Language Evolution
Posted on July 29, 2008
Filed Under Philosophy, Science |
There’s a great post up at Science Blogs on Language Evolution. Note that linguistic or symbolic evolution (yeah the whole meme thing) is unlike regular biological evolution. That’s because there is intentionality to it. I’ve long found linguistic evolution interesting - ever since I went though the math in Royal Skousen’s Analogical Modeling.
Skousen’s model never caught on but it actually modeled actual linguistic evolution quite well. He had two textbooks on the subject - but both are out of print sadly. The one that went through the math ended up being very similar to thermodynamics except that instead of some of the relations being exponential they were quadratic. (It’s been about 17 years since I last read it so I don’t remember the details on why) I used to have some code around that show how the encounter with new languages would change the pronunciation of existing languages. It was pretty impressive.
Peirce, famously, extends this to cosmology. So thinking about these issues actually predates the work of the last two decades by at least a hundred years.
Comments
I think language is hugely important for humans. But I’m not even sure what it would mean to say it is the primary driver of human evolution. Language and biology are simply different issues.
Clark, what does it mean that linguistic evolution has intentionality? I read that as saying that the changes in a language are executed to bring about predetermined goals; to produce French from Latin, for example. Obviously, I’m reading this wrong.
All I mean is that language has significance. That ‘aboutness’ or ‘directedness’ means that it’s not simply codes with mutations that are either successful or not. Which is not to say we can’t talk about success and success in new contexts. Just that one should be careful not to conflate the two. (And Skousen’s discussion of linguistic evolution actually doesn’t focus in on intentionality).
The way reproduction occurs is quite different as well obviously. Success in biological evolution simply means genes are copied over. (A bit trickier with non DNA heredity such as now is being discovered but you get the idea) Success in language means it’s used but the replication is a tad trickier simply due to the nature of language. We can talk about what replicated in biological evolution. What replicated in language?
How have books influenced human evolution? Computers? How will biotech influence human evolution? Those are the sorts of questions I’m asking when I consider whether active symbolic evolution is replacing passive biological evolution as the primary driver of human evolution.
Who lives and who dies (who is selected) is increasingly based on our symbol systems (medical ethics, for example) rather than our biological systems.
I’m not sure how one would measure that since human evolution is slow and the changes in society recent.
Very interesting ideas about language versus biological evolution. Another big difference, now that I think of it, is that biological species do not (normally) exchange genetic information, whereas languages are free to borrow from one another. This free flow of information allows for much more rapid evolution.
Perhaps a more rigorous—and more technically challenging—approach would be to look at grammatical evolution since (I think) grammar is less fluid.
Here is an attempt to measure the relative speed of symbolic evolution versus biological evolution by overlaying many experts’ assessment of the timing of important evolutionary events in human history and trending their average. The result is an exponential curve, with biological evolution driving the oldest changes and symbolic evolution driving the newest changes.
http://en.wikipedia.org/wiki/Image:ParadigmShiftsFrr15Events.svg
I confess I’m very skeptical of such charts for two reasons. One, they always come off as ad hoc to me. (i.e. by what criteria do we judge an event to be significant?) Second, it assumes a direction to biological evolution which is just plain wrong.
Brian, the problem of linguistic evolution is that while we have records of 20th century languages and then a few dead languages we don’t have much else. (And we’re loosing most languages despite the attempts of linguistic anthropologists to record data on them) Put an other way, we can infer differences from current languages and work backwards. But there’s precious little data from which to say too much.
Where we can do a lot is more in phonetic shifts especially when languages come in context as well as how that affects new vocabulary. We obviously can also examine languages that evolved fairly quickly which we have good records for. Especially English as it moved from being completely unlike it is now in the 13th century on up to Shakespeare’s time then into 19th century stabilization and to our current form. There were obvious huge influences from German, French and Latin. But then more subtle changes when the British empire took off.
You also have interesting cases to study as English became the dominate trade language supplanting Latin and perhaps French and German before. It’s had a huge impact on other languages even if most don’t adopt so much vocabulary wholesale the way English has. (Although there are more hybrid languages such as the form of French spoken in Quebec)
It’s not a field I’ve kept up on the past 15 years. So I can’t say too much. Every now and then I go back and read a bit on the debates about the (biological) evolution of language. But I’ve really not spent much time recently on linguistic development proper nor on the mathematical models for it. As I said, the last time I did was over 17 years ago.
Clark, help me in my ignorance. Certain aspects of language must be more fluid than others: phonetics and vocabulary, for example. Grammar (is this what you mean by “linguistic”?), from what I understand, is pretty well ingrained in kids at very early ages (2-3 yrs old), even though they don’t have the vocab to speak in complete sentences. Still, they “know” subject, verb, etc.—the basic structure. The Japanese might borrow words from English, but they would never borrow grammar, would they?
I appreciate the challenges of linguistic evolution (i.e., not having a record), but I’ve seen how biologists are overcoming analogous problems.
I just don’t know the answer to how often grammar is borrowed. I know that in pidgen language new grammars develop spontaneously. This is partially why folks like Chomsky think grammar is innate and that there are simply variations our cultures pick between. I also know English borrows grammar from French, Latin, German and then obviously its own roots. Indeed the whole problem in the early 20th century of teaching English was the attempt to enforce largely Latin grammatical rules on a language for which it was largely foreign.
But as I said, I’m just not the person to ask these questions of. As I said the best that can be done is somewhat similar to what biologists do. Look at structural similarities and make inferences from that. However DNA is simply much more fertile ground for that. (IMO)
Is it just plain wrong that there is direction in biological evolution? That depends on how you define “direction”. Is it direction when anatomies are selected for fitness in their environmental contexts, when environments are selected for fitness in their meta-environmental contexts? Given our set of meta-environmental laws, whatever they might be, it appears there is an extremely ancient trend of change, at least in our locale, toward increasing complexity.
Regarding the “ad hoc” criticism, the specific criteria for judging events to be significant appears to become of less practical consequence as more judges’ judgments are included in the calculation. For example, if you ask 1000 persons how many marbles are in a jar, the average of their judgments will probably be more accurate than most of their individual judgments. Of course, in the end, we’re making judgments about significant events using our common human anatomy, which is inclined to certain kinds of perceptions and processing, but your skepticism of such perceptions and processing also arises from your human anatomy.
I’m not sure you can say evolution leads to complexity in the fashion you are using it. Certainly complex forms, over time, develop. But there are still many bacteria which are not complex. Are they any less adapted for the environment?
To add, on the ad hoc issue I don’t think that resolves it. Certainly I believe as a Peircean that if inquiry is continued we’ll get closer to the truth. I don’t see any evidence that just throwing judges at the problem is sufficient. If only because (a) those making the judgements appear to be making it ad hoc so merely adding more ad hoc judgments doesn’t resolve the problem and (b) because I don’t think there’s been sufficient time to converge upon a stable answer.
Just saw this article in PNAS: http://www.pnas.org/content/105/31/10681.abstract?etoc
Didn’t have time to read it, but thought you might be interested. (If you don’t have access, let me know.)
Leave a Reply
Stemming from your past blog posts about Peirce’s symbols, would you agree that symbolic evolution is increasingly displacing biological evolution as the primary driver of human evolution?