1. INTRODUCTION
Skepticism has been offered a new resource. A belief is debased when it is given a basis that prevents it from being knowledge, such as wishful thinking or superstition. A debasing demon is an agent with a maximal ability to deny knowledge by debasing beliefs. Jonathan Schaffer argues that a debasing demon “threatens universal doubt, in the sense of threatening knowledge of any proposition p” (Schaffer Reference Schaffer2010: 232). Schaffer holds that a debasing demon is “more nightmarish than a deceiving demon,” since even if some knowledge is not vulnerable to deception, any belief can be debased (Schaffer Reference Schaffer2010: 234).
The skeptical threat from debasing merits scrutiny. We should concur with Schaffer about the metaphysical possibility that any and all beliefs are debased. For example, any propositions could have been believed on a non-justifying emotional basis, such as hoping for their truth. Any such non-justifying basis is very peculiar for beliefs that are conspicuously well justified and not emotionally involving, such as believing that every square is a rectangle and that something with mass exists. The existence of a demonic agent who can alter the internal causes of belief is very far-fetched. But peculiar and far-fetched possibilities are possibilities.
Possible debasing does not imply that we lack knowledge, any more than does possible non-belief. One or more further premises are needed to connect a debasing possibility with a lack of knowledge. The more reasonable the asserted connection, the better the argument. Schaffer introduces the debasing demon, but he does not assess the strength of skeptical arguments from debasing.Footnote 1
The main goal here is to assess the skeptical threat. A traditional conception of knowledge will be assumed: knowledge requires having a properly based belief and meeting the other traditional requirements of truth, justification, and some condition that solves the Gettier problem.Footnote 2
Some forms of debasing lack skeptical potential because they would be detectable. It can be apparent to us on reflection that we believe out of wishful thinking, emotional craving, or the like. Possible debasing by inducing this sort of basis poses no skeptical threat. For the beliefs that seem to us to be knowledge, no such bad basis is reflectively apparent to us.
In the most skeptically powerful sort of debasing possibility, the debasing is not introspectively noticeable. It is possible for our beliefs to have been surreptitiously made to depend causally on our hoping for their truth, or our fearfully taking them for granted, as the case may be. The emotive attitude itself could have been secretly induced and suppressed, or if we already had it, secretly made to sustain the belief, without any sign of that dependency. There might have been no clue available to us that a belief had an improper basis.Footnote 3 Hidden debasing is part of any strongest skeptical threat from debasing. Since the improper causing is not introspectively noticeable, no additional deception is required to prevent the beliefs from constituting knowledge. We can call debasing that is not introspectively discernible pure debasing. All of our beliefs might have been purely debased. A skeptical argument from debasing can rely on that.
Various further premises might be offered to link the possibility of pure debasing to a lack of knowledge. For example, because pure debasing is hidden from introspection, there is some initial rational force to reasoning along these lines:
Our beliefs are debased in metaphysically possible worlds that are introspectively the same as this one. As a result we cannot tell whether or not our beliefs are debased. If we cannot tell, then we do not have knowledge. So we lack knowledge.
For another example of a link from debasing to ignorance, Schaffer's argument from debasing asserts that knowing depends on having evidence available enabling us to know that the known beliefs are properly based. The argument can be rendered as follows:
Knowledge requires proper basing. If we know a proposition, then there is some way for us to get in an evidential position to know that the belief is properly based. Yet there is no way for us to get into an evidential position to know that the basing is proper. So we lack knowledge. (Schaffer Reference Schaffer2010: 234)
The pure debasing possibility is rationally secure. The other premises asserted by such arguments should receive critical attention. These two sorts of arguments rely on our being unable to know that our beliefs are properly based. So they can be refuted if we can have evidence that gains us knowledge that our apparently known beliefs are properly based. If so, then we can “tell” that the beliefs are not debased by knowing that this is so, and those who gain the evidence are “in an evidential position” to know of proper basing.
Many other connections of debasing to ignorance can be asserted, with varied plausibility. A more carefully formulated version will be offered in the next section. What is most important is that if evidence is available that enables us to know that our beliefs are properly debased, then any argument that relies on the debasing possibility relies on something refutable. A main thesis here is that such evidence is available to us. We can know that our beliefs are not debased.
Debasing is one way for us to lack knowledge. More broadly, we lack knowledge if we do not meet any necessary condition. A broader skeptical argument can take advantage of all of the traditional conditions. One of its premises asserts that every proposition either is a debased belief or fails to meet some other traditional necessary condition. The other premise says that this disjunction is sufficient for our ignorance. That premise is thoroughly credible. Our not meeting any one necessary condition clearly implies that we lack knowledge. Arguing in this way might seem to make a stronger case for skepticism than arguing from debasing on its own. The initial disjunctive premise, the only questionable one, would hold true if the failed necessary condition were not the one denying proper basing. A debasing possibility might still help by making more plausible the failure of one or more of the other necessary conditions.
We shall see that debasing possibilities do not help this argument either. A second main thesis here is that acknowledging the debasing possibilities does not enhance any rational defense of a premise asserting that some necessary condition for knowledge is not met. Debasing turns out not to give any new rational support to skepticism.
2. SKEPTICISM FROM DEBASING
We have no indication that we are in any real danger of having debased beliefs. We have no positive evidence for the existence of a debasing demon or any other sort of debasing mechanism. We should acknowledge, though, that all of our beliefs might have been purely debased, that pure debasing occurs with no introspectable indication of the debasing, and that debased beliefs are not knowledge.
These facts set up our central question: what is the rational strength of the skeptical threat from pure debasing? The argument is off to a good start with a premise asserting the metaphysical possibility of the pure debasing of all beliefs:
Debasing Premise: All of our beliefs might have been purely debased.
The rest of the argument must link the Debasing Premise to our lacking knowledge. However a skeptical argument gets from this premise to that conclusion, the argument is refutable if we can know that our apparently knowledge is properly based. If we can know that, then we can know that debasing is not preventing knowledge.
First we should set up a representative version of the reasoning. One way to capture what such arguments have in common uses two more premises. The first one asserts that the possibility of pure debasing deprives us of any way to know that it is not actually occurring.
Undetectability Premise: If all our beliefs might have been purely debased, then none of us is able to know that they are properly based.
The Undetectability Premise is the place where the argument is susceptible to refutation, if it can be established that we are able to know that our beliefs are properly based.
The second premise makes the argument valid. It asserts that we lack knowledge if we are unable to know that our beliefs are properly based:
Knowability Premise: If none of us is able to know that our beliefs are properly based, then we lack knowledge.
The three premises imply:
Skeptical Conclusion: We lack knowledge.
The Knowability Premise is another potentially objectionable component of the argument. But it is pretty modest. It asserts that our knowledge depends on some of us being able to know of proper basing. That claimed ability is not epistemically ambitious. Although the existence of such knowledge is not just obvious, it does not seem to be problematic. What is required is only that some of us are able to know certain causal facts about apparently known beliefs. We shall see ample rational grounds to affirm this ability to know.
We are about to see our representative skeptical argument refuted. The argument stands in for skeptical arguments from debasing in general. The refutation extends to any argument that depends on our not being able to know that our beliefs are properly based. Applied to our stand in, the refutation justifies the negation of the Undetectability Premise. We shall see that acknowledging the Debasing Premise offers no rational impediment to our knowing that our apparent knowledge is properly based.Footnote 4 The reasoning argues for our knowing approximately all that we apparently know. So it is more than we need to justify denying narrower Undetectability Premises in less ambitious skeptical arguments from debasing that conclude by asserting some less universal claim of widespread ignorance.
3. A REFUTATION OF DEBASING SKEPTICISM
The most familiar traditional skeptical arguments rely on possible deception: our external world beliefs might have been false, and thus not known, while we were deceived by appearances into thinking that they were true. One way in which these arguments have been opposed is by an appeal to fallible internal evidence. It is argued that the evidence is justifying enough to enable us to know what we apparently know despite the possibility that we are deceived.Footnote 5
Arguing from fallible internal evidence can likewise show that the possibility of debasing does not cast any rational doubt on our actual knowledge. In fact, the argument against debasing is stronger. Unlike pervasive deception, pure debasing would be a localized causal oddity. Unlike pervasive deception, pure debasing would leave the world very largely as our evidence indicates. As a result, we have particularly strong evidence that our apparently known beliefs are not debased.
Pure debasing would affect only the link between justification and belief. It would leave our evidence deriving from conscious states exactly as it is. So we have that evidence, whether or not we are purely debased. We can rely on that evidence to argue against the debasing. Pure debasing also leaves unaffected the truth of virtually all that we apparently know. These resources are enough to show that our apparent knowledge is not purely debased or otherwise improperly based.
Here is some available evidence that our apparent knowledge is based on the justification that we have for the apparently known propositions.Footnote 6
We have evidence from memory and perception for the existence of an extensive range of close correlations between our having justifying evidence for the beliefs that we apparently know and our forming and sustaining those beliefs. A few arbitrarily chosen examples: we believe that Jane Austen wrote Pride and Prejudice when we have recollections supporting that she wrote the book, we believe that we are sitting when we have sensory experiences giving evidence that we are seated, we believe that there are infinitely many primes when we have testimony or proof of this proposition, and so forth. These extensive and diverse correlations straightforwardly confirm an explanation by causation. The correlations give us inductive support for the conclusion that the justifying evidence causes the beliefs.
The metaphysical possibility of debasing does not defeat the induction. Debased believing is a contrary possibility that is compatible with the premises about the correlations. In inductive reasoning there are always such contrary possibilities. They show that inductive evidence is fallible. Their possibility does not defeat its actual support.
Furthermore, the correlations between our justifying evidence and our believing the supported propositions match closely some familiar associations of evidence that supports other propositional attitudes. The causal nature of these other evidence-to-attitude associations is not placed in question by the possibility of purely debased believing. They include the cases in which our evidence justifies our being puzzled about a proposition, or our being doubtful about it, rather than justifying our believing it. Economy and simplicity greatly favor the view that all of these associations are causal.
We have other evidence for the causation. Often, as we draw a conclusion from evidence that supports its truth, we have the marked impression that we are relying on the evidence to believe the conclusion. We take ourselves to believe the conclusions in deference to the reasons from which they are inferred. These impressions of reliance and deference, with their causal implications, are evidence that the reasons from which we infer our conclusions cause us to believe them.
Finally, the best explanations we have of our believing what we apparently know overwhelmingly favor our having proper bases. For example, suppose that during an apparent walk in the woods we experience the distinctive malodorous scent of a skunk. Thereupon we believe that we experience the scent. The proper basing hypothesis is that we believe this by deriving the belief from our experience of the scent. The main rival debased explanations are these: our believing the proposition derives instead from some bad inner basis such as our wishfully thinking or emotionally needing the proposition to be true, or the believing derives from the actions of some being with a brain-tampering capacity who gives the believing some bad basis, or we believe the proposition by uncaused chance.
The explanations involving debasing are much weaker. The alleged debased inner sources of the belief are preposterously incongruous with what introspection seems to show us about our psychology – we find nothing resembling any positive attitude toward that scent. The alleged intervening beings would be entities of a sort that our evidence about what is present in our world argues massively against. The entities would be inducing mental alterations for which we have no hint of a mechanism or a reason. The chance hypothesis has the belief fit with the olfactory experience in a way that is altogether too coincidental and it conflicts with our evidence from science and common sense about how such beliefs arise in the presence of the experience. Thus, among the rival accounts of why we believe that we experience the scent, the properly based explanation outstandingly outdoes the competition.
That is a case of apparent knowledge by introspection. The explanatory outcome is approximately as decisive in favor of a proper basing hypothesis for virtually all of the propositions that we apparently know from our other principal justifying sources, such as memory, perception, science, and formal reasoning. In each case the debased explanations are drastically weaker than some simpler and more coherent explanations of proper bases as origins of the beliefs.
Evidence such as this strongly supports the proposition that our apparent knowledge is properly based.Footnote 7 This support is not undercut by the fact that it is possible for any of our beliefs to be purely debased. That remains possible. It shows the evidence to be fallible. It is not contrary evidence about how things actually are. In the absence of any reason to believe that the debasing possibility actually obtains, it does not undermine the strength of the evidence for proper basing. Thus, our evidence strongly justifies the proposition that approximately all of our apparent knowledge is properly based (“approximately all” to allow us to make an occasional error about what we know).
We can readily see that we meet the other traditional conditions for knowing that the beliefs are properly based. First, the possibility of debasing gives no suggestion that we do not believe that our beliefs are properly based. Equally clearly, the debasing possibility is entirely independent of the truth of the proposition that they are actually properly based. Furthermore, the debasing possibility plainly does not increase the likelihood of our being in a Gettier case. In Gettier cases there is some epistemic accident in the relation between our justification and why the belief is true. In contrast, the debasing possibility is a possible causal disruption in the relation between our justification and our believing the justified proposition. So the debasing possibility does not even suggest that the belief is subject to any Gettier problematic sort of accident.
Finally, we have plenty of the sorts of justifying evidence just described for the proper basing of the particular belief that affirms proper basing. We have the impression of using our reasoning to accept the belief. This case of believing is enough like other associations of evidence and attitude that are evidently causal to give inductive warrant to drawing the same conclusion about it. Also, our believing that our apparently known beliefs are properly based is much better explained as depending on our reasoning to this conclusion than as depending on some emotional basis depending on some invasive mental alteration, or existing by sheer chance.
Thus, all of the traditional conditions are met for our knowing that our apparent knowledge is properly based.Footnote 8
The assertion made by the Debasing Premise is that our beliefs might have been debased without our having available any introspective indication of that. Ample justification is available in support of this proposition about a possibility. Conceivable scenarios apparently illustrate how pure debasing might have happened. Nothing defeats that appearance of possibility. So the Debasing Premise is justified. Thus we have support for the antecedent of the Undetectability Premise. This justification for the antecedent has given us no hint of a reason to deny what the evidentialist reasoning has defended: that our apparent knowledge is properly based.Footnote 9 That claim of knowledge is the negation of the consequent of the Undetectability Premise. So we have justification for the antecedent of the Undetectability Premise, along with justification for the denial of its consequent. This justifies the denial of the Undetectability Premise. The Undetectability Premise is an essential part of our representative skeptical argument from debasing and so that argument fails. More generally, the reasoning to establish that we can have knowledge that apparently known beliefs are not debased refutes any skeptical argument that rests its case on the debasing possibility. Our having available knowledge that apparently known beliefs are properly based decisively disqualified possible debasing from showing any actual ignorance.Footnote 10
Objection: For all that has been argued against debasing, the cited evidence itself is not secure from the activities of a debasing demon. Without invoking any radical internal skepticism that extends to our knowing the contents of currently conscious states, it remains possible that a debasing demon is giving us the misleading impression that we have the portion of the cited evidence that is beyond present awareness. The cited evidence could still be misleading, or mostly absent altogether, while our beliefs were being debased. So the possibility of actual debasing has not been excluded. Skepticism from debasing has not been refuted.Footnote 11
Reply: Refutation does not require excluding a possibility by denying its possibility. Providing knowledge that a premise in an argument is false is sufficient to refute the argument. A fallibilist anti-skeptical response offers knowledge of the denial of a skeptical premise, thereby refuting the skeptical argument without denying the possibility of skeptical scenarios. It is possible that not all of the cited proper basing evidence exists. It is possible that only the currently conscious portion of the evidence really exists while we are given the illusory impression of remembering that we have the rest as well. A best fallibilist response acknowledges this possibility. It does not strengthen the case for debasing. Either the unimpeached portion of the evidence is justifying, or not. First, there is reason to think that it is justifying. It includes the possibly deceptive impression that we have the rest of the non-conscious portion of our apparent evidence for proper basing. This might be evidence enough. A plausible principle asserts that if we have evidence E for the existence of evidence in support of X, then E is itself evidence for us in support of X.Footnote 12 If this principle is true, then the memory impression of having the rest of the evidence itself supports proper basing. So the unimpeached evidence might be enough to justify the denial of debasing.
Suppose, to the contrary, that either this evidence-for-evidence support exists but it is not strong enough to justify the denial of debasing, or that the principle is mistaken and the fragmentary conscious evidence that the possibility allows us does not really support the denial of debasing. If so, then it is noteworthy that this is a possibility of deception that goes beyond pure debasing. Some apparent evidence does not actually exist. This would be a version of skepticism argued not merely from debasing, but also from the possibility of further deceiving evidence. So it is not a skeptical threat from debasing itself. In any event, a fallibilist response addresses it in the familiar way: the undefeated apparent memory justifies the proposition that this deception is not happening. Where that evidence is not deceptive and the rest of the conditions for our knowing this are met, we know it. And our evidence from conscious states, prominently including other apparent memories, supports that those other conditions are actually met.
4. DEBASING-ENHANCED SKEPTICISM?
None of the traditional conditions on knowledge must be met by any propositions, except those that could not have been untrue. Any other proposition might not have been true, and any proposition might not have been believed, it might not have been justified, its believing might not have been properly based, and it might have been in a Gettier case. Even if skepticism derived entirely from debasing has been refuted, a debasing possibility still could enhance some reason to think that one of the other traditional necessary conditions for knowledge is not met. If so, then a debasing possibility helps to promote skepticism. In other words, it then strengthens our reason to believe the first premise in the following argument.
General Skeptical Argument (GSA)
Disjunctive Premise: Each proposition fails to meet one or another of the traditional conditions on knowledge.
Disjunctive Conditional: If the Disjunctive Premise is true, then we lack knowledge.
Skeptical Conclusion: We lack knowledge.
Given the traditional conditions for knowledge, GSA makes minimal skeptical assumptions. It requires only the failure of each proposition to meet some traditional necessary condition or other. If appeal to debasing strengthens any skeptical argument against knowledge as it is traditionally conceived, then it strengthens GSA.
Assuming that the traditional conditions are necessary, the Disjunctive Conditional is clearly true. It is the Disjunctive Premise that is rationally disputable. We have no need to rehearse existing skeptical disputations. They have been largely about whether we have adequate justification. Our new question is quite narrow: does possible debasing make it any more reasonable to think that one of the other knowledge-denying disjuncts actually holds for some range of propositions that we seem to know?
In a word, no. We have already seen this, in effect, when we considered how a justified denial of debasing relates to knowledge of that denial. We saw that the possibility of pure debasing is rationally unrelated to whether or not we meet the other traditional conditions. The propositional content of a purely debased belief is as well justified as is a properly based belief, and justified for exactly the same reasons. So the debasing possibility adds no new doubt about whether or not we have good enough justification to know. Debasing would not affect the existence of the belief, or its truth. So debasing possibilities do not support denials that these necessary conditions are met. Finally, the ways in which debasing might have occurred do not make more likely the presence of Gettier cases. Again, the relata involved in meeting the two conditions on knowledge are different. The relations are rationally independent. Debasing gives the believing some improper relation to our justification. In a Gettier case, the truth of the believed proposition is in some way too accidentally related to our justification. Thus, no matter what was our basis for believing a proposition, our justification for it could as easily have connected in a good way to the truth of the proposition. Our evidence supports that it is the good connection that holds in our apparent knowledge. Possible debasing casts no doubt on that.
Thus, debasing possibilities do not yield any new support for our failing to meet some other condition on knowledge. A refutation of the debasing disjunct itself is offered in the previous section. So no disjunct in the Disjunctive Premise in GSA is supported by the possibility of debasing.
5. CONCLUSION
Although all of our beliefs might have been purely debased, this possibility does not improve any argument for skepticism.Footnote 13
