Or is that just what Chomskytron programmed you to say?

Oct 13 2010 Published by under From the Melodye Files

I can't even explain how happy this comic makes me.  The poverty of the digits?  Chomskytron?  And just look at my lips!  I'm a mad hot bot, apparently.

Anyway, not what this post is about!  (But so excellent, I had to include).

I recently came by a fantastic little textbook written by Larry Trask, an acclaimed (and out-spoken) American linguist who specialized in the study of Basque, and was known to occasionally rage against Chomsky in The Guardian.  The subject of his text?  Historical linguistics.  A subject that, to be fair, I haven't read much about since being an impressionable young teenager, and discovering Merritt Ruhlen and protohuman language in the musty (dusty) stacks of the Glendale Public Library.  ("This sounds like historical fiction..." I remember thinking)  In any case, the Trask text has proved a wonderful refresher and I highly recommend it if you can find it on Amazon; it appears to be selling for under $5!

But to turn back the post at hand -- last week I wrote a piece on linguistic universals and language evolution, in which I suggested that the existence of linguistic universals is not good evidence for a universal grammar.  My argument was, in short, that linguistic universals might arise out of 1) overlap and borrowing between speech communities, 2) convergent evolution toward optimal communicative strategies, and / or 3) human constraints on perception and cognition (which are not rightly termed 'grammatical').  It seems likely that, taken together, these factors and others could account for the overlap found in comparative linguistics; certainly, I pressed, these kinds of similarities between languages provides a flimsy basis for positing 'deep structures.'

If I had had Mr. Trask to hand, I would have certainly quoted some of his heady facts and figures to bolster this argument.  Well -- better late than never!  What follows is a bit of light and enjoyable reading that speaks to last week's post (and others), and from which you will reemerge into the world, eyes sparkling, feeling altogether more clever and learned.  (Well, except for you, Avery... 😉 )

With no further introduction, Mr. Trask...

On multilingualism in the land before time

"Only very rarely, if ever, does a language find itself spoken in a completely isolated environment, with no contact at all between its speakers and the speakers of other languages.  Far more typically, the speakers of any given language have day-to-day dealings with the speakers of at least one or two other languages, and possibly with a larger number than this.  Indeed, for the larger part of human existence, the normal situation was probably for everybody to routinely learn and use two or three or four different languages.  This is still what we find today over most of the planet : it is thought that between 70 and 80 percent of the earth's population are bilingual or multilingual.  In the Amazon rainforest, in New Guinea, in much of Africa, in large parts of the Indian subcontinent, multilingualism is still the norm, and the same was true of Australia and much of North America before the European settlements largely destroyed the indigenous cultures and languages.  The state of affairs that we may now think of as typical, with a single language being spoken with some uniformity over hundreds of miles, is a relatively recent development in human affairs, and it is not at all representative of what has been going on during the past few millenia."

On convergence towards similar structures in 'Sprachbund'

"In some cases, centuries of contact between languages can lead to a particularly striking result: several neighboring but unrelated languages can come to share a number of structural properties with one another, properties which they do not share with their closest genetic relatives elsewhere.  ...[In the case of the Balkans], various innovations appear to have diffused across language boundaries, into several quite distinct languages which can in no way be regarded as forming a dialect continuum."

"One of the most famous examples of grammatical borrowing is found in the Indian village of Kupwar.  Kupwar is located on the boundary between the Indo-Aryan languages of nothern India and the unrelated Dravidian languages of the south.  Three languages are spoken in Kupwar: Urdu and Marathi (both Indo-Aryan) and Kannada (Dravidian).  Everyone speaks all three languages and switches among them constantly depending on the context.  Now, as a rule, Indo-Aryan and Dravidian languages have very different sentence structures, and elsewhere in the Indian subcontinent this is true of these three languages. In Kupwar, however, things are different.  ...The three languages have influenced one another so strongly that they have come to have identical sentence structures, structures which are different from what is found in the Urdu, Marathi and Kannada spoken elsewhere.  Indeed, one might almost suggest that the people of Kupwar speak only a single language, but that they speak it with three different vocabularies - and that is exactly what some linguists have suggested."

On variation within speech communities and among speakers

"No language is totally homogenous.  ...Even in a single locality, we can often find a substantial degree of variation.  For one thing, there is variation between social groups.  Women do not speak like men; middle class people do not speak like working class people; television newsreaders do not speak like disc jockeys.  For another, even within a single group, there is variation between individuals... For a third, even a single person doesn't always speak in the same way.  It's hardly likely that you speak in exactly the same manner when you're relaxing with friends in a pub or a bar and when you're being interviewed for a job."

On linguistic 'hypercorrection' as a function of social class

"[Labov found that] the second highest [SES] group, the lower middle class, uses more [prestigious' pronunciations in the most formal context than do the highest group, the upper middle class.  This phenomenon has turned up in a number of other such studies, and always it appears to be the second highest class which jumps above the highest class [in shifting their pronunciation] in the most formal context.  Labov calls this phenomenon hypercorrection... Very frequently, it seems, the members of the second highest class over-adjust their speech in very formal contexts towards prestige forms, as though they were particularly insecure about their speech and perhaps also about their social position."

6 responses so far

  • JS Allen says:

    This and your previous post remind me of a little puzzle I've been wondering about for a few years. Have you run across this one?

    In English, we use "will" to indicate future tense, and "have" to indicate past tense. We also use "will" to indicate desire, and "have" to indicate possession. For example, "I will go to the store", "I have gone to the store". In Chinese, the words yao (desire/future-tense) and you (possession/past-tense) serve the same dual purpose as in English. In Chinese, both senses of each word are written using the exact same character, so they aren't homonyms -- it's the *same* word used in two senses. And this usage appears even in ancient Chinese manuscripts. Other languages do the same. Latin didn't, but IIRC, at least one modern romance language brought back this convention.

    I don't believe a "common ancestor language" would be plausible to explain this, but would love to know if it's something experts have thought about?

  • I think there's something similar going on in Latin too -- at least some of the verb tense inflections are reduced forms of verbs.

    In the formal English of 19th century textbooks, the "borrowing" is person-dependent: in first person, "shall" is a neutral future-marker and "will" indicates volition; in second or third person, "will" is the neutral future-marker and "shall" indicates obligation. I don't know if other languages do that.

  • nopun says:

    'Child's play' is very apt name for this blog.
    I like these language games. thanks

  • JS Allen says:

    Thanks, Murray. Some of my British colleagues still use "shall", although it seems to be a bit sloppier than old English, in that they'll use it interchangeably with "will".

  • [...] paramétricos de Chomsky. O sea, la discusión sigue abierta y en algunos foros, encarnizada (Child’s Play) El aprendizaje del lenguaje es uno de los que podrían considerarse clásicos ejemplos de lo que [...]