I've been hypothesizing a bit - I'm wondering if any of you could point me in a helpful direction.
I learned in phonetics about the concept of 'prototypes' I think it was - where you have an idea (subconsciously) of the ideal frequencies of certain vowels, which 'magnetically' (I think my teacher used this word) gravitate towards these centers of the ideal in one's usage, based on what one hears and produces.
And also these vowels sort of self-organize on the vowel chart so that they are as distinguishable from each other as possible. So you wont get a language where there's [ i] [ I] [y ] [ Y] and [ o] as the only vowels, because they'd be clustered at the high front area with a random one at the back - what will happen is they'll be evenly distributed over the vowel frequency 'landscape'. I imagine this is because people reproduce sounds that distinguish them from other sounds, and the ones that are different enough to be easily understood and used, therefore kept in the language and passed on, are the ones that persist. And this makes vowel inventories self organize into something where vowels are discretish areas on the formant frequency chart, spread out pretty evenly. So you get languages with [ i], [ u], [ a] or [i ] [ e] [ u] [ o] [ a], where they are spread out and so more distinguishable from each other, to make usage and perception easier.
Then I've also read a bit about grammaticalisation, how new grammatical forms can emerge out of a system of non-grammatical items. So for instance in a creole, or even a normal language, an open class content word, with frequent usage, might become over time a grammatical item, such as the word 'one' turning into articles over time in lots of languages.
So this form coming out of nothing, or regulating chaos to create a neat discrete system, seems to be the result of iterated learning - generations of language learners and users that change the language through their choices of what to use and reproduce, which is affected by the ease of learning or reproducing the language data they have in their environment.
Is this concept found in other areas, or is their a name for it? I have a hypothesis that teaching for example, intoduces this sort of environment on the material culture that people produce, and I'd like to read up on the linguistic literature to see if the concepts can be applied to archaeology. Let me know if I'm just babbling on about crazy things...
[edit] sorry the square brackets I used for some vowels screwed with the formatting - hopefully fixed now!
