Abstract
It’s not implausible to think that whenever I have a justified noninferential belief that p, it is caused by a seeming that p. It’s also tempting to think that something contributes to the justification of my belief only if I hold my belief because of that thing. Thus, given that many of our noninferential beliefs are justified and that we hold them because of seemings, one might be inclined to hold a view like Phenomenal Conservatism, according to which seemings play a crucial role—perhaps the only crucial role—in the justification of our noninferential beliefs. But Phenomenal Conservatism seems to conflict, in a number of ways, with externalist accounts of justification. As a result, the attractiveness of the intuitions appealed to in support of views like Phenomenal Conservatism present something of a challenge to externalism. The purpose of this paper is to deal with that challenge by developing and defending an externalist-friendly account of the role of seemings in the formation and justification of our noninferential beliefs—an account that incorporates what is attractive in views like Phenomenal Conservatism. Because this externalist-friendly account is compatible with both externalist accounts of justification and the plausible elements of views like Phenomenal Conservatism, the challenge to externalism inspired by such views is thereby undermined.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Huemer (2006, p. 148, 2007, p. 30) defines Phenomenal Conservatism as follows:
-
PC
If it seems to S that p, then, in the absence of defeaters, S thereby has at least some degree of justification for believing that p.
Huemer also makes it clear (2001, pp. 99–100) that he thinks nothing other than seemings gives our beliefs prima facie noninferential justification. Tucker (2010) holds a very similar view.
Swinburne (1998, p. 20) defines the Principle of Credulity as the view that “other things being equal, it is probable and so rational to believe that things are as they seem to be”. He goes on to say (1998, p. 21) that “all and only” the beliefs formed in accord with seemings are “properly basic” (i.e., noninferentially justified). In later work (2001, pp. 141–49) he says that the only way our noninferential beliefs in contingent propositions can get prima facie justification—and the way in which they in fact do get such justification—is as a result of their contents seeming to us to be true.
Conee (2004, p. 15) says that according to Seeming Evidentialism (SE) “positive evidence is supplied by seeming truth”. He goes on (2004, p. 16) to suggest that, according to SE, only beliefs in propositions that seem true are rational or justified: “Concerning any thought, our epistemically rational basis for our initial thinking about its truth is our initial evidence on the matter. On the present version of evidentialism, SE, this consists in how things initially seem to us.” I should note that although Conee says things in defense of Seeming Evidentialism, it’s not clear that he endorses the view.
-
PC
Internalist accounts of a belief’s justification require that the person holding the belief is (actually or potentially) aware of at least some of the facts on which justification depends or supervenes; externalist accounts impose no such requirements.
It also presents a problem for internalist accounts of justification that deny that being based on a seeming that p is both necessary and, absent defeaters, sufficient for some degree of justification for a noninferential belief that p.
Those who don’t find attractive any of the elements of views like Phenomenal Conservatism won’t think that any serious challenge to externalism arises from such views. As a result, they won’t find EFARS to be helpful or needed. My target audience consists of those who think there is at least something attractive about views like Phenomenal Conservatism.
As already mentioned in the previous note, some philosophers don’t find Phenomenal Conservatism an attractive position. Although it is not my aim to defend Phenomenal Conservatism, I’ll mention one concern here and suggest a possible response. The concern is that we have justified noninferential non-occurrent beliefs and they don’t seem to be accompanied by seemings in support of them. One possible response for the Phenomenal Conservative (which would need to be developed further) is to say that a person’s noninferential belief that p is prima facie justified only if either (a) the belief is occurrent and it seems to that person that p or (b) the belief is non-occurrent but it was occurrent and when it was occurrent it seemed (at that time) to that person that p. For the purposes of this paper, I’ll be focusing on occurrent belief in the text.
See Huemer (2007, pp. 30–31).
In Bergmann (2006, pp. 114–116) I introduced the notion of a connector—a felt inclination to take one’s sensory experience as indicative of the truth of some proposition. The idea was that, in response to having a sensory experience, a person often has a felt inclination to take that experience as indicative of some proposition p and, as a result, the person believes p on the basis of that sensory experience and that connector. What I’m proposing above in the text is very similar to what I said in that earlier discussion in that it distinguishes the sensory experience from a seeming that inclines us to believe a certain proposition.
Thus, I think that the view that sensory experience has propositional content is a mistake that stems from the temptation to conflate the experience and the seeming. It’s the seeming that has the propositional content, not the sensory experience. But the sensory experience naturally produces that seeming and is constantly conjoined with it, so it is tempting (though mistaken) to assign the seeming’s content to the experience. My example, in the text, of the alien cognizer with the miraculously induced tactile sensations is intended to show that this temptation is a mistake: in that example, the alien has the sensory experience without the seeming and without the content of that seeming. I don’t have the space to defend this view further here, but see Bergmann (2006, pp. 121–132) and Alston (2005) for some relevant discussion.
As I understand the apprehension experience, it can be either illusory or veridical. It’s possible to have an experience that is phenomenally like the experience of having a veridical direct apprehension of a relation R holding among the conceptual constituents of p, and to have this experience even though relation R doesn’t in fact hold among the conceptual constituents of p. That’s why some possible apprehension experiences of such relations holding are illusory and some are veridical.
It’s possible that a desire that p can cause a seeming that p, which in turn causes a belief that p. In this case, the desire that p causes a state that does feel like a state whose content reveals how things really are. But the desire that p doesn’t itself feel like a state whose content reveals how things really are.
If we distinguish inclinations to believe to which we succumb from those to which we don’t succumb, and call the former ‘effective inclinations to believe,’ then (given that seemings cause beliefs via inclinations to believe) we can say that when a seeming that p causes a belief that p, it does so via an effective inclination to believe that p.
Huemer (2001, pp. 100–101, 2005, p. 100) discusses a case where, after one has a seeming that p and believes p on that basis, one later comes to have a contrary stronger seeming that ~p, which leads the person to believe ~p. This can occur even if one continues to have a weaker seeming that p. Tolhurst (1998, pp. 294–295) distinguishes prima facie seemings from all-things-considered seemings and notes we can have conflicting prima facie seemings.
For more on the nature of seemings, see Sect. 1 of Bergmann (forthcoming).
I’m focusing throughout on doxastic justification, not propositional justification. Doxastic justification is justification a belief has if it is formed the right way. Propositional justification is justification that a proposition p has relative to a person if that person has evidence such that, if she believed p on the basis of that evidence, her belief would be doxastically justified. A proposition p can be propositionally justified for a person even if she doesn’t believe it or if she believes it but not on the basis of the evidence that makes it propositionally justified (in which case it lacks doxastic justification).
What are epistemic appropriateness and epistemic fittingness? They are fittingness and appropriateness of an epistemic sort (I’ll be using ‘fittingness’ and ‘appropriateness’ more or less interchangeably). A mental act or event or state or response is epistemically appropriate if it is in accord with the standards or demands of epistemic rationality. Consider the parallel case with moral fittingness and appropriateness. If someone were to say to you that an action is morally appropriate or morally fitting you would understand her to be saying that the action is appropriate or fitting in a moral way—that it has some positive moral status in the circumstance. And if a person were to say that that an action is morally inappropriate, you would take that to count against its being virtuous or obligatory or morally good or right. There may be nuanced differences between being virtuous, obligatory, good, and right that involve more than just being morally appropriate. However, it would be strange to say that an action is entirely morally appropriate if you thought it lacked any of those moral properties. Similar remarks apply in the epistemic case. To say that a belief is epistemically appropriate or epistemically fitting is to say that it is appropriate or fitting in an epistemic way, that is has some positive epistemic status in the circumstance. And to say that a belief is epistemically inappropriate counts against its being rational or justified or warranted or an instance of knowledge. There may be nuanced differences between being rational, justified, warranted, or an instance of knowledge that involve more than just being epistemically appropriate in some way. However, a mental act or response or state wouldn’t be entirely epistemically appropriate if it lacked any of those epistemic properties.
Thus, Objectivity is opposed to views that (a) acknowledge that perceptual beliefs are (indirect) responses to sensory experience via seemings (i.e., the sensory experience causes the seeming which causes the belief) and (b) insist that such belief responses to sensory experience are epistemically appropriate so long as the experience causes the belief via a seeming with a content matching the belief’s content (and there is no contrary seeming).
For a more lengthy explanation of Objectivity, see Bergmann (2006, pp. 114–118).
Two points of clarification: First, to deny Contingency is to endorse Necessity, the claim that if B is a fitting response to E by itself, then, even if B could be an unfitting response to evidence that includes E and more besides, it couldn’t be an unfitting response to E by itself. Second, for almost any experience E, a cognizer can reasonably learn, without relying on E as evidence, that E is an indication of B’s truth. That’s a learned belief response to experience E. My focus in this paper is on unlearned automatic belief responses to experience. These are cases where the experience is (at some point in the believer’s cognitive development) treated as evidence indicative of the belief’s truth without the believer first learning by some other means that the experience is correlated with the truth of that belief. Some unlearned belief responses to experience are epistemically appropriate and some are not. See Bergmann (2006, pp. 116–118) for further discussion of learned and unlearned belief responses to experience.
Here I’ve specified sameness of belief type in terms of belief content. Given that many of our beliefs are indexical, it will often be more helpful to specify sameness of belief type in terms of belief character (so that beliefs with the same character but different content count as beliefs of the same type). See Kaplan (1989) for a discussion of the distinction between character and content.
Recall that if a human noninferential belief that p is based on sensory experience E, it is typically derivatively based on E. In most cases (in particular, those in which the belief that p is justified) it is directly based on a seeming that p, which in turn is based on E. In thinking about such cases, it may be helpful to keep in mind another version of Contingency, Contingency*, which replaces talk of a belief formed in response to E with talk of a seeming formed in response to E—noting that a seeming that p could be a fitting response to E for humans and an unfitting response to E for some alien cognizers. Contingency’s truth, as applied to humans, could then be explained using Contingency*.
See Bergmann (2006, pp. 118–130) for a defense of Contingency (there that defense is described as an objection to Necessity, the denial of what I’m here calling Contingency).
Reid says “no man can give a reason why the sensations of smell, or taste, or sound, might not have indicated hardness, as well as that sensation, which, by our constitution does indicate it.” (Reid 1764/1997, p. 57).
These two principles differ in a number of ways from the version of PC that Huemer explicitly endorses (see note 1 above). However, it’s not clear to me that Huemer would reject these principles. The main differences are (a) these principles are restricted to noninferential beliefs whereas PC is not, (b) these principles are explicitly presented as necessity claims whereas PC is not, (c) the focus here is on doxastic justification, not propositional justification, and (d) being based on a seeming is portrayed in these principles not only as sufficient but also necessary for noninferential justification (absent defeaters). Regarding (a), given that these principles are entailed by the corresponding principles that don’t include the word ‘noninferential,’ there can be no sensible objection, by a Phenomenal Conservative, to including that restriction. Regarding (b), Huemer says (2001, p. 103) that PC, as he understands it, is a necessary truth. Regarding (c), if you think about what it would be natural for Phenomenal Conservatives to say about doxastic justification, the two principles in the text seem to be in keeping with the spirit of Huemer’s version of PC that applies to propositional justification (see note 15 in this paper for the distinction between doxastic and propositional justification). Regarding (d), Huemer does insist that his version of PC is the only principle of noninferential justification, and concludes from this that the only thing that can override the evidence provided by a seeming is contradicting or undermining evidence provided by another seeming (see Huemer 2001, pp. 99–100).
Much of what I say in Bergmann (2006, pp. 116–118) about learned and unlearned belief responses to experiences applies as well to learned and unlearned seeming responses to experiences. In particular, some automatic unlearned seeming responses to experience are epistemically appropriate and some are not.
It’s true that it would be inappropriate for me to stipulate that Jill has no defeaters in this situation if the very fact that Jill’s seeming is an unfitting response to her olfactory sensation counts as a defeater for the justification of the belief held in response to that seeming. But that fact (i.e., that Jill’s seeming is an unfitting response) isn’t something that removes justification Jill’s belief previously had (so it’s not a defeater). Rather, that fact is something that prevents her belief from having any justification to begin with.
Thus, not just any seeming that p will confer justification on a noninferential perceptual belief that p based on that seeming. The seeming itself has to satisfy certain conditions (i.e., being appropriately caused by an experiential ground to which the seeming is a fitting response). This differentiation, between seemings that can confer justification and those that cannot, might lead one to wonder whether seemings are like beliefs in being capable of being either epistemically justified or unjustified (or whether they are, instead, like sensory experiences in not being capable of being either justified or unjustified). Those who defend Phenomenal Conservatism (e.g., Huemer 2001, p. 112) tend to say seemings are not capable of being either justified or unjustified. (Note that if Phenomenal Conservatives thought seemings could be either justified or unjustified, one would assume that they would think that only justified seemings could confer justification, in which case they should tell us what is required in order for a seeming to be justified, something they don’t do.) I see no need for EFARS to take a stand on the question of whether seemings can be either justified or unjustified. (It’s true that EFARS insists that only some seemings can confer justification; but they needn’t insist that this is because some seemings are justified and others aren’t.) So I’ll just note that EFARS is compatible with the Phenomenal Conservative view that seemings can’t be either justified or unjustified; likewise, it’s compatible with a view like Sosa’s (2007, Chap. 3), which says that seemings can be justified or unjustified.
What makes the causing appropriate? EFARS doesn’t need to take a stand on an answer to this question because EFARS is intended to be compatible with different possible answers to it. By way of comparison, note that one might take a partial stand on what the basing relation is by saying that “X is based on Y only if X is caused in the right way by Y” (without saying exactly what’s involved in being caused in the right way). In the same way, EFARS insists that both the seeming and the belief are appropriately caused (in noninferentially justified perceptual belief) but it needn’t take a stand on exactly what’s involved in being appropriately caused. The main reason for not taking a stand on this is that, as was suggested at the beginning and end of Sect. 1.2, EFARS is intended to be compatible with different versions of externalism (e.g., reliabilism, virtue theories, and proper functionalism), each of which might explain appropriate causation in different ways.
Does EFARS endorse the necessity claim in PCR—the claim that seemings are necessary for the justification of human beliefs? On this point, EFARS is silent and EFARS-ians can go either way in filling out their own views.
Here’s one reason you might be inclined to deny the necessity claim in PCR, even if you were an EFARS-ian. Suppose that, because of Contingency, you thought that it is possible for the beliefs of some alien species to be justified without their being based on seemings. And suppose you also thought that, unlike these aliens, humans are, in fact, (epistemically) supposed to base their beliefs on seemings with matching contents (i.e., the belief that p is supposed to be based on a seeming that p). Then the question arises: is it essential to being human that one’s beliefs are (epistemically) supposed to be based on seemings with matching contents? And, if so, is it essential to a particular human that she is human? You might think one could sensibly refrain from answering both questions in the affirmative. For similar reasons, you might also think that although each particular human is, in fact, (epistemically) supposed to base her beliefs on seemings with matching contents, it’s possible for things to change so that some of those who are in fact human are no longer human or so that it is no longer true that humans are (epistemically) supposed to base their beliefs on seemings. In that case, you could agree that it’s contingently the case that every justified human noninferential belief that p is caused by a seeming that p even while denying that it is (metaphysically) necessary for the justification of our noninferential beliefs that they are based on seemings. In virtue of endorsing that claim about “every” justified human noninferential belief, your view would be very much like and very sympathetic to the most plausible key claims of phenomenal conservatism. And in virtue of denying the claim about what is “necessary” for the justification of our noninferential beliefs, you would be denying the necessity claim in PCR.
Some externalists might not be moved by the charge that they’re committed to approving of this Absurd Speech. But other externalists will feel the force of it. In what follows, I offer a response that can be used by the latter group of externalists.
An additional concern about this objection inspired by Huemer is that it applies to his own view on the justification of memory belief. Huemer argues (1999, Sect. 4) that there can be two memory beliefs (one that p, the other that q), each of which is supported by a strong corresponding memory seeming and has no defeater, but only one of which is justified. He points out that if the memory seeming that q is due to the retention of a belief originally acquired through wish-fulfillment and the memory seeming that p was just at that moment produced in the subject by a deceiving demon, then the belief that p is justified but the belief that q is not. Now suppose that a single person, Susan, had both of these seemings, acquired in these ways and that she believed p but not q. Then Huemer’s 2006 “Absurd Speech” argument could be used to argue that, if Huemer’s 1999 view on the justification of memory belief is right, then Susan could sensibly say the following:
I have a strong memory seeming that p and a strong memory seeming that q. These two memory seemings appear to me to be equally reliable and, in general, seem alike in all epistemically relevant respects. However, I believe that p and I don’t believe that q. I have no reason to think that the memory seeming that p is any more reliable than the memory seeming q. I just believe p and not q, for no apparent reason.
Thus, Huemer’s “Absurd Speech” objection applies against his own view. This is especially strange because Huemer says (2006, Sect. 2) that this “Absurd Speech” objection captures the central internalist intuition. He then goes on to argue (2006, Sects. 3 and 4) that the main forms of internalism other than his own Phenomenal Conservatism (this includes direct acquaintance theories, internal state internalism, actual awareness internalism, potential awareness internalism, and introspectability internalism) conflict with the central internalist intuition. If I’m right here, Huemer’s own view also conflicts with this supposedly central internalist intuition. But then all of the positions standardly viewed as internalist positions conflict with what Huemer thinks is the central internalist intuition. Surely this shows that something is wrong with Huemer’s endorsement of the “Absurd Speech” argument and his characterization of it as an objection that captures the central internalist intuition.
That is, she could deny PCR not because she thinks the seeming that p isn’t necessary for the justification of a human noninferential belief that p (absent defeaters), but only because she thinks it isn’t sufficient. See note 32 for discussion of why this concession is detachable from EFARS.
A related objection is discussed in Bergmann (2006, pp. 163–168), though there the discussion isn’t limited in focus to human belief.
That is, its current justification depends on how it is currently formed or held or sustained, not on how it was originally acquired.
Given that I endorse J and J*, I need to replace the following account of justification given in Bergmann (2006, p. 133):
JPF: S’s belief B is justified iff (i) S does not take B to be defeated and (ii) the cognitive faculties producing B are (a) functioning properly, (b) truth-aimed and (c) reliable in the environments for which they were “designed”.
In its place, I offer this slightly altered and (I hope) improved account:
JPF*: S’s belief B is justified iff (i) S does not take B to be defeated and (ii) the cognitive faculties producing B are (a) functioning properly in response to all of S’s mental states, (b) truth-aimed, and (c) reliable in the environments for which they were “designed”.
Notice that although JPF* is consistent with both J and J*, it conflicts with mentalism, the view that if two possible subjects are exactly alike mentally and in terms of which of their mental states their beliefs are based on, then they are exactly alike justificationally (see Bergmann 2006, Chap. 3 for a discussion of mentalism and a critical assessment of the views of its proponents). JPF* conflicts with mentalism because what counts as a fitting (or properly functioning) response to a set of mental states is a contingent matter. Two possible subjects, S1 and S2, might have the same mental states even though what counts as a properly functioning response to mental state M for one differs from what counts as a properly functioning response to M for the other (perhaps because they aren’t members of the same species). JPF* implies that in such a case, S1 and S2 might differ justificationally despite being in the same mental states, contrary to what mentalism says.
In Bergmann (2006, pp. 63–64), I gave the following two examples of such beliefs. First, it seems that there could be rational beings to whom God reveals things by directly and reliably causing true beliefs to be formed in them, without using any mental states of the believers as intermediate causes of those beliefs. Such beliefs would be epistemically appropriate for them and yet not based on other mental states. Second, it seems that there could be alien cognizers who naturally form the belief that there is water nearby via a process that doesn’t involve any other mental states. It may be that water in the environment of these aliens causes in them the belief that there is water nearby and that it does so without using any other mental states as intermediate causes of those beliefs. Moreover, it may be that these beliefs are not only reliably formed but also formed in a way that is natural and healthy for members of the species of which these aliens are members. In this sort of case too, it seems possible for such beliefs to be epistemically appropriate even though they are not caused by or based on any previous mental states of the believer.
References
Alston, W. (2005). Perception and representation. Philosophy and Phenomenological Research, 70, 253–289.
Bergmann, M. (2006). Justification without awareness. New York: Oxford University Press.
Bergmann, M. (forthcoming). The dilemma for internalism and phenomenal conservatism. In C. Tucker (Ed.), Seemings and justification: New essays on dogmatism and phenomenal conservatism. New York: Oxford University Press.
Conee, E. (1998). Seeing the truth. Philosophy and Phenomenological Research, 58, 847–857.
Conee, E. (2004) First things first. In E. Conee & R Feldman (Eds.), Evidentialism: Essays in epistemology (pp. 11–36). New York: Oxford University Press.
Feldman, R. (1988). Subjective and objective justification in ethics and epistemology. Monist, 71, 405–419.
Greco, J. (2000). Putting skeptics in their place. New York: Cambridge University Press.
Huemer, M. (1999). The problem of memory knowledge. The Pacific Philosophical Quarterly, 80, 346–357.
Huemer, M. (2001). Skepticism and the veil of perception. Lanham, MD: Rowman and Littlefield Publishers.
Huemer, M. (2005). Ethical intuitionism. New York: Palgrave Macmillan.
Huemer, M. (2006). Phenomenal conservatism and the internalist intuition. American Philosophical Quarterly, 43, 147–158.
Huemer, M. (2007). Compassionate phenomenal conservatism. Philosophy and Phenomenological Research, 74, 30–55.
Kaplan, D. (1989). Demonstratives. In J. Almog, et al. (Eds.), Themes from Kaplan (pp. 481–564). New York: Oxford University Press.
Markie, P. (2004). Nondoxastic perceptual experience. Philosophy and Phenomenological Research, 68, 530–553.
Markie, P. (2005). The mystery of direct perceptual justification. Philosophical Studies, 126, 347–373.
Markie, P. (2006). Epistemically appropriate perceptual belief. Noûs, 40, 118–142.
Plantinga, A. (1993). Warrant: The current debate. New York: Oxford University Press.
Reid, T. (1764/1997). In: D. Brookes (Ed.), An inquiry into the human mind on the principles of common sense. Edinburgh: Edinburgh University Press.
Sosa, E. (2007). A virtue epistemology. Oxford: Oxford University Press.
Swinburne, R. (1998). Providence and the problem of evil. Oxford: Oxford University Press.
Swinburne, R. (2001). Epistemic justification. Oxford: Oxford University Press.
Tolhurst, W. (1998). Seemings. American Philosophical Quarterly, 35, 293–302.
Tucker, C. (2010). Why open-minded people should endorse dogmatism. Philosophical Perspectives, 24, 529–545.
Acknowledgments
Thanks to Stewart Cohen, Michael Huemer, Jonathan Kvanvig, Trenton Merricks, Michael Rea, Chris Tucker, and some anonymous referees for comments on earlier drafts. Thanks also to helpful comments from audience members when I presented this paper to the philosophy department at the University of Iowa, to the philosophy department at the University of Notre Dame, and at the Midwest Epistemology Workshop at the University of Nebraska—Lincoln.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bergmann, M. Externalist justification and the role of seemings. Philos Stud 166, 163–184 (2013). https://doi.org/10.1007/s11098-012-0037-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-012-0037-y