Goldman and reliabilism

Reliabilism is the main twentieth century rival to the justified true belief analysis of knowledge. It has two main points of contrast with the traditional version of the justified true belief analysis. First, rather than talking of justification, the third aspect of the analysis of knowledge is that the true beliefs are arrived at by a reliable process. (One feature of this is that 'reliable process' seems genuinely understandable independently of a prior grasp of knowledge. Thus it promises to shed light on knowledge from a reductionist standpoint.) Second, whether a belief acquisition process is a reliable process is an objective matter. So it is a third person rather than a first person matter. (I said that this was a contrast with a traditional version of the justified true belief analysis because whilst, traditionally, justification is taken to be justification for the knower, it need not be. It might be an objective matter. See McDowell.) So if reliabilism is true, one does not need to know that one knows in order to know.But what is meant by 'reliable process'? How reliable? If less than 100%, does this not re-introduce luck into knowledge? And can the third element of knowledge really be wholly external to the subject? Need they take no responsibility at all for their knowledge?

Reading

- Goldman, A. I. (2010) 'What is justified belief?' in Sosa et al (eds) Epistemology: an anthology, Oxford Blackwell or in Pojman, L. (ed) (1992) The Theory of Knowledge, Belmont, California: Wadsworth which is here.

Goldman, A. I. (1976). “Discrimination and Perceptual Knowledge,” Journal of Philosophy, 73: 771–791 here (Read the first and second sections as far as p778)