Francisco Javier Gil Martín Jesús Vega Encabo TRUTH AND MORAL OBJECTIVITY: PROCEDURAL REALISM IN PUTNAM’S PRAGMATISM

ABSTRACT. Drawing upon several questions raised in a recent exchange between Putnam and Habermas, this paper exposes Putnam’s pragmatist arguments against both moral skepticism and substantive moral realism and examines his realist account of moral objectivity in connection with the notion of truth. We defend that Putnam’s conception of moral objectivity can be better understood as a new version of a procedural realism. Furthermore, we discuss an internal tension and some problems that threaten the defense of this kind of moral realism.

1. Introduction Hilary Putnam and Jürgen Habermas have met on various occasions and their philosophies have been mutually enriched in such topics as the conception of truth, the notion of equality, the defense of democracy, and the rejection of sceptical and relativist trends. 1 The last episodes of this dialogue have revolved around some divergences on the explanation of moral objectivity. In a lecture delivered in Frankfurt in July, 1999, Putnam unmasked the dangers of holding a sharp distinction between norms and values (2001; 2002a, Ch. 7). In a few words, Putnam presents a sort of indispensability argument against the threat of losing content, if all we need to ground moral objectivity is obey the norms of “discourse ethics.” The argument can be summarized like this: the objectivity of norms presupposes the objectivity of some values. Therefore, some value 1

See the last two lectures of Putnam (1987; 1990a, pp. 217-231; 1990b; 1995; and 1994, Part III). On the other hand, see Habermas (1988, pp. 153-186; 1991, pp. 120-125; 1992, pp. 53-57; and 1999, especially the introduction and Ch. 5).

In: U. Rivas Monroy, C. Cancela Silva, and C. Martínez Vidal (eds.), Following Putnam’s Trail: On Realism and Other Issues (Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 95), pp. 265-285. Amsterdam/New York, NY: Rodopi, 2007.

266

F. Javier Gil Martín and J. Vega Encabo

judgments must be correct, even “true,” for the norms to have content and be valid at the same time. Habermas defends himself, claiming that his differentiation between norms and values doesn’t implicate a noncognitive status of the latter, and rejecting the frequent charges of empty formalism against Kantian ethics. His response insists on the idea that our moral point of view grows from within the ethical conditions of a community and its shared form of life. Values are not completely separate from norms; they are authorized by rational procedures and can also be subject to universality constraints. Even the formal procedures to solve moral problems embody a certain normative substance, such as the demand of full inclusion of the participants and their mutual perspectivetaking. Furthermore, Habermas suspects that Putnam’s view about the objectivity of value judgments tends to assimilate them to descriptive (even scientific) statements. The issue about truth has played a role in this clear misunderstanding (Habermas 2002; Putnam 2002b). But more significant is Habermas’ contention that Putnam should elucidate the priority of questions regarding justice over questions of good life, such as they are established in modern democracies.2 Among all the points raised in this open dialogue, we are going to focus on the understanding of moral objectivity in connection with the notion of truth. Both philosophers would be willing to refute a version of the so-called substantive moral realism. Habermas tries to secure his moral cognitivism with a constructive account of practical discourse, and renounces the application of the predicate ‘. . . is true’ to moral judgements. Putnam, on the contrary, likes speaking of the truth and falsity for evaluative and normative judgments and commits himself to a kind of moral realism. And so Habermas thought of Putnam as a “thoroughgoing realist.” But Habermas himself ought to take seriously both claims of Putnam. Given that norms and values interpenetrate, the objectivity of norms falls or stands with the objectivity of values; and given that issues of objectivity and truth are intimately connected, the defense of a certain kind of moral realism must be possible. In this paper, we argue that Putnam and Habermas should agree on a neutral kind of procedural realism. Naturally we do not claim that appealing to procedures exhausts their understanding of our moral point of view, at least in Putnam’s work. Nevertheless, certain pragmatist undertakings, together with a plain dismissal of a metaphysical realism of values and norms, lead them to emphasize the role of our standards and procedures

2

That was one of the contentious questions in the debate between Putnam and Habermas at Northwestern University, Evanston; November 13 th , 2002.

Truth and Moral Objectivity

267

of moral justification in order to account for the objectivity of moral statements. Our argument will run as follows. First, we will see that indispensability arguments are not designed to explain what the objectivity of a certain discourse domain consists in. Second, we will add that the appeal to the notion of truth does not even help in this task. Third, we will note that the rejection of the analysis of truth in epistemic terms still leaves open the possibility of a close link between justification and truth in explaining moral knowledge. We will conclude by arguing that moral procedural realism exposed in pragmatist terms only resists constructivist temptations by accepting primitive and underived obligations.

2. Indispensability Arguments (IA) in Ethics IA go very deep in Putnam’s philosophy. We are before an antiskeptical argument, with robust roots in Kantian thought (even with a transcendental flavour) and with a prominent pragmatist inspiration. It is antiskeptical, because it points up a cognitivist and objectivist account in different domains; Kantian, because it raises the question about the possibility of knowledge; and Pragmatist, because the indispensability is always seen from the priority of practice. To appreciate the scope of IA, it is necessary to answer (a) (b)

questions about the conclusions of the argument and questions about what grounds indispensability in every domain.

First, IA in Putnam’s version are designed to support cognitivism and objectivity; they are thus opposed to certain skeptical threats. Second, Putnam uses two distinct strategies to prove the motives for the indispensability in the normative case: the “companions in the guilt” and the inseparability arguments. Let us examine these questions closer. (a) How could we derive cognitivism from indispensability? Putnam thinks that indispensability justifies our treatment of some kind of statements as true or false (1994, p. 154; 2002a, p. 111). Nevertheless, we could suspect that this superficial fact is not enough to establish that they could be really true or false. What prevents us from thinking that we are under a mere illusion of cognitive content and an appearance of objectivity? But, from where could we ground the suspicion that we entertain an illusion? By the very premise of the argument, we cannot but place ourselves within the game of language we are considering, because

268

F. Javier Gil Martín and J. Vega Encabo

only within it are the statements viewed as indispensable. Dispensability could only be established by legitimizing a jump to a perspective external to the game. (When Quine presented the argument in philosophy of mathematics, claimed that no first philosophy could ever transcend science to put science itself into doubt.) The IA excludes an Archimedean point of view, from which we could discover the dispensability of concepts and forms of discourse that are deeply integrated in our scientific and daily practices. Such a detached point of view is adopted by non-cognitivist approaches in ethics that lead to revisionist descriptions of (the presumed cognitive content of) our ordinary moral convictions. Indispensability means that adopting an Archimedean perspective to criticize and revise moral language has consequences for the understanding that agents have of their own activities of giving reasons and asking for them. In other words, treating moral statements as true or false, as cognitively significant because they are indispensable in describing the language game of morals, is equivalent to eliminating any possibility for any revision that would make illusory our practices. So, from the internal point of view of our normative practices, it would be incoherent to consider moral judgments as lacking cognitive status. The IA defends cognitivism from the threats of an external, Archimedean skeptic. Putnam’s argument resembles a Kantian mode of thought; it can be seen as the result of a transformation in the way we ask certain questions. Instead of arguing about how we could cognitively have access to Platonic entities in order to secure the objectivity of mathematical knowledge, we ask about the function that mathematical statements play in science. Practical considerations on the indispensability of some statements suffice to establish the possibility of knowledge and the cognitive status of them. Otherwise, more entrenched areas of discourse would be affected. Indispensability concerns the intelligibility of our practices and gives reasons for legitimating a certain conceptual scheme without following the full path of a transcendental foundation. So, arguing for the indispensability amounts to a way of justifying the applicability of some concepts in practice. For example, it is indispensable in giving sense to the statements of physical science to accept the conceptual scheme of mathematics. We could also talk about a necessity derived from the impossibility to make sense of our practices without supposing the cognitive status of a certain domain of language. Although it is not about transcendental necessity committed to the unrevisability of those presuppositions, the Kantian roots of the argument

Truth and Moral Objectivity

269

are evident: we could not actually think without applying those concepts (Putnam 1994, pp. 245-263). (b) But we have not yet established the real indispensability of any conceptual scheme. The indispensability must be motivated and grounded in each particular case. For mathematical statements, the argument seems clear enough. Given the deep integration of mathematics in our physical objective knowledge of the world, to consider mere fiction the conceptual scheme of mathematics would also alter such knowledge and affect the very cognitive status of physical laws. 3 The same argument does not work for normative knowledge and moral knowledge. One difference concerns the scope of the agreement reached in mathematics and in morals. But it is not the only difference: normative predicates are not integrated in science or daily life in the same way as mathematical concepts are integrated in fundamental laws of physics. To overcome these difficulties, Putnam proposed two further arguments for the indispensability of normative discourse and against the subjectivist conception of normative concepts. (i) The “companions in the guilt” argument stressed the incoherence of being non-cognitivist in ethics and cognitivist in epistemology. “All values are in the same boat,” as Putnam often says. All values are predicates that guide actions; so their descriptive use is not detachable from an evaluative or normative one. If all predicates that are actionguiding would be “queer,” as Mackie contends, those predicates integrated in scientific practices (such as coherence, simplicity or explanatory power) would be also queer. The practice of science would be affected by the distortion of evaluative and normative language encouraged by expressivists and subjectivists. The notion of “fact” would be unintelligible unless the criteria for justifying the facts are subject to rational scrutiny and have cognitive status. We cannot build a scientific image of the world without essentially involving the reasons that guide the selection of hypotheses. Describing a scientific theory as simple or coherent is equivalent to saying that we are justified in accepting the theory and claiming that we ought to accept the theory. Therefore, either it has sense to talk about “facts involving duties” or we relinquish our claim that science provides us with objective facts. If “values” seem a bit suspect from a narrowly scientific point of view, they have, at the very least, a lot of “companions in the guilt”:

3

“Mathematics and Physics are integrated in such a way that it is not possible to be a realist with respect to physical theory and a nominalist with respect to mathematical theory” (Putnam 1979, p. 74). See also Putnam (2004, p. 66).

270

F. Javier Gil Martín and J. Vega Encabo

justification, coherence, simplicity, reference, truth and so on, all exhibit the same problems that goodness and kindness, from an epistemological point of view. (Putnam 1990a, p. 141)

(ii) This first argument is combined with a second one. The doubts casted on the objectivity of the normative are consequences of a mistaken and strict dichotomy between facts and values. The entanglement and interweaving of fact and values, facts and theories, facts and interpretations has become a recurrent theme in Putnam’s work during the last two decades. Thinking of facts in an empiricist or positivist vein unables us to see how value statements could be rationally justified. Ethics cannot be about matters of fact insofar as “facts” are observationally verifiable or reducible to observation predicates that would allow us to confirm factual statements by a simple test procedure. Obviously, evaluative and normative terms are not descriptive in this sense. The second part of Putnam’s argument is positive: if “value and normativity permeate all of experience” (2002a, p. 30) and if it is not possible to factorize in our epistemic or moral experience between the descriptive and evaluative (normative) components, then it is better to think about facts and values as entangled. The entanglement is shown by words as cruel, “used sometimes for a normative purpose and sometimes as a descriptive term” (2002a, p. 35), and it points to the fact that we are not actually able to describe such kind of terms without using the very same terms in a normative context. The normative point of view is unavoidable. This is how ethical concepts bear on the world. This way of seeing the fact/value entanglement leads us to the core of IA in the context of normativity. We said that indispensability disallows an external perspective from which to judge our moral life or to reject our normative practice as illusion. The primacy of practice is the expression of the indispensability. In fact, IA capture the idea that reasons have authority for us. Our life could be unintelligible without viewing ourselves as being subject to the authority of duties, norms and values. Normative language is indispensable because it is constitutive of the point of view of the participants. We cannot understand what it could mean to deny that there are normative truths, in the same way in which we cannot make sense of our thought without assuming that there are logical truths. It would be foolish to talk about what thinking signifies for us without presupposing the conceptual scheme of normativity. We cannot get rid of evaluative and normative vocabulary without risking the loss of the normativity as a whole. The normative problem does not rest solely on explaining the function of normative concepts, but on understanding how we are justified in our rational requirements. The

Truth and Moral Objectivity

271

main idea of the argument would be the following: all the efforts to explain normative concepts in terms of projections performed by the agents are condemned to lose the specificity of their normative authority, which manifests itself only from the first-person point of view. This fact also offers an answer to the possibility of normative knowledge in general. The Kantian roots reappear at this point: reflection on our practices clears the path to recognize the conditions that make possible our justified moral agency. Of course, instead of being transcendentally shaped, reflection arises from the conditions of the practice of inquiry and thereby, values and norms will not be infallibly justified. Revisability is not prevented, but we will not be able to know the conditions for it from the outside of the very normative stance.

3. Truth and Objectivity If the IA pervades the realm of normativity, then we have “reasons for rejecting the widespread belief that ethical judgments lack objectivity” (Putnam 2004, p. 3). But, what do we understand by the objectivity of normative language? Putnam will claim that moral statements can be taken as correct, incorrect, true or false, in the sense that they are “under rational control, governed by standards appropriate to their particular functions and contexts” (2002a, p. 33). However, Habermas contends that talking about “truth” in ethics compels us to embrace a substantive evaluative realism. Let us see briefly how Putnam’s notion of “objectivity without objects” uncouples the moral truths from this supposed bond with a substantive realism. There are logical reasons to claim the truth-aptness of moral statements and to state that the same notion of truth is applicable to any kind of statements. The difficulties found by the expressivist to account for conditionalization and for arguments using descriptive and evaluative statements are at the same time good reasons to uphold that moral statements are truth-apt. If the government succeeds in convincing the rebels to lay down the arms, the situation of the civilian population will improve; the government has succeeded in convincing the rebels to lay down the arms; therefore, the situation of the civilian population has improved. These facts can be accommodated within a minimal account of truth. The Equivalence Principle explains our use of the predicate ‘. . . is true’ within our languages. Accordingly, the judgment that it is true that murder is wrong is equivalent to the judgment that murder is wrong.

272

F. Javier Gil Martín and J. Vega Encabo

Someone could argue that this scheme accounts for superficial facts about the use of moral statements and that those superficial facts would disappear after providing a correct interpretation of moral statements. In other words, that moral language presents some features of the assertive mode of speech does not let us draw any conclusion about the realism/antirealism issue and the objectivity of moral judgments. What does it explain that we can assert, deny and ascribe truth or falsity in moral language? And what is meant by correctness and validity in ethics? The problem of truth for moral statements becomes interesting when it involves other theses about the value of truth and about how moral statements bear on reality. In the Dewey Lectures, Putnam displays an account of truth that provides some responses to these questions. He tries to block the deflationist’s use of the Equivalence Principle together with a verificationist theory of meaning. But he does not support this rejection by positing a substantive property that would make our assertions really true or false. Once the Equivalence Principle is accepted by logical reasons, the philosophical task is to give an account of what understanding is. And for this task, we need to escape the alternative between the verificationist and the metaphysical realist: on the one hand, the statements we use to assert something can transcend verification and, on the other, this does not require postulating a queer substantive property, a “substantive sort of rightness” (1999, p. 54) with a special metaphysical burden. Putnam then puts forward the following theses. First, we understand the predicate ‘is true’ by understanding what is asserted by the proposition from which we predicate truth; second, it is possible to formulate meaningful statements which are recognition-transcendent, because in asserting them we are just claiming that what is asserted is beyond our recognitional powers; and third, understanding does amount to seeing something in the marks and noises without positing something that goes beyond them. These theses articulate a new IA that points out that verificationists and metaphysical realists cannot do justice to “who we are and the sense that our practices have for us” (1999, p. 64). It is our self-understanding that would be affected by both positions. Suppose that understanding consists in knowing the assertability conditions of sentences. Then every change in the context of use would cause a change in meaning. But if this is so, we would distort the way in which understanding takes place for us. On the other hand, if we posit something beyond our practices to explain how understanding is possible, then we are trying to reflect a necessity external to our practices “and the

Truth and Moral Objectivity

273

ways of thinking internal to them” (1999, p. 63). We could even claim that learning a new set of concepts entails the ability “to see the face” of the activities that ground our own practices in other activities. If we cannot do that, then we are reaching the limits of the intelligibility with respect not only to the practice we are learning and evaluating, but also to our own practice. Correctness denotes the way a sentence in use “bears on reality.” Not to violate the priority of the normative viewpoint is a formal constraint on every notion of correctness. As such, it does not say anything about the way we understand how normative (and moral) concepts bear on reality. But we know that, given the intimate link between our language, our actions and the world, we are testing the sense and normative understanding of our own point of view when we inquire about the extension of our set of evaluative and moral concepts. Since not every statement “bears on reality” in the same way, we no longer need to explain the truth and objectivity of moral statements as necessarily performing a descriptive function. Moral statements are not made true or false by (non-natural) objects described by these statements. Habermas was therefore misguided when he objected that Putnam, in talking about the truth and falsity of value statements, was treating them as descriptions. Identifying the way moral language bears on reality is crucial to account for the objectivity of moral statements. We find valuable insights in this regard in Putnam’s recent book Ethics without Ontology. There he defends that logical and mathematical truths “bear on reality” without describing a set of special entities and properties. Something similar is argued for evaluative and normative statements; it’s possible to talk about objectivity without objects in logic, mathematics and normativity. We will not repeat the arguments against the descriptive use of logical and mathematical truths. We are interested in Putnam’s positive proposal about the objectivity of some statements considered as conceptual truths. The latter are not conceived as analytical, necessary and unrevisable, but as truths for which “it is impossible to make (relevant) sense of the assertion of its negation” (2004, p. 61; 1994, pp. 255-256). These truths are identified by way of interpretation into a body of beliefs; they are interwoven with other empirical beliefs. They are revisable as a result of deep changes in opinion about what we can “conceive”; certain intuitions can then disappear after taking into account alterations in the justification of those conceptual truths. It is our familiarity with what the justification is, for example, in the domain of logic that explains how one learns what logical truth is. Analogous considerations apply to mathematical truths and to methodological value judgments. Nothing prevents us from applying the same strategy for

274

F. Javier Gil Martín and J. Vega Encabo

evaluative and normative truths. Under the conditions of the ethical life in which we are involved as participants, it would make no sense to affirm the negation of certain evaluative and normative assertions. Of course, this kind of truths are again revisable as part of a whole of empirical and conceptual beliefs. And such a revision becomes possible because we are in fact familiar with moral justification. We have learned the standards and procedures of morals. Learning what moral truths are is internally connected to learning the procedures that help us to establish them. And among these procedures, those ones used to apply the truths in our individual and social life play a significant role. It is the familiarity with these procedures that it would explain that one cannot recognize certain truths but from a certain (normative) perspective. The comparison with logical and mathematical truths highlights why moral knowledge is subject to the same constraints of fallibilistic inquiry. Practical reasoning is guided by norms of truth and correctness. Evaluative and normative truths are revisable through learning processes, but they are not made true or false by objects and properties such as intrinsically non-natural values. Therefore, explaining how evaluative and normative statements bear on reality does not entail the defense of a substantive moral realism. We claim that it is closer to a version of a procedural realism, in which the objectivity of moral knowledge is explained in terms of the procedures we use to recognize evaluative and normative truths. It does not mean that all evaluative and moral truths are “conceptual” truths. But, regarding the normative point of view from within which these truths are available, some of them should have to be conceptually true if we want to preserve the whole world of normativity.

4. Truth and Justification Nevertheless, such a line of reasoning engenders obvious questions. Some of them are also prompted by the debate between Putnam and Habermas. Are both authors implicitly committed to an epistemic concept of truth? Are normative and moral truths identified with statements whose validity is established by procedures learned through inquiry? And are those procedures, in a sense, an essential aspect of normative and moral validation? The concepts of truth and justification are not linked by definition. In the last decade, Putnam has argued against the temptation to assimilate truth with warranted assertability even in ideal conditions. Truth is not definitionally the outcome of an ideal conduct of inquiry, because there

Truth and Moral Objectivity

275

can be truths that we cannot ever verify. Putnam’s objections to Habermas emphasized this point in the case of ethics. It was not only that we find difficulties making sense of an ideal speech situation without admitting “thick” ethical terms. It was also that “there is no reason to believe that the outcome of an ideal and sufficiently prolonged discussion of ethical questions would inevitably be correct” (2002a, p. 126). It must be noted, however, that Habermas has distanced himself from his previous Peircean account of truth. Criticisms by A. Wellmer and C. Lafont convinced him that “truth” is a property that statements cannot lose, and that the unavoidable epistemological link between truth and justification “must not be turned into a conceptually inseparable connection in the form of an epistemic concept of truth” (Habermas 1999, p. 52). The very notion of an ideal speech situation is also abandoned because it falls into the fallacy of misplaced concreteness in suggesting an state of definitive agreement which would make all further communication superfluous. The procedural conditions of discourse permit only to anticipate the argumentative force of those reasons that would convince us. Instead of evoking an end state as a regulative ideal, they favour an ongoing process of critical discussion by which objections are weakened and the acceptable becomes what survives criticism. Nevertheless, Habermas does trace a clear-cut distinction between the concept of validity at work in scientific and descriptive discourse and the validity of moral statements as ideal warranted assertability. Whereas truth claims in the case of empirical statements pragmatically presuppose a world of objects existing apart from our descriptions, moral validity claims lack such a realist commitment and in themselves hinge on the recognition by all the participants in the discourse. Moral correctness is conceptualized as a validity claim in a sense analogous to the truth, albeit without the same presuppositional commitments, and defined by the deontology of the just and the right, in terms of the norms and actions being equally in the interest of all possibly affected (Habermas 1999, pp. 271-318).4 Moral objectivity is thereby explained in a cognitivist and constructivist way. Value judgments do not respond to the same validity conditions. But they are candidates for embodiment in norms. And when values are discussed from an impartial point of view and are rationally accepted by all the participants in a moral discourse, then the content of the judgment can be authorized as obligating under universality conditions. In that

4

Obviously, the terms of the analogy were more convincing before dropping the epistemic notion of truth (Habermas 1991, pp. 11, 126, 130-131).

276

F. Javier Gil Martín and J. Vega Encabo

sense, they get entangled with moral norms. At the same time, Habermas accepts a weak notion of cognitive content for evaluative statements and insists that the meaning and significance of values can be rationally debatable in ethical discourses concerning the self-understanding and life-projects of persons and groups. Values are not objective just in the sense of being accepted within a community as a set of cultural conventions or customs. They deserve acceptance when subject to rational scrutiny and supported by good reasons. They are objective insofar as they are recognized intersubjectively by means of compelling reasons. Values have to be authorized in a practical context of rational discussion and inquiry. So Habermas is reluctant to consider them in naturalist terms, as sociological facts. Some convergences may not be less important than the differences between Putnam’s realist cognitivism about evaluative and normative judgments and the Habermasian account of ethical and moral discourse. We know that value judgments cannot be authorized or made true by reality itself. Habermas is willing to differentiate the assertoric mode of speech from the deontological mode as regards the validity claims. But his only reason is to secure that validity claims in the moral domain are unconditional and not to admit the existence of substantive “moral truths.” Moreover, rather than underestimating the cognitive status of value judgments, Habermas tries to hold a distinction internal to our normative discussion between the questions regarding justice and the questions regarding goodness, a distinction that he finds neglected in Putnam’s philosophy. But what is more important is that for both of them what is true (or valid) in ethics does not diverge from what is endorsed by the best reasons, even with regard to value judgments. As Putnam puts it: “In the case of ethics (unlike science), the true view cannot differ from the view for which there are the best reasons” (2002a, p. 175). He occasionally expounds that intimate relation by referring to how we judge objectivity in an ordinary sense, roughly equivalent to what Kant named the “enlarged mind” in the “maxims of common understanding”: Statements are called “objective” if their claim to truth is not dependent on idiosyncratic standpoints or on disregarding the standpoints and interests of others. A sufficient condition that an ethical claim be objective in this “ordinary” sense is that it be reasonable from the standpoint of an interest in the common welfare, where the common welfare is [. . .] something that is itself to be determined by intelligent discussion among persons who share this commitment. I wish to emphasize that this is something that I suggest only as a sufficient condition, and by no means a necessary one. “Value judgments” are not a homogeneous class, and different sorts of judgments possess different

Truth and Moral Objectivity

277

sorts of objectivity. If it is reasonable for those affected to accept an ethical claim of this sort after experimentation and discussion, it will be – so long as reasons to question it do not arise – what John Dewey called a “warrantedly assertable” claim. (Putnam 2002c, pp. 17-18)

Such a minimal condition is essentially procedural insofar as it links the objectivity of moral truths to their justification. In these terms, it is akin to what Habermas says about the deliberative process of the moral point of view. He emphasizes the role of the procedures in filtering the reasons given in support of the validity of moral norms. This filtering works by rationally grounding common interests in shared values and by examining the norms that incorporate these values: universal norms are worthy of recognition because they are in the common interest of all participants in the discussion and because they are equally good for all of them. That minimal condition actually applies to the ethical discussion about values, as we the affected deliberate to select our best values and to identify the common good from the point of view of what is the best “for us.” In other words, Putnam and Habermas agree on some minimal procedural conditions that would link the justification of our judgments with the consideration of what is an ethical truth (Putnam) or a valid norm (Habermas). Putnam would object that the procedural conditions do not exhaust the necessary conditions for the objectivity of ethical statements, because not every evaluative judgment work in the same way. But once we have given up postulating a realm of intrinsic moral facts or values, it is not easy to see how objectivity can be ascertained otherwise. A brief reflection about the reasons to disavow a substantive moral realism would clarify the inevitability of the so-called procedural realism.

5. Procedural Realism In The Sources of Normativity, Christine Korsgaard introduced a distinction between two kinds of moral realism, substantive and procedural. Any theory of normativity, also in ethics, has to explain how reasons have normative force without distorting the role they play in our life, seen from the first-person perspective. Traditionally, it has been thought that only some version of moral realism might reply to subjectivist and relativist arguments and in general to any sceptical position. But it is possible to identify different realist commitments. The

278

F. Javier Gil Martín and J. Vega Encabo

question about moral “reality” is one about how objective moral judgments are possible.5 Korsgaard (1996, p. 35) characterizes substantive moral realism as “the view that there are answers to moral questions because there are moral facts or truths, which those questions ask about.” According to Korsgaard, what distinguishes such realism is the way it explains the objectivity of moral statements. In a certain sense, it adopts some dubious metaphysical commitments. Putnam consistently rejects its distortion of our understanding of the objectivity of reasons: “I see the attempt [. . .] to provide an Ontological explanation of the objectivity of ethics as a[n] [. . .] attempt to provide reasons which are not part of ethics for the truth of ethical statements” (2004, p. 3). If the realist accounts for our moral “reality” in terms of the discovery of intrinsically normative entities, then the link between the standards and procedures prevailing in our moral language and the truths to be known would be severed. But our moral life involves per se what agents think from the inner perspective of their moral activities; there is not available a point of view from nowhere. The substantive realist strategy is pseudo-explanatory: on the one hand, it posits objects whose explanatory necessity is not obvious; on the other, it has to appeal to the self-evidence of normative truths, and then refuses to answer the normative question. The procedural realist proposes a different account for the objectivity of ethical judgments. He would defend three theses. The first one is very neutral and could accommodate a wide variety of alternatives, while the other two are more restrictive. (1) (2) (3)

There are correct and incorrect answers to normative (moral) problems; there are better or worse reasons; there are correct or incorrect modes of applying normative (moral) concepts. Moral statements can be treated as true or false, valid or invalid. Moral truths are not independent of the procedures to establish them.

The procedural realist notion of moral objectivity cannot be understood without invoking the way agents determine the existence of reasons for moral truths. It is not enough to claim that there are correct or incorrect answers; it is also required that what explains the objectivity be connected with our means to reach those answers. 5

In Putnam’s words: “As a being who makes value judgments every day, I am of course committed to the idea that there are true value judgments: what must be the case if there are to be true value judgments? In what kind of a world can there be true value judgments?” (1995, p. 43).

Truth and Moral Objectivity

279

Putnam staunchly upholds that there are no recognition transcendent facts in ethics. He also pays attention to the role played by procedures in explaining moral objectivity. In “William James’s Ideas,” written together with Ruth A. Putnam, he characterizes James’s ethics as an “imperfect procedural conception” (1990a, p. 225). The core of the position is that we cannot “discover” what truth or goodness is independent of our procedures. Rightness amounts to being the outcome of a self-correcting procedure. The success of a procedural rule cannot be established irrespective of the very same procedure and, given that our procedural rationality is imperfect, changes and revisions are possible. In short, Putnam’s explanation of moral objectivity has to do, at least in part, with the application and correction of procedural rules. The only constraints are to avoid conceiving procedures as algorithms and as not subject to fallibility conditions. Now, the thesis (3) is ambiguous and admits two different readings. (3a)

Normative truths are essentially dependent on rational procedures.

What it is to be a normative truth is to be the outcome of a procedure of justification. This thesis needn’t be interpreted as implicating a definitional link between truth and rationality. But in talking about conceptual truths Putnam has insisted on the inseparability of our conceptions about what a normative truth is and our judgements about what is reasonable or unreasonable. And in some places he has argued that the notion of truth in ethics is internal to the practical reasoning itself. If we ask about what it really makes sense from the moral point of view, we identify some truths that are not understandable without considering exercises in reasoning. But what kind of dependence are we talking about? If there is some necessity in the way moral truths connect with the reasons moral agents may articulate through the use of procedural rules, such a necessity derives from the fact that we could not make sense of our moral language without invoking the standards and procedures we use to justify our moral demands towards others and towards ourselves. Are we then faced with a weak transcendental necessity? (3b)

Normative truths are not completely independent of rational procedures.

Truth does not always transcend warranted assertibility. Truth is not radically non-epistemic. In Putnam’s terms, “truth does entail warranted assertibility if conditions are sufficiently good” (2002a, p. 108). But this thesis could hold simply that there must be a connection between the

280

F. Javier Gil Martín and J. Vega Encabo

truth and the justification of our moral claims if we are not to be condemned to skepticism. Some truths must be justifiable if we have any moral knowledge at all. Even if one is not very worried about skeptical challenges, to maintain that the very nature of moral truths entails that they are justifiable needs more argument, surely a kind of dependence as derived from (3a). Is it possible that normative truths transcend the best reasons we have for our moral claims? If we take seriously Putnam’s criticisms of the idea that the “outcome of an ideal and sufficiently prolonged discourse of ethical questions would inevitably be correct,” we would have to ask how to understand correctness if goodness and truth are nothing but “regulative ideals” independently of our imperfect procedures. There are two options here: either the procedure is not good enough or some standards of correctness are in fact separate from the procedures. This last option seems to commit us to some version of substantive realism. We should thus assume that our procedures are incomplete and revisable in such a way that we might identify those factors that would improve our methods. And then, what are the reasons to prefer some procedures? Are they independent of the procedure itself? In any case, it could be argued that this line of reasoning moves away from the real intentions of both Putnam’s insistence in procedures and his criticisms of Habermas’s discourse ethics. The issue of real concern is that of emptying moral claims of any content while preserving an extreme minimalist conception of ethics. The question is then how to fill ethical reasons with content while avoiding cautiously the postulation of intrinsically moral facts. (A) Our maxims frequently incorporate thick ethical terms. Habermas would accept that values play a role in formulating those maxims that arise from contingent forms of life and are subject to the universalization test. Again, this cannot be the issue at stake. Rather it is a matter of how to understand the validity of ethical maxims with thick terms regardless of the procedures that would lead us to judge them from a moral point of view. Putnam needs to claim that these kinds of evaluative judgments are in a sense objective before entering the discourse. An IA is here available. Although conditions in moral discussion, the rules of discourse ethics would be empty without incorporating thick ethical concepts; then, we have to presuppose that these concepts have some kind of cognitive content. Thick ethical vocabulary is inextricably linked to the very description of the procedures we use to discuss and evaluate our normative statements. If we would not assume that they are objectively correct or incorrect, then it would be in danger the validity and

Truth and Moral Objectivity

281

objectivity of the outcomes of our moral procedures. “Relativism of any kind with respect to value-terms cannot leave the objectivity of “norms” unaffected” (Putnam 2002a, p. 120; 2002b, pp. 308-309). But as we have seen, the IA does not say anything about what we understand by objectivity of evaluative judgements. It only shows that they must be treated as objective if we do not want to renounce the objectivity in the whole domain of ethics. Following Dewey at this point, Putnam would claim that it is to be subject to criticism and rational discussion what converts the valued within a community into the valuable.6 But this was precisely the point of the procedural realist. (B) Another possibility would be to recognize that our procedures embody certain substantive assumptions as, in Habermas’s case, the nonrestrictive inclusion of all the affected and the mutual exchange of perspectives in examining the moral correctness of maxims. Putnam conceives of these sorts of norms as non-formal conditions in our procedures (1990a, p. 224; 1994, pp. 154-155). Even if we concede that our conflicts of value and the discrepancies between the life-worlds of different communities are hard to solve, we may yet admit that we have in fact arrived at some agreements about these substantive conditions of our procedures. The question is then how to explain the emergence of such agreements. How do the most general truths about normativity appear as having normative authority for us? Putnam claims that it is not possible to derive principles from reason itself; normative truths are not a priori principles of reason. Nonetheless, he does defend the Kantian intuition that norms are “discovered not by mere trial and error, but by normative reflection on our practice” (1994, p. 168). Reflection engenders corrigible knowledge of those conceptual truths that are presupposed in our practices. These truths bring to light the presuppositions guiding the learning and experience of the participants in the ethical inquiry. Reflection takes place within the process of an open and cooperative inquiry. Conceptual truths about what is rational are not extracted by a transcendental reflection from the reason itself, but from the very process

6

In his reply at the conference at Santiago de Compostela, Putnam insisted that he is more interested in Dewey’s recovery of the Aristotelian idea that “the judge of the good life is the good man.” However, such an idea of a judicious or prudent man as a norm of moral judgments leaves the problem unresolved. According to Aristotle himself, it is the adequacy of the process of practical deliberation that offers genuine moral reasons. What is at stake here is how to properly build the procedures of moral reasoning. We find valuable what a good man does because it is the virtuous outcome of the forms of moral reasoning.

282

F. Javier Gil Martín and J. Vega Encabo

of research. So, the validity of a procedure cannot be unfolded in terms of being coherent with a priori norms of reason. Our standards and procedures are developed through criticism into the practices of moral agents. This pragmatist trend is crucial in understanding how we can improve substantively our procedures. Again, they are valuable insofar as they are accepted after a continuous process of inquiry and criticism. But that still does not answer our question about the agreement regarding the procedures and the set of abstract values embodied in very general procedures. If the agreement must grow from within the activities performed by moral agents, there must be some questions and solutions that are not in dispute. Putnam always argues from the impossibility of a global skeptical doubt, but in no place he expounds how it is possible that we really develop a moral point of view as an impartial means of solving ethical conflicts. It is not enough to charge contextualism with incoherence; it is also needed an explanation about why we ought to accept the conditions of a moral point of view. The IA should not assume that this could not be put into doubt. In fact, our recent past and our present places us before numerous cases in which an impartial moral point of view was not or is not accepted as the rational way to discuss moral conflicts. And we are yet confronted with difficulties to explain the improvement of procedures. Habermas would answer that we already presuppose, as a kind of anticipation facilitated by the very moral point of view, the inclusion in the discourse of people whose interests could be violated in the resolution of a conflict. He believes moreover that the mechanism for moral learning processes is internal to the pragmatic structure of rational discourse. We know that we cannot think of the success of the procedure irrespective of the procedure itself, and also that it is difficult to secure that reflection would lead us to the best procedures. To ask about the best conditions for the improvement of our procedures is not to adopt an external and transcendental point of view. The difficulties arise from the very rational acceptability of our practices of ethical valuation. Even if we suppose that there are some values and criteria that could be shared by different communities or ways of life, we still do not know what the correct resolutions of their conflicts are. And in any case, we could not guarantee that the deliberation would lead, in the domain of the procedures, to revise our standards of rationality toward the presuppositions of general inclusiveness and mutual perspective-taking. Or is it empirically true that societies which do not organize the inquiry in a democratic form are irrational? Or is it rather a moral truth,

Truth and Moral Objectivity

283

insofar as certain norms of inquiry can be considered as obligations and, even, moral obligations? What does it mean that to organize our inquiry by some principles, like those of discourse ethics, has normative authority? We tend to think that either we read the IA as a transcendental argument 7 or we need to posit some primitive obligations regarding the search of truth and the search of goodness. And it is not clear whether Putnam is willing to such a radical move.

6. Conclusion Cognitivism in ethics does not require a moral substantive realism that postulates intrinsically normative properties. Putnam and Habermas agree on this point. Having concentrated selectively on only one of the “legs of the table,” according to Putnam’s image of ethics as a system of interrelated concerns (2004, p. 28), we have maintained that their attempts to explain moral objectivity by recourse to procedures, viewed from a pragmatist perspective, are not far apart. The way we carry out the inquiry is in itself open, cooperative, fallibilistic. The objectivity of moral statements cannot be understood independently of the best reasons we have as moral agents. Nonetheless, the internal skeptic could bring up the difficulties to solve some ethical disagreements, especially when the discussants do not agree on the minimally good (substantive) procedures to solve practical problems in general. We cannot discover any necessity in the idea of rationality as the result of our involvement in an incessant and inclusive discussion. How do we accept the (rational, and even moral) obligations to conduct our inquiry under the principles of a democratic and unblocked scientific order? This is not a question about why we are rational or ethical beings at all. It is a question internal to some conflict, a question that has been answered either by appealing to our transcendental intimacy with obligations of rationality – as Habermas has suggested – or by being confident in the results of our historically shaped human experience. Undoubtedly, we have left aside one crucial point in Putnam’s thoughts about ethics. Maybe there are other ways of “bearing on reality” and responding the demands made by others. Moral life is shaped by the experience of certain demands made on us by other sentient life. There

7

That is the case of Habermas’ repeated arguments concerning the lack of “functional equivalents” and the “normative” content of the inescapable conditions of communicative praxis.

284

F. Javier Gil Martín and J. Vega Encabo

are even facts that correct our mistakes in reasoning, facts regarding “the cries of the wounded,” facts that generate also primitive moral obligations. Putnam (2004, p. 24) writes that, “for Levinas, the irreducible foundation of ethics is my immediate recognition, when confronted with a suffering fellow human being, that I have an obligation to do something.” The human being, an actual existing person, is then immediately valued. If we don’t have reasons to arise skeptical worries, this appearance of normativity seems enough to obligate me. But this is the philosophical position that authors like Nagel has identified as a modern version of substantive moral realism. Only stupidity of mind, as Samuel Clarke said, or corruption of customs can lead us to doubt about things so naturally plain and self-evident.

Acknowledgements We appreciate the comments and criticism of Hilary Putnam in response to our paper, presented at the Second Meeting on Pragmatism. Hilary Putnam’s Pragmatism (Santiago de Compostela, Spain, 13th -14 th May, 2004). We are indebted to Fernando Broncano and Carlos Thiebaut for helpful discussions. Cristina Lafont and Axel Müller provided us with insightful remarks and inestimable information. We are also grateful to David Jamieson for improvements to the English text. Javier Gil benefited from a postdoctoral grant at Northwestern University, funded by the Spanish Government (MECD). Northwestern University Department of Philosophy 1880 Campus Drive Evanston, IL. 60208-2214, USA e-mail: [email protected] Universidad Autónoma de Madrid Departamento de Lingüística, Lógica y Filosofía de la Ciencia Facultad de Filosofía y Letras Campus de Cantoblanco 28049 Madrid, Spain e-mail: [email protected]

Truth and Moral Objectivity

285

REFERENCES Habermas, J. (1988). Nachmetaphysisches Denken. Frankfurt: Suhrkamp. Habermas, J. (1991). Erläuterungen zur Diskursethik. Frankfurt: Suhrkamp. Habermas, J. (1992). Faktizität und Geltung. Frankfurt: Suhrkamp. Habermas, J. (1999). Wahrheit und Rechtfertigung. Frankfurt: Suhrkamp. Habermas, J. (2002). Werte und Normen. Ein Kommentar zu Hilary Putnams Kantischem Pragmatismus. In: Raters and Willaschek (2002), pp. 280-305. Korsgaard, C. (1996). The Sources of Normativity. Cambridge: Cambridge University Press. Putnam, H. (1987). The Many Faces of Realism. La Salle, IL: Open Court. Putnam, H. (1990a). Realism with a Human Face. Edited by J. Conant. Cambridge, MA: Harvard University Press. Putnam, H. (1990b). A Reconsideration of Deweyan Democracy. Southern California Law Review 63 (6), 1671-1698. Putnam, H. (1994). Words and Life. Cambridge, MA: Harvard University Press. Putnam, H. (1995). Pragmatism: An Open Question. Oxford: Blackwell. Putnam, H. (1999). The Threefold Cord: Mind, Body, and World. New York: Columbia University Press. Putnam, H. (2001). Werte und Normen. In: L. Wingert and K. Günther (eds.), Die Öffentlichkeit der Vernunft und die Vernunft der Öffentlichkeit, pp. 280-313. Frankfurt: Suhrkamp. Putnam, H. (2002a). The Collapse of the Fact/Value Dichotomy and Other Essays. Cambridge, MA: Harvard University Press. Putnam, H. (2002b). Antwort auf Jürgen Habermas. In: Raters and Willaschek (2002), pp. 306-321. Putnam, H. (2002c). Pragmatism and Nonscientific Knowledge. In: J. Conant and U.M. Żegleń (eds.), Hilary Putnam. Pragmatism and Realism, pp. 14-24. London & New York: Routledge. Putnam, H. (2004). Ethics without Ontology. Cambridge, MA: Harvard University Press. Raters, M.-L. and M. Willaschek, eds. (2002). Hilary Putnam und die Tradition des Pragmatismus. Frankfurt: Suhrkamp.