These are the briefest notes on the main features of synthetic epistemology.

Some aspects of synthetic epistemology are introduced through the work of other philosophers.

This is the second stage in the evolution of my philosophy from a primarily positivist philosophy (and hence perhaps one still resting too heavily on the standards of the enlightenment) towards one in which some of the considerations which lead to Romanticism are given due weight.

My first approach to this was to put "existential positivism" alongside "metaphysical positivism" (without getting very far in saying what that might be), but it seems to me that a closer integration might be realised and that a good place to start with this is epistemology.

The philosophes one might say, were deluded about the reliability of "reason" as a guide to truth and understanding, particularly but not exclusively in relation to matters of value and purpose. A radical Romantic response might jettison the whole rationalistic enterprise, but what we seek here is to put reason in its proper place and enquire what other sources of knowledge we may have.

Apart from the stimulus from Romanticism, epistemology has always been a problem domain for me because of the conflict I perceive between what I consider to be epistemological bedrock (the distinction between logical, empirical and evaluative sentences, various foundationalisms) and widely received attitudes among analytic philosophers in the second half of the twentieth century. A key element in this recent consensus is the idea that, by and large, one can and should take language as it is for the purposes of philosophy. It has long seemed to me that one cannot coherently suppose that language has any definite meaning, but must chose either from diverse prior meanings or from equally diverse possible future meanings in articulating an understanding of any matter. One must I believe, and in any case I do, chose ones language carefully and that choice should be made with some purpose in mind. There is therefore a creative element in a philosophical (or other) theory. The recognition that my epistemological ideas depend upon a purposive synthesis of language and methods rather than articulate in a given language objective epistemological truths has lead to my use of the term "synthetic epistemology"

Socrates is famous for his method of eliciting knowledge by questioning, and the associated doctrine (anamnesis?) along the lines that we are born with the knowledge which we have temporarily forgotten, needing only some hints to recall.

I have myself a poor memory, particularly for detail. My experiences and my reading are rarely memorised in an explicit form, but are digested and discarded. They affect in greater or lesser degree my conception of the world, but the details of what it was which caused that transformation are likely to be lost. If I need to retain details it can be done, but otherwise the chances of retention are small, and the effort involved in memorisation can be very large.

The effect of this manner of assimilation is that I find myself making judgements without knowing the source or nature of the knowledge on which the judgement is based, it is as if the knowledge were innate. My own experience is not exceptional, even though my own memory (which is rather bad) may not be typical. Probably the major part of our cultural knowledge (including almost the whole of our working knowledge of language) we know without having any trace of memory of how we came by it.

The relevance of this here is in the following points:

• En passant, that there is no reason to doubt that we have some innate knowledge (of how if not of that)
• the much of our best and most reliable knowledge is formed by sound judgement on the basis of extensive experience and has no rational basis

Leibniz is of interest here because of his ideas about the automation of reason.

The automation of reason is today becoming a reality. Not only men, but also machines, may perhaps now be said to "know"

What are the scope and limits of the kind of deductive knowledge which Leibniz envisaged, and, beyond this sphere in which reason reigns, what model there prevails?

In synthetic epistemology particular attention is intended to an idealised model of rational knowledge. The ways in which this breaks down and how things work beyond its realm are of interest.

There are several aspects of Hume's philosophy which are relevant here.

Firstly, in general tenor he was sceptical and has been regarded as an eminent and early positivist (before that term was invented).

Secondly, the distinction between "relations of ideas" and "matters of fact", a precursor of the analytic/synthetic distinction, as well as that between those two and on the one hand judgements of value, and on the other the speculations of metaphysics, had a central place in his philosophy.

His scepticism lead him to consider only the truths of logic and mathematics as known with certainty, and his negative views about the status of causality and the justification of causal or inductive inference (inter alia) presented to him a problem in need of resolution. His resolution of this problem we may find unsatisfactory, for it consisted in showing not why these kinds of inference are justified notwithstanding that they fail to be logically sound, but in explaining why we will continue to make them irrespective of their logical status.

This resolution, unsatisfactory though we may find it, is an element consistent with his general conception of philosophical method, which was inspired by the work of Newton and modelled on the methods of empirical science. It is for this reason that Hume's major work is entitled "A Treatise on Human Nature", he conceived of his philosophy as differing in subject matter rather than in method from the work of Newton. I guess this is naturalism.

Synthetic epistemology is not naturalistic, but is intended to be a synthesis rooted in an understanding of "human nature".