On 2016-06-20 22:57, Martin Escardo wrote: > [...] but the intuitionistically valid principle that not-not(every > type has an inhabitant or not). Or rather "for every type, not not (this type has an inhabitant or not)". -- /NAD