Exceptive phrases formed with but can occur with universal quantifiers, as in every student but John came, where it is entailed that John did not come, and that every other student came. They cannot, however, occur with other quantifiers (e.g. *some student but John came) or with definites (*the students but John came). Building on ideas in von Fintel (1993), Gajewski (2008, 2013) proposes a general framework to capture these facts: (i) but denotes a type of subtraction; (ii) alternatives to the complement of but are activated; and (iii) these alternatives are “used up” by a higher strengthening operator. Given this framework, exceptives become a testing ground for two general questions: how are alternatives computed, and what is the inventory of strengthening operators? I pursue the hypothesis that the analysis of but-exceptives can be achieved using “unexceptional” machinery familiar from other domains: the complement of but is focus-marked and alternatives are computed after Fox & Katzir (2011); these alternatives are used up by an exhaustivity operator (Exh) from the literature on scalar implicatures (e.g. Fox 2007). I propose that the distributional restrictions of but derive from an interaction of felicity constraints restricting the distribution of Exh and the distribution of particular determiners. I further extend the analysis to provide a preliminary account for differences between but-exceptives and exceptives formed with other than.