RESPONSIBLE BELIEVING

S. Joel Garver

CHAPTER THREE

Believing as One Ought:
Well-Established Practices

Section V

Let us connect these points up with what we have said so far concerning responsibility for belief, voluntary control over belief, deontological notions, showing justification, and justifying belief. In terms of voluntary control over belief we found that the only viable version was that of indirect influence. Responsibility for belief and prescriptive evaluation of belief are not a matter of the content of individual beliefs or effective control over them, taken individually. Rather responsibility and evaluation are seen more holistically as attaching to practices underlying belief-formation: both those specific practices that we have called "doxastic practices" and more general patterns of practice that are intertwined with and underlie doxastic practices.

These various kinds of practices involve [1] those "that bring influences to bear, or withhold influences from, a particular situation involving a particular candidate or set of candidates, for belief"; and [2] those "that affect our general belief-forming habits or tendencies" (1989b:138). In respect to the first category, we are concerned both with [a] what kinds of doxastic practices are engaged in, given a particular candidate for belief in a particular situation (e.g., use of logical practice or practices to determine if a particular formula is derivable from the given axioms) and also with [b] how doxastic practices are engaged in so that a particular belief is formed in a certain context, including how relevant non-doxastic practices affect doxastic engagement (e.g., allowing one's emotion of anger to cloud judgment of another's intentions).

In respect to the second category, we are concerned both with [a] general acquisition of doxastic practices and the influence of non-doxastic practices in that process (e.g., the ability to effectively interpret art objects) and also with [b] acquired patterns of engaging in certain doxastic practices (e.g., a judicious, discerning, and restrained trust of testimony).

In these ways then, the whole realm of belief-formation can arguably be taken to be under our indirect influence and, therefore, something for which we are responsible. Even in respect to those practices of which we have no real choice but to engage in (e.g., sense perception, memory, introspection), we have extensive choice as to how we engage in them. We have come to trust or distrust our eyesight, for example, in respect to certain objects, at certain distances, within certain contexts, given certain lighting conditions, and so on. The beliefs we form by sense perceptual practices are accordingly discerning. Even sense perception, then, though a compelling source of belief, is not insensitive or wholly promiscuous in producing belief. Likewise, we often know what kinds of memories are best taken to be reliable and which to distrust.

Furthermore, there are certainly prohibited abuses of even the most ingrained habits of belief. These habits can, given the appropriate situation and influences, be seriously dislodged from their more conventional moorings (though such behavior may sometimes be excusable given the contexts). For example, lack of sleep for more than two days can produce "travelers' psychosis," a disruption of ordinary adrenachrome levels producing an effect often symptomatically indistinguishable from an acute schizophrenic episode or an LSD induced "trip" (hallucinations, overly bright colors, voices, distortion of visual perspective, etc.; Luce and Segal 1967; Hoffer and Osmond 1966). Beliefs formed in such a situation would not be based on adequate grounds due, perhaps, to what an agent has done. Thus, neglectful failure to provide oneself with rest can issue in epistemic culpability.

Another example. Failure to take responsibility for actions and omissions combined with stress, harbored bitterness, exaggerated self-expectations, lack of social outlets and interaction, and so on can lead to forms of catatonic psychosis involving suppression of sense perception and distorted memory (Glasser 1965; Mowrer 1961; Adams 1970, 1973; Szasz 1960; Modrow 1992). Thus, neglectful failure to take appropriate means in solving personal problems can issue in epistemic culpability.

Again. Children who have been seriously abused (physically and/or sexually) show a remarkable ability for self-hypnosis by means of which they can desensitize themselves to the abuse. Nevertheless, under hypnosis suppressed memories of the abuse and even of actual physical pain can sometimes be recovered (false memories are also a psychological phenomenon that has been uncovered by research in this area; on various phenomena see Finkelhor 1984; Fredrickson 1992; Briere and Conte 1993; Terr 1992; Loftus and Ketcham 1994). Thus even introspection can be dislodged (the children had pain that didn't feel like pain at the time!).

While many of these cases are extreme and excusable given the situation (as lying to Nazis is excusable given the Jewish family hiding in your attic), these cases still represent what conditions we can be lead to by indulgence in what is epistemically prohibited. If these extremes can be reached, then the possibility of more minor (and perhaps more culpable) epistemic straying is quite palpable (for a fascinating, albeit dated, account of these problems see "The Occasions, Causes, Nature, Rise, Growth, and Remedies of Mental Errors" by 17th century English Puritan theologian John Flavel 1820:418-492).

Since responsibility for belief is widespread across and within doxastic practices we can sensibly speak of epistemic obligations. This is, in fact, what we regularly do in ordinary speech and to such an extent that prescriptive notions and responsibility dominate the spectrum of doxastic phenomena. That is also why we tend to approach cases in which responsibility is significantly lacking in terms of excuse or justified error. They are exceptions rather than the rule. Let us explicate epistemic obligations, then, in terms of doxastic practices.

The deontological conception that I put forward as best representing what might be said regarding responsibility for belief was the following:

We saw, however, that this cannot really count as justification since it does not coincide with a truth-conducive concept of epistemic justification--the sort of concept we are taking to be true. What I want to do now is to shift the focus of [DJ] away from analyzing under what conditions "S is justified in believing that p." I want to instead focus on analyzing under what conditions "It is rational to suppose that S is prima facie justified in believing that p" and, thus, the conditions that are appealed to in justifying one's beliefs. We are no longer concerned, then, primarily with being justified or having justification. Rather we are concerned with practical rationality, making practical ascriptions of justification, and the defense of beliefs against critics (particularly what I termed the "second line" of defense).

Alston's doxastic approach tells us that "It is rational to suppose that S is prima facie justified in believing that p" only when p is the output of a well-established doxastic practice. That is, it is in fact reasonable to suppose a subject to be justified in believing that p only when p is the output of a doxastic practice that is, in fact, well-established. We can, therefore, interpret "fulfilling intellectual obligations" in terms of engaging in well-established doxastic practices (what is permitted) and refraining from engaging in any unestablished or deviant practices (what is prohibited). Thus, doxastic evaluations and prescriptions are now firmly attached to practices.

Intellectual virtue requires acquiring doxastic habits that are well-established. Therefore, those practices which are associated with intellectual virtue may also, we are reasonable to suppose, be those that produce prima facie justified beliefs. Intellectual vice engages in unestablished and deviant practices. Therefore, those practices or ways of engaging in practices that are associated with intellectual vice may also, we are reasonable to suppose, be those that fail to produce prima facie justi fied beliefs.

I would also add our further condition at this point (the caveat regarding "properly engaging in a practice"). Doing what is intellectually permitted in respect to a particular belief requires not only engaging a well-established practice, but also engaging in it properly. That is to say, refraining from what is intellectually forbidden in respect to a particular belief includes not only refraining from engaging in unestablished or deviant practices, but also refraining from engaging in any practice improperly. Thus intellectual virtue requires not only engaging in the right practices, but engaging in them excellently as well.

It seems, then, that the deontological notions at work in some epistemologists' conceptions of justification may not be entirely useless. We will not apply them to whether or not a belief is actually in the state of having justification. That identification between deontological and truth-conducive justification is, as I have argued, a dead end. But we can apply [DJ] to the practical rationality of taking certain beliefs to be justified (which may, in some cases, result in not only justifying a belief but also showing it to be justified). To wit:

With [DJ'] in hand, then, we can see how the concerns of justification and epistemic prescriptions coincide. It is also clear what conditions must be appealed to if we are to rationally justify our beliefs: we must try to show that intellectual obliga tions have been met.

Of course, "showing" that intellectual obligations have been met is a rather slippery and subjective notion. What it takes to show one person may be very different from what it takes to show another and some people will doubtlessly remain unconvinced despite what you can show (e.g., James Ross relates a conversation with a well-known logician who told him "Jim, if you could prove that God exists from the law of noncontradiction, I'd give up the law").

Nevertheless, it is clear how one would go about the second line of defense in showing a belief to be rationally taken as justified--one would show that the belief stems from a certain practice and that practice is well-established. The practice in question provides consistent outputs; it is self-supporting (pragmatics, predictability, etc.); it is engaged in by many people over a long period; its outputs do not contradict those of other practices; it coheres with and is caught up in webs of mutual support with other established practices; and so on. How many arguments need to be given and how far one needs to push each line of argument depends on the critic.

The following points, at least, are clear. [DJ'] comports well with actual cases in which we single out a belief or beliefs for explicit prescriptive evaluation. Take the following accusation: "I don't know how you can believe that Samantha is the primary reason for her and Harry's divorce!" The blame implied in this belief might well be countered by a response of this sort: "I'm a good friend to both Samantha's sister and her son Roger--yes, the one who's in college--and while they both love Samantha dearly and want the best for her, even they can't help but hold her responsible." This is a defense of the first kind--an indication of the practice grounding a belief with an implicit assumption of its well-established reliability and proper engagement. It is an attempt to show the reasonableness of taking a belief to be justified within the framework of a shared practice.

Further challenge, if pursued, would likely be made with indications of the unreliability of Sam and Roger--inconsistencies, contradictions, half-truths, and the like. That would be an attempt to show that the practice that was engaged in is unestablished, deviant, or, at least, improperly engaged in. And that, if successful, would show the unreasonableness of taking the belief in question to be justified. Blame, then, is issued and answered with what amounts to an appeal to [DJ'], a concern with whether we can reasonably take the belief to be justified and, therefore, to be seen as blameless.

In terms of every day ascriptions of justification, applications of prescriptive language, and evaluations we are guided by what we rationally take to be appropriate grounds and practices. The concerns mentioned in connection with justifying belief are largely internalized and assumed so that in many areas we operate by an "innocent until a good reason is given to suspect guilt" policy. We also, as we shall see in later chapters, are guided by affective states--e.g. cognitive dissonance, disgust, satisfaction, and so on. Part of epistemic virtue for us, then, also consists in the practice of epistemic evaluation being well-established--which includes appropriate sensitivity to the issues we have discussed thus far.

Before moving on, however, let us make an important distinction between an epistemic norm and a more properly ethical one. A subject might engage a doxastic practice in an improper fashion (i.e., a case of engaging a practice improperly--an epistemic violation) but this is not the same as a case in which a subject engages a practice that is not the proper one to engage (i.e., a case of engaging an improper practice--a more ethical violation).

Consider the following cases. Ken and Floy have been married for many years and Ken is generally able to determine his wife's mood from subtle cues that are insufficiently overt for almost anyone else to detect, let alone interpret. The practice involved here, among other elements, requires prolonged exposure to another person accompanied by attention and affection. It is a well-established, justification-conferring practice. If, however, Ken tries to judge his wife's mood while he is angry with her, the practice will no longer produce beliefs that could reasonably be taken to be justified. Ken knows this very well and he has quite a bit of control over whether or not he will judge his wife's mood while he is angry. Nevertheless, he sometimes does so. The beliefs produced, however, cannot reasonably be taken to be justified because Ken has engaged the practice improperly. The obligations violated here are epistemic or doxastic obligations.

In the second case, the same sort of practice is being engaged by Professor Q in respect to a certain Joseph, her student. She has had prolonged exposure to Joseph, giving him a great deal of attention, and fostering affection towards him. It is, however, most improper for any professor to engage this practice in respect to her student. Nevertheless, Professor Q is not engaging in the practice improperly, but rather she is engaging in an improper practice. In fact, since it is a well-established practice, the resulting beliefs are to be taken as prima facie justified. The obligations violated here are not epistemic or doxastic, and though they are connected to belief-formation, they are more properly seen as ethical.

In general then, engaging in an improper practice can still lead to beliefs that are to be taken as justified when the practice that has been engaged in is well-established and engaged properly.

Let us address one final issue, then, by returning to an earlier discussion. Some time earlier in this chapter, two cases were used to show that deontological and epistemic justification do not coincide. We can examine them again briefly, using [DJ'] and focussing on whether or not we can attribute justification to the beliefs in question. In the first case, an individual was deontologically justified in perceptual beliefs that were caused by malfunctioning perceptual mechanisms. The beliefs, therefore, lack epistemic justification. On [DJ'], however, we could reasonably take the beliefs to be prima facie justified. That ascription, however, either [1] could be withheld (if the beliefs, by producing inconsistencies, gave an indication that perce ptual practices were wrongly engaged) or [2] could be overridden (if the belief, by producing inconsistencies gave an indication of a noncuplable malfunction of the mechanisms of the practice). This case indicates, therefore, that by [DJ'] reasonable ascriptions of justification do sometimes accurately track actual justification.

In the second case, beliefs based on reliable testimony were epistemically justified, though trusting the testimony was rashly done. In fact, if the trust given to the witness had been questioned, it would have mistakenly been found wanting. On the basis of [DJ'] we could not, therefore, help but be reasonable in taking the belief to lack prima facie justification. Assuming this was a sensitive situation in which a high level of trust needed to be established, by [DJ'] it would be reasonable to find the beliefs to lack prima facie justification since we would discover that the practice in question was engaged in improperly. Even if the practice had been properly engaged in, the subject would probably have never formed the beliefs in question. (Even if they had been formed, we would reasonably deem them unjustified by means of [DJ'] though we might excuse the subject from his illicit behavior.)

In accordance with [DJ'], then, we would not be reasonable to take the belief to be prima facie justified. This is an unusual case and that determination is, from a practical perspective, correct and wise. After all, in the actual situation, as specified, there would be no way to weigh the witness's trustworthiness. Discretion being the better part of valor, the one possessed of intellectual virtue would probably not acquire the beliefs in question. Nor would we want it to be reasonable to have acquired them, since that would encourage intellectual vice. This case also indicates that by means [DJ'] we do not track actual justification perfectly, a point which we have already admitted.

Let us conclude this chapter. The arguments for [DJ'] connect epistemic obligations to when we can reasonably take beliefs to be justified. Obligations can issue practically in actual blame and, in such a case, we are sometimes led to show the reasonableness of taking beliefs to be justified, that is, to justifying our beliefs. We rely on how well-established certain practices are (or can be taken or shown to be) and this points towards the importance of coherentist considerations in justifying belief.

Thus we have uncovered, perhaps, a major component of the appeal of deontological conceptions and coherentism to epistemologists as they analyze what it is for a belief to have justification. Their confusion arises from a conflation of what it is to be justified, what can be reasonably taken to be justified, and how to convince others to take them to be justified. They have confused justification with justifying. That confusion has hopefully been cleared up. As we proceed we will do well to keep in mind that it is the latter cases that are our concern--practical determinations of justification (what is reasonable and what can be shown to be reasonable). Our discussion of the basic framework of voluntary control over belief, deontological justification, holism, and ethical evaluation is essentially complete. Now we must turn to deeper examinations of the varying structures of doxastic practices and their connections to wider spheres of practice.