Cartesian Linguistics

by Noam Chomsky

  What of appropriateness? If language production were caused by the current circumstances of the speaker, no issue of appropriateness of linguistic actions could arise. One could, of course, ask if a causally determined ‘cognitive’ system that presumably evolved yields outputs that increase an organism’s chances of survival, or something like that. But thinking of the appropriateness of sentences produced by people (out loud or in thought) in this way indicates complete misunderstanding of the issue. Appropriateness of linguistic action poses a problem because human linguistic actions are uncaused and innovative. Humans must choose ‘what to say’ about some discourse circumstance and some task to which language contributes, and the choices are always open ones. It is wrong to think of the choice as solely one between telling the truth and telling a lie – changing, say, a sentence’s ‘is’ into an ‘is not’. That kind of choice could arise only where one is asked to (say) describe something, not (perhaps) complain to customer service about the amount of time they required you wait on the phone before they responded, or (perhaps) construct a fable or a parable to get a view across to an audience. And even if it were just a choice between truth and lie in carrying out a task of describing, unbounded ways to tell the truth are available, and again unbounded ways to lie. Given all this, the most one is likely to be able to do by way of showing how appropriateness is possible is to note that, first, judgments of appropriateness surely require drawing on resources that go well beyond the language faculty alone. They involve multiple other systems in the head. Second, one must keep in mind that in dealing with multiple modular systems ‘communicating’ with one another over their ‘interfaces’, it would hardly be surprising to find that dealing with the total output of a complex of systems is well beyond the scope of any science that human being seem to be able to construct. For dealing with the output of multiple systems operating together would require a way to deal with massive interaction effects. Since cosmologists have a hard time producing a theory that will determine the states at any given time of three point-mass ‘bodies’ moving with respect to each other, and find four or more completely out of range, we should hardly be surprised that theories of complex interactions between multiple mental systems are likely to remain out of the reach of the sciences now, and for any future where we humans have the cognitive resources we now have. Judgments of the appropriateness of what someone else says can require mobilizing all of one’s own mental resources. We manage quite well, to be sure. But we manage not because we are all well-trained members of a linguistic community, nor (obviously) because we are omniscient gods, nor (obviously again) because we have deterministic theories, but because we and the speaker are all organic creatures that have the cognitive resources humans do, and they use all that they have available and need to employ in ‘interpreting’ and understanding (as a person) what another says.

  Focusing again on Descartes, his observations apparently led him to abandon any serious effort to construct a science of mind. He tried to explain the apparent limitations on science’s capacity to deal with creativity not by appeal to the points made above, points that presuppose that sciences of the mind exist that meet the methodological requirements on science that he helped introduce, but by appeal to the idea that mind is just a different substance, Mind. In doing so he effectively placed the mind out of the reach of the most all-encompassing science he had managed to construct, his contact mechanics. Assuming as he did that the subject matter of that science is passive ‘body’, assuming further that his contact mechanics served as something like a theory of everything (he thought of his neurophysiology, for example, as a special case of it), and confronted as he was by the obviously non-passive creativity exhibited by human minds, ‘explaining’ creativity by introducing a non-body substance had the effect of putting the mind out of reach of science as constructed until then (but see note 24, above).

  Descartes himself made no obvious errors, given what was known at the time. The errors, it turned out, lay in his contact mechanics, but they were not to be revealed until a half-century later. As Chomsky points out in several places (Chomsky 1988a, 1996, and 2000), a crucial error was revealed by Newton when he pointed out that contact (and thus a contact mechanics) cannot explain gravitation.28 Describing and explaining gravitation requires postulating a force that is completely mysterious to someone wedded to a contact mechanics: ‘action at a distance’. Descartes’s contact mechanics failed, and along with its failure went the assumption that all physical action is by contact. His contact mechanics cannot serve as a science of everything. In fact, it cannot deal with elementary features of the physical world. Had he known, he might have come to the conclusion that there isn’t such a barrier between mind and ‘physical reality’ as he thought there was. For if ‘physical reality’ (the reality physics deals with) is not ‘body’ as he originally conceived it within a contact mechanics, one has to allow for what from the point of view of contact mechanics seem mysterious forces. If, moreover, these forces can only be understood by appeal to a formal mathematical theory, nature begins to look like something that we can no longer use common-sense intuition to understand. Science and its formal tools offer a better way, however unintuitive what they tell us might seem. If so, we should take seriously the possibility that our intuitions about ‘body’, those built into the ‘mind/body problem’ as understood by Descartes and since, are unreliable. And we should allow that ‘body’ for the sciences does, after all, countenance mental properties and events. The apparent barrier Descartes’s intuitive but wrong conception of body posed to constructing sciences of the mind disappears. We can begin to construct ‘physical’ sciences of mind that, methodologically speaking, meet all the conditions on natural scientific investigation. Descartes did not take this course, and for understandable reasons, given his assumptions.

  Unfortunately, even many philosophers and others after Newton have missed the implications of Newton’s discovery. They seem not to have noticed that there is no block to constructing ‘physical’ sciences of the mind. A block exists only if Descartes’s contact mechanics is correct, and its understanding of ‘body’ taken seriously. If the scientist must abandon Descartes’s conception of body and ‘physical theory’ ends up – as it does – postulating all sorts of forces and ‘entities’ that are mysteries for Descartes’s conception of ‘the physical,’ one might as well say that any science that meets the requirements of successful naturalistic science is a ‘physical’ science too. Following this line of thought to its conclusion, Chomsky (following Locke, Priestley, and others before them), often points out that there is no mind/body problem (1988a).29 There cannot be until some sciences of the mind and the physical world that meet the requirements on being good sciences show that their subject matters really are incompatible, as they were for Descartes. Until then, perhaps philosophers who are interested in a real problem as opposed to a Wittgensteinian Scheinstreit should abandon the so-called “mind/body problem,” for ever since Newton’s time, it has not existed.

  The problem creativity poses for the science of mind remains, of course. But as we have seen, it is dealt with by adopting an internalist and nativist strategy that recognizes that we have limited cognitive capacities, and by noting that any action – including speech – is the result of a massive interaction effect. We can – and in this case do – make sense of how some aspects of creativity are possible. But we certainly do not abandon the science of mind entirely, nor adopt a close analogue to abandonment, the methodological dualism favored by empiricists.

  III.3 A computational theory of mind

  One can understand why Descartes put the mind out of reach of the science of nature. But it put naturalistic research in a narrower and – surely – stifling box. Among other things, it left Descartes with no plausible account of how ideas could be innate. He noted in Comments on a Certain Broadsheet (CSM I, 303–4) that the innateness of ideas is like the disposition to develop certain diseases that one finds in some families. A disposi
tion to develop diseases demands scientific explanation by appeal to the processes of reproduction and inheritance. But placing virtually universal (across the human species) ideas and other mental conditions (sensory colors, sounds. . .) in the mind ruled out that ‘physical’ or naturalistic way of accounting for their universality and early acquisition. Calling on God – as Descartes sometimes did – to explain innateness in this domain is no help. He would have benefited from placing the science of mind in the domain of natural scientific research that he helped invent.

  Arguably, he was on the brink of constructing such a theory for the sensory qualities of vision – that is, for the characteristic ways in which our minds configure our visual experience in the ways familiar to anyone with vision (color and spatial volume at a moment – a retinocentric space). He argued – in part on poverty, in part on non-resemblance,30 in part on uniformity across the species grounds – that the sensory qualities of vision are both mental and innate.31 Given that he assumed that these are innate and universal, even though mental, he could and perhaps should have come to recognize that, as with disease, there must be some naturalistic explanation for how infants come to sense visually in the ways all humans do. There is, of course, no guarantee that naturalistic scientific research into a domain will lead to success, but poverty and uniformity observations surely offer reasonable grounds to assume that it will.

  Assuming innateness and mentality (‘being in the mind’), the work of poverty, universality, and non-resemblance observations is done. Thereafter, one proceeds in the way all scientific inquiry does: observe phenomena apparently within or relevant to the subject matter you are investigating, note differences, construct formal-mathematical descriptors of postulated entities and events, aim toward precision, put together hypotheses, aim to explain puzzles and anomalies, develop causal principles, detail algorithms, and the like. To see what Descartes had accomplished, and to point to what he could have reached for, I appeal to contemporary work in constructing a computational theory of vision – its aims and the tools it employs. There are several tasks for a theory of vision, but a central one is accounting for how ‘inputs’ in the form of firing rates of various cones in an array in the retina (ignoring rods, in order to focus on colored-shaped visual spaces), reacting to impacts of photons, are ‘processed’ in the various parts of the visual system (retina, lateral geniculate nucleus, and various brain visual areas) to yield ‘what one sees’ at a moment. The procedure – rather a complicated one32 – is captured by appeal to mathematical descriptions of the operations that the various parts of the visual system – descriptions that indicate how rather messy input is ‘processed’ to yield a colored volume with greatest resolution in the center (due to the fovea’s greater concentration of cones). Idealizing, one ends up with a volume/space where spatial positions are assigned by means of altitude, azimuth, and depth coordinates fixed on a momentary line of sight (which makes the volume retinocentric), and where each distinguishable point in the volume is assigned ‘color’ values of hue, brightness, and saturation. In effect, one ends up with a map of the finest color–space discriminations of which the human visual system is capable at a moment. This map can be used to describe ‘what is seen at a moment’ in color and spatial terms, for any normal human organism. All that is required is a specification of the ‘input’ values for the retinal array. The theory provides a much finer-grained description of the colors and (retinocentric) spatial positions than does anything like “red surface there now.” Generally, the science of vision can produce much better descriptions of qualia – rather, of possible total qualia spaces at a moment – than can the color and position terms of natural languages. It does not, of course, describe colors and experience of space “from the inside” – whatever that might mean. But given what it can do, that surely does not matter, and for any serious work on “what one can see,” it does – or aims to, when complete – do as well as humans are likely to be able to do.

  Chomsky’s linguistics has much the same aim: given an I-language specification of a person’s lexicon at a time,33 plus parameter settings, it is possible (in principle) to specify the set of linguistically expressible “perspectives” that that I-language has to offer. That is, it is possible to map “what can be meant” (in the linguistic technical sense of “mean” – i.e., possible SEMs. Naturally, that is a long way off, but it does indicate what the ultimate aim of a computational theory in another faculty aims towards. There is an important difference between SEM-mappings and retinocentric space color-volume mappings: in a way, the visual mapping does capture possible experiences, while SEM-mappings remain out of touch of consciousness. Both represent, however an ‘output’ level of a system, and the theory of the system says what they can be.

  It is certainly no fault of Descartes that he did not offer anything like a complete computational theory of vision – any more than that it was his fault that he did not anticipate Newton. But he did, I think, begin to move in what we now think of as the right direction. For one thing, he realized that vision depended on processing that can and must be represented mathematically; he was on the way to a computational theory. Second, on the determination of visual depth, he made a crucial observation: the visual system utilizes ocular convergence measurements (it utilizes much more, but this is a central contribution). He came to this view, in part, by noting that the blind can tell the distance from them of some object – that is, judge depth – by using a couple of sticks that they hold in front of them that touch at their tips. They – or, surely, their minds – calculate the distance from the angle that their hands and converging sticks subtend. Descartes also noticed that the eyes converge more when a person looks at a nearby object and less when it is further away. Noting the parallel between what the blind do and what the eyes ‘do,’ he drew the conclusion that “we perceive [depth] by means of our eyes exactly as we do by means of our hands” (Optics in CSM I: 169). In modern terms, he realized that the vision faculty computes depth by (in part) measuring the angle of convergence of the eyeballs at which their retinocentric images fix on a single focal point. The result is sensed depth, note, just as the colors of vision are sensed colors (where the blind cannot exploit the resources the retina provides). We humans use these sensed properties, not the properties of sensory organs (or hands, in the case of the blind), in order to navigate and identify objects. And yet these sensory features – as Descartes says – are features of the mind, not the body. It does seem, then, as if Descartes was launched on the way to producing a mathematical theory of vision. A completed theory could be claimed to actually tell us “what it is like to experience visually as a human being,” for it offers as fine a discriminative capacity as is available. Granted, it does not ‘give’ us what we ‘get’ when our system is operating and we happen to be looking at a field of flowers. But that is not its job. This has some bearing on the discussion in note 29.

  It is not clear why Descartes failed to note that by focusing on formal, mathematical descriptions of operations (as with depth), he had abandoned what he had to say about the body (“movements”) and its resources alone, and that he was already offering a computational theory of mind – of how the mind ‘computes’ the various visual sensations of which it is capable.34 If he had noted this, he would have had to acknowledge that a science of the relevant brain states and operations – those detailed by the formal mathematical theory of the vision faculty – should not take the form of a description of a hydraulic system with various valves and gatings (what we might now call an “implementational” description), which is how he described neural systems and their operations. That says nothing about what these items ‘do’ – what they produce or yield to the organism, and how. Rather, it can – and should – be a mathematical description of the operations that these brain states and events carry out and the mental sensory states they can assume, as Newton’s inverse square law is a mathematical description of gravitation and the relevant states of the interacting ‘bodi
es’ (point masses).

  IV Cartesian Linguistics: education and politics

  Although not in detail in CL, Chomsky does not shy away from pointing to empiricism’s implications for human nature, and at the attractions of this conception of human nature to those in power – those in government and (currently) especially those in corporations – and to the many ‘managers’ (including the majority of intellectuals) who aid those in power.35 Empiricism commits itself to the idea that central aspects of our minds and cognitive powers – our languages and our concepts – are lodged in a plastic part of the mind/brain. They commit themselves to the idea that much of what makes us human and distinct from other creatures is due to training and acculturation. Given this, he is blunt: those in positions of power who want to maintain their positions and the authority it gives them – the capacity to make decisions for others, and in doing so to serve their own interests – are attracted to the idea that those over whom they exercise their power are with regard to many of their essential features moldable clay, and must (and should) be molded in their best interests. Not surprisingly, the best interests of those to be molded always seem to turn out to be the interests of those in power. By believing this kind of immoral nonsense and the self-deception it requires, those in power try to justify themselves and their actions, at least to themselves and their friends.