Philosophical Investigations 25:1 January 2002

ISSN 0190-0536

How is Political Authority Possible?

Peter Winch

Political philosophy has always been a mëlange of political advocacy

and philosophical analysis. I am not saying that is a bad thing; there is
no doubt that the greatest work in political philosophy has been
written by philosophers in the service of some political cause to
which they have been deeply committed. This may even be
inevitable. Such overlapping concerns are nowhere more evident
than in the treatment of the nature of political authority. It is quite
evident in the work of, for instance, such writers as Plato, Hobbes,
Locke, Rousseau, and, in our own time, Hannah Arendt and Simone
Weil, that the accounts they have offered of `what authority is' are
inseparable from their evident views about the preferability of some
kinds of civil society over others. This situation suggests the
possibility that the question `what political authority is' gains its sense
from a context of particular political concerns, concerns which, of
their nature, will not be shared by everybody and will be opposed by
the concerns of others. I am not going to discuss that interesting
possibility here, but the issue which I am going to discuss illustrates
the general situation I have sketched and is consequently peculiarly
difficult to get hold of.
To put the matter crudely: those who have discussed the nature of
political authority have been exercised both about what can give a
person the right to that sort of power over another and also about
how exactly that power is to be understood. And these two apparently
distinct questions tend to run together.
There is a good example of this in Locke in a passage which raises
the issue which is the subject of my talk and gives us an important
hint how we should proceed. Locke believed that human beings are
naturally free. By this he meant roughly that they are subject to no
human authority which they have not themselves voluntarily
accepted. In defending this position he considers the objection that
children are `naturally' subject to the authority of adults, especially

ß Blackwell Publishers Ltd. 2002, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148,
USA.
Peter Winch 21
their parents, and acknowledges that his claim applies directly only to
adults. He goes on to explain:1
Thus we are born free as we are born rational, not that we have
actually the exercise of either; age that brings one brings with it the
other, too. And thus we see how natural freedom and subjection to
parents may consist together and are both founded on the same
principle.
His claim is not just that there is an analogy, but that there is a close
conceptual connection between rationality and freedom in respect of
their development with maturity. Age brings rationality and `with it'
freedom. Freedom is the right to decide one's own actions for oneself;
and only one who is rational can intelligibly be said to have the
power, and hence the right, of such decision. So much is not too
controversial.
But Locke is also making a much stronger, and more interesting,
claim: that the capacity for agency is at the same time a sort of right: in the
sense at least that it excludes the possibility that anyone else should be
entitled to decide on one's behalf in the absence of one's own rational
consent to such an entitlement.2
I reconstruct3 the reasoning behind this position somewhat as
follows: An action can properly be ascribed to an agent only in so far
as it flows from his or her will. This appears most clearly if one
considers those cases in which an action is ascribed to a person, A,
though physically carried out by B, on the grounds that its
performance flowed from A's, rather than B's will. (As when, for
example, A tricks B into unknowingly putting poison into a third
party's food.) But in this context to speak of `will' is to presuppose
practical rationality: A is presumed to have acted in the light of his or
her own assessment of the situation. So the primitive source of the
idea of agency is the immediate relation of actions to practical
rationality. In this context `rationality' means, roughly, an agent's
capacity to `account for' the action, by trying to show it as justified in

1. The Second Treatise of Government, New York, The Liberal Arts Press 1952, pp.
34±35.
2. There is a somewhat similar inference from a supposed fact about human nature
to a right in Locke's treatment of property: where he equates the fact that a certain
body is `my body' with my having a right of property in that body. This inference is
an essential foundation of Locke's attempted derivation of property from labor.
3. I am not suggesting that anything like this argument is actually set out by Locke.
My suggestion is that we can make better sense of his position if we see it in the light
of such an argument.

ß Blackwell Publishers Ltd. 2002

22 Philosophical Investigations
the light of the way he or she assessed the context in which the action
was performed. It would hardly make sense to demand such a
justification except of an agent thought of as entitled to act in the
light of such an assessment on his or her part. The possibility of
`calling someone to account' is a condition for ascribing to him or her
the kind of rationality involved in agency and, furthermore,
presupposes a right of autonomous decision.
Let us turn now to cases of acting on someone else's authority.
Here, the link between action and will is obviously complicated; it is
not, however, broken as it is when I am coerced by external force. I
still bear responsibility for it and can be called to account (although of
course the accounting may well be modified by the fact that the
action was done on the authority of another). This means, however,
that such authority cannot be imposed on me from without; I must
in some sense voluntarily accept it; otherwise the essential link with
my will is broken, and the actions I perform under such alleged
`authority' would no longer be in the relevant sense `mine'.
There is, however, a difficulty in seeing how this link with my
own will can be preserved when I obey someone else. The difficulty
is well brought out in an important passage from Hobbes's Leviathan,
which offers an analysis of the difference between `command' and
`counsel'.4
COMMAND is, where a man saith, do this, or do not this without
expected other reason than the will of him that says it. From this it
follows manifestly, that he that commandeth, pretendeth thereby
his own benefit: for the reason of his command is his own will
only, and the proper object of every man's will, is some good to
himself.
COUNSEL, is where a man saith, do, or do not this, and deduceth
his reasons from the benefit that arriveth by it to him to whom, he
saith it. And from this it is evident, that he that giveth counsel,
pretendeth only, whatsoever he intendeth, the good of him, to
whom he giveth it.
Hobbes states the distinction in terms of one of the central theses in
his account of human nature: that `the proper object of every man's
will, is some good to himself'. But the point he is making does not
depend on that and can be stated in terms that are, perhaps, likely to
win wider acceptance: viz. that a person's will is directed at the

4. Leviathan, Part II, Chapter XXV. 1988, by the way, is the four hundredth
anniversary of Hobbes' birth (Presumably, the date of Winch's paper ± Ed.).

ß Blackwell Publishers Ltd. 2002

Peter Winch 23
realization of his or her own projects (using this last word in the
widest possible sense).5 This has the advantage over Hobbes'
formulation that we are not tied to the doctrine that projects are
always directed at the good of the agent. From this point of view,
then, one who commands, according to Hobbes, claims only to be
pursuing his or her own projects, whereas one who advises, claims to
be promoting the projects of the advisee. So the person who accepts
advice does so in the belief that it will tend to promote his or her own
projects; whereas the person who obeys a command does so, or is at
least expected by the commander to do so, simply because the
commander wills it. That is to say, he is not supposed to weigh the
pros and cons of doing what is commanded in the light of its bearing
on his own projects; that the commander wills it is supposed to be
enough.
But in what sense `enough'? This cannot mean merely: enough to
ensure that what is commanded gets done. To see this, consider the
case of post-hypnotic suggestion. Here the hypnosis induced is
enough to ensure the performance of the action which the hypnotist
wills. But this is not a case of command and obedience (or not at least
the kind of case that we are interested in). We might, stipulatively,
mark the difference by saying that the hypnotist's will is the cause of
the subject's acting, but not the subject's reason for acting. And this
brings out the difficulty in what Hobbes says. How can the
commander's will be the subject's reason for acting, if the subject is
not supposed to be evaluating the pros and cons of obedience in terms
of his or her own projects? Many people would feel that to put the
matter thus would be to evacuate the term `reason' of all meaning:
that the parameters of a person's will are that person's own projects,
and own judgments as to how these projects are best to be pursued;
that to remove these from consideration is to eliminate the possibility
of the subject's retaining any sort of responsibility6 and, by the same
token, to attenuate any idea that the agent acts `for a reason' or that
the action springs from his or her will.
5. I do not mean to suggest that we ought to accept this weaker thesis; merely that it
is less obviously untenable than Hobbes' own. Indeed, analyses of reasons for action
widely accepted by present-day philosophers, in terms of `desires' (or `pro-attitudes')
and `beliefs', are a form of the same view. I shall come back to difficulties in the
weaker thesis later.
6. The issue we are concerned with is thus intimately bound up with questions
about the nature and extent of a citizen's responsibility for the acts of the state of
which he is a member.

ß Blackwell Publishers Ltd. 2002

24 Philosophical Investigations
Hobbes himself is of course by no means blind to such
considerations. Indeed his whole central argument, concerning the
reasons why, and the mechanism through which, a central
sovereign authority is established, appeals to them. People have
reason for accepting a sovereign authority: it is the only way of
establishing the order necessary for them to secure their own
existence and pursue their interests in any settled way. They
therefore voluntarily agree to surrender their individual freedom of
decision and give the sovereign the power `of the sword' to enforce
this agreement. Having done this, they then have a dual reason for
obeying the sovereign's commands: their own prior agreement and
their fear of sanctions.
Of course it is no news to anybody that Hobbes' position is beset
with all kinds of difficulty. Two of them (which are
interconnected) are particularly relevant to my present concerns.
First, the argument leads to a conclusion which seems inconsistent
with the definition of `command'. It turns out that the subject who
does what the sovereign commands does not after all do so merely
because the sovereign wills it (whatever the sovereign may think),
but because, for reasons of his own, he has agreed to do so, and
because it is in his present interests to do so (to avoid retaliation).
Hence, second, any obligation the subject is under to obey is
essentially limited: in particular, he cannot be under any obligation
to obey a command which will certainly lead to his own
destruction. Hobbes does of course recognize this though I do not
believe he realizes the extent to which it threatens his overriding
aim of ensuring that only the sovereign has the right to determine
what should and should not be required of the subjects. It seems to
imply not merely that the obligation to obey is a limited one, but
that the subject must be the ultimate judge of where that limit lies.
It was just such a situation, in which each individual is the ultimate
judge of what he or she is entitled to do, that Hobbes most feared
and wanted to preclude.
These implications severely compromise his account of `command'
and hence of the sovereign's authority. They are generated by his
assumption that this authority must initially spring from the individual
wills of the subjects. And he assumes this, I think, because he takes it
to be the only way in which the subject's wills can play enough of a
role in their obedience to the sovereign's commands for them to be
held accountable for it.
ß Blackwell Publishers Ltd. 2002
Peter Winch 25
Hobbes is not of course alone in thus linking the nature of the
subject's obligation with the source of the sovereign's authority. The
same pattern is repeated again and again in social contract theories of
the state. I want now to look at this aspect of the matter.
There are plenty of cases in which it is true that someone's
authority is created by the consent of those over whom it is exercised.
For instance, I am one of a party going on a sailing expedition.
Because sailing is an activity requiring the coordination of crew
members' actions, and because this coordination is best achieved if
certain decisions are made by a single person and implemented by all,
we agree together at the start of the trip to accept the authority in
sailing matters of a particular member of the party, designated
captain, for the duration of the trip. Here the captain's authority
derives from our consent; so too ± the other side of the coin ± does
our commitment to be guided by the captain's judgment rather than
our own in matters of navigation. The link with our responsibility
for our own actions, though stretched,7 is not broken, because our
acceptance of the captain's authority is itself something we have
rationally chosen as being necessary for the successful completion of
our freely chosen project. His authority is, moreover, limited as to its
duration and scope by the purposes for which we accept it. In acting
on the captain's authority each of us can be understood still to be
exercising his or her own will.
Countless human situations can be construed according to
something like this model. Sometimes only quite short-term
arrangements are in question, as in the previous example; sometimes
the arrangements are very long-term and the conferment of authority
approaches the permanent. Again, the conferment of authority can
sometimes, as in the example, be dated in an explicit act; but
sometimes it can grow up without any express agreement between
people as an outcome of habitual intercourse ± which we often
enough construe as quite adequately expressing the consent of the
parties. On the other hand, the observance of explicit forms of

7. When I say that this link is `stretched', I mean that it is so if it is conceived

according to the philosophical model which is under scrutiny here. In ordinary life I
do not think we should regard it in this way. The situation I have imagined is
absolutely typical for the way in which an agent's responsibility is commonly
understood. In the longish term, satisfactory coping with immediate contingencies
typically requires us to trust, and in that sense consent to, someone else's dealing with
them on our behalf.

ß Blackwell Publishers Ltd. 2002

26 Philosophical Investigations
agreement is not always a sure criterion of free consent. There are
various kinds of case where it is not: blackmail (explicit and
concealed, material and emotional); deception, bribery, etc., etc.)8
My chief concern is, of course, with political authority; and the
question is whether it too can be construed along these lines. It is
important to be precise about what we are asking. The question is not
whether some notion of consent is central to our understanding of
political authority; I do not believe that should be in doubt. The
question is rather whether political authority can be thought of as
derived from, or as originating in, the consent of those over whom it is
exercised. The importance of this distinction will become clear if we
think for a while about some criticisms made by Hume of Locke's
classical form of contractarian theory.9
Hume distinguishes two sorts of criticism. The first is an attack on
Locke's use of the notion of `tacit consent', based on an appeal to
history and to our common understanding of political life; the second
is `a more regular, at least a more philosophical, refutation of this
principle of an original contract, or popular consent',10 and I shall
have more to say about it later. But there is a point of great interest
that I want to consider first in the way Hume states his more informal
objections. He imagines a usurper banishing `his lawful prince and
royal family' and maintaining himself so efficiently in power for
several years by military force that there has been no question of
insurrection or overt opposition. Hume asks ironically of such a case:
Can it be asserted that the people, who in their hearts abhor his
treason, have tacitly consented to his authority and promised him
allegiance merely because, from necessity, they live under his
dominion?
He asks us to imagine that the deposed prince returns victorious and
re-establishes himself in power, received by the people `with joy and
exultation'.
I may now ask upon what foundation the prince's title stands? Not
on popular consent surely; for though the people willingly

8. On such questions see Simone Weil, `The Legitimacy of the Provisional

Government' in Philosophical Investigations, Vol. 10, No. 2, April 1987; and many
other places in her writings.
9. David Hume, `Of the Original Contract' in his Political Essays, ed. Charles W.
Hendel, The Library of Liberal Arts 1953, pp. 43±63. This essay seems to me one of
Hume's very best works.
10. Op. cit., p. 54.

ß Blackwell Publishers Ltd. 2002

Peter Winch 27
acquiesce in his authority, they never imagine that their consent
made him sovereign. They consent because they apprehend him to
be already, by birth, their lawful sovereign. And as to tacit consent,
which may now be inferred from their living under his dominion,
this is no more than what they formerly gave to the tyrant and
usurper.
I want to direct attention to the remarkable phrase: `They consent
because they apprehend him to be already, by birth, their lawful
sovereign'. It is remarkable because it occurs in the very middle of an
argument intended to refute the doctrine that the sovereign's
authority derives from the subjects' consent. In other words Hume is
distinguishing between (1) the question what is the nature of the
subject's obligation to a sovereign whose authority he recognizes as
legitimate and (2) the question of the origin of that legitimate
authority. This is no mere slip of the pen as we can see from another
famous essay of his, `Of the First Principles of Government',11 which
opens with the well-known observation:
Nothing appears more surprising to those who consider human
affairs with a philosophical eye than the easiness with which the
many are governed by the few and the implicit submission with
which men resign their own sentiments and passions to those of
their rulers.
The way in which Hume develops this remark shows that his interest
in the problem is not just, though it is in part, a `sociological' interest
in the mechanisms which allow some people to gain and retain power
over others. He is interested also in the more strictly `philosophical'
question how such power is to be understood, what is its nature. He
goes on to observe that the political domination of the many by the
few cannot be understood solely as the exercise of force, but requires
something on the side of the governed that he calls `opinion',
distinguishing between `opinion of interest' and `opinion of right'. The
first of these is a recognition of the `general advantage which is reaped
from government' and the second, amongst other things, a
recognition of the government's right to power. So we are back, as
far as the `right to power' is concerned, with the very same idea
expressed in Hume's discussion of the quoted example from the essay
`Of the Original Contract': that the people in his example who

11. Political Essays, ed. Charles W. Hendel, Bobbs-Merrill, The Library of Liberal
Arts, 1953, pp. 24±27.

ß Blackwell Publishers Ltd. 2002

28 Philosophical Investigations
welcome back their ruler from exile `consent because they apprehend
him to be already, by birth, their lawful sovereign'.
I believe that Hume here is pointing to the heart of the difficulty
about Locke's `tacit consent'. This difficulty is not that there is no
such thing, as some (like John Plamenatz) have been inclined to
maintain; for there obviously is. The difficulty is rather this. Locke
states12 his purpose as
To understand political power right and derive it from its original.
And although he does distinguish between legitimate and illegitimate
rëgimes, he quite clearly wants to include amongst the former many
(such as the Hanoverian settlement in Britain which he was especially
interested in supporting) where it was plainly not the case that each
citizen subject to the rëgime's authority had explicitly agreed to it. It
is this that forces him to the conclusion (magnificently lampooned by
Hume) that tacit consent to a government's authority in a region is
expressed through having `a lodging only for a week', `barely
travelling freely on the highway'; indeed `it reaches as far as the very
being of anyone within the territories of that government'.13
Now the civil authorities of any state will certainly take the view
that anyone within its territories is subject to their authority; and
most people would probably agree. But if a traveller is to be able to
consent to that authority, he or she must be able to recognize,
antecedently, where legitimate authority lies within the territories in
question. Given the existence of a legitimate authority, we might say,
anyone within the lawful bounds of its exercise is bound to defer to
it.
Locke's question is, precisely: what makes the exercise of power
and command an instance of legitimate authority? It may be said that
someone finding him or herself on alien territory can recognize
where authority lies through the behavior of the local inhabitants. It
may also be claimed that, in normal circumstances, such a person does
have an obligation to respect the authority which is locally
acknowledged. Both these claims seem to me true and important.
But they constitute a (largely unrecognized) difficulty for Locke. For
him, the judgment that someone is exercising legitimate authority
presupposes a disposition of the will, namely consent to that

12. Op. cit., Chapter II, p. 4.

13. Op. cit., Chapter VIII, p. 68.

ß Blackwell Publishers Ltd. 2002

Peter Winch 29
authority. It is not just, as it were, a piece of information which can
be established by observation. Yet it now looks as though that
voluntary consent cannot be given unless the one who exercises
legitimate authority is first identified as such. And that is precisely
what Hume brings out in the phrase that prompted these reflections:
`They consent because they apprehend him to be already, by birth,
their lawful sovereign'.
I shall return to this point shortly. But first I want us to take a look
at what Hume calls `a more regular, at least a more philosophical,
refutation of this principle of an original contract, or popular
consent'. His point can be quite briefly stated. The original contract
theory attempts to provide an explanation of how the citizen's will
can be bound by the will of the sovereign; it says that what makes this
possible is the citizen's prior agreement to be so bound. But now,
Hume objects, a question of exactly the same form can be raised
about this: how can the citizen's will be bound by his or her own past
promise? If practical rationality requires deciding on action on an
assessment of the best way to pursue one's own projects in the
circumstances (the present circumstances) one finds oneself in, then
allowing oneself to be constrained by one's own past promise is just as
irrational as allowing oneself to be constrained by the will of another.
In each case one is surrendering one's power to decide rationally for
oneself how best to act in the present situation. There are two points
to note here:
1. This objection (unlike the first one) is not immediately directed
at the account of the origin of the sovereign's authority, but at
the account of the citizen's obligation. That is, although (as I have
argued) Hume recognizes that the citizen's recognition of
obligation does indeed involve consent, he is arguing that the
obligation does not, any more than does the sovereign's
authority, derive from the citizen's consent.
2. The difficulty is similar to that which compelled Hobbes to
buttress the `obligation' created by the subject's original
agreement to the `covenant' with fear of the sovereign's
`sword':14
. . . covenants being but words and breath, have no force to oblige,
contain, constrain, or protect any man, but what it has from the
public sword . . .
14 Hobbes, Leviathan, Chapter 18.

ß Blackwell Publishers Ltd. 2002

30 Philosophical Investigations
Let me return to Hume's first objection in the light of this point.
Hume says of the citizens in the example I cited: `They consent
because they apprehend him to be already, by birth, their lawful
sovereign'. I will observe first that his point is a quite general one and
does not depend on the fact that his example is of a monarchical
regime. Elected officials may be said to acquire their authority from
the wills (votes) of the electors. But the nature of the authority which
they thus acquire must already be presupposed in the electoral
procedure. Something similar may even be said of the enactment of a
constitution: the definition of the roles of government officers
presupposes an understanding of the notion of government, and
hence of legitimate authority. No constituent assembly creates such a
notion.15
Now, `apprehending someone as being one's lawful (i.e.
legitimate) sovereign' is, as I have already noted, not just an
observational matter; it involves a disposition of the will. This is of
the utmost importance, because it undercuts the conception of
practical rationality in terms of which this whole set of issues has so
far been understood: the conception of an agent confronting the
world with certain desires and projects of his or her own and acting
according to his or her own assessment of the possibilities of success
in those projects given the facts of the environment. But it now looks
as though a practical disposition of the will is already presupposed in
the exercise of a certain kind of rationality. And the same thing has to
be recognized if we are to deal with the difficulty raised by Hume's
`more philosophical' refutation. Both the obligation of civil
obedience and the obligation of keeping promises require a certain
practical, or volitional, attitude to civil authority and promises;
otherwise these cannot figure in the requisite way as `reasons' for the
agent's behavior.
I will conclude by trying to indicate one direction in which this
conclusion points. I shall do this by taking up again a matter which I
touched on earlier: Locke's account of the rationale of the authority
exercised by adults (especially parents) over children. Remember that
he bases this on the fact that infants (in the quasi-legal sense of the
term) are not yet fully rational and therefore need to be directed in
their behaviour. The function of parental authority is thus largely

15. There is a tangle of difficult issues here which I cannot take any farther at this
point.

ß Blackwell Publishers Ltd. 2002

Peter Winch 31
protective; it enables children to mature, and this maturation of their
rational capacities is seen as a sort of natural process. There is, that is
to say, no recognition of the role played by the education of children
by adults in cultivating the latters' rational capacities, no recognition
of (what I take to be the fact) that human rationality, at least the type
of rationality required by the concept of political freedom, is a
cultural heritage, transmitted to children from the adult world into
which they grow.16 And as a matter of fact the empiricist
epistemology developed by Locke in An Essay Concerning Human
Understanding precludes any proper recognition of this.
To speak of rationality as a cultural heritage is to imply, amongst
other things, that it cannot be understood in abstracto but appears only
in a concrete context of institutions and traditions. Similarly consent,
one of the most important manifestations of practical rationality, is
recognizable only within such a context. I have already hinted at the
complexity of the notion of consent. We can apply that notion only
where we are in a position to recognize and assess the extraordinarily
multifarious circumstances which can qualify the judgment that
consent has been freely given. And such recognition and assessment
itself presupposes sensitivity to the difference made by different
varieties of cultural circumstance. Consent is something different, for
instance, in intimate personal relationships from what it is in the
realm of business dealings; it is different in the context of a religious
order from what it is in a sailing club; and so on ad infinitum. Most
importantly for our present purposes, what consent amounts to in
political contexts can be understood only by someone familiar with
the complex web of political relationships ± which differ of course
quite dramatically from one political culture to another ± in the
context of which it is exercised. And those relationships will quite
centrally involve an understanding of different kinds and degrees of
authority that go with particular political roles. Such an
understanding is one of the things that a child develops in the process
of growing into a culture; it can hardly be thought of as the result of
some natural process of maturation.

16. Locke does speak of the duty of parents to `inform' their immature off-spring.
But he is in no position to recognize the importance of training in what Professor
Anscombe felicitously calls `the practices of reason'. See G.E.M. Anscombe, Collected
Philosophical Papers, Vol. III, Basil Blackwell 1981. On this topic see too the important
essay, `Epistemology and Education' in Against Empiricism by R.F. Holland
(Routledge & Kegan Paul, 1980).

ß Blackwell Publishers Ltd. 2002

32 Philosophical Investigations
To understand how political authority is possible, we should try
first to understand the central role played in all education by responses
to adult authority on the part of children.17 This would make it easier
to recognize and acknowledge the very multifarious ways in which
practical rationality embodies from the outset habits of deferring to
and depending on the judgments and decisions of others (habits on
the prior existence of which the possibility of giving consent to
another's acting on one's behalf depends). We might then ask what
kind of influence is exercised on the habits thus formed by the
political structure of the society into the culture of which the children
are maturing.
So, having arrived at the threshold of my subject, I must
regretfully leave it.

Editorial Reference
D.Z. Phillips,
Dept of Philosophy,
University of Wales, Swansea
Swansea SA2 8PP

17. That is one of the themes of Wittgenstein's great work, On Certainty.

ß Blackwell Publishers Ltd. 2002