Summary of solutions to Laura Valentini’s paradox of ideal theory Author: Doctor Terence Rajivan Edward Draft version: 2 (25th September 2025) The paradox. Doctor Laura Valentini (now Professor) presents us with a paradox which starts with these three propositions: 1. Any sound theory of justice is action guiding. 2. Any sound theory of justice is ideal. 3. Any ideal theory fails to be action guiding. From (a) and (b), we can infer by conjunction introduction that any sound theory of justice is action guiding and is ideal. But from (c), this combination cannot be realized, because an ideal theory always fails to be action guiding. If you believe there can be a sound theory of justice, this is a paradox and you must reject at least one of the three propositions it seems. Solutions. The solutions below are not quite in chronological order of their appearance online. Most of them are mine, but first I present solutions from earlier literature. Note: I have engaged in some cut-and-paste from my own papers on Valentini. Note also: some other documents instead use (1), (2) and (3) to label the propositions giving rise to the paradox! G.A. Cohenite solution. Valentini anticipates a reaction from G.A. Cohen, but does not present it as a solution. There are two kinds of theories of social justice: those which aim to specify what a just society would be but without taking into account the extent to which the specification can actually ever be realized - they are fact-insensitive, in that they ignore facts which bear upon this question of extent (e.g. various facts of human nature); and those which aim to specify what a just society would be but while taking into account some or all such facts. G.A. Cohen recommends pursuing fact-insensitive theorizing about justice and so he would reject (a), Valentini thinks. There is no reason to think a sound theory of justice would be action guiding, because a sound theory is fact-insensitive and implementing it may require human beings (or other relevant beings) to be different from what they actually are. Valentini’s solution. Valentini implies that there are three rival criterions for when a theory counts as action-guiding, rejecting two. To the criterion that such theories must motivate actual agents, she says that it might be reasonable to expect compliance with the justice theory’s specifications, given our best account of human motivation, but this criterion wrongly expects us to motivate unreasonable agents too. To the criterion that such theories must be immediately applicable, she says that this is not rational, as any theory must be combined with actual information about the circumstances. She also points out how theories help select relevant information. The third criterion says that it must be possible to motivate agents (and not most unlikely?). Supporters of (c), as I understand them through Valentini (with some guesswork), say that an ideal theory will never meet this criterion. Valentini argues they can, thereby rejecting (c), giving as an example how John Rawls’s theory recommends equal basic liberties for adult citizens, thereby restricting legally acceptable institutions of the family: family law must not violate these liberty rights. (But here is a question: are the liberty rights for all adults or all sane adults? Can a society not roughly incorporate the legal restrictions Valentini is against through judging more people to be insane? Is the theory NOT actually action-guiding on family law, or to such a thin extent?) Lawford-Smith against (a). Lawford-Smith argues against (a) by arguing that “action-guiding” can be interpreted in two ways: but on one way (a) is trivially true when it is meant to be non-trivially true and on the other way (a) is false. The former way counts theories which are indirectly action guiding as action guiding and would count any theory of justice as action guiding, even fact-insensitive ones. A theory of justice provides a standard by which to assess real circumstances as just or unjust, which is action-guiding, because we can then take measures to move closer to ideal justice, as specified by the theory. (But why is (a) as a trivial truth a problem for someone who advocates the paradox? Because then they cannot endorse (c).) The interpretation of action-guiding which leaves (a) false is unclear to me at present, but Lawford-Smith says that the premise comes out false because a sound theory of justice has the task of telling us about the nature of justice, even if it fails to action-guide on this interpretation, and counts as sound if it does this task. Lawford-Smith against (b): for multiple ideals. Lawford-Smith observes that Valentini focuses on theories which are ideal in two senses: they specify an ideal of a just society that we should aim for and - the technical sense - they rely on false assumptions. Lawford-Smith argues that (b) depends on the assumption that there is a single ideal that we should be working towards, a theory of justice’s only being ideal if it specifies a single ideal to aim for. Instead there might be multiple ideals. We can specify possible world 1 as ideal with respect to freedom and possible world 2 as ideal with respect to equality but no world which is ideal for both and no way of saying one of the two is better. Nevertheless, our specifications can be action guiding, because a certain policy brings closer to both 1 and 2. But I suspect people will argue that this is the exception rather than the norm and even too rare to be an adequate solution. “A theory of justice which is not ideal, because rather than specifying a single ideal, specifies two ideals: it is action-guiding for 1 in 10 policies. That is not good enough as a solution.” Not sure. Lawford-Smith against (b): counterfactual truths. To repeat: Lawford-Smith observes that Valentini focuses on theories which are ideal in two senses: they specify an ideal of a just society that we should aim for and, the technical sense, they rely on false assumptions. Lawford-Smith argues that a theory of justice can be action-guiding without relying on false assumptions by specifying dispositions of ours: what we would do if not blinded by envy, etc. These are not false assumptions: they are true statements about ourselves, but ourselves as we would be in certain non-actual conditions. She proposes understanding Rawls and Ronald Dworkin as not actually relying on false assumptions. Not sure. Aren’t the decisions of ours that Rawls thinks we would make, if not blinded by envy etc., informed by social science, which features idealizing assumptions, in the sense of false assumptions? Economics uses false assumptions, such as that everyone has perfect information about what is in demand. The Moorean solution (from me!). The Moorean solution - if that is the right description - involves taking an ideal theory of justice, realizing it, and then saying, “Behold, we did it!” against the paradox. This solution does not identify a faulty proposition, but it is natural to conclude that (c) is false in light of the achievement. The cheap solution. Valentini writes, “First, a theory may be ‘ideal’ in a non-technical sense, insofar as it proposes an ideal of a fully just world towards which we should aim. Intuitively, it would seem that, unless a theory is ideal in this way, it cannot qualify as a normative theory of justice. Second, and more interestingly, a theory can be ideal in the technical sense of being designed under idealised, i.e., false, assumptions.” (2009: 337-338) But if she is working with the second sense, then (b) is contradictory, at least extending how we use “sound” in logic. (b) says that any sound theory of justice must be ideal, while in the second sense any ideal theory involves false assumptions. So (b) with the second sense of “ideal theory” entails that any sound theory of justice involves false assumptions. But a theory of justice that involves false assumptions cannot be sound. A sound theory, extending how we use the term “sound” in logic, must be entirely true. A simple-minded solution. Why do ideal theories in the relevant sense involve false assumptions? Otherwise they would be too complicated to be of use in guiding action. Once we appreciate this reason, we reject (c): that any ideal theory fails to be action guiding. A dilemma solution. What should we say about a theory which specifies an ideal and argues for it using idealizing assumptions, but also includes some account of how you apply the theory to less ideal reality? EITHER you do not count this as an ideal theory, in virtue of its how-to-apply account, in which case (b) is false. A “sound” theory of justice does not have to be ideal. OR you count it as ideal, in which case (c) is false: an ideal theory can be action guiding. The aesthete’s solution. I imagine a University of Oxford aesthete saying, “A good ideal theory has intrinsic value, even if it cannot guide action, owing to being too unrealistic. It is a utopian fantasy. The value is aesthetic in a broad sense. Imagine some writers depict utopian societies. Are the products not valuable in themselves?” (I presume the aesthete would get rid of (a) then.) The action-guiding contradiction solution. A rather strange attempt at a solution is to give up (c) as contradictory. If any ideal theory fails to be action guiding, owing to its being something that we cannot implement, then that is action guiding. We won’t try to implement an ideal theory or we won’t once we know that it is an ideal theory and that ideal theories fail to be action guiding. It does not guide us to do something but it guides us against doing something. From its being ideal, we can now infer that it is not doable. References Cohen, G.A. 2008. Rescuing Justice and Equality. Cambridge, Massachusetts: Harvard University Press. Edward, Terence Rajivan. 2022. A Moorean solution to Laura Valentini’s ideal theory paradox? Available at PhilPapers. https://philpapers.org/archive/EDWAMS-3.pdf Edward, Terence Rajivan. 2022. A cheap solution to Laura Valentini’s ideal theory paradox? Formerly available at PhilPapers (now missing, not at my request). https://philarchive.org/rec/EDWACS-2 Edward, Terence Rajivan. 2022. A simple-minded solution to Laura Valentini’s ideal theory paradox. Available at PhilPapers. https://philpapers.org/archive/EDWASS-2.pdf Edward, Terence Rajivan. 2022. An Oxford aesthete’s solution to Laura Valentini’s ideal theory paradox. Available at academia.edu at: https://www.academia.edu/93320014/An_Oxford_aesthetes_solution_to_Laura_Valentinis_ideal_theory_paradox Edward, Terence Rajivan. 2022. Laura Valentini’s ideal theory paradox and the action-guiding requirement. Available at academia.edu at: https://www.academia.edu/93560349/Laura_Valentinis_ideal_theory_paradox_and_the_action_guiding_requirement Lawford-Smith, Holly. 2009. Debate: Ideal Theory—A Reply to Valentini. The Journal of Political Philosophy 18 (3): 357-368. politicalnotmetaphysical. 2018. Laura Valentini, “On the Apparent Paradox of Ideal Theory.” Available at: https://politicalnotmetaphysical.wordpress.com/2018/01/16/laura-valentini-on-the-apparent-paradox-of-ideal-theory/ politicalnotmetaphysical. 2018. Holly Lawford-Smith–“Ideal Theory–A Reply to Valentini”. Available at: https://politicalnotmetaphysical.wordpress.com/2018/01/16/ideal-theory-a-reply-to-valentini/ Rawls, John. 1999 (revised edition). A Theory of Justice. Cambridge, Massachusetts: Belkap Press. Valentini, Laura. 2009. On the Apparent Paradox of Ideal Theory. The Journal of Political Philosophy 17 (3): 332-355.