Bertrand Russell, the famous British philosopher, gives some cogent criticisms of William James’s pragmatic theory of truth that amplify the critique given by W. Corduan in Reasonable Faith.[1] Yet before we address them, we should note that there is some debate as to just what James’s theory involved.

Some, like Mortimer Adler,[2] argue that James held to a correspondence view of truth, but thought that truth was verified pragmatically by its effects. Others, such as Corduan and Russell, argue that James made the pragmatic effects the very meaning or essence of truth. This charge seems most obvious if it can be shown that James argues that opposing religious doctrines can be true if they work for their adherents.[3] For instance, consider two sets of religious truth-claims:

Three Theravada Buddhist truth-claims

(1.0) Buddha alone is the supreme revelation of spiritual reality.
(1.1) Buddha teaches that belief in God or gods is irrelevant to spiritual liberation and probably harmful.
(1.2) Buddha’s teaching should be followed by all.

Three Orthodox Christian truth-claims:

(2.0) Christ alone is the supreme revelation of spiritual reality.
(2.1) Christ taught that belief in and commitment to God is essential for spiritual liberation.
(2.2) Christ’s teaching should be followed by all.

Given the correspondence view of truth, (1.0)-(1.2) are logically incompatible with (2.0)-(2.2). They cannot both be true. To say that propositions (1.0)-(1.2) and (2.0)-(2.2) are both true denies the correspondence view of truth, since both Buddha and Christ cannot be the uniquely supreme revelation of spiritual reality, given their disparate teachings and practices. Furthermore, pragmatically, one cannot follow both Buddha’s and Christ’s teaching on the salvific worth of the belief in God, since they contradict each other.

Russell says that James’s pragmatic view of truth requires that a belief is deemed true when its effect are good or when it 'works.' As Russell puts it, according to James, 'Truth happens to an idea; it is made true by events.'[4] Russell says that if this idea is to be useful (which is only fitting given the pragmatist’s view of truth) one must know two things before we know if a belief is true:

(3) What is good
(4) What the effects of this or that belief must be

We must know these things before we have determined whether a belief is true 'since it is only after we have decided that the effects of a belief are good that we have a right to call it "true."'[5]

Yet this is deeply problematic. One must measure beliefs by usefulness (or utility); yet in many cases we just don’t know the what the usefulness of a belief will be ahead of time. Russell gives the example of believing Columbus came to the new world in 1492. We can’t just look this up in a book; we have to determine its effects on us (or, as James puts it, it’s 'cash value'). But how can we know this ahead of time?

Added to this is the problem of knowing just what effects beliefs have produced after the fact. Beliefs do have consequences, no doubt; yet determining just what the causal connections are and whether or not they are beneficial may be difficult. Russell gives an example: 'It is far easier, it seems to me to settle the plain question of fact: 'Have Popes always been infallible?' than to settle the question whether the effects of thinking them infallible are on the whole good.'[6]

But what does it mean for an idea to 'work'? Russell and Arthur Lovejoy pointed out that James confuses two senses of an idea working. One view is that a theory is verified if it makes true predictions about reality. If I theorize that the 1994 baseball strike will be resolved by the end of 1994 and the players and managers reach a workable agreement by this time, the theory is verified. Yet this assumes a correspondence view of truth wherein a predictive proposition (“the strike will be settled by the end of 1994”) is verified by the actual occurrence of the thing predicted. But 'work' may simply mean that the individual effects of believing that the strike will be settled 'works' in the sense that the belief is deemed valuable to the individual. Certainly, optimism lifts the soul; however, optimistic thoughts are either true or false, no matter how happy they may make one.

Russell also notes that one’s estimation of the consequences of believing something (which is what makes the belief true) must also be true. Yet if the estimation itself is to be deemed true depends on whether it 'works,' ad infinitum. This is the problem of the infinite epistemic regress. One never lands on the truth because it always stands in need of being tested to see if it 'works.' If one holds to the correspondence view of truth as a necessary element of what truth means, one escapes this problem. A statement is true when it correspondence with or agrees with objective reality—whatever the consequences of that belief may be.

Russell also disputes James’s contention that the meaning of truth simply means its ability to produce desirable states of affairs. This ignores the common meaning of truth. Consider the difference between two statements:

(5) It is true that other people exist.
(6) It is useful to believe that other people exist.

If James’s view of truth—as Russell understands it—was correct, (5) and (6) would have the same meaning; that is, they would express the same proposition; the meanings of (5) and (6) would be just as exchangeable as the meanings of the following two propositions:

(7) Bill Clinton is pro-choice.
(8) Bill Clinton favors abortion on demand.

Statements (7) and (8) assert the same proposition, whereas (5) and (6) clearly do not. Therefore, the meaning of truth cannot be a belief’s usefulness, even though some beliefs are more useful than others (given a particular view of what constitutes usefulness).[7]

Furthermore, a belief may 'work' and not be true. Why is it true that Columbus existed? It is true because a certain person named Columbus lived in the fifteen century. This is the cause of my belief being true; the effects of this belief are irrelevant to its truth value. Russell notes that for James 'it might happen that "A exists" is true although in fact A does not exist.'[8] The belief in Santa Claus works quite well for millions of children—for a time, until they realize that the 'Santa Claus hypothesis' is not a plausible way to explain the arrival of their presents, etc. Yet James says, 'If the hypothesis of God works satisfactorily in the widest sense of the word work, it is true.' Russell’s comments, though he was an atheist, are on target:

This simply omits as unimportant the question whether God really is in His heaven; if He is a useful hypothesis, that is enough. God the Architect of the Cosmos is forgotten; all that is remembered is belief in God, and it effects upon the creatures inhabiting our petty planet. No wonder the Pope condemned the pragmatic defense of religion.[9]

Lastly, one can think of any number of useless truths, truths than have no pragmatic worth:

(9) The number of breaths I took on May 3, 1986.
(10) The number of hairs on Jerry Lewis’s head at noon on June 3, 1976.

Yet if truth is defined on the basis of its producing desirable results (its 'working'), statements (9)-(10) could not be considered to be true. Since we know that there are some true propositions (whether we know them to be true or not) corresponding to (9)-(10), the pragmatic view must be false. The pragmatist might respond that we can conceive of some possible use in believing (9)-(10), but this appears to be a desperate, ad hoc move. Surely, there are at least some truths that are useless in the pragmatic sense. Any possible use in believing (9)-(10) is incidental to their truth; it would not define them as true.[10]

For Further Reading

William James, Pragmatism and Four Essays from The Meaning of Truth (Cleveland, OH: Meridian Books, 1944).

William James, The Will To Believe and Other Essays in Popular Philosophy (New York: Dover Publications, Inc., 1956).

Bertrand Russell, A History of Western Philosophy (Forage Village, MA: Simon and Schuster, 1945).

Frederick F. Schmitt, Truth: A Primer (Boulder, CO: Westview Press, 1995), chapter 3.

Entries on “William James,” “Pragmatism,” and “Pragmatic Theory of Truth” in The Encyclopedia of Philosophy, 8 volumes (1967), edited by Paul Edwards.

Entries on “William James,” “Pragamatism,” and “Truth” in The Cambridge Dictionary of Philosophy, edited by Robert Audi (Cambridge, 1995).


[1] We should remember that there are several other pragmatic views of truth as well—such as given by John Dewey, C.S. Pierce, and modern pragmatists such as Richard Rorty—that are not identical to James’s view of truth, and so may not be subject to all the same criticisms.

[2] Mortimer Adler, Truth in Religion: The Plurality of Religions and the Unity of Truth (New York: MacMillan, 1990), 22-23.

[3] See Winfried Corduan, No Doubt About It (Nashville: Broadman and Holman, 1997), p.60.

[4] Bertrand Russell, A History of Western Philosophy (Forage Village, MA: Simon and Schuster, 1945), p.816

[5] Ibid., p.817.

[6] Bertrand Russell, Philosophical Essays (London, 1910), p.135; quoted in Gertrude Ezorskey, “Pragmatic Theory of Truth,” in Paul Edwards, ed., The Encyclopedia of Philosophy (1967) 6:428. Ezorskey’s article takes the view that James’s did not hold to a correspondence view of truth, but viewed expedience as the meaning of truth. On this see Gordon Clark, Thales to Dewey: A History of Philosophy (Jefferson, MD: The Trinity Foundation, 1989), pp.500-508, who likewise thinks that James abandoned the correspondence view.

[7] Russell, Philosophical Essays, Ibid.

[8] Russell, A History, pp.817-818.

[9] Ibid., p.818.

[10] See Frederick F. Schmitt, Truth: A Primer (Boulder, CO: Westview Press, 1995), pp.89-90.

All rights reserved. International copyright secured.
This resource is used by the kind permission of the author. It was previously available on the Denver Seminary website.