Megan Blomfield: What Was Hume’s Problem about Personal Identity in the Appendix?


What Was Hume’s Problem about Personal Identity in the Appendix?

Megan Blomfield

Hume’s theory of personal identity is the one thing in which he confesses to having made “considerable mistakes” in the Appendix to the Treatise (App. 1). There is little consensus, however, on what exactly was the source of his discontent. There is not time in this paper to discuss the multitude of opinions that have been given on the Appendix, but in what follows I will explain what I think was troubling Hume. I think that Hume finds that his explanation of how we attribute simplicity and identity to our minds fails once we are aware that the mind is a bundle of all our perceptions. I will then briefly discuss Pitson’s criticism of this sort of interpretation, concluding that Pitson’s objections are unsuccessful.

Hume’s bundle theory of the mind

Hume’s theory of mind essentially states that the mind is a bundle of perceptions. It is important, therefore, to understand exactly what perceptions are according to Hume. Perceptions are all that is present to the mind, whether sensing, thinking, reflecting or “actuated with passions” (Ab. 5). They are separated into impressions and ideas, which differ in terms of degree of “force and liveliness” (T, Impressions are those perceptions that have most “force and violence”, whereas ideas are the fainter perceptions we have whenever we reflect on other perceptions. I think that Hume seems to take an action theory of perceptions, in which they have an awareness of their objects built into them – a perception isn’t an image, rather it is an awareness of an image. Since a perception of a perception[1] is an awareness of a perception, to become aware of a perception is for an idea of it to occur. You are not aware of your perceptions unless you reflect upon them.

Hume delivers his account of personal identity in Treatise 1.4.6. In the previous section, he has rejected the Ancient Philosophy’s posit of a mental substance, or soul, in which the perceptions of the mind are said to inhere, and has argued that, in fact, each perception fits the definition of substance as “something which may exist by itself”. Each perception is different, and therefore distinguishable, from every other perception (and everything else in the universe). The imagination can therefore separate each perception from everything else and conceive it as separately existent. Hume concludes from this that each perception is in need of nothing at all - nothing such as a mental substance – to support its existence, but “may exist separately” from everything else in the universe (T,

In 1.4.6, Hume begins by challenging philosophers who believe in the ubiquity of self-consciousness: who hold that we have a constant awareness of the existence of a perfectly simple and identical self. Hume asserts that, when “I enter most intimately into what I call myself(T, he can never catch himself without a perception, and never observes anything but perceptions which are in permanent flux. There is no constant and invariable impression that would suffice to give rise to the idea of a simple and identical self. Hume argues that really we have no notion of self[2] distinct from our particular perceptions, and that the mind is therefore nothing but a bundle of perceptions, just like bodies are nothing but bundles of qualities. “There is no simplicity in [the mind] at one time, nor identity in different; whatever natural propension we might have to imagine that simplicity and identity” (T,

What remains, therefore, is for Hume to explain this “natural propension” to ascribe simplicity and identity to the different perceptions which constitute the mind; how it is that we form this idea of the simplicity and identity of a self, given that it cannot be derived from a simple impression (T, Hume attempts to do this analogously to the way he explains the origin of our mistaken idea of the distinct and continued existence of bodies in 1.4.2 – the idea results from the operation of our imagination upon our perceptions. Identity, for Hume, is the idea of an object that is uninterrupted and invariable over time. We are wont, however, to confuse the idea of identity with that of a succession of different objects connected by a close relation (T,, because our imagination has a natural and unavoidable tendency to attribute identity to the closely related objects. Hume thinks this is what takes place whenever we ascribe identity to an object which isn’t in fact unvaried and uninterrupted (T,

Therefore, since our perceptions are in reality distinct existences, we can only attribute the quality of identity to them because when we reflect upon them our ideas of them are given a “union in the imagination” (T, The ideas are given such a union by virtue of the universal principles that guide the imagination: the associative principles of resemblance and causation (Hume rejects contiguity). These principles give rise to passive general patterns that the mind wanders in accordance with, producing an easy transition of ideas. There are relations of resemblance between our past perceptions and the images we form of them in memory (T,, and relations of causation between our perceptions because they constitute the mind as a system in which they produce and destroy each other – impressions give rise to ideas which in turn give rise to further impressions (T, Our notion of personal identity therefore arises solely from the “smooth and uninterrupted progress of the thought along a train of…ideas”, due to the relations of causation and resemblance between the perceptions that the ideas are reflective perceptions of. This smooth and uninterrupted progress of the thought feels almost the same as perceiving one unvaried and continuing object – it feels as though the perceptions are connected by the relation of identity. This is why we attribute identity to our perceptions, even though it is really a “mistake” to do so (T,

To justify mistaken attributions of identity, we often invent some “new and unintelligible principle” that connects the objects together (T, This can be seen in the Ancient Philosophy’s posit of an underlying substance to connect the bundle of qualities that constitutes an external body, as discussed in 1.4.3. This is what Hume thinks is taking place when philosophers posit the existence of a “soul” or underlying mental “substance”. Since, as stated in, each perception is a distinct existence which can be considered separately from all other perceptions without contradiction or “absurdity” (T,, we never actually perceive a “real connexion” between them[3] (T, The positing by philosophers of an underlying “self” or “substance” which connects our perceptions is therefore seen to be a “fiction”.

The Appendix

In the Appendix, Hume refers back to his account of personal identity with unease. He now claims that he is caught in a “labyrinth” that leaves him the victim of sceptical doubts (App.10). “In short”, he says, the problem results from two principles that he claims unable to reject or render consistent (App.21):

P1 “All our distinct perceptions are distinct existences.”

P2 “The mind never perceives any real connexion among distinct existences.”

These principles, however, are not inconsistent with each other[4]. In fact, together they entail that:

P3 The mind never perceives any real connection among our distinct perceptions.

This strange claim of Hume’s, and his failure to explain clearly what was troubling him in the Appendix, has led to many interpretations of what his worry could have been. I think that it is important to look very closely at what else Hume says in the Appendix to try and figure out what his problem really was.

Just previously (App.20), Hume has stated that P1 entails that:

P4 Our perceptions “form a whole only by being connected together.”

It is P3 and P4 which I think give rise to the contradiction that Hume is worried about. Our perceptions do “form a whole” – they form a bundle to which we attribute simplicity and identity – so they must somehow be “connected together”. However, given P3, “no [real] connexions among [our distinct perceptions] are ever discoverable by human understanding”. Hume therefore, in his account of personal identity, had to find another way to explain how it is that our perceptions are “connected together”, involving the way in which we “feel a connexion” between our ideas of them. I think that Hume has discovered that the very explanation he gives to this effect defeats itself. In what follows I will explain how.

In his discussion in the Appendix, Hume begins by recounting his theory of personal identity with approval. He restates his denial of the “strict and proper identity and simplicity of a self or thinking being” (App. 10), reiterating arguments to support both his claim that all perceptions are distinct existences, and that the self is nothing but a bundle of these perceptions. He clearly retains his belief that the attribution of simplicity and identity to the perceptions in the bundle that constitutes the mind is a mistake.

It is only now, “having thus loosen’d all our particular perceptions”, that Hume announces his difficulty: “when I proceed to explain the principle of connection89, which binds them together, and makes us attribute to them a real simplicity and identity; I am sensible that my account is very defective” (App. 20). Hume here (89) refers us back to the point around where he claims that the associative principles of resemblance and causation must be the origin of our mistaken idea of a simple and identical self, and begins to explain how this could be the case. It appears to be something to do with the associative principles or the task he has assigned them that is therefore the problem. Now, I don’t think that it is the associative principles themselves that Hume is dissatisfied with. In the Abstract to the Treatise he insists that the principles of association are what, if anything, “can entitle the author to so glorious a name as that of an inventor” (Ab. 35). The principles of association are crucial to his entire project, and Hume’s use of them “enters into most of his philosophy”. If there was a problem with the principles themselves, then it seems unlikely that Hume would have confined his worries in the Appendix to his theory of personal identity. It must rather be the application of the principles to this particular problem that is “defective”.

Hume then repeats with satisfaction his claim that “the thought alone finds personal identity, when reflecting on the train of past perceptions, that compose a mind, the ideas of them are felt to be connected together, and naturally introduce each other”, asserting that “the present philosophy has…so far a promising aspect” (App. 20). However, he continues, “all my hopes vanish, when I come to explain the principles, that unite our successive perceptions in our thought or consciousness. I cannot discover any theory, that gives me satisfaction on this head”. Consciousness, Hume has just described as a “reflected thought or perception”; and it is thought which, he has just claimed, “alone finds personal identity, when reflecting on the train of past perceptions”. So, given Hume is worried about successive perceptions “in our thought or consciousness”, I think that his problem is clearly something to do with the reflective perceptions that we have when our imagination forms its mistaken attributions of identity[5].

This is how I think reflective perceptions of thought cause a problem for Hume: At, around the point which Hume refers us back to, he explains how it is that relations of resemblance between perceptions contribute to our attribution of identity to them, by talking about what we would observe if we were to look at the bundle of perceptions which constitutes the mind of another person. Hume argues that the relations of resemblance that are bound to hold between the person’s memory perceptions and the lower order perceptions which they are images of, will lead to the easy transition of our imagination along our ideas when we reflect on the other person’s perceptions. He then claims that “the case is the same whether we consider ourselves or others”. The case is not the same, however; for when we reflect upon the bundle of perceptions that constitute our own minds, we add perceptions to the bundle. To reflect – in thought or consciousness – on our own perceptions is to form ideas of them, and these ideas, which are also perceptions, are also members of the bundle that constitutes our mind.

Now, this isn’t a problem for explaining the vulgar’s mistaken idea of the simplicity and identity of their minds, because the vulgar aren’t aware of themselves as bundles of perceptions. The vulgar simply attribute simplicity and identity to their perceptions when they reflect on them in memory, without being aware of the reflective perceptions to which they are giving a union in their imagination. When philosophising, however, we become aware that the mind to which we attribute simplicity and identity is a bundle of perceptions that is “[the same with] all particular perceptions; [including those of] thought” (App.17, my italics). We also become aware of the existence of the ideas we form of our perceptions when reflecting on them to make our vulgar attributions of identity. We must therefore find a way of explaining how it is that we attribute simplicity and identity to these ideas – or reflective perceptions – and this is what Hume’s account involving the associative principles is unable to do.

The mind cannot perceive any real connection between these reflective perceptions and the other perceptions that constitute our minds. Therefore, since they are nevertheless a part of the bundle to which we attribute simplicity and identity, it must feel a connection between ideas of them. But this involves the mind reflecting upon these reflective perceptions, resulting in ideas or perceptions of an even higher order whose presence in the bundle we must also somehow account for. There will always be some remainder, some perceptions whose presence in the bundle Hume cannot account for. In giving his explanation of how the bundle of perceptions which constitutes a mind forms a whole, he defeats that very explanation – for the bundle he is trying to account for is no longer the whole bundle. This, I think, is Hume’s problem in the Appendix.

This interpretation of Hume’s worry explains why he cannot render the two principles consistent (since they necessitate his development of an explanation which defeats itself), and also explains why he thinks that the problem could be solved if (App.21):

S1 Our perceptions “[inhered] in something simple and individual.”

S2 “The mind [perceived] any real connection” among our perceptions.

If either of these were the case, then Hume would not have to find an explanation for how our perceptions form a connected whole which involved feeling connections between ideas of them. It is this explanation that leads Hume into a labyrinth. This also explains why it is that the problem Hume discusses in the Appendix seems to be fairly isolated. Normally, our mistaken idea of personal identity arises easily from our natural reflection on past and present perceptions in memory and thought. It is only when investigating this process, and thinking about the mistaken idea of the simplicity and identity of the mind given an awareness of it as a bundle of perceptions that Hume encounters a contradiction.

Pitson’s problem with this sort of interpretation

My view of the problem in the Appendix is very similar to the interpretation given by Ainslie (Ainslie, 2001). Pitson argues against Ainslie’s view, which he thinks is only plausible “given that one accepts the crucial claim that the beliefs for which Hume is trying to account in T, 1.4.6 are the product of philosophical reflection” (Pitson, 76-77). The problem of reflective perceptions only arises for “philosophical perspectives” on the belief in the simplicity and identity of the self as a bundle of perceptions. The vulgar do not reflect on themselves as bundles of perceptions, so Hume does not need to account for the inclusion of reflective perceptions in their mistaken ideas of a simple and identical mind. Pitson therefore concludes that this cannot be the contradiction Hume is referring to, because his discussion of personal identity in 1.4.6 and the Appendix concerns only the beliefs of the vulgar (Ibid., 79).

Pitson also argues that the appropriation into the bundle of reflective perceptions would only pose a problem for our attributions of synchronic unity, and since Hume refers in the Appendix to the trouble in accounting for the identity ascribed to successive perceptions in thought or consciousness (App.20), this cannot be the problem Hume is concerned with.

I think, in accordance with Pitson[6], that Hume is trying to account for the “natural propension” (T, of the vulgar (and philosophers) to believe in the simplicity and identity of the mind in 1.4.6. I also agree with Pitson that the interpretation I give of the Appendix doesn’t cause problems for Hume’s explanation of the origin of the vulgar’s mistaken ideas about personal identity. However, I think that it is Hume’s philosophical explanation of this belief that he is struggling with in the Appendix. I do not think Pitson gives any convincing reason why Hume’s discussion in the Appendix should be confined to the beliefs of the vulgar. In fact, I think that if Hume is concerned with the philosophical perspective in the Appendix this would explain why he retains his satisfaction with his assertion that “the thought alone finds personal identity, when reflecting on the train of past perceptions”. It is only once he comes to explain how we continue to attribute simplicity and identity to the mind given the awareness that it is a bundle of perceptions that the difficulty arises. This is one of the reasons why Hume’s problem seems fairly isolated even within his account of personal identity.

Similarly, I disagree that this interpretation fails because Hume states his problem concerns “successive” perceptions. I think that it is simply the case that, for Hume, the ascription of diachronic and synchronic unity to the perceptions constituting a mind is a temporal process – it is a judgement that takes place over time. We form our mistaken ideas of personal identity when reflecting on our perceptions in a “smooth and uninterrupted” train of thought or ideas (T, There are, therefore, a series of perceptions that Hume cannot appropriate into the bundle. This is why Hume says that it is the successive perceptions in our “thought or consciousness” that he has difficulty explaining. The problem is not necessarily confined to an attribution of synchronic unity. The attribution of simplicity to our perceptions might – in fact I think probably does – pose more of a problem for Hume than does the attribution of identity, but he doesn’t indicate that this is his particular problem in the Appendix. In fact the mistaken idea of the simplicity of the mind seems to be somewhat neglected in the Appendix – as it is in 1.4.6. The particular problem is the successive reflective perceptions involved in our vulgar attributions of identity and simplicity. Whether the problem they pose is confined to explaining our mistaken attributions of diachronic or synchronic unity is another question. Therefore I think Pitson fails to give any conclusive argument against this interpretation of the Appendix.


I have explained, and briefly defended, what I think Hume’s problem is in the Appendix to the Treatise. I think that this interpretation fits with what Hume says in the Appendix, and makes sense as to why the problem he refers to is isolated to, and even within, his discussion of personal identity – why he doesn’t seem to think it causes problems for his reasoning in the rest of the Treatise, or even for how it is that “the thought…finds personal identity, when reflecting on the train of past perceptions, that compose a mind”. It is only in his philosophical reasoning – when actually considering such an act of thought – that Hume runs into trouble.

Hume concludes that “if this be not a good general reason for scepticism, ‘tis at least a sufficient one (if I were not already abundantly supply’d) for me to entertain a diffidence and modesty in all my decisions” (App.10). Hume’s empiricism allows him to plead the “privilege of a sceptic” and announce the problem of explaining our philosophical attribution of simplicity and identity to our perceptions “too hard for my understanding” (App.21). Our natural propension to think ourselves possessed of a simple and identical mind remains, however – whether aware of ourselves as bundles of perceptions or not – and Hume therefore leaves us with yet another example of “the tendency for human nature to generate beliefs that cannot be shaken by sceptical arguments, but are not open to justification by metaphysical constructions either” (Penhelum, 42).



University of Bristol

Bristol, United Kingdom

About the Author



Ainslie, Donald. “Hume’s Reflections on the Identity and Simplicity of Mind” in Philosophy and Phenomenological Research, Vol. 62, No. 3, May 2001. pp. 557-578.

Garrett, Don. “Hume's Self-Doubts about Personal Identity” in The Philosophical Review, Vol. 90, No. 3, July 1981. pp.337-358.

Hume, David. A Treatise of Human Nature, ed. David Fate Norton & Mary J. Norton. Oxford; Oxford University Press, 2000.

Penhelum, Terence. Themes in Hume; the Self, the Will, Religion. Oxford; Clarendon Press, 2000.

Pitson, A. E. Hume’s Philosophy of the Self. London; Routledge, 2002, pp.32-49, 66-80.

Waxman, Wayne. “Hume’s Quandary Concerning Personal Identity” in Hume Studies, Vol. 18, No. 2, November 1992. pp.233-253.

[1] What I will call a “reflective perception.”

[2] Or mind. I think Hume uses ‘mind’ and ‘self’ interchangeably, and I will do likewise.

[3] By ‘real connection’ Hume seems to mean: “a connection between two objects such that the existence of one entails or is impossible without the existence of the other.” (Garrett, 350).

[4] A point “on which all seem to agree.” (Waxman, 1992, 234).

[5] Waxman also thinks this phrase (“our thought or consciousness”) seems likely to refer to the “associative imagination” rather than the mind itself (1992, 236), although he takes a different view as to what this implies is Hume’s main problem in the Appendix.

[6] And with Penhelum (107), but supposedly contrary to Ainslie.