THE NORMS OF THE MENTAL

Pascal Engel

in The philosophy of Donald Davidson, Library of Living Philosophers, ed. L. Hahn 1999, 447-459

 

 

            A central tenet of Davidson’s philosophy is that there is an “ irreducibly normative character”  in the concepts that we use to describe and explain thought and meaning, in the sense that ascriptions of intentional contents are governed by a “constitutive ideal of rationality” which has no “echo” in the realm of natural and physical facts. [1] This view plays a prominent role in his argument for anomalous monism. It is also essential to his theory of the interpretation of speech and language.[2]

But what does it mean to say that “norms enter in the study of mental phenomena”, that there is a “normative element” or that there are “ normative properties” in mental concepts[3] ? I would like to examine here what seems to be a striking feature of Davidson’s account of such norms : that they are, in a certain sense, elusive, and cannot be completely and precisely  spelled out. I shall try here to say what this normative features consist in.

             That there is a “normative character” to thought and meaning, or that norms “enter” in our mental concepts is often misunderstood, and has been for this reason a source of puzzlement : for in what sense could thoughts and meaning imply any prescriptions about what we ought to think or do, or about what it is valuable to think or do, as the word “normative” seems to imply? But the “normativity” of meaning and thought does not mean that there are norms of mental in the sense in which one can say that there are rules of etiquette, social norms of even linguistic rules. Davidson is quite clear on the fact that he rejects accounts of language and communication which rely upon such notions as those of rule, convention, or social norms in the sense that speakers would have, to make themselves understood or to understand others, to grasp such rules or to follow them. So “norm”, in his sense, is not to be understood as involving the existence particular  rules attached either to words or to concepts which would determine the correct use of these words or concepts. Meaning and thought are not “normative” if this means that using a word is to follow a certain prescription or rule.[4]  Nor does it imply normativity in the sense of giving a particular value to rationality. The rational norms are there, whether we like them or not, and in this sense they are not good or bad.  The normativity of meaning and thought pertains neither to particular rules nor to particular values, but to general principles of the interpretation of speech and thought. These are not, unlike particular rules, a matter of choice: they are compulsory because they are  a priori  requirements of the very task of making sense of others. Among these figure prominently the Principle of Charity -  that one could not interpret someone else if one did not suppose that most of the beliefs that we hold true by our lights are true for him, and that one could not understand someone’s beliefs and desires unless we could find in them an overall rational or coherent structure[5] , and in general the holistic requirements about thought and language.

            Davidson sometimes talks as if these general norms of interpretative charity were codified in the normative theories that we currently use: logic, decision theory, or general principles or laws used by such theories: the law of non contradiction, or the principle of maximizing expected utility . Sometimes he also alludes to specific epistemic principles, such as Carnap’s requirement of  “total evidence” for inductive reasoning (“give your credence to the hypothesis supported by all available evidence”) or the counterpart principle for practical reasoning (“perform the action judged best on the basis of all available relevant reasons”).[6] In such cases, these normative principles are indeed imperatives of a particular form, or something like rules for guiding beliefs or actions. But it is quite important here to see how general and unspecific such principles are and why they have to be so, for a number of critics have complained that they are too unspecific to serve as actual principles of interpretation: if most of someone’s beliefs are to be true, it does not tell us, as interpreters, which, and if an agent is to be  by and large  rational does not tell us how. But this is just as it should be:

            “The issue is not whether we all agree on exactly what the norms of rationality are; the point is rather that we all have such norms, and that we cannot recognize as thought phenomena that are too far out of the line. Better say: what is too far out of line is not thought. It is only when we can see a creature (or ‘object’) as largely rational by our own lights that we can intelligibly ascribe thought to it at all, or explain its behaviour by reference to its ends and convictions. Anyone¼ when he ascribes thoughts to others, necessarily employs his own norms in making the ascriptions. There is no way he can check whether his norms are shared by someone else without first assuming that in large part they are; to the extent that he successfully interprets someone else, he will have discovered his own norms (nearly enough) in that person.”[7]

 

            In this sense, it wrong to suppose that the general norms of interpretation             can be transcendent norms which have a universal validity, and which could be applied like models or idealized principles in science. For to take them as such would be to mistake them for general laws or regularities descriptive  of phenomena. For instance it is often said that the principles of rational choice making involved in Bayesian decision theories

are “normative” in the sense that they give us a picture of the perfectly rational agent, although they are not true of individual agents at the “descriptive” level, because they are oversimplified. As Davidson points out, this way of making the distinction between the normative and the descriptive is largely illusory: “Until a detailed empirical interpretation is given to a theory, it is impossible to tell whether or not an agent satisfies its norms; indeed without a clear interpretation, it is hard to say what content the theory, whether normative or descriptive, has.”[8] If the rational norms governing our mental concepts  had the form of specific, although in some sense idealized, principles such as, for instance, the Bayesian principle of maximization of expected utility, failures to apply these principles to human agents would imply clear failures of rationality on their part. But Davidson is quite clear that it is not easy to pin down empirically such failures, for rationality in general is a presupposition of any interpretation of human behaviour, and so cannot be tested at a purely descriptive level.

            The point can be illustrated though many examples of empirical work on the psychology of decision, within the framework of research pioneered by Davidson and his associates at Stanford during the fifties.[9] Here is one example which, as far as I know, is not used by Davidson, but which he could have used. Allais’ paradox seems to be a clear case where most agents violate the Bayesian norm of maximizing expected utilities (or, for that matter the “sure-thing principle”): where they are presented pairs of choices such that they should - if they respected the principle - choose choice a  over choice b , and choice c  over choice d , they reverse their preference in the second case (choosing d  over c). Discussing this example, the statistician Leonard Savage does not conclude that  agents are irrational. He tries to provide a point of view from which their mistake is understandable, reformulating the problem through a lottery with the choice between a corresponding number of tickets, and looking for a perspective which, in his own terms, “has a claim to universality”, where in the end the reversal of preferences disappears. This, comments Savage, is just what happens when someone who buys a car for a large sum of money is tempted to order it with a radio installed, finding the difference trifling, but who, when he reflects that if he already had the car, he would not spend the amount for a radio on it.[10] Even if the agent violates the Bayesian norm, understanding him involves the interpreter to find another universal and objective norm to which the agent can be responsive. Here the interpreter does not presuppose any transcendent norm; he tries to construct one by agreement with the agent. This emphasis on the fact that the norms of the mental are primarily those of the interpreter also explains why Davidson, after having formulated, in his first writings, the Principle of Charity as a principle of maximization of agreement, has finally preferred to talk of a maximization of understanding: the point is not to make those that we interpret as intelligent and as correct in their beliefs as possible, but to minimize unintelligible error.[11] This point is clearly emphasized in “A Nice derangement of Epitaphs” about “passing theories”, the theories that an interpreter formulates at a time for a given speaker:

            “ There are no rules for arriving at passing theories, no rules in any strict sense, as opposed to rough maxims and methodological generalities. A passing theory is like a theory at least in this, that it is derived by wit, luck, and wisdom from a private vocabulary and grammar, knowledge of the ways people get their points across, and rules of thumb for figuring out what deviations from the dictionary are most likely.”[12]

 

            If the norms of rationality which govern the interpretation of speech and action are so general and have such a high profile that they cannot consist in particular specific maxims, rules, or guides for actual interpretation, it seems to follow that the very notion of rationality cannot be, in McDowell’s phrase, “codified” in any formulas or set of principles from which one could derive strict prediction about the behaviour or thought of agents. [13] If there were such principles of formulas, they could be applied, like strict laws, to any particular case, and they could tell us what to do or think in a particular situation. But even the best entrenched logical or decision theoretic principles do not tell us what  we ought to believe or do in a particular case, or whether it is rational or not to do something in a given situation.[14] There is always a slack between the formulas and their application, which can only be remedied by guesswork or by what Aristotle called phronèsis. The case can, once again, be illustrated with a decision theoretic example. Bayesian Decision theorists claim that coherence in the probabilistic sense – non violation of  the axioms of the probability calculus - is a necessary condition of rationality; for otherwise a clever bookmaker could make a “Dutch book” against one such that the agent would systematically loose money on bets involving his beliefs. But it does not follow that an individual who is incoherent by this criterion – whose degree of belief violate the axioms of the probability calculus – is irrational. All that follows is that it is irrational to accept bets at odds that reflect one’s degrees of beliefs when these degrees are incoherent. An agent might have good reasons not to accept such bets, just as he might have good reasons to accept them, without those reasons being epistemic ones.[15]

            This “uncodifiability of rationality” can be evaluated also from other angles, which, according to Davidson, are intrinsically tied to the normative character of our mental concepts .  One is the causal nature of these concepts, which is “built into the concept of acting for a reason”: a reason is “a rational cause”, which can be spelled out as “a belief and a desire in the light of which the action is desirable”.[16]  But, as Davidson notoriously pointed out about examples of “deviant causal chains”, where an agent performs an act in conformity with his intention, but not in the “right” way, there is no automatic specification of the conditions in which the beliefs and the desires are “appropriate” for making the action rational. “What I despair of spelling out, says Davidson, is the way in which attitudes must cause actions if they are to rationalize the action.” [17] And he actually seems to imply that there is no such way. The second angle from which the uncodifiable nature of the norms of rationality is the holistic character of the rational patterns which must be presupposed to make an agent intelligible: the open-ended nature of the intentional contents which must be attributed is the mark of their rationality, but there is no clear limit to the extent of the pattern in question.  Finally the normative character of mental concepts is closely tied to the externalist features of mental contents: because contents cannot be individuated internally, but by reference to a variety of causal factors which cannot be systematized under laws. As Davidson says:

            “What I think is certain is that holism, externalism, and the normative feature of the mental stand or fall together. ¼There can be not serious science of the mental. I believe the normative, holistic, and externalist elements in psychological concepts cannot be eliminated without radically changing the subject.”[18]

 

            The distinctive character of our mental concepts is not simply that they are subject to norms. For there are also norms for our physical concepts, such as length or temperature, which Davidson once called  “constitutive or synthetic a priori”.[19] In physics too, there are norms of description and of explanation, for instance simplicity or explanatory power. What makes the normative element in the mental domain distinctive is that, unlike what happens in the physical case, where norms are employed about something which is essentially non minded, in the mental case “norms are being employed as standards of norms”.[20] We use norms to interpret patterns which are essentially normative. What does it mean? Here it seems to me that a comparison can be usefully made between Davidson’s and Wittgenstein’s views. For Wittgenstein there are rules of language, which record the use of words by speakers, and hence facts or regularities about these speakers -  for instance that speakers of English usually mean cats by using the word cat. But there are no further facts about what makes the rules correct. There is a correct application of the rules (Wittgenstein is not a “skeptic” about the existence of rules), but what makes it correct is not a special kind of fact (dispositional, psychological, or some platonistic super-fact).  Davidson, as I pointed out above, feels no attraction in talk of rules and of specific norms of language. But his view about the status of norms of rationality with respect to facts is quite similar to Wittgenstein’s. For Davidson, there is a correct way (in spite of the unavoidable amount of indeterminacy) of interpreting a speaker and of making what he thinks or does intelligible, but there is no fact about what makes this interpretation correct, for there is no way of codifying the rational patterns that one has to attribute to a speaker in a particular circumstance. Wittgenstein and Davidson are, among the philosophers of this century, those who have most insisted on the existence of a particular kind of norms, distinct from the usual kinds of norms - practical, epistemic or aesthetic – the interpretative norms. They both hold that such norms are distinctive, and that they cannot be reduced to facts.

            If the foregoing account of Davidson’s conception of the norms of rationality is correct, we should not expect any definition, nor any sort of specification of these norms. This is not, of course to deny that there are such standards, and that we can formulate them, for instance as principles of logic and of decision theory. The rules of logic are indeed those on which a theory of truth in Tarski’s style must rely, and the standards of Bayesian theory are those on which Davidson’s “unified theory” of meaning and action rests. But they offer no definition  of rationality in general, just as a Tarskian theory of truth offers no general definition of the concept of truth. Just as we can think of truth as Ramsey thought about probability, by setting up a testable method of measurement of degrees of beliefs, and by defining the conditions under which a pattern  of preferences can be “rational”, we can think of the norms of rationality themselves as the undefined structures which a theory of interpretation must rely upon.[21] Thus, although we cannot say what the norms are, we can say what they do.

           

            I suscribe to Davidson’s view about the necessary normative character of mental concepts, and I agree with him that it is the feature which stands in the way of a complete science of the mental as well as in the way of attempts to reduce what Sellars calls the “space of reasons” to the space of natural laws and physical concepts. But the thesis that these norms cannot be codified nor defined is not unproblematic. I can see at least three kinds of difficulties (although they are related). First, if the standards of rationality used by an interpreter to make ascriptions of mental contents are not transcendent norms fixed in advance, but his “own norms” which he can only supposed to be shared by others, don’t we run the risk of a certain amount of subjectivism in the interpretation procedure? Davidson’s account of interpretation certainly allows, and indeed requires, a large amount of convergence at the end of the process just as at the beginning, but how is this convergence acquired? Second, if we suppose that the interpreter is able, even by his own lights, to discern successfully rational from irrational patterns of thought or behaviour, where does he get this ability from? This problem is similar to a problem which has often been raised about the knowledge of meaning that the Davidsonian interpreter must have in order to frame his ascriptions of held-true sentences to others. For a central feature of  Davidson’s conception of interpretation is that the interpreters-speakers must know what they mean by their words, without interpreting themselves, in order to interpreter others and ascribe to them thoughts and meanings which are true by their own lights. But how do speakers know what they mean in the first place?[22] Why do they (rightly) think themselves authoritative about their own norms of rationality just as they (rightly) think themselves authoritative about the meanings of their own words? Even if the interpreter’s ability for discerning rational patterns is not strictly codified, it must come from somewhere, and it must have a certain shape. It cannot have been acquired by magic.[23] The third difficulty has to do with the move taken by Davidson from a formulation of the Principle of Charity as a principle of maximization of truth and rationality to a formulation in terms of a principle or minimization of unintelligible error: if the rational and true beliefs and desires ascribed by an interpreter must also be intelligible or explainable  (in the interpreter’s light) rational and true beliefs (so that it would be pointless to ascribe a vast majority of inexplicably correct beliefs), then it seems, if we do not want to say that there are no constraints at all upon such ascriptions, that there must be more particular constraints about what kind of contents are ascribable. When Davidson talks about the holistic character of interpretation, he always seems to have in mind very general (mostly logical) kinds of links within content, and does not mention the more particular or more local links that some contents might have; for instance, when a speaker can be interpreted as having thoughts about colours, or about shapes of objects, it may be presumed that he has some views about incompatibilities of colours, or about relationships between observational concepts of shape and colour. Making an agent intelligible must, at some point involve seeing the reasons why he holds a certain belief true, or his justification for holding-true certain sentences. But Davidson’s methodology of interpretation does not seem to give us more than a pattern of sentences held-true, without any hint about what might justify a speaker to hold certain kinds of sentences true. This is in perfect agreement with his denial that truth could be equated with the grounds, or the warranted assertability of sentences: what matters are the truth-conditions of sentences, not their assertion conditions.[24] Now, even if we grant that point, we might try to find less global constraints than those that Davidson envisages.

            Some writers, such as McDowell, have be tempted into reading into Davidson’s position “the irreducible subjectivity of propositional attitudes” and the impossibility of “a distinction between what makes sense and what could come to make sense to us”.[25] But Davidson’s denial that rationality is codifiable need not imply that it is not objective, and there cannot be a convergence in the norms that speakers use to interpret each other. Indeed there must, on his view, be such a convergence, since it there from the start. The way it is further reached is a matter of the extension of what he calls the basic process of triangulation upon which objective thought rests[26]. It is enough that the possibility of converge is not ruled out  a priori, and by definition it is not. So it seems to me to be a very misleading way of interpreting his views to read them in the way McDowell does.

            It does not seem to me, however, that one can answer in the same way the second and the third worry expressed above. The worry can be expressed again thus: given that the rationality requirements on interpretation are so general, and the interpreter has to rely upon his own sense of what is rational, how can he read the rational patterns in a particular case?  This is tied to the problem of holism. There are two kinds of cases to distinguish: holistic patterns within a particular attitude, say belief, and holistic patterns between particular attitudes.[27] I shall say more about the second than the first. The first kind of patterns are patterns such as those: in order for one to have beliefs about rain, one must have beliefs about clouds, about the condensation of drops in water saturated air, about wetness, etc. If we set aside the problem posed by the “etc.”, which has been at the center of many discussions about holism[28],  the holistic pattern within these beliefs is due to the way their components are related, and these components are concepts. Davidson’s account of interpretation, as well as Quine’s, tells us that in order to find basic rational patterns among beliefs, we shall have to rely on basic logical concepts, such as those of negation, conjunction, quantifcation and the like, but it tells us very little about other kinds of patterns within various types of concepts, for instance observational concepts (“wet”) vs theoretical concepts (“condensation”), demonstrative concepts (“this rain”) vs general descriptive concepts (“rain”). It is, however, legitimate to suppose that when an interpreter ascribes a particular kind of belief, involving a particular kind of concept, say an observational one, he takes it that there certain reasons which subjects typically have for having beliefs of such a kind, and that these reasons enter into what makes the ascription intelligible to the interpreter. In other words, what I want to suggest is that an interpreter, if he must use his “own norms” for ascribing contents, must not simply find a pattern of implications between sentences held-true, but must have a certain conception of the conditions of warrant or justification of such sentences, when they contain particular concepts. In other terms, certain norms must be presupposed about what justifies a certain kind of content, what justifies its rejection, or what inferences can be made about it. This line of thought might lead us to try to give accounts of particular concepts along the line of various forms of what is called “conceptual role semantics” or theories of “possession conditions” of concepts.[29] They might also help us to answer the objection voiced above, that the abilities of interpreters have to come from somewhere, for if mastery of a concept is a certain kind of ability or aptitude, the spelling out of the conditions under which the concepts are possessed should alos tell us what kind of abilities they presuppose. Such accounts should be described much more precisely, but I suspect that Davidson would find such them uncongenial for his own views, since they would have to assume much more individuation of content than he wants to assume in his theory of interpretation – the only kind of attitude assumed being that of holding certain sentences true. I think, nevertheless, that an account of the specific norms involved in the possession an the ascription of certain concepts can be extracted from his views.

            For that we can consider the other kind of holism mentioned above, between attitudes. Beliefs cannot be ascribed without ascribing other attitudes, such as desires and intentions, and intentions themselves cannot be ascribed unless desires and beliefs also ascribed. So an interpret must not simply have an idea of the rational pattern within one kind of attitude, but must also have an idea of the relationships between attitudes themselves. If the preceding line of thought is correct, he must have some reasons to ascribe contents which are belief-like, and be able to distinguish them from those which are desire-like, regret-like, and so on. In his essays on animal thought, Davidson goes further: famously, he holds that the very having of a belief must presuppose the having of the concept of belief.[30] If we set aside the problem of animal thought, what does it mean to “have the concept of belief”? In one sense of that term, it means being able to have beliefs about one’s beliefs, or to have reflexive beliefs. But in another, and as important sense, it means being able to identify a particular kind of state as being a belief. Now, the concept of belief is the concept of a state which is apt to be true or false, or, to use a familiar phrase, which “aims at truth”. Nothing, in this sense, can be a belief if it is not supposed to be true. According to this criterion, beliefs have a different “direction of fit” than desires: beliefs are states such that they should fit the world, whereas desires are such that the world should fit them. Now what I want to suggest is that this internal feature of the concept of belief, the fact that our beliefs “aim” at truth, is part of the “normative character” of this very concept. That beliefs are true or false according to whether their contents are true or false is part of why they are correct or not . The other part is that beliefs are justified according to whether the reasons for holding them are good or not. These are, indeed, platitudes, but they are integral to the concept of belief. The concept of belief is normative because it involves these basic norms of aiming at the truth and the justification of the contents of the corresponding state. Now, when Davidson tells us that belief is “by its very nature veridical” or that “most of the point of the concept of belief is the potential gaps it introduces between what is held true and what is true”[31], or that having the concept of belief is to have the contrast between subjective and objective, or to have the idea of an independent reality which is independent of my beliefs[32], he seems to me to point just to this feature. This feature of beliefs is also integral to the basic choice of Davidson’s interpretation procedure, to choose a particular attitude of taking certain sentences as true. If this is correct, then we have here an example of a particular kind of norm attached to a particular kind of concept. Indeed, this is not any sort of concept, for belief is central to most attitudes and concept formation. My suggestion is that we should try to give other similar normative conditions for various concepts.

            We could also express the foregoing remarks thus: truth is the internal goal of belief. This, of course, does not mean that there is any sort of intention or teleological character in this, but that truth is the norm, or the standard of appraisal of belief. This conclusion will be unwelcome by those who hold that the concept of truth is a concept which is necessarily extremely thin and “minimal” in its content, and Davidson is certainly a member of this family of theorists, although his views are specific. Davidson says that, although he does not agree with a “deflationist “ reading of truth, according to which there is nothing more to truth than the schema “ ‘p’ is true if and only if p “, still it is a “folly” to try to define this concept. He claims that the concept can be shown to be explanatory by the use to which it can be but in a theory of interpretation.[33] But nothing that he says seems to me to be inconsistent with the following. Since aiming at truth is a norm of belief, the activities in which we are engaged when we try to assess our beliefs, the activities of inquiring, justifying and giving reasons for these beliefs, are regulated by this norm. Interpretation would be impossible if it were not supposed that subject do not recognize this norm, which is thus part of the concept of truth. It does not follow that the norm is a convention of truthfullness, agreed to by the members of a community either tacitly or explicitly.[34] It is just that their attitudes could not be interpreted as attitudes of beliefs if they did not have this internal goal. If belief is thus central to our mental attitudes, then truth is indeed a norm of belief, and a norm – the basic norm -  of the mental.

            It is sometimes said that treating truth as a norm, as I suggest here, would imply some sort of appraisal, or some commandment to seek the truth.[35] It certainly implies no moral or ethical norm of the kind. Neither does it imply that truth is an essentially epistemic concept . It only implies that the norm of truth is among what I have called above interpretative norms. If to understand others is also to make intelligible their false of irrational beliefs, intelligibility requires grasp of the concept of truth. We can express it in the manner of John Donne: “Though truth and falsehood bee neare twins, yet truth a little older is.” (Satyre, II)

 

                       


REFERENCES

 

Bilgrami, A. 1992 Belief and Meaning, Blackwell, Oxford

Brandom, R. 1994 Making it Explicit, Harvard: Harvard University Press

Child, W. Causality, Interpretation and the Mind, Oxford: Oxford University Press

Davidson, D. 1980 Essays on Actions and Events, Oxford : Oxford University PressDavidson, D. 1980 Essays on Actions and Events, Oxford : Oxford University Press

                1984 Inquiries into Truth and Interpretation, Oxford: Oxford University Press        1984 Inquiries into Truth and Interpretation, Oxford: Oxford University Press

                 1985 “A New Basis for Decision Theory”, Theory and Decision, 18, 87-98

                 1986 “A Nice Derangement of Epitaphs”, in Le Pore 1986:433-440

                        “The Structure and Content of Truth”, Journal of Philosophy,

                        LXXXVII, 6, 279-326

                  1990 a “Representation and Interpretation”, in K.A. Mohyeldin Said,

                 W.Newton-Smith, R.Viale & K. Wilkes, Modelling the Mind, Oxford:

                 Oxford University Press, 13-26

                 1991 “Three varieties of Knowledge”, in A.Philipps Griffiths, A.J. Ayer,

                     Memorial Essays, Cambridge: Cambridge University Press, 153-166

               “The measure of the Mental”, ms, tr.fr. “La mesure du mental”, in

                        Engel 1994, 31-49

              1995 The Jean Nicod Lectures, Paris, unpublished

               1996 “The Folly of Trying to Define Truth”, Journal of Philosophy,

                                XCIII, 6, 263-278

Engel, P.  ed. 1994 Lire Davidson,Combas : L’Eclat

Foley, R. 1993 Working without a net , Oxford: Oxford University Press

Jackson, F. & Braddon-Mitchell D. 1996 Philosophy of Mind and Cognition, Oxford:

                                Blackwell

Le Pore, E. ed. 1985 Actions and Events: Perspectives on the Philosophy of Donald

                             Davidson, Oxford: Blackwell

                              1986 Truth and interpretation, : Perspectives on the Philosophy of

                                 Donald  Davidson, Oxford: Blackwell

Lewis, D. 1974  “Radical Interpretation”, Synthese, 23: 331-44

McDowell, J. 1979 “Virtue and Reason”, The Monist, 62, 1979, 331-350 repr. In McDowell 1998

                        1986 “Functionalism and Anomalous Monism”, in Le Pore 1985,387-398,

                           repr. in McDowell 1998

                       1998 Mind, Value and Reality, Harvard: Harvard University Press

Peacocke, C. 1992 A Study of Concepts, Cambridge Mass. : MIT Press

Rorty, R. “Is Truth a goal of Inquiry? Davidson vs Wright”, Philosophical Quaterly,

                vol. 45, N° 180, 281-300

Savage , L.J. 1954 The Foundations of Statistics,New York: Wiley (quoted after

                    the 1972 edition, New York: Dover Books)

Skorupski, J. 1990 “Explanation and understanding in the social Sciences”, in D. Knowles ed.

                        Explanation, Cambridge: Cambridge University Press, 119-134

Smith, B.C.  1998 On Knowing One’s Own Language”, in C. Wright, B.C. Smith & C.

                        McDonald, eds, Knowing our own Minds, Oxford, Oxford University,

                        Press

 

 



[1].  Davidson 1990a p.25 ,  1970, p. 223

[2]. See for instance Davidson 1990, p. 325 : What makes the task [ of providing a unified theory of thought, desire, and speech] practiciable at all is the structure the normative character of thought, desire, speech, and action imposes on correct attributions of attitudes to others, and hence to interpretations of attitudes to others, and hence on interpretations of speech and explanations of their actions ” . See also Davidson 1995, chapter 5 : “ The entire structure of the theory depends on the standards and norms of rationality ”

 

[3]. Davidson 1990 : 25 ; Davidson 1991 : 162

[4]. see in particular “Communication and Convention” in Davidson 1984, and Davidson 1986. A. Bilgrami (1992) is quite clear on the distinction between a “localist” and a “generalist” view of norms in their relevance to meaning and intentionality.

[5]. The first principle is sometimes called by Davidson the principle of Correspondence and the second the principle of Coherence (e.g. Davidson 1991 : 158)

[6].  “How is weakness of the Will possible” (169)  in Davidson 1980, p.41

[7]. Davidson 1990a, p.25

[8]. Davidson, 1985, p.89

[9].  see in particular his essays “Hempel on Explaining Action”, and “Psychology as philosophy” in Davdison 1980.

[10]. Leonard J. Savage 1954: 103-104 (The relevance of this example for considerations on normativity  in interpretation is pointed out by Skorupski 1990)

[11]. Davidson 1984 : xvii. Some writers have held that this amounts to a redefinition of the Principle of Charity as a Principle of “Humanity”, but I will not go into that.

[12]. Davidson 1986, p.446

[13]. Mc Dowell, 1979: 336 (repr. in Mc Dowell 1997:57-58). This view is  well put by Child 1994: 57-68 to whom I borrow the phrase “uncodifiability of rationality.”

[14].  This could be said to one of the numerous lessons of Lewis Carroll’s tale of Achilles and the Tortoise: Achilles tries to force the Tortoise to recognize the validity and applicability of a logical principle (the moduls ponens) to a particular case, but the Tortoise always refuses to see it applied in this case.  

[15]. See for instance, for further considerations on this point, Foley 1993 : 155-162

[16]. “Psychology as philosophy” (1974) , in Davidson 1980 : 233

[17]. “Freedom to Act”, in Davidson 1980 : 79

[18]. Nicod Lecture, lecture 5, see also Davidson 1996

[19]. Mental Events (1970), in Davidson 1980: 221

[20]. Nicod Lectures, V.; see also Davidson 1996.

[21]. Davidson 1996: 277-278

[22]. The feature is emphasized by Davidson 1987, and the problem well put by Barry C. Smith 1998: 415

[23]. This point is pressed by F. Jackson and D. Braddon-Mitchell 1996: 155-156

[24]. See, for instance, Davidson 1990: 307-308

[25]. Mc Dowell 1986: 396.

[26].  see in particular Davidson 1991; 1995.

[27].  In his 1995 Nicod Lectures(lecture 1) , Davidson calls them respectively “intra-attitudinal” and “inter-attitudinal” aspects of holism. the example about rain below is his.

[28]. Davidson is obviously not committed to holding that the “etc” means that all the other beliefs of a person have to be related to a single one.

[29]. For  account of the first kind , see R. Brandom 1994; for an account of the second kind, see Peacocke 1992

[30]. In particular “Thought and Talk” (1975) in Davidson 1984, and “Rational Animals” (1982), repr. in Le Pore,

 1985: 473-480

[31]. “A Coherence Theory of Truth and Knoweldge”(1986), in Le Pore 1986: 308

[32]. “Rational Animals”, ibid. p.480

[33]. Davidson 1996

[34]. David Lewis, in his theory of convention and of radical interpretation, has invoked such a convention.

see his (1974). But there is no need here to think of it as a convention.

[35][35]. Richard Rorty  (1995) suggests that taking truth as a norm implies this, and proposes his own deflationist version against Wright (1992) who advocated, in the same sense as here, such a view. He believes that such a view would imply what Davidson deplore, an attempt to “humanize” truth. My point is simply that we would loose the point of the notion of truth if we could not link it to the aims of our belief.