in: A House Divided: Comparing Analytic and Continental Philosophers, ed. by Carlos Prado, Amherst: Humanities Press 2003, pp. 235-258. (volume reviewed in: The Journal of Speculative Philosophy, Vol 18, no. 1 (2004). [online text])
Mike Sandbothe
Davidson and Rorty on Truth
Reshaping Analytic Philosophy for a Transcontinental Conversation1
In his essay “Analytic and Conversational Philosophy” included in this volume, Richard Rorty suggests that we replace the older distinction between analytic and continental philosophy with a new distinction between analytic and conversational philosophy. Rorty presents his replacement suggestion as a performative example of the metaphilosophical position characteristic of conversational philosophers. According to them, professional philosophers should stop trying to put the discipline on the secure path of a science and contribute instead to renewing our vocabulary to make it able to react in a more timely fashion to the challenges posed today by the arts, the sciences and politics.
In Rorty’s view, so-called continental philosophy is more inclined to open itself to such a purpose. In contrast to most representatives of analytic philosophy, continental philosophers in Europe and the United States appreciate the value of historical narrative, reconstructing the history of past changes in human thought and action. In another essay, Rorty offers the following future-oriented interpretation of the continental-style teaching typical in the Humanities: “By telling stories about past transformative encounters, members of these departments hope to put students in a better position to have similar encounters of their own, encounters some of which may help shove the World-Spirit along.”2 And in the same text Rorty comes to the conclusion that, “For all its pseudo-scientistic pretensions, and despite the countless dead ends it has backed itself into, twentieth-century analytic philosophy will also have transformative effects, and so will put our descendants in its debt.”3 In what follows I would like to take a look at the more recent history of analytic philosophy in order to point out the ways that history could potentially change our use of the concept of truth.
Since the nineteen-seventies Donald Davidson and Richard Rorty have been carrying on a widely regarded debate over the significance of the concept of truth for contemporary philosophy.4 Although this debate itself has to a large extent been carried out with the technical means of analytic philosophy, I offer my reconstruction of it in the narrative style typical of continental philosophy. My reconstruction was nevertheless understood by the Engerer Kreis as a piece of analytic philosophy. From a strictly analytic point of view that is, of course, false, since what I am offering here is not an independent contribution to the solution of the problem under discussion, but rather a story that I am telling in order to see what lessons we can take from it to bring about future transformations of common sense.
Davidson and Rorty share a common point of departure in their debate, for they both agree in relinquishing the explanatory notion of truth. According to that notion, truth is to be understood as a non-causal and atemporal relation between linguistic statements and extra-linguistic reality, supposed to explain why a statement is capable of consensus or coherence. What is at issue between Davidson and Rorty is the conception of truth that should replace the explanatory notion and what such a replacement implies for the philosophical relevance of “truth”. - My reflections are divided into three parts. In the first part I reconstruct the central arguments on which Davidson and Rorty base their critique of the explanatory notion of truth. In the course of that reconstruction, their shared basic concerns as well as certain differences between the two authors will become apparent. The way in which those similarities and differences are reflected in Davidson’s and Rorty’s investigations into non-explanatory uses of “true” is the subject of the second part of my paper. The third and final part is then concerned with showing how the difference that ultimately remains between Davidson’s and Rorty’s theories of truth is to be related to the metaphilosophical distinction analytic-conversational as well as to the sociological distinction analytic-continental.
TOWARDS A CRITIQUE OF THE EXPLANATORY NOTION OF TRUTH
The starting point for the discussion is Davidson’s 1974 paper On the Very Idea of a Conceptual Scheme. The centerpiece of that text is Davidson’s criticism of the “dualism of scheme and content”5. Davidson sees this dualism as an unquestioned basic presupposition of modern philosophy and finds that it lacks an “intelligible basis”6 for two reasons. The first reason lies in the fact that philosophers from Kant to Quine thought of a conceptual scheme as an instrument for ordering or appropriating a raw material one either desired to structure conceptually or needed to cope with technologically and to which the scheme was then applied. The idea of such a raw material, Davidson goes on to argue, depends on “the concept of an uninterpreted reality”7 and this concept is problematic, in that it is by definition incapable of further explication. The second reason why Davidson criticizes the dualism of scheme and content has to do with his claim that the idea of a conceptual scheme is self-contradictory. The clarification of this point is somewhat complicated.
According to Davidson, philosophers who make use of the idea of a conceptual scheme assume implicitly or explicitly that there are other conceptual schemes, from whose perspective the conceptual scheme that happens to be at issue can be described. On this assumption we are bound to accept a plurality of different and perhaps even incommensurable conceptual schemes and this, Davidson continues, leads to an aporia, since it divorces the relativistically articulated truth of a conceptual scheme from its complete or partial intertranslatability with other conceptual schemes. This uncoupling of truth and translatability conflicts however with Tarski’s Convention T, which to Davidson’s mind “embodies our best intuition as to how the concept of truth is used”.8
Davidson is appealing here to Tarski’s basic truth-theoretical insight that a theory of truth for a formal language can be reconstructed by translating every relevant sentence of that language into a sentence in a metalanguage of the form, “S is true if and only if p”, where S is a sentence of the formal language and p its translation in the metalanguage. Davidson applies Tarski’s insight to natural languages and goes on to make it applicable within a single natural language serving as its own metalanguage.
Davidson’s argument against the idea of a conceptual scheme draws its strength from the fact that, according to Convention T, the theory of truth for a natural language cannot be derived in any other way than by translating into another natural language or into a metalinguistic utterance of that same natural language. On these grounds Davidson concludes that the notion of a conceptual scheme that could be both true and untranslatable is self-contradictory. Since for Davidson translatability and truth are inseparable, the notion of a conceptual scheme whose truth value could be determined solely on the basis of the content schematized by it is unintelligible and hence must be abandoned.
Davidson’s criticism of scheme-content dualism in On the Very Idea of a Conceptual Scheme forms the basis of the arguments against the explanatory notion of truth formulated in his 1983 article A Coherence Theory of Truth and Knowledge. Taken in its standard version as a correspondence theory, the explanatory conception of truth assumes truth to consist in a non-causal and atemporal relation of correspondence between the relata “scheme” and “content”. In summing up his argument against the correspondence theory’s explanatory claim, Davidson appeals to Rorty. “As Rorty has put it,” he writes, “‘nothing counts as justification unless by reference to something we already accept, and there is no way to get outside our beliefs and our language so as to find some test other than coherence.’”9
Nor does Davidson let it go at that, for he also attacks attempts made by certain varieties of coherence theory to supply the notion of truth with an explanatory function. His criticism is aimed at coherence theories that operate on the basis of a scheme-content dualism. Unlike Davidson’s own version, these forms of the coherence theory are concerned not with the coherence of beliefs held by a single person, but rather with the coherence of systems of statements as such. In a nutshell, the point of Davidson’s argument is that on these premises coherence is not a criterium of truth at all. For him, the link between coherence and truth does not come into play until one ceases to understand statements as sentences that actualize a conceptual scheme and begins to see them as expressions of beliefs articulated by a person who takes those beliefs to be true.
Now that I have extracted Davidson’s two main arguments in A Coherence Theory of Truth and Knowledge against the correspondence theory and against explanatory conceptions of truth based on coherence theories, respectively, I shall go on to consider Rorty. We have seen that Davidson looks to Rorty for support in his criticism of the explanatory claims raised by correspondence theories of truth. Rorty in turn takes up Davidson’s critical reflections in Davidson, Pragmatism, and Truth (1986) and places them in the context of pragmatist reflexions on the problem of truth he finds in the works of William James and Charles Sanders Peirce.
An historical reminder serves as Rorty’s point of departure. He points out that Davidson’s thesis that “one could not use truth as an explanatory concept”10 goes back to James, who argued for the dismissal of the explanatory notion of truth by saying that from the failure of two thousand years of philosophy “to discover (…) the microstructure of the correspondence relation” the moral should be drawn “that there was nothing there to find”11. Peirce, too, Rorty continues, was already moving in the direction of this insight, but only went half the way toward realizing it. For Rorty, Peirce’s “pragmatist theory of truth”12, which appeals to a counterfactual, ideal point of convergence in the discourse of scientific inquiry in order to define correspondence in terms of consensus, represents a “half-way house between idealist and physicalist theories of truth”13. For unlike idealists and physicalists Peirce had abandoned the notion that truth must be conceived of as a relation between ontologically homogenous terms. Peirce claimed instead “that the ‘about’ and ‘true of’ relations can link utterly disparate relata, and that problems of ontological homogeneity need not arise”.14
To Rorty’s mind, the immanent problem with the conception that Peirce put forward and Rorty himself tried for a while to further develop in a constructive theory lies in the fact that to him the expression “ideal” used by Peirce to describe the ultimate result of scientific inquiry is itself just as fishy as the expression “corresponds” that it was meant to define or rather replace. The same goes for Peirce’s use of the expression “end”, which seems to Rorty to assume that scientific inquiry converge asymptotically. Modern philosophy of science since Kuhn, however, has shown that convergence in science is at best no more than “a local and short-term phenomenon”15 and not a defining characteristic of the course of scientific history as a whole, as Peirce imagines. On Rorty’s view, the reason why Peirce runs into these conceptual dead ends is that he lacked the insight that James would be the first to articulate and that Davidson would go on to ground argumentatively, namely that “not only was ‘true of’ not a relation between ontologically homogenous relata, but was not an analyzable relation at all (…)”.16
At first glance it may appear that Rorty’s contribution to Davidson’s position on the theory of truth consists wholly in making out historical precursors and thus situating Davidson within the pragmatist tradition. Yet that appearance would be deceiving. It is true that Rorty presents no further arguments against the explanatory conception of truth and rests content with merely summarizing Davidson’s arguments with a view to James and Peirce; in doing so, however, Rorty succeeds in rendering Davidson’s critique of the explanatory conception of truth both more specific and more radical.
Rorty’s specification consists in his having worked out more precisely than Davidson himself the connections that obtain among consensus theories, coherence theories and correspondence theories of truth on the one hand and between the explanatory conception of truth and scheme-content dualism on the other. Rorty’s radicalization of Davidson’s position consists in his having drawn from it the consequence that truth need no longer be taken seriously as a philosophical issue once the explanatory notion of truth has been dismissed – a stance beyond that which Davidson was prepared to take.
Rorty’s radicalization of Davidson is the logical result of his specification of Davidson’s position. Rorty’s specification consists in having shown that consensus theories of truth adhere implicitly to the correspondence model, as illustrated by the case of Peirce, since their conception of truth remains bound to scheme-content dualism. Thus, although Peirce begins by emphasizing the scheme side of the dualism at the cost of its content side, he ends up by reintroducing content as the surplus value meant ultimately to accrue to the scientific community when the process of inquiry has come to an end. In this light, Rorty’s radicalization can be seen to consist in carrying over his criticism of Peirce to coherence theories that try to secure something like “correspondence with reality”17 beyond the dualism of scheme and content – a correspondence relation, that is, which does without the confrontation of mind and world.
Whereas Davidson thinks that after having repudiated scheme-content dualism it still remains for philosophy to show how we can “have knowledge of, and talk about, an objective public world which is not of our own making”18, Rorty is of the opinion that the whole problem is done away with as soon as we drop that dualism, so that we just no longer need the philosophical answer to epistemological scepticism that Davidson persists in looking for. These differing assessments influence the positive contributions which the authors have made to the understanding of non-explanatory uses of “true”, and that is what I would now like to go on to show.
THE NON-EXPLANATORY USES OF “TRUE” AND THEIR RELATION TO EACH OTHER
We have already become acquainted with the core of Davidson’s concept of truth in the form it takes in his early paper On the Very Idea of a Conceptual Scheme, where he develops it along Tarskian lines. It is to those early reflections that Davidson harkens back in A Coherence Theory of Truth and Knowledge when he sums up his conception thus: “Truth, as applied to utterances of sentences, shows the disquotational feature enshrined in Tarski’s Convention T, and that is enough to fix its domain of application.”19 In the further course of that paper Davidson goes on to transform Tarki’s Convention T by using it as an instrument for working out truth theories for natural languages. To this purpose, he analyses the disquotational use of “true”, illustrating it by means of the figure of the field linguist introduced by Quine in Word and Object (1960).
In Quine’s work, the field linguist finds himself in a situation of radical translation; that is, he is confronted with the linguistic and non-linguistic behavior of a native, which he must try to translate without the aid of a lexicon or an interpreter. Davidson’s basic point and the one that takes him beyond Quine lies in his assumption that the development of a theory of truth for the foreign language forms the very condition of possiblity for developing a translation manual that could be used for deriving statements about the meaning of the sounds the native makes. Now, the point that Davidson is driving at with these considerations is to show that his “principle of charity”20 is a presupposition that both makes it possible to construct a theory of truth for a foreign language and at the same time guarantees that the beliefs of the speaker of that language are mostly true.
Davidson’s proof for this works in the following way. The principle of charity urges us “to interpret what the speaker accepts as true when we can”21. For Davidson, this principle guides us in the construction of a theory of truth for a foreign language in the simplest and most elementary cases22, in that it leads us to assume that the objects we perceive as the causal stimuli for the native’s utterance of a one-word sentence such as “Gavagai” are in fact those to which he means to refer. According to Davidson, this assumption allows us to employ Tarski’s Convention T when constructing a theory of truth for a natural language and to formulate sentences like: “The one-word sentence ‘Gavagai’ is true if and only if there is a rabbit.”
Davidson’s understanding of the field ethnologist’s situation differs crucially from Quine’s in that he, unlike Quine, does not assume a purportedly language-neutral stimulus pattern as the cause of the native’s utterance. Instead, Davidson gives a hermeneutic twist to Quine’s naturalized epistemology by determining the point of reference of the native utterance from the ethnocentric point of view of the interpreter, i.e. the field linguist.
As a theoretical move, this hermeneutic triangulation – which is at the very heart of Davidson’s theory of radical interpretation – makes it possible for him to describe the truth relation not only and not primarily as an internal coherence among the native’s own statements, but rather first and foremost as an external coherence between the native’s staements and those of the field linguist. That, in turn, allows Davidson to explain truth qua correspondence on the basis of and hence as secondary to truth qua coherence. That is the sense in which he tells us right at the outset of his paper that the “theory that I defend is not in competition with a correspondence theory, but depends for its defence on an argument that purports to show that coherence yields correspondence”.23
This ambitious claim points to Davidson’s view, mentioned earlier, that even after we have given up the explanatory conception of truth, it still remains the task of any serious theory of truth to give an answer to the sceptic. Davidson in fact believes himself to have completed just this task in A Coherence Theory of Truth and Knowledge. Davidson characterizes the central goal of that paper succinctly when he writes: “My slogan is: correspondence without confrontation. Given a correct epistemology, we can be realists in all departments. We can accept objective truth conditions as the key to meaning, a realist view of truth, and we can insist that knowledge is of an objective world independent of our thought or language.”24
In view of the realist rhetoric Davidson associates with his naturalization of Tarski’s theory of truth and with his own hermeneutic triangulation of Quine’s epistemology, Rorty’s interpretative strategy in Pragmatism, Davidson, and Truth consists in describing the consequences of Davidson’s deflation of the concept of truth in a vocabulary less fraught with epistemic assumptions. Thus Rorty begins by emphasizing that “‘the philosophy of language of the field linguist’ (…) is all the philosophy of language (…) anybody needs”.25 He then goes on to make clear that the sceptic would in all likelihood reply to Davidson’s realist rhetoric by saying that we need more than an account of the demands of linguistic field research if we intend to refute the possibility of our being in fundamental error about the way things are. Rorty puts the sceptical argument underlying this objection in a nutshell when he writes that the sceptic “will think that Davidson has shown no more than that the field linguist must assume that the natives believe mostly what we do, and that the question of whether most of our beliefs are true is still wide open”.26
According to Rorty, the only reply Davidson can in fact give to this objection is “that radical interpretation begins at home”.27 The simple lesson Rorty recommends that we learn from this retort is just that “since we already have (in dictionaries) a translation manual for ourselves, als well as (in encyclopedias) an autoethnography, there is nothing more for us to know about our relation to reality than we already know. There is no further job for philosophy to do”.28 But that, continues Rorty, is just what the pragmatist has always been telling the sceptic. And that, rightly understood, is just the direction in which Davidson, too, is moving; his attempt to save the old realist vocabulary of correspondence has merely prevented him from putting the matter in its proper terms. That is Rorty’s sense when he submits “that Davidson was a bit misleading in suggesting that he was going to show us how coherence yields correspondence. It would have been better to have said that he was going to offer the sceptic a way of speaking which would prevent him from asking his question, than to say that he was going to answer that question”.29
Rorty’s essay goes beyond simply suggesting how Davidson might better understand his own position; its most important contribution consists in having introduced further distinctions among the non-explanatory uses of “true” and thus having made distinctions explicit that Davidson implicitly assumes. In addition to the disquotational use of the concept of truth, two other uses of the adjective “true” need to be noted. These are the “endorsing” and the “cautionary use”, respectively.30 Both differ from the disquotational use in that they are not descriptive uses made of the adjective by the interpreter or the field linguist from an external perspective, but rather normative uses made by the speaker himself from an internal perspective.
Whereas the interpreter or linguistic ethnologist uses the adjective “true” in order to point out a causal connection “between the organism and its environment”31, the (native) speaker himself takes an internal perspective when he uses “true” either as “a term of praise used for endorsing”32 beliefs or as a cautionary term “in such remarks as ‘Your belief S is perfectly justified, but perhaps not true’”.33 From James’ point of view, what the the endorsing use comes down to is that “‘The true’ (…) is only the expedient in the way of our thinking (…)”.34 The cautionary use, by contrast, reminds us “that justification is relative to, and no better than, the beliefs cited as grounds for S and that such justification is no guarantee that things will go well if we take S as a ‘rule for action’ (…)”.35
On the basis of this distinction among the three non-explanatory uses of the concept of truth, Rorty shows that we should reject both the restriction of the concept to just one of them and the attempt to establish one use as foundational for the other two. With this move he succeeds in bringing the classic pragmatist account of truth, which in James was focussed exclusively on the endorsing and in Peirce on the cautionary use, up to the analytic standards set by Davidson. For to Rorty’s mind, Davidson “has given us an account of truth which has a place for each of these uses while eschewing the idea that the expediency of a belief can be explained by its truth”.36
This remark touches on a vulnerable point in Davidson’s realist rhetoric, albeit indirectly and by way of negation, for contrary to Rorty’s charitable interpretation, the realist orientation Davidson ascribes to himself in A Coherence Theory of Truth and Knowledge would seem to suggest the possiblity of reducing the normative uses of “true” to a correspondence with reality which in turn would result from a disquotational use of “true” independent of scheme-content dualism. Put succinctly, Davidson’s realist rhetoric would if taken literally lead to precisely the account of truth that Rorty described as a faulty perspective in the last part of the passage just quoted – a perspective he credited Davidson with having avoided.
In fact, in his Afterthoughts 1987, published on occasion of a reprint of A Coherence Theory of Truth and Knowledge in the 1990 volume Reading Rorty, Davidson followed Rorty’s suggestions regarding nomenclature and retracted his misleadingly realist rhetoric. Space prevents me from entering into the details here.37 Let it suffice to quote the statement of the Afterthoughts that bears centrally on Rorty’s Pragmatism, Davidson, and Truth. “In this paper (…) Rorty urges two things: that my view of truth amounts to a rejection of both coherence and correspondence theories and should properly be classed as belonging to the pragmatist tradition, and that I should not pretend that I am answering the sceptic when I am really telling him to get lost.”38 Davidson adds approvingly, “I pretty much concur with him on both points.”39
In his 1990 paper on The Structure and Content of Truth, Davidson reformulated the philosophical consequences of his theory of truth in the light of Rorty’s criticisms. Allow me to begin once more with a remark by Davidson: “We should not say that truth is correspondence, coherence, warranted assertability, ideally justified assertability, what is accepted in the conversation of the right people, what science will end up maintaining, what explains the convergence of single theories in science, or the success of our ordinary beliefs. To the extent that realism and antirealism depend on one or another of these views of truth we should refuse to endorse either. Realism, with its insistence on radically nonepistemic correspondence, asks more of truth than we can understand; antirealism, with its limitation of truth to what can be ascertained, deprives truth of its role as an intersubjective standard.”40
Davidson’s reformulation of his position must be understood against the backdrop of his radicalization of his earlier critique of correspondence and coherence theories of truth occasioned by Rorty. In The Structure and Content of Truth, Davidson summarizes the main argument against the correspondence account of truth given in A Coherence Theory of Truth and Knowledge as follows: “The usual complaint about correspondence theories is that it makes no sense to suggest that it is somehow possible to compare one’s words or beliefs with the world, since the attempt must always end up simply with the acquisition of more beliefs.”41 The passage continues in a self-critical vein: “This complaint against correspondence theories is not sound. One reason it is not sound is that it depends on assuming that some form of epistemic theory is correct; therefore, it would be a legitimate complaint only if truth were an epistemic concept. If this were the only reason for rejecting correspondence theories, the realist could simply reply that his position is untouched; he always maintained that truth was independent of our beliefs or our ability to learn the truth.”42
The argument against correspondence theories of truth we met with in A Coherence Theory of Truth and Knowledge, formulated there in response to an early remark by Rorty, does in fact miss the point. For a realist of the appropriate persuasion could easily dodge the objection that we cannot step out of our language in order to make statements of a one-to-one correspondence between language and reality, simply by pointing out that he did not say we could in the first place. Rather, he has insisted from the very first that reality can deviate from the picture we make of it in our language and that the possibility of such a state of affairs has nothing to do with the question of whether we can come to recognize that deviation or not, since the relation of correspondence, conceived in realist terms, is an nonepistemic and precisely not an epistemic relation.
This insight does not, of course, imply for Davidson that the realist is right. It does, however, show that we have to replace the false refutation of realism with a correct one. Thus Davidson goes on in The Structure and Content of Truth to say, “The correct objection to correspondence theories is not, then, that they make truth something to which humans can never legitimately aspire; the real objection is rather that such theories fail to provide entities to which truth vehicles (whether we take these to be statements, sentences or utterances) can be said to correspond.”43 Whereas false and superficial criticism begins with the epistemic question of whether we can come to know of the correspondence relation, correct and incisive criticism aims at the prior question of whether and how the terms of the relation can be described as such and indepedently of their relation to each other.
Now since, as Davidson emphasizes, the theoretical realist concedes that reference to individual things as the objects of our statements always supposes some conceptual “frame of reference”44 within which they are described, he is left with no way out other than to claim with Frege, “that, if true sentences correspond to anything at all, it must be the universe as a whole”.45 With that, however, the correspondence theory just becomes trivial, for “there is no interest in the relation of correspondence if there is only one thing to which to correspond, since, as in any such case, the relation may as well be collapsed into a simple property: thus, ‘s corresponds to the universe’, like ‘s corresponds to (or names) the True’, or ‘s corresponds to the facts’ can less misleadingly be read ‘s is true’”.46
So much for Davidson’s radicalized critique of the correspondence theory of truth. Davidson produces a side piece to it in the same paper by offering an equally radical critique of coherence theories. Thus he explains that coherence theories based on consistent sets of beliefs held by individual persons (rather than on sets of statements as such) do not yield a test for truth, for here “the obvious objection is that many different consistent sets of beliefs are possible which are not consistent with one another”.47
The crucial point for Davidson is that it is not sufficient to show “how beliefs are causally and logically related to each other”48 in order to construct a theory of truth for a natural language. Beyond that, it is a matter of working out “how the content of a belief depends on its causal connections with the world”.49 What Davidson means by that should be clear in light of the hermeneutic reinterpretation of Quine’s field linguist we considered in our review of A Coherence Theory of Truth and Knowledge.
On Davidson’s view of the matter, the relation between our beliefs and the world can be spelled out on the basis neither of a correspondence theory nor of a coherence theory claiming a non-causal relation of correspondence. It is rather a causal relation of conditioning which, in the external perspective of the field linguist, comes into view as soon as he uses his interpretation of the world in conformity with the principle of charity as a frame of reference for making the natives’ beliefs intelligible to himself.
To sum up, in the two-fold radicalization of his critique as we have described it, Davidson is following Rorty’s suggestion that he formulate the consequences if his theory of truth in a vocabulary that is neither realist nor antirealist. Yet in spite of his obvious proximity to Rorty’s views in this respect, in the last part of the text cited Davidson is nevertheless concerned to point out remaining differences between Rorty and himself. To wit, he says in conclusion that the empirical theory of truth formulated in causal terms in the external perspective of the ethnologist, does after all possess a certain explanatory force, albeit in a weakened sense. Davidson writes, “(...) a theory of truth is a theory for describing, explaining, understanding, and predicting a basic aspect of verbal behavior. Since the concept of truth is central to the theory, we are justified in saying truth is a crucially important explanatory concept.”50 With this thesis, which on the face of it might seem to deprive the debate of its common ground, Davidson has opened a further round of discussion. It is to this most recent phase of the debate that I now turn.
THE REMAINING DIFFERENCE BETWEEN DAVIDSON’S AND RORTY’S THEORIES OF TRUTH
Rorty’s reaction to Davidson’s thesis can be found in his 1995 paper entitled Is Truth a Goal of Inquiry? There Rorty begins by taking a step in Davidson’s direction when he writes, “I am quite willing to withdraw my 1986 claim that ‘true’ has no explanatory use (…).”51 Rorty is referring here to a passage immediately following the one cited above from the last part of Davidson’s Structure and Content of Truth. Rorty concedes that describing “the pattern of behavior necessarily exhibited by language users”52 may indeed serve an explanatory purpose. Unlike Davidson, however, he does not think that the explanatory power is primarily due to the use of the word “true” in such a description nor that “true” has any especially central role to play.
Here Rorty raises two objections. First of all, the word “true” can lay no claim to a privileged role in the process of describing those behavioral patterns, since an empirical theory of truth is “automatically a theory of meaning and of rationality (…)”53, as Davidson himself had shown. Secondly, Rorty insists against Davidson that there is no deep reason “why one of the words that we use to describe the pattern of behavior necessarily exhibited by language users (logical inference) should also be one of the words we use to caution people that they may be believing something that better-advised people would not believe”.54
What these two arguments of Rorty’s come down to is that the hermeneutic triangulation analysed by Davidson may shed light on the endorsing and cautionary uses of “true”, but that it posesses no explanatory virtues with respect to the concept of truth as such. Thus, even though the adjective “true” fulfills an important function in the empirical description of causal relations between sentences and their objects, the function of the concept of truth nevertheless continues to be purely descriptive, and not normative. Hence the formal disquotational concept “true” cannot furnish us with a substantial grounding of either the endorsing or the cautionary concept of truth.
Yet Davidson seemed to be implying precisely such a grounding. Rorty expresses that when he reconstructs the difference which to his mind separates him from Davidson, saying, “Davidson thinks it significant that we use the same word to designate what is preserved by valid inference as we use to caution people that beliefs justified to us may not be justified to other, better audiences”55. Rorty then goes on to add the critical remark that “as far as I can see, there is no deep reason why ‘true’ is used to do both of these jobs (...)”.
In his recent article Truth Rehabilitated, Davidson has replied to Rorty’s overview of the discussion and the arguments it is based on. Davidson calls Rorty’s representation of the issue into question, denying Rorty’s insinuation that he (Davidson) is in search of a deeper source of the various uses of truth and thus reaffirming their common stance on this point. With regard to the question of why we use the same word both to explain logical validity and to caution our fellow humans about their claims to justification, Davidson arrives in Truth Rehabilitated at the same conclusion as Rorty did before: “I doubt we can explain this in a philosophically interesting way.”56
Davidson, however, suggests an alternative way of specifying the remaining difference between himself and Rorty. Whereas Rorty had placed emphasis on the second argument of Is Truth a Goal of Inquiry? in order to work out that difference, Davidson accentuates the first of the two arguments. Now, in regard to that argument Davidson does concede “that truth is one concept among a number of other related concepts which we use in describing, explaining, and predicting human behavior”57. At the same time, however, he insists on the centrality he had claimed for the concept of truth in describing human behavior, saying, “All these concepts (and more) are essential to thought, and cannot be reduced to anything simpler or more fundamental. Why be niggardly in awarding prizes; I’m happy to hand out golden apples all round.”58
Davidson’s reply to Rorty’s deflation of basic philosophical concepts such as “truth”, “meaning” and “rationality” does in fact point to the crucial point of difference separating them. That point lies in their respective understanding of themselves as philosophers. Rorty calls attention to this situation in his twofold reaction to Davidson’s Truth Rehabilitated. His first reaction is to be found in the essay Davidson between Tarski and Wittgenstein (1998). The other was published together with Davidson’s Truth Rehabilitated under the title Response to Donald Davidson in a collection of articles edited by Robert Brandom, Rorty and His Critics (2000).
In both essays, Rorty takes a metaphilosophical approach. To this purpose he distinguishes two different traditional understandings of themselves that philosophers can have and which lie at the heart of the intellectual movement known as analytic philosophy. They are “philosophy as therapy and philosophy as system-building”.59 Rorty sees it as characteristic of Davidson that his philosophical work refuses to be categorized into either one of those camps, but rather bridges their difference. And that is precisely what distinguishes Davidson from Rorty: Rorty sees himself as a decidedly therapeutic philosopher, and this understanding of his own role determines his view of the project of formalization Davidson is persuing in the tradition of Tarski: “We therapists tend to think that we can keep most of the arguments while ignoring the project.”60
Davidson’s interest in the basic philosophical concepts of modernity is a descriptive one. Davidson is interested in making the golden conceptual apples of modern philosophy shine with glowing light by offering a sophisticated analysis reconstructing each concept in terms of its varying uses. Not so Rorty. He is “dubious about the concept of ‘concept’”61. He recommends that instead of keeping on polishing philosophy’s golden conceptual apples we should slice them up, retaining only those parts we are able to digest. Hence Rorty can articulate a stance on “truth” for which “the question is not whether we have exhausted the concept of truth, or gotten truth right. It is, rather, whether we have sorted out the various uses of the word ‘true’, decided which of them had better be discarded, and specified the functions performed by the remainder.”62
For Rorty, it is clear that the normative uses of “true” – i.e. the endorsing and the cautionary – “do not need much philosophical definition or explication”.63 Its endorsing or cautionary function classes the adjective “true” together with such simple and philosophically unassuming expressions as “‘good!’”, “‘right!’”, “‘false!’” “‘way to go!’” and “‘watch it!’”.64 Their seeming philosophical relevance only arises when we attempt to attribute exalted explanatory claims to them and to play them off of one another or to derive their various uses from some one supposedly fundamental use. If we follow Rorty, the same goes for the disquotational use of “true”. As soon as the trivial role the adjective “true” has to play in Davidson’s hermeneutic triangulation has been explicated, there ceases to be any need “to create new pseudo-problems in the course of dissolving old ones”.65
Thus, Rorty allows that the causal use of “true” differs from the justificatory use in that, given expressions of belief, in the one case we judge of truth claims from an internal perspective and in the other from an external perspective; yet he insists that the expressions of belief themselves as articulated by enduring sentences like “Snow is white” or “Gavagai” are the same for both perspectives. The same goes for the relations between the various expressions of belief that can be formalized on this basis.66 Rorty makes this point when he writes, “The systematic relations between linguistic expressions which are captured by the recursive character of a Tarskian truth definition are not different from the relations of being-frequently-inferred-from of which the radical interpreter, hoping to construct such a definition, must keep track.”67
If we take Rorty’s thought here seriously, then it implies that the one use of “true” that fundamentally takes us beyond justification – that is, the disquotational use – does so in a way such that the transition to a causal explanation of linguistic behavior leaves us with absolutely no connection to questions of justification. From Rorty’s point of view, a Tarskian theory of truth is nothing but a descriptive systematization of inferences that we have learned to make within the internal, normative perspective on the basis of relations of justification and recognition and which we have learned to question with the aid of the cautionary use of “true”.
The point of contention between Davidson and Rorty lies in the question of the relevance that each attributes to this formal systematization. Davidson is of the opinion that the formalization of our patterns of justification is a means to discovering the machine that produces those patterns in the first place – a machine that every speaker must already be making implicit use of and which could make our thought and speech explainable in terms relevant for questions of justification. Rorty, on the other hand, thinks that recursive formalization of natural languages is just a scientific abstraction after the fact, having no relevance for the production of speech itself. Davidson’s attempt to apply Tarski to natural languages is for Rorty nothing but the attempt to construct an “underlying order” where what we are dealing with is de facto just “a lot of confusing uses”68. Appealing to the late Wittgenstein in Davidson against Davidson’s Tarski, Rorty expresses the hope that, “Just as Wittgenstein got over his youthful, Tractarian, desire for structure, so maybe we can get over, if not Tarski on formalized languages, at least the desire to carry Tarski over into non-formalized languages.”69
How Davidson will reply to Rorty’s most recent suggestion is an open question. We can be certain, however, that the two philosophers’ debate on truth will continue to contribute to the clarification of the sense in which “truth” can be reinvented as a radically secularized concept of our vocabulary. With that remark I would like to return to Rorty’s suggestion of substituting “analytic-conversational for analytic-continental as a description of the most salient split amongst today’s professors (...).“70 Our observations have shown that Davidson’s ideas on truth are open to the purposes of conversational philosophy that Rorty favorizes over against those of analytic philosophy. The reason for this is that Davidson takes both of the traditions characteristic of analytic philosophy equally seriously, the system-building tradition stemming from the Tractatus and Tarski as well as the therapeutic tradition reaching from late Wittgenstein and Quine to Sellars and Brandom.71
In conclusion we can say that Rorty uses the analytic-conversational distinction in order to mark off a narrower conception of analytic philosophy from a broader conception of that enterprise. The narrower conception refers to the endeavor to make philosophy into a strict science and is expressed by the opposition between analytic and conversational philosophy. This distinction can be used as an internal specification of analytic philosophy in the broader sense of problem (dis)solving, encompassing both quasi-mathematical formalization of philosophic thought and the critical countermovement directed against scientistic pretensions. Rorty achieves this by distinguishing the therapeutic and the system-building traditions as two forms of problem (dis)solving within analytic philosophy in the broader sense. “There has been a certain amount of tension between the analytic philosophers who are interested simply in getting rid of pseudo-problems and those who want to give systematic explanations of the pseudo-ness of these problems in the form of analyses of the concepts used in their formulation.”72
Complementarily to its use within analytic philosophy, the analytic-conversational distinction can also be applied to continental philosophy. Within continental philosophy there are those concerned to make the narratives of the history of philosophy they work on useful for a conversation on the question of how we might succeed in “grasping our time in thought”, in Hegel’s famous phrase. Examples for this type of philosopher are Nietzsche and Kierkegaard as well as Hegel and Heidegger. On the other hand there are those who believe that fastidious reconstruction of past formations of thought will put philosophy on the secure path of science either by reducing it to pure philology or by way of continuing research into the supposedly a-historical foundations of knowledge discovered by thinkers such as Kant or Husserl.
Returning to Davidson and Rorty, the distinction between a broader and a narrower sense of “analytic” allows us to say that in Davidson’s case we are dealing with an analytic philosopher both in the broad and in the narrow sense of the word, while Rorty is an analytic philosopher in the broad sense alone. As for the continental penchant for story-telling, the comparison between Davidson and Rorty leads to a different result. Rorty uses narrative more or less exclusively in the strict sense, i.e. as a source of inspiration for future forms of common sense, while Davidson draws on the professional tradition of continental philosophy especially in its Kantian and Fregean form not only in order to supply his work as an analytic philosopher with a philologically broad and hermeneutically refined foundation, but also to give proof of his opinion that “modes of analysis and adherence to standards of clarity (...) have always distinguished the best philosophy (...).”73
The trend toward establishing closer ties between analytic and continental traditions that books like this one actively seek to support in an age of globalization cannot by itself eliminate the metaphilosophical differences remaining between Davidson and Rorty. For these differences are rooted in an understanding of the philosophic profession that Rorty’s analytic-conversational distinction refers to and which dominates both in the analytic and in the continental tradition. Theoretical work on a transcultural concept of philosophy is not enough to change that understanding, for that would need supplementing by political efforts in the institutions of philosophical teaching and research.
Rorty is sceptical in this regard. He believes that increasing professionalization gives analytic philosophers an advantage over conversational philosophers.74 But of course that need not necessarily be the case. At a time when philosophical thought is ever more greatly challenged by processes of transformation in media and technology,75 a truly professional professionalization would consist in widening the economy of philosophical issues step by step and in using the notion of conversational philosophy to bring the self-understanding of analytic philosophers up to date in such a way that tradition and innovation could be wed in a more timely fashion within the disciplinary matrix of philosophy.
The conversational reinterpretation of Davidson’s and Rorty’s theories of truth offered in this article is an attempt to use narrative means to work out paradigmatically what I see as analytic philosophy's potentially pioneering role in reshaping our future understanding of ourselves. In my view, both authors are contributing to the creative re-invention of our understanding of truth. In the face of a Zeitgeist for which “truth” has become an issue for academic philosophy and ceased being a serious point of reference for our cultural self-understanding, their re-invention represents an important transfer between tradition and our philosophical present. That transfer consists not in a founding philosophical act, but rather in the pragmatic suggestion that we secularize the complex language game of truth in intelligent ways. Davidson and Rorty are thus showing both science and common sense a path that leads between the Charybdis of realist exaltation and the Scylla of antirealist reduction of our understanding of truth.
1 Translated by Brady Bowman. The first draft of this paper was originally read in October 2001 at a meeting of the “Engerer Kreis” (“inner circle”) of the Allgemeine Gesellschaft für Philosophie in Deutschland (AGPD). The “Engerer Kreis” consists of selected members of the official association of German philosophers (founded in 1950) and has been meeting regularly since 1952. In the context of the Engerer Kreis, analytic philosophy has for the most part been treated with some disdain. Analytically oriented German philosophers founded an independent society in 1990, the Gesellschaft für analytische Philosophie (GAP), which now boasts more than 600 members and thus comes close to the AGPD in terms of size and influence.
2 Richard Rorty, “Analytic Philosophy and Transformative Philosophy”, Online-Manuskript, http://www.stanford.edu/~rrorty/analytictrans.htm.
3 Rorty, “Analytic Philosophy and Transformative Philosophy”, loc. cit.
4 I am presently in the midst of preparing a reader that will bring together the most important texts of this debate in German translation. The volume will be published next year by Suhrkamp.
5 Donald Davidson, “On the Very Idea of a Conceptual Scheme“, in Davidson, Inquiries into Truth and Interpretation, Oxford, Clarendon Press, 1984, pp. 183-198. (First published in Proceedings and Addresses of the American Philosophical Association, vol. 47, 1974, pp. 5-20, after having been held in Atlanta on the 28th of December, 1973, as the presidential address before the Seventieth Eastern Meeting of the APA.)
6 Davidson, “On the Very Idea of a Conceptual Scheme“, loc. cit. p. 198. See also Davidson, “A Coherence Theory of Truth and Knowledge“, in: Alan R. Malachowski (ed.), Reading Rorty. Critical Responses to Philosophy and the Mirror of Nature (and beyond). London, Blackwell, 1990, pp. 120-138, here p. 122. (Originally published in Dieter Henrich (ed.), Kant oder Hegel? Stuttgart, Klett-Cotta, 1983, pp. 423-438.)
7 Davidson, “On the Very Idea of a Conceptual Scheme“, loc. cit. p. 198.
8 Davidson, “On the Very Idea of a Conceptual Scheme“, loc. cit. p. 195.
9 Donald Davidson, “A Coherence Theory of Truth“, loc. cit., p. 123.
10 Richard Rorty, “Pragmatism, Davidson and truth”, in Richard Rorty, Objectivity, relativism, and truth. Philosophical Papers, Vol. 1. Cambridge, New York, Oakleigh, Cambridge University Press, 1991, pp.126-150, here p. 127. (Originally published in Ernest LePore (ed.), Truth and Interpretation. Perspectives on the Philosophy of Donald Davidson, Oxford and Cambridge, Mass., Blackwell, 1986, pp. 333-355.)
11 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 127.
12 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 129.
13 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 129.
14 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 131.
15 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 131.
16 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 132.
17 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 132.
18 Davidson, “A Coherence Theory of Truth”, loc. cit., p. 123.
19 Davidson, “A Coherence Theory of Truth”, loc. cit., p. 122.
20 Davidson, “A Coherence Theory of Truth”, loc. cit., p. 129.
21 Davidson, “A Coherence Theory of Truth”, loc. cit., p. 129.
22 The cases in question involve occasional sentences, i.e. sentences “whose causes of assent come and go with time and place”. (See Davidson, “A Coherence Theory of Truth”, loc. cit., p. 130.)
23 Davidson, “A Coherence Theory of Truth”, loc. cit., p. 120.
24 Davidson, “A Coherence Theory of Truth”, loc. cit., p. 120-21.
25 Rorty, “Pragmatism, Davidson and truth”, loc. cit., p. 132.
26 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 135.
27 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 135.
28 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 135.
29 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 138.
30 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 128.
31 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 134.
32 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 127.
33 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 128.
34 James, William, Pragmatism. A New Name for Some Old Ways of Thinking. Cambridge, Mass. and London, Harvard University Press, 1975, p. 106. For a defence against the misunderstandings this statement provoked among readers of Pragmatism, see James, “The Pragmatist Account of Truth and Its Misunderstandings”, in William James, The Meaning of Truth. A Sequel to “Pragmatism”, Cambridge Mass. and London, Harvard University Press, 1975, pp. 99-116.
35 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 128.
36 Rorty, “Pragmatism, Davidson and truth”, loc. cit., 128.
37 For a detailed reconstruction see my introduction to the forthcoming volume, Wozu Wahrheit? Schlüsseltexte der Davidson-Rorty-Debatte, edited by Mike Sandbothe, Weilerswist, Velbrück Wissenschaft, 2003.
38 Davidson, Donald, “Afterthoughts, 1987”, in Alan R. Malachowski (ed.), Reading Rorty. Critical Responses to Philosophy and the Mirror of Nature (and beyond), London, Blackwell, 1990, pp. 134-138, here 134.
39 Davidson, “Afterthoughts, 1987“, loc. cit., 134.
40 Donald Davidson, “The Structure and Content of Truth”, in: The Journal of Philosophy, Bd. LXXXVII, Nr. 6, Juni 1990, S. 279-328, hier: S. 309.
41 Donald Davidson, “The Structure and Content of Truth“, loc. cit., p. 302.
42 Davidson, “The Structure and Content of Truth“, loc. cit., p. 302-03.
43 Davidson, “The Structure and Content of Truth“, loc. cit., p. 304.
44 Davidson, “The Structure and Content of Truth“, loc. cit., p. 303.
45 Davidson, “The Structure and Content of Truth“, loc. cit., p. 303.
46 Davidson, “The Structure and Content of Truth“, loc. cit., p. 303.
47 Davidson, “The Structure and Content of Truth“, loc. cit., p. 305.
48 Davidson, “The Structure and Content of Truth“, loc. cit., p. 305, note 47.
49 Davidson, “The Structure and Content of Truth“, loc. cit., p. 305, note 47.
50 Davidson, “The Structure and Content of Truth“, loc.cit., p. 313.
51 Richard Rorty, “Is Truth a Goal of Inquiry? Donald Davidson versus Crispin Wright”, in Richard Rorty, Truth and Progress. Philosophical Papers, Vol. 3. Cambridge, New York, Oakleigh, Cambridge University Press, 1998, pp. 19-42, here p. 25, note 23. (Originally published in Philosophical Quarterly, vol. 45, 1995, pp. 281-300.).
52 Rorty, “Is Truth a Goal of Inquiry?”, loc. cit., p. 25.
53 Rorty, “Is Truth a Goal of Inquiry?”, loc. cit., p. 25, note 23.
54 Rorty, “Is Truth a Goal of Inquiry?”, loc. cit., p. 25, note 23.
55 Rorty, “Is Truth a Goal of Inquiry?”, loc. cit., p. 25, note 23.
56 Donald Davidson, “Truth Rehabilitated”, in Robert Brandom (ed.), Rorty and His Critics, Oxford und New York, Blackwell, 2000, pp. 65-74.
57 Davidson, “Truth Rehabilitated”, loc. cit., p. 73.
58 Davidson, “Truth Rehabilitated”, loc. cit., p. 73.
59 Richard Rorty, “Davidson between Wittgenstein and Tarski”, in: Crítica: Revista Hispanoamericana de Filosophía, vol. 30, No. 88, April 1998, pp. 49-71, here p. 49.
60 Rorty, “Davidson between Wittgenstein and Tarski”, loc. cit., p. 50.
61 Richard Rorty, “Response to Donald Davidson”, in Robert Brandom (ed.), Rorty and His Critics, Oxford und New York, Blackwell, 2000, pp74-80, here p. 77.
62 Rorty, “Response to Donald Davidson”, loc. cit., p. 77.
63 Rorty, “Is Truth a Goal of Inquiry?”, loc. cit., p.22.
64 Rorty, “Is Truth a Goal of Inquiry?”, loc. cit., p.22.
65 Rorty, “Davidson between Wittgenstein and Tarski“, loc. cit., p. 69.
66 Compare Rorty’s “Is Truth a Goal of Inquiry”, where Rorty attributes to Davidson the claim that “the pattern truth makes is the pattern that justification to us makes” (loc. cit., p. 25).
67 Rorty, “Davidson between Wittgenstein and Tarski“, loc. cit., p. 68.
68 Rorty, “Davidson between Wittgenstein and Tarski“, loc. cit., p. 65.
69 Rorty, “Response to Donald Davidson“, loc. cit., p. 74.
70 Rorty, “Analytic and Conversational Philosophy”, see p. # of this volume.
71 That Davidson’s thinking “spans this gap” (Rorty, “Davidson between Wittgenstein and Tarski”, loc. cit., p. 49) is again one of the reasons why the influence of his philosophy extends far beyond the bounds of the English-speaking world.
72 Rorty, “Davidson between Wittgenstein and Tarski”, loc. cit, p. 49.
73 Davidson, “Afterthoughts, 1987”, loc. cit., p. 137.
74 Cp. Richard Rorty, “A Pragmatist View of Contemporary Analytic Philosophy”, in William Egginton and Mike Sandbothe (eds.), The Pragmatic Turn in Philosophy, New York, SUNY 2002 [in print].
75 Cp. Mike Sandbothe, Pragmatische Medienphilosophie. Grundlegung einer neuen Disziplin im Zeitalter des Internet, Weilerswist, Velbrück Wissenschaft 2001.