3 Possibility modals: Which conditions make them possible?

Abstract

English has several modals that can express possibility: the periphrastic expression be able to (a 'quasi-modal' or 'semi-modal'), the modal auxiliaries can and could, and the modal auxiliaries may and might. On what basis do language users make a choice between them? There are arguably multiple factors involved. Some obvious ones are (morpho-)syntactic in nature. For instance, the members making up this set of options are not equally likely candidates to be hosts of a contracted negator: can't, couldn't, isn't/aren't/weren't able to are much more likely than mayn't, which is hardly ever used Pullum et al. 2002: 1611), or mightn't, which is less rare but still very uncommon (Palmer 1987: 17-18; Huddleston, Pullum and Reynolds 2021: 52). Looking at the left co-text, be able to is the only modal option to be encountered after the infinitive marker to and, at least in standard English, it is the only option to be used after an auxiliary: we can say, I think he'll be able to come tomorrow, but not I think { *he'll can / *he'll could / *he'll may / *he'll might} come tomorrow (