David Estlund is a leading political philosopher, and in Utopophobia: On the Limits (If Any) of Political Philosophy, he further cements himself as a crucial figure in ideal theory. His significant analytic skill is on display in this book: each argument (and there are many!) is carefully constructed and cleanly executed. This book stakes out clear territory for the value of ideal theory against the challenge brought by nonideal theorists. Estlund's goal is to reject “Utopophobia”: the idea that a conception of justice is defective if it sets such a high standard for compliance that it has essentially no chance of ever being met. He argues that such theories of justice could still be true.

While Estlund grants an “ought implies can” condition, in chapter 5, he dismisses what he calls “the human nature constraint.” This constraint posits that a normative political theory is false if it imposes standards that ignore human nature, that is, if it presents a theory that humans are motivationally unable to realize. He compellingly argues through two core thought experiments (Messy Bill and Professor Procrastinate, introduced at 28–29, though most fully developed in chapter 8) that lacking the motivation to do the right thing does not relieve one of one's moral obligations. He then argues for a social analog: if a society is required by justice to “build and comply” the relevant institutions to instantiate a just order, then we can reason about whether we are required to build based on whether we would comply after the institutions have been built. He nicely separates this proviso into concessive vs. nonconcessive requirements, and argues that nonconcessive arguments ought to have primacy.

Perhaps the most interesting idea in the book is that what Estlund calls “prime justice” involves both particular institutional arrangements and a list of behavioral obligations from the individuals in society. Estlund argues that since society is not a single agent, we need a resolution to the problem of plural obligation, which he calls the “plural requirement.” This is developed in most detail in part 4 of the book. Here he uses the thought experiment of Slice and Patch Go Golfing (33, 137–39, 167, 211–18, 220–21, 224–25, 227, 232, 234, 240–42, 251, 354–57) to set in motion intuitions about how a collection of individuals can have joint obligations even in the absence of individual obligations in the normal sense. He offers a rich account of the plural requirement, worthy of considerable study.

Lastly, and I think rather weakly, Estlund defends against “practicalism”—the idea that there is no value in understanding anything unless it has practical consequences. Here he argues that ideal theory is similar to mathematics. He adds that understanding true theories of ideal justice will also help orient our normative attitudes to better identify cases of justice and injustice, what he calls “informed concern” in the final chapter.

While I think that this is an impressive piece of philosophy, I am left thinking that it fails to get at the more interesting elements of the ideal/nonideal debate. Let us imagine that a philosopher-oracle determines the true nature of justice. We have uncovered the true principles. Even better, those principles uniquely identify a particular set of institutional arrangements and individual behaviors that would realize those principles if instantiated. If people fail to follow the dictates of justice, it could be because they are lazy, or worse, because they enjoy committing injustice. But perhaps more likely, they disagree. They think other values are more important to elevate, or believe that certain historical issues require redress, or contend that there is a better way of measuring the realization of the values implicated in the principles of justice. Failure to do one's duty could well stem from attention to other perceived duties. We could do more to enforce focus on the true principles of justice, but it is hard to do much in this direction and remain a liberal theory. Even if it were the case that we could identify the true theory of justice in some abstract mode of reasoning, it is impotent unless it has something to say about managing a diversity of sincerely held values, aims, and obligations that are ever-present in our formal and informal institutions. Examples like the Procrastinating Professor, Slice and Patch, and others sidestep these kinds of concerns by shaping our intuitions around the idea that failures of justice have to do with effort or our wish to avoid obligations. Indeed, Estlund's engagement of a concern of this sort appears in chapter 6, part B, though his focus is on competing motivations that might pull one away from social justice: his examples include favoring personal ties, rather than the more interesting challenge of disagreements about justice itself.

Political philosophy is more than moral normativity at scale. Motivated, sincere people who take their obligations seriously will still disagree, and how those disagreements are adjudicated has to be justifiable to all parties. While I am skeptical of the various candidates for principles of justice, I think that ideal theory is obviously valuable—just not in the way that Estlund proposes. Ideal theory is perfect for working out the relations between concepts and understanding the analytic consequences of particular commitments. This is a valuable contribution to nonideal theory, social critique, policy, and a worthy consideration for its own sake. Intellectual exploration is immensely valuable. But ideal theory can be made much more interesting if we take reasonable pluralism more seriously, making room for substantial diversity and exploring what kinds of principles and institutional realizations give people the room to be sincerely motivated by justice, but still to disagree.

Ideal theorists, including Estlund, often analogize their work to that of mathematics or sometimes theoretical work in the sciences (36, 203, 305, 310–11). Just as mathematicians do not need physical instantiations of their work for their proofs to be important, work in ideal theory does not need to look at real institutions to get at the truth about justice. But there is a quite serious disanalogy here: in pure math, the objects are agreed upon and have settled definitions. Proofs in pure math have their epistemic standing precisely because all of those methods and definitions are settled. But that is not the case in political philosophy. “Justice” is up for grabs, as is “reasonable,” “fair,” “reciprocal,” and a number of other crucial concepts. “Equal” might be well defined (thanks to mathematics), but what things should be equal is a core dispute. So, when we say something like “equal citizen,” even though those terms both have agreed-upon definitions, I very much doubt that there is agreement in the intension of that concept across political philosophers working today, let alone those in the past.

I do not raise such difficulties to say that we are out to sea, but I do think it means that we need to think more clearly about what ideal theory is doing. It is not the proofs of pure math. It is closer to work in string theory. There is a lot of elegant and interesting work there, and there is much we can learn from it, but we have no idea if it is even the right basic model. This is not to be defeatist, but to recognize that there is a great deal more work to be done.