Re: Pirahã

1

How did you like the book? I thought parts were fairly silly, but I liked it very much anyway.


Posted by: Thorn | Link to this comment | 03-22-12 6:57 AM
horizontal rule
2

I loved it. There are some very valid criticisms, but I was hooked and enamored the whole way through.


Posted by: heebie-geebie | Link to this comment | 03-22-12 7:00 AM
horizontal rule
3

Clearly the only acceptable course of action for Chomsky is to declare that the Piraha people are not human and demand their extermination.


Posted by: dsquared | Link to this comment | 03-22-12 7:27 AM
horizontal rule
4

The historical linguistics guy at Har/vard whose wife languished as a lecturer in Sanskrit despite lots of people pushing for her promotion was supposed to be super nice. He's retired now, and she got a job at UCLA where he's also affiliated.


Posted by: Bostoniangirl | Link to this comment | 03-22-12 7:28 AM
horizontal rule
5

(hmmm, that joke looks a lot more tasteless and horrible written down in black and white than while echoing in my brain. I still think that the tendency of NC and his epigones to make broad and categorical statements about the nature of humanity is dangerous and silly, but very sorry for that particular expression).


Posted by: dsquared | Link to this comment | 03-22-12 7:29 AM
horizontal rule
6

2: It found it totally captivating too.


Posted by: Thorn | Link to this comment | 03-22-12 7:29 AM
horizontal rule
7

Exterminate dsquared!


Posted by: heebie-geebie | Link to this comment | 03-22-12 7:31 AM
horizontal rule
8

Huh. Whoever's actually right, I'm really surprised that it's at the level where it's an important controversy. You have one language in the whole world that's alleged to lack recursion. And you have only one linguist who speaks it to witness to that claim. This doesn't mean that he's wrong or that Chomsky's right about universal grammar generally, but I wouldn't think one person's claims about Piraha would be enough to make people take it seriously as a counter-example to 'all languages have recursion' -- wouldn't that still be at the cold fusion of 'that's fascinating, let's see if anyone else can duplicate the research?'


Posted by: LizardBreath | Link to this comment | 03-22-12 7:35 AM
horizontal rule
9

How often does it come up in linguistics than other fields that someone's work is essentially unreproducible because no one else is an expert? Often enough that you can't discount that kind of research?


Posted by: heebie-geebie | Link to this comment | 03-22-12 7:39 AM
horizontal rule
10

Where's Cecily?


Posted by: heebie-geebie | Link to this comment | 03-22-12 7:39 AM
horizontal rule
11

(something happened to the last sentence of 8. If everyone can edit to make the reference to cold fusion comprehensible, I'd appreciate it.)


Posted by: LizardBreath | Link to this comment | 03-22-12 7:40 AM
horizontal rule
12

Also, the article talks about how he uses a previous missionary's transcription of the Pirahã dialogue as a unbiased second source of their language.

Another guy says that the second source shows rudimentary recursion. No one seems willing to say "Hey, the answer is 'sort of'."


Posted by: heebie-geebie | Link to this comment | 03-22-12 7:42 AM
horizontal rule
13

9: My guess is that there might be a fair number of languages in the "Only one academic speaker ever" category -- small group languages that are no longer spoken and one missionary in 1911 compiled a dictionary and grammar. I'm just surprised that anyone would use such a language as support for a revolutionary claim about the nature of language.


Posted by: LizardBreath | Link to this comment | 03-22-12 7:42 AM
horizontal rule
14

Oh, fuck Chomsky.


Posted by: Sifu Tweety | Link to this comment | 03-22-12 7:43 AM
horizontal rule
15

Who knew that dsquared would turn out to be a Dalek?


Posted by: One of Many | Link to this comment | 03-22-12 7:43 AM
horizontal rule
16

12: Sure, but saying that something doesn't exist in a set of notes seems like incredibly weak support for saying it doesn't exist in a language -- even if it unambiguously supported his position, that's not another person who's actively looking for recursion interacting with Piraha speakers.


Posted by: LizardBreath | Link to this comment | 03-22-12 7:44 AM
horizontal rule
17

I'm just surprised that anyone would use such a language as support for a revolutionary claim about the nature of language.

Why is this surprising? They seem like the most likely candidates to refute things that were taken as ultimate truths.


Posted by: heebie-geebie | Link to this comment | 03-22-12 7:45 AM
horizontal rule
18

8: the fact that this is controversial shows how ridiculously strong Chomsky's nativist claims are. He wants to say that there is a universal grammar module that includes recursion. He can acknowledge that this module will not develop given certain genetic abnormalities. (I think people have proposed some very specific ones for recursion and Chomsky has not gotten snippy about it.) But the claim that an environmental factor like the culture you are raised in is sufficient to block the development of a recursive language is too much for him.


Posted by: rob helpy-chalk | Link to this comment | 03-22-12 7:46 AM
horizontal rule
19

Oh, goody.

(1) There's a real tension between field linguists and theorists, and there always has been. In particular, the best field linguists aren't always the best theorists. Here, it ought to be troubling that Everett is the main authority on Pirahã, and a few more people should get real hands-on experience with the language and culture. It's agonizingly uncomfortable and demanding work, though, with little payoff--small wonder Everett started as a missionary.

(2) On the substantive point, recursion isn't necessarily a unitary property of languages. Some constructions have it, some don't. The idea that Pirahã has none at all, and that UG requires that it have it for every construction class, that's just strawman-on-strawman action. More savvy folks like Pullum and Jackendoff are going to turn out right. And what the hell, in Minimalism it isn't even clear what 'UG' even means any more--it's a 'perfect' interface property, so for all anyone knows it can vary in how much embedding it manifests.

(3) Making this about the personalities is a really stupid narrative frame, but journalists can never resist it. Annoyingly, neither can the participants. Chomsky himself has never been able to keep from responding when provoked--he's admitted as much before, and it's not a wonderful characteristic. Personally, he seems like a cipher to me (the New Yorker article on him years back seems to confirm this), so the idea that he's some sort of saint is baffling.

In any case, as a sociological point, both insiders and outsiders wildly exaggerate the influence of Chomsky and his current research program on the linguistic theory. There are tons of very healthy alternatives, and lots of heterodoxy even within mainstream generative grammar. (The old school generative semantics wars were a different story...)


Posted by: Man Suit | Link to this comment | 03-22-12 7:51 AM
horizontal rule
20

Do important (at least according to the Chomskyites) questions in linguistics actually hinge on the possibility of infinite recursion?


Posted by: Eggplant | Link to this comment | 03-22-12 7:52 AM
horizontal rule
21

Having recently read James C. Scott's The Art of Not Being Governed, I am now an expert on remote tribes. The argument in that book is that groups such as the Hmong, instead of being "uncivilized" and lagging in societal development, have deliberately avoided being subjects of a state by making themselves unreachable both in physical location and in cultural practices. The Hmong, if I remember correctly (this comes from a book called The Spirit Catches You and You Fall Down, which I read for a class maybe 10 years ago), have a language unlike any of their neighbors' as well. Scott doesn't say that current members of such groups necessarily commit to this consciously, but that they have inherited the culture. I wonder, assuming it's as unique as Everett says, if the Piraha language developed with the express intention of being incomprehensible.


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 7:52 AM
horizontal rule
22

I don't know what I'm talking about with linguistics, and I hold no brief for Chomsky. But whether or not grammar is universal, as far as I know out of thousands of languages studied, they all show recursion until Piraha. Regardless of the theory, that seems to me like awfully strong support for the presumption that recursion (as opposed to Chomskyian universal grammar generally) is a universal characteristic of human language.

Now you've got a guy claiming, on the basis of work that no one is in a position to replicate, that he's found a counterexample. When you have a presumption based on evidence as strong as that for the idea that recursion is universal (I could be wrong about the evidence, but that's what I believe it to be), one counterexample that can't practically be replicated doesn't seem like enough to shake it. It does seem like enough to send linguists into the jungle and learn Piraha to check, but if I were a linguist I wouldn't take the claim as a serious threat to any theory until another linguist Piraha speaker supported it.


Posted by: LizardBreath | Link to this comment | 03-22-12 7:54 AM
horizontal rule
23

How often does it come up in linguistics than other fields that someone's work is essentially unreproducible because no one else is an expert?

Depends what the subfield and the language are. For rare (meaning understudied, not necessarily that there are few speakers) languages, it's pretty easy to get crappy stuff published. This has been true for decades about research related to any signed language- no one on the peer review knows enough about it to be able to tell what's going on, and (until recently more, but still true now) many or most of the researchers aren't native or even fluent users, so they just make up stuff and everyone is impressed with how weird and new and interesting it all is.

Because of my experience with that, I'm pretty skeptical about one dude who is not a native user being able to make large-scale accurate claims about Pirahã. I am not saying he's a fraud, just that it's really easy to get things wrong the first time, and really similar things have happened before many times. (Hopi has no way to express tense, Eskimos have 5,628 words for snow, ASL has no contrastive sequence, etc etc).

On the other hand, N-C is a cranky old man who has totally dug himself into never changing his mind or considering new evidence about anything. In my studied opinion.

The corpus guy's work seemed the most likely to be right, to me, but I'm not a syntactician and I do not know Pirahã so who knows.

Anyhoo, I have to go teach, but I'll be back later.


Posted by: E. Messily | Link to this comment | 03-22-12 7:55 AM
horizontal rule
24

22 crossed with 19 et seq. 19 particularly sounds smart.


Posted by: LizardBreath | Link to this comment | 03-22-12 7:55 AM
horizontal rule
25

I wonder also if I'm not just falling into the "the latest interesting thing I've read explains everything!" trap.


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 7:56 AM
horizontal rule
26

Personally I prefer Nick Evans and Steve Levinson to Dan Everett.


Posted by: Nakku | Link to this comment | 03-22-12 8:00 AM
horizontal rule
27

Making this about the personalities is a really stupid narrative frame, but journalists can never resist it. Annoyingly, neither can the participants.

Agreed. The Piraha language, at least as presented by Everett, is deeply fascinating in its own right. It's very frustrating to keep on seeing it presented solely as a refutation of or support for Chomsky.

Do important (at least according to the Chomskyites) questions in linguistics actually hinge on the possibility of infinite recursion?

According to the Chomskyites, yes.



Posted by: Ginger Yellow | Link to this comment | 03-22-12 8:04 AM
horizontal rule
28

To expand on 20, I have the impression that Chomsky relies on infinite recursion to prove the unlearnability of language, which, if so, seems ridiculous. But, as always, my speculations are unencumbered by knowledge.


Posted by: Eggplant | Link to this comment | 03-22-12 8:04 AM
horizontal rule
29

||

I've just agreed to go to a midnight showing of The Hunger Games. This somehow involves leaving the house at 7:30 pm, which, alas, does not allow time after work to help my niece paint flames on her pants. (I could've said jeans, but pants is funnier.)

|>


Posted by: Sir Kraab | Link to this comment | 03-22-12 8:04 AM
horizontal rule
30

I have a friend who works in this field [her team is cited in the paper] so I've sent it to her for comments. She does theoretical stuff, and field research. It does seem like a field fraught with political difficulties. She needs access to tribe X that person Y has been working with because they do something odd with linguistic-thing-Z, which means horse-trading with Y on some other thing P, etc.

I wonder, assuming it's as unique as Everett says, if the Piraha language developed with the express intention of being incomprehensible.

There are deliberate cryptolects. Something I've always found fascinating.

http://en.wikipedia.org/wiki/Cant_%28language%29


Posted by: nattarGcM ttaM | Link to this comment | 03-22-12 8:06 AM
horizontal rule
31

Really, "infinite recursion" should be called "arbitrarily long recursion". I mean, please.


Posted by: heebie-geebie | Link to this comment | 03-22-12 8:10 AM
horizontal rule
32

I wouldn't even spot them "arbitrarily long". Do you think there are any people who regularly speak with more than a three or four deep recursion? Besides nosflow, I mean.


Posted by: Eggplant | Link to this comment | 03-22-12 8:14 AM
horizontal rule
33

My two favourite* things in that New Yorker article were the moment an audience at one of his lectures burst out laughing, to his astonishment: turns out a tiny orange kitten had emerged from under a curtain onto the stage behind NC, which he couldn't see but the audience could.

The other is that he and his wife enjoyed Law and Order. (The TV show not the oppressive social structure.)

*In the accurate sense that I recall no other things from the piece.


Posted by: tierce de lollardie | Link to this comment | 03-22-12 8:15 AM
horizontal rule
34

Is nosflow a sufficient counter-example? The whole problem with the Pirahã is that at most one other person understands them.


Posted by: heebie-geebie | Link to this comment | 03-22-12 8:16 AM
horizontal rule
35

33: The McSweeney's series of LOTR commentary by NC and Howard Z/nn is my favorite example of NC miscellany. I've got an excerpt taped to my office door.


Posted by: J, Robot | Link to this comment | 03-22-12 8:21 AM
horizontal rule
36

There are deliberate cryptolects. Something I've always found fascinating.

It is fascinating. Wouldn't we expect a child born to a group that spoke exclusively in cryptolect, and also moved really far away from anyone who spoke anything else, to, thanks to total immersion, demostrate an unusual pattern of linguistic development as immersion within that language superseded or at least modified the typical process? And wouldn't we therefore expect the language itself to demonstrate something similar?


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 8:27 AM
horizontal rule
37

The TV show not the celebrating oppressive social structure.


Posted by: CharleyCarp | Link to this comment | 03-22-12 8:32 AM
horizontal rule
38

||

For those following along with the Birdman-in-Clogs story yesterday, the Very Serious People (incl. Physicists, so science) who helped break the "story" at Wired now say that the Flying Dutchman's story is beginning to unravel, because none of his work or school references check out. "WHAT ELSE IS UNTRUE?" they ask dramatically. I can't imagine.

|>


Posted by: Gonerill | Link to this comment | 03-22-12 8:33 AM
horizontal rule
39

The idea that this is a 'controversy' really needs to be reined in, too. Whether it is or not depends on all sorts of other theoretical assumptions, some of which are pretty speculative. Sure, it's now a core tenet of Minimalism that all human languages manifest recursion. But (see above) recursion is a degree property. How much is needed, and where? The idea that it's somehow a death blow to UG if you don't find it maximally, everywhere, is just crazy. (I add: whatever UG now means, which is itself a seriously underexplored question.) So, anti-Chomskyians also need to cool it a bit.


Posted by: Man Suit | Link to this comment | 03-22-12 8:36 AM
horizontal rule
40

It's hard for me to read that article as its so annoyingly "journalistic." The whole point is to pick fights, not to inform.


Posted by: Unfoggetarian: "Pause endlessly, then go in." (9) | Link to this comment | 03-22-12 8:37 AM
horizontal rule
41

28: Actually it's a big deal trying to figure out what role learnability now plays in arguments over the structure of grammar. It used to be clear, but these days it's much murkier. There's an old argument that if you don't have something recursive already you can't *learn* it, so recursion at least must be innate. But to put it mildly, figuring out the recursive structure of thought is a lot harder than doing so for language.


Posted by: Man Suit | Link to this comment | 03-22-12 8:41 AM
horizontal rule
42

41.last: also, it might be totally irrelevant.


Posted by: Sifu Tweety | Link to this comment | 03-22-12 8:47 AM
horizontal rule
43

I am deliberately not reading about child language development while Mara's in the middle of it, but I really look forward to learning more when she's older. There are a lot of kids like her who have speech delays because of early neglect and I want to know whether her language development is like other kids with that background and how it differs from normal patterns of language development. But like I said, I'm not going to learn about it now because I don't want to intrude on her process or worry about her. She's doing fine.

But on the topic of potentially racist language comments, Lee and I were discussing last night how many of Mara's ways of talking (lack of possessive pronouns and possessives in general, "got" for "has," lots more that I'm not thinking of right now) fit AAVE patterns and whether there's a semi-subconscious "My toddler already learned not to do that!" kneejerk reaction to ebonics.


Posted by: Thorn | Link to this comment | 03-22-12 8:49 AM
horizontal rule
44

43: my undergrad advisor enjoyed reading Wittgenstein when his son was younger--might be a reasonable compromise.


Posted by: J, Robot | Link to this comment | 03-22-12 8:54 AM
horizontal rule
45

43: Is this where I confess that I'm not sure I've ever truly enjoyed reading Wittgenstein? Perhaps I should try again!

I should probably respond to the real post. I'm uncomfortable with assuming too much from one sole expert on the language, but that doesn't mean I'm comfortable with UG either.


Posted by: Thorn | Link to this comment | 03-22-12 9:02 AM
horizontal rule
46

44: Yes, I recall many happy hours playing Slabs and Pillars with my dad when I was young. The later lessons in genius, obsession, suicide, and madness were more unsettling.


Posted by: Man Suit | Link to this comment | 03-22-12 9:03 AM
horizontal rule
47

Argh, 45 to 44. Any linguistic theories on why I always, always do that wrong?


Posted by: Thorn | Link to this comment | 03-22-12 9:03 AM
horizontal rule
48

43: Huh. I wonder if a primary/exclusive speaker of AAVE would identify some toddler speech patterns as sounding like SAE -- if there'd be a symmetry, where what the toddler was doing was omitting grammatical subtleties generally, and depending on your native dialect you notice the missing subtleties in relation to other dialects that lack them. I don't think I quite said that clearly, and I certainly don't know if it's true, but I'd be interested to find out.


Posted by: LizardBreath | Link to this comment | 03-22-12 9:07 AM
horizontal rule
49

42: Pistols at dawn, sir! (Trying to get into the combative spirit of the original piece.)


Posted by: Man Suit | Link to this comment | 03-22-12 9:10 AM
horizontal rule
50

re: 36

I'm not sure I get what you are saying. Most cryptolects aren't that unusual linguistically. Some are just collections of slang or bits of deliberately obfuscated vocab; others are more or less normal complete languages but which are used for in-group conversations to exclude outsiders.


Posted by: nattarGcM ttaM | Link to this comment | 03-22-12 9:11 AM
horizontal rule
51

Making this about the personalities is a really stupid narrative frame, but journalists can never resist it. Annoyingly, neither can the participants.

Nicely recursive in itself!

Seriously, "humans have recursion, therefore suggesting they don't is suggesting they aren't human, therefore RACIST!!!" is one of the most stupid things I've heard...ever.

Has UG actually made any progress in the last however many? On the rare occasions I think about it I can't help remembering the dormitive power.


Posted by: Alex | Link to this comment | 03-22-12 9:13 AM
horizontal rule
52

48: I honestly don't know. Mara will say "I wan' go my grandma house!" and we attribute that to her speech not being as advanced as it eventually will be, whereas our adult friend who talks about being at "my mama hou'" is just speaking her dialect. Mara's youngest older sister, who was not raised with her but had a similar babyhood, has speech patterns very similar to Mara's. I wouldn't say either was raised to speak AAVE in a major way, though the sister (being raised by an aunt) is exposed to a lot more of it than Mara is with us.

One of my friends had a very opinionated kindergartener fostere daughter who'd fight with her when she'd say "You'll go in your daddy's car." because "I only have ONE daddy, so my daddy car!" She'd internalized the rule much more than any of Mara's siblings have, as they all code-switch to varying degrees.


Posted by: Thorn | Link to this comment | 03-22-12 9:15 AM
horizontal rule
53

I don't know anything beyond the merest pop version of linguistics, but I had the impression there was something to the argument that babies learn language much better than can be explained by the amount of data they have, so they have to be relying on some hardwired apparatus that tells them what human languages can possibly be like. And pidgins turning to creoles, that kind of thing -- if you give a kid language-like input, even if it's not a real language, it will generate something that looks like a conventional language out of that input.

I don't know that those claims either stand up or say much, but that's what I think of when people talk about grammar being innate.


Posted by: LizardBreath | Link to this comment | 03-22-12 9:17 AM
horizontal rule
54

The poverty of the stimulus thing has been kicked around a lot. It's been a long time since I studied this, but I seem to recall current [then-current, probably not now-current] thinking was that, 'actually, we get exposed to a fucking giant amount of language; so it's not really that amazing, given babies-as-experience-sponges'.


Posted by: nattarGcM ttaM | Link to this comment | 03-22-12 9:20 AM
horizontal rule
55

52: I'm the wrong person to theorize about this, because I'm really, really not a fluent speaker of AAVE. But something like the stressed 'been' to indicate that something's been happening over a long period of time, where AAVE does something that SAE doesn't -- I wonder if a toddler growing up in an AAVE household might not pick that up until later, and so would sound kind of SAE-ish until they did.


Posted by: LizardBreath | Link to this comment | 03-22-12 9:20 AM
horizontal rule
56

53, 54: Like I said, I'm really not reading up on this, but there's also a big difference between babies who are just sponges and babies who have interactions with people, and the latter part is really important. I'll never know what Mara's first two years were really like, but the general theory is that she spent her time either watching tv or watching grownups do whatever they were doing and that it was rare anyone would consistently interact with her in a way that required response from her. It wasn't like being in a Romanian orphanage or anything, but it definitely impacted her ability to speak and her attachment and trust for adults.

I really feel that she understands everything, probably at a muchhigher level than the average kid her age (and her teachers agree) but her expressive language is still low-normal. At 4, she has a hard time with things like "over" versus "under" that I suspect she classifies as "one of those place words" in the way someone learning a foreign language as an adult might, whereas the average baby probably picks it up from narrated play in interaction with the parent.

Last night at dinner she said, "Chicken is spicy and chicken is not spicy," which I was able to clarify was not a koan but her way of saying "Some/sometimes chicken is spicy and other/sometimes chicken us not spicy." While she'd have zero problem understanding that, it's not yet a construction she can make herself. (Her previous example was "Babies have vulvas and babies not have vulvas." and that led to a conversation about anatomy rather than grammar.)


Posted by: Thorn | Link to this comment | 03-22-12 9:32 AM
horizontal rule
57

53, 54: The poverty of the stimulus argument still gets a lot of play, since language displays some puzzling developmental features relative to other things children learn.

However, there have been three major trends that push back against it: (1) new grammars that use less recondite structures; (2) corpus linguistics (a la Pullum) which enriches our view of the data children are exposed to; and (3) much more powerful general mathematical learning theories that don't have the limitations of the ones originally criticized (which were crudely associationistic). Putting these together takes a lot of the sting out of the nativist argument, I suspect.


Posted by: Man Suit | Link to this comment | 03-22-12 9:36 AM
horizontal rule
58

At 4, she has a hard time with things like "over" versus "under"

Joey confused "yesterday" and "tomorrow" until about age 5.


Posted by: rob helpy-chalk | Link to this comment | 03-22-12 9:40 AM
horizontal rule
59

My 4yo still says "This Day" for "today" which often makes him sound like Gandalf.


Posted by: Gonerill | Link to this comment | 03-22-12 9:43 AM
horizontal rule
60

I wonder if the ability to find recursive grammar could be limited by a cultural preference - a fashion, basically - for short, simple sentences. That is, they could do relative clauses and so forth, but it would sound stupid to peers.


Posted by: Minivet | Link to this comment | 03-22-12 9:43 AM
horizontal rule
61

58: I think that's normal. I'm pretty sure Mara knows those and she definitely knows the days of the week in order and which day it is, but she doesn't do well with prepositions at all. "Upstairs" and "Downstairs" are often backwards unless she stops and thinks. Over/under, back/front, top/bottom, they're just not automatic to her yet. When she graduated from speech therapy, this was the one place where she was still way, way low. I've finally sort of started drilling her on them since six months later she still hasn't picked them up. "She" versus "he" is another problem for her, I think for similar reasons.


Posted by: Thorn | Link to this comment | 03-22-12 9:45 AM
horizontal rule
62

It was only in my mid-20s that I stopped saying "November" when I meant "September", and I still sometimes say "Updike" when I mean "Roth".


Posted by: nosflow | Link to this comment | 03-22-12 9:46 AM
horizontal rule
63

50

Admittedly I'm trying to shoehorn this Piraha story into my understanding of the Scott book, and I'm very much not an expert on anything in this discussion. But I was just thinking that it's possible Piraha culture as a whole came about as a purposeful attempt to simply walk away from and not participate in the larger culture of the region. The most obvious way a group could do this would be to get to the least geographically convenient location possible, in this case far away and in the middle of the jungle. Another method Scott discusses is to avoid a history, which the article claims the Piraha lack. Language can also be part of that strategy, but it would stand to reason that at least initially the group's language wouldn't be wholly original; instead it would be some kind of encryption of whatever they spoke before. Over time it could develop into something completely unique, but it would by necessity start out by upending conventions of syntax or whatever. So what I'm saying is that if a child was raised among people who spoke to him and to each other in this unusual way, and was also completely isolated from people who communicated in a more typical way, then he would develop the habits reinforced all around him rather than revert to what we expect to develop naturally. Even if Everett is correct and they don't exhibit a particular characteristic assumed to be universal in linguistics it doesn't therefore prove that that characteristic isn't universal, because the Piraha language didn't develop among a bunch of unsupervised babies. In other words, if the Piraha were all left-handed people Everett wouldn't conclude that some innate right-handedness wasn't a universal characteristic of human societies.

I am very open to the idea that I'm not making any sense.


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 9:49 AM
horizontal rule
64

Apologies for the wall of text. It's basically an attempt to say what 60 said.


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 9:51 AM
horizontal rule
65

How convoluted a sentence do you expect from someone used to communicating by whistling?


Posted by: Eggplant | Link to this comment | 03-22-12 9:53 AM
horizontal rule
66

I've finally sort of started drilling her on them since six months later she still hasn't picked them up.

Does she play with blocks/Lego/buildy-stuff generally? That seems like it'd be a useful way to force a preposition-heavy conversation. ("Where should I put the blue block? Over or under the red block?")


Posted by: LizardBreath | Link to this comment | 03-22-12 10:03 AM
horizontal rule
67

I think Chomskyan linguistics fits something MC John E once said (though he said it about analytic philosophy): it killed American linguistics, but since the massive extinction event it has speciated to fill the newly available ecological niches.

With the benefit of hindsight, we now know that the poverty of the stimulus argument proves very much less than was once thought, and can't bear the weight that Chomsky puts on it. It's true that humans can't just learn anything from scratch, but the big discovery of statistical learning theory is just that there needs to be some sort of in-born "bias", and that bias can be as simple as "prefer simpler grammars to more complicated grammars", rather than any kind of rigid universal grammar that only varies by setting a few parameters.

I'm also skeptical that any human language is actually recursive. They're sort-of recursive, in that you can go from simple sentences to more complicated sentences, but there's an obvious limit to how you can go from more-complicated sentences to even-more-complicated sentences. The sentences get obviously harder and harder to understand, which seems like an important piece of evidence.


Posted by: Walt Someguy | Link to this comment | 03-22-12 10:13 AM
horizontal rule
68

Agree with 51, and LB. But as for my own reading of the article, am I the only person to whom "recursion" seems like such a vague concept that it should be problematic?

I mean, the definition of it seems to be that you can make infinite arbitrarily large sentences by embedding clauses or phrases within each other. From the article,

Some of what [Gibson] found does support Everett. For example, he's confirmed that Pirahã lacks possessive recursion, phrases like "my brother's mother's house." Also, there appear to be no conjunctions like "and" or "or." In other instances, though, he's found evidence that seems to undercut Everett's claims--specifically, when it comes to noun phrases in sentences like "His mother, Itaha, spoke."

So "recursion" includes possessives, conjunctions and noun phrases. Based on other examples, it also includes subordinate clauses and prepositional phrases. So isn't that just everything in language that's not a verb? (OK, also nouns not used in noun phrases, and adjectives. But adjectives are different from nouns in the way that Pluto is a planet, amirite?)

According to the article and the Wikipedia page about recursion, it is supposedly unique to humans. That also seems like it would be hard to prove. Even if none of those dolphins and chimpanzees and gray parrots that have learned some rudiments of human languages ever used recursion - and with such a broad definition, that seems hard to believe - what about all those other "animal communication systems" out there that we don't think of as rising to the level of an actual language?

What do I know, I'm less of an expert than LB, the did article try too hard to be dramatic, and maybe this actually is rigorous and I'm misunderstanding it. This reminds me about an XKCD cartoon about how the only pure, hard science is mathematics, which now seems funny because math and linguistics both care about recursion.


Posted by: Cyrus | Link to this comment | 03-22-12 10:21 AM
horizontal rule
69

tr; dr


Posted by: JP Stormcrow | Link to this comment | 03-22-12 10:29 AM
horizontal rule
70

It's good to see that I'm no longer the only linguist around here, so I'm not expected to weigh in on stuff like this.


Posted by: teofilo | Link to this comment | 03-22-12 10:45 AM
horizontal rule
71

I wonder if the Piraha would have performed a mercy killing on me when I turned (let's say) seven and still couldn't whistle.



Posted by: peep | Link to this comment | 03-22-12 10:46 AM
horizontal rule
72

50: I too have heard the theory that the Piraha are the inheritors of a concerted and deliberate counter-cultural project. It sounds plausible to me in a way, though I'm really not sure how one would go about proving it. One problem would seem to be, though, that if this is just a really small population that has learned in the long term to suppress what would "normally" be naturally emergent traits, why more cracks in the Piraha's self-sufficiency and linguistic structure don't manifest themselves.

As for the article, it seems to really oversell the notion of a Chomskyan establishment, and to be particularly insulting to the scholars that it implies are acting as fawning acolytes defending their guru. The movie that's eventually made from all this seems sure to be super-annoying.


Posted by: Lord Castock | Link to this comment | 03-22-12 10:47 AM
horizontal rule
73

70: I think we're still waiting for you to weigh in and tell us what the right answer is.


Posted by: peep | Link to this comment | 03-22-12 10:47 AM
horizontal rule
74

73 to 63.


Posted by: Lord Castock | Link to this comment | 03-22-12 10:48 AM
horizontal rule
75

Aggghh! 72 to 63!


Posted by: Lord Castock | Link to this comment | 03-22-12 10:49 AM
horizontal rule
76

THE RIGHT ANSWER IS SMALLER GOVERNMENT LOL


Posted by: REPUBLICAN GRANDMA | Link to this comment | 03-22-12 10:49 AM
horizontal rule
77

I think we're still waiting for you to weigh in and tell us what the right answer is.

Aww.

Anyway, the article's quite good, and unlike most journalistic accounts of academic disputes the journalistic tone is actually pretty accurate in reflecting the vehemence of this one on both sides. I've generally been on Everett's side in earlier iterations of this, but it's not really my area of interest or expertise. It's definitely problematic that he's the only one with detailed knowledge of the language and that it's so hard for others to check his data, so I wouldn't put too much stock in him ending up being correct. I think the approach of Pullum and others in the middle as described in the article is probably more reasonable than going all-in with Everett.

It's definitely true both that Chomskyan theory has been very dominant over the past few decades in syntax, with its practitioners often being exceptionally nasty in theoretical disputes, and that there have always been alternatives alongside it, some of which have become pretty prominent recently. This is a good roundup of alternative theories, although it hasn't been updated in a while and may be a bit dated.


Posted by: teofilo | Link to this comment | 03-22-12 10:57 AM
horizontal rule
78

(2) corpus linguistics (a la Pullum) which enriches our view of the data children are exposed to

Which Chomsky hates hates hates and thinks is useless for some reason.

It's sort of annoying that this thread is happening on a day I have a midterm, but on the other hand I don't actually care all that much about language (or, anyhow, linguistics), so I suppose it's fine.

The invitation to duel aside I am curious where Man Suit comes down on the various debates about mentalese and mental recursion and executive function and, oh, you know, the brain and shit. But phrasing it like that is pretty much saying "please right seventeen thousand works describing your take on all extant literature for me to review and potentially disagree with at a later time", which I don't think grad students get to do.


Posted by: Sifu Tweety | Link to this comment | 03-22-12 11:04 AM
horizontal rule
79

The important thing is that we all agree that dinosaurs did not roar, they chirped.


Posted by: CharleyCarp | Link to this comment | 03-22-12 11:09 AM
horizontal rule
80

They might have whistled.


Posted by: heebie-geebie | Link to this comment | 03-22-12 11:14 AM
horizontal rule
81

I get Updike, Roth and Bellow all mushed together, mostly because I have read none of them. Fortunately, I can keep that mush separate from the Upton Sinclair Lewis mush.


Posted by: rob helpy-chalk | Link to this comment | 03-22-12 11:16 AM
horizontal rule
82

81: Wait, *none* of them? Dear god, go fix that now. And add some Cheever while you're at it.


Posted by: Man Suit | Link to this comment | 03-22-12 11:17 AM
horizontal rule
83

Sorta on topic: Here's former Fistful of Euros blogger Scott Martens on Chomsky: http://pedantry.blogspot.se/2003_07_20_archive.html#105916881448121437#105916881448121437


Posted by: David The Unfogged Commenter | Link to this comment | 03-22-12 11:20 AM
horizontal rule
84

It's interesting to finally see an issue on which Sifu and Man Suit disagree, btw.


Posted by: teofilo | Link to this comment | 03-22-12 11:20 AM
horizontal rule
85

77: Thanks, teo!

From the link in 77
Chomsky does not believe that human language evolved from any previous animal communication system, but sprang into being from nowhere.

That just sounds so crazy to me. Can Chomsky's views be summarized by John 1:1?


Posted by: peep | Link to this comment | 03-22-12 11:22 AM
horizontal rule
86

Also, regarding some of the comments upthread about whether the peculiar aspects of Piraha are due in part to intentional attempts at societal isolation, my understanding is that this is more or less what Everett argues. He also argues that those aspects of the language undermine UG, but these are really separate questions.


Posted by: teofilo | Link to this comment | 03-22-12 11:22 AM
horizontal rule
87

Can Chomsky's views be summarized by John 1:1?

In some sense, yeah. This is why arguing with him on some of these issues is so frustrating, and why tensions get so high in these debates.

It is increasingly true, though, that Chomsky's personal views aren't the be-all and end-all of "Chomskyan" linguistics, which has diversified into a variety of subfields and theoretical approaches. Not everyone has followed Chomsky on all of his (many) theoretical twists and turns.


Posted by: teofilo | Link to this comment | 03-22-12 11:25 AM
horizontal rule
88

Avoid recent Roth, however. And from what I have heard the late-era Updike stuff, too, although I've read none of it. Bellow annoys some people but I like him.

Here's what a New Yorker article on Everett and the Piraha from 2007 had to say about what is known about their origins:

The Pirahã were once part of a larger Indian group called the Mura, but had split from the main tribe by the time the Brazilians first encountered the Mura, in 1714. The Mura went on to learn Portuguese and to adopt Brazilian ways, and their language is believed to be extinct. The Pirahã, however, retreated deep into the jungle. In 1921, the anthropologist Curt Nimuendajú spent time among the Pirahã and noted that they showed "little interest in the advantages of civilization" and displayed "almost no signs of permanent contact with civilized people."


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 11:25 AM
horizontal rule
89

84: I don't know if we do. We might have overlapping opinions on a subject where there is nothing close to a scientific consensus (or even any particularly strongly supported theories).


Posted by: Sifu Tweety | Link to this comment | 03-22-12 11:28 AM
horizontal rule
90

Also, since it's not a linguistics question, disagreements tend to be fairly friendly and amenable to synthesis.


Posted by: Sifu Tweety | Link to this comment | 03-22-12 11:29 AM
horizontal rule
91

89: Fair enough.

90: Zing! There's a reason I left linguistics.


Posted by: teofilo | Link to this comment | 03-22-12 11:32 AM
horizontal rule
92

Is there a book that does for lingustics what, say, Skull Wars did for cultural and physical anthropology? Which is to say, make the discipline's major debates both discernible and interesting. I'd like to read such a book. For philosophy, too, now that I think about it.

(Yes, yes, there's much that's wrong with Skull Wars. Whatever.)


Posted by: Von Wafer | Link to this comment | 03-22-12 11:32 AM
horizontal rule
93

Cheever's short stories and the first Wapshot book are great.


Posted by: lw | Link to this comment | 03-22-12 11:33 AM
horizontal rule
94

92: This one is good, though very critical of basically everyone, especially the Chomskyans. It was written at what turned out to be the peak of Chomsky's influence, so it's not really reflective of what's going on now, but it's an accessible account of the history of the discipline.


Posted by: teofilo | Link to this comment | 03-22-12 11:34 AM
horizontal rule
95

85: Chomsky has often expressed skepticism about whether language is an adaptation. Denying adaptationism doesn't entail that it came from nowhere. It could have emerged gradually but not as an adaptation, or it could have emerged via saltation. But he likes to give the very strong impression that it marks an unexplained discontinuity from other animal communication systems. Nowadays this surfaces in the idea that, essentially, human language is a perfect or optimal solution to the problem of vocalizing thought.

Here's [PDF] an easy primer. Ignore the fact that one of the authors He-Who-Shall-Not-Be-Named (Mrc Hsr). Kennedy's lecture notes [PDF] on Minimalism may also be useful in getting at what Chomsky means when he says that language is an optimal solution to the problem of pairing meaning with sound.


Posted by: Man Suit | Link to this comment | 03-22-12 11:36 AM
horizontal rule
96

86: Thanks. The New Yorker goes into the differences between Everett and Chomsky in more detail than the CHE article does. I see now that my biggest error has been in assuming that all language must contain something of a cultural component, but evidently Chomsky denies it's a major factor? That seems bonkers to me. I agree with 92; I want to know more.


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 11:38 AM
horizontal rule
97

Which is to say, make the discipline's major debates both discernible and interesting. I'd like to read such a book. For philosophy, too, now that I think about it.

I think Appiah's Thinking it Through: An Introduction to Contemporary Philosophy is pretty good.


Posted by: Criminally Bulgur | Link to this comment | 03-22-12 11:40 AM
horizontal rule
98

|?

Sheila Kelley on Stalker (I watched it Tuesday). A extended quote of a quote:

[Strasberg] spoke of what he called "the blight of Ibsen", saying that Ibsen had taught most writers after him how to think undramatically. He illustrated this by an example. A man has been used to living in luxury finds he is broke and unable to face life -- he goes home and puts a bullet in his head. That, Lee said, any fair theatre person can lay out into a play. But it is not essentially a dramatic view of life. Chekhov is dramatic, he said, for this is how he treats related material: a man earns a million rubles and goes home and lies down on them and puts a bullet in his head.

Well, you say "Ibsen" I think "Joyce" who adored Ibsen. Then I thought of the other modernists, Mann, Proust, Gide and the scientists, Freud, Saussure etc.
(This will approach OT)

I then thought that underneath Joyce's (and other mods) art is a kinda clockwork world (w or w/o humans), a world that follows rules...but those rules are never completely accessible, are never knowable. They/It can only be modeled, approached with metaphor, simile, metonymy but never apprehended.

Does Godel mean the "Real is not rational" or that humans can never be rational?

Anyway, I am a little with Greenaway on Tarkovsky, "no patience for mysticism" Mysticism can be real pretty though, and at least it's better, more moral, more humble, than fucking science. Which means Chomsky's stubborn mystic reductionism makes more sense than the young assholes who collect data and put wires in peoples' brains to find the "truth."

|>


Posted by: bob mcmanus | Link to this comment | 03-22-12 11:42 AM
horizontal rule
99

94, 97: thanks!


Posted by: Von Wafer | Link to this comment | 03-22-12 11:43 AM
horizontal rule
100

93 and 99 both seconded.


Posted by: mark f the occasional deluker | Link to this comment | 03-22-12 11:45 AM
horizontal rule
101

Didn't Anne Patchett write Bel Canto? The wife of an old friend told me it was absolutely her favorite book in the world and thrust her copy into my hands. I proceeded to lose interest after ten pages, never read it, forget to return it, and feel really bad about all of this for years.

Now I will go read about the discourse of piranhas or something.


Posted by: Mister Smearcase | Link to this comment | 03-22-12 1:33 PM
horizontal rule
102

She did. I love that book, too. Apparently I have endless appetite for books where you take an action thriller premise, and then inch through it in a slow character study.


Posted by: heebie-geebie | Link to this comment | 03-22-12 1:34 PM
horizontal rule
103

Well, I'm done teaching! Glad to see you guys got everything sorted out.

My favorite story about a Chomsky lecture is the time when the interpreter (who had clearly not done enough, or any, prep) spent the entire time talking about some dude named Eugene and his theory. We* spent a long time asking each other who Eugene is and what his theory says.

*Me and my little cohort. Most of the audience was hearing and had no such confusion.


Posted by: E. Messily | Link to this comment | 03-22-12 1:41 PM
horizontal rule
104

Confidential to Heebie: "Universal Grammar" is often abbreviated as "UG".


Posted by: E. Messily | Link to this comment | 03-22-12 1:48 PM
horizontal rule
105

I saw Chomsky speak once (though about politics not linguistics) and remember thinking the ASL interpreter seemed not as good as I'd have expected.


Posted by: Unfoggetarian: "Pause endlessly, then go in." (9) | Link to this comment | 03-22-12 1:49 PM
horizontal rule
106

104. Won't tell a soul.


Posted by: md 20/400 | Link to this comment | 03-22-12 1:49 PM
horizontal rule
107

Confidential to Heebie: "Universal Grammar" is often abbreviated as "UG".

Named in honour of its originator, Ug the Caveman.


Posted by: chris y | Link to this comment | 03-22-12 1:51 PM
horizontal rule
108

Huh, when you say it outloud, "UG" sounds like yougee. Weird.


Posted by: heebie-geebie | Link to this comment | 03-22-12 1:56 PM
horizontal rule
109

No, Heebie, it sounds like "Eugene". Pay attention!


Posted by: E. Messily | Link to this comment | 03-22-12 2:03 PM
horizontal rule
110

I saw Chomsky speak once (though about politics not linguistics) and remember thinking the ASL interpreter seemed not as good as I'd have expected.

To be fair, that probably wasn't Chomsky's fault.


Posted by: E. Messily | Link to this comment | 03-22-12 2:05 PM
horizontal rule
111

Everything is Chomsky's fault!


Posted by: teofilo | Link to this comment | 03-22-12 2:43 PM
horizontal rule
112

Thanks for 95. What an interesting idea, that language is a solution but not an adaptation. I'm having a hard time putting my objection into words.


Posted by: lw | Link to this comment | 03-22-12 2:51 PM
horizontal rule
113

I've mentioned before that the last time I saw Chomsky speak, Cornell West tried to give him a hug before hand. It looked *really* awkward.


Posted by: rob helpy-chalk | Link to this comment | 03-22-12 2:58 PM
horizontal rule
114

Everything is Chomsky's fault!

It's that jerk Eugene.


Posted by: E. Messily | Link to this comment | 03-22-12 3:13 PM
horizontal rule
115

113: In West's defense, Chomsky does offer a sublime and funky love that is extremely craveable.


Posted by: Mr. Blandings | Link to this comment | 03-22-12 3:19 PM
horizontal rule
116

I think 115 means Brother Noam.


Posted by: mark f the occasional delurker | Link to this comment | 03-22-12 3:53 PM
horizontal rule
117

It's that jerk Eugene.

I think you mean Norman.


Posted by: One of Many | Link to this comment | 03-22-12 4:13 PM
horizontal rule
118

M/arc H/auser did a lot of really good work that was not faked, we hope.


Posted by: OPINIONATED MRC HSR'S COLLABORATORS, ALSO EVERYBODY | Link to this comment | 03-22-12 6:08 PM
horizontal rule
119

107: Ug the Caveman.

More formally, Mr.Rection.


Posted by: JP Stormcrow | Link to this comment | 03-22-12 6:23 PM
horizontal rule
120

Ug the ur-Caveman.


Posted by: fake accent | Link to this comment | 03-22-12 7:36 PM
horizontal rule
121

Reading the machine-transcribed captions on 117 is surprisingly similar to reading many linguistics papers.


Posted by: E. Messily | Link to this comment | 03-23-12 7:59 AM
horizontal rule