Skip to navigation – Site map

Does Rorty’s Pragmatism Undermine Itself?

James Tartaglia


Paul Boghossian and Hilary Putnam have presented arguments designed to show self-referential difficulties within Rorty’s pragmatism. I respond to these arguments by drawing out the details of the pragmatic account of justification implicit within Rorty’s writings, thereby revealing it to be a sophisticated form of relativism that does not undermine itself. In Section I and II, I motivate my strategy of attributing a positive position to Rorty in order to respond to detailed, analytical arguments such as those of Boghossian, and present an outline of this position, agreeing with Rorty’s critics that it can be justifiably classified as a form of relativism. Sections III to V concern the detail of Boghossian’s argument, in which I show that Boghossian’s contention that Rorty’s rejection of all absolute justification is inconsistent can be satisfactorily answered by explaining the differences between “epistemic systems” in terms of the different purposes they serve. Then in Sections VI to VIII, I further develop Rorty’s account of justification in order to answer Putnam’s charge that Rorty tries to say “from a God’s-Eye View there is no God’s-Eye View.” I reject Rorty’s own “social-reformer” response to this argument, but show that it can be satisfactorily answered by distinguishing two integrated components within Rorty’s pragmatism, one holistic and coherentist, and the other causal and social-evolutionary.

Top of page

Full text

  • 1 Rorty always refused the label except, curiously, in one of the last lectures he delivered before h (...)

1“‘Relativism’,” Richard Rorty once said, “is the view that every belief on a certain topic, or perhaps about any topic, is as good as every other,” immediately adding that “[n]o one holds this view” (Rorty 1982: 166). Given this understanding of “relativism,” it is not surprising that Rorty refused the label throughout most of his career, despite the fact that philosophers generally recognise a variety of considerably more nuanced forms of relativism, some of which Rorty himself would seem to have been committed to (see Miller 2002).1 Rorty’s usual attitude to the term “relativism,” however, was that it was a term of abuse used by foundationalist philosophers committed to the view that beliefs are justified by the representational relations they bear to the world, against pragmatists and other anti-foundationalists who hold that beliefs are justified holistically by societal agreement (Rorty 1999: xvi-xvii). The reason that anyone who denies that beliefs can be grounded upon something more solid than ongoing conversation is charged with relativism, according to Rorty, is that foundationalists assume that if our beliefs cannot be so grounded, then they must all be treated as equally valid. This, however, he regarded as simple scare-mongering rooted in an overestimation of the cultural significance of philosophy, for we need not become relativistic about physical science and democracy, for example, simply because epistemological theory cannot demonstrate their objective superiority to witchcraft and dictatorship; this kind of higher level theorising is irrelevant to our commitment to “real theories,” which is acquired instead through consideration of their “concrete advantages and disadvantages” (Rorty 1982: 168).

2Other philosophers, however, regard relativism not as a straw man, but as a real and worrying social phenomenon that has been encouraged by arguments against epistemological foundationalism; it is the felt need to counteract this influence which motivates Paul Boghossian’s Fear of Knowledge: Against Relativism and Constructivism. Boghossian thinks social constructivist theories of knowledge, of which his main example throughout the book is Rorty’s “epistemic relativism,” have exercised a pernicious influence on culture, by undermining the privilege traditionally accorded to scientific knowledge, and thereby disseminating the view that there are “many other, radically different yet equally valid ways of knowing the world,” such that we must accord “as much credibility to archeology as to Zuni creationism, as much credibility to evolution as to Christian creationism” (Boghossian 2006: 4-5; see also Blackburn 2005: ix). In order to undermine Rorty’s “epistemic relativism,” Boghossian develops a version of the “oft-repeated traditional objection,” that “any relativistic thesis needs to commit to there being at least some absolute truths; yet what a global relativism asserts is that there are no absolute truths. Hence, a global relativism is bound to be incoherent” (Boghossian 2006: 53).

3Whether or not he realised it, by adopting this argumentative strategy Boghossian was, in a sense, agreeing with Rorty, who himself had said that relativism could be refuted by “some variant of the self-referential arguments Socrates used against Protagoras” (Rorty 1982: 167). Remember, however, that Rorty denied that he or anyone else was a relativist: “such neat little dialectical strategies only work against lightly-sketched fictional characters,” he went on to say (ibid.). In this I am in almost full agreement with Rorty, since my defence of his pragmatism against Boghossian’s arguments, which will take up Sections III-V, will show that the “epistemic relativism” Boghossian targets is indeed “lightly-sketched,” and fails to do justice to the full resources of Rorty’s position; I say “almost full agreement” only because, as we shall see in Section II, there is no seriously disputing that Rorty’s position was a form of relativism. Rorty’s relativism, however, was not the unsophisticated kind that undermines itself, and this will become even clearer in Sections VI-VIII, when we turn to the best-known and most influential self-referential argument against Rorty’s pragmatism, first formulated by Hilary Putnam. Again, my tactic will be to show that when we are clear about the detail of Rorty’s position, we see that it has the resources to respond to this kind of objection.

4There are two main aims to this paper, then. The first is to answer the persistent suspicion among philosophers that Rorty’s pragmatism undermines itself, which has reached its most sophisticated expression with Boghossian’s arguments, but which remains most closely associated with Putnam’s argument. The second is to make Rorty’s pragmatic account of justification more explicit than Rorty himself ever did. The connection between the two is that it is only in drawing out the details of Rorty’s account that its resources for responding to the arguments of Boghossian and Putnam come into focus.

5Now this project may be of interest to those inclined to believe that Boghossian and Putnam pinpointed genuine self-referential problems with Rorty’s position, but to those less concerned by these arguments, and already sympathetic to Rorty’s pragmatism and overall metaphilosophical stance, it might seem entirely misguided to try to defend Rorty by engaging analytic philosophers like Boghossian on their own terms, and thereby burdening Rorty with an “account” of justification to defend. For surely, the thought goes, Rorty wanted to undermine the kind of “logic-chopping” epistemological debates engaged in by philosophers like Boghossian, and bring philosophy back to socially useful questions; if his positive comments on matters like the nature of justification were rather piece-meal and fragmented, then, this was because he wanted to avoid being pigeon-holed within such debates.

6This reaction, though understandable, reflects a certain ambiguity and to some extent confusion within Rorty’s own writings. For on the one hand, he continued to engage with and comment on contemporary analytic debates right up to the end of his career (see Rorty 2007, part III), but on the other, he was a critic of such debates, urging his colleagues to always connect up their thinking with matters of social practice. Rorty himself tried to reconcile these two stances with the idea that his interventions were justified so long as they promised social usefulness, even if this amounted to nothing more than the usefulness of closing off socially useless debates. However as I will argue in Sections VII and VIII, Rorty did not need to defend his pragmatism solely in terms of social usefulness, and in fact, was on stronger ground when he did not. As such, he had no reason to be shy about fully engaging with technical philosophical debates, and there is no reason for hesitancy about elaborating a detailed Rortian position in response to detailed objections such as those of Boghossian. Granted, if Rorty’s pragmatism were ever to win widespread societal ascent, the one long-term social effect of this might be to dampen-down interest in technical philosophical debates, as Rorty hoped. But appealing to this hoped-for effect was no way to win an argument.

7Boghossian reads Rorty’s discussion of the controversy between Galileo and Bellarmine in Philosophy and the Mirror of Nature (Rorty 1979: 327-33) as a defence of epistemic relativism, and uses the example as the basis for his own arguments. The controversy arose because Galileo had defended and developed Copernican astronomy, using the new type of telescope he invented to make observations of astronomical phenomena, such as the moons of Jupiter and the phases of Venus, that could not readily be accommodated within the Ptolemaic system. Galileo subsequently found himself accused of heresy, since certain passages in the Bible, such as the statement at Psalms 104: 5 that the Earth “can never be moved,” were regarded as divine endorsements of Ptolemy. He went to Rome to defend himself, where the Ptolemaic orthodoxy was defended for the Vatican by Bellarmine, a Cardinal and Aristotelian philosopher, whose reaction to the affair was considerably more moderate than some of his colleagues, the most notorious of whom, Cremonini, is reputed to have refused to look through Galileo’s telescope on the grounds that the Bible is a better source of evidence in astronomy (De Santillana 1958: 28-9). Bellarmine, however, was prepared to officially endorse Copernican astronomy as a useful mathematical device for simplifying astronomical calculations, so long as Galileo made it clear that the Earth did not literally revolve around the Sun. Galileo took the opposite line: he held that certain passages in the Bible were poetical, and hence were not to be taken literally.

8Rorty interprets this as a dispute over standards of evidence. Galileo and other new scientists were trying to limit the evidential scope of scripture; they wanted to keep religion and science separate, with those parts of scripture conflicting with science to be construed non-literally. Bellarmine and other churchmen, on the other hand, were trying to limit the evidential scope of the new science; they could see its power, but thought it would have to be construed non-literally whenever it conflicted with the word of God. The end result was that Galileo “won the argument” (Rorty 1979: 331), thereby setting up a clear demarcation between science and religion that curtailed the epistemic authority of religion. Rorty’s principal claim, however, is that Galileo did not win because Bellarmine was being “illogical or unscientific” (ibid.: 328), since the evidential standards which lead us to regard Bellarmine’s scriptural considerations as irrelevant to astronomy were not then extant. As such, there was no fact of the matter to determine that Galileo’s position was justified and Bellarmine’s was not at the time of the controversy, since there existed no wider “epistemic system,” in Boghossian’s terminology, or “grid,” in Rorty’s, to render both positions commensurable and decide in favour of Galileo. The achievements of scientists like Galileo led to the development of a new “grid” that counted Galileo’s position as rational and Bellarmine’s as irrational, but according to Rorty’s explicitly Kuhnian position, the paradigm shift from Scholasticism to modern science was not itself rational. Rorty later elaborated this view by arguing that such transitions are to be understood in terms of Darwinian evolution, since cultural evolution “takes over from biological evolution without a break” (Rorty 1999: 75); it is in this sense, then, that Galileo “won the argument,” namely that his ideas were found fruitful within the changing cultural environment of seventeenth century Europe, whereas Bellarmine’s ideas adapted less well, were marginalized, and then were gradually forgotten.

9This discussion illustrates what Boghossian calls Rorty’s “epistemic relativism,” because Rorty thinks the disagreement between Galileo and Bellarmine was not rationally resolvable: the arguments of each were justified relative to their own epistemic system, but not that of their interlocutor. Rorty consistently rejected this label for his position, but the reasons he gave were not compelling. One reason, which we have already encountered, was that he usually used “relativism” to denote only simple and self-refuting relativism about truth; but we are of course free to use the term more broadly so as to include any view that relativises truth or justification to an audience. Another reason he gave for rejecting the label was that he denied holding an epistemological position: “Not having any epistemology, a fortiori [the pragmatist] does not have a relativistic one” (Rorty 1991: 24). Again, however, this is simply a case of Rorty adopting an overly strict definition for polemical purposes; he is using “epistemology” to mean “foundationalist epistemology,” so that he can subsequently disclaim epistemological commitment, and thereby reinforce his call for an end to systematic epistemological research, which he considered socially useless (cf. Rorty 1979: 315). As I shall be trying to show throughout this paper, however, Rorty did have an anti-foundationalist epistemology which can profitably be pieced together from his various claims and counter-claims.

10The only substantial reason Rorty ever gave for denying that he was a relativist was his “ethnocentrism” (Rorty 1991: 23 & ff.), according to which we must endorse the epistemic norms of the contemporary liberal West, and reject the relativistic qualification that these are only our norms, with those of other societies counting as equally valid. This is because the latter, relativistic claim would be as much an attempt to “get outside our beliefs and our language” (Rorty 1979: 178) as the absolutist’s claim that the norms of some societies are objectively superior to others; it was with regard to this point that Susan Haack once memorably described Rorty as a “tribalist” rather than a relativist (Haack 1993: 92). Now ethnocentrism is certainly an integral part of Rorty’s position, for he never intimates equality between Galileo and Bellarmine’s astronomical beliefs, but rather endorses Galileo’s position on the grounds that we are his “heirs” (ibid.: 330). Nevertheless, there is also quite evidently a strong element of relativism to this position, given that Rorty thinks that a specification of the audience for a view, namely the fact that we are Galileo’s “heirs,” is required to establish that the view is justified. He reveals this relativity when he invokes historical and counterfactual considerations to remind us of the contingency of the standards of justification we apply. Thus even though Galileo was right, Rorty insists, the epistemic standards we use to determine this might have been different, and were not there to be appealed to at the time of the controversy, since these standards were not determined by an objective truth about rationality and the world, but rather by the contingent twists and turns of a historical conversation that could have gone differently. Thus he says, ethnocentrically, that as we find ourselves in the present day, Galileo’s arguments are justified not only to us, but to anyone who might consider them, but he adds the relativistic qualification that the arguments were not justified to audiences before Galileo “won the argument,” and strongly suggests that they would not have been justified at all had Bellarmine won instead. So long as it is borne in mind that Rorty’s position is both ethnocentric and relativistic, the latter element revealed most clearly by the commitment that Bellarmine’s argument is not “illogical or unscientific” except by reference to our justificatory “grid,” then, Boghossian’s label “epistemic relativism” is perfectly acceptable.

11Boghossian extracts the following formulation of Epistemic Relativism from Rorty’s discussion (Boghossian 2006: 73, 84-5):

(A) There are no absolute facts about what belief a particular item of information justifies. (Epistemic non-absolutism)

(B) If a person, S’s, epistemic judgements are to have any prospect of being true, we must not construe his [or her] utterances of the form
“E justifies belief B”
as expressing the claim
E justifies belief B
but rather as expressing the claim:
According to the epistemic system C, that I, S, accept, information E justifies belief B. (Epistemic relationism)

  • 2 I will capitalise “Epistemic Relativism” when discussing this formulation. Rorty would clearly acce (...)

(C) There are many fundamentally different, genuinely alternative epistemic systems, but no fact by virtue of which one of these systems is more correct than any of the others. (Epistemic pluralism)2

12Boghossian argues firstly that Epistemic Relativism is incoherent, since accepting that Galileo and Bellarmine’s arguments were relatively justified (Epistemic relationism) implies an acceptance of absolute justification (contra Epistemic non-absolutism), secondly, that it is incoherent because an acceptance of Epistemic non-absolutism implies that there are facts which favour some epistemic systems over others (contra Epistemic pluralism), and thirdly, that Rorty has failed to find a genuine case to support Epistemic pluralism, since Galileo and Bellarmine’s dispute took place within the same epistemic system. It is to these arguments that we shall now turn.

13Boghossian’s first argument (ibid.: 84-7) aims to show that relative justification presupposes absolute justification, and hence that Epistemic Relativism’s commitment to both (A) and (B) is incoherent. Boghossian premises the argument on a claim about the nature of epistemic systems, henceforth (ES). (ES) claims that an epistemic system consists of a set of epistemic principles which are themselves “just more general versions of particular epistemic judgements” (ibid.: 85). So, for instance, Galileo’s judgement that its visually appearing to him that Jupiter has moons justifies him in believing that Jupiter has moons, is just a particular application of the general principle that a visual appearance of something provides good prima facie justification for believing it to be the case. According to (ES), then, epistemic systems are entirely made up of general epistemic principles of this kind.

14The argument proceeds as follows. According to (B), an utterance of

“Copernicanism is justified by Galileo’s observations”

15is to be interpreted as expressing the claim:

According to the epistemic system C, that I, S, accept, Copernicanism is justified by Galileo’s observations.

16Given (ES), however, S’s acceptance of C amounts to an acceptance of a set of epistemic principles such as the following:

For any observational proposition p, if it visually seems to S that p and circumstantial conditions D obtain, then S is prima facie justified in believing p.

17Such principles do not state the conditions under which propositions are justified relative to an epistemic system, Boghossian argues, since they are the propositions that the epistemic systems themselves are composed of, and so they must state the conditions under which a proposition is absolutely justified. According to (A), then, all such principles must be false. The Epistemic Relativist is thus committed to the view that all epistemic systems are composed of “uniformly false propositions” (ibid.: 86). According to (B), however, for Galileo’s observations to even provide a relative justification for Copernicanism, S must accept epistemic system C. Consequently Boghossian concludes that Epistemic Relativism is “incoherent” (ibid.: 87): it requires us to reject all claims of absolute justification, and yet in order for our beliefs to be even relatively justified, we must accept the claims to absolute justification constitutive of our own epistemic system.

18Boghossian’s argument tries to show that all justification is absolute justification, since the Epistemic Relativist’s apparent alternative of justification relative to an epistemic system can be shown, through analysis, to be just another type of absolute justification. What the argument overlooks, however, is the possibility of another, novel conception of justification, one which is suggested at various points during Rorty’s discussion, and which clearly cannot be reduced to absolute justification. According to this conception, beliefs acquire their justification from their usefulness in achieving some purpose or set of purposes, and it is the possibility of these purposes varying from community to community that gives substance to the idea of justification being relative to differing epistemic systems. This explicitly pragmatic conception develops Rorty’s core view that justification is historical and social, rather than an a-historical relation between how beliefs represent the world and the world itself, by explaining differences in standards of justification between societies as a product of the differing needs and interests of those societies.

  • 3 The significance of this qualification will become clearer in Section VI, when we see that in addit (...)

19The Epistemic Relativist can use a pragmatic conception of justification to rebut Boghossian’s argument by qualifying (ES): they can accept the claim that epistemic systems consist of epistemic principles which are more general versions of particular judgements about justification, but must insist that these epistemic principles only state what we are justified in believing relative to certain purposes. This avoids Boghossian’s central charge that accepting an epistemic system implies accepting statements of absolute justification, by claiming instead that to accept an epistemic system is to accept a set of general principles which state what it is useful to believe for certain purposes. So, to introduce a concrete suggestion, the Epistemic Relativist can say that Galileo’s observations justify his belief about the moons of Jupiter relative to his epistemic system, and that accepting that system means accepting a set of relativised epistemic principles, such as that for the purposes of being able to predict and control the environment, a visual appearance of something provides good prima facie justification for believing it to be the case. What justifies adherence to the principle in the long term,3 then, is the success in prediction and control that such adherence provides.

  • 4 Some of Rorty’s later discussions lend support to this interpretation, as for instance when he says (...)

20This kind of position is hinted at a number of times in Rorty’s discussion, where he argues that before Galileo “won the argument,” there was no way to determine that Bellarmine’s concerns about “the impact of science on theology, the future of life on earth, and the like” were “extraneous” to the issue at hand (Rorty 1979: 327), and hence no way to determine that “getting the heavens right is a ‘scientific’ value, and preserving the church, and the general cultural structure of Europe, is an ‘unscientific’ value” (ibid.: 329). What lies behind Galileo and Bellarmine’s different standards of evidence, Rorty seems to be saying, is the different uses these standards subserve. From Bellarmine’s perspective, the advantages of Copernicanism in predicting planetary motions had to be weighed against the disadvantages of a forced abandonment of doctrine in reducing the authority of the church. That the latter outweighed the former for Bellarmine, whereas the opposite was the case with Galileo, is explained by the “shift in cultural climate” (ibid.: 332) taking place in the seventeenth century. This ties in perfectly with Rorty’s Darwinian conception of cultural evolution, which portrays Galileo’s ideas and the new epistemic principles underlying them as successful adaptations to the changing political and socio-economic environment of Europe. It suggests a changing pattern of behaviour, from a prior situation in which beliefs were counted as justified primarily in accordance with their usefulness in maintaining faith, a faith which helped people endure the material conditions of their lives, towards a new situation in which beliefs were counted as justified primarily in accordance with their usefulness in predicting and controlling the environment, an ability which promised to make people “the lords and masters of nature” (Descartes 1985: 142-3), and thus able to change and improve the material conditions of their lives.4

21If the differences between epistemic systems can be explained by their different purposes, rather than the different claims to absolute justification they make, then Boghossian’s attempt to demonstrate incoherence in the Epistemic Relativist’s commitment to both (A) and (B) is undermined. This is because once (ES) is qualified, the acceptance of an epistemic system required by (B) does not necessitate the acceptance of any epistemic absolutes to conflict with (A); in accepting different epistemic systems, Galileo and Bellarmine would simply be accepting different sets of epistemic principles which serve distinct purposes. These principles are ultimately to be understood as claims about the usefulness of beliefs in pursuit of the general aims of the system, and accepting claims about the usefulness of different beliefs for different purposes is perfectly compatible with rejecting all claims to absolute justification.

22Boghossian’s second argument (op. cit.: 89-91) aims to show that Epistemic Relativism’s commitment to both (A) and (C) is incoherent. According to (C), all epistemic systems are “on a par as far as their correctness is concerned” (ibid.: 90), such that for any two alternative epistemic systems, there can be no fact in virtue of which one system is more correct than the other. Given (ES), on Boghossian’s interpretation, this amounts to the claim that there can be no facts that favor the general judgements about absolute justification constitutive of one epistemic system, the epistemic principles of that system, over those of an alternative system. Since alternative epistemic systems would contradict each other about what is absolutely justified, however, Boghossian thinks that the rejection of all claims to absolute justification required by (A) will inevitably favor some systems over others, contrary to (C).

23To see this, suppose that C1 and C2 are alternative epistemic systems that contradict each other. So, for instance, according to C1,

For any observational proposition p, if it visually seems to S that p and circumstantial conditions D obtain, then S is prima facie justified in believing p.

24but according to C2,

It is not the case that for any observational proposition p, if it visually seems to S that p and circumstantial conditions D obtain, then S is prima facie justified in believing p.

25Since the Epistemic Relativist’s commitment to (A) requires them to reject all claims to absolute justification, they must reject the epistemic principle endorsed by C1, and thereby side with C2. But this conflicts with the Epistemic Relativist’s commitment to (C), according to which there can be no fact in virtue of which one system is more correct than the other, thus showing once again, according to Boghossian, that Epistemic Relativism is incoherent.

26This argument is also undermined by a pragmatic conception of justification, because if epistemic principles are general principles which state what we are justified in believing in order to achieve certain purposes, rather than what is absolutely justified, then the Epistemic Relativist’s endorsement of (A) does not imply that all epistemic principles are false. Consequently, the Epistemic Relativist is not committed to claiming that the principle endorsed by C1 is false, and can avoid the compromise to (C) which would result from conceding that one epistemic system is more correct than another. The Epistemic Relativist still needs to make sense of the idea that a subject can endorse one epistemic system and reject another, but this can be done without claiming that the principles of the rejected system are false. Instead, the principles of a system may be rejected, i.e. not adhered to, simply because they serve purposes that the subject does not prioritize. Thus if we interpret the observational principle of C1 as saying that observation justifies belief for the purposes of prediction and control, then C2 need not contradict this: someone who accepted C2 could grant that observation justifies belief for the purposes of prediction and control, but nevertheless choose not to act upon epistemic principles determined by these purposes.

  • 5 As an ethnocentrist, Rorty could only claim that Bellarmine and Galileo’s epistemic systems were eq (...)

27Once justification is relativised to purposes, then, the competing claims of alternative epistemic systems can no longer be represented as neat contradictory pairs according to which the rejection of a principle belonging to one system automatically counts in favour of the other. So, for example, if Bellarmine accepts the principle that Biblically based beliefs are justified, and so useful for his purposes, but Galileo rejects this principle, it does not follow that Galileo thereby endorses the claim that it is not the case that Biblically based beliefs are useful for Bellarmine’s purposes. Rather, Galileo’s rejection commits him only to the claim that it is not the case that Biblically based beliefs are useful for Galileo’s purposes. Since alternative epistemic systems serve alternative purposes, then, there is nothing incoherent in the idea of endorsing one system but rejecting another, while nevertheless claiming that they are all equally correct.5

28Boghossian’s third argument (ibid.: 103-5) aims to show that the controversy between Galileo and Bellarmine was not a clash between alternative epistemic systems, but rather a more mundane dispute within the same epistemic system. This gets to the heart of Boghossian’s misgivings with Epistemic Relativism, since the leading thought underpinning his discussion from the outset is that we have a solitary stock of ordinary and fundamental epistemic principles, rooted in observation, induction, and deduction, to which we can imagine no genuine alternatives, and which we must regard as absolutely justified on pain of incoherence.

29It is crucial for Boghossian to rule out any real life examples of alternative epistemic systems, because he thinks the case for Epistemic Relativism depends upon the possibility of encountering an alternative system that renders the epistemic principles of our own system doubtful. Only such an encounter would overcome the “blind entitlement” we have to our own epistemic system (Boghossian 2003; Boghossian 2006: 96-103), thereby obliging us to try to justify that system. Once our blind entitlement is overcome, the Epistemic Relativist can argue that any attempt to justify our own system over the alternative would inevitably be “norm-circular” (Boghossian 2006: 79), since we have to employ our own standards of justification, despite the fact that these are exactly the standards which would be put into doubt by an encounter with the alternative system. Thus Boghossian wants to cut off the Epistemic Relativist’s argument at the source by showing that we do not know of any alternative systems; he does not rule them out in principle, but does imply that all attempts to find or even imagine one have failed (ibid.: 103, 109).

  • 6 Boghossian falsely assumes that it is Bellarmine who is supposed to have refused to look through th (...)

30Unlike the previous two arguments, then, this third argument is specifically addressed to the issue of how best to interpret the actual historical dispute between Bellarmine and Galileo. It claims that the Epistemic Relativist cannot be right to interpret Bellarmine as employing a different epistemic system to Galileo, because outside of astronomy they employ exactly the same epistemic principles; Bellarmine “doesn’t divine what the Bible itself contains but rather reads it using his eyes” (ibid.: 103). Since Bellarmine relies on observation to find out what the Bible says, induction to predict it will say the same today as yesterday, and deduction to work out its implications for Copernicanism, Boghossian thinks it would be uncharitable to interpret Bellarmine’s refusal to accept the evidence of Galileo’s telescope6 as evidence that he adhered to an alternative epistemic system, for such a system would be incoherent. Rather, Bellarmine must be employing our own epistemic system, but believe that the Bible trumps observation because it is the word of God. Thus there is no clash of epistemic systems, only “a dispute, within a common epistemic system, about the origins and nature of the Bible” (ibid.: 105).

31The first problem with this argument concerns Boghossian’s claim that the dispute between Bellarmine and Galileo was the product of an underlying disagreement about whether or not the Bible is “the revealed word of the Creator” (ibid.: 105), since Boghossian makes no attempt to substantiate this claim against the standard view, attested to by Galileo himself, that the dispute primarily concerned Biblical interpretation. According to this view, Galileo was defending an Augustinian approach to Biblical interpretation according to which the Bible always spoke the truth, although its meaning was not always manifest. Thus Galileo thought that God had found it “necessary to attribute motion to the sun and rest to the earth, in order not to confound the shallow understanding of the common people and make them obstinate and perverse about believing in the principal articles of faith” (Galileo 1967: xxiv). In other words, God did not always have the luxury of being candid, and so given that “the holy Bible and the phenomena of nature proceed alike from the divine Word,” a premise which he shared with Bellarmine, Galileo argued that “nothing physical which sense-experience sets before our eyes, or which necessary demonstrations prove to us, ought to be called in question […] upon the testimony of biblical passages which may have some different meaning beneath their words” (Galileo, quoted in Seeger 1966: 271).

32Putting aside the dubious claim that Bellarmine and Galileo’s dispute concerned the “origins and nature of the Bible,” Boghossian might nevertheless be right that the dispute took place within a shared epistemic system. The problem with this, however, is that Galileo and Bellarmine seem to have agreed all the relevant facts. Bellarmine agreed that Copernicanism gave a better mathematical explanation of the observational data than the Ptolemaic system, but still insisted that it was only “a manner of speaking,” which although “enough for a mathematician” should not be interpreted as the literal truth (Bellarmine, quoted in De Santillana 1958: 99). Likewise, Galileo agreed that Ptolemy was supported by “the bare meaning of the words” of certain Biblical passages, but still insisted that the words might have a hidden meaning (Galileo, quoted in Seeger 1966: 271). What they seem to have disagreed about was not the facts, but the epistemic significance of the facts: Galileo thought observation and mathematical argument should determine how we interpret the Bible, and Bellarmine thought the manifest meaning of the Bible should determine how we interpret observations. This supports Rorty’s interpretation of the controversy as a clash of epistemic systems.

  • 7 Assuming the story is not apocryphal, a more charitable interpretation would be that Cremonini saw (...)

33This still leaves Boghossian’s claim that if Bellarmine did have an alternative epistemic system, then it was incoherent, since he used observation to find out about the Bible but refused to use it in astronomy. The problem with this, however, is that it attributes a straw-man position to the Epistemic Relativist, for there was never any suggestion in Rorty’s discussion that Bellarmine rejected observation outright, and even the Cremonini story is more plausibly interpreted as illustrating only Cremonini’s disregard for observational evidence in certain, specific circumstances.7 What Bellarmine was rejecting was not observation per se, but rather Galileo’s interpretation of particular observations that had been made using novel technology. These observations required considerable interpretation to lend support to Copernicanism (see Seeger 1966: 247-59), and the general view at the time was that observation favored Ptolemy; as Galileo himself conceded, it seemed that “in the Copernican doctrine the senses must be denied” (Galileo 1967: 253).

34Boghossian’s mistake is to assume that alternative epistemic systems must reject observation, induction or deduction outright, but the Epistemic Relativist need only claim that observation, induction or deduction could have different epistemic significances within alternative systems. Since Rorty makes a convincing case that this kind of difference is exactly what was driving the Galileo and Bellarmine controversy, and Boghossian fails to undermine this case, then Rorty’s suggestion of a real life example to motivate Epistemic Relativism stands undefeated.

35Boghossian’s argument is the most sophisticated and detailed attempt to date to prove something that many analytic philosophers have long suspected, namely that Rorty’s epistemic relativist position on justification is incoherent. However, the argument fails, and so until some further argument is presented, we may conclude that Rorty’s conception of justification is coherent; for all we have seen so far, it does seem to be, at the very least, a possible position on the nature of justification. The question we must now ask, then, is whether we have good reasons to think it is true, or to adopt a more anodyne and potentially less contentious formulation in the context of Rorty’s pragmatism, whether we have good reasons to adopt it.

36Now on the face of it, Rorty has a substantive case for his position, which is most fully developed in his critique of representationalism in Philosophy and the Mirror of Nature. The heart of this case is that the combination of Sellars’ arguments against the “Myth of the Given” and Quine’s arguments against analyticity, undermine any representationalist model of the relation between language and the world. Thus Sellars showed that representationalism confuses causation with justification when it locates the justification for our beliefs in the causal impact of the environment, while Quine showed that experience bears evidentially on our claims only as a collective whole, making it impossible to isolate the support the world provides for an individual belief. The combined effect of these arguments, according to Rorty, is to show that our beliefs are not justified through a quasi-mechanical transaction between mind and world that might be studied a priori by epistemologists, in order to determine the conditions of successful representation. Rather beliefs are justified by the outcome of large-scale social interactions that cannot in principle be predicted in advance; as such, there was never any prospect of philosophers devising a theory of knowledge capable of objectively adjudicating disputes such as those between Galileo and Bellarmine. Europe could only wait to see which side would win through, and thereafter would come to be retrospectively seen as better justified.

37Rorty’s alternative to represenationalism is an exclusively causal model of the relation between language and the world, such that the world can “cause us to hold beliefs” and can “decide the competition between alternative sentences,” but only “once we have programmed ourselves with a language” (Rorty 1989: 5-6). This “programming” is a “cultural evolution” (Rorty 1999: 38), akin to Darwinian natural selection, such that changes in vocabulary are understood as adaptations to the causal pressures exerted by the world in pursuit of changing social purposes, with language bearing “no more of a representational relation to an intrinsic nature of things than does the anteater’s snout or the bowerbird’s skill at weaving” (Rorty 1998: 48). Thus when Galileo looked through his telescope, there was a “brute physical resistance – the pressure of light waves on Galileo’s eyeball,” but there were “as many facts […] brought into the world as there are languages for describing that causal transaction” (Rorty 1991a: 81). For Rorty, then, causal pressures can be variously described by alternative vocabularies (“programs,” “epistemic systems”), with different vocabularies providing the causal pressures with different significances suited to different social purposes, but with no vocabulary being more or less intrinsically justified than any other, given that justification exists only within a vocabulary.

38Rorty’s main case for a pragmatic conception of justification, then, seems to be based on his case against representationalism and the need to provide an alternative. Thus he argues that beliefs cannot be absolutely justified in virtue of accurately representing the world, both because of Sellars’ and Quine’s arguments, but also because of the verificationalist reasoning, which is almost a mantra in Rorty, that “there is no way to get outside our beliefs and our language” (Rorty 1979: 178) to assess the adequacy of our descriptions to the world. Once we accept the “ubiquity of language” (Rorty 1982: xx), then, we need a conception of justification which is exclusively “intravocabulary,” as Robert Brandom puts it, to replace the representational model of justification as an “extravocabulary” relationlanguage bears to the world (Brandom 2000). This is provided by Rorty’s pragmatic account of justification, which we are now in a position to understand as comprising of two integrated components, one holistic and coherentist, and the other causal and social-evolutionary. According to the first, beliefs are justified by their coherence with other beliefs in accordance with the principles of an epistemic system; as Rorty puts it, they must “weave together with our other beliefs into a justificatory web” (Rorty 1999: 37). This first component ensures justificatory relations are kept thoroughly “intravocabulary.” According to the second, the holistic “justificatory webs” of an epistemic system are culturally selected for their ability to harness causal pressures in socially useful ways. Thus “extravocabulary” relations are also catered for by Rorty’s account, but these are the exclusively causal relations that hold between language-users and the world “once we have programmed ourselves with a language.”

  • 8 Fred Dretske once remarked, in response to a typical example of one of these disclaimers, that, “th (...)

39This is a subtle and original account of justification, albeit one which Rorty himself never presented as such, preferring to claim that there is “nothing to be said about justification in general” (ibid.: 38), in line with his nominalist and historicist conviction that fruitful discourse about topics like justification must always concern particular practices of justification within historically-embedded discourses. Nevertheless, Rorty actually had lots to say about “justification in general,” and once his various statements are pieced together, they clearly do add up an account, despite his disclaimers.8 The account, to repeat, is that beliefs are not justified individually by the world, but holistically by societal agreements, founded on coherence, which harness causal pressures to serve the social purposes selected for by cultural evolution.

40The difficulties for this account begin when we note that if justification is determined by societal agreement, and ultimately social usefulness, then for Rorty’s position to be consistent, it must be justified in the same way. Thus his argument cannot be that Quine and Sellars demonstrated objective truths about justification, namely that representationalism is untenable because it neglects both the holistic nature of justification, and the distinctness of causation and justification. Neither can it be that, as a matter of fact, it is impossible for us to “step outside our skins” (Rorty 1982: xix) by comparing our descriptions to the language-independent world. If Rorty were arguing this way, then he would be arguing, absurdly, that representationalism fails to accurately represent the facts about justification, and hence is an unjustified position. Rather, Rorty can only consistently argue that Quine and Sellars’ arguments, as well as his own “ubiquity of language” thesis and consequent intravocabulary conception of justification, are justified by societal agreement. The problem with this, however, as critics have not been shy to point out, is that Rorty’s position is manifestly controversial, and so cannot plausibly claim to have societal agreement on its side.

41This difficulty was provided with a focus by Hilary Putnam’s charge that Rorty attempts to say “from a God’s-Eye View there is no God’s-Eye View” (Putnam 1990: 25). Putnam’s point is far-reaching: if Rorty is saying that representationalism is objectively false because it presupposes an impossible stance, a “God’s-Eye View” from which language is compared to a language-independent world, then surely this very claim presupposes the same impossible stance. For it seems to be holding that representationalism fails to accurately represent human justificatory practices, which is something that could only be asserted, by Rorty’s own reasoning, from a “God’s-Eye View.” But if, on the other hand, Rorty is consistent in holding that justification is determined exclusively by societal agreement, then his claim could only be justified if there were societal agreement that society, rather than the world, justifies our claims. There is, however, no such agreement; as Simon Blackburn drolly observes, “Michael Faraday […] seemed to have designed instruments only for registering such things as electrostatic charge and never ones designed to detect the approval of his peers” (Blackburn 2005: 160). Blackburn’s point, of course, is that there is widespread societal agreement that it is the world, rather than the opinions of society, which justifies many of our beliefs. So unless Rorty is inconsistently speaking from the “God’s-Eye View” he urges us to abandon, and saying that representationalism is objectively false, it seems he must concede that societal agreement is against him, and hence that according to his own account of justification, his account is not justified. Thus the case in support of Rorty’s position is either inconsistent or undermines itself.

42Rorty made the following response to Putnam’s objection:

[M]y strategy for escaping the self-referential difficulties into which “the Relativist’ keeps getting himself is to move everything over from epistemology and metaphysics to cultural politics, from claims to knowledge and appeals to self-evidence to suggestions about what we should try. (Rorty 1998: 57)

43Rather than saying from a “God’s-Eye View” that there is no vocabulary-independent world to which our vocabularies answer, then, Rorty’s suggestion is that he is simply making a pragmatic, experimental suggestion for reform to our linguistic practices. He is willing to grant the point that there is no societal agreement in favor of his account, and hence that according to that account, it is not presently justified. However, he thinks this obstacle is easily dispensed with by distinguishing between “what is presently found convenient to say and what might be still more convenient to say” (ibid.: 57). Thus although society presently finds it convenient to say that the world, rather than society, determines matters of justification, Rorty, as a “pragmatic social reformer,” is suggesting that it would be even more convenient to adopt his account. Why? One reason is typical Rortian metaphilosophy: he thinks this will achieve the convenience of “avoiding fruitless disagreements on dead-end issues” (ibid.: 57), by which he means it would undermine the representationalist problematic of traditional, epistemology-based philosophy. However Rorty also has bigger targets in his sights, since he thinks that abandoning representationalism will help to promote a liberal, democratic and open society, by removing “a few more excuses for fanaticism and intolerance” (ibid.: 83). Representationalism creates such excuses, Rorty thinks, because a claim to represent the objective truth short-circuits conversation and overrides the need for compromise; that is why he thinks an account of justification which undermines such strategies promises greater linguistic convenience.

44It seems to me that this response was a miscalculation on Rorty’s behalf, for the following reasons. Suppose we take him at his word, interpreting him as a “pragmatic social reformer” who is proposing an amendment to our notion of justification because of the prospect of future social benefits; if society experiments with Rorty’s idea, he thinks, it will be better off, and so societal agreement will come to over to his position, thereby justifying it in the future. This immediately raises the question, however, of the status of the various arguments Rorty canvasses against representationalism, such as those of Quine and Sellars, and his trademark claim that we cannot “step outside our skins” (Rorty 1982: xix). For these are not arguments about social usefulness, and thus if Rorty was persuaded of his position only by the prospect of social benefits, it starts to look, as it does to Susan Haack, that he was a “cynic” who defended these arguments as “a ploy to persuade others less enlightened than himself by playing the game by their rules” (Haack 1993: 193). And cynical or not, a clear consequence of adopting the “social reformer” stance is that these well-known arguments are undermined: if only arguments about social usefulness can be appealed to without getting into “self-referential difficulties,” then since these arguments are not about social usefulness, they cannot support his position.

45The problems for Rorty’s “social reformer” response do not stop here, however, for once Rorty’s standard battery of anti-representationalist arguments is undermined, we no longer have any reason to believe that social usefulness, rather than representational accuracy, is germane to questions of justification. Thus somebody might be prepared, for the sake of argument, to take Rorty’s word for it that pragmatism about justification offers social benefits. Nevertheless, this person might also maintain, perfectly consistently, that although we would be better off if we became pragmatists about justification, we would be in error, since justification is determined by representational accuracy. In short, it is only on Rorty’s account, not the representationalist’s, that social usefulness determines justification, and so it is simple question-begging to appeal to social usefulness to justify his account over that of the representationalist. Rorty needs his stock of anti-representationalist arguments to avoid such question-begging, but his “social reformer” response undermines them.

  • 9 See Festenstein (1997: 109-44), for a detailed and informative discussion of the theoretical and so (...)

46Given that Rorty characterized his “social reformer” response as the same as “Dewey’s pragmatic justification of pragmatism” (Rorty 1998: 58), he was presumably well aware of, but thoroughly unperturbed by, this logical flaw to his argument. Perhaps he reasoned that if his pragmatism was adopted and found socially useful, then nobody would care about representational facts anymore, even if they did exist; they could be forgotten, as perhaps mortality might be forgotten in a society that found itself benefitting from a false belief in immortality. Perhaps Rorty was not trying to argue against representationalism, then, but simply expressing his preference for pragmatism, and inviting others to join him by advertising the benefits. This would be a very modest strategy, but for it to have any persuasive force at all, Rorty needed to make it plausible that his account would indeed be socially useful. This, as far as I can see, he singularly failed to do. Rather, he simply stated the benefits he envisaged, which are, basically, an end to epistemological debates and increased toleration. But as regards epistemological debates, Rorty’s account has caused massive controversy and spawned countless indignant responses, such as Boghossian’s book. On the face of it, then, it has fuelled the ancient debate between relativists and realists, rather than dampen it down in any way; it seems like just another contribution from the relativist side. And as regards toleration, many philosophers, such as Boghossian and Blackburn, think Rorty’s account promotes a socially deleterious, rather than beneficial, toleration of superstition and pseudo-science. Moreover it is not even clear that it does promote toleration once supplemented with ethnocentrism.9

47Rorty’s “social reformer” tactic, then, seems considerably more trouble than it was worth as a response to Putnam. But I shall end by suggesting that Rorty did not need it, and indeed, to argue for his pragmatism directly on the basis of envisaged social benefits was to sell his position short. For as we saw in Section VI, the link it makes to social usefulness is subtle: beliefs are justified holistically by societal agreements, founded on coherence, which harness causal pressures to serve the social purposes selected for by cultural evolution. The idea, then, is that beliefs which cohere with each other gain societal agreement, and are thereby justified. But although large-scale, holistic webs of belief harness causal pressures in socially useful ways, such that it is correct to say that justification in Rorty’s pragmatism is ultimately rooted in usefulness for a purpose, it is nevertheless not the case that usefulness directly justifies individual beliefs. Rather coherence does. As such, Rorty can base his justification for pragmatism on coherence rather than usefulness. This makes it legitimate for him to draw on Quine and Sellars’ arguments, and also allows him to concede that his position is currently controversial and not widely regarded as justified, without thereby speaking from a “God’s-Eye View.”

48To see this, consider the Galileo and Bellarmine case again. To echo Blackburn’s comment quoted above, Galileo used his telescope to detect the phases of Venus, not the approval of his peers, and the arguments he used to justify his position referred to astronomical data, rather than social benefits. However none of this precludes the possibility that an adequate philosophical account of what justified Galileo’s position must make reference to societal agreement and social benefits; these are not the kind of considerations it would be appropriate for Galileo to adduce, or that he needed to know anything about, but they might nevertheless provide the best explanation of what it means to say that this views were justified. The considerations it would be appropriate for Galileo to adduce in order to justify his views, on the other hand, are considerations of coherence: that the observations he had made of the phases of Venus were not what would be expected on the Ptolemaic model, but exactly what would be expected on the Copernican model, and thus that the new data he had collected cohered better with Copernicanism.

49To argue for his pragmatic account of justification, then, it seems Rorty has no more need to appeal to social usefulness than Galileo did. Rather, all he needs argue is that his account of justification coheres better with certain a priori considerations, such as the Sellars and Quine arguments, than representationalism does. If Rorty is right about this, and the arguments hold up, then this coherence may produce societal agreement, and ultimately social usefulness. If he is wrong, and the arguments fail, or simply fail to cohere with Rorty’s pragmatist conclusions, then the failure of Rorty’s account to cohere with the rest of our beliefs may be indicative of its failure to accurately represent the facts about justification. But either way, basing his argument on coherence does not involve Rorty in any “self-referential difficulties”: he does not need to presuppose his pragmatic account of justification to argue that it coheres better with the rest of our commitments than alternative accounts, and he does not need to appeal to a “God’s-Eye View” to argue that representationalism presupposes a stance from which to compare our representations to the world, and that the possibility of such a stance fails to cohere with other things we know about the world.

50At the beginning of this paper, we recalled Rorty saying that relativism could be refuted by “some variant of the self-referential arguments Socrates used against Protagoras,” but denying that his own position amounted to relativism. While the latter claim should be taken with a pinch of salt, as we have seen, we have nevertheless also found good reasons, in considering both Boghossian’s and Putnam’s self-referential arguments, to conclude that the sophisticated kind of relativism about justification which Rorty defended does not undermine itself, and is both coherent and argumentatively well-supported. It also has a distinct advantage over absolutist conceptions of justification, namely that justification, unlike truth, is naturally thought of as a social phenomenon rather than as an absolute notion; as Putnam puts it, “truth is supposed to be a property of a statement that cannot be lost, whereas justification can be lost” (Putnam 1981: 55). This is why we are liable to be unperturbed by the claim that Bellarmine’s view was once justified, even though it is not now, whereas only the most radical of postmodernists would countenance the parallel claim that Bellarmine’s view was also once true.

51However, even though Rorty’s view does have this important intuition on its side, it has others against it. For an example of this, remember that it was not only justification, but also truth, that Rorty wanted to keep exclusively intravocabulary, since he held that “the only point in contrasting the true with the merely justified is to contrast a possible future with the actual present” (Rorty 1999: 39). Thus we might, in Rorty’s “cautionary” sense of true, say that in 1616, Bellarmine’s view was justified but possibly not true; we later discovered that it was actually not true because Galileo “won the argument.” However if Galileo’s view was only justified after he “won the argument,” as Rorty insists, then given that the winning was contingent, it follows that Galileo’s view might never have been justified. But since truth cannot be asserted apart from justification, on Rorty’s view, it seems to follow that a future in which Galileo’s view was never justified is one in which it was never true; its being true would have forever remained an unactualized “cautionary” possibility. Now granted, in the actual world Galileo did win the argument, and so from our ethnocentric perspective, we can assert absolutely that the Earth has been orbiting the Sun throughout human history. But nevertheless, since Rorty holds that “the only criterion we have for applying the word “true” is justification, and justification is always relative to an audience” (Rorty 1998: 4), it is hard to see how he can avoid the consequence that since Galileo’s view might never have been justified, it also might never have been true, which, when we are talking about the Earth orbiting the Sun, is evidently a highly counterintuitive counterfactual. Barring some nimble manoeuvre, then, or a radical scepticism about counterfactuals (and Rorty did express such inclinations; see Putnam 2000: 85), it seems clear that Rorty’s account was bound to conflict with common sense. That, however, was how he seemed to like it; as Richard Bernstein once said, Rorty favored “ever new forms of dissensus, not epistemological consensus” (Bernstein 1991: 62). Rorty’s pragmatism was perhaps never likely to achieve its stated aim of dampening-down interest in epistemological debates, then, but properly understood, the novel and provocative account of justification it provides retains plenty of potential to enrich them.

Top of page


Bernstein R., (1991), The New Constellation, Cambridge, Polity Press.

Blackburn S., (2005), Truth, London, Allen Lane.

Boghossian P., (2003), “Blind Reasoning,” Proceedings of the Aristotelian Society, supplementary volume 77, 225-48.

Boghossian P., (2006), Fear of Knowledge: Against Relativism and Constructivism, Oxford, Oxford University Press.

Brandom R., (2000), “Vocabularies of Pragmatism: Synthezing Naturalism and Historicism,” in Brandom (ed.), Rorty and His Critics, Oxford, Blackwell.

De Santillana G., (1958), The Crime of Galileo, London, Heinemann.

Descartes R., (1985), The Philosophical Writings of Descartes, vol. 1, trans. J. Cottingham, R. Stoothoff & D. Murdoch, Cambridge, Cambridge University Press.

Dretske F., (1982), “Review of Philosophy and the Mirror of Nature,” International Studies in Philosophy 14, 96-8.

Festenstein M., (1997), Pragmatism and Political Theory: From Dewey to Rorty, Oxford, Polity Press.

Galileo Galilei, (1632/1967), Dialogue Concerning the Two Chief World Systems – Ptolemaic and Copernican, trans. S. Drake, Berkeley, University of California Press.

Haack S., (1993), Evidence and Inquiry: Towards Reconstruction in Epistemology, Oxford, Blackwell.

Miller C. B., (2002), “Rorty and Moral Relativism,” European Journal of Philosophy 10, 354-74.

Putnam H., (1981), Reason, Truth and History, Cambridge, Cambridge University Press.

Putnam H., (1990), Realism with a Human Face, Cambridge, MA, Harvard University Press.

Putnam H., (2000), “Richard Rorty on Reality and Justification,” in R. Brandom (ed.), Rorty and His Critics, Oxford, Blackwell.

Rorty R., (1979), Philosophy and the Mirror of Nature, Princeton, Princeton University Press.

Rorty R., (1982), Consequences of Pragmatism, Minneapolis, University of Minnesota Press.

Rorty R., (1989), Contingency, Irony, and Solidarity, Cambridge, Cambridge University Press.

Rorty R., (1991), Objectivity, Relativism and Truth: Philosophical Papers, vol. 1, Cambridge, Cambridge University Press.

Rorty R., (1998), Truth and Progress: Philosophical Papers, vol. 3, Cambridge, Cambridge University Press.

Rorty R., (1999), Philosophy and Social Hope, London, Penguin.

Rorty R., (2007), Philosophy as Cultural Politics: Philosophical Papers, vol. 4, Cambridge, Cambridge University Press.

Rorty R., (2011), An Ethics for Today, Columbia, Columbia University Press.

Seeger R., (1966), Galileo Galilei, His Life and His Works, Oxford, Pergamon Press.

Top of page


1 Rorty always refused the label except, curiously, in one of the last lectures he delivered before his death (Rorty 2011), where he identifies himself as a relativist without explanation, but also without amending his position in any way; perhaps he had decided it was better to embrace the label and make it his own, rather than issue yet more jaded and ineffectual denials.

2 I will capitalise “Epistemic Relativism” when discussing this formulation. Rorty would clearly accept (A) and (B), although he might dispute (C), since he sometimes seems as sceptical as Boghossian about there being genuine alternatives to the epistemic standards of the contemporary West (see Rorty 1991: 29-31; and Rorty 1999: 276-7). Nevertheless, he is committed to the view that there were once genuinely alternative epistemic systems, namely those of Galileo and Bellarmine, which is all the epistemic pluralism Boghossian needs.

3 The significance of this qualification will become clearer in Section VI, when we see that in addition to the social-evolutionary component of Rorty’s account appealed to here, there is also a holistic, coherentist one.

4 Some of Rorty’s later discussions lend support to this interpretation, as for instance when he says that, “science and religion are both respectable paths for acquiring respectable beliefs, albeit beliefs which are good for quite different purposes” (Rorty 1999: 36; see also Rorty 1982: 192-5).

5 As an ethnocentrist, Rorty could only claim that Bellarmine and Galileo’s epistemic systems were equally correct before Galileo “won the argument.”

6 Boghossian falsely assumes that it is Bellarmine who is supposed to have refused to look through the telescope (op. cit.: 60, 103-5); this discrepancy has no bearing on the argument.

7 Assuming the story is not apocryphal, a more charitable interpretation would be that Cremonini saw no need to look through the telescope because he knew beforehand that whatever he saw would have to be interpreted in accordance with the Ptolemaic system; he did not doubt Galileo’s testimony, only his methodology of prioritising mathematical economy over Biblical authority.

8 Fred Dretske once remarked, in response to a typical example of one of these disclaimers, that, “this sounds to my ear like a theory of knowledge. Not a very well articulated theory to be sure, but a theory nonetheless” (Dretske 1982: 97). The more you look at Rorty, though, the more articulation you can uncover.

9 See Festenstein (1997: 109-44), for a detailed and informative discussion of the theoretical and social implications of Rorty’s conception of ethnocentricism.

Top of page


Electronic reference

James Tartaglia, « Does Rorty’s Pragmatism Undermine Itself? », European Journal of Pragmatism and American Philosophy [Online], IV - 1 | 2012, Online since 23 July 2012, connection on 23 September 2017. URL : ; DOI : 10.4000/ejpap.801

Top of page

About the author

James Tartaglia

Keele University

Top of page


Licence Creative Commons
Author retains copyright and grants the European Journal of Pragmatism and American Philosophy right of first publication with the work simultaneously licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Top of page
  • Logo Associazone culturale Pragma
  • Logo DOAJ – Directory of Open Access Journal
  • Logo École des Hautes Études en Sciences Sociales – EHESS