Eric Vogelstein's response to Silver and an evaluation of Vogelstein's critique

In an interesting response characteristic of what we have called ‘the standard defence’ (of RE), Eric Vogelstein charges that Silver's reasoning depends on the truth of the following principle:

• Principle of Testimonial Defeat (PTD)  If I believe proposition P in the basic way, then if I hear testimony that ~P, I ought to believe neither P nor ~P (or equivalently, P and ~P act as defeaters for each other) unless I have a defeater for ~P other than P (in which case I ought to believe P), or a defeater for P other than ~P (in which case I ought to believe ~P).Footnote ⁹

But Vogelstein asks us to consider two counter-examples to PTD. First, suppose a person S receives testimony from you (and forms the corresponding belief in the basic way) that all his memory beliefs are false. In such a case, Vogelstein thinks that the only beliefs S holds that could effectively counter this testimony are based on S's ‘memory of what the world is (contingently) like’ – and that S consequently ‘[has] no candidate defeater for your testimony other than the very beliefs that your testimony calls into question’.Footnote ¹⁰ Second, suppose a moral realist receives testimony from a ‘sociopath’ that there are no moral truths. In such a case, ‘[the moral realist's] only defeater for the belief that moral truths are fictitious is [his] belief that there indeed exist moral truths’.Footnote ¹¹ PTD, says Vogelstein, would wrongly entail that the memory beliefs in the first example and the moral beliefs in the second are defeated – and this is because PTD fails to accommodate the intuition that basic beliefs can vary in strength. Hence, he says, PTD is false and should be rejected in favour of the following:

• Principle of Testimonial Evidence (PTE)  If I believe a proposition P in the basic way, then if I hear testimony that ~P and have no further defeater for P or ~P, I ought to weigh the strength of my inclination to believe that P against the strength of my inclination to believe that ~P (based on that testimony) in order to determine whether to believe P, ~P, or neither P nor ~P.Footnote ¹²

Vogelstein maintains that, according to PTE, one's inclination to believe a proposition p (i.e. the strength of one's disposition to believe p) counts as evidence for p. Thus, he says, when one is faced with two conflicting basic beliefs and has no candidate defeater for either of them, that person should adopt the belief that is associated with the stronger inclination (and hence, the stronger evidence). Consequently, Vogelstein thinks that in Silver's example, it may well be that Faith's inclination to believe that Christianity is true is stronger than her inclination to believe that Victor's testimony is true; if so, Faith does not acquire a defeater for her religious beliefs. And so the facts of religious pluralism need not threaten the rationality of basic religious belief.

Vogelstein's critique of Silver is flawed in several ways. First, note that his purported counter-examples to PTD are not really counter-examples at all. Consider the first example, where someone (call him Bill) tells another person (call him John) that all of his (John's) memory beliefs are false. Here, it seems that John would have reasons – not entirely dependent on the strength of his memory beliefs – to consider such testimony to be doubtful, at best. For example, John might reasonably believe that Bill could not possibly know that all his memory beliefs are false. Or, John might realize that this one person's testimony contradicts all of his experiences. These beliefs do involve memory, but do not acquire their epistemic status entirely on that basis. The objective probability that someone's testimony could at once be true and contradict all of one's experiences is extremely low, and could only take root in highly specific contexts, such as cases of complete amnesia or a mad scientist bringing it about somehow that all of one's memory beliefs are false. These reasons count against the credibility of Bill's testimony.

Now in all likelihood, John simply finds himself strongly inclined to believe that most of his memory beliefs are true. But even if inclinations to believe a proposition count as relevant evidence, as Vogelstein suggests, it is important to ask: ‘Is the epistemic force of my inclination offset or even overridden by stronger evidence to the contrary, such that the inclination can no longer epistemically support the relevant belief?’. In the ‘memory case’, above, we think that the fact that John is probably very strongly inclined to believe that most of his memory beliefs are true certainly counts as relevant evidence. And the reason for this is captured by the ‘principle of credulity’, that things are probably as they seem to be, in the absence of (relevant and sufficient) counter evidence.Footnote ¹³ Now since John has been given no relevant counter evidence, he might reasonably appeal to the principle of credulity and argue that his inclination to believe the proposition about his memory beliefs counts as evidence in favour of that belief. Such an appeal is not an appeal to John's own ‘memory of what the world is (contingently) like’ – for the principle of credulity is a general (and some say, a priori) principle about what is evidence for what.

(Things are very different, however, in the case of Faith and Victor. True, it is possible that Faith's inclinations count as evidence in the relevant way. But because Faith has received counter-evidence from a trustworthy source against the proposition that Christianity is true, her inclinations (which have also been challenged) are not sufficient for internal rationality in Plantinga's sense (the relevant sense here). If Faith had reasons or evidence that counted against Victor's testimony, then his testimony might not constitute relevant counter evidence. But as the example is set up, Faith has no such reasons; her situation is epistemically symmetrical, as far as she can tell, with that of Victor. But then from an epistemic point of view, Faith will have no reason to think that her religious beliefs are true and those of Victor are false. And the verdict advocated by Silver – namely, that Faith give up her religious beliefs – would be the epistemically appropriate move to make.)

Hence, in the ‘memory case’ above, John has negative reasons against Bill's testimony that all of his (John's) memory beliefs are false and a positive reason (based on the principle of credulity) for thinking that many if not most of his memory beliefs are true. These reasons together constitute, in the language of PTD, ‘a defeater for ~P other than P’. Thus, the potential defeater provided by Bill's testimony is ‘neutralized’ (to use Silver's term) immediately and does not become an actual defeater for John. So Vogelstein is mistaken when he says that ‘by PTD, [John] ought to believe neither that [his] memory beliefs are true nor that they are false’.Footnote ¹⁴ PTD has no such consequence. Hence, this first purported counter-example fails to hit its mark.

In the ‘sociopath case’, says Vogelstein, ‘my only defeater for the belief that moral truths are fictitious is my belief that there indeed exist moral truths’.Footnote ¹⁵ This, however, looks more like an interesting but irrelevant psychological fact about Vogelstein than an indication of the epistemological case with respect to belief in moral realism. Given all the literature defending moral realism via argument and evidence, there are plenty of candidate defeaters for the sociopath's testimony and the relevant disagreement need not be adjudicated on mere intuition.Footnote ¹⁶ Furthermore, it is questionable whether the sociopath's testimony should be taken seriously in the first place. Perhaps he is simply blind to moral realities, as some people are colour blind or tone deaf. What we really need is a sufficient reason to think that moral realism is false; simple testimony against moral realism amounts to little more than gainsay when there are respectable arguments to the contrary. So here again, there is ‘a defeater for ~P other than P’. Specifically, this defeater involves the conjunction of (1) the fact that there are good arguments for moral realism and (2) the fact that one person's testimony against moral realism is, given (1), of little epistemic value.

To be fair, Vogelstein is really suggesting that if the moral realist were to base his disagreement with the sociopath on ‘intuition alone’, PTD would issue the ‘counter-intuitive’ result that he is not rational in accepting moral realism. But the example fails to show this. If the sociopath's testimony is not to be taken seriously, then because PTD concerns cases where a proposition (P) and counter testimony (~P) ‘act as defeaters for each other’, the example itself is not relevant to PTD (and therefore does not show us a consequence of PTD). For, it is hard to see how a statement that is not to be taken seriously need act as a defeater for anything. On the other hand, if the testimony is to be taken seriously, then it serves as a potential defeater for the belief in moral realism. And the moral realist cannot appeal solely to his relevant inclinations in an effort to save his moral beliefs from defeat, since he has a good reason to think that those inclinations do not track truth.

Moreover, the sociopath's inclinations to believe that moral realism is false effectively make things epistemically symmetrical, as far as either party can tell. To appeal either to the very intuitions themselves, or to the fact that one holds one's moral beliefs ‘in the basic way’ will not help to break the symmetry. The moral realist needs to provide some reason or evidence (even if the sociopath won't be convinced)Footnote ¹⁷ – beyond his mere intuitions – in order rationally to maintain his moral beliefs. But then Vogelstein is mistaken when he says that ‘it certainly appears as if [the moral realist who bases his moral beliefs on intuition alone is] justified in maintaining moral conviction when confronted with the sociopath's testimony’.Footnote ¹⁸ Pace Vogelstein, PTD does not have dramatically counter-intuitive results; this second purported counter-example also fails.

Since PTD at this point remains largely intact, Vogelstein's PTE appears unmotivated. Moreover, PTE is flawed in at least two ways. First, PTE includes the provision that ‘if I have no further defeater for P or ~P’, then ‘I ought to weigh the strength of my inclination to believe that P against the strength of my inclination to believe that ~P’. But Vogelstein doesn't tell us when and under what conditions one ‘has no further defeater’ and can rely on mere strength of intuition in adjudicating what to believe. In our discussion of the above examples, we saw that there were epistemically significant reasons (independent of mere inclinations to believe) which were relevant in determining how the conflicting basic beliefs were to be adjudicated – and these reasons functioned as ‘further defeaters’ in the relevant sense. Vogelstein has yet to present a plausible case in which one in fact ‘has no further defeater’.

In general, liberal estimates about how frequently such cases occur seem to be mistaken. And even if Vogelstein were to present such a case, the claim that inclinations to believe a proposition are sufficient to ward off defeat, even in the face of serious counter-testimony, is implausible. Such testimony, after all, calls into question whether those inclinations are tracking truth. Second, PTE seems to suggest that a person could rationally hold basic beliefs in the face of counter-testimony simply by announcing that she ‘has no further defeater’ and claiming a stronger inclination to accept them than the counter-testimony. But this allows basic beliefs of this sort to be effectively immune from defeat, and reduces epistemology to dogged psychological prejudices.

But isn't it still true that PTD ‘entails the highly implausible proposition that basic beliefs do not vary in strength’?Footnote ¹⁹ We concede that basic beliefs vary in strength. However, it seems that PTD could be amended in some way to accommodate this fact. And in any case, this concession will be of little help to Vogelstein's critique of Silver. For, it is reasonable to suppose that basic beliefs (however strongly held), on reflection, can be supported by epistemically significant reasons. These reasons are independent of mere inclinations to believe, and bear importantly on how one ought to adjudicate between conflicting basic beliefs.

Plantinga himself acknowledges something very much like this, as follows:

Notice that Plantinga does not simply pronounce that your belief (p) that a counsellor's seducing a client is wrong is rational so long as your inclinations favour p. Instead, he also lists several reasons which, even if they won't convince those who demur, bear importantly on the rationality of continuing to hold p in the face of disagreement.

Plantinga and religious pluralism

Given the flaws we've identified in Vogelstein's attempt to defend RE, Plantinga's emphasis on the use of epistemically significant reasons to support a ‘contested’ basic belief seems more plausible. This improved defence may be characterized by means of the following argument.

1. (1) If a person S (i) becomes aware (or more fully aware) of the facts of religious pluralism and their implications with respect to her particular religious beliefs,Footnote ²¹ and (ii) reflects more carefully on the epistemic support that her beliefs enjoy (via certain of her experiences) and more strongly believes that these beliefs are likely to be true, and (iii) forms and maintains her beliefs in the right way, then S's beliefs can be warranted in spite of S's awareness of the facts of religious pluralism.

2. (2) Therefore, S's awareness of the facts of religious pluralism needn't defeat her religious beliefs.

However, in order for S's greater conviction in (2) to be rational, she must take it that her religious beliefs are not epistemically inferior to, or on a par with, the religious beliefs of those who disagree with her. Here, Plantinga agrees:

The fact that she thinks she has a source of warranted belief denied to the other person functions as an epistemically significant reason for believing that the relevant beliefs are not on an epistemic par. Plantinga then makes the important admission: ‘If the believer concedes that she doesn't have any special source of knowledge or true belief with respect to Christian belief … then, perhaps, she will have a defeater’.Footnote ²³ In the remainder of the paper, we argue that absent the use of relevant evidence, etc., a person in Faith's situation will indeed have ‘nothing not available to those who disagree with her’ and will have a defeater for her belief.

Silver's example revisited

Victor's testimony provides Faith with a defeater, not only for her belief that Christianity is true, but also for her belief that she has a special source of religious knowledge. Hence, Faith cannot legitimately cite her belief that she has the IIHS (or something similar) so as to break the apparent symmetry between the relevant beliefs. And so it seems thus far that Plantinga's defence of RE will fare little better than Vogelstein's. But perhaps there is some other way for Faith to reasonably deny that Victor's testimony has greater or symmetrical credentials in relation to her religious beliefs. For example, Faith might think that she has better grounds (given by her relevant experience) in support of her beliefs than she has (given by Victor's testimony) in support of Victor's beliefs.

Now it may be that another's testimony is insufficient to defeat the beliefs one forms from one's own lived experiences. Still, that Victor is entirely truthful and a friend of Faith's gives Faith more reason for taking his testimony seriously than would be the case if the testimony came from a stranger or even an acquaintance. Also, a thorough knowledge of the facts of religious pluralism implies an awareness that there are millions of epistemically virtuous people who report profound and transformative experiences upon which they come to form mutually exclusive religious beliefs. This spreads the quality of testimony across a larger quantity of testifiers, through different religious traditions. Even if Faith were to amass testimony from many epistemically virtuous Christians, there would, in all probability, be just as many (or more) honest and sincere individuals who could provide counter-testimony. And this sufficiently casts doubt on Faith's belief that her religious beliefs are epistemically better grounded than Victor's; at the very least, it makes it likely that there exists epistemic parity between hers and Victor's beliefs.

But if Faith's religious beliefs are on a par with (or inferior to) those of Victor, her beliefs cannot be rationally maintained. Given the above discussion, it is very hard to see how Faith's beliefs could escape the symmetry and avoid defeat apart from, at a minimum, some argumentation, evidence, or inference to other beliefs.Footnote ²⁴ But if evidence, etc. thus becomes necessary, then we have secured the first of our two central theses: that basic religious belief is subject to defeat.

One might then object: ‘Even if basic religious belief is initially defeated, once the defeater is itself defeated, such belief can again be held in a properly basic way.’ To show the inadequacy of this suggestion, we need to briefly articulate two kinds of responses to defeaters.

First, there is what we'll call the ‘Drain-O’ response (because it works like Drain-O in un-clogging a kitchen sink): once the ‘garbage’ (the defeater) is disposed from the pipes, as it were, the sink functions normally as before and no longer needs the Drain-O. Second, there is the ‘table-leg’ response, following the manner in which a table leg upholds a table. Here, the presence of what effectively defeated the defeater is continually necessary for the once-defeated-belief to be rational again.

For example, suppose Jim mentions that he saw Sue at the dance last night, and you form the belief (p) that Sue was there. Then another colleague of yours, Paul, reports to you that Jim was out of town on business last night, and that, since she usually avoids social dances, Sue was probably not at the dance. Given that you believe Paul to be trustworthy (and that you don't know anything one way or another about the trustworthiness of Jim), you realize that you have a defeater for your original belief (p), and so now form the belief that Sue probably did not go to the dance. But suppose you come to recognize that Paul is intoxicated. Your reason for doubting Jim's testimony is no longer well-grounded and you are now rational in believing p in the same manner in which you were when you first formed this belief. Your belief that Paul is drunk does not itself support your belief p, but instead (like Drain-O) ‘flushes out’ Paul's testimony and makes it unable to defeat p.

By contrast, in ‘table-leg’ cases, the belief is not rational apart from the defeater-defeater. And this, we have been arguing, is the case with respect to Faith: in the absence of relevant evidence, etc., Victor's testimony gives her a defeater. The continued rationality of her religious beliefs depends on that evidence (the defeater-defeater) – even if she doesn't or needn't have such evidence continually before her mind.