On Compromise and Rotten Compromises (review)

According to Machiavelli, there are not one but two moralities: a private (or Christian) morality, prescribing how individuals are to deal with other individuals, and a public (or pagan) morality, dealing with the interaction of collectivities, such as nations and their rulers. As Isaiah Berlin famously argued, it was Machiavelli's great discovery that these two moralities are truly incommensurable—more specifically, that you cannot derive the latter from the former. The moralist statesman who tries to do so will render his country ill service.

Margalit's latest book must be situated against the background of this Machiavellian claim. He insists that the individualism of rational-choice theory and, more generally, of liberalism founders on the incommensurability of private and public morality, and on the irresolvable dilemmas with which politicians are typically confronted. Put differently, the world of politics knows of no ideal solutions acceptable to all parties involved: we can only strive for the best compromise possible. But contemporary anglophone political philosophy, taking private morality as its model, is a stranger to compromise. Margalit discerns (in my view, correctly) that this feature is its main weakness.

Still, Margalit distinguishes between "compromise" and "rotten compromise"—between compromises that private morality can endorse and those that go beyond it. He defines a rotten compromise as one with an "inhuman" regime, guilty of cruelty and of crimes against humanity. According to him, such compromises are not permissible. He proposes a number of useful distinctions when exploring the domain of (rotten) compromise. For example, he distinguishes between "politics as economics and politics as religion" and demonstrates that the former must remain blind to the true nature of compromise. Even more interesting is his distinction between "anemic and sanguine compromise," where the latter, unlike the former, requires that both parties "recognize" each other, meaning that each must be willing and capable of picturing what the world looks like when seen from the perspective of the other. Margalit convincingly demonstrates that recognition involves not only compromise but also the trajectory leading up to it.

In the second half of his book, Margalit turns to World War II for examples of rotten compromises. Consider the "Blood for Trucks" case—Eichmann's offer to the Allies of 1 million Jews for 10,000 trucks. The deal never came off, but it would have been a rotten compromise, since it was a compromise with the absolute evil of Nazi Germany. On the other hand, had the compromise been made 1 million Jews would have been rescued from a grisly death in the concentration camps. Or think of Yalta. Margalit wavers between calling Stalin's regime "inhuman" and merely "menacing"; but assuming it to be the former, the Yalta conference obviously produced a rotten compromise. He goes on to say that, then, "we may ask a different question: can 'political necessity' excuse rather than justify concluding a rotten compromise? . . . We did it. It is really bad. But there was no alternative." Both cases suggest that a simple ban on rotten compromises cannot be our last word on the subject. Margalit accepts the challenge, but he stops short of limiting the cases. For example, who could have blamed Churchill for concluding an alliance with Stalin against Germany in, say, January 1941, supposing an alliance to have been imperative at that moment for Britain's sheer survival? What else could he have done but opt for that rotten compromise?

Moreover, in warfare such situations are the rule rather than the exception, and so the ban on rotten compromises would, in practice, be lifted in wartime; rotten compromises would of necessity be "excusable" (in Margalit's terminology). Of course, there may be times when no rotten compromise is possible, simply because there are no "inhuman regimes" around. Think of eighteenth-century Europe. At such times, Margalit's ban on rotten compromises would be pointless. In sum, then, either violating the ban is excusable (ex post facto) or the dilemma does not present itself; and, if so, there is little rotten about rotten [End Page...