Vous êtes sur la page 1sur 16

Mikael M.

Karlsson
Roots of Legal Normativity

1. The question to be discussed here under the title “Roots of Legal Nor-
mativity” is the question how the passing of a law can create a binding norm, or
in more traditional language, how laws acquire their so-called “binding force”.
Right away, this implies a certain limitation of domain; for not all norms, even
legal norms, have binding force, although it may be maintained that a certain
“normative pressure” attaches to norms of all sorts. Some norms amount to re-
commendations which are not binding upon their addressees. I will be interested
here in the sorts of norms which are binding – or are said to be binding – upon
their addressees, whether or not these happen to be legal norms. But I will be
specifically interested in legal norms which are of that kind, and in the basis for
their being of that kind.
My procedure in this paper will be to consider, and to propose answers to, a
certain series of questions: First, “What is it to be bound or obligated to act or
refrain from acting in certain ways?” Second, “What is it to be ‘bound by a
norm’?” Third, “What, if anything, is distinctive about legal norms in this
connection?”

2. It may be helpful to recall immediately the distinction emphasized by H. L.


A. Hart between being obliged and being obligated 1. The latter, but not the
former, may be equated with having an obligation or being bound. Although
Hart’s point turns in part around a trick of English, the distinction corresponds to
– and is arguably the same as – the Kantian distinction between hypothetical and
categorical imperatives 2. Kant is, we should notice, willing to speak of
imperatives in both cases; and an imperative is the presentation of an action (or
the willing of an action) as necessary, in one way or another. The aspect of
1 H. L. A. Hart, Concept of Law (Oxford: Oxford University Press, 1961), pp. 79-88
and passim; and cf. Hart’s earlier article “Legal and Moral Obligation” in Essays in Moral
Philosophy, A. I. Melden ed. (Seattle & London: University of Washington Press, 1958),
pp. 82-107, esp. pp. 95-99.
2 For the Kantian views described here and in the following paragraphs, see
Immanuel Kant, Grundlegung zur Metaphysik der Sitten, Akademie-Ausgabe, vol. IV
(Riga, 1785), esp. pp. 413-440.

Analisi e diritto 2000, a cura di P. Comanducci e R.


Guastini
2

necessity in whatever may present itself as necessary – whether a conclusion,


say, or a belief or an action – is a creature of reason, on Kant’s account.
In the case of a hypothetical imperative, an action is presented as necessary
only hypothetically: it is necessary just in case we happen to want something
which can be obtained (or which we think can be obtained) only by acting in a
given way; it is, in other words, a necessary condition (or perceived as such) for
reaching a certain goal 3. Reason then tells us that given that we want such-and-
such, we must do so-and-so; and here we could as well say, “given that we want
such-and-such, we are obliged to do so-and-so.” The necessity here is the child
of reason, but only given a certain inclinational hypothesis. Notice, though, that
while we may be obliged to do so-and-so given that we want such-and-such, we
are not necessarily (and not usually) obligated to do so-and-so; because, for one
thing, we need not want such-and-such, and perhaps in many cases we shouldn’t
want it 4.
In the case of a categorical imperative, an action is presented as necessary
non-hypothetically, or unconditionally. That is, it is necessary independently of
whatever it is that we may happen to want. The necessity in such a case is the
child of reason alone, or so Kant thinks 5. In such a case, it is appropriate to say
that we are bound or obligated to cleave to a certain course of action.
But it is difficult, Kant admits, to understand this categorical kind of practical
necessitation. He suggests several approaches – basically two – that might be
taken, and these are meant to be complementary (or even, at bottom, identical).
On the formal approach, volition or action contrary to a categorical imperative
involves a kind of contradiction. Admittedly this is not a straightforward logical
contradiction – not a contradiction of theoretical reason – but a practical
contradiction in which the will is brought into conflict with itself. As a matter of
common and familiar experience, our reason pressures us to avoid
contradictions; and this pressure seems unrelated to wants or desires in any
straightforward sense, so the formal approach helps to make the idea of
categorical necessitation appear plausible. Unfortunately, however, few have
been able to grasp the relevant sense of contradiction. John Stuart Mill, for
example, remarks in one place that Kant:
… [lays] down a universal first principle as the origin and ground of moral
obligation; it is this: –“So act, that the rule on which thou actest would admit of

3 Cf. Aristotle on hypothetical necessity in Physics II,9 and Parts of Animals I,1:
639b21-640a10, 642a1-13, 642a31-642b4. For an excellent discussion see John M.
Cooper, “Hypothetical Necessity and Natural Teleology” in Philosophical Issues in
Aristotle’s Biology, Allan Gotthelf & James G. Lennox, eds. (Cambridge: Cambridge
University Press, 1987), pp. 243-274.
4 I do not imply here that if we had to want such-and-such we would necessarily be
obligated to do so-and-so.
5 What else than reason, Kant seems to have thought, given that inclination is
irrelevant?
3

being adopted as a law by all rational beings”. But when he begins to deduce from
this precept any of the actual duties of morality, he fails, almost grotesquely, to
show that there would be any contradiction … in the adoption by all rational
beings of the most outrageously immoral rules of conduct. All he shows is that the
consequences of their universal adoption would be such as no one would choose
to incur 6.
Kant also approaches categorical necessitation from the material point of
view. Here he says that as rational beings we necessarily value rationality and
thus necessarily recognize our own value or worth – which he calls dignity
(Würde) – and that of all other rational beings (since they are all exactly like
ourselves in the relevant respect). This kind of valuation is not the child of (or
any form of) inclination – that is, of desire or fear – but of reason alone 7. From
the point of view of a rational being, the worth of a rational being is truly its own
worth, and not the worth that is imparted to it by someone’s desiring or fearing it.
If we want to label the attitude which we take towards things valued in this way,
we can use the term respect (Achtung) 8. Now, the fact that certain things have
worth of this kind, Kant wants to say, entails that we must act towards them in
certain ways, and must not act towards them in certain other ways. So, whatever
else one might want to say about it, this approach represents obligation, or
categorical necessity, as resting upon value, but upon value that is in no way
dependent on anything that we may happen to want; it rests upon no inclinational
hypothesis. Sometimes Kant seems to say that it rests upon no hypothesis at all,
but this is evidently not the case. It rests upon a hypothesis that we must accept
as rational beings, which is, perhaps, no mere hypothesis, but a hypothesis
nonetheless 9.

3. Now those who have written about norms and values have generally fallen
into one of two camps: those who think that norms rest upon values, and those
who think that values rest upon norms 10. I have made Kant out as someone who
belongs to the former camp, at least as regards practical necessitation (“im-
peratives”) of both the hypothetical and categorical kinds, i.e. being obliged to

6 John Stuart Mill, Utilitarianism in Collected Works, X: Essays on Ethics, Religion


and Society (Toronto: University of Toronto Press, 1969), p. 207.
7 Cf. Kant, Grundlegung, p. 401n.
8 Thus, for example, we respect consistency; we don’t typically desire it. Nor do we
typically fear inconsistency.
9 It is a hypothesis in the sense of being a premise needed in order to draw a
conclusion of practical necessity. Kant does seem to think that there are some inclinational
hypotheses which must also necessarily be accepted by rational beings: they must want
happiness, for example. This complication will not be further considered here.
10 Neither of these “positions” comes to anything without a great deal of further
specification. This formulation provides only indicative slogans.
4

act or to refrain from acting in certain ways, and being obligated or bound to
follow or to avoid certain courses of action. Roughly speaking, I follow Kant in
the matters so far discussed. I will now make a stab at presenting this kind of
position more formally, without worrying further about the extent to which I am
representing Kant. Representing Kant is not my objective; talking about norms
of action and obligation is, with special attention (after a while) to legal norms.
So first let me offer an account of being obliged to act:
s is obliged to a (in circumstances c) =def.: From the point of view of some
system V of values or goods, a and only a is logically open to s as a course of
action (in circumstances c) 11;
and correlatively,
s is obliged not to a (in circumstances c) =def.: From the point of view of some
system V of values or goods, a is not logically open to s as a course of action (in
circumstances c).
There are a number of things that must be said about these proposed
definitions.
First, “s’s being obliged” refers to an objective condition of s and not to a
subjective condition. That is, s may be obliged to a without thinking or
recognizing that he is. He may, for example, subscribe to a system V, but not
realize that from the point of view of V, a and only a is open to him as a course
of action. He may not realize, for instance, that the airline is on strike, and thus
that if he values getting to Nice today, he is obliged to take the train. Again, s
may not accept V. It may be that if he values winning the chess game, he is
obliged (in the sense described) to castle now; but he may not care about
winning the chess game and so not be obliged to castle. Conversely, s may feel
or suppose that he is obliged to a when in fact he is not. He may, for instance,
accept V and think falsely that from the point of view of V, a and only a is
logically open to him as a course of action.
Second, these definitions make being obliged V-relative. What s is obliged to
do from the point of view of V, he may not be obliged to do from the point of
view of some other system of values. This relativity is independent of any system
of values that he happens to hold. Thus, in our previous example, we may say
that the chess player is V-obliged to castle, where V is a system within which
winning the chess game is a good thing. But there will be other value-systems
(e.g. the one to which the chess player subscribes, in the example given) with
respect to which he is not so obliged.
Third, in continuation of the last point, it is important in each case to refer to
a system of values or goods, within which, presumably, the goods or values are

11 In an unpublished paper, “Law and Morality: On the Ties That Bind” (first
presented by invitation at a Ratio Juris seminar, held in Bologna on 15 March 1996, under
the title “My View on the Relationship Between Law and Morality”), I distinguish
between value-obligation, which is what is spoken of here, and rule-obligation. In the
present paper, rule-obligation is not further discussed.
5

prioritized or weighted. For despite the V-relativity of being obliged, we do want


to be able to say what s is actually obliged to do (in a given set of circumstances)
from a certain personal perspective (whether or not subscribed to by s); and
unless such a perspective considers goods and values systematically, such
questions will go unanswered. We would get instead only an endless run of
hypotheticals 12: Is s obliged to take the train? Well, if getting to Nice were the
wanted thing, then, yes, he’d be obliged to take the train; but if getting to La
Bouexiere were the wanted thing, then, no, for no train runs by there and he’ll
have to drive; however, if he isn’t concerned with traveling at all, but is looking
for a good meal, then he has only to walk out the door and turn left, no train
again; but if the wanted thing is running into Julie, he’ll be obliged to take the
train to Paris, for she’ll be on that train later on; and so on.
Fourth, there are many situations in which there is no a which is the only
action logically open to an agent from the point of view of some system of
values V. Normally, agents have plenty of options. If the airline is not on strike, s
may perhaps get to Nice either by flying or by taking the train. So even if he
values getting to Nice today, he is neither obliged to take the plane nor obliged to
take the train. He may, however, be obliged to do one or the other. We may say,
then, that agents are normally obliged only to choose some course of action from
among a certain set of alternatives.
I do not take any of these things to be disadvantages of the proposed
definitions; rather the reverse. I think that we need to understand what it is to feel
or think oneself to be obliged (subjective obligation) by reference to what it is (in
the objective sense) to be obliged. The V-relativity of being obliged simply
stresses what Kant called its hypothetical character. It is an advantage of these
definitions, in my view, that they bring out this fact (as I consider it) of relativity.
It is again an advantage, I think, to see that a course of action can be seen as
hypothetically necessary (actually, and not only prima facie) only from a
systematic value perspective. And lastly, it is valuable to realize (as Kant’s own
discussion hinders us from doing) that one is rarely obliged to follow any single
course of action.
Since the proposed definitions describe being obliged as an objective
condition, it remains to connect up s’s being obliged to a with s’s having a
subjective motive to a. This seems to me not too difficult to do. A person who is
V-obliged to a, in the sense defined, is not necessarily motivated to a. And he
may likewise be motivated to a, while being obliged not to a. On the other hand,
if he consciously subscribes to a certain V, and believes that from the point of
view of V, a and only a is logically open to him as a course of action in the
circumstances in which he finds himself, then this state of mind constitutes a
motive for him to a. It would not necessarily follow, of course, that he a’s, for
any number of reasons; but probably we can say that he will a, other things being

12 Perhaps these are prima facie obligations; but very weak and uninteresting ones.
6

equal. The state of mind in question is here understood as s’s supposing himself
to be obliged to a in the circumstances in which he believes himself to be.
Normally, let us notice, when we say that s was obliged to a, it is assumed
that s in fact a’d 13, that he did so for the reason that he subscribed to some V and
supposed that from the point of view of V that a and only a was logically open to
him as a course of action 14. So, “s was obliged to a” carries with it a number of
important “conversational implicatures” concerning s’s state of mind. It was
these implicatures which led Hart to conclude that when we say that someone
was obliged to do something, this is
… often a statement about the beliefs and motives with which an action is done:
B was obliged to hand over his money may simply mean, as it does in the …
[case of his being confronted by a gunman], that he believed that some harm or
unpleasant consequences would befall him if he did not hand it over and he
handed it over to avoid those consequences 15.
This is fair enough; in essence, I am agreeing with it here. But Hart has no
other, or further, account of being obliged. We are thus left with his idea that “…
the statement that a person was obliged to obey someone is … a psychological one
referring to the beliefs and motives with which an action was done” 16. By contrast,
my account of being obliged does not refer to beliefs and motives, but to values
and their implications in a given set of circumstances; it is not, on my account, a
“psychological” concept at all. In my opinion, Hart took the normal assumptions
attaching to a concept for parameters of the concept itself. But the normal
implicatures are defeasible in context and can thus be seen for what they are.

4. So much, for the moment, for being obliged. Let us now turn to an account
of having an obligation to act: If we follow up on the Kantian suggestions made
earlier, and if we hope to provide an account which parallels the account of being
obliged that we have just considered, it seems that we will have to make room
for values or goods of a special kind: those whose status as goods is not
dependent upon their being wanted but is in some sense intrinsic to them. These
are things that we ought to value, whether or not we do; and things towards
which it is appropriate for us to have an attitude of respect, as opposed to desire
or fear. Let us call these essential goods or values. On the roughly Kantian
position developed here, unless it is possible to provide a philosophical place for
goods of this kind, it will not be possible to provide an account of having an
obligation or being bound, as opposed to being obliged. And so, if we are

13 Cf. Hart, Concept, p. 81.


14 Or, in the case of a present situation: normally assumed that s will in fact a and
will do so for the reason that he subscribes to some V and supposes that, from the point of
view of V, a and only a is logically open to him as a course of action.
15 Concept, p. 80.
16 Concept, p. 81.
7

interested in binding norms – imagining that there are such and that at least some
legal norms are of this kind –, our account of such norms will be rooted in
whatever account we can give of essential goods or values.
The general formula proposed is this:
s has an obligation to a (in circumstances c) = def.: From the point of view of a
system Ve of essential values or goods, a and only a is logically open to s as a
course of action (in circumstances c).
Correlatively,
s has an obligation not to a (in circumstances c) = def.: From the point of view of
a system Ve of essential values or goods, a is not logically open to s as a course of
action (in circumstances c).
Now most of the observations we made earlier about being obliged apply
here as well: Having an obligation is here described as an objective, non-psycho-
logical condition of s. Even with regard to some particular obligation, we are
invited to consider essential values from a systematic perspective, and not just
one by one. Typically, more than one specific course of action will be logically
open to s from the point of view of Ve, so an agent will be bound only to choose
one way of acting from a set of alternatives. And so on. But what of Ve-
relativity? We said earlier that the fact that being obliged was always relative to a
given system V of values or goods was a reflection of the hypothetical character
of this form of practical necessitation. But in the present case we were hoping to
capture the non-hypothetical character of being bound or having an obligation;
and that would seem to imply – would it not? – that being bound should not be
Ve-relative in the way that being obliged was V-relative.
Well, the definitions just proposed do make having an obligation Ve-relative
in the sense that being bound is said to rest upon a system Ve of essential values
or goods. But in the case of the V-relativity of being obliged, there is the
omnipresent possibility of many different V-systems: What s is obliged to do
from the point of view of a certain V, we said, he may not be obliged to do from
the point of view of some other system of values. So we may ask in the present
case whether there is likewise more than one system of essential values or goods.
And the answer to this question must evidently be no, if we hope to use our
definitions to capture the non-hypothetical character of being obliged. Perhaps
alternative systems of essential values could be described, but that would not
show that such alternative systems actually existed. If there is only one actual
system of essential values and goods, then being bound, although Ve-relative in a
certain way, would not be Ve-relative in the way that being obliged was V-
relative. For if, in a certain set of circumstances c, a and only a is logically open
to s as a course of action from the point of view of a system Ve of essential
values or goods, then there will be no other evaluative point of view – with
regard to essential values, that is – from which some alternative course of action
would be logically open to s. Either s will have an obligation to a in
circumstances c, or he will not. Thus, we may not know quite how to
characterize essential values or goods; and we may not know how to justify the
8

proposition that there are such goods and values; but we can say that if there is to
be any point to recognizing them, there must evidently be only one system of
such values.
But here it might be objected that surely there must be alternative systems of
essential goods and values, if there are any such values at all. For let us imagine
that human life is an essential value: it is surely an obvious candidate. Now, it
might be argued, Jones’ life is an essential good for Jones (I mean here not that
Jones thinks of his own life as being essentially valuable to him, for he may not;
I mean that it really is essentially valuable for him, whatever he may think). But,
it might be added, Jones’life may not be an essential good for Smith. Or at least
it may be less valuable for Smith than for Jones, and this would already yield
different systems of essential goods or values.
If there is an answer to this, it has got to be, I think, that Jones’ life is not, as
such, an essential good. Essential goods and values will all be generic:
independent, as goods and values of this kind, from any individual instantiation 17.
So, human life may be an essential good, but Jones’ life will not be. Knowledge
may be an essential good, but my knowledge about the composition of the moon
will not be. Happiness may be an essential good, but my happiness will not be. We
should note that individuals may nevertheless be the beneficiaries of the
obligations which rest upon essential goods; for if human life is an essential good
then presumably my life, along with Jones’and Smith’s, is not to be trifled with, for
it is an instantiation or manifestation of this good. Indeed, in this sense we may say
that an individual’s life is essentially good or valuable, and there will be an
important point to saying this; but we are nevertheless not to say that what is thus
essentially good is an essential good. A system Ve of essential values or goods will
thus, on this account, be an ordering of generic goods; and the objection just
considered does not give us any reason to insist that there must be more than one
such ordering 18.
In order for our account to appear plausible as an account of being bound, we
will evidently have to make the additional claim that essential goods and values
are in an important sense incommensurable with ordinary, non-essential goods
and values. Otherwise, the fact that from the point of view of a system Ve of
essential values or goods, a and only a is logically open to s as a course of action
(in circumstances c) would not close the question as to what is practically
necessary for s. The point of view of Ve might always be compared on an equal
footing with the point of view of some system V of non-essential goods or
values. What is necessary from the former point of view would then be seen as

17 This point is related to Alan Gewirth’s idea about generic consistency, cf. his
Reason and Morality (Chicago: University of Chicago Press, 1977). But I believe that,
with respect to Gewirth, I reverse the order of explanation.
18 Of course, nothing said here yet excludes by argument the possibility of alternative
systems of essential values or goods; I have only sought to make plausible, in the face of a
fairly obvious objection, that such a system might well be unique.
9

not necessary from the latter, and what is concluded from the point of view of V
could conceivably overweigh what is concluded from the point of view of Ve. It
is true that an agent may always consider what he is to do in a given situation
from these different perspectives. But in order to capture the notion of being
bound, it will have to be maintained, I think, that the two perspectives do not
yield conclusions which can be weighed against one another on a common scale.
Essential goods may perhaps be weighed against one another in this way, and
non-essential goods weighed against other non-essential goods. But essential and
non-essential goods will be incommensurable. I will not pursue this important
issue further here 19.
It may seem that, despite all we have said, our account of having an
obligation is open to a criticism famously directed by H. L. A. Hart at John
Austin’s theory of obligation, which Hart labels the “theory of coercive orders”:
A orders B to hand over his money and threatens to shoot him if he does not comply
[Hart says]. According to the theory of coercive orders this situation illustrates the
notion of obligation or duty in general. … The plausibility of the claim that the
gunman situation displays the meaning of obligation lies in the fact that it is certainly
one in which we would say that B, if he obeyed, was ‘obliged’ to hand over his
money. It is, however, equally certain that we should misdescribe the situation if we
said, on these facts, that B ‘had an obligation’ or a ‘duty’ to hand over the money 20.
Now the gunman in Hart’s example threatens B’s life, which even if not an
essential value is presumably a manifestation of such; and for this reason, not
handing his money over to the gunman is evidently an option closed to B from
the point of view of essential values or goods. So it looks as if, on the theory
proposed above, the gunman will cause B to have an obligation or a duty to hand
over his money. But this, Hart claims, is wrong; B is, to be sure, obliged to hand
over his money but it is not his duty – that is, he has no obligation deriving from
the threat – to do so. Must not our account then founder on the same point as
Austin’s?
Well, we should notice that Hart here overlooks the fact that a particular
obligation is typically an obligation to someone, the recipient 21 r of the
obligation 22. The recipient may be oneself, another, oneself among others,
19 I have treated this matter in more detail in my “Norms and Values”, a paper
originally presented at an international “Journée d’étude sur Normes et Valeurs” held by
the Centre for Research on Logic and its History, University of Rennes I, 1 April 1995.
This paper is presently unpublished.
20 Hart, Concept, p. 80.
21 The recipient of an obligation is not the same as what is often called the
“addressee”. The addressee is the party who is obligated (or said to be such); the recipient
is the party to whom the addressee is obligated (or said to be obligated). As discussed
later, a third party (neither the addressee nor the recipient) may be the beneficiary of an
obligation.
22 This involves a deliberate simplification. I have elsewhere maintained that, while
many of our obligations or duties are obligations to someone, we can also have
10

everyone, etc. The reason why Hart’s counterexample to Austin seems so


convincing is that B would surely be misdescribed as having an obligation or a
duty to the gunman; and this is in fact what Austin’s theory requires, so Hart’s
point is indeed telling against that theory. However, on the account proposed
here, the obligation or obligations created by the gunman’s threat would hardly
be to the gunman; B would arguably have an obligation or duty to himself to
preserve his own life, and obligations to his dependents, loved ones and friends
as well 23. Once we specify the recipients correctly, we see (or at least make
plausible) that the gunman can, through his threat, create duties, though not to
himself; so Hart’s example does not tell against the present theory.
This does bring out, however, an important disanalogy between being obliged
and having an obligation: the former does not have a recipient, the latter
typically does 24. Therefore, having an obligation, although importantly related to
being obliged, is not well understood as a variety of being obliged. The recipient
or recipients of any given obligation will be anyone who has a privileged interest
or a stake in the essential values in question in any given case 25. Insofar as a
given obligation is an obligation to someone, a recipient r, we might bring this
into our formulation, thus:
s has an obligation (to a recipient r) to a (in circumstances c) = def.: From the
point of view of a system Ve of essential values, a and only a is logically open to
s as a course of action (in circumstances c, which include r’s having a privileged
interest in the relevant essential values).

5. Given what we have thus far learned – or may claim to have learned – about
being bound or obligated, we may now go on to ask what it is to be “bound by a
norm”. What we have discussed thus far are particular obligations: those we may
have in (and which in important part derive from) particular sets of circumstances
to perform individual actions. But a norm – or at least a general norm, the sort of
norm upon which I wish to focus in the sequel – has the character of a rule.
A general rule of this kind, even a binding norm, need not be exceptionless. I
have argued elsewhere that general norms typically have a certain scope – a
domain outside of which they do not apply – and a certain power, or strength

obligations or duties with respect to something (e.g. works of art or natural objects) which
are not duties to anyone. Austin’s theory requires the former; and this is also the sort of
obligation which Hart is attempting to capture. So I here ignore the latter sort of duty, for
the purpose of simplification; but see my “Náttúran sem skepna” in Náttúrusyn, Róbert
Haraldsson and Thorvardur Árnason, editors (Reykjavík: Sidfrædistofnun, 1994).
23 He might even have an obligation with respect to human life which was not an
obligation to anyone (see previous note).
24 Although in some cases, the recipient may be indefinite.
25 Correctly specifying the needed notion of interest would be a matter of some
subtlety, and is not attempted here.
11

with respect to other norms; and I have argued in the same place that it is not
always possible to build a specification of scope or strength into the rule itself,
which, as stated, may therefore exhibit a certain defeasibility 26. I will assume
here, without argument, that a binding general norm may be defeasible in this
way.
We have thus far said that an agent is bound, or has an obligation, to act in a
certain way in a given set of circumstances when that way of acting, and only
that way of acting, is logically open to him in those circumstances from the point
of view of a system of essential goods or values. The obligation derives from the
system of values, or rather the way in which that system applies to particular
circumstances.
Now if we give such an account, the binding aspect of a general norm would
seem also to have to derive from a system of essential values, given which only
certain modes of action will be logically open to an agent. A general norm will
merely sum up, as it were, the forms of action which, from the point of view of a
system of essential values, are exclusively open to us – or, in the case of binding
general prohibitions, not open to us – in a manner which is relatively
independent of particular sets of circumstances. In essence, it is not the general
norm which binds us, but rather the system of essential values; the norm merely
reflects the manner in which we are generally bound by those values.
Thus, the Mosaic commandment, “Thou shalt not kill” amounts, on the
account given here, to a claim that killing is incompatible with, and thus
excluded by, essential values, in a manner which is relatively independent from
particular circumstances. Or at least this commandment is grounded in, and
depends for its “binding force” upon that claim.

6. But surely we cannot think that binding legal norms – laws – are merely
summaries of what courses of action are open or closed to us, in a circumstance-
independent manner, from the point of view of a system of essential values or
goods. The legislator is not merely reporting or reiterating our obligations; he is
making law, and is binding us through his act 27. How can this be done? If the
approach taken here has any merit to it whatsoever, we will have to say that it
can only be done by tying a positive rule (i.e. a rule which has been set) to a
system of essential values in some appropriate way.
I can think of at least two ways in which this can be done. The first way is to
set a rule for action which is backed by a practice designed to bring non-com-
pliance into conflict with a system of essential values: For example, a tyrant
26 See my article, “Defeating the Inference from General to Particular Norms”, Ratio
Juris 8 (December, 1995) 3, pp. 271-286.
27 I do not suggest here that the legislator can make whatever binding rules he wants.
I believe, indeed, that there are firm and important limits to what norms can be set that
will have the “binding” force of law. But this does not negate the present point.
12

might levy taxes and might institute the practice of killing the children, parents
or loved ones of those failing to pay. The idea under discussion here is in most
respects what is envisioned by those who think of the power of legal norms as
“coercive”. But two things must be said which are not usually said by “coer-
civists”, and both of which are important. First, we have specified that the
coercive practice must bring non-compliance with a set rule into conflict with a
system of essential goods or values, if the rule is actually to be binding. So, if a
rule is set forbidding overtime parking in downtown Rennes, and this rule is
backed by a practice of levying a fine of 100 francs upon those who do not
comply, then this rule, although coercive, will not be binding; for there are no
essential values or goods which must be sacrificed for non-compliance. If one
does not want to be fined, one may indeed be obliged to comply with the rule;
but one is not bound, or obligated, to do so; at least this is so if the force of the
rule derives solely from the conditional sacrifice of a non-essential good.
Moreover, it will not be apparent, if the fine is supposed to tell the whole story,
why it would be inappropriate to decide simply to park overtime as often as one
wanted and consider the fine to be just a “cost of doing business”. The tyrant’s
tax law is another story, for there non-compliance creates a conflict with an
essential good –human life– and one is thus obligated to pay the tax. The conflict
with essential goods in this case is also what would make it inappropriate simply
to decide to absorb the consequences of non-compliance as a cost of doing
business. But coercive theories generally do not make the distinction between
essential and non-essential values in this connection, and thus fail to distinguish
being coercively obliged from being coercively obligated.
But even if non-compliance with a set rule is brought, by artifice, into
conflict with essential goods or values –thus creating the obligation to comply–
the result would not be the one sought by the coercivist. For legal obligations are
supposed to be obligations to the body politic: to the state, to the people, to the
prince, or perhaps to the law itself. But this is not what happens when a rule is
coercively backed in the way here discussed. To continue with the tyrant’s tax
law: one would be obligated to pay the tax; but the obligation would be to one’s
parents, children and loved ones, and perhaps to oneself. It would not be to the
tyrant, to the state or to the body politic. These might be the beneficiaries of the
obligation, but they would not be the recipients of it. The point being made here
is essentially the same that was made earlier in replying on behalf of the present
account to Hart’s criticism of Austin.
I take it, then, that the coercivist approach does not yield a satisfactory
account of legal obligation or the “binding force of law”.
However, there is a second way in which set rules might be connected up
with a system of essential goods or values so that they could be seen as
“binding” in a more appropriate manner; and so that they could be seen,
moreover, as binding legal creations, and not merely as reports concerning the
options which a system of essential values closes, or leaves exclusively open, to
13

an agent. This is where the legislator sets rules which further essential goods and
values through the medium of organized, civil society; indeed, the rules thus set
may be seen as partially constitutive of civil society, understood as an artificial
institution ordered to this end. This last sentence sounds a bit pompous, so let us
explain what is meant in terms of examples.
Suppose, very simply, that a legislator imposes, by law, a tax on luxury cars,
intended to support expanded and improved health care. Let us imagine that this
is lex imperfecta: there is no penalty for not paying the tax. Nevertheless, the rule
requiring payment of this tax is set according to recognized procedures and
addressed to all citizens. Now I maintain that this rule can have the status of a
binding general norm; that those buying luxury cars can have an obligation to
pay this tax. Since I have made the norm lex imperfecta, we cannot suppose that
the obligation could have a coercive basis. The root of the obligation would be
the fact that the actions required by this legal norm help to manifest essential
goods, in this case, better health (I take it that health is a convincing example of
such a good, if we countenance such goods at all). But the vehicle for doing this
is created by the legislator in setting the rule; serving better health in this way
was not a pre-existing option. Perhaps there was no pre-existing option. Here, an
essential good is being furthered through the medium of civil society, which is
first and foremost a certain sort of organization, constituted by rules. These rules
may serve essential goods in ways that nothing else does; in that case, we may,
for that very reason, have an obligation to follow them. Or they may be
destructive of essential goods; in that case, we may, for that very reason, have an
obligation not to follow them. If civil society is on the whole ordered to the
furthering of essential values and goods, we may, for that reason, have an
obligation to follow rules of civil society – laws – which do not themselves serve
essential goods, in order to preserve the wholesome instrument of civil society.
In a reasonably good society, we have a prima facie obligation to be law-abiding.
A binding legal norm so conceived does not merely constitute a report of
what options for action are open or closed to us from the point of view of a
system of essential values, for it creates new options for action which are to be
tried against such a system of values.

7. I have claimed, then, that what makes laws binding is their furthering
essential goods or values; this might be called the proper connection between
legal norms and values of this kind. I have not proven that there are any essential
values; nor have I addressed the question by what evidence we might establish
which goods or values are essential, if any are (for this might be disputed even
among those who acknowledge the existence of such values). Those questions lie
beyond the scope of this paper. What I am urging here is primarily that if
someone does not recognize a distinction between essential and non-essential
goods, or some closely corresponding distinction, then he will not be able to
14

make out the difference between being obliged and being obligated. He will, I
believe, ultimately have to deny that any such distinction exists; and he will
thereby be unable (though perhaps gladly unable) to account for the “binding
force” of certain legal norms.
It is worth noticing that one need not venture into Kantian metaphysics in
order to countenance a distinction between essential and non-essential goods.
John Stuart Mill, an uncompromising empiricist and utilitarian, makes out just
such a distinction in his Utilitarianism. Mill thinks, in a very un-Kantian way,
that all goods are “pleasures”. But he says:
It is quite compatible with the principle of utility to recognise the fact, that some
kinds of pleasure are more desirable and more valuable than others. It would be
absurd that while, in estimating all other things, quality is considered as well as
quantity, the estimation of pleasures should be supposed to depend on quantity
alone.
If I am asked … what makes one pleasure more valuable than another … except
its being greater in amount, there is but one possible answer. Of two pleasures, if
there be one which all or almost all who have experience of both give a decided
preference … that is the more desirable pleasure. If one of the two is, by those
who are competently acquainted with both, placed so far above the other that they
… would not resign it for any quantity of the other pleasure which their nature is
capable of, we are justified in ascribing … a superiority in quality, so far
outweighing quantity as to render it, in comparison, of small account 28.
Here, Mill not only frames a distinction akin to our distinction between
essential and non-essential goods –these kinds being incommensurable with one
another and the former more valuable than the latter– but also advances a thesis
about the evidence that would be material to identifying essential goods. And
Mill maintains that, on that evidence, there are indeed such goods, or “higher”
pleasures, as he calls them. For, he says:
… it is an unquestionable fact that those who are equally acquainted with, and
equally capable of appreciating and enjoying, both, do give a most marked
preference to the manner of existence which employs the higher faculties. Few
human creatures would consent to be changed into any of the lower animals, for a
promise of the fullest allowance of a beast’s pleasures; no intelligent being would
consent to be a fool, no instructed person would be an ignoramus, no person of
feeling and conscience would be selfish and base … 29
In the language used in the present paper, Mill recognizes intelligence, know-
ledge and moral virtue as essential values. Without making any attempt here to
assess the merits of Mill’s particular position, the example should make clear that
the distinction between essential and non-essential goods is neutral between
widely divergent moral and metaphysical theories.

28 Mill, Utilitarianism, p. 211. Cf. Aristotle, Nicomachean Ethics X, 3, esp. 1173b30-


1174a10.
29 Mill, ibid.
15

8. In the last section, I mentioned some important questions which are not
treated in this paper. For my present efforts are devoted mainly to mapping out a
position in such a way that we can see more clearly which projects have to be
pursued by anyone who finds the overall approach tempting. But before closing
there is one further issue that I do want to address. In § 4 above, we described
essential goods as “things that we ought to value, whether or not we do; and
things towards which it is appropriate for us to have an attitude of respect.” And
then we went on to define obligation in terms of goods of that kind. Now this
might well strike someone as circular, or almost circular: we define what we
ought to do in terms of what we ought to value. The puzzling notion of obligation
seems to be explicated in terms of obligation.
To those antecedently persuaded that an “ought” can never be derived from
an “is”, this should come as no surprise. But it may nevertheless appear that the
thesis of this paper is ultimately trivial. I do not think that it is. What is true, I
think, is that all “oughts” come down to rational necessitation – what is
rationally necessary in some sense 30 – just as Kant thought. There are
nevertheless differences between hypothetical and categorical necessity, and
between practical and theoretical necessity 31.
Now, when we say that there are things that we ought to value whether we
happen to value them or not, we are, I believe, relying upon a different variety
– or at least a different instance – of rational necessity than that which pertains to
what we are obligated to do (the categorical necessity of a certain form of
action). What we are saying depends upon the idea that there are certain goods
which are really more valuable than the general run of goods and whose value is
incommensurable with ordinary (non-essential) goods. These goods are goods in
themselves, not goods which depend upon anyone’s desiring or appreciating
them. In short, the basis for talking about what we ought to value is the idea that
there are essential goods. We can admit that there are such goods while at the same
time disagreeing, or being uncertain, about which goods have this special kind of
value; that is, we can see that something might really be good in this way without
our realizing that it is 32. Now it wouldn’t make much sense to assert: that (1) “x has
a certain value in itself (independent of our attitudes) which makes it superior to
all of the things that we value in the ordinary way (its value cannot be measured
in terms of ordinary goods)” and to go on to say at the same time (2) “I have no
reason to value x”. Something’s having value of this kind – essential value – is a
reason for valuing it, and if I recognize (or, less tendentiously, believe) that
something has essential value then I, myself, have a reason to value it (the belief
constitutes such a reason). The necessity of valuing something – that one ought
30 Note that I do not claim here all all necessity comes down to rational necessity; in
other words, I leave open the possibility that there are forms of necessity not aptly
categorized as “oughts”.
31 As Kant himself emphasized.
32 Or, we can think that something is good in this way when it is not.
16

to value it, that it is appropriate to value it, that valuing it is non-optional –


comes from the connection in reason between something’s having essential value
and there being a reason to value it. The argument to what one is obligated to do
depends upon a different connection, namely the possibilities of actions that are
open to us, system Ve of essential values or goods. I maintain therefore that the
position advanced in this paper, while it may conceivably be quite wrong, is
neither circular nor trivial.
Notice that there is no argument pertaining to non-essential values (here
construed as attitude-dependent) parallel to that just given for essential values.
For one might very well happen to value or cherish something, say Coca-Cola,
without thereby having any reason to cherish it. One could not necessarily be
faulted for one day ceasing to cherish it, as one could be faulted for ceasing, or
failing, to cherish an essential value. Non-essential values are optional for the
individual, 33 as essential values are not.
This having been said, I close my present discussion concerning the roots of
legal normativity. If someone were to complain that this paper raises more
questions than it answers, he might well be right; but that will not distress me if
at least some of the questions are thought to be interesting and significant ones.

33 There may be various important exceptions to this rule, but this appears to me to
be the general rule for non-essential values as it is not for essential values.

Vous aimerez peut-être aussi