The discussion highlights the need to distinguish between perceptions and the experiences implicated by perceptions, noting that Coliva’s framework makes perception irrelevant to justified belief, except for being the contingent means by which we are furnished with experiences that are the real source of justified belief. It then addresses two issues concerning the problem of cognitive locality. The problem concerns what enables us rationally to suppose that our perceptual experiences mostly put us in touch with reality. The issues addressed are: (1) whether, assuming that there is a problem of cognitive locality, Coliva’s Moderate position adequately addresses it; and (2) whether Coliva gives us enough to make sense of the claim, central to the Moderate position, that certain background presuppositions are constitutive of empirical rationality.
Annalisa Coliva’s (2015) book is an ambitious and wide-ranging treatment of fundamental problems in epistemology. I come at the issues it raises from a perspective that is somewhat different from hers and also from that of proponents of the main positions on perceptual justification with which she engages. Though my remarks will be coloured by that perspective my aim here is not to defend it. I shall highlight one point concerning the outlook that Coliva shares with her opponents, and raise two issues arising from her account of perceptual justification and extended rationality.
The point to be highlighted has to do with a slide that it is all to easy to make from a common sense thought—that we can be justified in believing things because of our perceptions—to a further, more theory-laden thought—that our sensory experiences, along with other factors, can justify our beliefs. I discuss this in Section 2.
The issue is whether, assuming that there is a problem of what Coliva calls cognitive locality, the Moderate position could adequately address it. I consider this in Section 3.
[A] belief about specific material objects that P is perceptually justified iff, absent defeaters, one has the appropriate course of experience (typically an experience with content that P) and it is assumed that there is an external world (and possibly other general propositions, such as “My sense organs work mostly reliably”, “I am not the victim of massive cognitive deception”). (Coliva 2015: 34)
It is crucial to the Moderate position that the tacit presupposition of certain background assumptions is constitutive of empirical rationality. The second issue to be raised is whether Coliva has given us enough to make sense of this idea.
2 Perception, Experience, and Justified Belief
I begin with the common sense thought referred to above. The thought is that when, for instance, you look at something, say the contents of your fridge, you can, and normally will, thereby acquire beliefs about the fridge that are justified. This thought is apt to lead epistemologists, and does lead Coliva and those with whom she engages, into a further, more theory-laden thought that is certainly not part of common sense and is disputable. To see this we need to attend to the distinction between perception and experience.
It is agreed on all sides that perception is a relational notion. Clearly, when you see inside your fridge you stand in a relation to the contents of the fridge. It is also agreed on all sides that perception, at least typically, implicates sensory experiences. Looking in the fridge you have a course of visual experience. A widely held view is that such a course of experience is not essentially related to anything in your current surroundings. Although as you look your visual experiences will depend causally on the layout that you see, the same experiences might have been implicated in a perfectly hallucinatory counterpart of your perception—an episode in which you visually perceive nothing but have a course of experience just like that which you received when looking into your fridge. If this is right then, although seeing the inside of the fridge is essentially relational, it involves having an experience that is ‘inner’ in the sense that it is not essentially related to anything presently in your surroundings. This view has in recent decades been challenged by a number of formidable philosophers who deny that perceiving involves having an experience that is inner in this sense.1 These philosophers—relationalists about perceptual experience—claim that the visual experiences implicated in the case described are themselves essentially relational. It’s not just that they use the term “experience” where others use “perception”. They differ substantively from non-relationalists because they deny that perception is a matter of having an inner experience contingently related to the objects perceived, and they hold that perceiving implicates—indeed consists in—an experience that is unmediated acquaintance with the objects perceived. If they are right about this then any such experience is different in kind from any perfectly hallucinatory experience.2
I take it that those who hold to the conceptions of perceptual justification (warrant) critically discussed by Coliva are not relationalists. I also take it that in declaring McDowell’s disjunctivism, and by implication his relationalism, to be “extremely problematic” (2015: 45) Coliva indicates that she aims to develop her own view without a commitment to any such view. She shares with the two accounts of perceptual justification that she criticises—liberalism and conservatism—the idea that, roughly speaking, experiences, conceived non-relationalistically, can be justifiers of beliefs in that, along with other factors, they serve to justify beliefs. This is the further thought to which I alluded at the start—the thought that takes us beyond the common sense thought that frames Coliva’s discussion. The common sense thought is that perception can supply justification for beliefs. The further thought is that sensory experiences, under the operative conception, can justify beliefs.
Those who endorse the further thought may accept that when you look in the fridge and see some butter you are caused to be in a position in which you are justified in thinking that there is butter in the fridge, but for them it is not constitutive of your being justified in so thinking that you see the butter that is in the fridge. For you could have been no less justified in thinking so if you had been hallucinating butter sitting in a fridge. Expressed in more general terms, the claim is that there can be parity with respect to the justification of the belief that an F is before one between a subject whose sees an F, recognizing it to be an F, and one for whom it merely looks as if an F is there and who on that account believes that an F is there. I’ll call this claim Parity.
I take Parity to be problematic though it is widely accepted and, indeed, widely taken to be a datum that any plausible theory of perceptual justification should accommodate.
In the following passage Coliva objects to this in view of a problem about what she calls cognitive locality.
a belief about specific material objects that P is perceptually justified iff, absent defeaters, one has the appropriate course of experience (typically an experience with representational content that P). (Coliva 2015: 21)
This is not a rejection of Parity in general, but the passage expresses unease at the idea that two subjects who, on having visual experiences such that it looks just as if a red cube is before them, believe that a red cube is before them, would be on a par with the respect to the justification of that belief simply on account of being on a par with respect to their current visual experiences and the absence of defeaters. For Coliva two subjects on a par in the specified respects need not be on a par with respect the justification of believing that a red cube is before them because it could be that only one of the subjects has a reason to hold that his or her experiences are at least mostly caused by causal interaction with physical objects. Lacking such a reason, Coliva suggests, “we [would] remain confined within the realm of experience” and if we were so confined we would not have sufficient warrant for any beliefs concerning what lies beyond the realm of experience.
It seems that as long as we are merely concerned with perceptual experiences and do not avail ourselves of some externalist story to provide an account of why they should, at least mostly, put us in touch with material objects out there, we remain confined within the realm of experience which, as such, is not sufficient to get us outside of it, that is, to warrant beliefs about material objects. This does not mean we should think of experience in terms of sense data, conceived of as mental entities or as a “veil”, which wraps us up. Rather, the point is that even if, metaphysically, experiences were individuated by reference to material objects they allegedly represented, and even if it were the case that we were lucky enough to have mostly veridical experiences, there would be no subjectively available reason to hold that they are in fact, at least mostly, caused by causal interaction with physical objects. If everything looks internally just the same to one, why would one be warranted in believing, upon having a hand-like experience, “Here is my hand” rather than “I am hallucinating a hand” or “I am dreaming of having a hand”, and so on? In such a predicament, it would seem natural to think that if we can surpass our cognitive locality, it would be because of some extra information that—as implicit as it may be—allows us to consider our perceptual experiences as bearing on physical objects. (Coliva 2015: 25–26)
For those who adhere to Parity perceiving an object can have only causal relevance to whether any belief about it is justified. By their lights perceiving the object explains how one comes by the experiences that are essential sources of the justification of those beliefs, but is inessential to the justificatory status of those beliefs. Most epistemologists are not troubled by the idea that it is thus inessential. I think we ought to be troubled by it in view of the role of perception in enabling us to have demonstrative thoughts about perceived objects and the role of recognitional abilities in enabling us to have knowledge about those objects and beliefs about them the justification of which depends on that knowledge. (I set out this more fully in Millar 2011.)
Another reason for taking the further thought to be problematic is that, arguably, our practices of epistemic evaluation engage with epistemic standings of beliefs arising non-inferentially from perception at the level of success concepts—knowledge and perception, rather than belief and experience that is non-relationalistically conceived. We readily think in these cases of what people know from what they see or otherwise perceive. The notion of experience, conceived non-relationalistically, is a relatively sophisticated one. I think it has little if any role to play in actual epistemic evaluation. Taking experiences to contribute to the justification of belief amounts to a commitment to a disputable body of philosophical theory. It does not drop out of the common sense thought that perception enables one to have justified beliefs. That was the point to be highlighted.
3 Overcoming Cognitive Locality
The first issue is whether, assuming that there is a problem of cognitive locality, the Moderate position could adequately address it. We are to make sense of the standing of the relevant background assumptions in terms of what Coliva calls Extended Rationality. Central to the Moderate position is that the relevant assumptions need not be warranted either evidentially or in the style of the entitlement that is central to the Conservative position. The challenge facing Coliva is to show that relying on such assumptions is legitimate (“within our epistemic rights”, Coliva 2015: 33) despite the fact that one lacks any warrant to accept them.
We are to think of Moore as part of a mistaken account of a warrant for (iii). It is said that if Moore were to articulate a warrant one has for (iii) then one would need to have a warrant for (i). Suppose that this warrant is visual-experiential. The problem is that having a visual-experiential warrant for (i) would seem to presuppose that one has the very warrant for which Moore was supposed to account, that is, a warrant for (iii). In Coliva’s picture the explanation for this depends crucially on her endorsement of the assumption that the warrant that is supplied by perception is provided by the subject’s current sensory experiences. The idea seems to be that the assumptions that are integral to experiential justification are such that
(i) Here is my hand.
(ii) If there is a hand here, then there is an external world.
(iii) There is an external world. (Coliva 2015: 29)
(a) constrains which background assumptions have to be in the picture that, according to (b) is required. We are to suppose that the assumption that there is an external world satisfies both (a) and (b). The same applies to the assumptions (i) that for the most part one’s sense organs function reliably, and (ii) that one is not subject to massive deception.
(a)only if they were true would our experiences be reliable indicators of states of the external world, and
(b)only if they form part of an implicit or tacit background picture of the world that is presupposed by our modes of belief-formation can our experiences serve their warranting function.
This reading of Coliva’s framework fits well with her concern with cognitive locality—the problem of how we may transcend the confines of our own sensory experiences. Suppose that on having a certain course of experience one forms a belief about one’s surrounding. On Coliva’s view, forming such a belief would warrantedly transcend the confines of one’s then current visual experiences, if the experiences had contents appropriate to the content of the belief, there were no defeaters, and one (implicitly or tacitly) took for granted—presupposed—appropriate, highly general background assumptions that satisfy (a) and (b). But now, assuming that the problem of cognitive locality is a real one, the question arises whether the Moderate position really addresses it.
The issue now is, “Why is this not just as much of a problem for the Moderate position?” How can it be that presuppositions that for all we know are false can, along with satisfaction of the other requirements of the Moderate position, put us in a position in which we have a reason to believe that a red cube is before one that would not be available to someone who merely satisfied the requirements of the Liberal position. How one reacts to this will depend on one’s conception of having a reason.
…there would be no subjectively available reason to hold that [one’s experiences] are in fact, at least mostly, caused by causal interaction with physical objects. …In such a predicament, it would seem natural to think that if we can surpass our cognitive locality, it would be because of some extra information that—as implicit as it may be—allows us to consider our perceptual experiences as bearing on physical objects. (Coliva 2015: 25)
Coliva is committed to Parity in general—the claim that there can be parity with respect to the justification of the belief that an F is before one between a subject whose sees an F, recognizing it to be an F, and one for whom it merely looks as if an F is there and who on that account believes that an F is there. For according to the Moderate account two subjects to whom it looks just as if a red cube is before them, and who are otherwise the same, will be on a par with the respect to the justification of the belief that a red cube is before them provided only that they lack defeaters and take for granted the relevant presuppositions. If one of the pair additionally sees a red cube, recognizing it to be a red cube, then by the account that person is no better off than the other with respect to justification. Many will not be troubled this conclusion. They may acknowledge that in the latter case—the case of recognition—the subject has knowledge that a subject who merely satisfies the requirements of the Moderate position will lack. I find the conclusion troubling because I think that perception of a red cube can furnish us not just with knowledge that a red cube is there but with the knowledge concerning that red cube that it is something that one sees to be a red cube. The latter knowledge is access to a reason to believe concerning the red cube that it is a red cube. This is a reason that need not be available to one who satisfies the Moderate requirements on experiential justification. It is constituted by a truth that one can know if suitably equipped (for discussion, see Millar 2011).
4 The Status of the Background Assumptions
As Coliva expresses it, the principal challenge to the Moderate position is posed by the initially plausible thought that “our basic assumptions are neither warranted nor warrantable” and, accordingly, “are not epistemically rational” (Coliva 2015: 120). Her proposed solution is to challenge the move from “not warranted” to “not epistemically rational” with the help of the claim that the basic assumptions “are constitutive of epistemic rationality itself” (Coliva 2015: 129). The upshot is that we are not irrational in presupposing the various assumptions on which our experiential justifications rest, even though these assumptions are unwarranted and not warrantable, because presupposing these assumptions is constitutive of what empirical rationality is.
As a point of method, it seems right that an account of rationality or justified belief should mesh with our actual modes of belief-formation and with our practices of evaluating beliefs with respect to whether they are justified. Clearly, though a question arises as to what these practices are. It is indisputable that we form beliefs in response to what we perceive and that the beliefs we form in this way are sensitive to the courses of experience that our perceptions implicate. This is a modest elaboration of what I called the common sense thought. But when Coliva talks of the deliverances of the senses I take it that she is not speaking of anything that depends essentially on perception. As I take her to view the matter, the deliverances of the senses are experiences that might or might not have arisen from perception. As already remarked, it seems to me that our evaluations of beliefs arising non-inferentially from perception operate at the level of success concepts like those of perception and knowledge, rather than in terms of concepts of sensory experience and belief. I might add that Tim Williamson strikes me as being spot-on when he makes the more general claim that “justification is primarily a status which knowledge can confer on beliefs that look good in its light” (Williamson 2000: 9). There is, then, a sub-issue about how we characterize the relevant practices, even granted the methodological point.
I take it that the notion of epistemic rationality shared by skeptics and non-skeptics alike … does not hang in the air, but depends on our practices. In particular, the notion of epistemic rationality depends on the basic practice (or method) of producing, assessing and withdrawing from ordinary empirical beliefs, such as “Here’s a hand”, “This wall is red”, and so on, interpreted as being about mind-independent objects, based on the deliverances of our senses. Now, if … that practice rests on assuming—with no warrants—that there is an external world, that our sense organs are mostly working reliably and that one is not the victim of a lucid and sustained dream, or otherwise disconnected from causal interaction with physical objects, then those assumptions are constitutive of epistemic rationality itself. (Coliva 2015: 129)
Although Strawson frames his main point as a claim about meaning it seems clear that, in keeping with the times at which he wrote, his claim is meant to be as much about what constitutes being reasonable in relation to the type of case he considers. The problem for his view, as it applies to the inductive justification of a belief that every case of f is a case of g, held on the basis he describes, is that by his own account the belief will be justified only if there being a suitable number of observed cases of f that are all cases of g, and no observed cases of f that are not cases of g, is, indeed, evidence that every case of f is a case of g. Plausibly, that condition will be satisfied only if the world is a certain way and, in particular, only if it is such that not easily would it be that there is a suitable number of observed cases of f that are all cases of g, and no observed cases of f that are not cases of g, and yet it is false that every case of f is a case of g.3 Whether the world is that way is not settled by reflection on what we count as justified belief.
We often talk … of justification, good grounds or reasons or evidence for certain beliefs. Suppose such a belief were one expressible in the form “Every case of f is a case of g”. And suppose someone were asked what he meant by saying that he had good grounds or reasons for holding it. I think it would be felt to be a satisfactory answer if he replied: “Well, in all my wide and varied experience I’ve come across innumerable cases of f and never a case of f which wasn’t a case of g.” In saying this he is clearly claiming to have inductive support, inductive evidence, of a certain kind, for his belief; and he is also giving a perfectly proper answer to the question, what he meant by saying that he has ample justification, good grounds, good reasons for his belief. It is an analytic proposition that it is reasonable to have a degree of belief in a statement which is proportional to the strength of the evidence in its favour … So to ask whether it is reasonable to place reliance on inductive procedures is like asking whether it is reasonable to proportion the degree of one’s convictions to the strength of the evidence. Doing this is what “being reasonable” means in such a context. (Strawson 1952: 256–257)
The question now is whether a somewhat analogous problem does not arise for Coliva’s account. Suppose, for the sake of argument, that we count a person as being visual-experientially justified in believing something provided that she satisfies the requirements of the Moderate position. It clearly does not follow that beliefs that would count as being justified according to this standard would for the most part be true. Whether they would for the most part be true depends on how the world is, and that is not settled by reflection on what we count as experientially justified. This will be a problem for Coliva if whether an empirical belief is justified depends on how the world is—more specifically, if it depends on whether the belief was formed or is sustained in a manner such that beliefs that were formed or sustained in that manner would for the most part be true. Clearly, it is open to Coliva to deny that this condition is a necessary condition on empirically justified belief. She is surely committed to denying this, and in this she is not out of step with many epistemologists. If you think that theoretical rationality has to do solely with the internal organization of our beliefs and our sensory experiences, non-relationalistically conceived, then you might be undisturbed by supposing that justified belief should be conceived in a similar spirit. But, by contrast, you might, and I think you should, suppose that rationality has a great deal to do with having a grip on reality, though our various modes of perception, and having our beliefs constrained by the realities that our perceptions disclose. That conception of rationality strikes me as being more in keeping with our evaluative practice. It fits with the idea that for a belief to be justified is for it to be well-founded in a sense that depends on how the world actually is.4
It is fitting that I should conclude by acknowledging that the issue of the status of background assumptions is a problem for all of us and merits far more attention than it has generally received. It does not simply go away on the perspective that I favour. Coliva deserves credit for having given it such a prominent place in her book.
Bonjour L. (1992). “Problems of Induction.” In Dancy J. and Sosa E. (eds.), A Companion to Epistemology, 391–395. Oxford: Blackwell, 1992.
Millar A. (2011). “How Visual Perception Yields Reasons for Belief,” Philosophical Issues 21 (The Epistemology of Perception): 332–351.
Millar A. (2014). “Reasons for Belief, Perception, and Reflective Knowledge,” Proceedings of the Aristotelian Society, Supplementary Volume 88: 1–19.
I use “perceptual experience” only for sensory experiences implicated in episodes of perception. The term “sensory experience” applies to both perceptual experiences and hallucinatory experiences that are not perceptual.
Perhaps Strawson is committed to denying this. Laurence Bonjour remarks that by Strawson’s lights “the notion of reasonable belief and the correlative notion of strong evidence must apparently be understood in ways that have nothing to do with likelihood of truth, presumably by appeal to standards of reasonableness and strength of evidence that are accepted by the community and embedded in ordinary language” (Bonjour 1992: 393).
For some relevant discussion, see Millar (2014).