Instability and Knowledge
Consider a story that raises issues not of fixedness but instability.
Recall Sally, whose beliefs are as accurate and comprehensive as it is humanly possible for them to be. She has true beliefs about the basic laws of the universe as well as complete as possible information about its history, and using these laws and information, she can explain virtually everything that has happened, is happening, and will happen. Moreover, she has accurate and comprehensive beliefs about how it is that she came to have such accurate and comprehensive beliefs.
Now retell the Sally story so that a demon has instilled these beliefs in her, and every two seconds the demon replaces her true beliefs with massively mistaken beliefs for two seconds. Her belief system is thus the analogue of a blinking light. For two seconds she has maximally accurate and comprehensive beliefs; for the next two seconds she is massively deceived; two seconds later, she once again has accurate and comprehensive beliefs; and so on. During the two-second intervals in which Sally has accurate and comprehensive beliefs, does she have knowledge?1
The story is bizarre even by the standards of the evil demon genre. Indeed, so bizarre that it may not be intelligible. Is it really possible for states to blink on and off from moment to moment and still be genuine beliefs? Or is there a degree of volatility that states must not exceed if they are to qualify as beliefs?2
Suppose, however, we agree at least for the moment to waive this concern and grant that these are genuine beliefs. During the two-second interval in which her beliefs are completely accurate and comprehensive, does she then have knowledge? To be sure, she will have these true beliefs only for a brief time, but this would not seem to preclude knowledge automatically. It is not impossible to have fleeting moments of insight. I may suddenly understand a logic proof, but I then get interrupted and when I return to the proof, the insight is gone. This does not necessarily mean that I did not have knowledge for a moment.
More exotic examples may be found as well. Think of mystics who claim that some knowledge by its very nature is such that it cannot be sustained. It is too fragile. When every circumstance is right, powerful insight is possible, but when the smallest detail is altered, insight is lost.
If there is mystical knowledge of this sort, it cannot help but be of short duration. The qualification “if” is needed, because the point here is not that the claims of professed mystics are legitimate but rather that their claims to knowledge should not be dismissed as inherently incoherent.
In everyday situations, if one has true but unstable beliefs and is unaware of the source of the instability, this gap in one’s information can be used to explain why one lacks knowledge. This is not Sally’s situation, however. During the intervals in which the demon is enlightening as opposed to deceiving her, Sally has complete as possible information about the world, including information about how the demon is providing her with all this information and also how this demon will soon be deceiving her. She is fully aware, in other words, of how accurate and comprehensive her beliefs are but also how ephemeral they are.
Demon and brain-in-the-vat hypotheses typically make use of the familiar narrative technique in which the audience is given information that the characters of the story lack. The audience is aware that the characters are being massively deceived, but the characters themselves are not. Sally by contrast does not lack information about her situation that is available to the audience. We the audience thus don’t have the standard rationale for denying she has knowledge. On the contrary, we are aware that her information is vastly superior to ours, if only for brief, intermittent periods. Why then shouldn’t she be thought of as having knowledge?
On the other hand, intuitions about cases as bizarre as this are apt to vary. It is thus more important to account for the sources of intuitions than insist on any particular one as correct.
Insofar as there is a pull in the direction of knowledge in this kind of case, this is explainable by Sally’s having as close to perfect information as it is possible for humans to have, but insofar as there is pull away from knowledge, this too is explainable. Her blinking states are so strange that we may be reluctant to regard them as beliefs, but then, since knowledge requires belief, we also may be reluctant to grant she knows.
It is possible, of course, to retell the story so that the alternating moments of enlightenment and ignorance are extended from two seconds to, say, two weeks or two months or even two years. If so, we are likely to be more comfortable in regarding them as genuine beliefs, but correspondingly it will then also seem less problematical to grant that she has knowledge during the periods of enlightenment.