possible. And it is precisely this desire to minimize such a trade-off that makes contractarianism such a promising means of justifying the principled exceptions to constraints on harming others required by some cases of self-defense. For it seems obvious that if a set of bargainers agrees to certain constraints on behavior because, for example, adherence to those constraints is expedient in fulfilling their desire to maximize overall well-being, and if the bargainers also can reasonably foresee that on a few particular occasions, their very adherence to those constraints should happen to threaten or undermine overall well-being, then it seems an exception to those constraints would be warranted on those particular occasions. Furthermore, as we have just seen, this exception is grounded in the very principle that grounded the constraints in the first place
namely, the bargainers’ desire to maximize overall well-being. Any constraints agreed
to by a set of bargainers, it seems, will thus be qualified in such a way as to allow for certain principled exceptions. This attribute – call it ‘principled defeasibility’ – would therefore seem to be a necessary feature of any successful contractarian account; no wonder, then, that contractarianism seems a promising route to take when trying to meet Kagan’s challenges. And certain cases of self-defense certainly provide a good illustration of this: we would not expect our bargainers to remain constrained to adhere to restrictions against harming others, even if such measures might be their only means of protecting themselves from the attacks of people who are breaking these very restrictions. Such an exception would be warranted as a result of the bargainers’ desire to minimize, in this case, the trade-off posed by the threat of free-riders. In this respect, then, contractarianism seems to be the best choice among the available alternatives; neither Kantian theories, virtue theories, nor divine-command theories seem to offer as plausible a justification of principled exceptions to constraints as contractarianism can.11 I will have more to say about the precise way in which my contractarian account will ground constraints and their principled exceptions in a common source in chapter three below.
11 Act-consequentialist theories, of course, offer an easy means of justifying exceptions to secondary rules, or ‘rules-of-thumb’, but of course these are not constraints in the relevant sense of the term. At any rate, consequentialist theories are of necessity extremist theories – namely, extremist theories that do not recognize any constraints on the pro tanto reason to promote the greater overall good. Hence the specification that contractarianism is the best candidate, among the available alternatives, for the task of providing a moderate morality that will meet Kagan’s two specific challenges and hold up against his general charge that options cannot be defended.