The Standard Picture and Statutory Interpretation

Abstract

The Standard Picture holds that the contribution to the law made by an authoritative legal pronouncement is directly explained by the linguistic content of that pronouncement. This essay defends the Standard Picture from Mark Greenberg’s purported counterexamples drawn from patterns of statutory interpretation in U.S. criminal law. Once relevant features of the U.S. rule of recognition are admitted into the analysis—namely, that it arranges sources of law hierarchically, and that judicial decisions are sources of valid law—Greenberg’s counterexamples are revealed as only apparent, not genuine. The legal norms that result from the patterns of interpretation he identifies can be directly explained in terms of the linguistic contents of authoritative pronouncements: judicial decisions. Furthermore, those norms can be understood as modifications of the valid norms contained in their originating statutes because judicial decisions are permitted ‘explanatory intermediaries’ of statutes by the rule of recognition.

The Standard Picture, so named and thoroughly critiqued by Mark Greenberg, is the view that the contribution to the law made by an authoritative legal pronouncement is directly explained by the linguistic content of that pronouncement. In more positivistic terms, the view is that the valid legal norms of a system are directly explained by the linguistic contents of the sources of valid law. This essay will argue that Greenberg’s purported counterexamples to the Standard Picture drawn from patterns of statutory interpretation in U.S. criminal law are in fact only apparent, and not genuine, counterexamples to that picture. While I agree with much of what legal positivists have said in regard to the Standard Picture’s theoretical advantages over Greenberg’s own Moral Impact Theory, and while I will briefly touch on those issues here, my primary concern will be Greenberg’s examples drawn from legal practice. If those examples do not cast genuine doubt on the Standard Picture, then Greenberg’s Moral Impact Theory loses much of its motivation.

The essay is in four parts. First, I will rehearse Greenberg’s overview of the Standard Picture and clarify certain points. Second, I will discuss Greenberg’s objections to the Standard Picture, including the patterns of statutory interpretation he identifies as counterexamples (or at least, as kinds of cases the Standard Picture is supposed to have difficulty accommodating). Third, I will argue that the Standard Picture in fact fits those patterns of statutory interpretation very nicely once relevant features of the U.S. rule of recognition are admitted into the analysis, and I will respond to potential objections to my account. Finally, I will remark in closing a heretofore unacknowledged theoretical benefit of the Standard Picture: it may afford positivists a promising answer to Ronald Dworkin’s objection from theoretical disagreement, along lines I will indicate.

I. The Standard Picture

Greenberg says that adherents of the Standard Picture—a group he thinks includes almost all legal theorists (more on this below)—typically do not expressly endorse or even articulate it, but rather take it as a common and obvious presupposition. For this reason, he says, it is inapt to characterize it as a fully formed doctrine or theory; rather, it is an organizing scheme or (as its name suggests) a picture, which he does his best to sketch. ¹ Be that as it may, the view Greenberg proceeds to sketch is a sufficiently developed position to merit the name of ‘theory,’ and I will refer to it as such. In particular, it is a theory of the content of the law, that is, the relationship of the authoritative legal pronouncements (or sources of law) to the rights and duties that arise from them (i.e. the valid legal norms).

The central idea of the Standard Picture is that “the content of the law is the meaning of certain legal texts (or utterances).” ² The “key point” of the view is that “what is authoritatively pronounced becomes a legal norm—or, equivalently, becomes legally valid—simply because it was authoritatively pronounced.” ³ The Standard Picture maintains that “the content of the law is primarily constituted by linguistic (or mental) contents associated with the authoritative legal texts”; ⁴ and again, “[t]he Standard Picture holds that it is an ordinary linguistic (or mental) content associated with an authoritative legal pronouncement that constitutes a legal norm.” ⁵ (Greenberg parenthetically mentions “or mental” in these formulations, he says, because the Standard Picture is sometimes relaxed to include mental states beyond speaker’s intent, which he allows may be a part of linguistic content. He has in mind “the content of a legislature’s intention to achieve particular legal effects by enacting a statute.”) ⁶ In short, the Standard Picture is “the layperson’s idea that the law is what the code or law books say[:] … the linguistic contents of the authoritative legal pronouncements are the contents of the legal norms.” ⁷

That is the general idea. Greenberg provides a more detailed specification in the form of what he calls the ‘Core Model’ of the Standard Picture. The Core Model comprises three theses, the first of which is a more technical specification of the “key point” noted in the previous paragraph, and the second and third of which are consequences of that driving thesis. ⁸

The first thesis is Explanatory Directness. It asserts that the authoritativeness of a pronouncement (by which Greenberg means here its legal standing, not its moral authority) is prior in the order of explanation to the obtaining of a valid legal norm. Moreover, the content of the valid legal norm is explained directly by the authoritative pronouncement—that is, there are no “explanatory intermediaries” between the pronouncement and the norm. ⁹ It is not the case, for instance, that the pronouncement changes people’s expectations or changes what is morally required, which then explains why a resultant norm is legally valid.

Next, the Linguistic Content thesis specifies that the way in which the authoritative pronouncement directly explains the content of the valid legal norm is via its own linguistic content: the content of the valid legal norm is the linguistic content of what was pronounced. Greenberg says he uses “linguistic content” rather than “meaning” in articulating this thesis because the latter could be mistaken as standing for general upshot or significance. ¹⁰ (Greenberg allows that linguistic content may include speaker’s intent as well as semantic content. He is agnostic on the precise nature of the mechanism by which we understand linguistic content; whatever its nature, he says, it is evidently systematic and reliable.) ¹¹ The Linguistic Content thesis is meant to be a straightforward consequence of the Explanatory Directness thesis: “If the content of the legal norm created by an authoritative pronouncement is not the linguistic content of the pronouncement, the explanation of the norm’s legal validity cannot be simply that it was said or meant (since, by hypothesis, it wasn’t).” ¹²

The third and final thesis of the Standard Picture’s Core Model is Atomism. It says that the content of the law as a whole is the amalgamation of all of the legal system’s individual valid legal norms, that is, “individual legal norms are explanatorily prior to the content of the law as a whole.” ¹³ Contrast this with a kind of legal holism that would purport to explain the validity of any individual legal norm by derivation from the total content of the law. (This aspect of the Standard Picture will be the least important for purposes of our discussion.)

Those are the core views (or assumptions) Greenberg attributes to adherents of the Standard Picture. And, as remarked above, Greenberg thinks the Standard Picture is all but taken for granted in legal philosophy: he identifies only Ronald Dworkin’s law as integrity and his own Moral Impact Theory (see next paragraph) as theories that hold that the content of the law depends in the first instance on the content of morality, as opposed to the meaning of authoritative legal pronouncements. ¹⁴ But there would appear to be conceptual space between those alternatives. A theorist could hold, for instance, that the valid legal norms of a system depend on the meanings of the authoritative pronouncements, but only after those meanings have first been filtered through some manner of interpretive methodology characteristic of the legal system in question. Such a view would arguably run afoul of the Core Model’s Explanatory Directness and Linguistic Content theses, and if so would not, any more than Greenberg’s or Dworkin’s, constitute a Standard Picture view. ¹⁵ (We will touch on views of this kind in Parts III and IV.)

At all events, Greenberg thinks the Standard Picture a fatally flawed account of the content of the law; we will come to his reasons for thinking so momentarily. He offers in its stead his own ‘Moral Impact Theory.’ The Moral Impact Theory holds that “[t]he content of the law is that part of the moral profile created by the actions of legal institutions in the legally proper way.” ¹⁶ By “moral profile” Greenberg means all moral obligations, powers, privileges, and permissions; what makes those obligations (powers, etc.) ‘moral’ is that they are all-things-considered practical obligations (powers, etc.). ¹⁷ As my interest is primarily in Greenberg’s reasons for rejecting the Standard Picture, I will say little about the Moral Impact Theory. It is, however, important to appreciate that Greenberg offers the Moral Impact Theory as a theory of the content of the law, just as the Standard Picture is a theory of the content of the law. That is, the Moral Impact Theory is not a theory of law in the broad sense that H.L.A. Hart’s legal positivism or Dworkin’s law as integrity is a theory of law; in particular, it does not contain any account of legal systems or legal institutions (and nor does the Standard Picture). In the case of the Standard Picture, we may say simply that it is a theory of the valid legal norms of a system. The Moral Impact Theory is also a theory of a legal system’s valid norms, but it is so only derivatively: it provides an account of the valid norms of a system by derivation from the total content of the system’s law, which is derived in turn (and in part) from the content of morality. That is to say, the Moral Impact Theory rejects the Atomism thesis of the Standard Picture’s Core Model (indeed, it rejects all three theses of that model).

II. Greenberg’s Objections to the Standard Picture

The rest of this essay will focus on features of U.S. legal practice, specifically patterns of statutory interpretation, that Greenberg alleges do not fit the Standard Picture. I should be clear, however, that Greenberg does not mean these examples to refute the Standard Picture, but rather to show that it is “not obviously true” and cannot be taken for granted. ¹⁸ He also levels a theoretical critique to the effect that the Standard Picture is uncongenial to something he and others take to be an important fact about the nature of law (namely, that law is supposed to create morally binding obligations—more on this below). In later work Greenberg attempts to show that his own Moral Impact Theory both satisfies that theoretical desideratum and also better fits these statutory interpretation examples. ¹⁹ Here I will largely bracket Greenberg’s theoretical critique. If his statutory interpretation examples can be accommodated by the Standard Picture and, importantly, can be accommodated without relying on controversial claims that might themselves weaken the Standard Picture’s plausibility, then at least a significant part of Greenberg’s motivation for rejecting the Standard Picture will be mitigated.

I should say at least a word about Greenberg’s theoretical critique, however. It proceeds from the “bindingness hypothesis,” the hypothesis that law is supposed to create morally binding obligations. ²⁰ From that starting point, Greenberg’s argument runs as follows: The only way the Standard Picture could create morally binding obligations is if there were a general moral obligation to obey authoritative legal pronouncements. Yet the emerging theoretical consensus is that there is not any such general moral obligation to obey the authoritative pronouncements of contemporary legal systems. Consequently, the Standard Picture cannot create morally binding obligations. Since law is supposed to create morally binding obligations, this “gives us serious reason to doubt SP [the Standard Picture].” ²¹

Other positivists have ably extolled the theoretical virtues of the Standard Picture in comparison to Greenberg’s Moral Impact Theory, and I agree with much of what they say. ²² So here I will only comment rather peremptorily that I think Greenberg’s theoretical critique of the Standard Picture is dubious. First, I see no reason to accept Greenberg’s claim that law is supposed to create morally binding obligations. Law that does not create morally binding obligations may be morally defective, but it is not defective as law. Second, even if law were supposed to create morally binding obligations, I fail to see why it couldn’t do so legal norm by legal norm, without depending on either a general moral obligation to obey the law (a “morality-on-law” direction of dependence, in Greenberg’s terms) or a stipulation that legal obligations must first be moral ones (a “law-on-morality” direction of dependence, and Greenberg’s approach). ²³ Greenberg briefly mentions this possibility, the alternative that “bindingness could be achieved in a pronouncement-by-pronouncement way, rather than by relying on morality-on-law or law-on-morality dependence.” ²⁴ To this alternative he responds that the morally relevant circumstances are so complex and unpredictable that the linguistic content of the authoritative pronouncements is bound to diverge from the content of morality. But I do not see why that should be. ²⁵ Third, even were we to grant both Greenberg’s claim that law is supposed to create morally binding obligations and his claim that it cannot reliably do so pronouncement by pronouncement, why not simply conclude that the law is to some significant extent defective (defective as law now, for the sake of argument)? Greenberg’s strategy is to offer an alternative theory of the content of the law that does not require us to take such an uncharitable view. But if Greenberg’s account of the mechanism by which valid legal norms arise—i.e., via an authoritative pronouncement’s impacting the society’s moral profile—is significantly more opaque than the mechanism described by the Standard Picture, then why should we prefer a view on which the content of law is non-defective but mysterious to one on which the content of the law is defective but readily identifiable?

That will conclude my discussion of Greenberg’s theoretical critique. Without further ado, I turn now to those patterns of statutory interpretation Greenberg identifies as problematic for the Standard Picture. The first involves the rule of lenity, the interpretive principle that criminal statutes are to be interpreted narrowly in favor of the defendant (in particular, ambiguities are to be resolved in the defendant’s favor). Here we have the legal norm arising from a criminal statute bearing a different relationship to its source than that of legal norms arising from non-criminal statutes, Greenberg says. ²⁶ He claims the Standard Picture has no explanation for why a different aspect of the linguistic content of this particular kind of authoritative pronouncement should constitute the law than in the case of other kinds of pronouncements. Greenberg’s second purported counterexample to the Standard Picture is the mens rea canon, which on the occasions it is applied interprets a criminal statute as if it contained a mens rea requirement even though none is written. Greenberg observes that “[i]t would be a strain to argue that mens rea requirements are somehow part of the linguistic content of criminal statutes, whatever their wording.” ²⁷ The third pattern of statutory interpretation Greenberg identifies as problematic for the Standard Picture is that courts interpreting statutes based on the Model Penal Code (MPC) (a reform model statute originally drafted by the American Law Institute in the 1960s) often interpret them in accordance with the MPC drafters’ commentary and in ways that depart from the linguistic content of the statute itself. ²⁸ For instance, according to Greenberg, although §5.01 of the MPC requires that “a defendant be aware of circumstance elements of an offense in order to be guilty of an attempt to commit that offense,” courts often interpret statutes based on this section as requiring only the mens rea required for the underlying offense, which is what the commentary recommends. ²⁹

Those are the problem cases of statutory interpretation Greenberg raises for the Standard Picture. Greenberg anticipates, and attempts to forestall, some potential responses on behalf of the Standard Picture. One is simply that the courts in these cases get the law wrong insofar as they depart from the linguistic content of the statutes in question. Greenberg, for his part, thinks the courts in these cases are getting the law right. Such a disagreement about the true identity of the valid norm is irresolvable without recourse to more theoretical considerations, Greenberg says, and adverts to his bindingness hypothesis. ³⁰ Another potential response on behalf of the Standard Picture is that the courts in these cases are identifying the enacting legislature’s intentions. If that were the case, then the law would still be directly explained by what the authoritative pronouncements in question meant, if not by their semantic content. This response is unsuccessful, Greenberg says, for:

[E]ven given generous assumptions about how legislators’ attitudes combine to constitute the legislature’s intentions, it is wildly implausible, as an empirical matter, that legislators have the intentions that they would need to have in order to accommodate my examples … i.e., intentions that correspond to the way in which the statutes are interpreted. ³¹

I happen to agree with Greenberg that the courts in his example cases are not identifying legislative intentions, and furthermore that they are not getting the law wrong by applying the statutes in the ways that they do. ³² Nevertheless, I do not consider these kinds of cases counterexamples to the Standard Picture, or even see them as difficult for the Standard Picture to accommodate. In the next section I will give a different kind of response to these cases on behalf of the Standard Picture.

III. Accommodating Greenberg’s Problem Cases

a) A Standard Picture Account of Greenberg’s Cases

As we discussed in Part I, Greenberg’s professed interest is in the content of the law, as opposed to legal systems or their institutions or procedures. I think this self-imposed blinkering distorts Greenberg’s perspective. A theory of legal norms does not float free of a theory of the legal system in which those norms are the valid ones. Insofar as Greenberg’s own Moral Impact Theory endeavors to be freestanding in this way, it is unsatisfactory; it is fairly evident, for instance, that Greenberg needs a theory of legal institutions and their procedures, as these are implicated by his theory of the content of the law (“that part of the moral profile created by the actions of legal institutions in the legally proper way”). But I will set aside criticism of Greenberg’s own theory. I want to focus instead on showing that the Standard Picture fits U.S. legal practice very nicely, once we presuppose a certain background theory of that legal system (a positivist one) and take into account certain facts relevant on that theory.

There are two relevant aspects of the U.S. legal system’s rule of recognition that, once admitted into the analysis of the relationship of statutes to the legal norms that arise from them, allow us to see that those norms do arise from their statutes in accordance with the Standard Picture. The first is that the rule of recognition not only identifies the sources of valid law in the system, but it also arranges them hierarchically. ³³ The segment of the hierarchy of sources of law that is relevant to Greenberg’s problem cases is that, when a dispute is brought to court that turns on the application of a statutory provision, the court is empowered to determine authoritatively whether the provision applies to the fact-situation. As the court has the final word on whether the statutory provision applies—as Joseph Raz puts it, it has the authority to make a “binding application” of the provision—the court’s decision in this way ‘trumps’ the statute. ³⁴ It follows that if a court applies a statutory norm to a fact-situation to which that norm’s application is uncertain, the court thereby further specifies, and modifies, the norm. ³⁵

The second relevant aspect of the U.S. rule of recognition, implicit already in the first, is that judicial decisions are sources of valid law. In the kind of case just mentioned, if the court is one of precedential authority, then its decision will constitute a new valid legal norm in the jurisdiction, i.e. that the statutory provision does or doesn’t (depending on the court’s decision) apply to fact-situations of this kind. If the court is not one of sufficient authority, then its decision results in a norm that binds only the parties to the dispute. Although the court’s decision has legal validity and thus so does the norm it pronounces, that validity reaches only to the parties to the dispute and so the norm of the decision will form no part of the content of the law in the jurisdiction going forward.

Bearing in mind these background facts of the U.S. legal system, we can now return to Greenberg’s alleged counterexamples to the Standard Picture. The first, recall, involved cases in which courts follow the rule of lenity, or the interpretive principle that ambiguities in criminal statutes are to be resolved in the defendant’s favor. Greenberg contends the Standard Picture cannot account for the fact that a different aspect of the linguistic content of the criminal statutes at issue in these cases constitutes the law than in the case of other statutes. That is not so. The Standard Picture maintains that legal norms are directly explained by the linguistic contents of authoritative legal pronouncements. When the interpretive principle of the rule of lenity is applied in a case, the resulting legal norm is that the statutory provision in question does not apply to conduct of the kind the defendant engaged in. That norm is directly explained by the linguistic content of an authoritative legal pronouncement: the court’s decision.

To stop there is to rely solely on the second feature of the U.S. rule of recognition discussed above, i.e., that judicial decisions are sources of valid law. In a limited sense, that response is adequate to meet Greenberg’s objection; after all, it directly explains the norm resulting from an application of the rule of lenity in terms of the linguistic content of an authoritative legal pronouncement. However, the authoritative legal pronouncement with which this response directly connects the norm that results from an application of the rule of lenity is a judicial decision. Yet Greenberg is surely right to insist that there is a connection between the norm that results from an application of the rule of lenity and the statutory provision itself. Greenberg’s objection to the Standard Picture is that this connection cannot be explained solely in terms of the linguistic content of the statutory provision. Can it nevertheless be explained consistently with the Standard Picture?

I contend that it can. I propose that the connection between statute and eventual norm be made in terms of the linguistic content of the other authoritative legal pronouncement we have already identified: the judicial decision that applies the rule of lenity. Cases in which the rule of lenity is applied are hard cases in Hart’s sense: the criminal prohibition at issue neither clearly applies to the defendant’s conduct, nor is the defendant’s conduct clearly beyond the reach of that prohibition. If read in a broad sense, the key term of the crime’s definition encompasses the defendant’s actions, while if read in some available narrower sense, it does not. Cases in which the valid norm’s application is unclear because the fact-situation falls in the “penumbra of uncertainty” of that norm are hard cases, Hart says. ³⁶ In such cases, courts are inevitably faced with a choice among alternatives left open by the existing norm. ³⁷ In making their choice courts inescapably perform a “rule-producing function,” even if they disclaim, or even if they are unaware, that they are doing so. ³⁸ In short, in such a case the court must exercise its discretion. Consequently, the legal norm that results from the court’s decision is properly understood as a further specification of the original statutory norm. ³⁹

Making this connection from, on the one hand, the norm the court creates in applying the rule of lenity in a hard case to, on the other hand, the linguistic content of the originating statutory provision requires invoking the first feature of the U.S. rule of recognition discussed above. That feature was that the rule of recognition arranges sources of law hierarchically, and specifically that a judicial application of a statutory norm trumps the statute and may thereby modify the norm. ⁴⁰ (I say it may modify the norm because, in an easy case, the valid norm clearly applies on its own terms, so the court’s application of the norm does not alter it.) In the light of this fact about the rule of recognition, the judicial decision in question is not just an additional authoritative pronouncement. It is an intervening—or to use Greenberg’s word, intermediary—authoritative pronouncement between statute and norm. The Standard Picture itself, as merely a theory of the content of the law, is silent concerning any relative authority of legal pronouncements vis-à-vis one another. But once we recognize that the criteria of validity undergirding those pronouncements arrange them into an ordering of authority, then the possibility of higher-order pronouncements intervening to explain the norms that originate in lower-order pronouncements presents itself.

Much the same can be said in response to Greenberg’s second and third purported counterexamples to the Standard Picture. His second counterexample, recall, was the mens rea canon, which when applied imputes a mens rea requirement to a statutory prohibition that does not expressly include one. It would be a “strain,” Greenberg says, to argue that the resulting mens rea requirement was part of the linguistic content of the statute. ⁴¹ But the defender of the Standard Picture needn’t make that argument. The mens rea requirement that is imposed is part of the linguistic content of an authoritative legal pronouncement—again, the decision of a court. Moreover, the court in such a case has been presented a fact-situation to which the statutory provision would seem to apply, except the defendant did not possess the particular bad intent we would expect before imposition of this criminal liability. This is a hard case, to which the statutory norm neither clearly applies nor clearly does not. Thus, as in the case of the rule of lenity, a court is exercising its discretion in applying the mens rea canon. The norm the court pronounces is properly seen, then, as a modification of the original statutory norm.

Greenberg’s final problem case was that courts sometimes interpret Model Penal Code statutes in ways suggested by the MPC drafters’ commentary but that depart from the statutory text itself. His particular example was that courts often interpret MPC § 5.01 on criminal attempts as requiring only the mens rea to commit the underlying offense (as the commentary recommends), rather than as requiring knowledge of all of the circumstance elements of the crime (as the statutory text would seem to require). Here again we have a hard case in which the court must exercise its discretion. Indeed, this is the converse of Greenberg’s previous example, the mens rea canon, which saw courts reading mens rea requirements into statutes and thereby restricting criminal liability; now we see courts reading mens rea requirements out of statutes and thereby expanding criminal liability. Thus, here again the norm constituted by the linguistic content of the court’s decision can be seen as a further determination of the statutory norm.

That completes my Standard Picture account of Greenberg’s alleged problem cases. Unlike the possible defenses of the Standard Picture mentioned at the end of Part II, it does not claim either that the courts got the law wrong in these cases or that they were identifying legislative intentions. Instead, it backstops the Standard Picture—an account of the valid norms of a system—with certain relevant structural features of the criteria of validity by which those norms are determined. It is neglect of such background facts of the legal system that makes these cases appear problematic to Greenberg in the first place. If one attempts to isolate the content of the law from the legal system in which that content has its home, then one is liable to end up with a warped view of that content.

b) Objections and Replies

I will now consider three potential objections to my proposed way of accommodating Greenberg’s purported counterexamples to the Standard Picture.

1. The courts in Greenberg’s example cases are identifying the content of the law

One potential objection to my account is that many lawyers and judges would agree with Greenberg that in a case in which the rule of lenity is applied (to take one of his examples), the court is doing what the law requires, that is, the court is identifying a preexisting valid norm to the effect that the defendant’s conduct was permitted. Is that fact not a strike against the plausibility of my account, according to which the court is exercising discretion? Central to my response to this potential objection, as it is to the whole of my account of Greenberg’s problem cases, is the distinction between identifying the law and applying it to a discrete case. Greenberg and the practitioners imagined by this objection elide that distinction, but it is not any the less viable for that. ⁴²

Provided one does not already accept an anti-positivist view for independent reasons, I do not see any compelling reason to think the norm that the defendant’s conduct was permitted was already determined prior to the court’s decision in a case applying the rule of lenity. Rather, in my view, the governing statutory norm was indeterminate on the question—that is why the rule of lenity was invoked. Some positivists might entertain the possibility that the rule of lenity is a criterion of legal validity; on such a view, while the statutory language itself was ambiguous, its correct interpretation, and thus the valid norm, was dictated by the rule of recognition. ⁴³ But the rule of lenity is not a criterion of validity in the U.S. system. In some cases in which the definition of a crime is ambiguous, courts apply the rule, but in others they do not. ⁴⁴ If application of an interpretive principle like the rule of lenity is ad hoc—as even the most ardent defenders of the traditional canons of construction, Bryan Garner and the late Justice Antonin Scalia, concede that it is ⁴⁵ —then that principle does not enjoy convergence and acceptance by legal officials from an internal point of view and consequently is not part of the rule of recognition. ⁴⁶ At the conclusion of a case in which lenity is a relevant consideration, the rule of recognition will pick out the court’s decision as a source of valid law irrespective of whether the court applied the rule of lenity or even considered it. ⁴⁷ A positivist might suggest, alternatively, that in applying the rule of lenity the court was doing what the law required because the rule of lenity is itself a valid norm. But the rule of lenity clearly is not valid in the full sense that the statutory norm is valid, for the reason just specified: its application to any given case is legally optional. Even Scalia and Garner find this a mischaracterization: a canon of construction “is not a rule of law but one of various factors to be considered in the interpretation of a text.” ⁴⁸

In short, the fact that lawyers and judges feel as though they are answering the question presented in the way required by the law does not entail that their answer (or any answer) is so required. In attempting to follow a rule we naturally take our decision to be determined by the rule. For practitioners the distinction between identifying a rule and applying it may make little difference, but as theorists we ought to be mindful of it. As I remarked in passing, Hart’s claim that in hard cases courts exercise discretion is not the psychological claim that judges feel as though they have a choice in applying the valid norm. It is rather the claim that when as a matter of accepted language use (perhaps among other social-fact considerations) a fact-situation falls in the valid norm’s ‘open texture’ or ‘penumbra of uncertainty’, either answer to the question presented is legally permissible. Greenberg’s purported counterexamples to the Standard Picture are just such hard cases, I contend, so the courts deciding them are exercising discretion, not merely identifying the content of the law.

2. This account represents a hollow victory for the Standard Picture

At the end of Part II, we mentioned two potential responses on behalf of the Standard Picture anticipated by Greenberg: that the courts in his example cases are getting the law wrong, or alternatively that the courts are really identifying legislative intentions. There I agreed with Greenberg that neither of these candidate defenses succeeds. Greenberg also anticipates my own kind of defense of the Standard Picture, in a way. He considers the possibility that the defender of the Standard Picture could accommodate his problem cases “by appealing to (the linguistic content of) authoritative pronouncements that specify how criminal statutes are to be interpreted.” ⁴⁹ Now, Greenberg certainly is not thinking of the judicial decision that applies the statute as the authoritative pronouncement contemplated by this response; he talks about courts following such an auxiliary authoritative pronouncement that gives instruction on statutory interpretation (presumably another statute), not themselves making such a pronouncement. Be that as it may, the reason he gives for dismissing this response applies equally to my own response. Greenberg grants that a legal system might arrange authoritative pronouncements hierarchically. But, he says, that prospect offers no aid or comfort to the Standard Picture:

[I]f SP is correct, the content of a criminal statute is constituted by the ordinary linguistic content of the statute.… It would be a major setback for the defender of SP to fall back to the position that although the legal norms contributed by statutes are not constituted by the ordinary linguistic contents of the statutes, this is consistent with SP because there are higher-level authoritative pronouncements … that require that statutes not be interpreted in accordance with their ordinary linguistic content. ⁵⁰

This ‘fallback’ position, as Greenberg calls it—that a norm that originates from a statute but departs from its linguistic content can be explained in terms of a relevant higher-order pronouncement—counts as a “major setback for the defender of SP” only if we accept the alleged implication of the Standard Picture stated in the opening sentence of his rebuttal: “the content of a criminal statute is constituted by the ordinary linguistic content of the statute.” But that statement is not supported by Greenberg’s original conception of the Standard Picture. ⁵¹ That conception was more general: the contents of legal norms are constituted by the linguistic contents of authoritative legal pronouncements. Greenberg now pitches the Standard Picture in a higher, more particularized register, insisting on a one-to-one correspondence not from legal norm to (some) authoritative legal pronouncement, but from legal norm to the particular statute to which it ultimately traces its lineage. Yet the Explanatory Directness thesis of the Standard Picture’s Core Model is not violated by the presence of any explanatory intermediary between statute and norm. It is violated only by the presence of an explanatory intermediary that is not itself an authoritative legal pronouncement. If the intermediary is an authoritative legal pronouncement—a judicial decision, for instance—then the Explanatory Directness thesis still holds good, for there remains an explanatorily direct relationship between the resultant legal norm and the linguistic content of an authoritative pronouncement.

Again, Greenberg is justified in wanting an account of the connection between the norm resulting from a judicial application of a statute, on the one hand, and the statute from which it originates, on the other. But the Standard Picture, as merely a theory of the content of legal norms, does not purport to provide that account, so its absence does not qualify as an objection to that picture. Such an account can be provided, as we have seen, only once we bring into the analysis certain relevant aspects of the U.S. rule of recognition.

3. This account cannot differentiate between right and wrong judicial decisions

A third potential objection is that, on my account, the explanatory role played by the originating statute is the same in all cases: namely, the norm that results from the court’s decision is explained directly by that decision and only indirectly by the statute. But it seems a criterion of adequacy, for any theory of the content of law, that it can distinguish the role a statute plays in a case rightly decided from its role in a case wrongly decided.

It is true that I have claimed that the norm that eventuates in each of Greenberg’s problem cases is explained directly by the judicial decision and only indirectly by the originating statute. But that characterization is apt only because these are hard cases: ones falling in the penumbra of uncertainty of the statutory norm. As Hart explains, in such cases courts inevitably exercise discretion, or a rule-making function. But the existence of hard cases at the margins of rules does not mean that there are not also easy cases. In such cases the fact-situation falls in the statutory norm’s “core of settled meaning,” which the court is “not free to depart from.” ⁵² That is to say, in such a case the court does not enjoy discretion: it can get the decision wrong.

The paradigmatic explanatory role played by a statutory norm in an easy case is a justificatory one: the court decides as it does for the reason that the statutory norm clearly requires the result. I said parenthetically before that, in an easy case, the statutory norm clearly applies on its own terms, so the court’s application of the norm does not modify it. That assumes the court applies the norm correctly. If it does, then the norm remains directly explained by its originating statute, which clearly encompassed the fact-situation. We might say that the judicial decision correctly applying the statute also directly explains the resulting norm, but the explanatory work that decision does is secondary. We could characterize it as causal, rather than justificatory: if the court had decided otherwise (i.e. wrongly), the resulting norm would be different. If a court misapplies a statutory norm in an easy case, on the other hand, the situation is reversed. The justificatory connection between the statute and the eventuating norm is broken, and the norm that results is directly explained only by the judicial decision. The only remaining, indirect explanatory connection between statute and norm is a formal, causal one: the statute supplied the norm the court (mis-) applied, so the statute remains the formal head of authority of the norm that results from the court’s decision.

The explanatory role played by statutes in hard cases, finally, is as I specified in my account of Greenberg’s problem cases. The hard case falls in the penumbra of uncertainty of the statutory norm, so that norm’s application to the facts is indeterminate and the court must make a choice. Once the court has made its choice, the resulting norm is explained directly by the judicial decision and only indirectly by its originating statute. There is no possibility of a direct, justificatory relationship between statute and eventual norm in such a case because the statute does not clearly require one application or another. Of course, we remain free to judge that the court got the decision wrong. But that conviction must find its justification in extralegal reasons, for the valid norm governing the case did not dictate a correct result.

* * *

Greenberg’s purported counterexamples to the Standard Picture drawn from patterns of statutory interpretation are premised on a demand that the legal norm that ultimately arises from a statutory provision be directly explained by the linguistic content of that provision itself. But that demand is not warranted by the Standard Picture. Once we take into account certain features of the U.S. rule of recognition—namely that it organizes sources of law into a hierarchy of authority, and that judicial decisions are sources of valid law—Greenberg’s examples are revealed as only apparent, and not genuine, counterexamples to the Standard Picture. The contents of the legal norms that result from the patterns of interpretation he identifies can be directly explained in terms of the linguistic contents of authoritative legal pronouncements: judicial decisions. Furthermore, those norms can be understood as modifications of the valid norms contained in their originating statutes, as judicial decisions are permitted ‘explanatory intermediaries’ of statutes by the rule of recognition.

IV. Implications for the Objection From Theoretical Disagreement

In closing, I want to note briefly a connection between the Standard Picture debate and another ongoing debate in Anglo-American jurisprudence: that over Ronald Dworkin’s objection from theoretical disagreement.

There are two kinds of possible disagreement about the law, Dworkin says: empirical and theoretical. ⁵³ Whereas an empirical disagreement is one over whether agreed-upon criteria of legal validity have been satisfied in the case of a particular norm, a theoretical disagreement is one over the identity of those criteria of validity themselves. Dworkin contends that theoretical disagreements among judges are common in the U.S. and U.K. legal systems. If so, that would be a problem for legal positivism. For according to Hart, the criteria of validity compose a legal system’s fundamental rule (its rule of recognition), the existence of which is supposed to be constituted by the social fact of the convergent behavior and attitudes of the system’s officials. ⁵⁴ But if the criteria of validity are what they are purely in virtue of empirical facts about official behavior and attitudes, then what room is there for judicial disagreement over the identity of those criteria? Would not such disagreement itself demonstrate that the criteria being disputed do not enjoy official convergence? Yet judges persist in offering their differing opinions in such cases as the decision required by the law, which (according to Dworkin) is to say the one “required by a correct perception of the true grounds of law [i.e., the true criteria of validity].” ⁵⁵ So those criteria must not, Dworkin concludes, be determined solely by empirical facts after all.

The instances of alleged theoretical disagreement Dworkin points to in support of his claim are cases in which judges disagree over the proper methodology for interpreting the source of law. His leading examples are two U.S. statutory interpretation cases in which the judges disagree over the right canon of construction or principle of statutory interpretation to follow. ⁵⁶ To date, Dworkin’s positivist respondents have granted him his presupposition that these cases indeed represent disagreements over the identity of the valid norm. Some interpretive method is needed, the thought goes, in order to identify the valid norm from the statutory language, and that interpretive method is what is being disputed in these cases. Consequently, positivists have been obliged to cast about for reasons that these official disagreements over the criteria of validity do not disprove positivism. Brian Leiter, for instance, has argued that the judges involved in theoretical disagreements are either in error or being disingenuous if advancing an interpretive principle or method that is not in fact converged upon. ⁵⁷ Scott Shapiro disagrees; he thinks theoretical disagreement too prevalent for an error theory to explain away, so he has abandoned Hart’s convergence account of the criteria of validity altogether and devised an alternative positivist theory of law: the planning theory. ⁵⁸

In this essay I have defended the view that statutory norms are explained directly by their statutes’ linguistic contents. ⁵⁹ Interpretive principles, like the rule of lenity and mens rea canon, are serving not to identify valid norms but rather to guide their application to difficult fact-situations with respect to which those norms are, on their own, indeterminate. (Indeed, I argued explicitly in Part III.b that the rule of lenity cannot be a criterion of validity in the U.S. system because its application to any given case is a matter of judicial discretion.) If that is right, then cases in which judges disagree over canons of construction or principles of statutory interpretation—like Dworkin’s purported examples of theoretical disagreement—are not really disagreements over the correct criterion for identifying the valid norm, at all. They are rather disagreements over the best application of that norm to an unforeseen and difficult fact-situation, and the best reasons therefor. Disagreements over the application of a valid norm, unlike disagreements over the criteria of validity, pose no threat to legal positivism.

Obviously, there is much more to say in this connection. This essay has aimed primarily to present a plausible account of Greenberg’s problem statutory interpretation cases that is consistent with the Standard Picture. If it has succeeded, then as these concluding remarks have indicated, there may be an additional theoretical benefit for positivists to reap.