Akeel Bilgrami’s Introduction to Noam Chomsky’s “What Kind of Creatures Are We?”

by Akeel Bilgrami

9780231175968These lectures present a lifetime of reflection by a scientist of language on the broader implications of his scientific work. The omnibus title of the lectures, “What kind of creatures are we?” conveys just how broad the implications are meant to be. They cover an impressive range of fields: theoretical linguistics, cognitive science, philosophy of science, history of science, evolutionary biology, metaphysics, the theory of knowledge, the philosophy of language and mind, moral and political philosophy, and even briefly, the ideal of human education.

Lecture 1 presents with clarity and precision, his own basic ideas in theoretical linguistics and cognitive science (both fields in which he has played an absolutely central founding role) recording the progress achieved over the years but recording much more strenuously how tentatively those claims to progress must be made and how a very large amount of work remains to be done even in the most fundamental areas of study. Changes of mind over these years are also recorded, some of the most striking of which occurred only in the last decade or so.

The lecture begins by motivating the question its title announces, “What is Language?” It behooves us to ask it because without being clear about what language is, not only will we not get the right answers to other questions about various specific aspects of language (perhaps cannot even correctly frame those specific questions), but because we won't get close to investigating or even plausibly speculating about the biological basis and evolutionary origins of language.

A tradition that goes back to Galileo and Descartes recognized the most fundamental feature of language that then got its most explicit articulation in Humboldt, which as cited by Chomsky is: “Language is quite peculiarly confronted by an unending and truly boundless domain, the essence of all that can be thought. It must therefore make infinite employment of finite means, and is able to do so, through the power which produces identity of language and thought.”[1] Darwin too is cited as repeating this in a more elementary form in the context of evolutionary concerns about language: “The lower animals differ from man solely in his almost infinitely larger power of associating together the most diversified sounds and ideas.” It is worth noting that there are three fundamental features observed here by Humboldt and Darwin. First the claim to an infinite power residing in a finite base, second the link of ideas with sound, and third the link of language with thought. All of them are gathered in what Chomsky declares at the outset as the ‘Basic Property' of language: “Each language provides an unbounded array of hierarchically structured expressions that receive interpretations at two interfaces, sensorimotor for externalization and conceptual-intentional for mental processes.” The hierarchical-structural element speaks to the first feature, the sensorimotor interface to the second feature and the other interface to the third feature.

What will account for this Basic Property is a computational procedure. The philosophical significance of this is two-fold: a theory of language is necessarily a generative grammar and, moreover, the theory is necessarily about an object that individual human beings possess, internal to the individual subject and its mentality (i.e., intensional elements). It is not a theory about externalized utterances, nor is it, therefore, about a social phenomenon. The nomenclature to capture this latter distinction between what is individual/internal/intensional and what is externalized/social is I-language and E-language respectively. It is I-languages, which alone can be the object of scientific study, not E-languages.[2] And though such study is eventually to be redeemed in a biological account, until that eventuality the science captures the phenomena at a level of abstraction from the biology and speaks at the cognitive level of the computational power that satisfies the Basic Property.[3]

A different, more general, task is to discover the shared underlying features of all I-languages, which is determined again by the biological properties with which human beings are endowed (a theme whose wider significance for cognition in general, is discussed again in Lecture 2). This more general task is undertaken with a view to discovering the biological endowment that determines what generative systems can serve as I-languages; in other words what are the possible human languages.

Chomsky, then, points out that as soon as the study of generative grammars addressing the Basic Property of language was seriously undertaken, some surprising puzzles emerged, with far-reaching implications. One is the “structure dependence” of linguistic operations: in all constructions, in all languages, these operations invariably rely on structural distance rather on the computationally far simpler notion of linear distance. Language learners know this automatically, without instruction. There is support for this from evidence from experimental neuroscience and psychology. The result follows from the assumption that the order is simply not available to the operations that generate the structured expressions that are interpreted at the conceptual-intentional interface, for thought and organization of action. That follows in turn from the very natural assumption that I-languages are generative systems based on the most elementary computational operation, which is order-free. These and numerous other considerations provide substantial evidence that linear order is ancillary to language, not involved in core syntax and semantics. The same is true of the various external arrangements of sign language, which is now known to be remarkably like spoken language in its structure, acquisition, use, and even neural representation. Presumably these external properties reflect conditions imposed by the sensorimotor system. The option of using linear order does not even arise for the language learner. Linear order and other arrangements are relevant to what is heard, i.e., externalized, not to what is thought, which is interior.

He then points out that these conclusions accord well with the little that is known about the origin of language. The sensorimotor system “appears to have been in place long before language emerged,” and there seems to be little specific adaptation for language. Cognitive properties of far deeper kinds than those possessed by apes, or presumably non-human hominins, are intrinsic to language. Apes have gestural systems adequate for signing and auditory systems adequate for perception of speech, but unlike human infants, they interpret speech just as noise, and even with extensive training cannot achieve even rudiments of human sign language. Aristotle had said that language is ‘sound with meaning', but these considerations just outlined, suggest to Chomsky that the priorities in the slogan might be reversed and language would be better understood as ‘meaning with sound'. In case this comes off as Platonist (something that was zealously propagated by Jerrold Katz), it must be kept firmly in mind that for Chomsky, ‘meaning' here is intended as a thoroughly psychological (eventually biological) category and thus not at all reified in Platonist terms.

Such conclusions, in turn, fuel Chomsky's longstanding claim that language is not to be understood as it everywhere is among philosophers, anthropologists, and others, as in some defining way tied to communication. If externalization of language is secondary, and the tie of language to thought is primary, then communication cannot be central to any answer to the question this lecture asks: What is language? Indeed, as he says, there is reason to think that most of language/thought is not externalized at all. If one firmly understands that language is not designed by human beings but part of their biological endowment, then as an object of study, whether scientific or philosophical, there might have to be considerable shift in our methodological approaches.

The quotation from Darwin that Chomsky cited with approval had it that what is fundamental about language is a ‘power of associating together the most diversified sounds and ideas.” Except for the fact that, as we have mentioned, sound (along with other modes of externalization) has been demoted, Chomsky's own theoretical account of the Basic Property takes this point in Darwin for its word –though perhaps not the exact word since ‘associating' isn't exactly right in describing the central operation that the account posits. Associating happens, after all, even in classical conditioning (bell, food) and Chomsky has famously repudiated behaviourist accounts of language. Moreover associations between two objects, as even non-behaviourist psychologists understand association, may imply that the order of the objects is important in a way that the far greater weight put on the forms suited for semantic interpretation at the conceptual/intentional interface (rather than the sensorimotor interface) establishes it is not. So moving away from Darwin's misleading word ‘associate' for what Darwin himself wants to say, what Chomsky has in mind rather is to make central that we are unique in possessing the capacity to ‘put together' ideas and syntactic elements. And this fundamental conception of language is echoed in the theoretical account of the Basic Property, in which the crucial operation is given the name MERGE, which can operate externally on two distinct objects to create another, or it can operate internally from within one object to create another, yielding automatically the ubiquitous property of “displacement” (phrases heard in one place but understood also in a different place) in the form appropriate for complex semantic interpretation.

These are called External and Internal MERGE respectively and respect for simplicity in scientific method, applicable in linguistics as anywhere else, dictates that we keep the basic operation down to this minimum and not proliferate operations in accounting for the computational power that grounds the Basic Property. Working through some examples to present how language design is at its optimal if we stick to this methodological injunction, Chomsky presents changes in his own view, such as on the phenomenon of ‘displacement' which he once saw as an ‘imperfection', but which now, if one correctly keeps to the simplest methodological assumptions as just mentioned, is something that is simply to be expected.

The lecture concludes with a bold attempt to exploit these last methodological points to bring two seemingly disparate questions together: what account shall we give of the Basic Property? and how and when did language emerge? This confluence of simplicity of assumptions in accounting for the Basic Property and the accompanying claim of the optimal design of language may help to give substance to what is the most plausible hypothesis on the limited evidence we possess about the origins of language, viz., that language emerged not gradually, but suddenly (and relatively recently). Such a sudden ‘great leap forward' it may now be speculated was perhaps caused by a ‘slight rewiring of the brain that yielded MERGE in its simplest form, providing the basis for unbounded and creative thought', hitherto unpossessed.

Lecture II consolidates some of these conclusions by first elaborating on another central theme in his work: the limits of human cognition.

There is a locution we have all used frequently: ‘the scope and limits of…. ‘. Chomsky takes it very seriously and gives it a crucial twist in elaborating his understanding of our cognitive abilities. These abilities, which in their scope are wider and deeper than those of any other creature we know, are so partly because they are also subject to limits, limits owing to our nature or, as the title suggests, the kinds of creatures we are—in particular, the fact that our cognitive abilities have a biological basis.

We had already implicitly come across this point in the first lecture, though it was restricted there to the human ability for language, in particular. The theoretical account of language presented there presupposed this notion of limits, i.e. presupposed that we are genetically endowed with innate structures that afford us our unique capacity for language, structures which at the same time constrain what language is for us, what possible I-languages there are. It is for the characterization of these innate structures that the technical term ‘UG' is intended; and it is within the framework of the scope and limits set by this genetic endowment, that language as a computational power is explained in the generative account summarized above.

What is true of language is just a special case of a perfectly general set of scopes and limits that come from the fact of being creatures with a biology. The idea seems to raise no controversy when it comes to physical ability: what makes us suited to walk limits us, so that we are not suited to slither like snakes.[4] Chomsky thinks that it is a prejudice to deny that what is obvious in the case of such physical abilities is not obvious (as the incessant controversies around innate ideas would suggest) in the case of cognitive abilities. To possess some cognitive abilities necessarily means that other cognitive abilities may be missing, cognitive abilities that other sorts of minded subjects could conceivably possess. It is only if we ignore the fact of our biology when we study human cognition that we would contrive to deny these limits. And the second lecture proceeds to look at the question of such limits on our cognitive abilities quite generally beyond the specific domain of language, though returning at various points to draw conclusions about language again.

It explores the methodological upshot of this idea of cognitive limits by first recalling a distinction Chomsky made almost five decades ago between ‘problems' and ‘mysteries'. Invoking, Peirce's understanding of scientific method and scientific growth that appeals to the concept of abduction which puts limits on what count as ‘admissible hypotheses', he argues that innate structures that are determined by our genetic endowment set limits to the questions that we can formulate. The questions we can tractably formulate are called ‘problems', but given the limits within which their formulation is so much as possible, there will be things that escape our cognitive powers and to the extent that we can even think them we will, given our current conceptual frameworks and knowledge, find ourselves unable to formulate them in a way that a tractable form of scientific inquiry of them can be pursued. These he calls ‘mysteries'. The omnibus title of the lectures “What kind of creatures are we?” is directly addressed by this since other sorts of creatures, with a different biological endowment from ours, may be able to formulate problems that remain mysteries for us. Thus, for Chomsky, if not for Peirce, (who, in speaking of admissible hypotheses, may have given less of a determining role to the fact of our being biological creatures[5]) the distinction between ‘problems' and ‘mysteries' is an organism-relative distinction.

It is a very important part of this methodological picture that we should learn to relax with the fact of our cognitive limits and the ‘mysteries' that they inevitably force us to acknowledge. The lecture, along with the companion essay in this volume, ‘The Mysteries of Nature”, traverses vital moments in the history of science to draw this methodological lesson.

One crux moment is when Newton overturned the contact-mechanical assumptions of the Early Modern science that preceded him and posited a notion of gravity that undermined the earlier notions of matter, motion, and causality which were scientific consolidations of our commonsense understanding (presumably determined by the cognitive limits of our biology) of the world of objects. He points out that with Newton a new framework emerged in which –by the lights of those limits—something inconceivable was being proposed. Newton himself admitted to this inconceivability, even calling it an absurdity, and nobody since Newton has done anything to redeem things on just this score. Rather the absurdity has simply been subsumed into our scientific picture of the world. Newton never let it deter him, constructing explanatory laws, ignoring the lack of a deeper underlying understanding that would, if we had it, make sense of what were, by these admissions on his (and others') part, described as an ‘occult' force. It was sufficient to construct intelligible theories of the world. And to do so, it was not necessary to find the world intelligible in the deeper sense that our cognitive limits frustrate.

Subsequent thinkers (Priestley, in particular, comes through as a most shrewd and comprehending commentator) made explicit this methodological outlook and drew consequences for issues in the philosophy of mind that vex philosophers today, but which, were they to take in what Priestley had to offer, might make them reconsider what they present as the mind-body problem or ‘the hard problem' of consciousness. Philosophers have a tendency to stamp some issue as uniquely ‘hard' and rest complacently in that frustrated register. Chomsky appeals to precisely this history to show first of all that there is nothing unique about finding something ‘hard' in just this way. Thus, for instance, what the introduction of ‘gravity' did in physics was conceived to be just as hard in the aftermath of Newton, including by Newton himself.[6] The significance of this to the so-called mind-body problem is that it puts into doubt whether it can any longer –since Newton– even be formulated coherently. The initially anxiety-inducing introduction of something ‘mysterious' like ‘gravity' eventually became essential to our understanding of material bodies and their acting upon each other without contact and so it simply got incorporated into science, indeed the new common sense of science. From this, we should, if anything, conclude philosophically that everything is immaterial, so nothing clear can remain of a mind-body problem. In a memorably eloquent reversal of Ryle's slogan, he says that far from the ghost having been sent to oblivion, the machine was discarded and the ghost remained intact. As for consciousness, the philosopher's tendency to require that much of our mentality be conscious, a tendency explicit in philosophers as different as Quine and Searle, is brought into question by looking at the operations of the rule-bound abilities of both language and vision. Chomsky feels particularly strongly about this since even much of our conscious thought interacts with aspects of mind that are hidden from consciousness, and so to restrict oneself to what is conscious would hinder a scientific understanding of even the conscious mind.

Given his concern with a scientific account, he is concerned too to show that some ways of thinking about language, and thought more broadly, are not scientifically sound. There is, in particular, an extended discussion of the atomic elements of computation. Invoking points established in the previous lecture, he points out that these are misleadingly described as ‘words' and as ‘lexical' items in the literature because –as they feed into the conceptual-intentional interface, which has been shown to be primary in contrast with the sensorimotor interface—they are not constructed by the processes of externalization. Even more startling for philosophers is the claim that, except for some explicitly stipulative exceptions in mathematics and the sciences, they don't have any referential properties and are not to be thought of as bearing constitutive relations to mind-independent objects in the external world. I-language, which is the only scientifically accountable notion of language, thus, is thoroughly internal. This point is explored via a discussion of historical views, such as those of Aristotle and Hume, and via a discussion of examples of such atoms, ranging from the relatively concrete such as ‘house' and ‘Paris' to relatively abstract such as ‘person' and ‘thing'. Reference or denotation is shown by these discussions to be too contextual to bear scientific study and should be seen as relevant to the use that language is put to rather than a constitutive aspect of language itself. All this leads to a different taxonomy than is found among philosophers, relegating almost all of what they have in mind by ‘semantics' to pragmatics.

These conclusions are relevant to the question of the origin of language. Animal signals to each other are caused by direct links they have to objects in the external world. There is no understanding them if one left these causal links out, whereas, the burden of the discussion above was to show precisely that there are no such constitutive causal links to a mind-independent reality for the atoms of human computation. This gives further reason to conclude that the kind of creatures we are, possessed of the kind of powers for language and thought we possess, should get an evolutionary account of the sort presented in the first Lecture rather than what Chomsky, citing Lewontin in this lecture, describes as the ‘storytelling' about gradual evolution from our creaturely ancestors, a mode of explanation that one would only indulge in if one does not pay enough prior and scientific attention to the nature of the phenotype being explained. It is storytelling partly also, as Lewontin is cited as saying, because of the ‘tough luck' of not having access to any evidence on which these explanations could be based. They are hidden from human cognitive access, another form of our limitation.

Thus limits on our cognition are inevitable for a variety of reasons, chief among which is the taking seriously of the sheer fact that we are biological creatures. Unlike Locke, Priestley, Hume, Russell, Peirce, and Lewontin, who are among the heroes of this lecture, Hilbert most explicitly (“There are absolutely no unsolvable problems”) and much of contemporary philosophy more implicitly deny that there are mysteries, thereby denying a truism based on this sheer fact. What is fascinating is that Chomsky, having presented all this, takes an interesting combination of attitudes towards it. On the one hand, the very idea of cognitive limits that lands us human beings with ‘mysteries', which other sorts of subjects may find perfectly tractable, is a commitment to what philosophers call a realist metaphysics. As he says, “Given mysterian truisms, what is inconceivable for me, is no criterion for what does not exist”. But on the other hand, taking his cue from Newton, his attitude, once this is acknowledged, is thoroughlypragmatist. Just because what we study, the world, may not be ultimately intelligible, does not mean that we should be inhibited from striving to produce intelligible scientific theories of the world. Even the concept of free human action, he says, which may go beyond any of the concepts we possess (crucially, determinacy and randomness) might one day, he says, be scientifically tractable, though we are far from anything like that understanding at present. This is quite different from the attitude of Kant, who declared freedom to be thinkable but never knowable. Like Peirce and before him Newton, and unlike Kant, Chomsky does not want his own mysterianism and his own insistence on the limits of our cognitive powers, to place, as Peirce once put it, “roadblocks on the path to knowledge”.

Lecture III lifts the restriction on our natures, considered in terms of individual capacities (for language and cognition), and considers us as social creatures, seeking to explore what is the common good and which political and economic arrangements promote or thwart it.

The Enlightenment figures large in the pursuit of these questions, though what Chomsky has in mind by the Enlightenment is capacious including the familiar ‘liberal' figures of Adam Smith[7] and Mill as well as those in a broadly Romantic tradition such as Humboldt and Marx. And its interpretation is capacious too, stressing not only the side of Adam Smith that is often suppressed by most of his liberal and radical critics as well as his conservative devotees, but also stressing principles that allow it to be seen as a precursor of a later anarchist tradition in Europe as well as John Dewey in America.

The starting point of these inquiries is in fact individualist and has ties to the earlier lectures. Even within their biologically determined limits, the creative capacities that each individual possesses (and which were discussed in the first lecture in the specific domain of language) are precisely the sort of thing whose full development makes individuals flower as subjects. The social question of the common good necessarily comes in when one asks what sorts of institutions hinder such development within the individual. Social frameworks such as capitalism that stress self-interest hinder rather than encourage the development of individual capacities. Adam Smith's vivid excoriations of what the division of labour does to destroy our creative individuality and Dewey's harsh words on the shadow cast by corporate interests on just about every aspect of public and personal life are both invoked to establish this. The tradition of anarchism (from Bakunin to Rocker and the anarcho-syndicalism of the Spanish Civil war period) combines socialist ideas with the liberal principles of the classical Enlightenment to construct an ideal –of cooperative labour, workers' control of the workplace and the means of production, and a social life revolving around voluntary associations– that, if implemented, would sweep away the obstacles to the goal of human development which come from both free market capitalism and Bolshevik tendencies to a ‘red bureaucracy'. Dewey's ideas on education reveal how by contrast with much of the contemporary practice found in educational institutions, the goal of human development can best be pursued from an early age.

There are touching descriptions of how many of these ideals were central to the activism of a wide range of grass roots movements from the early radical parliamentary tradition in seventeenth century England to the ‘factory girls' and artisans that Norman Ware wrote of in his study of the industrial workers in the American tradition to the liberation theologians in the Catholic tradition of Central America. These longstanding democratic labour traditions are contrasted in some detail with a different understanding of democracy in a tradition that begins in America with Madison's ‘aristocratic' strictures on who may govern and updated in the vision of Walter Lippmann's ideas of democratic rule by the ‘expert', the American version of Leninist vanguardism, ensuring –as Chomsky makes clear with a glance at the results of polls on various important issues such as healthcare– that what the people want is almost never what gets on the agenda of ‘democratic' politics. This latter understanding of democracy, of course, dominates the practice of societies and governments in much of the Western world and Chomsky is keen to point out that even at its worst, it never lets up on the claimto be pursuing high sounding ideals of the common good, showing how the common good is universal in a quite paradoxical way: it is preached as applying to all, even as it is everywhere violated by those said to be representing all and who mostly pursue the interests of a few.

Given the fundamental starting point in human creativity and the importance of its unhindered flowering, Chomsky's leaning towards anarchism is not surprising and his way of putting the point has always been to ask, as he does in this lecture again: any form of coercion that hinders it can never be taken for granted. It needs a justification. All arrangements that have coercive power, including centrally the state, must always be justified. The default position is that they are not justified –until and unless they are. And given the contingency of the ‘shoals of capitalism' (his phrase) in all corners of the world, there is indeed a justification of a notion of the state that protects the vast numbers who are pushed to the margins of society[8] (echoing Adam Smith himself who thought only the state could alleviate the oppressive life that industrial capital forces upon labour), very different from the actual state in most societies which, as Dewey is cited as saying, largely do the bidding of corporations and in doing so remove the socialist element from anarchism and allow only the libertarian element –as a result of which democracy becomes ‘neo-democracy' (to match ‘neo-liberalism') in which if one suffers in poverty it is because, as Hobbes might have put it, one has chosen to do so. Thus to turn one's back on this and to justify the state as offering protections for those who suffer under capitalism, far from contradicting anarchism, is a consistent application of its principles in historical contingencies, a point that Chomsky presents with a marvelous metaphor which he says he has borrowed from the Brazilian rural workers' movement and extended –the metaphor of an ‘iron cage', whose floors one tries to extend as one tries to reduce the coercive power of the state, even as the cage protects one from the destructive forces outside the cage, forces which render us weak and impoverished and alienated, to say nothing of rendering our planet uninhabitable.

I have tried, as best I can, to summarize a book whose intellectual complexity and power and whose breadth of knowledge and originality cannot possibly be captured in a summary –so, an exercise and duty that may not, in the end, aid the reader at all. But what I will say, without pause or condition, is that there was such pleasure and instruction in the exercise that I could do no better than ask the reader to study the book for herself –not only for the qualities I have just mentioned, but for its utter seriousness of purpose regarding the deepest questions in philosophy and science and, above all, its vast humanity.



[1] (For all references, see the lectures that follow from which the quotations are taken.) On the relation between language and thought, Chomsky, though he now thinks it to be even closer than he himself once did, does not think it is necessary to assert something as strong as ‘identity' between them, as Humboldt in this cited passage, does. Descartes and Darwin who also figure in his discussion of the relation did not go that far.

[2] Though Chomsky mentions E-languages by way of contrast with I-languages, he doubts the coherence of the very idea and therefore whether they exist. In a number of his essays, he is critical of the most basic assumptions philosophers make about their coherence, in giving accounts of them.

[3] In making this point about study at a level of abstraction with a view to an eventual account in terms of the brain, he points out how the approach is no different in the scientific study of language than it is, for instance, in insect navigation. In other work, Chomsky cites some progress that might have been made in the inquiry into biological underpinnings, but also cites how there may also be some fundamentally wrong assumptions being made by brain scientists about what the object of study is. On this last point, see his reference to Gallistel's work in Lecture 2.

[4] I owe this example to Carol Rovane. See Carol Rovane and Akeel Bilgrami, “Mind, Language, and the Limits of Inquiry” in James McGilvray (ed), The Cambridge Companion to Chomsky, Cambridge University Press, 2005

[5] This should be qualified by pointing out that Chomsky, at the end of this lecture, actually discusses an argument in Peirce that appeals to biological considerations, in particular evolutionary considerations based on natural selection (which he finds completely fallacious). This would suggest that Peirce was himself somewhat ambivalent about whether or not to see his overall methodological claim regarding admissible hypotheses and limits on them as owing to our biology.

[6] And before Newton, motion was considered to be ‘the hard problem' by William Petty and others.

[7] Chomsky was the first to stress this side of Adam Smith many decades ago, a side of him that has been pursued in some detail much more recently in scholarship by Emma Rothschild and commentary by Amartya Sen.

[8] One might add that there are issues on which the state can be justified because it might protect not just the marginalized and impoverished but everyone from their folly and doom, issues such as those of the environment, for instance, and more generally protect citizens from the cultural detritus and psychological desolation (issues of ‘alienation', in a word) that afflict capitalist societies.

* * *

Akeel Bilgrami is a philosopher of language and of mind, and the author of Belief and Meaning, Self-Knowledge and Resentment, and Politics and the Moral Psychology of Identity (forthcoming), as well as various articles in Philosophy of Mind as well as in Political and Moral Psychology. Some of his articles in these latter subjects speak to issues of current politics in their relation to broader social and cultural issues. He has two upcoming books, “What is a Muslim?” and “Gandhi the Philosopher”. Bilgrami is currently the Sidney Morgenbesser Professor of Philosophy at Columbia University in New York.

Like what you're reading? Don't keep it to yourself!
Share on Facebook
Facebook
Tweet about this on Twitter
Twitter
Share on Reddit
Reddit
Share on LinkedIn
Linkedin
Email this to someone
email