Obligations, Interpretivism and the Legal Point of View


Nicos Stavropoulos

The Legal Significance of Politics

Some political action is legally relevant: it has an impact on the law, roughly defined as the change in legal rights and obligations (powers, privileges, immunities or liabilities) that obtains, without further political action, as a result of some such action. All legally relevant action taken together constitutes legal practice. Politics plausibly includes much more than action that has an impact on legal rights and obligations, and one of the tasks of legal theory is to specify which kinds of action are legally relevant and what it is, exactly, for such action to have impact on the law. Even if we confine our attention to standard examples—enacting statutes, adopting regulations, deciding cases—some hard philosophical questions arise. What precise aspect of this kind of action is legally relevant? Why so—what gives a given aspect of institutional action legal relevance?

On a version of legal positivism that I shall call the orthodox view, the legally relevant aspects of institutional practice do not include any substantive normative facts such as facts about what is right or good or valuable or what we have reason to do or want or believe. (For convenience, I will call facts other than substantive normative facts “nonnormative.” As I will be using those terms, that people ought to have a chance to participate in political decision-making, if true, is a (substantive) normative fact. By contrast, that someone asserts or believes or has some other attitude towards the claim that people ought to participate in that way, or that “democracy” means the rule of the people, are all, if true, examples of nonnormative facts.)

On the orthodox view, institutions shape the law by communication. The fact that a certain statute was enacted and what its text means together explain the impact of its enactment on the law. What the text of a statute says, including what it logically entails, corresponds to one or more new norms that the statute created or determines the extent to which its enactment modified preexisting norms. The law is the set of norms produced in that way (perhaps together with some norms implicitly endorsed by institutions in their settled practice).

On some competing views, institutional communication is only part of the explanation. The law further includes norms that justify institutionally produced norms, or excludes institutionally produced norms that fall below some threshold of decency. On such views, it is not true that history and language alone explain the legal impact of institutional action. Substantive normative considerations, plausibly narrowly moral ones (facts about what we owe to each other), are part of the explanation.

Confining our attention to the way in which these two kinds of view differ, however, is misleading. This is but one way in which the orthodox view contrasts with alternatives regarding the question of whether substantive normative considerations must figure in the explanation of the legal impact of politics. There are other, more fundamental ways. Suppose a theory agrees with the orthodox view that the only thing that matters to the legal impact of an enactment is what the enacted text means, but adds that that is so because government is morally bound to give citizens fair notice of how it plans to treat them. Legal rights and obligations are only those announced by institutions, because government would otherwise wrong its citizens. On this view, the complete explanation of the legal impact of politics includes reference to moral considerations. But the place of those considerations in the explanation is more basic. Morality, on this view, does not directly determine legal impact. It explains why the other, nonnormative, aspects of political action have legal relevance.

The more fundamental controversy brings into play the often overlooked second of the questions identified above, of what gives legal relevance to certain aspects of the practice. It concerns the place, in the complete explanation of the legal impact of politics, of a substantive normative explanation of why certain aspects of the practice should have the relevance that the theory assigns to them. This issue will be my focus in this essay.

The orthodox view has it that the explanation must begin in nonnormative elucidation of what is legally significant in legal practice and why, by the lights of the practice itself. From that “legal point of view” we notice that the practice assigns to some aspects of itself—institutional directives—a certain kind of binding force. The unqualified normative question of why those parts of the practice should have the contemplated role comes last, and concerns whether the law in fact has the force it claims for itself—whether the law so understood ought to be obeyed. Interpretivism, by contrast, begins by rejecting the claim that nonnormative elucidation of the point of view of the practice determines the correct explanation of the legal relevance of institutional practice. Rather, the explanation of how legal practice matters to legal rights and obligations is an ordinary explanation of the normative effects of action by appeal to the substantive normative considerations that give it that role. For interpretivism, the unqualified question of why legal practice matters must come first. If the orthodoxy is right on that matter, interpretivism cannot get started.

The existence of this fundamental controversy undermines the significance of a familiar argument offered in support of the orthodox view and against its competitors. It is a familiar orthodox thesis that the law claims authority. For the law to claim authority is for law-shaping institutional action to have a certain logic, namely to assign to itself the characteristic role of authoritative directive in the practical reasoning of subjects. An authority directs others to take action for the reason that the authority so directed—because it said so. It therefore guides action by means of replacing by authoritative directive the reasons that apply to subjects independently of the directive. Legal institutions claim authority in that they purport to regulate domains of action by their directives, so it must be at least on the cards that their directives might play the practical role just described. But institutional directives wouldn’t even be eligible candidates for that role if, of necessity, legal rights and obligations were determined even in part by substantive considerations of the sort that the directives were meant to replace—e.g., by the moral principles that governed the relevant domain of action. So theories that make the identification of law dependent on substantive normative considerations of that kind are untenable because they contradict the very logic of legally relevant institutional action.

This argument from law’s claim to authority presupposes that the fundamental controversy between the orthodox approach and interpretivism has been resolved in favor of the orthodoxy. But if it has, it’s over for interpretivism. The argument from the claim to authority is redundant.

The Normative Significance of Action: The Formal Approach

The fundamental controversy with which I am concerned also arises in connection with the explanation of some other phenomena, including making a promise, decision or request—actions or attitudes which are generally understood to result in some distinctive obligations or to have some other distinctive normative significance or impact, or at least to be capable of so doing. The relevant theoretical choices are posed particularly clearly in these domains, so I will explore them in some detail in relation to promising.

The first issue, which arises in all these cases, concerns the choice between two approaches to the explanation of how a given kind of normatively significant action produces its distinctive normative effects. On the first, formal approach, each kind of such action has its own characteristic logic. Investigation of the logic of the action is nonnormative in character. Taking promising as a familiar example, this investigation tells us how the action looks from the inside. This is not empirical investigation of the psychology of promisors, but conceptual reflection about what it is to take such action, in terms of the attitudes the action implies (which may diverge, at least in certain ways, from those that actually obtain). It explores promising from the perspective of the agent, explaining what it is to occupy that position and to take the action in question. The availability of this perspective allows the theorist to avoid substance. He can set out to describe, not what genuinely matters in the action of promising and therefore what explains why it should normally result in obligations on the part of promisors, but rather, suspending judgment, describe how the action presents itself, including what normative relevance the action, broadly understood to include the attitudes that it implies, assigns to itself. This is not to say that the explanation is merely epistemic, aiming only to describe how promising looks. Rather, it aims to define the subject matter of promising—what promising is and how it works—by appeal to our conceptual scheme, our shared understanding of promising. To do so, it must make explicit the criteria we generally follow, usually unreflectively, in making judgments and drawing inferences that involve the concept of promising; or to articulate the conception that we implicitly possess, which guides such judgments and inferences. This is only a rough characterization of the task. There is much controversy as to what exactly conceptual analysis consists in. Typically, analysis is presented as an explanation of what is implied in saying “I promise to f.” But the word “promise” is not essential, and arguably analysis has further tools in addition to reflection about the meaning of characteristic utterances through which one promises, e.g., investigation of common reactions to hypotheticals that involve the concept of promising. (Would we say that an act of assuring others that one will take some action in the future, which failed to induce expectation of performance, binds as a promise? How about an assurance given without an intention to perform? And so on.) One thing seems clear: analysis must aim to derive the complete explanation of its target concept from convergence in the applications, judgments and inferences that competent speakers or thinkers are disposed to make or draw. A theorist who defends some analysis must think that the analysis is made true by these considerations, even though he needn’t interrogate actual speakers in order to construct it.

Promising, it is generally agreed, is action that normally results in the promisor being under an obligation to perform another action. The complete explanation of that phenomenon, running from the promising act to the obligation, must include an account of what precise action the promisor becomes obligated to perform; of what it is about the action of promising that has that effect—what is the aspect of the action that is relevant to the production of the obligation; and why—what makes that aspect relevant. You say to your friend Peter: “I will read your draft and send comments by Friday; I promise.” Suppose that several things are true at the time. You firmly intend to do what you said you will do and you mean to convey that fact. You also understand that Peter wants assurance that you will read the draft and send the comments, you aim to give him such assurance and you approve of him counting on that assurance in making further plans. All of that matches Peter’s understanding of the situation. You also intend that you come to be under an obligation to do what you said you will do regardless of the assurance sought and given, the expectations you believe Peter would be justified in forming and the prospects of his making plans on that basis. You take yourself to have come under an obligation, namely to read the draft and send Peter the comments, just by saying that you will do so. There is a small discrepancy, however, between each party’s understanding of what you’ve committed to do. You take yourself to have committed to sending comments at any time Friday—by midnight, at the latest—whereas he takes you to have conveyed that you commit to send the comments by close of business. This is only a small sample of colorable candidates for some role in the explanation of your coming to be under some determinate obligation by virtue of your speaking as you did. At first sight, any of the candidates might be relevant. Which, if any, are in fact relevant? What makes them relevant?

It is an essential part of the formal approach that nonnormative reflection on the common understanding of the concept of promising can answer all questions such as these to which an answer exists; as for those it can’t, they admit of no determinate answer. A prominent example is provided by Raz (1972, 1977, 1978). On Raz’s view, conceptual reflection shows that someone who understands what promising is, and acts sincerely, promises when he conveys an intention to undertake, by the very act of conveying that intention, an obligation to perform some other action, namely the action the agent said he will take. This means that the act of promising as such consists in the conveying of the intention just described. Other candidate attitudes or other factors—e.g., the intention to give assurance, the fact that the recipient sought such assurance and so on—may be present in standard examples but are ultimately irrelevant. The key factor, the conveying of an intention to be bound by that very act, entails that the promisor regards his own action, itself, as a new reason for the other action (Raz 1978: 137). It entails that the promisor wants to direct his own future action by his say-so.

This explanation implies a principle to which promisors are tacitly committed, which, if valid, would give the act of promising as explained (conveying the intention just described) the contemplated normative significance (the role of an obligation to perform). (On a social-practice version of this claim, the principle is a social norm that promisors accept and expect others to follow when they promise.) It is important to be clear on the relation between the attitudes and the standard in play, specifically regarding what explains what. The view in discussion is that the attitudes explain the standard, not the other way around. It’s not a standard, otherwise constituted, against which a promisor’s attitudes are assessed (or else the theory could not confine itself to elucidating others’ attitudes, as opposed to critically assessing them). Rather, the standard is what it is in virtue of the attitudes, which therefore cannot correct them. The standard is to be extrapolated from the attitudes. It’s because promisors have a certain intention or expectation or consider something or other relevant that they follow a standard on which those considerations are relevant to promissory obligation.

This nonnormative part of the explanation anticipates and shapes the second, normative part. The nonnormative part entails how things must be if an act of communication, assumed to be successful as such, is to achieve its intended normative effects, i.e., if the obligations that it was intended to produce in the specified way are actually to obtain. Only at this point are we able to move to the second stage of the complete explanation of promising, where we consider whether things are so, and therefore whether promising does impose on promisors the obligations they purport to assume. The second stage is substantively normative in character. We consider at this stage whether promises so understood ought to be kept. To continue with the example, this is to investigate whether the conditions, under which people would get to place themselves under obligations by conveying an intention thereby to do so, actually obtain. In this part, we would be looking to justify the principle to which promisors are tacitly committed. We would do so if we could establish that it is good in some specific way that people should get the power to bind themselves by their say-so under those conditions. It is only at this stage that we consider the question of why promises (understood as the conveying of the intention described earlier) should matter, as a question about their genuine normative relevance.

The formal approach naturally splits the explanation of promising into two parts: the first, nonnormative part concerning what force promisors mean to secure for their action and therefore the conditions under which they might succeed in securing it; and the second, normative part concerning the question of whether they do so succeed. Obviously, success is narrowly defined. Promisors succeed only when, as a result of their action, an obligation to perform comes about in the way they aimed at, as that is explained by the formal part of the account.

This means that the formal approach presents to normative scrutiny a package of candidate obligations, whose existence and content depend only on the considerations identified by nonnormative analysis—on Raz’s suggestion, on a promisor’s linguistic action alone. Promissory obligations so defined would obtain if substantive normative considerations ratified the package as is. This would be to assign to promisors normative powers to bind themselves by their say-so (Raz 1977). The only other alternative is to reject the package altogether. It is conceivable that promising might never succeed: it might turn out that the normative power it presupposes does not really exist. If so, we would have discovered that promissory obligation does not exist.

This view need not deny that promising may have, perhaps often or even always, further normative effects including obligations, by virtue of other considerations that give the act that role (for example, because the act normally induces expectations on the part of promisees that it would be wrong to frustrate). But these effects would be irrelevant side effects, not the effects of promising as such. That may also be the case with someone who uses the language of promising while his actual attitudes fail to conform to the analysis. On this view, analysis presents, as essential to promissory obligation, the specific intention, not merely to come under an obligation but to do so just because of the act of conveying the intention; in other words a claim of a power to bind oneself by one’s say-so. Absent the claim, the action would not count as promising and therefore distinctively promissory obligation could not obtain.

The Normative Significance of Action: The Substantive Approach

The other, substantive approach, does not expect to find defining conditions in convergent applications, judgments or dispositions to draw inferences. It treats standard examples of promising, understood to include colorable candidates for some role in the explanation of promissory obligation, and intuitions about the act’s binding nature, only as a starting point. It then asks a substantive normative question: what is it about these paradigm acts and surrounding attitudes that might make it the case that the acts produce obligations?

Importantly, this approach need not deny that people have intuitions, not only to the effect that promising binds, but also specifically about how promising binds (what aspect of promising, broadly understood, is relevant to the explanation of how the act produces obligations and why). However, and even on the assumption that there is a unique pattern of such intuitions shared among competent users of the concept of promising, the approach treats them as hypotheses, which theory checks and may correct, rather than as constraints that theory must respect on pain of changing the subject.