Strawson's book is concerned with Hume's views on the mind as they emerge from his account of personal identity in Treatise 1.4.6.Footnote ¹ The book is in three parts. The first is devoted to background remarks about Hume's philosophy more generally in which Strawson reaffirms the sceptical realist interpretation familiar from his previous writings on Hume; the second examines the ‘bundle’ account of mind or self associated with Treatise 1.4.6 and argues that Hume does not in fact identify the mind with a bundle of perceptions; finally, in the third part Strawson takes up the question of the nature of Hume's problem with his account of personal identity as expressed in the Appendix to the Treatise and claims that Hume identifies here an insoluble difficulty with what he has said about the mind itself. The discussion throughout is lively and provocative though at crucial points in the argument Strawson relies on claims about Hume's philosophical position which are at least questionable. (It is also fair to say that Strawson neglects important aspects of the historical context which shed light on Hume's account of the self; in particular, the debate between Clarke and Collins about the nature of the soul).

There is an issue of general importance for our understanding of Hume which arises in the first part of Strawson's discussion. This concerns the nature of Hume's scepticism and its relation to his account of meaning. The gist of what Strawson says here is that while Hume regards propositions about the ultimate nature of reality as being empirically undecidable, he does not reject them as meaningless. In fact, according to Strawson, Hume never questions the idea that something other than our experiences does in fact exist. This, as Strawson notes, might be thought to contradict Hume's view of meaning which apparently requires that any significant idea must derive from experience in the form of an impression. Indeed, this is Hume's ‘first principle’ (T 1.1.1.12) which allows him to say in the Abstract of the Treatise that ‘when he [the author] suspects that any philosophical term has no idea annexed to it …he always asks from what impression is that intended idea derived? And if no impression can be produced, he concludes that the term is altogether insignificant’ (Abstract §7). This strongly suggests that, for Hume, any term for which we are unable to provide a corresponding impression is thereby meaningless. According to Strawson, however, Hume does not claim to exhaust the content of our concepts in this way: rather, he is concerned with how far they are empirically warranted and to that extent employable in philosophy. Strawson therefore disputes those interpretations of Hume that represent him as moving from an epistemological claim to the effect that our knowledge of mind is limited to the content of our experiences, to the corresponding semantic claim that these experiences are all that can be meant by the concept of mind, to the ontological conclusion that there is nothing more to mind than the existence of these experiences.

Strawson further claims that Hume's scepticism commits him to rejecting the above move from an epistemological claim to a metaphysical one. The distinction Strawson has in mind is between the ‘mitigated’ or ‘moderate’ scepticism which Hume endorses at various points in his writings, and the more dogmatic (or 'extravagant') scepticism which he elsewhere condemns. The former, Strawson suggests, allows that physical objects may be supposed to exist independently of our perceptions, or that causal power may exist in addition to the regularities we experience, even if their nature remains unknown to us. Strawson further claims that Hume takes for granted that these things do exist while also holding that the ultimate causes of experience are inexplicable. There are questions here about the nature of Hume's scepticism which take us beyond the particular topic with which Strawson's book is concerned. But it seems fair to say that while the moderate (‘modest’, ‘mitigated’) scepticism which Hume advocates in a number of places (T 1.4.3.10, 1.2.5.26 n.12, EHU 12.4, 25) might leave open the possible existence of things which lie beyond our experience or understanding, it should also discourage us as philosophers from taking for granted that such things do exist. If Hume's moderate scepticism prevents him from claiming to know that experiences are all that exist it also seems clear that it should prevent him from taking for granted as a philosopher the existence of things other than our experiences.

The other issue which arises here concerns the nature of Hume's theory of ideas. Once more, space does not allow a proper consideration of an issue which has been extensively discussed in the debate on the ‘New Hume’ (i.e. Hume understood in accordance with the sceptical realist interpretation). An obvious question which arises here is how Strawson's view of Hume's theory of ideas can be justified in the face of Hume's own appeal to the priority of impressions to ideas in elucidating the meaning of certain key philosophical ideas (for example, space and time, T 1.2.3; causation, T 1.3.2-14; the mind, T 1.4.5.3-4; and the self T 1.4.6.2). There are also places in which Hume appears to be quite unequivocal about the power of this principle to resolve controversies about the meaning of ideas. Apart from the passage in the Abstract of the Treatise where Hume ties the significance of a term to the existence of an impression corresponding to the idea involved, there is a very similar passage in the first Enquiry where the link between term, idea and meaning is again made to depend on the existence of a prior impression (EHU 2.9).

Strawson does adduce further considerations for the claim that Hume accepts that there is more to the mind than its experiences or ‘perceptions’, contrary to the view of Hume as a ‘bundle theorist’ for whom the mind just is a collection or succession of experiences. According to Strawson, we may prove this view to be mistaken in light of the fact that Hume holds both that ‘the essence of mind [is] equally unknown to us with that of external bodies’ (T Intro. §8) and also that ‘the perceptions of the mind are perfectly known’ (T 2.2.6.2). It seems doubtful whether so much weight can be placed on these two quotations. Strawson returns to the former throughout his book. He finds a comparison here with Locke's concept of real essence as something unknown to us, of which we have no real idea, but to which a word like ‘gold’ carries a ‘tacit reference’. Hume's references in the Treatise to essence, however, surely fail to amount to a theory which bears comparison with Locke's distinction between nominal and real essence. In the context of the Introduction Hume's remark about the unknown essence of mind arises from the parallel he draws between his projected ‘science of man’ and the ‘natural philosophy’ of scientists like Newton. In each case our notion of the powers and qualities involved can be formed only on the basis of experience and this prevents us from arriving at the ‘ultimate principles’ which underlie the particular effects we are able to observe (Introduction §11). This anticipates Hume's reference in T 1.1.5.6 to the association of ideas as a ‘kind of ATTRACTION’ whose role in the mental world may be compared with that of gravity in the natural world. Hume says here that the causes of attraction are ‘mostly unknown, and must be resolv'd into original qualities of human nature, which I pretend not to explain’. But this is scarcely tantamount to recognising that the mind itself must be something more than its perceptions. It is also difficult to see how much we should read into Hume's reference to our perceptions themselves being ‘perfectly known’. This presumably cannot be taken to apply to the relational qualities of our perceptions; yet it is the problem of saying more about the principles which underlie these relations that Hume has in mind when he refers to the unknown essence of mind.

In Part 2 of his book Strawson turns to Hume's conception of the mind as a bundle of perceptions and suggests that Hume comes to recognise that this view of the mind is inconsistent with his philosophical commitments and presuppositions. Strawson takes Hume's remark in his first Enquiry that ‘it is evident that there is a principle of connexion between the different thoughts or ideas of the mind’ (3.1) as confirmation of this understanding of his final position in the Treatise. It is hard to see the justification for this since it is evident from the context that Hume's ‘principle of connexion’ is an allusion to the association of ideas and not a reference to some ‘real’ underlying connection unavailable to experience. A particular target of Strawson's is the ‘no ownership’ interpretation which ascribes to Hume the view that perceptions (or experiences) can occur in the absence of a separate subject. Given the apparent incoherence of the idea that experiences may occur independently of a subject to which they belong, Strawson draws the conclusion that whatever may be Hume's target in T 1.4.6 it cannot be the belief in subjects of experience. But is there anything here for the bundle theorist to dispute? Any of our perceptions or experiences belongs to an experiencer in the form of a bundle or system of such experiences. The question is whether, as Strawson is claiming, the subject of experience must be ontologically distinct from the experiences themselves and, additionally, whether Hume himself accepts this requirement.

According to Strawson, when Hume claims that experiences (or ‘perceptions’) may exist separately (T 1.4.5.5) he means only that if experiences do exist separately then there is a separate subject for any such experience. This interpretation of Hume finds little support in the text of the Treatise (and, indeed, appears to be contradicted by Hume's apparent acceptance of the possibility that a perception may exist separately from a mind – T 1.4.2.40). Hume makes it clear that whatever we may be able to conceive so far as the separate existence of our perceptions is concerned, there are ‘experiments’ which enable us to infer that all or our perceptions in fact depend on our physiology (T 1.4.2.45). This opens the possibility, at least, that the subject of experience is an embodied self though one constituted by its perceptions rather than some separately existing soul or substance. As further evidence against the view of Hume as an ontological bundle theorist Strawson refers to Hume's view that ideas ‘remain’ in the mind after the corresponding impressions cease even if we happen not to be consciously attending to them. How can ideas continue to exist in the mind if all perceptions are conscious occurrences? One possible reply is that talk of ideas as remaining in the mind should be understood as referring to mental dispositions and ones which depend on the existence of corresponding non-dispositional properties. Hume refers to ideas which are not ‘in fact present to the mind, but only in power’ (T 1.1.7.7); and his subsequent account of the part played by brain traces in the relations of ideas (T 1.2.5.20) identifies a categorical basis for this kind of ‘power’: one which is consistent with his empiricism and the avoidance of the ‘occult qualities’ associated with the doctrine of substance.

When Strawson turns to the Appendix to the Treatise he notes that Hume's problem there is often taken to concern his account of how we come to believe in a persisting self – one in which he appeals to the principles of association. This, Strawson claims, cannot be Hume's problem since he pronounces himself satisfied with this account. Strawson presumably has in mind Hume's comment on his summary in the Appendix of the argument by which he had arrived in T 1.4.6 at his conclusion that ‘the thought alone finds personal identity’, namely, ‘The present philosophy so far has a promising aspect’ (Appendix 20). But everything depends on how we understand Hume's remark at the beginning of this paragraph about his account of personal identity being ‘very defective’ in respect of his attempt ‘to explain the principle of connexion, which binds them [particular perceptions] together, and makes us attribute to them a real simplicity and identity’. This seems to be a reference to Hume's attempt to explain our belief in the identity of the mind on the basis of the associating principles among our perceptions unless we read more into the phrase ‘principle of connexion’ than is conveyed by ‘associating qualities’. It is striking that, contrary to what we should expect on Strawson's interpretation, Hume does not appear to retract in the Appendix his account in T 1.4.6 of the mind or self as a bundle or system of perceptions. On the contrary, at Appendix 15 he repeats the claim that in reflecting on himself he perceives nothing but perceptions and concludes ‘Tis the composition of these, therefore, which forms the self'. Hume goes on to suggest that the annihilation which some people believe to follow death and which destroys the self is nothing but the extinction of all our perceptions; and, he adds, ‘These therefore must be the same with self, since the one cannot survive the other’ (17). It is in respect of this view of the self that Hume appears to find that he is ‘attended with sufficient evidence’ (20).

There are several further points that might be made in favour of the view that Hume is not expressing despair in the Appendix about his earlier account of the self. For example, in Part 4 of the Dialogues Concerning Natural Religion we find the human soul being described as ‘A composition of various faculties, passions, sentiments, ideas’. It is surprising, at least, that if Hume really had discovered a fatal flaw in his account of personal identity he should then reintroduce this conception of the mind as one of the few matters which is not in dispute among the protagonists in the Dialogues. Strawson suggests that the central problem for Hume is that on his account of the self in T 1.4.6 our perceptions are just bundles with no real unity. But Hume's reference in T 1.4.6.19 to the mind as a system of different perceptions already indicates that the mind or self is more than just a bundle – rather, our different perceptions influence each other and are able to give us a concern for our past as well as our future pains and pleasures. In fact, Hume anticipates here the further aspect of personal identity to be dealt with in Book 2 of the Treatise (and to which he alludes at T 1.4.6.5). He provides there a complex account of the passions in terms of relations among our perceptions in a way which reflects the view of the mind in T 1.4.6. Since the gist of Treatise Book 2 is contained in the later Dissertation on the Passions it is difficult to accept that Hume wishes for a greater unity to the mind than this account of the relations among its perceptions would provide.

It is an indication of the value of Strawson's book that there are many other aspects of his discussion of Hume which might be pursued including his subtle treatment of the phenomenological aspects of experience and their relation to self-awareness. Nor can it be denied that Strawson raises important difficulties for the ascription to Hume of a broadly reductionist view of the mind or self which is retained both in the Appendix and also in his later writings. There are, however, various critical points at which Strawson's reading of the relevant texts may be disputed though this provides a debate to which many more contributions will be made thanks to the stimulus provided by his book.