What Theories of Knowledge are Theories of
In this article I try to answer two questions: First, what is it that we’re trying to do in formulating a theory of knowledge? One (unhelpful) answer to this question that we’re trying to construct a true theory of knowledge. But what are theories of knowledge theories of? Now, in general, attempted answers to a question of the form “What are theories of x theories of?” face a dilemma: The answer that they are theories of x is uninformative, but the answer that they are theories of x according to some particular account of what x is is illegitimate, because it is question-begging. Second, can we learn anything from the fact that Gettier cases are difficult to solve? I believe the answer to the second question is “yes”, and that inquiring into why that is will give us a clue to resolving the dilemma we face in trying to answer the first. I begin by explaining what a Gettier case is, and explain why I think there is no agreed-upon response to such cases. I then examine the notion of a theory of knowledge, and the issue of what makes different such theories genuinely rival theories of the same thing. I then give my own account in terms of the notion of a paradigm example, arguing that theories of knowledge should be theories of how paradigm examples of knowledge sources work. Along the way, I answer objections to my views.
Up until Gettier first presented a Gettier case, it was commonly assumed that knowledge was justified true belief. In order to know something, you must believe or think that it is true: even if intelligent life exists elsewhere in the universe, you can’t count as knowing that there is if you fail to believe there is. Furthermore, what you believe must be true: In order for you to know that there’s intelligent life somewhere out there, there has to be intelligent life somewhere out there. Finally, your belief must be justified: Believing in an alien civilization because you have observed their spaceships through a telescope is one thing, forming that belief after slipping and hitting your head on the sidewalk is quite another—even if, in the second case, you happen to be right. In his paper, Gettier showed that while these conditions might be necessary for knowledge, they are not sufficient. One example of a Gettier case would be this: Jones knows that there’s going to be a promotion in his company. A manager tells Jones that Smith is going to get it. Also, Smith tells Jones that he’s feeling lucky because he just found a penny, which he shows to Jones and puts in his pocket. Jones thus forms the belief that the man who will get the promotion has a penny in his pocket. Unbeknownst to Jones, the manager is wrong. Jones will get the promotion, and though he’s forgotten about it, he too has a penny in his pocket. So Jones’ belief is both justified and true, and yet most people think that Jones doesn’t know that the man who will get the promotion has a penny in his pocket. However, there is no agreed-upon solution to cases of this sort. Some add more conditions to their definition of knowledge, some modify their account of justification, while others do still other things. Everyone feels that something is wrong with the traditional account, but we disagree about how to fix it.
What could account for this lack of consensus? We can answer that question by asking a superficially unrelated one: What is it that makes different “theories of knowledge” rival theories of the same phenomenon—as opposed to being non-rival theories of different, though perhaps very similar phenomena? For example, what would make reliabilism and internalism rival theories of knowledge, as opposed to two non-rival theories—one of beliefs produced by reliable belief-forming mechanisms, and the other of beliefs that one can tell “from the inside” to be true? My suggestion is that in that the answer is “nothing.” They are “rival” theories only in the sense that they are verbally different definitions of the same word—’knowledge’—but nothing in either theory ensures that that word has the same meaning or reference in both cases. These definitions may both be right without any inconsistency, because they can apply to different things.
At this point one could object that there is a common core to all attempted accounts of knowledge. Knowledge, whatever else it may include, must include justified true belief as an essential part, and whatever it is that includes justified true belief is just what knowledge is. This is the common element that makes the different accounts rivals. My response is that there is really no more agreement in philosophy over the nature of justified true belief than there is over knowledge. There are different philosophical theories of justification, belief, and truth, but we shall focus exclusively on the first. Some philosophers think that justification requires having conclusive evidence, though most do not. Among those who do not, it is not agreed whether there is some particular degree of justification that is required for knowledge, nor, if there is, what it might be. It is also not agreed whether it requires having evidence or not, nor whether, if it does, that evidence must be internally accessible to the knowing subject. The point is that, if I’m right to question the idea that the dispute over the nature of knowledge constitutes a genuine disagreement, due to the fact that the different accounts don’t have enough in common, appealing to the notion of justified true belief does nothing to help matters.
The key to resolving our problem lies in the notion of a paradigm example. In giving a paradigm example, one answers the question of what something is not by offering a description or definition of it, but by making someone conscious of the thing itself. If someone wants to know what chess is, I can take them to a chess match so they can watch people play, and later on I can teach them to play themselves. (I construe the process of play broadly, to include consulting rule-books and similar things.) The games that they observe, or participate in, serve as paradigm examples of chess. If different people who had never played chess, or who were still learning to play, disagreed about what chess was or what it entailed, one could appeal to such paradigms to resolve their disagreements. Because they are not descriptive, paradigm examples enable one evade the dilemma we faced in trying to answer questions of the form “What are theories of x theories of?”. They explain what x is, not by representing x as being a certain way, but by making one conscious of examples of x, which are themselves that way. I think this applies, in particular, to theories of knowledge. What, though, would be the paradigm examples in this case? I think they would be some of our most commonly used sources of knowledge, such as perception, memory, reasoning, and testimony. In developing a theory of knowledge, then, what we should do is give an account of how these sources manage to do what they do—how vision and memory work, how people go about inferring some beliefs from others, and how people manage the business of sorting out good testimony from bad. Equally importantly, we should investigate the methodologies used in science and mathematics, which are more historically recent extensions of the other sources. If these sources are the paradigm examples of knowledge, which serve as the true “common core” that all theories of knowledge (perhaps unwittingly) aim to capture, we would do better to determine the nature of knowledge by inquiring into them rather than by considering thought experiments which by themselves have failed to settle the issue. It could be objected that this account is circular. Can we identify “sources of knowledge” as sources of knowledge independently of some prior account of what knowledge is? As it turns out, I think we can. Consider what would happen if you were to ask an ordinary person (i.e., a non-philosopher) how they know various things. “How do you know that it’s raining outside?”–“I can see that it is.” “How do you know what you had for breakfast?” —“I remember eating corn flakes.” We can easily imagine similar exchanges. If you asked them what knowledge is directly, I predict that they would be unable to say much of anything informative. (Of course, I might be wrong. If the reader suspects that I am, they can try asking a real person and see what happens.) Thus, I think the ordinary person would simply appeal to these sources, without giving or even having in mind any definition or description of what knowledge itself is.
In sum, I have argued that the two questions with which we began can be answered by related sets of considerations. The difficulty of finding a solution to Gettier cases led us to consider what makes different theories of knowledge genuine rivals. Our answer was that there must be a common element, but it could only consist in paradigm examples of knowledge—such commonly acknowledged sources as perception, memory, reasoning, and testimony. This gave us an answer to the first question, which was that theories of knowledge should try to settle what knowledge is by analyzing these sources, rather than thought experiments—of which Gettier cases are a prime example. Ironically, the very sort of thought experiment which motivated our quest for an account of a proper theory of knowledge has led us to the conclusion that we have little need of such cases to further pursue our project. We have, so to speak, thrown away our ladder after having climbed up on it.
Gettier, Edmund. “Is Justified True Belief Knowledge?”, Analysis, 23: 121–123, 1963 [Available online here].
What do you think of this article? Discuss it on Scholardarity’s message board.
Books You Might Like
PARENT PAGE: Philosophy