Tuesday, September 24, 2013

When UG?

MP makes the working assumption that whatever happened to allow FL to emerge in its current state happened recently in evo time. This, in turn relies on assuming that precursors of us were without our UG (though they may have had quite a bit of other stuff going on between the ears, in fact, they MUST have had quite a bit of stuff going on there). This assumption was recently challenged by Dediu and Levinson (D&L). Here's an evaluation of their paper by Berwick, Hauser and Tattersall (BHT) (here). BHT argue that there is no there there, a feature, it appears, of much of Levinson's current oeuvre (see here). They observe that the evidence for the quick time frame is sorta/kinda supported by the archeological record, but that such evidence can hardly be dispositive as it is not fine grained enough to address the properties of "the core linguistic competence" of our predecessors as this "does not fossilize." However, such that exists does appear to (weakly) support the envisaged timeframe proposed (roughly 100,000 years). Indeed, as BHT note, D&L misrepresent an important source (Somel et. al) which, concludes, contrary to D&L that: "There is accumulating evidence that human brain development was fundamentally reshaped through several genetic events within the short time space between the human-Neandertahl split and the emergence of modern humans."

So take a look. The D&L paper got a lot of play, but if BHT are right (that's where my money is) then it's pretty much a time sink with little to add to the discussion. You surprised? I'm not. But read away.


  1. Why base so much of linguistics on an assumption that is a) at best weakly supported, and b) such that the evidence for it can't be professionally evaluated by people with training in linguistics? Directing attention towards the idea of FLN also entails directing it away from the classic generative task of explaining how languages can be learned as fast as they are, at exactly the moment when various advances in mathematical understanding seem to offer some promise of making real progress on this, in a framework that most of out intellectual neighbors understand (e.g. the Kwiatkowski ety al CCG learner - if I could figure out how to make CCG manage the exciting aspects of Icelandic, Kayardild and Dinka, I'd probably 'convert' to it).

    1. I don't think that this is what MP does, at its best. Moreover, these old questions need not. Be abandoned by anyone. The program is one of unification, at least for me, of the various apparently disparate modules. This gets some slight motivation from the evo timeline, but also makes perfect sense without this.

    2. But the best tends to get mixed up with the not-so-best, and the prominent place of what are essentially speculations generates a lot of turbulence and headwinds (at least from where I sit).

    3. BTW, if anyone is interested, there's a *long* version of the rebuttal to D&L that I hope might appear sometime (> 40 pp). D&L is in far graver shape than one can say in a few pages. Second, off-topic a bit but in reply to Avery's enthusiasm for Kwiatkowski et al.: I agree this is a very nice demo...but I wouldn't jump ship quite yet. For one thing, it's a jumped-up probabilistic version of Siskind's 1996 approach, which dovetails OK w. any LFG or (then GB) approach. Don't get me wrong. It is a major advance on this, in my view. And it's nice that it works w/ CHILDES examples. But there's no reason I can see why one can't apply the same brute force 'try all possible word-object combos" and then all "syntax-LF combos" as long as one has some systematic lf underlying it all. I guess I will try to write about this in more detail.

    4. @ Robert: I am interested in reading the longer version, where does one get a copy?