Knowledge in Theory and Practice

As human beings, we all hold beliefs about which we feel so certain as to call them “knowledge” (in the sense of “knowledge-that”). Despite our confidence in these beliefs, in some future we might encounter information showing that they’re mistaken. We would then reevaluate these beliefs and say that they were never really knowledge; but even though they were not knowledge, we could still have been justified in calling them knowledge at the time.

Consider this case: Mary says that she knows that Valentine’s Day will be a on Wednesday because that’s what she saw in the calendar. Unbeknownst to Mary, the calendar she consulted was one for the previous year; this year’s Valentine’s Day won’t fall on a Wednesday. Later on, Mary realizes her mistake and corrects her beliefs. She now looks back on her prior belief and concludes that it didn’t really qualify as knowledge.

In the case above, Mary thought she knew that Valentine’s Day will be on Wednesday because she thought that she had all the relevant information; it wasn’t until later that she realized this wasn’t the case.

When Mary first called her belief “knowledge”, the label was expressing a strong sense of certainty regarding the belief in question. When she later judged that it had never been real knowledge, she was using “knowledge” to mean a belief that meets certain objective criteria, such as being true.

To differentiate between the two, we might call the former “knowledge-in-practice” and the latter “knowledge-in-theory”.

Knowledge-in-theory is about knowledge as an ideal. A belief counts a knowledge-in-theory when it meets a set of objective criteria, such that if we had knew all of the relevant information regarding that belief, we would be able to judge for certain whether or not the it qualifies as knowledge. By and large, this sort of definitive judgement is only possible in hypothetical cases, such as the one above, where by simply reading the case we can take the information it provides as objectively and infallibly true as far as that case is concerned. In the above case, for instance, when we are told that Mary checked the wrong calendar, we are to take this as an established fact.

Outside of such cases, we are constantly faced with situations in which we can never be sure that all our information is correct or that we have all the necessary information to judge whether or not a belief is true or justified, like how Mary didn’t know that she was consulting the wrong calendar. This is where knowledge-in-practice comes in. When dealing with matters about which we can’t (yet) be absolutely certain, we can still treat beliefs about them as knowledge when we feel they carry a high degree of certainty, such that it seems they are very likely to be true.

The collective means by which a person decided when their beliefs can be considered knowledge-in-practice are what we might call our “sense of knowledge”. Ideally, this sense of knowledge aims at making our knowledge-in-practice as close as possible to knowledge-in-theory. This means that what we take as true should be true, and what is true we should take as true. Put another way, our sense of knowledge should not lead us to believe something which is false, nor to fail to believe something which is true (assuming that we have enough information to justify the belief in question).

Of course, we don’t begin our lives with a sense of knowledge already capable of this. Rather, our sense of knowledge must be refined over time, becoming ever more reliable. This is done in large part through trial and error.

As we go through our lives, we use various methods to form beliefs. Later on, we might encounter new evidence with disproves or supports any given belief. If a method consistently gives us beliefs for which we find more support later on, we come to regard that method as a reliable source of true beliefs. Conversely, if we find that a method often gives us false beliefs, we come to regard that method as a poor source of true beliefs, and either stop making use of it, or make corrections to make it more reliable. In the above case, for instance, we might assume that when Mary consults a calendar in the future, she will make sure that it corresponds to the correct year. In this manner our sense of knowledge approaches, though never quite achieves, the ideal of perfect knowledge.