Norms constraining intended meanings as antidotes to the foregoing infelicities

3.4.1 Substantive constraints

3.4.1.1 Norms for avoiding substantively infelicitous results

Substantive constraints on rule-makers' determinations, although quite important practically, are relatively unproblematic jurispruden- tially.

Substantive constraints such as one denying the authority of any legal rule that is absurd or grossly unjust (by the interpreter's standards) are different from epistemological principles that help interpreters discover what the legal rule-makers intended ought to be done. That some result would be absurd or grossly unjust can be evidence - often strong evidence - that the rule-makers did not intend that result. Nonetheless, as we have stressed, rule-makers can intend absurd or unjust results, even if not under those descriptions. That is, it is pos­sible that: (1) rule-makers wish to require only what is just and not absurd; (2) rule-makers intend that X be done; and (3) X is unjust or absurd. Simply put, rule-makers can intend unjust or absurd results because they make mistakes. And a norm that directs interpreters to disregard intended absurd or unjust results operates as a constraint on the rule-maker's power to determine authoritatively what ought to be done rather than as an aid to understanding what the rule-makers did in fact determine.

Substantive constraints on the rule-makers' determinations can func­tion as absolute limits, much as do ordinary constitutional norms that limit the authority of governmental actors. Alternatively, they can function as do artificial evidentiary presumptions by directing inter­preters to resolve uncertainties about the rule-makers' intentions in favor of certain outcomes. Thus, if the rule-makers' intentions are not clear, such substantive norms might direct interpreters to resolve the ambiguity in favor of the result best supported by the evidence regard­ing the rule-makers' intentions, which evidence includes the fact that one result is more just or wiser than the other.^[33]

Both types of substantive constraints, however, must be the product of a decision-making body that possesses higher authority than the rule-maker thereby constrained. That is so because both absolute and presumptive constraints on a rule-maker result in departures from that rule-maker's intended meaning and thus depart from that rule-maker's determination regarding what ought to be done. And the norms that are the product of this higher authority decision-making body - be it a constitutional ratifying body or ultimately, insofar as the substantive norms are preconstitutional norms on which the constitutional norms ultimately rest, the citizens who accept the norms - must mean what their authors intended them to mean, whether the authors are the con­stitutional ratifying bodies or the people themselves ‘authoring' norms by accepting them.

3-4-1-2 Norms for effectuating specific policies

The common law is replete with doctrines that direct judges to dis­regard the intended meanings of documents that are otherwise sup­posed to be canonical statements regarding parties' legal rights and duties. These include the parole evidence rule for the interpretation of contracts, the doctrine that an ambiguous contract shall be construed against the party who drafted it, presumptions against disinheritance, and many others.

Similarly unproblematic are substantive constraints on statutory inter­pretation that derive from the higher law of the Constitution. The doctrine of lenity, for example, supposedly effects the policies of due process and ex postfacto law clauses that seek to protect actors against being charged with crimes based on non-obvious (to ordinary people) intended meanings of criminal statutes. The doctrine directs the courts to give criminal statutes the most restricted meaning where more than one meaning is possible. In other words, the courts are to imagine that criminal statutes are drafted by hypothetical legislators who seek to limit the scope of criminal liability perhaps more than did the actual legislators who enacted those statutes.

As we said, none of these doctrines that constrain the interpretation of canonical legal texts on behalf of substantive policies requires depart­ing from the ordinary human practice of seeking authors' intended meanings. Any of us can imagine that a document was drafted by someone other than its actual author and determine what it would have meant in such a case. It takes no special legal training to answer the question of what would ‘I'll make him an offer he can't refuse' mean if, instead of Don Corleone making the statement, it were made by our sweet, good-natured real estate broker.

There is an interesting problem here, however. If courts are directed to imagine hypothetical authors who differ from the actual authors along certain dimensions, there have to be additional constraints added to keep from completely undermining the ability of actual authors, whether private or legislative, to have their intended meanings heeded. This is because any symbols can be employed to communicate any intended meaning. Thus, as an example, if the doctrine of lenity instructs courts to ‘interpret' criminal statutes narrowly, what stops them from interpreting those statutes ridiculously narrowly? It would not be the actual intended meaning of the actual author, for we have already dispensed with that in invoking the doctrine of lenity. Nor would it be the actual meaning of the words; for, as we emphasized in the preceding chapter, actual words - certain shapes and sounds - do not carry a meaning if there is no author intending a mean­ing by those words. So the doctrine of lenity and the various other similar doctrines that, in service of substantive policies, constitutional and non-constitutional, direct judges to disregard the actual intended meanings, need to supplement the characteristics of the target hypo­thetical authors beyond those we have mentioned. So if the hypotheti­cal legislators of criminal statutes are supposed to be motivated to limit the scope of criminal liability, judges need to know by how much.

The most obvious way to avoid such results is to assume that the hypo­thetical legislature or will drafter uses, say, standard English - perhaps definitions listed first in a designated dictionary - and standard gram­mar. Doing so will in most cases put limits on the hypothetical authors that will prevent courts from undermining all criminal liability in the name of lenity and undermining the ability of drafters of private docu­ments, such as wills and contracts, to accomplish their ends.

One doctrine that directs courts to disregard actual intended meanings for substantive policy reasons but that raises a worry of a different kind is the doctrine of interpreting statutes to avoid having to resolve constitutional questions - the so-called Ashwander doctrine in United States constitutional law.^[34] The problem with Ashwander is not that it directs courts to look to what a hypothetical legislature would have meant by a statute rather than to what the actual legislature meant. The doctrine of lenity does that as well. The problem is that, at least arguably, although the doctrine of lenity is a doctrine commanded by the Constitution, the authority of which is superior to that of the leg­islature, the Ashwander doctrine does not implement any high-order legal norm. Interpreting to avoid constitutional issues is not itself a constitutional command. So when a court following Ashwander dis­regards the actual intended meaning of a statute and ‘interprets' as if the statute were authored by a hypothetical legislature intending the statute to steer clear of any constitutional limits, the court is exercising only the power it has as a court to interpret statutes and not the power it possesses to strike down unconstitutional statutes.

3.4.2 Procedural constraints

Procedural constraints on authorities' determinations may dictate the form that rule-makers' determinations must take and may deal with situations in which rule-makers have only apparently but not really determined what ought to be done. The norms that dictate form reflect the rule of law value of the accessibility of law. The norms that handle cases of failed law reflect both the rule of law value of accessibility and the more substantive value of consistent policy.

3.4.2.1 Norms of form

We are quite familiar with certain formal constraints on rule-makers' determinations that must be satisfied before those determinations are deemed to have legal effect. For example, we have norms covering what counts as a ‘vote' by a legislator to enact a rule, which ‘vote' not only signifies that the voting rule-maker's intention will be counted but also marks the moment in time at which the relevant intention must exist. One who votes ‘aye' when the roll is called is counted as in favor of the proposed bill for purposes of determining whether the bill has become law, even if she is subjectively opposed to the bill, and even if, moreover, she believes ‘aye' means ‘opposed'.

Notice that, in the case of the rule-maker who votes ‘aye' thinking it means no, if that person provides the necessary vote for passage, the law is deemed to have been passed even though in fact a majority of the rule-makers intended that it not pass. In such a case, the law is not what the (majority of) the rule-makers determined ought to be done. When the law comes to be applied, what does it ‘mean'?

We could have a norm that provided that, in the case just described, the law means what the rule-makers voting in favor intended to mean by it, except that for the person mistakenly voting aye, the law means merely what she thought it would mean if passed. She herself did not

intend that meaning because she did not intend for the law to exist as law.

We need some such procedural norm because we have a gap to bridge: the gap between what the (majority of) rule-makers determined ought to be done - nothing - and what our norm regarding voting provides is the case, namely, that a law was passed. If we say that because of the mistaken and pivotal ‘aye' vote a law was passed that has no applica­tions, we undermine the norm regarding what count as ‘aye' and no votes. And without that norm, there will be continued uncertainty regarding what laws exist, an uncertainty that undermines the rule of law value of the knowability or accessibility of the law.

It is a short step from familiar procedural norms about voting to some similar but perhaps less obvious procedural norms. Consider the fol­lowing procedural norms that might constrain the rule-maker's deter­minations. (1) All texts shall be interpreted as if they were written in the standard English of the date of enactment, with respect both to the meanings of the words used and to grammar and punctuation. (2) All texts more than 100 years old shall be deemed of no legal effect (or, more limitedly, of no legal effect if the standard English meanings of any of the text's words have changed in the 100-year period).

Norm (1) reflects the rule of law value of legal accessibility. It bars authoring legal texts in the rule-makers' private code or in a different language from the populace. It thus prevents recondite law. We discuss norms such as norm (1) more fully in the next chapter when we con­sider textualism as a theory of legal interpretation.

Norm (2) reflects the same value - that recovering rule-makers' inten­tions becomes more and more difficult as the moment of enactment recedes further and further into the past. With norm (1) in existence, norm (2) is perhaps less important but surely not unnecessary, espe­cially if modified as indicated in parenthesis.

Norms (1) and (2) should be contrasted to proposals that might appear to be similar. For example, some have proposed that statutes be given the meaning their words would standardly carry if authored at the time of interpretation or application (see Alexander 1990; Aleinikoff 1988). One impetus behind such a proposal is to make law's mean­ing even more accessible than it would be under norm (1). There are also substantive concerns about policy obsolescence that underlie that proposal. This form of ‘updating' of statutes, however, has a markedly different effect from norm (1) on the rule-makers' ability to carry out their role. Norm (1) forces rule-makers to consult the dictionaries and grammars of their time in order to maximize their ability to effectuate their determination of what ought to be done. The ‘updating' proposal, on the other hand, reduces their control to that of selecting the lan­guage (English) and the marks but then leaves the translation of those marks to the fortuity of subsequent changes in the language.^[35]

Norms (1) and (2) are purely procedural norms that attempt to rec­oncile the role of rule-makers - to determine what ought to be done - with the rule of law value of legal accessibility. Norm (2) effects the reconciliation by restricting the temporal scope of the rule-makers' authority. Norm (1) effects the reconciliation more or less well depending on the rule-makers' skill in drafting so as to communicate their determinations accurately according to the grammar and dic­tion of the time. Where the reconciliation is imperfect - where the rule-makers fail to draft so that their determinations are conveyed in standard English - the law will not be what they determined ought to be done. Rather, the law will be a product of their determination regarding marks or sounds and the independent process of codifying standard English meanings and grammar. It will be to some degree ‘mindless', in that the codification of meanings and grammar will not be a reflection of anyone's determination of what ought to be done in the world beyond dictionaries and grammar books. Unlike the ‘updat­ing' proposal, however, norm (1) does not deprive rule-makers of con­trol over effectuating their determinations. It demands linguistic skills but not linguistic prescience.

3.4.2.2 Norms for failed law

The procedural norms in this category are somewhat different from the norms of form. Instead of dealing with how to make the rule­makers' determinations more accessible, they deal with the problem of what to do when the rule-makers appear to have determined an issue but actually have not. Such cases of failed law, as we call them, are phenomena associated with multi-member legislative bodies, though they occur as well in other multi-member rule-making bodies, such as administrative boards, appellate courts, and constitutional ratifiers. They occur when the individual rule-makers, despite having agreed on the language of their legislative text, mean different things by it. In other words, norms for failed law deal with the problem of conflicting intended meanings within a multi-member rule-making body.

When the general legislative norm is that rule-makers' determinations have the force of law only if a majority (or, in some cases, a supermajor­ity) of the rule-makers concur, and members of the (apparent) majority have made different and potentially conflicting determinations regard­ing what ought to be done despite having agreed on the words of a text, then it is possible that there is no majority determination having the force of law. Multi-member legislative bodies as such have no intentions regarding what ought to be done. Rather, they have only an aggregation of the individual intentions of the members. Aggregation of individual intentions is not a problem if there is a majority with identical intentions. Where those individual intentions come apart, however, the passage of a law may be illusory in this sense: despite the appearance of legislative (majority) agreement that there be such a law, every possible intended meaning of that law would have been opposed by a majority of the legislative body. Put differently, no majority deter­mined any possible meaning for the law.

A norm of form such as one that irrebuttably assumes that all laws are intended to mean what they mean in the standard English of the date of enactment can save some laws from the failure to command a majority of consistent individual intentions. Such norms will not handle all cases, however, for even standard English will leave open the possibility of ambiguity. And in such instances, by application of Arrow's theorem (Arrow 1963), we face the possibility of majorities in favor of a law and its language but against all possible meanings of that language.^[36]

Consider this stylized representative case. Rule-makers A, B, and C vote on term T. T is vague and can mean W, X, and Y or W, X, and Z. A votes ‘aye', intending W, X, and Y. B votes ‘aye', intending W, X, and Z. C votes no. A would have voted no had he been presented with B's definition as his only choice. B would have voted no had he been presented with A's definition as his only choice. Moreover, a truncated version of T, one that covered only W and X and neither Y nor Z, would likewise have been rejected. It is clear in this case that we have no majority of the authorities in favor of any possible meaning of T, even though we have a majority in favor of T itself.

Or consider a simple case of ambiguity, such as would occur if there were an ordinance forbidding ‘meetings by the bank'. A voted for it to prohibit gatherings by the river. B voted for it to prohibit gatherings by the town's financial institution. C voted against it, as would have A or B had ‘bank' clearly meant what the other intended by it.

One way to handle such cases of failed law is to bite the bullet and say that, despite appearances to the contrary, those legal texts are not laws. They are only apparent laws.

Alternatively, one might decide that having apparent laws on the books is undesirable for two related reasons. First, the existence of appar­ent laws produces uncertainty in those subject to the law. In many or most cases, it will be difficult for the average subject to determine if a particular law is real and meaningful or only apparent. The uncertainty implicates the rule of law value of accessibility.

Second, the existence of apparent laws and the consequent uncertainty makes it difficult for legal authorities to carry out their functions. For if they are uncertain about the content of the existing laws, they will be severely handicapped in deciding how best to legislate, prosecute, and adjudicate.

It might be desirable, therefore, to have norms that direct official inter­preters to breathe meaning into laws that are actually meaningless - in effect, to adopt the language of the existing laws but to ‘reauthor' those laws so that the laws reflect the interpreters' determinations of what ought to be done (within the constraints of the norms of form). Such norms would make the official interpreters into the primary rule-makers, though constrained by the earlier rule-makers' choice of language.^[37] Although many commentators propose such norms, they frequently make the mistake of conflating actions taken in pursuance of such norms with interpretation of the statutes or constitutional provisions that exhibit the failure of law. Those statutes or constitu­tional provisions are not being interpreted, however, for interpretation would conclude that they are meaningless marks. Rather, the statutes and constitutional provisions are being reauthored by a new set of legal rule-makers.

Again, it should be emphasized that, as with substantive norms, these norms of form and of failed law must be the products of a decision­making body that possesses higher authority than the rule-maker to which they apply. That is so again because these norms lead to departures from the norm-constrained rule-maker's intended meaning regarding what ought to be done. And, just as with the substantive norms, these norms of form and of failed law themselves mean what the higher authority decision-maker intends their meaning to be.

3.4.2.3 Levels of generality of rule-makers’ intentions

It is a common observation that the intentions of legal authorities can be described at various levels of generality. Thus, rule-makers may intend a law to accomplish specific results R in order to further a more general purpose P, which purpose furthers a still more general purpose ‘P', which purpose furthers Goodness and Justice. The rule­makers obviously think those various intentions are consistent, which is why they passed the law in question. However, in the view of those interpreting the law, the intentions may turn out to be inconsistent and inconsistent at any level. Thus, R may not in fact further P, P may not in fact further ‘P', and ‘P' may in fact be inconsistent with Goodness and Justice. So some people believe that this raises the ques­tion: If the interpreters are to give effect to what the authoring rule­makers intended, at what level of generality should that intention be described?^{[38] [39]}

Some believe that this question can be answered only by reference to an interpretive norm chosen because of its anticipated good results, and that the question cannot be answered in the absence of such a norm. We believe that view to be mistaken, as we made clear in the preceding chapter. A norm that directs interpreters to correct rule-makers' mis­takes regarding how their actual intended meanings square with their more general purposes is one that threatens completely to undermine the rule-makers' role of determining what ought to be done. Because rule-makers always intend to achieve Goodness and Justice - to ‘do the right thing' - if they are acting legitimately, the interpreter can substi­tute his own views about what Goodness and Justice require for any specific intended meaning of the rule-makers and still claim to be hon­oring their (most general) intent. For the interpreter will undoubtedly believe that, had the rule-makers been disabused of all their mistakes of fact, mistakes of means-end reasoning, and mistakes in reasoning about values, they would have enacted (intended to mean) what the interpreter would have enacted (intended to mean). In effect, a norm allowing the interpreter to disregard the more specific intended mean­ings in favor of more general purposes will convert the interpreter into the rule-maker (and, of course, make the interpreter's determination subject to being corrected by a subsequent interpreter, and so on) (see Epstein 1992). As we said earlier, in our discussion of Lessig's theory of interpretation,¹¹ if the rule-makers' mistakes are always to be corrected by interpreters, then there will be no rule-makers: if the interpreters are tantamount to the rule-makers, there can never be rule-makers for interpreters to interpret.

One might reply that, although a norm directing interpreters to follow the rule-maker's most general purposes would undermine that rule­maker's role, a norm directing interpreters to follow the purposes characterized at some intermediate level of generality would not undermine that role. The interpreters could correct the rule-maker's specific intended meaning by reference to somewhat more general purposes, but not by reference to Goodness and Justice.

There are two problems with such a response, however. First, there is no one way to describe the levels of generality or to count them. It is thus impossible for a norm to specify the precise level of generality interpreters should look to in following the rule-maker's purposes.

Second, and more important, the choice to correct the rule-maker's specific intended meaning in light of the rule-maker's more general ^[40] purposes but not in light of the rule-maker's most general ones seems quite arbitrary. In some cases, for example, the rule-maker might spe­cifically intend a meaning that is inconsistent with a more general pur­pose she has, but which is in fact consistent with Goodness and Justice. In such a case she has done the right thing, although she has made two mistakes in reasoning from means to ends that happily cancel each other. The hypothesized norm directing interpreters to follow the rule-maker's purposes characterized at an intermediate level of generality would result in the interpreters acting contrary to Goodness and Justice when the rule-maker, having made some lucky mistakes, actually intends a policy that is consistent with Goodness and Justice.

Nonetheless, if there were no fact of the matter regarding what the rule-maker intended in the sense of at what level of generality to char­acterize that intent, we would have no choice but to construct a norm that would prescribe the level of generality for interpreters to follow. In our view, however, there is a fact of the matter. This facticity about what the rule-makers determined ought to be done in the face of the various levels of generality at which their purposes could be described is what makes formal legal rules - rules that are opaque to their back­ground purposes - possible. If there is a fact of the matter about what the rule-makers determined ought to be done, then that fact, possibly as filtered through norms of form and norms for failed laws, should guide interpreters. Otherwise, interpreters are not interpreting, and the role of the rule-making authorities does not exist.

Thus, we believe that, despite their having purposes of varying degrees of generality that can come apart, the rule-makers' intended meanings are not a function of the interpreter's choice. Rather, as we argued in the preceding chapter, there is a fact of the matter regarding intended meaning; and the rule-makers' intended meaning is, obviously, a func­tion of the rule-makers’ choice, not the interpreter's. Moreover, there needs always to be the possibility, if not the certainty, for intended meaning and more general purposes to come apart in order for there to be determinate rules that perform the morally demanded settlement function.

It is always possible that the search for the rule-maker's intended meaning will reveal in various infelicities. And it is always possible that higher-order norms will help minimize or negate such infelicities. Substantive higher-order norms can combat substantive infelicities, such as rules whose intended meanings are absurd, unjust, or obsolete.

The higher-order norms could themselves be determinate rules: for example, ‘No statute shall be effective one hundred years after enact­ment'. Or they could be standards: for example, ‘No statute shall be given effect if it is absurd, unjust, or obsolete'. In the latter case, the higher-order norm delegates decision-making authority to some other institution - the interpreting one - to determine what is absurd, unjust, and not obsolete. The attempted settlement by the body whose rule is deemed absurd, unjust, or obsolete will be undone. The interpreting institution may itself at least partially settle the meaning of absurd, unjust, or obsolete, or it may just leave the meanings of those standards unsettled except insofar as the particular rule is concerned.

Procedural higher-order norms can deal with rules that have no intended meaning (‘failed laws') or have an inaccessible meaning. Application of these higher-order norms is not interpretation of the rules. Interpretation is the recovery of the rule-maker's intended meaning. Higher-order procedural norms do not aid in that endeavor and are not meant to do so. They direct ‘interpreters' to construct rules out of materials that the rule-makers have provided. The meaning that the ‘interpreter' will assign is not the statute's intended meaning but a meaning that is the product of the marks on the page that the legisla­ture produced coupled with a higher-order norm.

In the next chapter, when we examine textualism, we argue that norms of form such as the hypothetical norm ‘read as if in standard English' are less attractive than they might appear. The other high-order norms we have mentioned may or may not be worth having. We take no position on those. But even if they are, neither they nor the infelici­ties they are meant to address cast any doubt on our central claim - namely, that interpretation of legal rules is a search for the rule­maker's intended meaning and is no different from the interpretation of ordinary demands and requests.^[41]