The main rival to the vulnerability model, its theoretical foil, is the voluntaristic notion of special responsibilities according to which, "special responsibilities derive their moral force from the fact that they have been voluntarily self‑assumed" (13). If voluntarism is the correct account of the origin and basis of special moral responsibilities, then a moral agent's voluntary and uncoerced consent is a necessary condition for the acquisition of such responsibilities, and, therefore, individual moral agents cannot have any special moral responsibilities without first giving their consent to undertake them. According to the voluntaristic model, it is the agent’s consent, rather than vulnerability and dependency of the moral patients to the moral agents that gives moral force to these special moral responsibilities.
The vulnerability model, on the other hand, assumes that at least some special responsibilities can arise without the agent's consent, and that consent even in cases where it is present does not explain the content nor the moral force of these obligations. As Goodin explains it, “If one party is in a position of particular vulnerability to or dependency on another, the other has strong responsibilities to protect the dependent party. These responsibilities both precede and constrain any bargain between the parties over what rights and duties they may voluntarily assume. Thus, it is vulnerability rather than some voluntary act of will which gives rise to special responsibilities of the most basic kind” (39).
The voluntaristic model assimilates special responsibilities to promises and contracts by assuming that the source of the moral obligations associated with these responsibilities is the agent's voluntary consent. Goodin argues that "it is wrong to suppose that all special responsibilities are necessarily self‑assumed" (30). The voluntaristic model does not adequately explain many types of special responsibilities which we commonly acknowledge in which the voluntary consent of the agent is neither a necessary nor a sufficient condition for the creation nor ascription of these special responsibilities.
As Goodin notes, by explaining all special responsibilities as deriving from promises and similar forms of self‑assumed obligations, the voluntaristic model gives both too strong and too weak an account of special responsibilities. It is too strong since: "[d]ischarging our special responsibilities, especially insofar as these responsibilities are seen to have been voluntarily self‑assumed, is ordinarily regarded as a matter of justice" (16). If I promise to do a thing for you, then I have established an obligation which entails a reciprocal right of the promisee to demand the fulfillment of my obligation. Promises, like other types of consensual obligations between moral agents, entail correlative rights which create moral claims or entitlements. As H. L.A. Hart has put it, "By promising to do or not to do something, we voluntarily incur obligations and create or confer rights on those to whom we promise; we alter the existing moral independence of the parties' freedom of choice in relation to some action and create of moral relationship between them" [quoted by Goodin 1985, 30‑31].
However many special responsibilities do not create correlative rights on the part of their beneficiaries; only those in which both the object and the addressee of the responsibility is a moral agent or person do so. Thus the voluntaristic model is too strong in this respect. On the vulnerability model, B can have a moral responsibility towards A without it being the case that A has a right against B, in cases where B is a moral patient for whose own sake one is acting, or where B is a moral agent who is the object or beneficiary of a responsibility, but not also its addressee. The vulnerability model predicts that moral agents can acquire special moral responsibilities even in cases where the moral patients concerned do not have the status of right-holders, and indeed, may not even be the sorts of things to which it is possible to ascribe rights.
The voluntaristic model, on the other hand, assumes that special obligations can only arise once the agent has volunteered to undertake them, and have given their consent to other moral agents to whom these responsibilities are owed. Since can only meaningfully give one’s consent to other moral agents, the voluntaristic model is too weak to explain our moral intuitions about many kinds of special moral responsibilities, for instance, those for the welfare of nonhuman animals. While I believe that Goodin is correct in denying Hart's thesis that all special duties arise from previous voluntary actions, it does not follow from this that none do. In fact, many special responsibilities contain voluntaristic aspects and well as other aspects, such as vulnerability and reciprocity which are needed to explain the precise character of these special moral relationships. It is important in analyzing real cases to distinguish between the specific content of the obligations that flow from the responsibility, for instance, what sort of benefits are to be bestowed, from the grounds for ascribing the responsibility to a particular bearer, on the one hand, and directing it towards particular objects or beneficiaries, on the other. Different ethical principles may be involved in each of these separate conditions: the vulnerability model, might account only for the direction of a particular substantive responsibility upon a certain class of beneficiaries, but not also for its being directed towards a specific individual member of that class, nor for its being ascribed to a particular member of the class of potential benefactors.
Thus, perhaps Goodin overstates his case by suggesting that the vulnerability model alone can provide a single, coherent account for all special responsibilities, any more than the voluntaristic model can provide such an account. Rather it may be that the precise character of special moral responsibilities in various contexts of moral action can only be accounted for by a "mixed" or pluralistic account of the origins and scope of these responsibilities, and that only one of the elements of such mixed accounts derives from considerations of vulnerability. In ethics, as in many other domains of inquiry into human action, there may be no single explanatory principle which will account for all of the relevant features of the data. But before we conclude that this is the case, we should examine some examples of special moral responsibilities in various domains and show how Goodin develops the vulnerability model for each of them, and how this approach helps to explain their characteristics.