk<strong>in</strong>g<br />
skrumpen förtorkad rynkig rough surface structure<br />
skrutt dåligt pejorative<br />
skrymsle litet utrymme dim<strong>in</strong>utive<br />
skrymt spökeri troll pejorative<br />
skrynkla oönskat veck rough surface structure<br />
skryta tala överdrivet talk<strong>in</strong>g<br />
skrål pratande talk<strong>in</strong>g<br />
skråma m<strong>in</strong>dre ytligt sår rough surface structure<br />
skrälla ljud sound<br />
skrälle gammalt nedslitet destruction<br />
skräna röster talk<strong>in</strong>g<br />
skräp värdelös destruction<br />
skräppa skryta talk<strong>in</strong>g, pejorative<br />
skrävla skryta talk<strong>in</strong>g, pejorative<br />
skröna lögnaktig talk<strong>in</strong>g, pejorative<br />
skröplig svag ömtålig destruction<br />
skval ljud vatten sound, wetness<br />
skvaller löst prat talk<strong>in</strong>g, pejorative<br />
skvalp ljud sound<br />
skvalta stänka wetness<br />
skvimpa skvalpa små rörelser movement, wetness<br />
skvätt liten mängd vätska wetness<br />
slabba vätska slarv wetness, pejorative<br />
slabbertacka sladdertacka talk<strong>in</strong>g, pejorative<br />
slack alltför stor slakhet slackness<br />
sladd tamp long th<strong>in</strong> form<br />
sladda okontrollerat glida q or s movement<br />
sladdra o<strong>in</strong>tressant skvaller talk<strong>in</strong>g, pejorative<br />
sladdrig saknar styvhet slackness<br />
slafs kladdande wetness, pejorative<br />
slag hård träff ljud ton beat<br />
slak mjuk böjlig slackness<br />
slam vätska wetness<br />
slammer starkt buller sound<br />
slampa lösaktig slarvig pejorative<br />
slams värdelöst prat pejorative<br />
slamsa smal sladdrig remsa slackness, long th<strong>in</strong> form<br />
slana smal böjlig slackness, long th<strong>in</strong> form<br />
slang långt böjligt slackness, long th<strong>in</strong> form<br />
slank smal long th<strong>in</strong> form<br />
Appendix 1 Sida 11
slankig alltför mjuk böjlig slackness<br />
slant långt, spö long th<strong>in</strong> form<br />
slapp alltför mjuk slackness<br />
slarvig <strong>in</strong>te noggrann och ordentlig pejorative<br />
slas slö slarvig pejorative<br />
slasa gå tungt och hasande pejorative, walk<strong>in</strong>g<br />
slask blöt sörja sopor wetness, pejorative<br />
slatt bottenskvätt wetness, pejorative<br />
slattrig slapp, pratig talk<strong>in</strong>g, pejorative<br />
slejf band long th<strong>in</strong> form<br />
slem segt sekret wetness<br />
slicka flytande kletigt wetness<br />
slidder o<strong>in</strong>tressant skvaller talk<strong>in</strong>g, pejorative<br />
sliddrig sladdrig slackness<br />
slimmad <strong>in</strong>sydd long th<strong>in</strong> form<br />
sl<strong>in</strong>ga krök v<strong>in</strong>dl<strong>in</strong>g long th<strong>in</strong> form<br />
sl<strong>in</strong>ka förflytta sig snabbt q or s movement<br />
sl<strong>in</strong>ka prostituerad pejorative<br />
sl<strong>in</strong>kig sladdrig slackness<br />
sl<strong>in</strong>ta plötsligt glatthet q or s movement<br />
slipa yta slät gnida smooth surface str<br />
slipprig hal kladdig wetness, q or s movement<br />
slips band long th<strong>in</strong> form<br />
slira okontrollerat glida q or s movement<br />
slisk äckligt kladdigt wetness, pejorative<br />
slita häftigt rycka q or s movement<br />
slits smal öppen ränna long th<strong>in</strong> form<br />
slok odåga pejorative<br />
sloka slapp avlång böjlig slackness<br />
sludder otydligt tal talk<strong>in</strong>g, pejorative<br />
slugga boxas vilt q or s movement<br />
slum förfallet pejorative<br />
slunga kasta stor rörelse q or s movement<br />
slur<strong>in</strong>g soppa wetness<br />
slurk liten vätska wetness, dim<strong>in</strong>utive<br />
slusk mycket smutsig ovårdad pejorative<br />
sluss vatten wetness<br />
slutta luta nedåt yta potential movement<br />
slyna ouppfostrad slarvig pejorative<br />
slyngel ouppfostrad pejorative<br />
slå slag beat<br />
släde medar long th<strong>in</strong> form<br />
slägga stor tung hammare beat<br />
slänga kasta vårdslöst häftigt q or s movement<br />
slänt sluttn<strong>in</strong>g potential movement<br />
släntra gå långsamt utan mål walk<strong>in</strong>g<br />
slät fri från ojämnhet yta smooth surface str<br />
slätt vidsträckt plant smooth surface str<br />
slö kraftlös håglös pejorative<br />
slödder föraktad ouppfostrad pejorative<br />
Appendix 1 Sida 12
slösa för mycket misshushålla pejorative<br />
smack ljud sound<br />
smal r<strong>in</strong>ga bredd narrow form<br />
smash snabbt hårt slag beat<br />
smask ljud sound<br />
smatt prång dim<strong>in</strong>utive<br />
smatter ljud sound<br />
smegma sekret wetness<br />
smet kladdig röra wetness<br />
smetana tjock grädde wetness<br />
smicker överdrivet pejorative<br />
sm<strong>in</strong>k krämig konsistens wetness<br />
smisk aga beat<br />
smita avlägsna sig obehagligt secrecy<br />
smock rynkn<strong>in</strong>g i rutmönster rough surface str<br />
smocka hårt slag beat<br />
smolk smutspartikel pejorative<br />
smuggla olovligt secrecy<br />
smula liten partikel torr lös dim<strong>in</strong>utive<br />
smussel i smyg olovligt secrecy<br />
smuts fläckar oönskade pejorative<br />
smutt liten klunk dim<strong>in</strong>utive<br />
smutt prång dim<strong>in</strong>utive<br />
smyga röra sig tyst och smidigt secrecy<br />
små r<strong>in</strong>ga utsträckn<strong>in</strong>g dim<strong>in</strong>utive<br />
smäck slag med handflata beat<br />
smäcker slank elegant narrow form<br />
smäda omdöme talk<strong>in</strong>g<br />
smäkta <strong>in</strong>tensiv längtan mental feel<strong>in</strong>g<br />
smäll aga beat<br />
smälta flytande wetness<br />
smärt smal smidig narrow form<br />
smärta plågsam känsla mental feel<strong>in</strong>g<br />
smätta ljud sound<br />
smörj stryk beat<br />
smörja mjukt kräm wetness<br />
snabb quickness<br />
snabel mycket lång utdragen long th<strong>in</strong> form<br />
snacka prata talk<strong>in</strong>g<br />
snafs smuts pejorative<br />
snagga mycket kort small end form<br />
snappa kort snabb rörelse quickness<br />
snar snabb quickness<br />
snarka ljud sound<br />
snarpa läte sound<br />
snarra skorra sound<br />
snaskig sölig kladdig wetness, pejorative<br />
snatta m<strong>in</strong>dre värde pejorative<br />
snattra läte sound<br />
Appendix 1 Sida 13
snava falla snubbla walk<strong>in</strong>g<br />
snegla titta ögon gaze<br />
snibb hörn trekantigt small end form<br />
snicksnack struntprat talk<strong>in</strong>g, pejorative<br />
(snigel) lång slemmig wetness, long th<strong>in</strong> form<br />
sniken girigt orättmätig pejorative<br />
snilja tråd long th<strong>in</strong> form<br />
snip spetsig form small end form<br />
snirkel starkt böjd l<strong>in</strong>je spiral w<strong>in</strong>d<strong>in</strong>g form<br />
snitsel smal remsa long th<strong>in</strong> form<br />
snitt liten f<strong>in</strong>are dim<strong>in</strong>utive<br />
sno långsmalt long th<strong>in</strong> form<br />
sno snabbt quickness<br />
snobb överdrivet fåfängt pejorative<br />
snodd garn long th<strong>in</strong> form<br />
snofsa vard. slang<br />
snok långt smalt long th<strong>in</strong> form<br />
snopen besviken mental feel<strong>in</strong>g<br />
snopp smal ända small end form<br />
snor slem wetness<br />
snorkel rör long th<strong>in</strong> form<br />
snorkig oartig pejorative<br />
snubba tillrättavisn<strong>in</strong>g talk<strong>in</strong>g<br />
snubbe vard. slang<br />
snubbla falla walk<strong>in</strong>g<br />
snudig fl<strong>in</strong>k quickness<br />
snugga liten pipa dim<strong>in</strong>utive<br />
snultra långsträckt long th<strong>in</strong> form<br />
snurra vrida cirkelformig w<strong>in</strong>d<strong>in</strong>g form<br />
snusa hörbart andas sound<br />
snusk smuts illaluktande pejorative<br />
snut vard slang<br />
snutt liten kort dim<strong>in</strong>utive<br />
snuva slem wetness<br />
snyfta stötvisa andetag sound<br />
snylta ytnyttja pejorative<br />
snyta fräsa sound<br />
snål överdrivet pejorative<br />
snäcka spiralvridet w<strong>in</strong>d<strong>in</strong>g form<br />
snäppa ljud sound<br />
snärj högt tempo quickness<br />
snärp fågelläte sound<br />
snärt ända slag beat<br />
snärta nedsätt pejorative<br />
snäsa irriterad tillrättavisn<strong>in</strong>g talk<strong>in</strong>g<br />
snöd fåfäng tarvlig pejorative<br />
snöpa stumpa, vanställa pejorative<br />
snöre garn long th<strong>in</strong> form<br />
snörvla ljud sound<br />
Appendix 1 Sida 14
spackel degartat wetness<br />
spad vatten wetness<br />
spade platt th<strong>in</strong> form<br />
spagat åt var sitt håll separation<br />
spaljé spjälor long th<strong>in</strong> form<br />
spalt lodrät long th<strong>in</strong> form<br />
spana blicken gaze<br />
spant balk long th<strong>in</strong> form<br />
spark hård stöt beat<br />
sparre bjälke long th<strong>in</strong> form<br />
sparris stam long th<strong>in</strong> form<br />
spasm ryckn<strong>in</strong>g q or s movement<br />
spat klyvbart glansigt th<strong>in</strong> form, light<br />
spatel platt avlångt long th<strong>in</strong> form<br />
spatt tokig överdrift jocular<br />
spe illvilligt förlöjligande jocular<br />
spegel plan smooth surface str<br />
speja blicken gaze<br />
spektakel förargelse jocular<br />
spene utväxt long th<strong>in</strong> form<br />
spenslig tunn smal long th<strong>in</strong> form<br />
sperma vätska wetness<br />
speta litet tunt vasst long th<strong>in</strong> form<br />
spets tunn ända avsmalnande long th<strong>in</strong> form<br />
spett lång stång spets long th<strong>in</strong> form<br />
spex parodisk dråplig jocular<br />
spigg taggar long th<strong>in</strong> form<br />
spik spetsig p<strong>in</strong>ne long th<strong>in</strong> form<br />
spila spjäla long th<strong>in</strong> form<br />
spilkum liten skål dim<strong>in</strong>utive<br />
spill blir över pejorative<br />
spilla vätska wetness<br />
sp<strong>in</strong>del långa ben long th<strong>in</strong> form<br />
sp<strong>in</strong>kig mycket tunn long th<strong>in</strong> form<br />
sp<strong>in</strong>na garn long th<strong>in</strong> form<br />
sp<strong>in</strong>na ljud sound<br />
spira smal stav long th<strong>in</strong> form<br />
spiral kurva vriden w<strong>in</strong>d<strong>in</strong>g form<br />
spola vatten wetness<br />
spole avlångt long th<strong>in</strong> form<br />
spol<strong>in</strong>g ouppfostrad pejorative<br />
spont utstående parti long th<strong>in</strong> form<br />
spotsk hånfullhet attitude<br />
spott vätska wetness<br />
spurt ökn<strong>in</strong>g av farten q or s movement<br />
sputum slem wetness<br />
spydig elaka kommentarer attitude, talk<strong>in</strong>g<br />
spån tunn avhyvlad bit th<strong>in</strong> form<br />
Appendix 1 Sida 15
spång smal long th<strong>in</strong> form<br />
späd liten dim<strong>in</strong>utive<br />
spänta stickor long th<strong>in</strong> form<br />
spö smal böjlig käpp long th<strong>in</strong> form<br />
spjut lång smal stång hård spets long th<strong>in</strong> form<br />
spjuver skämtar luras jocular<br />
spjäla ribba long th<strong>in</strong> form<br />
spjälka sönderdela destruction<br />
spl<strong>in</strong>ta klyva separation<br />
split oenighet separation<br />
splits ändar av tågvirke separation<br />
splitter små vass slagits sönder destruction, dim<strong>in</strong>utive<br />
spraka knastra gnistra sound, light<br />
spralla kroppsligen livlig q or s movement<br />
spratt skämtsam lura jocular<br />
sprattla rörelser q or s movement<br />
sprej f<strong>in</strong>fördelat vätska wetness, separation<br />
spreta åt olika håll separation<br />
spri smäckert rundhult long th<strong>in</strong> form<br />
spricka långsmalt brott long th<strong>in</strong> form<br />
sprida fördela utbredn<strong>in</strong>g separation<br />
spr<strong>in</strong>ga lång mycket smal öppn<strong>in</strong>g long th<strong>in</strong> form<br />
spr<strong>in</strong>ga förflytta sig walk<strong>in</strong>g<br />
spr<strong>in</strong>kler vatten, sprider wetness, separation<br />
spr<strong>in</strong>t p<strong>in</strong>ne long th<strong>in</strong> form<br />
sprits strut form long th<strong>in</strong> form<br />
spritta plötsligt rycka q or s movement<br />
sprudla bubblande välla fram q or s movement, wetness<br />
sprund smal öppn<strong>in</strong>g long th<strong>in</strong> form<br />
spruta häftigt stråle vätska q or s movement, wetness<br />
sprutt fart vard. q or s movement<br />
spränga stor kraft splittras separation<br />
sprätt överdrivet pejorative<br />
sprätta kr<strong>in</strong>gkastande separation<br />
spröjs list long th<strong>in</strong> form<br />
spröt långt, smalt long th<strong>in</strong> form<br />
stabbig kort kraftig short form<br />
stake lång käpp long th<strong>in</strong> form<br />
stamma tala stötigt talk<strong>in</strong>g, iterative<br />
stampa stöta hårt ljudligt sound, iterative<br />
stappla gå ostadig walk<strong>in</strong>g<br />
stav långt smalt long th<strong>in</strong> form<br />
stepp ljudliga slag sound, beat<br />
stick smärtsamt vass bodily feel<strong>in</strong>g<br />
sticka tunn vass flisa long th<strong>in</strong> form<br />
Appendix 1 Sida 16
stift smalt spetsigt long th<strong>in</strong> form<br />
stig smal long th<strong>in</strong> form<br />
stilett stickvapen smal long th<strong>in</strong> form<br />
stim ljud röster talk<strong>in</strong>g<br />
st<strong>in</strong>g smärtsam vass bodily feel<strong>in</strong>g<br />
st<strong>in</strong>t blick gaze<br />
stirra titta gaze<br />
stirrig upprörd virrig mental feel<strong>in</strong>g<br />
stock stam long th<strong>in</strong> form<br />
stoft små partiklar f<strong>in</strong>e gra<strong>in</strong><br />
stoj högljutt prat talk<strong>in</strong>g<br />
stolle tok pejorative<br />
stolpe påle long th<strong>in</strong> form<br />
stubb uppstickande strån short form<br />
stubbe uppstickande rest short form<br />
studsa rörelse q or s movement, iterative<br />
stulta tulta walk<strong>in</strong>g<br />
stump återstående bit short form<br />
stumpa liten dim<strong>in</strong>utive<br />
stura tjura attitude<br />
stursk trotsig fräck attitude<br />
stybb f<strong>in</strong>fördelat spill f<strong>in</strong>e gra<strong>in</strong><br />
stycka hugga i bitar separation<br />
stylta stång long th<strong>in</strong> form<br />
stympa avhuggn<strong>in</strong>g short form<br />
styng smärtsam vasst bodily feel<strong>in</strong>g<br />
styv ej lätt böjs stiffness<br />
stång avlängt rakt long th<strong>in</strong> form<br />
stånka läte talk<strong>in</strong>g<br />
stänka vätska wetness<br />
stöddig stor kraftig alltför självsäker attitude, pejorative<br />
stön ljud sound<br />
stöppla stöta q or s movement<br />
stör stång long th<strong>in</strong> form<br />
stöt rörelse kraftigt q or s movement<br />
stötta stolpe long th<strong>in</strong> form<br />
stöv hudavlagr<strong>in</strong>g f<strong>in</strong>e gra<strong>in</strong><br />
strak rak stel stiffness<br />
stram spänd stiffness<br />
streamer långsmal remsa long th<strong>in</strong> form<br />
streck kortare l<strong>in</strong>je long th<strong>in</strong> form<br />
stretch töjbar long th<strong>in</strong> form<br />
strigel läderrem long th<strong>in</strong> form<br />
stril ljud av vätska long th<strong>in</strong> form, sound, wetness<br />
strimla liten långsmal bit long th<strong>in</strong> form<br />
strimma långsmalt band ljus long th<strong>in</strong> form, light<br />
stripa rak hårtest long th<strong>in</strong> form<br />
stropp ögleformad round form<br />
Appendix 1 Sida 17
strosa lugnt promenera walk<strong>in</strong>g<br />
strul besvärligt pejorative<br />
strunt värdelöst pejorative<br />
strut konformigt long th<strong>in</strong> form<br />
strutta gå knyckigt walk<strong>in</strong>g<br />
strå stjälk long th<strong>in</strong> form<br />
stråk bandliknande long th<strong>in</strong> form<br />
stråle ljus long th<strong>in</strong> form, light<br />
stråt väg bana long th<strong>in</strong> form<br />
sträcka längd long th<strong>in</strong> form<br />
sträng spänd tråd long th<strong>in</strong> form<br />
sträv yta ojämnhet rough surface str<br />
strö f<strong>in</strong>fördelat material f<strong>in</strong>e gra<strong>in</strong><br />
ström r<strong>in</strong>nande vatten long th<strong>in</strong> form, wetness<br />
strössel avlånga korn f<strong>in</strong>e gra<strong>in</strong>, long th<strong>in</strong> form<br />
ströva långsamt gå walk<strong>in</strong>g<br />
svabba tvätta långa garnändar wetness, long th<strong>in</strong> form<br />
svacka större fördjupn<strong>in</strong>g bent form<br />
svada tala länge talk<strong>in</strong>g<br />
svaja svänga fram och tillbaka q or s movement<br />
svalla röra sig häftigt q or s movement<br />
svamla prata strunt talk<strong>in</strong>g. pejorative<br />
svank <strong>in</strong>åtböjt bent form<br />
svans långsmal long th<strong>in</strong> form<br />
svassa tillgjort pejorative, walk<strong>in</strong>g<br />
svepa rörelse q or s movement<br />
svett vätska wetness<br />
svicka tapp short form<br />
svida brännande smärta bodily feel<strong>in</strong>g<br />
svikt böjlighet bent form<br />
sv<strong>in</strong>g rörelse q or s movement<br />
svirvel vridbar q or s movement<br />
svischa ljud sound<br />
svulstig alltför pejorative<br />
svämma vatten q or s movement, wetness<br />
sväng rörelse q or s movement<br />
svära yttra talk<strong>in</strong>g<br />
svärm flygande m<strong>in</strong>dre q or s movement<br />
sväva rörelse q or s movement<br />
tradig långtråkig pejorative<br />
traggel ständig tröttsam upprepn<strong>in</strong>g pejorative<br />
trampa rörelse walk<strong>in</strong>g<br />
trams dumt prat pejorative, talk<strong>in</strong>g<br />
trasa sönderrivet destruction<br />
traska gå vårdslöst klumpigt walk<strong>in</strong>g, pejorative<br />
trassel virrvarr krångel pejorative<br />
tratt strutformig pip tratt long th<strong>in</strong> form<br />
trava röra sig walk<strong>in</strong>g<br />
Appendix 1 Sida 18
tredsk envis ovillighet bad mood<br />
trilla falla fall<strong>in</strong>g<br />
trilla forma runt round form<br />
trilsk enveten motsträvig bad mood<br />
tr<strong>in</strong>d klotrund form round form<br />
tripp kortare dim<strong>in</strong>utive<br />
trippa gå små lätta hoppande walk<strong>in</strong>g, dim<strong>in</strong>utive<br />
trips liten dim<strong>in</strong>utive<br />
trissa liten rund round form, dim<strong>in</strong>utive<br />
trist enformig nedstämd bad mood<br />
troll ful klumpig enfaldig pejorative<br />
tromb häftig luftvirvel q or s movement<br />
tromla roterande cyl<strong>in</strong>drisk q or s movement<br />
tross l<strong>in</strong>a long h<strong>in</strong> form<br />
trubbel besvär problem bad mood<br />
trubbig <strong>in</strong>te avsmalnande short form<br />
trudelutt glad melodi(stump) short form<br />
trulig trumpen bad mood<br />
truls liten slarvig pejorative<br />
trumla roterande q or s movement<br />
trumma slag<strong>in</strong>strument sound<br />
trumpen butter och missnöjd bad mood<br />
trumpet blås<strong>in</strong>strument sound<br />
trunk stor augmentative<br />
truta skjuta fram läpparna long th<strong>in</strong> form<br />
tryne nos long th<strong>in</strong> form<br />
tråd tunt utdraget long th<strong>in</strong> form<br />
tråg avlångt fyrkantigt long th<strong>in</strong> form<br />
tråkig enformig negativ bad mood<br />
trål strutformig long th<strong>in</strong> form<br />
tråna känna stark längtan mental feel<strong>in</strong>g<br />
trång m<strong>in</strong>sta dim<strong>in</strong>utive<br />
träda långsamma steg walk<strong>in</strong>g<br />
träns snöre long th<strong>in</strong> form<br />
träsk vatten wetness<br />
träta irriterad ordväxl<strong>in</strong>g bad mood, talk<strong>in</strong>g<br />
tv<strong>in</strong>g sammanhålln<strong>in</strong>g putt<strong>in</strong>g together<br />
tv<strong>in</strong>na sno ihop trådar putt<strong>in</strong>g together, long th<strong>in</strong> for<br />
vrak skadat odugligt destruction<br />
vrede stark känsla förorättad bad mood<br />
vrensk själsligt motsträvig bad mood<br />
vresig ovänlig missnöje bad mood<br />
vret liten dim<strong>in</strong>utive<br />
vricka kraftigt vrida q or s movement<br />
vrida runt rörelse q or s movement<br />
vrål skrik talk<strong>in</strong>g<br />
vrång sur motsträvig bad mood<br />
vräka handlöst kasta q or s movement<br />
Appendix 1 Sida 19
vräkig överdrivet pejorative<br />
vränga vända fel pejorative<br />
vrövel struntprat talk<strong>in</strong>g, pejorative<br />
Appendix 1 Sida 20
Appendix 2<br />
Some <strong>in</strong>terjections for 9 different languages. (Swedish and English are written <strong>in</strong> ord<strong>in</strong>ary orthography)<br />
'Pejorative'<br />
Swedish Icelandic English Polish Hungarian F<strong>in</strong>nish Ososo Malagasi Slovenian<br />
usch [Oj] oh [O] [o:] [h¨i] [fjO] [dE] (nasal [fOI]<br />
hu [Ojbjak:] pooh [Oj] [u:] [h¨h¨h] [pçO]<br />
E) [fEI]<br />
"t" [|] [u:] phew [ox] [jOj] [hui] [œh] [ts]<br />
blä<br />
ugh<br />
[juj] [|]<br />
whistle<br />
ha<br />
ooh<br />
[fui]<br />
[uœk]<br />
håhå (jaja)<br />
tut [|]<br />
[|]+gesture whistle<br />
tvi<br />
boo<br />
öh<br />
äh<br />
bah<br />
asch<br />
äsch<br />
isch<br />
uh<br />
urrk<br />
fy<br />
bu<br />
tss
'positive' (appreciation, joy etc)<br />
Swedish Icelandic English Polish Hungarian F<strong>in</strong>nish Ososo Malagasi Slovenian<br />
tjo(hej) [vau] oh [O] [oh] [ah] [ah] - -<br />
ah [jipiO] ah [m:] kiss<strong>in</strong>g [m:] sigh<br />
åh<br />
wow<br />
sound + [jip:i]<br />
haha(ha)<br />
yippee<br />
gesture<br />
hihi(hi)<br />
[|]<br />
m:<br />
'surprise'<br />
Swedish Icelandic English Polish Hungarian F<strong>in</strong>nish Ososo Malagasi Slovenian<br />
oj [ha] oh [O:] [je:] [o:] [Ou] [a] [Oho:]<br />
åh<br />
ah<br />
[o:] [m:] [aj]<br />
(positive)<br />
ä<br />
whoops<br />
[h¨]<br />
[Ox]<br />
åhå<br />
[O:]<br />
hoppsan<br />
[OI]<br />
(negative)<br />
[jOI]<br />
(negative)<br />
'other bodily and mental feel<strong>in</strong>g'<br />
Swedish Icelandic English Polish Hungarian F<strong>in</strong>nish Ososo Malagasi Slovenian<br />
aj [aj] oh [Ox] [au] [aj] [Ou] [aj] [aIs]<br />
('pa<strong>in</strong>') [Ou] ow [O] [jaj]<br />
[euo:]<br />
[ux]<br />
[au] ouch<br />
[juj]
- -<br />
shiver<strong>in</strong>g<br />
sound<br />
[br:] shiver<strong>in</strong>g<br />
sound<br />
[O]<br />
[Oj]<br />
brrh<br />
ugh<br />
[ı]<br />
(voiceless)<br />
[br:]<br />
[hm] - [hm] - - -<br />
hm<br />
[ahem]<br />
hm:<br />
m:<br />
- -<br />
- - - - work<br />
songs<br />
hm<br />
(<strong>in</strong>gressive)<br />
brr<br />
burr<br />
('freez<strong>in</strong>g')<br />
hm<br />
('thoughtfulness')<br />
åhej('åhå')<br />
('jo<strong>in</strong>t<br />
effort')<br />
[nu]<br />
[huhuh] expiration - -<br />
phew [ux] [´:]<br />
[h¨]<br />
[F]<br />
[´f:]<br />
sighs<br />
- -<br />
[ajE] - - - - [euo:]<br />
[Oia]<br />
- [aha:] - [aha:]<br />
- [aha]<br />
[jOj]<br />
- aha<br />
haha<br />
oho<br />
<strong>in</strong>halation - - - [h¨i] silence - [Ox]<br />
[at·Cu:] atishoo [apCik] [haptsi] - [itC·imi] - [a·tCix]<br />
[tCixah]<br />
puh<br />
pust<br />
('exhaustion')<br />
ojojoj<br />
vojvoj<br />
håhåjaja<br />
('compla<strong>in</strong><br />
-<strong>in</strong>g')<br />
aha<br />
('sudden<br />
<strong>in</strong>sight')<br />
iiih<br />
('fear')<br />
atjo<br />
('sneeze')
- [m:] (with<br />
ris<strong>in</strong>g <strong>in</strong>tonation)<br />
[n´mn´m]<br />
yummy [njamnjam] - [m:] smack<strong>in</strong>g<br />
noise<br />
[m:]<br />
[nam]<br />
[namEnamE]<br />
swear<strong>in</strong>g<br />
- - - - maybe<br />
imitation<br />
of animal<br />
mums<br />
namnam<br />
m:<br />
('good<br />
taste')<br />
grr<br />
('anger')<br />
[gr:]<br />
[Oh] (with<br />
coarse<br />
voice)<br />
'commands to animals'<br />
Swedish Icelandic English Polish Hungarian F<strong>in</strong>nish Ososo Malagasi Slovenian<br />
schas [S:] shoo [S:] [hœS] (to [S:] [Cu:] [uS:] [pC:]<br />
('go +gesture<br />
<strong>in</strong>sects)<br />
(common<br />
away') [tvi:tvi:]<br />
[Sits]<br />
command<br />
[hOuhOu] (to<br />
to several<br />
sheep)<br />
animals<br />
"march")<br />
pull pull<br />
- - - - - - -<br />
(to hens)<br />
('come<br />
here')<br />
[gibagiba gib]<br />
(voiceless<br />
[b]) (to<br />
sheep)<br />
[putaputaput]<br />
(to hens)
'commands to people'<br />
Swedish Icelandic English Polish Hungarian F<strong>in</strong>nish Ososo Malagasi Slovenian<br />
jaja [fOs:] now now [njE] (=no) [œlœ] (verb [kai]<br />
aja(baja) [svei] no no<br />
negation)<br />
ajaj [sveiDr] naughty<br />
(mild<br />
naughty<br />
warn<strong>in</strong>g<br />
to<br />
children)<br />
sch [¨s˘] sh [tC:] [tCit] [S:] gesture [S:] -<br />
(be quiet)<br />
[kuS:] +gesture<br />
pst [hai] here whistl<strong>in</strong>g [he:] [pst] (<strong>in</strong> family [e] -<br />
(soft call [¨u] (=du) hey<br />
play) name<br />
for<br />
[EnE]=you<br />
attention)<br />
pst or<br />
whistl<strong>in</strong>g<br />
(to<br />
prostitute)<br />
bu [ba] boo [bu:] [hu] [b¨] [dja] [a] (coarse -<br />
(to scare<br />
[bux] [bu]<br />
voice)<br />
some-<br />
[´] (coarse<br />
body)<br />
voice)<br />
vyss vyss [bi¨mbi¨m] hushaby - - - [S:] - -<br />
lull lull [s¨s¨˘s¨s¨˘]<br />
(quietly)<br />
(putt<strong>in</strong>g s<strong>in</strong>g<strong>in</strong>g<br />
sb to<br />
sleep)
_ -<br />
[¨u] (=du) [EwE] (not<br />
to adults)<br />
[Ei]<br />
ohoj<br />
hoho<br />
(louder<br />
call for<br />
attention)