Monday, 28 December 2015

John Pollock on Percepts and Description

John Pollock writes:

When I see an object and make a judgement about it, I do not usually think of that object under a description – not even a description like ‘the object I am seeing’…my visual experience involves what we might call a ‘percept’ of an object…” [1986]

If the above is true, then how does Pollock know that he's seeing an object? How has he distinguished it from its surrounding objects or even from the extended spatial mass in front of him?

The object forces the judgement, as it were; though judgements aren't entirely constituted by the object. That is, there would be no a posteriori judgements about the object if there had been no a priori judgements which have offered up the object as an individualised particular.

Pollock continues by saying that

[p]ercepts are not descriptions, so this is an example of a nondescriptive mental representation. A percept can only represent an object while that object is being perceived.”

It depends on how strong we take Pollock’s phrase “under a description” to be. Is he referring to a vocal or even sub-vocal description of the object? If so, vocal or sub-vocal descriptions may not be needed in order to have some form of description of the object. The concepts he applies may be instantaneous or even a priori.

In order to have a percept of an object that object needs to have been individuated in some way. Indeed by saying that a percept is "of an object", Pollock has implied that such individuation has already been carried out. How can a percept be of an object unless the subject has distinguished it from that object's surrounding landscape?

The same problem can be seen with his use of the term “representation”. Isn’t it the case that representations are representations of something? Aren’t they about something? Perhaps percepts, in Pollock’s book, aren't of - or about -anything. That would be fair enough; although he also uses the words “non-descriptive representation”. And surely representations represent individuated objects. A percept, on the other hand, could be deemed not to be of - or about – anything.

In that case, imagine facing a white wall with one’s eyes wide open. Whilst facing - rather than looking at - the wall, a person may be thinking about things that aren't at all related to the white wall in front of him. However, the white wall would still be part of his overall mental state at that time. However, he wouldn't be having thoughts about the white wall. Perhaps we can say that he's not even looking at it; even though sensory data from the white wall are entering his consciousness. In that case, part of the overall mental state (which includes thoughts which aren't about the white wall) would still include the white wall. We can say that the white wall is an accompaniment to his cognitive activities at that time. Perhaps this is what a percept is. That is, he would have a white wall percept without him knowing that it's a white wall percept and without that percept being about - or of - the white wall. The white wall percept wouldn't even be an image of the white wall; for the same reasons given about representations not being intentionally directed.

If percepts are as I've described them, then one may be able to give descriptions of one’s percepts; though as soon as one did so, they would no longer be percepts. In fact, a genuine percept may not even be remembered in order to describe it.

Remember the white wall percept: if his thoughts were elsewhere, he couldn’t give an after-the-fact description of the white-wall percept because that percept wouldn't have even entered his memory. How could it be in his memory if at the time he was thinking about things which had no relation to the white wall in front of him? And, of course, Pollock doesn't want his percepts to be descriptive anyway. Percepts are non-cognitive, on my reading of Pollock’s term.

I agree with one thing that Pollock says vis-à-vis percepts: they only occur in the presence of the objects that cause them. There is just a causal non-epistemic and non-cognitive relation to the white wall. Indeed, by saying that percepts only occur in the presence of the objects which cause them, Pollock appears to concede the point that they're are non-cognitive backgrounds to thoughts which are about or of things other than themselves. That is, Pollock acknowledges the causal relations that are required for percepts.

Representations or images, on the other hand, don't need direct causal contact with the objects the representations or images are about; though, of course, their causal ancestry can be traced.


Pollock, John. (1986) 'Epistemic Norms'.

No comments:

Post a Comment