Introspective Knowledge
In everyday contexts it sounds odd to say of someone that she believes she has a headache, and equally odd to say she knows. This has encouraged some philosophers to conclude that it is inappropriate to talk of either belief or knowledge in such cases.1
An alternative diagnosis, however, is that the “owner” of a headache is in such a superior position compared with others with respect to determining whether she herself has a headache that it is usually enough to say of her that she has a headache and unnecessary, and hence odd sounding, to add that she believes or knows this.
What exactly is this superior position? Some philosophers maintain that first-person beliefs about headaches are infallible; S cannot believe that she has a headache without understanding what a headache is, but insofar as she understands what a headache is, she cannot be mistaken about whether she has one.2 Others disagree, arguing that belief acquisition always involves background assumptions and hence it is always possible for things to go wrong, but S’s beliefs about her own headaches do at least have a positive presumptiveness about them.3 Still others make the point that headaches are among the psychological states that automatically reveal themselves to their owners. Timothy Williamson’s term for such states is “luminous,” while Roderick Chisholm’s is “self-presenting.”4
Whatever their other differences, however, virtually everyone concedes that S has an awareness of her own headaches that others lack, an awareness that involves a special access to the headache’s defining characteristic, namely, its unpleasant quality. Her beliefs about her own headaches are thus different in kind from the beliefs that others can have about her having a headache. It is this that accounts for the much tighter than usual link between belief and knowledge.
Recall the beetle in the box story, in which S has a true belief that there is a beetle inside the box but lacks knowledge, but not because she is unaware of available information. It is the world, not S, that has the defect. The world is so informationally impoverished with respect to the box that it is not possible for anyone to know that there is a beetle inside.
The beetle story is the reverse of cases involving awareness of one’s own headaches. There are no important truths about the box that others have and S lacks, and yet she seems not to know that there is a beetle inside the box. By contrast, if S has a true belief that she has a headache, there can be any number of truths about the headache that others have and she lacks, for example, ones about its etiology. These truths are nonetheless unlikely to be grounds for denying that she knows, given that she has a special access, an access that others of necessity lack, to its characteristic unpleasantness.
In all but a few unusual situations in which knowledge is blocked, when S has a true belief P but we are nonetheless inclined to say she does not know P, we are aware of something important about the situation that she doesn’t “get,” but with respect to her own headaches, we realize that the key feature of the headache is precisely the feature she gets all too well. First-person beliefs about headaches thus tend to be instances of narrow knowledge. Little information beyond awareness of the headache itself is relevant to whether she has adequate information, not even information that is noteworthy with respect to other purposes. Suppose the headache is being caused by a tumor that if left untreated would cause her serious health problems. We the audience are aware of this, but S is not. This is clearly a consequential truth, but we are nonetheless not likely to regard it as precluding her from having adequate information about the fact that she has a headache. Like other cases of narrow knowledge, little matters beyond the truth of her belief.
More precisely, this is so with respect to first-person, present tense beliefs about headaches. Beliefs about one’s own past and future headaches, even if true, do not so readily qualify as knowledge, because these beliefs, being about the past and future, do not involve a special access to the defining unpleasant quality of the headache. So, additional information can be relevant, just as it is in assessing whether one knows that someone else has a headache.5
Nor need first-person, present tense beliefs about other kinds of psychological states be cases of narrow knowledge. Think of states less closely connected with a distinctive sensation and more closely associated with telltale patterns of behavior, for example, jealousy, envy, and fear, and states involving propositional attitudes (what one believes, hopes, intends, wants, and so on). There is more room for observers to think that the owner lacks key information about these kinds of states, and hence more room to deny that she has knowledge.
A simple way to illustrate this is to consider the role that others can play in getting the owner to be aware that she is in such a state. Consider envy. Others can observe envious patterns in S’s behavior and point them out to her. If S herself has been unobservant of these patterns, she can be genuinely surprised to discover that she is envious. Correspondingly, when she is not envious, she can have a true belief to this effect and yet not know it, because once again she may not have been sufficiently attentive to her own behavior.