16th February 1999(1,739 words)
Was Hume right to think of the self as a bundle of perceptions? Is there any sensible alternative to thinking of it in this way? If we think of the self as Hume thought of it, is its survival of any importance?
Hume's basis for discussing the self comes after he address what he sees as the inadequacies of discussions of personal identity. In fact, it is clear that Hume wishes to restrict the use of the term identity to an extremely small amount of occurrences. In the physical world, his notion of identity demands that a substance is totally unchanged to claim an identity through time. He certainly rejects any notion that identity can be issued according to usage of a substance. Whilst it might perhaps seem a pedantic point to define the term 'identity' so closely that it no longer has any use, by doing so, it prevents easy mistakes being made when discussing less obvious cases of identity. An example of this could be the ageing desk, which slowly has part by part replaced. Each stage seems to suggest that there is no change in identity, according to a looser definition, yet if the later desk is compared to the original, it would be seen that there is a clear difference. For this reason Hume seems to be justified in calling a search for identity a grammatical, rather than philosophical problem. For this reason, when Hume looks towards the person, he is attempting to show what a person is at a specific time, rather than looking for an identity across time. It is only with there reservations accepted that Hume then moves to talk of why it seems common sense to discuss personal identity, what he calls a form of fiction. He does not believe in an over-ridding framework which unites what he identifies as the self at various temporal points.
This study must therefore discuss the notion of the self that Hume puts forward. The first point to make of this, is something which I believe is its greatest flaw. This is that Hume's notion of the self as purely a mental state. He makes little reference to the importance of the body as a site for a particular set of mental activities to be housed. This point will be addressed later, particularly as an important alternative to Hume's model. The next point is the specific nature of the metal state that Hume puts forward as being the self. He suggests that the self is a bundle of ideas and impressions (which are 'lively ideas') which present themselves to the mind at any one time. He seemingly rejects any background structure to the mind, instead suggesting purely that it is made up of these perceptions. It must be stressed here that such bundles include notions such as emotions and memories. It would be wrong to interpret Hume as suggesting that there is no temporal linkage between bundles, instead he is insisting that this linkage is not strong enough to be sure that there is a common identity. He suggests that what many interpret as personal identity is instead a sequence of impressions and ideas held together by certain relations. In his treatment of explaining this misconception, he notes that there does seem to both resemblance and causation linking bundles at different times. Later bundles can resemble earlier bundles and there can be seen through experience to be direct causal relationships between pairs of perception bundles. Perhaps Hume is therefore simply unhappy in the use of the word 'identity' to describe these linkages. He wishes to use 'identity' only in its purest form and not slip into the errors of sloppy language. This does not mean that he cannot see a form of link. Because I believe that Hume does see linkage between bundles, it seems that there could be some importance linked to survival of the self in Hume's system. This also will be discussed later.
As noted above, the first major problem with Hume's system of self is the absence of linkage between self and body. There are two reasons for this being a difficulty. Firstly, it seems difficult to place a spatial identity on a mind without the use of reference to a body. This does not mean that I am demanding for a model in which mind has a spatial dimension, but only that it is related to specific things with a spatial dimension. Otherwise it could easily be suggested that a single bundle could be acting on more than one body, or that bundles could mix, because there is no spatial constraints on where a mind can act. It does seem implied that Hume doesn't believe in non-temporal links between his bundles of perception, except of those indirect links, such as through communication, this problem of not confining the mind to a body is hard to reject.
A more obvious problem with omitting the importance of the body is that of a more basic notion of the body itself being part of a self. There is a distinction to be made between mind and body, but this does not mean that the self cannot be an amalgamation of the two. I think that Hume is perhaps too quick to judge the self as purely a mental notion. A lot of what is implied in being person is to have physical characteristics. In fact, it is further clear that many of the mental characteristics are dependent on the functioning of the body. This is more than simply the use of sensory organs which give many of the perceptions, but also the functioning of the brain, with different areas responsible for perception functions, as is seen in the cases of stroke victims. By denying the role of the body, Hume makes one of the central errors of Locke. Not only is the mind dependent on the body to form impressions and ideas, but it is also dependent on it to express them. Without such functions the mind would not be able have ideas or impressions. The structure of an individual's body with therefore be a part of what it is to be them, part of their self.
The second major problem is that of the temporal dimension of Hume's self. Although perceptions include memories, he does not try and show any common identity between bundles of different times. There are links, but to Hume, it would be fictitious to associate these with sameness. Further, it might be suggested, that these links, not seemingly strong, might make it possible for links between bundles from different bodies to be associated together. The links involved are resemblance and causation and so it might be possible for bundles, which have no spatial constraints to affect others elsewhere. This seems to contradict much of what is thought of continuity of the character within the person through time. In reality, I do not believe that such a possibility would fit into Hume's model, as his point seemed to be, that there is some strength in traditional notions of temporal continuity, but that he believed that the constant dynamics meant that the different occurrences of perceptions could not exhibit sameness. Opponents of such a view point to some things which do seem to show a degree of sameness that they would accept to show common identity over time. The most obvious cases are Locke's use of the memory and Parfit's use of character. The problem still remains that both of these factors are still dynamic and as with the altered desk, comparing the mind of a child with an old man, who Locke and Parfit might say share an identity, could show no sameness in any of their mental characteristics. Indeed the old man could have lost all memory of his childhood and the character would be expected to be dramatically different. Noting that the criteria that Hume demanded for identity seem far stricter than others, suggests that such alternatives would not be acceptable.
The last point to address is whether according to Hume's model survival is of any importance. The answer to this depends greatly on the interpretation that is given to Hume's discussion. If Hume is read to suggest complete lack of connection between the bundles of perception over time, it could be suggested that one bundle would have no care towards any future bundle as they would be as different as any bundle in another body. A more generous interpretation, might however suggest that although Hume rejects a common identity between bundles, he does accept connections, through resemblance and causation factors. This connection might give rise to an acceptance that a self would be more interested in bundles which share these connections that those that do not. This suggest that in as much as a self would look towards the future, it would see its own future more connected to those future bundles, even if it could not claim sameness. It might therefore see some importance of the survival of such bundles. This question can therefore be answered in two ways according to how Hume is interpreted.
Before concluding, it must be noted that after writing his model of personal identity in the Treatise Book I, he admits his failure to fully explain a workable system in his Appendix to the Treatise. He does not, however, try to create a replacement system. This does not mean, however, that the original model does not raise some interesting points as to how identity and the self are to be judged, particularly over time.
In all, the strict nature with which Hume uses the term identity means that it is unsurprising that he is limited in scope for usage. The dynamics of mind and body would suggest that he might be trapped into ignoring temporal links in description of identity. He does, however, make an important mistake in omitting the body from his system and confining it to bundles of perception. His recognition of the connection, but not sameness of bundles that others would associate with personal identity, does show that the system could be applied to problems such as whether a Humean self would look towards the future. There are many other forms of personal identity that can be offered, but working within Hume's tight definition suggests that something as changeable as the self would be hard to show some constant common thread over time.