Marry Me a Little. How Much Precision is Enough in Law?

Télécharger en PDF

Sommaire de l'article

Jose J. Moreso

Dans cet article trois sous-idéaux de l'idéal de l’État de droit sont présentées : la précision – le langage du droit devrait être précis –, la justice formelle – l'obligation de traiter les cas égaux de façon égale – et la « defeasibility » – le droit devrait faire de la place pour des exceptions aux règles générales. L'article montre diverses façons selon lesquelles ces sous-idéaux sont en tension. Nous pouvons introduire des concepts classificatoires avec des frontières nettes dans le droit, mais il faut alors sacrifier la justice formelle dans une certaine mesure. Nous pouvons également introduire des concepts métriques dans le droit avec l'attribution de numéros à certaines propriétés, préservant la précision et la justice formelle, mais cette stratégie n'est pas viable à l'égard de concepts vagues de manière combinatoire. Ces concepts requièrent de faire de la place aux exceptions. Dans les dernières sections, je défends la valeur du vague dans le droit, contre la récente d'objection formulée par Asgeirsson (2015) selon laquelle, dans ces cas, la valeur ne résulte pas du vague, mais à partir de la multidimensionnalité incommensurable.

Stephen Sondheim, Marry me a little (Off-Off Broadway, 1980).


1. The ideal of precision


When I studied law in Spain, the Criminal Code (art. 10. 12) classified as an aggravating circumstance the fact that the crime was committed at night*. Obviously, this was one of the favourite examples of vagueness for our Criminal Law Professors. In Spain, at noon it is not night and at midnight it is, but it is impossible to ascertain the instant during the twilight when the night starts. “Night” expresses a vague concept, a concept which has borderline cases. By contrast, it seems that other expressions, like ‘being married’, are not vague. Every person is either married or not married, tertium non datur. There are no borderline cases of being married. However, I also remember, in the lectures of Roman law, a case – called the Spanish abandoned wife – narrated by Cicero, which may cast some doubt on the alleged immunity from vagueness of “marriage”:[1]

And what of a case that really happened, within our fathers’ recollection, of the head of a family coming from Spain to Rome, and leaving in the province his wife with child: at Rome he married another wife, without having sent notice of divorce to the first, and afterwards died intestate, when each woman had borne a son; was it but an ordinary dispute that thereupon arose, involving as it did the civil rights of two citizens, the boy born of the second consort, and his mother? She, if it were held that the first wife could be divorced only by using some specific formula, and not by marrying again, would be regarded as being in the position of a concubine.

There are concepts whose reference comes in degrees. People are, for instance, rich or adult to the extent that they have a certain amount of money or are a given number of years old. Amancio Ortega is, no doubt, rich and adult. These concepts have borderline cases and give, as it is well known, rise to sorites arguments. Sorites arguments have the following structure:

(1) Amancio Ortega is rich

(2) If s, who has n $, is rich, then s’, who has n-1 $, is also rich.

Therefore, John Doe, who has only 1 $, is rich.

The reiterated application of premise (2) for a number of times that equals Ortega’s dollars leads us, step of modus ponens by step of modus ponens, to the conclusion. The intuitive idea behind premise (2) is that small changes of this sort do not affect the truth-value of propositions which are expressed by sentences containing vague gradable adjectives such as “rich”. This phenomenon is sometimes called, in the literature on vagueness, the tolerance of this kind of predicates.[2] We can substitute in (2), “s” for “Amancio Ortega” and we obtain:

(2’) If Amancio Ortega, who has 69.2 billion dollars, is rich, then Jane Roe, who has 69.2 billion – 1 $, is rich.

Given that the antecedent of (2’) is true, the consequent is; and the consequent can be the antecedent of other conditionals like (2’) until we reach the absurd result. Nonetheless, all of us know that someone with only 1 dollar is not rich – in fact she is extremely poor.

There are also concepts that have borderline cases, but where our uncertainty on the application of the concept to certain objects is not, or not only, a consequence of this scalar or linear structure. The concept of cruelty and the concept of marriage display another kind of vagueness, different from vagueness by degrees: combinatory or multidimensional vagueness, as it has been called.[3] In those cases, we are not certain as to which properties, and within what limits, are jointly necessary and sufficient for the application of the concept. These concepts are multidimensional. Torturing babies for fun is cruel, sure. However, it is not so clear whether the death penalty is cruel or not.[4] Marriage does not seem a vague concept, but – Cicero’s case apart – in the last years we have witnessed in several jurisdictions the emergence of the question whether same-sex couples can marry or not.[5]

The law takes into account rich and adult people, cruel punishments, marriage and so on. But the law does not deal with the corresponding predicates in the same way, and the value of precision changes depending on which predicates are used. Precision in law is often thought as being part of our ideal of rule of law. For instance, John Rawls in A Theory of Justice put the question in the following terms:

Now the connection of the rule of law with liberty is clear enough […]. But if the precept of no crime without a law is violated, say by statutes, being vague and imprecise, what we are at liberty to do is likewise vague and imprecise. The boundaries of our liberty are uncertain. And to the extent that this is so, liberty is restricted by a reasonable fear of its exercise. The same sort of consequences fellow if similar cases are not treated similarly, if the judicial process lacks its essential integrity, if the law does not recognize impossibility of performance as a defense, and so on.[6]

At least three sub-ideals are included in this Rawlsian presentation of the ideal of rule of law:

(1) Precision: The language of the law should be precise.

(2) Formal justice: The law should treat like cases alike.[7]

(3) Defeasibility: The law should make room for justified exceptions to general rules.

It is a main aim of my contribution to argue that these sub-ideals are in tension and that it is not possible to achieve them without making some sacrifices. In (2) I shall present the introduction of classificatory concepts with sharp boundaries in law as a way to provide precision, and I shall argue that this requires sacrificing formal justice to some extent. In (3) I shall consider the introduction of metrical or quantitative concepts in the law assigning numbers to certain properties, in which case the law achieves precision and formal justice; however, this strategy is not viable with respect to every concept that we need in order to regulate human behaviour. In (4) I shall argue that in some cases of combinatory vagueness, mainly those where we use evaluative concepts, neither sharp boundaries nor metrical concepts are useful, and that here vagueness has value, at least if we intend to take exceptions seriously. In (5) I present an objection to the argument recently put forward by Asgeirsson[8] to the effect that in those cases the value arises not from vagueness, but from incommensurate multidimensionality. Finally, in (6) I outline the conclusions of my investigation.


2. Sharp Boundaries


The natural language predicate “adult” expresses a vague concept. A child who is two years old is not adult, a woman who is thirty years old is, and since there are borderline cases, we can construct a sorites series. For instance:

(1) Julia, who is 2 years old, is not adult

(2) If s, who is n years old is not adult, then nor is s’, who is n years +1 day old.

Nonetheless, the reiterated application of premise (2) leads to the conclusion that Lidia, who is fifty-three years old, is not adult: an obviously false conclusion.

In this kind of cases, the law tends to use vocabulary with sharp boundaries. In Spain, for instance, the legal age comes at 18 years.[9] Given that it is important to be certain about who can vote in the political elections, who can be prosecuted for a criminal offense, who can make contracts and wills, who can drive cars, who can buy alcoholic beverages or cigarettes and so on, the law constructs the concept of legal age, a concept with sharp boundaries.[10] In this sense the concept of legal age neither is tolerant nor licenses a sorites argument. Someone who is exactly 18 years old is legally adult, while someone who is 18 years minus one day old is not. Legal age does not come gradually, it comes suddenly.[11]

We can say that in these cases, even if the boundary is arbitrary, it is very important to have a boundary.[12] Legal attributes are often of this sort. In order to live in a well functioning legal order it is necessary to be sure of who is legally adult, a citizen, married, a legal heir, a judge, a property owner, guilty, and so on. Because the law ascribes to these conditions an array of legal rights and duties it is convenient to know with certainty who meets a certain legal condition. Not all of the corresponding concepts are as precise as the concept of legal age. Legal concepts often have negative conditions of application which allow one to nullify a will or a marriage and it is possible for the application of these negative conditions to have borderline cases: for instance, a marriage is voidable if either member of the couple got married under duress, and it is not difficult to imagine borderline cases of duress. However, the vocation of the concepts associated with legal conditions is to avoid borderline cases: they are in the search of sharp boundaries.[13]

Nonetheless, the aim of precision goes against formal justice. For someone whose 18th birthday is today is legally of age and another person, who is only a day younger, is not. So the law will have treated similar cases differently. In case the latter person were to commit a crime today, she could not be prosecuted in a criminal trial and would instead be subjected to the jurisdiction of juvenile justice (under Spanish law), while the former alone could legally buy a bottle of wine in a supermarket; and so on.

3. Law by numbers

Sometimes, however, precision and formal justice – treating like cases alike – can both be achieved. Tax law is the paradigmatic example.[14] In income tax law, for instance, we avoid the use of concepts with blurred boundaries, like “very rich”, “rich”, “poor”[15] (and so on), while we also eschew the use of concepts with sharp boundaries, such as “people with an annual gross income between 500.000 and 1.000.000 euros”, “people with an annual gross income between 300.000 and 500.000 euros” and so on. Given that money is measurable, here we can introduce metric concepts and correlate quantities of money which should be paid as income tax to quantities of money which are income derived of any source.

It is often held that the substitution in science of classificatory concepts by metric concepts is due to the fact that the worldly objects of our life do not have the sharp boundaries that are suggested by our classifications. But, as Hempel remembered: “this way of stating the matter is, at least, misleading. In principle every one of the distinctions just mentioned [Hempel mentioned ‘long and short, hot and cold, liquid and solid, living and dead, male and female’] can be dealt with in terms of classificatory schemata, simply by stipulating certain precise boundary lines”[16].

Some of the reasons adduced by Hempel in favour of the introduction of quantitative concepts in our scientific theories are also valid when given in favour of their use in some branches of law, like tax law: a) by means of metrical concepts it is possible to differentiate cases which are put together in our classifications, “[and] in this sense a system of quantitative terms provides a greater flexibility and subtlety”[17]; b) while in the case of classificatory concepts only a limited number of categories are available, in the case of metrical concepts we have infinite possibilities and, moreover, we do not need the introduction of new terms[18]; c) “a characterization of several items by means of a quantitative concept shows their relative position in the order represented by the concept”[19] and it allows us to compare them; d) “Greater descriptive flexibility also makes for greater flexibility in the formulation of general laws”[20] – here I benefit from the ambiguity of “general laws” because it is obvious that this sentence is applicable not only to scientific general laws, but also to normative general laws.

It is hardly surprising that measurement theory has inspired the so-called degree theories of vagueness, which assign degrees of truth to propositions[21].

However, it is possible that in tax law there remains room for inexact concepts: for instance, are the meals’ tickets in the restaurant of the company of an employee part of her gross income as payments in kind? It seems that not all the concepts that are included in tax law are susceptible of (elimination through) measurement.

Be that as it may, income tax law is an approach to the use of concepts which allows us to get both precision and formal justice. We have all the precision that is afforded by the infinite series of natural numbers and we have tolerance, given that a small increment in income corresponds to a small change in legal consequence – the quantity of money to be paid, in this case – thereby treating like cases alike and unlike cases unalike. Nonetheless, this solution is only possible if we are dealing with one-dimensional predicates (e.g. those that allow the income to be measured in quantities of euros). When we have multidimensional predicates (like ‘cruel’, ‘reasonable’ and so on) the operation is more difficult. And if these concepts display incommensurate multidimensionality it is impossible. In the next section I shall deal with multidimensional predicates.

It would not be a good idea to convert the concept of legal age into a metrical concept, even if it is obviously possible. Whereas it is very convenient to correlate the quantity of money that is obtained in one year with income tax duties, it seems implausible that we should make the right to vote depend on age, by giving different weight to each vote depending on the age the voter has, or that we should allow the quantity of bottles of wine that can be bought depend on the buyer’s age. I mean that it is not always convenient to convert a classificatory concept, with sharp boundaries, in a metrical concept. In cases where stability is important, we cannot afford the elasticity of metrical concepts.


4. Unbestimmte Rechtsbegriffe


The concept of rich is vague but not multidimensional, because its application, in normal contexts, only depends on the quantity of money which a person has. The concept of elegant is vague because it is multidimensional, for when predicated of a person it might refer to a fine and proper way of dressing, speaking, walking, eating, drinking and so on. Someone is elegant or not depending on the presence and the combination of all these features. If we intended to precisify the concept of elegant, for instance by specifying certain kinds of dresses, we would probably exclude kinds that are elegant and include ones that aren’t.[22] In these cases it is not possible to generate a metrical concept, since we have several dimensions and their combinations: the way of walking can ruin a splendid dress and the way of speaking can save a non-convenient combination of clothes. In this sense, these are cases of extravagant vagueness,[23] because two people who know the relevant facts and the language can strongly disagree about the correct application of the concept of elegant.[24] For instance, for our tastes, the various actors who interpreted the character of James Bond in the movie are elegant, but they are elegant in a very different way, the elegance of Sean Connery is very dissimilar to the elegance of Daniel Craig.

In law we have a lot of multidimensional concepts, like the concept of reasonable, neglect, excessive, cruel, proportional, due care and so on. Usually these concepts figure as negative conditions of other main concepts, as defeaters.

Paradigmatic cases are justifying and excusing conditions in criminal law and invalidating conditions in private law.[25] We can consider the regulation of freedom of speech in the European Convention of Human Rights (ECHR) enshrined by article 10:

1. Everyone has the right to freedom of expression. This right shall include freedom to hold opinions and to receive and impart information and ideas without interference by public authority and regardless of frontiers. This article shall not prevent States from requiring the licensing of broadcasting, television or cinema enterprises.

2. The exercise of these freedoms, since it carries with it duties and responsibilities, may be subject to such formalities, conditions, restrictions or penalties as are prescribed by law and are necessary in a democratic society, in the interests of national security, territorial integrity or public safety, for the prevention of disorder or crime, for the protection of health or morals, for the protection of the reputation or rights of others, for preventing the disclosure of information received in confidence, or for maintaining the authority and impartiality of the judiciary.

In the first paragraph the freedom of expression is established. In the second paragraph certain negative conditions are introduced often through open and multidimensional concepts, “necessary in a democratic society”, “in the interests of national security”, “public safety”, “protection of health or morals” and so on. The application of these concepts is controversial and a huge part of the work of the ECtHR in Strasbourg consists in ascertaining if the legislation and activity of the member States have respected the freedom of expression. Usually States argue that the second clause of the article 10 gives discretion to apply the limitations in the so-called margin of appreciation. But, is the application of these multidimensional concepts discretional?

Now I want to present the Spanish doctrine in Public Law, following the German Public Law, on the so-called indeterminate legal concepts (Unbestimmte Rechtsbegriffe). This doctrine intends to justify the judicial control over the regulations of administrative bodies and their application of concepts like reasonable, proportional, excessive and so on. The government departments and administrative agencies always argue that, to the extent that these concepts are indeterminate they defer to the administrative discretion and, therefore, their decisions are not judicially reviewable. This doctrine distinguishes within these concepts a core of certainty, where the application of the concept is clear, and a penumbra of doubt of borderline cases. This is, as it is well-known, the way Hart[26] introduced his doctrine of the open texture of law. More surprisingly, it is also the way in which Philip Heck,[27] more than one hundred years ago, introduced his distinction between Begriffskern (the core of the concept) and Begriffshorn (the halo of the concept): “A nucleus of certain meaning is surrounded by a gradually fading halo of meaning”. Hart’s conclusion in these cases of regulation by administrative bodies is as follows “In these cases it is clear that the rule-making authority must exercise a discretion, and there is no possibility of treating the question raised by the various cases as if there were one uniquely correct answer to be found, as distinct from an answer which is a reasonable compromise between many conflicting interests”.[28]

However, this is not the doctrine of the indeterminate legal concepts. For instance, in a decision of the Spanish Court of Cassation (Tribunal Supremo) about the legal validity of the imposition of a fine to a real estate company it is argued:[29]

All the activity of the Administration in the field of penalties results thus regulated even if it is dealing with the quantification of fines – it is not imaginable that two different fines can be equally fair to the same infraction –, given that even if indeterminate legal concepts should be taken into account, with the margin of appreciation called by their halo of difficulty, the application of such concepts is a bounded behaviour.

This doctrine started in German Public Law doctrine in the 50’s of the past century[30] and now is widely accepted by the majority of countries in Continental Law. In the most famous textbook (and deservedly so) on Administrative Law in Spain, García de Enterría and Fernández put it in the following terms:

This is the essence of the indeterminate legal concept: the indeterminacy of the sentence is not translated in a indeterminacy of the applications of it; these applications only allow us a “unity of right solution” in each case, grasped through an activity of cognition, thus able to be objective, and not through an activity of mere volition. [31]

Indeterminate legal concepts have incommensurate multidimensionality,[32] they display extravagant vagueness. But there are at least two accounts of how their application works:

(1) The discretion account: These concepts have a big halo of penumbra and, in this zone, their application is discretional.

(2) The one-right answer account: despite their indeterminacy, when applied to concrete cases, these concepts determine one and only one right solution.

Let us consider the Eight Amendment of the American Constitution prohibiting cruel and unusual punishments, or article 5 of the Universal Declaration of Human Right, according to which “no one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment”. Is the punishment established in the Lex Pompeia of old Roman Law, the poena cullei, a cruel, inhuman or degrading punishment? The poena cullei, the punishment for the crime of parricide, consisted in drowning the culprit in a leather sack together with a cock, a dog, a serpent, and a monkey. I guess that all of us consider the poena cullei a cruel punishment.

However, we can wonder whether another punishment, the punitive coma, is also a cruel punishment or not. The punitive coma appears first as a fiction punishment in a Philip Kerr’s crime novel (1992) and, ten years after, becomes a serious and intriguing proposal:

Let us suppose that it were possible to punish offenders by giving them an injection that would instantly induce a state of coma. Let us further suppose that it were possible, perhaps by administering additional injections on a regular schedule, to maintain them in such an unconscious state for a period of months, years, or even decades. This presents interesting possibilities. If prisoners could be placed into comas, they could be packed tightly into a very limited space, with none of the deleterious side effects that are currently associated with prison overcrowding. For where there is no consciousness, there can be neither inmate stress nor disciplinary infractions. Drugs and gangs and rape and assault, indeed, most or all of the ongoing problems that plague the staff of warehouse prisons, would immediately disappear. It might even be possible to warehouse unconscious prisoners in coma bays more cheaply than containing them in warehouse and supermax prisons. Even if the drugs used to induce and maintain the comas were not cheap, these coma-bay prisons could be operated with a reduced staff of medical and security personnel. Indeed, if such an injection were available, it would be a humane and rational solution to the nightmarish social problem of an uncontrollable prison population.[33]

This is a controversial case. Despite the argumentation of Oleson, I consider that the punitive coma is actually a cruel, inhuman and degrading treatment. Depriving human beings of communication and interaction with other human beings is, in my view, cruel. Be that as it may, with this example we can realize the difficulty of application of this type of concepts. If you adopt the discretion account, you can argue that in this case the law cannot settle this hard question. This is Hart’s account.[34] In a paper that was recovered and published recently,[35] a lost essay written by Hart for a lecture during his stay as Visiting Professor at Harvard Law School in November 1956, Hart seems to prefer an account of vagueness in law as multidimensional vagueness[36]:

This position, namely that we are able to distinguish the leading features of a clear case and then borderline cases where some but not all of the features are present, is characteristic, it seems to me, of definition in this field. I prefer this way of putting the semantic situation to just saying that we have a continuum which stretches over a wide area and that we distinguish something which fades gradually into other notions because this metaphor of a continuum does not bring out the fact that we do, as well as recognize the vagueness at the boundary of such notions as discretion, also recognize clear or simple cases, and if we could not do this we should not be able to use the term in communication with each other.

Vagueness’ accounts in the philosophy of language seem to support the discretion account. On the one hand, we have the semantic theory of vagueness which holds that vagueness is one feature of our conceptual scheme, “vague predicates are both partially defined and context sensitive”[37] and for this reason these predicates have borderline cases. In such cases speakers have discretion over whether to include the relevant object in the reference of the predicate. On the other hand, we have the epistemic theory of vagueness according to which there is always a precise point that draws the line between the right application of a vague predicate and its complement, even though we are not able, in a irremediable way, to draw that line. Our predicates are not tolerant predicates, even though we are not able to assign truth-values in the penumbra of the predicate.[38]

It seems that neither the semantic theory nor the epistemic theory are compatible with the one-right answer account. And this account is not only the account accepted by our doctrine of indeterminate legal concepts, it is also the Dworkinian account. For Dworkin, concepts such as reasonable or cruel are not vague, they are concepts which admit of different conceptions[39] because they are essentially contested[40] or interpretive concepts.[41] However, these conceptions are comparable among them and it is possible to ascertain which of them put the notion under the best light. But, can the philosophy of language provide some support for the one-right answer account? I shall intend to show how this account can be understood in line with two different approaches to vagueness: the contextual approach and the supervaluationist approach. Here, the similarities between these approaches are more relevant than the differences and, for this reason, they will be presented as a unique view.

For the contextual approach, a sentence like “Punishment P is cruel” expresses a proposition whose truth-values vary in accordance with the circumstances of evaluation. The circumstances of evaluation include a parameter that is able to point out the (admissible) sharpening of the vague predicate, a function capable to make a vague predicate precise through reasonable standards of precisification.[42] There are some doubts as to whether the contextual account collapses into supervaluationism[43]. For this approach,[44] a vague predicate fails to divide things precisely into two sets, its positive and its negative extensions. When this predicate is applied to a borderline case, we will get propositions which are neither true nor false. This gap reveals a deficiency in the meaning of a vague predicate. We can remove this deficiency and replace vagueness by precision by stipulating a certain arbitrary boundary between the positive and negative extensions, a boundary within the penumbra of the concept. Thus, we get a sharpening or completion of this predicate. However, there is not only one, but many possible sharpenings or completions.

In accordance with supervaluationism, we should take all of them into account. For supervaluationism, a proposition p -containing a vague concept – is true if and only if it is true for all its completions; it is false if and only if it is false for all its completions; otherwise it has no truth-value – it is indeterminate. A completion is a way of converting a vague concept into a precise one. So now we should distinguish two senses of “true”: “true” according to a particular completion, and ‘true’ according to all completions, or supertrue. If a number x of grains of sand is in the penumbra of the concept of a heap, then it will be true for some completions and false for others that x is a heap and, therefore, it will be neither supertrue nor superfalse.

Completions should meet some constraints. In particular, propositions that are unproblematically true (false) before completion should be true (false) after completion is performed. In this way, supervaluationism retains a great part of classical logic. Thus, for instance, all tautologies of classical logic are valid in a theory of supervaluations, “x is a heap or x is not a heap” – a token of the law of excluded middle – is valid, because it is true in all completions independently of the truth-value of its disjuncts. Another idea suggested by Fine[45] and useful here is the notion of penumbral connections. For instance, the Spanish Constitutional Court decided that the punishment in a prison of solitary confinement for 15 days is neither an inhuman nor a degrading treatment.[46] Even if someone considers that this is a borderline case of application of inhuman or degrading term, she should accept that “If 15 days of solitary confinement is not a degrading treatment, then 10 days is not either”. A sharpening establishing that 15 days of solitary confinement is not a degrading treatment, but 10 days is degrading, is an inadmissible sharpening. The last sentence expresses a penumbral connection of the concept inhuman or degrading treatment. Penumbral connections constrain our completions or sharpenings of extravagant vague concepts.

When the language of the law uses extravagant predicates that exhibit multidimensional incommensurability, not all the ways of precisifying the predicate are equally admissible. The only admissible ones are those which meet previous assignments of meaning, respect penumbral connections, fit relevant legislation and precedent, endorse canons of interpretation in force in a certain community and so on. In this sense, the range of admissibility becomes more and more reduced. In fact, a unique admissible sharpening in each context is not even necessary in order to accept that these elements are sufficient to adopt, for the judicial application of law, the one-right answer account. For instance, even though an instruction included in a transport regulation such as:[47]

(ins) Move suspicious packages away from crowds,

is very generic and has borderline cases, when placed in a concrete context it can become very clear. An abandoned backpack in a train station is suspicious and should be moved. A school bag next to the school bus is, normally, not suspicious and should not be moved.

Schiffer[48] put in the following terms the situation we find in the judicial application of the law:

When someone says something we judge not to have a determinate truth-value, it’s no big deal: we judge that what the guy said has no determinate truth-value and move on. Judicial Necessity is the fact that that is a luxury judges don’t have. More exactly, Judicial Necessity applies to federal judges who must decide cases involving the interpretation of legal texts; it’s the fact that a judge hearing such a case never has the option of not deciding the case because it’s indeterminate whether the law in question applies to it. Even if a judge knows that the relevant law has no determinate application to the case she is hearing, she must still officially “decide” either that the law does apply to the case or that it doesn’t apply to it.

In this sense the very well-known prohibition of non liquet integrates also the context where the judicial decisions are taken. This is clearly not sufficient to endorse the one-right answer account. It would be necessary to deal with intricate issues on the nature of law and the place of legal reasoning inside the practical reasoning. Here I limit myself to remembering, with Dworkin[49], that the judicial discretion account cannot be considered the winner by default:

It is a popular thesis that in very hard cases at law, when the legal profession is split about the right answer, there actually is none, because the law is indeterminate on the issue. This “no right answer” thesis cannot be true by default in law any more than in ethics or aesthetics or morals. It does not follow from the fact that no knock-down argument demonstrates that the case for the plaintiff is, all things considered, better or worse than the case for the defendant that the plaintiff’s case is not, in fact, actually better or worse. Since the no-right-answer claim about law is a legal claim-it insists that no legal reason makes the case for one side stronger than the case for the other-it must rest on some theory or conception of law.

I would like to add that the use of evaluative concepts in law that exhibits multidimensional vagueness, often as defeaters which allows us to introduce justified exceptions in the application of generic rules, honours and respects the rule of law. Here again having a bit less of precision increases our capacity to take justified decisions by treating like cases alike.


5. Instrumental Values


It is natural to consider that law should not be vague, since vagueness is seen as a defect of our language and, therefore, of our legal language too.[50] Maybe it is inevitable, but we would be better without it. This seems the intuition of Hart when remembering us that ‘we are men, not gods’:

If the world in which we live were characterized only by a finite number of features, and these together with all the modes in which they could combine were known to us, then provision could be made in advance for every possibility. We could make rules, the application of which to particular cases never called for a further choice. Everything could be known, and for everything, since it could be known, something could be done and specified in advance by rule. This would be a world fit for “mechanical” jurisprudence.

Plainly this world is not our world; human legislators can have no such knowledge of all the possible combinations of circumstances which the future may bring.[51]

It seems that Hart regarded the imagined possible world as a better world than ours. And in our world the use of vagueness has a remedial value, it allows us a more flexible application of rules. This function is fulfilled through the use of generic and abstract expressions, which are able to be applied to particular cases with flexibility and accurateness.

Remedial or not, some authors[52] have argued that vagueness is valuable in law, because sometimes it is a means to achieve some good aims. Recently Asgeirsson[53] argued that this instrumental value is actually played by a feature different from vagueness and that, even if this feature – incommensurate multidimensionality – entailed vagueness, given that not everything that is entailed by something that is instrumentally valuable has value, we cannot conclude that vagueness has value.

Sometimes it has been argued that is not vagueness which is playing this instrumental and valuable role, but it is some phenomenon close but distinct to vagueness, like generality.[54] But the more compelling argument is due to Asgeirsson. The skeleton of the argument in favour of the value of vagueness in law can be characterized as follows:

P1. Applying the law in a flexible, fine-grained and sensitive to the particularities of the case is good.

P2. Vagueness in the law is a necessary means to applying the law in a flexible, fine-grained and sensitive to the particularities of the case.

P3. Value transmits (in a pro tanto way) from ends to means.

Conclusion. Vagueness in the law is (sometimes) instrumentally good.[55]

Asgeirsson accepts premises P1 and P3, but he rejects the conclusion because he believes that there are no good reasons to endorse the premise P2. And there are no good reasons to accept P2 because P2 presupposes a very controversial necessitation principle in the logic of value: closure under necessary consequence.[56] Something like:

Value p, p ® q ½¾ Value q.

One proxy in deontic logic is responsible for the well-known deontic paradoxes (as the paradox of derived obligation and the paradox of good Samaritan).[57] One possible formulation of the paradox of derived obligation is the following: “It is forbidden to steal this bicycle” (that amounts to “It is obligatory not to steal this bicycle”) entails “if you steal this bicycle, you ought to go to the café de la Pompeu”, and one possible example of the paradox of good Samaritan: “If you ought to help this abandoned child, then it is obligatory to abandon the child”.

There are a lot of reasons for rejecting necessitation in the practical domain. However, one of the most important is that it makes it impossible to understand the remedial dimension of morality and law.[58] Remedial virtues such as courage always presuppose that the realization of a valuable action entails (because it presupposes) a previous bad situation. Courage is only triggered in circumstances of serious danger or important risk, but the value of the virtue of courage is not transmitted to the circumstance of danger, that obviously lacks value. The action of a fireperson rescuing an old lady of the flames is a valuable virtuous action, but the fact that there are flames in a building is a fact not only without positive value, but of negative value.

My suggestion is that Asgeirsson’s argument presupposes this kind of remedial scenario. Vagueness is bad. It is a necessary evil because our world is not the world that Hart imagined: a world with a finite number of properties and where we knew all the combinations between them would be a better world than our world. Now, incommensurate multidimensionality has this remedial function: in our non-ideal world it makes it possible to take practical decisions in a way that is closer to the way adjudicatory decisions are taken in the ideal world. Vagueness is entailed by incommensurate multidimensionality, but vagueness is not valuable, it is only the enabling circumstance which calls for the remedial solution.

However, it is possible to challenge this approach.[59] Waldron has argued that evaluative predicates in law (such as “reasonable” or “proper” or “appropriate”, and also “inattentive” or “aggressive”) “are action-guiding, thought guide in the first instance the element of practical deliberation involved in the exercise of agency […]”.[60] They too guide practical reasoning (and action based on that reasoning), but they provide additional structure and channeling for the practical deliberation that they elicit”. Here we have an additional argument. The use of evaluative arguments is not only good for more elastic legal adjudication, it is also good because it invokes[61] “people’s capacity for practical deliberation in a way that realizes the dignity of human agent”. This feature does not seem remedial. If practical deliberation is part of our constituency as moral agents, and practical deliberation is triggered by the use of multidimensional evaluative concepts, then the function of these concepts in law is not remedial. It is the expression of our nature as moral actors. And the Hartian legal world with a finite number of features and combinations among them is not better than our world. In fact it is worse. It is worse because in that world there is no exercise of practical deliberation, it is a world in which rules are mechanically applied.

If the flexibility and adaptability of law have remedial nature, then one world, as the Hartian imagined world, without adaptability and flexibility but with a finite number of features and combinations among them is better than our world. Nonetheless, if the flexibility and adaptability of law have no remedial nature, then the argument against the value of vagueness is more questionable, because our world with evaluative vague concepts is better than a world without them.

Perhaps there is some way to retrieve some kind of principle which is different from necessitation, and that is able to ground the value of vagueness. We need a way to order the possible worlds and a relationship of being more or equally valuable referred to worlds.[62] If the Hartian imagined world is better than our world then the value of the presence of evaluative concepts in law is only remedial and, therefore, vagueness has no value. If, on the contrary, our world is better than the Hartian imagined world, then the value of the inclusion of evaluative concepts is not remedial, it is a necessary way to trigger our deliberative capacity as rational agents, and then, maybe, the situation of deliberative capacity as rational agents can transmit value to vagueness.

My consideration cannot be seen as a way to defeat the argument of Asgeirsson.[63] After all Asgeirsson only casts doubt on the value of vagueness, entailed by a, in his view, valuable feature: incommensurate multidimensionality. However, I intend to show that behind this entailment there is a question of whether incommensurate multidimensionality has a remedial value or not. The answer of this question determines the answer on the value of vagueness.


6. Conclusion: Rule of Law in Lesbian moulding


I have argued that the ideal of rule of law contains sub-ideals in tension. Not always can we achieve precision, formal justice and defeasibility at once.

Sometimes our main objective is precision, because we (judges, administrative bodies, police and, in general, all of us) need to identify in a stable way the legal conditions of people in order to act in accordance with the law. In these cases, the law uses concepts with sharp boundaries, non-epistemically vague predicates applicable to the objects, as it were.[64] In this way precision is privileged. In other cases, a minority of cases, it is possible to introduce metrical concepts, maintaining precision and warranting formal justice.

Nonetheless, in a lot of cases, we need both a general rule and a set of defeaters, formulated in general and evaluative terms,[65] and here precision is in some sense sacrificed. In this way formal justice and defeasibility are honoured and respected. Whether this sacrifice necessarily remits to the discretion of the adjudication organs of general rules is more controversial and I presented certain arguments in order to show how taking decisions in accordance with rules including evaluative defeaters and guiding the conduct of the decision-makers are compatible.

At the end, I tried to show that behind the criticism of Asgeirsson[66] of the idea of the value of vagueness, it lies the assumption that the presence in law of evaluative defeaters has only a remedial value and, then, it is more acceptable that vagueness could have no value. However, if the value of the presence in law of evaluative defeaters is apt to trigger our deliberative capacity as moral agents, then the value of these predicates is not remedial and, maybe, the value of vagueness can be restored.

A well-known passage of Aristotle[67], see also Waldron[68], calling our attention to this text) can be interpreted both as an acknowledgment of our imperfect abilities and of the necessity of remedial flexibility and as an acceptance of the superiority of the flexibility over rigidity, like actually it is my preference:

When the law speaks universally, then, and a case arises on it which is not covered by the universal statement, then it is right, where the legislator fails us and has erred by oversimplicity, to correct the omission-to say what the legislator himself would have said had he been present, and would have put into his law if he had known. Hence the equitable is just, and better than one kind of justice-not better than absolute justice but better than the error that arises from the absoluteness of the statement. And this is the nature of the equitable, a correction of law where it is defective owing to its universality. In fact this is the reason why all things are not determined by law, that about some things it is impossible to lay down a law, so that a decree is needed. For when the thing is indefinite the rule also is indefinite, like the leaden rule used in making the Lesbian moulding; the rule adapts itself to the shape of the stone and is not rigid, and so too the decree is adapted to the facts.


J. J. Moreso

Professeur de Philosophie du Droit à l’Université Pompeu Fabra, Barcelona.




Alston William P., « Vagueness », in P. Edwards (ed.), The Encyclopaedia of Philosophy, 8, New York, Macmillan, 1967, p. 218-221.

Aristotle, The Nichomachean Ethics, D. Ross transl., L. Brown rev., Oxford: Oxford University Press, 2009.

Asgeirsson Hrafn, « On the Instrumental Value of Vagueness in the Law », Ethics, 125, 2015, p. 425-448.

Asgeirsson Hrafn, Forthcoming. « Can Legal Practice Adjudicate Between Theories of Vagueness », in G. Keil and R. Poscher (eds.), Vagueness and the Law. Philosophical and Legal Perspectives, Oxford, Oxford University Press, 2015.

Atienza Manuel, « Sobre el control de la discrecionalidad administrative. Comentarios a una polémica », Revista española de Dere ho administrative, 85, 1995, p. 5-26.

Bachof Otto, « Beurteilungsspielraum, Ermessen und unbestimmter Rechtsbegriff im Verwaltungsrecht », JuristenZeitung,1955, p. 97-102.

Burns Linda, Vagueness. An Investigation into Natural Languages and the Sorites Paradox. Dordrecht, Kluwer, 1991.

Cappelen Herman, « The Creative Interpreter: Content Relativism and Assertion », Philosophical Perspectives, 22, Philosophy of Language, 2008, p. 23-46.

Cargile James, « The Sorites Paradox », British Journal for the Philosophy of Science, 20, 1969, p. 93-202.

Cargile James, « An Investigation into Natural Languages and the Sorites Paradox », Philosophical Books, 34, 1993, p. 22-24.

Cicero, De oratore, trans. E.W. Sutton, Cambridge, The Loeb Classical Libarary, Harvard University Press, 1967.

Dummett Michael, « Wang’s Paradox », Synthese, 30, 1975, p. 301-324.

Dworkin Ronald, Taking Rights Seriously, London, Duckworth, 1977.

Dworkin Ronald, Law’s Empire, Cambridge, Harvard University Press, 1986.

Dworkin Ronald, Justice in Robes, Cambridge, Harvard University Press, 2006.

Ehmke Horst, “Ermessen” und “unbestimmte Rechtsbegriffe” im Verwaltungsrecht, Tübingen, Mohr, 1960.

Endicott Timothy, Vagueness in Law, Oxford, Oxford University Press, 2000.

Endicott Timothy, « The Value of Vagueness », in A. Marmor and S. Soames (eds.), Language and Law, Oxford, Oxford University Press, 2011, p. 14-30.

Endicott Timothy, « Vagueness and Law », in G. Ronzitti (ed.), Vagueness: A Reader, Dordrecht, Springer, 2011, ch. 7.

Endicott Timothy, « Law and Language », in E.NZalta (ed.), The Stanford Encyclopedia of Philosophy, spring 2014 édition, URL:

Enoch David, « Epistemicism and Nihilism about Vagueness: What’s the Difference? », Philosophical Studies, 133, 2007, p. 285-311.

Fernández Tomás-Ramón, Arbitrariedad y discrecionalidad, Madrid, Civitas, 1991.

Fine Kit, « Vagueness, Truth and Logic », Synthese, 30, 1975, p. 265-300.

Foot Philippa, « Virtues and Vices », in Virtues and Vices and Other Essays in Moral Philosophy, Berkeley, University of California Press, 1978, p. 1-18.

Gallie William B., « Essentially Contested Concepts », Proceedings of Aristotelian Society 56, 1955-6, p. 167-198.

García de Enterría Eduardo & Fernández Tomás-Ramón, Curso de Derecho Administrativo, I, 13th ed., Madrid, Civitas-Thomson Reuters, 2013.

Goguen Joseph A., « The Logic of Inexact Concepts », Synthese, 19, 1969, p. 325-373.

Gottlieb Paula, « Are the Virtues Remedial », The Journal of Value Inquiry, 35, 2001, p. 343-354.

Heck, Philip, Gesetzesauslegung und Interessenjurisprudenz, Tübingen, J.C.B. Mohr, 1914.

Hempel, Carl Gustav, Fundamentals of Concept Formation in Empirical Science, Chicago, Chicago University Press.

Kamp Hans, « Two Theories about Adjectives », in E. Keenan (ed.), Formal Semantics and Natural Language, Cambridge, Cambridge University Press, 1975 p. 123-155.

Kaplan, David, « How to Russell a Frege-Church », The Jorunal of Philosophy, 72, 1975, p. 716-729.

Keefe Rosanna, Theories of Vagueness, Cambridge, Cambridge University Press, 2000.

Kerr Philip, A Philosophical Investigation. London, Chatto & Windus, 1992.

Kölbel Max, « Vagueness as Semantic », in R. Dietz and S. Moruzzi (eds.), Cuts and Clouds: Vagueness, its Nature and its Logic, Oxford, Oxford University Press, 2010, p. 304-325.

Korsgaard Christine, « Aristotle and Kant on the Source of Value », Ethics, 96, 1986, p. 486-505.

Hansson Sven Ove, The Structure of Values and Norms, Cambridge, Cambridge University Press, 2000.

Hansson Sven Ove, « Ideal Worlds –Wishful Thinking in Deontic Logic », Studia Logica, 82, 2006, p. 329-336.

Hart Herbert L.A., The Concept of Law, Oxford, Oxford University Press, 1961.

Hart Herbert L.A., « Legal Responsibility and Excuses », in H.L.A. Hart, Punishment and Responsibility, Oxford, Oxford University Press, 1968.

Hart Herbert L.A., « Discretion », Harvard Law Review, 127, 2013, p. 652-665.

Hyde Dominic, Vagueness, Logic, and Ontology, Aldershot, Ashgate, 2008.

D. Lewis, « General Semantics », in Philosophical Papers, vol. III, Oxford, Oxford University Press, 1983, p. 189-232.

Machina Kenton F., « Truth, Belief and Vagueness », Journal of Philosophical Logic, 5, 1976, p. 47-78.

Marmor Andrei, The Language of Law. Oxford, Oxford University Press, 2014.

Moreso Jose J., « Legal Defeasibility and the Connection between Law and Morality », in J. Ferrer Beltran and G.BRatti (eds.), The Logic of Legal Requirements. Essays on Defeasibility, Oxford, Oxford University Press, 2012, p. 225-237.

Moreso Jose J., « The Uses of Slippery Slope Argument », in T. Bustamante and C. Dahlman (eds.), Argument Types and Fallacies in Legal Argumentation, Dordrecht, Springer, 2015, ch. 4

Oleson James C., « The Punitive Coma », California Law Review, 90, 2002, 829-901.

Parejo Alfonso Luciano, Administrar y juzgar: dos funciones constitucionales distintas y complementarias, Madrid, Tecnos, 1993.

Poscher Rolf, « Ambiguity and Vagueness in Legal Interpretation », in L.MSolan and P.M. Tiersma (eds.), The Oxford Handbook of Language and Law, Oxford, Oxford University Press, 2012, ch. 9

Prior Arthur N., « The Paradox of Derived Obligation », Mind, 63, 1954, p. 64-65.

Prior Arthur N., « Escapism. The Logical Basis of Ethics », in A.IMelden, Essays in Moral Philosophy, Seattle, University of Washington Press, 1958, p. 135-146.

Rawls John, A Theory of Justice, Cambridge, Harvard University Press, 1971.

Rawls John, A Theory of Justice, 2nd edition, Cambridge, Harvard University Press, 1999.

Raz Joseph, The Concept of a Legal System, Oxford, Oxford University Press, 1970.

Ross Alf, « Tû-Tû », Harvard Law Review, 70, 1956-1957, p. 812-825.

Sáinz Moreno Fernando, Conceptos jurídicos indeterminados, interpretación y discreción judicial, Civitas, Madrid, 1976.

Sánchez Morón Miguel, Discrecionalidad administrativa y control judicial, Tecnos, Madrid, 1994.

Schauer Frederick, Playing by the Rules. Oxford, Oxford University Press, 1991.

Schiffer Stephen, « A Little Help from Your Friends », Legal Theory. 7, 2001, p. 42-432.

Schiffer Stephen, « Philosophical and Jurisprudential Issues on Vagueness », in G. Keil and R. Poscher, Vagueness and the Law. Philosophical and Legal Perspectives, Oxford, Oxford University Press, 2015.

Sorensen Roy A., Blindspots, Oxford, Oxford University Press, 1991.

Sorensen Roy A., Vagueness and Contradiction, Oxford, Oxford University Press, 2001.

Sorensen Roy A., « Vagueness Has no Function in Law », Legal Theory, 7, 2001, p. 387-417.

Sorensen Roy A., « Vagueness », in E.NZalta (ed.), The Stanford Encyclopedia of Philosophy, 2012, (Spring 2016 Edition, forthcoming, URL:

Soames Scott, « What Vagueness and Inconsistency Tell Us about Interpretation », in A. Marmor and S. Soames (eds.), Language and Law, Oxford, Oxford University Press, 2011, p. 31-57.

Soames Scott, « Vagueness and the Law », in A. Marmor, The Routledge Companion to Philosophy of Law, London, Routledge, 2012, p. 95-108.

Waldron Jeremy, « Vagueness and the Guidance of Action », in A. Marmor and S. Soames (eds.), Language and Law, Oxford, Oxford University Press, 2011, p. 58-82.

Williamson Timothy, Vagueness, Routledge, London, 1994.

Williams Bernard, « Which Slopes are Slippery? », in Making Sense of Humanity and Other Philosophical papers, Cambridge, Cambridge University Press, 1995, ch. 18.

Wright Crispin, « On the Coherence of Vague Predicates », Synthese, 30, 1975, p. 325-365.

Wright Crispin, « Language-Mastery and the Sorites Paradox », in G. Evans and J. McDowell (eds.), Truth and Meaning, Oxford, Oxford University Press, 1976, p. 223-247.


Remonter en haut de page