cannot pose a threat to our well-being, or to the overall good, in the state of nature. It is a community’s very acceptance of a set of agreed-upon restrictions itself, and nothing else, which opens up the possibility of such cases. Obviously, then, the adoption of certain agreed-upon restrictions brings with it – due to the very nature of the community that will arise should these new rules be implemented – the requirement that certain exceptions be built into the structure of the restrictions. Thus, we see that any such account permits – indeed, requires – the possibility of principled exceptions to at least certain restrictions in order for the entire apparatus to be effective. And these exceptions are grounded in the same principle that grounded the restrictions in the first place – in this case, the bargainers’ self-interest-maximizing motive. Without these exceptions, as we noted above, the set of restrictions will serve merely as one elaborate mechanism to be exploited by, and serve the interests of, violators and free-riders.
The defeasibility of these restrictions will take two forms. First of all, we will permit sanctions for the violators, as a means of punishment for the sake of deterrence and social protection. This feature of contractarian accounts – that they allow for exceptions to the rules for the sake of punishing violators – is widely recognized. But furthermore, we must also build in reasonable exception clauses, allowing (for example) for proportional harm to attackers in self-defense. C. Minimizing the trade-off: the ‘principle of defeasibility’
So far, then, we have established that a certain feature of contractarianism – namely, the threat of free-riders that arises once a community adopts the contract – yields the necessity of allowing for certain principled exceptions to the restrictions that make up the contract – the kinds of exceptions required both in the case of sanctions for wrong- doers and in some cases of self-defense. But which exceptions can be allowed, and which cannot? How do we determine which exceptions, in other words, are ‘principled’? Do the bargainers need to enumerate, for each restriction on behavior they agree to adopt, a list of scenarios in which the restriction may be defeated or overridden, as part of the terms of the contract? This seems as burdensome as it is ad hoc. Instead, I want to argue that in addition to agreeing to whichever restrictions they happen to adopt, the bargainers need simply agree to one further principle that will generate all the exceptions they need, rather than specifying each individual acceptable exception. This principle might be