Prof. Norman Swartz, Copyright © 1997.
October 10, 1997
Department of Philosophy
Simon Fraser University


Notes on Knowledge
Part One




An overly strong analysis

Some authors have offered the following two conditions as being necessary for "x knows that P":
    x knows that P, if and only if
    1. P is true.

    2. x is able to prove the truth of P to other persons.
Such an analysis will not do. It is idiosyncratic and mistaken.

Here are two counter-examples, cases of x's knowing that P, but where x is unable to prove to anyone else that P is true:
    1. Alice is two years old. She knows, but is quite unable to prove to anyone, that a dog has run off with her favorite Teddy Bear.

    2. Butch has witnessed a robbery and knows who the robber is. But Butch is unable to prove what he claims to know. For Butch has a long history of lying and is known to be an unreliable witness. The police will not take him seriously any longer.


The standard analysis of knowledge

On the standard account of knowledge, due to Plato (427-347 B.C.), there are (at least) three necessary conditions:
    x knows that P, only if

    1. P is true.

    2. x believes (on the basis of e) that P.

    3. x has good evidence e (or reason) to believe that P.
A number of comments and elucidations are in order.
  • Question: Why add the parenthetical qualification to condition #2 above?

    Answer: Consider the case of Sherlock Holmes and Watson. Both Holmes and Watson believed that what's-'is-name committed the crime, both had the same evidence, yet Holmes knew and Watson did not. What accounts for the difference? Simply this: Holmes, but not Watson, recognized the evidence as pointing to what's-'is-name. Watson, unlike Holmes, didn't see the evidence as evidence.

  • Objection: "But surely one can know all sorts of false things. For example, I know that it is false that Russia and Germany were allies in 1944 in the Second World War."

    Reply: The claim
    that it is false that Russia and Germany were allies in 1944 in the Second World War
    is true. To say of a false claim (proposition) that it is false, is to assert something true. No one knows (that it is true) that Russia and Germany were allies in 1944 in the Second World War.

  • Objection: Shouldn't the 1st condition read "x knows that P is true"?

    Reply: The analysis would then become circular. We do not want to use "know" in any of the necessary conditions for knowledge. Consider this analogy (parallel case):

      y is a square if and only if

      1. y is closed

      2. y has four, straight sides

      3. y has equal sides

      4. y can be divided into four smaller squares

    Clearly, we do not offer the foregoing as an analysis of "y is a square".


Under what conditions is one justified in saying "I know that P"?

The conditions for being justified in claiming to know are weaker than the conditions for knowing. x is justified in claiming to know P (i.e. it is proper, or appropriate to say "I know that P") when:

  1. x believes (on the basis of e) that P.

  2. x has good evidence e (or reason) to believe that P.
I.e. when the 'psychological' and the evidential conditions for knowledge are satisfied.

Objection: "But the foregoing two conditions are too weak. Those two conditions could be satisfied, but the person might not (actually) know: the claim might – for all that (i.e. in spite of the evidence) – be false."

Reply: Of course one might make a mistaken claim. There are few, if any, guarantees in life. We do the best we can. Sometimes we claim to know something and, for whatever reason, it turns out that we do not. But shall we refrain from ever claiming to know, simply because on some (few) occasions we 'get it wrong'? We can 'slip up' in almost anything we do or try to do. Why should we impose restrictions on making claims to know that are so much higher than on other things we do?

Sometimes a person will claim to know something and it turns out that she does not. So what? We all occasionally make mistakes (i.e. 'get it wrong'). It's no 'big deal'. The possibility of occasional error should not deter us from making knowledge claims. Yes, I might (in some exaggerated sense of "might") be wrong about there being a hand on the end of my right arm, but I am justified in claiming to know that there is a hand on the end of my right arm.



Skepticism (also spelled "scepticism") is the philosophical theory that knowledge, either in general or of a certain class of propositions, is impossible.

     General skepticism

General (or unrestricted) skepticism is the philosophical theory that no knowledge whatever is possible.

Persons who subscribe to general skepticism must be careful in stating their position. They will fall into a self-refuting position if they say that they know that there is no knowledge. In order to avoid self-contradiction, they will have to claim no more than that they strongly believe that there is no knowledge (and that, perhaps, they also have good reasons for that belief).

     Restricted skepticisms

Note the plural noun "skepticisms". There are several different kinds of restricted skepticism. A restricted skepticism is a philosophical theory to the effect that knowledge is impossible, not in general, but in certain limited areas of concern/inquiry. A person who subscribes to a restricted (or limited) skepticism might, for example, assert that knowledge is impossible concerning (one or more of) the following areas:
  • factual matters (but is possible in mathematics and logic)

  • the future

  • universal propositions (e.g. "all human beings are genetically programmed to age")

  • unobserved entities (e.g. tables and chairs in a room where there are no living things)

  • the theoretical entities of science (subatomic particles, information, electrical charge, etc.)

  • other minds. (Skeptics who argue that there are no other minds - that they alone constitute the universe; that all [seeming] 'others' are nothing more than the creations of their own minds - are known as 'Solipsists'.)

  • the material world. (All that exists are minds, and all [seeming] material objects – e.g. chairs, tables, cars, clothes, even human bodies – are nothing more than bundles of sensations 'in a mind'. The principal exponent of this theory was Bishop George Berkeley [pronounced "Barkly"] (1685-1753) for whom Berkeley [pronounced "Berkly"] University in California is named.)

  • whether we are dreaming.

  • whether we are being completely deceived. (Descartes' version involved a malign supernatural intelligence; the more contemporary version would have us being nothing more than brains in a vat being fed electrical signals by a mad scientist.)

  • the time of the creation of the Universe. ("The universe was created at precisely 5:07 PDT this very morning, with everything - yellowing newspapers, newsreels, tree rings, fossils in the ground, mildew on rooftops, etc., etc. – just as it would have been had the universe actually been much older and had evolved through time to its present state.")

  • the consciousness of other species. ("What is it like to be a bat?" – Thomas Nagel)


Case Study: Am I dreaming?

Two different questions (which a number of persons, including some philosophers, have failed to distinguish from one another) are:
  1. How can I now know whether I am dreaming?

  2. How can I know whether I am dreaming now?
The answer to the first question is:
    There are no really good tests for determining RIGHT NOW whether you are dreaming. But, it is also terribly – indeed, excessively – demanding to want, or expect, immediate satisfaction of your looking for knowledge.

    There are very few items of knowledge that it is possible to establish 'on the spot' so to speak. For example, if the question were, "How can I know now whether I have diabetes?", the answer (for most persons) is: "You can't".
The answer to the second question is:
    Wait awhile. If you wake up, then you are dreaming now and you will come to know – in a hour or two (maybe a bit longer, maybe a bit shorter) – that you were dreaming now. If, however, you don't wake up, then (either you have died while waiting or) you were not asleep now.

    In general, to know what is happening now, often (perhaps usually), involves our doing something in the future.

    If you want to know whether you have diabetes now, i.e. today, you'll need to have certain blood tests, and you probably won't get the results for another day or two after that. But if those results are positive, then you will know at that time that you had diabetes now (i.e. today, three days earlier than when the results are given to you).
Objection: You've admitted that the test – waiting to see whether you wake up or not – might not work: you might die while waiting to see whether you wake up or not.

Reply: The possibility of our dying can (unfortunately) never be eliminated. But take a parallel case. A friend asks you who won the World Soccer Championship in 1951. Neither of you has the slightest idea, but you know how to go about finding the information. You tell your friend to ask for help from the Reference Librarian.

Now of course, and perhaps unfortunately, your friend might die en route to the Library. Even so, the technique you commended was exactly right, and works almost all of the time. Why should anyone think that a proper answer to such a question would have to be a technique that could never fail? Only if one is beguiled by Descartes' insistence on absolute certainty.

All of us have known, since childhood, how to tell dreams from reality: we wake up from dreams. The answer is very simple. It is only a piece of bad philosophizing that makes that correct answer seem suspicious. What ought to be suspicious is the claim that we can't tell dreams from reality. Of course we can; but not always – or even usually – 'on the spot', i.e. immediately. But the same is true for having diabetes, and no one is thus persuaded that we can't tell when a person has diabetes. Of course we can; it just takes a bit of time and a blood test.

Continue -  Part Two

Return/transfer to Norman Swartz's Philosophical Notes

Return/transfer to Norman Swartz's Home Page