This paper intends to examine the words starting with given in the Oxford English Dictionary, and if possible categorise them, according to shared meaning. It will examine briefly the academic background to Sound symbolism along with contemporary attitudes towards the study of the relationship between sound and meaning. It aims to draw conclusions from the results of the categorisation of the words with the view to deciding whether it provides any evidence for phonesthesia.
Saussure is considered to be the founder of modern scientific linguistics. In relation to Sound symbolism, he had said that “the entire linguistic system is founded upon the irrational principle that the sign is arbitrary” (Saussure, 1983: 131). By this he means that the words used to denote concepts could be any words at all. He believed that there was no identifiable pattern or relationship between the word and its referent (http://encyclopedia.laborlawtalk.com/Sound%20Symbolism). He furthers this by explaining that because he considers the word to be arbitrary it can only hold meaning in relation to other words. These ideas are the basis on which much of the current literary thought stems from.
Phonesthesia is a linguistic term relating to the idea that phonesthemes can contain inherent meaning. It is a branch of linguistics that stems from Sound symbolism. Sound symbolism is the study of the relationship between of an utterance and its meaning and consists mainly of three types of sound meaning which were originally formed by Von Humboldt in the 18th Century; these are onomatopoeia, clustering and iconism. Onomatopoeia refers to the use of the letters of the alphabet to imitate a sound in the real world. It can therefore only apply to sounds or to words which denote sound.
Clustering is concerned with words which share sounds that have meaning at vowel and consonant level; Magnus (1997) refers to this as the Semantic Association. An example of Semantic Association given by Magnus is that if in any language the given word for house begins with an or a or even an then by the process of clustering it would be expected that a disproportionately large number of words concerning housing would also being with that letter, or of . So therefore in English we have house with the other house words including hut, home, hacienda, and hovel. It is important to recognise that in a language where the house set or words is not denoted by a and by a for example, that the clustering will occur around a different letter. Therefore the meaning of both the and clusters will be different in the non English language. Iconism is exhibited by words which have the same referent but only differ in their sound.
The theory of Iconism presumes that each vowel and consonant sound has its own inherent meaning. Furthermore it states that the meaning of these sounds is universal; therefore the consonant sound will hold the same inherent meaning in all languages where it is spoken (Magnus, 1997). This is challenged by Shisler (1997) who says that most sound symbols are found to language specific. He does, however, give generalised statements regarding the universality of sound symbolism. He gives examples as not being specific to the phonestheme such as but as being specific to the generalised phonetic sound type. For example he says that abrupt sounds are often represented by plosives. This view is in line with the view of Hans Marchand (1969) who provided the first extensive list of English phonesthemes.
He states that certain sounds are connected with certain areas of meaning. He calls this the “imitative principle” (Marchand: 1969: 398). Universality of the inherent meaning in phonesthemes is shown to be incorrect by Diffloth (1994: 112). In this paper it can be seen that in the language of Bahnar, the vowels used in words denoting things that are big and small are high and low vowels respectively. This is entirely opposite to the system in English where a high vowel denotes something small and a low vowel denotes something big. Therefore it is not possible that Iconism is a language universal. The analysis of the group of words with an phonestheme does not concern the onomatopoeic section of thought on sound symbolism. It is mainly concerned with the other two branches of sound symbolism: Clustering and Iconism.
In the analysis of words beginning with the phonestheme, it was found that it was possible to categorise the words into several groups of related meaning. Also, it was found that it was possible to further subcategorise the groups. The dataset for the phonestheme was taken from the pages of the O.E.D. online (http://dictionary.oed.com). The dataset consisted of a total of around one thousand words which all started with the phonestheme . It was found that it was possible to group the words into six categories of meaning. These categories do not include all of the words given in the initial dataset and at most they represent only a fraction of the words considered. With this in mind it offers little evidence for phonesthesia. How can a phonestheme which by definition, is a sound cluster which is directly associated with a specific meaning be associated with more than one meaning?
Moreover if, in relation to Iconism, a phonestheme possesses an inherent meaning the expectation would be that it would be apparent in all manifestations of the phonestheme. Of the sample taken, only a comparably small number of words could be categorised. For the theory of Iconism to be correct it should be possible to categorise the entire sample. Also the categories created would have to be of related meaning so the overall meaning of the phonestheme was still preserved. The six categories were: words which expressed something to do with the face, words which expressed deviousness, words which evoked the emotions of being warm and cosseted, words which expressed anger or annoyance, words expressing snobbery, the qualities of a person who is “highly strung” and lastly words that expressed things to do with snow.