Full Idea
An unmodified reliabilism does not accommodate defeaters, and surely there can be defeaters against reliably produced beliefs?
Gist of Idea
Simple reliabilism can't cope with defeaters of reliably produced beliefs
Source
Thomas Grundmann (Defeasibility Theory [2011], 'Defeaters')
Book Reference
'Routledge Companion to Epistemology', ed/tr. Bernecker,S/Pritchard,D [Routledge 2014], p.160
A Reaction
[He cites Bonjour 1980] Reliabilism has plenty of problems anyway, since a generally reliable process can obviously occasionally produce a bad result. 20:20 vision is not perfect vision. Internalist seem to like defeaters.