I thought I would transfer the discussion from Education to here, so that we keep the UG discussion in one place.
Good idea to bring it over here. I am going to try to be brief, but not curt, since there is so much to respond to.
Okay, but would it be fair to say that according to Chomsky:
a) Language is separate from other forms of cognition and therefore a language universal is not merely something that there are physical mental constraints on such as speaking in ways that the human vocal tract is not able to produce?
b) That a language universal is exactly that: universal, and not merely likely.
I don't know whether Chomsky made fine distinctions as you do in a) but some psycholinguists do. It's a huge question, and personally, I think the answer is still unknown.
I have to say that as far as I know, language
universals for Chomsky and for the disciples I knew have always been "universal". The classic example is the so-called center- or self-embeddings.
From Wiki: In linguistics, center embedding refers to the process of embedding a phrase in the middle of another phrase of the same type. This often leads to difficulty with parsing which would be difficult to explain on grammatical grounds alone. In theories of natural language parsing, the difficulty with multiple center embedding is thought to arise from
limitations of the human short term memory. In order to process multiple center embeddings, we have to store many subjects in order to connect them to their predicates.
- The man that the actress loves arrived.
- The man that the actress that her child watched loves arrived.
- The man that the actress that her child that an artist painted watched loves arrived.
- The man that the actress that her child that an artist that I heard painted watched knows loves arrived.
It's clear that the increasing number of nested clauses causes confusion. This is very likely to be similar in all languages, or a universal constraint on syntactic structures, since we are all human. There are other examples. They get very arcane quickly.
The police arrested John and Harvey.
*Who did the police arrest and John? (or...John and.)
The police arrested John and who?
In the list on “phonological universals”, we have:
Here are some phonological universals concerning vowel systems:
Symmetry
• Vowel systems tend to be symmetrical.
• The minimal vowel system includes /i a u/. All known languages are said to have these three vowels, or slight variations of them.
Rounding
• Back vowels tend to be rounded.
• Front vowels tend to be unrounded.
Can they really be said to be “universals”? And again, are we talking about what is physiologically easier or an actual property of the “language faculty”?
I guess they could be said to be if there were few exceptions to be found. These are more along the lines of stuff a phonetician would be interested in. I studied phonetics mainly to figure out how to accurately describe sounds for students, and I never looked into putative universals.
Furthermore, is it true that Chomsky argues that:
c) human languages are not related to animal communication systems.
and
d) that syntax is what sets human language apart from animal communication systems.
What I know is that human natural languages across an amazing spread of families exhibit similarities in many areas when scrutinized properly. The huge problem that linguists face is that their data is human intuitions about well formedness, not terribly scientific without huge amounts of data.
I have delved into the phonology of a few American Indian languages, Athabaskan, Swahili, Ewe, Igbo, Kanuri, Chinese, Thai, Japanese, Russian, French, Spanish but no animal languages.
I just listened to a woman on the Big Picture Science podcast who is teaching dolphins words using their own whistle and click sounds generated via computer, and recognition software to detect if they happen to get it right. Her main interest beyond linguistics of dolphin language is the sociology or etiquette of dolphins. Very cool stuff.
In my opinion, what sets human language apart from animal communications and thought is a very tricky question. First, we have the vocal mechanism for a wide range of production of sounds. We have an ability to engage in abstract thinking, which is probably needed to form and infer syntactic rules. If I had to guess, I would think that stone tool making, sophisticated hunting and fire maintenance became so important to an improved life style at some point that it required more than rudimentary communications to maintain one's immediate group in safety and comfort.
There is what can be called syntax of animal communications, but it is much different qualitatively from that of human languages.
Well, it could be that they are arguing that grammar is not separate from semantics or that language is not separate from broader human cognition. I could be wrong, though.
I don't know, but the concept sounds foreign to me. Some have tried to treat semantics in a similar way to how syntacticians deal with grammar, but it's a very tough nut to crack.
Yes, I remember reading about the wug test in Pinker’s books Language Instinct and Stuff of Thought. To be honest, I am not really sure how the wug test is supposed to be any kind of evidence for Universal Grammar or deep structure.
The notion of "deep structure" can be pushed too far as a matter relating to human performance. My own experience was that this was not something that should ever be done. Leave that to the psychologists.
All the child has to do is recognize that there is a convention for plurals and apply it. Why would we expect a child to otherwise apply a “random sound” to pluralize something? And why would we expect the /z/ or the /es/ sounds to appear when they are more difficult to produce than a /s/? I’m not sure what the point of that is given that children don’t pluralize church as church/z/ or church/s/. I doubt it has anything to do with deep structure.
That's all? Here we have an organism that is receiving noises from a hole in the face of a huge organism who feeds it. Often noises are accompanied by eye contact and loud, threatening or soft cooing sounds. This noisemaker utters a string of sounds and the organism applies a sophisticated linguistic "convention" properly for producing a syntactic plural suffix. That's all?
I'm not concerned about "difficult to produce". That has nothing to do with it. I think I am beginning to understand why you are having trouble with universals.
The underlying plural for English nouns is the morpheme /+əz/. In surface structure, what we produce or hear, this can become either [-s], [əz], or [z] depending on what
sound the singular noun has in final position, i.e. according to a phonological rule.
There is no universal rule that predicts plural suffixes in languages, and there is no phonetic principle that makes [tops] easier to pronounce than *[topəz]. There are plenty of languages that, in principle, would pronounce the plural of [bag] as */baks/, using an assimilation rule of phonology to devoice the /g/ to /k/ when adjacent to the unvoiced /s/--again, hypothetically.
The point is that the child has made a sophisticated linguistic generalization about a phonological rule in deciding that /wugz/ would be preferable to */wugəz/ or */wugəs/ in his response to the request for a plural of [wug]. Ease or difficulty of pronunciation is not an issue nor can his correct derivation and pronunciation be claimed to be a random guess.
As for "evidence for deep structure"? I'm not sure what kind of evidence one would be looking for. It's just how phonology works, whether or not the brain actually does a stepwise derivation is not an issue that I would ever argue. I do know that generative phonology works for describing language morphology, and the experts who deal with speech production, language acquisition and FMRI brain studies may or may not find such research helpful.
In fact, Gleason herself seems to be bemused by the claims of innatists over the wug test when all she was looking to investigate was the ages at which children apply certain morphological rules in English:
Your pdf won't let me copy/paste. What she said was, that her study showed that children
clearly do have rules. She didn't set out to confirm or make claims about innateness, but it seems she actually fell into an area that makes it difficult to avoid such claims.
http://ltprofessionals.com/journalpdfs/Vol9no1/Murphy-GLEASONinterview.pdf
Yes, but what can also complicate language acquisition at later ages is more complicated lives of adults who largely will not live their day to day lives in a second language which is something that young children can do. For adults, more of their lives have been encoded in their first language, and they have friends and relatives who speak in those languages.
Sure. Lots of factors, but the LAD is a very compelling claim to me when I see children progressing in sentence formation. I have always felt that perhaps for the child learning his first language, it is really tantamount to learning about the real world. The linguistic symbolism for objects and relational concepts is reality in a way that it can perhaps never be as clearly felt again using different symbols. I am not wedded to this idea though.
Trying to learn a new language at a later age, is first of all, by no means impossible, but second of all, could be complicated by other factors than simply the UG switch turning off. The most salient example of something appearing to be determined by some “critical period hypothesis” is accent, which may come down to physiological rather than some hypothesized dissipation in the UG-posited “Language Acquisition Device” (the latter we have no evidence for).
It is rare for an older learner of a second language to achieve total mastery of the accent, but not unheard of. It seems like puberty is the cutoff point, with some exceptions of course.
However, you seem to have produced two pieces of evidence of what UG supposedly accounts for:
a) that children learn morphological rules without being explicitly taught
b) that after a Critical Period (puberty) the UG switches off, and hence adult second language learning becomes more difficult
Yet, adult learners also make over-generalizations of grammar. A university student of mine today wrote the sentence “We eated the cake” by over-generalizing the past tense rule. I expect there are large corpora of similar mistakes among second language learners.
Children "acquire" rules through generalizations from language input provided by speakers in their environment. Sure, mistakes are made by all speakers. Transcripts of conversations are horrendously full of all kinds of errors. I still can't get the meaning of
hoi polloi right. I keep using it to mean the wealthy! So what? We generalize, we overgeneralize, and then we use words like "who'd of thunk it?" Some humorous analogy of [sink, sunk] and [think, thunk]?
The objections are really that:
1) UG fails to provide explanations for language acquisition that cannot be explained in other ways.
2) That it is thus unnecessary.
3) That it is unfalsifiable
OR
4) trivial
1) I suppose you could posit huge memory banks that randomly spit out every utterance until the correct one makes it past a special pleading filter in the brain, or sumpin'. Be my guest.
2) Meh! Ok.
3) The best test for falsifiability is actual research with human informants. There is such a thing as an ill formed sentence. In fact, there are ill formed bilingual code-switched utterances in bilingual communities.
4) Could be trivial. Most pursuits are in the grand scheme of things. Occam's Razor is after all a noble yardstick. I find E=R*I trivial now, but it sure did take a lot of effort and working with electronics to get there.
It took me an entire year in grad school until I finally figured out what the problem was these guys were trying to solve. Once I caught on, I was hooked. My dissatisfaction with the field of linguistics was that I didn't see it going much beyond anecdotal analysis, and churning out papers full of arcane stuff no one would ever read.
That is, I saw theoretical syntax as a field in search of a theory, and I was not a Chomsky, a Lakoff, or a Haj Ross. I was not a
whiz kid. I figured that I could not do any better at solving this problem than they had, so after 7 years I got out and changed careers. I don't regret the experience, nor do I regret getting out.
