Rorty, Davidson, and Representation

1 Rorty, Davidson, and Representation Steven Levine To be published in The Wiley-Blackwell Companion to Rorty Draft, please do not cite without permission. 1. Introduction In his paper “Rorty and Anti-realism” Frank Farrell argues that Rorty’s engagement with Davidson is born of a somewhat cynical attempt “to show that his position, radical though it may seem, is largely motivated by considerations that come out of the work of one of the most honored of contemporary analytic philosophers” (Farrell 1995: 155). In essence, Farrell accuses Rorty of riding Davidson’s coattails, of using his prestige to advance his own more dubious philosophical (or anti-philosophical) program. But, Farrell argues, because their views are so divergent, Rorty being an anti-realist, Davidson a realist, Rorty can only use Davidson in this way by significantly misrepresenting him. If we get straight about Davidson’s views we can see that Rorty’s reading of him as providing support for an anti-realist form of pragmatism not only gets Davidson wrong, but precisely backward. I would not deny that Rorty sometimes misreads Davidson, or that there are some significant differences between them, some which we will discuss below. But in this paper I would like to show that the affinity between the two thinkers is far greater than interpreters like Farrell allow. To put it simply, Rorty and Davidson are united in their attempt to get beyond certain dichotomies—between scheme-content, subject-object, and realism-anti-realism—that structure not only twentieth century analytical philosophy but modern philosophy generally, and both attempt to do so by rejecting the concept of representation. If we jettison the concept of representation, they think, we jettison the concept that makes the dichotomies between scheme and content, subject and object, realism and antirealism meaningful. 2 Now Farrell and many others question this interpretation. First, as we mentioned above, they argue that while Davidson rejects the concept of representation, and so the metaphysical form of realism that relies on it, he is nonetheless a realist. Second, they argue that Rorty is not able to get beyond the dichotomies, thinking that he simply occupies one side of them; the scheme, subject, or anti-realist side. This claim is embedded in the common charge that Rorty is a relativist, or a subjectivist, or a linguistic idealist. In this paper I mostly focus on evaluating the argument that Rorty is not able to get past the dichotomies—although I will have cause to comment on the first one as well. I do not contest that for most of his career Rorty, even though he claimed to disown the dichotomies, nonetheless espoused a view that was structured by them. As I try to show, from the seventies till the mid-nineties Rorty was committed to versions of the scheme-content and subject-object distinctions, and this pointed his view in an anti-realist direction. But in his late work, Rorty, through a reading of Davidsonian triangulation, changed his view in a significant way. In the context of a continued antirepresentationalism, Rorty comes to admit that there is a ‘truth in realism’ and that our relations to the world are not just causal but also normative. I argue that with these admissions Rorty comes to articulate a view that can, with some plausibility, be said to have gotten beyond the dichotomies of scheme and content, subject and object, and realism and anti-realism. But I also want to show that the breakthrough here is fleeting, that Rorty does not fully accept the consequences of his own change in view. 2. Davidson and Anti-Representationalism In Philosophy and the Mirror of Nature Rorty unfolds a general critique of the concept of representation, which he argues underlies modern philosophy, where epistemology is thought of as first philosophy. The “general theory of representation” (Rorty 1971: 3) offered by philosophy so understood is comprised of two general claims. 3 First, this theory claims that representation of the world is achieved through non-causal intermediaries, whether epistemic intermediaries—ideas, sensations, sense data, etc.—or sentencelike items conceived of as a medium that stands between self and world. In either case, representations are the product of an encounter between two distinct ontological realms, mind and world. The theory of representation claims we are able, a priori, to parse out what in such representations is owed to the mind and what is owed to the world. As such, it is committed to the scheme-content distinction. Second, the theory makes the essentialist claim that there is a class of privileged representations that accurately represent things because, through a type of confrontation, they are gripped by the nature of the object itself. It is these representations that correspond to the world’s single intrinsic nature, and it is these representations that must be called on to make true statements about the world. There are two types of representations that are potentially privileged in this way: intuitive representations that are given to the mind, and conceptual representations that, in being the product of the mind, are a priori and necessary. In the central chapter of Philosophy and the Mirror of Nature the critique of representation takes the form of a story about “how the notion of two sorts of representations—intuitions and concepts—fell into disrepute in the latter days of the analytic movement” (Rorty 1979: 168). Davidson does not have the starring role in this story, rather Sellars and Quine do. They have this starring role because of their shared holism, their view that “justification is not a matter of a special relation between ideas (or words) and objects, but of conversation, of social practice” (1979: 170). If justification is a social practice in which our assertions are only answerable to others in the space of reasons, then we “have no need to view it as accuracy of representation” (1979: 170). And if justification is not based in accuracy of representation then we have no reason to try to pick out privileged representations—whether intuitive and given, or conceptual and necessary—that correspond to the world’s intrinsic nature. But without these privileged representations the 4 dichotomies between what is given to the mind and what is a product of the mind, what is contingent and what is necessary, cannot even so much as be formed. But while Sellars and Quine respectively go much of the way toward a satisfying critique of the theory of representation, they each come up short. While Sellars dispatches the Given, he holds onto a tacit distinction between the contingent and the necessary, the empirical and the structural. Conversely, while Quine gives up on the distinction between facts and a priori meanings, contingent and necessary truths, he holds onto a conception of the Given. For Rorty, it is Davidson who is able to finally liberate us from the representationalist picture that ‘holds us captive’. He does this by providing a holistic anti-representationalist theory of meaning in which there is no place for the scheme-content distinction. So by “deepening and extending the lines of thought traced by Sellars and Quine” (Rorty 1991: 1) Davidson culminates “the holist and pragmatist strains in contemporary analytic philosophy. These motifs, in turn, are the culmination of a long struggle (which extends far outside the boundary of ‘analytic’ philosophy) against Platonic and religious conceptions of the world” (1991d: 117). Davidson’s importance extends far beyond analytical philosophy because in giving us a picture of belief, meaning, and knowledge that dispenses with the concept of representation, Davidson allows us to see that the philosophical picture that depends on it, i.e., epistemology as first philosophy, is optional and can simply be dropped. In so doing, he also gives us the means stop being gripped by the anxieties generated by this picture, namely, conceptual relativism and epistemological skepticism. So Davidson is perhaps the key figure that Rorty calls on to forward his prophetic desire to leave philosophy (so understood) behind. But as we shall see, in so extensively engaging with Davidson Rorty’s path out of philosophy becomes less direct than he might have hoped. 5 3. The Attack on Representation: Epistemic Intermediaries Davidson has a two-pronged attack on the concept of representation.1 The first concerns representations conceived of as epistemic intermediaries—sensations, sense data, the uninterpreted given—while the second concerns representations conceived of as thoughts or sentences that correspond to, or are made true by, the facts. These two attacks, in turn, undermine two distinct versions of the scheme-content distinction. According to the first, the scheme side of the distinction is comprised of concepts and the content side is comprised of given sensory or intuitive experiences. Here we have the classical Kantian version of the distinction. The attack on representations as sentences made true by the facts, on the other hand, is an attack on a version of the distinction in which the scheme side is comprised of linguistic representations and the content side is the world itself. Here we have the linguistic-analytical version of the distinction, where a conceptual scheme is equated with a language. In this section we’ll look at the first attack, coming back to the second in the next section. In his paper “A Coherence Theory of Truth and Knowledge” Davidson unfolds a general argument again the idea that our beliefs can be grounded in epistemic bases or intermediaries that stand outside the scope of our beliefs—in sensory experience or the Given. Indeed, the distinguishing mark of Davidson’s coherence theory is “that nothing can count as a reasons for holding a belief except another belief” (Davidson 2001c: 141). Or as Rorty puts it, in a passage cited approvingly by Davidson: “nothing counts as justification unless by reference to what we already accept, and there is no way to get outside our beliefs and our language so as to find some other test than coherence” (Davidson 2001b: 141, Rorty 1979: 178).2 The idea that sensation can provide a reason for holding a belief is attractive because it is seems clear that it is through sensations that we are primarily related 1 Here I follow the interpretive schema found in Ramberg 2001. Davidson later came to express regress at calling his view a coherence theory (see Davidson 2001c: 155). But this is not because he gives up this notion of justification, but because he thinks calling a theory a ‘coherence theory’ lends credence to the idealist or anti-realist idea that the world is dependent on the mind, an idea Davidson rejects. 2 6 to the world, and that we are sometimes aware of such items, making them candidates to be justifiers. But this attractive idea cannot be made good. Davidson’s first argument for why this is so is that if the justification that sensation provides depends on an awareness of sensation, then we have not made good the idea of sensation being the justifier, as awareness of sensation is just another belief. So we have not gotten outside the circle of beliefs. But why can we not just say that sensations justify beliefs directly, without the intervention of an awareness of sensations? The problem here is that sensation, in not being a propositional attitude, is not the type of item that can transmit epistemic warrant to belief, “for that requires that both relata have propositional content” (Davidson 2005: 136). The relation between a sensation and a belief cannot be logical, since sensations are not beliefs or other propositional attitudes. What then is the relation? The answer is, I think, obvious: the relation is causal. Sensations cause some beliefs and in this sense are the basis or ground of those beliefs. But a causal explanation of a belief does not show how or why the belief is justified. (Davidson 2001c: 143) In Rorty’s Sellarsian language Davidson here provides a critique of the Myth of the Given by untangling “the basic confusion contained in the idea of a ‘theory of knowledge’,” i.e., the “confusion between justification and causal explanation” (Rorty 1979: 161). It is through sensory experience that we are causally confronted by the world, but the notion of experience at play here is not itself justificatory and hence not one that is relevant, epistemically or semantically, to the propositionally structured beliefs that comprise knowledge. So “although sensation plays a crucial role in the causal process that connects belief with the world, it is a mistake to think that it plays an epistemological role in determining the content of those beliefs” (Davidson 2001a: 46). We don’t need an account of sensation as playing a role in determining the content of belief because Davidson endorses an externalism in which beliefs gain their content, via learning mechanisms, through their direct causal relations to the objects they are about. “[I]n the simplest and most basic cases, words and sentences derive their meaning from the object and circumstances in whose presence they were learned. A sentence which one has been conditioned by the learning 7 process to be caused to hold true by the presence of fires will (usually) be true when there is a fire present” (Davidson 2001a: 44). It is important to characterize what Davidson is saying here correctly. The point is not that there are given epistemic intermediaries, neutral and specifiable in worldindependent terms, which are then somehow causally related to worldly items.3 Davidson’s point is rather that the very characterization of a belief’s content is intrinsically—because of the direct causal link between belief and object found in the learning situation—world-involving. But if this is true then we have beliefs and we have the world. What we don’t have are intermediaries—neutral and world-independent representations—that are the touchstones through which beliefs gain meaning.4 What is the upshot of this picture? For Davidson and Rorty—who follows Davidson closely here—the epistemic neutralization of sensation is meant to lead us to abandon the ‘third dogma of empiricism’, the dogma that there is a distinction between scheme and content. The sensory Given supports a scheme-content distinction because if there is an “element in the mind untouched by conceptual interpretation” (2001a: 41) then we have the means to imagine “the unsullied stream of experience” as able to be “variously re-worked by various minds or cultures” (2001a: 41). So the sensory Given provides us with a common reference point which could in principle be taken up by distinct conceptual schemes or languages. Here we have the two essential ingredients of schemecontent dualism: the notion of an “organizing system and something waiting to be organized” (Davidson 1984: 189), and there being “something neutral and common that lies outside all schemes” (1984: 190). But if, as Rorty puts it, there is “no need to postulate an intermediary between the physical thrust of the stimuli upon the organ and the full-fledged conscious judgment that the properly programed organism forms in consequence” then there would be “no need to split the organism up into a receptive wax tablet…and an ‘active’ interpreter of what nature has there 3 Davidson argues that these two features of the Given are separable because it is possible to think of an item as world-independent, a source of evidence that can be “specified without reference to what it is evidence for” (Davidson 2001a: 42) without thinking that it is neutral or un-interpreted by a scheme. Davidson thinks Quine has a view like this. In this paper I treat the Given as having both features. 4 For more on this see Child 1994. 8 imprinted…The suggestion that our concepts shape neutral material no longer makes sense once there is nothing to serve as this material” (Rorty 1982: 4). In other words, if we don’t have the sensory Given then we don’t have something neutral and world-independent that stands ready in the mind waiting to be organized by distinct conceptual schemes. But if there is nothing neutral and world-independent standing ready in the mind then there is also nothing that mediates the relation of belief and world. So in undermining conceptual relativism Davidson also undermines epistemological skepticism insofar as global skeptical doubts are predicated on the idea that one could be systematically out of touch with the way things are due to the fact that there might be a gap between the epistemic intermediaries of which one is immediately aware and the world. However, if there are no epistemic intermediaries that stand between belief and world, but only direct causal connections, then we have not created “a gap no reasoning or construction can plausibly bridge” (Davidson 2001a: 43). Rorty urges Davidson to accept a certain meta-philosophical interpretation of this argument against the skeptic. Davidson should not think of himself as providing an answer to the skeptic, for this entails accepting the validity of the idea that knowledge can be secured only by our accurately representing the world, and so entail accepting the framework of epistemology as first philosophy. Rather, he should think of himself as developing “a way of speaking” about truth, belief, and knowledge “which would prevent [the skeptic] from asking his question” (Rorty 1991e: 138). In an “Afterward,” written after “A Coherence Theory of Truth and Knowledge,” Davidson accepts Rorty’s framing of the issue: “I agree with Rorty to this extent: I set out not to ‘refute’ the skeptic but to give a sketch of what I think to be a correct account of the foundations of linguistic communication and its implications for truth, belief, and knowledge. If one grants the correctness of this account, one can tell the skeptic to get lost” (2001d: 157). One can tell the skeptic to get lost because on Davidson’s account “general skepticism about the deliverances of the senses cannot even be formulated, since the senses and their deliverances play no central theoretical role in the account 9 of belief, meaning, and knowledge” (Davidson 2001a: 45). If one accepts it then “[a]nswering the global skeptic will no longer be a challenge, the search for epistemological foundations will be seen as pointless, and conceptual relativism will lose its appeal” (2001a: 47). 4. The Attack on Representation: Sentences Let’s now move to the other powerful concept of representation available to the philosopher: the notion of a sentence made true by corresponding to the facts. Davidson also provides a powerful critique of this idea, thinking that we can give no sense to the idea of there being anything to which representations correspond, and so no sense to the idea of truth as correspondence. Here, once again, Davidson and Rorty converge. Based on Davidson’s critique of epistemic intermediaries one might think that he rejects the correspondence theory of truth because without them one has no point of reference to get ‘outside’ one’s beliefs so as to compare them with the world. But while Rorty often says things like this, this is not Davidson’s argument. This is because he thinks that this kind of argument takes for granted the correctness of the view that truth is an epistemic concept, one where it is internally tied to justification—whether in ideal circumstances, at the end of inquiry, etc. But Davidson emphatically rejects epistemic theories of truth—whether pragmatic, coherentist, or internally realist. Davidson’s argument against the correspondence theory is not predicated on assuming that anti-realism is right, but in finding the idea of correspondence to be incoherent insofar as there “is nothing interesting or instructive to which true sentences might correspond” (Davidson 1990: 303). Because he argues the world is not cut up into propositionally structured blocks to which our sentences correspond, truth cannot be correspondence between sentences and these blocks.5 In light of this: 5 This follows from C. I. Lewis’ notion that one cannot locate the part of reality to which a sentence corresponds and Frege’s slingshot argument that all true sentences refer to the same thing. Although these arguments are controversial we shall grant them. 10 The correct objection to correspondence theories is…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. As I once put it, ‘Nothing, no thing makes our statements true’. If this is right, and I am convinced it is, we ought also to question the popular assumption that sentences…can properly be called ‘representations’, since there is nothing for them to represent. If we give up facts as entities that make sentences true, we ought to give up representations at the same time, for the legitimacy of each depends on the other. (Davidson 2001e: 184) Those who claim Davidson for the realist camp argue that this is not dispositive in showing that he not a realist.6 For even though it is clear that Davidson rejects building-block accounts of meaning where meanings are confronted with worldly facts—thinking instead that satisfaction relations between word and world are derived from a holistic Tarski-style theory of truth rather than being its basis—he nonetheless claims that if meaning and belief cohere in the right way that we can account for the correspondence relation. As he put it in “A Coherence Theory of Truth and Knowledge”: “My slogan is: correspondence without confrontation. Given the 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” (2001c: 138). But things are not so simple. For after this passage was written Davidson expressed regret at ever calling his view a correspondence theory and in light of that says: “To the extent that realism is just the ontological version of a correspondence theory, I must…reject it” (Davidson 2001e: 185). And looking back on his prior position, he says this: All I had in mind in calling my position a form of realism was a rejection of anti-realism; I was concerned to reject the doctrine that either reality or truth depends directly on our epistemic powers. Once more however, I must make clear that I am neither accepting nor rejecting the objectivist realist slogan that the real and the true are ‘independent of our beliefs’. (2001e: 185) This is how Rorty would put this point: in rejecting the idea that there is anything for representations to represent, and therefore in rejecting representations, Davidson means to challenge the philosophical picture which forces one to accept or reject the objectivist-realist slogan about the 6 Many early readers of Davidson read his semantics as a type of realism. See McDowell 1998. For readings of Davidson as a realist that directly compare him with Rorty, see Baghramian 1990 and Farrell 1995. 11 independence of the real. Davidson is not best thought of as a realist, but as one who is trying to question the terms upon which the very distinction between realism and anti-realism is formed. But while Davidson rejects the correspondence theory of truth, and the concept of representation that it supposes, he nonetheless thinks that truth is objective. Here is how Davidson puts it: “I can believe it is now raining, but this is because I know that whether or not it is raining does not depend on whether I believe it, or everyone believes it, or it is useful to believe it; it is up to nature not me or my society or the entire history of the human race” (Davidson 2000: 16). But if truth is objective how is his view distinguishable from the correspondence theorist, and from the view expressed by his own prior realism? His answer is this: “What saves truth from being ‘radically nonepistemic’ (in Putnam’s words) is not that truth is epistemic but that belief, through its ties with meaning, is intrinsically veridical” (Davidson 2001d: 156). While truth cannot be equated with what is justified or what we rationally believe at a given point, as epistemic theories hold, it is related to belief because most of our beliefs must be true. While our mutually believing something does not make it so (because truth is objective), belief creates a presumption that it is so. Of course the question emerges: why think that belief, because of its ties with meaning, is intrinsically veridical? This thesis falls out of Davidson’s theory of meaning. The basic idea at the heart of this theory is that a Tarskian theory of truth, suitably modified, can stand in for a theory of meaning. To formulate what it is for an agent to understand and use a language we don’t need to posit private meanings or ideas, rather we can, in an extensionist fashion, lay out their grasp of a sentence’s truth conditions. Davidson’s way of empirically confirming the theory is to imagine a field linguist in a totally foreign culture trying to interpret the language of a native speaker. In this case he does not know what the native believes or what they mean by their utterances. So to begin interpretation the interpreter needs to latch onto something simpler than these, which Davidson thinks the interpreter can find in “information about what episodes and situations in the world cause an agent to prefer that one rather than another sentence be true” (Davidson 1990: 322). 12 The interpretation of the native, the carving out of intelligible patterns in their utterances, happens through observing their prompted assent to true sentences, and interpreting those sentences in light of two assumptions: that the sentences they assent to are in general caused by the same features of the world that the interpreter’s matching sentence are caused by, and 2) that these sentences display a degree of logical coherence with the other sentences they assent to. Here the interpreter’s observations are governed by two principles of charity: a principle of correspondence and a principle of coherence.7 In both cases the interpreter reads “some of his own standards of truth into the pattern of sentences held true by the speaker. The point of the principle is to make the speaker intelligible, since too great deviations from consistency and correctness leave no common ground on which to judge either conformity or difference” (Davidson 2001c: 148). The thesis that ‘belief is of its nature veridical’ follows directly from the necessary use of these principles of charity. Rorty perspicuously glosses the argument thusly: [I]f one thinks of ‘meaning’ in terms of the discovery of the speech dispositions of foreigners rather than in terms of mental essences (ideas, concepts)…we will never reach the limiting case of a foreigner all or most of whose beliefs must be viewed as false…We will not reach this case (so the Davidsonian argument goes) because any such translation scheme would merely show that we had not succeeded in finding a translation at all. But (to extend Davidson’s argument a bit) if we can never find a translation, why should we think that we are faced with language users at all? (Rorty 1982: 5-6). This is a powerful argument: If an interpreter, after a sustained and good-faith effort, cannot find—in light of the principles of charity—translations or interpretations for a range of the native’s utterances and actions, if they cannot find some of what the native says and does intelligible, then they have no reason to think that they are dealing with an agent at all, a being whose behavior (including speech behavior) displays a rational and intelligible pattern. Conversely, if an interpreter can understand some of what the native says and does then it is impossible, because of the principle 7 The principle of correspondence is named in such a way as to mislead us into thinking that Davidson has a building block account of meaning, one based on identified atomic correspondence relations. But as we mentioned above we do not arrive at a native’s T-sentences by noting single causal links between sentences and bits of the world, rather, we work ourselves in the whole system of the native’s nested propositional attitudes by moving back and forth between occasion sentences and standing sentences till an intelligible pattern emerges. 13 of correspondence, that they could “discover the speaker to be largely wrong about the world. For the interpreter interprets sentences held true (which is not to be distinguished from attributing beliefs) according to the events and objects in the outside world that cause the sentence to be held true” (Davidson 2001c: 150). Hopefully it is clear how this applies to the conceptual relativism that follows from acceptance of the linguistic-analytical version of the scheme-content distinction. A complete failure of translation or interpretation of a different linguistic scheme cannot be sustained because complete failure just signals that there is no scheme there to be translated. If this is not to be the conclusion then the interpreter must find at least some of the native’s speech intelligible. But if an interpreter is to find some of the native speech intelligible, and therefore some of it unintelligible, then they must already have a significant baseline of shared belief with the native, giving them a basis to work their way into the native’s system of belief. Because “[w]e can make sense of differences” in meaning and belief “only against a background of shared belief” (Davidson 1984a: 200) interpretive differences do not signal incommensurability but simply the open and temporal nature of interpretation. The thesis that ‘belief is of its nature veridical’ is also necessary to complete the picture of belief, meaning, and knowledge that prevents the skeptic from asking their question. While skeptical questioning can get a grip with respect to single beliefs, or single classes of belief, they cannot get a grip with respect to our global system of belief. They cannot because with the belief is veridical thesis, we have an assurance that our beliefs “are not mistaken in the main” (Davidson 2001c: 146), and therefore that global skeptical doubts are otiose. But Rorty does not think that this is enough to forestall the skeptic from asking their question. For the skeptic “will think that Davidson has shown no more than that the field linguist must assume that the natives believe mostly what we do, and the question of whether most of our beliefs is true is still wide open” (Rorty 1991e: 135). Davidson argues that reflection on the very concept of belief gives a subject who applies the principles of charity to others reason to think that 14 most of their own beliefs are true. But Rorty thinks that the skeptic will find this argument questionbegging. To address the point, we must again show that the skeptic’s assumptions are based on a contestable philosophical picture, one where our access to the world is mediated by representations. If we instead come to “see language in the same way we see belief—not as a ‘conceptual framework’ but as a causal interaction with the environment described by the field linguist” then it will be “impossible to think of language as something which may or may not (how could we ever tell?) ‘fit the world’. So once we give up tertia, we give up (or trivialize) the notions of representation and correspondence, and thereby give up the possibility of formulating epistemological skepticism” (Rorty 1991e: 139). So to tell the skeptic to get lost we must eliminate the concept of representation and therefore the whole framework in which skepticism makes sense. 5. Two Differences Between Rorty and Davidson Two differences between Rorty and Davidson emerge at this point. First, Davidson thinks that the determination of content by the radical interpreter is not only causal but also normative. The causal account of the content of belief given above is the starting point of Davidson’s view of belief, not its terminus. While interpretation begins by tracing connections between sentences that are assented to as true and the world such sentences are about, the point of doing so is to work one’s way into the native’s network of interlocking propositional attitudes, such that one grasps not only what prompts them to assent to sentences but also what they find to be a reason for what. As McDowell says: “Davidson’s field linguist aims to work into an appreciation, as from within, of the norms that constitute the language she investigates: the specific sense of when it is right to say what according to which that language-game is played” (McDowell 1996: 152). The field linguist not only utilizes norms in their interpretive practice, but they mean to find norm-governed behavior, rational behavior, in the ‘object’. This distinguishes the field linguist from the physicist, whose practice is norm- 15 governed, but for whom there is no expectation of finding rationality in their object (see Davidson 2001f: 215). For most of his career, Rorty, in a revisionary spirit, shrugged off this point, thinking that the radical interpreter fixes meanings simply through descriptions of the causal interchange of the native and world. “There is nothing especially normative about my effort to translate, since all I am doing is finding a pattern of resemblances between my linguistic behavior and the native’s. I am trying to mesh her behavior with mine by finding descriptions of what she is doing that also describes what I sometimes do” (Rorty 1999: 583-4). The idea that in radical interpretation we try to gain the inside view of the native countenances the idea that there is an inside view to get, that there is some divide between minds, which have a norm-governed inside view, and all other physical things, which don’t. But if radical interpretation simply tries to describe patterns of behavior then we can see that psychology and its vocabulary is continuous with the vocabularies of the rest of the special sciences, and consequently, that the mind is not a special object that stands outside of the rest of nature. The second difference between Rorty and Davidson concerns how the ‘belief is of its nature veridical’ thesis bears on the concept of truth. For most of his career Rorty interpreted this thesis as saying that interpreter and native agree in most of their beliefs not because most of those beliefs get reality right, but simply because the principle of charity demands it: [W]hen we remember that Davidson will have no truck with the idea that truth consists in correspondence to, or accurate representation of, reality, we realize that he is not saying that our minds are, thanks to God’s or Evolution’s contrivance, well suited to the task of getting reality right…He is rather saying that most of anybody’s beliefs must coincide with most of our beliefs (because to ascribe beliefs in the first place one must invoke the Principle of Charity) and that to reject that mass of shared belief (as perhaps not corresponding to reality) is to bring back a tangle of uncashable and useless metaphors—those used to state the scheme-content distinction. To say, as Davidson does, that ‘belief is in its nature veridical’ is not to celebrate the happy congruence of subject and object but rather to say that the pattern that truth makes is the pattern that justification to us makes. (Rorty 1998a: 25) Here Rorty interprets Davidson’s view in light of his coherentism and ethnocentrism: to say that belief is of its nature veridical, and so that most of our beliefs are true, is to say that the native mostly agrees with us concerning the justification of our beliefs. It is because others mostly agree 16 with us about justification that we hold beliefs to be veridical and true, not because they get reality objectively right. While we ought not see the mass of our shared beliefs as diverging from the way things are (which brings back the global skepticism based on the scheme-content distinction) we cannot explain this through the congruence between subject and object but through the convergence of shared justifications. In contrast, and as we saw above, Davidson thinks that truth is objective and that our agreements in justification depend on our beliefs getting reality right.8 In light if this he poses a challenge to Rorty’s pragmatism: From the fact that we will never be able to tell for certain which of our beliefs are true, pragmatists conclude that we may as well identify our best researched, most successful, beliefs with the true ones, and give up the idea of objectivity. (Truth is objective if the truth of a belief or sentence is independent of whether it is justified by all of our evidence, believed by our neighbors, or is good to steer by.) But here we have a choice. Instead of giving up the traditional view that truth is objective, we can give up the equally traditional view (to which the pragmatists adhere) that truth is a norm, something for which to strive. I agree with the pragmatists that we can’t consistently take truth to be both objective and something to be pursued. But I think they would have done better to cleave to a view that counts truth as objective but pointless as a goal. (Davidson 2000: 67) The dilemma is this: either accept that truth is not a goal of inquiry, but is objective; or that truth can be a goal of inquiry, but is not objective. This dilemma has bite because, as Davidson points out, Rorty often says both that truth cannot be a goal of inquiry and that there is no difference between what is true and what is justified (see Davidson 2000: 74n). The latter follows from this thought: “[I]f something makes no difference to practice, it should make no different to philosophy. This conviction makes [pragmatists] suspicious of the distinction between truth and justification…I cannot bypass justification and confine my attention on truth: assessments of truth and assessments of justification are, when the question is about what I should believe now, the same activity” (Rorty 1998a: 19). Based on this it seems that Rorty has a ‘theory’ of truth, i.e., that it is justification. 8 Neil Gascoigone identifies this as the deepest difference between Davidson and Rorty (Gascoigone 2008: 95). This difference is softened by the moves that Rorty makes late in his career. 17 But Rorty’s view of truth is, by his own admission, not stable.9 For he came to see, even before his dialogue with Davidson on truth, that any belief might be justified and still be false. The so-called ‘cautionary use’ of the truth predicate, which is used “in such remarks as ‘Your belief that S is perfectly justified, but perhaps not true’” (Rorty 1991e: 128), is meant to mark this point. So one can’t reduce truth and justification (even in ideal circumstances) after all. For most of his career Rorty accounted for the gap between justification and truth by saying that it was “a gap between the actual good and the possible better.” On this account, “to say that what is rational for us now to be believe may not be true, is simply to say that someone may come up with a better idea” (Rorty 1991a: 23). Here Rorty continues to reject the idea that truth is an objective or absolute concept. But in Truth and Progress Rorty takes a new tack, one more in line with Davidson’s view. This can be seen in a criticism that Rorty makes of James’ epistemic theory of truth, which equates truth with ‘what it is better for us to believe’, with beliefs that work in meeting all the local norms of justification. James would, indeed, have done better to say that phrases like ‘the good in the way of belief’ and ‘what is better for us to believe’ are interchangeable with ‘justified’ rather than with true. But he could have gone on to say that we have no criteria of truth other than justification, and that justification and betterness-to-believe will always be relative to audiences (and to ranges of truth candidates)…Granted that ‘true’ is an absolute term, its conditions of application will always be relative. For there is no such thing as belief being justified sans phrase—justified once and for all…[and] there are no beliefs that can be known to be immune to all possible doubt. (Rorty 1998: 2) Notice that Rorty here concedes that “[t]ruth is…an absolute notion in the following sense: ‘true for me but not for you’ and ‘true in my culture but not in yours’ are weird, pointless locutions. So is ‘true then, but not now’” (1998: 2). So now true does not just signal the ‘possibly better’. But this concession is Janus-faced: while truth is absolute, it is precisely this absoluteness that makes it unsuitable to be a goal of inquiry, and makes the concept indefinable and ultimately far less important than we had thought. If truth is absolute it cannot be a goal for inquiry, for a “goal is 9 As Rorty points out, he ‘swings back and forth’ between a view that reduces truth to justification and a minimalism about truth that holds that there is much less to say about the concept than previously thought (see Rorty 1998a: 21). 18 something that you can know that you are getting closer to, or farther away from. But” if truth is absolute “there is no way to know our distance from truth, nor even whether we are closer to it than our ancestors were” (1998: 4). And if truth is so distanced from human practices then clearly it also cannot be defined. As we know from Plato’s Meno to even begin giving a definition requires already having some, perhaps implicit, grasp of the concept to be defined. But if truth is genuinely absolute then we don’t have this grasp. Rorty’s conclusion is that “the very absoluteness of truth is a good reason for thinking ‘true’ indefinable and for thinking that no theory of the nature of truth is possible. It is only the relative about which there is anything to say” (1998: 3). These ideas comprise Rorty’s minimalism about truth. Davidson agrees with Rorty that truth is not definable by simpler and more transparent concepts. But “this does not mean we can say nothing revealing about it: we can by relating it to other concepts like belief, desire, cause, and action” (Davidson 1996: 265). We could put the point in a pragmatist vein by saying that we can provide an account of how the concept of truth is used, how the concept is applied in connection with other concepts in making the speech and action of other subjects intelligible. Rorty recognizes that truth is a concept that is necessarily employed by the radical interpreter. But he argues that the same could be said for other concepts that we use to describe, predict, and interpret behavior. If tracing the pattern of truth in an interlocutor’s behavior is also tracing the pattern of belief, desire, cause, and action, then why, Rorty asks, should Davidson’s theory of radical interpretation be called a theory of truth rather than “‘a theory of complex behavior’ or ‘a theory of justificatory behavior’” (Rorty 1998a: 25n)? Why, he asks “does it seem important to Davidson to think of a Tarski-type theory for a natural language as a truth-theory for that language rather than simply as a way of predicting regularities in the behavior of speakers of that language? 19 Why, given our agreement on the indefinability of ‘true’, does Davidson object to my saying that there is nothing much to said about truth” (Rorty 2000: 370)?10 6. Rorty’s Dichotomies Before answering these questions in the next section, we must first highlight a tension in the above account: Rorty’s point about the descriptive nature of interpretation and his reading of the ‘belief is of its nature veridical’ thesis pull apart. On the one hand, Rorty’s account of radical interpretation is of a purely descriptive practice in which the native’s beliefs are treated as naturalistically describable ‘habits of action’ developed through the their causal interaction with the environment. On the other hand, Rorty’s reading of the ‘belief is of its nature veridical’ thesis makes radical interpretation an intrinsically normative affair, one in which we expand the solidarity of the community through our coming to agree with the native about justifications. Here we do not come to an agreement about habits of action, but about beliefs understood as sentence-like items that gain their content and epistemic status in the inter-subjective space of reasons. This tension is the result of Rorty’s adherence to the Sellarsian injunction to avoid the myth of the Given, the myth that there are given states, in this case beliefs, that are outside of the space of reasons, the space that confers content, but that nonetheless have sematic purport. To avoid this idea we must see beliefs either “from the outside as the field linguist sees them (as causal interactions with the environment) or from the inside as the pre-epistemological native sees them (as rules for action). To abjure tertia is to abjure the possibility of a third way of seeing them—one which somehow combines the outside view and the inside view, the descriptive and the normative attitudes” (Rorty 1991e: 139). 10 In “Pragmatism, Davidson, and Truth” Rorty’s central claim was that Davidson did not think that truth was an especially important concept. This is the essence of Rorty’s ‘pragmatist’ reading of Davidson. In “Is Truth a Goal of Inquiry?” Rorty concedes, in light of various writings where Davidson insists that truth is the central concept of his theory of radical interpretation, that his interpretation was off base. In light of this, Rorty no longer claims to be giving an interpretation of Davidson, but simply to be disagreeing with him about the importance of truth. 20 I agree with John McDowell that the dichotomy between these two points of view structures Rorty’s thinking generally.11 What I want to argue here is that this dichotomy leads Rorty to accept the dichotomy between scheme and content. My claim is not that his view is structured by either the Kantian or language-analytical versions of the dichotomy discussed above, but rather, in Rorty’s own words, by “a certain picture—the picture which Davidson calls ‘the dualism of scheme and content’ and which Dewey thought of as ‘the dualism of Subject and Object’. Both pictures are of disparate ontological realms, one containing beliefs and the other non-beliefs” (1991e: 129). Rorty takes it that he is a critic of this picture, as the “picture of two such realms permits us to imagine truth as a relation between particular beliefs and particular non-beliefs which is non-causal in nature” (1991e: 129). But it is my claim that Rorty, despite his self-understanding, accepts a fundamental dichotomy between these two ontological realms, and therefore accepts a version of scheme-content or subjectobject dualism.12 Rorty thinks that he can’t be committed to a distinction between two ontological realms because if there are no representations but only causal relations between self and world then we can’t answer the question of whether our representations correspond to something independent of us or not. “I have argued that, if one is content to think of the relations between human organism’s, their beliefs, and the rest of the universe, in merely causal terms, rather than dragging in representational relations in addition to causal ones, questions about realism and anti-realism will not arise” (Rorty 1995: 193). Questions about whether the world is, or is not, independent of our minds or vocabularies—which for Rorty is equivalent to the question of whether there are two realms—will not arise because if there are only causal relations between mind and world we cannot make sense of either “the picture of the mind projecting structure onto an unstructured world” or “the idea of the 11 See the first Afterward in McDowell 1996, ‘Davidson in Context’. I think it is questionable whether it is best to call this a dichotomy of scheme and content rather than just a dichotomy of subject and object. However, Rorty calls it a scheme-content dichotomy and I will follow him in this. I am grateful to Neil Gascoingne and Carl Sachs for helping me think through this issue. 12 21 world projecting structure onto, or into, language” (1995: 192). We cannot make sense of these pictures because if we have “relations of justification holding between beliefs and desires, and relations of causation holding between these beliefs and desires and other items in the universe, but no relations of representation” (Rorty 1991c: 96) then nothing intentional or semantic moves across the self-world divide in either direction. Based on this, Rorty rejects the whole set of optical metaphors—projecting as well as mirroring, reflecting, or shadowing— and thus to reject the question Which come first, subject or object? This means rejecting the question Whose contours were there first Language’s or the Word’s? Whose contours are reflecting whose? Rejecting such questions seems to me the cash value of rejecting what Davidson calls ‘the dualism of scheme and content’. (Rorty 1995: 192) In rejecting such questions Rorty avoids two opposite yet co-dependent views: a linguistic idealism in which the world is both representationally and causally dependent on minds or vocabularies; and a metaphysical realism in which there is a single way the world is, the World initself, that can only be represented by a single privileged vocabulary. Both views depend on the concept of representation. But the question of whether one thinks that there are representational relations that bridge the divide between the two ontological realms is orthogonal to the question of whether one thinks there are two such realms. One might think that there are no representations that connect the two realms, but nevertheless that there are two realms. We can see that Rorty has a picture that includes two realms if we consider his account of how the self gets in touch with the world: [I]f we have causal relations (like that between the opening of the door and the acquisition of a belief) holding between the World and the Self, as well as relations of justification (‘being a reason for’) internal to the Self’s network of beliefs and desires, we do not need any further relations to explain how the Self gets in touch with the World, and conversely. We can tell an adequate story about the progress of human inquiry (in all spheres—logic and ethics as well as physics) by describing the continual reweaving of system of belief and desire. This reweaving is made necessary by the acquisition of new beliefs and desires—e.g., the sort which are caused to occur in human beings by such events in the World as the opening of doors. (Rorty 1991d: 120) For Rorty, we should not ask where the new beliefs and desires come from, they just pop up, and when they do the self must accommodate them by changing their web or network of prior beliefs and desires on pain of contradiction or tension (see Rorty1991c: 93). If we wish to justify new beliefs 22 we should not point to the items in the world that they are purportedly about, for “aboutness is not a matter of pointing outside the web. Rather, we use the term ‘about’ as a way of directing attention to the beliefs which are relevant to the justification of other beliefs” (1991c: 97). So intentional, rational, and justificatory relations are internal to the network of belief and desire, and the reweaving of these relations (which is justification) happens when a new belief is caused to occur by an event or object in the world. In saying that a reweaving is brought about by a new belief that is caused by an event or object in the world we are only saying that this event or object is the occasion of the reweaving, not that the specific form the reweaving takes is rationally responsive to the event or object that occasions it. This is because while “there is such a thing as brute physical resistance—the pressure of light waves on Galileo’s eyeballs, or the stone on Dr. Johnson’s boot,” there is “no way of transferring this nonlinguistic brutality to facts, to the truth of sentences” (Rorty 1991b, 81). If our relations to the world are strictly causal relations then the world is mute, i.e., it cannot speak and offer reasons to us. Only other subjects in the space of reasons can do that.13 So here we have a view structured by two sets of relations that are not themselves related: semantic, intentional, and justificatory relations between beliefs in the space of reasons; and causal relations between habits of action and world. To abjure representations is to rule out putting these relations into a single picture, a picture that is both normative and descriptive. It instead requires having two pictures of two distinct realms. One might think that my rendering of the two realms is different than Rorty’s. His rendering concerns beliefs and non-beliefs; whereas mine concerns beliefs and their semantic and epistemic relations to other beliefs, and habits of action and their causal relations to objects. So the second realm, one might say, contains beliefs, just rendered as habits of action, and so is not a realm of ‘non-belief’. But if we can genuinely give naturalistic descriptions of habits of action then they are to be seen as pieces of nature in causal relation to other pieces of nature. In this respect, the second realm is not a realm of belief, but is a realm of causally ordered 13 For more on the idea of the world’s ‘muteness’ see Levine 2010 and Brandom 2011. 23 natural objects, which includes our habits.14 We have here, as McDowell points out, a dichotomy between Reason and Nature, which entails a scheme-content and subject-object dichotomy.15 I said in the introduction that Rorty’s endorsement of these dichotomies pointed his view in an anti-realist direction. I can now say why. Because there is no way of transferring the nonlinguistic brutality of the causal order to facts, Rorty thinks “the hardness of fact…is simply the hardness of the previous agreements within a community about the consequences of an event” (1991b: 80). He thinks this because when “something causal happens…as many facts are brought into the world as there are languages for describing the causal transaction” (1991b: 81). So what fact there is depends on the language game that is being played by a community, and this can’t be determined by the causal transaction itself. The world—here understood as the mundane world of causally interacting objects in space and time rather than the ineffable World-in-itself—cannot determine what language game is being played because it does not project its structure onto, or into, the language.16 What determines the language game that is being played is not the world but the usefulness of the game to advance some of the purposes of language users. But which language games are more useful to advance certain purposes is not something that can be determined a priori, rather it requires action and experimentation—which requires entering into causal interaction with 14 For the classical pragmatists, habits involve a non-conceptual embodied sense. For Rorty, in contrast, they are strictly causal mechanisms. 15 See McDowell 1996: 153. Rorty could deny this conclusion by saying that the two pictures he gives are not pictures of distinct ontological realms. Because he accepts Davidson’s thesis that reasons can be causes—which is just “the claim that a given event can be described equally well in physiological and psychological, non-intentional and intentional terms” (1991d: 114)—he can say that describing beliefs in terms of their rational relations and describing them in terms of their causal relations is just describing the same event from two points of view. I cannot adequately address this complex issue here. My brief response is this: Rorty thinks that it is a deep assumption of the Quine to Davidson tradition that the use of a certain vocabulary does not entail ontological commitment. But I would like to say that if one really disowns the scheme-content distinction then one can no longer neatly separate ‘epistemological’ questions (questions about the standpoint from which we access the world) and ‘ontological’ questions (questions about the nature of the world accessed). I take this to be one of the upshots of Davidsonian triangulation, where subjectivity, inter-subjectivity, and objectivity come as a package. Part of the story I am telling is how, in accepting Davidsonian triangulation, Rorty comes to realize this as well. 16 If the mundane world was thought of as projecting its structure into language such that only a single privileged vocabulary was able to correctly capture it (let’s say the vocabulary of micro-physics), then the mundane world would simply be the ‘World-in-itself’, one posited by a metaphysical type of scientific realism. 24 the world. Perhaps here we have found the place where the causal deliverances of the mundane world have a say in determining the vocabulary that is used to talk about it. But this is not the case, because for Rorty the upshot of our thinking of “linguistic behavior as tool-using” is an account “of language as a way of grabbing hold of causal forces and making them do what we want, altering ourselves and our environment to suit our aspirations. The pragmatist thus exalts spontaneity at the cost of receptivity, as his realist opponent did the reverse” (Rorty 1991b: 81). What language game we play is not determined by the causal deliverances of the mundane world because such deliverances can’t constrain the spontaneity through which we describe and re-describe our wants and purposes, which, in turn, entails different vocabularies and different ways of linguistically grabbing hold of the world’s causal forces. While the world’s causal order perhaps rules out certain games as one’s that simply don’t ‘work’, it does no positive work in determining the language game we use to cope with it. Here, we have the essence of Rorty’s anti-realism. His anti-realism is not articulated by the thought that the existence of the world is dependent on minds or vocabularies. Because he includes the outside view of belief Rorty can genuinely say “most things in the universe are causally independent of us” (Rorty 1998b: 86). But what he can’t say is that these causally independent things have a genuine say in determining the vocabularies we use to talk about them. In the language that Rorty comes to use, our freedom to use a vocabulary to describe and re-describe things is not answerable to the mundane world we are describing or re-describing, it is only answerable to other subjects in the space of reasons.17 And our use of a vocabulary is not answerable to the world because, in not being related, the realm of non-beliefs (including habits)—which is the realm of objectivity and necessity—can have no authority with respect to the realm of belief—which is the realm of subjectivity and freedom. 17 For his rejection of the language of answerability see Rorty 1998c and 1998d. For a critique, see McDowell 2000. 25 7. Triangulation and Truth Let’s come back to Rorty’s question of why it is that Davidson calls his theory a theory of truth. The basic reason is that “[w]ithout a grasp of the concept of truth, not only language, but thought itself, is impossible” (Davidson 2000: 72). But why is this the case? To answer this question we must briefly discuss Davidsonian triangulation. In a series of papers beginning in the eighties Davidson transformed his theory of radical interpretation into a theory of triangulation. In radical interpretation the field linguist triangulates between the native’s utterances and the world. But in this arrangement the interpreter already understands a language and already has thoughts that display intentionality. The theory of triangulation, in contrast, tries to give a general account of how belief, thought, or intentional mindedness emerges überhaupt through the triangular interaction between two creatures and the world. So with triangulation we get an account of the conditions of possibility of thought, and not just a theory of interpretation.18 Specifically, entering into a communicative triangle with another creature about the world makes thought possible because it allows a creature: 1) to distinguish between themselves and the world, between subjectivity and objectivity; and 2) to ascribe to their empirical thought a determinate content or meaning. Let’s briefly describe how these two capacities are made possible by triangulation.19 Davidson takes that for a creature to have thought or belief “demands” that they appreciate “the contest between true belief and false, between appearance and reality, mere seeming and being” (2001f: 209). A creature grasps this distinction when they grasp the possibility that their thought can be in error. Error is essential for grasp of the distinction between appearance and reality because in being different from what one takes it to be, the world holds itself out as independent, objective, and 18 For more on the relation of triangulation to radical interpretation, see Ramberg 2001. For how triangulation is a type of transcendental argument, see Sachs 2009 and the literature cited there. 19 Davidson’s theory of triangulation is very complex, involving pre-communicative and communicative stages. Here, I only discuss, in the most rudimentary way, the communicative stage. 26 separate from one’s perspective on it. But how does one come to grasp the possibility that one’s thought or belief can be in error? Davidson argues, in a Wittgensteinian fashion, that a grasp of error, and consequently of objective truth, requires interpersonal communication: The source of the concept of objective truth is interpersonal communication. Thought depends on communication. This follows at once if we suppose that language is essential to thought and we agree with Wittgenstein that there cannot be a private language. The central argument against private languages is that, unless a language is shared, there is no way to distinguish between using the language correctly and using it incorrectly, only communication with another can supply an objective check. If only communication can provide a check on the correct use of words, only communication can supply a standard of objectivity in other domains…We have no grounds for crediting a creature with the distinction between what is thought to be the case and what is the case unless the creature has the standard provided by a shared language; and without this distinction there is nothing that can clearly be called thought. (Davidson 2001f: 209-10) So a creature grasps the distinction between what is thought to be the case and what is the case, i.e., grasps the concept of objective truth without which there cannot be thought, when they come to be able to act in accord with standards of objectivity provided by a shared language.20 But Davidson asks, “why should an interpersonal standard” rendered through the shared use of a language “be an objective standard, that is, why should what people agree on be true?” (2001f: 212). Here the other part of Davidson’s account of triangulation, the one concerning determinate content, becomes important. All living creatures classify the world by responding differentially to stimuli. But, Davidson argues, patterns of differential responsiveness to stimuli cannot deliver a theory of determinate content, a theory that can give an answer to the question of what exactly the creature is responding to. For what is the criterion by which it is said that that a creature’s response to things is similar to their past responses? This criterion cannot be derived from the creature’s responses; it can only come from the responses of an observer to the responses of the creature. And it is only when an observer consciously correlates the responses of another creature with objects and events of the observer’s world that there is any basis for saying the creature is responding to those objects or events rather than other objects or events. (Davidson 2001f: 212) 20 Although Davidson invokes Wittgenstein here, it is important to point out that for him such shared standards are not provided by rules, but by overlapping empirical theories of truth. 27 For a creature’s thoughts or words to have a determinate content they must not only respond differentially to stimuli but must consciously correlate those responses with the observed responses of another creature. This correlation is necessary because until a “triangle is completed connecting two creatures, and each creature with common features of the world, there can be no answer to the question whether a creature…is discriminating between stimuli at the sensory surfaces or somewhere further out, or further in” (2001f: 212).21 So triangulation is necessary for objective content because through their conscious correlations two creatures are able to identify the ‘common cause’ of their responses and place it in “an objective location in a common space” (2001b: 119). Now the indeterminacy of the stimulus has been overcome, and our thought and speech have a determinate objective content. The upshot of the account of objective truth and of determinate content is that there is a twoway constitutive dependence between inter-subjectivity and objectivity: While our concept of what is objectively true or false is dependent upon our taking part in intersubjective communication, for such communication to have content “requires and assumes knowledge of a shared world of objects in a common time and space” (2001b: 118). As such, the concept of objective truth does not just apply within the communicative relation subject to subject, for there would be no such relation if it did not triangulate on a shared world in which common causes are placed. Because of this the ‘belief is of its nature veridical’ thesis cannot be read as only applying to out shared beliefs, it must also apply to the world those beliefs are inextricably about. 8. Changes in Rorty’s View Late in his career, through an exchange with Bjorn Ramberg concerning Davidsonian triangulation, Rorty came to realize precisely this: the ‘belief is of its nature veridical’ thesis does not just apply to 21 Davidson thinks that his perceptual externalism by itself cannot handle the problem, developed by Quine, of whether stimuli is proximal or distal. This is one reason he develops the theory of triangulation, and is one reason why his causal account of content is only the starting point for his theory of meaning and not its terminus. 28 the convergence of justifications but to the fact that the beliefs that we converge on (generally) get things right.22 [W]hen Davidson argues that most of anybody’s beliefs must be true, he is not just saying (as I sometimes have been tempted to construe him) that most of the beliefs of anybody whom we can treat as a language-user must accord with most of our own beliefs. He is saying that most of what anybody says about whatever they are talking about gets that thing right. (2000: 374) What Rorty realized is that it is wrong “to go from criticism of attempts to define truth as accurate representation of the intrinsic nature of reality to a denial that true statements get things right” (Rorty 2000: 374). So Davidson’s theory is rightly named a theory of truth because in triangulation my interlocutor and I generally get things right not only about each other’s beliefs but about the objective world itself. What are the ramifications of Rorty’s admission that we need a notion of ‘getting things right’? Rorty correctly names two. First, it undermines Rorty’s long held doctrine, laid out above, that we only have causal relations to the world and not intentional, semantic, or justificatory ones. For if our mental states only have determinate content through our causal relations to stimuli being taken up into the communicative triangle, then the semantic terms ‘true’ and ‘refers’ cannot just signify intra-linguistic relations, as Rorty had always thought. Instead, the normative relations that they signify should be seen as “hovering over the whole process of triangulation” rather than just locating them “at one corner of the triangle—where my peers are” (2000: 379). But if this is so, then word-world relations cannot just be causal. But since Rorty’s anti-representationalism is still in place we cannot say that they are correspondence relations to ‘reality-in-itself’. Rather they signify 22 See Ramberg 2000 and Rorty 2000. In this paper I only have space to note the consequences of the exchange between Ramberg and Rorty for our topic and not the details of the exchange. Briefly, the exchange concerned the issue of normativity and mind. Ramberg was intent to convince Rorty that Davidson accepts the indeterminacy of translation thesis not because he thinks that the vocabulary of psychology is irreducible, or that minds are ontologically unique, but because the normative ‘vocabulary of agency’ is inescapable due to its necessary use in triangulation. When through this exchange Rorty came to see that his interpretive focus on “the famous Brentinian irreducibility of the intentional” prevented him from understanding that Davidson’s real concern was “the inescapability of the normative” (Rorty 2000, 370-1), he came to accept the normativity of the mental, which led to his new reading of the ‘belief is veridical’ thesis. For more on this see Stout 2007, Levine 2008, and Sachs 2009. 29 relations between an expression and an object as that object is rendered within the communicative triangle, and so “are neither causal nor representational” (2000: 374). Second, the admission that we need a notion of ‘getting things right’ leads Rorty to recognize that there is a ‘truth in realism’ in addition to a ‘truth in pragmatism’: What is true in pragmatism is that what you talk about depends not on what is real but on what it pays you to talk about. What is true in realism is that most of what you talk about you get right. Would there be snow if nobody ever talked about it? Sure. Why? Because according to the norms we invoke when we use ‘snow’ we are supposed to answer this question affirmatively. (If you think that that glib and ethnocentric answer is not good enough, that is because you are still in the grip of the scheme-content distinction. You think you can escape the inescapable, cut off one corner of Davidson’s triangle, and just ask about a relation called ‘correspondence’ or ‘representation’ between your beliefs and the world.) (2000: 374) Traditional realism involves two thoughts, one modest, one presumptuous.23 The modest thought concerns the independence of reality; that how matters stand with the world, and what beliefs about it are true, are settled independently of what we think. The presumptuous thought is that even though exactly which beliefs about the world are true is a question settled independently of our thought, nevertheless our beliefs are able to capture in their net a substantial portion of the truth. The ‘truth in realism’ pointed at by Rorty involves both of these thoughts: snow would exist even if we never talked about snow (the modest thought), and our talk about snow is able (and is in fact likely) to get snow right (the presumptuous thought). But Rorty’s version of either thought is not that of traditional realism. He grounds the independence claim about snow not in a correspondence relation but in the idea that the content of the concept of snow is articulated by norms that include the idea of snow as something that is independent of one’s thought. The central place of truth in triangulation dictates this result. And he grounds the presumptuous thought not in the idea that we accurately represent or map the world, but in in the veridicality of belief thesis. If most of our beliefs are true then they are able to capture in their net a substantial portion of the truth. The ‘truth in pragmatism’, in contrast, involves two ideas: 1) that what we talk about is determined by what pays, i.e., what satisfies our interests and purposes, rather than what is really 23 Here I follow Crispin Wright’s way of characterizing realism. See Wright 1992: 1-2. 30 real; and 2) that truth is not a goal of inquiry, even in light of the fact that most of our beliefs get things right. First, the real cannot determine what it is that we talk about because there are not norms within the communicative triangle to talk about it, only the particular things upon which we triangulate. The concept of reality-in-itself has no content. As such, Rorty continues to espouse the radical idea that “[n]o area of culture, and no period of history, gets Reality more right than any other. The differences between areas and epochs is their relative efficiency at accomplishing various purposes” (2000: 375). All areas of culture, and all periods of history, mostly get things right due to the fact that belief is of its nature veridical. But if we can’t talk about Reality insofar as the concept has no content then there is no way of determining which areas of culture or periods of history are closer to, or further away from it. Second, even though most of our beliefs get things right and are therefore true, truth is still not a goal of inquiry. This is because “the connections between the concept of truth and those of meaning and belief has nothing whatever to do with the question of whether, or how, we can tell when a belief is true” (Rorty 1998a: 23). For truth to be a goal of inquiry we would need to know what truth was and what it would be to achieve it. But if truth is absolute then we cannot know either of these two things. So although due to the connections between truth, meaning, and belief most of our beliefs are true, we can never know with certainty which beliefs those are. As Davidson puts the point: “We know many things, and will learn more; what we will never know for certain is which of the things we believe are true. Since it is neither visible as a target, nor recognizable when achieved, there is no point in calling truth a goal” (Davidson 2005: 6). And as Rorty puts it, “since truth swings free of justification, belief, success and everything else save meaning, truth cannot be a goal or a value” (Rorty 1998e, 69n). In this he and Davidson remain in agreement before and after his change in view. 31 9. Conclusion As Jeffrey Stout says, the admission that there is a truth in realism, is “as surprising as any in the entirety of Rorty’s published writing” (Stout 2007: 17). It is surprising not only because of Rorty’s insistent rejection of the rhetoric of objectivity, but also because of his long held claim that the realism-anti-realism distinction depends on representationalism. But now Rorty accepts some notion of objectivity, and he accepts that the realism-anti-realism distinction has some purchase, even after the rejection of representationalism. Let us take these is reverse order. First, instead of flatly rejecting the realism-anti realism distinction Rorty now argues that there is some truth on both of its sides. There is a truth in realism and a truth in pragmatism, which, we could say, expresses the truth in anti-realism. Here Rorty breaks with his usual way of dealing with a philosophical dichotomy, not by simply finding it to be the vestige of an optional philosophical picture that ought to be left behind, but by appropriating the partial truth that each of its sides express. Here we have a more dialectical, less eliminativist, approach. I do not mean to suggest that this approach come to predominate in Rorty’s late work. But it does show that Rorty’s late work involved a trajectory toward a quite different meta-philosophical position than the one that he espoused for most of his career. Second, Rorty now accepts some notion of objectivity, of getting things right. But we have to be careful in how we parse his new found commitment to this ideal. Here is how Stout puts it: I am tempted to sum up what he is doing…by saying that he is trying to formulate a nonnarcissistic pragmatism, a pragmatism that can do justice to the objective dimension of inquiry. For he is describing inquiry as a human practice that answers to human interests, but also as portraying as an expression of human interests the distinction between getting one’s subject matter right and merely holding beliefs about that subject matter (or holding beliefs that one’s peers would let one get away with holding). (Stout 2007, 17) But does the admission that belief is of its nature veridical entail that we have an interest in getting things right in addition to achieving agreement with our fellows? I have my doubts because as we saw above, Rorty, even after his change in view, remains resistant to the notion that having true 32 beliefs about something is a distinct goal of inquiry. But how can we have an interest in getting things right over and above arriving at agreement with our fellows if getting things right cannot be a goal of our inquiries? Stout recognizes the force of the question and consequently argues that Rorty, having admitted that we have an interest in getting things right, should go all the way and admit truth to be a goal of inquiry. But I think this gets things hermeneutically backwards: Rorty’s continued dismissal of truth as a goal of inquiry should cast doubt on the notion that Rorty thinks that we have an interest in getting things right. While Rorty accepts that most of our beliefs in fact get things right (because belief is of its nature veridical), he does not think that we have an interest in getting things right over and above agreement with our fellows. This conclusion is buttressed by the fact that in an exchange with Stout that took place after the engagement with Ramberg, Rorty backslides altogether on his willingness to utilize “the rhetoric of objectivity” (Rorty 2010, 422). Here is what he says: “As I see it the whole point of pragmatism is to insist that we human beings are answerable only to one another. We are answerable only to those who answer to us—our conversational partners. We are not responsible either to the atoms or to God, at least not until they start conversing with us” (Rorty 2010, 423). Stout chalks this up to the fact that the rhetoric of objectivity is not consistent with Rorty’s prophetic persona, the persona Rorty adopts when rejecting wholesale the claims of the philosophical tradition (see Stout 2007, 9-10). While this is true, I think there is a deeper reason for the backsliding, namely that Rorty does not seem fully cognizant of what is required by his own change in view, especially with respect to the dichotomies. For example, in a response to Davidson written contemporaneously with his response to Ramberg, Rorty interprets his dictum that semantic relations are ‘neither causal nor representational’ in this way: The point of this doctrine [of triangulation] is that you cannot get along with just holistic inferential relations between beliefs and statements (as coherence theorists tried to do) nor with atomic relations of being-caused-by (as realists fixated on perception still try to do). You have to play back and forth between causation and inference in a way which does not permit any of the corners of a triangle to be independent of any of the others. (Rorty 2000a, 78) 33 So to say that the relations between our thoughts and utterances and the world are neither causal nor representational is not to say that these relations are somehow causal and inferential, natural and rational, at one and the same time. It is just to say that we work back and forth between causation and inference such that a belief enters two distinct streams of relation, a causal one and a normative one. While none of the corners of the triangle, self, other, and object, are independent of one another insofar as without all three determinate content would not be possible, such content does not depend on there being semantic relations that bridge the divide between cause and norm, nature and reason. So the dichotomy between the two disparate ontological realms, which underlies Rorty’s version of the realism/anti-realism distinction, remains in place. So like with the change in his meta-philosophical view, we should not say that Rorty definitively overcomes the dichotomies. What we should say is that his work involves a trajectory to do so. In one respect this is not surprising. It is not surprising because it was always Rorty’s stated view that the dichotomies between scheme and content, subject and object, realism and anti-realism should not get a grip on our thinking. Seeing a truth in realism, and accepting that words have both causal and normative relations to the world, brings Rorty’s actual view far closer into alignment with this stated view. So instead of seeing Rorty’s change in view as expressing a radical break with his prior work, we could see it as part of Rorty’s living up to his deepest insight. 34 Works cited: Baghramian, M. (1990), ‘Rorty, Davidson, and Truth’, Ratio 3 (2). Brandom, R. (2000), Rorty and his Critics. Malden, MA: Blackwell Press. ——(2011), ‘Linguistic Pragmatism and Pragmatism about Norms: An Arc of Thought from Rorty’s Eliminative Materialism to His Pragmatism’, in Perspectives on Pragmatism: Classical, Recent, and Contemporary. Cambridge: Harvard University Press. Child, W. (1994), ‘On the Dualism of Scheme and Content’, Proceedings of the Aristotelian Society, New Series, Vol. 94. Davidson, D. (1984), ‘On the Very Idea of a Conceptual Scheme’, in Truth and Interpretation. Oxford: Oxford University Press. ——(1984a), ‘The Method of Truth in Metaphysics’ in Davidson 1984. ——(1990), ‘The Structure and Content of Truth’, Journal of Philosophy 87 (6). ——(1996), ‘The Folly of Trying to Define Truth’, Journal of Philosophy 93 (6). ——(2000), ‘Truth Rehabilitated’, in Brandom 2000. ——(2001), Subjective, Intersubjective, Objective. Oxford: Oxford University Press. ——(2001a), ‘The Myth of the Subjective’, in Davidson 2001. ——(2001b), ‘The Second Person’, in Davidson 2001. ——(2001c), ‘A Coherence Theory of Truth and Knowledge’, in Davidson 2001. ——(2001d), ‘Afterthoughts’, in Davidson 2001 ——(2001e), ‘Epistemology and Truth’, in Davidson 2001. ——(2001f), ‘Three Varieties of Knowledge’, in Davidson 2001. —— (2004), ‘The Problem of Objectivity’, in Problems of Rationality. Oxford: Oxford University Press. ——(2005), ‘Seeing Through Language’, in Truth, Language, and History, Oxford, Oxford University Press. Farell, F. (1995), “Rorty and Anti-Realism,” in Rorty and Pragmatism, ed. Saatkamp. Nashville: Vanderbilt University Press. Gascoigne, N. (2008), Richard Rorty. Cambridge: Polity Press. 35 Levine, S. (2008), ‘Rorty, Davidson, and the New Pragmatists’, in Philosophical Topics 36 (1). ——(2010), Rehabilitating Objectivity: ‘Rorty, Brandom, and the New Pragmatism’, in The Canadian Journal of Philosophy 40 (4). McDowell, J. (1996), Mind and World. Cambridge. MA: Harvard University Press. ——(1998), ‘Anti-Realism and the Epistemology of Understanding’, in Meaning, Knowledge, and Reality. Cambridge: Harvard University Press. ——(2000), ‘Towards Rehabilitating Objectivity’, in Brandom 2000. Ramberg, B. (2000), ‘Post-Ontological Philosophy of Mind: Rorty versus Davidson’, in Brandon 2000. ——(2001), ‘What Davidson Said to the Skeptic or: Anti-Representationalism, Triangulation, and the Naturalization of the Subjective’, in Interpreting Davidson, ed. Kotatko, Pagin, and Segal. Stanford: CSLI Press. Rorty, R. (1979), Philosophy and the Mirror of Nature. Princeton, NJ: Princeton University Press. ——(1982), ‘The World Well Lost’, in The Consequences of Pragmatism. Minneapolis: University of Minnesota Press. ——(1991), Objectivity, Relativism, and Truth. Cambridge: Cambridge University Press. ——(1991a), ‘Solidarity Or Objectivity’, in Rorty 1991. ——(1991b), ‘Texts and Lumps’, in Rorty 1991. ——(1991c), ‘Inquiry as Recontextualization’, in Rorty 1991. ——(1991d), ‘Non-Reductive Physicalism’, in Rorty 1991 ——(1991e), ‘Pragmatism, Davidson, and Truth’, in Rorty 1991. ——(1991f), ‘Representation, Social Practice, and Truth’, in Rorty 1991. —— (1995), ‘Response to Farrell’, in Rorty and Pragmatism, ed. Saatkamp. Nashville: Vanderbilt University Press. ——(1998), Truth and Progress. Cambridge: Cambridge University Press. ——(1998a), ‘Is Truth a Goal of Inquiry? Donald Davidson versus Crispin Wright’, in Rorty 1998. ——(1998b), ‘Charles Taylor on Truth’, in Rorty 1998. ——(1998c), ‘Robert Brandom on Social Practices and Representations’, in Rorty 1998. 36 ——(1998d), ‘The Very Idea of Human Answerability to the World: John McDowell’s Version of Empiricism’, in Rorty 1998. ——(1998e), ‘Davidson Between Tarski and Wittgenstein’, Critica, 38: 49-71. ——(1999), ‘Davidson’s Mental-Physical Distinction’, in Donald Davidson (Library of Living Philosophers), ed. L. E. Hahn. Peru, IL: Open Court Press. ——(2000), ‘Reply to Ramberg’, in Brandom 2000. ——(2000a), ‘Reply to Davidson’, in Brandom 2000. ——(2010), ‘Comment on Jeffry Stout’s Democracy and Tradition’, in ‘Pragmatism and Democracy: Assessing Jeffry Stout’s Democracy and Tradition’, in J. Springs (ed.) Journal of the American Academy of Religion, 78: 418-425. Sachs, C. (2009), ‘Natural Agents: A Transcendental Argument for Pragmatic Naturalism’, in Contemporary Pragmatism 6 (1). Stout, J. (2007), ‘On Our Interest in Getting Things Right: Pragmatism without Narcissism’, in New Pragmatists, ed. C. Misak. Oxford: Oxford University Press. Wright, C. (1994), Truth and Objectivity, Cambridge, Harvard University Press.