The reasonability threshold in feasibility as ‘conditional probability’: A defense

The concept of feasibility is located at the junction between facts and norms, or between empirical political science and normative political theory. Since ‘ought’ implies ‘can’, it is commonly agreed that our political obligations are limited to acts and outcomes that fall within the feasible realm (Gilabert and Lawford-Smith 2012). An influential understanding of feasibility is the so-called conditional probability-account, which defines feasibility as “reasonable probability of success, conditional upon trying” (Brennan and Southwood 2007). In recent years, the conditional of the conditional probability-account has been challenged by authors who offer competing views of what trying involves (Wiens 2016, Estlund 2016), or point to cases where agents seem unable to try in the first place (though had they tried, they would most likely have succeeded) (Stemplowska 2016, 2017). Considerably less attention has been granted to the reasonability threshold, which states that outcomes are feasible only if success is reasonably probable to follow from our attempts. My paper highlights a dilemma that the probability threshold raises for defenders of the conditional probability-account. The problem is that what degree of probability of success is reasonable seems to depend on the risks and gains involved in specific cases. If the alternatives to attempting an act are truly disastrous, even a very low probability of success appears to suffice. A 1/10 chance of surviving a jump from the 5th floor may render jumping reasonable if the only alternative is to be caught by a fire raging the building. But given that the chances of surviving the jump fall well below the threshold set by the reasonability criterion under normal circumstances (where there is no fire but, say, a chance to catch the last bus home if we skip the stairs), this either suggests that we should, in cases like these, undertake an act that is unfeasible for us, or that what is feasible or not will depend on the available alternatives. This apparent dilemma has led some thinkers to abandon the reasonability threshold, opting instead for a scalar or comparative account of feasibility (Gilabert and Lawford-Smith 2012), or to replace conditional probability by a different understanding (Southwood 2023). Contrary to those responses, my paper will defend the reasonability threshold. I will argue a) that feasibility, to fulfil its intended role in normative political theory, should be understood in threshold rather than scalar terms, and b) that reasonable probability of success is the most plausible understanding of such a threshold. In the paper, I will defend this view against the objections from counterintuitive implications outlined above, pointing to different ways to escape the apparent dilemma. Further, I will argue that accepting the reasonability threshold does not render feasibility dependent on normative considerations in a way that produces a “moralized” understanding of this concept.