John McDowell’s Mind and World (1996) has become a chief point of reference not only for present-day revisionist Kantians of a moderately naturalising bent but also for those in the wider Anglophone philosophical community seeking a new way forward from the problems that beset their analytic forebears.[1] Chief among them is a version – with sundry variants – of the eminently Kantian problem as to how intuitions are 'brought under' concepts, or how we can attain knowledge of the world through the mind’s capacity to synthesise the data of passive sensory uptake with its own actively shaping or knowledge-constitutive power. From Quine to Davidson and Rorty, they have typically sought to defuse this problem by rejecting any version of the Kantian dualism between scheme and content, analytic and synthetic judgements, or 'truths of reason' and 'matters of fact'. This they have done most often in the name of a radical empiricism which treats such distinctions as merely the product of an old (presumptively discredited) attachment to 'metaphysical' ideas about meaning, knowledge, and truth.
Hence Quine's argument for doing away with the two central 'dogmas' of Carnap-style logical empiricism, namely (1) the analytic/synthetic dualism and (2), the closely related belief that statements can be checked off one-by-one for their truth as concerns real-world (observable) states of affairs or their consistency with the supposed logical ground-rules or a priori 'laws of thought'.[2] Rather we should think in holistic terms of the totality of knowledge at any given time as a man-made 'fabric' extending all the way from empirical observations at the 'periphery' to putative logical truths-of-reason at the centre, but with all such items – logic included – ultimately open to revision should this be required, as for instance by some startling new development in the physical sciences. However, as Davidson then pointed out, this left Quine in the grip of yet another such dogma, i.e., a relativised version of the scheme/content dualism whereby the truth-value of statements was thought of as dependent on their role within this or that currently favoured ontological framework, belief-system, conceptual scheme, or whatever.[3] Rorty went a good way further along this road by enlisting Davidson as an ally in the drive to rid philosophy of all such residual Kantian elements by combining a purely causal or physicalist theory of cognitive input - 'the pressure of light waves on Galileo's eyeball, or the stone on Dr. Johnson's boot' - with a thoroughly holistic theory of belief wherein such inputs exerted no constraint on the range and variety of possible interpretations brought to bear by diverse cultural communities.[4] Yet of course (as McDowell rightly remarks) this amounts to just another, more drastic version of the same old Kantian dualism. Thus Kant's great problem in the First Critique – that of explaining how sensuous intuitions are 'brought under' adequate concepts – is now pushed to a point where the causal and rational components of knowledge-acquisition are conceived as existing in realms quite apart, such that any attempt to unite them must amount to a species of downright category-mistake.[5]
McDowell finds this argument wholly unsatisfactory, 'both in itself and as a reading of Davidson' (p. 148). It requires that the two perspectives in question – beliefs as 'causal interactions with the environment' and beliefs interpreted 'from the point of view of the earnest seeker after truth' – should be prevented from getting into conflict by treating them as different (incommensurable) language-games or ways of interpreting the mind/world relation. That is to say, any putative causal 'input' must be thought of as impinging on our sensory apparatus at such a basic level that it exerts absolutely no constraint upon the higher-level process of belief-formation or the kinds of adjustment we make in order to achieve the maximal coherence with the rest of our standing beliefs and ontological commitments. For Rorty, indeed, 'there seems no obvious reason why the progress of the language-game we are playing should have anything in particular to do with the way the rest of the world is' (p. 151). But in that case – as McDowell justifiably concludes – Rorty is about as far as can be from the straightforward, commonsense-pragmatist outlook that he claims to uphold against the bother-headed worries of philosophers locked in the endless sterile debate between realism and anti-realism. For the default attitude of most people commonsensically immune to those worries is one that takes it pretty much for granted that the truth or falsehood of our various beliefs is determined by the way things stand in reality, rather than reality somehow corresponding to the way things appear from our particular linguistic or cultural perspective. Thus the trouble with Rorty's drastically dualist conception is that it 'severs what we want to think of as responsiveness to the norms of inquiry from any connection with [the] unproblematic notion of gettings right' (p. 150).
The same may be said about Quine's and Davidson's positions, at least in so far as they leave themselves open to a plausible reading on Rortian terms. In both cases there is a claim to get over the problems of old-style logical empiricism by adopting a truly (more radically) empiricist approach which renounces the delusive Kantian distinction between analytic and synthetic judgements, or Humean 'truths of reason' and 'matters of fact'. Hence Quine's well-known espousal of a naturalised (or physicalist) epistemology which rejects all recourse to intensional, modal, or other such 'opaque' concepts that cannot be cashed out directly in terms of a purely extensionalist scheme with no ontological commitments beyond those required for application of the first-order quantified predicate calculus.[6] Hence also Davidson's equally well-known statement that '[i]n giving up the dualism of scheme and world, we do not give up the world, but re-establish unmediated touch with the familiar objects whose antics make our sentences and opinions true or false'.[7] For this is not so much to overcome that dualism as to leave it all the more firmly entrenched by adopting a causal ('unmediated') theory of direct knowledge-by-acquaintance, and thus pre-emptively excluding such knowledge from the justificatory 'space of reasons' or the realm of critical-evaluative thought. Thus – according to McDowell – 'Davidson resolves the tension he finds in Quine in the wrong direction, and the result is precisely to leave us with the philosophical problems he wants to eliminate' (p. 138). Chief among these is the problem of explaining how a causal account of belief-acquisition along Quinean physicalist lines can possibly make the required distinction between rationally justified items of belief and those arrived at merely through exposure to a range of 'unmediated' sensory promptings.[8] It is for want of this distinction – so McDowell maintains – that Quine and Davidson both lean over into a coherentist or holistic theory of rational warrant where beliefs are justified solely in virtue of their hanging together with the entire going range of beliefs-held-true at any given time. Where Davidson differs from Quine is only in his more explicitly drawing the consequence: that 'we cannot make sense of thought's bearing on the world in terms of an interaction between spontaneity and receptivity', since 'if we go on using [these] Kantian terms, we have to say that the operations of spontaneity are rationally unconstrained from outside themselves' (p. 139).
This brings us to the heart of McDowell's argument for returning to Kant in quest of an answer to these problems in the wake of old-style logical empiricism. It is an argument that turns crucially on the notions of 'spontaneity' and 'receptivity', terms that figure in the First Critique at the point where Kant seeks to explain how sensuous intuitions may be 'brought under' concepts (or experience be rendered rationally intelligible) without engendering a vicious regress. That regress threatens on account of the gap – the difference of categoric status – between 'intuitions' and 'concepts'. After all, it appears that something more is required (some additional mediating term) in order to achieve the otherwise impossible passage from a realm of as-yet preconceptual experience to a realm of knowledge where judgement is exercised through the bringing to bear of conceptual understanding on the deliverances of sensibility. Kant talks here about the 'schematizing' power of the productive imagination, a faculty that is supposed to accomplish this passage by somehow linking intuitions and concepts through a power vested in the human mind, but one whose operations he fails to specify with any degree of clarity. Thus in one famously obscure sentence he refers to this power of imagination as 'an art buried in the depths of the soul', a phrase that has since given rise to much speculation among commentators – Heidegger especially – who seek to press beyond the limits of Kant's critical philosophy to its concealed hermeneutic or depth-ontological dimension.[9] It was also the starting-point of various developments in the German Idealist tradition after Kant which took such passages as justification for claiming (like Fichte) that objective reality was a construct or 'posit' of the world-constituting Ego, or again (like Schelling) that our representations were aspects of an all-encompassing dialectics of nature wherein all the stages and forms of human knowledge could find their appointed place.[10] And of course these two schools of thought – 'Subjective' and 'Objective' Idealism – were then taken up and purportedly transcended (Aufgehoben) in Hegel's attempt to assimilate everything (history, philosophy, art, politics) to a grand dialectical schema leading to the advent of Absolute Knowledge as a perspective atop all the limited or partial viewpoints achieved to date.
As we shall see McDowell has his uses for Hegel though a version of Hegel – like his version of Kant – suitably revised and rendered fit for consumption by Anglophone philosophers with no taste for such giddy metaphysical excesses. Still he accepts that there is a very real problem about Kant's idea of knowledge as consisting in the union of sensuous intuitions with adequate or corresponding concepts. Indeed it is the same kind of scheme/content dualism that has continued to haunt analytic philosophy in the post-Quinean line of descent. This is why McDowell suggests that we should switch attention to those other passages in Kant where the emphasis is more on 'spontaneity' and 'receptivity' as terms of address which may help to prevent that dualism from getting a hold in the first place. The following passage is typical enough and brings out his main points of disagreement with Quine, Davidson, and Rorty.
The conceptual capacities that are passively drawn into play in experience belong to a network of capacities for active thought, a network that rationally governs comprehension-seeking responses to the impacts of the world on sensibility. And part of the idea that the understanding is a faculty of spontaneity – that conceptual capacities are capacities whose exercise is the domain of responsible freedom – is that the network, as an individual thinker finds it governing her thinking, is not sacrosanct. Active empirical thinking takes place under a standing obligation to reflect about the credentials of the putatively rational linkages that govern it. There must be a standing willingness to refashion concepts and conceptions if that is what reflection recommends. No doubt there is no serious prospect that we might need to reshape the concepts at the outermost edges of the system, the most immediately observational concepts, in response to pressures from inside the system. But that no-doubt unreal prospect brings out the point that matters for my present purpose. This is that although experience itself is not a good fit for the idea of spontaneity, even the most immediately observational concepts are partly constituted by their role in something that is indeed appropriately conceived in terms of spontaneity. (pp. 12-3)
In this passage – and many others like it – McDowell claims to avoid a whole range of Kantian and post-Quinean dilemmas by way of a topographic metaphor which redraws the 'boundary' between experience and concepts so as to emphasise their areas of mutual overlap. Thus experience is 'open to reality' precisely in so far as it receives impressions from a 'world independent of our thinking', yet a world which all the same 'is not to be pictured as outside an outer boundary that encloses the conceptual sphere' (p. 26). That is to say, the precondition for attaining knowledge of an objective (mind-independent) world is that experience should always already be informed by concepts or modes of evaluative judgement that belong just as much to that world itself as to the mind that seeks to comprehend it. Thus 'the idea of conceptually structured operations of receptivity puts us in a position to speak of experience as openness to the layout of reality. Experience enables the layout of reality itself to exert a rational influence on what a subject thinks' (ibid).
There is a certain ambiguity about the placement of the word 'rational' in this sentence, depending on whether one takes it to mean 'what a subject [rationally] thinks to be the case in reality', or - as the grammar more strongly suggests – 'how the subject is duly influenced in her thinking by the rational structure of reality'. This latter (distinctly Hegelian) reading is one that seems most natural for a good many passages in McDowell's book and which indeed follows from his basic premise that epistemology goes off the rails – falls into various endemic dualist fallacies – when it allows the least possibility of distinguishing passive from active capacities of mind, or 'receptivity' from 'spontaneity'. Yet Kant himself is quite unable to explain how this dualism might be transcended, except through that vague appeal to the schematizing power of Imagination as an 'art buried in the depths of the soul'. Hence - as I have said - the subsequent history of radically divergent interpretations, from Fichte's idea of a world-constituting Ego which 'posits' reality as a projection of its own demiurgic powers, to the kinds of neo-Kantian (and logical-empiricist) construal which leave no room for the exercise of thought in its active or reflective capacity.
It seems to me that McDowell is right in suggesting that analytic philosophy has artificially narrowed its sights and created all sorts of needless, self-inflicted dilemma by ignoring developments in the 'other' (continental) tradition of thought. But it also seems to me that he has hitched his argument to some of the weakest passages in Kant and ignored just those later developments which engage most closely and critically with the various problems bequeathed by Kantian epistemology. 'If we can rethink our conception of nature', he writes, 'so as to make room for spontaneity, even though we deny that spontaneity is capturable by the resources of bald naturalism, we shall by the same token be rethinking our conception of what it takes for a position to deserve to be called »naturalism«' (p. 77). In other words, we can have the best of a modern (naturalised) epistemology while avoiding the reductively physicalist account which – as in Quine and Davidson – creates yet another drastic dualism between the realm of causally acquired and explainable beliefs and the justificatory 'space of reasons'. But in saying that we need to 'rethink our conception of nature' so as to 'make room for spontaneity' McDowell runs perilously close to endorsing the kinds of putative 'advance' upon Kant that were essayed by subjective and objective idealists like Fichte and Schelling. So it is hardly surprising – given this re-run of episodes from the history of post-Kantian thought – that McDowell should make his cautious gestures toward Hegel as offering an escape-route from all these Kant-induced puzzles and perplexities.
Still one may doubt that Hegel provides much of a solution once the issues have been set up in this way. McDowell brings him in as support for the idea that Kant's theory of knowledge can be naturalised, relieved of its surplus transcendental baggage, and thus restored to the community of human meanings, interests, and concerns. In short, this is a distinctly Wittgensteinian version of Hegel which locates the conditions for knowledge and experience in shared language-games or cultural forms of life, and which seeks thereby to talk philosophy down from its self-inflicted dilemmas. All the same, McDowell is prepared to push pretty far with the Hegelian doctrine of Absolute Idealism since he thinks – improbably enough – that it brings 'reality' back into the picture and, moreover, manages to dispense with the 'transcendental framework' which vitiates Kantian epistemology. Thus 'in spite of his [Kant's] staunch denials, the effect of his philosophy is to slight the independence of the reality to which our senses give us access' (p. 44). This results – McDowell thinks – from Kant's failure to follow through consistently on his own most important insight, namely the 'unboundedness' of the conceptual sphere or its active role at every stage in that process whereby reality is taken up into experience and experience in turn becomes the basis for reliable knowledge of the world. So Kant's successors from Fichte to Hegel were justified in their claim that he had 'betrayed' the project of Absolute Idealism by conceding the existence of an ultimate (noumenal) reality outside and beyond that sphere. Only if the latter is taken to encompass every aspect of thought, knowledge, and experience can philosophy be saved from the sceptical outcome – the divorce between mind and world - that results from any form of scheme/content dualism.
On this point McDowell agrees with Kant's Absolute Idealist critics: that 'we must discard the supersensible in order to achieve a consistent idealism', one which effectively 'frees Kant's insight so that it can protect a commonsense respect for the independence of the ordinary world' (p. 44). That is, we can (and should) reject Kant's distinction between phenomenal experience and noumenal reality since this is where a whole company of latterday thinkers – whether anti-realists or framework-relativists – are able to insert their sceptical wedge and argue for the sheer impossibility of crossing that divide. But we can do so only on condition of accepting that the realm of concepts is indeed 'unbounded' in the sense (quite simply) that there is 'nothing outside it', i.e., no element of 'reality' that is not taken up into knowledge and experience via the conceptual sphere. With this recognition, McDowell writes, 'we have arrived at a point from which we could start to domesticate the rhetoric of that philosophy' (p. 44). In other words, we can continue the naturalising process that Hegel applied to Kant and temper the language of Absolute Idealism by referring it back to the everyday conditions of situated human being-in-the-world. This would be a 'naturalism of second nature', one that acknowledged – in Hegel's spirit but also like Aristotle before him – that reality is always known in and through the various communal projects and enquiries that make up an ongoing tradition or cultural form of life. Thus '[e]ven a thought that transforms a tradition must be rooted in the tradition that it transforms . . . The speech that expresses it must be able to be intelligibly addressed to people squarely placed within the tradition as it stands' (p. 187).
These reflections derive mainly from McDowell's reading of Hans-Georg Gadamer and other writers in the German 'hermeneutic' line of descent from Schleiermacher and Dilthey.[11] What they provide – so he believes – is a means of re-thinking epistemological issues in the wider space of a philosophy freed to consider the various interests, values, and concerns that constitute the 'horizon of intelligibility' (Gadamer's phrase) for all knowledge and experience. Equally useful in this regard is the notion of Bildung – 'spiritual development' or 'self-cultivation' – applied both to individuals and the cultures or the communal traditions in which they participate. For it is, McDowell thinks, a signal advantage of such largely Hegel-inspired approaches that they help to wean us off the narrow (Cartesian) conception of knowledge as involving the transactions of a solitary mind with so-called 'external reality'.
Here again McDowell has recourse to Wittgenstein and in particular to the latter's idea that 'commanding, questioning, recounting, chatting, are as much a part of our natural history as walking, eating, drinking, playing'. By 'our natural history', McDowell suggests,
he [Wittgenstein] must mean the natural history of creatures whose nature is largely second nature. Human life, our natural way of being, is already shaped by meaning. We need not connect this natural history to nature as the realm of law any more tightly than by simply affirming our right to the notion of second nature. (p. 95)
This is where we arrive, according to McDowell, at the end of that passage from Kant, via Hegel, to a naturalised version of Absolute Idealism that can take Wittgenstein's point on board as a last farewell to all the chronic dualisms that have worked such mischief in epistemology and philosophy of mind. Among recent varieties perhaps the most damaging was the phenomenalist notion of sense-data as somehow presented to the mind in a raw state and hence in need of being worked up into structures of intelligible thought.[12] This approach 'aims to overcome anxiety about a gap between experience and the world by constructing the world out of experience, still conceived in just the way that gives rise to the anxiety' (p. 94). Thus McDowell agrees with Quine in roundly rejecting any such attempt – as by Carnap and the Logical Empiricists – to put together what philosophy had torn asunder by adopting a ground-up constructivist method wherein sense-data are taken as the basis for a full-scale account of knowledge and experience.[13] However, as we have seen, he considers Quine (and Davidson likewise) to have stopped well short of any adequate alternative theory for repairing the damage thus inflicted. Rather they should have followed the route that leads through and beyond Kant's philosophy to Hegel's Aufhebung of the Kantian dichotomies and thence – in a further naturalising move - to the idea of 'second nature' as pretty much expressing what the Absolute Idealists had to say minus their unfortunate metaphysical excesses.
There are several problems with McDowell's argument that make this a difficult (if not impossible) position to sustain. One is the fact that Kant's epistemology is a complex system of interrelated claims about experience, knowledge, and their transcendentally justified conditions of possibility which cannot be prised apart and applied in a selective fashion as McDowell seeks to do. This is not to deny – as I have argued above – that those claims are themselves deeply problematic and open to challenge on various grounds. Chief among them are his framing of the noumenon/phenomenon distinction and his insistence that noumenal 'reality' cannot by very definition fall under the concepts and categories of understanding, despite Kant's perforce having to apply such categories (e.g., that of causation) when he attempts to explain how the noumenal somehow relates to phenomenal experience.[14] So McDowell is far from alone in concluding that Kant's philosophy is in need of a pretty thorough overhaul and that the noumenal (or 'supersensible') could well drop out to the benefit of his other, more valuable insights. But unfortunately these are just the insights – in McDowell's reading – that depend most heavily on the whole co-implicated structure of Kantian argument. Thus, for instance, it is hard to make sense of Kant's (or McDowell's) claim for the conjoint operation of 'spontaneity' and 'receptivity' unless with reference to Kant's conception of the mind as predisposed to accomplish such acts of synthesis through its openness to 'supersensible' ideas concerning noumenal reality. These include the teleological idea – taken up at greater length in the Third Critique – of mind and nature as exhibiting a kind of harmonious purposive adjustment which transcends the furthest bounds of conceptual knowledge, yet in the absence of which such knowledge would be wholly unattainable.[15] McDowell is understandably reluctant to be drawn in this direction and indeed goes out of his way at several points to make that reluctance plain. But even then it is hard to make out what is left of Kant's claims – and McDowell's case in defence of them – if those claims are recast in a naturalised (detranscendentalised) idiom that effectively denies their validity.
Now it may well be (as many critics have thought) that these passages in Kant are obscure to the point of defying rational commentary. Certainly they go far beyond anything that would count as an adequate epistemological argument –or one with any claim to serious attention – for most Anglophone philosophers currently working on issues of mind and knowledge. Hence their strong appeal to exegetes like Heidegger whose main purpose in reviewing the tradition of 'Western metaphysics' is to point out just such regions of obscurity and exploit their potential for revisionist commentary in the depth-hermeneutic mode.[16] McDowell gives such thinking a wide berth despite his attraction to Gadamer's more moderate or traditionalist brand of hermeneutical enquiry. After all, it would run clean against the main purpose of his book to suggest that these problems with Kant might be overcome - or simply set aside - by taking an altogether different view of what philosophy ought to be about. At least he has to maintain the position that 'receptivity' and 'spontaneity' are useful terms with which to approach certain genuine (indeed urgent) philosophical issues, whatever the difficulty of holding them apart from other, more dubious aspects of the Kantian enterprise. Thus McDowell cannot go along with any 'strong'-revisionist reading, like Heidegger's, which treats those issues as merely symptomatic of a deep-laid aberration in the history of Western thought, nor again with any briskly dismissive attitude – such as Rorty's – which regards them as pointless and tedious worries that we should have got over by now.[17] In fact this is just what Rorty suggests in his essay on McDowell where the latter is cast as one of those fretful analytic types who have failed to follow through on the 'linguistic turn' and are still hung up on the same old pseudo-problems that have vexed philosophy from Descartes down.[18] From this point of view McDowell is headed in the right direction when he offers a brief passage of tribute to Robert Brandom's 'revelatory' reading of Hegel as just what is required to break the spell of Kantian transcendental illusion and bring us out safely on the far side of all that vexatious philosophical talk (p. ix).[19] If he had only taken the lesson to heart – so Rorty believes – then McDowell might have saved himself a deal of misplaced anxiety and also a large amount of desperately wiredrawn Kantian exegesis.[20]
As I say, McDowell is in no good position to accept this kind of therapeutic advice since he does think that there is a 'problem of knowledge' that merits serious attention and can best be approached via Kant with some help along the way from Hegel, Wittgenstein, Gadamer, Brandom, and others. But there is still a deep tension – as Rorty shrewdly perceives – between McDowell's express Kantian commitments and his need all the same for these further points of reference in order to propose a reading of Kant that avoids (or at any rate tries to avoid) the above-mentioned dilemmas. This conceptual strain shows up most clearly when McDowell invokes Gadamer's Hegelian appeal to 'tradition' as a means of restoring knowledge to the context of a lifeworld wherein those dilemmas supposedly cannot arise. 'Understanding', he writes, 'is placing what is understood within a horizon constituted by a tradition, and . . . the first thing to say about language is that it serves as a repository of tradition' (p. 184). And again: 'a natural language, the sort of language into which human beings are first initiated, serves as . . . a store of historically accumulated wisdom about what is a reason for what' (p. 126). To grasp this point is also to see that any problems encountered with Kant's theory of mind and knowledge are problems only when viewed in artificial (philosophically-induced) isolation from any such background context. No doubt it is the case – especially for Kant – that this tradition 'is subject to reflective modification by each generation that inherits it', and indeed that 'a standing obligation to engage in critical reflection is itself a part of the inheritance' (p. 126). Nevertheless (and here McDowell follows Gadamer) such criticism has to start out from within the tradition – that is to say, from a standpoint sufficiently informed by prevailing values and beliefs – in order to count as rationally justified for members of any given community.
Of course this argument has provoked much debate, often on the grounds of its implicit conservatism when applied to ethical, social, or political issues.[21] That is to say, such thinking will tend to privilege the claims of tradition or communal assent over the freedom to challenge those claims through an exercise of rational-evaluative judgement that requires the capacity to stand back from them and thereby achieve a genuinely critical perspective. This is the objection that Habermas raises against Gadamer in making his case for a theory of 'communicative action' which envisages precisely such a transformation of tradition-bound beliefs and values through the capacity and freedom to criticise existing communal norms.[22] It is also the chief area of dispute between followers of Wittgenstein – such as Peter Winch – who adopt a communitarian approach based on the appeal to cultural 'forms of life' and opponents who argue that this approach leads to cultural relativism and hence (in some cases) to a placid endorsement of sundry irrational or morally repugnant customs and practices.[23] There is no room here for a detailed review of the various arguments and counter-arguments brought up by parties to this long-running debate. My point is that McDowell finds himself very awkwardly placed since he sets out to defend both a Kantian commitment to the exercise of rational freedom (or autonomy) in matters of moral-intellectual conscience, and – somehow consistently with this – an appeal to thinkers such as Hegel, Wittgenstein, and Gadamer on the tradition-based or communally-sanctioned character of all thought and judgement.
Now it may well be argued that there is no necessary conflict here in so far as we can take the latter point (i.e., the sheer impossibility of breaking altogether with communal norms of reason) while none the less preserving a significant margin of freedom for the expression of dissident or heterodox views. Indeed it is hard to deny that this joint process must be in play when critics reject some more-or-less salient item of locally accepted belief and yet manage to elicit understanding – and maybe assent – from members of their own cultural community. However the issue is more pressing for McDowell since he stakes so much on Kant's claims for the exercise of intellectual-moral freedom as a defining feature of human personhood, while none the less rejecting any version of Kant's appeal to the 'supersensible' realm. This leaves him with the difficult task of explaining how such freedom can possibly be derived from – or rendered compatible with – the process of knowledge-acquisition whereby sensuous or phenomenal intuitions are 'brought under' adequate concepts. For, as Kant sees it, there is just no way that these two faculties (cognitive understanding and practical reason) can be thus run together without either depriving knowledge of all determinate content, or on the other hand embracing a determinist outlook which denies any space for the exercise of freely-willed, autonomous agency and choice.
Kant's 'solution' – his requirement that we simply must treat them as realms utterly apart – has seemed to many critics a desperate measure and one that leaves the problem very firmly in place. McDowell thinks that it is not, after all, so damaging and can best be got over by adopting a naturalised (broadly Hegelian) approach. That is, we can preserve the most important aspects of Kantian epistemology and ethics if we build in just a few revisionist concessions which prevent that otherwise disastrous rift from opening up in the first place. This requires (1) that we abandon all talk of a noumenal (unknowable) reality-beyond-appearances along with a realm of purely 'supersensible' ideas, and (2) that we emphasise Kant's appeal to the joint operations of 'spontaneity' and 'receptivity', rather than his notion of raw 'intuitions' which are somehow (impossibly) supposed to match up with 'concepts' of understanding. This latter notion, according to McDowell, is one that has caused endless trouble not only for Kant and his more orthodox exegetes but also for Quine, Davidson, and other latterday crypto-dualists. Its source is always some version of the 'Myth of the Given', or the mistaken idea – regrettably present in some passages of Kant – that knowledge must involve the bringing-into-relation of sensory 'content' and conceptual 'scheme' through some further (mysterious) operation of mind.[24] Once abandon that idea – so McDowell thinks – and we will be free to pursue Kant's other, more promising line of thought.
However there are still deep problems with this Kantian conception and with McDowell's ambitious claims on its behalf. Indeed McDowell states the problems more than once with great clarity and force, even though he goes on each time to suggest that they are not after all so serious if viewed in that alternative light. Thus:
[the] trouble shows up again here, in connection with impingements on spontaneity by the so-called conceptual deliverances of sensibility. If those impingements are conceived as outside the scope of spontaneity, outside the domain of responsible freedom, then the best they can yield is that we cannot be blamed for believing whatever they lead us to believe, not that we are justified in believing it. (p. 13)
McDowell holds this to be a pseudo-dilemma which we force upon ourselves – and on our reading of Kant – if we fail to take his oft-repeated point about the jointly operative roles of 'receptivity' and 'spontaneity' in every act of cognitive or rational-evaluative judgement. For we shall then be hard put to discover any space for the exercise of 'responsible freedom', given that those sensory-perceptual 'impingements' are taken as already subject to 'conceptual' processing of some kind, and must thus be assumed to occupy the whole space of cognitive judgement in so far as such judgement is assumed to consist exclusively in the bringing of phenomenal intuitions under adequate concepts. Thus the only way out of this all-too-familiar impasse is to follow Kant (or McDowell's Kant) in redefinining 'spontaneity' as that which encompasses both the mind's power to receive and conceptualise the 'deliverances of sensibility', and its freedom to reflect 'responsibly' on those same deliverances. Otherwise we are back, so to speak, at square one with a further series of yet more troublesome dichotomies between intuition and concept, experience and knowledge, or judgement conceived as passive 'receptivity' and judgement conceived as the exercise of our capacity for 'spontaneous' thought and reflection.
McDowell's way of handling this problem most often takes the form of conceding its existence – even its intractable nature – while declaring nevertheless that there must be some solution (and one to be had from Kant) if we can just make suitable adjustments to our sense of what that solution should entail. Hence his suggestion – to repeat – that 'although experience itself is not a good fit for the idea of spontaneity, even the most immediately observational concepts are partly constituted by their role in something that is indeed appropriately conceived in terms of spontaneity' (p. 13). But it is hardly nitpicking to comment that 'experience' has to be thought of – after Kant – as already subject to some form of conceptual processing, or – in Davidsonian parlance – as 'under some description' or other. So if 'experience itself' is not a 'good fit' (whatever precisely this means) for the 'idea of spontaneity', then it is hard to see why that idea should apply to 'even the most immediately observational concepts'. All that is happening here – to put it bluntly – is that McDowell is indulging in a bit of stipulative redefinition which requires us to accept it as only 'appropriate' that those concepts should be brought within the sphere of spontaneity, and thus freed up for the exercise of responsible (rational-evaluative) judgement. But this involves such a tortuous process of argument – and such a strained reading of Kant – that one has to ask why McDowell has chosen this particular route through difficult and at times treacherous terrain. For there is always the alternative (as Rorty is on hand to remind him) of skipping the whole sad business with Kant and marching straight on – with Brandom's assistance – to Hegel and a happy issue out of all his philosophical perplexities.[25]
Now we have seen why this is not in fact an option for McDowell, given his commitment to a theory of knowledge that conserves more of the Kantian conception – and a stronger emphasis on its critical aspect – than anything to be had from a Hegel 'naturalised' in the wholesale Rortian fashion.[26] But we have also seen how McDowell has recourse to other, more apparently attractive options for talking Kant's philosophy down from the heights of transcendental illusion and restoring it to the lifeworld or the various contexts of human knowledge and experience. Among his sources – as I have said – are readings of Hegel (such as those of Brandom and Gadamer) which stress that dimension in different ways while stopping well short of Rorty's radical-revisionist approach, and also the Wittgenstein-derived idea of communal usage as the furthest one can get by way of ultimate grounds or justifications. What he hopes to find is a viable account of 'second nature' that would allow due space for the exercise of critical or rational-evaluative thought (thus avoiding any kind of complacent traditionalism), yet at the same time incorporate the lessons learned from these and other broadly communitarian thinkers. Thus the question becomes: 'how can spontaneity permeate our lives, even to the extent of structuring those aspects of them that reflect our naturalness – those aspects of our lives that we share with ordinary animals?' (p. 65).
According to McDowell the chief reason for many philosophers' unease with this idea is the grip on their minds of a different variety of naturalism, one with its origins in early modern science and its upshot in the various dualisms – overt or covert – adopted under pressure from the scientific worldview. On this account a 'naturalised' epistemology would explain knowledge entirely as a matter of belief-states arrived at through the action of various physical stimuli or causal inputs assumed to act directly on our sensory receptors and thereby trigger the appropriate response. In other words, it would leave absolutely no room for those justificatory processes – belonging to the 'space of reasons' as opposed to the causal-explanatory realm – which make all the difference between human and nonhuman (animal) forms of cognitive endeavour. Thus '[t]he thought is that the freedom of spontaneity ought to be a kind of exemption from nature, something that permits us to elevate ourselves above it, rather than our own special way of living an animal life' (p. 65). However this thought has the unfortunate consequence – as McDowell sees it – of reintroducing all those bad old dualisms that have cast such a long shadow over philosophy from Descartes, via Kant, to Quine and Davidson. It is then a short step to Rorty's yet more drastic proposal, namely that of a total divorce between the realm of causal stimuli (where photons and suchlike impact on our sensory receptors with various physically determinate effects), and the realm of belief or interpretation (where those impacts are held to exert no constraint upon the range of admissible theories or descriptions).[27]
Such is the result of that progressive 'disenchantment' of nature – McDowell's term – brought about by the methods of the physical sciences from Galileo down. So it is not hard to see why other philosophers should adopt an opposite (strongly reactive) position and reject any kind of naturalised epistemology that purports to explain knowledge by reference to a causal theory of belief-acquisition lacking all normative values or justificatory criteria.[28] However this position has its own liabilities, among them a tendency to flip straight over into something very like Rorty's all-out dualist view of the matter. That is to say: since causal explanations run out at such an early stage – somewhere between our sensory receptors and basic cognitive processing activities – therefore we should view them as pretty much irrelevant to whatever goes on at more advanced stages of interpretation, belief-adjustment, theory-construction, etc. Davidson's view is less extreme on the face of it, holding - in McDowell's words - that 'experience is causally relevant to a subject's beliefs and judgements, but has no bearing on their status as justified or warranted' (p. 14). However it is then hard to see how those 'beliefs and judgements' could possess any genuine truth-evaluable content, restricted as they are (on Davidson's account) to a ground-floor level of stimulus-response where they cannot be thought of as 'justified' or 'warranted' in higher-level (epistemological) terms.[29] Davidson may try to get around this problem by defining 'belief' differently in various contexts of argument so that it straddles the line – or can be made to perform double service – as between the causal-explanatory and the rational-evaluative realms. However, as McDowell rightly suggests, we should perhaps be suspicious of his 'bland confidence that empirical content can be intelligibly in our picture even though we carefully stipulate that the world's impacts on our senses have nothing to do with justification' (p. 15).
This is why McDowell thinks that there is a great deal at stake on the idea of 'second nature' as a means of healing the rift between causes and reasons, sensory inputs and rational outputs, or - to adopt his preferred terminology – Kantian 'receptivity' and 'spontaneity'. This would be a naturalised epistemology, to be sure, but one that nevertheless managed to avoid the reductionist perils of treating knowledge as purely and simply the upshot of our causal interactions with the world or our direct responses to physical stimuli. It would be 'naturalised' in the sense of accepting – like Aristotle – that we are physical creatures impinged upon by various circumambient sensory promptings but able to take those promptings up into our own life-histories and the wider realm of shared human interests and concerns. Following Gadamer it would point the way down from those rarefied heights of transcendental reflection where Kant on occasion leaves us stranded and recall us to the lifeworld – or horizon of mutual understanding – wherein all our cognitive and ethical enquiries have their beginning and their end.[30] It would also take Wittgenstein's point about 'language-games' and communal 'forms of life' while none the less conserving a space - unlike Wittgenstein's more 'complacent' exegetes - for the exercise of critical-reflective reason as itself one crucial and defining component of our own, distinctively human lifeworld.[31] Above all, it would explain what sets us apart from other (non-human) animal species without in the process going so far so far as to sever all links between 'us' and 'them' by adopting a Cartesian dualist view on which 'they' come out as mere automata devoid of sentient being, and 'we' as creatures unthinkably split between a realm of pure rational consciousness and a realm of brute physical existence. For this is just a product of the sharply dichotomised (since thoroughly 'disenchanted') naturalistic worldview which inspired Descartes in his solitary project and which has left its mark on so many subsequent chapters of epistemological enquiry.
'We can avoid the dilemma', McDowell writes, and do so by conserving our proper self-image as rational enquirers both responsive to causal inputs from the physical world and responsible for the exercise of evaluative judgement with respect to those same deliverances. Thus:
[w]e do not need to say that we have what mere animals have, non-conceptual content, and we have something else as well, since we can conceptualize that content and they cannot. Instead we can say that we have what mere animals have, perceptual sensitivity to features of our environment, but we have it in a special form. Our perceptual sensitivity to our environment is taken up into the ambit of the faculty of spontaneity, which is what distinguishes us from them. (p. 64)
There are several other passages in the book that make this point with slightly different phrasing, though always to the effect that we humans are both like and unlike non-human animals in certain crucial (philosophically salient) respects. For instance: '[i]t is the spontaneity of the understanding, the power of conceptual thinking, that brings both the world and the self into view. Creatures without conceptual capacities lack self-consciousness and – this is part of the same package – experience of objective reality' (p. 114). However there are several problems here, quite apart from the standard 'animal-rights' objection on grounds of speciesism, anthropocentrism, the arrogant claim to 'know' for sure what animals can or cannot possess in the way of concepts, self-awareness, 'experience of objective reality', and so forth. One is the fact that McDowell's idea of 'spontaneity' is itself just another, more roundabout means of smuggling back that same old dualism which fixes a gulf between human and animal modes of being, and hence (by implication) between the mental and physical aspects of human existence. After all, it is only by imposing a stipulative (Kantian) idea of what shall count as 'experience' or 'objective reality' that McDowell can hold this distinction in place and define what is human – as opposed to merely animal – through recourse to the notion of 'spontaneity'. But this is just to say – despite his denial in the above-quoted passage – that we do have 'something else as well', that is, something in addition to the 'non-conceptual content' which we possess in common with the animals and which (presumably) consists in their and our shared capacity for reacting appropriately to various sorts of ambient physical stimuli. Perhaps we are saved from other, more outright and challengeable ways of stating the difference, as for instance by the claim that 'we can conceptualize that content and they cannot'. Perhaps it allows us to make more room for the commonalty of human and nonhuman-animal experience by speaking of a 'shared perceptual sensitivity to features of our environment', despite the crucial rider that 'we have it in a special form'. But here again this looks very much like a case of McDowell having his cake and eating it, or exploiting whatever is to be had by the appeal to a (quasi-)naturalised epistemology while retaining the idea of 'second nature' – 'spontaneity' in a somewhat less Kantian guise – as a hedge against the kind of naturalistic approach that would threaten his project at source.
It may seem odd that McDowell should have so much to say with regard to the difference between humans and 'mere' animals in the context of an argument otherwise chiefly concerned with epistemological issues. But the relevance of these passages will perhaps become clearer if one considers how closely they connect with his whole attempt to discover a third-way alternative path beyond all the vexing dualities of post-Kantian philosophy of mind and knowledge.
Thus McDowell doesn't want to be misunderstood as 'debunking animal mentality', as adopting 'a reductive conception of biological imperatives', or as in any sense seeking to deny 'that they ["mere animals"] can be, in their ways, clever, resourceful, inquisitive, friendly, and so forth' (p. 182). Rather, '[t]he point is just that dumb animals do not have Kantian freedom', the latter taken as uniquely characteristic of human animals in virtue of their possessing just that attribute (i.e., 'spontaneity') which pertains to them alone as 'second nature', and which thus raises them above such merely 'biological imperatives'. So there is a sense in which McDowell's epistemological case stands or falls with his case for the partial (strictly limited) measure of continuity between nonhuman-animal and human modes of cognitive dealing with the world. What he needs to establish by means of these parallel arguments is that the human (conscious and reflective) way of being-in-the-world is sufficiently in touch with – but also sufficiently distinct from – the kind of physically responsive being-in-the-world that is held to typify animal existence. Thus 'no one without a philosophical axe to grind can watch, say, a dog or a cat at play and seriously consider bringing its activities under the head of something like automatism'. All the same – more crucially for McDowell's case – 'we can deny Kantian spontaneity [to such animals] while leaving plenty of room for the self-movingness that is plain to the unprejudiced eye in such a scene' (p. 182). And it is just this kind of commonality-with-difference that McDowell requires in order to explain by analogy how the human capacity for knowledge both depends upon and radically transforms the animal capacity for 'merely' responding – in however sensitive or resourceful a manner – to a range of circumambient causal stimuli.
However his argument also requires an appeal to that idea of human cognitive capacities as a kind of 'second nature' which provides the necessary bridging concept between these two domains. And it is here that McDowell makes his brief venture into regions of post-Kantian 'continental' thought – especially the tradition of philosophical hermeneutics represented by Gadamer – in quest of an approach that might give substance to this claim while retaining what is crucial (for his own purposes) in Kant's theory of mind and knowledge. Thus McDowell's chief debt to Gadamer has to do with the latter's 'remarkable description of the difference between a merely animal mode of life, in an environment, and a human mode of life, in the world'. (p. 115). That is to say, it concerns the way in which certain (basically naturalistic) modes of description can assume a quite different salience when translated from the one context to the other, or from talk about features of a physical 'environment' – such as applies most aptly in the case of mere animal existence – to talk about a 'world' of distinctively human reasons, interests, meanings, values, priorities, and so forth. 'The point of this', McDowell writes, 'is that it shows in some detail how we can acknowledge what is common between human beings and brutes, while preserving the difference that the Kantian thesis forces on us' (p. 115). But his phrasing here is enough to suggest – as I have argued above – that there is an unresolved tension of some sort between McDowell's avowedly Kantian commitments and his wish to avoid their more awkward implications by adopting a hermeneutic approach. After all, what can it be that 'the Kantian thesis' so powerfully 'forces upon us' if not the drastic bifurcation of realms between animal and human, causal and rational, or physically-explainable and normatively-justified modes of knowledge and behaviour? Or again: why introduce this appeal to Gadamer unless – as so often in recent 'continental' debate – for the sake of talking philosophy down from its high-flown critical pretensions in the face of opposed (e.g., Wittgensteinian) claims for the priority of communal 'tradition' or cultural 'forms of life'?
McDowell's answer is seemingly quite straightforward: that it ties in with the idea of 'second nature' as encompassing both our existence as 'natural beings' and – compatibly with that – our difference from the animals in matters of knowledge or rational belief-formation. 'Of course', he cautions, 'it had better not be that our being in charge of our lives marks a transcendence of biology; that looks like a version of the rampant platonist fantasy' (p. 115). But we can still have a kind of 'naturalised platonism' – McDowell's curious phrase – which effectively gives us the best of both worlds by holding the crucial distinction in place while acknowledging the claims of our partial commonality with nonhuman modes of sentient being. Where 'rampant platonism' envisages a realm of abstract Ideas or rational structures 'independent of anything merely human, so that the capacity of our minds to resonate to it looks occult or magical', naturalised platonism holds on the contrary that 'the demands of reason are essentially such that a human upbringing can open a human's eyes to them' (p. 92). In short, it is a philosophy of Bildung – of self-cultivation through the norms and values of an acquired cultural tradition – as developed by those otherwise diverse thinkers (among them Aristotle, Hegel, Wittgenstein, and Gadamer) whom McDowell here calls to witness. Such is that 'naturalism of second nature' which he thinks of as a viable escape-route from many of the problems that currently beset analytic philosophy in the wake of Quine and Davidson.
However this is a very long way from Kant, or from any interpretation that could claim fidelity to the main precepts of his epistemology and ethics. For of course it is Kant's single most important claim that philosophy should be free to judge, criticise, and reach an independent verdict on all issues concerning knowledge and truth, or all matters properly pertaining to the exercise of reason in its 'public' (intellectually accountable) role.[32] Thus there is no item of received wisdom – whether scientific, philosophical, moral, or religious – that warrants endorsement merely on the grounds of its traditional status or its playing some role in the system of inherited communal beliefs. Rather, everything should be open to question from the standpoint of critical philosophy, even if the philosopher has sometimes to concede (in exchange for this rightful prerogative) that such freedom of expression may be subject to limits in the case of publications intended for a wider readership.[33] Still this concession on Kant's part can best be seen as a sop to the censors – theologians and state officials – and as carrying more than a hint of self-protective irony. At any rate the main thrust of his argument in the three Critiques is to vindicate the freedom of intellectual conscience, of rational-reflective enquiry, and – above all – of autonomous judgement in the ethical and socio-political realms. Indeed, it was largely in response to these claims that there developed a strong counter-movement (beginning with Herder and continued through the line of hermeneutic theorists from Schleiermacher and Dilthey to Gadamer) which emphasised the role of tradition or cultural context as a formative element in all such spheres of judgement.[34]
Needless to say the issue is more complex – and the positions adopted more varied and nuanced – than I have managed to convey here. For one thing there is the question (central to Gadamer's debate with Habermas) as to whether the hermeneutic enterprise should be viewed as inherently conservative or tradition-bound.[35] Otherwise put, it is the issue of how thinking can attain a critical standpoint vis-à-vis tradition if the act of criticism is itself bound up with the so-called 'hermeneutic circle', or the need to interpret existing beliefs (so to speak) 'from the inside' before such criticism can make any sense to members of the cultural community concerned. Certainly Gadamer goes out of his way to reject any claim that this argument entails an attitude of passive or unquestioning adherence to customary ways of thought. Also there are writers on Wittgenstein, sensitive to just this charge, who roundly reject the idea that talk of 'language-games' or communal 'forms of life' goes along with a failure (or an ideologically motivated refusal) to criticise received beliefs.[36] Besides which – to complicate matters yet further – one must take account of those recent defenders of Kant (Onora O'Neill among them) who deny that there exists any sharp conflict between the Kantian and at least some versions of the communitarian approach to issues of ethics and epistemology.[37] So I should not wish to claim that McDowell's appeal to a broadly hermeneutic tradition – reaching back as it does via Gadamer and Wittgenstein to Hegel and thence to Aristotle – necessarily clashes with his Kantian stress on the exercise of reason as involving the capacity to stand back from traditional beliefs and subject them to critical scrutiny. Nevertheless there is a genuine tension at just this point in his argument and one which shows up most plainly when he adduces the concept of Bildung (of self-cultivation as a 'naturalism of second nature') in order to domesticate Kant's more 'transcendental' or metaphysically extravagant claims.
The trouble is that 'second nature' has a meaning which stubbornly clings to the phrase despite such efforts to redeem it for a better purpose. That is, it tends always to put us in mind of those standing beliefs, dispositions, customs, practices, etc., which are 'naturalised' just to the extent that they form the taken-for-granted background of our everyday life. McDowell does his best to rehabilitate the notion by denying that any such conflict need arise between the sorts of thinking that typically occur in that context (where we are not self-consciously aware of them) and the kinds of thought-process that engage our attention when reflecting – in a more critical mode – on the validity of our various claims to knowledge or the justification of our various social, ethical, or communal practices. In this respect he agrees with Gadamer: that it is only through a false abstraction from the lifeworld of situated human understanding that philosophers can force such an artificial choice between tradition-based (hermeneutic) modes of enquiry and critical-evaluative projects of the kind espoused by thinkers like Kant and Habermas. Here again it is 'second nature' that sets us apart from the non-human animals, those for whom it is scarcely intelligible to suppose that they might be 'born at home in the space of reasons'. For in the case of human beings we can readily conceive that they are 'born mere animals', but are then 'transformed into thinkers and intentional agents in the course of coming to maturity' (p. 125). Moreover, what enables this transformation to occur is the process of Bildung wherein language takes pride of place, since by acquiring language a human being also acquires the resources and capacity for rational thought which signal her emancipation 'from a merely animal mode of living into being a full-fledged subject, open to the world' (p. 125).
However it is still hard to see why such progress should go any further than initiating the 'subject' into a lifeworld of pre-given values, beliefs, or traditional (acculturated) habits of thought. Indeed McDowell effectively concedes as much when he writes – more in Wittgensteinian than Kantian vein – that through being introduced into a language, the human creature is also introduced 'into something that already embodies putatively rational linkages between concepts, putatively constitutive of the layout of the space of reasons, before she comes on the scene' (ibidem). The sheer awkwardness of phrasing here – as so often in McDowell's book – suggests that his argument is under strain from conflicting ideas about just what constitutes the 'space of reasons' (the realm of critical-evaluative thought) as opposed to whatever is handed down by tradition, by commonsense wisdom, shared social custom, and the like. At any rate that space would seem to be drastically narrowed by his very un-Kantian concession that the 'rational linkages between concepts' are already in place – established by communal warrant – 'before she [i.e., the subject-initiated-into-language] comes on the scene'. Nor is the strain taken off by his assertion that those linkages are 'putatively constitutive of the layout of the space of reasons'. For any Kantian resonance of the term 'constitutive' is promptly cancelled – or rendered otiose – by the notion that they might all the same be merely 'putative', or (what appears to be implied) count as 'rational' only from the standpoint of one brought up within the relevant tradition or communal form of life.
The paragraph goes on in much the same way without resolving the issue, or rather – I would suggest – by refusing to face it and continuing to mix broadly Wittgensteinian with vaguely Kantian directives. Thus:
Human beings mature into being at home in the space of reasons or, what comes to the same thing, living their lives in the world; we can make sense of that by noting that the language into which a human being is first initiated stands over against her as a prior embodiment of mindedness, of the possibility of an orientation to the world. (p. 125)
This passage seems to me symptomatic of the failure, on McDowell's part, to provide any argument that would make good his central claim to have redeemed Kant for the purposes of a naturalised (post-Quine-and-Davidson) approach, while none the less retaining a Kantian sense of its properly rational or critical-evaluative bearings. Rather what it shows is yet another displaced variant of the dualism between causal explanation and the 'space of reasons', a dualism starkly present in Kant – despite McDowell's claims to the contrary – and still very evident (as he himself argues) in thinkers like Quine and Davidson. Where McDowell's argument differs is in shifting the locus of causality from 'nature' to 'second nature', or from an overtly determinist concept of causal explanation to a theory which seeks to avoid that consequence by appealing to the freedom notionally vested in our coming-to-maturity as denizens of a lifeworld or inheritors of a communal tradition.
Not that we can hope for any straightforward return to Aristotle's 'innocent' conception of human beings as creatures whose 'rationality is integrally part of their animal nature', and whose relationship to the natural order need not be 'fraught with philosophical anxiety' (p. 109). Such thinking came easily to Aristotle since he lived long before the advent of a modern, scientifically informed worldview in which the 'realm of law' would be reconceived as existing quite apart from the 'realm of meaning', or the sphere of distinctive (humanly-intelligible) purposes, needs, and desires. Indeed '[i]t would be crazy' – McDowell flatly declares – 'to regret the idea that natural science reveals a special kind of intelligibility, to be distinguished from the kind that is proper to meaning' (ibidem). All the same we may reasonably hope to turn back the tide of post-Galilean disenchantment by allowing that, even if 'the spontaneity of the understanding cannot be captured in terms that are apt for describing nature on that conception', still 'it can permeate actualizations of our animal nature', thus pointing a way beyond those same philosophical difficulties 'while fully appreciating what makes them gripping' (ibid). Such would be a 'postlapsarian or knowing counterpart of Aristotle's innocence', an outlook that accepts this historically unavoidable separation of realms, but which yet holds out against the kinds of disabling dualism that have left such a mark on recent philosophical thought. In other words – as McDowell puts it – we can fully accept 'the great step forward that human understanding took when our ancestors formed the idea of a domain of intelligibility, the realm of natural law, that is empty of meaning', while none the less 'refus[ing] to equate that domain with nature, let alone with what is real' (ibidem).
I must confess to finding this passage opaque to the point of making no sense on any possible construal of its terms and logical structure. After all, what precisely can it mean to speak of a realm of 'natural law' – that whereby the physical sciences have attained their present high status – while refusing to grant that it equates with 'nature' or indeed with what is 'real' from some ultimate (unspecified, maybe noumenal) perspective? Then again: how should we interpret McDowell's claim that '[s]ince we are setting our faces against bald naturalism, we have to expand nature beyond what is countenanced in a naturalism of the realm of law' (p. 109)? Presumably this is the 'second nature' – the realm of 'postlapsarian' innocence regained – which allows us somehow to reintegrate our modern worldview (after such knowledge, what forgiveness?) with our need for a conception of human enquiry responsive to values that find no place within the 'disenchanted' realm of law-governed natural necessity. But McDowell's suggested way of answering that need is that we redefine 'nature' in such terms as to conjure up memories of Schelling and the kind of quasi-pantheistic Naturphilosophie that followed so closely after Kant.[38] Thus: '[i]f we can rethink our conception of nature so as to make room for spontaneity . . . we shall by the same token be rethinking our conception of what it takes for a position to deserve to be called »naturalism«' (p. 77). There is just room – again through the presence of a saving grammatical ambiguity – to interpret McDowell as saying that the 'spontaneity' in question is that which belongs to 'second nature', i.e., to human beings and their active capacity for gaining knowledge of the natural world. But there is also the suggestion – here and elsewhere – that this 'spontaneity' is somehow intrinsic to nature itself as revealed through a deeper (interactive) process of reciprocal exchange between mind and world. I have remarked already on the echoes of Fichte and his world-positing ego that sometimes emerge to rather startling effect from McDowell's meditations on this theme. What the above-cited passage brings out – to likewise startling effect – is the way that his argument sometimes leans over into a strain of crypto-Schellingian nature-philosophy where the mind-world dualism is simply collapsed (so to speak) in the opposite direction. In other words there is more than a suspicion that McDowell is engaged in a largely unwitting repetition of episodes from the history of German idealism during the period directly after Kant. And this suspicion is reinforced by his idea that Hegel – or a suitably 'naturalised' Hegel – can be brought in by way of damping down the chronic oscillation between 'subjective' and 'objective' idealisms which marked that tempestuous chapter of thought.
Now there is – though McDowell seems strangely unaware of it – a whole rich tradition of more recent continental philosophy that engages these issues at far greater depth and with a much greater degree of historically informed understanding. I am thinking here chiefly of Husserl's phenomenological meditations on scientific knowledge and its relation to the lifeworld of human interests and concerns. This is the topic of his great book The Crisis of the European Sciences and Transcendental Phenomenology, a work – unmentioned by McDowell – which should be a standard reference-point for anyone attempting such a project.[39] Thus Husserl offers some acute and far-reaching observations concerning the rise of modern post-Galilean physical science, the progressive mathematization of nature, the encroachment of a narrowly positivist outlook in other (e.g., historical, sociological, and philosophic) domains of thought, and the reactive turn toward irrationalist doctrines among thinkers in the Nietzschean line of descent. Moreover, he pursues these philosophico-historical investigations always in the hope of reintegrating science with the lifeworld of shared ('traditionalised') knowledge and experience, even while conceding – as against any form of likewise reactive cultural relativism – that scientific knowledge has its own distinct standards of objectivity, rigour, and truth. This connects in turn with Husserl's wider project of transcendental phenomenology, that is to say, his lifelong concern to redeem the discourse of Kantian philosophy in such a way as to retain its epistemo-critical dimension yet at the same time preserve its ethical (normative-evaluative) content, and also – what Kant signally failed to do – avoid the drastic separation of realms between knowledge and human interests.[40] The above sentence could just as well stand as a description of what McDowell himself sets out to achieve in Mind and World. But he does so – it seems to me – under the self-imposed handicap of starting out from a reading of Kant (or certain passages in Kant) that is locked into various long-familiar problems and which takes no stock of later 'continental' developments which are highly germane to his purpose.
This takes us back to my earlier point: namely, that McDowell wants to think beyond the false dilemmas (as he sees them) of logical empiricism, but fails in this endeavour – like Quine and Davidson before him – precisely on account of a residual attachment to the logical-empiricist paradigm. Indeed this objection, or something very like it, is one that McDowell also brings against Gareth Evans, his early colleague and mentor whose work is acknowledged handsomely at various points in the book but whose thinking he now finds regrettably subject to a dualist conception of mind and world.[41] Thus Evans never quite presses through with the Kant-derived insight that experience is always already within the conceptual 'space of reasons', and therefore (in McDowell's words) that 'we must not suppose that receptivity makes an even notionally separable contribution to its co-operation with spontaneity' (p. 51). On the contrary, according to Evans, perception has its role in an 'informational system' whose content is strictly 'non-conceptual' and is hence available only in the form of data for higher-level processing, or as 'input to a thinking, concept-applying, and reasoning system' (cited p. 49).[42]
Of course this goes against McDowell's entire line of argument concerning the constitutive role of Kantian 'spontaneity' in even our most basic acts of perceptual or experiential grasp. 'In Evans's account of experience', he writes,
receptivity figures in the guise of the perceptual element of the informational system, and his idea is that it produces its content-bearing states independently of any operations of spontaneity. It is true that the content-bearing states that result count as experience, in the somewhat Kantian restricted sense that Evans employs, only by virtue of the fact that they are available to spontaneity. But spontaneity does not enter into determining their content. So the independent operations of the informational system figure in Evans's account as a separable contribution made by receptivity to its co-operation with spontaneity. (p. 51)
Now in a sense this criticism is well taken and pinpoints the chief problem not only with Evans's account but with every version of the scheme/content dualism, whatever the various refinements of detail or the number of mediating levels envisaged between raw sensory 'inputs' and full-fledged conceptual knowledge. It is especially a problem, no doubt, for those thinkers in the wake of logical empiricism (Quine, Davidson, and Evans among them) who have inherited this sharply dualist way of thinking and expended much effort in trying to think their way beyond it. But it is also a problem for Kant and – even more – for McDowell's version of Kant as a philosopher who managed to signpost that route despite certain prevalent misreadings. Thus what McDowell calls the 'somewhat restricted' construal of Kant (more specifically, Kant's idea of 'experience') to be found in Evans's work is in fact just the construal invited by those passages in the First Critique where Kant defines knowledge (or understanding) as the capacity to bring sensuous intuitions under adequate concepts. In this respect Evans is true enough to Kant, as likewise when he attempts to fill the problematic gap between those disparate orders of 'experience' by multiplying intermediate stages, themselves just as problematic since required to perform the same (impossible) task of reconciliation. McDowell thinks to halt this vicious regress at source by switching attention from Kant's dualist talk of intuitions and concepts to his far more promising idea of the reciprocal interplay between 'spontaneity' and 'receptivity'. But here also – as I have argued above – the proposed synthesis either falls apart into just another, somewhat more roundabout version of the same dualism, or veers toward regions of speculative thought (such as those first explored by Fichte and Schelling) which are scarcely what McDowell has in mind.
Of course this leaves a third alternative, namely the idea that these issues might be resolved by some combination of Aristotle with Wittgenstein and Gadamer, or a broadly communitarian approach which incorporates elements of practical wisdom with an appeal to 'language-games' or 'forms of 'life' as the furthest one can get in the way of justificatory or evaluative grounds. However – as I have said – this idea breaks down when it comes to explaining how any such approach could possibly account for the challenge to established (traditional) ways of thought posed by radically new scientific theories or even by changes to our customary modes of ethical or socio-political thought. McDowell implicitly admits the problem when he refers – no doubt via Quine's 'Two Dogmas' – to Neurath's famous metaphor of the ship which has to be refurbished plank-by-plank while afloat rather than given a thorough overhaul in dry dock.[43] For Quine this metaphor indicates the need to think of science in holistic terms as a total 'fabric' or 'web' of beliefs where no single item stands or falls on a putative 'crucial experiment', but where pragmatic adjustments can always be made at some point between the logical 'centre' and the observational 'periphery' so as to conserve this or that cherished item. Thus one might, for instance, plead hallucination in order to discount a recalcitrant empirical result or even – at the limit – opt for revising the logical 'laws of thought' so as to accommodate anomalous data such as wave-particle superposition or quantum nonlocality. Or again, as allowed for by the Duhem-Quine thesis, one might always adduce the presence of various implicit 'auxiliary hypotheses' which could be taken to have played some covert role in the formulation of the theory.[44] The trouble with this is that it fails to explain how theory-change could ever come about given the inherently conservative bias – the pragmatic inclination to resist radical change – which Quine identifies as the chief psychological factor in deciding which beliefs to hang onto and which to adjust under pressure from 'recalcitrant' results.[45]
McDowell brings in this metaphor from Neurath/Quine at just the point where he is discussing certain problems with Aristotelian ethics. Thus 'ethical thinking is under a standing obligation to reflect about and criticize the standards by which, at any time, it takes itself to be governed' (p. 81). So much for the familiar Kantian (deontological) objection to any theory of ethics, such as Aristotle's, which rests too comfortably on received ideas of civic or communal virtue. All the same, though Aristotle 'may be less than duly sensitive to [this] obligation in the case of ethics', still it is 'implicit in the very idea of a shaping of the intellect, and that is what "practical wisdom" is' (ibid). And again:
it is a key point that for such reflective criticism, the appropriate image is Neurath's, in which a sailor overhauls his ship while it is afloat. This does not mean that such reflection cannot be radical. One can find oneself called on to jettison parts of one's inherited ways of thinking; and, though this is harder to place in Neurath's image, weaknesses that reflection discloses in inherited ways of thinking can dictate the formation of new concepts and conceptions. But the essential thing is that one can reflect only from the midst of the way of thinking one is reflecting about. (p. 81)
It may seem perverse to pile such a weight of significance onto McDowell's (or, for that matter, Quine's and Neurath's) choice of illustrative metaphor. Still it does have a definite load-bearing role in his and their theories which helps to explain why McDowell goes to such literalistic lengths to spell out its relevance here. However the metaphor tends to kick back by refusing to perform as required. Thus McDowell has to concede – albeit in parenthesis – that his 'radical' reading is one that is 'harder to place in Neurath's image', since the whole point of the ship-metaphor (as understood by Quine also) is to suggest a more gradualist or piecemeal conception of scientific theory-change.
The same can be said more generally of that whole hermeneutic/communitarian approach that McDowell sees as a promising escape-route from the various residual problems with Kant, but which in fact merely shelves them in favour of a theory (summed up in his phrase 'second nature') with deeply conservative implications.[46] This comes out in a passage where McDowell remarks (again) on the need for 'reflective thinking', for 'standards of self-scrutiny', as a result of which certain traditional values or beliefs may come to strike us as 'hav[ing] hitherto unnoticed defects, such as parochialism or reliance on bad prejudice' (p. 81). He then appends a footnote commenting that '»bad prejudice« is not a pleonasm', since (according to Gadamer) 'prejudice, so far from being necessarily a bad thing, is a condition for understanding' (p. 81n). Now of course there is a sense of the word 'prejudice' – roughly speaking, 'the background of implicit normative assumptions presupposed by every act of thought or judgement' – according to which this claim is necessarily (almost trivially) true. On the other hand it is often pushed much harder by philosophers – e.g., some Wittgensteinians and hermeneutic theorists like Gadamer – who adduce it in support of a traditionalist case for setting sharp limits to the scope and remit of critical-evaluative thought.[47]
McDowell is well aware of this and indeed takes issue at one point with what he sees as a mistaken reading of Wittgenstein, one that interprets the 'space of reasons' in a purely 'communitarian or »social-pragmatist« style'. After all, he protests,
[i]f there is nothing to the normative structure within which meaning comes into view except, say, acceptances and rejections of bits of behaviour by the community at large, then how things are – how things can be said to be with a correctness that must partly consist in being faithful to the meanings one would exploit if one said they are thus and so – cannot be independent of the community's ratifying the judgement that they are thus and so. (p. 93)
I agree with McDowell that the consequence of such thinking is 'intolerable' since it opens the way to an outlook of fully-fledged cultural relativism with regard to ethics, the physical sciences, historical enquiry, and everything else. On the other hand I cannot agree that this is to get Wittgenstein wrong since it ignores 'his own conception of what is to be done in philosophy, his "quietism", his rejection of any constructive or doctrinal ambitions' (ibidem). As McDowell sees it the result of such 'quietism' must be to leave our everyday (non-philosophical) beliefs and values firmly in place, among them what he calls our 'commonsense conception of the objectivity of the world, the reality that our command of meaning enables us to think and talk about' (ibidem). I have suggested already that realism of this kind – belief in the reality of an objective and (largely) mind-independent world – is the default attitude of most people and even of most philosophers in their time off from epistemological problems. What is not so clear is why McDowell should think that Wittgenstein's 'quietism' – his desire that philosophy should 'leave everything as it is' – comes out on the side of that commonsense-realist view. Quite the opposite, in fact: more often it conduces to a cultural-relativist outlook in which the truth-value of various beliefs is thought to be determined not by the way things stand in reality, but rather by their role in some given 'language-game' or communal 'form of life'. Indeed Wittgenstein is the chief philosophical source for a good many current varieties of anti-realism, from the so-called 'strong programme' in sociology of knowledge to the more technical version to be found in the writings of Michael Dummett.[48]
Later on McDowell has some pertinent thoughts about the claim by some exegetes – Bernard Williams and Jonathan Lear among them – that Wittgenstein's philosophy retains a certain element of 'vestigial transcendentalism', that is to say, a residually Kantian appeal to the conditions of possibility for knowledge and experience.[49] In their view, this should be enough to placate any worries concerning its supposed relativist implications while not giving way to the kind of 'full-blown transcendental idealism' that would offer reassurance but at too great a cost, namely 'by affirming . . . that we cannot be fundamentally wrong about the world we think about, since it is constituted by us' (p. 159). Thus it would simply be a matter - in Wittgenstein's phrase – of 'how we go on' in some given situation – talking, reasoning, acting, behaving, evaluating evidence, etc. – quite apart from any ultimate (transcendentally warranted) 'conditions' for knowledge and experience in general. However McDowell is not happy with this scaled-down compromise solution since it still carries an excess weight of superfluous philosophical baggage. After all, what betrays Kant into a form of 'idealist' thinking is precisely that 'the constituting of this harmony between world and mind is supposed to be a transcendental operation of mind: not, of course, the empirical mind, which is in constituted harmony with the world, but an off-stage transcendental mind' (p. 159). Where the Wittgensteinian account so radically differs – according to McDowell – is in finding no room for such pointless appeals to an a priori order of intelligibility behind or beyond the straightforward self-evidence of everyday commonsense realism. Thus 'we should be struck by the thought that there is nothing in Wittgenstein's picture to do the constituting of the harmony', in which case – happily – 'the appearance of a piece of transcendental philosophy is hard to maintain' (ibidem).
This seems to me a massive evasion of the central problem and a sign that McDowell is no further on toward resolving the epistemological and ethical issues than those Wittgensteinians who placidly endorse the 'form-of-life' argument as a welcome release from all our philosophic perplexities. In the end he cannot have it both ways, playing off the claims of Kantian critical-evaluative thought against those of Wittgensteinian 'quietism' while hoping to steer a middle course between them which involves no damaging compromise on either side. His book is important – or so it seems to me – more for these symptomatic blind-spots than for anything it offers by way of philosophical (or indeed 'post-philosophical') solutions. Thus it shows very clearly how philosophy is condemned to run into the same dilemmas over and again if it loses touch with some significant chapter of its own formative history. In McDowell's case – as in so much recent Anglophone debate – it is a complicated chapter with branching plots that divide after Kant and then go their own way with just occasional glimpses across or efforts to mend the rift. The main factors here are (1) the sharp distinction standardly drawn between 'genuine' (analytic) philosophy and 'mere' second-order historical enquiry into its sources, development, or alternative paths not taken; (2) the related analytic mistrust of 'continental' epistemology as a species of thinly disguised psychologism; and (3) the belief that any lessons to be learned from Kant have to do with those matters that find a place within the current analytic prospectus and most definitely not with those aspects of his thought that have preoccupied philosophers in the line of descent from Fichte, Schelling, and Hegel to Husserl, Heidegger, and (God forbid!) Derrida. In other words it is the upshot of that same old split between, on the one hand, a school of thought that takes its lead from the chiefly logico-semantic issues raised in Kant's 'Transcendental Analytic' and, on the other, a school whose preferred starting-point is the 'Transcendental Aesthetic', along with all its deeply problematic claims regarding the modalities of experience, imagination, knowledge, and judgement.
As I have said there are signs that this prejudicial mindset is beginning to lose its grip, as for instance in Dummett's (albeit rather grudging) attempt to rehabilitate Husserl, and likewise in McDowell's return to Kant – or a suitably naturalised version of Kant – as pointing a way forward from the post-Quine-and-Davidson impasse.[50] But in both cases, McDowell especially, this renewed sense of openness to 'continental' thought goes along with a persistent narrowing of focus that still excludes pretty much everything outside the chosen few texts (or the select few passages) for redemptive treatment. Thus Dummett sticks closely to the points of convergence/divergence between Frege and Husserl, while McDowell considers only those 'problems from Kant' (and those partial or hinted-at solutions) which lend themselves to adequate re-statement in the idiom of present-day analytic thought. What neither seems able to envisage is the idea that certain deep-laid dilemmas – notably those handed down by Logical Empiricism – might have received an altogether more adequate and resourceful treatment by certain philosophers in the 'other' (post-Kantian continental) tradition.
This emerges most plainly in those passages of McDowell's book where he admonishes Kant for an overly formal (or 'transcendental') approach to such issues yet fails to acknowledge how subsequent thinkers – Husserl chief among them – not only criticised Kant on the same grounds but carried their criticism to a more advanced stage of constructive philosophical engagement.[51] Hence the impression he often gives of venturing like Robinson Crusoe into unknown territory where, as it turns out, somebody has gone before and left unmistakable footprints. One such passage has to do with Kant's idea of the 'transcendental unity of apperception', that is to say, his notion of 'the merely formal persistence of the I, in the "I think" that can "accompany all my representations"' (p. 103). McDowell's point – in line with his naturalising approach – is that this mysterious je ne sais quoi 'had better be only an abstraction from the ordinary substantial persistence of the living subject of experience', and had therefore better not be 'something free-standing', something 'which we might hope to build on in reconstructing the persistence of the ordinary self' (ibidem). Now of course this is precisely Kant's own point when he objects to Descartes's fallacious argument from the 'merely formal' condition of possibility for thought and self-consciousness in general to a substantive conclusion regarding the cogito and its necessary (rationally self-evident) mode of existence.[52] Or rather: McDowell agrees with Kant that it had better be 'only an abstraction', but differs in counting that abstraction just a piece of otiose 'transcendental' machinery, an item which can and should be dropped in favour of the 'ordinary self' and its 'substantial persistence [as] the living subject of experience'. However there is nothing in McDowell's book to suggest that later philosophers – notably Husserl – took up this issue where Kant left off and pursued his critique of Cartesian 'rational psychology' to very different philosophical ends.[53] That is, Husserl went further than anyone – certainly much further than McDowell – in seeking to explain how our various modalities of experience, knowledge, and judgement could be both apodictically grounded in constitutive acts of consciousness and at the same time inseparably tied to our existence as physically embodied creatures in whom there could exist no Cartesian divorce between 'mind' and 'world'. It is a measure of the tunnel-vision syndrome afflicting so much recent analytic philosophy that Husserl's entire life's work on just this topic should rate not a mention from McDowell while he devotes many pages of strained exegesis to passages in Kant that often scarcely serve his argumentative purpose.
I shall cite just one further example – a striking instance of the kind – in order to illustrate this curious feature of McDowell's book. 'Why', he asks in relation to Kant, 'can there not be a free-standing idea of formal subjective continuity?'.
The answer is this: the idea of a subjectively continuous series of states or occurrences in which conceptual capacities are implicated in sensibility – or, more generally, the idea of a subjectively continuous series of exercises of conceptual capacities of any kind, that is, the idea of a subjectively continuous series of 'representations', as Kant would say – is just the idea of a singled out tract of a life. The idea of a subjectively continuous series of 'representations' could no more stand alone, independent of the idea of a living thing in whose life these events occur, than could the idea of a series of digestive events without its appropriate kind of continuity. But in the absence of a serious notion of second nature, this exploitation of the concept of life, which is a quintessentially natural phenomenon, to make sense of a unity within the domain of spontaneity, which by Kant's lights has to be non-natural, is not within Kant's grasp. (pp. 103-4)
From one point of view – the 'naturalised' perspective that McDowell acquires from Aristotle, Hegel, Wittgenstein, and Gadamer – this passage represents an acute piece of critical commentary and a pinpoint diagnosis of Kant's failure to make good on his transcendental claims. From a different, more 'continental' (and especially Husserlian) standpoint it looks like a typical instance of what goes wrong when philosophers renounce some unattainable goal – such as Kant's on McDowell's understanding of it – only to flip straight over to the opposite extreme, that of a naturalism wholly divorced from the realm of rational or critical-evaluative thought. The point comes out with almost comical force when McDowell compares the indissoluble nexus between concepts (or Kantian 'representations') and the 'living thing' in whose experience they figure with that between 'digestive events' and the organs or bodily processes through which they occur. This digestive metaphor seems to have been sparked by McDowell's reference to the 'tract of a life' as the only context within which thought and experience can make sense or be 'singled out' as possessing the kind of intelligible continuity which Kant misguidedly sought to establish through his transcendental mode of argument. Certainly it makes more vivid his claim that 'life' is 'a quintessentially natural phenomenon', a claim whose literalness may seem compromised by McDowell's adjacent talk of 'second nature', but which is brought back to earth with a bump by the image of concepts as making their progress through the digestive tract of experience.
My point is not so much to belabour this inept choice of metaphor but rather to emphasise the persistence in McDowell's work of a mind/body dualism which he tries very hard to get over but which at times breaks surface in a strikingly overt or desublimated form. This goes along with his failure to reconcile the various perplexing antinomies (e.g., that of normative or rational-justificatory versus naturalised or causal-explanatory theories of knowledge) that McDowell considers to have skewed the discourse of philosophy from Kant to Davidson and Quine. What he wants to do is put the 'body' back into our conceptual experience of the world so that any such dualism – whether between mind and body or 'concept' and 'intuition' – can be seen as just a product of that lingering Cartesian prejudice which prevented Kant from following through on his own best insights. Thus, according to McDowell,
[i]f we begin with a free-standing notion of an experiential route through objective reality, a temporally extended point of view that might be bodiless so far as the connection between subjectivity and objectivity goes, there seems to be no prospect of building up from there to the notion of a substantial presence in the world. (p. 102)
But there is no more prospect of 'building up' toward this integrated theory of knowledge and thought as aspects of our rationally-justified mode of human being-in-the-world if mind and body continue to be treated in so sharply dichotomous a fashion. Once again McDowell might have learned some useful lessons from Husserl, as well as from later work in the phenomenological tradition, including Merleau-Ponty's extensive writings on the way that modes of bodily-perceptual experience enter into even our most 'abstract' forms of representation or reasoning.[54] Kant has no guidance to offer here – despite McDowell's claims to the contrary – since his thinking is riven by all those persistent dualisms which have proved so vexatious in recent analytic debate. Nor does it help (again pace McDowell) to shift attention from Kant's problematical talk of 'intuitions' and 'concepts' to his apparently more promising idea of the intimate relation between 'spontaneity' and 'receptivity' as powers of mind that necessarily conjoin in every act of thought and judgement. For as we have seen this generates all the same problems, whether pushed to the extremes of metaphysical speculation (as by Kant's Subjective and Objective Idealist successors), or supplied with a naturalising gloss which seeks to hold those tendencies within analytically respectable bounds.
Crossings between the realm of understanding (where the rule is that intuitions must be brought under adequate concepts) and the realm of reason (where no such rule can apply since here it is a matter of regulative ideas which guide the quest for knowledge while themselves transcending the limits of conceptual grasp). Thus Kant spends much of his time in the First Critique pointing out the sorts of confusion that inevitably arise when philosophers overstep those limits, as for instance by seeking to attain determinate knowledge of matters – such as noumenal reality, the self, freewill, and other 'supersensible' ideas – whose very nature it is to elude or frustrate the most strenuous efforts of empirical-conceptual understanding.[55] Yet it is also Kant's claim, paradoxically, that these confusions are indeed inevitable and cannot be remedied simply by applying better rules for the conduct of disciplined thought. For it is just his point – in the 'Paralogisms of Pure Reason' – that we should have no incentive to explore the scope and limits of human knowledge were it not for this innate tendency to raise questions that tempt understanding into regions beyond its power to survey or comprehend.[56] McDowell wants to keep well clear of those regions since they offer just the kind of seductive metaphysical or 'transcendental' temptation which, in his view, led Kant sadly astray and should therefore be avoided by anyone hoping to redeem Kant's project for present-day constructive purposes. However – as I have argued – McDowell's own attempt at a revisionist or naturalised reading is one that reveals all the stresses and strains that Kant so exhaustively diagnosed in those cautionary sections of the First Critique devoted to the various antinomies of pure reason. Or rather, it shows how such antinomies arise when philosophy fails to heed the lesson of its constant proneness to forms of paralogistic reasoning.
In McDowell's case they emerge most sharply when he strives to reconcile a naturalistic approach with a theory of knowledge that retains the Kantian commitment to modes of normative, justificatory, and critical-evaluative thought which stand outside the space of natural (causal-explanatory) description. This is why McDowell is so insistent that '[t]he lack of freedom that is characteristic of merely animal life is not enslavement to the practical as opposed to the theoretical, but enslavement to immediate biological imperatives' (p. 117). For he can then take this claim as justification for the idea that there is another kind of naturalness – the realm of distinctively human 'second nature' – wherein the antinomy can find its solution without any need to follow Kant into dubious regions of thought. Thus McDowell writes, again with reference to Gadamer:
[e]mancipation into the 'free, distanced orientation' brings intentional bodily action onto the scene no less than theoretical activity. The picture of full-fledged subjectivity that is in play here is not a picture of that dubiously intelligible kind of thing, an observer and thinker that does not act in the world it observes and thinks about. (p. 117)
He then goes on to cite early Marx – from the 1844 Paris Manuscripts – on alienated labour and the counterposed need for an active, humanly fulfilling relationship to nature which transcends the 'merely animal' mode of existence imposed by capitalist forces and relations of production.[57] Here again there is a curious sense that McDowell is rediscovering long-forgotten features of some remote intellectual landscape which have actually been explored – in much greater depth and detail – by thinkers belonging to that 'other' (post-Kantian continental) tradition. The passage is worth quoting at length:
Marx complains memorably of a dehumanization of humanity in wage slavery. The part of human life that should be most expressive of humanity, namely, productive activity, is reduced to the condition of merely animal life, the meeting of merely biological needs. And although it is freedom that gives its distinctively human character to human life, wage slavery restricts freedom to the merely animal aspects of what are thus only incidentally human lives . . . Marx sums up his vision of what a properly human life would be in a striking image: without alienation, 'the whole of nature' is 'the inorganic body of man'. We can point up the convergence with Gadamer by glossing the image like this: the world is where a human being lives, where she is at home. (p. 118)
It is unsurprising that Gadamer 'does not note the parallel' given – what McDowell himself fails to note – the strongly traditionalist cast of his hermeneutic project and its lack of room for any critical-emancipatory dimension such as that kept open by Marx and other 'left' theorists (Habermas among them) in the broadly Marxian line of descent.[58] Nor is the distance between them appreciably closed by McDowell's rather anodyne gloss on Marx as saying merely that 'the world is where a human being lives, where she is at home'. For this comes down to the homely Wittgensteinian doctrine that philosophy should leave things as they are, that what makes sense can do so only by the standards (or 'criteria') of some given communal life-form, and hence that we err – begin to talk nonsense by those same communal standards – when we presume to raise issues or to broach critical concerns 'external' to the life-form in question.[59]
In short there is no place for Marx's talk of 'dehumanization', 'wage-slavery', or a 'properly human life' in a discourse which finds human beings 'at home' with received (traditional or currently prevailing) ways of thought. Here again McDowell seems torn between the claims of a critical outlook with its source in Kant's famous dictum Sapere aude ('Have the courage to think for yourself!') and a communitarian doctrine of 'second nature' whose lesson – whether derived from Aristotle, Wittgenstein, or Gadamer – is to counsel an attitude of wise acquiescence in existing values and beliefs. What that doctrine serves to accomplish is the passage from a realm of 'merely animal' (directly biological) needs and imperatives to a realm where their function is in effect taken over by naturalised language-games or forms of life which can always be relied on to conserve what is vital to the continuance of communal tradition. This is not to say – McDowell cautions – 'that a merely animal life is a constant struggle, whereas a distinctively human life is peculiarly easy' (p. 118). After all, '[i]n Marx as in Gadamer, the point is not that a human life is easy but that it is distinctively free. And that is the same as the fact that it is lived in the world, as opposed to coping with an environment' (ibid). But this point is itself perhaps made too easily – with too little sense of the problems entailed by any workable definition of terms such as 'freedom', 'world', and 'environment' – when glossed in a fashion that takes for granted the naturalness of talking in just this way about distinctively 'human' attributes and values. One may agree with McDowell that Kant went off in some unfortunate directions when he sought to provide a transcendental grounding for those same attributes and values. However one may doubt that any naturalising strategy such as McDowell proposes can resolve the dilemmas thrown up by Kant's project while also keeping faith – as he wishes to do – with its rational-evaluative or critical-emancipatory promise.
I have argued that this project miscarries for various reasons. Some of them are essentially problems with Kant that resurface in McDowell's argument, while others are problems that Kant himself addressed yet which continue to exert a vexing grip on the discourse of recent analytic philosophy, especially when thinkers such as Quine and Davidson claim to have naturalised them out of existence. Most of all it is McDowell's very patchy dealing with continental philosophy after Kant – in particular his failure to engage even briefly with Husserl – which accounts for his proneness to fall back into modes of thought that have a long prehistory (and have received more intensive critical treatment) in that 'other' tradition. Nevertheless Mind and World is clearly an important book in so far as it signals the growing awareness among at least some Anglophone philosophers that there is much to be gained by abandoning this manichean view and exploring various paths closed off since the heyday of Logical Empiricism.