It is well-known that excellent critics need not be excellent artists, and vice versa. At least in some cases, it seems, an analogous divergence exists between the writing and reviewing of academic books. Fodor is without question an outstanding reviewer – clear, perceptive, incisive, witty and suitably outrageous given a decent opportunity. At the same time, his own books have markedly declined in quality. That at any rate is the conclusion one must draw from comparing LOT 2 to Fodor's seminal (if, in my view, mistaken) The Language of Thought of 1975. The divergence is no coincidence. Having a jolly good rant against a single target is often a boon when it comes to presenting and evaluating the work of others within relatively strict word-limits. But it can rarely sustain a whole book in a rewarding fashion. And it certainly cannot sustain a series of three books (plus a raft of articles). LOT 2 follows Concepts: where Cognitive Science went wrong (1998) and Hume Variations (2003) in its single-minded attack on what Fodor calls ‘pragmatism’. By this he means any position that links concepts and concept-possession (the capacity for conceptual thought) to inferential rules and abilities rather than to inner goings-on. Displaying his trademark penchant for subtlety and understatement, Fodor pronounces pragmatism to be ‘perhaps the worst idea that philosophy ever had; there will be much more of that as we go along’ (9). No kidding! Fodor seems to think that he can make his claims and arguments more compelling by using derogatory terms for his targets (‘monsters’, ‘disaster’), peppering them with ‘obviouslies’, ‘surelies’ etc., and by repeating them ad nauseam. At one point he inflicts the ultimate indignity on his readers by repeating a clumsy diatribe in CAPITAL LETTERS (39). Add to that incessant interludes, mainly dialogues between the author and a figure called ‘snark’. These are obviously meant to be humorous, yet they are in fact laboured and self-indulgent. Worse still, they distract from whatever line of argument one might otherwise discern.

As regards content, the book is very uneven. Like many of his compatriots, Fodor fancies himself as a hobby biologist who can refute contemporary evolutionary theory, principally by showing that it is at odds with his own articles of faith. Thus he takes Patricia Churchland to task for writing ‘looked at from an evolutionary point of view, the principle function of nervous systems is to get the body parts where they should be in order that the organism may survive’. To this, Fodor reacts by maintaining that ‘nothing at all is known about how cognition evolved’, proudly flaunting his ignorance of paleontology and cognitive archaeology (8–9 + n). He goes on to complain: ‘Pace Churchland, it simply couldn't be true that the Ur-form of thinking is thinking what to do or where to go’ (13). It is unclear what, if anything, Fodor means by ‘the Ur-form of thinking’. But it couldn't have any bearing on Churchland's claim about the evolutionary function of nervous systems. Perhaps Fodor has a different theory about the selective advantage conferred by nervous systems (sustaining buoyancy in water?!). More probably, like his hero Chomsky, he repudiates the Darwinian theory of natural selection altogether. Such ideological aberrations notwithstanding, however, Churchland's statement is simply a truism of contemporary biology. And that discipline can lay a substantially more secure claim to scientific respectability than the axe-grinding speculation that Fodor practices under the label ‘cognitive science’.

While philosophical accounts of the mind must not contradict biological knowledge, they cannot simply be derived from it. And as a philosopher, Fodor is a heavy-weight, especially when it comes to being a fly in the ointment. Just like some theist philosophers detect arguments in favour of God's existence around every corner, Fodor detects arguments against pragmatism everywhere along his chosen path. Another problem is that the forces of darkness that Fodor labels pragmatists are a very motley crew indeed, including diverse and often incompatible views about truth, meaning and concept-possession. Nor is Fodor excessively fussed about whether anyone has ever held the views he attacks. Nevertheless, some of his targets are genuine, and some of his arguments are pertinent, ingenious and pursued in a resourceful fashion. Thus Fodor is right to point out that ‘acting on plans’ cannot simply precede the ability ‘to think about the world’ (13). In so far as the former involves the capacity for practical reasoning, it clearly requires the capacity to have beliefs. But Fodor begs the question against his opponents in assuming that the capacity ‘to think about the goal of a plan’ is independent of the capacity ‘to think the plan’, i.e. ‘to think of [the goal] as the goal of the plan’. Even the former must amount to the capacity of thinking about a possible state of affairs (whether it be a yet unexecuted action or it as yet unachieved goal), without holding it to be actual. That is a very advanced capacity, which goes way beyond the capacity to believe something to be the case. As far as temporal priorities are concerned, the biological genesis of such a capacity can only be explained in the context of an emergent capacity to devise and execute plans for action. For there is no evolutionary premium on the mere entertaining of possibilities. As far as conceptual priorities are concerned, it is unclear what sense it could make to ascribe thinking about possibilities to creatures that are incapable of practical reasoning, and to ascribe that latter capacity to creatures incapable of acting for reasons. Fodor's objections ignore a possibility that is central to his pragmatist target: the proper role of our idiom of propositional attitudes and practical reasoning lies in describing and explaining complex behavioural capacities; the applicability of the former and the capacity for the latter are two sides of the same coin; and that coin was gradually minted for the need to predict and explain increasingly flexible and intelligent behaviour.

The same difficulty bedevils his other circularity arguments against pragmatism. Contrary to Fodor, the latter does not tacitly take conceptual thought for granted. Instead, it contrasts fully-fledged conceptual thought – intentional states like beliefs (judgements) or desires – with qualitatively distinct more basic cognitive (especially perceptual) and conative phenomena. And it maintains that the difference lies in the complexity of the behavioural capacities of the subject. As a result it has no difficulties granting that there are intermediate stages between the total absence of mental phenomena and full-fledged thought and rationality. Fodor, by contrast, simply takes for granted that full-blown representation must somehow have been present all along, and that the apparent stages of onto- and phylo-genesis are merely qualitative extensions of pre-existing mental capacities that are completely independent of behvioral capacities.

It hardly comes as a surprise, therefore, that LOT 2 not only reverts to a strong nativism that Fodor had abandoned in the meantime, but adopts a position even more radical than that of LOT 1. It isn't just that there are empirical reasons for thinking that primitive concepts are innate, ‘it's true and a priori that the whole notion of concept learning is confused. Punkt’ (130). The argument is based on the assumption that the only way in which concept learning could proceed is by induction: learning a concept C ‘is coming to believe that C applies to cs. But beliefs are constructs out of concepts, not the other way around’ (139). Yet this is precisely what Fodor's opponents deny. They regard concepts as abstractions from beliefs. More precisely, the idea is something like this. Imputing concepts is something we do when perceptual thought and the activity which it guides become so complex that discrimination becomes deliberate enough to qualify as classification and when planning becomes so complex that inferences and hence inferential relations have to be imputed. Even if Fodor's argument weren't question-begging, it wouldn't be valid. As he himself acknowledges, concept learning as he describes it isn't the only way in which concepts could be acquired. Fodor holds that it is the only cognitive way of acquiring concepts, but fails to justify that claim. And even if he were right on this point, he would not entitled to state that even a concept like CARBURATOR is innate to us in the same way as having 10 fingers (146). Indeed, there is a good reason for rejecting that statement. Fodor himself grants that it is sufficient and may well also be necessary for having the concept C that the subject be ‘able to think of something as (a) C’. Obviously, however, in our case that ability is not innate in any sense whatever: it is neither present at birth nor automatically acquired irrespective of input. Without experience, indeed, without linguistic instruction, humans cannot even discriminate carburators, let alone think of them as carburators.

The aforequoted passage ending with Punkt is one of several in which Fodor makes claims that he himself occasionally describes as a priori or conceptual. Yet at the same time unsubstantiated denials of the feasibility of conceptual analysis, etc., are a kind of background noise in the book. When the issue cannot be circumvented, he insists on his claims being ‘metaphysical’ or ‘ontological’, yet without explaining what that means (14). It cannot mean that his claims are borne out by empirical science, not just because that wouldn't be metaphysics, but also because Fodor's whole approach tends to be at loggerheads with plausible mainstream opinion in the relevant hard sciences – notably neurophysiology and evolutionary biology.

LOT 2 features important accounts of the development of Fodor's thought as well as interesting claims (both negative and constructive) about various issues in cognitive science. But it does not furnish an argument against pragmatism that is remotely compelling. And this is where my own anti-anti-pragmatist rant ends.