William James presents an account of truth according to which the truth of our thoughts and utterances, even of those that are about the past, is sensitive both to our interests and to how our inquiries go in the future.  This preference-sensitive and future-directed notion of truth has struck many as wildly revisionary, but James' position is actually quite plausible if one understands how his account of truth is intertwined with his account of intentionality. James claimed that his point of contention with his opponents was not over the claim that an idea's truth consisted in its agreement with reality, but rather over the nature of this "agreement" (PR 96), and his views on this subject resulted from his attempt to give an account of the "intentional" character of our thoughts. James' forward-looking account of intentionality (or "knowing") compares favorably the 'causal' and 'resemblance-driven' accounts that have been popular since James' day, and it is only when his remarks about truth are placed in the context of his account of intentionality that they come to seem as plausible as they manifestly did to James. Furthermore, once one understands how James' account of intentionality and truth are connected, one can understand how James is able to allow 'subjective' elements into his account of truth in a way that remains perfectly compatible with commonsense realism.
Truth and Propositions
First of all, to properly understand James' position one must notice a crucial difference between James' usage and that of both his early critics and many contemporary philosophers. Philosophers have tended to follow writers like Frege and Russell in understanding our thoughts and utterances in terms of the propositions they express (propositions which are, in turn, individuated by their truth conditions).  James, on the other hand, often intends his talk of thoughts and utterances to be about non-intentional phenomena (variously, strings of sounds, brain processes or just "flat experience").  It is the question of how these realities come to be true or false that interests James.
According to James, the word "proposition" seemed "expressly invented" to foster the confusion between truth as a property of opinions and truth as a property of the facts which the opinions assert , and he thought that trying to understand our thoughts and utterances in terms of propositions only leads to confusion. As he put it, propositions were "mongrel curs that have no real place between realities on the one hand and the beliefs on the other."  Talk of propositions causes trouble for the pragmatist precisely because it allows one to talk of truth while taking the question of intentionality to be already settled.
Furthermore, thinking of beliefs in terms of propositions is not confined to those who make explicit use of the term "proposition." Even those who do not talk explicitly of propositions often treat beliefs, ideas and utterances in a way that causes similar problems for the pragmatist. If by a belief one means, say, the belief that snow is white, then one is treating beliefs as if they already had an interpretation, and in doing so one makes any pragmatic account of their truth impossible.
When James talks of what makes a believed sentence true he has in mind not what makes an interpreted sentence true, but rather what both gives the sentence its interpretation and makes it true. It is only the answer to this more complex question that requires the cooperation of both the world and our interest-driven investigations. The truth or falsity of a proposition, or an interpreted sentence, (unless it is itself about our investigations or interests) is completely independent of our interests and inquiries. As a result, if one takes James' topic to be the truth of interpreted sentences rather than uninterpreted ones (or about "propositional attitudes" rather than "beliefs"), one cannot but hear James' remarks about truth as implausibly revisionary. Once James' topic is understood properly, it becomes clear that the real debate between him and his "intellectualist" opponents is as much about meaning as it is about truth. If one has an intellectualist conception of meaning (that is, if one takes sentences to have their meaning independently of our interests and actions), then the claim that truth is at all dependent upon our interests will seem obviously false. The disagreement about meaning thus manifests itself in a disagreement about truth, and so James often talked about his dispute with the intellectualist in terms of truth rather than meaning.
His intellectualist opponents, however, usually failed to see this. Even a comparatively sympathetic critic such as Pratt insisted on talking about the intellectualist conception of truth in a way that simply presupposes the intentional relation between a thought and its object. For instance, Pratt writes:
The intellectualist's meaning of truth is so simple, so commonplace, so close at hand, that the pragmatist has overlooked it. By the truth of an idea the intellectualist means merely this simple thing, that the object of which one is thinking is as one thinks it. Is there anything hard about this, anything meaningless, anything 'metaphysical' or abstract? 
By James lights, such an account leaves the 'hard' question about truth completely unaddressed. We can admit that one has truth when the object of which one is thinking is as one thinks it, but the question remains, just what determines which object, if any, one is thinking of?
James' remarks about truth are bound to seem confused as long as one thinks of beliefs and other psychological states in terms of propositional attitudes. While he would never have put it this way himself, his remarks on truth relate to how our merely 'psychological' states can come to be 'propositional attitudes' in the first place.
The Problem of Intentionality
The truth of our thoughts and utterances depends both upon what they are about, and upon how the world is. The pragmatic account of truth gets started with the question how our utterances and ideas come to be about what they are. James explicitly discusses intentionality in terms of how our ideas can "know" realities outside of them , and he claimed in his Principles of Psychology that "the relation of knowing is the most mysterious thing in the world."  James' serious concern with this problem was stressed again in his review of Royce's The Religious Aspect of Philosophy where he presents his feelings about the problem in a particularly vivid way.
The more one thinks, the more one feels that there is a real puzzle here. Turn and twist as we will, we are caught in a tight trap. Although we cannot help believing that our thoughts do mean realities and are true or false of them, we cannot for the life of us ascertain how they can mean them. If thought be one thing and reality another, by what pincers, from out of all the realities, does the thought pick out the special one it intends to know? (CER 386)
Indeed, many of his discussions of what is involved in an idea's being 'true' are just as much about intentionality as about truth.  When most contemporary philosophers talk of truth, they are talking about a property of assertions, judgments or propositions, not something that can be attributed to words or images. James, on the other hand, talks of truth, not just a property of assertions and judgments, but also of mental images and even names.  In particular, a name is 'true' if it guides us to and helps us cope with the object that it is 'true of.' When James speaks of what makes an idea true, then, he frequently has in mind what relates it to its object.
James was unsympathetic to Royce's idealistic theory of intentionality and he wanted to give a naturalistic account of how our words and ideas came to possess their intentional character. The following passage is representative of this hope.
What that something is in the case of truth psychology tells us: the idea has associations particular to itself, motor as well as ideational; it tends by its place and nature to call these into being, one after another; and the appearance of them in succession is what we mean by the 'workings' of the idea. According to what they are, does the trueness or falseness which the idea harbored come to light. These tendencies have still earlier conditions which, in a general way, biology, psychology and biography can trace. This whole chain of natural causal conditions produces a resultant state of things in which new relations, not simply causal, can now be found, or into which they can now be introduced -- the relations namely which we epistemologists study, relations of adaptation, of substitutability, of instrumentality, of reference and of truth. (MT 96)
The question of how our thoughts come to be about realities outside of them is still with us today, and James' account will here be contrasted with two alternatives, various forms of which have held a central place in Anglo-American philosophy since James' time: resemblance-driven accounts and 'causal' accounts. If one compares the three types of account, one thing quickly becomes clear: while James' account of intentionality underwrites his claims about truth, if one presupposes either of the other two accounts, James remarks about truth are (if intelligible at all) plainly false.
Causal and Resemblance Accounts
Take, for instance, accounts of intentionality which try to explain it in terms of resemblance. These come in roughly two forms: the 'classical' resemblance accounts attributed to the earlier empiricists (where ideas refer to outside realities by resembling those outer realities) , and the 'logical' resemblance accounts attributed to writers such as Frege and Russell (where ideas refer to outside realities by being associated with sets of descriptions uniquely true of those realities) . In spite of their differences, both accounts tie what, if anything, our ideas pick out exclusively to facts about our own present non-desiderative psychological states. As a result, neither our desires nor what we do in the future can have anything to do with what we are thinking about now. Just what, if anything, our thought is about will be settled at the moment we think it, and whether or not it is true will be entirely 'up to the world.' If one presupposes a resemblance account, then, the truth of our thoughts will be independent of future-directed and interest-relative notions such as the 'end of inquiry' or what 'works for us.'
Much the same happens if one presupposes the type of 'causal' accounts of intentionality which explain it in terms of what has been aptly called "the photograph model."  According to such accounts, just as a photograph of James is a photograph of James in virtue of its being at the end of a causal chain connecting it to James himself, our ideas know realities outside of them in virtue their standing at the ends of causal chains with (instances of) those realities as their sources. Since such causal relations are 'backward looking,' they will be settled at the moment of thought or utterance and thus independent of our interests and subsequent inquiry. What our ideas are ideas of will be no more sensitive to our interests and future inquiries than what our photographs are photographs of. Accordingly, the propositions our utterances express will be settled at the time of utterance, and whether or not they are true will be (once again) entirely 'up to the world.' 
On two of the most popular approaches to the problem of intentionality, then, the pragmatist's conceptions of truth will seem obviously false. If what we are talking about is settled at the time of utterance, the pragmatist's claim that truth is sensitive to our interests and inquiries can only be understood as a rejection of commonsense realism. On the other hand, the suggestion that a thought is true if it is 'expedient in the long run' or that its truth can understood in terms of some idealized limit of inquiry follows quite naturally from James' account of intentionality, and in spite of its radically different starting point, James' account of intentionality can ultimately accommodate the intuitions which may initially seem to go against it. This should become clear once his account is presented.
James' Account 
The cornerstone of James' account of intentionality is the picture of the relation between perception and its objects associated with his radical empiricism (in particular, what he refers to as his "doctrine of pure experience"). As James puts it, to perceive an object, or to know it immediately, "is for mental content and object to be identical."  In perception "[t]he external and the internal, the extended and the not extended fuse and make an indissoluble marriage" (ERE 265). Whether James' radical empiricism amounted to a type of phenomenalism or a type of 'direct realism' according to which there were no intermediaries (representations, sense data, etc.) between perceptions and the objects perceived will not be discussed here, since whichever way one interprets him, perceptual reference becomes relatively unproblematic on his account. Even if one is not fond of the details of James' radical empiricism, the basic idea that our perceptual contact with the world has a crucial role in determining how our ideas come to be about the world is extremely plausible.
James extends his account of perception-based reference by arguing that one's ideas can know objects outside of one's perceptual field by being capable of leading one through a series of experiences that would terminate in an actual percept of the object referred to.  For instance, James considers his dog-idea to be "cognitive" of his real dog because "the actual tissue of experience is constituted [so that] the idea is capable of leading into a chain of other experiences . . . which go from next to next and terminate at last in vivid sense-perceptions of a jumping, barking, hairy body" (MT 74, ERE 101).
It should be stressed that, on James' account, our being led to the referent does not merely indicate what we had always been thinking about; rather, the leading relation is supposed to be constitutive of the intentional one. The objects our practices ultimately end up 'operating' on are the ones our ideas refer to, and "knowledge of sensible realities . . . is made . . . by relations that unroll themselves in time."  As James puts it:
The percept here not only verifies the concept, proves its function of knowing that percept to be true, but the percept's existence as the terminus of the chain of intermediaries creates the function. Whatever terminates the chain was, because it now proves itself to be, what the concept 'had in mind'. (MT 64, ERE 31, italics James')
On James' account, then, the 'direction of fit' between what we are referring to and our attempts to locate such referents may seem to be the opposite of what is commonly supposed. What we are referring to seems to tailor itself to our investigations rather than the other way around. Unlike contemporary causal accounts where our idea of, say, the Washington Monument refers to the Washington Monument because it is at the end of a causal chain extending back to the Monument itself, James' account suggests that my Washington-Monument-idea is 'cognitive of' the Washington Monument because, if asked to locate its referent, I would eventually track the Monument down.
The Problem of Error
The account of intentionality outlined immediately above seems to face an obvious objection. If what we are thinking about is constituted by the identification process, it seems impossible for us to misidentify the objects of our thoughts. We can occasionally misidentify the referents of our terms, but it is unclear how James' account could allow for such mistakes. Consider, for instance, James' explanation of how his hat-idea is cognitive of his actual hat:
I think of my hat which a while ago I left in the cloak room . . . this idea-hat, will presently determine the direction of my steps. I will go retrieve it. The idea I have of it will last up to the sensible presence of the hat, then will blend harmoniously with it. (ERE 264-5)
James here claims that he is thinking of his actual hat because his actual hat is what his hat-idea would lead him to go pick up. However, if James accidentally picked up Santayana's hat, then he picked up the wrong hat and made a mistake. Just because he picked up Santayana's hat, it should not follow that his hat-idea was of Santayana's hat all along. However, if the leading process actually creates the function, this result may seem unavoidable.
Fortunately, the resources for giving an account of error are present as early as in "The Function of Cognition,"  where James claims not only that a feeling "knows whatever reality it resembles and either directly or indirectly operates upon," but also that "if it resembles without operating, it is a dream; if it operates without resembling, it is an error."  Successful reference requires not only that we act upon an object, but also that it be the one that 'most completely resembles' our idea. The requirement that an idea 'most completely' rather than merely "partially" resemble what it refers to is elaborated in an example much like the Santayana's hat case described above.
Among such errors are those cases in which our feeling operates on a reality which it does partially resemble, and yet does not intend: as for instance, when I take up your umbrella meaning to take my own. I cannot be said here either to know your umbrella, or my own, which latter my feeling more completely resembles. 
James' account thus incorporates both our causal interaction with the world and resemblance into his account of intentionality. What an idea operates on is simply the thing it actually leads us to track down and act upon, but what the idea should operate on and thus what we should track down is tied to what it most completely resembles. If we find a new object that more completely resembles our original idea (and 'blends more harmoniously' with our general set of ideas) we will be willing to treat that new object as what we had been referring to all along and treat the older object that only partially resembled our idea (and only appeared to blend harmoniously) as a misidentification. 
It is this requirement that the object blend harmoniously with our general set of ideas that allows James' future directed account of reference to accommodate the intuitions behind the more backwards-looking causal accounts. One of the important aspects of our ideas that plays a part in resemblance is the belief that we are talking about the hats, dogs, and umbrellas that we were acquainted with in the past. James is thus able to account for the importance of past usage by stressing that agreement with such usage is important to us now and will continue to be so in the future. Past causal history is not taken as a primitive in the explanation of reference, but is itself incorporated into his forward-looking account. 
It may seem, however, that this appeal to resemblance takes the bite out of James' theory. If working upon an object only leads to reference if the object "most completely" resembles the relevant idea, it begins to look as if resemblance is doing all the work, and that James' account is ultimately no more than a resemblance theory. Once we account for error by allowing resemblance to fund a distinction between what we should and what we merely do pick out, it may seem that truth and intentionality aren't really dependent upon our inquiries and interests at all. To think this would, however, be a mistake. While we can make sense of misidentification in terms of disagreements with future identifications that more completely agree with our ideas, it may still be undetermined what future inquiry, even in its most idealized sense, should pick out. There is a significant difference between saying that resemblance determines what we are talking about, and saying that our investigations do, but that such investigations are constrained by resemblance. In particular, if our idea is a 'thin' one that is satisfied equally well by many objects, nothing will be referred to on the resemblance account, while the pragmatist account can allow there to be many different and equally correct functions that future inquiry can create. Resemblance may constrain how future investigations should turn out, but they need not determine how they should. Most potential developments may be ruled out, but more than one path may remain open. We can understand mistakes as violating the resemblance constraints while still allowing that, as inquiry develops, there often remains a non-trivial amount of 'free play' compatible with all the existing constraints. It is within this free play that our interests and even accidental aspects of our inquiries can help determine what is true.
The Contribution of our Interests
The requirement that their results blend as harmoniously as possible with existing ideas constrains how our investigations should go, but there are still a number of ways in which new ideas can be harmoniously blended, even to the most 'ideal' extent, with the old. The opposing thought that it would still be determined, even in the absence of fully specific ideas, what an 'ideal' investigation should pick out can come from at least two sources: a belief in a 'ready-made world' and a belief in a 'ready-made' psychology. According to James, however, neither the world nor our psychology is 'ready made' in the relevant sense.
James has little sympathy with the suggestion that the world already has an 'essential' structure that an ideal inquiry must capture, and the belief that the world is, in this sense, "ready made" (PR 123) is the target of Pragmatism's seventh chapter. The following passages are fairly representative of his views:
What shall we call a thing anyhow? It seems quite arbitrary, for we carve out everything, just as we carve out constellations, to suit our own human purposes. . . We break the flux of sensible reality into things, then, at our will. We create the subject of our true as well as our false propositions. . . . you can't weed out the human contribution. Our nouns and adjectives are all humanized heirlooms, and in the theories we build them into, the inner order of arrangement is wholly dictated by human considerations, intellectual consistency being one of them. (PR 122)
I am a natural realist. The world per se may be likened to a cast of beans on a table. By themselves they spell nothing. An onlooker may group them as he likes. He may simply count them all and map them. He may select groups and name these capriciously, or name them to suit certain extrinsic purposes of his. Whatever he does, so long as he takes account of them, his account is neither false nor irrelevant. If neither, why not call it true? It fits the beans-minus-him, and expresses the total fact, of beans-plus-him. 
James is certainly right to think that the distinctions we make are a function of our interests. Indeed, even if there were a description of the world which was somehow 'metaphysically privileged,' if such a description did not serve our particularly human interests, then there would be no reason to think that the one designed to serve our interests would be seriously compromised by failing to map on to such an interest-independent description. What our terms refer to is partially a function of our interests, and people with different interests may thus come to mean different things by their words.
This, however, leads to another line of thought that might seem to reduce the significance of James' position. If our classificatory interests were themselves shared and predetermined, one could grant that our classification system was a reflection of our interests and still insist that it has always been settled how our terms should be applied. If it was always settled what our classificatory interests were, then it would always be settled what our terms referred to, and the actual process of investigation would have a merely confirmatory rather than constitutive role in determining what we meant. Our interests would be relevant, but no substantive role would seem to be left for future inquiry. One can see such a picture of our antecedent interests in Kripke's assertion that, when scientists originally claimed that whales were not fish, they corrected standard usage (rather than merely changed it) because an interest in natural kinds was "part of the original enterprise" of classification.  However, this assumption that the scientist's interests have always been our own would certainly be challenged by James, who insisted that the drive to classify things 'scientifically' is an "altogether peculiar and one-sided subjective interest" that was "invented but a generation or two ago," and that "few even of the cultivated members of the race have shared." 
One should agree with James's claim that the "original enterprise" of classification need not commit its participants to a preference for scientific taxonomies. Scientific taxonomies are not always the most useful, and if a speaker has a practical interest that is served better by another type of taxonomy, there is little reason to think that he should be committed to the scientific one. We can see a particularly vivid example of such a speaker in Melville's Moby-Dick, where Ishmael, in the face of precisely the 'correction' that Kripke focuses upon, makes the following statement:
In his System of Nature, AD. 1766, Linnaeus declares, 'I hereby separate the whales from the fish.' But of my own knowledge, I know that down to the year 1850, sharks and shad, alewives and herring, against Linnaeus's express edict, were still found dividing the possession of the same seas with the Leviathan. The grounds upon which Linnaeus would fain to have banished the whales from the waters, he states as follows: 'On account of their warm bilocular heart, their lungs, their movable eyelids, their hollow ears, penem intrantem feminam mammis lactantem' and finally 'ex lege naturae jure meritoque'. I submitted all this to my friends Simeon Macey and Charley Coffin, of Nantucket, both messmates of mine in a certain voyage, and they united in the opinion that the reasons set forth were altogether insufficient. Charley profanely hinted they were humbug. . . . I take the good old fashioned ground that the whale is a fish, and call upon holy Jonah to back me. This fundamental thing settled, the next point is, in what internal respect does the whale differ from other fish. Above, Linnaeus has given you those items. But in brief, they are these: lungs and warm blood; whereas, all other fish are lungless and cold blooded. 
Prior to the discovery that whales had internal structures different from those of 'other fish,' one might have been able to believe both that whales were fish, and that fish were a biological kind. This new discovery upsets the 'harmony' that may have existed among the beliefs of various individuals. James describes the emergence of such disharmony below.
The individual has a stock of old opinions already, but he meets a new experience that puts them to a strain. Somebody contradicts them; or in a reflective moment he discovers that they contradict each other; or he hears facts with which they are incompatible; or desires arise in him which they cease to satisfy. The result is an inward trouble to which his mind till then had been a stranger, and from which he seeks to escape by modifying his previous mass of opinions. He saves as much as he can, for in this matter of belief we are all extreme conservatives. (PR 34)
However, it isn't predetermined how people will react to such disharmony and we can expect different people to act differently. The fisherman and the scientists will react to the disharmony in different ways, ways determined largely by the differing importance that different beliefs play in their lives. Each must adopt either the idea that whales are not fish, or the idea that fish do not all have the same sort of internal structure. Which new belief will disturb the individual's existing set the least will depend upon their interests and can thus vary from speaker to speaker. For Linnaeus, the idea that fish are not a natural kind would be very disruptive. For Ishmael, the idea that whales are not fish would be. As James puts it:
This new idea . . . preserves the older stock of truths with a minimum of modification, stretching them just enough to make them admit the novelty, but conceiving that in ways as familiar as the case leaves possible. . . . We hold a theory true to the extent that it solves this 'problem of maxima and minima.' But success in solving this problem is eminently a matter of approximation. We say this theory solves it on the whole more satisfactorily than that theory; but that means more satisfactorily to ourselves, and individuals will emphasize their points of satisfaction differently. To a certain degree, therefore, everything here is plastic. (PR 35, italics mine)
As Linnaeus and Ishmael illustrate, different speakers can have different explanatory and classificatory interests, and in this particular case it seems best to say that Linnaeus and Ishmael simply mean different things by the word "fish." Their coming to mean different things by the term results from the 'plasticity' that James stress above. What compromise will be the most satisfactory can vary from person to person. 
The Contribution of the Future
The considerations immediately above suggest how James' account of intentionality can allow our interests to help determine whether or not a belief such as "whales are fish" is true, but how does it allow future developments to contribute to truth? Linnaeus's and Ishmael's preferences may have led them to mean different things by "fish," but (given the preferences that each had) how their investigations into the nature of "fish" would go seem predetermined. The actual process of investigation seemed only to make explicit commitments of theirs that were already in place.
However, Linnaeus and Ishmael have the theoretical commitments they do about "fish" picking out a natural or a functional kind because of their practical commitments as biologists and fishermen. Speakers without such practical commitments may have no way of forming a firm opinion about what one should say when one finds out that whales have a different internal structure than most fish.  Speakers may, however, acquire practical commitments that settle such issues over time, or may acquire the relevant 'theoretical' commitments themselves during the course of inquiry. If a young Linnaeus was trying to choose between becoming a fisherman or a scientist he would have the possibility of developing two entirely different sets of commitments and thus turning out to mean two entirely different things by his word "fish." This potential variability of the commitments we can take on allows the actual course of events to contribute to what we mean by a terms in a fashion that is not predictable from previous usage.
As our earlier discussion of resemblance and harmony suggested, it is an essential part of James' position that one's stock of previously held truths be part of what one must take into account when one searches for new ones. Because old truths play this role, the order in which one takes on new commitments by adding 'new truths' to one's stock of beliefs can affect which truths one can reach 'absolutely.' Until such commitments are made, it is not completely determined how one's inquiries should go.
The following case should illustrate how the order in which one acquires new truths can determine the eventual outcome of inquiry. Imagine a speaker, Edwin, who is part of an isolated community inhabiting an island on which the class of birds and the class of flying things are coextensive. He has a term "ave" that he applies to these locally co-extensive classes, and he has beliefs such as, "all and only aves can fly," and "all aves are living things."  Both of these beliefs are true of the "aves" on the island, but it is clear that one must be given up if, say, a plane lands on the island and he encounters a non-living flying thing. Edwin's set of ideas that previously seemed to blend harmoniously will no longer do so, and there may be no way to tell in advance how to preserve "the older stock of truths with a minimum of modification" (PR 35) and thus whether he should apply "ave" to planes or not.
In particular, the mode in which Edwin first encounters planes may determine the number and entrenchment of the 'old truths' favoring each of the two alternatives when it becomes clear that some of his 'ave'-beliefs must be given up. For instance, if he first sees planes flying high in the sky, his belief that all flying things are aves will lead him to call the planes "aves." Edwin's application of the term "ave" to planes may thus already be one of the "previous truths of which every new inquiry must take into account"  when he first sees one land and realizes that it is not a living thing. As a result, he may give up the belief that all aves are living things, and "ave" would 'absolutely' mean flying thing. No further experience on Edwin's part could show him that he was wrong to have applied the term the way he did. On the other hand, if Edwin first sees planes on the ground, his belief that all aves are living things will lead him to believe that planes are not aves, and this will be the new truth that subsequent inquiry will have to accommodate. This additional truth may cause him, on discovering that planes can fly, to reject the belief that all flying things are aves, and thus "ave" would come 'absolutely' to mean bird. The order of appearance of novel phenomena may affect which new truths we accept, and thus the vanishing point towards which our inquiry will approach. 
The way in which novel phenomena appear often makes how one should settle such issues seem so obvious that one fails to notice that one could have ever settled things differently. In each of the cases described above, Edwin will not see himself as having changed what he meant by "ave" in any way. He will describe himself as having discovered that he had always been wrong in thinking that all aves were living things, or as having discovered that he had always been wrong in thinking that all flying things were aves. According to James' theory, he is perfectly entitled to understand things this way. 
It should also be noted that while such cases are handled well by James' account of intentionality, they cause non-trivial problems for most others.  The claim that there really must be an inquiry-independent answer to what Edwin initially meant by "ave" seems indefensible, and if it can be unsettled what even our best inquiries should pick out, then any theory suggesting that it must be so settled will be in trouble. Since they understand intentionality in a way that makes it independent of subsequent inquiry, resemblance-driven and 'backwards-looking' causal theories require the extensions of our terms to be settled in just such a fashion. 
James Account of Truth
With James' picture of intentionality in place, his account of truth naturally follows. If the process of inquiry determines what our ideas refer to, then it too, and not just the world, helps determine whether our thoughts and statements are true. Cases such as those of Edwin or Ishmael illustrate how the claim that truth is tied to what would be believed at an ideal limit of inquiry can remain a substantive one. It has frequently been suggested that such appeals to an ideal limit let the older intellectualist conception of truth in through the back door, but this is not so. As James himself puts it, "like the half-truths, the absolute truth will have to be made, made as a relation incidental to the growth of a mass of verification-experience, to which the half-true ideas are all along contributing their quota" (PR 107). As a result, we need not agree with Suckiel's claim that "the hypothetical and normative notion of absolute truth as the best possible account of experience, an account which is retroactively true even before it is held, renders otiose reference to any beliefs which are actually held."  We can admit that "the 'absolutely' true . . . is that ideal vanishing-point towards which we imagine that all our temporary truths will some day converge" (PR 106-7), and still insist that the way in which we come across, and our attitudes towards, these 'temporary truths' will affect what this 'vanishing point' will turn out to be. The best possible account of experience is not out there waiting for us to discover it, it is essentially tied to our own accounts and itself changes as those accounts change. Once held, it applies retroactively, but there is no sense in which it was settled to be true before it was actually held.
It should be noted that once we understand how James' theories of intentionality and truth are connected, we can see how his account of truth makes room for the contributions of our preferences and interests without entailing the type of subjectivism that it has so often been accused of leading to. Our interests determine what we talk about and thus what we use our words to refer to, but once meanings are in place, it is the world which determines what is true. The subjective contribution to truth has to do with determining thought and utterance content, not with whether contentful thoughts and utterances are satisfied by the world. It thus does not, as his critics often suggested, allow for the possibility that "P" could be true even though not-P. 
This answer to the charge of subjectivism allows the James' account of truth to be understood as applying to all truths. In particular, it avoids what I will here refer to as "bifurcationist" attempts to defend James from charges of subjectivism. These accounts try to defend James by restricting the scope of his theories and leaving interest and inquiry independent conceptions of cognitive matters in place in those areas where subjectivism seemed threatening. Such bifurcationist moves free James from the charge of subjectivism only at the expense of stripping his theories of their revisionary potential.
Such a bifurcationist move is, for instance, popular in readings of James' "The Will to Believe." This early essay of James' is of interest precisely because it suggests how our 'passional' nature can contribute to what is epistemically rational for us to believe. Unfortunately, in order to avoid the subjectivism about epistemic justification that this view might seem to lead to, James is often interpreted as leaving epistemic rationality entirely to his intellectualist opponents and merely arguing that believing what is in our interest can occasionally be prudentially rational. Rather than challenging traditional conceptions of epistemic rationality, then, James is instead treated as arguing that epistemic rationality is not the only game in town, and that other factors can contribute to what is rational to believe.  This temptation to read James as leaving unchallenged traditional conceptions of epistemic and cognitive matters strips his writings on belief justification of most of their interest, and similar dangers are presented by many readings of his account of truth.
For instance, Ayer tries to defend James from the charge of subjectivism by arguing that James' claims about the contribution of our interests to determining what was true were meant to apply to only a restricted range of truths, namely those of morals and theology.  Subjectivism is, presumably, meant to be comparatively harmlessness here because one's intuition that there are 'objective' facts in these areas is considerably weaker than it is about, say, the number of apples presently in one's refrigerator. Ayer's account 'saves' James by effectively claiming that the particularly 'pragmatic' aspects of his theory of truth did not apply to most of our statements about the world. On such a reading, however, James comes out as little more than an empiricist with an unusually forgiving conception of a range of statement that most other empiricists viewed as problematic. The pragmatist turns out to have nothing new to say about the truth of most of our statements about the world around us.
In a similar fashion, the interpretation suggested here avoids subjectivism without having to accept H.S. Thayer's claim that James' account "might be more accurately described as a theory of pragmatic truth rather than a pragmatic theory of truth."  Thayer's account distinguishes 'cognitive' and 'pragmatic' truth, leaving cognitive truth to the intellectualist and arguing that James was providing an analysis of the latter notion. According to Thayer, for an idea to be "pragmatically true" it must:
1. Be cognitively true.
2. Be compatible with the older body of truths.
3. Work. It must provide some satisfaction of a need or purpose. 
"Cognitive" (as opposed to "pragmatic") truth is to be understood in terms of traditional notions of correspondence, and so Thayer's James is able to avoid charges of radical subjectivism by pointing out that an idea cannot be pragmatically true unless it has already satisfied the traditional requirements of cognitive truth. To be pragmatically true, and idea must at least be true in the ordinary sense.
There are, of course, serious exegetical difficulties with maintaining that James both held such a crucial distinction between cognitive and pragmatic truth and yet, even in the face of the barrage of criticisms that followed the publication of Pragmatism, never managed to flag it clearly.  However, for present purposes, the most serious disadvantage of Thayer's account is that it deprives James' view of much of its interest.
Cognitive truth is, after all, truth, and by handing cognitive truth over to James' intellectualist opponents, Thayer allows that pace James, they have a working conception of truth. Indeed, if cognitive truth is one of the three ingredients of pragmatic truth, then James' criticisms of the emptiness and unclarity of traditional correspondence accounts would rebound onto his own. If cognitive truth is an ingredient in pragmatic truth, then to the extent that cognitive truth is unclear, pragmatic truth will be as well. Pragmatism was supposed to convert "the absolutely empty notion of a static relation of 'correspondence'" into that of "a rich and active commerce" between our thoughts and "the great universe of other experiences." (PR 39) However, if pragmatic truth is understood in terms of cognitive truth, such a conversion would seem to be impossible. Things may be added to the intellectualist conception of truth to get the pragmatist one, but the intellectualist conception itself would remain unchanged and unchallenged.
By making cognitive truth an independent ingredient in pragmatic truth, Thayer leaves the pragmatist nothing positive to say about the crucial notion of cognitive truth. However, the main source of interest in James' account was precisely its promise to show how our interests and investigations played a role in determining what was cognitively true. Any attempt to preserve James' account from subjectivism by arguing that it was not, after all, meant to be an account of cognitive truth seriously compromises its potential to challenge what James considered the "vicious intellectualism" in philosophical thought about cognition.
Fortunately, the account of James presented in the preceding sections requires no such distinction between cognitive and pragmatic truth. Thayer's conditions (2) and (3) are not something to be added to cognitive truth, they are precisely what determines cognitive truth. Cognitive truth is not a primitive in terms of which we can understand pragmatic truth, rather it is the pragmatic process of making truths that underwrites cognitive truth. 
By connecting it to his account of intentionality, then, one is able to make James' account plausible without having to claim that it was only meant to apply to a restricted range of truths or special type of truth. Our interests and investigations have the potential to contribute to the cognitive truth of any of our beliefs and assertions.
Rather than as presenting an unintuitive and undermotivated theory of truth, then, James should be understood as recognizing that an account of the truth of our thoughts and utterances should include an explanation of how they come to be about the world in the first place. While I've only been able to sketch it here, James' account of intentionality stands up well to its better-known rivals, and in light of it, his account of truth preserves both its plausibility and its bite. James' account of truth allows subjective factors to contribute to the truth of our beliefs, but does so in a way that is not overly subjectivistic. Our interests and investigations help determine whether or not our thought and utterances are true by helping to determine just what we are talking and thinking about.
* I would like to thank Robert Brandom, James Campbell, James Conant, Richard Gale, Peter Hare, John McDowell, Ram Neta, Stuart Rosenbaum, Hans Seigfried, H.S. Thayer and members of the audience at the 1997 SAAP meeting for comments on earlier drafts of this paper.
1 This combination is captured well in his claim that the truth is "the expedient in the way of our thinking." (Pragmatism 106, original in italics.) James' works will be referred to hereafter with the following abbreviations. "PR" for Pragmatism, "MT" for The Meaning of Truth, "ERE" for Essays in Radical Empiricism, "PP" for The Principles of Psychology, "VRE" for The Varieties of Religious Experience, and "CER" for Collected Essays and Reviews. All six are cited from the Harvard editions. Ralph Barton Perry's The Thought and Character of William James, Vol. 2, is here referred to as "RBP II."
2 Witness the extensive literature since Russell on "propositional attitudes."
3 "Ideas are so much flat psychological surface unless some mirrored matter gives them cognitive luster" (MT 106).
4 MT 151. James did, however, occasionally use the term himself (see for instance PR 122). See Russell, Our Knowledge of the External World (New York: Routledge, 1914/1993), p.62, for a discussion of propositions of the sort James was suspicious of.
5 Letter to Kallen quoted in MT 305. See also RBP II 485, where he refers to propositions as "mongrel figments."
6 J. B. Pratt "Truth and its Verification" reprinted in D. Olin (ed), William James' Pragmatism in focus. New York: Routledge. Italics Pratt's.
7 James considers "a thought knows a reality" to be equivalent to its being "an incomplete 'thought about' that reality" or that reality being "its 'topic'" (MT 28).
8 PP 212. The problem comes up again at the beginning of the chapter on truth in Pragmatism.
9 A book in which Royce presented what James called an "original proof of Idealism" based on the question, "How can a thought refer to, intend, or signify any particular reality outside itself?" (CER 384, 385).
10 For instance, James "A Word More about Truth" (MT 78-89) is primarily about how our thoughts come to be about the world.
11 "Names are just as 'true' or 'false' as definite mental pictures are." (PR 102, italics James'.)
12 James is, of course, very critical of the idea that such resemblance could be either necessary or sufficient for intentionality. If it were sufficient, we would expect each egg to be 'about' every other egg (MT 21). If it were necessary, most of us would be unable to have thoughts about, say, a clock's inner workings (PR 96).
13 The attribution of such an account to Frege has been criticized by Evans (see his The Varieties of Reference (OUP, 1982), ch. 2). Nevertheless, whether or not Frege actually held such a theory, 'descriptive' accounts of intentionality became quite popular in part because of his influence.
14 This term is from Evans, The Varieties of Reference, ch. 3. For a good example of such a causal account, see Michael Devitt's Designation (Columbia UP, 1981). Recent enthusiasm for such accounts has been inspired, of course, by Saul Kripke's Naming and Necessity (Harvard UP, 1980) and Hilary Putnam's "The Meaning of 'Meaning'" (reprinted in his Mind, Language, and Reality, Cambridge UP 1975). Though it should be noted that neither Putnam nor Kripke defend causal theories quite as crudely non-intentional as the ones they inspired.
15 This is why one should not (as Robert Burch does in his "James and the 'New Theory of Reference'" (Transactions of the Charles S. Peirce Society v. XV, 1979)) overemphasize the undoubted similarities between James' account and contemporary causal theories. While Burch is right in thinking that James anticipates many aspects of contemporary causal accounts, he downplays just how serious a difference James' orientation towards the future makes. An idea's past causal history is, according to James, important precisely because such history would be important to us in the future. It is this essential reference to the future that allows James' account to underwrite his claims about truth.
16 This essay presents only a brief sketch of James account of intentionality, since its main purpose is to argue for the importance his account of intentionality has for his account of truth. I hope to give a fuller presentation of his account of intentionality elsewhere.
17 MT 36. See also MT 35, 61-2.
18 "A percept knows whatever reality it directly or indirectly operates on and resembles; a conceptual feeling, or thought, knows a reality, whenever it actually or potentially terminates in a percept that operates on or resembles that reality, or is otherwise connected with it or with its context" (MT 28, italics James').
19 ERE 29, Italics James.'
20 Which he published in 1884 and later described as "the fons et origo of all my pragmatism" (RBP II, 548).
21 MT 26, italics removed. "Feeling" being used by James more-or-less interchangeably with "idea," "thought" or "state of consciousness" (see MT 13-14).
22 In a letter to C.A. Strong, James reiterated the importance of resemblance to the cognitive process, and also stresses that this resemblance was never intended to be exclusively sensuous or imagelike but was meant to be viewed as 'logical' as well (RBP II 545-6).
23 MT 26. Ideas refer to those objects which they intend, and James' talk of intentionality in terms of "intentions" makes its future-directed character more vivid.
24 James makes this clear in annotations made to an article by John E. Russell. Russell asks how James could have known that "the subsequent experience did so fulfill and satisfy that antecedent idea as to make it a true idea" (quoted in ERE 278). James' comments on this passage ("Do you mean "know for sure" or do you mean "feel as if he knew"? If the former, no one ever knows for sure" (ERE 278)) illustrate that, even if the knowing relation is made, there is still a difference between thinking that one has found the object of one's thought, and actually having found it. 25 Similar considerations will allow James to incorporate his more social picture of language into his basically individualistic framework. Considerations of space will not, however, allow to pursue this topic further here.
26 James to Dickinson S. Miller, August 5, 1907. The Letters of William James, volume 2. (Boston: The Atlantic Monthly Press, 1920), p.295f. The bean analogy is found again in VRE 346. Thanks to Hans Seigfried for drawing these last two passages to my attention.
27 Kripke, Naming and Necessity, p.138.
28 PP 1236.
29 Melville, Moby-Dick, The Library of America, p.936.
30 And while these differences can involve 'theoretical' commitments, emotional commitments can also play a role in choosing between two possible resolutions that are on equal footing intellectually. As James puts it, "My position is that, other things equal, emotional satisfaction counts for truth -- among the other things being the intellectual satisfactions" (Letter to Perry quoted in RBP II 475).
31 As James put it, "the moment you pass beyond the practical use of these categories . . . to a merely curious or speculative way of thinking, you find it impossible to say within just what limits of fact any one of them apply" (PR 90).
32 I.e., an equal number of his beliefs favor each of the two alternatives. We can assume that the rest of Edwin's community is in a similar position. Indeed, the example here is adapted from Mark Wilson's "Predicate Meets Property" (Philosophical Review, XCI, No, 4, Oct 1982), and Edwin is a member of his community of "Druids".
33 PR 118, "previous truths" originally in italics.
34 In this respect, linguistic development is much like case law, which, notoriously, has this property. (James compares the way new truths must accommodate the old with the development of law in PR 116.)
35 Some of the consequences of, and objections to, such a view of linguistic development are discussed in my Semantic Norms and Temporal Externalism, Ph.D. Thesis, University of Pittsburgh, 1996.
36 And Wilson argues that such cases occur "in virtually every case of enlargement of our world view through scientific progress" (Wilson, "Predicate Meets Property", p.572).
37 As a result, non-forward-looking accounts must, in these cases, depart from the way we typically understand ourselves and not only claim that what we mean is often indeterminate, but also claim that what we mean changes over time (and that we just falsely believe otherwise). Rather than seeing our language as a continuously evolving practice, they must treat speakers as using a series of indeterminate languages which they mistakenly assume to be the same. That it does not commit one to such claims is a substantial advantage that James' account of intentionality has over its rivals. These issues are discussed in greater detail in my Semantic Norms and Temporal Externalism, Ph.D. Thesis, University of Pittsburgh, 1996.
38 Suckiel, The Pragmatic Philosophy of William James, (Notre Dame, Notre Dame University Press. 1982) 115.
39 For perhaps the most influential expressions of such criticisms, see Russell "William James' Conception of Truth" in his Philosophical Essays (Routledge, 1910/1994), and Moore "William James' 'Pragmatism'" in his Philosophical Studies (Routledge: 1922).
40 Examples of such readings of James and a sketch of an alternative account can be found in my "James on Prudential Arguments and the Will to Believe" (Proceedings of the Ohio Philosophical Association, 1997, pp.97-108), and I hope to provide a fuller account of James' views in this area some time in the near future.
41 See his The Origins of Pragmatism (San Francisco: Freeman, Cooper & Company, 1968), pp. 191, 213, 323-4, and his Introduction to Pragmatism and the Meaning of Truth (Cambridge: Harvard UP, 1978), p. xxx. For criticisms of Ayer's attempt to restrict the scope of James' theory in this way, see James Campbell's "Ayer and Pragmatism" in Hahn, (ed.) The Philosophy of A.J. Ayer (La Salle: Open Court, 1992).
42 Thayer, Meaning and Action Indianapolis, Hackett: 1981. p. 540. See also his introduction to MT.
43 Thayer, Meaning and Action, p. 546. See also his introduction to MT.
44 For a discussion of these and other exegetical difficulties, see Suckiel, The Pragmatic Philosophy of William James, pp. 118-19, and especially Seigfried William James' Radical Reconstruction of Philosophy (Albany: SUNY Press, 1990), pp. 289-91.
45 On this issue I am entirely in agreement Seigfried, op. cit. These criticisms should not, of course, obscure the debt I owe Professor Thayer both for his written work, and for his comments on an earlier draft of this paper.
Back to Henry Jackman's Home Page