Thursday, April 27, 2017

How biological is biolinguistics?

My answer: very, and getting more so all the time.  This view will strike many as controversial. For example Cedric Boeckx (here and here) and David Berlinsky (here) (and most linguistics in discussions over beer) contend that linguistics is a BINO (biology in name only). After all, there is little biochemistry, genetics, or cellular biology in current linguistics, even of the Minimalist variety. Even the evolang dimension is largely speculative (though, IMO, this does not distinguish it from most of the “seripous” stuff in the field). And, as this is what biology is/does nowadays, then, the argument goes, linguistic pronouncements cannot have biological significance and so the “bio” in biolinguistics is false advertising. That’s the common wisdom as best as I can tell, and I believe it to be deeply (actually, shallowly) misguided. How so?

A domain of inquiry, on this view, is defined by its tools and methods rather than its questions. Further, as the tools and methods of GG are not similar to those found in your favorite domain of biology then there cannot be much bio in biolinguistics. This is a very bad line of reasoning, even if some very smart people are pushing it.  In my view, it rests on pernicious dualist assumptions which, had they been allowed to infect earlier work in biology, would have left it far poorer than it is today. Let me explain.

First, the data linguists use is biological data: we study patterns which would be considered contenders for Nobel Prizes in Medicine and Physiology (i.e. bio Nobels) were they emitted by non humans. Wait, would be?  No, actually were. Unraveling the bee waggle dance was Nobel worthy. And what’s the waggle dance? It’s the way a bee “articulates” (in a sign language sort of way, but less sophisticated) how far and in what direction honey lies. In other words, it is a way for bees to map AP expressions onto CI structures that convey a specific kind of message. It’s quite complicated (see here), and describing it’s figure 8 patterns (direction and size) and how they related to the position of the sun and the food source is what won von Frisch the prize in Physiology and Medicine. In other words, von Frisch won a bio Nobel for describing a grammar of the bee dance.

And it really was “just” a G, with very little “physiology” or “medicine” implicated. Even at the present time, we appear to know very little about either the neural or genetic basis of the dance or its evolutionary history (or at least Wikipedia and a Google search seems to reveal little beyond anodyne speculations like “Ancestors to modern honeybees most likely performed excitatory movements to encourage other nestmates to forage” or “The waggle dance is thought to have evolved to aid in communicating information about a new nest site, rather than spatial information about foraging sites” (Wikipedia)). Nonetheless, despite the dearth of bee neurophysiology, genetics or evo-bee-dance evolutionary history, the bio worthies granted it a bio Nobel! Now here is my possibly contentious claim: describing kinds of patterns humans use to link articulations to meanings is no less a biological project than is describing waggle dance patterns. Or, to paraphrase my good and great friend Elan Dresher: if describing how a bunch of bees dance is biology so too is describing how a bunch of Parisians speak French.

Second, it’s not only bees! If you work on bird songs or whale songs or other forms of vocalization or vervet monkey calls you are described as doing biology (look at the journals that publish this stuff)! And you are doing biology even if you are largely describing the patterns of these songs/calls. Of course, you can also add a sprinkle of psychology to the mix and tentatively describe how these calls/songs are acquired to cement your biological bona fides. But, if you study non human vocalizations and their acquisition then (apparently) you are doing biology, but if you do the same thing in humans apparently you are not. Or, to be more precise, describing work on human language as biolinguistics is taken to be wildly inappropriate while doing much the same thing with mockingbirds is biology. Bees, yes. Whales and birds, sure. Monkey calls, definitely. Italian or Inuit; not on your life! Dualism anyone?

As may be evident, I think that this line of reasoning is junk best reserved for academic bureaucrats interested in figuring out how to demarcate the faculty of Arts from that of Science. There is every reason to think that there is a biological basis for human linguistic capacity and so studying manifestations of this capacity and trying to figure out its limits (which is what GG has been doing for well over 60 years) is biology even if it fails to make contact with other questions and methods that are currently central in biology. To repeat, we still don’t know the neural basis or evolutionary etiology of the waggle dance but nobody is lobbying for rescinding von Frisch’s Nobel.

One can go further: Comparing modern work in GG and early work in genetics leads to a similar conclusion. I take it as evident that Mendel was doing biology when he sussed out the genetic basis for the phenotypic patterns in his pea plant experiments. In other words, Mendel was doing biogenetics (though this may sound redundant to the modern ear). But note, this was biogenetics without much bio beyond the objects of interest being pea plants and the patterns you observe arising when you cross breed them. Mendel’s work involved no biochemistry, no evolutionary theory, no plant neuro-anatomy or plant neuro-physiology. There were observed phenotypic patterns and a proposed very abstract underlying mechanism (whose physical basis was a complete mystery) that described how these might arise. As we know, it took the rest of biology a very long time to catch up with Mendel’s genetics. It took about 65 years for evolution to integrate these findings in the Modern Synthesis and almost 90 years until biology (with the main work carried out by itinerant physicists) figured out how to biochemically ground it in DNA. Of course, Mendel’s genetics laid the groundwork for Watson and Crick and was critical to making Darwinian evolution conceptually respectable. But, and this is the important point here, when first proposed, its relation to other domains of biology was quite remote. My point: if you think Mendel was doing biology then there is little reason to think GGers aren’t. Just as Mendel identified what later biology figured out how to embody, GG is identifying operations and structures that the neurosciences should aim to incarnate.  Moreover, as I discuss below, this melding of GG with cog-neuro is currently enjoying a happy interaction somewhat analogous to what happened with Mendel before.

Before saying more, let me make clear that of course biolinguists would love to make more robust contact with current work in biology. Indeed, I think that this is happening and that Minimalism is one of the reasons for this. But I will get to that. For now let’s stipulate that the more interaction between apparent disparate domains of research the better. However, absence of apparent contact and the presence of different methods does not mean that subject matters differ. Human linguistic capacity is biologically grounded. As such inquiry into linguistic patterns is reasonably considered a biological inquiry about the cognitive capacities of a very specific animal; humans. It appears that dualism is still with us enough to make this obvious claim contentious.

The point of all of this? I actually have two: (i) to note that the standard criticism of GG as not real biolinguistics at best rests on unjustified dualist premises (ii) to note that one of the more interesting features of modern Minimalist work has been to instigate tighter ties with conventional biology, at least in the neuro realm. I ranted about (i) above. I now want to focus on (ii), in particular a recent very interesting paper by the group around Stan Dehaene. But first a little segue.

I have blogged before on Embick and Poeppel’s worries about the conceptual mismatch between the core concepts in cog-neuro and those of linguistics (here for some discussion). I have also suggested that one of the nice features of Minimalism is that it has a neat way of bringing the basic concepts closer together so that G structure and its bio substructure might be more closely related. In particular, a Merge based conception of G structure goes a long way towards reanimating a complexity measure with real biological teeth. In fact, it is effectively a recycled version of the DTC, which, it appears, has biological street cred once again.[1] The cred is coming from work showing that one can take the neural complexity of a structure as roughly indexed by the number of Merge operations required to construct it (see here). A recent paper goes the earlier paper one better by embedding the discussion in a reasonable parsing model based on a Merge based G. The PNAS paper (Henceforth Dehaene-PNAS) (here) has a formidable cast of authors, including two linguists (Hilda Koopman and John Hale) orchestrated by Stan Dehaene. Here is the abstract:
Although sentences unfold sequentially, one word at a time, most linguistic theories propose that their underlying syntactic structure involves a tree of nested phrases rather than a linear sequence of words. Whether and how the brain builds such structures, however, remains largely unknown. Here, we used human intracranial recordings and visual word-by-word presentation of sentences and word lists to investigate how left-hemispheric brain activity varies during the formation of phrase structures. In a broad set of language-related areas, comprising multiple superior temporal and inferior frontal sites, high-gamma power increased with each successive word in a sentence but decreased suddenly whenever words could be merged into a phrase. Regression analyses showed that each additional word or multiword phrase contributed a similar amount of additional brain activity, providing evidence for a merge operation that applies equally to linguistic objects of arbitrary complexity. More superficial models of language, based solely on sequential transition probability over lexical and syntactic categories, only captured activity in the posterior middle temporal gyrus. Formal model comparison indicated that the model of multiword phrase construction provided a better fit than probability- based models at most sites in superior temporal and inferior frontal cortices. Activity in those regions was consistent with a neural implementation of a bottom-up or left-corner parser of the incoming language stream. Our results provide initial intracranial evidence for the neurophysiological reality of the merge operation postulated by linguists and suggest that the brain compresses syntactically well-formed sequences of words into a hierarchy of nested phrases.
A few comments, starting with a point of disagreement: Whether the brain builds hierarchical structures is not really an open question. We have tons of evidence that it does, evidence that linguists a.o. have amassed over the last 60 years. How quickly the brain builds such structure (on line, or in some delayed fashion) and how the brain parses incoming strings in order to build such structure is still opaque. So it is misleading to say that what Dehaene-PNAS shows is both that the brain does this and how. Putting things this way suggests that until we had such neural data these issues were in doubt. What the paper does is provide neural measures of this structure building processes and provides a nice piece of cog-neuro inquiry where the cog is provided by contemporary Minimalism in the context of a parser and the neuro is provided by brain activity in the gamma range.

Second, the paper demonstrates a nice connection between a Merge based syntax and measures of brain activity. Here is the interesting bit (for me, my emphasis):

Regression analyses showed that each additional word or multiword phrase contributed a similar amount of additional brain activity, providing evidence for a merge operation that applies equally to linguistic objects of arbitrary complexity.

Merged based Gs treat all combinations as equal regardless of the complexity of the combinations or differences among the items being combined. If Merge is the only operation, then it is easy to sum the operations that provide the linguistic complexity. It’s just the same thing happening again and again and on the (reasonable) assumption that doing the same thing incurs the same cost we can (reasonably) surmise that we can index the complexity of the task by adding up the required Merges. Moreover, this hunch seems to have paid off in this case. The merges seem to map linearly onto brain activity as expected if complexity generated by Merge were a good index of the brain activity required to create such structures. To put this another way: A virtue of Merge (maybe the main virtue for the cog-neuro types) is that it simplifies the mapping from syntactic structure to brain activity by providing a common combinatory operation that underlies all syntactic complexity.[2] Here is Dehaene-PNAS paper (4):
A parsimonious explanation of the activation profiles in these left temporal regions is that brain activity following each word is a monotonic function of the current number of open nodes at that point in the sentence (i.e., the number of words or phrases that remain to be merged).
This makes for a limpid trading relation between complexity as measured cognitively and as measured brain-wise transparent when implemented in a simple parser (note the weight carried by “parsimonious” in the quote above). What the paper argues is that this simple transparent mapping has surprising empirical virtues and part of what makes it simple is the simplicity of Merge as the basic combinatoric operation.

There is lots more in this paper. Here are a few things I found most intriguing.

A key assumption of the model is that combining the words into phrases occurs after the word at the left edge of the constituent boundary (2-3):
…we reasoned that a merge operation should occur shortly after the last word of each syntactic constituent (i.e., each phrase). When this occurs, all of the unmerged nodes in the tree comprising a phrase (which we refer to as “open nodes”) should be reduced to a single hierarchically higher node, which becomes available for future merges into more complex phrases.
This assumption drives the empirical results. Note that it indicates that structure is being built bottom-up. And this assumption is a key feature of a Merge based G that assumes something like Extension. As Dehaene-PNAS puts it (4):
The above regressions, using “total number of open nodes” as an independent variable, were motivated by our hypothesis that a single word and a multiword phrase, once merged, contribute the same amount to total brain activity. This hypothesis is in line with the notion of a single merge operation that applies recursively to linguistic objects of arbitrary complexity, from words to phrases, thus accounting for the generative power of language
If the parsing respects the G principle of Extension then it will have to build structure in this bottom up fashion. This means holding the “open” nodes on a stack/memory until this bottom up building can occur. The Dehaene-PNAS paper provides evidence that this is indeed what happens.

What kind of evidence? The following (3) (my emphasis):
We expected the items available to be merged (open nodes) to be actively maintained in working memory. Populations of neurons coding for the open nodes should therefore have an activation profile that builds up for successive words, dips following each merge, and rises again as new words are presented. Such an activation profile could follow if words and phrases in a sentence are encoded by sparse overlapping vectors of activity over a population of neurons (27, 28). Populations of neurons involved in enacting the merge operation would be expected to show activation at the end of constituents, proportional to the number of nodes being merged. Thus, we searched for systematic increases and decreases in brain activity as a function of the number of words inside phrases and at phrasal boundaries.
So, a Merge based parser that encodes Extension should show a certain brain activity rhythm indexed to the number of open nodes in memory and the number of Merge operations executed. And this is what the paper found.

Last, and this is very important: the paper notes that Gs can be implemented in different kinds of parsers and tries to see which one best fits the data in their study. There is no confusion here between G and parser. Rather, it is recognized that the effects of a G in the context of a parser can be investigated, as can the details of the parser itself. It seems that for this particular linguistic task, the results are consistent either a bottom-up or left corner parser, with the latter being a better fit for this data (7):
Model comparison supported bottom-up and left-corner parsing as significantly superior to top-down parsing in fitting activation in most regions in this left-hemisphere language network…
Those findings support bottom-up and/or left-corner parsing as tentative models of how human subjects process the simple sentence structures used here, with some evidence in favor of bottom-up over left-corner parsing. Indeed, the open-node model that we proposed here, where phrase structures are closed at the moment when the last word of a phrase is received, closely par- allels the operation of a bottom-up parser.
This should not be that surprising a result given the data that the paper investigates. The sentences of interest contain no visible examples where left context might be useful for downstream parsing (e.g. Wh element on the left edge (see Berwick and Weinberg for discussion of this)). We have here standard right branching phrase structure and for these kinds of sentences non-local left context will be largely irrelevant. As the paper notes (8), the results do “not question the notion that predictability effects play a major role in language processing” and as it further notes there are various kinds of parsers that can implement a Merge based model, including those where “prediction” plays a more important role (e.g. left-corner parsers).
That said, the interest of Dehaene-PNAS lies not only in the conclusion (or maybe not even mainly there), but in the fact that it provides a useful and usable model for how to investigate these computational models in neuro terms. That’s the big payoff, or IMO, the one that will pay dividends in the future. In this, it joins the earlier Pallier et al and the Ding et al papers. They are providing templates for how to integrate linguistic work with neuro work fruitfully. And in doing so, they indicate the utility of Minimalist thinking.
Let me say a word about this: what cog-neuro types want are simple usable models that have accessible testable implications. This is what Minimalism provides. We have noted the simplicity that Merge based models afford to the investigations above; a simple linear index of complexity. Simple models are what cog-neuro types want, and for the right reasons. Happily, this is what Minimalism is providing and we are seeing its effects in this kind of work.
An aside: let’s hear it for stacks! The paper revives classical theories of parsing and revives the idea that brains have stacks important for the parsing of hierarchical structures. This idea has been out of favor for a long time. One of the major contributions of the Dehaene-PNAS paper is to show that dumping it was a bad idea, at least for language, and, most likely, other domain where hierarchical organization is essential.
Let me end: there is a lot more in the Dehaene-PNAS paper. There are localization issues (where the operations happen) and arguments showing that simple probability based models cannot survive the data reviewed. But for current purposes there is a further important message: Minimalism is making it easier to put a lot more run of the mill everyday bio into biolinguistics. The skepticism about the biological relevance of GG and Minimalism for more bio investigation is being put paid by the efflorescence of intriguing work that combines them. This is what we should have expected. It is happening. Don’t let anyone tell you that linguistics is biologically inert. At least in the brain sciences, it’s coming into its own, at last![3]

[1] Alec Marantz argued that the DTC is really the only game in town. Here’s a quote:
…the more complex a representation- the longer and more complex the linguistic computations necessary to generate the representation- the longer it should take for a subject to perform any task involving the representation and the more activity should be observed in the subject’s brain in areas associated with creating or accessing the representation or performing the task.
For discussion see here.

[2] Note that this does not say that only a Merge based syntax would do this. It’s just that Merge systems are particularly svelt systems and so using them is easy. Of course many Gs will have Mergish properties and so will also serve to ground the results.
[3] IMO, it is also the only game in town when it comes to evolang. This is also the conclusion of Tatersall in his review of Berwick and Chomsky’s book. So, yes, there is more than enough run of the mill bio to license the biolinguistics honorific.


  1. Dualism anyone?

    It's a good thing no GGer has ever made any claims about how human language is unique among animal communication systems.

    Less sarcastically, as a phonetician, it's always been obvious to me that language is rooted in biology. Phonology and syntax are more abstract, but to the extent that they're about human language, there's still a more or less direct connection to a biological foundation.

    But it's not clear to me what emphasizing the bio in biolinguistics gets you (or any other proponent) other than bureaucratic demarcations. Are you arguing that linguistics departments should be housed in biology departments? That linguists should be competing directly with birdsong and bee-dance researchers for grant funding?

    The fact that Merge seems to make nice, relatively easily tested predictions about brain activity seems to me to be less biological and more computational. Or are you arguing that the methods used to measure brain activity are what makes this kind of study biological?

    Getting back to the sarcasm (though perhaps not completely), what happened to your skepticism about statistics? You wrote, in the linked post, that "It seems that even in the hands of experts the tools systematically mislead.... Even more amusing, it seems that lack of statistical training leaves you better off."

    How confident are you that the many steps between the computational concept of Merge and "linear" increases in certain kinds of brain activity in certain brain regions have successfully avoided being systematically misled by (presumably) expertly handled regression modeling?

    1. Disciplinary cross talk is always complicated. What is required is that the core concepts of one discipline be made accessible in a usable way to the other. One of the nice features of Minimalist syntax is that it has stripped things down to the basics: the key is unbounded hierarchy and the measure of syntactic complexity can be measured in the number of Merge steps. This is the core idea, I believe. And this idea has legs when applied to cog-neuro questions. Why? Because it is simple and straightforward and focuses on a key (maybe THE key) feature of language: hierarchy and movement dependencies. This is an analytic gran that cog-neuro can investigate, and it is doing so. There are finer grained claims, but frankly the tools of neuroscience are too blunt to identify these. The fillip that minimalism has given to bioling in cog-neuro domain has been by zeroing in on this feature of lang and making it easy for others to use. This is really no mean accomplishment. And, yes, it is bio relevant. We are seeing cog-neuro people finding central features in brain signals. This is good for questions of localization and even for questions of organization (stacks, yes stacks!). So, yes, this is a big step forward.

      As for the stats: I am of course wary. But, heh, maybe this time it's been done right. I leave this for others to judge.

    2. And, yes, it is bio relevant.

      How, exactly? What implications are there for biology? Is it all and only that there it makes predictions that are tractable in cognitive-neuroscience?

      It's trivially true that human language is biological, since it's human language. Ditto for cognitive-neuroscience, since the neurons of interest are in brains.

      But you pretty clearly want the biological implications of linguistics to be non-trivial. You haven't made the case here, at least not yet. And it's never been clear to me exactly what this case could be, even in principle.

    3. This probably won't help but I will give it one more try:
      Linguistics is to biology what Mendel's genetics was to biochemistry. Linguists have discovered that human linguistic capacity rests (in part) in internalization of Gs with specific properties, an important one being hierarchical recursion. We know this to be the case. This sets a job for biology, two in fact. First, in what way is it internalized in brains. Second, how could it have evolved? Thus, ling puts questions on the biological agenda. Moreover, these questions are currently being vigorously addressed (hence my discussion of the Nelson et al paper and the excellent feedback by, a.o. Jon Brennan). As for the second item on the agenda, some think that current speculation has moved the issue forward, including Tattersall and Dawkins. Are they right? I dunno. But they are right to think that the issue is important and that current ling speculation is a step towards solving the evolving problem.

      My point was that this entirely analogous to what happened in biochemistry wrt Mendelian genetics. The latter placed a question on the agenda for biochemists to solve. And they understood that they had to solve it. Read the last line of the original Watson and Crick paper to see this. Pari pass with linguistics, Gs and FL/UG. So, in exactly this sense, ling is part of bio.

      Last point: I did not take this to be a controversial position. Hence I welcome your "It's trivially true that human language is biological, since it's human language. Ditto for cognitive-neuroscience, since the neurons of interest are in brains." Quite right. What astounds me is that this seems to be contentious. Where we might differ is that I think that the interactions have become more interesting of late and I believe that this is in part due to the rise of Minimalism. I agree with Jon and Tim that many of the questions being asked could have been asked before (though not all if Jon is right), but they were not. I suspect that this is because our conception of G obscured the key fact of interest: unbounded hierarchy and this has been the cynosure of current ling discussion. So I think that Minimalism has facilitated a comity that was long overdue.

      So does ling have bio implications? Sure. There must be some way that the brain codes for hierarchy of the kind we know exists. Are biologists looking at this? Yes. As for evolving, well I think even here MP has made a small contribution by refocusing attention on what could have evolved and removing attention from what just "popped in" in a flash (Merge). This too is a step forward, I believe.

      I am sure this won't satisfy you. But that'w what makes the world such a fun place: disagreement over fundamental concerns.

    4. I think it satisfies me more than you expect it to. This all might just be a disagreement about terminology.

      I agree that it's an interesting problem to try to figure out how linguistic structure is implemented in actual brains. I'm not a syntactician, so my knowledge is limited, but I find the idea that minimalism makes this more tractable fairly compelling.

      I find the evolutionary issues somewhat less compelling, mostly just because it seems less likely that we'll make much progress on that front. I don't doubt that evolution is true, nor that something interesting happened to give humans the linguistic capacities we have. I'm a lot more skeptical that we'll ever be able to figure out exactly what that interesting happening was.

      Ultimately, it's still not clear to me why either of these cog-neuro or evolutionary implications show that there's a benefit to adding the bio- prefix to linguistics. Two different scientific fields can feed each other interesting results and fertilize interesting questions without one subsuming the other, right?

    5. There is a pretty large (maybe dominant) conception of linguistics as an enterprise unrelated to bio. Languages are not grounded in brians on this view nor do they supervene on mental structures of any linguistic interest. The bio prefix distinguishes the present enterprise from that one and invites active cooperation. And it works both ways. We think atht what we have found has implications for bio and hope that they will return the favor. One big argument against certian conceptions of the brian (e.g. Connectionism) is that it cannot serve for linguistic cognition. At any rate, we think there is and can be more fruitful cross talk. This is not what you aim for if your conception of language eschews any bio roots.

  2. I find the relationship between the two halves of this post confusing (perhaps even nearly contradictory ...)

    I agree wholeheartedly with the first half of the post, down to and including the point that "Whether the brain builds hierarchical structures is not really an open question". The questions that do remain open include:
    (1) What sort of hierarchical structures are they exactly? e.g. Do they look more like the minimalist sort, the TAG sort, the CCG sort, perhaps even more like the plain CFG sort, etc.?
    (2) How do those hierarchical structures get built in real time? e.g. Is it more like bottom-up, top-down, left-corner, etc.?

    I haven't read the paper yet so perhaps there's more I'm missing, but given this point from the first half of the post, I find it hard to make sense of the latter parts of the post that take these results to be evidence for "merge-based grammars". On a broad interpretation of "merge-based", this seems true, but it doesn't actually touch on either of the open questions in (1) or (2), it only relates to the point which (you and I both agree) is essentially settled. If there's some evidence for bottom-up or left-corner structure-building, great, that's some progress on (2). But is there evidence in support of merge-based grammars in any narrower sense, i.e. evidence that minimalists have something like the right answer to (1)?

    I think this is closely related to the point that Jon Brennan was making in a comment on the other post. I worry that this kind of use of the term "merge" paves the way for bad motte-and-bailey arguments that there is neuro evidence for something with the specific properties minimalists talk about (for example, forming unordered two-element sets).

    1. "is there evidence in support of merge-based grammars in any narrower sense, i.e. evidence that minimalists have something like the right answer to (1)?"

      No: there is evidence that hierarchy matters and that it is the kind of hierarchy that a simple combinatoric system that involves Merge would give you. Does it argue for an MP view of things vs a CFG or something else that chunks words into phrases? Not really. But, IMO, Merge isolates the relevant property very cleanly. It concentrates on unbounded hierarchy abstracting away from the idiosyncratic features of the elements being combined and their size. It requires that there be some "bottom-up" combination of open nodes (no Gs that involve "sinking" will fit as cleanly). This is good. And important. If you look at the paper you will note that it argues against theories that eschew hierarchy altogether (such theories are sadly all too common). The paper argues that this will not do; the brain seems to track hierarchy in roughly the way linguists have said that it should for the last 60 years. Could this have been done in a non MP based G? Sure. Did using Merge simplify things? I think so: again because it allows for basically a simple mapping from merges to indices of brain activity (given some simple ancillary assumptions about stacks). So, is MP necessary? No. Is this an argument for MP? Yes insofar as MP focuses on unbounded hierarchy. Is this an argument for MP over other "formalisms"? Not if they too have unbounded hierarchy. Should we expect cog-neuro to be able to adjudicate among competing G hypotheses? No. The methods are too blunt. Does this matter? Not to me. Right now, finding the signature properties of hierarchy building is a very good general project. I don't worry much about "bad motte-and-bailey arguments" because I don't think that cog-neuro is currently in a position to tell us much about the fine details that GGers care about. But who cares? I want to find out how the brain does language. To do it it will need to confront constitutency and hierarchy. MP provides a simple way of thinking of the issues, a very simple system (Merge, merge, merge again and again and again). Is this simple system enough? Not but it is at least this. So, if you don't demand,and that cog-neuro solve your syntax problems (which it cannot) this kind of work is really just what a bio linguist would order.

    2. Could this have been done in a non MP based G? Sure. Did using Merge simplify things? I think so: again because it allows for basically a simple mapping from merges to indices of brain activity

      I'm not sure I see the simplifying role merge played. It seems to me that the same work could have been done if we'd never even gotten past context-free grammars. I'm happy to agree to disagree on this, however, since the work didn't in fact get done until now.

      But isn't there a difference between whether the idea of merge facilitated the line of thought that led to this work being done, and whether this work provides evidence that human grammars are merge-based? It seems that, as you say, the relevant methods are too blunt (at least right now) to do the latter.

    3. I agree (as usual) with Tim's points.

      The idea that counting nodes might index processing goes back at least to Chomsky & Miller 1963; it's also essentially the intuition behind Minimal Attachment (Frazier makes this explicit in a tragically under-cited 1985 position paper.) You end up with the same intuition if you follow the trace from the classic automata used by Nelson et al. (e.g. by counting the number of reduce steps in the bottom-up parser). None of these angles seem to be particularly Minimalist-inspired.

      Where do Nelson et al. come in? Well, Hale's work has long argued that automata are a reasonable model for mediating between Gs and processing data; my own work has shown that parsing traces have success modeling brain activity. Nelson et al. bring ECoG into the mix, which offers some absolutely exquisite detail about how these parser operations are distributed in time and space. This is exactly what we need to move forward to test finer-grained aspects of these models. I love it!

      Last comment: I think the productive move here is to split, not to lump. Identifying Merge with parse-steps, node count, stack depth, or whatever relevant property of some incremental parse state seems to downplay the point that Merge (and other auxiliary assumptions of the MP) defines a special set of representations for mapping meaning to sound. Neural evidence that the brain traffics in *just these sorts* of representations when parsing is pretty interesting! But Nelson et al. don't show that. Doing so would require defining a non-minimalist G, a minimalist G, parsing both of them, and seeing which best fits brain data. (spoiler alert: first round goes to the minimalist G! doi: 10.1016/j.bandl.2016.04.008)

  3. A little too much worry about the sociology of science in here for my taste. So you would like to lump the GGers in with the biologists, but split them off from the "languists" (not the topic of this post, I know). I think biologists don't think GG should be classified as biology because it largely doesn't employ methods, concepts, tools and arguments from modern biology. And GGers don't see "languists" as engaged in the GG enterprise for similar reasons, that they largely don't employ methods, tools, concepts and arguments from GG. What's at stake here? Some version of a unity-of-science argument? (Which I don't think you believe anyway.)

    1. No, I would like to allow that GG has bio implications and that it is fruitful to consider these. Its findings pertain to bio and looking to ground the high level descriptions biologically is both reasonable and a desideratum of the enterprise. I ma regularly told that there is nothing bio going on here. I don't see why not and if someone could show me that what I am doing is at right angles to these questions of bio and cognition I would consider this a serious criticism of the GG enterprise. Hence my reluctance to concede the point.

      As for languists, their interests are partially related to mine but not quite. The demarcation is purely in service of seeing how much of what languists do I need to worry about (e.g. how relevant is typology to study of FL/UG?). There is no unity of science argument, just a there are many roads to Rome argument and abstract linguistics is relevant to biology just as much as bee dances and Mendel's genes. In fact, as regards the latter, I take it as a good model of what we should be hoping to find in brains (hence the discussion of the PNAS paper).

    2. Linguistics is certainly relevant to biology, but if that's the case then why are linguists only engaging with neurobiology and genetics in the most timid ways possible? When neuroscientists try and figure out the neural code for spatial and conceptual navigation, for instance, they go way beyond correlational analysis of oscillatory entrainment to external stimuli (as in Ding et al and much other work). They also look at what's going on inside the rest of the brain, examining cross-frequency couplings (like phase-amplitude coupling), spike time coordination, cellular analysis, etc.

      If Norbert was genuinely proud of the current linguistics-biology relationship, then why not blog more about biology? Take this recent study by Constantinescu et al ( They show that the neural code which has long been implicated in spatial navigation may also be implicated in navigating more abstract representations, such as conceptual space (recent work also points to the same code being implicated in navigating auditory space, too).

      This work should be of exceptional interest to linguists. If this is how the brain interprets basic relations between conceptual representations, then we should probably put aside the Jabberwocky EEG studies and eye-tracking experiments for a little while (important though they may be) and engage in these sorts of emerging frameworks.

      Instead of claiming that some region of interest in the brain (or some oscillatory band) is responsible for some complex process (e.g. "semantic composition is implemented via gamma increases", "syntax is represented in anterior Broca's area", and "my Top 10 Tom Cruise movies are stored in the angular gyrus"), exploring the neural code is of much greater importance and urgency. This is something Gallistel actually stressed at CNS recently.

      Final implication for Merge, the "Basic Property", and other rhetorical and computational constructs: The Constantinescu study actually reflects a more general trend in neurobiology these days. Things that were once deemed highly domain-specific are now being understood to implement much more generic computations, and the only domain-specific things left are the *representations* these computations operate over. In other words, good luck trying to find the "neural correlates of Merge" if you only have your Narrow Syntax glasses on.

  4. How shall we decide whether linguistics is part of biology? The argument seems to be the following:

    (1) You can't base the decision on a comparison of the tools and methods both fields are using, but

    (2) You *can* base the decision on whether similar work won a Nobel Prize at some point, though it remains unexplained why the Nobel committee should have the power to decide what is biological [actually physiological/medical, which isn't the same thing] and what is not.

    Except that (2) is specious given that many linguists don't believe that bee communication has much of anything to do with linguists, so we must rewrite it as:

    (2) Linguistics is part of biology because work on bee communication, which many (most?) linguists consider nonlinguistic, has won a Nobel Prize for physiology or medicine (not biology).

  5. If the argument is about the use of behavioral data, then there is a perfectly valid field to which linguistics belongs--namely, psychology/cognitive science (I refuse to get embroiled in a debate about the difference between the two). If you want to tie that to something, well, biological, as Dehaene et al. appear to be doing, great--that's psychobiology or cognitive neuroscience or whatever you want to call it. As you're phrasing it, though, *everything* psychological is biological--personality theory, behaviorism, positive psychology, etc.

    Look, I grant that disciplines are always fuzzy at the edges. At my institution, people working on birdsong, bee waggling, etc., would be called something like behavioral biologists, and we have some in psychology and some in biology. Nobody fusses about it too much.

    I think the distinction is easier: Suppose you are looking at behavior/cognition on its own, without any direct reference to cells, genes, hormones, whatever. That's a perfectly valid thing to do because sometimes it isn't useful to be more reductive, just s it's not useful to reduce all chemistry to physics. In that case, you're doing psychology or cognitive science.

    If you're looking at cells, genes, hormones, whatever without reference to behavior/cognition, then you're doing biology, not psychology/cognitive science.

    To the extent that GGers are looking just at GG, that's not biology. The fact that it is implemented in the brain doesn't change that. When they're *looking* at the implementation in the brain, it is, or it's a hybrid between psychology and biology.

  6. When we talk about phonetics, a large part of the discussion is in terms of anatomical features, and how they are transduced from lexemes (production) or to lexemes (perception). That is a kind of bio that is obviously, measurably, and directly involved in language use and structure...

  7. Norbert writes in a thread above: "There is a pretty large (maybe dominant) conception of linguistics as an enterprise unrelated to bio. Languages are not grounded in brians on this view nor do they supervene on mental structures of any linguistic interest. The bio prefix distinguishes the present enterprise from that one and invites active cooperation."

    This seems a pretty feeble straw man of most conceptions of linguistics that come to mind. Take Zipf, Hockett, Ohala, Lieberman, Dor & Jablonka, Jackendoff, Fisher, Millikan, Hurford, Emmorey, Tomasello, Foley, Croft, Bybee, Evans & Levinson, Christiansen & Chater — all of these seriously engage with biology, and for none of them you could reasonably claim that they see linguistics as "an enterprise unrelated to bio". They may or may not adopt your preferred framing of the major issues at stake, but they certain do see language as grounded in brains, supervening on mental structures, and shaped and constrained by human biology.

  8. Paradoxically, what Minimalism in Norberts description appears to aim for is absolutely minimal contact between language and biology. The goal is to characterize a single hinge point (e.g., Merge) and hang everything off of that. In contrast, many of the approaches mentioned in my comment above see a more multifaceted relation between biology, cognition, and language. They see language in coevolution with physiological and anatomical features; they study how neurobiological principles and processes shape and constrain language structure; they use animal models to study a range of traits that seem involved in language as part of human behaviour; they look at epigenetics to shed light on language disorders; they work out parallels and differences between biological and cultural evolution; and so on.

    So: is linguistics biologically inert? Far from it. Is minimalism the only game in town when it comes to interfacing language and biology? Far from it. Linguistics would benefit from less territoriality and more collaboration. We don't need bio-prefixes and other facile forms of virtue signalling; we need serious work from a variety of perspectives to contribute to solving the formidable puzzle posed by language as a part of human behaviour. As I've pointed out before, Minimalism can only benefit from this: the more explanatory work is offloaded to other approaches, the more truly minimalist its own targets become.

    1. Actually Norbert's Minimalism wants to make contact where contact is worth making. Norbert believes that some fruitful contact can be made with the euro world around Merge (and that it is being made). He also thinks that it is possible to do interesting work in FLW and that some of this is being done. What he does not believe is that there is much interesting contact to be made wrt Merge if one's interest is explaining how it evolved piecemeal via natural selection. Not gonna happen.

      Moreover, linguists have a role to play in getting bio to touch more closely with core work in syntax: to show how far one can get regarding ling phenomena using a Merge based theory. If one takes GB as offering a plausible account of FL/UG properties we can rephrase this as how much of GB can be recovered on a merge based system? IF it is all (I doubt this) then much of what we take as complex will have actually been the by-product of one (simple) innovation in interaction with other pre-ling cognitive capacities. The job then is to outline these capacities which, in concert with Merge, gave us what we see. Were tis kind of theory available then we would have learned a lot about how FL arose and why it has the properties it has. Would there be nothing left to do. OF COURSE NOT! THERE IS ALWAYS MORE TO DO. But some questions would have been answered and we could concentrate on others. That's how we get progress. The only real question, IMO, is whether this picture is redeemable. I still think it is.

    2. This comment has been removed by the author.

    3. Is FLW the same as FLB? Or does it entail some different? (I'm asking because I never saw the former before the last couple of posts/comments; I've only seen the FLN/FLB dichotomy in the literature and other discussions).

  9. Like all the recent Ding et al and Martin & Nieuwland work, the Nelson et al (Dehaene) study contains no theory. It's just a re-description of the data, a la Ding et al's "sentences = 1Hz, phrases = 2Hz" business. Why Nelson et al actually found gamma decreases at phrasal boundaries is not discussed in the paper, you'll notice. Typical of contemporary neurolinguistics - too much data, not enough theory. What are the intrinsic computational properties of the gamma pockets Nelson et al found, for instance? Instead of asking such questions (which demands a more serious engagement with the neurobiology literature), they simply use their data (in a pretty lazy way) to support their own pre-existing higher-level conceptions of Merge. Theoretically speaking, I don't think the field of neurolinguistics has progressed all that much since the early days of "Look, Broca's area lights up!"

    These studies have a (perhaps sensible) urge to keep the story as simple as possible (for now), and so we get really biologically implausible claims from people like Hagoort and Bastiaansen about particular, discrete oscillatory bands being responsible for complex cognitive processes, e.g. "gamma = semantic composition" and "beta = syntactic processing". This is not a joke, these are genuine claims from the literature. If your definition of "biological engagement" is simple mapping complex processes to even more complex biological oscillatory signatures (and not decomposing either the computations or investigating the brain signatures in any sophisticated depth) then yes, linguistics is currently great at it. But if you want a kind of story which will crucially satisfy researchers across *all* domains of the cognitive and biological sciences (and not just Mr Dehaene) then we're far away indeed.

    It's worth stressing that Nelson et al. did *not* investigate the neural basis of the merge operation (for that, see this paper: They rather explored the basis of phrasal labeling and transfer, which could be seen as part of the merge operation *complex*, but not part of any generic set-formation operation.

    What is needed are genuine linking hypotheses between biological levels of analysis, like what we provide here concerning the language 'oscillogenome':

    I also try and elaborate on these concerns here:

  10. Just a quibble about Von Frisch's Nobel Prize: he received it jointly with Tinbergen and Lorenz, and my understanding is that they received it not so much for the specific discoveries such as the waggle dance, but for developing methods for scientifically studying behavior and thus for founding a new science, 'ethology'.

    The 1973 press release explicitly mentions that this provided a way out of behaviorism. In that light, a Nobel Prize for Chomsky in the 1960s might have been well justified.

    I agree with most of this post - although I'd prefer replacing 'GG' with 'linguistics' throughout. Why not be inclusive when you can? In fact, I suspect that the tribal attitudes within much of linguistics have a lot to do with the lack of recognition from outside.

  11. This comment has been removed by the author.

  12. This comment has been removed by the author.

  13. I read a article under the same title some time ago, but this articles quality is much, much better. How you do this.. lgals9 elisa