Further biological/linguistic parallels?
macw at mac.com
Sun Mar 26 21:09:49 UTC 2006
I have found your comments regarding parallels between genetics and language stimulating. But I am having trouble with the details of your current claims. I agree that protein folding is one of the most beautiful examples of emergence in biology. In fact, my intro college bio textbook (Campbell, Reece, Mitchell, p. 71) says, "The function of a protein is an emergent property resulting from exquisite molecular order." The book then goes on to describe the four levels of emergent structure. Beautiful stuff that can clearly serve as a model for emergentist thinking of relations across structure. On that level, I follow your thinking.
But there is nothing here about Darwinian selection of proteins. So, this level of analogy with your ideas about the emergence of the lexicon doesn't go through. This issue illustrates a general concern regarding accounts of linguistic structure as emergent phenomena. In the end, there is no doubt that all language structure is a result of emergence, albeit across many diverse time scales supported by many different mechanisms. But a proposal for a specific emergentist mechanism can be either right or wrong. I don't see how viewing the lexical inventory of a language in terms of protein folding is going to work.
You could well decide instead to use another, more selectionist, biological mechanism as his analogy. That would at least allow us to keep out analogies straight. But, then we need to have some evidence that the "form/meaning mapping is usually very crisp, and based on the internal symmetries of the used portion of the phonological inventory of the language." Is this basically the insight of Paget (1930) regarding sound symbolism in Indonesian? If so, getting this to work universally is going to be a pretty big project.
I like the idea of getting Tom G. involved in this discussion, but maybe first we can get the details of the emergentist proposal clarified a bit.
--Brian MacWhinney, CMU
More information about the Funknet