Arguing About Knowledge

Duncan Pritchard, Ram Neta, Duncan Pritchard, Ram Neta

• Publisher: Routledge
• Year: 2020
• ISBN:
  9781000154986
• Edition: 1
• Topic: Philosophie
• Subtopic: Geschichte & Theorie der Philosophie

What is knowledge? What are the sources of knowledge? What is the value of knowledge? What can we know? Arguing About Knowledge offers a fresh and engaging perspective on the theory of knowledge. This comprehensive and imaginative selection of readings examines the subject in an unorthodox and entertaining manner whilst covering the fundamentals of the theory of knowledge. It includes classic and contemporary pieces from the most influential philosophers from Descartes, Russell, Quine and G.E. Moore to Richard Feldman, Edward Craig, Gilbert Harman and Roderick Chisholm. In addition, students will find fascinating alternative pieces from literary and popular work such as Lewis Caroll, Jorges Luis Borges and Paul Boghossian. Each article selected is clear, interesting and free from unnecessary jargon. The editors provide lucid introductions to each section in which they give an overview of the debate and outline the arguments of the papers. Arguing About Knowledge is an inventive and stimulating reader for students new to the theory of knowledge.

Part One
What is knowledge?

INTRODUCTION TO PART ONE

Ram Neta andDuncan Pritchard

CLEARLY, THE ISSUE OF WHAT KNOWLEDGE IS — i.e. how best to define this notion — is central to the theory of knowledge (epistemology). Until relatively recently, it was generally thought that knowledge could be 'decomposed' into three conditions (this is why the classical account of knowledge was known as the 'tripartite' account of knowledge).

The first condition is the truth conditionand it demands that if you know a proposition then that proposition must be true. This condition captures the key element of knowledge that knowledge involves being right.Of course, one might think that one knows a proposition and yet that proposition turns out to be false — i.e. one turned out to be wrong after all — but this just shows that one didn't reallyknow that proposition at all.

The second condition is the belief conditionand it demands that if you know a proposition then you believe that proposition. This condition captures a second key element of knowledge, which is that it is you who is getting things right — it is, after all, you who is believing a true proposition.

These first two conditions are relatively straightforward, and most epistemologists even today agree that knowledge has these conditions. By themselves, however, they are insufficient for knowledge. After all, one could come to believe a true proposition purely by luck. Imagine, for example, that you form a belief that you will win the lottery in this week's draw purely on the basis of wishful thinking. Suppose, further, that this belief turns out to be true, such that you do indeed win the lottery. Did you know thatyou were going to win? Surely not, since it was just a matter of luck that your belief was true. In contrast, suppose you form the belief that you are going to win the lottery in this week's draw on the basis that you have fixed the lottery to ensure that you win. In such a case you surely did know that you were going to win. What, then, is it that marks the difference between the two cases?

According to the tripartite account of knowledge, the difference is that in the second case, but not the first, your true belief is justified.There are various different accounts of what such justification might involve, but the general idea is that when one is justified in believing a proposition this is because one is able to offer good grounds in favour of what one believes. This, then, is the third condition — known as the justification condition— that the tripartite account lays down for knowledge.

We include an example of a tripartite account of knowledge in the readings for this section, due to A. J. Ayer (1910-89) (Chapter 1), to give you a flavour of the view. Rather than talk of belief in the target proposition, he talks of 'being sure' that the proposition is true, and rather than talk of having a justification for one's belief in the target proposition, he talks of 'having a right to be sure' that the proposition is true. The basic point is the same, however, in that what is required for knowledge on this view is that one has a true belief and that one is in addition able to offer adequate grounds in favour of what one believes (i.e. what one is sure of).

By incorporating the justification condition, the tripartite account of knowledge is able to deal with the cases just mentioned. In the first case, for example — where one's belief is based on wishful thinking — one clearly lacks good reason for what one believes, since one cannot acquire such reasons in this way. One thus lacks justification and hence it is no surprise on this view that one also lacks knowledge. In contrast, in the second case one clearly does have good grounds in favour of one's belief — one knows that one has fixed the lottery draw in one's favour after all — and hence one's belief is justified and thus, since it is also true, it is a case of knowledge. There is therefore good reason to suppose that knowledge just is justified true belief, just as the tripartite account of knowledge predicts.

In an extremely short and seminal article, however — which is reprinted in this section — Edmund Gettier (1927-) (Chapter 2) conclusively demonstrated that the tripartite account of knowledge, despite its strong prima facieplausibility, was completely unsustainable, at least in its current form. What Gettier noticed was that it is part and parcel of the tripartite account of justification to allow that one's belief can be justified even if one's belief is false. This is entirely in accordance with intuition, since one might well have good reason to believe a proposition even if that proposition is false (as when one is deceived by a temporary trick of the light, for example). Given this feature of justification, however, Gettier realised that it was relatively easy to construct counterexamples to the tripartite account of knowledge.

These examples have a general two-part structure. First, you take a belief which, while justified in the relevant sense (i.e. supported by good reasons), would ordinarily be such that it would, in those circumstances, have been false. Second, you add a twist to the case such that the belief is true anyway, though for reasons that are completely unconnected with the agent's justification.

Consider the following example that Gettier offers. Suppose that our hero, Smith, has applied for a job but has good grounds for believing that another candidate, Jones, will get the job (perhaps the head of the appointing panel intimated this to him). Since Smith also has good grounds for supposing that Jones has ten coins in his pocket, he infers that the person who will get the job has ten coins in his pocket. Given how Smith has arrived at this belief, it is certainly justified. As it happens, it is also true, in that the person who gets the job does indeed have ten coins in his pocket. Crucially, however, it is not Jones who gets the job but Smith, and the reason why the belief that Smith inferred is nonetheless true is that, unbeknownst to him, it so happens that Smith has ten coins in his pocket. Smith thus has a justified true belief. Nevertheless, it is surely not knowledge, since it is merely a matter of luck that his belief is true in this case (after all, he has no inkling that he has ten coins in his pocket, still less that he will get the job). The tripartite account of knowledge is thus in serious trouble.

As this case illustrates, the examples that Gettier offers are quite complicated. They needn't be, though. Here is a different case to the ones that Gettier himself offers in his article, but which illustrates the same point.¹ Suppose that our protagonist comes down one morning and forms her belief about what the time is by looking at what the reliable grandfather clock in the hall says. Suppose further that our hero has excellent grounds in support of this belief. For example, she knows first-hand that this clock is very reliable at telling the time, and that the time is roughly as the clock says that it is. Her belief is thus justified. Crucially, however, she is in fact forming her belief in a way that would have ordinarily led to a false belief, because the clock is broken. However, as it so happens, our hero is looking at the clock at the one time in the day when it is 'telling' the right time. Clearly, however, one cannot come to know what the time is by looking at a stopped clock. Our agent thus has a justified true belief which is not knowledge, and hence the tripartite account of knowledge is unsustainable.

Various sorts of responses were offered to Gettier's counterexamples to the tripartite account of knowledge in the ensuing literature. Many of them proceeded by trying to find an additional clause that one could add to the tripartite account in order to avoid the problem. A good example of this tendency can be found in the article by Keith Lehrer (1936-) reprinted in this section (Chapter 3). What Lehrer adds to the tripartite account is, very roughly, a clause to the effect that the agent in question should not form his belief on the basis of any false proposition.² For example, in the 'Smith and Jones' example given above, one could argue that the problem lies in the fact that Smith is basing his belief that the person who will get the job has ten coins in his pocket on the false proposition that Jones will get the job. Notice, however, that it isn't so clear how a strategy like this will work in Gettier-style cases where there is no obvious inference, such as the 'stopped clock' case.

Assuming that Lehrer's proposal does not work, where do we go from here? What is interesting about Lehrer's proposal is that, as noted above, it works on the assumption that the tripartite account of knowledge is at least roughly correct. A more radical way of responding to the problem that Gettier raised, however, would be to conclude that the tripartite account of knowledge was incorrect in a more fundamental way.

For example, some have argued that we should drop the justification condition altogether and opt for a different sort of epistemic condition. For instance, one kind of proposal contends that all there is to knowledge is a true belief that is formed in a reliableway — i.e. formed in such a way that it regularly leads you to the truth. Notice that what is different about this approach when compared with the tripartite account is that one's true belief could be reliably formed in the relevant sense — and hence amount to knowledge — even though one is unable to offer any grounds in favour of that belief. The move towards a reliability condition on knowledge rather than a justification condition is thus a radical move indeed.

The problem facing so-called 'reliabilist' accounts of knowledge is to give a compelling specification of what such reliability consists in. It is in this regard that Robert Nozick's (1938-2002) account of knowledge — extracts from which are reprinted in Chapter 4 — is so significant. Nozick argues that the sense of reliability that is in question is one which ensures that our belief appropriately 'tracks' the truth. That is, what we want when an agent knows is for that agent not just to have a true belief, but also to have a belief which tracks the truth in the sense that had what the agent believed been false (or true, but in a slightly different way), then her belief would have altered accordingly.

An example will help illustrate this. Consider again the Gettier-style case considered above involving the stopped clock. That the agent in question happens to form a true belief obviously doesn't suffice for knowledge, as we noted above, but what in addition is required? Nozick's account of knowledge offers a neat explanation, for in this case had what the agent believed been false — had, for example, the time been slightly different but everything else remained the same — then she would have formed a false belief (because the time on the stopped clock would remain the same). The problem with her belief is thus that although it happens to be true, it is not sensitiveto the truth (i.e. it doesn't track it).

This is not to suggest that Nozick has resolved the problem of how to define knowledge, since objections have since emerged even to this account. The significance of the view, however (and the reason why it is collected here), is that it offers a clear statement of one influential position which is radically different from the sort of proposals that were common before, and in the immediate aftermath of, Gettier's famous paper. It thus should offer you a sense of the sort of theories of knowledge that are currently 'live' in the contemporary literature.

Study questions

1. Why does knowledge of a proposition entail having a true belief in that proposition? Why isn't this enough for knowledge?
2. What is the general structure of Gettier-style cases? Construct a Gettier-style example of your own.
3. Do Gettier-style examples essentially involve the agent basing her belief on a false assumption? Try to formulate a Gettier-style example of your own where the agent is not basing her belief on a false assumption.
4. What does Nozick mean when he says that in order for a belief to qualify as knowledge it must 'track' the truth? Explain how Nozick's proposal is meant to deal with the Gettier-style cases.
5. Is justification necessary for knowledge, do you think?

Further reading

A good place to start when it comes to learning more about this topic is Pritchard (2006: chapters 1-4), which offers an accessible introduction to some of the central issues that are relevant here. See also Steup (2005) for an excellent and readable overview of the contemporary literature on the different definitions of knowledge. For an in-depth study of the various responses to Gettier-style problems that have been offered in the literature, see Shope (1983). For a more accessible overview of the problem posed by Gettier cases and its impact on the contemporary epistemological literature, see Hetherington (2005). All of these sources contain lists of useful further readings.

Notes

1 This example is actually due to Russell (19...