Posted: Aug 23, 2010 1:45 pm
Hmmm. This is a very interesting question. But formulated as it is, it would seem to imply that language development is an aim-directed process (which probably wasn't the intention of the OP?). In this form, it seems similar to asking what is the point of us having five fingers (rather than, say, six or four). The proper way of looking at this would be the evolutionary history, not the number of fingers it would be useful for us to have at this stage. The same is true of language. The principle of economy that the OP invokes is a very important one, but it can only work on what is given at each stage of the evolution of a particular language (and what is given depends on its particular history, which in turn is linked with social and cultural history, contacts with other languages etc. as well as constraints and possibilities afforded by its language material).
So I prefer the reformulation of the question at the end of the OP, which is essentially "why do any languages retain gender/noun classes"? (The same could be asked for different verb conjugations.) I can't pretend to answer this one, but I suspect at least part of it is the general context of the morphology of a given language. If you have a largely synthetic language with a lot of declination going on and this is necessary in order to indicate syntactic relations in the sentence, then series of inflections (and groups of series of inflections) used to indicate this will tend to persist. Such a group is usually thought of as indicating a grammatical gender, but in reality the importance of these inflections is simply in their difference from one another which marks the different syntactic functions a word may take. That's why the declination systems may be so resilient: preserving these differentiating markers is useful for structuring utterances. Of course, a language may transition from synthetic to analytic, using word order and prepositions to indicate syntactic relations. This is what happened in English, and as the morphology of the noun got simpler, the grammatical gender was "eroded" (all according to the economy principle: you don't generally need to encode the same information twice; although epepke is right in pointing out that redundancy is always present in linguistic communication, up to a point, and it could be argued that it is often "useful" since it helps avoid ambiguities and misunderstandings, or even conveys additional information - such as emotional charge, or aesthetic effects ... but I digress.)
How genders "started" is a distinct question to the previous one ... we could answer it if we knew more about how language itself started. FWIW, I'd venture that it has to do with gradual ordering into structural paradigms (declinations, conjugations) of the initial variability of linguistic material, since language certainly didn't start out as a fully-fledged system it is now (well, I say certainly - I assume, of course, that it wasn't given us ready-made by the gods!).
So I prefer the reformulation of the question at the end of the OP, which is essentially "why do any languages retain gender/noun classes"? (The same could be asked for different verb conjugations.) I can't pretend to answer this one, but I suspect at least part of it is the general context of the morphology of a given language. If you have a largely synthetic language with a lot of declination going on and this is necessary in order to indicate syntactic relations in the sentence, then series of inflections (and groups of series of inflections) used to indicate this will tend to persist. Such a group is usually thought of as indicating a grammatical gender, but in reality the importance of these inflections is simply in their difference from one another which marks the different syntactic functions a word may take. That's why the declination systems may be so resilient: preserving these differentiating markers is useful for structuring utterances. Of course, a language may transition from synthetic to analytic, using word order and prepositions to indicate syntactic relations. This is what happened in English, and as the morphology of the noun got simpler, the grammatical gender was "eroded" (all according to the economy principle: you don't generally need to encode the same information twice; although epepke is right in pointing out that redundancy is always present in linguistic communication, up to a point, and it could be argued that it is often "useful" since it helps avoid ambiguities and misunderstandings, or even conveys additional information - such as emotional charge, or aesthetic effects ... but I digress.)
How genders "started" is a distinct question to the previous one ... we could answer it if we knew more about how language itself started. FWIW, I'd venture that it has to do with gradual ordering into structural paradigms (declinations, conjugations) of the initial variability of linguistic material, since language certainly didn't start out as a fully-fledged system it is now (well, I say certainly - I assume, of course, that it wasn't given us ready-made by the gods!).