In this chapter we turn to practical questions, or rather to theoretical questions about practical matters. First, what is a practice (section 4.1)? Some philosophers have argued that to engage in a practice is, in some sense, to follow a rule. Others have puzzled, however, about what exactly it is to follow a rule, and how rules can be thought of as determining both what it is right to do and what we do in fact do (section 4.2). These questions lead to others about the role that judgment plays in applying general rules or concepts to particular objects or cases. What is it to judge that something is thus-and-so? What accounts for the unity and structure of judgments, assertions, or assertoric sentences? What role does judgment play in practical activities, including that of speaking a language (section 4.3)?
After considering the nature of practices in general we turn to the specific character of ethical and political practices. A fundamental question in normative ethics is whether the basis of ethical action lies in virtue, duty, or the greatest good for the greatest number (section 4.4). A fundamental question in metaethics, or philosophical reflection about the nature of ethics, is whether it is possible to derive an “ought” from an “is” (section 4.5). Another metaethical question concerns the nature of good and evil, traditionally thought of as the highest, if not the most fundamental, ethical concepts (section 4.6). What the good is to ethics, justice is to political philosophy. In section 4.7 we consider competing conceptions of the nature of justice and related questions as to whether justice requires democracy, communism, or both. We conclude the chapter by discussing the relationship between justice and law: is justice nothing more than what the law happens to decree, as legal positivists hold, or is justice a higher standard to which the law ought to conform, as proponents of natural law theories have traditionally held? This debate bears on the kind of force that existing laws have, a question that can be generalized to include the kinds of force that other types of laws have, including laws of nature and laws of logic (section 4.7).
Discussion of a special kind of “practical” reasoning or rationality, as distinct from theoretical reason in general, begins with Aristotle's introduction of phronesis, or a special capacity of reasoning about what one should do, in the Nicomachean Ethics, book 6. For Aristotle, phronesis or practical wisdom is to be distinguished both from art and from philosophical wisdom, or knowledge of things that are “highest by nature” and unchanging, in that it concerns one's own particular circumstances and the best way of acting within them (Aristotle, 1984: 1800–1 [1140a24–1141a23]). Beginning in the Groundwork of the Metaphysics of Morals, Immanuel Kant theorized the “metaphysics of morals” as based in the laws governing free action, as parallel to the “metaphysics of nature” which theorizes the laws of nature (Kant, 1785: 4:387–8). For Kant, by contrast to Aristotle, however, the foundation for properly moral action is not to be found in the specifics of particular situations or any empirical facts, but in a moral law that is universal and independent of any particular psychology, situated interest, or inclinations (Kant, 1785: 4:389–90). According to Kant, this yields a categorical imperative which, unlike various hypothetical imperatives, applies universally and unconditionally to all finite rational beings and which mandates that one act in accordance with a maxim that can itself be willed to be a universal law (Kant, 1785: 4:419–21).
In the early twentieth century, C. S. Peirce repeatedly articulated, in differing versions, what he called the “pragmatic maxim,” which enjoins an inquirer who is evaluating the truth of an assertion to consider only the totality of “practical” consequences for action and behavior that would follow from the acceptance of that truth (see, e.g., Peirce, 1905). In modified form, the maxim became the basis for the classical pragmatism defended by William James, and later applied (under the label “instrumentalism”) to questions about education and democracy by John Dewey. For the classical pragmatists, the imperative to reflect on “practical” consequences in this sense extends, as well, to the practices of scientific and philosophical inquiry, bringing metaphysical speculation back down to earth and allowing for essentially empty or primarily verbal questions to be resolved by means of the consideration of the actual implications for practice of adopting any given answer.
In the twentieth century, discussion of praxis and practical philosophy has more often involved a consideration of inherently social or collective practices rather than simply individual action, or its maxims or principles. Recent and contemporary pragmatists have argued for replacing accounts of our knowledge of the world as primarily theoretical with an understanding that situates this knowledge as the outcome of collective, cultural, or social practices of inquiry and reasoning. For instance, in Philosophy and the Mirror of Nature, Richard Rorty argues for a deflationary view of our self-understanding and for a replacement of accounts in terms of more or less accurate representational “descriptions” of the world with accounts in terms of variable practices of justification adopted locally within communities of “peers” (Rorty, 1979: 99). Along partially similar lines but in a more theoretical register, Robert Brandom argues in Making it Explicit for an “inferentialist” pragmatism that sees the semantic content of terms and expressions as inherited from their role in practices of “giving and asking for reasons.”1 On this view, the task of articulating semantic contents and their relationships amounts essentially to making explicit commitments that have already been undertaken largely implicitly in the course of ordinary linguistic practices of claiming, justifying claims, and attributing more or less justification to others. For Brandom, it is these practical activities of regarding expressions in terms of the primitive “proprieties” instituted by a particular linguistic practice that ultimately ground rationality and make possible the more articulate structure of explicitly logical norms and principles (Brandom, 1994: 117).
In Being and Time, Heidegger argues against a Cartesian conception of the “external world” as something set over against an individual subject and capable of being represented by it. By contrast with this sort of picture, Heidegger argues that there is a basic structure of “being-in-the-world” by which Dasein is always already situated and involved in specific kinds of projects and concerns. Within these projects, objects are often primarily unconcealed or given, in the context of activity, as “ready-to-hand” [Zuhanden] rather than as “present at hand” [Vorhanden] objects of explicit or theoretical cognition or representation. Some interpreters (e.g., Dreyfus, 1991) have found in Heidegger's discussion of this primarily pre-theoretical structure of being-in-the-world an analogue to Aristotle's conception of phronesis as an irreducibly situated and local skill of practical coping, and have read Heidegger as reversing the priority that Aristotle himself accords to theoretical knowledge over practice. Relatedly, some commentators have drawn from Heidegger's suggestions in Being and Time an account of the intentionality of thoughts and cognition as arising from a more basic level of practical agency, involving primarily activities of handling things in the course of projects and everyday activities (Okrent, 1988: 6; Haugeland, 2013). On the other hand, others have argued that the primacy of the Zuhanden over the Vorhanden in the structure of Being-in-the-World is only a relatively minor theme in the context of the whole analysis of Being and Time, and that especially in view of his theory of the origin of temporality in the structure of the individual Dasein and his critique of the “inauthentic” mode of the activity of the “they” [das Man], Heidegger should not be read as a “social pragmatist” in any overall sense.2
Many of those who have suggested accounts of intentionality, meaning, or mental content as based in social or collective practices have drawn centrally on an idea of linguistic practices that some have seen as suggested by the late Wittgenstein. In apparent response to the problem about rule-following that he introduces in the Philosophical Investigations (see section 4.2), Wittgenstein suggests that ‘following a rule’ must be understood as a practice (Praxis) that is, as such, essentially public (and so could not be done ‘privately’) (Wittgenstein, 2009 [1953]: 201–1). Wittgenstein suggests that such activities as making reports, giving orders, or playing games of chess are similarly essentially public practices in the sense of “customs” [Gepflogenheiten], “usages” [Gebräuche], or “institutions” [Institutionen] and that the understanding of language may itself be understood as the mastery of a “technique” [Technik] (Wittgenstein, 2009 [1953]: 199). Drawing on these suggestions, some commentators have seen Wittgenstein as proposing a “social theory” of the acquisition of knowledge or the structure of linguistic content (Bloor, 1993; Kripke, 1982), while others have combined these considerations with ones derived from Heidegger to support claims for the irreducibility of praxis as an alternative to metaphysical explanation or speculation (Dreyfus, 1991; Braver, 2012). However, it is not clear that Wittgenstein is here offering anything like a theory of practice or an account (anti-metaphysical or not) of our activities, especially given the strong critique he levels in the Investigations against the idea of philosophy as issuing in theories (see, e.g., Wittgenstein, 2009 [1953]: 126–30). Rather, it appears likely that, as he sometimes suggests, he is instead simply offering a set of relatively simple “language-games” as objects of comparison in order to clear away confusions rooted in a tendency to misunderstand the workings of our own, much more complex language (Wittgenstein, 2009 [1953]: 119, 122, 130–2).
For those who do seek to account for knowledge, linguistic meaning, semantic content, or intentionality in terms of a more basic level of social practices, or those who seek to replace accounts in terms of the former with descriptions in terms of the latter, there are several prima facie questions that arise. One of the most difficult of these is that of the origin of linguistic normativity: What is the basis in practice for the character of linguistic expressions and performances that makes them correct or incorrect, or (in the case of assertions) true or false? (See also section 1.2, and section 4.4 of this chapter.) Even if assertions are treated (as J. L. Austin suggests) simply as one kind of linguistic performances among others, each with certain conditions of “felicity” or “infelicity,” it is puzzling how these conditions could depend simply on the typical regularities of a specific group, community, or cultural practice and still leave open the possibility of criticizing claims that are made in conformity with this practice as false or incorrect (Austin, 1962). And even if (as is suggested, for instance, by Brandom) the normative features of fully articulate language are inherited from a basis of prior activities of rewarding or punishing individual community members, it is unclear how to account for the possibility that these prior activities, themselves, may be rationally criticized.
There are also questions about how more sophisticated or cognitive linguistic contents are structured by a more basic level of practices. As we have seen, some philosophers, such as Wittgenstein, seem to point to a level of irreducibly linguistic practices, while others, such as Heidegger, seem to suggest that the practices in which content is ultimately grounded need not be linguistic in nature and may involve simply everyday activities of handling objects and coping with the world. Either way, there appears to be a question about how linguistic content originates from the non-linguistic. If our practices “always already” involve such linguistic activities as making articulate claims and raising questions, then we seem to presuppose here what needs to be explained. But if we hold that there is an origin of language itself from pre-linguistic or non-linguistic practices, this also needs to be explained. Along similar lines, for those who hold that linguistic practices or their constitutive rules are instituted or constituted conventionally, for instance by means of initial decisions or designations, there is the question of how such an initial decision or convention could take place, unless it presupposes the very linguistic articulation that is to be explained (cf. Livingston, 2012, ch. 1). Lurking behind each of these questions is the difficult one of the origin of language itself, the question of how language emerges from within a world that at first has no language-speakers or linguistic agents.
Many in twentieth-century and recent philosophy have also appealed to notions of practice and practical activity in sociopolitical theory and critique. Beginning with Marx's materialist insistence on the primacy of material conditions and concrete modes of the practical organization of social and economic relations over abstract or ideal thought, the subsequent Marxist tradition has developed various conceptions of the role of practice and its relation to philosophical theory. In History and Class Consciousness, Georg Lukács theorized the development of a revolutionary consciousness by the proletariat as the outcome of a dialectical praxis arising from awareness of social relationships as constituting the structure of the world as it is experienced and known. Subsequently, Louis Althusser (Althusser, 2005 [1965]) conceived of the “scientific” theory of dialectical materialism as itself, like all theoretical activity, the outcome of a specific kind of theoretical praxis, one which ultimately is capable of transforming the world by making concrete what is at first thought only abstractly. In later writing, Althusser (1971) began to develop a critical theorization of the role of what he called the “Ideological State Apparatus,” or the set of instituted mechanisms and structures, for instance prisons, schools, and armies, by which the state maintains itself and ensures the perpetuation of the existing class structure.
Outside the direct lineage of Marxism, others have sought to develop the socio-political implications of a thinking of practice. In a variety of works, Foucault treated the history and development of particular social institutions such as asylums and clinics as arising and maintaining themselves through a complex set of relations of discursive and non-discursive practices and developments of power. In The Human Condition, Hannah Arendt considers the historical foundation of the distinction between private and public realms of activity in the ancient Greeks and criticizes the emphasis placed in modern thought, since Descartes, on abstract mentality and individual thinking over the realm of public and social practice (Arendt, 1998 [1958]). In recent years, a series of “communitarian” thinkers such as Alasdair MacIntyre (MacIntyre, 2011 [1981, 2007]), Michael Walzer (Walzer, 1983), and Michael Sandel (Sandel, 1998 [1982]) have called for a corrective to the tendency of contemporary political theory to favor abstract and general structural solutions without taking due consideration of the particularities of specific communities and systems of value. In a different theoretical context, Giorgio Agamben has argued for a “coming community” to be characterized by the liquidation of all fixed identities and the development of the radical and pure praxis grounded in the re-appropriation of language itself as the fact of “linguistic being,” over against its alienation in forms of abstract communication and information in contemporary capitalist life (Agamben, 1993: 71–83).
For all of these positions which accord practice, practices, or praxis an essential role in the description or critique of contemporary life, there arise questions about the nature of (what is here called) a “practice” itself. Must the practices that structure a given community or social whole be carried out consciously and with the full agreement of the relevant community, or are they just as often hidden, obscured, or ideologically displaced or concealed in terms of ultimately false motivations or justifications? Is there a category or type of practice (say economic, ideological, or scientific) that determines in the most basic sense the character of the social situation as a whole, or are all existing “practices” (in some sense or another) relevant to this determination? Can practices be distinguished as better or worse with reference to some antecedently given social ends, or are ends themselves determinable only within, and thus relative to, specific practices? How can we think about the possibility of large-scale changes in predominant or prevailing practices, and under what conditions can this change lead to progress or to the positive transformation of society?
Finally, and perhaps most fundamentally, there are questions about what kind of unity a “practice” has to begin with. Are all practices, understood simply as regularities of behavior, necessarily determined by a set of specifiable rules? Or does the dimension of “practice” point just as centrally to regularities or patterns of activity that can be structured much more by habit, inheritance, or intuition? To what extent and in what way does the idea of an (essentially public, collective, or linguistic) practice impose the assumption of an already-constituted collective subject, a “we” which (it then seems) can only act as “we” do, or can only continue with our collective life in the ways that are already familiar or legitimate for “us”?
One reason this question is important is that it bears on our understanding of the difference between ourselves and other animals. As Derrida has pointed out, philosophers have generally assumed that we alone are capable of acting responsibly, the behavior of all other animals consisting in nothing more than automatic or semi-automatic reactions to stimuli. Derrida challenges this dogma not only by seeking to show that nonhuman animals act responsibly, but also by showing that there is a reactive core to much of what passes for responsible human action, for instance when we automatically deploy customary linguistic idioms. Derrida's aim is not to show that there is no such thing as genuine responsibility, but rather to prompt us to think more carefully about what genuine responsibility would consist in. He suggests, in particular, that responsibility begins when instead of simply following customary rules we question or even suspend them and risk acting in a manner that might seem to be irresponsible by the lights of prevailing norms (Derrida, 2011 [2001–2]).
Philosophical accounts of meaning, justification, and ethical practice prior to the twentieth century have often presupposed the idea of constitutive rules which govern or constitute the activities of (correctly) speaking, understanding, or acting morally. While Kant, for example, treats concepts in general as rules under which particular objects or phenomena fall, he also suggests in the Critique of Pure Reason the existence of a special faculty of judgment which comprehends the rules of this subsumption of particulars under universals (see section 4.3), and understands moral action as action determined in accordance with the supreme rule of the categorical imperative, itself a direct outcome of the moral law (Kant, 1998 [1781/1787]: A132–3/B171–2 and 1785: 4:419–21). For Kant, since the moral law springs from our own rational nature, it is at the same time a law that binds us and one that we give ourselves independently of natural laws that affect us from without, allowing for the possibility of our freedom or autonomy.
Early analytic philosophers, in close connection with the idea of a practice of logical or grammatical analysis of language into semantically simple parts and their rule-governed interrelations, proposed a conception of logically structured language itself as essentially a system of rules. For Frege, for example, the laws of logic are objective (though non-spatiotemporal) realities that govern the process of inference in natural (or any) language insofar as this regular process preserves truth. Wittgenstein, in the Tractatus Logico-Philosophicus, suggests that philosophical errors can be avoided by developing a rigorous logical syntax that would specify precisely how each sign in a symbolic language is to be used (Wittgenstein, 2001 [1921]: 3.325). According to the rather different conception developed by Carnap in The Logical Syntax of Language (1937), any particular language consists of a vocabulary of symbols and a freely chosen corpus of rules of formation and transformation for introducing sequences of signs. Later, Quine would influentially criticize this conventionalist picture, arguing (for example) in “Truth by Convention” (Quine, 1976 [1936]) and “Carnap and Logical Truth” (Quine, 1960 [1957]) that the conventionalist stipulation of logical rules would itself presuppose – and so cannot explain – the logical rules purportedly stipulated (see also Livingston, 2008, ch. 5).
In the Philosophical Investigations, the later Wittgenstein criticizes his own earlier view that “if anyone utters a sentence and means or understands it, he is thereby operating a calculus according to particular rules” (Wittgenstein, 2009 [1953]: section 81). The skein of passages in the Philosophical Investigations that is sometimes discussed as the “rule-following considerations” begins with the question of what is actually called “the rule according to which [one] proceeds” (PI 82), and progressively deepens the questions of the actual nature of (what are called) “rules” for the use or understanding of linguistic expressions and what it means to follow them.
One problem is that even a simple rule, e.g. for the coordination of words to pictures, can itself, if written down, be interpreted in any number of ways (PI 86). More generally, no symbolic expression of a rule determines uniquely how the rule itself is to be interpreted – what it is to be interpreted as demanding – in each of an infinite number of cases. For there will always be alternative interpretations that are possible at any particular step, and cannot be ruled out in advance (except by means of another symbolic expression, which could itself be variously interpreted). The general situation leads to what Wittgenstein states as “our paradox” at PI 201: “…no course of action could be determined by a rule, because every course of action can be brought into conflict with it.” In particular, the natural idea that the formula that expresses the rule itself contains each of the infinite number of cases of application itself has its foundation in a picture of the rule as (as it were) a “rail laid to infinity” (PI 218), with the infinite applicability of the rule corresponding to the infinite length of the rail. This idea is closely connected with the idea of identity as a predicate or property that all things bear to themselves, and thereby suggests a conception of rules themselves as self-identical items capable of repeating themselves ad infinitum through the diversity of particular cases (PI 214–16; cf. Livingston, 2003). But this conception is to be rejected, for even if it were the case that the rule had this structure of infinite self-repetition, it would still remain mysterious how it guides our action or judgment in each case (PI 219).
In an influential interpretation and development of Wittgenstein's rule-following argument, Saul Kripke suggests that the argument poses a “skeptical paradox” about rules and meaning. Kripke considers in particular the case of computing sums according to the familiar mathematical function denoted by the ‘+’ symbol. He considers the possibility of my encountering a “bizarre skeptic” who holds, with respect to a sum I have not computed before (say, 68 + 57) that the answer I should give should not be 125, but rather 5. For perhaps in the past I have always used ‘+’ not to denote plus but rather quus, where x quus y is x + y if x and y are both less than 57, but 5 otherwise. The hypothesis is bizarre, but as Kripke argues, it is consistent with all of my past behavior: given that I have never performed this computation before, there is no fact I can cite about how I have behaved in the past that suffices to rule out the skeptical possibility that I have been computing in accordance with quus all along. Moreover, there is also apparently no fact about any internal mental representation, phenomenological experience, internal dispositions, or subjective item that could rule out the skeptical hypothesis either, for any such item would itself, as finite, be incapable of determining all of the infinitely many cases. Generalizing these considerations to the question of how the meaningful usage of words is itself determined, Kripke draws the radical skeptical conclusion that appears to be implied: “There can be no such thing as meaning anything by any word. Each new application we make is a leap in the dark; any present intention could be interpreted so as to accord with anything we may choose to do” (Kripke, 1982: 55).
Various interpreters have proposed solutions to this skeptical problem about meaning, either on Kripke's formulation or Wittgenstein's original one. Kripke himself argues that the problem can be given a “skeptical solution” in terms of essentially public and intersubjective practices. Somewhat analogously to the way Hume proposes a solution for skeptical worries about the real metaphysical existence of causal relations in terms of the “habit” or custom that constrains us to suppose them, Wittgenstein, on Kripke's reading, proposes a “skeptical solution” in terms of essentially public and intersubjective practices of calculating and intending, and of expecting the calculations and intentions of others to accord with our own (Kripke, 1982: 92–3).
With this, Kripke joins a number of communitarian readers of Wittgenstein who interpret him as holding that the regularities or institutions of a community or its characteristic practices provide grounds for a positive account of what is involved in following rules or in meaning (see also, e.g., Bloor, 1993 and Brandom, 1994 for readings broadly of this kind). Others (e.g. Maddy, 1984) have proposed various forms of naturalist resolution, appealing, for example, to the possibility that rule-following may have an ultimate basis in computational capacities implemented or ‘hard-wired’ into the neurobiological structure of the brain. However, given the structure of Wittgenstein's argument, both types of solution face substantial additional problems. For the communitarian reader, for example, it will be necessary to explain how positive accounts in terms of “practices,” if these are themselves understood as based in or constituted by rules or similar ‘regularities,’ do not simply replicate the problem of how a finitely statable rule determines an infinite number of cases. Similarly, naturalist resolutions according to which rule-following ultimately has an internal basis in the computational architecture of the brain will have to answer the objection that this just reproduces the original problem, with an essentially finite item or structure being seen as determining a “disposition” to behave “in accordance with” the rule in an infinite number of particular cases.
More broadly, Wittgenstein's rule-following problem can also be seen, through its connection with the question of how meaning is itself determined, as intimately linked with more explicitly social and political questions of rules, laws, institutions and force. In The Theory of Communicative Action, for example, Jürgen Habermas understands Wittgenstein's positive description of rule-following as a “practice” (see section 4.1) as pointing toward an irreducibly situated context in which language, intersubjective action, and meaning are inherently entangled. Adopting terminology from Husserl, Habermas calls this context the “life-world,” and opposes it to the more abstract and “rationalized” context of generalized and institutional rules and techniques separated from particular communicative situations (Habermas, 1984 [1981a, 1981b]). In The Differend, Jean-François Lyotard considers the problem of finding just solutions when systems of law or administration impose themselves on communities and practices characterized by different, and possibly incommensurable, language games, so that no single set of procedural rules may justly be considered universally applicable across these differences.
More generally, in view of Wittgenstein's rule-following problem, it appears relevant to reconsider the form and force of abstract and “universalizing” systems of law, technology, and abstract practice in relation to the concrete forms of life that they purport to envelop and regulate, up to and including the systematic forms and characteristic practices of contemporary “global” capitalism. In connection with this, it may be possible to read Wittgenstein's rule-following considerations as implying a broad critique of this interconnected set of practices, institutions, and regulations, or of the assumption of automatically guaranteed force and effectivity which typically characterizes the application of abstract rational and procedural rules within them (Livingston, 2012: chs. 6 and 10).
Beginning at least with Aristotle, philosophers have often considered the structure of a judgment to mirror the subject/predicate structure of a predicative sentence: something (the predicate) is judged of or about something else (the subject). Early in the Critique of Pure Reason, Kant described analytic judgments (drawing on Leibniz's earlier view of the structure of all true judgments) as those in which the predicate is “contained within” the concept of the subject and synthetic ones as those in which this is not the case (Kant, 1998 [1781/1787]: A6–7/B10–11). Kant summarizes what are for him the twelve forms of judgment in a table which then provides the basis for a deduction of the “categories” or pure concepts of the understanding; these are the a priori concepts which allow any object whatsoever to be thought (A 70/B 95, A 80–1/B106–7, A92–3/B125–6). In the first Critique, Kant also discusses the power of judgment [Urteilskraft] as the “faculty of subsuming under rules, i.e. of determining whether something stands under a given rule… or not” (A 132/B 171). Kant's account of the power of judgment involves a “schematism” which explains the conditions under which the categories can be applied to spatiotemporal appearance by means of a kind of mediating representation that is at once intellectual and sensible (A138/B177), as well as a doctrine of the general form of all synthetic a priori judgments as conditioned by the possibility of this application of (atemporal) concepts to temporally given intuitions (A155–8/B193–7). Some years later, in the Critique of the Power of Judgment, Kant considered the particular problems of the application of the power of judgment or subsumption to aesthetic judgments, including judgments of taste, connecting these as well to the rational structure of teleological unity that characterizes the systematic organization of nature.
In the nineteenth century, Bernard Bolzano drew a distinction that Kant himself does not generally observe but which proved decisive for a variety of later approaches, including neo-Kantianism, Husserlian phenomenology, and Frege's project in the Begriffsschrift (and after). This is the distinction between the temporal act of judgment and its ideal or atemporal content. According to the distinction, the content – what is judged in a judgment – must always be sharply distinguished from the individual, temporal act of judging itself. Whereas the act takes place at a specific time and is carried out by a particular agent, the content is largely or wholly independent of time and the agent and can in principle be judged by anyone at any time. This distinction became the basis for Husserl's distinction, drawn in Ideas 1, between “noetic” content or meaning immanent to an intentional act on the occasion of its performance and ideal “noematic” content that can be shared or instantiated by any number of concrete acts (Husserl, 1983 [1913]: 214). The distinction between the psychological act of thinking and the ideal content of what is thought also became a key weapon in Frege's arsenal against the claims of psychologism, or the attempt to reduce the normativity and legality of logic to psychological laws or occurrences, as Frege accused Husserl of doing in his early Philosophy of Arithmetic. In the late essay “The Thought,” Frege clarified his picture of judgment in terms of a (perhaps metaphorical) distinction among three “realms”: the “first” realm of spatiotemporal objectivities, the “second” of psychological or subjective acts, and the “third” realm of ideal contents which are non-spatiotemporal but nevertheless objective, capable of being “grasped” in any number of distinct concrete acts of judgment or assertion.
In the Begriffsschrift, Frege introduced into his symbolism for logical inference a two-part symbol for judgment, the (now-familiar) “turnstile,” ‘|-’. In the symbolism, according to Frege, while the horizontal line or “content stroke” binds the symbols that follow into a unified whole, the vertical line or “judgment stroke” serves to show that what follows expresses a (genuine) judgment as opposed to a mere content that is not judged or asserted (Frege, 1997 [1879]: 52–3). In a private summary of his accomplishments written in 1906, Frege points to the discovery of the judgment stroke as one of the greatest of these, in that it expresses the “dissociation” of assertoric force – the difference between merely entertaining or contemplating a content and believing, asserting, or judging it – from the logical structure of predication itself (Martin, 2006: 75).
However, questions arise here about the relationship between content and force itself. One is the question of whether an indication or expression of judgment has any legitimate role in a purely logical notation to begin with. After all, the fact that something is judged, believed, or asserted would appear to characterize temporal and psychological acts rather than logical contents, and thus to play no legitimate role in a purely logical notation for expressing inferential relations among contents. Second and relatedly, it is not clear that an expression of the fact of judgment is relevant to a characterization of logical inference at all. For inferential relations hold between contents whether or not anyone actually carries out the inference or actually makes the relevant judgments (both worries are sketched in Martin, 2006: 85–6).
Frege also argued for a partially transformed conception of the structure of the predicative sentence itself. Rather than simply being a combination or synthesis of two individually significant terms, a “subject” term and a “predicate” term (as on Aristotle's conception), the simple predicative sentence has, for Frege, the unity of a general function and its determinate value. Thus, rather than construing the sentence “the apple is red” as the synthetic unity of a subject term standing by itself for the apple and a predicate term standing by itself for the universal or property of redness, Frege understands it as an instance of the general predicative function “x is red.” The function is signaled by the “concept” term “red,” with the apple here being supplied as a determinate value for the variable in the “object” place (see section 2.1). The predicative function itself can then be understood, on analogy with an algebraic function of a single variable, as a function or rule taking each determinate object to one of the two truth-values, True or False, and the account can easily be extended to handle relational sentences (e.g. “Socrates is older than Plato”) as involving propositional functions with two or more variable places. Frege accordingly argues strenuously for a fundamental distinction between concept terms, which serve to predicate, and object terms, which fill the variable “place” left open by a conceptual function, and maintains that it is essential that one never mistake one kind of term for the other.
However, Frege's conception also raises additional metaphysical and semantic questions. One is that of the actual reference of concept-terms. On Frege's analysis, concepts are, as we've seen, inherently “gappy” or unsaturated. Concept-terms do not signify objects by themselves, but rather stand for functions that themselves need to be completed by objects in order to determine a truth-value. But since, according to Frege, concept-terms themselves do have reference, there apparently must be, in reality, actual “gappy” referents – and it is then obscure what these could be or how they can be related to the familiar objects of experience. Another, and perhaps more deeply indicative, problem is noted by Frege himself in “On Concept and Object” (Frege, 1997 [1892]). It follows from the distinction between concept and object that the term for a concept can never be used with sense in the object-position in a sentence: to do so would be to violate the very rules that establish the possibility of saying anything about anything at all. Accordingly, it must be impossible meaningfully to say anything about a concept, even (for instance) to claim that “the concept horse is easily acquired.” Moreover, for the same reason, it is impossible even to state or describe the difference between concepts and objects themselves by saying that a concept is one kind of “thing” while an object is another. Frege answers the problem by suggesting that certain logical or categorical distinctions such as this one cannot be straightforwardly described, but must instead be indicated by means of what he calls “elucidations.” While some (e.g. Conant, 2002) have seen this as anticipating Wittgenstein's distinction, in the Tractatus, between the realm of factual description or “saying” and the status of distinctions or categories of logical form that can only be “shown,” it is not clear that this resolves the problem of the status of categorical or metaphysical distinctions of logical type itself.
Further questions can be raised about the structure and force of concepts as (truth-determining) functional rules. If a concept is indeed conceived as a rule for determining a truth-value, how are such rules constituted or instituted? If their constitution takes place at some time, how does it nevertheless succeed in determining truth-values that hold timelessly for each of an infinite number of sentences and objects? Most broadly, what then exactly is the relationship between the general rule and its application to a determinate case which yields the distinction, in such a case, between a true and a false predication or (in an older vocabulary) that between a successful and an unsuccessful “subsumption” of an individual object under a general term?
In his posthumously published book, Truth and Predication, Donald Davidson argues that many of these problems can effectively be solved by explicating and adapting the apparatus that Tarski developed for the introduction and clarification of the structure of a truth-predicate for a language (see section 1.4). On this solution, the ancient problem of the unity of the predicative sentence is solved, for a particular language, by the reconstruction, in the radical interpretation of a natural language, of a finite vocabulary of primitive concept and object terms and a recursive grammar that allows the language to generate an infinite number of sentences thereby evaluable in terms of their truth-conditions. For Davidson, the crucial consideration in resolving the metaphysical problems pointed to by Frege's account of concepts as functions is that not only artificial formal languages of the kind Tarski considered, but also natural languages themselves, must be capable of characterization in terms of a recursive structure of semantic rules if they are to be able to generate an infinite number of sentences evaluable as true or false. However, one can still evidently raise questions about this structure of rules itself. As Davidson himself argues in Truth and Predication, the application in interpretation of the concept of truth and falsity to the utterances of an initially unfamiliar language would seem to require as a presupposition a more general and non-language-specific concept of truth, whose structure then cannot be given simply by means of the Tarskisan structure of a truth-predicate for any specific language but must rather be seen as indicated, without being defined, in the actual practice of interpretation itself.
In a series of lecture courses in the years leading up to the publication of Being and Time in 1927 , Martin Heidegger challenges on “ontological” grounds the distinction between temporal acts and ideal contents, in both its Husserlian and Neo-Kantian versions (cf. esp. Heidegger, 2010 [1925–6] and 1985 [1925]). For Heidegger, the crucial problem is that in both versions, the distinction presupposes an ultimately temporal distinction between the realm of spatiotemporal and experiential objects, which arise and perish in time, and the “timeless” or eternal realm of ideal contents, and provides no real explanation for the latter.
Although Heidegger does not suggest returning to a psychologistic picture of the relationship of act and content, he argues that this temporal distinction of two (or, as Frege would put it, three) realms has never been clarified and that its problems in fact return us directly to the ancient Platonic problem of participation, or of the relationship of (what Plato thought of as) the ideality of forms to real spatiotemporal existents. At the actual bottom of this relationship as well as the contemporary accounts of judgment (whether as a psychological act or as an ideal form of contents) is, Heidegger argues, the phenomenon of truth itself, here understood as “unconcealment” or “presencing” rather than as linguistic truth-functionality (see section 1.4). For Heidegger in Being and Time, in particular, explicit and linguistic judgments or “assertions” have their deeper ground in a prior, hermeneutic structure of disclosure that is not in itself linguistic, but according to which entities appear “as” something or in some way. Clarification of this deeper phenomenon of truth as hermeneutic presencing offers to put the theory of judgment on a new footing and show in a different way what is presupposed ontologically in the predicative structure of sentences. In particular, Heidegger suggests, it offers to clarify the deeper relationship between Dasein as the kind of being capable of truth and unconcealment and the structure of time as it is given in experience and thought.
Each of these problems and questions about the theory of judgment in its various contemporary forms might be related to the general problem introduced by Wittgenstein about rules and rule-following. If, in particular, judgment remains, for Frege, Tarski, and Davidson as much as for Kant, a phenomenon of rule-following or of the application of a concept as a general rule to a particular case, we can ask the question, with Wittgenstein, of how the “ideal” structure of the rule as such is actually constituted and how the force of its application “in principle” to infinitely many cases of conceptual subsumption or judgment is actually maintained. Once we have seen the depth of the problem posed by Wittgenstein's considerations for this conception of a rule and its application, moreover, we can see its implications as not simply theoretical but also as directly practical and even political. In State of Exception, Giorgio Agamben considers the interrelated problems of the “normative” force of laws within a particular, constituted juridico-political regime and of the more general “force of law” as such (see section 4.8). He argues that there is always a constitutive gap, in such a regime, between laws and their interpretation or application to particular cases, and that this gap can only be filled, and the assumed normative force of laws maintained, by the implicit or explicit invocation of a “state of exception” in which norm and force coincide, for instance in an executive power that can under special circumstances directly issue decrees which, though they circumvent the usual process of law, are immediately in force.
From this perspective, Agamben argues, the problem of law and application is “put on a false track” (Agamben, 2005: 39) when it is seen in terms of Kant's conception of a faculty of judgment, understood as a power of determining the application of given rules to particular cases. For whereas on this conception (as well as, we have seen, on the Fregean, Tarskian, and Davidsonian ones), the rules are understood as pre-given and as needing only to be applied in order thereby to be in legitimate force, in reflecting critically about the force of law we must consider, as well, how such rules are constituted, instituted, learned, or understood to begin with. And this question, Agamben argues, is not simply a logical one but, more deeply, a practical one: how does an agent or subject assume the linguistic position from which she can “successfully” (by our lights) apply rules or laws (linguistic or juridical) to new cases? And how are rules and norms for such “successful” judgment constituted or instituted to begin with? Put this way, the problem rejoins that of the relationship between act and content – or that of the distinction of content from force. It may also be thought thereby to rejoin Heidegger's questions about the temporality of the “ideal” and its relation to that of particular objects or cases (see Livingston, unpublished).
So far in this chapter, we have considered the general nature of practices, rules, and judgments. We now turn our attention to specifically ethical practices, rules, and judgments. The term “ethics” can be used to refer either to social practices, rules, and judgments (in the sense of the “ethos” of a culture: what Hegel calls Sittlichkeit) or to individual practices, rules, and judgments (ēthos in the sense of what pertains to a person's character) (cf. Aristotle, 1984: Nicomachean Ethics 1103a). Normative ethics is prescriptive rather than descriptive. Instead of telling us what a particular society's or individual's practices, rules, and judgments happen to be (a topic for the social sciences to address), it seeks to determine what they should be. Any attempt to ground a normative ethical claim on the authority of prevailing practices would need to account for the justification of the practices in question. The wrongness of murder, for example, cannot be derived from the near-universality of coercive laws prohibiting murder. Rather, for a normative ethicist, the moral rightness of such laws has to be independently demonstrated. But what if, contrary to the normative ethicist's expectations, it is impossible to provide such independent demonstrations? What if, apart from the relative perspectives of particular individuals and societies, there are no absolute, or objective, ethical values?
Contemporary relativists, such as David Wong and Gilbert Harman, argue that the truth or falsity of ethical statements is in fact relative to those who assess them. For Wong, moral relativity is compatible with a limited type of objectivity. On his view, moral statements have determinate truth values within particular social contexts. Such truth values may vary from society to society, but the statements themselves retain their cognitive import across societies (Wong, 1984). For Harman, moral judgments are relative to “moral frameworks” in much the same way that determinations of motion are relative to reference frames in the theory of special relativity (Harman and Thomson, 1996: 3). Once a particular framework has been selected, it fixes the truth values of moral statements. A potential problem for this analogy, though, is that in relativity theory it is possible to provide a general framework that reconciles the various appearances across reference frames. Even in the limited context of the special theory of relativity, the relativity of rest and motion doesn't carry over to a relativity of causal determination: the order of causes and effects holds universally. It is open to a critic of moral relativism to argue that the fixity of causal determination across reference frames is analogous to the fixity of fundamental moral facts across reference frames.
Some contemporary philosophers have rejected moral relativism while trying to accommodate the relativist's insight that norms express subjective assessments. Expressivists hold that no facts in the world can make an ethical (or, typically, any other) evaluation or prescriptive statement true. One type of expressivism is emotivism, a position advocated by A. J. Ayer in Language, Truth and Logic. For Ayer, ethical evaluations are mere “pro-” or “con-” attitudes with no genuine cognitive import: in saying that something is good or bad in an ethical sense, I do not really make a statement about the properties of the thing, but merely express my attitude toward it. Another type of expressivism is “prescriptivism” (Hare, 1991 [1952]). On this position, ethical judgments have imperative or prescriptive force in that they are ways of requiring or prescribing that something be done. Thus, although they are still not directly descriptive, they are not just expressions of emotions, and can be supported by reasons that are potentially universal.
For relativists and expressivists alike, actions thus have no intrinsic (good or bad) value, and there are no absolute values on the basis of which we can assess their relative value. In one way or another, the basis of ethical valuation lies in the perspective and activity of those who judge. Those who ground ethics in virtue take a different approach. Virtue ethics is one of the oldest and also newest normative ethical theories. Its oldest proponents include Plato and Aristotle; its more recent ones include G. E. M. Anscombe, who revived it at a time when versions of deontology and utilitarianism were regarded as the primary alternatives to expressivism and relativism.
According to Aristotle, the basis of normative ethics lies in virtue. An action is ethically right if it conduces to the acquisition of virtue, or if it is the kind of action that a virtuous person would perform. Virtue itself is excellence of character (in a sense of character that can be applied both to individuals and to societies). Excellence consists in the regularly successful performance of a thing's proper function. For both Plato and Aristotle, moreover, the proper function or virtue of a human being is (in some sense) to live rationally, both as an individual and as a member of a political community. Specific virtues, such as courage and wisdom, are aspects of virtue as a whole. Socrates, who equates virtue with knowledge, affirms the unity of the virtues: to possess any one virtue is, necessarily, to possess them all (Plato, 1997: Protagoras 329d.). Aristotle also affirms the unity of the virtues, but unlike Socrates he distinguishes virtues of character (such as courage) from intellectual virtues (such as wisdom) (Aristotle, 1984: Nicomachean Ethics 1144b). Aristotle also argues that individual virtues of character admit of degrees between opposite vices (as courage is a mean between cowardice and rashness). If virtue were knowledge, the possession of an individual virtue, like the possession of virtue as a whole, would seem to be an all-or nothing thing: one would either be fully courageous or not at all courageous. Perhaps, however, Socrates could respond that degrees of virtue and partial virtues correspond to degrees of right opinion and partially correct opinion.
One attractive feature of virtue ethics is its emphasis on the role that rational and practical judgment plays in ethical evaluation. To appreciate this feature, it is helpful to consider the difficulties that arise for other kinds of positions when they equate ethical action with rule-following of some sort.
Deontologists locate the basis of normative ethics in the concept of duty (deon in ancient Greek). To have a duty is to be morally obliged to act in a particular way. Divine command theorists ground the authority of genuine duties in the will of God. An obvious difficulty for this account is the potential inscrutability of God's will. Another has to do with arbitrariness: if God were to issue an arbitrary command, why should we regard it as morally binding? In response to this worry one could argue that God's intrinsic moral goodness wouldn't allow him to issue arbitrary or random commands: if there is a God who issues moral commands, we may assume he has good reasons for issuing them. But then it is no longer the bare will of God that grounds genuine duties; it is the good reasons God has for willing as he does. If we ourselves possess reason, though, we should be able to formulate the same ethical commands that we would regard as binding if we knew that they had been issued by a just God.
By this line of reasoning we arrive at Kant's alternative to divine command theory (see section 3.8 for his distinction between moral theology and theological morality). The basis of Kantian deontology is pure practical reason. Practical reason is pure when it determines the will independently of empirical incentives. It does so by issuing a categorical (i.e., unconditional) imperative (see section 4.1). The primary formula of the categorical imperative is: “Act as if the maxim of your action were to become by your will a universal law of nature” (Kant, 1785: 4:421). To determine whether a proposed action is morally permissible we must evaluate the rule we would be following in acting that way. We must not get distracted by the particular circumstances we find ourselves in unless they somehow bear on our understanding of the rule we would be following in acting a certain way. In particular, we must not subvert our true duty by pursuing ends, however well-intentioned, whose means involve actions whose maxims cannot be universalized (such as telling a white lie for the sake of not hurting someone's feelings). Kant points toward a more situated or “intuitable” understanding of ethical duty with the second formulation of the categorical imperative, namely, “So act that you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means” (Kant, 1785: 4:429). Taken together with the first, this version yields a third formulation: “Act in accordance with the maxims of a member giving universal laws for a merely possible kingdom of ends” (Kant, 1785: 4:438–9).
Despite the greater specificity of the second and third formulations, Kant's critics have focused on the shortcomings of the first. In Utilitarianism, John Stuart Mill claims that Kant's universalizability test fails to prohibit even the most vicious acts, for there is nothing intrinsically impossible about a world in which everyone lies, cheats, steals, murders, etc.3 We can only determine the wrongness of such acts by evaluating their consequences. Mill further criticizes Kant for failing to acknowledge that he thinks like a “consequentialist” rather than a deontologist when he considers what would follow from the universal adoption of a particular maxim. Mill himself defends an unabashed version of consequentialism, namely, utilitarianism, a position first advocated in detail by Jeremy Bentham. Utilitarians make the “principle of utility” or, equivalently, the “greatest happiness principle” the basis of normative ethics. This principle tells us that, in any situation, the right act to perform is the one that will bring about (or that seems likeliest to bring about) the greatest good for the greatest number. Like Bentham, Mill defines goodness hedonistically as the enjoyment of pleasure and absence of pain (thereby incurring Moore's charge of committing the naturalistic fallacy: see section 4.6). Mill disagrees with Bentham, however, about whether the pleasures and pains felt by all sentient creatures should be counted equally in our reckoning of the greatest good. For Mill (but not for Bentham) some specifically human pleasures (such as the pleasure of philosophizing) are qualitatively superior to animal pleasures and should be weighted accordingly in our reckoning of the greatest good for the greatest number. This disagreement continues to divide contemporary utilitarians, Peter Singer being among the most prominent defenders of Bentham's view that all sentient creatures should be regarded as having equal moral standing (Singer, 1975).
Another modification that Mill makes to the basic principle of utility involves the form of its application to particular situations. Since the principle itself does not seem to determine who should be included in our utilitarian reckoning, and also might allow certain kinds of actions which most of us do not intuitively consider to be even possibly good or right (such as a murder that might seem to bring about greater total good), Mill argues that the principle of utility doesn't directly tell us which acts to perform in concrete ethical situations but rather indirectly applies to acts by telling us which general rules to follow. With its emphasis on rules, Mill's “rule” utilitarianism is closer to Kantian deontology than is Bentham's version of the doctrine (so-called “act” utilitarianism). Unlike Kant, Mill allows for exceptions to rules in extraordinary circumstances. According to Mill, every reasonable ethical theory must allow for exceptions, but the fact that his version of rule utilitarianism is less strict than Kantian deontology threatens to reopen the problem that made him switch from act utilitarianism to rule utilitarianism in the first place. After all, who is to decide when customary rules should be suspended?
Henry Sidgwick deals with a related problem in The Methods of Ethics. While defending utilitarianism as the right ethical doctrine, he argues on utilitarian grounds that it might not be the right one for most people to follow. That is, if most people made the principles of utilitarianism the basis of their actions the result might actually be less good than if they followed ethical principles more consistent with common sense (Sidgwick, 1981 [1874, 1907]). Sidgwick concludes that it could be better for committed utilitarians not to seek converts to their way of thinking. In Ethics and the Limits of Philosophy, Bernard Williams criticizes the ethical or metaethical elitism of this position (Williams, 2010 [1985]: 122).
Beyond deontology and utilitarianism, T. M. Scanlon locates the basis of normative ethics in the social contracts that regulate human behavior. What makes an action right or wrong, for Scanlon, is its compatibility or incompatibility with social rules that no one could “reasonably” violate. Scanlon advocates a substantive conception of reasonableness that is supposed to differ from Kantian reason in being culturally grounded and therefore heteronomous rather than autonomous (Scanlon, 1998). Yet many of Scanlon's ethically determining reasons look a lot like Kantian reasons (especially if we focus on reasons based on the second and third formulations of the categorical imperative). Given the obvious similarities between contractualism and deontology, on the one hand, and between deontology and rule utilitarianism, on the other, it is not surprising that Derek Parfit takes advocates of all three positions to be “climbing the same mountain” (Parfit, 2011).
Instead of trying to reconcile deontology and utilitarianism, some virtue ethicists and communitarians have argued that both theories face common, perhaps insuperable, problems. One question is whether deontology or utilitarianism can adequately capture the orientation toward the good that is arguably essential to ethical action and character. For Alasdair MacIntyre, for instance, the problem with purportedly universal liberal ethical theories such as deontology and utilitarianism is that they lack any substantive (or “thick”) conception of the common good (beyond that of individual pleasure, for utilitarians) and fail to recognize their own cultural specificity. Substantive conceptions of the good vary from community to community, while ethical rules derive their binding force from the larger social practices in which they are embedded. According to MacIntyre, Aristotle was more sensitive to these facts than either Kant or Mill.
Charles Taylor, another communitarian, finds similar insights in Hegel. According to Taylor, Hegel privileges Sittlichkeit, social mores, over individual Moralität. In the Phenomenology of Spirit Hegel shows how a Kantian-inspired “drive to absolute freedom” manifested itself in the Reign of Terror during the French Revolution. For Taylor, the moral that Hegel drew from this “disaster” was that human freedom must be concretely embodied in social practices (Taylor, 1979: 118, 100). According to Slavoj Žižek, however, Hegel's deeper point is that freedom can be effectively concretized only by passing through the universalizing stage of absolute freedom, a passage that makes it impossible to return to the type of community Taylor favors. Despite Hegel's critique of the “formal indeterminacy” of the categorical imperative, he shares Kant's conception of the universality of practical reason, and he fully endorses revolutionary struggles for freedom. In a more liberal vein, Robert Pippin argues that Hegel's response to Kant consists not in rejecting the universality of practical reason but in calling attention to the various ways in which practices of mutual recognition establish new norms – an interpretation that makes Hegel more of a contractualist than either a deontologist or a communitarian (Pippin, 2011: 92).
For Slavoj Žižek, the strength of the categorical imperative lies precisely in its “emptiness,” that is, in its non-specificity about which maxims we ought to follow. Far from being a shortcoming, the impossibility of “deducing” our duty means that we must take responsibility for identifying something as our duty in the first place (Žižek, 2012: 127–8). Jacques Derrida makes a similar point when he argues that genuine responsibility cannot be reduced to bare calculation. To borrow a phrase from John Rawls, there is no “decision procedure in ethics” (cf. Rawls, 1951; Rosen, 2014). Instead of following such a procedure, we can only “calculate with the incalculable” (Derrida, 2002: 244). For both Derrida and Žižek, ethical judgment comes into play not only at the level at which we deal with particulars (the distinctive feature of Aristotle's conception of phronesis, as noted in section 4.1), but at the level at which we identify rules as rules, or formulate the maxims we would be following in performing particular actions.
One fundamental question of metaethics is whether it is possible to derive an “ought” – a judgment about what we should do, or about what should be the case – from an “is” – a claim about what is or how things, in fact, are. In A Treatise of Human Nature David Hume denies this possibility. From merely descriptive premises about the way things are (for example, that a child is hungry) one cannot derive prescriptive statements about the way things ought to be (for example, that the child should be fed). Prescriptive conclusions require at least some prescriptive premises (for example, the premise that hungry children ought to be fed). But according to Hume, traditional ethical theorists have failed to notice this fact. Their arguments typically involve an “imperceptible” shift from descriptions to prescriptions without any acknowledgment, let alone justification, of the transition (Hume, 2000 [1738]: 302 [3.1.1]).
Hume is generally characterized as the first modern philosopher to explicitly call attention to the is/ought distinction, but its roots can be traced back to the rise of modern science in the seventeenth century. As Alexandre Koyré observes, the replacement of the medieval conception of a hierarchical cosmos with the modern conception of an infinite, homogeneous universe involved a “devalorization of being, the divorce of the world of value and the world of facts” (Koyré, 1957: 2). While some modern philosophers, like Hume, responded to the new worldview by questioning the existence of a world of values, others, like Kant, distinguished the world of facts (nature) from the world of value (the “kingdom of ends”) (Kant, 1785). For Kant, unlike Hume, moral value has its source in reason rather than in sentiment, making it possible to identify imperatives that are categorical (in the sense of being independent of hypothetical sentiments or inclinations). In a more naturalistic vein, contemporary Kantian “constructivists” have tried to preserve Kant's idea that reason is a source of ethical value while downplaying or rejecting the metaphysical and theological framework in which he embeds his moral theory (e.g., Rawls, 1980, O’Neill, 1990; Korsgaard, 1996).
In Principia Ethica, G. E. Moore argued by means of an “open question argument” (see section 4.6) that any attempt to define “good” in terms of natural or factual characteristics or properties must fail and that any ethical position that assumes such a definition thereby commits a “naturalistic fallacy.” Along partly similar lines, Wittgenstein argued in the 1929 “Lecture on Ethics” that, although one can express judgments of relative or instrumental value with factual descriptions (e.g. that increased food production would lead to more children being fed), a book containing all the facts about the world would nevertheless contain nothing that is an expression of absolute value or a statement of what is good in an absolute or unqualified sense. Such a book could only include facts, and these are all (according to Wittgenstein) of equal value: no fact or set of facts – nothing in the world – could establish or prove that some other fact or set of facts is intrinsically of more or less value, or more or less to be desired, in an absolute sense, than any other.
Against these positions, ethical naturalists claim that ethical properties such as goodness and evil can be reduced to natural properties or facts. Proponents of this view include Philippa Foot and the so-called “Cornell [moral] realists” Nicholas Sturgeon, Richard Boyd, and David Brink (Foot, 2001; Sturgeon, 1988; Boyd, 1988; Brink, 1989). Sturgeon argues that skepticism about unobservable ethical properties should be no more plausible than skepticism about many of the theoretical entities that physicists posit. Both types of skepticism rest on outdated empiricist prejudices. Neo-Aristotelians such as Foot further suggest that ethical properties supervene on natural functions: ethical virtue can be treated as an aspect or consequence of the good natural functioning of an organism, according to the distinctive purposes this functioning has. In advocating these positions, contemporary ethical naturalists seek to reclaim a key feature of the pre-seventeenth century picture of the hierarchical cosmos, namely, its teleological conception of life. Whether and how such attempts to “re-enchant” nature can be reconciled with a post-Darwinian understanding of evolution is a contested question.
Heidegger suggests that the true problem of modernity lies not with the disenchantment of nature, but rather with the interpretation of being as value itself. The roots of this interpretation go back to the very picture of nature that the neo-Aristotelians seek to revive. Pre-Platonic Greek thinkers conceived of being as physis, or (roughly) coming-into-presence. Plato and Aristotle reinterpreted physis as striving for the good, conceived either as a single transcendent form (Plato) or as a multiplicity of forms corresponding to the multiplicity of forms of life (Aristotle). Aristotle's teleological interpretation of nature gave rise to the hierarchical world-picture that prevailed up until Descartes relocated its foundation in the thinking subject. From here it was a short, inevitable step to Hume's moral subjectivism and Nietzsche's identification of will to power as the true source of value. According to Nietzsche, this path from Plato to “the death of God” is the history of nihilism, understood as the decline of the highest values. But according to Heidegger, Nietzsche's attempt to create new values – and to valorize processes of becoming rather than being – unwittingly replicates the very nihilism he seeks to overcome. For Heidegger, the appropriate response to nihilism is to rethink the relationship between being and nothing (see section 3.7).
Whether Heidegger has a positive ethics is a contested question. According to Emmanuel Levinas, Heidegger's exclusive focus on the forgetting of being itself involves a forgetting of the good. In Being and Time, Heidegger represents intersubjectivity as the experience of being-with-others in the world, but he doesn't thematize the ethical transcendence that occurs in a face-to-face encounter with another person (see section 3.8). Levinas concludes that Heidegger has no true conception of moral obligation toward others (Levinas, 1969 [1961]: 68). Instead of rejecting the is/ought distinction, Levinas embraces it. On his quasi-prescriptivist account, the prescriptive force of “ought” claims derives neither from a first-person (or “subjective”) point of view, nor from a third-person (or “objective”) point of view, but from a second-person (or properly ethical) point of view. Obligation has its source in the claim “Thou shalt not kill” which another person makes on us.
Peter Singer argues that descriptivism and prescriptivism pose symmetrical problems. If we accept descriptivism, we preserve the factual status of “ought” claims, but at the cost of not being able to account for their obligatory force. If instead we accept prescriptivism, we then save obligation at the cost of not being able to distinguish genuinely binding moral prescriptions from prescriptions that are arbitrary in form and content (Singer, 1973). These problems could perhaps be resolved by considering the fact/value distinction as a spectrum that admits of varying degrees rather than as a stark difference in kind. Hilary Putnam makes a suggestion along these lines, enabling him to defend a softer version of ethical realism (as he likewise defends softer forms of scientific realism) (Putnam, 1987; see sections 3.1 and 3.3).
If we could close the gaps identified by Singer (whether in the manner proposed by Putnam or in some other way) we would still be faced with another that might be thought of as a meta-level equivalent of the gap between “is” and “ought.” This is the motivational gap opened up by the question “Why be moral?” It has a meta-level status insofar as it challenges the idea that we ought to do what we are ethically obliged to do. Philippa Foot and Kai Nielsen are among those who have argued that no completely satisfactory answer can be given to this question (Foot, 1972; Nielsen, 1989). Like every other form of skepticism, moral skepticism is difficult, if not impossible, to refute. In the Republic, Glaucon asks Socrates whether there are external circumstances that could lead one to prefer an unjust to a just life (Plato, 1997: 1001–2 [Republic, 360e–362c]). Socrates rejects this possibility on the grounds that a just life is intrinsically superior to an unjust life. If living a good life is equivalent to living a just life, then no further response needs to be given (or can be given) to the question “Why be just?” The most we can do is illuminate the relationship between goodness and justice.
In Book VI of the Republic, Socrates tells Glaucon that the Good – or the Idea or Form of the Good – exceeds (epekeina) being in dignity and power. He goes on to draw an analogy between the Good and the Sun: just as the Sun illuminates visible objects and so enables us to see them with our eyes, so the Good “illuminates” the forms, enabling us to intuit, or “see,” them with our minds. Plato further suggests that the Good is the cause of the reality of the forms, as well as of the sensible particulars that participate in them. This idea was developed further by the third-century neo-Platonic philosopher Plotinus. For Plotinus, the Good – or, equivalently, the One – gives rise to the Many (and to Evil) through a series of “emanations.” The first emanation involves the passage from the One-Good to the divine Intellect (nous) which contemplates the multiplicity of intelligible forms residing within it. A second emanation involves a further passage from the Intellect to the World-Soul, whose object is the world as a whole. The division of the World-Soul into the multiplicity of embodied souls within the world represents a still further emanation or “specification.” At the bottom rung of this metaphysical ladder is bare matter, which Plotinus represents as pure Multiplicity and Evil. Multiplicity and Evil are states of privation rather than positive realities in their own right. Through their desire for material objects, embodied souls become mired in Multiplicity-Evil, but by a reverse process of education/elevation (a process Plato describes in the Symposium) they can redirect their desire toward the forms and the One-Good.
Instead of placing the One-Good beyond the Divine Intellect, Augustine argues that the God of Christianity is the ultimate source of unity and goodness. The relation between God and the world is one of creation rather than emanation. Augustine agrees with Plotinus that evil represents a privation of the good rather than a positive reality, but he locates the source of this privation not in matter but in sin. We are fundamentally good insofar as we are created by God, but we freely choose evil through perversity of will (see section 3.6). Redemption from evil is possible, but it ultimately depends upon God's grace rather than our freely turning away from evil.
Aquinas represents unity and goodness (as well as truth) as transcendentals, or trans-categorial properties. In his view, the cause of evil isn't the perversity of our wills but our misperception of the true good. He therefore agrees with Socrates and Aristotle that all desire is directed toward the good as it is conceived by the desiring agent. The source of evil volitions lies in the discrepancy between the actual good and the apparent good.
In Principia Ethica, Moore criticizes the “naturalistic fallacy,” or the mistake of thinking that the good can be defined on the basis of some natural (or even non-natural positive) property (Moore, 1993 [1903]). His basis for the critique is an “open question” argument: given any positive property or combination of properties X, it would be possible for something to be X but for it still to be an open question whether that thing is good or not. This is sufficient, according to Moore, to show that “good” cannot be defined in terms of any such positive property or combination thereof. Given this indefinability, according to Moore, truths about what is or is not good can only be known, ultimately, through a non-discursive kind of intuition. On this basis, Moore argues against a wide variety of hedonistic, utilitarian, and consequentialist theories of ethics.
In the Nicomachean Ethics, Aristotle challenges Plato's thesis about the unity of the good. Goodness consists in excellence of functioning, so there must be as many kinds of goodness as there are kinds of proper functioning (see section 4.5). For a human being, whose function consists in the exercise of rationality, the good is eudaimonia, happiness or flourishing, the highest form of which involves a life of philosophical contemplation.
Another question is whether knowledge of the good presupposes knowledge of the law or vice versa. In the Republic Socrates suggests that philosopher-kings should model the laws of the city on the Form of the Good (Plato, 1997: 1107–1108 [Republic 484c–d]. In the Critique of Practical Reason Kant reverses this line of argument. Since our knowledge is limited to phenomena, we have no cognitive access to a transcendent Form of the Good. Our conception of the highest good (conceived as virtue combined with happiness) must be derived from our prior understanding of the fundamental moral law that reason gives us (the “categorical imperative”). In Religion within the Boundaries of Mere Reason Kant further argues that human beings are “radically” evil in the sense that we have an innate propensity to subordinate our respect for the moral law to the egoistic principle of self-love. Despite this propensity, we can (and should) reverse our fundamental priorities to put respect for the law before the principle of self-love. Kant denies that any human being is diabolically evil in the sense of willing evil for its own sake. Anyone who professes to have willed evil for its own sake (as Augustine seems to do in his Confessions) must be confused about his true motives (Kant, 1996 [1793]: 84). The majesty of the moral law is such that it cannot but produce in us a feeling of respect, even when we are tempted (as we always are to one degree or another) to subordinate this feeling to pathological incentives based on self-love.
In Beyond Good and Evil and The Genealogy of Morals Nietzsche criticizes Kant's moral psychology. He develops not so much an alternative moral psychology as a psychology of morality itself. Moral philosophers such as Kant, Nietzsche argues, have mistakenly identified moral values with values in general. In antiquity, words signifying goodness were used by the nobility to indicate their approval of anything that either expressed or contributed to their own flourishing. Anything weak or harmful they called “bad.” Only later, through a “slave revolt” that privileged the perspective of the weak, did the idea of goodness take on a moral connotation. Nietzsche associates this transition with the rise of Judaism and Christianity. What had previously been regarded as good was now dubbed “evil.” Nietzsche's positive project (“beyond good and evil”) is to bring about a “transvaluation of values” that will liberate us not (pace Kant) from the perversity of evil but from the perversity of morality. To overcome morality isn't to become evil in Kant's sense of the term. In Nietzsche's view strict Kantian duty already harbors within it an element of evil that takes the form of self-directed cruelty (Nietzsche, 2008 [1887]: 47). Freud discerns a similar form of self-directed cruelty in ordinary moral experience and in pathological forms of “moral masochism” (Freud, 1991 [1924]).
In Eichmann in Jerusalem: A Report on the Banality of Evil Hannah Arendt argues that certain forms of evil result from a sheer lack of moral reflection rather than from deliberate cruelty. Adolf Eichmann was one of the chief administrators of the Holocaust, yet during his trial he came across to Arendt not as unthinkably evil, but as unthinkingly evil. In his testimony he professed to have followed a modified version of Kant's categorical imperative according to which he should always act in such a way as if the Führer would approve of his actions (Arendt, 2006 [1963]: 135–7). For Arendt, this travesty of Kant's moral philosophy shows that Eichmann was motivated not by cruelty but by blind obedience to authority. Like Nietzsche, she denies that there are transcendent moral values that can justify or condemn our actions. Instead of advocating a standpoint “beyond good and evil,” as Nietzsche does, she argues that we must create shared values in the absence of a transcendent Good. In The Human Condition, she criticizes Plato for conflating the highest object of desire – correctly identified in the Symposium as the Form of the Beautiful – with the (alleged but spurious) Form of the Good (Arendt, 1998 [1958]: 226). Like Aristotle she rejects Plato's thesis that politics can be based on the philosopher's insight into the nature of a One-Good. Only together can a plurality of citizens collectively forge workable conceptions of justice (see section 4.7).
Arendt's denial of the reality of the One-Good is echoed in the work of Alain Badiou and Slavoj Žižek. Žižek agrees with Kant that determinate conceptions of the good must be based upon a prior disclosure of a moral law or categorical imperative. He goes further than Kant, however, in emphasizing the emptiness (or “purity”) of the moral law (see section 4.4). The categorical imperative tells us that we should always act in a formally lawful manner, but it leaves it up to us to determine the law's actual content. Kant tries to fill this gap by specifying a list of “maxims” that can be deduced a priori from the categorical imperative, but Žižek (following Hegel) regards this as a futile endeavor. Every genuine act is to some extent a leap in the dark that can only retroactively posit the maxim from which it supposedly follows. Such a leap is not simply “beyond good and evil” in Nietzsche's sense; taking his cue from Kant rather than from Nietzsche, Žižek characterizes it as diabolically evil. A true act (that is, an autonomously performed act) is diabolically evil not because it flouts the law but because it flouts pre-given conceptions of the good. Countervailing efforts to specify and follow determinate conceptions of the good – or to follow pre-set moral maxims – are so many ways of protecting ourselves from the traumatic encounter with the law. Kant's failure to equate true autonomy with diabolical evil explains his strange proximity to the Marquis de Sade, an affinity noted by Horkheimer and Adorno (2002 [1944]) and the French psychoanalyst Jacques Lacan (2006 [1966]). For Žižek, Sadean cruelty represents not the “truth” of Kantian morality (as it implicitly does for Nietzsche, and explicitly does for Adorno and Horkheimer) but rather a “symptom” of Kant's failure to grasp the implications of his conception of the law's transcendence of the good (Žižek, 2012: 817).
Badiou, whose philosophy of multiplicity is predicated on a rejection of the One, likewise rejects the concept of radical evil. In his view, good and evil are both to be understood in relation to “truth procedures” that name events. To name an event is also to leave something within it unnamed – something “subtracted” from the situation in which the event is designated. The subtraction of the unnamable can be equated with the transcendence of the good. Evil, by contrast, manifests itself as, or in, the insistence on naming the unnamable element, on refusing to subtract it from a truth procedure (Badiou, 2009 [2006]: 113–19), and thus in totalizing the procedure. In his updated version of Plato's Republic Badiou equates to agathon (Plato's term for “the Good”) with Truth, and he represents justice as “immanent Truth” (Badiou, 2012: 59).
In a fragmentary text that Heidegger characterizes as “the oldest saying of Western thinking,” the pre-Socratic philosopher Anaximander suggests that it is diken (just) that things judged adikia (unjust) should die or perish (Heidegger, 2002 [1950]: 242). To know what Anaximander means by this, we would have to know what he means by dīke (justice). But this is not so easy. Heidegger criticizes a number of German translations of the passage, beginning with one by Nietzsche. Nietzsche's version makes Anaximander say that unjust things must perish because “they must pay a penalty” [sie müssen Buße zahlen] for their injustice. As Heidegger points out, however, there is nothing in the Greek text that corresponds to the concept of a penalty. Nietzsche projects onto Anaximander's text his own inherited understanding of justice as vengeance. According to Heidegger, a more attentive reading shows that Anaximander thinks dīke as “jointure” or “fit” (Fuge). Jointure (i.e., justice) is the way in which things abide together in the presencing of their being, while “dis-jointure” (adikia, Un-Fuge, injustice) is the way in which things cling to such abiding. This interpretation may be no less speculative than Nietzsche's, but it invites us to think of justice independently of legal frameworks and conceptions of value.
In Specters of Marx, Derrida argues that justice should be conceived not only in terms of ontological jointure but also in terms of ethical dis-jointure. Ethical dis-jointure is to be understood not in terms of maintaining or abiding in presence, but in Levinas's sense as the opening of the relation to the personal other (Derrida, 2006 [1993]: 26). The problem is how to maintain these two conceptions together.
In the Republic, Socrates defines justice as the harmony that results when every part of a soul or city performs its proper function. Justice in the soul is an ethical virtue, while justice in the city is a political virtue. Plato's ideal republic is an aristocracy ruled by philosophers whose souls are ruled by reason. The only conceivably just democracy would be a democracy of true philosophers, but since the majority of people are ruled by appetite this is for Plato an untenable possibility. By contrast with this, Aristotle argues that since there is no guarantee that any ruler will be ruled by reason, we should evaluate constitutions, or forms of government, by weighing the potential benefits of having good rulers against the potential harms of having bad rulers. He concludes from this that, all other things being equal, an egalitarian or quasi-democratic constitution is preferable to all other forms of government (Aristotle, 1984: 1834 [Nicomachean Ethics, 1160a–1160b]). In The Social Contract, the eighteenth-century philosopher Jean-Jacques Rousseau offers a more ringing endorsement of democracy. For Rousseau, sovereignty lies not in the will of the majority but in the general will, which Rousseau defines as the consensus that emerges after conflicting private interests have canceled each other out (Rousseau, 1997 [1762]: 60). How exactly one goes about determining the general will is nevertheless a difficult question, one that notoriously plagued the French revolutionaries when some of them sought to implement Rousseauian political principles.
Kant agrees with Rousseau that the people are the ultimate source of sovereignty, and that existing laws should express the general will. Unlike Rousseau, however, he rejects the idea that justice requires democracy. What justice requires instead is the separation of the legislative and executive branches of government that is the characteristic form of a true republic. From this perspective, Rousseauian democracy is inherently despotic, since it assigns the legislative and executive powers to the people as a whole (Kant, 1795: 325). Kant thus reconceives the general will as the collective commitment of citizens to live under civil laws (Kant, 1797: 409). While Kant doesn't grant citizens a right to revolt, he suggests that an implicit collapse of an existing form of government entitles the people to reconvene as a new constitutional assembly. Such an implicit collapse occurred in France when Louis XVI convened the Estates General (Kant, 1797: 481). When the members of the Third Estate subsequently declared themselves the National Assembly they were acting not as revolutionaries but as individuals thrown back into a state of nature. In such a predicament, it was their moral duty to give the French people a new constitution.
In A Theory of Justice, John Rawls argues that there are two fundamental principles of justice. The first states that every citizen is entitled to as much liberty as is compatible with the liberty of everyone else. This principle is somewhat analogous to Kant's conception of the general will as a legal analogue of the categorical imperative. Rawls's second principle states that everyone should have an equal opportunity to pursue social and economic goods, and that social and economic inequalities are justifiable only when their overall distribution is more advantageous to the least advantaged members of society than any other distribution would be. Together, these principles of freedom and equality define the basic framework of Rawls's conception of “justice as fairness” (Rawls, 1999 [1971]: 53 and ch. 1).
This theory has been used by Rawls and others to articulate and defend the principles of political liberalism and the modern welfare state. For Rawls's critics, however, these positions have significant shortcomings. As we saw in section 4.4, communitarians regard the principles of Kantian deontology as overly formal and too far removed from substantive conceptions of the good. Michael Sandel raises similar objections against Rawls's principles of justice. According to Rawls, when reasoning about the fundamental principles of justice we need to abstract from our actual social situation, as if a “veil of ignorance” kept us from knowing what that situation would turn out to be. For Sandel, such a power of abstraction is both psychologically impossible and ill-advised, since our core identities are social through and through (Sandel, 1998 [1982]). Other critics target Rawls's defense of the modern welfare state. For Robert Nozick, the only just state is a “minimal” state, one that leaves individuals free to pursue social and economic goods regardless of what inequalities subsequently emerge (Nozick, 1974).
For Marxists, by contrast, true justice could flourish only under communism, a form of social organization characterized by radical equality. Plato himself promotes two forms of communism: a general form, in his vision of a minimal republic composed only of citizens whose souls are ruled by reason, and a restricted form, in his characterization of the way of life of the philosopher-kings who rule the “luxurious” city. Alain Badiou and Slavoj Žižek argue that there is an essential link between capitalism and modern liberal democracy: liberal-democratic states as well as positive theories of them such as Rawls's depend essentially on the constitutive inequalities that capitalism introduces and promulgates. In their view it is thus futile to promote genuine social and economic equality within the confines of such a state. Instead, we should build upon or reconceive Marx's political vision of a “dictatorship of the proletariat.” While it is no longer possible to identify the proletariat simply with the exploited working class in a narrow sense, we can today look instead to those who have no social or economic standing within existing economic and political frameworks (what Jacques Rancière calls “the part of those who have no part” [Rancière, 1999]). Žižek acknowledges the catastrophic character of previous attempts to establish a communist dictatorship (from Stalinism to Maoism to the Khmer Rouge). But he argues that it would be a mistake to repudiate the ideal of communism (to do this would be equivalent to repudiating justice), or to conflate communism's past failures with the inherent evils and injustices of fascism. Arendt conflates the two when she represents Stalinism and Hitlerism as equivalently “totalitarian” forms of government, tracing the roots of totalitarianism back to the influence of Rousseau's idea of the general will on the French Revolution (Arendt, 1976 [1951]; Žižek, 2008; Arendt, 1991 [1963]: 311).
Even if we were persuaded that a truly communist society that maximized liberty was the most just, we might nevertheless wonder how this bears on the Aristotelian question of the benefits and harms of various forms of government or types of state. It might be argued, for instance, that although the ideal of communism itself represents the best possible form of society, the danger of its collapsing into totalitarian or dictatorial forms of state apparatus means that it should nevertheless be avoided. Yet there will always be something strange about a theory of justice that permits us to settle for something second-best (or least worst). G.A. Cohen raises a similar objection against Rawls's pragmatic toleration of social and economic inequalities despite Rawls's implicit acknowledgment that justice requires true equality (Cohen, 2008). In defense of communism, Badiou and Žižek similarly refuse to settle for the least worst. While acknowledging that future attempts to actualize the communist ideal may fail, Žižek expresses the hope that they will “fail better” than past attempts (Žižek, 2008: 210).
Many political theorists have suggested that law and coercive force are inherently and necessarily related. In Leviathan, Hobbes argues that fear is both the chief motive for constituting a body politic and the instrument by which the sovereign ensures that his subjects obey the laws he enacts (Hobbes, 1996 [1651]). Sovereignty is undermined when the fears felt in the state of nature are revived (when subjects no longer feel protected by the sovereign from internal and/or external threats) and when subjects no longer fear breaking the law themselves. Paradoxically, then, the Hobbesian sovereign uses fear to ward off fear, and in so doing claims a monopoly on it. This doctrine presupposes the traditional claim that in order to be effective the law must have coercive force. Aquinas distinguishes the law's “directive” force (its explicit prohibitions and permissions) from its power of enforcement (Aquinas, 2002 [1265]: 147). “Positive” laws enacted by rulers should be based on “natural” laws that God decrees and which human reason can apprehend. Natural law has internal coercive force in the sense that it commands us to be virtuous, i.e., to follow the law of our own free will. By contrast with this, positive law has external coercive force in that it commands us to obey the law, whether or not we are virtuous. External enforcement consists both in incentivizing people to obey the law and in punishing those who violate it.
Kant, who draws a similar distinction between internal duties of virtue and external duties of right, claims that the state's duty to punish criminals is so sacrosanct that the sovereign's “right to grant clemency” should seldom, if ever, be exercised (Kant, 1797: 477). Clemency, or pardon, undermines the law from above, just as revolution undermines it from below. Kant denies that subjects in a political/juridical community have a right to revolt, but he does grant them a right to protest against any existing laws that they regard as unjust. As Jeremy Waldron observes, Kant's claim that even unjust laws should be obeyed makes him a legal positivist of some sort (Waldron, 1996). More commonly associated with Jeremy Bentham, legal positivism is the view that the force of positive law is independent of natural law. This raises a question: if justice is a function of natural law (or vice versa), on what does the authority of positive law rest?
One response to this problem is to say that the force of law rests on nothing more than coercive force or power itself. The expression “force of law” could then be taken in two different, albeit related, senses: first, in the sense of the force that establishes law (what Walter Benjamin calls “law-constituting violence [Gewalt]”); second, in the sense of the force that enforces existing law (what Benjamin calls “law-preserving violence”) (Benjamin, 1996 [1921]). In “Force of Law,” Derrida questions whether such a distinction can be rigorously sustained. If all law rests on force, then the supposed transition from the law's constitution to its legitimate enforcement would, in a sense, never take place. Every purportedly established law would stand on the threshold between constitution and preservation, and could ultimately be legitimated only in terms of the (originally non-legitimate) force that put it in place to begin with. Along these lines, Carl Schmitt (Schmitt, 2004 [1922]) argued that the legitimacy of every sovereign juridical regime depends ultimately on an act of founding decision that is neither legitimate nor illegitimate in terms of that regime itself, a power of original constitution that must remain reserved to the sovereign power and can again be invoked in times of crisis or emergency. For partially similar reasons, legal positivists often represent laws as useful fictions. More broadly, what Montaigne and Pascal call the “mystical authority of the law” – or what we today might call the law's “mystique” or “aura” (a term that Benjamin uses in another context) – itself may be seen as having a fictional structure (Derrida, 2002). Bentham, the first legal theorist to codify the principles of legal positivism, did so in the context of proposing an entire theory of socially constructed fictions.
Instead of appealing to legal fictions, H. L. A. Hart traces the force of existing law to custom, or the general practice of following socially accepted rules (Hart, 2012 [1961]). Such rules needn't be based on natural law, Hart argues; they simply must be generally accepted and acted upon. Thus law exists because we collectively recognize it as existing. On this position, our recognition of it is a function of our actual practices rather than our theoretical commitments, and the force of law is the force of our inertia in maintaining these practices. In principle, however, we could transform it at any time.
Just as we can question the force of positive or natural law, so we can raise questions about the force of laws of nature. One important difference between natural law and laws of nature is that the former tells us what we are practically necessitated to do (what we ought to do), while the latter tell us what we (and other things in nature) are theoretically necessitated to do (such as “obey” the law of gravity). In section 4.3 we discussed Agamben's observation that there is always a gap between a practical law and the cases to which it applies. A similar gap pertains to the relationship between laws of nature and their cases. This gap arises not because there are always exceptions to general rules, but because the lack of exceptions to uniform regularities doesn't suffice to characterize them as law-governed. At issue is not just the epistemological problem of induction made famous by Hume, namely, that no matter how many times events of type A have been succeeded by events of type B, we can never be sure that the next A-type event will be followed by an event that isn't of type B. At issue is a further metaphysical question about the kind of necessitation (or force) that laws of nature are supposed to have.
This puzzle is often posed in terms of possible worlds. Consider two possible worlds, W1 and W2, in both of whose entire histories every A-type event is followed by a B-type event. In both worlds, there are no exceptions to this pattern. Suppose, however, that in W1 the regularity exists “just by chance,” while in W2 it exists because a law of nature necessitates that A-type events are always followed by B-type events. In what does this difference consist?
One response to this question is to deny that there is any difference at all. A law of nature may be nothing more than a uniform regularity, in which case it was meaningless for us to distinguish the situations at W1 and W2 in the first place. Perhaps we were misled by the idea that so-called “covering” laws of nature “force” events to happen in regular ways. This response has a Humean flavor in that it calls into question the very idea of necessitation in nature, whether conceived as causal necessitation or as nomological necessitation (i.e. necessitation based in law in a basic sense). However, David Lewis offers an alternative response that is no less Humean in inspiration. Developing a suggestion first made by Frank Ramsey, Lewis argues that we can distinguish contingent generalizations from laws of nature by identifying the latter with the true statements we could logically deduce from fundamental axioms if we knew all of the facts that there were to know and could organize them in the simplest available logical system (i.e., a system with the simplest axioms) (Lewis, 1973: 73).
Contra Lewis, D. M. Armstrong argues that the relation of nomological necessitation holds between universals. If it is a genuine law that all Fs are Gs, this must be because the relation of necessitation holds between F-ness and G-ness (Armstrong, 1983: 85). Lewis objects that this account fails to explain what type of necessity links the universals. Armstrong has simply shifted the mystery of primitive modality (in this case, an unanalyzed conception of necessity) from particulars to universals (Lewis, 1999 [1983]: 40).
What about the force of the (“normative”) laws of logic? Aristotle identifies three: the law of identity, the law of non-contradiction, and the law of the excluded middle. It is reasonable to think of such laws as even more constraining than laws of nature in the sense that they apply to a wider range of possible worlds (perhaps to all). Unlike laws of nature, laws of logic (like moral and political laws) have directive or normative force: they determine not how we actually think, but how (in some sense) we ought to think. Once again, though, we could ask whether it makes sense to ask whether in one of the two worlds it just so happens that people think in accordance with the laws of logic, whereas in the other world they “obey” them. Moreover, there are questions about the sense in which logical laws are normative at all: if, for instance, the law of non-contradiction tells me to avoid believing or thinking contradictions, we can legitimately ask the further question of why we should do so. Any claim that belief or thought of a contradiction will lead us to think something false seems to presuppose that the law is not just normatively binding but also descriptively true, i.e. that there are (in the world) no contradictory facts or states of affairs. But it is unclear how this can be established independently of the law itself. Finally, we may wonder if it makes sense to speak of necessary or ineluctable laws of logic at all, given that it is possible to specify a variety of alternative logical systems, including intuitionist logics that suspend the law of excluded middle, dialectical logics that suspend or modify the law of identity, and even dialetheic logics that suspend the law of non-contradiction.