Sunday, 4 February 2007

Moore On Sense Data

One of the central questions in the philosophy of perception is: what are the objects of perception? One way to interpret the locution ‘objects of perception’ is in terms of what it is that we are immediately aware of in the act of perceiving. In the early 20th century the traditional answer to this question was ‘sense data’ (henceforth SD). Thus, Moore defines SD as that which we ‘directly apprehend’ in perception [p. 46 ff*]. A striking feature of Moore’s account of SD is that he appears to remain open to the possibility that SD are mind-independent [p. 31, 43-5]. For example, Moore opines that SD may continue to exist even when we do not perceive them [p. 31], in which case their existence would not be dependent on the mind of the perceiver. In this regard, Moore’s conception of SD differs slightly from that of most contemporary philosophers who view mind-dependence as a fundamental feature of SD (a fact which Moore [p. 40-3] himself acknowledges).

Moore points out that when we combine (1) the assumption that in perceptual experience we only directly apprehend SD, with (2) the standard view of SD as mind-dependent, we appear to quickly run into epistemological difficulty [p. 48]. The former claim makes SD the basis for our perceptual beliefs while the latter allows for the possibility that SD only exist in the mind of the perceiver without any corresponding physical objects. Together, they appear to invite scepticism regarding our belief in the existence of physical objects since said belief is based on something that is compatible with there being no physical object. Significantly, the above line of reasoning rests on the equating of mind-dependence with object-independence (and by ‘object-independence’ I mean the thesis that SD (can) exists independently of any corresponding physical object). However, it is conceivable that SD could be simultaneously mind-dependent and object-dependent (for example, we may conceive of SD as arising out of some type of interaction between the mind and object). Moore seems to conflate mind-dependence and object-independence [p. 48], though he does avoid making the inverse mistake of assuming that mind-independence equals object-dependence (more on the latter claim below).

At times Moore seems to lean toward the view that SD may actually be identical to the physical surfaces of objects, but at no point does he ever endorse this view. Such reluctance on the part of Moore seems warranted since identifying SD with the surfaces of objects falls prey to one of the challenges that led philosophers to posit SD in the first place—namely, the problem of hallucinations. Since in the case of hallucinations, unlike illusions, there is no object present at all, there is nothing for SD to adhere to. Thus, even if he were to concede that SD are mind-independent, if he is to account for hallucinations, Moore must maintain that SD are object-independent (as he does in fact do). Another question that Moore appears to remain neutral on is whether SD are physical or non-physical entities. If the former, then Moore seems forced to identify them with the surfaces of objects (a prospect we have already deemed unattractive). However, if he regards them as non-physical then Moore seems faced with the difficulties of explaining what exactly they are and how could they interact causally (if at all) with our sensory faculties. In sum, it remains unclear whether Moore regarded SD as mind-dependent or mind-independent, or physical or non-physical. But this much seems clear: whatever they are, SD must (by Moore’s lights) be object-independent [p. 50].

Question: Does Moore ever come to a final position on whether (1) are SD mind-independent or mind-dependent or (2) are SD physical or non-physical?

*All citations refer to the first chapter of Moore’s Some Main Problems of Philosophy.



The following helpful comments were provided by Alan Millar:

[In response to paragraph 2]:

It is important to distinguish between contingent and necessary dependence. It might be that the sense-data with which we are acquainted depend on the features of the objects around us in that the sense-data with which we are acquainted would not be as they are unless brought about as a result of the causal impact of things in our surroundings with features that the sense-data in some way reflect. This kind of dependence is contingent. The idea is that as a matter of contingent fact one would not usually be acquainted with the sense-data with which one is acquainted unless … . This allows for the possibility that it is possible to be acquainted with sense-data, as in cases of hallucination, even though my being acquainted with those sense-data does not depend causally on what is around me. But when Moore takes seriously the view that sense-data are mind-dependent he has in mind a different kind of dependence. The idea is that sense-data depend essentially on the mind of the subject who is acquainted with them. If that is so, then it would not be possible for there to be such data but for the mind of the person acquainted with them.

Was your idea that if sense-data were object-dependent then they would pose no epistemological problem concerning our knowledge of the physical world?

[In response to (1)]:

As I recall he continued to take seriously the idea that they are essentially mind-dependent and non-physical. You might like to check out what I think is his last essay on the matter. This is 'Visual Sense-Data' in C. Mace (ed.) British Philosophy in Mide-Century (Allen and Unwin, 1957).


Benj Hellie said...

Alan is right -- in 1957's 'Visual sense-data', Moore concludes that the sense-datum is not "identical with that part of the opaque object's surface which you are seeing" (this is p 136 of the Swartz reprint).


Thanks for the info Benj, I'll be sure to check it out!