In his 1981 essay “Knowledge” published in “Philosophical Explanations” Robert Nozick deals with two major problems of epistemology: firstly what is knowledge or what does it mean to say that someone knows a proposition p? Secondly, what can we say to the skeptic who holds the view that we cannot know anything?
It was widely agreed among philosophers that knowledge is justified true belief. This standard analysis was defeated in 1963 by the brief but powerful essay “Is justified true belief knowledge?” by Edmund Gettier. Using thought experiments Gettier developed counterexamples to the standard analysis of knowledge by describing situations in which we would not say that someone knows "that p" although he has a well justified and true belief about p. In the following years philosophers tried to substitute or add to conditions of the standard analysis in order to find a theory which is strong enough to rule out Gettier counterexamples. This is what Nozick tries. After explicating his account of knowledge he shows how it can handle the skeptic’s objections.
In this essay I will first give a description of Nozick’s truth-tracking-theory of knowledge and what this means with regard to philosophical skepticism. After this I evaluate Nozick’s account and show that his theory is not as strong as it looks. My thesis is that Nozick overlooks that knowledge is more than true belief which varies with the truth value of p. Nevertheless, Nozick draws attention to an important connection which is in itself not sufficient for knowledge but which should be attended to by developing a theory of knowledge.
Contents
1 Introduction
2 Robert Nozick’s Theory of Knowledge
2.1 The Conditions of Knowledge
2.2 The “possible world” Account
2.3 Skepticism and Nonclosure
2.4 The Skeptic’s Contradiction
3 Critical Assessment of Nozick’s Theory of Knowledge
3.1 The Cost of Giving Up the Closure Principle
3.2 The Circularity of the “possible world“ Approach
3.3 Truth-tracking is not enough: The Missing Link between Belief and Facts
4 Conclusion
Bibliography
1 Introduction
In his 1981 essay “Knowledge” published in “Philosophical Explanations” Robert Nozick deals with two major problems of epistemology: firstly what is knowledge or what does it mean to say that someone knows a proposition p ? Then secondly, what can we say to the skeptic who holds the view that we cannot know anything?
It was widely agreed among philosophers that knowledge is justified true belief. This standard analysis was defeated in 1963 by the brief but powerful essay “Is justified true belief knowledge?” by Edmund Gettier. Using thought experiments Gettier developed counterexamples to the standard analysis of knowledge by describing situations in which we would not say that someone knows that p although he has a well justified and true belief about p. In the following years philosophers tried to substitute or add to conditions of the standard analysis in order to find a theory which is strong enough to rule out Gettier counterexamples. This is what Nozick tries. After explicating his account of knowledge he shows how it can handle the skeptic’s objections.
In this essay I will first give a description of Nozick’s truth-tracking-theory of knowledge and what this means with regard to philosophical skepticism. After this I evaluate Nozick’s account and show that his theory is not as strong as it looks. My thesis is that Nozick overlooks that knowledge is more than true belief which varies with the truth value of p. Nevertheless, Nozick draws attention to an important connection which is in itself not sufficient for knowledge but which should be attended to by developing a theory of knowledge.
2 Robert Nozick’s Theory of Knowledge
2.1 The Conditions of Knowledge
Nozick takes over the first two conditions of the standard analysis of knowledge and his idea is to find a list of conditions which must be met to accredit knowledge to someone.1 S knows that p if and only if:
1. S believes that p.
2. It is true that p.
We would not say that someone knows something if he does not believe it or if his belief is wrong. Yet what should we add to get a satisfying account of knowledge? Nozick refines his theory by examining the following Gettier example:
“Two other people are in my office and I am justified on the basis of much evidence in believing the first owns a Ford car; though he (now) does not, the second person (a stranger to me) owns one. I believe truly and justifiably that someone (or other) in my office owns a Ford car, but I do not know someone does. Concluded Gettier, knowledge is not simply justified true belief.”2
The justified belief just coincides by good luck with the facts. We would not say that the office man does know that someone in the office owns a Ford car. Nozick points out that in the case of the stranger not coincidentally owning a Ford car the man would still believe that someone in his office owns a Ford car. His believe is not sensitive to the truth, because facts and beliefs do not vary together.3 Nozick now makes the following and for his theory crucial argument: If someone believes something irrespective of whether it is true or not, this belief cannot meaningfully count as knowledge. This initiates Nozick to stipulate a third condition, which helps to rule out Gettier examples:
3. If p weren’t true, S wouldn’t believe that p.
This makes sure that the belief can vary appropriate to the facts, but obviously this subjunctive condition “tells us only half of the story about how [a] belief is sensitive to the truth-value of p”4, because not every problem case can be ruled out with (3). Consider the following example:
“[T]he case of the person in the tank who is brought to believe, by direct electrical and chemical stimulation of his brain, that he is in the tank and is being brought to believe things in this way; he does not know this is true.”5
This person’s belief is still not sensitive to the truth because the manipulators could make him believe that he is in his student residence in Leuven while he is not. We cannot say subjunctively that if he were in a tank he would also believe this. Nozick consequently adds a fourth condition:
4. If p were true, S would believe that p.
This means that S does not only believe p when p is true, but that S would also believe that p (if it were still the case) in slightly different circumstances.
In a nutshell Nozick argues that we can assign knowledge to someone “when he not only does truly believe p, but also would truly believe it and wouldn’t falsely believe it”6. Knowledge is not comprised only of true beliefs, but is also to an extent conditional. If conditions (1) – (4) hold then a person’s belief “tracks the truth that p”7. A belief tracks the truth if and only if it varies with the facts: if the facts are the same the belief would be the same, but if they were different, the belief would be different. Nozick proposes a connection between S’s belief that p and the fact that p.
The concept that a belief has to be not only actually true but also true in slightly different circumstances is not easy to grasp. Nozick uses a “possible worlds”8 account to elucidate this scheme.
2.2 The “possible world” Account
What does it mean if a subjunctive conditional is true and how can we decide this? In order to illustrate this let us take a look at the first subjunctive: “If it were not the case that p, S would not believe that p”. In our actual world p is true. The subjunctive demands us to consider possible worlds in which p is not true but which are otherwise very similar to the actual world. They differ just slightly in their circumstances and are equal in all major respects. The truth value of the subjunctive is determined by ascertaining if S believes not-p throughout all those possible worlds that are close to the actual and in which p is not true. If S does not believe p in these close worlds, which Nozick calls the p-neighborhood, the subjunctive holds true.9 The subjunctive conditional tells us in general what would be, or would have been the case, if p were not true.
More concretely, suppose it is raining and S believes this. Would S still believe it was raining if this was actually not the case? To find an answer we have to think about the possible worlds (PW) that are close to the actual.
PW 1: Equals the actual world except it is not raining.
PW 2: It is not raining but S gets wet because some children spatter S with water guns.
PW 3: It is not raining but instead of walking down the street S is driving in a car.
PW 4: It is not raining but aliens from Alpha Centauri try to fool S by following him with their airplane and simulating rain.
Let us assume that these are all (relevant) possible worlds; normally there is an infinite number of possible worlds. Obviously (PW 4) is distant to the actual world, and hence it plays no role in determining the truth value of the subjunctive in question (at least unless mean aliens do actually deceive S).10 (PW 1) – (PW 3) describe the p-neighborhood; that is all possible worlds which are close to the actual. It seems to be plausible to say that S would not believe that it is raining in all of these and therefore that the subjunctive holds true.
The following scheme explains the idea at a glance for the other subjunctive (the fourth condition): “If p were true, S would believe that p”. In this scheme “S would believe that p” is abbreviated as “q”.
Abbildung in dieser Leseprobe nicht enthalten
It would be a misunderstanding to think Nozick claims that the possible worlds exist in a real or metaphysical way; the account is in fact a metaphor to illustrate what it means that a condition is subjunctive. What does Nozick’s truth-tracking-theory of knowledge now counter to the skeptic who doubts everything?
2.3 Skepticism and Nonclosure
(Cartesian) Skeptics describe different situations in which our beliefs are false. Suppose some scientists from Alpha Centauri kidnapped your brain and put it into a vat. By sending electrochemical impulses to your brain so that your experiences are of the sort you would have if your brain had not been kidnapped they can manipulate your beliefs as they want. The scientists can let you believe everything and you would not notice that these are not your real thoughts. Another example would be the Cartesian evil demon who deceives us by controlling our thoughts and beliefs. Let us call such a situation, as according to Nozick, a skeptical possibility “SK”11. In what a way does SK threaten our knowledge of ordinary facts?
Consider the following example:
Abbildung in dieser Leseprobe nicht enthalten
Nozick’s third condition demands:
3. If p weren’t true, S wouldn’t believe that p.
To have knowledge that I am in my student residence in Leuven I must be truth sensitive: if I were not in Leuven I would not believe that I am in Leuven. Yet this is exactly the situation which SK describes: if you were a brain in a vat on Alpha Centauri it is possible that you would still think that you were in Leuven since the scientists from Alpha Centauri can let you believe whatever they want. Therefore I do not know that I am in Leuven.
Furthermore because I can never know that I am not a brain in a vat the Skeptic argues that I cannot know anything.12 In general they attack the following argument:
Premise 1: I am in my student residence in Leuven.
Premise 2: If I am in Leuven then I am not a brain in a vat on Alpha Centauri.
Abbildung in dieser Leseprobe nicht enthalten
Conclusion: I am not a brain in a vat on Alpha Centauri.
The Skeptic argues backwards. I cannot know that I am not a brain in a vat and hence I cannot know that I am in Leuven because if I would know that, it would entail that I do know that I am not a brain in a vat. It therefore looks like we have to disprove that we are a brain in a vat or that an evil demon deceives us in order to be able to gain knowledge. How can I know that I am not such a brain or that I am not being deceived by an evil demon?
Instead of trying to find an answer to this question Nozick admits that we can never know whether we are in a SK or not because we cannot track the truth on this question.13 The situations are constructed in such a way that it is not possible to disprove them, and we have to accept that. There is no case in which my belief that not-SK can be truth-sensitive. If it were the case that I am in a SK in a close possible world I would still believe that I am not because the mad scientists could and would ensure that my experience is identical to that which it would be in the real world.
However there is no reason to give up the truth-tracking-theory of knowledge. Nozick claims that it is possible not to know whether we are in not-SK situations and simultaneously to have knowledge. Nozick argues that the Skeptic does a “short step”14 by ignoring the subjunctive character of his conditions of knowledge and that he is mistaken in the premise that knowledge is closed under known logical implication.
What does the latter, which we can term the closure principle, mean?
The principle of Closure
Abbildung in dieser Leseprobe nicht enthalten
If S knows that p, and S knows that p entails q, then S knows that q.
Let us consider again the above example: I believe that I am in Leuven. If knowledge would be closed under known logical implication then in condition (3) my not being in Leuven would entail my not believing that I am in Leuven. But this is not what (3) says, because it is a subjunctive condition. It says something about a situation which would occur if p were false, within a close possible world. Normally, in the closest worlds, if I am not at home I will be at university, in a shop or something like that and then I would not believe that I am at home. In these cases I can satisfy the necessary conditions because my belief varies with the facts.
[...]
1 Nozick 1981: 172.
2 Nozick 1981: 173.
3 Nozick 1981: 174.
4 Nozick 1981: 174.
5 Nozick 1981: 174.
6 Nozick 1981: 176.
7 Nozick 1981: 176.
8 Nozick 1981: 173.
9 Nozick 1981: 173 and 178.
10 Nozick 1981: 173.
11 Nozick 1981: 178.
12 Nozick 1981: 179.
13 Nozick 1981: 179.
14 Nozick 1981: 180.