David Hume did not discover learning.

A while back I was reminded (by Donald Clark) of David’s Hume’s use of the term ‘impressions’ in his Enquiry Concerning Human Understanding, and of its possible similarity to my use of ‘affective context’ in describing the process by which we learn.

Whilst I think it is unlikely that I am the first person to describe learning in this way, after re-reading Hume I do not think he and I mean the same things:

In section 2 of his Enquiry, entitled ‘Of the Origin of Ideas’ he sets out to describe the difference between what he calls "impressions" and "thoughts or ideas". In summary, he concludes that both ideas and impressions are types of perception – but that our ideas are “less lively” than our impressions.

This is promising: almost all of the examples Hume gives are of an exaggeratedly affective nature, such as the following:

“A man in a fit of anger, is actuated in a very different manner from one who only thinks of that emotion. If you tell me, that any person is in love, I easily understand your meaning, and form a just conception of his situation; but never can mistake that conception for the real disorders and agitations of the passion. When we reflect on our past sentiments and affections, our thought is a faithful mirror, and copies its objects truly; but the colours which it employs are faint and dull, in comparison of those in which our original perceptions were clothed.”

It sounds as if he is saying – roughly – that ‘as we experience the world we have emotional reactions to it. I’m going to call those emotional reactions ‘impressions’. These reactions leave a trace in our memories similar to the initial reaction – which I’m going to call ‘ideas’.’

At this point it does sound as if Hume is describing something similar to the Affective Context model in which memory encodes experience according to our affective reaction to events.

What is unclear however is whether, when he says ‘perceptions’ he includes the affective reaction, or rather if he imagines that our perceptual experience can somehow be separated from our affective reactions and stored or compared without reactive/affective elements.

It is not until much later – part 2 of section 5 of the Enquiry – that it seems clear that he does not take this position. Describing the way in which ideas may be associated according to their resemblance he says:

“We may, therefore, observe, as the first experiment to our present purpose, that, upon the appearance of the picture of an absent friend, our idea of him is evidently enlivened by the resemblance, and that every passion, which that idea occasions, whether of joy or sorrow, acquires new force and vigour.”

In essence Hume is saying that a picture of our friend bears a resemblance to our idea of him – and that only then are our affective reactions triggered.

This is significantly different from the Affective Context account. He is saying that our impression can be compared to our idea, which in turn arouses a reaction. And it is here that I believe he is wrong: it would be more accurate to say that the picture gives rise to a set of reactions (an impression) which can then be compared to our stored set of reactions (our idea). On this explanation there is no need to add on the feelings that this comparison might evoke – since the feelings are themselves essential to the comparison process.

So I don’t think Hume discovered learning; in fact I think that in his writing it is possible to see that the outlines of the Cartesian corruption which today has led us so far astray. ‘I think, therefore I am’ is no more true than ‘I feel, therefore I am.’ – but it established a false primacy of reason that only now, with the help of neuroscience, can we begin to unravel.