Moral Epistemology and Social Constructivism’s Ethical Turn

Moral Epistemology and Social Constructivism’s Ethical Turn Jonathan Havercroft University of Southampton 21 January, 2016 In recent years, some social constructivists have made a turn towards ethical theorizing. The most prominent work in this field is the collection of essays and journal articles that has grown out of Richard Price’s Moral Limit and Possibility in World Politics. There have been several other important contributions to this turn. Mark Hoffman has explored the question of whether a constructivist ethics is even possible, and has concluded that it is, but that because “of the lack of substantive moral beliefs embedded in constructivist theory, we can only derive constructivist ethics from its assumptions about the nature of social processes.” Antje Wiener has developed a theory of contestation partially to answer questions about the normativity of norms – i.e. how do we know if international norms are good? In addition to these theoretical developments, constructivists have begun to incorporate clearer normative dimensions to their empirical research. Barnett and Wiess’s recent edited volume uses empirical studies of humanitarian agencies in order to grapple with the many ethical dilemmas that these agencies confront in daily activities. Neta Crawford’s most recent volume draws upon a very rich empirical analysis of the U.S.’s post-9/11 conflicts to answer the normative question of who is morally responsible for the deaths of civilians in these conflicts. In part, the impetus for this ethical turn grows out of these scholars’ desires to square their empirical research with their normative commitments. For example, Price, in citing his own work and that of his constructivist peers, writes that “while it has thus opened up convincing space for taking seriously the role of moral change in the study and practice of international relations, this literature has not offered its own normative or prescriptive defences of particular changes as good – such positions are often not articulated let alone rigorously defended.” In a similar vein, Christian Reus-Smit writes “that International Relations scholarship ought to confront, in a systematic rather than unreflective fashion, the praxeological question, ‘how should we act?’” . And Kathryn Sikkink in reflecting upon her own work observes: “Some of the most important intellectual disagreements I have had with very diverse individuals in my lifetime were so deeply felt exactly because they were simultaneously ethical and empirical. But I was often unable even to identify this and explain cogently how I combined ethical judgment with the results of my empirical research to arrive at strongly held positions.” One could say, following Marx, that up until this moment social constructivists had only interpreted the world, however now they want to change it. The other impetus for this development has been from the external critique by scholars such as Mervyn Frost who has argued: “For the task of IR theory according to constitutive theorists is to reveal our global international social order to be a human construct within which are embedded certain values chosen by us and to show how this construct benefits some and oppresses others. . . . However, in practice, constitutive theorists have done very little of this kind of theorizing. They do not for the most part tackle the question ‘What would it be ethical to do in the circumstances’” . Neta Crawford has made a similar point, arguing that “constructivists have little to say about what to do.” So, both within and without constructivist circles, there has been a demand for research in international norms to move from the description and explanation of the dynamics of these norms towards ethical prescriptions about what international actors should do. In many ways, the constructivists have answered this charge forcefully and decisively. The Price volume draws upon rich empirical analysis to make a set of normative interventions in pressing political issues ranging from the rights of indigenous peoples to humanitarian intervention to the rights of migrants. While this volume and other research in normative constructivism has been well received by international ethicists, there has also been some concern that normative constructivists lack clear evaluative criteria with which to determine if a norm is good or bad. Toni Erskine, in her assessment of the Price volume, has argues: “In other words, this constructivism is able to say that norms matter in a way that is extremely valuable; however, it is less equipped to say why certain norms are more or less just or ethical than others, or why their emergence or transformation constitutes moral progress or regress.” Erskine echoes similar critiques made about constructivism prior to the Price volume by Hoffman that “constructivism lacks a fundamental moral core” and Rengger who argues that Price’s version of constructivism is not “clear and explicit about its normative commitments and the reasons it has for them.” What all three of these critics are saying is that constructivism lacks a clear moral epistemology. Moral epistemology refers to the meta-ethical question: how does one know if an action is right or wrong? So, the general critique of the constructivists’ turn to ethics is that they are lacking a clear set of guidelines by which to determine whether or not a principle is moral. From the empirical social scientist’s perspective meta-ethics can seem like arcane debates about angels dancing on the heads of pins. But for those working in the world of ethics, to accuse a scholar of lacking a moral epistemology is as serious a charge as accusing an empirical researcher of selecting a case on the dependent variable, conflating causation with correlation, confusing the direction of the causal arrow or lacking a clear hypothesis. Without an account of how one knows whether an action is right or wrong, from the ethicist’s perspective, a normative claim is simply an assertion, lacking in any scholarly merit. Price’s response to Rengger and Erskine’s concern about the lack of a clear moral yardstick is two-fold. First, in responding directly to Erskine, Price argues that the primary value of constructivism to normative IR “is less in extended normative arguments as such grounding general moral commitments . . . and more in offering perspective concerning possibilities for their realization and structural costs, and leverage when brutal trade-offs are confronted between moral commitments that cannot simultaneously be realized.” So, the primary contribution of constructivist research to IR scholarship is its use of empirical discoveries to clarify what is at stake in various normative dilemmas and thereby clarifying the best possible courses of action. However, Price is not willing to surrender the entire enterprise of defining and defending the evaluative criteria used in normative judgments to international ethicists such as Erskine. After responding to Erskine’s charge that a constructivist ethic is inherently conservative because it is dependent upon existing moral norms, Price concludes that “[f]urther development of just how the interpretive analytics of the type [Erskine] seems to champion within normative theory would deal with this would seem to offer one avenue for further conversation.” So, while Price does not develop a clear moral yardstick, he does concede that this is one gap in the current move towards a normative constructivist research agenda that should (or at the very least could) be filled in. The purpose of this article is to suggest that Stanley Cavell’s work on moral argumentation provides constructivist scholars with theoretical tools to develop a moral epistemology that is consistent with their other theoretical commitments. Cavell is a contemporary American philosopher who has published widely in the fields of epistemology, ethics, political philosophy, aesthetics and philosophy of language. He is not someone who is usually associated with the discipline of international relations, but because his work draws heavily upon the philosophy of Wittgenstein his work is attentive to the role of social norms and identities in normative argumentation. In terms of the normative constructivist enterprise, there are four contributions that Cavell’s work makes to the debate between Price and his interlocutors. First, Cavell clarifies why there is an impasse between Price and Erskine on the issue of moral epistemology – because of social scientist’s confusion over the is/ought distinction. Second, Cavell reconceptualizes ethics not as the search for universal criteria by which we can judge actions as right or wrong, but as a process of moral argumentation in which participants work out their moral commitments and their relationships to others. As such, Cavellian ethics is attentive to the constitutive powers of social norms and the ways in which identities shape beliefs and behaviors. Third, Cavell’s work on ethics focuses upon the grammatical structure of moral argumentation. This grammar of moral argumentation provides a useful heuristic for scholars to analyze and assess moral arguments. Fourth, his attentiveness to the norms of conversation and the means by which participants in a moral argument can assess competing claims provides normative constructivists with the “moral yardstick” that they and their critics are seeking. This article will proceed in three parts. In part one, I will explore why constructivists have explicitly rejected the available approaches in normative IR. I will argue that foundationalist approaches to ethics are not a good fit with social constructivism, and that while anit-foundationalist approaches are a better ontological fit, constructivists remain skeptical of the ability of these approaches to provide a means of evaluating norms. In part two I argue that because of constructivists’ claims that moral values are socially constructed it stands in a unique position to overcome the false dichotomy in the social sciences between facts and values. But in order to do this, constructivists must be clear about how it is possible to derive moral claims from factual descriptions. I make this argument by reviewing Searle’s and Cavell’s critiques of the is/ought problem and suggesting that their work offers some useful clarifications about the nature of norms for social constructivists. Finally, in the third part I draw upon Cavell’s analysis of morality to argue that his understanding of the grammar of moral argumentation is useful for constructivists because it provides an analysis of how to construct moral arguments that are likely to succeed. I. The Limits of Existing Normative Frameworks for Constructivism One of the criticisms leveled by all three participants in the International Theory roundtable on the Price volume is that the move towards a normative constructivism was blurring the boundaries between normative and empirical International Relations scholarship. It is noteworthy that there were frequent calls for a clearer division of labour, and the demand that boundaries not be blurred. Price has rightly resisted this call, because constructivism does have important contributions to make to both empirical and normative scholarship. I will address the specific issue of drawing a boundary between normative and empirical scholarship in the next section. But a related concern to this line of critique raised by both Erskine and Rengger is that Price and other constructivists should draw upon existing normative approaches. The problem with this critique is that some of the approaches that are suggested as possible candidates – Habermasian discourse ethics , Rawlsian political constructivism It is important to bear in mind that social constructivism in international relations is distinct from the political constructivism in the philosophy of Kant in (O'Neill 1989, 1996). Political constructivism is a philosophical attempt to offer justifications for moral or political principles in light of a plurality of authoritative sources for these justifications. It is an attempt to construct a ‘thin universalism’ that retains traditional universalism’s commitment to objective and authoritative justification of principles without being foundationalist. Instead, moral and political constructivists attempt to develop procedures for constructing principles of justice that are universal and objective. See 06:07:13 While there are some commonalities between political constructivism and social constructivism (see ), the central difference between the two approaches is political constructivism’s commitment to developing universal and objective principles of justice. Social constructivists are not committed to uncovering such principles, and have expressed skepticism about whether or not this search for universal principles is useful in international politics. , and Nussbaum and Sen’s capabilities approach, 06:07:13– do not fit with the ontological assumptions of social constructivists. This is because most of these theories have foundationalist ontologies, where as social constructivism is committed to a variety of anti-foundationalism. On the other hand constructivists have expressed reservations about more anti-foundationalist approaches to normative IR such as constitutive theory, post-structuralism and pragmatism, because these approached lack clear criteria for evaluating norms. It is difficult to reconcile constructivist ontological commitments with ethical theories that have strong foundations such as Rawls’ political liberalism, Habermas’ discourse ethics, the capabilities approach, and neo-Kantian approaches to international justice. Reus-Smit argues that neo-Kantian approaches to international ethics – such as those developed by Rawls and O’Neill – engage in “a form of philosophical inquiry characterized, first and foremost, by logical reasoning from first principles.” This approach does not mix well with constructivist scholarship because “‘Facts’ are chosen selectively to undergird the preferred line of moral reasoning, and often voluminous amounts of relevant empirical research and theory are ignored.” The consequence of this is that many approaches to international ethics make moral judgments about general phenomena– such as poverty, war, human rights, etc. – without seeing how particular contexts may make particular courses of action impossible or perhaps even undesirable. Reus-Smit is primarily directing his critique at and Both of whom deduct their international ethics from first principles. He is silent on the question of whether or not Walzer’s casuistic approach (2006) is sufficiently context sensitive to overcome his charge that international ethicists in general make the facts fit their ethical principles. Sikkink is critical of the capabilities approach pioneered by Sen and Nussbaum 06:07:13 for similar reasons. She critiques Sen and Nussbaum for believing “that they must start from scratch in inventing their central list of rights and capabilities.” There is already a dense existing set of international human rights principles, developed over 50 years through the deliberations of over 150 countries and thousands of human rights NGOs. The existing consensus on human rights “provide a more legitimate source of general principles than any I or any other individual or group of researchers could invent.” From this perspective, normative theorizing on international politics, by ignoring empirical considerations, appears to be doing little more than re-inventing the wheel. It should be noted that not all neo-Kantians grapple with questions of international norms in this way. Allen Buchanan has critiqued political philosophers of human rights on similar grounds as Sikkink and Reus-Smit. He argues that the focus of political philosophers on providing a justification for human rights tends to neglect the ways in which existing human rights regimes have been developed through processes of practical reasoning. And this neglect of already existing deliberative processes often leaves political philosophers own justifications of human rights incompatible with existing human rights regimes. His work calls upon political philosophers to engage more directly with existing human rights regimes in developing their justifications for human rights. See In addition to Reus-Smit’s and Sikkink’s charges that the existing modes of normative theorizing are insufficiently empirical, constructivists have complained that normative theorizing is often impractical. For instance, Price has been critical of attempts by constructivists to draw on Habermasian discourse ethics because ideal-speech situations “would seem to be empirically rare if not indeed theoretically impossible for some versions of constructivism.” None of these approaches to normative theorizing sufficiently incorporate into their approaches analyses of power imbalances and the constraining effects of existing social structures that are the mainstay of constructivist theory. The lack of normative theorizing by constructivists is complemented by the failure of normative theorists to adequately take into account the empirical conditions that govern human action in world politics. Consequently, normatively inclined international relations scholars and empirically inclined social constructivists talk past each other rather than to each other. Price, Reus-Smit and Sikkink reject the above approaches to international ethics partially because of their foundationalism. There is, however, a second set of approaches to normative IR that are a better fit for a normative constructivism such as constitutive theory, pragmatism and post-structuralism. These approaches do not cleave to first principles or universalizable moral positions in order to make judgments about human conduct. Constitutive theorists such as Frost use historical analysis of settled norms in the practices of states in order to examine the moral responsibilities of states. Cochran has drawn upon the work of pragmatists such as Dewey and Rorty to develop an anti-foundational approach to normative IR that uses Dewey’s scientific method and Rorty’s concepts of solidarity and irony to develop an ethically rich, yet normatively informed method for the English School’s middle ground ethics. Post-structuralist IR scholars such as Connolly and Campbell have drawn upon the techniques of normative technique developed by Derrida and Foucault to critique existing social norms and practices in world politics. While all three of these approaches to normative IR are closer to the ontological commitments of social constructivism, it is notable that when constructivists such as Price, Reus-Smit, and Sikkink consider these approaches they find them wanting in different ways. In the case of constitutive theory, Price argues that in terms of the reconstruction of the ethical criteria, the ethical yardstick by which we judge the validity of ethical claims evolves as the norms of the global practice shift. So in this case the primary means for determining the validity of an action is its consistency to the ethical norms that the participants in a give practice already subscribe to. From a thick constructivist perspective, this shifting normative horizon may not be a problem. However, Price expresses concern that Frost’s approach does not (for example) answer the question: “what should be the criteria in terms of which states recognize one another”. For a thinner constructivist such a Price, constitutive theory does not offer clear criteria by which to evaluate norms. Pragmatist normative IR as pioneered by Cochran argues that one way of avoiding the pitfalls of moral relativism is to focus less on “moral principles in all their universality” and more on “the empirical facts as they are”. She suggests drawing upon Dewey’s method of inquiry to evaluate international practices. This method revolves around asking two key questions when confronted with an ethical dilemma: “what social values are in conflict and what possibility exists for an integrative value or principle to emerge that would be capable of generating shared problem-solving with improved effects?” While the contributors to the Price volume do not draw upon Cochran’s work or pragmatic IR, it is worth noting that Reus-Smit simultaneously describes this approach as “excellent” and as narrowly conceiving normative IR as “deducting from first principles”. This strikes me as a mischaracterization of Cochran’s work and pragmatism in general, as the whole point of this normative enterprise is to begin ethical inquiry from the facts and norms as they exist. However, for reasons that remain unclear a dialogue between constructivism and pragmatic IR has yet to take place. Finally, while constructivist and poststructuralist approaches to IR both emerged out of a critique of the hegemony behaviouralist methods in IR, there are significant tensions between these two approaches, especially when it comes to normative theorizing. Poststructuralist scholars have taken IR scholars to task for assuming that the proliferation of liberal norms represents moral progress often arguing that there are other forms of domination hidden in the constructivist research agenda. Consider, for instance Conversely, constructivists have argued that the poststructuralist focus on critique rarely leads to a productive discussion of what should be done. Personally, I think that all three of these approaches have some important resources to offer a normative constructivism. And they constitute a better fit for constructivists than the more foundationalist approaches discussed above. However, when constructivists have considered constitutive theory, pragmatism, and post-structuralism, there has been some hesitancy to adopt these approaches. More foundational approaches to ethics seem too abstract and divorced from the empirical realities that social scientifically oriented constructivists grapple with in their research; the more anti-foundational approaches do not provide the constructivists with a clear enough means for evaluating specific norms. Whether or not constructivists such as Price, Reus-Smit, and Sikkink in reaching this conclusion have sufficiently understood how these theories make normative evaluations in order to justify their not adopting these approaches is a question I do not want to tied up in here. What I will argue in the rest of this paper is that Cavell constitutes a theoretical resource for normative constructivism that can help social constructivists respond to the moral epistemology question. Cavell is a fellow traveller of poststructuralists, constitutive theorists and pragmatists. He shares the quasi/anti-foundational orientation of these other normative approaches. Indeed his earliest published work was and engagement with pragmatism. One crucial difference between Cavell and these other approaches, however, is that whereas a pragmatist such as Rorty argues that the search for criteria for making moral evaluations is a pointless metaphysical exercise , Cavell argues that the search for shared criteria by which we make shared judgments is part of what makes us human. There are significant tensions between Cavell and Rorty. See And While Cavell shares the anti-foundational critique that the search for transcendental criteria to evaluate normative claims will end in failure, he does believe that there is a kind of normativity built into our language and our forms of life that do provide us with the means for assessing our conduct. In the remainder of this article I will explore this dimension of Cavell’s thought, to show how it can offer constructivists a moral epistemology. II. The fact/value distinction and the impasse in constructivist ethics While one reason for constructivism’s ethical impasse is the lack of a fit between existing approaches to international ethics and social constructivists’ ontological commitments, a second, more significant obstacle is a blurring of the lines between explanatory and normative approaches to international politics. In philosophy, the division between the normative and the empirical is referred to as the “is/ought” distinction, whereas in the social sciences it is often referred to as the “fact/value” distinction. Putnam’s book provides a useful history of the debate over the fact/value dichotomy in philosophy and develops a pragmatic argument that the fact/value dichotomy is not sustainable because values and normativity are embedded in all human experience. While the fact/value distinction and the is/ought distinction are not identical, their shared function is to remind scholars to not confuse discussions of fact with normative assessments of values. Stephan Mulhall points out that to elide the is/ought problem with the fact value distinction assumes “that every evaluative judgment is a judgment of obligation or duty, and that every descriptive judgment is essentially non-evaluative.” While this is an important point to keep in mind, because this article is discussing international ethics, the evaluative statements under consideration here are judgments of obligation or duty, and as such it is possible to use the two terms interchangeably in this context. See In the case of the is/ought distinction, the basic point is that it is impossible to derive an “ought” – an evaluative statement – from an “is” – a factual description. The upshot of these two distinctions is that ethical considerations are separated from more “scientific” approaches that attempt to explain how some part of the world functions. One possible response to the constructivist ethics “paradox” is that there is no paradox: constructivists study the ethical positions of other actors, but in studying the values of others they need not evaluate those positions. In a review article on constructivist scholarship Finnemore and Sikkink argue that “modern constructivists” maintain a strict social science epistemology (2001: 395). Scholars committed to such a strict social science epistemology might not be concerned with developing a normative research paradigm. But it is worth noting that both authors contributed chapters to the Price volume on constructivism and ethics (Sikkink 2008; Finnemore 2008). So there is a desire for a normative research program in constructivism, and that one of the acknowledged obstacles to such a program is the rigid fact/value distinction that informs modern social scientific epistemology. In this sense, constructivists are not much different from those who study public opinion in domestic politics. They might ask how specific opinions shape behaviour, they might consider how people come to hold the opinions that they do, and they might even be interested in cataloguing the specific opinions that individuals hold, but they do not concern themselves with evaluating specific opinions. From this perspective constructivism is firmly on the “is” side of the “is/ought” divide. I, however, think that this move forecloses the possibility of a constructivist ethics too quickly, for two reasons. First, the fact that many of the constructivists discussed above think that there is a normative dimension to their work and express befuddlement or dismay at the fact that their work has not been able to support a normative agenda reveals at the very least that scholars see their work as having a normative dimension. So, there is a desire for the descriptive “is” of constructivist research to somehow translate into a set of prescriptive “oughts.” The question these scholars trip over is: how do you move from the factual to the prescriptive? Second, it is not clear that the divide between “is” and “ought” is as rigid as scholars make it out to be. Critical IR scholars have always been skeptical about the claim that social science is value neutral. And in philosophy, scholars have long questioned the rigid divide between “is” and “ought.” Philosophers trace the “is-ought” problem to David Hume’s Treatise of Human Nature, where Hume argues that it is impossible to derive an “ought” (i.e. a normative prescription) from an “is” (i.e. a factual description). His distinction claims that knowledge of how the world is cannot lead to conclusions about how the world ought to be. In the 20th century this concern was revived in the meta-ethical writings of G. E. Moore and R. M. Hare. Moore chided moral philosophers for committing what he termed the ‘naturalistic fallacy’ – i.e. denoting naturalistic qualities for non-naturalistic terms such as “the good”. R. M. Hare insisted that prescriptive statements, such as “you ought to do X,” could not be derived from descriptive statements. The “is/ought” thesis created a clear demarcation between moral philosophy and empirical observation. As the “is/ought” thesis was gaining influence in moral philosophy a parallel movement was gaining influence in the social sciences. Attempts to study society, culture, and politics “scientifically” led scholars to demand that social research be conducted “objectively.” Early social scientists such as Max Weber did this by clearly demarcating the study of social facts from the study of social values. One of the effects of Weber’s founding of modern social science was that the study of social facts was to be clearly demarcated from questions of social values. See for an interpretation of Weber’s methodological writings that argues that Weber’s founding of social science was anything but value-free. This demarcation between facts and values also presupposes that it is impossible to move from the scientific study of social facts to their normative evaluation. It is this divide between fact and value in the social sciences that is the stumbling block over which attempts to develop constructivist ethics continually trip. While constructivist scholars study norms, their work does not assess the content of those norms from an ethical perspective. Constructivist scholars may choose to study disarmament because they are horrified by war, or conventions against torture because of strong commitments to human rights, but these personal beliefs only explain the reason why they feel compelled to study one subject in a universe of thousands of possible topics. Implicit in the fact/value distinction is the belief that social scientific work need not generate any normative prescriptions in and of itself. However, two things point against maintaining such a rigid boundary between normative and empirical questions in constructivist research. First, constructivist scholars are themselves expressing dissatisfaction with this division. They believe that their work should generate prescriptions and that they should be able to differentiate between good norms and bad norms. Second, in philosophy the rigidity of the distinction between “is” and “ought” has been challenged consistently over the past 50 years. By exploring some of the more prominent critiques of the is/ought distinction, we can see how IR constructivists can move beyond the normative impasse that they currently finds themselves in. The criticism of the is/ought distinction in moral philosophy revolves around two issues. First, John Searle contended that the fundamental claim of the is/ought distinction – that one cannot derive an “ought” from an “is” – is simply untrue. Searle offered the example of a promise as a case where an evaluative statement “You ought to keep promises” could be derived from the descriptive statement “You promised to do X.” Searle argued that in the case of “social facts” – i.e. facts that are socially constructed by institutions – it is possible to derive moral obligations from factual descriptions. To support this claim Searle provided the example of a promise; the fact that one has made a promise creates an obligation to uphold that promise. Second, Stanley Cavell argued that the ontology presupposed by the is/ought distinction – namely that reality is a realm of pure facts – ignores the way in which our social world is not simply a collection of brute facts, but rather a shared world of moral meanings. This second point dovetails nicely with the social constructivist ontological assumption that the social world is constructed through shared interpretations of social meanings and values. While Searle’s argument that it is possible to derive an “ought” from an “is” demonstrates that it is possible to develop normative conclusions from empirical descriptions, Cavell’s approach is more useful for our purposes here because he turns to the structure of moral argumentation to show how specific moral arguments must be constructed. Cavell’s critique of the is/ought problem drew upon the ordinary language philosophy pioneered by J. L. Austin. Ordinary language philosophy developed in response to the dominance of logical positivism in Anglo-American philosophy in the first half of the 20th century. Broadly speaking, logical positivists believed that all knowledge was based on logical inference from analytic statements based upon observable facts. Austin argued that while the positivists had been successful at analyzing this type of logical statements, they had missed a whole category of human statements that he termed “performatives” – i.e. statements in which a person does an action. One of his famous examples is of a person saying, “I do” at a wedding. Unlike analytic statements, whose truth-value is analyzed through logic and observation, performative utterances are successful or unsuccessful depending upon the context in which they are performed, and whether or not the performance of the utterance follows the existing protocols of that social institution. Cavell’s critique of the is/ought distinction examines the performative utterances used in ordinary human practices to demonstrate how we derive obligations from facts all the time. For Cavell, there is no gap between factual statements and evaluative statements. The difference between fact and value lies in how individuals perform their utterances – in whether they make statements or evaluations. What is striking to Cavell is that ought statements add nothing to the content of the factual statements that support them. If this is the case, then why use an “ought” statement? Cavell contends that what differs between a factual and an evaluative statement is not the content but the mode of presentation. And that choosing to use an “ought” instead of an “is” statement implies that the person to whom the “ought” statement is directed has a choice between alternative courses of action. In pointing out this feature of moral discourse Cavell is making a broader claim about the entire enterprise of morality. What holds the is/ought problem in place in analytic philosophy is the assumption that the purpose of morality is to deduct general principles about what people ought to do. The underlying assumption behind this picture of morality is that for morality to be rational participants in moral discourse must be able to reach agreement on moral principles. Yet Cavell argues that the rationality of moral argumentation does not lie in its ability to generate consensus, but in a shared commitment by its practitioners to certain modes of argumentation. The effect of this distinction is that Cavell protects moral discourse from the argument that the inability to generate consensus about moral conclusions means that rationality in morality is impossible. It also means that it is possible for participants in a moral debate to not reach a conclusion, to agree to disagree, without such failure to reach agreement undermining the entire moral enterprise. As Cavell points out, even if the participants in a moral dialogue do not come to an agreement, they have agreed upon shared modes of argument in order to engage in moral argumentation in the first place. And these shared modes of argumentation place constraints on what may be pertinent and how participants in a moral dialogue might respond to one another. In elaborating upon how moral argumentation is rational, Cavell argues that only certain types of statements are moral. What makes a statement moral is how it is presented and how the interlocutor responds to the initial challenge. In his analysis of moral argumentation in the Claim of Reason, Cavell identifies three grammatical features When I use the term “grammar” in this article, I am using it in the sense that Wittgenstein does in the Philosophical Investigations. For Wittgenstein, grammar refers to the rules that govern what expressions make sense (and which ones do not) in a given context. These rules are embedded in the everyday practices that form the background of linguistic meaning (Wittgenstein, 1953: §371, §373). of arguments: the mode of presentation, the reasons used to support a moral claim, and the elaboratives used to justify, excuse, or defend a certain action. Each of these features of moral argumentation contains non-arbitrary rules about how an interlocutor must respond if he or she is to be morally competent. In unpacking this grammar of moral discourse, Cavell hopes to respond to those who think that in the absence of principles that assure consensus about moral actions, the practice of morality is arbitrary, perhaps even relativistic or nihilistic. The first feature of Cavell’s grammar of moral argumentation is the mode of presentation. Most approaches to ethics attempt to develop a set of general principles or rules and then assess particular cases against these rules. What these approaches neglect are the ways in which particular cases are presented. When a person uses an “ought” statement, she is attempting to persuade her interlocutor. And this act of persuasion is what makes the statement a moral one as opposed to an analytical one. The mode of presentation simultaneously offers a moral evaluation, and it establishes the moral position of the actor. In an early work on social constructivism Kratochwil discusses the role that “must” and “ought” play in the development of international norms. See (Kratochwil, 1988). If the actor must keep an obligation and he fails to do so, he will have also failed morally. The second feature of Cavell’s grammar of moral argumentation is the “basis of support.” Cavell argues that built into the grammar of moral argumentation are morally relevant reasons for challenging a person’s actions. He contends that when a person confronts another person’s actions, one questions the actions by questioning how the actor’s actions fit with that actor’s cares and commitments. Cares and commitments are the two different modes of support that a person in a moral argument may offer in defense of his action. Commitments refer to pre-existing obligations that a person has to pursue a given course of action. For example, in the case of promises, a person who has said “I promise to do X,” has made a commitment to upholding that promise. Conversely, when Cavell discusses an agent’s “cares” he refers to those things that an agent either does, or ought to be concerned about. These cares are embedded in the identity of the moral agent. Agents can care about any number of things from the well being of others, to social harmony, to environmental sustainability. The third feature of Cavell’s grammar of moral argumentation is the elaborative. An underappreciated aspect of moral argumentation is the way in which individuals defend actions that have been challenged by others. How one defends one’s failure to uphold a previously stated care or commitment is morally significant. Cavell calls the defenses – i.e. the excuses, explanations, justifications, etc. – that we use whenever our conduct is challenged “elaboratives.” In moral argumentation, how one uses an elaborative to defend one’s conduct is critical. Elaboratives “serve to acknowledge your awareness of what you have done. Without the expression of that awareness, even the extreme defense is incompetent.” Others may not accept our defenses or excuses, but to not offer them would mean that we do not care about the opinions of those who confront us over our actions. When these three elements of moral argumentation – modes of presentation, modes of support, and elaboratives -- are put together, Cavell presents us with a grammar of moral arguments that shows us what we do (and by extension what we cannot do) when we engage in moral arguments. Moral arguments begin when one person confronts another over his conduct. The mode of presentation shapes the nature of the encounter. The basis of support constructs the identity of the moral agent with respect to the moral act. Finally, elaboratives point to the ways in which one can counter the charges of one’s interlocutors. In the next section I will consider how these three features of moral argumentation fit with the ontological assumptions of social constructivism. III. How to evaluate norms: Cavell’s moral methods To return to the puzzle of moral epistemology with which this article opened: how does Cavell help constructivists answer Toni Erskine’s question about a moral yardstick? Modern moral philosophy has tended to seek out universal principles with which to judge particular actions. The most obvious examples of this are deontological and utilitarian approaches. Deonotologists ground their ethical claims in the categorical imperative, where as utilitarians judge actions by whether or not they maximize utility. Yet both approaches have also invited refutation via numerous thought experiments. For example, critics of utilitarianism have provided examples where punishing an innocent person is permissible if it maximizes utility. For a history of this criticism of utilitarianism see Conversely, critics of Kant often point to how his commitment to the duty not to lie can lead to morally dubious situations wherein one is obliged to tell a potential murder the whereabout of the murderer’s intended target. According to Kant, one is not permitted to lie even if it saves a life. Cavell’s own work on moral philosophy observes that this tendency to seek out universal rules upon which all can agree and which can be applied in all contexts leaves both utilitarianism and deonotological approaches to ethics open to these kinds of refutations. Furthermore, because such theories aim at comprehensively judging all actions, any possible refutation of a universal principle makes it seem as if the whole approach to ethics, and perhaps even in the most extreme cases the ethical enterprise itself, open to moral skepticism. This is because traditional approaches to ethics seek out rational foundations that can provide a basis for which all rational participants to an argument can come to an agreement about a moral judgment over an action. When rational agreement does not occur, then either one of the participants must not have reasoned correctly, or there is something wrong with the foundation. The philosophical challenges to utilitarianism and deontological approaches (discussed above) are instances of philosophers demonstrating a flaw in the foundation of the different approaches. Exposing these types of flaws in foundational approaches to ethics leaves ethics as a whole open to moral skepticism. If one assumes that ethics must rest upon some foundation, some comprehensive doctrine, and these foundations are shown to be incapable of covering all cases and generating universal agreement, then it is possible to conclude that no rational agreement about moral principles is possible. Cavell argues that the methods used in foundational approaches to moral judgment are what set the enterprise up for failure. When moral philosophers conceive of the task as identifying formal structures for developing first principles and then aligning an individual’s thoughts and actions with these principles, she misunderstands what the activity of morality is. Consider for example Cavell’s critique of an early essay by Rawls, “Two Concepts of Rules”. In this essay, Rawls set out to defend utilitarianism from a standard critique: that there are instances where either breaking a promise or punishing an innocent person would be acceptable if the action maximized utility in some way. Rawls responds to these challenges to utilitarianism by arguing that both promising and punishing are social practices. Rawls defines practices as “any form of activity specified by a system of rules which defines offices, roles, moves, penalties, defenses and so on, and which gives the activity its structure”. Rawls argues that these types of practices can have a utilitarian justification, while using a system of rules that would not permit utilitarian justifications for doing things that violate our moral intuitions such as breaking promises and punishing innocents. Rawls argues that there are two distinct concepts of rules at work when utilitarians examine practices. The first conception is what Rawls calls the summary view, as it sees rules as helping an individual decide in each particular instance how to apply the utility principle to a particular action. One makes these kinds of decisions by summarizing what past experience has shown is the likely consequence of a particular action, and examining the expected utility of the predicted outcome. On the other hand the practice conception of rules imagines that specific social practices such as promising and punishment can have utilitarian justifications, while the practices themselves consist of rules that would not permit utilitarian justifications for acting against the practice. So we can imagine that the practice of promising is justified in a society because a society in which people keep their promises is better than one in which promises are regularly broken, however specific instances in which breaking a promise cannot be justified because it goes against the rules of the practice of promising. In drawing this distinction Rawls moves utilitarianism away from the pitfall of trying to judge every individual action according to the utility principle, and instead grounds morality in social practices. While Cavell is largely sympathetic with the shift that Rawls undertakes, he is still critical of the attempt to ground utilitarianism in rules. Cavell argues that activities such as punishing and promising do not operate according to rules in the way that Rawls claims they do. In order to defend rule utilitarianism in the “Two Concepts of Rules” essay, Rawls draws an analogy between social practices and games such as baseball. Cavell argues that this analogy leads Rawls to make a crucial mistake. Cavell argues: “No rule or principle could function in a moral context the way regulatory or defining rules function in games. It is as essential to the form of life called morality that rules so conceived be absent as it is essential to the form of life that we call playing a game that they be present.” To underscore how different playing a game is from morality, Cavell considers the case of how an argument that might be acceptable in excusing a broken promise, would not be acceptable in playing a game. I might try to excuse a person for missing an engagement by claiming something else came up, but I cannot ask for another pitch if I swing (analogous to a promise) and miss, but the sun was in my eyes. In the first case making an excuse is part of the form of life of promising (n.b. it does not follow that any excuse I make for breaking a promise will be accepted); but in the second case asking for another pitch when the sun was in my eyes would call into question my very competence as a ballplayer. The crucial difference is that in games the rules settle in advance what it is acceptable or unacceptable for the players of a game to do. Morality is different. As Cavell argues, “That moral conduct cannot be practiced in that way, that you cannot become a moral champion in that way, and that no one can settle a moral conflict in the way that umpires settle conflicts, is essential to the form of life we call morality.” Cavell’s point is that the “form of life” of our moral actions such as promising and punishment, not the institutional rules, provide the foundation for moral judgments. Cavell takes the term form of life from Wittgenstein. Wittgenstein introduces the term form of life early in the Philosophical Investigations when he writes: “The word ‘language-game’ is used here to emphasize the fact that the speaking of a language is part of an activity, a form of life.” Wittgenstein introduces this term to underscore that our language is always interwoven with our cultural and social practices. In order to say something like “I promise” competently, we must have a tacit understanding of what the activity of promising entails, and the consequences (including the likely social sanctions) of breaking a promise. The grammar of moral argumentation discussed in the previous section is what constitutes the form of life of moral argumentation. For a person to act morally he or she must have cares and commitments and use elaboratives in order to justify, acknowledge, excuse and apologize for his or her actions. For Cavell, knowledge in morality (i.e. moral epistemology) consists precisely in being able to elaborate upon your actions. “To know what you are doing is to be able to elaborate the action: say why you are doing it, if that is competently asked; or excuse or justify it if that becomes necessary.” This type of knowledge cannot be gained from identifying rules and using these rules to judge the morality of particular actions. Cavell argues that morality must be reconceived as a social activity, as one in which we engage in moral argumentation to work out who we are. This understanding of morality contradicts some core assumptions of many traditions of ethical theorizing. Cavell argues that: “what makes moral argument rational is not the assumption that there is in every situation one thing which ought to be done and that this may be known, nor the assumption that we can always come to agreement about what ought to be done on the basis of rational methods.” Cavell argues that both of these assumptions – that moral theory can tell us the one right course of action, and that it can create agreement in society about what ought to be done – is precisely what sets ethical theorizing up for disappointment. In the everyday ethical dilemmas that we face, whether in our interpersonal relationships, or in the difficult moral tradeoffs that confront heads of state, there is no single right course of action, and no rational method that will attain the assent of all who are affected by an ethical decision. While insisting that these are two firm limits on ethical theorizing, Cavell rejects the tendency to assume if morality cannot point to the single correct course of action or provide reasons to which any rational actor could assent, then morality is a pointless enterprise limited to justifying the self-interests of those who moralise. Instead, Cavell insists that the rationality of moral argumentation “lies in following the methods which lead to a knowledge of our own position, of where we stand; in short to a knowledge and definition of ourseleves.” For Cavell, morality is about working out who “we” are. There is significant overlap here between Cavell and Richard Rorty who also insists that the source of moral authority comes from within the community, and who calls for “sentimental education” in order to expand our “we” reference. For an exploration of how expanding our “we”-reference can contribute to normative IR see It is an inter-subjective enterprise, and its concern is with the first person plural. My conduct only comes into question when my actions affect others. Norms exist to guide and judge the conduct of individuals in terms of their memberships in groups. These norms do not have some transcendental origin. Instead the group develops them over time as disputes arise over the behaviour of individuals. The reference point against which the rightness or the wrongness of any particular action is judged is the identity of the group. For Cavell, moral argumentation is co-constitutive the identity of the group and the criteria by which that group judges the conduct of its members. But how exactly does one go about finding the criteria by which a group judges its members? What are the methods that lead to a knowledge of ourselves? And, most crucially for the central concern of this article, how do these methods help constructivist scholars adjudicate between norms? Cavell himself does not explicitly state what this methodology is. But by pulling together the different threads of his writing on morality in The Claim of Reason and drawing recent scholarship on moral reasoning that draws upon Cavell, I think that there are three distinct sets of methodological tools that Cavell provides to help constructivists in their ethical turn. The first are concepts of grammar of moral argumentation, outlined in the previous section. The second are the norms of conversation that provide a procedural way of validating how norms are generated. The third is a moral yardstick that validates moral claims against the identity of the group. Built into Cavell’s analysis of the three core features of moral argumentation – the mode of presentation, the supporting reasons, and elaboratives – is a method that enables individuals to judge the validity of particular moral claims. The grammar lays out what kinds of moral arguments are plausible and implausible. Understanding how arguments work enables constructivist scholarship to proceed along two lines. First constructivists can deepen both their empirical and normative evaluation of norms through a discourse analysis that pays close attention to the grammar of moral arguments made by representatives of states and non-state actors. While much empirical work already implicitly examines how states comply with and contest international norms on the basis of a state’s implicit cares and commitments, elaboratives are an under explored dimension of moral argumentation. Cavell’s call to pay attention to the different elaboratives states use – e.g. the excuses, apologies, justifications and acknowledgments states give when they violate norms – can reveal the different moral dilemmas states confront and whether other members of the international community find the elaborations acceptable or believable. Second attentiveness to the grammar of moral argumentation can help normatively oriented scholars construct arguments that are more likely to be persuasive to international actors. In addition to Cavell’s analysis of the grammar of particular moral claims, his work also rests upon an understanding of justice being grounded in conversation. Conversations operate according to norms, and only moral principles that develop out of conversations that abide by these norms of conversation can be considered valid. While Cavell’s approach has significant differences with Habermas’ discourse ethics and Rawls’ political liberalism, like them his approach to validating claims is largely procedural. How a norm has been developed is as important as its content, because the process of norm generation ensures that the norm is accepted by those affected by it. Anthony Laden, one of Cavell’s interpreters, describes Cavell’s procedure of moral argumentation as a social theory of reasoning. By social theory he means that reason is an ongoing activity between two or more people who “continually attune themselves to one another”. This picture is in contrast to philosophy’s traditional picture that sees reasoning as an activity that makes judgments by aligning an individual’s thoughts with formal structures of logic. The traditional picture sees reasoning as a deductive process, whereas the social picture sees reasoning as inter-subjective and dialogical. Laden argues that social reasoning is shaped by certain characteristic norms. Reasons are offered by participants in this norm-governed conversation with the hope of developing attunement between the conversation’s participants. Drawing upon psychological and linguistic analysis of conversations, Laden argues that conversations involve four constitutive norms. When any of these norms is violated, conversations break down and either end altogether, or morph into another mode of speaking such as commanding, lecturing, insulting, etc. First conversations require speaking with others rather than speaking to, at, or past others. Speaking with others means that “[p]articipants in a conversation thus must reach or rely on certain forms of agreement about the meanings of the terms of reference and be able to find routes to agreement where it is lacking”. Second conversations do not have a clear purpose. If one or both participants engages in a dialogue with some clear end in mind, such as a sales pitch or a negotiation, the dialogue morphs into a strategic interaction where the parties are attempting to get something out of each other. These kinds of interactions are not wrong, but they leave the participants in the dialogue as two individuals seeking out the satisfaction of their self-interest. Conversely, the aimlessness of conversation ends up creating attunement between the participants, which in turn leads to constructing a first person plural subject out of the participants of the conversation. According to Laden, “these first two norms shape the kind of unity that conversations tend to produce.” The other two norms concern how the participants relate to each other in conversing. The third norm is equality. Conversations require that the participants speak with rather than at others. One-way forms of address create structures of domination and command, whereas the egalitarian nature of conversation sustains the “we-ness” of the group generated by conversation. Finally that participants in a conversation must mean what they say. Laden calls this the norm of sincerity, and argues that if a conversation is to sustain the sense of “we-ness” between participants, each speaker must speak genuinely for him or herself. While the grammar of moral argumentation lays out how people must phrase a moral claim in order for others to respond to it, and the norms of conversation establish the rules by which one must carry out a moral dialogue, how do the participants in a moral conversation assess the claims that they are presented? Since conversations construct plural subjects, Laden suggests that the evaluation of specific claims is carried out “in terms of the point of the plural subject in question, and our continued membership in it.” The guiding question that Laden suggests any member of any group asks when evaluating a reason is “what problem does the association solve?” In international politics we can think of evaluating norms by reference to the community or inter-governmental organization that generates the norm. Security norms tend to emerge out of organizations and regimes whose ultimate goals are an end to violent death. As such norms and regulations that make war and death less likely will be preferred to those that make war and death more likely. If a regime or organization is established with the aim of solving a collective action problem – such as reducing green house gas emissions by states – then preference will be given to norms that solve “coordination problems in stable and mutually beneficial ways”. Finally, if the issue is one of how to respect the cultural differences of a diverse membership – consider here international regimes that are established to protect the political, economic, social and cultural rights of individuals and minorities – then the norms cannot impose values on others who reasonably disagree, and decisions must be made in accordance democratic deliberative processes. The point is that moral arguments emerge in groups, and at their core they refer back to deliberations on what kind of group its members wish it to be. Laden refers to this as the Hegelian approach to evaluating reasons, because it draws upon Hegel’s claim in The Philosophy of Right “Thus, the disposition . . . is to have self-consciousness of one’s individuality within this unity as essentially which has being in and for itself, so that one is present in it not as an independent person but as a member.” that participation in some institutions gives its members new practical identities – i.e. “new senses of who we are, how we value ourselves, and what makes our actions worth undertaking.” It is worth remembering that Hegel’s understanding of how membership in institutions shapes the identities of its members is the same process as the one that conventional constructivists use to explain how states internalize norms. Both Wendt’s discussion of how a state’s identity towards other states is affected and affects its culture of anarchy, and Risse and Sikkink’s theory of the process of norm socialization describe processes whereby the states membership in a larger community leads to a shift in how the state perceives its identity. While these are both explanatory theories of how the identity of a state shifts so that it complies with the norms of its relevant community, the normative evaluation of the norm and the community can be teased out of this process. The relevant questions to ask are: What is the end that the community was set up to serve (i.e. security, coordination of action, or fostering respect for difference)? Does the contested norm help fulfill this end? There are, crucially, ways in which a contestation of a specific norm might also call the very validity of the entire group into question. So, how a community generates a norm, is as important to that norm’s validity as whether it complies with the end of the group. Therefore, international norms that are generated through processes that are contrary to the norms of conversation – for example compelling a state to adopt a norm through coercion, or deception – are invalid. How norms are generated, and how states consent to them is just as important as the ends that a norm seeks to achieve. Finally, states are always members of numerous, often competing international communities. Each community will have different ends that will come into conflict with each other. A state, for example, may face pressure from one international institution that demands open access to domestic markets in order to promote free trade, while at the same time be a member of an international regime that seeks to protect traditional cultures by placing limits on who can produce a traditional product. In this case the ends of free commerce and protection of traditional cultures can come into conflict, and so one legitimate reason for rejecting a norm is that it might impact a states ability to fulfill its obligations to other groups or institutions of which it is a member. Conclusion: The Availability of Cavell for Social Constructivism’s Ethical Turn The reason constructivists have not developed a normative agenda is because of the is/ought problem. Constructivists inherited the rigid distinction between fact and value that has led many scholars to believe that you cannot derive an “ought” from an “is”. I have drawn upon Cavell’s argument that the difference between evaluative and descriptive statements is their mode of presentation to suggest how constructivists might be able to derive normative arguments from their empirical analyses of international politics. By this I mean that Cavell’s approach to ethics focuses on what are grammatically possible arguments. And this moral grammar can be deployed in a discourse analysis of statements made by international political actors on normatively contested issues. Furthermore, because Cavell ground his ethics in practices of social reasoning, the procedures by which norms are generated can be used to test the validity of norms. Finally, Cavell’s work provides a means of evaluating the substance of normative claims testing specific norms against the ends of the community in which the norm operates, and against potential conflicts from other communities of which a state is also a member. In their analysis of international norm dynamics Martha Finnemore and Kathryn Sikkink describe the process of political change as an “ought” becoming an “is”. The challenge that constructivists who wish to engage in normative argumentation face is how to bring about movement in the opposite direction – from the world as it is to the world as it ought to be. I have argued that the reason constructivists have had difficulty developing an ethical research agenda is because most constructivist scholarship is implicitly captivated by the fact/value distinction. Developing a constructivist ethics requires clarifying the relationship between empirical and normative statements. While more traditional approaches to the social sciences accept a rigid divide between facts and values, the social constructivist ontology is well constituted for normative research. The foundational constructivist assumption is that world politics is socially constructed. An implication of this is that the difference between facts and values is not a difference between matter and ideas, but between how ideas are presented in everyday discourse. As such, the rigid divide between empirical work and normative work that informs most social scientific research does not hold for social constructivists. Given this the challenge that a constructivist approach to ethics faces is how to engage in normative research. I have argued in this essay that rather than seek out universal ethical principles and attempt to apply them to particular cases, constructivists should focus on the grammar of everyday moral argumentation. As we have seen Cavell’s analysis of the three core features of moral argumentation, his grounding of moral arguments in the norms of conversation, and his evaluation of normative claims in the ends of a particular community enables him to analyze what individuals are doing in a particular moral argument. The central insight of this approach is that because the difference between factual and evaluative statements lies in their mode of presentation, the rigorous empirical analysis that constructivists engage in is fertile ground out of which they can develop morally persuasive arguments about how the world ought to be. PAGE 1