RESPONSIBLE BELIEVING

S. Joel Garver

CHAPTER FOUR

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

Section I

Bhagavad-Gita

If we are to understand voluntary control over belief and responsibility for belief, we will also need to understand the activities involved in coming to beliefs, how belief-formation is connected to non-doxastic practices, and what it is to engage a doxastic practice properly. To this end I will lay out, as clearly as I can, how Alston conceives of doxastic practices. That will be the first part of our discussion in this chapter.

While Alston gives us a lot of general direction regarding belief forming practices, he is short on specifics of particular practices (with the notable exceptions of perception and mysticism) and how doxastic practices interrelate with non-doxastic practices. In the second part, then, I will explain some specifics of perceptual doxastic practice by drawing on Alston's account along with Alan Millar's discussion of belief and experience. The third section will be a brief digression into the question as to whether we possess grounds for all our beliefs, particularly beliefs that function as presuppositions of doxastic practices or of inferential patterns. The fourth section will conclude the chapter with an overview of Humean suggestions regarding intellectual habits, various discussions of the role of trust and testimony in knowledge (Ross 1995), and Mark McLeod's theories regarding interpersonal and unique person practices (1993).

All of these points will lay a groundwork for a discussion of the interrelation of various practices--doxastic and non-doxastic. For that discussion, which will occupy my fifth and sixth chapters, I will be using Foucault's explication of the connection of knowledge to power and desire and, in the seventh chapter, Michael Stocker's connection of emotions and knowledge. The fifth chapter will conclude by taking various examples--particularly that of domestic violence--in order to explicate Foucault's views in some depth.

Alston's basic definition of a doxastic practice is "the exercise of a system or constellation of belief-forming habits or mechanisms, each realizing a function that yields beliefs with a certain kind of content from inputs of a certain type" (1991:155). While Alston tends to go with the prevalent terminology of "mechanisms", for my part I shall prefer "habits", "activities", "practices", "processes" or the like. Our analysis of indirect influence (character foresight, and so on) tends away from mechanistic and reductionistic terminology in favor of these terms that, in various ways, bring out the more organic, personalist, and responsible aspects of belief-formation.

Doxastic practices are individuated on the basis of their distinctive sorts of inputs, outputs, functions, and--if present--overrider systems. For instance, the perceptual belief forming practice is taken by Alston to be "a very wide function that takes inputs of the type: an experience of the sort S would be inclined to take as a case of X's appearing O to S, and yields outputs of the correlated type: a belief of the form 'X is a O'" (1991:156). There are, however, wider and narrower ways of individuating practices, though there seems to be no one correct system of individuating practices short of a clear understanding of the neural substrata that underlie the practices (assuming there's even one answer to that question). A number of subpoints will further explain this general approach to doxastic practices.

First, Alston distinguishes between practices that are "generational" and those that are "transformational". In generational practices beliefs are produced by nondoxastic inputs (e.g., experiences, perceivings), while in transformational practices beliefs are produced from other beliefs as inputs. Certain cases of sense perception might well be generational, while cases of strictly logical inference are transformational. Those practices which have mixed inputs (doxastic and nondoxastic) are grouped with generational practices. The distinction is necessary since generational and transformational practices have distinctive characteristics. For example, generational practices typically give us access to reality and use conceptual schemes. Transformational practices, however, can involve purely hypothetical data and enable us to perform operations upon them that are neutral to conceptual frameworks. (Alston 1991:157)

Second, doxastic practices, Alston insists, also have an evaluative side. He writes concerning the vast majority of our belief forming habits:

Moreover, our doxastic practices include what he terms "overrider systems." The justification attributed to a belief on the basis of a practice can also, by the very same practice, be overridden--by a capacity to see either that there are reasons to doubt the veracity of the belief in question (rebutters) or that there are reasons to doubt that the basis of the belief yields justification in the way which the practice authorizes (underminer). For instance, in order to determine whether there a re any overriders to

This discovery of overriders draws on other beliefs from the perceptual practice in question (and possibly from other practices) that are sanctioned by the perceptual practice for that purpose. In light of some of our earlier discussion, we can also note that underminers may well include both reasons to think that the mechanics of practice have failed in producing justification (e.g., there is reason to think that the blow to the subject's head is altering her perception) and reasons to think that the practice has been engaged in improperly (e.g., there is reason to think that the subject is allowing his emotions to cloud his judgment of the situation).

For a belief to have unqualified justification (as opposed to merely prima facie) it need not be determined that there are no overriders, but simply be the case that there are none. Nevertheless, overriders, it seems, ought to be the sorts of reasons that are accessible. There is a close relationship, I have suggested, between the rationality of taking a belief to be (prima facie) justified and justifying that belief. Both of these conditions arise out of the nature of the doxastic practices that are rationally taken to confer justification on the belief in questions. Overriders, however, are also part of the nature of these doxastic practices. Furthermore, we are not simply interested in the rationality of taking a belief to be prima faci e justified. We also care about the rationality of taking a belief to be unqualifiedly justified. In terms of criticism and defense of beliefs this issues into a concern for being able to show that it is rational to take a belief to be unqualifiedly justified. If that is case, however, then there must be some access to overriders and that access is part of the nature of doxastic practices themselves.

A number of ambiguities enter here. There is a question whether an overrider must be something accessible to an individual subject, regarding the justification had by his own beliefs; whether the overrider must be accessible by means of the doxastic practices he engages in when forming the beliefs in question or merely by the practices he more widely engages in; and whether the overrider must be accessible by means of the doxastic practices in the way he actually engages in them or in the way they ought to be engaged in or as they are actually engaged by his social context as a whole. Furthermore, what if an overrider is accessible via someone else's telling the subject?

Though these are ambiguities into which Alston does not always delve (but see 1989:177-179), his considered view appears to be that overriders are confined to the subject's own knowledge and justified belief. Still, even that is open to qualification: is it only the subject's actual store of knowledge when he forms a belief that can override justification, or can overriding knowledge be picked up later through, e.g., interaction with others? If so, does the belief switch from being unqualifiedly justified to being only prima facie justified since there were no overriders when the belief was formed but there are once the further knowledge is obtained?

Such questions are important, especially for our discussion of obligations in respect to belief. For instance, what if an overrider is something that can be possessed by the members of a doxastic community as a whole (though not necessarily each and every member of that community)? Then the existence of a socially established overrider which is, nevertheless, inaccessible to the subject through no fault of her own, may render excusable an otherwise impeachable belief or process of belief-formation.

Considerations like those mentioned in the last few paragraphs (as well as my discussion of overriders in Chapter Three) lead me away from Alston's position in the context of doxastic practices. Rather than limiting overriders to the perspective of the individual subject, I suggest that overriders (within a doxastic practice approach) be seen more socially. I think, then, that overriders should include any knowledge or justified beliefs that are actually or potentially possessed by the members of a doxastic community by means of a shared practice (where the beliefs would constitute reasons for taking a prima facie justified belief to be false or to be formed in such a way that the usual justification-conferring practice failed).

Third, doxastic practices are mutually dependent. Some examples: reasoning takes its premises from other practices; the overrider system for sense perception is essentially dependent on memory and inference; practical use of beliefs obtained within one practice depend upon memory and reasoning for further use; some perceptual beliefs require inferences about our experiences; conceptual schemes used in recognition abilities draw on other beliefs from memory and reasoning; we must use many practices when trying to establish the validity of one since that one must cohere with those other practices to be well-established. (1991:159-160)

Fourth, there is an irreducible plurality of practices. While practices may cohere and work in a unity, there is a fundamental diversity of them. The mode of justification particular to one practice cannot be usefully and informatively reduced to some underlying mode of justification that is common to all. (1991:162-3)

Fifth, while some doxastic practices may be purposively acquired at some point (e.g., interpretation of art objects?), most practices--and the most fundamental--are ingrained habits long before any reflective critique is possible. How this might affect responsibility for our doxastic life is a topic that has been addressed to some degree, but will be taken up more extensively below. It is a problem akin to the Aristotelian problem of how someone raised without the virtues--which by their nature are best acquired in youth--can be held responsible for his incontinence or viciousness as an adult. (1991:163)

Sixth, our doxastic practices are involved with wider spheres of (non-doxastic and mixed) practice. Alston does not explicate this point in general or in great detail, but does give it specific treatment in the context of theistic belief (sense perception too). He writes, "the practice of forming perceptual beliefs about God is thoroughly interwoven with orienting oneself to God in attitude, feeling, and action" (1991:163). It will be the burden of this and following chapters to show the extent to which this is generally--perhaps without exception--the case across practices. It will turn out, in fact, that rather often the "outputs" of non-doxastic or mixed practices can serve as "inputs" or "grounds" in the formation of beliefs within doxastic practices (e.g., it might very well be the case that trust is among the grounds or inputs or habits of belief-formation in the doxastic practices that are associated with testimony and taking something on authority).

Seventh, doxastic practices are "thoroughly social: socially established by socially monitored learning, and socially shared" (1991:163; emphasis mine). This is another point which will be of great interest. If we are at all responsible for our doxastic lives, then we must have some choice in the way we acquire, maintain, and engage doxastic practices. But how is such choice possible if our doxastic lives are largely bequeathed to us by our culture? This is a question that parallels the nagging question in postmodern political and social theory as to how political change and resistance is possible if Foucault, Deleuze, Lyotard, Baudrillard, or others are at all correct in their assessment of the roles of power, desire, commodification, or the death of meaning. The answers and critiques given within and against postmodernism may prove helpful for our question as well.

Eighth, Alston says that "We should not suppose that doxastic practices are immutable. They can and do change" (1991:163).

Finally, "each practice possesses its own distinctive set of foundational presuppositions" (1991:164). For example, it is among the distinctive presuppositions of sense perception that there is a world of physical objects and that sense perception itself is generally reliable.