This piece is a response to an article by Professors Baude and Sachs published in the Northwestern University Law Review. It expands upon the conversation begun by Professor Barzun’s prior article published in the Stanford Law Review.
The philosopher of science Ernest Nagel once observed that “[t]here is, in general, little intellectual nourishment to be found in rebuttals to rejoinders to replies.” That is especially true when there are other able critics ready to carry the mantle. But, although Professor Eric Segall in a recent essay raises powerful objections to the argument Professors William Baude and Stephen Sachs advance in Grounding Originalism, he gives them a pass on what seems to me the key issue. Segall concedes that the originalism Baude and Sachs offer “might be our law in some academic, theoretical sense”; however, he insists that even if so, originalism does little work “on the ground” in the practical business of courts deciding cases. There, political ideology dominates.
Segall may be right about that. Originalism may only be a “rhetorical device” judges deploy. But law is a rhetorical practice through and through, so it matters (to me, at least) what arguments are advanced and what stories are told. Like Baude and Sachs, I care about whether there is an “official story” of American law and, if so, what it is, what purpose it serves, and how it relates to what we call “law.” This brief response thus presses them on their claims about law in this “theoretical sense.”
In my initial critique of the novel brand of originalism Professors Baude and Sachs had advanced, both separately and together, my objection was not so much that their theory was wrong but rather that it depended on conceptual claims and normative commitments they refused to acknowledge. Their response and further elaboration of their theory in Grounding only partially remedies those defects. They now explicitly align themselves with H.L.A. Hart’s theory of law, which usefully focuses the debate. But they still fail to reckon with the difficulties Hart’s model of law poses for their approach. In fact, their somewhat creative interpretation of Hart further confirms my initial observation that their true jurisprudential ally is Ronald Dworkin. It also points to another feature of the “deep structure” of law that Baude and Sachs mostly ignore.
Below I expand upon and clarify these claims. Part II briefly recapitulates the debate between Baude and Sachs and me. The next two Parts each address one of the two core issues that divide us. The first issue is about American judicial practice, and the question is, how originalist (in their distinctive sense of that term) is our practice? The second issue is about the relationship between judicial practice and “law,” according to H.L.A. Hart’s theory of law. There the question is, can originalism still be our law even if no judges recognize that fact? Part V concludes by suggesting that Baude and Sachs’s use of Hart’s theory reveals the way in which both legal practice and theory is interpretive in a technical sense of that term I will explain.
II. “Previously on . . .”
Baude and Sachs each initially offered their theories of originalism separately, but because each stressed the way in which his approach was premised on a “positivist” understanding of law, their common approach was dubbed the “positive turn” in originalism. The basic argument proceeded in two steps. First, they suggested that scholarly debates over the proper methods of constitutional and statutory interpretation, including originalism, could be resolved, or at least advanced, by looking to our actual, “positive” law. Rather than defending originalism (or any other interpretive approach) by reference to normative claims about what would lead to good results or would be democratically legitimate, or to conceptual claims about what, for example, the act of “interpretation” necessarily requires, we should look to “our law.”
The second step was to say that if you take this approach, what you see is that our law requires originalism as a method of constitutional interpretation. More specifically, it required a particular version of originalism, which Professor Sachs had dubbed “original-law originalism” (herein “OLO”). According to this doctrine, rules or methods for interpreting statutes and the Constitution qualify as law if and only if they either were the law at the time of the Founding or were changed pursuant to a “rule of change” that can itself trace its lineage back to the Founding.
My main objection to the so-called “positive turn” was that its advocates purported to remain agnostic on the question of which positivist theory of law was the right one, whereas it seemed to me that they could not remain so agnostic since different positivist theories pick out different social facts in determining what counts as law. Professor Mark Greenberg made the same point in specific reference to their joint article on the “law of interpretation.” Moreover, deciding which theory of law to rely on requires them to engage in precisely the kind of normative argument and conceptual analysis they seemed inclined to avoid—a point I demonstrated by applying their approach to three of the best-known positivist theories of law in order to show the difficulties such application necessarily entailed.
Although Baude and Sachs continue to talk about the lack of any need to “solve jurisprudence,” they effectively concede Greenberg’s and my point by aligning themselves with Hart’s philosophy of law. So that’s some progress. But we also both argued that Hart’s theory was a bad fit for them. I made two arguments. First, courts do not seem to apply OLO when deciding cases. Second, even if some courts do so, for Hart, the existence of law requires a judicial consensus as to the criteria of legal validity, and yet there is no such consensus around the statutory and constitutional interpretive methods under discussion. Therefore, such methods cannot be part of the rule of recognition.
Baude and Sachs have still not met those essential objections, though they have clarified their strategy for doing so. First, they argue that courts need not establish the pedigree of interpretive rules back to the Founding if Founding law itself treated interpretive rules as a form of customary law. Second, they insist that interpretive disagreement among judges is not a problem for Hart’s theory because it may be that some judges are simply mistaken about what the rule of recognition—or, what they call the “official story”—properly requires. In fact, all judges might be so mistaken such that there exists “global error” as to what the rule of recognition demands.
We can thus state fairly crisply the issues on the table. Once we wade through discussions of “dephlogisticated air,” felonious sandwiches, Pteropus fruit bats, the Territory of Alaska, Al Capone’s nephew, and endangered elephants, there remain two, core issues that divide us. The first issue is whether the central requirement of OLO, namely the demand that constitutional and interpretive rules and methods trace their pedigree back to the Founding, is a feature of judicial practice today; that is, do courts explicitly or implicitly demand (even if only indirectly) satisfaction of the OLO rule? They think they do; I think they do not. The second issue is, even if the OLO rule is not a feature of judicial practice, can it still be part of our “positive law,” according to Hart’s theory of law? They think it can be; I think it cannot be. Or, to put it more precisely, Hart’s theory only allows for such a profound disconnect between judicial practice and the “law” under a creative reinterpretation of that theory. I consider each of these issues in turn.
III. Do Judges Practice Original-Law Originalism?
In my earlier article, I considered first the possibility that OLO was part of our rule of recognition. Since Hart defines the rule of recognition as a practice of courts and officials, I pointed out that in most of the cases they relied on in support of their claim that judges practice OLO, the Court made no effort to trace the pedigree of an interpretive rule or method back to the Founding. I also observed that it was easy to find cases where the Court explicitly employs an interpretive method and only bothers to cite a few cases in support. I offered these cases as evidence of what I took to be a banal observation, namely that courts do not today, nor have they ever, required interpretive rules to prove an originalist pedigree (or any other sort of pedigree) before invoking them. Thus, OLO is not part of our rule of recognition. Courts are not methods-originalists.
I then suggested that a more plausible explanation for why courts cite only a handful of cases (if any) when invoking interpretive methods is that they treat interpretive rules as deriving from judicial practice itself—in other words, as a form of customary or common law. Baude and Sachs now respond by saying that that view is in fact perfectly consistent with OLO if, but only if, such a customary practice was the law at the timing of the Founding. Here is the key passage:
Customary law was itself a well-recognized kind of law at the Founding, and the Founders’ law likely included the possibility, within certain fields, of legal “evol[ution] by slow accretion.” . . . We see such customary fields, together with any limits on their potential evolution, as possible contingent implications of our legal system’s official story.
By contrast, Barzun proposes customary law as an alternative to our view, requiring confidence that it would still be compatible with American legal norms regardless of whether there had been such law at the Founding—apparently, even if the modern use of customary law in a given field were entirely the product of some mid-century judicial power grab. That, we think, is a much taller order, at least as judged by our existing legal culture.
Let’s be clear about what the issue here is. We are now agreed that courts treat canons of construction and other interpretive methods as a form of customary law, so that those methods do not themselves require a pedigree back to the Founding; one must only establish that the custom of treating them as customary traces back to the Founding. So the question is, do courts today apply something like the following rule?: In order for Interpretive Method X to be validly invoked by a court, that canon or method must either establish a pedigree back to the Founding or derive from judicial custom that is itself traceable back to the Founding.
This is an interesting question, and any reader of this Essay, or at least any domestic one, is probably well-equipped to help answer it. Anyone who has read this far in an essay about abstruse questions of constitutional theory and jurisprudence is likely either to be an expert (or expert-in-training) on what the rule of recognition in the United States is (i.e., what all or nearly all courts treat as law) or someone whose view on the matter is actually constitutive of the rule of recognition. So, what do you think, dear reader? If you discovered that at the time of the Founding the law did not treat canons of construction and other interpretive rules as amenable to common-law development in spite of the fact that many of the canons and interpretive methods courts use today had entered our law through precisely such a process, would your judgment as to the “legality” or judicial propriety of those methods be materially altered?
In answering this question, though, do not be misled by the false dichotomy Baude and Sachs (once again) draw between a pedigree-to-the-Founding rule, on the one hand, and a “mid-century judicial power grab,” on the other. For one thing, almost by definition, if a legal doctrine is plausibly characterized as a “power grab,” then it will not be plausibly understood as having been properly derived from custom. More important, the contrast drawn misstates the burden their theory must meet. To say that OLO is part of the rule of recognition is to say that our current practice treats having a pedigree back to the Founders’ law as a necessary condition of legal validity, which means that even if the customary features of an area of law developed slowly throughout the nineteenth century, it would still be inconsistent with OLO and so not valid law.
Even with this clarification, my question is not meant to be rhetorical. I suspect some readers with originalist inclinations may answer in the affirmative—such a discovery might indeed make a big difference to them. But my bald, speculative guess is that a good many readers would not care much at all. These readers may understand our system of law to bottom out in a common-law tradition of legal evolution and judicial practice, so that the legal status of customary law does not critically depend on proving that the founders approved of its use in some particular domain—especially when it comes to methods for interpreting statutes and the Constitution.
If I am right that the opinions of judges might vary on this question, then that would seem to be enough to establish my negative claim that OLO is not our law. Remember that in order for a rule to be part of the rule of recognition, all or nearly all judges must accept it as a criterion of legal validity. So, you might think that, unless virtually all judges agree that the use of customary law is only valid if the law at the Founding authorized its use in that particular domain, Baude and Sachs have failed to show that OLO is part of our rule of recognition. Where there is disagreement, there can be no rule of recognition.
IV. Can Positive Law Exist in “Exile”?
You might think that, but you’d be wrong; Baude and Sachs have a response to that line of reasoning. Maybe OLO is not itself part of our rule of recognition at all. Instead, it achieves the status of “law” by satisfying the criteria that that rule of recognition provides. In Hart’s terms, OLO is a “subordinate” rule of our legal system. I will first summarize and explain this response and then show why it still places more weight on Hart’s theory than it can bear, even under a charitable reading.
A. Original-Law Originalism as a Valid Subordinate Rule
The argument goes like this: the rule of recognition in the United States is what Baude and Sachs call “the official story” of American law, which includes things like: “We treat the Constitution as a legal text, originally enacted in the late eighteenth century,” and “Actors in our legal system don’t acknowledge, and indeed reject, any official legal breaks or discontinuities from the Founding,” and the other apparent truisms about American legal practice they list in Grounding. Because the vast majority of judges would accept these truisms as accurate statements about the fundamental assumptions of our legal practice, they constitute—or at least “help guide us toward”—our rule of recognition.
The official story, as recounted by Baude and Sachs, nowhere includes an explicit requirement to the effect that “no method of constitutional or statutory interpretation is legally permissible unless it was part of Founding law or produced by a rule of change ultimately authorized by Founding law” (i.e., the OLO rule). But that’s not a problem because that rule is a mere subordinate rule (in Hart’s terms) entailed by, or derived from, the official story, which itself provides the ultimate criteria of legal validity. The OLO rule follows from the fact that courts do not recognize any “official legal breaks” since the Founding, so it counts as part of “our law” even if many courts—perhaps all courts—fail to do so. Courts today may simply be mistaken in their failure to see that OLO is entailed by the criteria implicit in the official story. Thus, as Baude and Sachs put it, “[a]n originalist rule can still be a legal rule, even if no court applies it—so long as the legal system still recognizes an official story with that result.”
Under this view, then, OLO is part of our law in the same way that the U.S. tax code is part of our law. Both become law through application of parts of our rule of recognition. The only difference is that OLO is part of our law because it satisfies a part of our rule of recognition that is unwritten (namely, the official story) and so must be inferred from judicial practice, whereas the federal tax code became law by satisfying the criteria for passing legislation laid out in Article I of the Constitution.
Once OLO is understood to be a subordinate rule, rather than part of our rule of recognition, then it is easy to see why judges could be in “global error” about it. Imagine, for instance, that the rule of recognition provides that “whatever rule written on Paper X is law,” but no judge has ever noticed that, on Paper X, Rule Y was written in faint ink in tiny script. Rule Y would still be “valid law” according to the criteria specified in the rule of recognition, even though no court recognized it. Or, to continue the two prior examples, just as it is conceivable that all judges might misapply some technical provision of the tax code, so, too, is it possible that all of them would err in failing to apply OLO when employing rules of statutory or constitutional interpretation. The possibility of such “surprise” as to what the law requires follows from the “hierarchical structure” of a legal system, where certain rules derive their legal status from still other rules, until you get back to the ultimate rule of recognition.
B. Hartian Positivism and Judicial Practice
It’s a clever argument and one I did not fully appreciate the first time around. Moreover, it insightfully exploits a deep tension that runs throughout Hart’s theory and which has divided scholars of Hart’s work. But even on the reading of Hart more favorable to Baude and Sachs’s approach, it is hard to see how his theory can support the claim that judges are in “global error” about such fundamental aspects of our law.
The tension in Hart revolves around the question of whether law is better understood as a genuine feature of practical reasoning or is instead merely a brute sociological fact about a given society. Traditionally, natural lawyers had argued for the first view, whereas legal positivists argued for the latter. Hart tried to forge a middle path between the two by showing the way in which it could be both. On the one hand, for a legal system to exist, legal officials must treat the system’s ultimate “rule of recognition” as a genuine guide or standard for rendering decisions; that is, they must take the “internal” point of view with respect to it. Austin’s “habits of obedience” are not enough. On the other hand, what the rule of recognition happens to be in any given society is entirely a question of fact. So, as long as that society’s courts and officials treat it as providing the authoritative criteria for legal validity, and as long as the general populace complies with their orders, law exists there.
It’s not entirely clear, though, how Hart thinks one should ultimately answer the question of whether any particular rule in a given society qualifies as a law: should one, as Baude and Sachs maintain, apply the criteria of the rule of recognition to the rule and so render an answer from the “internal point of view” (law-as-practical reasoning)? Or should one simply observe the behavior of courts and officials to see whether they are enforcing the rule against private citizens and thus presupposing its satisfaction of the rule of recognition (law-as-brute-fact)?
To continue with the example from above, when we see that judges have ignored Rule Y written on Paper X, should we conclude that the judges have failed to properly apply the rule of recognition, “whatever rule is written on Paper X is law,” or should we instead say that the rule of recognition is really “whatever rule is written on Paper X is law unless the rule is not visible under normal conditions” (or any rule that would explain the judicial behavior)?
Under the latter view, it should be clear that Baude and Sachs can get little help from Hart. The reason is that according to this view, Hart’s theory is a work of “descriptive sociology,” as he famously put it in the preface to The Concept of Law, so making judgments as to what rules exist in a legal system is fundamentally an empirical inquiry that looks to official behavior. True, it is a sociological approach that treats the language officials use as part of the sociological data that needs to be explained and in that way it does not restrict itself to the “external point of view.” But Hart’s critique of that more behavioristic approach was that it would “miss out a whole dimension of the social life” of those who live under the rules examined. The observer would fail to see, for instance, that a red light is not only a sign that people stop their cars, but, by their lights, a reason for them to do so. But nowhere does Hart ever suggest that such an “external” observer would fail to see a society’s rules because no one was even conforming their conduct to them.
This reading—which we might call the “sociological” reading—is supported by Hart’s discussion of “rule-scepticism” in Chapter Seven of The Concept of Law. There he again pursues a middle path, rejecting the extreme skepticism of the legal realists, who denied that rules could ever constrain, while also denying that every case either does or does not fall under a rule’s scope. Rather, due to the inherently “open-textured” nature of language, some applications of general rules were, in his view, simply indeterminate. In such cases, “all we can profitably offer in answer to the question: ‘What is the law on this matter?’ is a guarded prediction of what courts will do.” In other words, pace Baude and Sachs, in such cases—i.e., hard cases, where courts disagree—all we can do is to adopt the “external” point of view.
But let’s adopt a reading of Hart more charitable to Baude and Sachs. As I have written elsewhere, The Concept of Law is ambiguous on these questions, and there is definitely evidence that Hart aimed to give an account of how the law could generate genuine, even if not moral, obligations. It seems most likely that Hart was conflicted on this question—a fact to which his biographer, Nicola Lacey, has reported that his journals attest. And if that’s the case, then it would seem that answering questions about the validity of legal rules should be responsive to “internal” or “legal” arguments as to whether the rule properly satisfies the rule of recognition. So, in my view, Baude and Sachs are picking up on something real in Hart.
Even on this more normative reading of Hart, though, it is hard to see how his theory can do the work they require of it. The reason is that for law to figure in practical reasoning in the way that Hart envisioned, there must be a limit to how far it can deviate from official practice. So it is hard to see how Hart’s concept of the rule of recognition could generate a “constitution in exile.”
To see why, consider that for Hart, the essential function of a rule of recognition is to remedy various “defects” that plague a system governed only by “primary rules of conduct.” In particular, having such rules cures the problem of uncertainty as to which rules are the official rules of society. When officials accept a rule that effectively says, “we will enforce as law any rule that passes X test or possess Y features,” it makes it clear to people which rules they must pay attention to at the risk of being punished for their violation. Plainly, a rule of recognition as abstract and amenable to competing plausible interpretations as the “official story” will not do much to remedy such uncertainty.
This point is reinforced by a revision Hart made to his own theory, in which he emphasized the conventional nature of the rule of recognition. When Ronald Dworkin pointed out that not all rules are “social rules” in Hart’s sense (that is, rules whose existence depends on the fact that others treat them as rules), Hart conceded the point. He nevertheless insisted that even if not all rules were social, the rule of recognition was. More specifically, the rule of recognition is a conventional rule insofar as officials treat the fact that other officials apply it as itself a reason for them to apply it, too—just as the fact that people drive on the right side of the road is a reason for you to do so as well. But if the rule of recognition is so uncertain in its implications that no official could have confidence that its application would produce similar behavior, what would be the point of such a convention?
Perhaps a concrete example will illustrate the point. Hart is famous for defending the view that the Nazis had law. He thought the issue important because he wanted to stress that the law, and the obligations it imposes, were distinct from morality and the obligations it imposes. Recognizing that difference mattered to Hart because he thought it clarified the nature of certain practical dilemmas the existence of law raises. So, for instance, you can understand more clearly the difficulty that postwar German courts faced when considering cases of the so-called “grudge informer” (people who had informed on someone during the Nazi regime so that he or she would be imprisoned or killed by the regime) if you see that the two sources of obligation—legal and moral—are in conflict in such cases.
But one can easily imagine making the case, on Baude-Sachsian grounds, that the “Third Reich,” and all of its actions, were not legal under German law. You would point to the fact that the Rules Enabling Act of 1933, which purported to authorize Hitler’s government to make laws (including those inconsistent with the constitution), was improperly enacted because the Reichstag did not have the 432 members necessary to establish a quorum. Since the statute that delegated to Hitler the power to make law was itself a legal nullity, all the rules of the Third Reich promulgated pursuant to that authority were legally null and void.
Now Baude and Sachs might happily allow for this interpretation. They would say that their whole point is that what the law is in any given country at any given time is a contingent question of social fact. So it’s an “empirical” question whether the legal argument above is a correct. It all depends on what the rule of recognition was in Germany at the time and what the law required, as assessed from the “internal point of view.” Maybe the “official story” that German courts and other officials accepted at the time entailed that after 1933 Hitler was acting ultra vires; or maybe it entailed that his actions were properly authorized; or maybe the Fuhrer’s edicts were only treated “as if” they were law; or maybe the official story recognized an official legal break from the Weimar regime in 1933, as historians now do. Any of these scenarios is possible—it all depends on contingent facts about official attitudes and assumptions, so you have to “look and see” what German law really required.
Maybe so. But the point is that Hart would have had none of this. Why not? The reason is not because he clearly assumed that Nazi laws satisfied whatever the German legal system’s rule of recognition at the time, though that is true (after all, Hart could have been unaware of this particular procedural defect). Nor is the reason that he understood the rule of recognition to be an exclusively sociological phenomenon observable by looking to judicial conduct alone, though that may also be true. (After all, we are accepting for the sake of argument the more normative reading of Hart here—one that ascribes to him the ambition of showing how the existence of a rule of recognition can generate rules of “law” that are “valid” and so impose genuine legal obligations).
Instead, the reason Hart’s theory is inconsistent with the idea that judges or officials could be in “global error” about fundamental aspects of its law is that if such law existed entirely in “exile,” it would be incapable of playing the role in practical reasoning Hart envisioned for it in the sorts of ethical dilemmas he thought his positivist account of law helped to clarify. If the people in a society could only come to know the law by engaging in esoteric debate as to what the “deep structure” of their legal system entailed, such law would not have even have a prima facie claim to obedience when it conflicts with basic moral duties. The reason is that such law would be effectively unknown to those living under it, and that’s the key issue for Hart. Indeed, a society governed by such “positive law” would suffer from precisely the same defect as a society consisting of only “primary rules” of behavior without any rule of recognition at all, namely profound uncertainty as to which rules were the official rules of the group. Thus, although the postwar German courts Hart criticized happened to invoke “natural law” as their basis for punishing grudge informers, the problem would have been precisely the same had they instead based their decisions on a judgment about what the “deep structure” of German law had really required at the time. In both cases, the courts would be essentially punishing people retroactively for conduct they had no reason to think was illegal at the time.
So yes, for Hart, as for Baude and Sachs, the question of whether the Nazis had law was an “empirical” question. But it was not a question of what the “official story” of German judicial practice was or a question of what subordinate unwritten rules could be derived from the “deep structure” that that story reflects and embodies. It was simply a question of whether the judges and officers of the Third Reich carried out Hitler’s orders, implicitly treating them as legally authoritative. And we know all too well what the answer to that question is.
V. Constructing Originalism
To all of this Baude and Sachs might offer something like the following response:
We are not saying that ours is the only possible interpretation of Hart’s account; we are just saying that our interpretation is logically consistent with the essential features of his theory of law and even seems compelled by it if one takes seriously the idea that the rule of recognition provides the criteria of legal validity—the ultimate test of which laws “exist” in that society. Hart himself may well have been conflicted in his ambitions, and there may be tensions in his book that reflect such internal conflict. But our interest lies not in Hart’s psychology nor in intellectual history but in legal theory and in what that theory can tell us about our practice. So it hardly matters, for instance, that Hart himself put little stock in the capacity of “ordinary legal reasoning” to resolve disputes about the proper application of the rule of recognition in a way that our analysis assumes. That is what a proper application of his theory in fact requires.
Indeed! I think that’s exactly the right response to make. What is more, there is something appealing in the Baude-Sachsian version of Hart. It is a Hart much more willing to see judges as struggling to “find” the law than the actual Hart was and, relatedly, one less constrained by the metaphysical skittishness of the ordinary-language philosophers who surrounded and influenced him. So what if you can’t see (or feel or touch or even read—after all, it’s unwritten) the “deep structure” of the law, and so what if not everyone agrees about what it is or what it requires—that does not necessarily mean it does not exist or that it’s not worth trying to figure out what it entails. You can’t see gravity either, but that doesn’t mean it doesn’t exist.
But as is perhaps obvious by now, in going this route, Baude and Sachs would be confirming the point I tried to make in my initial critique, namely that they are really closet Dworkinians. True, they continue pretending as if Ronald Dworkin never existed by rarely citing and never engaging with his work. But they share much more in common with him, substantively and methodologically, than they do with Hart. As I previously pointed out, like Dworkin, they think law can exist in the face of judicial disagreement; like him, they take seriously the idea that judges can discover the law; and like him, they privilege the “internal” perspective of lawyers and judges when determining what the law is.
Now, with their inventive interpretation of Hart, Baude and Sachs reveal yet another resemblance to Dworkin. Dworkin once wrote that “no firm line divides jurisprudence from adjudication or any other aspect of legal practice.” For Dworkin, both judges and legal theorists engage in what he called “constructive interpretations” of legal practice. Judges constructively interpret some area of doctrine when they (1) make judgments about what sources and materials constitute the relevant area of law, (2) extract from those materials some underlying purpose or principle that best fits and justifies them, and then (3) apply that rationale to the case at hand. Legal philosophers do the same thing at a higher level of generality. They draw conclusions about what principles best fit and justify legal practice as a whole.
Baude and Sachs are constructively interpreting Hart’s theory of law. They discern in it (not entirely unfairly) an underlying purpose to render intelligible, and even plausible and attractive, the practice of judges identifying rules as either valid or invalid by reference to some publicly recognized source of authority (i.e., a rule of recognition) and using those rules to determine who wins and loses in the cases before them. Baude and Sachs then apply this rationale to our own legal practice in order to show what taking it seriously might imply for our legal practices. The result is their account of the “official story” and, derivatively, their account of the original-law originalism they think the official story entails.
I will close this Essay by suggesting that Baude and Sachs’s act of Dworkinian constructive interpretation reveals two important points. First, it reminds us that the “deep structure” of legal reasoning is not only deductive and hierarchical in the way Baude and Sachs both preach and practice. They see law as essentially a hierarchy of authorities and even treat Hart himself as an authority, deflecting any challenges to his theory as effectively involving matters above their pay grade.
But their creative use of Hart reminds us that legal theory, like legal practice, is also interpretive and holistic. Yes, lawyers routinely trace titles to properties and require legal rules to be properly authorized by higher-order rules. But in harder cases, they also rely on analogies between cases to discern underlying purposes in the law or to reconcile seemingly inconsistent legal authorities. Similarly, there is, of course, nothing wrong with drawing on an expert or authority in some intellectual domain to support one’s arguments when one has not adequately investigated the underlying issues oneself. But philosophical texts are nearly always susceptible to competing interpretations, requiring the reader to make difficult interpretive judgments. The same is true of legal texts and authorities. Yet Baude and Sachs downplay this aspect of our practice, treating the difficulties involved as mainly, if not exclusively, “empirical” in nature.
The second, related point that their constructive interpretation nicely reveals is also why Baude and Sachs would reject this characterization of their argument. Dworkin insisted that constructive interpretations necessarily involve what they want to deny, namely reliance upon the interpreter’s own evaluative judgments. The contrast between Baude and Sachs’s invocation of an “official story” and Dworkin’s concept of “constructive interpretation” is in this way telling. Both “official” and “constructive” convey something of a formal or artificial quality, but only the latter term, in its association with such legal doctrines as “constructive possession” or “constructive knowledge,” conveys the further fact that its artificiality is meant to serve an underlying purpose.
The view that interpreting practices necessarily requires interpreters to make value judgments is controversial and I cannot defend it here. But if Baude and Sachs admit that Hart’s account is at the very least ambiguous on the matters discussed above, which seems impossible to deny, on what basis do they choose their own reading? Perhaps because it makes the best sense of Hart’s theory overall? “Best” in what sense? For what purpose? The same question could be asked of their particular gloss on the “official story.” Why not a looser understanding of what “continuity” to the Founding demands?
Ultimately, some criteria for selecting one theory, or interpretation of a theory, over another must be applied. The question is whether those criteria include moral or political considerations or whether, instead, they are exclusively epistemic or theoretical. Dworkin and other legal philosophers have argued that when it comes to theorizing about legal practice the former sort of criteria dominate, and properly so. Even if that’s not always true, it seems likely to be true here since Baude and Sachs seek to employ Hart’s positivism in service of claims about how the Supreme Court should conduct its business—a domain in which the political stakes are high. On this view, then, they are making an implicit judgment as to the purpose of having an “official story” in the first place. But what is that purpose? They do not say. If they could answer that question, then we might start getting somewhere.
Baude and Sachs might deny all of this and insist that their criteria for choosing legal theories (or interpretations of those theories) are purely epistemic and theoretical, not moral or political. But then we might worry that they are either proceeding in bad faith or (more intriguingly) suffering from a mild form of false consciousness. I doubt it’s the former, which makes the latter possibility more likely. If that’s true, then legal theory may require us to go deeper than either Hart or Dworkin were willing to venture.
 Ernest Nagel, Fact, Value, and Human Purpose, 4 Nat. L. F. 26, 26 (1959).
 William Baude & Stephen E. Sachs, Grounding Originalism, 113 Nw. U. L. Rev. 1455 (2019); Eric J. Segall, Originalism Off the Ground: A Response to Professors Baude and Sachs, 34 Const. Comment. 313 (2019).
 Segall, supra note 2, at 326.
 See id.
 Id. at 313.
 Baude & Sachs, supra note 2, at 1458.
 Charles L. Barzun, The Positive U-Turn, 69 Stan. L. Rev. 1323, 1378 (2017).
 See Baude & Sachs, supra note 2, at 1463.
 See Barzun, supra note 7, at 1355–57.
 Baude & Sachs, supra note 2, at 1463.
 William Baude, Is Originalism Our Law?, 115 Colum. L. Rev. 2349, 2351 (2015) (referring to the “positive turn” in originalist theory); Stephen E. Sachs, Originalism as a Theory of Legal Change, 38 Harv. J.L. & Pub. Pol’y 817, 819, 835 (2015).
 Baude, supra note 11, at 2351.
 Id. at 2351, 2392; Sachs, supra note 11, at 823–35.
 Baude, supra note 11, at 2392–93; Sachs, supra note 11, at 874–75.
 Sachs, supra note 11, at 874–75.
 Baude, supra note 11, at 2355; William Baude & Stephen E. Sachs, The Law of Interpretation, 130 Harv. L. Rev. 1079, 1131 (2017); Sachs, supra note 11, at 845; Their theory of originalism is very similar to what Professors McGinnis and Rappaport have dubbed “methods-originalism.” See John O. McGinnis & Michael B. Rappaport, Original Methods Originalism: A New Theory of Interpretation and the Case Against Construction, 103 Nw. U. L. Rev. 751, 783 (2009). But, unlike Baude and Sachs, McGinnis and Rappaport have not offered their theory as a descriptive one about what our law today requires.
 Barzun, supra note 7, at 1339, 1341–42; see Baude, supra note 11, at 2362 (making the comparison to methods-originalism).
 Mark Greenberg, What Makes a Method of Legal Interpretation Correct? Legal Standards vs. Fundamental Determinants, 130 Harv. L. Rev. Forum 105, 117–18 (2017) (observing that “an inquiry into legal interpretation cannot afford to set aside the question of how the content of the law is determined at the fundamental level,” because “prominent theories of law have extremely different implications for legal interpretation”).
 Barzun, supra note 7, at 1342 (“It matters which theory of law underwrites the [positive turn], and defending any of them requires engaging in conceptual or normative argument.”).
 Baude & Sachs, supra note 2, at 1462–63. That theory is laid out in H.L.A. Hart, The Concept of Law (3d ed. 2012).
 Cf. Greenberg, supra note 18, at 114 (observing that “Hart’s theory is an unfortunate choice for theorists who claim that longstanding debates over legal interpretation can be resolved by looking to the law of interpretation”).
 Barzun, supra note 7, at 1347–52.
 Id. at 1352–60. Again, Greenberg raised a similar objection. See Greenberg, supra note 18, at 115 (explaining that “no theory of legal interpretation that is not widely accepted can be part of the rule of recognition”).
 See Baude & Sachs, supra note 2, at 1483.
 Id. at 1473–74.
 Id. at 1473.
 Id. at 1461.
 Id. at 1462.
 Id. at 1465.
 Id. at 1467.
 Id. at 1469.
 Id. at 1471.
 See Barzun, supra note 7, at 1347–51.
 Hart, supra note 20, at 110 (describing the rule of recognition as a “complex, but normally concordant, practice of the courts, officials, and private persons in identifying the law by reference to certain criteria”). As the passage quoted indicates, Hart was not always consistent as to whose practice matters for establishing a rule of recognition. It is another of the ambiguities referred to below. See infra note 50. For simplicity, I followed Baude and Sachs’s lead in assuming, for present purposes, that judges or courts are the relevant community for our purposes, Barzun, supra note 7, at 1346, and I continue to do so here. At times, Baude and Sachs suggest that they may differ with Hart as to the scope of the recognitional community and that perhaps lawyers and legal scholars ought to be included as well. See Baude, supra note 11, at 2370 (“[I]t may ultimately be important to look to official practice beyond judges.”); Sachs, supra note 11, at 883 (acknowledging the significance of—without actually answering—the question, “Does the best positivist theory identify law through the conventions of ordinary people, or through the practices of lawyers, judges, and officials?”). Such an extension of the recognitional community would fit well with my suggestion below, infra Part V, that Baude and Sachs are engaged in a constructive interpretation of Hart’s theory.
 Barzun, supra note 7, at 1349–50.
 Baude and Sachs also argue that citation practices do not constitute sufficient evidence of the “deep structure” of our practice, because sometimes courts do not cite a source when its authority can be safely assumed. See Baude & Sachs, supra note 2, at 1478–82. Fair enough, but that just pushes the issue back to the question of whether courts are in fact assuming what the authors say they are, namely that the interpretive methods they employ are valid because those methods can trace a pedigree back to the Founding. See infra note 39 and accompanying text.
 Baude & Sachs, supra note 2, at 1483 (alteration in original) (footnotes omitted).
 Baude & Sachs, supra note 2, at 1483. Sachs drew the same contrast in his initial article. See Sachs, supra note 11, at 866–67; Barzun, supra note 7, at 1350.
 Baude & Sachs, supra note 2, at 1474.
 Hart, supra note 20, at 110 (distinguishing between the rule of recognition of a legal system and “subordinate” rules of the system).
 Baude & Sachs, supra note 2, at 1477–78.
 Id. at 1478. The “guiding toward” formulation may be preferable because the “official story,” according to Baude and Sachs, does not consist of prescriptive or rule-like provisions. See id. at 1468.
 Id. at 1474.
 Id. at 1477.
 Id. at 1473.
 Id. at 1465 (“This hierarchical structure makes it possible for the correct ground-level legal rules to surprise us.”).
 Baude and Sachs thus fairly point to an imprecise statement in my article. I had said that under Hart’s view, “a rule no court applies cannot be law,” Barzun, supra note 7, at 1360 n. 228, but they correctly note that that is not true of subordinate rules. Baude & Sachs, supra note 2, at 1473–74. I should have specified more clearly that a rule that no court applies cannot be part of the rule of recognition. I had assumed (incorrectly, it now turns out) that they understood OLO to be part of the rule of recognition itself.
 The issue has traditionally been framed as whether Hart’s ambition was to explain the normativity of law or whether, instead, he intended to limit himself to a purely sociological analysis. Compare Scott Hershovitz, The End of Jurisprudence, 124 Yale L.J. 1160, 1168 (2015) (“Hart invites us to derive a normative statement (that is, a claim about what you are legally obligated to do) from descriptive statements about the social practice among legal officials around here.”), and Veronica Rodriguez-Blanco, Peter Winch and H.L.A. Hart: Two Concepts of the Internal Point of View, 20 Canadian J.L. & Juris. 453, 460 (2007) (interpreting Hart as seeking to explain how “rules give reasons for actions”), with Brian Leiter, Beyond the Hart/Dworkin Debate: The Methodology Problem in Jurisprudence, 48 Am. J. Juris. 17, 38 (2003) (“Hart has nothing to say about the normativity of law in the main text of The Concept of Law, beyond a refutation of the Austinian account.”), Stephen Perry, Hart on Social Rules and the Foundations of Law: Liberating the Internal Point of View, 75 Fordham L. Rev. 1171, 1173 (2006) (“Since the internal point of view is nothing more than an attitude that a standard is binding, Hart is not offering an account of the normativity of law that looks to its (potential) reason-givingness.”), and Scott J. Shapiro, What is the Internal Point of View?, 75 Fordham L. Rev. 1157, 1166 (2006) (“Hart did not intend for the internal point of view to provide an explanation for the reason-giving nature of social rules and law.”).
 Compare Thomas Aquinas, Treatise on Law: Summa Theologica, Questions 90-97, at 10–11 (1996) (defining law as “an ordinance of reason for the common good, made by him who has care for the community, and promulgated”), with John Austin, The Province of Jurisprudence Determined 147 (David Campbell & Philip Thomas eds., 1998) (“Every positive law, or every law simply and strictly so called, is set by a sovereign person, or a sovereign body of persons, to a member or members of the independent political society wherein that person or body is sovereign or supreme.”).
 Hart, supra note 20, at 116.
 Austin, supra note 51, at 147 (emphasis omitted).
 Hart, supra note 20, at 116.
 Baude & Sachs, supra note 2, at 1475 (“But whatever the criteria of validity may be, those criteria are to be applied from the internal point of view, using the ‘characteristic vocabulary’ of ‘it is the law that . . . ,’ and so on.”) (quoting Hart, supra note 20, at 102).
 Hart, supra note 20, at vi.
 Id. at 90.
 Id. at 124–54.
 Id. at 130 (emphasizing the need to “escape this oscillation between extremes” as to the determinacy of rules).
 Id. at 128.
 Id. at 147.
 Baude & Sachs, supra note 2, at 1470 (“The essence of the internal/external distinction is not simply to accept that ‘[i]t is what the Court has been doing that is our law’—for it matters greatly why a court is doing what it’s doing, and what kind of grounds it can cite in support.”) (emphasis omitted) (quoting Richard Primus, Is Theocracy Our Politics?, 116 Colum. L. Rev. Sidebar 44, 51–52 (2016)).
 Charles L. Barzun, Inside-Out: Beyond the Internal/External Distinction in Legal Scholarship, 101 Va. L. Rev. 1203, 1223–24 & n. 65 (2015) (offering textual support from The Concept of Law for this view).
 Nicola Lacey, A Life of H.L.A. Hart: The Nightmare and the Noble Dream 228 (2004) (observing that Hart’s notebooks reveal that he “struggled with the concept of legal obligation” and that he thought that the concept of legal obligation “would be the linchpin of his delicate middle way between Realism or crude positivism and natural law: the idea of law as generating genuine obligations rather than merely forcing compliance, those obligations however falling short of moral obligations”).
 See Hart, supra note 20, at 110 (explaining that once a rule of recognition is in place, “a subordinate rule of a system may be valid and in that sense ‘exist’ even if it is generally disregarded”).
 Stephen E. Sachs, The “Constitution in Exile” as a Problem for Legal Theory, 89 Notre Dame L. Rev. 2253 (2014).
 Hart, supra note 20, at 92.
 Id. at 94.
 Id. at 255.
 Id. at 256; see Leslie Green, Introduction, in id. at xxii (“Not all rules, [Hart] now admits, are practice rules, but conventional rules are and they form the basis of law.” (emphasis omitted)).
 Ronald Dworkin made a similar criticism of Jules Coleman’s nearly identical strategy for handling disagreement. Coleman distinguished between agreement as to the content of the rule of recognition and agreement over its proper application, arguing that Hart’s theory only required the former, which could exist at a very high level of generality (i.e., something akin to Baude and Sachs’s “official story”). Jules Coleman, The Practice of Principle: In Defense of a Pragmatist Approach to Legal Theory 116–17 (2001). In response, Dworkin pointed out that this strategy “eviscerates the idea of convention itself” because, among other problems, it would seem to allow for rules of recognition (e.g., “decide the case properly”) that are so abstract that they could not plausibly constitute conventional rules. See Ronald Dworkin, Thirty Years On, 115 Harv. L. Rev. 1655, 1660–61 (2002) (reviewing Coleman, supra).
 H.L.A. Hart, Positivism and the Separation of Law and Morals, 71 Harv. L. Rev. 593, 619 (1958) (criticizing a German court for “declaring a statute established since 1934 not to have the force of law”); Hart, supra note 20, at 210 (making a similar point).
 Hart, supra note 20, at 210.
 Hermann Goring, who served as presiding officer of the Reichstag, said the quorum requirement was satisfied because he did not count the communists, thereby reducing the needed number to 378. According to the historian Richard Evans, “[t]his was a high-handed decision that had no legitimacy in law whatsoever.” See Richard J. Evans, The Coming of the Third Reich: A History 351–52 (2003).
 Cf. Baude & Sachs, supra note 2, at 1491 (empirical question); id. at 1475 (“internal point of view”).
 Id. at 1472.
 Id. at 1463 (“The thing to be explained strikes us as a feature of our society, and so it seems fair to ‘[l]ook and see’ what our society’s practices actually are.”) (quoting Leslie Green, Introduction, in Hart, supra note 20, at xlv).
 Hart, supra note 20, at 210.
 Cf. Baude & Sachs, supra note 2, at 1476 (“The deep structure of our legal system is a question of present law. . . .”).
 Cf. id. at 1463 (endorsing a “generally Hartian version of positivism”).
 Cf. Hart, supra note 20, at 274 (noting the “familiar rhetoric” from judges that there are no “legally unregulated cases,” but asking rhetorically, “how seriously is this to be taken?” and observing that many judges have acknowledged that “there are cases left incompletely regulated by the law where the judge has an inescapable though ‘interstitial’ law-making task”).
 See Barzun, supra note 64, at 1214–15 (discussing this aspect of ordinary-language philosophy and its influence on Hart).
 Cf. Stephen E. Sachs, Finding Law, 107 Calif. L. Rev. 527, 581 (2019) (“Unwritten law can be found, as well as made; the brooding omnipresence broods on.”).
 Or does it? See Bas van Fraassen, The Scientific Image (1980) (expounding an anti-realist philosophy of science, dubbed “constructive empiricism,” which denies that science does or should deliver the truth about unobservable entities or forces).
 Dworkin is not mentioned or cited once in either Grounding Originalism or Professor Sachs’s recent article, Finding Law, supra note 84. The omission in the latter piece is particularly glaring since Dworkin is more associated with the view that judges (at least in some sense) discover the law, even in hard cases, than just about any other modern legal philosopher.
 See Barzun, supra note 7, at 1385.
 Ronald Dworkin, Law’s Empire 90 (1986).
 Id. at 52, 65–67.
 Id. at 90.
 See, e.g., Baude & Sachs, supra note 2, at 1471 (“In modern societies, law is a hierarchical and structured normative practice.” (emphasis omitted)); id. at 1463–64 (acknowledging that “[i]t might turn out that the Hartian account is generally wrong,” but insisting that that is “bigger game” and that they are “satisfied to show that our account of the law is generally consistent with the most-accepted theory of positivism”).
 For my own effort to identify a form of interpretive holism that differs in important ways from Dworkin’s particular brand of it, see Charles L. Barzun, Justice Souter’s Common Law, 104 Va. L. Rev. 655 (2018), and Charles L. Barzun, Three Forms of Legal Pragmatism, 95 Wash. U. L. Rev. 1003 (2018).
 Compare Dworkin, supra note 88, at 52–53 (explaining that constructive interpretation involves “imposing purpose on an object or practice in order to make of it the best possible example of the form or genre to which it is taken to belong”), with Baude & Sachs, supra note 2, at 1465 (arguing that “originalism can be a correct descriptive account of our legal system, even if few people would currently describe our system that way” (emphasis omitted)).
 Cf. Baude & Sachs, supra note 2, at 1487 (“We believe we’ve put forward the best account of the official story of our constitutional law.”).
 See, e.g., Dan Priel, Toward Classical Legal Positivism, 101 Va. L. Rev. 987 (2015) (arguing that the best defense of legal positivism is a normative one); see also Charles L. Barzun, The Forgotten Foundations of Hart and Sachs, 99 Va. L. Rev. 1, 32 n.130 (2013) (collecting sources in the methodology debate); Leiter, supra note 50 (describing the debate and endorsing a descriptivist, but non-conceptual approach to jurisprudence).