Log in

No account? Create an account
delirium happy

Just keep on trying till you run out of cake

Previous Entry Share Next Entry
Scandinavian languages are like firebellied toads
delirium happy
One of the (many) things that tends to annoy me about the British education system is the enforced narrowness of the education. On the one hand, it's good to be able to get a degree without having to take classes in PE or geography, for instance, but on the other hand, there are so many things beyond physics that I'm interested in and would like to study. Naturally, I do try to learn about other stuff whenever I get the chance; for instance, I'm currently reading a book about the evolutionary history of fishes. While I do frequently manage to boggle friends by knowing stuff, time, energy and useful guidance are all in short supply.

One of the consequences of this is that when I have thoughts or ideas about things I've picked up, I have no idea if they've been thought over before and rejected by more knowledgable people than I, or if they're generally accepted common knowledge, or if I'm actually having original thoughts. I mostly suspect that it's a combination of the first two. This is not necessarily a bad thing, in terms of getting a mental workout, but please do forgive me (especially the several of you reading this who know much more linguistics than I do) if what I'm talking about here is either obvious or rubbish.

What I've been pondering recently, in a half-baked sort of way, is the connection between evolutionary biology and historical linguistics. I've always been aware that the two diswciplines are superficially similar (for instance, the prevalence of tree like structures) but it isn't something that I've given all that much thought to. On closer inspection, though, I realise that what I know of evolutionary biology influences the way I think about historical linguistics. For instance, I have very clearly in my head the idea that if we were to have complete knowledge, there would be no places in the language tree where one point split into two or more branches. One language always splits off fractionally before the other. I've never seen this idea expressed in anything I've read of linguistics, and it may not even be applicable, given that there's much more scope for borrowing between languages than there is for gene flow between (non-bacterial) species. That's the mental picture I have in my head for an idealised language family tree though. The cladistic approach makes sense to me.

As another example, I'm as wary of phrases like "English evolved from German" as I am of ones like "people evolved from chimpanzees", and yet I've heard the former from linguists but never the latter from biologists. Ironically enough, I think the problem here may be one of language. In the biological example, "people" and "chimpanzees" clearly both refer to points at the end of the evolutionary tree, and if you go back through time to find the last common ancestor then you're clearly no loner at the specific point that is labeled "chimpanzees". I think of "English" and "German" in the same way and as such I'd say "English and German share a common ancestor; that this language may have been spoken in what is now Germany, and may be refered to as "Old German" or some such (I haven't a clue if it actually is, but it could be) doesn't really matter to me. It would be equally as valid, though, to reserve the name "German" for the paraphyletic group of all germanic languages spoken in what is now Germany. And in that sense, saying that English evoled from German makes perfect sense and is equivalent to saying that birds evolved from reptiles. My mind rejects this though, because it likes to picture things in terms of clades.

So given this, I decided to try to take the connection further, deliberately this time. One of the nice things about much of evolutionary theory is that it doesn't depends on any sort of mechanism, but only on more basic concepts such as heritability. That Darwin predaed Crick and Watson speaks volumes. This caused the birth of the concept of the meme (and no, I don't mean anything to do with quizilla here). Richard Dawkins was trying to make the point about evolutionary theory being independant of genetics, or even biology, and realised that similar ideas could be applied to thoughts, knowledge, ideas and so on.

In a sense, one could look at language itself, or individual languages as particularly successful meme-complexes, and just analyse them in terms of memetics, but that wouldn't be all that helpful. To pull an analogy from physical sciences, nobody tries to solve the Schrodinger equation to know what happens when you mix hydrogen and oxygen at a high enough temperature.

So what can you do? Well, I know (read: read in wikipedia) that the concept of mass lexical comparison is similar to techniques used in phylogenetics, and yet the idea that is embraced by biiologists and taxonomists is poo-pooed by many linguists. Why is this? Damned if I know. I suspect that one factor in it may be the difference in dealing with 4 DNA bases compared with many hundred different phonemes used in the whole spectrum of human language. It still seems somewhat silly on behalf of the linguists to me though. It may not give perfect results, but it's a start.

Then there's the actual speciations themselves. (I wonder, is there an equivalent word for the evolution of a new language?) The first thing thing that amuses me here is that both disciplines have difficulty with their basic definitions. What is a species? What is a language? The oft quoted definitions of the ability to interbreed and mutual intelligibility are both somewhere between wrong and incomplete. At the end of the day, something is a species if smart people say it is, and a language if a different set of smart people say it is.

But given that, what causes one language to split in two? I must confess, I'm not actually sure about that, but I'd presume that a big factor would be geographic isolation. Picture a bunch of Saxon, pottering around Saxony, speaking Saxon. Then a bunch of them go off and invade England, never send letters home, and before you know it, a few hundred years have passed and we're going on holiday to Germany and hoping that speaking really loudly will make them able to understand us. But wait a second, that's a classic allopatric speciation.

So how far can you go? Does language change have any sort of element of "survival of the fittest"? Well, clearly there's some sort of adaption to the environment; it's hardly surprising that Inuktitut has more snow-related words than Swahili. And equally obviously, a language which couldn't evolve new words would be at a stern disadvantage in these days of computers and technology. But these sort of environment-influenced vocabularies will surely not have a whole lot of influence on the language's overall structure.

Languages, though, are intimately tied to cultures. The language a person speaks closely affects the way that they view the world, and that certainly can have a notable influence. I'm curious as to which is the chicken and which is the egg; does the nature of the culture influence the development of the language, or does the nature of the language influence the development of the culture. Most probably a little from column a and a little from column b. Either way, it's easy to conceive of how the languages that thrive and develop will be those associated with a warlike or imperialist culture.

We have a Europe full of the highly migratory and aggressive Indo-Europeans, with only a tiny Basque pocket isolated up in the Pyrenese (well yes, the Finns, Lapps, Estonians, Hungarians, Maltese, Turks etc. aren't Indo-European either, but they are all from migratory groups). While there are undoubtedly other factors involved in this wide gepgraphic distribution, I do wonder if the language itself could in any way be seen as having anything to do with it.

Then what about synpatric speciation? What of language splits without any sort of geographic separation? Do we get that? Where there's one language spoken by two populations in the same place, with very little contact between the two, eventually evolving into separate languages? Sure we do. A language might go one way when used in religious ceremonies and a different way when used secularly. Or we might get a difference between legal language and colloquial language. Or we can just get isolated populations, speaking their own version of a lnaguage, such as Ebonics.

And so on and so forth. I'm sure that people with more knowledge than myself could make even more comparisons. Pidgins and hybridisation. Genetic and linguistic drifts. This, that and the other. I remember reading something, a while ago, which I've taken to heart, about the importance of reading outside of your own field. If you only read inside your own field then you mostly end up just recycling the same old ideas which everyone else has already had, whereas if you read outside your field, you stand a chance of seeing new and inovative ways of applying ideas to different areas; your world view gets broadened and it becomes easier to think outside of the box. I wish I could remember where I read it, though I suspect that it was either Terry Pratchett or Richard Feynman. It seems to apply fairly well in this case, though. Or at least, it seem that way to an ignorant physicist who really doesn't know what she's talking about.

  • 1
(Deleted comment)
In a purely descriptive sense, if someone says something like "it is me" and it doesn't sound somehow weird to them (aside from having it drilled into their heads that it's wrong), then it's grammatical. What's happening now is that very few people under the age of 70 (at least in the U.S.) genuinely use "It is I" without having to remind themselves consciously that it's the prescriptively "correct" usage.

When 2nd language speakers outnumber the native speakers in a population, it tends to result in a creole forming: this is why exploitation colonies in the Caribbean developed creole versions of French, English, etc. while in the Americas, which were settlement colonies with low L2-to-L1 ratios, that didn't happen.

Definitely check out Cavalli-Sforza's books: a good one is The Great Human Diasporas. It talks about evolutionary biology, linguistics, and how populations spread through the world.

I shall completely defer to froia_arme on matters of linguistics, since she's severl orders of magnitude more knowledgable on the subject than I am, and her recomendations will no doubt be much better than anything I could come up with.

In terms of evolution, I'm not really sure what to recommend, since I have a hard time remembering where I picked up any of the various things that I happen to know; they just form a big jumble of facts swimming about in my head. Using the very scientific method of "glancing up at my bookshelf", I'd pick out "Frogs, flies and dandelions" by Menno Shilthuizen, which is a book about speciation and is quite fascinating, and "Genome" by Matt Ridley, which is primarily a book about genetics, but covers a fairly wide range, including evolutionary stuff.

Whoa. It's hard to know where to even begin replying to this. Your intuitions are indeed correct. To hit upon a few highlights: we assume binary branching in linguistic speciation, the ancestor of both German and English is a reconstructed language called Proto-Germanic, and you should check out Tandy Warnow's website for some papers on using phylogenetic techniques to investigate linguistic family trees.

As for what causes languages to split: no language is monolithic; the saying is "a language is just a dialect with an army and a navy." There's always a continuum of dialects, and the changes (=mutations) that build up in various dialects through chance, through contact with other languages, etc. keep building up until the dialects are no longer mutually intelligible. There is some sort of Darwinian pressure in that there is the vague prediction that more easily learnable things in language replace less easily learnable ones, but there is always a huge debate over what's more learnable.

The idea that language affects thought is mostly a fallacy; the idea was popular many years ago (Edward Sapir and one of his students, a guy named Whorf, were big proponents of it, so it became known as the Sapir-Whorf Hypothesis) but we're now pretty sure that the effects of language on thought are quite weak. The two most commonly cited "examples," the Eskimo words for snow and the Hopi concept of time, are both myths.

As for the above commenter: I would highly recommend reading Luigi Luca Cavalli-Sforza's work.

I don't suppose you have any recomendations for things I could read about the Sapir-Whorf Hypothesis, do you? To me, the idea that language influences thought seems intuitive. For instance, I've read that Welsh has a different boundary between blue and green to English, and would presume that that would influence the way that its speakers viewed colours. Or to use a slightly less trivial example, I've also heard that it isn't really possible to understand some of the formalities of Japanese society without understanding the formalities of the language.

I'm very happy to accept that the people who have put a whole lot more time than me into thinking about the issue are probably right, but I would be interested to know just where I went wrong.

The idea of learnability is also an interesting one that I hadn't thought of, and will no doubt spend some time pondering upon.

For color words specifically: there was an article in Scientific American about a year ago talking about the classic work in the field, done by Berlin & Kay. To preview their conclusions, it's precisely *not* true that the number/boundaries of color terms in a language affects its speakers' ability to identify colors; it just affects their labeling. (But there is the effect that if someone has a name for a color, they can remember it better.) There's bound to be stuff about Sapir-Whorf in any book on linguistics that you find. Pinker definitely treats it in The Language Instinct.

As for learnability: it's been claimed that nearly all linguistic change arises from imperfect acquisition by children. This can't be entirely true (people's grammars & pronunciations aren't totally fixed in concrete at a certain age) but it is largely right.

Oooh. That's really interesting. Thank you.

Wheeeee! I understand this thread from the point of having studied it in the psychology of language modules I've done. Now I feel good and intelligent. And this comment probably undid that, but I felt the need to squee.

  • 1