A theory of Justice: Revised Edition

partial compliance theory. Thus granting that adherence to whatever con-
ception is acknowledged will be imperfect if left completely voluntary,
under what conditions would the persons in the original position agree
that stabilizing penal devices can be employed? Would they insist that a
person can be required to do only what is to his advantage as defined by
the thin theory?
It seems clear, in the light of the contract doctrine as a whole, that they
would not. For this restriction amounts in effect to general egoism which,
as we have seen, would be rejected. Moreover, the principles of right and
justice are collectively rational; and it is in the interest of each that
everyone else should comply with just arrangements. It is also the case
that the general affirmation of the sense of justice is a great social asset,
establishing the basis for mutual trust and confidence from which all
normally benefit. Thus in agreeing to penalties that stabilize a scheme of
cooperation the parties accept the same kind of constraint on self-interest
that they acknowledge in choosing the principles of justice in the first
place. Having agreed to these principles in view of the reasons already
surveyed, it is rational to authorize the measures needed to maintain just
institutions, assuming that the constraints of equal liberty and the rule of
law are duly recognized (§§38–39). Those who find that being disposed
to act justly is not a good for them cannot deny these contentions. It is, of
course, true that in their case just arrangements do not fully answer to
their nature, and therefore, other things equal, they will be less happy
than they would be if they could affirm their sense of justice. But here one
can only say: their nature is their misfortune.
The main point then is that to justify a conception of justice we do not
33. See Foot, pp. 99–104.
504
The Good of Justice

---

have to contend that everyone, whatever his capacities and desires, has a
sufficient reason (as defined by the thin theory) to preserve his sense of
justice. For our good depends upon the sorts of persons we are, the kinds
of wants and aspirations we have and are capable of. It can even happen
that there are many who do not find a sense of justice for their good; but
if so, the forces making for stability are weaker. Under such conditions
penal devices will play a much larger role in the social system. The
greater the lack of congruence, the greater the likelihood, other things
equal, of instability with its attendant evils. Yet none of this nullifies the
collective rationality of the principles of justice; it is still to the advantage
of each that everyone else should honor them. At least this holds true so
long as the conception of justice is not so unstable that some other
conception would be preferable. But what I have tried to show is that the
contract doctrine is superior to its rivals on this score, and therefore that
the choice of principles in the original position need not be reconsidered.
In fact, granted a reasonable interpretation of human sociability (provided
by the account of how a sense of justice is acquired and by the idea of
social union), justice as fairness appears to be a sufficiently stable con-
ception. The hazards of the generalized prisoner’s dilemma are removed
by the match between the right and the good. Of course, under normal
conditions public knowledge and confidence are always imperfect. So
even in a just society it is reasonable to admit certain constraining ar-
rangements to insure compliance, but their main purpose is to underwrite
citizens’ trust in one another. These mechanisms will seldom be invoked
and will comprise but a minor part of the social scheme.
We are now at the end of this rather lengthy discussion of the stability
of justice as fairness. The only further point to note is that congruence
allows us to complete the sequence of applications of the definition of
goodness. We can say first that, in a well-ordered society, being a good
person (and in particular having an effective sense of justice) is indeed a
good for that person; and second that this form of society is a good
society. The first assertion follows from congruence; the second holds
since a well-ordered society has the properties that it is rational to want in
a society from the two relevant points of view. Thus a well-ordered
society satisfies the principles of justice which are collectively rational
from the perspective of the original position; and from the standpoint of
the individual, the desire to affirm the public conception of justice as
regulative of one’s plan of life accords with the principles of rational
choice. These conclusions support the values of community, and in reach-
ing them my account of justice as fairness is completed.
505
86. The Good of the Sense of Justice

---

87. CONCLUDING REMARKS ON JUSTIFICATION
87. Remarks on Justification
I shall not try to summarize the presentation of the theory of justice.
Instead I should like to end with a few comments about the kind of
argument I have offered for it. Now that the whole conception is before
us, we are in a position to note in a general way the sorts of things that
can be said on its behalf. Doing this will clarify several points which may
still be in doubt.
Philosophers commonly try to justify ethical theories in one of two
ways. Sometimes they attempt to find self-evident principles from which
a sufficient body of standards and precepts can be derived to account for
our considered judgments. A justification of this kind we may think of as
Cartesian. It presumes that first principles can be seen to be true, even
necessarily so; deductive reasoning then transfers this conviction from
premises to conclusion. A second approach (called naturalism by an
abuse of language) is to introduce definitions of moral concepts in terms
of presumptively non-moral ones, and then to show by accepted proce-
dures of common sense and the sciences that the statements thus paired
with the asserted moral judgments are true. Although on this view the first
principles of ethics are not self-evident, the justification of moral convic-
tions poses no special difficulties. They can be established, granting the
definitions, in the same fashion as other statements about the world.
I have not adopted either of these conceptions of justification. For
while some moral principles may seem natural and even obvious, there
are great obstacles to maintaining that they are necessarily true, or even to
explaining what is meant by this. Indeed, I have held that these principles
are contingent in the sense that they are chosen in the original position in
the light of general facts (§26). More likely candidates for necessary
moral truths are the conditions imposed on the adoption of principles; but
actually it seems best to regard these conditions simply as reasonable
stipulations to be assessed eventually by the whole theory to which they
belong. There is no set of conditions or first principles that can be plausi-
bly claimed to be necessary or definitive of morality and thereby espe-
cially suited to carry the burden of justification. On the other hand, the
method of naturalism so-called must first distinguish moral from non-
moral concepts and then gain acceptance for the definitions laid down.
For the justification to succeed, a clear theory of meaning is presupposed
and this seems to be lacking. And in any case, definitions become the
main part of the ethical doctrine, and thus in turn they need to be justified.
Therefore we do better, I think, to regard a moral theory just as any
506
The Good of Justice

---

other theory, making due allowances for its Socratic aspects (§9). There is
no reason to suppose that its first principles or assumptions need to be
self-evident, or that its concepts and criteria can be replaced by other
notions which can be certified as non-moral.
34
Thus while I have main-
tained, for example, that something’s being right, or just, can be under-
stood as its being in accordance with the relevant principles that would be
acknowledged in the original position, and that we can in this way replace
the former notions by the latter, these definitions are set up within the
theory itself (§18). I do not hold that the conception of the original
position is itself without moral force, or that the family of concepts it
draws upon is ethically neutral (§23). This question I simply leave aside.
I have not proceeded then as if first principles, or conditions thereon, or
definitions either, have special features that permit them a peculiar place
in justifying a moral doctrine. They are central elements and devices of
theory, but justification rests upon the entire conception and how it fits in
with and organizes our considered judgments in reflective equilibrium. As
we have noted before, justification is a matter of the mutual support of
many considerations, of everything fitting together into one coherent
view (§4). Accepting this idea allows us to leave questions of meaning
and definition aside and to get on with the task of developing a substan-
tive theory of justice.
The three parts of the exposition of this theory are intended to make a
unified whole by supporting one another in roughly the following way.
The first part presents the essentials of the theoretical structure, and the
principles of justice are argued for on the basis of reasonable stipulations
concerning the choice of such conceptions. I urged the naturalness of
these conditions and presented reasons why they are accepted, but it was
not claimed that they are self-evident, or required by the analysis of moral
concepts or the meaning of ethical terms. In the second part I examined
the sorts of institutions that justice enjoins and the kinds of duties and
obligations it imposes on individuals. The aim throughout was to show
that the theory proposed matches the fixed points of our considered con-
victions better than other familiar doctrines, and that it leads us to revise
and extrapolate our judgments in what seem on reflection to be more
satisfactory ways. First principles and particular judgments appear on
34. The view proposed here accords with the account in §9 which follows “Outline for Ethics”
(1951). But it has benefited from the conception of justification found in W. V. Quine,