Bergmann’s Dilemma: Exit Strategies for Internalists



Yüklə 131,13 Kb.
səhifə1/3
tarix08.12.2017
ölçüsü131,13 Kb.
#14713
  1   2   3

Bergmann’s Dilemma: Exit Strategies for Internalists

Forthcoming in Philosophical Studies

By Jason Rogers and Jonathan Matheson
ABSTRACT: Michael Bergmann claims that all versions of epistemic internalism face an irresolvable dilemma. We show that there are many plausible versions of internalism that falsify this claim. First, we demonstrate that there are versions of “weak awareness internalism” that, contra Bergmann, do not succumb to the “Subject’s Perspective Objection” horn of the dilemma. Second, we show that there are versions of “strong awareness internalism” that do not fall prey to the dilemma’s “vicious regress” horn. We note along the way that these versions of internalism do not, in avoiding one horn of the dilemma, succumb to the dilemma’s other horn. The upshot is that internalists have many available strategies for avoiding dilemmatic defeat.
I. INTRODUCTION: A DILEMMA FOR INTERNALISM
Internalism is the genus for a certain class of views about epistemic justification. The defining characteristic of this genus, according to Michael Bergmann (2006a), is its claim that there is an awareness requirement on justification. This awareness requirement is the requirement that, for a belief to be justified for some person at some time, the person holding the belief at that time be “aware (or at least potentially aware) of something contributing to its justification” (Bergmann 2006a, 9).1 Put more precisely, internalism about justification endorses the following thesis:

The Awareness Requirement (AR): S’s belief B is justified only if (i) there is something, X, that contributes to the justification of B—e.g., evidence for B or a truth-indicator for B or the satisfaction of some necessary condition of B’s justification—and (ii) S is aware (or potentially aware) of X (Bergmann 2006a, 9).
The conditions mentioned in AR can be further specified in various ways. Specific versions of internalism will endorse AR and differentiate themselves by specifying either the sort of awareness mentioned in AR’s condition (ii), or the sorts of things that can be the “justification-contributors”2 mentioned in AR’s condition (i), or both.

Bergmann claims that all versions of internalism face an irresolvable dilemma.3 Every version of internalism maintains that a necessary condition for some belief’s being justified is that the subject possess actual or potential awareness of some justification-contributor for the belief (as per AR, above). The required awareness will be either “strong awareness” or “weak awareness.”4 If the required awareness is strong awareness, then, Bergmann argues, internalism “has vicious regress problems leading to radical skepticism” (Bergmann 2006a, 14). If the required awareness is weak awareness, then, alleges Bergmann, internalism is vulnerable to the “Subject’s Perspective Objection” (SPO). In either case, the argument goes, we should not endorse internalism.

It will be helpful to present this dilemma a bit more formally, so as to better consider its premises. The following presentation of the dilemma is Bergmann’s:

A Dilemma for Internalism


  1. An essential feature of internalism is that it makes a subject’s actual or potential awareness of some justification-contributor a necessary condition for the justification of any belief held by that subject.

  2. The awareness required by internalism is either strong awareness or weak awareness.

  3. If the awareness required by internalism is strong awareness, then internalism has vicious regress problems leading to radical skepticism.

  4. If the awareness required by internalism is weak awareness, then internalism is vulnerable to the SPO, in which case internalism loses its main motivation for imposing the awareness requirement.

  5. If internalism either leads to radical skepticism or loses its main motivation for imposing the awareness requirement (i.e., avoiding the SPO), then we should not endorse internalism.

  6. Therefore, we should not endorse internalism.

Premise I of the dilemma simply follows from our description of the awareness requirement as an essential feature of internalism.5 We accept premise V for the sake of argument.6 Premise II follows from the fact that strong awareness and weak awareness are exhaustive and exclusive kinds of awareness. Briefly,7 strong awareness is awareness that involves “conceiving of the justification-contributor [mentioned in AR’s condition (i)] . . . as being in some way relevant to the justification or truth of the belief [B]” (Bergmann 2006a, 13; emphasis in original), where this “conceiving” involves a certain kind of conceptualization of the justification-contributor8—it involves the application to that contributor of the concept being in some way relevant to the appropriateness of B.9 In contrast, weak awareness is awareness that does not involve this conceptualization.10 Having accepted the preceding premises for the foregoing reasons, we will focus our attention in what follows on premises III and IV of the dilemma. We will argue that each premise is false, showing in the process that both internalists who construe AR as requiring only weak awareness and internalists who construe AR as requiring strong awareness can avoid Bergmann’s dilemma.



II. REJECTING PREMISE IV: WEAK AWARENESS INTERNALISM AND THE SPO

What reasons are there for thinking that premise IV of the dilemma is true? Bergmann offers several considerations, all of which turn on the nature of the SPO and the nature of weak awareness.

We begin with some words about the SPO. The SPO is an objection often leveled by internalists against externalist theories of epistemic justification. The SPO claims the following:

The Subject’s Perspective Objection (SPO): If the subject holding a belief isn’t aware of what that belief has going for it, then she isn’t aware of how its status is any different from a stray hunch or an arbitrary conviction. From that we may conclude that from her perspective it is an accident that her belief is true. And that implies that it isn’t a justified belief.11
To take just one example, Laurence Bonjour offers something like the SPO against reliabilist theories of justification in his presentation of the well-known case of Norman the clairvoyant.12 After describing that case, Bonjour writes (in language very close to the preceding formulation of the SPO) that

Norman will in fact not go wrong in accepting the belief [resulting from his clairvoyant powers], and it is, in a sense, not an accident that this is so: it would not be an accident from the standpoint of [a] hypothetical external observer who knows all the relevant facts and the laws. But how is this supposed to justify Norman’s belief? From his subjective perspective it is an accident that the belief is true. And the suggestion here is that the rationality or justifiability of Norman’s belief should be judged from Norman’s own perspective rather than from one which is unavailable to him (Bonjour 1985, 43-44; emphasis in original).


Using the SPO in this way, Bonjour takes himself to have provided a counterexample to externalist theories of justification that attribute justification to Norman.

The central idea behind the SPO appears to be that it is a necessary condition for (doxastic) epistemic justification that the subject holding a belief be aware in some way of what that belief “has going for it,” in a way that makes it the case that it is not just an accident, from the subject’s perspective, that his belief is true. A belief’s truth is an accident from a subject’s perspective, we suggest,13 if it is not the case that there is something of which he is aware in a way that makes it reasonable for him to suspect that the belief is true. Given this understanding of the objection, internalists may press the SPO against externalist views about justification that fail to require such awareness. In addition, they may also use the SPO to directly motivate their own views. Internalists may suggest that if a subject is appropriately aware of the justification-contributor for the relevant belief, as AR requires, then the subject is aware of something that makes the truth of the belief non-accidental, from his perspective. If this is correct, then internalism meets the SPO in a way that versions of externalism do not. This is allegedly some reason to accept internalism over externalism.

Premise IV of Bergmann’s dilemma denies that any version of weak awareness internalism (hereafter, WAI) has this reason in its favor.14 It contends that versions of WAI fail to meet the demands of the SPO. The reason that they allegedly fail is that the awareness of the justification-contributor that versions of WAI require is not strong enough—it is not sufficient to make it the case that the subject who possesses said awareness of the justification-contributor is aware of something that the belief has going for it in a way that makes it the case that the belief’s truth, from the subject’s perspective, is non-accidental. The possession of merely weak awareness supposedly allows that it is just an accident, from the subject’s perspective, that his or her belief is true.

Consider, first, what Bergmann calls “nonconceptual” versions of WAI. Nonconceptual versions of WAI require, for the justification of a subject’s belief, that there be some justification-contributor for the belief of which the subject is weakly aware, and they allow that such awareness may consist in an awareness that does not involve conceiving of the justification-contributor at all.15 According to Bergmann, it is precisely this fact about nonconceptual versions of WAI that makes them vulnerable to the SPO. He contends that,

[S]ince the awareness required is nonconceptual, a person can have the required awareness of [the justification-contributor] without conceiving of [it] in any way—without categorizing it according to any classificatory scheme. But then [a subject] can be nonconceptually aware of [the justification-contributor] without conceiving of [it] as relevant at all to the appropriateness of his belief . . . [I]f [a subject] does not conceive of [the justification-contributor] as something relevant to the appropriateness of his belief, it is an accident from [the subject’s] perspective that his belief is true (Bergmann 2006a, 20).
From the alleged fact that all nonconceptual versions of WAI allow cases of justified belief where it is an accident, from the subject’s perspective, that the belief is true, it follows that they are vulnerable to the SPO.

Similar comments apply to “conceptual versions of WAI. While conceptual versions of WAI require that the awareness of the justification-contributor involve conceiving of that contributor in some way, they still do not require that the awareness involve conceiving of the justification-contributor as in any way relevant to the appropriateness of the pertinent belief.16 Bergmann’s objection to nonconceptual versions of WAI thus applies here as well: “If [a subject] does not conceive of [the justification-contributor] as something relevant to the appropriateness of his belief, it is an accident from [the subject’s] perspective that his belief is true.” And then the SPO straightforwardly applies.

That said, if the reasoning of the preceding two paragraphs succeeds, and both nonconceptual and conceptual versions of weak awareness internalism do allow that, possibly, a subject’s belief is justified while it is an accident from his perspective that the belief is true, then both versions of WAI succumb to the SPO, and premise IV of Bergmann’s dilemma is true.17 No version of WAI could then avoid the dilemma.

We will argue, however, that the reasoning of the aforementioned paragraphs does not succeed. It is not the case that all versions of WAI allow that, possibly, a subject’s belief is justified while it is an accident from his or her perspective that the belief is true. There are versions of WAI that can avoid the SPO and escape Bergmann’s dilemma.

Consider, as an example, a version of nonconceptual WAI that appeals to “seemings” as justification-contributors.18 The idea behind such a view might go as follows. Certain propositions seem true to us, where a given proposition’s seeming true to us consists in our being spontaneously inclined to form a belief in the proposition in response to being immediately acquainted with (i.e., immediately weakly aware of) some object of awareness.19 These spontaneous inclinations are identified by the view as “seemings.” We emphasize that seemings are mere inclinations to form beliefs, by which we mean only that they are themselves nonconceptual psychological states—they are neither instances of concept-application nor of outright belief. Such seemings may be in certain respects conceptual—for example, they may arise as the results of noticing certain conceptual relationships (e.g., “inclusion” relationships) between the constituents of propositions under consideration—but the important point is that a subject hosting such seemings need not additionally conceive of the seemings themselves, or of the first-order objects of awareness that cause the seemings, as justifying any particular belief(s). A subject may “host” a seeming without conceiving of—or even having higher-order awareness of—the seeming itself in any way. Likewise, the seeming may result for the subject as a result of merely weak awareness of some object of first-order awareness.20 This being the case, an individual can be in a state wherein he hosts the seeming that some proposition is true while remaining in a state of weak—and even nonconceptual—awareness of that seeming, or while having no higher-order awareness of the seeming at all, and all while remaining in a state of weak awareness concerning the object of first-order awareness that gives rise to that seeming.

But despite the fact that all of these states have been described as states of (at most) weak awareness, it is readily apparent that they can make a relevant difference from the subject’s perspective—specifically, they can make it the case that, from the subject’s perspective, the relevant belief is appropriate to hold. A subject hosting these mental states concerning a particular belief is reasonable (assuming he possesses no justification defeaters) in expecting that belief to be true,21 and so the belief’s truth, from his perspective, is non-accidental. There is a clear subjective difference, for example, between the belief that inexplicably (from his own perspective) pops into Norman the clairvoyant’s head as a result of the operation of his clairvoyant powers, on the one hand, and a subject’s belief in some proposition as a result of the proposition’s actually seeming true to him upon his understandingly considering it and being weakly aware of some object of awareness (e.g., conceptual inclusion relationships), on the other. After all, in the latter case, the proposition seems true to the subject—to use the language of Plantinga (1993), the subject feels “pulled toward” believing it.22 That said, the subject who forms a belief in such a case does not (assuming a non-deviant causal chain) do so in such a way that it is an accident, from his perspective, that his belief is true. Given that the proposition seems true to him, it is not surprising or accidental, from his perspective, that it is true. If this is right, then (and this is the crucial point) one’s merely being in a state like the one described is sufficient to make it the case that it is not an accident, from one’s own perspective, that one’s belief in the relevant proposition is true. Importantly, no (higher-order) awareness of or conceptualization of the seeming itself is required for the subject’s belief to be relevantly “non-accidental” from his perspective—it is enough that the belief seem true to the subject (by virtue of his hosting the relevant seeming).23 But then the view just described is a version of WAI that avoids the SPO.

One response to this defense of WAI may be to deny that the view defended is really a version of internalism, since, as just presented, it does not even require awareness of all of the relevant justification-contributors (it does not require higher-order awareness of the seeming itself, for example). But it is worth noting, first, that lacking such a requirement is entirely consistent with Bergmann’s characterization of internalism. The view just presented requires awareness of something contributing to the justification of the relevant belief (namely, the object of weak awareness that causes the seeming), and so conforms to Bergmann’s formulation of internalism’s AR—it needn’t require awareness of everything contributing to the belief’s justification.24 And second, we can readily add to the view the requirement that the subject also have some weak (higher-order) awareness of the seeming itself in order for his particular belief to be justified. There can be no qualms about this latter view’s internalist credentials—it is full-blown weak awareness internalism. Still, this view also clearly avoids the SPO. It remains intuitively the case that the seeming still makes a difference (plausibly an even greater difference, given the subject’s higher-order awareness of it) from the subject’s perspective. The believed proposition still seems true to the subject, and so is not accidentally true from his perspective, and this is so even while the subject does not conceptualize the seeming, or the object of first-order weak awareness, as justifying the belief in question. This is enough to show that there are plausible versions of WAI that do not succumb to the SPO. So, premise IV of Bergmann’s dilemma is false.

Though we have just argued against Premise IV on the basis of a view appealing to “seemings” as justifiers, we note that this appeal is not essential to our argument. Versions of internalism that deny that seemings are justifiers, but that appeal to other appropriate mental states, can employ our general strategy of argument for defending themselves against Bergmann's dilemma. The central idea behind this strategy is that there are some internal states which are such that merely being in those states is enough to make it the case that it is no accident, from one's own perspective, that the relevant belief is true; no higher-order conceptualization of such states as justifying the belief in question is required.25 We find it plausible to think that seemings are such states, but we also think it is plausible that there are other kinds of internal states that share this feature—i.e., that are such that one’s merely being in those states is sufficient to put one in a significantly better epistemic situation regarding the truth of the relevant belief than is Norman regarding his belief. Apparently appealing to direct acquaintance states as states with this feature, Richard Fumerton plausibly remarks that “there is surely an enormous difference between the intellectual satisfaction one achieves [by being in such states, as compared with] . . . being in a belief state that [like Norman's] may or may not have the relevant causal pedigree.”26 In our view, both direct acquaintance states and seeming states are kinds of internal states to which the weak awareness internalist can appeal as making such an “enormous difference.” And, since these states do make this difference, versions of WAI that require such states for justification are invulnerable to the SPO. Again, this falsifies premise IV of Bergmann’s dilemma.

We close this section with some brief considerations about “accidentality.” Earlier, we suggested that what it is for it to be “an accident, from the subject’s perspective, that his belief is true” is for it to be the case that the subject lacks the kind of awareness of some justification-contributor that makes it reasonable for him to suspect that his belief is true. Both Bonjour and Fumerton consider the case of Norman as a case of such accidentality. One thing worth noting, though, is that Norman’s case is not a case wherein he is aware of some justification-contributor for his belief that the President is in New York City, albeit in an allegedly insufficient way. Norman simply has no awareness whatsoever of a justification-contributor for his belief.

This being the case, internalists are not committed to claiming that the accidentality to which externalist views are specially subject is an accidentality having to do with an inappropriate sort of awareness of justification-contributors. Rather, they may instead claim that the bad sort of accidentality infecting cases like Norman’s is a result of the subject’s having no awareness whatsoever of anything contributing to his belief’s justification. Such internalists could then employ this notion of accidentality in pressing the SPO against externalists. And even the very weakest version of internalism could avoid this sort of accidentality (and thus this “revised” SPO) simply by requiring some mere awareness of a relevant justification-contributor, and nothing else. Non-accidentality would be achieved, according to such views, when one simply has some awareness of a relevant justification-contributor(s), as one does not in cases like Norman’s. We point this out in order to suggest another way in which versions of WAI might avoid Bergmann’s dilemma. If the preceding considerations are right, then Bergmann’s characterization of what is required to avoid the “accidentality” of the SPO may simply be wrong-headed, and there will then be many more versions of WAI that can meet the SPO.



III. REJECTING PREMISE III: STRONG AWARENESS INTERNALISM AND “VICIOUS REGRESS”
In this section, we move to dispute premise III of Bergmann’s dilemma for internalism in order to show that versions of strong awareness internalism (SAI) can also avoid it. That premise claims:

III. If the awareness required by internalists is strong awareness, then internalism has vicious regress problems leading to radical skepticism.


Why might one think that versions of SAI lead to vicious regress? Recall that SAI requires that the justification-contributors of a belief be conceived of as related to the truth or justification of that belief in order for them to actually justify it. As Bergmann notes, such conceptualization of a justification-contributor can itself be justified or unjustified (Bergmann 2006a, 17).27 Any plausible version of SAI will hold that only justified conceptualizations will be sufficient. So, there must be some justification for the relevant concept-application. That said, suppose that we are considering whether some subject, S, is justified in holding some belief, B, given his awareness of some potential justification-contributor, X1. According to SAI, S’s belief will be justified only if there is a justified occurrence of A1:

A1: S’s application to X1 of the concept being in some way related to the appropriateness of B.


However, S’s conceptualization of X1 in this way, A1, is itself justified only if there is some justification-contributor for A1, X2, of which S is aware, together with a justified occurrence of A2:

A2: S’s application to X2 of the concept being in some way related to the appropriateness of A1.


But like A1, A2 is also in need of being justified, which will require a further (justified) concept-application, A3, and so forth (Bergmann 2006a, 18). It thus appears that there are at least three regresses threatening versions of SAI. First, there is a regress of instances of concept–application; second, there is a regress of justification-contributors needed to justify those concept-applications; third, there is a regress involving the increasing complexity of the concepts that are being applied.28 Bergmann’s challenge to versions of SAI focuses on the third of these regresses—he explicitly says that the regress problem for SAI that he has in mind in his premise III is a “complexity regress,” and that “the issue is the complexity of, not the number of, beliefs.”29 That being so, we will be focusing on falsifying this premise of Bergmann’s dilemma by showing that versions of SAI can avoid the complexity regress, though we will also suggest ways in which such views might also avoid the other two regresses.

In fact, we will first address the regress of justification-contributors, which we will call the “contributor regress.” We do this in order to locate a particular justification-contributor that will do substantial work in our later defense of versions of SAI from the complexity regress.

In seeking to stop a similar regress of justification-contributors, many defenders of foundationalism appeal (plausibly, we think) to experiences as regress-stopping conferrers of justification. This appeal to experiences is made for a number of reasons. First, experiences are not themselves the type of thing that can be epistemically justified or unjustified, so if experiences can justify beliefs, they will be suitable regress-stoppers—they will be able to confer justification upon a belief without themselves needing to be justified. And second, it seems that experiences are otherwise well-suited to play a justificatory role.30 Intuitively, experiences can provide good reasons to believe propositions. When one has an experience as of there being a dark green triangular shape within one’s visual field, for example, one seems to have at least part of a good reason to believe that there is a dark green triangular shape within one’s visual field.

A defender of SAI cannot be satisfied with the mere appeal to experiences as regress-stoppers, however. According to SAI, in order for an experience to justify a belief for an individual, that individual must also conceive of that experience as relating to the truth or justification of that belief. So, regarding the above example, one’s having an experience as of there being a dark green triangular shape in one’s visual field is itself insufficient to justify one’s belief that there is a dark green triangular shape in one’s visual field. One also needs to conceive of that experience as relevant to the truth or justification of that belief.

We can represent this central idea of SAI formally. Suppose that our subject, S, has the aforementioned experience and belief. Let us call the experience ‘E1’ and the belief ‘B1’:

E1: S’s experience as of there being a dark green triangular shape within S’s visual field.


B1: S’s belief that there is a dark green triangular shape within S’s visual field.
In addition to having E1, SAI requires that S also conceive of E1 as being related to the truth or justification of B1 in order for E1 to actually act as a justifier of B1 for S. That is, SAI requires a justified occurrence of C1:

C1: S’s application to E1 of the concept being in some way related to the appropriateness of B1.


For ease of expression, let us call the appropriate relation mentioned in C1 the “justifies” relation, so that C1 is S’s conceiving of E1 as justifying B1.31 As indicated above, it is precisely at the point of requiring a justification for C1 that SAI encounters the contributor regress problem. SAI will maintain that S must be justified in conceiving of his experience, E1, as justifying his belief, B1, in order for that experience to act as a justifier for that belief. But now it seems that the regress is on its way, for E1 by itself is no longer sufficient for stopping the contributor regress, as was originally hoped.

In our view, the defender of SAI has the resources to respond to this concern. Even if S must be justified in conceiving of E1 as justifying B1, it is possible that the justification for this conceptualization can itself have a regress-stopping justification-contributor. As we have noted, any suitable regress-stopper will be such that it confers justification while not itself being subject to being justified or unjustified. And we have already seen that experiences can be such as to confer justification while not being subject to being justified or unjustified. We might thus think of experiences as being ‘ultimate evidence’.32 That said, the defender of SAI might be able to appeal to some further experience had by S that could justify S’s conceiving of his initial experience as justifying the belief in question. What is needed is an experience that justifies S in conceiving of his object-level experience as justifying his belief.

But are there such experiences? Quite plausibly, there are. The “seemings” mentioned earlier in this paper constitute a kind of experience with which many people are familiar. Many people experience such seemings upon considering allegedly “self-evident” propositions, such as the proposition that every golden trumpet is a trumpet. When one considers this proposition and it seems true to one, one hosts a certain difficult-to-describe experience which we earlier called “an inclination to believe.” Such seemings might be—indeed, seem to us actually to be—present exactly where the defender of SAI needs them. For there seem to exist seemings regarding evidential relations. Sometimes it seems to us that something of which we are aware, X, is evidence for (i.e., reason to believe) a particular proposition, p. Such experiences can be described as its seeming to us to be the case that X justifies (is evidence for) p.

Supposing that such experiences are available to the defender of SAI, let us return to our consideration of the example of S, above. Suppose that, in addition to S’s having E1 as data, S also has an experience, E2, in which it seems to S that E1 justifies B1:

E2: Its seeming to S that E1 justifies B1.
Plausibly, E2 can play some role in justifying S in conceiving of E1 as justifying B1. Indeed, such seemings appear to be potential providers of foundational justification.33 E2 can play some role in justifying S in conceiving of E1 as justifying B1, and yet E2 itself is not subject to being justified or unjustified. E2 thus appears to be an available regress-stopper for the contributor regress.

But unfortunately, the work for the defender of SAI with respect to this regress does not end so readily. E2, though it is not itself in need of being justified, must, according to SAI, be justifiably conceived of as justifying C1 in order for it actually to justify C1 for S. What is required, then, is a justified occurrence of C2:

C2: S’s conceiving of E2 as justifying C1.
An experience analogous to E2 could do the justificatory work here, but the worry is that this would only lead to an infinite regress of further requisite experiences, thus continuing the contributor regress. In order to prevent such a regress, one thing that the defender of SAI seems to need is an experience that justifies some belief while at the same time justifying the higher-order conceptualization of itself as justifying this belief, thus stopping the need for any further experience.

We believe that such an experience is available to the defender of SAI. Suppose that S considers the universal proposition that for all p, its seeming to one that p justifies one’s belief that p.34 Suppose, further, that it seems to S (as it does to us) that this universal proposition is true.35 S thus has the following experience, E*:

E*: Its seeming to S that its seeming to one that p justifies one’s belief (or conceptualization) that p.36
Recall that what we are looking for here is a justification-contributor that justifies C2—S’s conceptualization of E2 as justifying C1—for S. The first relatum of C2, E2, is simply its seeming to S that E1 justifies B1. The second relatum, C1, is just the conceptualization by S that E1 justifies B1. It appears, therefore, that E* can justify S in conceiving of this first relatum as justifying the second, i.e., in conceiving of E2 as justifying C1, which is to say that it can justify C2 for S. For notice that the propositional contents of the relata specified in C2 are the same. S’s experience E* can justify C2 because (substituting the contents of C2’s relata, E2 and C1, for ‘p’ in the above schema) E* gives S a reason for taking its seeming to S that E1 justifies B1 (the content of E2) as a reason for believing (or conceptualizing) that E1 justifies B1 (the content of C1). So, E* justifies C2.

But, given the requirements of SAI, in order for E* to stop the contributor regress, it must also be the case that E* justifies the conceptualization of itself as justifying C2. That is, E* must justify C3:

C3: S’s conceiving of E* as justifying C2.
It may not be immediately obvious how E* could do this. But there is a way. Notice that C3, spelled out explicitly, amounts to the following:

C3’: S’s conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p] as justifying [S’s conceiving of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1)].37


In other words, what S needs to justify C3 is a reason for taking the first relatum of C3’, its seeming to S that its seeming to one that p justifies one’s belief that p, to justify his conceiving of [its seeming to S that E1 justifies B1] as justifying his conceptualization of E1 as justifying B1. But this reason is already available to S in the form of E*. For E* justifies S in taking it that seemings justify beliefs when those seemings and beliefs have the same contents (this is what allowed E* to justify C2). But then E* justifies S in taking the experience mentioned in the first relatum of C3’, its seeming to S that its seeming to one that p justifies one’s belief that p, to justify his belief that its seeming to one that p justifies one’s belief that p. And then, transparently, this justified proposition that [its seeming to one that p justifies one’s belief that p] justifies him in conceiving of the seeming that E1 justifies B1 as justifying the belief that E1 justifies B1—the very “conceiving” mentioned in the second relatum of C3’—for that conceptualization is just an instance of the more general schema. In other words, E* justifies S in conceiving of the experience mentioned in the first half of C3 as justifying the conceptualization mentioned in the second half, which is to say that it justifies C3 for S. So E* justifies C3 in just the way the defender of SAI should hope. No additional experience is needed; the contributor regress stops here.38

This way of stopping the contributor regress provides a general strategy for all versions of SAI concerned by it. The strategy involves locating a particular justification-contributor that both non-inferentially justifies some belief and non-inferentially justifies the conceptualization of itself as being related to the truth or justification of that belief. It is this feature of the justification-contributor that allows for the avoidance of the contributor regress. The particular justification-contributor need not be a seeming (or even an experience) of the sort that we have identified; it can be any justification-contributor that, according to the particular version of SAI, has the requisite features.

Having thus defended the preceding version of SAI from the contributor regress, our central concern now is whether the view can avoid the “complexity regress” leveled against versions of SAI in premise III of Bergmann’s dilemma. It initially appears that it cannot. For even if it is the case that E* justifies C2 as well as C3, and so on, the view, as a version of SAI, requires in every case that E* additionally be conceived of as justifying whatever it allegedly justifies. In other words, taking C3 as our example, it must be that there is an additional justified conceptualization, C4:

C4: S’s conceiving of E* as justifying C3.

And even if E* can justify C4, there will then need to be an additional justified conceptualization, C5:

C5: S’s conceiving of E* as justifying C4.

And so on. Since ‘C3’ and ‘C4’ serve as mere placeholders referring to earlier levels of conceptualization, these additional conceptualizations are increasingly complex.39 The worry is that this complexity will increase to such an extent that S cannot possibly have the requisite conceptualizations, given his finite conceptual abilities, and thus will ultimately lack the justification he needs. Can the preceding version of SAI avoid this regress in complexity, and if so, how?

To answer this question, we should first consider the explicit contents of C4 and C5.40

C4’: S’s conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p] as justifying p justifies one’s belief that p] as justifying [S’s conceiving of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1)]>.
C5’: S’s conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p] as justifying {S’s conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p] as justifying p justifies one’s belief that p] as justifying [S’s conceiving of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1)]>}.
A careful examination of C3’, C4’, and C5’ reveals that each one simply includes an additional iteration of the clause, “S’s conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p] as justifying,” before the clause, “S’s conceiving of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1).” (The reader should note that these two clauses together constitute the whole of C3’.) This is the crucial point to notice, for these additional clauses are ultimately “superfluous,” in the way that the following examination of the way in which E* justifies C4’ indicates.

E* justifies C4’ in the following manner. E* justifies S in taking it that seemings justify beliefs (again, this is what allowed it to justify C2). But then E* justifies S in taking the experience mentioned in the very first clause of C4’, its seeming to S that its seeming to one that p justifies one’s belief that p, to justify his belief that its seeming to one that p justifies one’s belief that p. The question is then whether this can justify S in “conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p] as justifying [S’s conceiving of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1)].” But the clause, “S’s conceiving of [its seeming to S that its seeming to one that p justifies one’s belief that p],” simply mentions experience E* all over again. By our prior reasoning, we already know that, given S’s justified belief that its seeming to one that p justifies one’s belief that p, this experience justifies S’s belief that its seeming to one that p justifies one’s belief that p. And then all that we are left with in C4’ is the question of whether or not this latter justified belief justifies S in “conceiving of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1).” But that is the very same question asked in our examination of C3’ above, and we already know the answer (‘yes’). In other words, C4’, like every conceptualization beyond C3’, simply redundantly mentions E* over and over again until returning to the issue of taking E* to justify the final conceptualization mentioned in C3’ (i.e., the conceptualization by S of (its seeming to S that E1 justifies B1) as justifying (S’s conceptualization of E1 as justifying B1)).

This fact allows the defender of SAI to say at least two things about the complexity regress. First, the apparent complexity of conceptualizations is a trivial sort of complexity. Each conceptualization beyond C3 simply redundantly reiterates E* over and over again and then claims that E* justifies S’s conceptualization of [its seeming to S that E1 justifies B1] as justifying [S’s conceptualization of E1 as justifying B1]. And reasoning our way to the conclusion that this latter claim is correct is not difficult (as shown above). So, in effect, S can readily see, via that reasoning, that E* will trivially justify all conceptualizations above C3, even though such conceptualizations are ever increasing in complexity.41

This much shows that the “complexity” of such additional conceptualizations does not in principle prevent S from being justified in making them, or from being justified in taking his experience to justify them. But there remains a more pressing worry behind the complexity regress: although there is no problem with complexity that prevents S, in principle, from being justified in making these further conceptualizations, it still might be the case that the increasing complexity of these conceptualizations is such that, ultimately, finite minds (such as S’s) are unable to grasp them. That is, it might be that, even though the added complexity is redundant or superfluous (in the sense we have articulated), it nonetheless results in conceptualizations that quickly become too large and complicated for S’s finite mind to grasp. And, in that case (the worry goes), there would be some level of conceptualization at which S simply could no longer conceive of E* as relevant to that conceptualization’s justification, simply because he could not grasp the conceptualization’s content. This failure of S to conceive of E* as relevant to the justification of that conceptualization would have SAI render the verdict that this conceptualization is unjustified for S. And this lack of justification would affect S’s conceptualizations at earlier levels, so that SAI would ultimately judge that S is unjustified in his initial, object-level belief. The specter of skepticism appears to loom.

But here is where the very same reasoning that could in principle justify S in conceiving of E* as relevant to his additional conceptualizations can once more play a role, thereby allowing him to avoid the complexity regress entirely. That reasoning will justify S in the general belief that E* justifies all conceptualizations beyond C3. From this general belief, S can then readily infer any number of de re beliefs about conceptualizations at levels “higher than” C3, even if S cannot explicitly grasp the contents of these conceptualizations due to their complexity. This will allow S to form justified beliefs about each higher level of conceptualization—including the belief (required by SAI) that E* justifies that conceptualization—without his having to form increasingly complex de dicto beliefs explicitly including those conceptualizations’ contents. A version of SAI could thus maintain the requirement that S form additional beliefs regarding increasingly higher-level conceptualizations without falling prey to Bergmann's charge that this requires S to form unintelligibly complex beliefs.  Instead of requiring that S justifiably believe de dicto that E* justifies C24, for example (where, again, ‘C24’ is a mere placeholder for C24’s content), a version of SAI could require simply that S justifiably believe, of C24, that E* justifies it.42  This latter belief can apparently be justified for S without his having to explicitly entertain C24's content, since he can use the reasoning mentioned above to note that whatever additional content C24 has will be trivial (in our aforementioned sense) and will be justified by E* (as will any higher-level conceptualization beyond C3).43 So, the increasing complexity is harmless on all fronts. First, the increasingly complex propositions (e.g., that E* justifies C24) are still justified for S by E*, and second, it needn’t be that S form de dicto beliefs in these propositions and thereby grasp explicitly their contents. That being the case, versions of SAI can avoid all elements of the complexity regress leading to radical skepticism; premise III of Bergmann’s argument is false.44

Having shown this much, we acknowledge that there remains another regress confronting versions of SAI that many readers will likely think is bad enough (perhaps bad enough to make it the case that SAI still leads to radical skepticism, even if not in the way that Bergmann claims). There is still a regress of instances of belief or conceptualization. That is, it appears that versions of SAI end up requiring that a subject actually hold an infinite number of beliefs or conceptualizations in order to be justified. That might seem to be an implausible and, more importantly, unsatisfiable requirement. If this requirement is unsatisfiable, then one might worry that the SPO has not in fact been avoided.

There are at least three plausible responses to this problem. A first option for versions of SAI is to simply embrace the requirement that the subject actually hold an infinite number of beliefs or conceptualizations. The best hope for plausibility for such versions of SAI is an appeal to actual tacit beliefs.45 There is some plausibility to the claim that we already have an infinite number of such beliefs;46 if that is right, then the idea that we might (have to) come, via the preceding reasoning, to hold many more such beliefs is not obviously problematic. But whether this appeal is ultimately satisfying is something that we will not go on to consider in any more detail here. This is because we think that there are two alternative options for versions of SAI that more straightforwardly avoid the “infinite beliefs” regress.

As we have already seen, E* is capable of justifying all of the (infinitely many) requisite beliefs or conceptualizations. Experiencing E* and going through the reasoning described above justifies S in the general belief that E* justifies all of the conceptualizations beyond C3. As we noted, S can then readily infer any number of (justified) de re beliefs about conceptualizations at levels “higher than” C3 from this general belief (namely, that they are each justified by E*). That being so, one way for versions of SAI to avoid requiring an infinite number of actual beliefs or conceptualizations (even if they are de re) is to require only that the subject actually have two justified de re beliefs or conceptualizations regarding all of the conceptualizations at levels “higher than” C3.47 One of these would be a justified de re belief, of the conjunction composed of the infinitely many propositions expressing the evidential relationships between E* and the increasingly-complex conceptualizations, that it is true—i.e., a de re belief concerning the conjunction E* justifies C3 and E* justifies C4 and E* justifies C5 . . . and E* justifies Cn and E* justifies Cn+1 (etc.). The other would be a justified de re belief that the truth of this conjunction is relevant to the truth or justification of the pertinent object-level belief B1 (the belief whose justification is under scrutiny). Let’s consider each of these in turn.

We have already seen how one can be justified in forming the relevant de re beliefs concerning each of the individual conjuncts (i.e., concerning the proposition that E* justifies C3, and so on for each conceptualization beyond C3). Going through the reasoning described above justifies the subject in believing each of these conjuncts. Once one realizes this, it seems that one can justifiably believe (at least de re) the proposition that is the conjunction of these infinitely many propositions with their ever increasing iterations of complexity. Since the justification of each conjunct is not independent, and since all but the first conjunct simply adds “trivial” content (in the sense described above) to the previous conjunct—content which the subject who performs the aforementioned reasoning recognizes as trivial—there is no worry here that the ever-growing number of conjuncts will make it such that the subject will at some point lack justification for believing the conjunction. So, going through the reasoning described above can justify the subject in believing of this infinitely long conjunction that it is true.

Further, understanding what the content of each conjunct is like (even though S cannot explicitly grasp (de dicto) the content of the more complex propositions) justifies S in the de re belief that this conjunctive proposition is relevant to the truth or justification of the pertinent belief (the belief whose justification is under consideration). That is, S recognizes that each conjunct claims that the experience E* justifies the subject in conceiving of E* as justifying one of the infinite number of requisite conceptualizations (from C3 onward), and that each conjunct is ultimately (in the way shown by the reasoning about “trivial” complexity) concerned with the justification of the subject’s conceptualization of E1 as relevant to the truth or justification of B1 (the object-level belief). This being so, each conjunct is transparently relevant to the truth or justification of this belief, and the subject, having gone through the above reasoning, will be able to recognize this. Having done so, the subject will likewise be able to recognize that the conjunction composed of these conjuncts is, ipso facto, itself relevant to the truth or justification of the object-level belief. So, both the belief, of the conjunction, that it is relevant to the truth or justification of the object-level belief, and the belief, of the conjunction, that it is true, will be held by the subject with justification.

Any view requiring such justified beliefs or conceptualizations seems to readily avoid the SPO. This is because once one justifiably believes, via the above reasoning, of this infinitely long conjunction, that it is true and that it is relevant to the truth or justification of the pertinent belief, it would not seem like an accident from the subject’s perspective that the relevant object-level belief, B1, is true. Bergmann’s worry with such views is that they allegedly succumb to a vicious regress, but, as we have seen, this is not so for the view just described. So, we appear to have a version of SAI that avoids the SPO and does not succumb to any of the regresses.

A final, alternative version of SAI that perhaps even more straightforwardly avoids the “infinite beliefs” regress appeals to potential beliefs. Bergmann himself notes that there may be versions of what he calls “potential strong awareness internalism” that require for object-level justification only that the subject be able to ascend to higher levels of justified conceptualization upon reflection. Of this third class of views, he says that they manage “to avoid requiring for justification the actual possession of an infinite number of . . . beliefs.”48 Such versions of SAI could thus avoid the “infinite instances” regress while also employing the strategies presented earlier in this section to avoid both the complexity and contributor regresses. Importantly, those earlier strategies not only do avoid these other regresses, but they also present the very reasoning by which any subject could ascend to higher levels of justified conceptualization upon reflection, precisely as a version of potential SAI would require. Such reasoning shows that the conceptual ascent does not require an infinite number of experiences or impossibly powerful conceptual abilities—that is, it shows that (typical) subjects are in principle able to meet the requirements of versions of potential SAI.49 That said, versions of potential SAI can avoid all three regresses.

One might worry at this point that any version of SAI that only requires potential belief or conceptualization of the justification-contributor as relevant to the truth or justification of the pertinent belief is subject to the SPO, since such views will not require that the subject actually conceive of the justification-contributor as relevant to the belief in question. It seems to us, however, that such a view avoids the SPO for the same reasons that the earlier discussed versions of WAI avoid it. On the view currently under discussion, the pertinent belief, B1, still seems true to the subject, and, as we have seen, this alone is sufficient for making the truth or justification of B1 non-accidental from the subject’s perspective. But the view under discussion, as a version of SAI, adds even more to the subject’s perspective, putting him in an even better epistemic position regarding the pertinent belief. Not only does the relevant proposition seem true to the subject, but it also seems to the subject (by virtue of his having experienced E*) that its seeming so is good reason to think that the proposition is true. This second-order seeming makes it definitively not an accident from the subject’s perspective that his belief is true or justified. So, such a version of SAI clearly avoids the SPO while also avoiding the “infinite beliefs” regress.

Having shown this much, we note again that the strategy employed in our defense of SAI is, like the strategy employed in our defense of WAI, perfectly general. That is, it need not appeal to seemings as justifiers in avoiding any of the preceding regresses. A defender of SAI could build on Fumerton’s work, for example, and utilize an epistemology which conceives of direct acquaintance as a provider of justification. Thus, although Fumerton himself embraces a version of WAI, it is possible to use elements of his view to defend SAI as well. One would simply need to substitute direct acquaintance states for seeming states in the preceding defense, in a way analogous to the substitution made in defense of versions of WAI.50 Many specific versions of internalism could thus be defended using our defensive strategy. Any version of SAI that employs this strategy appears to falsify premise III of Bergmann’s dilemma.51



Yüklə 131,13 Kb.

Dostları ilə paylaş:
  1   2   3




Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©genderi.org 2023
rəhbərliyinə müraciət

    Ana səhifə