The Philosophy Resource

Teaching resources for philosophy students and teachers

Gettier cases

Recall the tripartite analysis of knowledge, according to which a subject S knows that p if:

S knows that p if:

  1. S believes that p.
  2. P is true.
  3. S is justified in believing that p.

As we mentioned, epistemologists are looking for a definition of knowledge that is necessary and sufficient. Conditions (1), (2) & (3) are individually necessary for knowledge; are they jointly sufficient? According to Edmund Gettier, they are not. In his well known 1963 paper ‘Is Justified True Belief Knowledge?’ Gettier constructed a series of counterexamples to the tripartite analysis. That is, cases where intuitively the subjects do not seem to know that p but have justified, true beliefs that p. Any case that functions as a counterexample to the tripartite analysis is called a Gettier-case, but it is unlikely that Gettier was the first to construct such counterexamples. They are also found in the work of Bertrand Russell and Alexius Meinong decades earlier and arguably in the work of classical Indian philosopher Sriharsa as far back as the 1100s.

Gettier Cases: Counterexamples to the view that knowledge is justified, true belief.

Click on the tabs below to view some of these cases. Cases (1) & (2) are Gettier’s own cases, case (3) comes from Bertrand Russell.

Smith and his mate Jones both apply for the same job. Smith has the following two beliefs about Jones:

(A) Jones will get the job, and;
(B) Jones has ten coins in his pocket.

From (A) & (B) Smith justifiably infers, and so is justified in believing, that (C) the man who will get the job has ten coins in his pocket. But it is actually Smith who gets the job. That is, (A) is false. But by luck it just so happens that, without his realising, Smith also happens to have ten coins in his pocket. So (C) is true. And Smith is justified in believing (C). But, intuitively, it seems to us that Smith does not know (C).

So the case shows that one can have a justified, true belief that is not a case of knowledge. So the three conditions stipulated by the tripartite analysis are not jointly sufficient for knowledge.

For a “dramatisation” of this case watch this YouTube video

Suppose that Smith believes the following false proposition (D) that:

(D) Jones own a Ford.
- From this false belief (D), Smith justifiably infers, and so is justified in believing:
(E) that either Jones owns a Ford or Brown is in Barcelona;
(F) that either Jones owns a Ford or Brown is in Boston;
(G) that either Jones owns a Ford or Brown is in Brest-Litovsk.

Smith is justified in believing (E), (F), and (G) because they are what are called ‘disjunctions’. Disjunctions, roughly, are sentences that contain the word ‘or’. All that is needed for a disjunction to be true is that one of the disjuncts (the items either side of the word ‘or’) is true. Since Smith believes that Jones owns a Ford, it is reasonable (i.e. justified) for him to infer that (E), (F) and (G) are also true.

But imagine that by luck Jones happens to be in Barcelona. So it turns out that Smith has a justified, true belief (E). But, intuitively, it seems that Smith does not know (E), despite his justified, true belief.

My favourite example of a ‘Gettier’ case comes from Bertrand Russell. It goes something like this:

Suppose I need to know the time (I am about to go to an appointment with my doctor) and I look at the clock on my mantelpiece. It tells me it’s 12.15. I form the justified belief that it’s 12.15 and I go to see my doctor on time. My belief that it was 12.15 was true. However, when I return home, I notice that the clock on my mantelpiece says it is 12.15. My clock must have stopped! But how lucky it was that it stopped at exactly 12.15 the time it was when I happened to look!

My belief that it was 12.15 was true, and justified, but we don’t seem to be inclined to say it was knowledge.

A diagnosis

Notice that I have written ‘lucky’ in bold in each of the three scenarios. What seems to be the problem with the tripartite analysis is that it allows for lucky knowledge. But, as we saw in Step 1, Knowledge is incompatible with luck. The whole point of introducing the justification condition was to try to accommodate that thought, to rule-out lucky knowledge. What makes Gettier cases so problematic is that they show that the justification condition was not up to the task assigned to it.

Go to…

Step One

Go to…

Step Two

Go to…

Step Three

Go to…

Step Four

Go to…

Step Five