The third prong of attack on contributory generalism involves asking for an appropriate epistemology. How are we to tell, from what we can discern case by case, that this feature will function in the same way wherever else it appears? Ross, our paradigm generalist, holds that we start with the recognition that this feature counts in favour here, but that we can immediately tell (by a process which he calls ‘intuitive induction’) that it must count in favour everywhere. The question is how this i