Friday, December 5, 2014

The verdict is in regarding Evans' book

Ok, the verdict is in. remember when I wondered (here) if Alun Anderson accurately reported the content of Vyvyan Evans’ book (The Language Myth))? Well he did! It seems that Evans actually made the arguments Anderson attributes to him. How do I know? Norbert must have read the book, you are thinking. Nope. Easier. Evans outlines his views here and it is actually as confused and uninformed as Anderson’s review suggests. In fact, it may even be worse than Anderson’s review suggests.  This paper has no clue. Not even a scintilla of one. The piece is less than unenlightened, it is aphotic. And it is precisely for this reason that I cannot recommend Evans' short paper highly enough. It is a pedagogue's dream. How so?

Well first it is very short. In fact, it’s hard to believe that someone could cram so much misunderstanding into so short a format. Clearly Evans does have talent. It is so replete with the standard confusions that one need send an eager student no further than to this short piece for an example of the kinds of conceptual errors critics of GG seem drawn to (here a nice metaphor of moths and flames seems called for, but I will show strength of literary character and resist). This short piece exhibits every possible mistake. It’s a godsend. Moreover, because it is both short and so replete with misinformation, there is really no need to ever buy the book (so don't!). It is inconceivable that the book could possibly provide more illustrative examples of miscomprehension than is packed into this dense eight pages. So both cheap and serviceable. Perfect.

BTW, those thinking of using this material for teaching purposes (and I am not joking, you really should consider making criticism of these views part of the graduate degree process (even undergrads would benefit)) might also consider using Anderson’s excellent review. It too, as I indicated previously, is thoroughly misguided and misinformed, bit sometimes it’s nice to have two versions of the same song.

As I’ve already (pre)reviewed the Evans' material in the earlier post on Anderson’s reviewed, I will only add a few cursory remarks here:

1.     The confusion between Greenberg and Chomsky Universals is evident throughout. It’s amazing how hard this simple distinction has been for critics to grasp. But this piece is not only conceptually myopic, it is also intellectually lazy. Here’s a piece of critical etiquette: If one wants to sink the GG ship one needs to go after the trophy cases. There are many of these in GG and they are not hard to find. However the observation, yet again, that languages really really look different is not one of these. We know. We’ve always known. We even have GGers who have worked on (gasp!) Salish. So when making a case, discuss the serious work. Failure to do so is a sure fire indicator that nothing of interest has been discussed.

2.     There is a difference between two oft used metaphors for describing innate structure: (i) UG is a blueprint, (ii) UG is an operating system. The difference between the two caused quite a bit of confusion within biology almost as much as it is doing in linguistics (or so I remember Gould telling us in his discussion of the homunculus problem). Blueprints invite the search for Greenberg Universals (every house has 2 bathrooms, 2 bedrooms, a kitchen and a foyer: find them!). Operating systems not so much. Windows is not OS, though they can do many of the same things, they are different systems with different formal requirements. In particular, there are few surface indications of what operating system you are using. In contrast, it is actually quite easy to tell whether some set of blueprints are those of the apartment/house/building you are looking at (I speak from some personal experience here). The GG conception of UG is more like operating system than blueprint. But really we don’t need these metaphors anymore (or at least we should not rest content with them): UG is a function that given PLD yields Gs. Your job, Mr Phelps, should you agree to take it, is to describe the fine structure of this function. In so doing GGers have discovered some very subtle features of this function and critics, to be serious, must address them in detail. Though I am a big fan of hand-waving (especially on sultry summer days) it generally fails to advance the topics of interest. Evans piece is an excellent illustration of this truism.

3.     All the same non-Chomskys make their appearance. Everett is cited (yet again) as showing that Gs need not be recursive and so disconfirming UG and Tomasello is trotted out to represent the idea that a penchant for co-operative behavior is all we need to get human language.  Starlings are schlepped out to argue that one finds recursion in non-humans and Neanderthals make an appearance to testify language facility pre-dating homo sapiens. All of this asserted with nary a hint of skepticism (or apparent knowledge that each one of these claims has been contested and all are likely to be false). Wouldn’t it be nice sometime if someone (e.g. Tomasello, Evans, Anderson) explained to anyone how it is that a penchant for co-operation yielded island effects or the ECP or Principle C, or …you get it. I understand that many other species display quite a bit of co-operative behavior without appearing to be grammatically competent. There is a simple reason why accounts purporting to explain the intricacies of language based on co-operation are never forthcoming, and a moments reflection tells us what this is: there is no route from the former to the latter that does not need to engage with the exactly the questions that Evans begs. And this is evident throughout: just like Anderson, faithful describer of his views, Evans prefers to abstract away from facts and arguments enjoying instead the broad oh-my-god-how-nutty-this-all-looks summary conclusions.

4.     Did I say ‘competent’ above? Here’s another great feature of the Evans’ piece. He seems not to understand the competence-performance distinction as he keeps talking about “speech” rather than linguistic competence. Evidence for the former is not in and of itself evidence for the latter. Birds “speak” but are not syntactically competent. Not even starlings.

5.     Oh yes, the Neo-Darwinian synthesis makes its gradualist appearance here to argue against the emergence of an FL about 100 kya (where is Gould now that we really need him) as does the assertion that being non-localized in the brain (there is no “spot specialized just for language” indicates that the idea that there is a language module is over the top.  I assume that similar reasoning could show us that there is no electrical system in a car. After all it is everywhere.

There is a lot more in this slender cornucopia of misapprehension and I could go on, but you should enjoy this yourself.  I recommend this short précis of Evans work to all looking for a paper to hand out to a class that concisely embodies all the mistakes that it is possible to make about the Chomsky program in GG. Though appalling, the piece is very useful.

One last exhortation: linguists should make it their business to loudly criticize this junk at every opportunity. It is getting wide distribution. I got it from Aeon and it was picked up in The Browser. It has just the ingredients to make it big: another one of those Chomsky-has-been-proven-wrong memes that seem so popular. So, criticize this in all venues, especially where non-linguists gather. Consider it part of your linguistic public service.


  1. This comment has been removed by the author.

  2. I just started reading the article; apparently the author doesn't just bungle the linguistics, but also the biology. He claims that "our bodies are language-ready too: our larynx is set low relative to that of other hominid species, letting us expel and control the passage of air" - perhaps true with respect to other hominids, but thanks to Tecumseh Fitch’s insights, there are several mammal species with permanently lowered larynges (including a couple species of definitely not-talking deer), and more species that lower the larynx when vocalizing.

    Not only does the author misunderstand the data supposedly speaking against the language/speech innateness hypothesis, but he makes mistakes about data that would otherwise have worked in favor of his argument. Resoundingly sloppy all around.

  3. Yeah, I actually read the book, which had me apoplectic in the tube, and poor Anson exhausted by my continual explosions of ire. The book is all that this article is, just writ larger. I was going to write a review somewhere, but am saving it for a longer piece.

    1. Good. And make sure that everyone and anyone you know understands just how bad this stuff is. This gives flat earthism a good name.

  4. It is quite heartwarming to see [read] how those who claim to be badly offended by my allegedly unpleasant use of language speak to express their disagreement. Explosions of ire, eh? And nothing of the like was felt when reading Chomsky/McGilvray's "The Science of Language"? As for resoundingly sloppy all around: seemingly no one here is offended when Chomsky calls nematodes insects or tells us that his program requires that we accept things we know do not make any sense...

    Even though I predict that, once again, Norbert will hide behind the cowardly promise not to speak to me again, I cannot rest to ask: Enlightened One, you taunt: "Wouldn’t it be nice sometime if someone (e.g. Tomasello, Evans, Anderson) explained to anyone how it is that a penchant for co-operation yielded island effects or the ECP or Principle C, or …you get it." - It would be even nicer if YOU would finally tell us how all the things Tomasello&co keep ignoring are encoded in the genome. So can you be so kind and finally provide details?

  5. My explosions of ire were to my boyfriend on the tube! But I did wonder if they way I was feeling reading Evans was like the way you feel reading Chomsky!

    1. If your feelings were caused by the belief that Evans is factually * wrong [*insert whatever adjective covers how wrong you think he is] then NO - that has virtually nothing to do with my 'feelings' towards SoL. True enough one will be hard pressed to find a volume that is sloppier when it comes to describing the achievements of your field [and frankly it surprises me that none of the readers of this blog is commenting on that, leaving open the interpretation that you [pl] are unable to see the obvious]. What I find offensive is Chomsky's tone when he talks about the work of others. According to him, academia is populated with irrational dogmatists and crazy theorizers. In fact, an uninitiated newcomer [the target audience of SoL!] must walk away with the impression that anyone who disagrees with Chomsky suffers from a severe mental disorder - an attitude perfectly mirrored by Norbert on this blog: it is never enough to say 'X is wrong'; X always needs to be insulted to the bone [like "This gives flat earthism a good name" above].

      I encourage you to read the comments on the piece Norbert so generously linked to. You will notice how much cyber-ink is spilled for name-calling and invective. And how little of substance Evans' critics actually have to say. Where are the positive proposals that Evans should have considered? Where are links to the work on genetics by generativists? Where are links to work showing how generativists deal with challenges from non-generativist linguists you got too busy to outline on this blog a while ago?

  6. Ok, so the aetiology of the feelings are different. My problem was that there was so much wrong in Evan's article, and in the book, that misrepresents the field and its results. I'll probably not read the comments, as I find name calling as tedious as you do. But if you want a response by me to challenges raised by non generativists, have a look at the debate between myself an Goldberg (in Mind and Language and on Lingbuzz) and if you want a fairly ecumenical overview of current syntax, with a lot of links to references where generative work has been tested via psychological and other techniques, have a look at the paper entitled Syntax by me on Lingbuzz coming soon to WIREs Cognitive Science open access. On positive proposals that Evans could have considered, I think the major lacuna is dealing with the empirical successes of generative grammar in understanding the way that form links to meaning in languages in general - this is a point that Norbert makes a lot, but it's one that Evans doesn't touch on in his book or article (I'm thinking about facts about languages such as some of those I point out in that Syntax paper.)

    On why I think Evans is wrong, here are a few initial thoughts, just on the article:

    #1: linguists don't use the term 'instinct' as a scientific term: it's a shorthand Pinker came up with. We talk of an innate capacity triggered by and partly shaped by experience. This is rhetorical misdirection.

    #2 Evans misunderstands the difference between typological similarities/differences between languages and proposals about the structure of the human capacity for language. Like saying that because frogs look different from goats, they're not both built of proteins.

    #3 recursion: we don't know starlings have recursion. That's a basic mathematical mistake, since the reported behaviour is compatible with non recursive grammars using a counting stack.

    #4 the claim about recursion is a claim about the human language capacity, not about languages. If Piraha lacks clausal embedding, that's irrelevant to the claim.

    #5 the stuff on acquisition is hugely controversial and the argument Evans makes shows a misunderstanding of probability theory. Actual tokens are too rare to generalise from by Zipfs law, but kids do generalise. Charles has made this point very forcefully and there doesn't seem to have been any counterargument to his point presented in the non-generativist acquisition literature:

    #5 Evans seems to think that it's a consequence of modularity that a structural fact about humans (eg a system that allows language) must be anatomically localised. But the nervous system is a structural fact about human beings. Hardly localised.

    ... tbc

  7. ...
    #6 Evans discusses evidence for double dissociation from the 1990s and says it doesn't hold up. He should have read more up to date literature such as curtiss's work

    #7 On DNA: Current theory takes the language capacity to be an interaction between genome, experience and physical/computational law. I don't see how any claim can be made about how much information can be stored in DNA. This is an argument from incredulity, but that isn't, of course, a real argument.

    #8 ditto the arguments on evolution. This is not an uncontroversial area! Evans should read Tattersall or others to see how the proposals about language evolution are potential scenarios

    #9 Evans in the article or book gives not one example of how 'cooperation' explains facts about human syntax that are well known ( and well explained. None have ever been given. Unclear why we should swap an empirically successful theory of human grammatical capacity for one with absolutely no explanation for the vast array of facts we know about the syntax of individual languages, or relations between them.

    #10. I think I was somewhat shocked by the lack of discussion of recent literature in especially the book. Evans at one point talks about `recent' work, and refers to work in the late 80s and early 90s, over 20 years ago. I just thought this was sloppy scholarship.

    Turning to the book, I'm pretty sure it's a very difficult task writing a pop science linguistics book, so I'm willing to forgive a bit of sloppiness or inaccuracy, but there was just so much of that, and the actual arguments were so weak. At some point, I actually stopped feeling outraged while reading the book, and just began to feel a bit depressed at just how much of it was wrong. By the time I was halfway through, I was thinking of writing a response, but by the time I'd finished, it just seemed so feeble that it felt that there were more worthwhile things to do - some linguistics perhaps.

    1. David, I think maybe you ought to consider drafting up that response and trying to get it to run somewhere prominent. (Is it Slate or somewhere else that now has the fairly well informed language column now? Or in the form of a 17-point list to get onto Buzzfeed..) I think you would find I and others around here would be happy to help out.

      I say this with a sense of urgency as I was just in an exchange on someone's Facebook thread where they said they would give this article to their psycholinguistics class (endorsing the contents, not trying to draw out criticism) and in the meantime they got someone else to join up saying they would too! Suffice it to say pointing out the easy flaws in the article didn't dissuade them. If there was something a bit more carefully written up .. firmly but dispassionate etc.. then they could be circulated as a pair. Maybe the second could be the "answer key" for the critical thinking exercise...

  8. Thank you for the elaboration. I will have a bit more to say later; just one brief comet for now. You say:
    #1: linguists don't use the term 'instinct' as a scientific term: it's a shorthand Pinker came up with. We talk of an innate capacity triggered by and partly shaped by experience. This is rhetorical misdirection.

    We agree that rhetorical misdirection should be avoided. But why have you not raised this point with McGilvray/Chomsky? Take your copy of SoL, go to p. 178, last paragraph. There McGilvray tells his readers that Chomsky is "treating language - its growth/development and its internal operation - as an 'animal instinct' introduced by mutation into the human species"

    McGilvray is THE authority on anything Chomsky and he tells his readers that "Professor Chomsky reviewed the edited and commented text and made many suggestions." [p. 3]. So can you blame Evans, or others, that they believe Chomsky and those following his lead treat language as an instinct?

  9. The writers of Aeon must be L1 speakers of some language without citations... was wondering about the source(s) for the claim about specific language impairment just being a sensory (or is it motor? Evans says one, than another) impairment. If I understand correctly, this would rule out the very notion, since SLI can be diagnosed only in the absence of gross sensory or motor impairments.

    1. One way to find out about sources would be to ask the author: v.evans AT

  10. This comment is a bit tangential, but it's relevant to the mention of linguistic public service.

    Norbert started this blog to rectify the fact that "linguistics outsiders (and even practitioners) know [so little] about the foundations and results of the Generative Enterprise initiated by Chomsky in the mid 1950s".

    I've really enjoyed following this blog, both for its debunking of positions taken by those who refuse to play the game and for all of the other linguistically interesting discussion that isn't directly related to said debunking. However, one thing that I haven't seen discussed—perhaps I've missed it?—is the implication(s) of the foundation of the Generative Enterprise that takes each speaker to have a different grammar, or I-language. In fact, I've only ever seen this extensively discussed by three linguists in a public venue. (Though perhaps I'm just not looking in the right places.)

    At any rate, Lauren Squires discusses this in a post on Language Log. It also occasionally comes up on Gretchen McCulloch's All Things Linguistic blog (e.g., here). And it also seems to be something that John Rickford has dedicated a decent amount of time and work to. A most recent example is this post from the Stanford News where he points to a particular example of the severe implications that linguistic prejudice can have. (For those who don't have time to read the post, it discusses the lack of credibility that the all-white jury attributed to witnesses who were speakers of AAVE in the Trayvon Martin case, which probably contributed to the verdict that was reached.)

    In addition to this particularly pernicious example of the lack of justice in the Trayvon Martin case, there are many ways in which prescriptivism and linguistic prejudice/discrimination manifest themselves, including in the K-12 education system, college applications, job applications, etc..

    So it seems to me that being more vocal against prescriptivism and linguistic prejudice/discrimination is a very valuable public service that linguists should be engaging in (for hopefully obvious reasons).

    I realize that the primary audience for this blog is linguists and related academics, but I think there is still valuable discussion that could be had on this blog. I, for one, would find it particularly interesting to see discussion about various ways to motivate the concept of I-language and the associated implications that there is no such thing as 'correct' grammar—at least not like we're taught in English class—and that there is thus no reason to discriminate against someone based on their language. I say this both as someone who hopes to teach linguistics some day and as someone who can anecdotally report from conversations with non-linguist friends that it is not at all a trivial task to motivate the concept of I-language, much less get folks to see the implications thereof.

    I suspect part of the reason this is so hard is because of the fact that it is beat so thoroughly into people's head as they go through English classes in the K-12 system (and even in university) that there is such a thing as 'correct' grammar and that there is such a thing as 'English'.

    (continued ... )

  11. ( ... continued)

    For those of you who have the (mis)fortune of teaching introductory linguistic courses, what sorts of things do you do to motivate the concept of I-language? How do you discuss the implications of this regarding prescriptivism and linguistic prejudice and discrimination?

    I suspect that being very intentional about trying to teach this in introductory courses is one of the best things that linguists could be doing as a public service, given how pernicious the repercussions can be. And I also suspect that prescriptivism and linguistic prejudice/discrimination won't be going away all that easily without having many more linguistically-informed K-12 English teachers.

    I am currently at a university where a lot of education majors are required to take an introductory linguistics course. In my opinion, one of the most valuable things that people in this course could walk away with is an understanding of these implications of the I-language concept. (Not to burst anyone's bubble, but I'm not sure how valuable a constituency test—or whatever you teach them in an introductory course—is outside of further linguistic study and research.)

    So anyway, I think it would be really interesting to see more discussion of this on this blog. How can linguists be intentional about teaching this in introductory courses? Which ways of motivating these things work? Which ways don't? How can we make these things "stick" after the test/semester?

    And, for extra credit: how can we be intentional about tackling linguistic prejudice/discrimination outside of the classroom, too?

    Of course, I hope all of the great linguistic discussion and the debunking continues, too. This blog has been very fun to follow. So thank you, Norbert! But perhaps this could be incorporated into the blog, too.

    1. I feel like I'm going to wait for Norbert to give a fully-fledged response to this. But my feeling was always that human Language - that is, I-Language - is a property of an organism, ultimately a biological property of that organism (properties of minds result from properties of brains). Analogously, a leg or an arm or a digestive system are biological properties of organisms. Each leg or arm or digestive system is unique to that individual, it has its own unique configuration of cells, proteins, etc. Regardless, they all share common properties specified by the commonalities of the human genome. Likewise with I-Languages - they develop as "organs of the mind" in the same way that conventional organs develop - according to a highly innate and articulate genomic plan (Universal Grammar), but are of course sensitive to environmental factors that could in some cases dramatically change its development (e.g. why I speak English and not Japanese).

    2. So to follow up, would somebody claim that there is a "correct" leg or digestive system?

    3. I think the analogy makes sense, but I doubt that it makes sense to most people. Most non-linguists (that I encounter) do in fact seem to think that there really is such a thing as 'correct' grammar and that it is whatever we get taught in English class.

      A related fact is that there are stereotypes about speakers of certain dialects. For example, a common stereotype of speakers of southern dialects in the US is that they are uneducated. Similarly, a common stereotype of speakers of AAVE is that they are either uneducated or lazy (for not speaking 'proper' English), or both. And these linguistic stereotypes can lead to linguistic prejudice and discrimination, which can sometimes (often?) have some rather pernicious effects.

      This is where I think linguists can and should be doing something. I think we could be much more vocal against prescriptivism and linguistics prejudice/discrimination. But what I also think might be more valuable than just being vocal against these things is trying to engender actual understanding of the relevant concepts in others, which is why I asked some questions about how to motivate the relevant concepts and implications.

      If the average non-linguist understood the concept of I-language, perhaps they would find themselves asking the very same question that you raise—namely, how could it possibly make sense to say that there is such a thing as 'correct' grammar if it doesn't make any sense to say that there is such a thing as a 'correct' leg (much less be prejudiced against someone for not having something that doesn't exist).

      (And hey, maybe this would have the added side-effect of people coming to have a better understanding/appreciation of what we study and no longer asking us how many damn languages we speak ... :P )

    4. I will leave it to the linguists to answer the questions about 'correct grammar' and dialects of English. But I think, unfortunately, the fact that Dan Everett has been accused of racism for the descriptive claims he made about Piraha suggests that at least some people believe there IS a correct [or better?] UG and individuals who do not obtain it are inferior [why else would the claim that a particular language does not have recursion be interpreted as racist?].

      Also notice that we may not use the term 'correct' for the digestive system [and hence blame someone for being lactose intolerant]. But we try to 'fix' [we call it treat or medicate] deviations from what we consider the proper functioning of these systems. So calling language a biological organ [and tying it as closely to the essence of the human mind as Chomsky does] certainly has its problems and I agree with Adam: it would be good if those defending this view of language would address these problems.

    5. I'm just wrapping up an intro course, that doubled as a history of linguistics, and that way, we reinforced the underlying concepts of descriptivism long before applying them to prescriptivism.

      a) The concept of not judging different languages goes back to the anthropologists of the early 20th century. That is to say, to analyze without prejudice.

      b) Saussure's concept of language (grammar, we'd say) as a system is still essentially held, even if we understand the system differently. This implies that no system is inherently better than any other. So people don't speak wrong, they just use different systems correctly. (To which people sometimes say, "their system is messed up", but that leads to the rest).

      c) Even before that, phoneticians in the late 19th century (Swift, famously), noted that certain features that were prestigious in one dialect (say, r-dropping) were unprestigious in others. With facts like this, coupled with Saussure's, we can conclude there is no linguistic criterion for ranking any system (i.e. language variety) as better than any other.

      d) Sociolinguistic study and historical study find that the fate of a language is inextricably linked to the fate of its speakers. Again suggesting there is no linguistic reason to prefer one variety to another. Notably, we also find the truism that when people talk about the qualities or lack thereof of a language variety, they're really talking about the qualities or lack thereof of the speakers of that variety.

      e) Now we get to I-language. But you see that, even without this concept, there's ample reason to discount prescriptivism. But the notion that languages don't exist per se, but are sort of circles drawn around people whose I-languages (whose systems) are close enough to permit understanding... that hammers the point home. Each of us has our own system, and so we can't say that there's a right one, because a) whose would it be, and b) how could we tell WE were the ones with it?

      I will add this. We also read a fascinating essay by David Foster Wallace in defense of prescriptivism. Essentially, he admits that prescriptive rules are arbitrary and cruel, and unfair to people who acquired a different variety. However, he points out that that doesn't mean there should be no standards at all, giving the analogy to tax evasion (that common behavior isn't right behavior). That's a faulty analogy, of course, but a cleverer argument than some pretentious nonsense about preserving rational thought like you usually hear, and it's done politely, so it fosters class discussion. He continues with a stronger point: As an English teacher he feels obligated to hold students to prescriptive rules, because people will judge you for your language use, and he wants them to be able to be judged positively. With this I sympathize.

      I personally liken prescriptive rules to fashion rules, and in that sense they do have some value... but at the same time, the sharp limits to their value become instantly clear.

      Overall, though, the case against prescriptivism has had bricks laid for it since long before I-languages were even thought of, and I find the case more persuasive when we build upon those bricks.

    6. Adam, you are of course correct. It's really hard for people not to be prescriptive but the official view, which I think is right for roughly the reasons William above notes, is that each I-langauge is as good as any other and everyone IN FACT has a unique one. They overlap sufficiently, but identical they are not. Which is better? Wrong question.

      One thing though is interesting: it is interesting that the I-languages seem to clump into "languages." So the I-languages are not ALL over the place but seem to find common peaks in the hypothesis space. So though not identical, they overlap quite a bit. That is interesting for it suggests something about the hypothesis space. However, this observational fact has zero normative weight.

      Language is a big deal. It's a trait unique to humans and language differences are very visible. Like every thing else along these lines (dress, manners, hair length, girth), language usage can be invidiously compared and contrasted for the benefit of the powerful. Everything that can be recruited to benefit the powerful will be so recruited. That's an axiom of mine. So language will be too, and you are right to warn that this has NO scientific standing. I doubt we can stop it, but we can point out that it hangs on a cultural rather than a descriptive basis.

      Last point: one thing that is worth doing to help dispel this is to note (1) that no language is more or less complex than another. All are rule governed and the kinds of rules are more or less the same form. (2) that much that we consider "low" today was high yesterday (think double negation) and (3) much that is considered low is high in another language. This can be therapeutic, even if the pull to judgment is very powerful.

    7. @Christina: It's worth being careful here about conceptual distinctions. Presumably there is a 'correct' UG at least insofar as there is a fact of the matter as to what constitutes whatever it is that allows us to do language. This is different from particular I-languages or grammars—or, as Norbert is fond of abbreviating, Gs—that people acquire as they grow up. It is in this sense that there is no 'correct' grammar for something like 'English' (which doesn't exist).

      I would also appreciate it if you didn't put words in my mouth. I don't think there are any (conceptual) problems with such an understanding of language. (Though there are certainly problems and puzzles that remain to be solved, of course.) But by saying that I thought William's analogy made sense, I was not saying that I endorse the analogy wholeheartedly. I'm sure you know as well as I do that argument by analogy almost always breaks down at some point. Thanks!

      @Andrew: Thanks for the discussion! I'll check out the David Foster Wallace essay. It sounds like an interesting way to start a discussion about this. Regarding your points (a) and (b), I'm wondering if this is a point that you made in your class with regard to idiolects that aren't mutually intelligible or if you were able to successfully make the point concerning idiolects that are mutually intelligible, too. Based on my experience from conversations with non-linguist friends, it's much easier to make the point that one ought not to judge another person for speaking, say, 'Marathi' as opposed to 'English' than it is to make the point that one ought not to judge another person for speaking some non-standard variety of English (say 'AAVE') as opposed to 'standard English'. Even if they see the logic in the English vs. Marathi case, it's for some reason very hard to translate that to the AAVE vs. standard English case.

      Lastly, regarding your analogy to fashion rules, I don't think I agree that prescriptive rules have value. I think it's fair to say that they serve a societal function. As Karthik Durvasula (who sometimes comments here) has pointed out to me, they can play a role in allowing one to adjudicate between different job applicants that you might not otherwise be able to adjudicate between insofar as you can use conformity to prescriptivist rules as a proxy for how much the person cares about the job that they are applying for. I think this is a fair point, but I also think it highlights why viewing this as a value—rather than a function—is highly dangerous. For independent reasons (one such independent reason being lack of access to adequate education), there are many who might not know what the prescriptivist rules are that they ought to conform to when applying for a job. So to see conformity to prescriptivist rules as valuable contributes to devaluing the non-conforming applicant. Perhaps this is just arguing semantics (in the non-linguistic sense :P ), but I do think that viewing conformity to such rules as a value rather than as serving a function makes it much easier, for example, to view speakers of AAVE that are called as witnesses in a court as much less credible than speakers of standard English.

      @Norbert: I agree that what can be recruited to maintain the status quo, will be, which is why I think a particularly valuable public service that linguists could (and should) be rendering is trying to disarm this as a tool that can be so recruited. Thanks for your points (2) and (3)! I think those could be useful tools for spreading the word to non-linguists. I'm not so sure about (1), though. I think that saying all languages are rule governed and that the rules are more or less the same is only going to be a convincing argument to a linguist. Having said this myself to a few non-linguist friends, I'm often met with skepticism, and I'm not sure how to make the point stick, so to speak, without trying to teach them a non-trivial amount of linguistics.

    8. @Adam: I could "retaliate" and ask you not to put words into my mouth and we might be off to an entirely unnecessary dispute. So instead, let me say: thank you for reminding us all how right Chomsky is when he tirelessly stresses the inherent ambiguity of language. You misunderstood my comment. I did not mean to suggest you see any conceptual problems with the UG approach. You said far too little about your views to allow anyone to make such a judgment. So it did not occur to me what i said could be interpreted that way.

      I was referring to the problems that arise when one accepts the UG approach. These are problems any form of biological essentialism faces and they have been raised in the philosophy of biology some time ago. So I was referring to your concern about the implications that linguistic prejudice can have. If one assumes LF is part of the genetic make-up of a person just as for example skin colour is, it is only a small step towards linguistic prejudice (remember that people have accused Everett of racism because of his claims about the Piraha language). So I certainly think that anyone who defends the UG view of language ought to take a stand on the biological essentialism issue. If you do not think this ought to be a concern I apologize for misunderstanding your concerns.

    9. @Christina: I understood you as raising biological essentialism as a conceptual problem for endorsing such a view of language. Perhaps I shouldn't have used the word 'conceptual', but, in any case, I don't see this as a problem for this view of language, whatever sort of problem it is. In fact, it seems to be quite the opposite, which is why I am interested in figuring out how linguists can better convey this to non-linguists. An I-language understanding of language—which I have no reason to doubt as being the correct understanding of language—entails that linguistic prejudice and discrimnation informed by prescriptivism are irrational and unfounded.

      To say a bit more about this, it's again worth being careful about conceptual distinctions that linguists make. We use 'UG' to mean whatever it is that allows us to do language, and a particular I-language (or grammar) is what results after a child has been subjected to enough linguistic data from her environment. Linguists use LF to refer to a particular (syntactic) level of representation, so your usage of that term in your comment is—to me at least—confusing and seems to be a category mistake. We assume that UG is part of the genetic makeup, not LF. (Or did you mean FL, which we use more or less interchangeably with UG?)

      At any rate, as you point out, philosophers, critical race theorists, and feminists have rightly criticized biologically essentialist positions about race and gender. In the case of the feminist critique, for example, one thing that is pointed out is that gender roles and relations are a function of societal organization and culture (nurture) and not a function of genetics (nature). And so it is rightly argued that contemporary gender roles are not a necessary result of genetics (nature). But this doesn't mean that there isn't a fact of the matter about the difference in genetic makeup.

      An I-language understanding of language makes room for an analogous sort of argument. There is a fact of the matter about the genetics that allow us to do language—namely, UG—but everyone acquires a different I-language given the linguistic data that they are subject to (nurture). And so to discriminate on the grounds of nurture—that is, on the grounds of someone's particular I-language that they have acquired—is unwarranted.

      So it seems to me that this position is not open to such a critique. Indeed, it seems to be quite the opposite.

      One last thing: you point out the fact that Everrett has been accused of racism as if it were evidence that this position on the nature of language is biologically essentialist. I'm a bit leery of commenting on this because I am too young to have seen the whole debacle as it took place. I tried searching the internet for these accusations, but I have not been able to find the content of the accusations, just that they were made.

      Nonetheless, I can imagine these accusations arising precisely because people were not careful about these conceptual distinctions. Everrett has touted Pirahã as falsifying UG. But linguists just use 'UG' to mean whatever it is that allows humans to do language. And it seems relatively reasonable to postulate that whatever allows us to do language is also what distinguishes us as a species. So, at face-value, a linguist might take Everrett as claiming that Pirahã speakers lack UG and interpreting this as a claim that these speakers aren't fully human. And I can see some quite easily construing that as racist.

      (continued ... )

    10. ( ... continued)

      Let me reiterate that I am not familiar with the actual content of any of these accusations, so this is not meant to be a defense of any who did call Everrett racist. It is just to point out that it is relatively easy to imagine how it could've happened and to highlight the importance of being careful about these conceptual distinctions.

      I think this also reinforces the importance of the question I initially asked. How can we ensure that non-linguists have at least a rough understanding of these concepts in order to be able to see that they render linguistic discrimination irrational?