Book Review

The Bull in the China Shop: A Discussion of an Ambiguity Within Pettit’s Theory of Freedom as Discursive Control

Steven Youngblood

Pettit, Philip, A Theory of Freedom: From the Psychology to the Politics of Agency, Cambridge, Polity Press, 2001. ISBN: 0745620930

In Philip Pettit’s “A Theory of Freedom”, Pettit claims that being free to do something is being held responsible for what we do; so whatever theory of freedom we develop must allow the agent to be held responsible for the free actions that they do. In this paper I am going to examine Pettit’s claim about what a satisfactory theory of freedom would require, and discuss several ambiguities within the theory. However, within this reading two major interpretations may be taken: the first of which suggests that freedom can only be freedom when there is a moral ‘ought’ involved; the second is a more generous reading in which freedom includes all realms of responsibility.

To establish this connection, Pettit begins by discussing the “ought/can” condition, and its relevance to an adequate account of freedom:

The supposition is that we live in a world where we continually address ‘oughts’ to one another… and that we think of freedom just as the ‘can’ which such an addressing of ‘oughts’ presupposes… but we will be able to recognise that sometimes it is appropriate to address ‘oughts’ to others, sometimes not, and we will think of freedom as the capacity that makes the difference between such cases.[1]

Our ability to address ‘oughts’ to one another, and holding one another responsible for our respective actions, relies on the supposition that we are free (and able) to do what we ought. Responsibility suggests freedom; inasmuch as the ‘oughts’ we hold each other for imply the ‘cans’ of freedom. In this line of thought there exists a problem that we have left freedom to the realm of the moral; it takes the seemingly neutral ‘can’ of freedom and implicates it as essentially moral, deriving from the ‘ought’ of responsibility. This seems to be an ambiguous statement from Pettit, as while there are traditionally two kinds of ‘ought’ statements, Pettit only seems to rely on one kind for his argument. The idea that ‘ought’ implies ‘can’ is traditionally a moral claim; but there is another type of ‘ought’ statement, to which no moral status can be ascribed, an example of which might be “If you want to play the guitar well, then you ought to practice”. The problem is that throughout Pettit’s book, he uses ‘ought’ in a strictly moral sense, suggesting that all instances of ‘ought’ are moral, and thus useful in determining freedom as fitness to be held responsible.

Also, a majority of the examples that Pettit gives to restrict freedom are examples to do with morality. Coercion against one’s person is a limiter of freedom according to Pettit:[2] If I am being held up by a mugger, and he says, “Give me your money or I’ll kill your wife”, then you are in a state of coercion, thus limiting your freedom:

If I make a threat of exercising violence against you unless you do something, I may go on to discourse with you about why you should take that threat seriously and perform the action required. But I will no longer be discoursing with you from the baseline of all the considerations that were relevant prior to the threat…I will have restricted discourse, at best, to the issue of why you should take the threat seriously and act accordingly.[3]

So coercion is a threat against freedom, and as we should avoid limiting freedom, we ought not coerce. Another threat to freedom for Pettit is the great republican evil of domination[4]. When someone is being dominated, then he or she is under constant threat of having his or her life interfered with in some way or another:

Suppose that a person is manifestly vulnerable to obstruction or coercion or manipulation at the hands of another – in particular, to forms of interference that are not guided by the person’s avowable interests – because of having a lesser degree of relational or social power… they are ‘dominated’ in the way in which an employee may be dominated by an employer in a tough labour market… It is a commonplace of received lore that such a person will not be able to speak out in a forthright and free way – or act on a basis that such speech might justify – but must always have an eye out for what will please the powerful and keep them sweet.[5]

This is perhaps the most severe threat to freedom for Pettit: when one is being dominated, then one is not free. Domination is wrong, so we ought not do it… again a moral example used to back up Pettit’s account of freedom as fitness as responsibility. By limiting all of his freedom-restrictors to the realm of the moral, he seems to have also limited freedom to the very same realm.

I will now consider the ambiguities in Pettit’s argument. If we look at the first extract, then by considering the first interpretation (which suggests that freedom can only be freedom when there is a moral ‘ought’ involved), we can see that an appropriate (or suitable) theory of freedom will be one that makes the difference between cases to which we ascribe ‘oughts’, and cases to which we do not.[6] What this implies is that we need an account of freedom that will satisfy the condition of ascribing appropriate ‘cans’ to suitable ‘oughts’. Consider the following statement: ‘You ought to save a child from a burning building’. Is this a moral statement? The presence of the word ‘ought’ can be argued to fulfil this condition. But what if you are unable to save the child, due to the smoke blinding you and rendering you unconscious? Or, what if you are in a wheelchair and the child is on the second floor? Failure to save the child on these grounds is hardly cause for blame, but why not? Because in these instances, there is no corresponding ‘can’ to the ‘ought’, which means that we shouldn’t be held accountable for our actions. Perhaps if we alter the statement: ‘you ought (if you are able) to save a child from a burning building’. By adding in this condition of ‘can’, we now have a way of denoting moral responsibility to people for their actions (it would be irrational to blame somebody for not saving the child because they were in another city, say). Thus ‘to be able to’ implies freedom.

But what of instances in which we have a ‘can’ but no moral ‘ought’? Consider now the following statement: ‘I can play the guitar’. Is this a moral statement? No, it is not—there is no moral ‘ought’ that can be sensibly applied to this statement that can make it a moral one: ‘I ought to play the guitar’ does not denote moral responsibility, nor does ascribing moral equivalents (such as ‘I <should> play the guitar). At most, these are statements of intent or desire; I may feel that I should play the guitar to keep in practice or to occupy my mind away from philosophy; yet this statement will not have moral ascription; it is amoral.

And now, if we consider together both moral and amoral statements, we will be able to formalise their differences. Compare the statements ‘you ought to (if you are able to) save a child from a burning building’; with ‘I can play the guitar’. The former statement is decidedly moral, while the latter is not a moral statement. We can say that each statement has a ‘morality condition’ (to coin a phrase), with which we determine whether a statement is moral or amoral. A formalised example of a moral statement might look like this:

  1. (fM+) <You ought to (if you are able) save a child from a burning building>

…where (fM+) is where the morality function has been met, thus (1) is a moral statement. If a statement does not imply morality (that is, something for which I can not be held morally responsible for), then we can say the following:

  1. (fM-) <If I want to be good at the guitar, I ought to practice>

…where (fM-) is where the morality function does not apply to a statement. There is not the right kind of content in this statement to allow a moral ascription.[7]

If we accept what Pettit says, then we can say that a proper account of freedom will be one that differentiates between cases with a positive morality function (fM+) and a negative morality function (fM-). However, it does not seem to me that Pettit can make this claim. By suggesting that “…we think of freedom just as the ‘can’ which such an addressing of ‘oughts’ presupposes”,[8] then we are effectively restricting freedom to the realm of the moral, which thus implies that cases with a negative morality function (fM-) are not the sort of thing to which freedom applies. I have already shown though that there are cases of ‘can’ to which no ‘ought’ apply (e.g. I can play the guitar). To say that freedom does not apply in these instances does not make sense. I do not believe that Pettit wants to say that I am not free to play the guitar, but this seems to be what this interpretation leads to.

So let us now look at a different interpretation of Pettit’s idea. We saw above that a literal analysis of Pettit’s condition of freedom suggested that freedom was restricted to the realm of the moral. This meant that only moral responsibility has been considered in acts of freedom: one is only free when performing an act that meets the “ought/can” distinction. Perhaps a more generous interpretation of Pettit, in which we consider not just the moral sphere of responsibility, but instead all spheres of responsibility, which includes the amoral capacity for action. On this reading of Pettit, we are no longer restricting talk of freedom to the realm of the moral, but are opening it up to include new aspects of responsibility to which we can apply a negative morality function, yet still be free. Consider the statement ‘I ought to practice the guitar’. Now that we have weakened the responsibility condition we can look at whether this is now a free action or not. Is this a responsible action? It is not clear from the phrasing of that question whether it is or not – the term ‘responsible’ in that context still makes it sound like a moral question, and thus doesn’t feel as though it’s the right kind of question you can ask about my playing the guitar… but that doesn’t mean it’s wrong, rephrasing the statement might help: ‘Steven is responsible for making the guitar make a noise’. While we have moved from passive to active tense, the statement shows us an aspect of responsibility that is not restricted to the realm of the moral.

The fact of denoting responsibility to me implies that I must have been free to do that action. This is language that we use in our everyday discourse. Statements such as “mum is responsible for the mess in the kitchen”, and “the referee was responsible for the football loss” all denote responsibility in an action performed by somebody; the mother and the referee are responsible, and we say that they are responsible because they are free.

So far so good – by weakening the condition for responsibility from the realm of the moral to include all spheres of responsibility, we have been able to give an account of freedom in actions both moral and amoral. So is this a satisfactory solution to the ambiguity in Pettit’s theory? By changing the responsibility condition we have been able to provide an account of responsibility in actions with a negative morality function. However, there is still a problem with looking at it this way. When we arrange responsibility statements in such a way as to avoid morality but still imply freedom, we leave ourselves open to ascribing freedom to a whole range of things, including inanimate objects. For example, statements with a negative morality function such as “the bull was responsible for destroying the china” and “the meteor is responsible for ruining this year’s corn harvest” can, under this theory, imply that the bull and the meteor are responsible for the damage that they caused. So it seems that we are able to ascribe freedom to both the bull and the meteor. This interpretation causes problems for Pettit, as he says that animals are not the sorts of thing that can be free:

We do not hold animals responsible for the thing they do, for we do not think that they make evaluations of their options such that we might hold them to the values involved. If freedom means fitness to be held responsible, then we are not going to credit them with the enjoyment of freedom… we will not think that freedom has any application to the mode of agency exercised by such animals.[9]

So there seems to be a contradiction in Pettit’s theory of freedom. In saying that anything that is responsible is by definition free, then it is possible to ascribe freedom to objects that we normally wouldn’t consider as free. The meteor is free because it is responsible for crashing into the cornfields; the bull is free because it is responsible for breaking all of the china.

Perhaps a way around this problem could be to say that there is a difference in responsibility between the bull and the referee. The bull being responsible for the destruction of the china is a causal claim, which has no bearing on moral responsibility in the strictest sense,[10] while the referee being responsible for the football outcome is a claim of responsibility, so there must be some other factor involved in her being held responsible. But are we entitled to say this; can we just make a claim that the bull isn’t really responsible for destroying the china – yes, it may have caused the damage, but can we blame the bull for his actions? I think the best way that Pettit can answer this criticism is by his introduction of the theory of freedom as discursive control.

When we describe persons as free… we say that in their agency as persons – in the agency allowed to them by their standing relative to others – they are fit to be held responsible; they do not act under pressure or duress or coercion or whatever… If one’s standing among persons is to make one free, therefore – if it is to leave one fit to be held responsible for thing – then it must be richer than [freedom as rational control]. It must rule out, not just the obstruction of choice, but also a myriad of other coercive and quasi-coercive ways in which people may intrude upon a person or inhibit them.[11]

Pettit here is outlining the requirements that he feels an adequate theory of freedom would need to fulfil. An individual will be free so long as they are able to be fit to be held responsible for thing that they do, so we need an account of freedom that will allow an individual to legitimately be held responsible for something; and this theory of freedom must be able to distinguish between ascribing freedom to the responsibility acts of bulls, and the responsibility acts of referees.

There is one sort of interaction, and one sort of influence, that paradigmatically meets the requirement expressed. That is the interaction that occurs when people attempt to resolve a common, discursive problem – to come to a common mind – by common, discursive means… Specifically, it refers to the sort of turn-taking involved in the attempt to resolve a problem by reference to what all parties regard as inferentially relevant considerations or reasons.[12]

It is here that Pettit first talks about freedom as discursive control. It is a theory of the free person, inasmuch as in order to count as free, the person must be free in relation to other people, not just in terms of rational or volitional control.[13] Persons must be discursively active in order to count as free, and thus fit to be held responsible:

An agent’s freedom as a person will naturally be identified, according to the line of thought we have been following, with the form of control that people enjoy within discourse-friendly relationships. An agent will be a free person so far as they have the ability to discourse and they have access to discourse that is provided within such relationships… For the discursive influence to which a person may be subject, consistently with retaining discursive control, will leave them fully fit to be held responsible for what they decide and do; it will be consistent with their counting as a fully free person.[14]

So here we start to see how Pettit might answer the problem of distinguishing between responsibility actions of bulls and meteors, and responsibility actions of mothers and referees. Bulls do not have the ability to discourse (with other bulls or any animal), so are not free, and thus they not fit to be held responsible. Referees and mothers have the ability to discourse, and so are fit to be held responsible for their actions; thus meaning that they are ultimately free. So mum and referees are discursively free, while bulls and meteors are not.

It seems that Pettit’s intention that freedom be the deciding capacity that makes the difference between moral actions and amoral actions is somewhat ambiguous; on a literal reading it appears that one is only free in cases of moral activity, though not in cases of action to which no moral ascription applies. The problem, however, is that on a more generous reading he still implies that discursive freedom is required for moral responsibility, as all of the situations that limit freedom are situations that fall under the sphere of moral actions, such as coercion and domination. It appears then that Pettit needs to take note of these ambiguities in order to provide a cogent argument for freedom. To do this, he will need to account for the two types of ‘ought’ statements, and incorporate them into his theory.

[1]. Pettit (2001 pp. 18-19 italics mine)

[2]. (2001 p. 73)

[3]. (Pettit 2001 p. 74)

[4]. (2001 p. 78)

[5]. (Pettit 2001 p. 78)

[6]. Pettit (2001 p.19)

[7]. Though this wouldn’t technically be true on utilitarian terms – if playing the guitar well maximises utility, you are then compelled to practice.

[8]. Pettit (2001 p. 19)

[9]. Pettit (2001 p. 24)

[10]. The bull isn’t held to the same values that we as humans are, thus can’t be held responsible in the same way

[11]. Pettit (2001 pp. 55-56)

[12]. Pettit (2001 p. 67)

[13]. Pettit (2001 p. 66)

[14]. Pettit (2001 p. 70)