unchallenged control over a fixed territory. Or we might call a regime legitimate that was
simply “lawful,” in the sense that it came to power and continues to govern according to the
generally accepted rules of its state, or if it refrained from the persecution or deliberate
impoverishment of its subjects (or of particular groups of subjects).(23) Or we might, with
the majority of contemporary social scientists (and following Weber), call “legitimate” those
regimes that are accepted or approved of by their subjects in certain distinctive ways.
The satisfaction of any of these criteria might, of course, be said to confer on a regime the
kind of moral legitimacy we have been discussing. But while concerns about a state’s
stability or lawfulness or about institutional racism and persecution are clearly moral
concerns, they are concerns bearing more obviously on what I have been calling the state’s
justification than on its legitimacy. That a state is stable and lawful and refrains from
persecution shows that it is good (or, at least, not bad) in certain ways, but it does not
obviously show that the state has the kind of special moral relationship with any particular
subjects that gives it a right to rule them. And international recognition, considered alone,
plainly tracks the moral legitimacy of states at best irregularly. What, though, of the last,
Weberian conception of legitimacy? Its popularity and its apparent similarity to the Lockean,
consent-based conception of legitimacy warrants a slightly more extended consideration of
this proposal.
One proponent of the Weberian view, Charles Taylor, distinguishes between two senses of
“legitimacy” as follows. On his preferred use, legitimacy “is meant to designate the beliefs
and attitudes that members have toward the society they make up. The society has
legitimacy when members so understand and value it that they are willing to assume the
disciplines and burdens which membership entails. Legitimacy declines when this willingness
flags or fails.”(24) Worries about the possibility of a contemporary “legitimation crisis” are
often understood in this way–that is, in terms of the special difficulties faced by today’s
industrial democracies in maintaining or generating the attitudes of allegiance, loyalty, or
identification on which their “legitimacy” (in this first sense of the word) depends.(25) Taylor
contrasts legitimacy in this first, “attitudinal” sense with what he calls “the seventeenth
century use of the term not to describe people’s attitudes, but as a term of objective
evaluation of regimes.”(26)
The majority of the social scientists writing about legitimacy during the second half of this
century have, like Taylor, identified legitimacy with members’ positive beliefs, attitudes,
perceptions, or other “favorable orientations” toward their society or its regime. In this, as I
have noted, they mostly take themselves to be following Weber, who famously attempted to
analyze the legitimacy of power solely in terms of people’s belief in its legitimacy.(27)
The most familiar criticism of this analysis of legitimacy points to its quite obvious
circularity. But this is not a particularly difficult problem to repair, for Weber can be easily
corrected to say more carefully (as he himself sometimes does) that the extent of a regime’s
legitimacy is equivalent to the extent to which its subjects regard its directives as obligatory
or authoritative, or regard the regime as lawful, exemplary, morally acceptable, or
appropriate for the society. Legitimacy is then just understood as the “reservoir of loyalty on
which leaders can draw,”(28) the subjects’ beliefs in the regime’s authority (or their feelings
of allegiance, trust, or other attachment) that will typically produce compliance and support
(or at least guilt feelings on occasions of noncompliance and nonsupport).(29)
There are, however, more serious problems facing attitudinal accounts of legitimacy. One is
that such accounts make judgments of legitimacy turn out to be about the wrong thing. Just
as subjectivist accounts of moral judgment implausibly understand my judgment that an act
is wrong, say, as a statement that I have negative feelings about that act–so that the “moral
judgment” oddly turns out to be about me instead of about the act–so attitudinal accounts of
political legitimacy make judgments of legitimacy too much about subjects and too little
about their states. To call a state legitimate is surely to say something about it, about the
rights it possesses or the scope of its authority. The attitudes of a state’s subjects can at best
be part of what argues for its legitimacy, not that in which its legitimacy consists.
It will not do, however, in response to this problem, to simply shift our focus onto the
properties of the state that produce feelings of allegiance or support, so that legitimacy can
be redefined as “the capacity of the system to engender and maintain the belief that the
existing political institutions are the most appropriate ones for the society.”(30) For there is a
second and much deeper problem with all accounts of legitimacy that thus centrally refer to
subjects’ beliefs or attitudes: no plausible theory of state legitimacy could maintain that a
state has the rights in which its legitimacy consists–rights to exclusively impose and
coercively enforce binding duties on its subjects–simply in virtue of its subjects’ feelings of
loyalty or its own capacities to generate such feelings. Surely by now the history of human
oppression has taught us how often people come to feel obligated toward and believe in the
rights of those who simply wield over them irresistible power, with no more moral authority
over them than such power yields. Attitudinal accounts of state legitimacy appear to
disregard such lessons. On such accounts states could create or enhance their own
legitimacy by indoctrination or mind control; or states might be legitimated solely by virtue of
the extraordinary stupidity, immorality, imprudence, or misperceptions of their subjects.
31b