RESPONSIBLE BELIEVING

S. Joel Garver

CHAPTER FOUR

Seeing Isn't Believing:
(And Other Thoughts on Doxastic Practices)

Section II: Part A

Let's turn now to examine the basic structure of a number of specific practices. In these examinations we are concerned with the actual structure of the various practices that are--at least sometimes--taken to be justification-conferring. As I talk about the justification conferred by these practices, I am taking the practices to be reliable, and therefore, I am taking my examinations to reflect the structure of actual justification. Whether any particular belief produced by these practices is reliably formed is a separate question, involving the presence of overriders, whether we are rational to take the practice as reliable and the like.

While one might not want to endorse Quine's views in all respects, he is, I think, correct on one fundamental point: epistemology is rightly preoccupied with the "relation between the meager input and the torrential output" in belief-formation (1969:83). The role of our examination of doxastic practices is to provide something of a structural analysis of that relation--or, rather, those various relations. In our exposition of the structure of doxastic practices we shall have to account for the nature of the inputs and outputs and the various relations that hold between them as embodied in the practices (functions). Our account, however, will not be completely "naturalized" into psychology or sociology but will include essentially normative concerns.

To begin, let's consider simple perceptual belief-forming practices. Alston holds, in contrast with many epistemologists, "that perceptual beliefs are sometimes based solely on an experiential presentation, and are sometimes prima facie justified thereby" (1991:81). What he has in mind are cases of perception where a certain object X presents itself visually to a subject as being O and the subject comes to hold the belief that "X is O" and the subject holds that belief solely on the basis of the visual presentation. Such beliefs are said to be purely immediately justified. Various attempts have been made to discount this possibility. Alston considers these arguments in turn.

First, some might argue that perceptual beliefs as a rule are justified on the basis of beliefs concerning perceptual experience. For instance, on this view the belief "There is a brown cow in that field" is based on the belief "I am having a brown cow presented to me." Alston's objection is that this is simply falsified by experience. Seldom do we, when forming perceptual beliefs, also form any beliefs concerning perceptual experience, let alone are perceptual beliefs justified in virtue of experiential beliefs. (1991:82-83)

Second, some might argue that perceptual beliefs must be justified on the basis of perceptual cues. For instance, on this view the belief "That sculpture is at the far end of the gallery" is based, in some fashion, on various perceptual cues, e.g., the size of the retinal image of the sculpture, its apparent brightness, shading, texture, density and the like. There are several ways in which such perceptual cues might come into the basis of the belief. [1] There might be an unconscious inference made from premises regarding the cues to the concluding belief. [2] The connections between the cues and belief are purely causal. [3] Some sort of "subdoxastic" cognition or taking account of the cues is involved. Alston thinks there is no suffi cient reason to take account of these possibilities since their operation is "maximally hidden" and so epistemically unimportant. (1991:84-84)

Third, some might argue that justified perceptual beliefs (involving attributions of nonsimple sensory predicates to external objects) are based on certain assumptions (i.e., beliefs) concerning what is adequate to ground those beliefs in a justification-conferring fashion. These assumptions are what Alston calls "adequacy assumptions." For instance, on this view

On such a view, whatever is involved in moving from adequacy assumptions to perceptual beliefs (involving nonsimple sensory-predicate attribution) may satisfy the "based on" relation without being an explicit inference.

Alston has a number of criticisms of this view, but his most telling criticism suggests that the view confuses what must be true for a belief to be justified with what must be (justifiably) believed. If the basis of my belief that this is a house must include not only [A] a certain visual experience but also [B] the belief that this visual experience is a reliable indication of the presence of a house, then those elements, [A] and [B], taken together as the basis for a belief must, in fact, be reliably indicative of the truth of that perceptual belief. If, however, the adequacy assumption expressed by [B] was correct, then [B] need not be added to [A] in order to form a reliable basis for the belief in question. Alston identifies this as an example of the fallacy of "supposing that what it takes for a condition, C, to be sufficient for P...must itself be part of any sufficient condition for P. If the adequacy assumption is correct, then all that is in fact needed for my belief that this is a house to be justified is that it be true that my visual experience is a reliable indication of the presence of a house and not that I believe it to be so." (1991:84-86)

Fourth, some might argue that perceptual beliefs are justified on the basis of what Alston calls "contextual beliefs." Alston identifies three types of contextual beliefs: those that concern the setting, positional beliefs, and normality assumptions. Beliefs concerning setting would come into play in those situations in which the look of a certain object does not serve to identify it, but awareness of the setting does form part of the basis for an identification. Positional factors concern beliefs regarding the angle of view, distance, and the like. Normality assumptions include beliefs, for example, about the constancy and persistence of the physical objects of perception through time.

In the cases of positional beliefs and normality assumptions, we are not generally aware of them when we have them. Thus it seems unlikely that they play any doxastic role in the basis of our beliefs. It may well be that the various factors expressed in positional beliefs and normality assumptions must be in fact true in order for our beliefs to have an adequate basis, but that does not imply that the beliefs in question form even part of the basis. Beliefs concerning the setting obviously do sometimes play a role in the basis of our beliefs, but the question is whether they must in all cases of justified perceptual belief. Again, such beliefs may tell us more about what we think needs to be true in order for the basis of our beliefs to be adequate than actually forming part of the basis. These background beliefs, however, may function somewhat differently than adequacy assumptions. While such assumptions concern what would in general be an adequate basis for a certain belief, background beliefs are not so general and will vary from context to context. In this fashion, then, they may play an important role in overriding what has been taken to be an adequate basis. (1991:88-91)

From this discussion of the doxastic practice of sense perception, Alston's case emerges for the possibility of perceptual beliefs being justified immediately on the basis of experience. The various arguments given against immediate justification of perceptual belief have failed to demonstrate that no perceptual belief can be immediately justified. Thus immediate justification of perceptual beliefs appears to be quite possible, and if Alston's points are correct regarding background beliefs, adequacy assumptions, and the like, then immediate justification probably occurs in a wide variety of cases.

Nevertheless, I find it difficult to attribute the intuition against immediate justification on the part of many philosophers to mere confusion or philosophical mistake. In the following paragraphs we will be considering several points--in addition to or as variations on those already noted--that may contribute to the intuition against immediate justification. Even if Alston has argued correctly and these points do not count decisively against immediate justification, they do raise important problems and considerations concerning the shape of doxastic practices, forms of "inference", the nature of experience, the interpenetration of practices, and so on. If these points are relevant and central in respect to sense perception, then they will loom large in all those practices tied to sense perception (event memory, interpersonal and unique person practices, various interpretive activities, etc.) and may have parallels or analogs in practices less tied to sense perception (introspection, the moral sens e, fact memory, etc.).