From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel

1. The Argument from Illusion

The argument from illusion infers from the fact that we can have illusion which is phenomenological indistinguishable from perception to the conclusion that the state of appearance — subjective phenomenal awareness — is fundamental to all cognitive state which deserves basic epistemological status. The following is a brief formation of the argument from illusion:

• (1) Our sense perception can be deceptive: it can appear to one exactly as if things were a certain way when they are not.
• (2) A deceptive case can be experientially indistinguishable from a veridical case.
• (3) One’s phenomenal awareness is the same in both deceptive and veridical cases. In other words, perception and illusion include the same state, namely appearance.
• (4) In illusion, one’s phenomenal awareness falls short of the fact. The objects of subjective experience cannot be facts but appearances.
• (5) Likewise, in perception, the objects of experience are not facts but appearances.

Here, the argument employs an unorthodox method by explaining standard situations in terms of non-standard ones, that is, explaining perception in terms of illusion. One main motivation underlying this method is to isolate a concept of epistemological justification — as I shall explain, an internalist conception, according to which epistemological appraisals depend essentially on what is internal to a perceiver. On this view, an agent’s epistemic status is determined solely by his internal mental conditions such as what he is consciously aware of, what he takes to be true, and what he deems reasonable. Given that these internal conditions are completely the same, there seems no ground to attribute different epistemological entitlements. For instance, if an agent having an illusory experience about a blue vase is in exactly the same subjective phenomenal states as he would be were he to perceive one, then he is no less entitled to assert “There is a blue vase” than he were in a genuine perceptual situation. In this sense, two phenomenally indistinguishable mental states are said to share the same epistemological status.

The internal conception of epistemological entitlement makes clear why the analysis must start from a failed case rather than a successful case. There is an important type of epistemological evaluation that can be made intelligible only in a failed case.¹ Basically, a failed perceptual case is a situation in which there is a split between the inner mental conditions and how things are in the world, e.g., when someone has apparently good reason for his belief which happens (or turns out) to be false. In such a case the internal condition is met but not the external condition; and since the external condition is out of one’s control, the epistemological status seems to depend crucially on whether the person takes up his responsibility in a blameless manner. If he does not commit any mistake on his part, he must deserve entitlement of some sort. This type of epistemological entitlement is too important to be ignored and, moreover, it boasts a major theoretic attraction — it can be attributed to both illusory and perceptual subjects: whereas an illusory subject enjoys this entitlement, a perceiving subject receives extra credit on top of it. The argument from illusion hence attributes basic epistemic entitlements to appearance and derivative ones to perception.

2. Two Types of Epistemological Justification

The argument from illusion suggests that appearance, as the unit of all states of phenomenal awareness (veridical and deceptive alike), occupies a basic epistemological standing. On this view, one obtains this basic epistemic status simply by having the appearance that things are thus and so. When S has the appearance that P, he is entitled to believe or assert that P, whether or not it is a fact that P. Given that S believes that P on the basis of his appearance that P, he is epistemologically responsible. Thus, when S’s belief turns out to be false, he is blameless, since there is nothing S can do to improve his epistemological situation: S is in exactly the same appearance state as he would be were it a fact that P. S’s belief is, in this light, reasonable. We may therefore call this normative status associated with appearance reasonable entitlement.

There are of course other cognitive states that deserve epistemological statuses. The idea of factive states has gradually attracted considerable philosophical attentions. Roughly speaking, a factive state is a state in which a subject perceives, or “takes in,” a relevant fact. As Wittgenstein writes,

“I know” has a primitive meaning similar to and related to “I see.” . . . “I know” is supposed to express a relation, not between me and the sense of a proposition (like “I believe”) but between me and a fact. So that the fact is taken into my consciousness (Wittgenstein 1969, § 90).

Seeing, knowing, and remembering are typical factive states, states whose existence implies the obtaining of relevant facts. For instance, that one remembers that it snowed yesterday entails that it snowed yesterday; one knows that there is a blue vase entails that there is a blue vase. Factive states are not basic mental states — at least they are not as minimal as appearances are supposed to be. Factive states, however, are central to mental states since they indicate a “matching relation” between mind and the world (Williamson, 2000: 40).

Wittgenstein provides a vivid picture of the matching relation between mind and world, when he addresses the immediate connection between meaning and facts. He states, “When we say, and mean, that such-and-such is the case, we — and our meaning — do not stop anywhere short of the fact; but we mean: this—is[—]so” (Wittgenstein 1951, § 95). This stance is in direct contrast with a traditional picture of mind, according to which mind and objects are made of difference substance such that what one sees is not normal physical objects but something representing them. Wittgenstein denies such gap between mind and the world. McDowell elaborates this to the effect that in veridical experience the content of thinking is a fact; in his words, a perceiving subject has the fact “in view.” He suggests, “To paraphrase Wittgenstein, when we see that such-and-such is the case, we, and our seeing, do not stop short of the fact. What we see is: that such-and-such is the case” (McDowell 1996, 29).

There are various accounts of the matching relation; for present purposes, we need only a very mild version that what we are experiencing (or thinking, in general) and what is the case can, in principle, be in agreement. The matching relation points toward the perceptual contact (perceptual success) between mind and world, and is therefore fundamental to the possibility of thought, language, and action — the matching relation must be presupposed in any account of the contentfulness of thought, the acquiring of language, and the practical reason for action. In this paper I will begin with practical reason, which in my view is the best way to illustrate the matching relation that underlies factive states.

Factive states, so understood, enjoy a certain type of entitlement. When one is in a factive state, that is, when a fact is taken into one’s consciousness, the obtaining of the fact is constitutive of his epistemological entitlement — it enables him to make a relevant assertion which precludes the possibility of falsehood. This feature is absent in reasonable entitlement: one can have reasonable entitlement even when one’s belief turns out to be false. For example, when one forms a belief on the basis of appearance alone, one is reasonably entitled to his belief, but being reasonably entitled does not guarantee the belief to be true. Let’s call the type of entitlement one enjoys when one is in a factive state factive entitlement.

The argument from illusion of course would not deny the characterization of factive states and the relevant entitlements. Nevertheless, it would insist that reasonable entitlements has explanatory priority over factive entitlements — factive entitlements have to be understood in terms of reasonable entitlements. In the following I will try to show why the order of explanation should be reversed, by considering some issues about practical reasons.

3. Two types of reasons for action: belief and fact

In order to explain the contrast between reasonable and factive entitlements, I will start with a similar distinction between two types of practical reasons. It is usually claimed that what constitute reasons for action are beliefs rather than facts. Compare the following two cases.

• (i) S believes correctly that it is raining, and he takes an umbrella on the way out.
• (ii) S believes that it is raining — in fact, it is not raining — and he takes an umbrella on the way out.

In the first case, the reason for S’s action of bringing an umbrella is obvious: he knows the fact that it is raining. The fact (or, more precisely, S’s being in this factive state) explains and justifies his action. In the second case, S’s reason for action is not fact but belief — he believes that it is raining and thus performs the same action in the absence of fact.

The question concerning us is, “in the two cases, does S have the same reason for action?” It is tempting to reply that S has the same reason for action, for he has the same belief in both cases, even if the belief has different truth-value in the two situations. The idea is that one acts in accord with one’s belief and whether the belief is true is a further question: Given the same belief, the agent would perform the same act. On this theory, what explain action is belief rather than fact; or alternatively put, belief is the proximal reason for action, while fact distal.

An immediate problem with this approach is that it can explain the sameness of the cases but not their difference, since it implies that the two actions do not have essential difference — they are the same type of actions caused by the same reason (namely the same belief). What makes the two cases different is something accidental: the belief in the first case happens to be true, and its being true does not play a role in the rational explanation of the action. On this account, belief exhausts the explanation and leaves no room to truth in practical reason. This consequence is perplexing because believing is basically a take-true attitude. S believes that Prozac can reduce depression only if S takes it as true that the medicine can actually cure his disease. For what is essential to the explanation is that he has the belief whose truth rationalizes his action. Belief rationalizes action only in an elliptical way; facts provide the ultimate source of justification for action.

This point can be further supported by the following fact: in a deceptive case, the subject may have reason of some sort, but he does not have the reason he thinks he has. The reason he thinks he has is the fact-related reason, i.e., the reason that he can have when he is in a standard factive situation — the situation in which he thinks he sees the fact and acts accordingly. In general, a practical explanation in terms of belief presupposes a practical explanation in terms of fact: a belief-rationalization makes sense only if a corresponding fact-rationalization is in place.

What reason does S have, given that he does not have the fact-related reason for action? In the deceptive case S thinks he acts on a fact-related reason but he doesn’t; nevertheless, he acts according to his belief, and his belief is supported by his phenomenal state in exactly the same way the belief in a veridical case is supported by the phenomenal state. Accordingly, S’s action is deemed reasonable. In other words, the justification of an illusory subject’s action comes from a (prospective) fact-related reason via phenomenal indistinguishability. In sum, both the explanations of veridical and misleading cases make reference to the fact-related reason; thus, belief-related reason relies on fact-related reason for its intelligibility. In this sense fact-related reason is said to be more basic than belief-related reason.

4. Factive and reasonable entitlements

The relation between fact-related and belief-related reasons for action can shed light on the relation between factive and reasonable entitlements. Again, let us consider the following contrast:

• (i) S believes that it is raining because he sees it.
• (ii) S believes that it is raining because he has a mere appearance which is indistinguishable from seeing that it is raining.

In the first case, S’s perceptual belief is justified because he perceives the fact. His being in this factive state explains and justifies his belief. In the second case, S’s reason for his belief is not fact but the mere appearance that it is raining — he believes that it is raining on the basis of the appearance but in the absence of fact.

The question concerning us is, “in the two cases, does S have the same entitlement or reason for belief?” It is tempting to reply that S has the same entitlement for belief because he has the same appearance state in both cases, except that in the first case the appearance happens to be veridical. The idea is that one forms a belief in accord with one’s appearance and whether the appearance is veridical is a further question: Given the same appearance, the agent is equally entitled to form the same belief.

An immediate problem with this view is that it can explain the sameness of the cases but not their difference, since it construes the two perceptual beliefs as essential the same — they have the same content that is based on the same appearance. What makes the two cases different is something accidental: the appearance in the first case turns out to be veridical and its being veridical is external to the entitlement of perceptual belief. On this account, appearance alone determines epistemic entitlement, in which veridical experience does not play a role. The consequence is confusing because the ultimate source of justification for perceptual belief traces back to veridical experience, i.e., experience directly connected with what is the case. The point of the epistemological appraisals of perceptual experiences is to reflect the extent of a subject’s sensitivity to the world surrounding him, and the point goes missing if the evaluation is done in a way that is indifferent to veridical experiences. Furthermore, the contentfulness of perceptual belief seems to presuppose veridical experience. According to a widely accepted theory of thinking, the content of perceptual belief is determined by its normal causal connection with the relevant features of the world, and this connection can be located or established only in a context of successive veridical experiences.²

This point can be further supported by the following fact: in case (ii), S may have entitlement of some sort, but he does not have the entitlement he thinks he has. The entitlement he thinks he has is the factive entitlement, that is, the entitlement that he can have only when he is in a standard factive situation — a situation in which his belief is based on the fact he has in view. What entitlement does S have to his belief, if he has no factive entitlement? In deceptive case S thinks he has a factive entitlement (he thinks he sees the fact) but he doesn’t; however, his belief is based on his appearance in exactly the same way that a belief in a perceptual case is. In this light, S’s belief is regarded reasonable. In brief, the justification of an illusory subject’s belief comes from a (prospective) factive entitlement via phenomenal indistinguishability. Thus, both the explanations of veridical and misleading cases make reference to factive entitlement. It follows that reasonable entitlements depend on factive ones for their intelligibility. In this sense factive entitlement is said to be more basic than reasonable entitlement.

5. Concluding Remarks

In this paper we adopt an approach combining Wittgenstein’s elucidation of factive mental states and his construal of the identity relation between what can be thought and what is the case, according to which factive entitlements are shown to be explanatorily prior to reasonable ones. Hence the argument from illusion offers only a partial notion of epistemic credit and thus fails to confer fundamental epistemological standing to the state of appearance.