Reply to Danny Frederick’s Review Essay of Nozick’s Libertarian Project (“NLP”), in Reason Papers Vol. 36, no. 1 (henceforth “RP”)
I wish to start by thanking Danny Frederick for investing his time in reading and carefully critiquing my book. If all critics were as meticulous and fair as Danny has been to me, there would be far more constructive engagement between philosophers, and much less of theorists fruitlessly talking past each other. I concentrate below on which I regard as the key points raised in his Review Essay.
In NLP (20-29), I reconstruct as a deductive proof what I take to be Nozick’s argument for libertarian rights in Anarchy, State, and Utopia (see especially ASU, 29-34 and 48-51). This consists of 5 premises, leading to the conclusion that, “The use of force or coercion against innocent persons (those not engaged in aggression or fraud against other persons) interferes with their rational agency and is therefore morally impermissible.” I note in the introduction that despite the fact that Nozick does not present his ideas in this fashion, I have elected to do so because it “makes his premises explicit and allows us to assess the overall strength of his reasoning.” (NLP, 5). Subsequently, when I actually detail Nozick’s argument, I caution further that “Philosophy is not a branch of mathematics. These premises are certainly controversial, and cannot be conclusively demonstrated.” (NLP, 29).
I am also quite open about my view that Nozick’s argument is built on a foundation of certain widely shared and deeply felt moral intuitions. In this vein, I quote Bertrand Russell’s observation that ethical reasoning starts with “the kind of proposition of which proof is impossible, because it so simple or so obvious that nothing more fundamental can be found from which to deduce it.” (see NLP, 21).
Danny’s critique of my (Nozick’s) argument for libertarian rights focuses on the transition from Premise 4 (“Persons are inviolable because they are rational agents”) to Premise 5 (“Persons have a right to exercise their rational agency without interference, subject only to the equal rights of other rational agents”). I offer two reasons why an agent who accepts the truth of (4) should also accept (5). The first draws on the notion that, “Because the special moral status of rational agents is rooted in their autonomy, appropriate deference [to this characteristic] requires that they be permitted to live the life they choose, so long as they do not infringe the equal rights of others” (NLP, 27).
The second reason draws on the appeal of “moral impartiality,” meaning that there are no special privileges or exemptions that apply to some persons, but not others. So, those who accept the inviolability of persons generally, “may not, without contravening a basic moral principle, demand respect for their own autonomy while denying equal respect to other persons” (NLP, 28). I further explain that while many egalitarians may purport to accept Premise 4 while nonetheless rejecting 5, they do so on the basis of faulty logic. For example, they draw an indefensible distinction between the stringency of property and other sorts of rights. (Ibid). Accordingly, although they may not realize this, if they assent to Premise 4, they cannot consistently reject Nozick’s argument for libertarian rights.
Danny is satisfied with neither justification. He objects to the first on the grounds that Premise 4 does not entail 5 because “Friedman’s argument for (4) showed only that rational agency and overridable middling side-constraints are correlated. We are looking to the second part of his argument to provide the because.” (RP, 137; his emphasis). Danny is right that (4) does not logically imply (5) in some strict sense, but I nevertheless believe that (4) provides a reason for accepting (5), at least as far as “reason” is usually understood in moral discourse.
Suppose “Joe” is a classical utilitarian who believes accordingly that the wanton infliction of pain on sentient creatures is wrong. I happen to encounter him standing idly by while his small child continuously cries out in intense pain, which he could be immediately alleviated by a trivial exertion. What better reason could I give Joe for acting, other than pointing out to him that his inaction violates what he acknowledges to be a valid moral principle?
Analogously, I think that a person who accepts (4) is committed to respect the exercise of rational agency in the same way that the classical utilitarian is bound to promote pleasure and minimize pain. If an agent accepts that other persons have moral status only because they are rational agents, then at least is she wishes to act rightly, the appropriate response to this fact is to not gratuitously impede the exercise of this attribute. If she lacks good will, then I am afraid no argument will suffice.
As noted, my second argument for the claim that an agent holding Premise 4 should also accept Premise 5 rests on the notion of moral impartiality. Here, Danny asks first about those who reject (4), “Friedman is supposed to be explaining why rational agency grounds side-constraints; he must address his arguments to those who doubt or deny (4), not just to those who already accept it.” (RP, 137; his emphasis). I think he is demanding too much of me.
Premise 4 is built-up from previous premises, all of which ultimately rest on Kantian notions of respect for persons, and particularly his idea that they may never be used simply as a means of accomplishing objectives not of their own choosing. Committed utilitarians and those enamored with “social justice” are in the grip of other, inconsistent intuitions, and will thus never accept (4) or (5).
Danny further contends that my appeal to moral impartiality does not give even those egalitarians who claim to accept (4) an adequate reason to endorse (5). (Ibid). In response to my argument that (as I purport to show in Chapters 2 and 3) egalitarians will not be able to draw a principled distinction between acceptable and unacceptable redistributions, he observes that I draw such lines in Chapter 6, so my argument “falls apart.” But, the redistributions I endorse there are consistent with (4), while those proposed by our egalitarian friends are not.
For instance, I argue in Chapter 6 that if no other means are available, coercive taxation to support the innocent poor can be justified under either Nozick’s Lockean proviso or by the demands of moral pluralism. The latter approach is consistent with (4) because the “inviolability” referenced there is not absolute. Conversely, I don’t believe that an egalitarian can commit to even a defeasible notion of inviolability, while at the same time endorsing massive social engineering projects that do not target the truly needy, and fail to discriminate between the blameless and the irresponsible.
This brings me to Danny’s objection that my reconstruction of Nozick’s argument for rational agency does not produce a version of side-constraints that accurately encompasses or defines libertarian rights. (RP, 138). He notes that respect for rational agency is not coextensive with a prohibition on using persons solely as a means [which Nozick identifies as proposition “p” at ASU, 34, i.e. “a strong statement of the distinctness of individuals”], and argues with various examples that “focusing simply on exercises of rational agency, autonomy, or free choice will not get us to (p) or to side-constraints that mark the bounds of permissibility.” (Ibid.). Because Danny contends that p is superior to my formulation of the partial libertarian side-constraint [1], he offers a friendly amendment along critical rationalist lines. (see RP, 139).
The issues Danny raises are interesting and complex, and they cannot receive here the detailed treatment they deserve. Clearly, reasonable libertarians can disagree about this, but to my eyes the “never simply as a means” and the “separateness of persons” formulations of the libertarian constraint, like Rand’s NAP, operate at too high a level of abstraction to function as adequate guides to action. I suspect that this concern is what prompts Nozick to ask “in virtue of precisely what characteristics of persons are there moral constraints on how they may treat each other or be treated?” (ASU, 48), which leads him to offer the argument I describe in NLP. In any case, I am not convinced that what I take to be Nozick’s statement of the libertarian constraint fails to competently resolve any of the examples Danny presents, bearing in mind the interpretation given in Chapter 6.
With respect to Danny’s suggested critical rationalist alternative (RP, 139-40), I would just say that I question whether Nozick is committed to the idea that it is exclusively through “pure reason” that “persons discover who they are and then live their lives accordingly.” I fail to see why, for Nozick, experience should not play a vital role. In fact, although Danny is proposing a very different meta-ethics, the “framework for utopia” described in Part III of ASU seems very much in the spirit of Danny’s proposal, including Nozick’s concern that his framework protect people’s right of exit from communities that no longer meet their needs. (see ASU, 307-8).
Finally, this brings me to my effort to improve upon Nozick’s defense of the coercion employed by the minimal state in the provision of national defense and domestic security (see NLP, 89-100). With respect to the former, I argue that a relatively just and peaceful state threatened with foreign aggression may permissibly compel all citizens to pay their fair share of taxes for military defense. Since all rational agents in such a polity benefit from this public good, would-be free riders commit what I call a “passive form of aggression.” (Ibid., 95). Accordingly, it is not that their right to these resources is overridden by the necessity of preserving rational agency; rather, the state may prevent them from culpably refusing to pay. I believe that Danny does not object to this argument.
With respect to domestic security, as Danny rightly notes, the issue is different. Very briefly, my argument here is that the risk to the exercise of our rational agency comes not from potential free riders, but from the possibility that too many individuals and private protective agency would insist on enforcing their own conceptions of justice, by means of their own preferred legal rules and procedures, leading to widespread, violent chaos. Thus, I claim that in (and only in) states “that operate substantially in accordance with F. A. Hayek’s construal of the rule of law” the government is entitled to “exclude private PAs and independents from the unauthorized administration of justice.” (NLP, 96).
In such communities all citizens are required to forego their personal conception of rights and their preferred procedures, but they do so as members of a state whose role is strictly limited to promulgating and enforcing only abstract and neutral rules that are consistent with Hayek’s political ideal. Moreover, in such a polity, it is possible to reform laws that are shown to unduly limit freedom. Here again, the argument is not that the individualist anarchist’s right to enforce justice is overridden by the demands of rational agency, but that no such right exists.
Danny questions why I need to invoke the idea that citizens in a society governed by the rule of law have “collectively forgone” their individual notions of justice. Why not simply rest my argument on the paramount value of rational agency? As Danny puts it, “If side constraints do not permit an exception for state monopoly provision of internal security, they will not fulfill their function of securing our rational agency.” (RP, 142).
As I hope is clear by now, I agree with Danny’s point, but I am afraid that, as stated, it concedes more than I would like. I do not wish to say that the Hayekian minimal state is engaging in objectionable coercion that is nonetheless acceptable as the price of preserving our rational agency. Rather, my idea is that the individual anarchist would be acting wrongly if he insists on enforcing his own rights. Just as financing the military defense of a relatively peaceful state requires all citizens to pay their fair share of taxes, the rule of law requires all members to accept, subject to reform efforts, the state’s monopoly of law enforcement. A contrary judgment would cede to the independent a special privilege surrendered by his fellow citizens. (see NLP, 98-9).
[1] “Partial” because it does not cover paternalistic aggression. See ASU, 34.
Hi Mark,
You want to argue from
(4) persons are inviolable because they are rational agents
to
(5) persons have a right to exercise their rational agency without interference, subject only to the equal rights of other rational agents.
One problem is that the vagueness of the word ‘inviolable’ leaves it unclear what (4) means. Of course, it means that there are side-constraints on how persons should be treated; but which side constraints and how strong are they? There is a spectrum of collectivist views which could claim to acknowledge that persons are inviolable because they leave in place some, more or less weak, side-constraints.
Some people of a libertarian leaning may interpret ‘persons are inviolable’ to mean the same as (5), in which case (5) trivially follows from (4). Indeed, the nub of your argument appears to be the claim that possession of rational agency explains (5).
But why should possession of rational agency engender a right to exercise rational agency which is limited only by other people’s rights to exercise their rational agency? I possess hands. Does that give me a right to use my hands? Even if we assume that it does, it does not follow that my right to use my hands is limited only by the equal rights of other people to use their hands. If it did, I would be entitled to smash up someone’s house while he is away, so long as no-one else is around, because there is no exercise of anyone else’s hands that could interfere with what I am doing.
Further, what does it mean to exercise one’s rational agency? The dominant conception of rational agency involves instrumental rationality, which means using something as a means to an end. Do I have the right to use things as means to my ends subject only to the equal right of others to do the same? No, I do not: I normally have no right to use things as means to my ends if other people have property rights in those things. My exercise of my instrumental rationality is limited, not only by other people’s rights to exercise their instrumental rationality, but also by other people’s property rights.
You and I are broadly in agreement about what moral rights persons have and the importance of extensive private property rights for freedom, prosperity and a moral order. We also agree that the moral status of persons, which makes rights and freedom important, is inseparable from rational agency. Where I think you err is in trying to obtain substantive conclusions about the rights that people have simply from the fact that they possess rational agency.
In contrast, I try to account for moral rights (including property rights) as conditions for the flourishing of persons. The account draws upon:
• facts about persons, particularly their fallibility and their capacity for critical as well as instrumental rationality;
• facts about the world, particularly the fact that resources are scarce, which means that flourishing depends upon efficiency;
• facts about economy, particularly the facts that extensive and secure private property rights are essential for prosperity, for progress and for the critical experimentation that is essential for self-discovery.
Your explanation, in NLP, of the legitimacy of a minimal state which provides national defence is, I think, along the right lines: I would only want to tweak it here and there. What I could not understand, as I said in my Review Essay, was why you did not provide a similar explanation of the legitimacy of government provision of internal security, instead of offering an obscure discussion about ‘collective foregoing.’ You say in the last paragraph of your Response to my Review Essay: ‘I agree with Danny’s point, but I am afraid that, as stated, it concedes more than I would like.’ You then go on to say why; but what you say seems only to repeat what I said, so I cannot see what I am supposed to be unacceptably conceding. I suspect that it is the idea of ‘collective foregoing’ that is causing the trouble. We agree that a minimal state’s provision of internal security, financed by equitable taxation, does not violate any rights of the citizens. But you then express that point in the following terms: ‘the rule of law requires all members to accept, subject to reform efforts, the state’s monopoly of law enforcement.’ That may sound like a ‘collective foregoing.’ But it is false. There is no connection between
X does not violate any rights of the citizens
and
all the citizens accept X (either individually or collectively).
A proposition can be true even though none of the citizens accepts it; and a proposition that all the citizens accept may be false.
As an aside, I would like to make a couple of more or less logical points. First, Bertrand Russell (one of the very few great philosophers of the twentieth century) set out on a passionate quest for certainty. He thought that knowledge should be organised as a deductive system with indubitable premises and impeccable rules of inference from which theorems could be proved. He gave voice to that conception of knowledge in ‘The Elements of Ethics,’ which you quote. He thought that mathematics was the branch of knowledge in which most progress had been made, but he also thought that many accepted mathematical ‘proofs’ were abysmal. He set out to derive the whole of mathematics rigorously from simple and self-evident logical axioms. He then made what he called an ‘unpleasant’ discovery: ‘It appeared that, from premises which all logicians of no matter what school had accepted ever since the time of Aristotle, contradictions could be deduced’ (My Philosophical Development, Unwin, 1959, p. 58). In order to remove the contradictions, he had to introduce axioms which were highly complex and counterintuitive, leading him to say: ‘if we are to believe in the truth of pure mathematics, it cannot be solely because we believe in the truth of the set of premises. Some of the premises are much less obvious than some of their consequences, and are believed chiefly because of their consequences’ (‘Logical Atomism,’ 1924, reprinted in his Logic and Knowledge, ed. R. Marsh, George Allen and Unwin, 1956, p. 325). Thus, the old Euclidean idea of a body of knowledge being deduced irrefragably from self-evident axioms was jettisoned in favour of the idea of knowledge consisting of hypotheses which are accepted tentatively and which are evaluated in terms of their consequences (hypothetico-deductivism). Unfortunately, most contemporary philosophers seem ignorant of this development, as they continue to talk in a Euclidean way. It is why I talk, for example, of explaining the legitimacy of the state (from hypothetical premises which are held open to criticism and possible improvement) rather than arguing for or justifying the legitimacy of the state.
Second, you say (above): ‘Committed utilitarians and those enamored with “social justice” are in the grip of other, inconsistent intuitions, and will thus never accept (4) or (5).’ Some people might construe that as saying that there can be no rational assessment of competing ethical views: one simply has to make a choice. That, however, is a mistaken view. We can (and do) criticise utilitarian and ‘social justice’ theories by pointing out their unwelcome consequences. What I think you are saying, though, is that we should not expect such criticism to prompt the people who hold such theories to cease to hold them. I think that is too strong. Some people do respond to rational criticism by giving up the criticised theory. Some attempt to criticise the criticism or to replace their old theory with an improved version which surmounts the criticism. Those are rationally commendable moves. But some people either just ignore the criticism or attempt to deflect it with ad hoc manoeuvres, such as redefining terms in a way which reduces the content of their theory or trying to deflect the criticism by introducing hypotheses which are untestable, and so on. Those people can sometimes be brought to recognise their folly if we criticise their ad hoc manoeuvres, and they might then be led to give up their view. But you are quite right that, if people are committed to a view come what may, then rational criticism will be powerless to shift them. There is no such thing as a compelling argument.
Best wishes,
Danny
Thanks Danny for the additional thoughts and argument. I’m content to leave the discussion where it stands. I believe anyone who has read (i) the relevant portions of ASU and (ii) my book, (iii) your critique, (iv) my response, and (v) your reply, should be able to reach their own conclusions regarding these questions.
Pingback: Friedman and Frederick on Nozick; Sadler on Plato | Policy of Truth