The areas of onomatopoeia and sound symbolism have not been central in linguistics, and Swedish, as well as many other languages, lacks a thorough description of its sound symbolic characteristics. This paper reports part of a larger study of sound symbolism in Swedish, where the emphasis is on phonesthemes of initial consonant clusters. I will use the term sound symbolism as a general term for an iconic or indexical relationship between sound and meaning including the relationship between speech sound and other sound. Onomatopoeia is then a special case of sound symbolism. A phonestheme can be described as "the grouping of similar meanings about similar sounds" (Bolinger, 1965) or as "a phoneme or cluster of phonemes shared by a group of words which also have in common some element of meaning or function, though the words may be etymologically unrelated" (Householder, 1946).

In order to find phonesthemes, words were excerpted from Svenska Akademiens Ordlista (SAOL, The Wordlist of the Swedish Academy) and Svensk ordbok (SOB, Dictionary of Swedish). To qualify as a phonestheme at least two words with the same initial consonant sequence and similar meaning should be found in the lexical material, and the likeness should not come from trivial morphological relatedness like derivation of e.g. nouns or adjectives. As an example fjant, fjanta, fjantig are counted as forms of the same root. Kladd, kladda och kladdig are another example.

Likeness in meaning was searched for in the form of words, "key words", in definitions or in definitions of words in definitions. A key word is always a word which is written in the lexicon entry of the word under consideration and is, in most cases, a word in the definition of the actual word. Examples of key words are: ljud (sound), rörelse (movement), spetsig (pointed), ljus (light), äcklig (nasty), vatten (water), tjock (thick), klibbig, (sticky), ohyfsad (rude), löjlig (ridiculous). With an attitude of accepting doubtful words rather than discarding them, about 1 250 words — in the sense of root morphemes — were registered for further analysis. (According to Nusvensk frekvensordbok 4 (Allén et al, 1980) the number of lexical morphemes in Swedish amounts to around 8 300.)

Of the 37 initial consonant clusters of Swedish (Sigurd, 1965) all but one are used for sound symbolism, but they are used for different purposes and to different extents. The 37 clusters are: bj-, bl-, br-, dr-, dv-, fj-, fl-, fn-, fr-, gl-, gn-, gr-, kl-, kn-, kr-, kv-, mj-, nj-, pj-, pl-, pr-, sk-, skr-, skv-, sl-, sm-, sn-, sp-, spj-, spl-, spr-, st-, str-, sv-, tr-, tv-, vr-. The unused one (dv- ) has a lexical frequency of 3 morphemes. The following diagram shows which initial consonant clusters have the largest number of sound symbolic root morphemes.

The cluster with the highest number of sound symbolic root morphemes is sl- (70), followed by sn- (63), kn- (62), and by kr- (55) sp- (52) etc.

Naturally, the clusters differ in main semantic content: Sl- stands for the meanings 'pejorative' (24), 'wetness' (12) 'long thin form' (12) 'slackness' (11); sn- stands for 'pejorative' (13), 'sound' (10), 'long thin form' (10); kn- stands for 'sound' (15), 'round form' (15), 'pejorative' (10); kr- means 'winding form' (17), 'pejorative' (10), 'sound' (8); sp- typically means 'long thin form' (23).

Considering the fact that different clusters have differing lexical frequencies — a cluster like sl- is very high while a cluster like pj- is very low — even the proportional frequencies are interesting. The clusters with the highest percentages of sound symbolic root morphemes are: fn- (100%), kn- (81%), gn- (77%), spr- (74%), pj- (71%) and spj- (67%). Of these clusters, all but kn- have a low lexical frequency. Kn- is the initial consonant cluster which is most frequent both in absolute numbers and percentually.

As can be noticed in the description above of the semantic content of the five most frequent clusters of Figure 1, some meanings are recurring, and they recur throughout the whole material. The most recurring meanings, ranked according to absolute frequency, are 'pejorative', 'sound', 'long thin form', 'quick or strong movement', 'wetness' and 'talking'.

The following table shows which the most frequent initial consonant clusters are for the five most common meanings.

The table shows the following: The cluster sl- is the most dominating, in absolute numbers, both for the meaning 'pejorative' and for 'wetness'. (Many of the 'wetness' words also have the semantic feature 'pejorative'.) Fl- is the largest both absolutely and percentually for the feature 'quick or strong movement'. The clusters that are most frequent, in absolute numbers, show a strong tendency to end with l. The clusters that are most frequent proportionally show a weak tendency to end with j. In both groups, all clusters begin with a voiceless obstruent. Sl-, kl-, and sp- are lexically frequent clusters, fl- and kn- are intermediate. All these five are found in the left column. The clusters of the right column are all lexically infrequent clusters, except for fn-.

There are clear tendencies for different initial consonant clusters in Swedish to carry different sound symbolic meanings. Lexically small clusters are utilized to a large extent. This is in accordance with the discussion of Hinton, Nichols, Ohala (1994) about unusual sounds being used for sound symbolism. As shown above, pj- is the cluster with the highest percentage of pejorative root morphemes. Also, other two-consonant clusters containing a /j/ phoneme are pejorative and to quite a high degree (Abelin, forthcoming). Similar results for cluster types were obtained for other semantic features: 'light' is dominated by voiced plosives + /l/ or /n/: bl-, gl-, gn- (dl-, bn-, dn- are not allowed in Swedish), 'bad mood' is constrained to clusters with /r/ in second poistion gr-, tr-, vr- and 'wetness' is dominated by: sl-, kl-, sp-, sm-, skv-, spr-, sn-, bl-, pl-, sv-, i.e. /s/ or initial unvoiced plosives are preferred (the only exception is bl-). It is not only the initial consonant clusters which can carry sound symbolic meaning, but also the final clusters and vowels. The combined effects are studied in Abelin (forthcoming).