Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

How (and If) Law Matters

Responding to Mark Greenberg, How to Explain Things with Force

Mark Greenberg’s deep and thoughtful review of The Force of Law1×1. Mark Greenberg, How to Explain Things with Force, 129 Harv. L. Rev. 1932 (2016) (reviewing Frederick Schauer, The Force of Law (2015)). flatters me in two ways. Of minimal importance is Greenberg’s generous appraisal of the book and the issues it seeks to put on the agenda of legal theory. Much more valuable is the review’s challenging critical engagement with several of the book’s major themes. In many places Greenberg’s critique is on target and has caused me to rethink some of my arguments and the ways in which I expressed them.2×2. I offer just two among a larger number of examples. First, in arguing for the respectability (and not necessarily the all-in correctness) of a nonessentialist view of concepts generally and the concept of law in particular, Schauer, supra note 1, at 3–4, 35–41, I place excess and faulty weight on a prototype or core-and-fringe picture of concepts. As Greenberg properly points out, such accounts are not inconsistent with the core cases having essential properties. Greenberg, supra note 1, at 1945–46, 1946 n.16. This is not the case with cluster or family resemblance concepts, as I discuss below in Part II of this response, but Greenberg is correct to maintain that nothing in an essentialist account is inconsistent with concepts having core and penumbral applications, to use H.L.A. Hart’s preferred term, H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 Harv. L. Rev. 593, 607 (1958). Second, Greenberg correctly observes that the distinction between moralized and nonmoralized accounts of coercion, Greenberg, supra note 1, at 1941 n.10, is more relevant to thinking about the nature of law than The Force of Law acknowledges. Too much of the book treats the literature on coercion in criminal law, contracts, torts, and morality, see generally Alan Wertheimer, Coercion (1987) (developing a theory of coercion across a range of legal contexts), as being largely about a different topic, but Greenberg persuades me that my attempted separation of the two different coercion literatures was a mistake. But in other places there remain interesting disagreements, and delving more deeply into two of them — first, the relationship between the law’s ability to provide reasons for action and the existence (or not) of a moral obligation to obey the law; and, second, the nature and role of conceptual analysis in jurisprudential inquiry — may enable us to move a bit further down the path of understanding the phenomenon of law.

I. Law’s Reasons

At the heart of Greenberg’s analysis is his claim that law can adjust our moral calculus (or “moral profile,” as he puts it) by providing morality-based reasons for action that did not exist absent the law.3×3. Greenberg, supra note 1, at 1964. And he maintains that the manner in which law adjusts our moral reasons is different from and not reducible to the venerable claim that there is a (prima facie) moral obligation to obey the law.4×4. Id. at 1963–64. I will address these two aspects of Greenberg’s argument in turn.

A. Two Concepts of Reasons

Greenberg and I (and pretty much everyone else) agree that people can have law-independent moral reasons for action.5×5. See, e.g., Joseph Raz, Practical Reason and Norms 36 (1990) (describing decisions made on the “balance of reasons”). But when law enters the picture, Greenberg argues, the moral landscape changes in at least two ways.6×6. Greenberg, supra note 1, at 1958–71. First, law may provide an epistemic guide to what these law-independent moral reasons are and how they apply to the case at hand. In an analysis that resembles (but is not identical to) Donald Regan’s idea of indicative (as opposed to intrinsic) reasons for action (or “indicator-rules”)7×7. See Donald H. Regan, Reasons, Authority, and the Meaning of “Obey”: Further Thoughts on Raz and Obedience to Law, 3 Canadian J.L. & Juris. 3, 7–11, 13 (1990); Donald H. Regan, Authority and Value: Reflections on Raz’s Morality of Freedom, 62 S. Cal. L. Rev. 995, 1003–18 (1989). and David Enoch’s related conception of triggering reasons,8×8. See David Enoch, Reason-Giving and the Law, in 1 Oxford Studies in Philosophy of Law 1 (Leslie Green & Brian Leiter eds., 2011). Greenberg argues that the law might serve as a sort of proxy, helping to identify and apply moral reasons in the face of informational, cognitive, or motivational limitations.9×9. Greenberg, supra note 1, at 1963–64. The law on this account does not provide a new reason but, instead, indicates to the puzzled, the confused, or the weak-willed what they should do in their application of the law-independent moral reasons they actually have. I might believe, for example, that I should drive safely but might not know on a strange road just what speed is safe. And so when I see a sign telling me that the limit is forty miles per hour, the law tells me what is in fact safe. The law indicates or points to the correct course of action, but it does not create the reason to drive safely. The law simply helps us to identify and apply the law-independent reason. Similarly, I might be aware of my tendency to drive unsafely even though I know I should not. And so I follow the speed limit because it compensates for what Greenberg nicely refers to as “motivational” deficiencies.10×10. Id. at 1964.

I have no quarrel with this aspect of Greenberg’s picture of how law might make a difference in our reasoning processes. Nor do I disagree with his analysis of the other ways in which law might alter the terrain of our moral reasoning. Law might, for example, operationalize, institutionalize, or make more specific our abstract moral reasons. We have a good moral reason, say, to make things pleasant for those with whom we share the sidewalks and the parks, and the legal requirement that we clean up after our dogs gives concrete manifestation to the abstract requirement. Moreover, Greenberg is surely correct in his lengthier and important picture of how law can operationalize or institutionalize a cooperative “scheme” whose reason-giving capacity is independent of the law.11×11. See id. at 1965–70.

Greenberg has thus performed a valuable service in illustrating how law might provide second-order adjustments to our first-order moral calculus, but a question — my question — remains about whether law actually, as an empirical matter, does what Greenberg claims it has the ability to do. And in order to address this question, we need to distinguish between two different senses of the word “reason” and two different concepts of reasons. The sense that Greenberg assumes — and the one that dominates the literature in moral, legal, and political philosophy — understands a reason as a justified basis for doing something.12×12. E.g., id. at 1933 (referring to “moral reasons” and “morally relevant factors”). I have a reason, for example, not to lie, not to commit assault, and to come to the aid of people in distress. And so does everyone else. From a utilitarian perspective, we might say instead that I have a reason to take those actions that would produce the greatest aggregate welfare. And, again, from a utilitarian perspective, so does everyone else.

In another sense, however, to have a reason is actually to be motivated in a certain way.13×13. See Raz, supra note 5, at 36 n.* (distinguishing the reasons that are relevant to a decision from the reasons that an agent in fact considered). When someone asks why I am carrying an umbrella and I respond that the weather forecast predicted rain, my statement of a reason is an explanation of what in fact motivated me, what reasons I in fact took into account.14×14. Thus, a common dictionary definition of “reason” is “[t]he basis or motive for an action, decision, or belief.” Reason, Webster’s II New Riverside University Dictionary (1984). But a reason in this motivational sense need not be a good reason, and thus not a reason in the sense assumed by Greenberg. And a reason in the first sense might not actually motivate some agent, or even be recognized by that agent. Thus, if someone who engages in animal cruelty — promoting dog fights, for example — says he does so because the activity is profitable or because it gives him pleasure, he is giving his reason in this latter sense for doing something, even if in the former sense it is a bad reason or simply no reason at all. And thus when the law imposes a requirement of reason giving on a decisionmaker,15×15. See, e.g., K.W. ex rel. D.W. v. Armstrong, 789 F.3d 962, 974 (9th Cir. 2015) (describing due process basis for requirement to give reasons); Murphy v. Astrue, 245 Fed. App’x 602, 603 (9th Cir. 2007) (applying statutory requirement that administrative law judges give reasons for denial of Social Security disability benefits). the decisionmaker is required to describe her motivations — to state the actual cause of the outcome she has reached. The legal requirement to give a reason for a decision is accordingly satisfied by offering a descriptively accurate account of the decisionmaker’s motivations, even if these are substantively poor motivations. The requirement to provide a reason is thus analytically distinct from a putative requirement to have (or not have) a reason of a certain kind.

This distinction between reasons as motivation-independent good reasons and reasons as motivations explains much of what might appear as a disagreement between Greenberg and me. Greenberg is interested in the former, and he tells us how law can provide reasons for action.16×16. Greenberg, supra note 1, at 1932–33. But in The Force of Law I am principally interested in the latter — that is, in the law-produced motivations that people actually have. Assume for the moment that there is — as Socrates, John Locke, John Rawls, and many others have insisted — a prima facie moral obligation to obey the law.17×17. See Plato, The Trial and Death of Socrates (John M. Cooper ed., G.M.A. Grube trans., Hackett Publ’g Co. 3d ed. 2000); John Locke, Two Treatises of Government 330–32 (Peter Laslett ed., Cambridge Univ. Press student ed. 1988) (1689); John Rawls, Legal Obligation and the Duty of Fair Play, in Law and Philosophy 3 (Sidney Hook ed., 1964). The best historical overviews of arguments for the prima facie moral obligation to obey the law come from two scholars, A. John Simmons and M.B.E. Smith, who deny the existence of such an obligation. See generally A. John Simmons, Moral Principles and Political Obligations (1979); M.B.E. Smith, Is There a Prima Facie Obligation to Obey the Law?, 82 Yale L.J. 950 (1973). But even so, some, many, most, or even all people might not recognize and comply with that obligation, just as some, many, or most people do not recognize and comply with their obligation — their reason — not to lie. Those people would be wrong, given our assumption, in not doing so — in not recognizing and complying with their prima facie obligation to obey the law — but the wrongness of their nonrecognition of this actual obligation and actual reason (in the first sense) would be entirely consistent with the empirical fact of their nonrecognition.

The Force of Law is substantially concerned with this empirical issue. And thus one of the book’s pervasive claims, supported by various forms of empirical and experimental data, is that obedience to law qua law, whether for reasons of actual obligation to obey the law because it is law, or for the kinds of reasons that Greenberg illuminates, or simply because of irrational rule fetishism, is in reality less common than many people and many theorists believe. Once we understand what it is to follow the law because it is the law, or just what it is to have our reasons for action affected in the ways that Greenberg describes, it turns out that such behavior is less common than is often supposed. And if this is so, then we have located an explanation for why coercion appears to be so pervasive in actual legal systems.

The biggest disagreement between Greenberg and me is thus about whether we have a disagreement. Greenberg believes that I disagree with the ability of law to affect our reasoning in the way that he develops. But I do not. Rather, I am doubtful that law actually does affect the reasoning processes of actual people in the ways he rightfully claims is possible. Even if I agreed with every word of his analysis, which may not be that far from the truth on this point, I could still — and do — argue that what can happen, and what should happen, does not in fact happen. My project is thus more empirical and contingent than Greenberg’s, and if Greenberg errs, he errs only in assuming that my project is his.

B. Moral Reasons and the Moral Obligation to Obey the Law

Greenberg’s project is, as just explained, more normative than mine, yet even within the province of the normative he stresses that the way in which law can adjust our moral calculus or moral profile differs from the question whether there is a (prima facie) moral obligation to obey the law. As Greenberg puts it, “the law can alter morally relevant factors, thus generating moral reasons, . . . without people internalizing legal norms or believing that they have a moral obligation to obey the law.”18×18. Greenberg, supra note 1, at 1933. But even if we remain in Greenberg’s domain of the “can” — rather than my preferred domain of the “is” or the “does” — it is still less clear to me than it is to him that law’s alteration of the morally relevant factors is independent of the existence of a moral obligation to obey the law. And this is so even if one agrees with Greenberg, as I do, that law can alter the moral profile in the manner he posits.

Greenberg several times describes the idea of a general obligation to obey the law as a “fetish,”19×19. Id. at 1933, 1962–64. but Socrates, Locke, Rawls, and their intellectual heirs do not argue that people should engage in fetishes. Rather, they claim that the very fact of law provides a moral reason for action, whether because of the social contract, or because of an obligation of reciprocity or fair play, or because of moral obligations to participate in the nationwide cooperative enterprise we label as law.20×20. On this last basis for legal obligation, see Gerald J. Postema, Coordination and Convention at the Foundations of Law, 11 J. Legal Stud. 165 (1982). But the basic idea is that the legal provenance of a directive provides a content-independent reason to follow it — a reason distinct from the content of what the particular directive is doing on this occasion, or even does more generally.21×21. On content independence, the classic analysis is H.L.A. Hart, Commands and Authoritative Legal Reasons, in Essays on Bentham 243, 254–66 (1982). References to additional discussions of content-independent reasons can be found in Frederick Schauer, Thinking Like a Lawyer 62 n.6 (2009). Uncharitably, we might say that Socrates is telling us why we have a moral reason to fetishize law. Charitably, we might say that to have a moral obligation to obey the law is simply not to look at whether some law is a good law, but to treat its very law-ness as a good moral reason. If this is to fetishize law, then so be it.

Greenberg argues that law can change our moral profile — the array of moral reasons we accept — in a different way. Even if we reject a general moral obligation to obey the law, the fact of law might make a difference “in particular cases” because it would give people good reason to “participate” in a law-created cooperative “scheme.”22×22. Greenberg, supra note 1, at 1965–68. But we know that people participate in many cooperative schemes that exist apart from or outside the law. After an automobile accident, for example, a Good Samaritan may stand in the road to direct traffic away from the accident. Here the law is absent, but there is still a good reason to obey the instructions of the Samaritan.

The question then is whether, sanctions apart, people would have a better or additional reason to follow the same instructions in the same circumstances from someone acting with the authority of the law. Perhaps the fact of law should make no moral difference, but Greenberg appears to suggest otherwise: that the law has created the scheme should make a moral difference. But if it does, it is because people are treating the law-ness of the directive as morally relevant, and it is here that the difference between Greenberg’s picture and the idea of a moral obligation to obey the law begins to evaporate. If the fact of law makes a moral difference, it is because the subject now has an additional or a heightened reason because of the law, and it is not clear that this idea is very different from saying that the fact of law creates distinct or heightened obligations.

A potential response to this argument is that law could create the very behavior just described. The law might, for example, require people to act as Good Samaritans. If that were so, then the moral reason to follow the Samaritan’s directions would be a product of the law having altered the moral profile of its subjects. But now suppose that the subject believes the Samaritan’s directions to be mistaken. Under these circumstances, should the fact that the Samaritan is clothed with the authority of law make a moral difference? Perhaps the answer is yes, but if that is the case then it remains difficult to see why this is not the same as having a moral reason to treat the fact of law as morally relevant to our obligations, which seems very close to saying that we have a defeasible moral obligation to follow the law.

Greenberg might now respond that the difference between his form of law-created moral relevance and the moral obligation to obey the law is the difference between a moral obligation to obey all law and an obligation to obey a particular law-created cooperative scheme. But participating in a “scheme” is different from making an entirely case-by-case decision about what to do. If we participate in a scheme, we do what the scheme tells us to do even if our judgment is that it would be better to do something else on a particular occasion. That is what makes a scheme a scheme. And thus it turns out that the purported difference between the moral obligation to obey the law and the moral reason to participate in a law-created cooperative scheme is entirely a matter of degree. If we can imagine the moral obligation of a Norwegian to obey all of the laws of Norway but not to obey the laws of another nation, then Norwegian law is simply a (relatively) particular scheme of just the kind that Greenberg describes. Yes, an obligation to obey all of the laws of all nations regardless of the circumstances of that nation’s law-creation would indeed be a fetishization of law, but no one who has argued for a moral obligation to obey the law just because it is the law has taken the obligation to exist at this level of generality. And once we particularize the obligation to particular nations or particular legal systems, then Greenberg’s claim is simply that the obligation should exist at the level of a particular scheme rather than at the more general level that is the nation’s entire legal scheme. This might be a plausible distinction in particular times or places, but it is less different from the standard claims of a moral obligation to obey the law than Greenberg imagines.

II. On the Enterprise of Characterizing Law

Greenberg correctly observes that questions of jurisprudential methodology play a significant, even if preliminary, role in The Force of Law. Some of these questions are ones about the focus of modern jurisprudence, and Greenberg believes that I may have exaggerated the lack of attention to coercion in the contemporary jurisprudential environment. But this dispute, if it is a dispute at all, cannot usefully be resolved here, or for that matter anywhere else. The parable of the Blind Men and the Elephant reminds us that perception is partial, and it is common or inevitable to overemphasize the importance of that part of the whole that we happen to have seen. Neither Greenberg nor I are immune from such distortion, and so it seems wise to let readers determine how the jurisprudential terrain appears to them.

Of more interest, however, is the question of what it is to examine the nature of a phenomenon, including the phenomenon of law. Most of modern jurisprudence focuses on the task of attempting to say something interesting about the nature of law itself, whenever and wherever it may appear, and not so much about the features of law in particular legal systems or legal cultures. And those who seek to apply philosophical approaches to this question have often, especially recently, engaged in debates about the methodologies that might be deployed in addressing the question of the nature of law. And because some philosophers of law believe that the (or a) task of the philosophy of law is to identify the necessary properties of law — something that law must have in order to be law23×23. See, e.g., Scott J. Shapiro, Legality 9 (2011) (arguing that “to discover the law’s nature” is “to discover its necessary properties, that is, those properties that law could not fail to have”). — a significant part of The Force of Law is devoted to arguing that we may learn more about the phenomenon of law if we focus not on its necessary properties but instead on its typical ones, recognizing that to be typical is not to be necessary. What makes this methodological issue important in the book is my claim that coercion may be important — and jurisprudentially important — even if it is not strictly necessary of all possible legal systems in all possible worlds, and thus not a property that law could not fail to have.

Greenberg takes issue with some of my characterizations of the character of law, and of what it is for law to have a character (or nature), but here again the differences between us may be less than they seem. Although Greenberg offers some sound and useful technical corrections to my characterizations of what it is to inquire into the nature of a phenomenon such as law, in the final analysis he appears to agree that using the tools of traditional philosophical conceptual analysis may not help us say much of interest about law or even about the nature of law. And he appears to agree as well that the items “picked out” by the term “law” might be such a “heterogeneous collection of items” that trying to generalize about them would not be a fruitful enterprise.24×24. Greenberg, supra note 1, at 1949. As against theorists who believe that the major task of jurisprudence is using the tools of conceptual analysis to say interesting and valuable things about the nature of law, that is to use conceptual analysis to discover the necessary properties of law, Greenberg and I are, albeit in different ways, united in our skepticism.

But now things become more interesting, for there may be two ways in which Greenberg’s concessions (if that is what they are) do not go far enough. First, although he acknowledges that the word “law” may pick out too heterogeneous a group of instances to be the fruitful object of conceptual analysis, much the same may apply to the subsets of the full set that is picked out by the word “law.” There is, I suppose, a nature of “blond dentists living in Nebraska,” but that nature might be a composite of multiple natures (blond, dentist, and Nebraskan) or might simply say nothing of any interest. Insofar as there are things in the world — blond dentists living in Nebraska, for example — whose nature we wish to understand, our understanding may not be very much (or at all) assisted simply by listing the characteristics of the categories that may happen to intersect in particular instantiations.

If this is so about blond dentists living in Nebraska, then much the same might be true of law, even though “law” is one word and “blond dentists living in Nebraska” is five. Greenberg observes, properly, that the word “law” itself is overinclusive, encompassing things — the laws of physics, for example — far removed from the social phenomenon that is the object of jurisprudential inquiry. But even if the word “law” is overinclusive in the way that Greenberg suggests, the subset of that broad array that jurisprudes tend to study might still be almost as heterogeneous as the full set of instances picked out by the word “law.” That is, even the category of items that is the focus of jurisprudential inquiry might have little more of an essence or nature than the category, however real it might be, of blond dentists living in Nebraska. If so, probing the nature of the category — the jurisprudentially interesting subset of items picked out by the word “law” — although technically possible, might again be neither interesting nor illuminating.

Moreover, it is not even clear what kind of category the word “law” or its relevant subsets designate. Greenberg notes very briefly my mention of family resemblance concepts,25×25. Id. at 1951. but The Force of Law’s discussion treats more extensively than he suggests the possibility that law might be a family resemblance or cluster concept whose nature cannot be captured by anything we think of as an essence, even accepting the distinction between the essential and the necessary. Ludwig Wittgenstein famously, and controversially, used games and the word “game” as an example of a family resemblance, and the idea of a cluster concept, as developed by Max Black and John Searle, is more or less the same idea.26×26. See the discussion in Schauer, supra note 1, at 38–39, 172 n.10. And thus the warning — not a definitive conclusion — in The Force of Law is that we ought not too quickly to assume that the category even legal theorists designate as “law” has any more of an interesting nature or essence than the category encompassed by the word “game.”

Even more important is considering just what we are doing when we attribute a property or quality to a class of individuals. When we attribute a quality to law — whether that quality be coerciveness, the union of primary and secondary rules, or even the kind of purpose that Greenberg and others identify — we might simply be making an attribution of a quality to a category of particulars, as when we say that Volvos have the quality of reliability, or that mosquitoes have the quality of being disease-bearing. Greenberg passes by my suggestion that the qualities of law may be of this variety — the category law may be a generic, in modern philosophical terminology27×27. See id. at 39–41. — but it may be that the entire enterprise of trying to say interesting things about the nature of law presupposes (or at least tolerates) law being a category of this variety. This conclusion is not inconsistent with one of the generic properties being the kind of purpose that Greenberg usefully develops, but neither is it inconsistent with coerciveness being of the same variety. And thus when Greenberg says that “Schauer is on to something important when he suggests that a property can be part of the nature of law even if there are legal systems that lack that property,”28×28. Greenberg, supra note 1, at 1953. he punctuates what seems to me the important methodological agreement between us. And when this idea, regardless of the label we use to identify it, is applied to the coerciveness of law, this, in the final analysis, is what The Force of Law is directed principally at demonstrating.


* David and Mary Harrison Distinguished Professor of Law, University of Virginia.