Brandom on Facts, Claims, and Deflationism about Truth

Transcription

1 Brandom on Facts, Claims, and Deflationism about Truth Bernd Prien Philosophisches Seminar WWU Münster Abstract: In this paper, I want to do three things: First, I will elucidate Brandom s understanding of the concepts claim and fact and their relation to one another. This will in part be done by looking into Brandom s distinction between explication and explanation, in part by showing that Brandom distinguishes between two phases of the explication of claiming. The crucial difference between these phases is that the first one invokes only our use of expressions, while the second also invokes the physical facts embedding our practices. Second, I will outline a possible objection against explicating claimings partly in terms of the concept fact and reject it as mistaken. Finally, I will argue that because of his understanding of the relation between the concepts claim and fact, Brandom cannot, in the end, be a deflationist about truth. In this paper, I want to discuss Brandom s understanding of the concepts claim and fact and their relation to one another. I will try to shed light on Brandom s thesis that one has to explicate facts in terms of claims 1 and claims partly in terms of facts 2, as well as on his thesis that this twofold explicatory relation between the concepts claim and fact gives rise to a symmetric explanatory relation, i.e., a reciprocal sense-dependence. In order to do so, I will in section 1 explicate the concepts explication and explanation in general, before I get to the explicatory and explanatory relations between the concepts fact and claim in section 2. In section 3, I will outline a possible objection against Brandom s thesis that facts have to be invoked in order to explicate claimings. Proponents of this objection hold that pragmatists can only invoke our practices of using expressions in the game of giving and asking for reasons to explain what claims are and, moreover, that what a fact is has to be understood in terms of possible claims, and not the other way around. Even though I think that this is a natural objection to make, I will reject it as mistaken. In section 4, I will argue that because of the relations he sees between the notions claiming and fact Brandom cannot, at the end of the day, be a deflationist about truth. If facts are explicated as that which is claimed by true claims, one has to be a deflationist about truth. If one, however, also explicates what claims are partly in terms of facts, one has to have a robust notion of truth. 1 Explication and Explanation Probably the most interesting and also the most condensed formulation of Brandom s 1 Cf. MIE 622: Facts are just true claims. 2 Cf. Brandom 2000a, 163: [W]e can also only understand the notion of a vocabulary [i.e., claimings] as part of a story that includes facts.

2 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 2 understanding of the concepts fact and claim appeared in an exchange between him and Habermas over Making It Explicit (MIE). In his reply to Habermas, Facts, Norms, and Normative Facts, Brandom endorses the pragmatist explanatory commitment, which insists that the category of facts can in principle only be made intelligible in the context of an account of the practice of claiming. (2000b, 369) He then goes on to explain: (*) There is an asymmetry here, but it is intended to be understood as an explicative asymmetry, rather than an explanatory asymmetry. The concepts of claiming and of facts mutually involve one another, in the sense that one cannot deploy the concept of claiming without already at least implicitly having brought the concept of facts to bear as well. But the involvement may, at least to begin with, be only implicit. That is, I claim that one can make explicit various crucial aspects of claiming without yet having talked about facts even though one of the things one will discover by doing that is precisely that fact-talk is already implicit in claiming-talk. (2000b, 369) The following two points contained in this passage will be of particular relevance for this paper: First, Brandom here explicitly distinguishes between explanation and explication, though he says nothing about the nature of this distinction. Secondly, he implicitly distinguishes between two phases of the explication of claiming in the last sentence of this quote. In the first phase, one make[s] explicit various crucial aspects of claiming without yet having talked about facts, while in second phase, one does talk about facts in order to make explicit the fact-talk [that] is already implicit in claiming-talk. In this section, I will discuss what Brandom means by explanation and by explication in general before I discuss, in section 2, the explication of claiming in particular, as well as Brandom s distinction between two phases within this explication. In this context, I will address the questions of why we need, on the one hand, to invoke the concept claim in order to understand what facts are and, on the other hand, the concept fact in order to understand what claimings are. 1.1 Explication When considering explications in general, it should, first of all, be noted that different concepts can be the target of an explication: One can explicate, e.g., facts, claimings, and inferences. To explicate something (or to make it explicit) means that we put something we otherwise only do in explicit words. We describe, specify, or codify what we do. Probably the best known example of this is the explication of inferences in terms of conditionals. We can implicitly treat an inference from p to q as good by drawing it ourselves and expecting others to draw it, and we can make this doing explicit by endorsing the conditional claim if p then q (cf. BSD 2.4). As becomes clear in the course of BSD, there are different ways in which one can specify the practice of drawing an inference. Inferences can be made explicit both in normative and modal vocabulary. Thus, the inference from p to q can be codified by saying if p

3 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 3 is the case, q must be the case as well (modal vocabulary) and by saying if you endorse p, you should endorse q as well (normative vocabulary). Even though the nature of the relation between these to ways of explicitating inferences is of great importance for Brandom, I will not discuss it here. I only want to note that according to Brandom, the normative and modal vocabularies make explicit different features (aspects) of our practice of inferring (cf. BSD 181). There is a third way explicating inferences, the explication in terms of conditionals. In this case, however, we merely make explicit an aspect common to both the modal and the normative case (cf. BSD 46, fn. 6). Thus, explicitation in terms of conditionals does not really constitute a third, independent way of making inferences explicit. Brandom s discussion of the different ways in which the act of inferring is specified or codifies raises the following problem: It might be doubted that an explicitation in modal terms has enough in common with an explicitation in normative terms to warrant to subsumtion of both cases under a common concept. As Brandom himself acknowledges, the expression If this is copper, it must conduce electricity clearly does not describe the corresponding act of inferring. However, he still thinks that modal vocabulary specifies acts of inferring in an indirect way because the above expression conveys in a certain sense that the corresponding inference is good. However, I am going to ignore these difficulties here because, first of all, the central concern of this paper will be with the explication of claimings, rather than inferences. Secondly, the kind of explication at issue in that case is exclusively that of describing the practice of claiming in normative terms. As I will explain in the next section, Brandom proposes to explicate claimings in terms of the game of giving and asking for reasons and in terms of the practice of rational rectification. These are explications in normative vocabulary because describing these practices consists in formulating the norms we otherwise simply follow in the practice of claiming. 1.2 Explanation In the passage (*) quoted above, Brandom implicitly distinguishes explication from explanation by speaking of an explicative asymmetry and an explanatory symmetry between the concepts fact and claim. In MIE, Brandom explicitly draws this distinction: The key point is that explicitation is not explanation. Proprieties of inference are not explained in terms of something more primitive by being expressed in the explicit form of claims by the use of conditionals; (MIE 112) 3 Having discussed explications above, I will now turn to the notion of explanation. To my knowledge, Brandom nowhere says in general terms what an explanation is. According to the passage just quoted, however, an explanation explains in terms of something more primitive. This suggests that in an ex- 3 Though this remark appears in the context of a discussion of Frege, it seems to reflect Brandom s own position.

4 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 4 planation, one more primitive concept is taken to be clear or understood and another concept is made clear or intelligible in terms of it. This interpretation is also confirmed by two cases in which Brandom distinguishes between two complementary orders of explanation, namely inferentialism vs. representationalism and objectivism vs. subjectivism. In my discussion of these cases, I will only be interested in the meaning of explanation, and not so much in the defensibility of Brandom s choice of a particular order of explanation. Let us start with Brandom s distinction between representationalism and inferentialism which constitute two complementary orders of semantic explanation. 4 Representationalism is the thesis that the notion of representation should be assumed as given or understood and that the notion of a valid inference should be explained as something necessarily truth preserving. Inferentialism, the complementary order of explanation, takes the notion of inference as given and explains representation in terms of it. What emerges from this case is that an explanation takes one concept as understood, in this case either representation or inference, and aims at elucidating another concept in terms of it. This interpretation is confirmed by the second case I want to consider, Brandom s distinction between an objectivist and a subjectivist order of semantic explanation in BSD 6.4: The former begins with the way the world is construed here as what really follows from what and what is really incompatible with what (BSD 194f) which is then taken to define the goal of inquiry (BSD 195), so as to explain the subjective notion of incompatibility. The subjectivist order of semantic explanation, on the other hand, seeks to make the notion of objective modal relations intelligible in terms of this process [of rectifying and amplifying commitments] (BSD 195). In this case, the subjective notion of incompatibility is taken to be understood. In the case of subjectivism and objectivism, Brandom stresses that he endorses neither order of semantic explanation. He defends the thesis (which he calls objective pragmatism ) that incompatibility in each sense is in principle fully intelligible only in terms of its relation to the other. (BSD 196) Consequently, the explanatory relation between these two concepts is symmetric, they mutually involve one another, i.e., they are reciprocally sense-dependent: One can fully understand either concept only if one also understands the other (cf. TMD 194f). The upshot for the notion of explanation in general is this: That one concept is explained in terms of another means that the latter is antecedently intelligible and that our understanding of the former concept is derived from it. Understood in this way, the concept explanation is a rather broad umbrella-term in that it allows many different kinds of relations between concepts, for example semantic ones, but also relations of explication. 4 cf. MIE 2.III.5: Inferentialism and Representationalism, p. 93f.

5 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 5 As I will now explain, certain explicatory relations can thus turn out to be cases of explanatory relations between concepts as well. 1.3 The Relation between Explication and Explanation Though explication and explanation need to be distinguished, there seems to be following connection between them: In the case of inferentialism, one aims to explain representation in terms of inference. This, however, is done by showing that representational locutions should be understood as making explicit [my emphasis] certain features of communicating by claiming the interpersonal giving and asking for reasons. (MIE 496) So, the concept inference, insofar as it figures prominently in the description of the game of giving and asking for reasons can be used to explain the concept representation because the latter makes explicit certain aspects of this game. Given that we understand the concept inference in a broad sense, we also understand the game of giving and asking for reasons and thus the concept representation because it makes explicit aspects of the latter. In the case of the concepts claim and fact, there seems to be a similar connection between explication and explanation: As I will discuss in more detail below, Brandom defends the view that the concept fact has to be explicated in terms of claims and the concept claim partly in terms of facts. Because of these two asymmetric explicatory relation he concludes that the concepts fact and claim stand in a symmetric explanatory relation: they are reciprocally sense-dependent, i.e., one can grasp either concept only if one also grasps the other. Thus again, we have an explanatory relation that is due to explicatory relations. Though I would argue that explanatory relations are sometimes due to explicatory ones, I do not think that this is always the case. If one concept is defined in terms of another, there is an explanatory relation between them, but no explicatory relation. Also, not every explication yields an explanatory relation between concepts: Though inferences can be made explicit in terms conditionals, conditionals do not explain what inferences are. 2 Facts and Claims So far, I have discussed what Brandom means by explication and explanation. The main topic of this paper is Brandom s explication of the act of making a claim. This explication will turn out to be an explication in normative terms, i.e., an explicit formulation of the norms constitutive of something being a claim. Within this broad topic, 5 my main in- 5 A comprehensive discussion of the explication of claiming would have to cover more or less everything said in MIE as well as the practice of rational rectification. The latter practice plays a central role in ch. 6 of BSD and in Sketch of a Program for a Critical Reading of Hegel, but it is also briefly described in MIE.

6 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 6 terest will lie in Brandom s distinction between two phases of this explication. This distinction is at least implicit in Brandom s remark in passage (*) that one can make explicit various crucial aspects of claiming without yet having talked about facts (2000b, 369) which part of the explication I will here refer to as its first phase. The second phase becomes necessary because one will discover that fact-talk is already implicit in claiming-talk. (2000b, 369) What I will refer to as the second phase makes this implicit facttalk explicit. I will now describe more concretely what the two phases of the explication consist in, and why exactly the second phase is necessary (2.1 and 2.2). Following that, I will describe what I take to be the main difference between the two phases (2.3). To conclude this section, I will point out that MIE is mostly concerned with the fist phase of the explication of claiming, while the second is only present in a very rudimentary form (2.4). 2.1 Phase I of the Explication of Claiming The first phase of Brandom s explication of claiming consists in his description of the game of giving and asking for reasons. The framework within which this game is played consists of the following three different kinds of material inferential relations between claims: commitment-preserving ones, entitlement-preserving ones, and relations of incompatibility. In the case of empirical assertions we moreover have to take into account relations of non-inferential entitlement to claims, as well as practical commitments following from doxastic commitments. The basic move in the game of giving and asking for reasons consists in endorsing a claim. This has two kinds of consequences, the first of which in turn comes in three flavors: In accordance with the three kinds of inferential relations just mentioned, the speaker is committed to endorse further claims, she is entitled to endorse further claims (provided she was entitled to the original one), and she looses entitlement to claims she otherwise had. The second kind of consequence of endorsing a claim is that the speaker undertakes a conditional task responsibility to provide justification for her claim if appropriately challenged. This responsibility is to be discharged in accordance with the entitlement preserving inferential relations and non-inferential entitlements to claims. A particularly important feature of the game of giving and asking for reasons consists in the fact that different speakers have different views about the inferential relations between claims, i.e., about what commits a speaker to what, what entitles a speaker to what, and what rules out what. This forces scorekeepers to keep two books on the commitments and entitlements of a speaker, one based on the infrerential relations the speaker acknowledges and one based on the inferential relations the scorekeeper herself acknowledges. This feature leads to a social dimension, i.e., a specifically interpersonal aspect of the game of giving and asking for reasons. Our practice of using assertions is such that a scorekeeper is entitled to draw a conclusion from the claim of a speaker according to the

7 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 7 inferential relations she herself acknowledges, even if the speaker does not acknowledge them. Moreover, the scorekeeper is entitled to ascribe the conclusion thus arrived at to the speaker and to criticize him for being committed to this conclusion, even though the speaker would not have drawn that conclusion. 6 I have now given a brief sketch of the norms that constitute the game of giving and asking for reasons. These norms could be called fundamental discursive norms in contradistinction to inferential norms determining what follows from what. 7 Though Brandom does not address this question, it seems that the fundamental discursive norms are known a priori, while the inferential norms are only known empirically. This concludes the first phase of the explication of claiming. Though this is not immediately relevant for Brandom s understanding of the concepts fact and claim, it should be noted that social aspect of them game of giving and asking for reasons, in terms of which claimings are made explicit, can in turn be made explicit in terms of the notion of representation. [R]epresentational locutions should be understood as making explicit certain features of communicating by claiming the interpersonal giving and asking for reasons. (MIE 496) 8 The main concern of this paper is the explication of claiming, the first phase of which has just been sketched in briefest outline. In order to understand the relation between the concepts fact and claim, however, it is important that Brandom also explicates the concept fact as what is claimed by a possible true claim. As Brandom puts it in Holism and Idealism in Hegel s Phenomenology : That facts can be the contents of assertions, judgments, beliefs that they are claimable, thinkable, believable is an essential feature of them. One does not know what a fact is unless one understands that they can be stated. (TMD 197). Sometimes, Brandom uses the slogan Facts are just true claims (MIE 622), which is prone to be misunderstood because it seems to imply that there would be no facts if there were no speakers who could endorse claims about them. Brandom has repeatedly explained that this is not what he has in mind. There is no referencedependence between fact and claim, but only a sense-dependence: to understand what a fact is, one has to understand that it can be claimed. 9 6 If we consider the special case of substitution inferences between singular terms, we get the practice of de re ascriptions Brandom describes in MIE, ch. 8.II. 7 This distinction is drawn in Prien (2010). Similarly, Loeffler distinguishes between sui generis norms governing our interpretive practice and objective semantic norms governing inferences (2005, 34). 8 It is not quite clear in what sense the assumption of a representational relation between singular terms and objects specifies or codifies the game of giving and asking for reasons. This clearly is not a case of an explication in normative terms. It rather seems that the assumption of a representational relation between singular terms and objects specifies or codifies the game of giving and asking for reasons insofar as its social aspect can be made intelligible by making this assumption. Therefore, this kind of explication seems to be similar, though not identical to the explication of inferences in modal vocabulary. 9 At this stage, Brandom has only established a dependence of the sense of fact on the sense of claim.

8 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 8 That Brandom views the above consideration as an explication of facts in terms of the concept claim (even though officially only doings can be made explicit) can be seen from the following quote: Immediately before the passage (*), Brandom formulates the thesis that facts can in principle only be made intelligible in the context of an account of the practice of claiming (2000b, 369) and then goes on to call this an explicative asymmetry, rather than an explanatory asymmetry. To conclude this subsection, I will briefly restate its two main results: First, the game of giving and asking for reasons explicates the act of claiming in normative terms. (As we will see, this is only a partial normative explication, or only its first phase.) Second, Brandom uses the concept claim to explicate what facts are. We thus have two explicatory steps, first claim in terms of the game of giving and asking for reasons, and second, fact in terms of claimings. 2.2 Phase II of the Explication of Claiming In the passage (*) quoted above, Brandom writes that one can make explicit various crucial aspects of claiming without yet having talked about facts, though one will discover that such an explication must remain incomplete because fact-talk is already implicit in claiming-talk. (2000b, 369) Having discussed the first, non-fact-involving phase of the explication above, I will now turn to the second, fact-involving phase. The reason why Brandom thinks that fact-talk is already implicit in claiming-talk (2000b, 369) is indicated in his paper Vocabularies of Pragmatism. There, Brandom writes that it is a consequence of the Quinean point with which we began that we can also only understand the notion of a vocabulary [i.e., claimings] as part of a story that includes facts. (2000a, 163) The Quinean point mentioned here consists in Quine s criticism of attempts by Carnap and other logical positivists to divide the explanatory labor addressed to linguistic practices between meanings and beliefs. (2000a, 156). As Brandom understands Quine s critique of the logical positivists, the latter have claimed that for each assertion, there is a privileged set of inferences constitutive of its content. The other, non-privileged inferences are then to be justified empirically. Quine argues that these attempts are misguided because there are no analytical inferences that are immune from revision in the face of recalcitrant experience. Instead, all inferences are subject to a standard of correctness set by the way the world is. After this relatively abstract argument that inferences are subject to a standard set by the way the world is I will now describe the actual practice by means of which we accept the way the world is as a standard of correctness of our inferences. This is the practice of rational rectification which Brandom briefly mentions in MIE and which assumes a more prominent role in some of his papers on Hegel and in BSD. This practice is governed by the following norm: Suppose, I acknowledge the following circumstances and conse- Later, the sense-dependence will turn out to be reciprocal.

9 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 9 quences of application for my concept acid : I am entitled to apply acid to a liquid if it tastes sour and if I apply acid to a liquid, then I am also committed to the claim that it will turn litmus paper red. Now, suppose the world actually contains a sour tasting liquid that does not turn litmus paper red. Then, if I taste this liquid, I will acquire an entitlement to apply acid to it and, after actually applying acid, a commitment to the claim that the liquid turns litmus paper red. On the other hand, if I dip litmus paper into the liquid, I perceptually acquire an entitlement to the claim that the liquid does not turn litmus paper red. Thus, the inferential relations I acknowledge for acid lead to incompatible commitments. When something like this happens, the world has, as it were, shown me that my concept acid, characterized by the above circumstances and consequences of application, is inadequate because there are, as a matter of fact, sour tasting liquids that do not turn litmus paper red. 10 If I find myself with incompatible commitments because of the circumstances and consequences of application of one of my concepts, the norm of rational rectification applies and puts me under an obligation to modify my inferential commitments so as to avoid this incompatibility (cf. MIE 332; BSD 184f; 2005, 146). This is what I want to call the norm of rational rectification. It is another norm constitutive for expressions to count as assertions. Therefore, formulating this norm is part of making explicit what claims are. As I understand Brandom s Quinean point, something is web of beliefs with inferential relations among them only if the inferential relations are to be modified if the web turns out not to correspond to the way the world is. This means that a web of beliefs related by inferences determined by communal agreement, i.e., inferences for which there is no ratification-transcendent standard of correctness, would not really be an inferentially related web of beliefs at all for which reason I have put these terms in scarequotes. If our practice was such that in the acid case described above there would be no obligation whatsoever to modify our inferential commitments, we would not really be talking about the world. In such a case, our expressions would not really be assertions expressing facts in the objective world. I suppose that something like this consideration stands behind Brandom s thesis that [i]t is a fundamental feature of our understanding of our concepts that they incorporate objective commitments. (MIE 53) This second phase of the explication of claimings makes explicit how fact-talk is already implicit in claiming-talk (2000b, 369), as Brandom puts it in the passage (*).Since claimings are explicated in terms of inferences, and inferences, in virtue of the practice of rational rectification, are subject to standards set by the way the world is, one has to in- 10 Concepts can only turn out to be inadequate with respect to the way the world is if their content is determined both by their circumstances and consequences of application, i.e., if one adopts a two-sided inferentialism. If one adopts a one-sided semantics, the content of a concept cannot lead to incompatible commitments in the way described above. In my view, this is the ultimate reason why Brandom insists on a two-sided semantics.

10 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 10 voke the concept fact in order to make fully explicit what claims are. 2.3 The Difference between the Two Phases of Explication As the passage (*) I began with with shows, Brandom distinguishes between two phases of his explication of claiming, such that the first phase does not invoke facts, while the second does. I have now proposed a more concrete interpretation of what these two phases amount to: The first phase of the explication consists in a description of the game of giving and asking for reasons, while second phase consists in a description of the practice of rational rectification. Both phases of the explication of claiming consists in describing our assertional practices by formulating the norms constitutive of it. On the basis of this proposal, we can now state more specifically which resources each phase invokes. In the first phase, the practice consists merely of acts of claiming, challenging, attributing deontic statuses, etc., or rather norms governing these acts. In the second phase, however, the practice also involves an environing physical world and a determinate way the world is. Thus, the first phase of the description of our practices only invokes our actions or the way we use expressions, while the second also invokes facts going beyond our use of expressions, i.e., objective physical facts. For this reason, Brandom says that the practice of rational rectification is solid, corporeal, thick, or lumpy, and that it does not fall entirely on the word -side of the word/world gulf (cf. MIE 332; MIE 632; BSD ch. 6.1). This means that in codifying these latter practices, it is not sufficient merely to describe speech-acts and the norms they are subject to. Rather, one also has to invoke environing physical facts in which these performances of speech-acts are embedded. Therefore, a vocabulary VP-sufficient to describe the practice of rational rectification differs significantly from the minimal vocabulary VPsufficient to describe the game of giving and asking for reasons, in that the former also has to talk about physical facts. 11 In any case, the necessity of invoking environing facts to explicate claimings is a very important point because it means that according to Brandom, a pragmatist account of assertions, as it is maybe most naturally understood, fails: One cannot explicate claimings exclusively by describing our acts of endorsing them; instead, one also has to invoke environing facts to do so. 2.4 MIE and the Two Phases of the Explication of Claiming When reading MIE, it is easy to get the impression that, according to Brandom, the game of giving and asking for reasons constitutes a complete explication of claiming and that a second phase is not required. This is due to the fact that practically everything said in MIE belongs to the first phase of the explication, apart from 3 pages at the very end of ch. 11 I think that the qualification minimal is necessary because a stronger vocabulary would also be sufficient to specify the practices.

11 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 11 5 where Brandom discusses the practice of rational rectification. Moreover, looking back at the argument of MIE, Brandom recapitulates the two steps of the first phase of the explication described above: [Step 1] What it is for something to state or express a fact is explained in normative terms [i.e., in terms of the game of giving and asking for reasons], and what it is for something to be stated or expressed is explained in turn by appeal to that practice. [Step 2] So what it is to be a fact that is, true claim is explained in normative terms. (MIE 624) He then draws the following conclusion: In this order of explanation[ 12 ], normative notions such as commitment and entitlement [ ] are more fundamental than the nonnormative properties they enable discursive practioners to express explicitly. (MIE 625) As the considerations advanced in this paper should make clear, this recapitulation can only apply to the first phase of the explication of claiming. However, since no hint of the necessity of a second phase is to be found in the vicinity of the above quotes, it is easy to get the impression that, according to MIE, no second phase is necessary. It is therefore tempting to conclude that, at the time he wrote MIE, Brandom thought the first phase would already constitute a complete explication of claiming and only later became fully aware of the necessity of the second phase. Even if this should be true, it has to be noted that there are at least traces of the view that a second phase of the explication of claiming is necessary. For one thing, Brandom describes the practices of rational rectification at the end of ch. 5, though not under that name. However, this is not more than a trace of the second phase because the relation of the norm of rational rectification to the norms constituting the game of giving and asking for reasons remains completely unclear. The second trace is that Brandom repeatedly makes the Quinean point in MIE (cf. 483f and 634), which he takes to imply that we can also only understand the notion of a vocabulary [i.e., claimings] as part of a story that includes facts. (2000a, 163). It is probably this that Brandom has in mind when he writes in Holism and Idealism in Hegel s Phenomenology that he has argued for the reciprocal sense-dependence of the concepts fact and claim in MIE (Cf. TMD 196f, and fn. 32, on p. 383). As far as I can see, in MIE Brandom argues explicitly only for the thesis that one has to understand the concept claim in order to understand the concept fact, but not the other way around. 12 As I have argued above, an explicatory relation sometimes constitutes an explanatory one. Brandom seems to be relying on this fact here because, strictly speaking, he should have said that normative notions are prior in the order of explication, rather than in the order of explanation. Similarly, when Brandom speaks of a pragmatist explanatory [my emphasis] commitment, which insists that the category of facts can in principle only be made intelligible in the context of an account of the practice of claiming. (2000b, 369) Strictly speaking, this is an explicative asymmetry, as Brandom puts it in the next sentence, but this explicative relation also gives rise to an explanatory relation.

12 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 12 3 The Objection from Pragmatist Idealism In this section I will introduce a pragmatist version of idealism and formulate an objection defenders of this position would raise against Brandom s explication of claiming or, more specifically, its inclusion of a second phase (3.1). I will then argue that this objection is mistaken because it ignores the inferentialist context of Brandom s explication of the concepts fact and claim (3.2). 3.1 Sense-Data Idealism and Pragmatist Idealism Defenders of a position I want to call pragmatist idealism would object to the inclusion of a second phase in the explication of claiming because it invokes facts to explicate claimings. Pragmatist idealism holds that the assertional practices have to be describable in a vocabulary that only makes reference to actions and norms governing them. The meta-vocabulary may not, on the other hand, make reference to environing physical facts. To understand the motivation behind this objection, it is useful to compare pragmatist idealism to classical sense-data idealism. Clasically, idealism is the thesis that primarily only mental entities, i.e., representations exist and that facts exist only insofar as they are represented. 13 Classical sense-data theories of cognition assume that we have access only to our sense impressions and that we should understand what physical facts are in terms of these sense-impressions, e.g., as that which gives us the impressions. In a pragmatist version of idealism, the role classically played by representations is played by our linguistic practices. 14 Therefore, pragmatist idealism requires of a pragmatic meta-vocabulary used to describe our practices that it only invoke actions and norms governing them, but not environing physical facts. This requirement is in turn motivated by the view that we are immediately acquainted only with our actions, in particular with our linguistic actions and the norms governing them. On the other hand, we are not so acquainted with the environing physical facts because our access to these facts is always mediated by the language we use. This observation then leads to the view that we should understand what facts are exclusively in terms of possible true claims. In my view, pragmatist idealism is a very natural and intuitive position and I have argued that it is easy to get the impression that Brandom endorses this position in MIE. In fact, however, he only adopts this position for the first 13 Instead of this ontological reading, this thesis can also be given an epistemological reading according to which we ultimately have knowledge only of our representations. Brandom, following Sellars, rejects this way of thinking because it presupposes the independent intelligibility of looks-talk, which in fact is dependent on is-talk (cf. BSD 12). 14 Thus, there is the following crucial difference between the classical and the pragmatist version of idealism: Whereas in the classical case, the ontological or, respectively, epistemological basis is had by an isolated individual thinker, in the pragmatist case, this basis is most plausibly understood as something involving a several individuals.

13 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 13 phase of his explication of claiming, which indeed culminates in the slogan Facts are just true claims (MIE 622). Pragmatist idealists would thus reject the second phase of Brandom s explication of claiming because we do not have access to a reality independent of language and thus not to the thick practice of rational rectification. It is argued that we really only have access to our speech-acts and that we should therefore understand what facts are in terms of these speech-acts and that we cannot in turn invoke facts to explain what speech-acts are. 15 In terms of the distinction between two phases of the explication of claiming this means that pragmatist idealism accepts the first, but not the second of these phases. As I will argue in a second, the objection from pragmatist idealism is mistaken. But if it is accepted, there seem to be two possibilities: A pragmatist idealist can either accept Brandom s argument that a complete explication of claiming has to invoke facts and conclude that such an explication cannot be given, or she can try to reject the Quinean point or the consequences Brandom draws from it, thus opening up at least the possibility that an explication of claiming without yet having talked about facts (Brandom 2000b, 369) could be complete. However, for reasons of space, I will not pursue these matters further, here. 3.2 The Rejection of Pragmatist Idealism I will now show that the argument in support of pragmatist idealism sketched above is mistaken. I would concede, at least for the sake of argument, that we are in some sense acquainted to our speech-acts, in a way we are not acquainted to facts in the world around us. What I doubt is that this is a necessary condition for an act or fact to play a role in practices that are drawn upon to explicate something. This is because the view that the terms in which practices are described have to be directly, intuitively intelligible to us does not go well with an inferentialist approach to content, in this case the contents of concepts such as inference, claim, and fact. According to inferentialism, our understanding of a concept does not consist, at least not exclusively, in our acquaintance with objects falling under it, but rather in our (at least practical) grasp of the inferential relations it stands in. It is a consequence of interentialism that we can get clearer about two dark concepts by noting an inferential relation between them. We do not have to invoke concepts that are somehow antecedently intelligible to get ahead with our understanding of other concepts. Thus, for an inferentialist, the fact that we are not directly acquainted with facts on the world-side of the word/world-gulf does not mean that we cannot invoke these facts in the description of our practices. Doing so just means that we start from the concept fact to see how it is inferentially related to other concepts, such as claim, in order to get clear 15 Apparently in this vein, Price argues that Brandom should have stayed resolutely on the word -side of the word/world gulf (2009, 15).

14 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 14 about the inferential roles played by both of these concepts. 4 Brandom s Deflationalism about Truth I have argued that Brandom is entitled to invoke facts in order to explicate claimings. As far as I can see, Brandom only argues that one has to do so, but nowhere that one can. Against Brandom, however, I would argue that his deflationism about truth and reference can only be upheld in the context of the first phase of the explication of claiming. In ch. 5 of MIE, Brandom defends a deflationist account of true, according to which this word is just a pro-sentence-forming operator by means of which we can form expressions that anaphorically refer to other claims and endorse them. Thus, by using the expressions What she just said is true and Everything the oracle says is true the speaker refers to one particular claim or, respectively, to a set of claims and endorses them. The word true is called a pro-sentence forming operator in analogy with pronouns: Just as the pronouns he, she, it refer anaphorically to singular terms and inherit their content, so expressions formed by means of the word true refer anaphorically to claims and inherit their content. In think that a deflationist account of truth is inevitable in the first phase of the explication of claiming because the explication starts from the practice of giving and asking for reasons and aims at explicating what claims are in terms of these practices and, in a second step, what facts are in terms of claims. Therefore, one cannot invoke facts and claims as independent relata in order to explain the notion of truth as some sort of correspondence between them. In the second phase of the explication, however, a deflationist account is bound to appear incomplete because of the thickness of the practices invoked there. In the second phase, we have at our disposal both the explication of claims in terms of the game of giving and asking for reasons and the description of the practice of rational rectification which involves physical facts. If we confine ourselves in this context to say about the expression true only that it serves as a pro-sentence forming operator, we leave out something crucial. In this context, we also have to say about the expression true that it is properly applied to a claim if and only if the fact expressed by it really obtains. In practice, this also seems to be Brandom s view. In his Sketch of a Program for a Critical Reading of Hegel, where a detailed elaboration of the second phase of the explication of claiming is provided, Brandom says that, according to Hegel, any set of determinate empirical claims [ ] not only must omit some claims that are true and may contain some claims that are not true, it must contain some claims that are not true. (139, my emphasis) Here, I am not concerned with Brandom s or Hegel s claim that any set of beliefs must contain falsehoods. I want to point out that Brandom speaks of the truth and falsity of claims and that the context of this passage makes clear that the point of

15 Bernd Prien: Deflationism and the Two Phases in Brandom s Explication of Claiming 15 using the expressions true and false is not to refer anaphorically back to earlier claims. The point is to express the adequacy or inadequacy of systems of beliefs with regard to the way the world is, which can show up in the experience of error when we acquire incompatible commitments by what we take to be correct applications of concepts. The upshot of these considerations is that Brandom should be a deflationist about truth only in the first phase of his explication of claiming, but not in the second phase. Moreover, it seems that Brandom agrees with this view, at least in practice, because in the passage just quoted, which belongs to the second phase of the explication, he seems to employ a non-deflationist notion of truth. 5 References Brandom, Robert (1994): Making It Explicit. Cambridge, Mass. (MIE) (2000a): Vocabularies of Pragmatism. In: Robert Brandom (ed.): Richard Rorty: The Philosopher Meets his Critics. Oxford pp (2000b): Facts, Norms, and Normative Facts: A Reply to Habermas. In: European Journal of Philosophy 8, pp (2002): Tales of the Mighty Dead. Cambridge, Mass. (TMD) (2005): Sketch of a Program for a Critical Reading of Hegel, in: International Yearbook of German Idealism 3, ed. by K. Ameriks & J. Stolzenberg, Berlin, New York: de Gruyter, 2005, pp (2008): Between Saying and Doing. Oxford. (BSD) Loeffler, Ronald (2005): Normative Phenomenalism: On Robert Brandom s Practice- Based Explanation of Meaning, European Journal of Philosophy 13, pp Price, Huw (2009): One Cheer for Representationalism?, to appear in R. Auxier (Ed.), The Philosophy of Richard Rorty, Open Court, Library of Living Philosophers XXXII, here quoted from Accessed 5 October Prien, Bernd (2010): Brandom on Communication, Reference, and Objectivity, forthcoming in International Journal of Philosophical Studies.