Rough Theory

So it’s been a hot and smoky weekend in Melbourne. The cool change has just come through – not much help unfortunately, I think, for those on the massive firefront. But a signal for me to shake off my heat-induced sluggishness, and get a bit of thinking done.

I’m well and truly past my self-imposed deadline for writing something substantive on the reading group discussion of the debate between Pinker & Jackendoff and Chomksy, Hauser & Fitch, over the evolution of the language faculty – the trajectory of which is conveniently outlined at Language Log. I’ve hesitated to post in part because I was trying to work out a way to break through what seemed to be the main issue that arose in the reading group discussion: the perception that these articles were highly technical pieces, written by and for specialists, such that deciding between the various “they said-they said” arguments would be essentially impossible for a lay reader. I wanted to work out whether there were some way to approach these readings that could at least minimise this reaction – since the reaction, after all, tends to become a self-fulfilling prophecy.

What I do below is therefore not summarise the debate – as I suspect that even a summary of these already fairly condensed and economical texts would mire us in minutiae. Instead, I provide some suggestions that might help someone read the debate a bit more easily – mainly by locating the various empirical skirmishes in the context of what I take to be the overarching theoretical conflict that motivates the empirical battles. I’ll say at the outset that I very much doubt the reading framework I outline is the only – let alone the best – way of working your way into these texts. I offer it more as an example of how I personally went about trying to make sense of this discussion, without a specialised background in any of the scientific fields referenced in these texts: hopefully, your personal path through these texts will substantially improve on mine.

I want to start where Chomsky, Hauser & Fitch begin: by noting that the purpose of these articles – a point agreed by all sides in the debate – is to make a case for the value of interdisciplinary work when investigating the evolution of the language faculty. This point is more important than it may seem, particularly to non-specialist readers: it tells us that, in spite of first impressions, these articles will not assume that readers have any specific disciplinary background – they may assume a sound scientific knowledge of some sort, but they won’t be assuming a socialisation into any particular scientific discipline: the nature of interdisciplinary work is that you cannot assume such things. This therefore holds out hope that, in principle, a non-specialist reader ought to be able to make sense of these debates.

Where I’d like to go next is to make the suggestion that, in the beginning, readers bracket the empirical skirmishes. This may sound a bit perverse, as these empirical conflicts make up the overwhelming majority of the exchange – and, in fact, mark the least contested points of contact between the two sides: the existence or nature of any theoretical argument is disputed within these texts; there is more consensus over where the empirical fault lines lie. Nevertheless, I suspect we’ll find more light if we step back a bit from the empirical heat, and take a closer look at the strange, half-denied theoretical debate that runs through these articles.

I’ve characterised the theoretical debate as “half denied” for a specific reason: Chomsky, Hauser & Fitch don’t admit that their position is motivated by any specific theoretical perspective. Instead, their repeated claim is that they are posing the only possible scientific questions one could pose about the evolution of language at this moment in time – a point to which I’ll return in a moment. Pinker & Jackendoff then argue: no, these aren’t the only possible scientific questions that could be posed – and, in fact, you have only posed these questions because you are presupposing the validity of a particular linguistic theory: the Minimalist Program. Chomsky, Hauser & Fitch then say: we are doing no such thing, our questions have nothing to do with the Minimalist Program – they are instead the only possible questions a scientist could ask. Pinker & Jackendoff then come back and say: actually, a scientist could ask many other questions, if not already inclined to believe the Minimalist Program were true: here, look! – we’ll show you some…

Essentially, then, in venerable academic tradition, the debate boils down to a set of “did not!” – “did too!” exchanges over whether Chomsky, Hauser & Fitch are pushing the Minimalist Program on the sly. The empirical exchanges all fit within this context – which is why so many of the empirical debates are not primarily about what the facts are – not about who did what, when, where, why and how – but about which specific facts matter, and in what ways they matter, for understanding the evolution of human language. I am suggesting, in other words, that this is not an empirical quarrel, but a philosophical one.

Because what I’ve written is somewhat long, I’ll tuck the main body below the fold. I’ll have to apologise in advance for the length, and for what will almost certainly be inadequate copy editing – I’m writing this on borrowed time, so to speak, and I haven’t been able to give this piece the thorough proofing it undoubtedly desperately needs.

Back to the argument: from here I need to become very cautious. Pinker & Jackendoff provide what was, to my non-specialist eyes, a quite handy (if scathing) outline of the Minimalist Program in their first article, as well as an interesting (but sketchy) discussion of a possible alternative approach in their second article. Because I am not otherwise familiar with these traditions – and because Chomsky, Hauser & Fitch refuse to dance with Pinker & Jackendoff at any length around this issue, explicitly at least – I am reluctant to outline specific claims about the Minimalist Program or its competitors, as it would be easy to have gotten a distorted impression from the materials provided in this exchange.

What I propose to do is sidestep the issue for the moment – bracketing whether the Minimalist Program is a good, bad or indifferent contribution to linguistics. I would like instead to suggest some of the reasons I’m sympathetic to Pinker & Jackendoff’s claim that Chomsky, Hauser & Fitch are proposing a vision for research into language evolution that must be highly theoretically motivated. Pinker & Jackendoff are reacting, I believe, to elements that also caused me to pause when I was reading the Chomsky, Hauser & Fitch pieces. Specifically, I found myself being repeatedly thrown out of the text by what seemed – again, to very inexpert eyes – to be quite arbitrary assertions about the nature of scientific investigation, research into evolution, and the nature of the human language faculty.

Chomsky, Hauser & Fitch do always, of course, hold their arbitrary choices out to the test of empirical accuracy: they are consistently very careful to emphasise the hypothetical and tentative character of their proposals (although they are not always so tentative in dismissing potential alternatives). I am not accusing them of stacking the empirical deck or pre-determining the outcomes that will fall out of their research.

Nevertheless, they seem to behave as though no reasonable scientist could ever bring themselves to disagree with their specific framing of this research field. And I find myself perplexed, either because, in some cases, I am aware that there are quite reasonable people who in fact would approach the issue in a different way, or I can easily – even with a non-specialist background – think of possible alternative visions for this research process. It therefore becomes a puzzle why Chomsky, Hauser & Fitch work so hard to package their specific research program as though it has fallen quasi-naturally out of the best possible application of the scientific method to problems of language evolution.

I read Pinker & Jackendoff’s articles as motivated by a similar confusion. They note the arbitrariness of some of the choices made by Chomsky, Hauser & Fitch, and then ask: given that we cannot make sense of the positions taken by Chomsky, Hauser & Fitch based on some neutral application of science to problems of language evolution, how can we best make sense of the specific program of research into language evolution that they seize upon – and of the alternative research programs they either reject or fail to discuss? Pinker & Jackendoff argue that Chomsky, Hauser & Fitch’s position makes most sense if they had arrived at it by asking themselves what vision of linguistic evolution would result if the Minimalist Program were true. Pinker & Jackendoff may well be wrong – certainly Chomsky, Hauser & Fitch deny this charge. But it certainly seems, to my admittedly inexperienced self, that Chomsky, Hauser & Fitch have not offered a suitable and persuasive competing explanation.

To provide a bit more grist for discussion, what I want to try to do next is highlight a few of the kinds of claims that puzzle me in the Chomsky, Hauser & Fitch articles. I should indicate at the outset that many of these claims, if articulated just a slight bit differently – with a recognition, for example, of their own idiosyncratic or contestable character – probably wouldn’t throw me out of the text. As formulated, however, the claims make me somewhat nervous, for reasons I’ll discuss under the following loose categories:

(1) Chomsky, Hauser & Fitch’s understanding of the criteria for “scientific” research;

(2) their terminological and ontological distinction between narrow and broad language faculties;

(3) their somewhat idiosyncratic vision of how evolutionary processes shape the development of the language faculty.

A quick note before diving in: my sense on reading this exchange is that Chomsky, Hauser & Fitch leave the reasoning behind these three elements of their framework somewhat underdetermined, in the sense that their stated explanations for privileging certain hypotheses, and downgrading or not discussing others, do not appear fully to account for the positions they take. For this reason, I would suggest, Pinker & Jackendoff have found it plausible to argue that something else – some insufficiently explicit set of criteria – must be operating in the background to motivate the choices Chomsky, Hauser & Fitch make. Given Chomsky’s current research interests, I don’t find it implausible that the Minimalist Program should provide these background criteria, but I regard this as less important than getting a clearer view of the apparent arbitrariness of some of the choices expressed in the Chomsky, Hauser & Fitch articles. I’ll apologise in advance for what will no doubt be a highly abbreviated and condensed outline of these issues.

(1) What makes a program of research “scientific”?

This point may be the most difficult for me to express and substantiate clearly – I’ll apologise in advance if I don’t communicate my points well, or if I have simply misunderstood the stakes of the debate. What concerns me in Chomsky, Hauser & Fitch’s texts, though, is an apparent slide from fairly unproblematic standards for scientific investigation – empiricism, for example, and falsifiability – to a much narrower set of standards, which are then presented as though they automatically follow from a scientific commitment to empiricism and falsifiability. Thus, for example, Chomsky, Hauser & Fitch argue:

Science progresses by stating and testing falsifiable hypotheses, and the hypotheses more conducive to progress are those that are most readily falsifiable (“strong”), because a falsifiable hypothesis that repeatedly resists falsification is likely to be true. (2005 p. 193)

I may simply be reading this sentence oddly but, to me, there is a gap between the content of the first clause – that science progresses by stating and testing falsifiable hypotheses – and the second – that the easier a hypothesis is to falsify, the better that hypothesis is, from a scientific point of view. To some degree, this may come down to what Chomsky, Hauser & Fitch mean by a “readily” falsifiable hypothesis. If this means simply that the hypothesis makes clear predictions, I see no problem.

In places, however, Chomsky, Hauser & Fitch suggest that they might also understand the notion of “readily” falsifiable to encompass things like the technical feasibility of designing an experiment, based on current tools and techniques. So, for example, they dismiss hypotheses that would seem to require historical evidence, on the grounds that fossil records are not adequate, and they dismiss hypotheses that would seem to require reverse engineering from current functions, on the grounds that current function does not cast adequate light on evolution (which, as a side point, seems a problematic contention to me: is reverse engineering from current function not commonly used for evolutionary research?).

These may be quite important critiques, but it is important, I think, to recognise that they don’t actually “connect” with the “definitional” issue for scientific research: hypotheses that might currently appear testable only by unavailable fossil evidence, or by problematic reverse engineering techniques, might in fact make predictions that are quite strong, quite falsifiable – once we obtain relevant data or become clever enough in constructing new experimental techniques. This issue is highlighted by the evidence that Chomsky, Hauser & Fitch put forward as their “gold standard” model for the sorts of research we ought to do: the comparative studies of animal communication, many of which rely on extremely clever experimental techniques (some of which Chomsky, Hauser & Fitch outline in passing) that have been developed only recently. Many of the hypotheses this data can now confirm or refute would have appeared untestable in practice a short time ago.

Chomsky, Hauser & Fitch’s own data therefore suggest what, I believe, the history of science also easily shows: that currently untestable – but strong, falsifiable – hypotheses can be a major driving force in scientific progress. Ruling out hypotheses on the grounds that Chomsky, Hauser & Fitch cannot think of a means of testing them now is not an application of the scientific method.

In fairness, I should note that Chomsky, Hauser & Fitch do move back and forth between the more and less problematic dimensions of this claim: at times (their discussion of the vagueness of the term “communication” in discussions of adaptation, for example) they clearly are pointing out a severe indeterminateness in the hypothesis that would make it very difficult to open the issue to experimental investigation (cf. 2005 pp. 185-86). At other times, however, I think this clarity becomes blurred, and they move into something closer to an argument from lack of imagination: they cannot personally see how a hypothesis could viably be tested, so they rule it out of serious consideration (cf. 2005 pp. 184, 206).

Chomsky, Hauser & Fitch are of course well within their rights, of course, to say that their empirical investigations are ready to go, and so we might as well pursue their lines of enquiry while waiting for other experimental techniques to be developed. My point is simply that this is neither as global nor as authoritative a claim as implying that opposing positions that cannot currently be empirically tested are for this reason alone somehow intrinsically unscientific.

(2) What is distinctive about the human language faculty?

My concern here is more basic, and I will state it very briefly. Chomsky, Hauser & Fitch propose a terminological distinction between what they call the “broad” language faculty, which comprises all the systems and mechanisms that contribute to the language faculty that are shared either (a) with other animals or (b) with other cognitive faculties within humans, and what they call the “narrow” language faculty, which comprises only those systems and mechanisms that are unique to humans and to the language faculty. Into the narrow language faculty, Chomsky, Hauser & Fitch place recursion, while also holding out the possibility to add other elements (or to conclude that there is no specific human language faculty), pending the analysis of further empirical data.

My concern with this discussion is that, in places, Chomsky, Hauser & Fitch are quite clear that the distinction between the narrow and broad language faculties is not meant to be an ontological one, but only a matter of terminological convenience and clarity. They note, for example:

The distinction itself is intended as a terminological aid to interdisciplinary discussion and rapprochment, and obviously does not constitute a testable hypothesis. (2005 p. 181)

Having in a sense “cocooned” this distinction by defining it as beyond the reach of empirical verification, Chomsky, Hauser & Fitch then – to me – appear to try to have their cake and eat it too: they freight what they initially characterise as a “terminological aid” with a set of strong ontological claims, arguing that the components of the broad language faculty are subject to essentially different evolutionary pressures from the components of the narrow language faculty. The broad language faculty, in their account, can be subject to a wide variety of adaptational pressures – including selection for the ability of components of the broad language faculty to enhance communication. The narrow language faculty, strangely, cannot be subject to pressures to ability to enhance communication – in part because Chomsky, Hauser & Fitch conceptualise the narrow language faculty as an abstract computational system that is specialised more to provide an internal language of thought, than to provide a means of communication with other humans (2002 pp. 1570-71). They then propose – as a hypothesis subject to empirical verification – that this recursive computational system, initially evolved for other purposes, might have become implicated in communication relatively late in human evolutionary history:

One possibility, consistent with current thinking in the cognitive sciences, is that recursion in animals represents a modular system designed for a particular function (e.g., navigation) and impenetrable with respect to other systems. During evolution, the modular and highly domain-specific system of recursion may have become penetrable and domain-general. This opened the way for humans, perhaps uniquely, to apply the power of recursion to other problems. (2002 p. 1578)

They thus conceptualise the narrow language faculty as a kind of “emergent” phenomenon – a sort of evolutionary Athena burst from the head of Zeus full-formed – or as a “spandrel” – an accidental evolutionary byproduct not shaped by direct adaptational pressures, but emerging as a structural consequence of evolutionary forces selecting for some other function entirely. They then rebuff attempts by Pinker & Jackendoff to suggest the ways in which adaptational pressures might have shaped recursion or other human-specific elements of the language faculty (gradually and in tandem with other dimensions of the language faculty), by strongly reasserting their definitional distinction between broad and narrow language faculties, insisting that Pinker & Jackendoff honour that distinction in their own work, and arguing that Pinker & Jackendoff’s critique is confused. To me, these reactions to Pinker & Jackendoff appear to beg the question – they seem to presuppose the ontological validity of a terminological distinction, when this ontological validity is precisely what I take Pinker & Jackendoff to be trying to contest.

My point in drawing attention to this discussion is not to make any strong claim that Chomsky, Hauser & Fitch are wrong, but more to wonder – as, I think, Pinker & Jackendoff do – why this somewhat strange vision of the narrow language faculty is regarded as so important, so exciting? Chomsky, Hauser & Fitch of course subordinate their claims to empirical verification – they aren’t interested in putting forward a vision of language evolution in defiance of the evidence. Yet it is clear that they find it particularly compelling and important that the aspect of the language faculty that is unique to humans should somehow have emerged without adaptational pressures selecting for communication – that it sit outside normal adaptational processes, even though it remains a product of evolution. Why?

Pinker & Jackendoff obviously believe that the most convincing explanation is that this vision of language evolution “fits” with the Minimalist Program. Chomsky, Hauser & Fitch deny this, and put forward their own explanation for preferring this vision of the evolution of the narrow language faculty – which is the topic to which I will now turn.

(3) What is required in order to provide an account of the evolution of the human language faculty?

The reading group members will know that I have always been troubled by Chomsky’s apparent attempts to “exceptionalise” the evolution of human language. My problem is not with specific claims about uniquely human linguistic abilities, but with Chomsky’s apparent reluctance to admit that the same garden-variety evolutionary processes that bring about a host of other exceptional traits in all life forms, might plausibly be able to bring about an exceptional human-specific trait like language. From this perspective, I find the articles by Chomsky, Hauser & Fitch reassuring in their attempt to take seriously the need for an explanation of how the human language faculty could have evolved. At the same time, the articles still seem to express some views about evolution that appear – admittedly from a non-specialist’s perspective – somewhat idiosyncratic – and that, in perhaps a more sophisticated way, may echo some of those aspects of Chomsky’s earlier work that I found troubling.

I’ll start by picking up the issue left hanging at the end of the previous section: if Chomsky, Hauser & Fitch maintain that the Minimalist Program is not motivating their vision that the narrow language faculty is an “emergent” phenomenon, what do they claim is motivating it? Their main argument seems to hinge on the notion that their understanding of the language faculty would provide an easier problem for evolution to solve:

…a minimalist FLN [narrow language faculty] would be easier to implement neurally, easier to code genetically, and easier to evolve… (2005 p. 184)

I’m not clear, though – and here I’d love to have persons more expert than I in these matters chime in – whether this a very normal conception of how evolution works. Several issues worry me: first, is it reasonable to assume the sort of “three birds with one stone” perspective suggested here – is it reasonable to expect, for example, that something easier to code genetically would necessarily also be easier to implement neurally, or to evolve? Is it reasonable to expect that, just because something is “minimalist” from our point of view, it would necessarily provide a “small target” for evolutionary processes? Is it reasonable to assume that evolution would “prefer” a solution that is “easier”? Reading these sorts of passages, what keeps flashing into my mind is Crick’s comment: “God is a hacker”. Evolution modifies existing systems with the tools ready to hand – it is a bricoleur, not an architect. This means, among other things, that evolutionary processes can generate quite elegant design solutions – but these solutions unfold within, and are conditioned by, a specific context, rather than a blank slate… Chomsky, Hauser & Fitch read to me as though they are trying to come as close as possible to a blank slate, without absolutely breaking with the concept that the language faculty has evolved.

This approach seems to lack – if this term isn’t too ironic given the context – faith in the ability of evolutionary processes to evolve complexity – as though Chomsky, Hauser & Fitch feel they need to lend evolution a hand, by conceptualising a way to make its job as easy as possible. It is in this way that they explain why they don’t want the narrow language faculty to have evolved under adaptational pressures selecting for communication: surely it would be easier for evolution, they suggest, if the narrow language faculty could emerge, not from small steps, but from one giant leap. And wouldn’t it also be easier to investigate empirically, they ask? This then positions them, as discussed above, to dismiss hypotheses that do rely on adaptational pressures over time – that do propose that the human language faculty evolved in a series of small steps – on the grounds that these hypotheses are both more difficult for evolution to fulfill, and for us to study.

What troubles me about all of this is that I see no particular reason to assume that evolution would politely organise its inner workings for the convenience of our investigations, and I do see reason to assume – as I thought scientists routinely did in studying the evolution of other complex systems – that adaptational pressures can explain the gradual evolution of complex systems over time. If the same point had put a slightly different way, I suspect I would find myself more open to the concept: if Chomsky, Hauser & Fitch had been less dismissive of gradualist conceptions of language evolution, for example, but had said, e.g., given that we’re not in the position to study such things directly at the present moment in time, why don’t we at least try to rule out this competing hypothesis, and see where that leaves us in the end.

That they don’t do this leaves me with a residual disquiet – a disquiet that is not entirely dispelled by Chomsky, Hauser & Fitch’s commendable commitment to empirical testing of their hypotheses. I find myself, at the end of this exchange, in the strange position of not having much of an opinion one way or the other on Pinker & Jackendoff’s competing claims – but nevertheless tilting in favour of their critique of Chomsky, Hauser & Fitch. I offer these thoughts very much as an amateur in all of the areas under discussion, and I am sure there are gaping chasms in what I’ve written above. At the very least, hopefully this discussion can provide some kind of guide to non-specialist readers who may originally have been intimidated by this debate: armed with perhaps a chosen selection of the concepts covered above, hopefully you will find it easier to improve on my provisional reading of these texts…