James' Naturalistic Account of Concepts and his 'Rejection of Logic'
Logic is viewed by many as inseparable from rationality, and James' 'rejection of logic' in A Pluralistic Universe has been viewed as the most flagrantly 'irrational' strand in his philosophy.  Indeed, the late date of A Pluralistic Universe (the lectures were given in 1908) may tempt some to write it off as inessential to James' larger philosophical vision. Nevertheless, James' views on logic flow from currents that run deep in his philosophy, and these 'naturalistic' currents can be traced back to works as early as 1879's "The Sentiment of Rationality." These currents are crucial to understanding James' later work, since when viewed in light of the psychological naturalism developed in The Principles of Psychology, James' so-called 'rejection of logic' can seem both plausible and, crucially, rational.
James' rejection of logic is not simply an unintended consequence of other views that can be found in his writings, it is a commitment that he takes on explicitly in passages such as the following:
I have finally found myself compelled to give up the logic [of identity], fairly, squarely, and irrevocably. It has an imperishable use in human life, but that use is not to make us theoretically acquainted with the essential nature of reality. Reality, life, expedience, concreteness, immediacy, use what words you will, exceeds our logic, overflows and surrounds it. If you like to employ words eulogistically, and so encourage confusion, you may say that reality obeys a higher logic, or enjoys a higher rationality. (PU 96-7)
Passages such as this have led many to view James as suggesting that the law of identity was not true, and that as advising us to give up on using logic. It is thus not surprising that his writings on logic were met with incomprehension and disappointment by many of James' contemporaries. 
However, James 'anti-logical' writings, while perhaps not as happily put as they could be, pick out something very deep and important that runs throughout his philosophy. In particular, James's target is not so much logic, as it is a certain attitude towards our concepts. If logic (particularly the logic of identity) fails to apply to reality, the problem is not so much with logic itself as it is with our attitudes towards the conceptualizations of reality upon which our logic is applied. Logical inferences are only applicable to conceptualizations of reality, and our conceptualizations may not (for certain theoretical purposes) adequately reflect reality's actual structure.  James' claim is that logic can take concepts that have evolved to cope with reality on a practical level, and derive a theoretical picture that grotesquely distorts reality. In such cases the rational thing to do is to "subordinate logic ... [and] throw it out of the deeper regions of philosophy to take its rightful and respectable place in the world of simple human practice" (PU 97). The claim that logic will not always lead us to the truth is not the same as the claim that its laws are not true. James defends the former claim, but he is not committed to the latter.
James' naturalistic understanding of concepts is not a late appearance in his philosophy, it goes back at least as far as works like "The Sentiment of Rationality" and The Principles of Psychology. James' view has always been, as he puts it in both of these early works, that concepts are 'teleological instruments' with which partial aspects of things (which "for our purpose" we regard as essential aspects) are used to represent the whole.  James takes such conceptualizations to be indispensable because they allow us to make sense of experience by breaking it up into kinds about which general inferences can be made:
Each concept means a particular kind of thing, and a far more efficient handling of a given bit of experience begins as soon as we have classed the various parts of it. Once classed, a thing can be treated by the law of its class, and the advantages are endless. (PU 98.)
We of course need a stable scheme of concepts, stably related with one another to lay hold of our experiences and to co-ordinate them withal. The immutability of such an abstract system is its great practical merit; the same identical terms and relations in it can always be recovered and referred. (PU 105.)
There are, of course, many ways to divide experience in to kinds, and it should be stressed that James views the 'essential' properties that our concepts pick out as having as much to do with our interests as with the world itself:
There is no property ABSOLUTELY essential to any one thing. The same property which figures as the essence of a thing on one occasion becomes a very inessential feature upon another.... But as I am always classifying it under one aspect or another, I am always unjust, always partial, always exclusive. My excuse is necessity -- the necessity which my finite and practical nature lays upon me. My thinking is first and last for the sake of my doing, and I can only do one thing at a time.the only meaning of essences is teleological, and that classification and conceptions are purely teleological weapons of the mind. The essence of a thing is that one of its properties which is so important for my interests that in comparison with it I may neglect the rest. (PP 959-61)
Concepts are thus not heavenly forms that we somehow grasp or intuit. Nor are they forced upon us by a 'ready made' world that has essential properties of its own.  Rather, they are natural simplifications/adaptations that we develop in order to make sense of our experience, and thus cope with our current environment. As James puts it, the concepts under which we characterize a given object "characterize us more than they characterize the thing" (PP 961). Concepts of the less naturalistic sorts should automatically match (perhaps even determine) reality's structure, but there is no guarantee that concepts as naturalistically understood will do so.
James always held this more naturalistic view of concepts, but its consequences for the extent of the applicability of logic are only fully developed in A Pluralistic Universe and Some Problems of Philosophy.  In these works, James explains that our concepts are themselves independent objects of experience.
Concepts are realities of a new order, with particular relations between them. These relations are just as much directly perceived, when we compare our various concepts, as the distance between two sense-objects is perceived when we look at it. Conception is an operation which gives us material for new acts of perception. (PU 122.)
James further expands on how our concepts can collectively make up self-standing models (or "maps" (SPP 43)) which we can inspect, and in terms of which perceptual experience can be understood.
The 'rationalization' of any mass of perceptual fact consists in first assimilating its concrete terms, one by one, to so many terms of the conceptual series, and then in assuming that the relations intuitively found among the latter are what connect the former too. To 'explain' means to co-ordinate, one to one, the thises of the perceptual flow with the whats of the ideal manifold, whichever it be. (SPP 41-2, see also p. 33.)
The map which the mind frames out of [our concepts] is an object which possesses, when once it has been framed, an independent existence. It suffices all by itself for the purposes of study. The 'eternal' truths it contains would have to be acknowledged even were the world of sense annihilated. (SPP 43.)
As James later puts it, in order to be successful, the models of reality constructed with our concepts need only do justice to those aspects of reality that they are used to cope with.  Our concepts have been developed and selected through our history for their usefulness, not necessarily their 'truth' (though the two will not be entirely unrelated), so the types of inferences that our concepts license may not be true of everything (or always true of anything) that they are applied to. A simple model that allows one to make the right decision most of the time is often more useful than a more 'truthful' model that is too complex to be used effectively in actual practice. 
Indeed, James argues that we frequently use different models to cope with different aspects of reality, and while this practice is useful, it would inevitably lead to contradictions if the models were all viewed as true theoretical descriptions of reality.  This is one of the sources of James' 'intrumentalism.' Our models are useful instruments to cope with experience but their theoretical incompatibility prevents them from being viewed as absolutely true descriptions of reality. As James famously put it "Common sense is better for one sphere of life, science for another, philosophic criticism for a third; but whether either is truer absolutely, Heaven only knows" (PR 93). James' 'instrumentalism' does not stem from a prudishness about 'unobservables' (as if we had a single coherent theory of the world, but refused to commit ourselves to the existence of the theoretical entities postulated in it). Rather, it derives from a recognition that we have a number of indispensable yet incompatible models of the world no single one of which is adequate for all of our purposes, and no two of which could be 'absolutely true' together. 
Furthermore, these conceptual models of reality are often metaphorical in character, and James is very sensitive to the analogical and metaphorical nature of many of our concepts. We often understand novel ranges of experience by analogy with other experiences that we are more familiar with. This 'metaphorical' form of understanding is a very powerful tool for comprehending not only novel experiences, but also things as familiar as our own minds. Indeed, James was very aware of our tendency to understand 'abstract' phenomena such as the mind in terms of 'concrete' metaphors relating to our practical interactions with the physical world. As he puts it "To deal with moral facts conceptually, we have first to transform them, substitute brain-diagrams or physical metaphors, treat ideas as atoms, interests as mechanical forces, our conscious 'selves' as 'streams' and the like."  These 'concrete' metaphors are, according to James, essential to our understanding precisely because human cognition evolved not in the context of having to solve theoretical problems about comparatively abstract objects, but rather in the context of practically coping with our concrete environment. Concrete objects and "things of the sort we literally handle, are what our intellects cope with the most successfully," and this suggests that "the original and still surviving function of our intellectual life is to guide us in the practical adaptation of our expectancies and activities" (PU 111).
However, while importing the inferential structure of one domain into another is often a successful way of coping with experience,  it can occasionally misdirect our thinking. If an analogy that is successful for certain practical purposes is treated as a literal reflection of reality, then all of the inferential transitions licensed in the primary domain would be licensed in the metaphorical one. Loosing sight of the differences between the two domains can lead reasoning astray, and James argues that while genius is "the power of seeing analogies", not making allowances for the differences between the two domains is "the common fallacy in analogical reasoning" (PU 71). If we were to uncritically tease out all of the 'logical consequences' of our metaphorically structured concepts, we would frequently be led into error.  Consequently, while metaphorical concepts are useful, indeed indispensable, they should be used with caution outside of their 'everyday' practical use."
Concepts are, then, for James, simply tools with which we practically cope with our environment, and this conception of concepts (and their resulting limitations) is radically at odds with that of his more 'logical' opponents.
To understand how James' account of concepts is tied to his views on logic, we should remember that the main target of the 'anti-logical' lectures in A Pluralistic Universe is not monism, but intellectualism. James characterizes intellectualism as "The treating of a name as excluding from the fact named what the name's definition fails positively to include."  Intellectualism requires not that concepts exclude from themselves everything that they do not include. Rather, it requires that they exclude from the realities to which they are applied everything that they do not include. Consequently, it presupposes a very close match between the nature of our concepts and the nature of the world they purport to represent.
According to James, intellectualism has as its source "the faculty which gives us our chief superiority to the brutes," namely, our power "of translating the crude flux of our merely feeling-experience into a conceptual order" (PU 98). James claims that whenever we conceive a thing, we define it,  and intellectualism involves taking concepts to capture reality so well that the inferential patterns flowing from our definitions become the measure of reality itself. James traces this tradition of 'abusing' our concepts back to Socrates and Plato:
Intellectualism in the vicious sense began when Socrates and Plato taught that what a thing really is, is told us by its definition. Ever since Socrates, we have been taught that reality consists of essences, not of appearances, and that the essences of a thing are known whenever we know their definitions. So first we identify the thing with a concept and then we identify the concept with a definition, and only then, inasmuch as the thing is whatever the definition expresses, are we sure of apprehending the real essence of it or the full truth about it. 
If the inferential consequences that flow from our concepts' definitions reflect the 'essence' of reality, then logic (by being able to tease out these inferential consequences) would be "an adequate measure of what can and cannot be."  Logic is able to determine the structure of, and relations between, the models we construct to understand the world, and if we can assume that the structure of these mental models is isomorphic to the structure of the world,  then such logical investigations would reveal the structure of the world as well.
This assumption that our concepts match reality, coupled with the use of logic to determine just what is, and is not, included in our concepts, leads the intellectualist to deny the reality of seemingly obvious features of experience. James' complaints about the 'verbal' nature of Lotze's, Royce's and Bradley's arguments relate to how they all rely on the properties of words rather than things (PU 31-3). Such arguments, James would insist, properly draw conclusions about the nature of our concepts, not about the nature of the reality conceived. In such cases, concepts, "first employed to make things intelligible, are clung to even when they make them unintelligible." 
James claims that ordinary logic "substitutes concepts for real things" (PU 67). If (as the intellectualist supposes) the structure of our concepts 'mirrors' the structure of reality, then conclusions logically derived from the structure of our concepts should also be true of the reality conceived. If, however, the conceptual order does not mirror the order of reality, no such conclusions follow. Of course, from James' naturalistic perspective, there is little, if any, reason to think that such a mirroring relationship must exist. If concepts are effective but imperfect instruments we developed to cope with reality, there is no a priori reason to think that the structure of these tools must be completely isomorphic to the structure of what they work on.
It is this potential gap between concepts and what they represent that lets James see a distinction, invisible to most of his contemporaries yet essential to his own position, between logic and rationality. James is not here making the now familiar claim that logic is concerned simply with truth, while rationality is concerned with a wider range of human concerns (not the least of which is utility). This familiar claim is associated with a popular reading of James' philosophy, particularly his Pragmatism and "The Will to Believe."  The suggestion that we should understand his distinction between logic and rationality in terms of a distinction between truth and utility might also seem supported by James' claim that "rationality has at least four dimensions, intellectual, aesthetical, moral, and practical" (PU 54-5).  However, James' claim that there are at least four dimensions of rationality does not in itself suggest that there are forms of rationality that are not truth-directed. Indeed, such a reading of James would suggest that truth was the exclusive concern of intellectual rationality, and thus that the aesthetical, moral and practical dimensions of rationality have no business with truth. This would be a very un-Jamesian concession to his rationalist opponents.  In any case, James is clearly talking about rationality in all its dimensions when he claims right before the passage quoted above that any hypothesis that makes the world appear more rational "will always be accepted as more probably true than an hypothesis that makes the world appear irrational" (PU 54). Consequently, such passages give us no reason to think that James' distinction between logic and rationality should be understood as mirroring the comparatively uncontroversial distinction between truth and utility. 
Rather than relying on a division between truth and utility, James' distinction between logic and rationality is best seen as drawing on the potential differences between the conceptual order and the reality that it is supposed to represent. Rationality is concerned with reality and truth, while logic is concerned with the inferential relations between our concepts. If (as the intellectualist assumes) our concepts capture the structure of reality, then there will be no room for a conflict between logic and rationality. However, if (like James) one feels that our concepts are a practically adequate, but nevertheless imperfect reflection of reality, then there will be room for a conflict between logic and rationality. James sees his intellectualist opponents' uncritical use of conceptual logic as leading them to conclusion that are radically out of touch with any robust sense of the real. Such philosophers, in virtue of being "loyal to the logical kind of rationality" end up being "disloyal to every other kind" (PU 94). The contradictions which can follow from the unrestricted use of conceptual logic point to a dilemma that James takes his opponents to simply ignore.  Namely, we must either "give up the logic of identity" or "believe human experience to be fundamentally irrational", and while "neither is easy", "we must do one or the other."  When he faces up to the dilemma, James has no doubt about which horn to grab.
That secret of a continuous life which the universe knows by heart and acts on every instant cannot be a contradiction incarnate. If logic says that it is one, so much the worse for logic. Logic being the lesser thing, the static incomplete abstraction, must succumb to reality, not reality to logic. Our intelligence cannot wall itself up alive, like a pupa in its chrysalis. It must at any cost keep on speaking terms with the universe that engendered it. (PU 94)
What James chooses to preserve, it should be noted, is not only the legitimacy of naïve perceptual experience, but also the assumption that the world we experience is fundamentally rational. If a conceptual treatment of perceptual reality ("when radically and consistently carried out," (SPP 46)) leads to the conclusion that perceptual reality is not real at all, this simply illustrates our concepts' inability to adequately capture (for the purposes of theory) the reality perceived. A proper 'sense of reality' is crucial when making inferences with our concepts. If they seem to be leading us astray, that may be a good indication that they in fact are.
Nevertheless, the use of concepts is essential to coping with reality, and James is certainly not suggesting that we try to get by without them. If concepts had a purely theoretical function, then their leading to contradictions might suggest that they should be given up. On the other hand, if (as James insists) they have primarily a practical function, and their use leads to no practical problems, then there is no reason for us not to keep using them. James thus advocates a 'pragmatic' approach to the use of our concepts. Use them when they help us understand reality (as they typically do) but discard them whenever they seem to lead us to contradiction and confusion.
Since it is only the conceptual form which forces the dialectic contradictions upon the innocent sensible reality, the remedy would seem to be simple. Use concepts when they help, and drop them when they hinder, understanding. (SPP 53)
This paradigmatically pragmatic attitude towards our concepts is firmly grounded not in a lack of concern with truth or rationality, but rather in a naturalistic attitude towards concepts and their limitations.
James can thus be understood as making a type of 'Wittgensteinian' point. Our concepts are fine for their 'everyday' use, but if the inferential moves they license are applied indiscriminately, they can lead us to the sorts of contradictions and paradoxes characteristic of philosophy.  Nevertheless, the fact that theoretical contradictions can be derived from a concept does not, in itself, require that it be given up in practical life.  Just because a map fails for some purposes, it does not follow that it can't be used for others, and just because a set of maps are not jointly consistent, it doesn't follow that we can't benefit from using them.
Our experience and upbringing will provide us with a practically indispensable set of concepts. Nevertheless, the 'theory of the world' (better, 'theories of the world') that these concepts embody may lead to apparent contradictions if all of the inferences that they make available are followed through. Inquiry may result in a new concepts (or reinterpretations of old concepts) that license new sets of inferences that 'work' better with our environment.  Concepts are still directed towards aspects of our environment even when they license inferences that are not entirely true of them. An improved set of concepts may more accurately capture the relations between the phenomena the original set of concepts was developed to cope with. Consequently, by taking the inferences licensed by one's current set of concepts to simply reveal 'how things are', intellectualism can actually block inquiry by preventing a new and more adequate set from being developed.  Since we have no guarantee that our current concepts perfectly reflect the structure of reality,  James would argue that we are free to reject the inferential consequences of those concepts if they seem radically at odds with common sense. 
James' rejection of conceptual logic is thus deeply connected to his naturalism about concepts and the limitations of human conceptualization. There is no reason to think that an intellect "built up of practical interests" (PP 941) need develop concepts that precisely mirror the structure of reality. Our concepts may be flawed from the point of view of pure theory, but in absence of a more adequate set (and in face of the fact that they work fine for practical purposes), giving them up is neither a realistic nor a rational option. The concepts are not only practically useful, but may serve as a theoretical base camp that will serve until we find a more theoretically adequate set. James' rejection of logic can thus be understood as reflecting a type of anti-rationalism, in that it undermines the 'rationalist' program that extends from Plato right though to 20th century 'conceptual analysis.'  Nevertheless, it is not a form of irrationalism. That is to say, it is not committed to the claim that life or the world is fundamentally irrational. Indeed, just the opposite is the case. James rejects logic precisely because of his commitment to the belief that life and reality must be rational. If reality didn't have to be rational, the contradictions that conceptual logic presents us with could be accepted as literally and unproblematically true.
James position is, then, not so much that we should give up logic, but rather that we should give up the assumption that we are rationally obligated to endorse all of the apparent logical consequences of all the claims that we accept.  This view is ultimately motivated not in terms of any commitment to the irrational, but rather from James' fundamentally naturalistic approach to the mind and its powers.
1 A number of James works will be cited throughout this paper and The Will to Believe, The Principles of Psychology, Pragmatism, A Pluralistic Universe and Some Problems of Philosophy will here be referred to as WB, PP, PR, PU and SPP respectively. All references are to the Harvard editions of his work. I'd also like to thank Jim Campbell, Richard Gale and audience members at the 1999 meeting of the Society for the Advancement of American Philosophy for comments on earlier version of this paper.
2 Other candidates include Pragmatism's purported equation truth with what is expedient to believe, and his purported claim in The Will to Believe that we are rationally entitled to form any belief that makes us happy. I don't believe that any of these attributions of defenses of irrationality to James are, ultimately, justified, and I have argued against these other two purported strands elsewhere. (See my "James' pragmatic account of intentionality and truth" (Transaction of the Charles S. Pierce Society, Winter 1998, Vol. XXXIV, No. 1. ), and "Prudential Arguments, Naturalized Epistemology, and the Will to Believe" (Transaction of the Charles S. Pierce Society, Winter 1999, Vol. XXXV, No. 1.)
3 For the negative reaction of James' contemporaries to his rejection of logic, see Perry's The Thought and Character of William James, v. II pp. 594-7.
4 As he also puts it: "logic, giving primarily the relations between concepts as such, and the relations between natural facts only secondarily or so far as the facts have been already identified with concepts and defined by them, must of course stand or fall with the conceptual method. But the conceptual method is a transformation which the flux of life undergoes at our hands in the interests of practice essentially and only subordinately in the interests of theory." (PU 109) This is not to suggest that reality need have a single privileged structural description that any theoretical concepts must try to fit. Such a view would be out of line with the pluralistic 'humanism' outlined in James' Pragmatism. However, even if there are many different conceptual systems that match structures that reality 'actually' has, there may still be conceptual systems that match none of these.
5 PP 961-2. For similar remarks in "The Sentiment of Rationality", see WB 62.
6 As they might seem to do in theories ranging from Aristotelian/Medieval theory of phantasms to version of the "causal theory of reference" in which the world's essential structure determines the content of our terms (e.g.: Michael Devitt's Designation (New York, Columbia University Press, 1981)). For a discussion of the similar ambitions of such theories, see Rorty's Philosophy and the Mirror of Nature (Princeton: Princeton University Press, 1979), especially chapter four. A similarly naturalistically motivated suspicion of taking our concepts too seriously was also characteristic of Dewey's philosophy. (For a useful discussion of this, see Jim Campbell's The Community Reconstructs (Chicago: University of Illinois Press, 1992), especially chapter five.)
7 Though remarks about the difficulty about capturing experience with 'static' concepts go all the way back to The Principles of Psychology (e.g. PP 442).
8 If one is driving from New York to Boston, one doesn't want a map that shows every road, alley and cow path between those two cities, and information about the terrain; vegetation and population are even less necessary. A map that had every such detail would typically be less effective in guiding one between the two cities than the less 'cluttered' maps we typically use. Different maps server different purposes, and there is no reason to think that just because a feature is absent on a particular map that that feature must be missing in the reality depicted.
9 For instance, the primitive 'model' of the world found in frogs treats all small flying objects as things to be eaten, and while the actions endorsed by this model are not always optimal (eating BBs etc) it works often enough for frogs to survive in their environment. Such issues are discussed in more detail in relation to James in my "Prudential Arguments, Naturalized Epistemology, and the Will to Believe".
10 For a contemporary version of the suggestion that our use of multiple (often conflicting) models to understand the world, and of the problems this can lead philosophers to, see Lakoff and Johnson, Philosophy in the Flesh.
11 His instrumentalism is thus closer to the view developed by writers such as Nancy Cartwright in her How the Laws of Physics Lie (New York: OUP 1983), rather than the sort defended by the classical positivists or in Van Fraassen's The Scientific Image (New York: OUP 1980). James was not alone in his pessimism about current theories. Consider, for instance, the following remark from his contemporary, Henry Adams: "Forty years ago, our friends always explained things and had the cosmos down to a point, teste Darwin and Charles Lyell. Now they say that they don't believe that there is any explanation, or that you can choose between half a dozen, all correct." (Letters of Henry Adams, Boston, 1938, II 407-8). For a discussion of James' instrumentalism, see my Review of Pihlström's Structuring the World: The Issue of Realism and the Nature of Ontological Problems in Classical and Contemporary Pragmatism." (Transactions of the C.S Peirce Society, Spring 1998, Vol. XXXIV, No. 2).
12 PU 111. This is another respect in which James anticipates some of the claims about the metaphorical character of cognition worked out in more detail in Lakoff and Johnson's Metaphors we Live By (Chicago: University of Chicago Press, 1980) and their Philosophy in the Flesh.
13 James claims that our conception of 'the whole world' will inevitably have such a character: "We can invent no new forms of conception, applicable to the whole exclusively, and not suggested originally by the parts. All philosophers, accordingly, have conceived of the whole world after the analogy of some particular feature of it which has particularly captured his attention. Thus, the theists take their cue from manufacture, the pantheists from growth." (PU 9)
14 Treating ideas as objects is a notorious case of this, and the fact that we typically conceptualize experience in terms of concrete bounded objects is part of the reason why James thinks that it will be so difficult (if not impossible) for us to come up with adequate conceptualizations for phenomena which are not 'static' (SPP 51, 54-5). For further examples of how the uncritical acceptance of our metaphors can lead us astray, see Lakoff and Johnson, Metaphors We Live By, and Philosophy in the Flesh.
15 PU 32. See also PP 36, 52. James may not be entirely consistent with his use of this term, and for a discussion of the many senses of "intellectualism" in James, see Gale, The Divided Self of William James (New York: Cambridge University Press, 1999), 294-5.
16 SPP 47. This might seem like a stretch to some, but it is entirely natural if viewed as a consequence of the then prevalent idea that all categorization is in terms of sets of necessary and sufficient conditions. If categories did really work this way, then all concepts would at least involve 'implicit' definitions in terms of the necessary and sufficient conditions that they embody. For a discussion of the popularity of this conception of concepts and categorization, and a criticism of its empirical accuracy, see Lakoff, G. Women, Fire and Dangerous Things, Chicago, University of Chicago Press, 1987.
17 PU 99. See also: "The conceptual order into which we translate our experience seems not only a means of practical adaptation, but the revelation of a deeper level of reality in things. Being more constant, it is truer, less illusory, than the perceptual order, and ought to command our attention more." (SPP 42)
18 PU 101. See also, "Logic can extract all its possible consequences from any definition, and the logician is often tempted, when he cannot extract a certain property from a definition, to deny that the concrete object to which the definition applies can possibly possess that property." (PU 99)
19 An assumption that seems present in Plato, Aristotle, the Medievals, Kant and many contemporary philosophers as well. Our perception of our concepts' structure was not always be understood as clear, but it was still presupposed that a clear perception of our concepts would also provide one with an understanding of the structure of the world.
20 "It is but the old story, of a useful practice first becoming a method, then a habit, and finally a tyranny that defeats the end it was used for. Concepts, first employed to make things intelligible, are clung to even when they make then unintelligible." (PU 99.)
21 This popular reading remains, nevertheless, a mistaken one. For a discussion of this, see my "Prudential Arguments, Naturalized Epistemology, and the Will to Believe" and "James' Pragmatic Account of Intentionality and Truth."
22 An observation that leads him to remark that: "to find a world rational to a maximal degree in all these respects simultaneously is no easy matter. The rationality we gain in one coin we thus pay for in another; and the problem accordingly seems at first sight to resolve itself into that of getting a conception which will yield the largest balance of rationality rather than one which will yield perfect rationality of every description" (PU 54-5).
23 Though still a concession that many commentators have been willing to make on his behalf. For a discussion of this, see my "James' Pragmatic Account of Intentionality and Truth", and "Prudential Arguments, Naturalized Epistemology, and the Will to Believe."
24 Which is not to say that one couldn't draw a distinction between logic and rationality in this way. Prudential and truth-directed rationality need not always give the same advice about, say, what to believe. Nevertheless, while James is clearly aware of the importance of prudential rationality, it is not prudential rationality that he is talking about when he contrasts logic and rationality.
25 For instance, "Few philosophers have had the frankness fairly to admit the necessity of choosing between the 'horns' offered. Reality must be rational, they have said, and since the ordinary intellectualistic logic is the only usual test for reality, reality and logic must agree 'somehow'." (PU 96.)
26 PU 96. James further claims, "I must squarely confess that the solution to the problem impossible, and then either give up my intellectualistic logic, the logic of identity, and adopt some higher (or lower) form of rationality, or, finally, face the fact that life is logically irrational" (PU 95), see also PU 108-9.
27 Ethical concepts may be a good example of this. Ethical inferences that work 'locally' lead to trouble and contradictions when applied 'globally.' Deontological and consequentialist 'moves' typically agree with our intuitions and each other in everyday cases, but quickly come into conflict if expected to provide a true theoretical description of 'ethical reality.'
28 "[T]he immediate facts don't sound at all, but simply are, until we conceptualized name them vocally, the contradiction results only from the conceptual or discursive form being substituted for the real form. But if, as Bergson shows, that form is superimposed for practical ends only, in order to let us jump about over life instead of wading through it; and if it cannot even pretend to reveal anything of what life's inner nature is or ought to be; why then we can turn a deaf ear to its accusations." (PU 121, italics mine.)
29 Though James is pessimistic about our ever finding a set of concepts that would capture aspects of reality such as time and change (See SPP 51, 54-5). For a discussion of this, see Gale, The Divided Self of William James.
30 This is a familiar line of thought in the work of Khun and Feyerabend. See, for instance Feyerabend, P.K. Against Method. Thetford: Thetford Press Limited (1975) and Kuhn, T. The Structure of Scientific Revolutions, Chicago: The University of Chicago Press (1962). One might also usefully compare James' position to that proposed by Mark Wilson in his "Can We Trust Logical Form?" (JPhil XCI, no. 10, 1994). Wilson distinguishes the apparent semantics of a language, the implicit picture that emerges from conventional linguistic training of how the units of language correlate to the world and the logical relations between then, with the working semantics of the language. The working semantics emerges later as "agents discover that inferential pathways validated in their original apparent semantics sometimes lead to unhappy results, whereas other, officially unsanctioned, deductions generally lead to success" (520). The apparent semantics are the models we inherit, the working semantics are those that emerge from new experience, not just a priori reflection on the apparent semantics.
31 Indeed, James would argue that we have good reason to believe (relating to the incompatibility of the various models) that they don't.
32 Compare Wilson, "Can We Trust Logical Form?", 530.
33 For a discussion of this program and its ambitions, see, once again, the fourth chapter of Rorty's Philosophy and the Mirror of Nature.
34 This suspect assumption is something like what Wilson has referred to as the "the moral imperative of first-order logic." ( "Can We Trust Logical Form?" 527)