Two Kinds of Justifications, or Why Desert is not Really an Important Issue
Philosophers have long been debating whether a belief in determinism (metaphysical, not social) is compatible with the belief that humans have free will. There are three broad general views on this issue: you can consider that both beliefs are inconsistent and reject either of these, or you can argue for the so-called compatibilist view according to which determinism does imply the negation of free will. Which view is adopted has obvious implications for a range of moral issues, in particular those related to the attribution of responsibility and desert. This is the topic of a recent book co-authored by Daniel Dennett and Greg Caruso (see also this video for instance).
My objective here is not to speak about the book, which I have found interesting though the format (basically, a dialogue between the two philosophers) is not particularly well-adapted, as it leads to many repetitions and occurrences where both authors are talking past each other. What has interested me is the way Dennett – who defends a compatibilist position – rejects the idea of basic desert while in the meantime arguing that it still makes sense to say that someone deserves what is happening to her. Basic desert is the idea that individuals are responsible for their choices in such a way that they are metaphysically “entitled” to be positively or negatively retributed based on them. In particular, basic desert is backward-looking: the fact that you have done x with consequences y justifies that you receive (or are punished) z. It is relatively obvious that someone (like Caruso) who is a determinist and rejects free will must reject this idea as meaningless. Compatibilists like Dennett also reject it, because it is in tension with determinism. But compatibilists (contrary to determinists) are more likely to accept a different notion of desert as meaningful. On this notion (that I will write “desert*”), one deserves* to be punished or to receive a benefit for her acts conditional on the fact that this whole sequence (the act and the retribution) takes place within a justified system of institutions and rules that are constitutive of the way deservedness is ascribed within the related social practices.
This last sentence may be opaque, so let’s take a simple example. For a compatibilist, Lebron James doesn’t deserve the money he has won during his career and his related privileged social status because he cannot claim any responsibility for his natural talents, starting with his advantageous physical attributes. True, James has worked hard for his achievements – economists would say that his level of effort has been tremendously high. But James is not responsible for his “choice” of effort either because everything was already fully determined. And of course, there is a more mundane form of social determinism: James is just lucky to have been born in a society that gives so much importance to sports achievements and performances. Nonetheless, it can be said that James deserves* the benefits he has received because the latter follows from a set of rules that are constitutive of social practices in which everyone (or at least people living in the same society as him) are taking part. Within this set of rules, to deserve z (e.g., earning big money) is just to act in some ways (e.g., training hard and playing better than others) that lead to specific outcomes (e.g., winning titles). It is part of the definition of deservedness* within this system of rules. Of course, deservedness* not only depends on the existence of the appropriate set of rules. It is also conditional on the fact that this set of rules, and the related set of social practices, are themselves properly justified from the moral point of view. It could be argued for instance that the whole system of rules constitutive of free-market economies and thanks to which James has been able to convert his talents and performances into money and status is unfair. Or (less plausibly) it could be argued that the way the NBA organizes its competition and determines the winners is unfair. But the point is that deservedness* has meaning and the attribution of desert* is justified as long as the social practices corresponding to this notion are themselves justified.
Now consider a related example. Rawls (whose theory of justice famously also builds on the claim that natural talents and social opportunities are not deserved in the basic sense) has been insisting that principles of justice only apply to the basic structure of the society, i.e., the institutions that determine the allocation and distribution of primary goods. Individual behavior cannot be evaluated as “just” or “unjust” in the same sense and based on the same criteria. on Rawls’s account, individuals can act wrongly by not respecting the institutions that are justified by the principles of justice, but not unjustly. An important implication is that it is perfectly consistent in the Rawlsian system to accept an egalitarian principle of justice, such as the difference principle, while not exerting the required action in light of this principle (for instance, by exercising the required effort) if not properly incentivized to do so. This has been criticized by several philosophers, most prominently by Jerry Cohen, who regards this as an inconsistency: “if you’re an egalitarian, why are so rich?”, as Cohen famously asked.
But is this really inconsistent? There is no real inconsistency if we realize that there are two levels of justification as above.[1] Suppose that you accept Rawls’s principles of justice. Consider a basic structure S that “sufficiently” satisfies the two principles to be regarded as justified from the moral point of view. Within this basic structure, rules determine who gets what depending on outcomes, status, efforts, and so on. The benefits people are receiving by taking part in the social practices defined by these rules are justified according to S. Consider again the case of Lebron James. James is rich, far more than most of the persons on the planet, including his American fellows who are living in the same basic structure as him. James would not have achieved what he has achieved if he had not been incentivized as he has been (the counterfactual is admittedly impossible to falsify, but it seems reasonable). But – who knows – maybe James is egalitarian in its heart. On the Rawlsian account, there is no inconsistency: James did not act wrongly by keeping his money for himself because he has just acted in conformity with a system of rules S that, we assume, is justified by the principles of justice.
This idea that there are two different (or more) levels of justification has been actually articulated early by Rawls in one of his first papers, “Two Concepts of Rules”. In this article, Rawls (who had not arrived at the point of rejecting utilitarianism yet) argues that we should distinguish between the justification of a practice and the justification of an action taking place within a practice. The former (e.g., the practice of promise-keeping) can be justified for instance by the general benefits it brings to a society that adopts it, while the latter is justified as part of the practice itself, irrespective of its good or bad consequences.
This idea – which of course can be disputed – is helpful to clarify many confusing and confused debates about deservedness and more generally meritocracy. There are many studies that indicate that not only social origins, but just brute luck largely explain individual outcomes. That surely goes against the claim that basic desert has any normative relevance. That does not imply however that we should stop talking of desert*, or that egalitarian ideals are incompatible with an incentive-based market economy where some persons earn more than others. Whether Lebron James deserves* his money and his status or whether a rich businessman acts wrongly by not giving 10% of his income to efficient charities is not the real issue at stake. It is rather whether these behaviors take place as part of social practices that can be properly justified, in a Rawlsian or any other relevant perspective.
[1] I owe this idea to a paper by Chris Melenovsky, “Incentives, Conventionalism, and Constructivism”, published in Ethics in 2016.