Specializations > Historical Linguistics

Extent of parallelism between evolution and language change

(1/7) > >>

We all use the metaphors, and they're useful for explanatory purposes. But how well do the metaphors really apply (are we just picking relevant aspects?), and is it entirely metaphorical?

We might have discussed this a while back, but I'm thinking about it again.

The inconsistencies in the metaphor are probably due to applying it loosely. It isn't "languages" that evolve any more than "species" evolve. It's the individual idiolects and individuals in the population. In both fields we might say vaguely that a species/language has evolved, but that's a simplification. And once we recognize that, the metaphor seems to almost apply perfectly to language, perhaps even beyond metaphorically.

As I see it, there are three remaining differences:

* Genetic inheritance is "nature", while language input is "nurture" (and we're not talking about UG / language faculty here obviously, because we don't have any relevant data on that ever changing)
* Multilingualism is common, as is adult learning; individuals in a species don't change genetically mid-life and also don't typically represent cross-species DNA
* The "reproduction" aspect is not asexual (one donor) or sexual (two donors) but broadly involves input from many speakers, although certainly the mother and father do typically contribute significantlyBeyond these points, almost everything else seems to line up.

There's another potential objection:

* Natural selection doesn't operate on linguistic form for survival as it does in speciesThat's partly true. But:
1. It does to some basic degree-- expressiveness and complexity balance out-- grammaticalization and change. Some is random, though.
2. Most importantly, language is just one property of individuals, like hair color. Certainly there may be some broad reasons why blonde or brown hair is related to evolution, but at least after these genes exist it seems somewhat random-- language is equal, I'd say. If you speak English or speak Spanish, it's about the same as having different hair colors (except for nature vs. nurture, see above). The crucial point here is that it eliminates the thought "but wait, if a language is spoken by more people, doesn't it mean it's a better language?" -- no, not at all, it's just one of the traits of those people, and others likely contribute to why that language is carried along for the ride.

Any other objections you can think of?

I agree that there is a type of selection that acts on language, to do with its learnability and complexity which are balanced out through iterated learning and self organization. 

The metaphor comes from the fact that they are both similar processes but acting on different materials - there will of course be difference because they are acting on different things.  'Evolution' simply means change over time in a population with variability.  Language's 'population' is of words and rules, and speakers are the engine that learn and use language, which changes it over time.

Both language and species are participating in this process of change over time through selection because of their variability.  Therefore they both exhibit this change over time.

You might be interested in this article which talks about the self-organization of language, and specifically how complexity can arise in an artificial language system:

See what I mean (in http://linguistforum.com/index.php?topic=76.msg237#new)?

If we stick to the code prejudice we will have to imagine all sorts of clever (for they ARE clever, as Daniel's theory about the French word salir was) theories to account for its changes to adapt to new situations.

But if we become aware of how human communication seems to work, then these problems disappear just as the fist disappears when we open our hands (Zen maxim).

Sorry Guijarro I don't understand what you mean!  Can you clarify?


--- Quote from: Corybobory on December 22, 2013, 04:26:09 AM ---You might be interested in this article which talks about the self-organization of language, and specifically how complexity can arise in an artificial language system:

--- End quote ---

That's kind of what I first thought of when I first read djr's post. I thought of an experiment where small robots were given two basic features:

* playing football (no, not handegg)
* communication
These robots were not explicitly given a language, so every time the data was reset the game would begin with all the robots being unable to communicate or play as teams, chaotically going all over the place and appearing almost random.

Over time the robots would learn from each other and each team eventually converging on a single, mutually intelligible language. Each time it would be different, random, emerging from the co-operative interaction of the robots

Naturally, this is a highly simplified model and experiment of language, but what I find intriguing is how the language in this case is an emergent result of interaction.

I can't remember the exact details, but it was always this emergent property of language that really intrigued me. Why would human language be any different? Would we have developed such "complex" language if we were not also a very social animal?


[0] Message Index

[#] Next page

Go to full version