We recently read Susanna Schellenberg’s (2010) paper, “The Particularity and Phenomenology of Perceptual Experience”. Here Schellenberg develops an account of perceptual experience that accounts for what she calls the “particularity” of perceptual experience, while also accommodating the possibility of two experiences being subjectively indistinguishable despite involving perceptual relations to distinct environments (including where one of the experiences is hallucinatory). She discusses various construals of the “relational particularity” that needs accommodating, but at the end of the day she argues that what is needed is “that the object of perception is a constituent of the perception” (p. 24), and that the best way to accommodate this is to hold that the particular object one is perceiving makes a constitutive difference to the accuracy conditions (and thus the content of) the experience.
She argues that the “austere relationalist” (which includes many disjunctivists) can accommodate particularity but not subjective indistinguishability. And she argues that the “austere representationalist” (which includes most representationalists about perceptual experience) can accommodate subjective indistinguishability but at the loss of accounting for particularity. This leads Schellenberg to develop an interesting alternative position, according to which perceptual experiences involve de re modes of presentation of particular objects. In cases of hallucination, these de re modes of presentation are gappy. Subjectively indistinguishable experiences of distinct objects will involve distinct de re modes of presentations, but will have in common a type of de re mode of presentation. This is not strictly speaking a shared content, and so Schellenberg’s view requires that we abandon the relatively popular thesis that phenomenally identical experiences share a kind of intentional content. (I will often use the expression “phenomenally identical” instead of “subjectively indistinguishable, since to use the latter strikes me as being compatible with views that treat the commonality between hallucination and perception as being merely epistemic–a view that Schellenberg appears to disfavor in her discussion of disjunctivism).
I want to consider the degree to which we should expect a proper theory of perceptual experience to accommodate what Schellenberg calls “relational particularity”, in contrast and addition to “phenomenal particularity”. But first, the Schellenberg paper elicited in me some (possibly naive) concerns about just what is supposed to be at issue in such debates.
On Individuating Experiences
Schellenberg presents the issue in part as one concerning how experiences are individuated. [In what follows always read "experience" as "perceptual experience", where hallucinatory experiences are also perceptual experiences--this seems to be Schellenberg's usage.] I take it as a reasonable thesis that any particular perceptual experience is a token physical event that can be individuated in a variety of ways. What is at issue, then, is how one ought to individuate experiences qua experience. But even then, what is at issue is unclear. We can type experiences by their phenomenal character, and this seems like a natural way to type them. We also take perceptual experiences to be ways of making contact with worldly objects and events, and we might type experiences in a correspondingly world-involving way. Schellenberg mentions these two views, and ultimately sides with the world-involving way of individuating experiences. But given that these seem to both be perfectly reasonable ways of individuating experiences, what criteria determines whether we have found the correct way of individuating experiences qua experiences? Without some guidance on even what one is aiming to do when one offers an account of “how perceptual experiences are individuated”, it is not even clear to me that these two views need to be seen as competing. One might appeal to the concept PERCEPTUAL EXPERIENCE, but given the state of the debate I’m pessimistic that this will give us a clear verdict. One might appeal to causal/explanatory usefulness considerations, but I suspect that this favors the non-world-involving views.
I think Schellenberg (and others, such as Martin, who often makes reference to “fundamental kinds” in discussing what perceptual experiences are) owe us some guidance on what considerations would support treating perceptual events as being individuated one way rather than another. The disjunctivist who denies the possibility of phenomenally identical experiences of distinct objects (and between hallucinations and perceptions) is on sturdier ground here, simply because to individuate by phenomenal character will turn out to also be to individuate in terms of relations to particular objects. [I'm ignoring some possible complexities in exegesis of the writings of disjunctivists!] There are not two competing and viable options on their metaphysics. But given Schellenberg’s acceptance of a common phenomenology between experiences of distinct particulars, she must allow that there exists an interesting kind to be picked out in terms of phenomenal identity, and that individuating events as falling under this kind will be insensitive to the particular objects of perception. So what makes it correct to say that *perceptual experiences* are not to be individuated in this way, but in some other way that involves the worldly objects of perception?
Do Particulars Enter Into the Accuracy Conditions of Perceptual Experiences?
In developing a “third way” between austere relationalism and austere representationalism, Schellenberg develops a view into which particular objects enter into the intentional contents of perceptual experiences. The above worries would undermine one motivation for that project. But Schellenberg also thinks that considering just the accuracy conditions of experiences themselves would lead us to the view that those accuracy conditions are sensitive to facts about the particular objects that the subject is perceiving.
I wish Schellenberg would have argued more for this thesis. She gives an example of perceiving a coffee cup1, and then having a qualitatively indistinguishable experience of a distinct coffee cup2. She then states that the accuracy conditions of the subject’s experiences would change in such a case of switching. But surely many would simply disagree with this intuition.
I again find myself with some methodological question marks. Schellenberg’s project requires that we have a clear sense of the idea of the content of a perceptual experience that does not depend on the assumption that such content supervenes on phenomenal character. In considering such contents then, we cannot pose the question of how the world appears to be to the subject, given how things phenomenally seem. We have to do something else. How might this go?
We might start with the things we say about perception. We say things such as “Brian saw the albino alligator” or “Liz heard Jill come home last night”. It is clear that we often talk about the particular things and the particular events that subjects perceive. This might provide some inclination for thinking that it would be a mistake to characterize perceptual content as having only general or existentially-quantified content.
But we must be careful in moving from observations about the things we say to claims about perceptual content. For one thing, most philosophers are willing to make a distinction between the content of a perceptual experience and the content of perceptual judgments. Even if the things we say about perception are reports of mental contents associated with perceptual experiences, it would be too hasty to assume that such reports reflect perceptual content proper rather than perceptual judgments. Furthermore, such reports might be reports about the perceptual relations that hold between perceivers and objects, and not reports that are purely about the perceptual contents of the perceiver.
Compare also to the distinction between de re versus de dicto propositional attitude ascriptions. Kevin believes that today’s winner of Jeopardy is smart but he has no way of singling that woman out via anything other than a definite description. If Brian and I are acquainted with this woman, Jennifer, then it can be appropriate to say to Brian that “Kevin believes that Jennifer is smart.” Here the use of the proper name is a way to convey to Brian whom Kevin’s belief is about. The use of the name should not be taken to provide information about how Kevin thinks about that woman, or even that he is in a mental state that has a particular (Jennifer) as part of its content.
Similarly, even if perceptual contents themselves were merely existentially-quantified contents, it would not be surprising that in making reports about such contents we would make reference to particular objects that satisfy those contents.
[Michael Tye (2009) gives an interesting case involving mirrors in his most recent book that could support the idea that particulars enter into perceptual content. But I will save discussion of that case for another time.]