Uncertain Knowledge

A patient came into Emergency Psychiatric Services (EPS) with complaints of suicidal and homicidal thinking. The patient reported a past history of multiple suicide attempts and harm to others. She presented as depressed. Nevertheless, at the end of the dispositional, the attending assessed the patient as malingering. Asked by others on the health care team how that assessment was made, the psychiatrist was at a bit of a loss. "It's subtle," he said.¹

What would make this a case of the psychiatrist "knowing"?

The standard account of knowledge as justified true belief places emphasis on justification, where justification involves being able to give reasons as to why one believes what one does, thus ruling out lucky guesses. Karyn Freedman offers an original and captivating argument that a person can have knowledge of something and yet not be able to give justification for it. She draws upon the work of Robert Brandom, who argues that a case exists (of chicken-sexing) where one can know that some proposition p is true even if she cannot give reasons for her belief that p. Freedman agrees, but argues that Brandom's theory has much wider application than he supposes. She considers victims of trauma and argues that, given the widespread occurrence of sexual violence against women, knowledge without reasons is an everyday phenomenon.

The alternative theory of knowledge that Freedman employs is called "reliabilism," and the idea is that true beliefs that are reliably formed count as knowledge even if the knower cannot say how he or she knows. Reliable belief formation comes from sensory experience, cognitive processes, and so on. To count as knowledge, all we need is that our true beliefs be produced by a reliable method such as reliable cognitive processes.

Reliabilism is contentious as a theory of knowledge, but it is compelling in many ways. Consider the example above, where it seems plausible to say that the psychiatrist knows that p, where p = "This patient is malingering," without being able to give reasons. He has a gut feeling, but is unable to articulate the source of the intuition. His belief that p counts as knowledge if it is true and if it has been reliably formed. The reliability, in this case, most likely rests on the attending's long-term observation of patients, sensory data that he is taking in on a subconscious level, memory of other patients who say they are depressed or suicidal, and so on.

This inability to give reasons, though, turns out to be a difficulty for the rest of the health care team; many of those present in EPS are students who need to understand what the attending is basing his claims on. It seems that some things can only be learned through direct experience and not by being taught facts or theories; residents will eventually know how to tell the malingerers from the mentally ill but they, too, may know that p without being able to justify it.

Reliably formed knowledge without access to reasons occurs in other areas of life as well. Some people can look at a recipe and declare it an excellent one while their partners are unable to discern a good recipe from a lousy one. Asked how one identifies a good recipe, one might be at a loss to explain. Are some people just lucky, or have they been able to form beliefs about combinations of spices, textures, and tastes in a reliable way, on which they can draw for future recipe choices?

But this type of example is, perhaps, trivial in the face of trauma and mental illnesses. What we want is the application of Freedman's reliabilism to psychiatry. Freedman reminds readers of the extensive literature that suggests that victims of traumatic events or experiences block those experiences from memory. But, she points out, those experiences are nevertheless informational: they frequently shape and modify the sorts of beliefs a victims holds—beliefs about herself, her perpetrator(s), her place in the world, and even the world itself. Even though she may not have access to how those altered...