Application of the rules requires the use of one of two methods: deduction or analogy. When using deduction, the lawyer determines whether the facts of the situation are or are not described by the factual predicate of a rule and thus whether the legal consequence imposed by the rule does or does not apply to the situation. When using analogy, the lawyer determines whether the facts of the situation are or are not like those described by the factual predicate of the rule and thus whether the legal consequence imposed by the rule does or does not apply to the situation. In applying either of these methods, the lawyer uses rules to determine the rights and duties that exist in the situation and thereby completes the legal reasoning process.

As will be seen, however, deduction and analogy often produce indeterminate results. That is, the lawyer cannot determine with sufficient certainty whether the facts of the situation are described by, or are like, the situation described in the factual predicate of the rule. In that case, the lawyer must refer to the underlying policies to ascertain the rights and duties that exist in the situation.

The application of policies to a situation involves the use of methods quite different from those used to apply rules. Rather than comparing the situation of the case to that described in the factual predicate of the rule, the lawyer determines whether imposing the legal consequence described in the rule would further the underlying policies to a greater extent than not imposing the legal consequence. The application of policies is thus neither deductive nor analogical in form. It involves a process of weighing policies and assessing the relationship between ends and means.

The relative extent to which the lawyer relies on the rules and the policies is to some degree a matter of choice. For example, the lawyer may take the position that disputes should be settled by the application of rules and that policies will be consulted, if at all, only when there is extreme uncertainty concerning the applicability of the rule. Alternatively, the lawyer may take the position that it is always preferable to seek the result that would further the underlying policies, unless (or, perhaps, even if) that result would be contrary to very clear language of the rules.

The choice between rules and policies is profound. In applying rules, the lawyer seeks consistency with past decisions, regardless of whether the result seems desirable. In applying policies, the lawyer seeks the most desirable result. Rules refer the lawyer to prior decisions, whereas policies refer the lawyer to future consequences. Rules require the use of the logical methods of deduction and analogy, whereas policies require the use of an empirical method for assessing ends-means relationships and a normative method for weighing values.

The appropriate role of rules and policies is perhaps the most debated issue in modern American jurisprudence. The debate proceeds on both descriptive and normative grounds.

The descriptive question that lawyers debate is whether in fact courts resolve issues primarily by the application of rules or by the application of policies. The orthodox theory is that judges apply rules through the logical processes of deduction and analogy, turning to policies only in the occasional hard case. The competing theory is that judges in reality intuit the best result, that is, the result that is most satisfactory to them as a matter of policy, and only then do they turn to the rules to explain and justify the result they have reached on other grounds. In this view, the judge may even have the sensation of following the rules, but the interpretation of those rules as the judge applies them is guided by a prior intuition about the most desirable resolution. In this way, the rules can seem to produce the correct result. Indeed, some argue that the application of the rules always requires some reference to the underlying policies.¹

The normative question that lawyers debate is whether rules or policies should be given primacy in deciding disputes. For reasons described in chapter 11, the orthodox position again is that courts should give primacy to rules, referring to the underlying policies only to resolve indeterminacies or to decide whether the rules should be changed. Policy application is thus treated as a supplement to rule application.

In this chapter, the application of rules is treated as involving either deduction or analogy.² In the next two sections, these two methods are described, followed by a discussion of some of the ways in which each method is supplemented by the application of policies. The principal concern here is thus the application of rules. A fuller discussion of the application of policies is reserved for chapter 6.

Legal reasoning is indeterminate because the elements of the factual predicate are stated in such general terms that the lawyer cannot determine with certainty whether they include the facts of the client’s situation.³ Thus, it is often possible to characterize the facts of the client’s situation in more than one way and thereby to derive two or more alternative minor premises. The conclusion of the legal reasoning process depends entirely upon which minor premise the lawyer selects.

Consider the example of the client who punched his neighbor. Few would doubt that a punch in the face satisfies the elements of a battery, including the element of a touching. Thus, the only plausible minor premise would include an assertion that the client did cause a touching.

Assume now, however, that the client, rather than punching the neighbor in the face, had tugged on the neighbor’s necktie, pulled on his shirtsleeve, pushed his hat off his head, or knocked a box of pizza out of his hand. Or consider less tangible contacts, such as blowing air on the neighbor, causing sound waves to reverberate in his ear, or sending electromagnetic radiation through his body.

The problem is that the term touching is so general that lawyers may differ over which, if any, of these examples constitutes a touch. Whether the client has satisfied the elements of a battery in each case depends upon how the lawyer characterizes the client’s conduct.

The problem of indeterminacy is especially difficult if the factual predicate incorporates a legal standard, such as “reasonableness” or “good faith.” Rather than stating the facts that must be present, standards merely characterize the facts at a high level of generality. Whether the characterization applies may depend upon innumerable circumstances that cannot be, and are not, specified in the factual predicate. Thus, standards are so general that their language is particularly unlikely to decide a case with complete certainty.

For example, one tort law rule states that a person is negligent if he or she fails to exercise reasonable care. Driving a car at sixty miles per hour is unreasonable if the road passes through a school zone and children are present but may be entirely reasonable if the road is a deserted four-lane highway and the driver is rushing a coronary patient to the hospital. In short, standards do not create a sharply defined category of facts giving rise to the legal consequence, and thus they can intensify the indeterminacy of a legal rule.⁴

Because the problem of how to characterize facts arises from the generality of the language of the rule, the difficulty can sometimes be addressed by finding a more specific rule that defines the elements with greater precision or illustrates them. Recall that, in step 3, the lawyer has organized the rules into a framework in which specific rules are categorized as subrules of more general rules.

One type of subrule illustrates a general rule or an element of a general rule. Thus, for example, the lawyer may identify a case holding that knocking a box of pizza out of the victim’s hand was an offensive touching. If the current case involves a defendant who knocked a hat off the head of the plaintiff, the lawyer for the plaintiff will argue that, by analogy, knocking a hat off one’s head should be treated the same as knocking a box of pizza out of one’s hand. This use of analogy is discussed further below.

Another type of subrule defines an element of a more general rule. Thus, for example, the lawyer may identify a rule defining the element of a touching. Such a rule may state that a touching occurs when the defendant causes physical contact with the plaintiff’s person or with something closely identified with the plaintiff’s person. This rule adds a degree of specificity to the word touching.

Once a more specific rule is identified, the lawyer again performs syllogistic reasoning, now using the specific rule as the major premise. Assume that the defendant has tugged on the plaintiff’s necktie. The new major premise in this syllogism is that one who causes physical contact with the plaintiff’s person or something closely identified with the plaintiff’s person commits a touching. The minor premise might be that a necktie is closely identified with the plaintiff’s person. The conclusion then would be that pulling on the necktie constitutes a touching.

A similar syllogism might be used in an effort to decide whether knocking a box of pizza out of the plaintiff’s hand is a touching. In this case, however, even the more specific rule may be too general. It simply may not be self-evident to the lawyer whether the box is “closely identified” with the person. The lawyer may look for an even more specific rule defining the elements of the specific rule, such as the phrase closely identified, and, if there were such a rule, the lawyer would again use it syllogistically to decide about the box. If no more specific rule exists, the lawyer will be in a quandary over how to determine whether the box is closely identified with the plaintiff’s person and, thus, whether striking the box is a touching.

In short, identifying more specific rules at times may seem to solve the problem of indeterminacy. Lawyers often find, however, that even the most specific rule is sufficiently general to leave room for doubt. Something other than formal reasoning is necessary to reach a conclusion.

A second canon is that every word or phrase in a statute is to be given effect. Thus, if an interpretation can be reconciled with the language of a statute only by assuming that a certain word is redundant or without effect, the court will generally reject that interpretation.

A third canon, sometimes expressed in the Latin maxim Expressio unius est exclusio alterius (meaning “The expression of one thing is the exclusion of another”), is that the language of a rule is presumed to be exhaustive. Thus, if a statute lists certain instances to which it applies, the court will presume that the legislature meant to exclude any instance not listed.

A fourth canon, known as the doctrine of ejusdem generis (a Latin phrase meaning “of the same kind”), states that when a statute contains a list of instances to which it applies followed by general language indicating other instances, the statute will be presumed to apply to other instances only if they are of the same type as those listed. For example, a statute that authorized an investigator to inspect “books, papers, and other records” would authorize inspection only of other records that were like books and papers. A similar canon is that of noscitur a sociis (a Latin phrase meaning “let it be known by its associates”), under which a vague word is to be interpreted by reference to accompanying words.

Another canon, which is also frequently summarized by a Latin maxim—Generalia specialibus non derogant (meaning “The general does not derogate from the particular”)—provides that a specific statute prevails over a more general one. For example, if one statute provides that claims of negligence must be brought within three years and a second statute provides that malpractice claims against physicians must be brought within one year, the latter statute would probably govern a negligence suit brought by a patient against a surgeon. Because malpractice actions against physicians are a narrower category of lawsuits than negligence actions generally, the malpractice statute is likely to be considered more specific and thus the governing statute.

A final linguistic canon is that statutes in pari materia (a Latin phrase meaning “on the same matter” or “on the same subject”) shall be interpreted with reference to each other. For example, a term used in different statutes relating to taxation should be interpreted in harmony.

Other canons of statutory interpretation are categorized as substantive. For example, one such canon is that a statute will not be interpreted in a way that leads to an absurd result. Another such canon is that an ambiguous statute shall be construed so as to be consistent with the Constitution. A third such canon is that an ambiguous statute shall be construed against the government. Thus, a criminal statute will be interpreted, in doubtful cases, so as not to apply to the accused. Still another substantive canon provides that statutes in derogation of the common law shall be construed narrowly. Thus, a statute that alters common law rules will be interpreted, in doubtful cases, so as not to apply.

First, a canon of interpretation is itself a form of rule and, as such, may be too general to lead to a single conclusion. For example, in the case of the doctrine of ejusdem generis, lawyers may disagree concerning the features that the listed items have in common.⁵ Thus, in the illustration just given, one lawyer may conclude that the kinds of records that the investigator may inspect include only those that are in printed form on paper, which would exclude microfilm and computer discs, whereas another lawyer may conclude that they include any record that is in printed form whether on paper or not, which would include microfilm but not computer discs. Each of these characterizations accurately describes books and papers, but each has a somewhat different scope and, if used, would give a different meaning to the term other records. Thus, the rules of interpretation are very often victims of the same problem of generality that they are intended to address. For that reason, the rules of statutory interpretation can be indeterminate.

Second, canons of statutory interpretation may be indeterminate because they are sometimes contradictory on their face. Indeed, Professor Karl Llewellyn, who taught at Columbia Law School from 1925 to 1951 and then at the University of Chicago, wrote a well-known article in 1950 in which he attempted to show that, for each of numerous canons of statutory interpretation, one could state another canon that seemed to contradict it, at least in part.⁶ For example, Llewellyn noted that one commonly stated canon to the effect that “if language is plain and unambiguous it must be given effect” was contradicted by another canon stating that a literal interpretation will not be adopted if it would “lead to absurd or mischievous consequences or thwart manifest purpose.”

One explanation for this phenomenon is the underlying tension in statutory interpretation concerning the extent to which a lawyer should consult extrinsic sources in interpreting a statute. Put another way, conflicts between rules of statutory interpretation are rooted in a dispute about the relative primacy to be afforded the language of the rule and the underlying policies. As discussed in chapter 2, a lawyer may attempt to interpret a statute by examining its language alone or by consulting extrinsic sources. If the lawyer decides to consult extrinsic sources, there are at least two types of extrinsic sources that might be consulted. One is the legislative history, which would be researched in an effort to interpret the statute in accordance with the intent of specific legislators. Another is the subsequent understanding of what is just or wise, which could lead to an interpretation of the statute in accordance with current notions of justice or good policy.

Each of these theories of statutory interpretation is reflected in a variety of canons of statutory interpretation. Consider, for example, the two canons taken from Llewellyn’s article. The first theory—that interpretation is based on the language alone—is reflected in the canon that plain and unambiguous language is to be given effect. The competing theory—that interpretation may be based on extrinsic sources—is reflected in the second canon. Under the second canon, the literal language is not to be followed if it would “thwart manifest purpose” (i.e., run contrary to legislative intent) or lead to “absurd and mischievous consequences” (i.e., contradict current notions of justice or sound policy). Thus, the first canon tends to restrict the lawyer to the text alone, whereas the second encourages the lawyer to look outside the language of the statute at legislative history and common notions of justice and sound policy.

The rules of statutory interpretation, in other words, are based on competing theories of statutory interpretation, that is, on different conceptions of the relative importance of rules and policies. Because the theories point in opposite directions, the rules also point in opposite directions. Thus, the result reached may depend upon which rules of statutory construction are applied.

The principal nontextualist or nonformal technique for addressing the indeterminacy of rules, whether embodied in statutes or judicial decisions, is the use of policy judgments. The lawyer must decide whether the policies underlying the rule would be furthered by characterizing the facts as satisfying or as not satisfying the elements of the rule.

The first type of judgment required is a judgment about the relationship between ends and means. In the battery example given earlier, one end is to preserve the peace. The lawyer must decide whether that end would be served by requiring the client to pay compensation for, say, pushing a hat off someone’s head.

In theory, the extent to which a particular means will further a given end is an empirical question. Controlled experiments to determine the effects of a particular rule, however, are rarely performed. Thus, lawyers and courts generally use a combination of experience and intuition to make judgments about the relationship between ends and means. In making the judgments, lawyers may reach conclusions about which reasonable people can differ.

The second type of judgment required is a judgment about the relative importance of policies. This judgment is required because all rules represent a compromise among a number of opposing policies. For example, the rule defining a battery may be based on the policy of preserving the peace, but it is also based on the opposing policy of discouraging litigation over trivial offenses. The lawyer must decide which policy is to prevail in a given situation. Again, such judgments may yield conclusions about which reasonable people may differ.

The lawyer, in effect, makes two judgments: which policy to prefer and which means are effective in furthering that policy. In the next subsection, the nature of policy judgments is described, and in the subsection thereafter, the discussion centers in general terms on how these judgments are combined to decide a case. These topics are discussed in further detail in chapter 6.

Assume that the lawyer must apply the policies to the question of whether pushing a hat off someone’s head is a battery. One may be tempted to structure the reasoning process in a syllogistic form. To simplify matters, assume momentarily that there is but one policy—preserving the peace. In this proposed syllogism, the major premise is that one commits a touching by causing a physical contact with something closely identified with the person of another such that the contact is likely to cause a breach of the peace. The minor premise is that pushing a hat off someone’s head is likely to cause a breach of the peace. The conclusion would be that pushing the hat is a touching.

Using policies to decide cases, however, is not truly deductive. This is so because policies do not state a major premise from which one can reason deductively; policies are not general rules describing an entire category of cases that have a particular legal consequence. Rather, policies are the ends for which rules are the means. Thus, policies by their terms are stated as absolutes, as goals to be sought in all circumstances by any means, whereas legal rules by their terms create a specified legal consequence only in the limited circumstances set forth in the factual predicate.

In using policies to complete the legal reasoning process, the lawyer is not reasoning from a general rule to a specific case. Rather, the lawyer is reasoning from a specific case toward at least one, and probably several, policy goals. This form of reasoning requires the lawyer to make empirical judgments about the relationship between ends and means and value judgments about the relative weight of policies.

Policies, of course, can be stated at different levels of generality. A policy of protecting individual freedom, for example, can be stated more specifically as the policy of protecting free speech and more specifically still as the policy of protecting newspapers against libel actions. As policies are stated at lower levels of generality, they become limited in their scope of application and begin to take on the character of rules. Thus, as policies become more specific and rules become more general, the distinction between the two dissolves. This idea will be discussed in greater detail in chapter 6. At the moment, suffice it to say that policies can be stated at different levels of generality and that, as will be shown below, lawyers can manipulate the level of generality in order to create arguments for or against the application of a rule.

The lawyer cannot simply decide that one policy is more important than the others and then propose any result that furthers that policy. The problem with such an approach is that a particular result may not further the preferred policy enough to justify the cost to the competing policies. In predicting the result that a court is most likely to reach, the lawyer chooses the result estimated to provide the most benefit, taking into account the relative weight of the competing policies and the extent to which each result will further or impede those policies.

For example, the rule creating liability for a battery, as already noted, is based on policies that include deterring violence and avoiding wasteful litigation. Assume that the court decides that the former policy is more important than the latter. Using that assumption, the court might well be inclined to adopt a rule that all physical contacts result in liability for a battery, in order to discourage any physical contact that could even remotely lead to violence. Yet two objections to that rule exist. First, it would encourage too much litigation—the morning commute on a crowded subway could spawn dozens of lawsuits. Second, many physical contacts would never lead to violence. Thus, the benefit of such a rule in preventing violence would be small, while the cost in terms of the policy of avoiding wasteful litigation would be great. Accordingly, despite the greater importance of preventing violence, the court might decide to impose liability only for those touchings likely to provoke violence, such as touchings that a reasonable person would find offensive. If avoiding litigation were to become a relatively more important policy, then the court might require an even greater likelihood of violence before imposing liability. For example, the court might require outrageous conduct before imposing liability, thus avoiding liability in all but the most extreme cases. In combining the judgments, then, the lawyer considers both the importance of the policies and the degree to which they will be furthered or impeded by each result.

Although reasonable persons may differ about the relative weight of policies and the degree to which particular results will further or impede a given policy, policy judgments are not wholly indeterminate. As will be discussed in chapter 6, policy judgments are made in a particular context that includes the historical setting, the individuals who are making the judgments, the facts to which the judgments will be applied, and prior judicial decisions. This context often constrains, even if it does not wholly determine, judgments about the weights of policies and the relationships between ends and means.

Ideally, the situations on one side of the line should be qualitatively different from those on the other side. This qualitative difference justifies treating the situations differently. In practice, however, it is often characteristic of line drawing that, no matter where the line is drawn, there will be situations on either side that appear not to differ qualitatively, that is, the difference will seem to be one of only very slight degree. Given that the situations on either side differ only slightly, the precise location of the line will seem arbitrary.

In extreme cases, the choice of the side of the line on which to place the factual situation will seem easy. To return to the example of the landowner’s duty to warn about hazards, at one extreme it may seem obvious that a twenty-foot deep pit is so dangerous that the court should impose on the landowner the minimal burden of warning about the pit. At the other extreme, it may seem equally obvious that a one-inch deep depression in the ground poses so little danger that no discernible benefit would result from imposing a duty to warn. The twenty-foot-pit case and the one-inch-depression case seem clearly to belong on different sides of the line. The situations are qualitatively different, and this difference justifies their having different legal consequences.

As cases arise, however, involving pits of varying depths, some of the situations on opposite sides of the line may begin to resemble each other rather closely. At some point, the distinction between two cases on opposite sides of the line may seem arbitrary because there is no qualitative difference between them.

Nevertheless, the line must be drawn somewhere. The alternative would be to allow one policy to prevail all of the time. The landowner either would have no duty to warn of any hazard, no matter how dangerous, or would have a duty to warn of any condition on the land, no matter how trivial. Because both of these rules are highly undesirable, the drawing of a line that may seem arbitrary in close cases is tolerated, as long as the location of the line appears to reflect a qualitative difference between at least most cases on either side of the line and to yield a desirable result in most situations.

The metaphor of balancing seems to suggest that deciding a case requires the lawyer to make only the first type of judgment—a judgment about the relative importance of policies. And, indeed, most lawyers, to the extent that they think about it at all, probably do think of balancing tests as just that: a stark weighing of one policy against another.

In fact, however, balancing generally requires both types of judgments. The lawyer is not simply weighing, for example, free speech against the state’s desire to suppress speech. If that were so, then one policy would outweigh the other, with the result that free expression would be either always protected or never protected.