Full Idea
An empirical coherence theory needs, for the beliefs of a cognitive system to be even candidates for empirical justification, that the system must contain laws attributing a high degree of reliability to a variety of spontaneous cognitive beliefs.
Gist of Idea
Empirical coherence must attribute reliability to spontaneous experience
Source
Laurence Bonjour (The Structure of Empirical Knowledge [1985], 7.1)
Book Reference
Bonjour,Laurence: 'The Structure of Empirical Knowledge' [Harvard 1985], p.141
A Reaction
Wanting such a 'law' seems optimistic, and not in the spirit of true coherentism, which can individually evaluate each experiential belief. I'm not sure Bonjour's Observation Requirement is needed, since it is incoherent to neglect observations.