Linguistics from 50,000ft.

June 13, 2006 2:34 PM

What are the major unanswered questions in linguistics?

I'm familiar with what linguists do on a day-to-day basis, but I'm looking to get a sense of the big picture and how it differs from the big picture in the philosophy of language and logic. Is there agreement on the big picture or are there several different big pictures with competing research programs?

Some examples of big questions (some of which have been answered, of course):
Biology: What is the chemical structure of DNA? How does DNA create organisms?
Biochemistry: How could life arise from chemical soup?
Neurobiology: How are memories created, stored, and recreated in the brain?
Physics: How do we reconcile quantum physics with relativity?
Psychology: Is development more nature than nurture?
Philosophy: What is knowledge? What is the relationship between the mind and brain?

posted by ontic to science & nature (21 comments total)

Some linguists like to argue about the Altaic Theory
posted by thirteenkiller at 2:43 PM on June 13, 2006

isn't there a big fight between chomsky, lakoff, and pinker? or is that more philosophy of mind/neuropsych?
posted by goethean at 2:52 PM on June 13, 2006

hey, wow. the first hit is to languagehat's blog .
posted by goethean at 3:06 PM on June 13, 2006

Just off the top of my head:

Basque
Will we ever get to FAHQT?
How did protolanguages develop?
A whole slew of undeciphered languages; my favorite is the Phaistos Disc but that's just me.
Do animals use language?
posted by TheRaven at 3:20 PM on June 13, 2006

TheRaven's list is excellent, except that I'm not sure what he means by "Basque," and his link is borked, so I guess I'll never know. (If the question is "Was Basque developed and implanted by aliens?" the answer is "Certainly!")
posted by languagehat at 3:26 PM on June 13, 2006 [1 favorite]

Here's a random list from the perspective of mainstream generative linguistics (which most of the questions mentioned above aren't really tailored to). Also, "do animals use language [in the way that humans do]" is not particularly unanswered in linguistics -- the answer is "no". Some animal language researchers (who tend not to be linguists) may disagree. Some researchers might (implicitly or explicitly) remove the bracketed part and play with what they mean by language when addressing this question.

Why are there island constraints? How does ellipsis work? What are the semantics of conditionals (that link's actually pretty awful)? Are there functional bases for constraints in the grammar (couldn't find any good links for this; by "functional" I mean grounded in constraints on language processing or production, i.e. psycholinguistics/phonetics)? Where is the boundary line between semantics and pragmatics? (In fact, where are the boundary lines between phonetics, phonology, morphology, syntax, semantics, and pragmatics?) Why is there redundancy in language (e.g. languages that use both case marking and word order to indicate grammatical function)?

Perhaps the largest unanswered question is how all of linguistic theory corresponds to whatever the brain is doing when it processes language. This is what, if we knew the answer to it, we'd probably be done.

are there several different big pictures with competing research programs?

At the largest level, not really. In subfields there are sometimes competing theories -- e.g. there are 4 or 5 competing views of syntax that are popular somewhere (minimalism, which is a descendent of government and binding theory, lexical functional grammar, head-driven phrase structure grammar, tree adjoining grammar, functionalist syntax), as well as a few more that have got lost somewhere along the line. But there is actually quite a bit of agreement among these if you don't get too caught up in the details.
posted by advil at 3:45 PM on June 13, 2006 [1 favorite]

That list is by no means exhaustive or representative, by the way. Linguistics is an extremely young field!
posted by advil at 3:49 PM on June 13, 2006

Isn't there something called the "horrible experiment" or some such, whereby a bunch of newborns would be left to grow up in isolation to see if they would spontaneously develop a language?
posted by StickyCarpet at 3:56 PM on June 13, 2006

Can we demonstrate (weaker than proof, but stronger than "look at those big brains-- they must have") that Neanderthals had language (or did not)?

How has language, over evolutionary time, shaped the human phenotype?

Is the origin of language associated with an abrupt change in the human genotype? What subsequent changes has language selected into our genes?

Are all contemporary languages lineal descendants of the first? Are any?
posted by jamjam at 4:23 PM on June 13, 2006

We still don't know have an adequate theory of the relationship between socialization and acquisition. A lot of sound and fury, a raft of theories that don't explain enough, or explain too much, etc.

As for the "horrible experiment," it's been done, sort of. See Russ Rymer's Genie or the PBS NOVA episode based on the book. In principle, this experiment addresses the lacuna above. In practice, there is no way to determine this matter experimentally, since it is clear that the neurobiological faculty is, minimally, paramaterized by socialized learning processes.

The greatest question of all, of course, is what is the relationship between language and thought. All linguistics is in some sense oriented to this question eventually, and despite the polemics of schmucks like Stephen Pinker, we really have only the faintest idea.
posted by fourcheesemac at 5:22 PM on June 13, 2006

I recommended this in another thread, but Derek Bickerton's Language and Species is one of the most interesting books I've ever read on the nature of the "big question" in linguistics.
posted by fourcheesemac at 5:24 PM on June 13, 2006

As for the "horrible experiment," it's been done, sort of. See Russ Rymer's Genie...

As much sympathy as I have for Genie, I don't think she fits the criteria. She might not of been spoken with or even spoken to but the whole time she was strapped to that potty seat in a back bedroom she was exposed to speach. And there was no one else like her there, for her to attempt speech with.
posted by StickyCarpet at 6:47 PM on June 13, 2006

Great stuff everyone. I suppose I actually meant the more cognitive science side of linguistics than the historical/etymological side. But thanks for both (and more of both).
posted by ontic at 6:59 PM on June 13, 2006

How modular is language in the brain?

Not only, as advil asks, "how all of linguistic theory corresponds to whatever the brain is doing when it processes language," but how does it process multiple languages?
posted by kmel at 7:54 PM on June 13, 2006

I think you may be thinking of a story from Herodotus for the "horrible experiment". He tells of a king who wanted to know the first language. This king had two children raised along with no one speaking to them. Their first word was the Phrygian word for bread, thus the king said the first language was Phrygian. I don't remember where this is in his histories, but google helps pull up a wikipedia entry for the Phrygian language which mentions the story.
posted by bleary at 8:40 PM on June 13, 2006

Is there a Universal Grammar? See also Nicaraguan Sign Language.
posted by sundress at 9:15 PM on June 13, 2006

As much sympathy as I have for Genie, I don't think she fits the criteria. She might not of been spoken with or even spoken to but the whole time she was strapped to that potty seat in a back bedroom she was exposed to speach. And there was no one else like her there, for her to attempt speech with.

Actually, the point Rymer ends up making. In no sense was Genie the controlled "wild child" experiment S. Curtiss and her linguist colleagues wanted her to be, and in any case the very conception of "innateness" was so rudimentary then.

The case is dealt with quite interestingly in the Bickerton book I recommended upthread, along with other "wild child" claims. There is simply so much better evidence for innateness, critical periods, paramaterization, the naturalistic basis of word order in all new languages, and the majority of the worlds existing languages, etc. in primate studies, brain studies, in the structure of language grammars themselves, in the process by which new languages have been observed to emerge, and in an examination of the evolutionary record than could ever be garnered from in vivo experimental studies. Far more interesting are the many close descriotions we now have of normal childhood language socialization in different cultures (work for which Elinor Ochs won a MacArthur, as well as the aforementioned NSL research by the Senghases, which documents a classic creolization process and the normalization of grammar within one generation of children seemingly deprived of "normal" linguistic input, and the subsequent elaboration of that grammar by subsequent generations into a full scale language.

Good times, all this.
posted by fourcheesemac at 4:09 AM on June 14, 2006

My general sense is that linguistics has not yet developed into a science with anything near to having any facts (save phonetics maybe). As a result, things are pretty much wide open with not a lot of consensus on any front.
posted by bluesky43 at 5:12 AM on June 14, 2006

linguistics has not yet developed into a science

That's ridiculous, unless you think physics and chemistry are the only sciences. Linguistics is as much a science as anything that deals with humanity can be (no controlled experiments, no 100% certainty, etc.). Many linguists don't seem to grasp the scientific method and think using pseudo-mathematical formulas is an acceptable substitute (Chomskyans, I'm lookin' at you), but that's another matter.
posted by languagehat at 6:51 AM on June 14, 2006

no controlled experiments

Actually, there are plenty of controlled experiments, and the field is moving even more in an experimental direction (in that people who don't specialize in pyscholinguistics now typically have to understand how to run some basic experiments). I think the reason there seems to be so little consensus is just that linguistics as a science is so young, and turned out to not have a ready solution waiting to be found (like other cognitive sicences). Actually, I also do think there's a lot of consensus, but as linguistics isn't taught in middle school or whatever, not a lot of people know about it.
posted by advil at 10:20 AM on June 14, 2006

linguistics has not yet developed into a science

Are you joking? Modern cognitive science is nearly founded on the findings of linguistics. Not all science is conducted in the form of "controlled experimentation," a layperson's view of the nature of the enterprise, though there is plenty of controlled experimentation, as advil says. The formal description and formal analysis of grammar is a very rigorous enterprise, subject to constraints of replication, hypothesis testing, and reanalysis. Linguistics also includes a good deal of rigorous social scientific methodology.
posted by fourcheesemac at 2:21 PM on June 14, 2006