Forgotten Languages Full: Phonosemantics Models and Phonaesthemes

Phonosemantics Models and Phonaesthemes

The idea that individual vocal sounds have meaning is not a popular idea amongst linguists which follow conventional linguistic theories, that is, those theories that cannot account for the results obtained when the sound symbolism hypothesis is applied. Thus, no surprise. Mind that the term "sound symbolism" is a confusing one and can be interpreted à lá Voronin (who talks about a sub-field of linguistic iconism -- phonologic as opposed to syntactic iconism) or in line with Jakobson's view, who considers the term more concerned with Peirce’s ‘icon’ rather than ‘symbol’.

Testing the phonosemantics hypothesis requires highly complex non-linear models quite beyond the naïve conventional linguistic models. This is so because experience is neither simple, not linear, something already signaled by Lakoff when he explicitely writes about the possibility of polysemy, the role of metaphor and metonym.

In her thesis, Margaret Magnus wrote that the 14 experiments she ran showed that word meanings are decomposable into various components, some of which are arbitrary and some not. She adds that since no word can function without all these components, it follows that all word meanings are in part arbitrary and in part predictable from their form. The new non-linear models do favor this statement and, furthermore, they show that even those word meanings termed "arbitrary" are not that arbitrary as she assumed. Magnus' asks as to whether the semantics of words (other than the Concrete Nouns) shifts with it, and the new phonosemantics models do give an answer: yes, they do. The new models include mechanisms to account for phonaesthemes (Firth 1930), defined as frequently recurring sound/meaning pairings that are not clearly contrastive morphemes. An example is the English onset gl- (Wallis 1699, Bloomfield 1933), which, like other phonaesthemes, is relatively infrequent in English, except among words with meanings related to ‘vision’ and ‘light’ (see Benjamin K. Bergen).

Again, the extent to which phonaesthemes play a role in the synchronic mental organization of language remains an open question for those conventional linguists, be it because the concept of mental synchronicity needs be revised, or be it because what turns those linguists "conventional" is precisely their unability to accept anything that lies beyond their fences. As Bergen correctly states, phonaesthemes' cognitive status remains controversial in part because they do not fit well into linguistic theories that view compositionality as a defining characteristic of morphological complexity, such as so-called ITEM-AND-ARRANGEMENT models (Hockett 1954).

In this report we refine our models (NodeSpaces-3316 and 3321) by incorporating the ability to encode some specific statistical associations within the linguistic environment in order to test Bergen's hypothesis that phonaesthemes that are well represented should yield new forms over time for which there is no cognate in closely related languages (taking into account that, for the established cognate sets, words can easily have different vowel quantity).

Additionaly, we also incorporate in our models Harris & Campbell (1995) proposal of syntactic correspondence sets, by means of which they compare sentences from different languages which have the same propositional meaning, and the grammatical morphemes of which correspond to each other systematically (turning therefore these sentences into cognates).