Pursue attractive models, don't self-analyse

Aesthetics of scientific model choice

What criteria should the positive scientist consider when building a model, according to Comte in J.S. Mill's book?

1. respect the facts and relationships as currently observed

2. give as much satisfaction to the essential inclinations of our intelligence as we can afford collectively

Mill unpacks these "essential inclinations" to mean:

1. our instinctive predilection for order and harmony and classifying phenomena systematically (Levi Strauss makes a similar point about the essence of the human spirit)

2. our aesthetic feelings

So Comte is saying that model choice is also partly an aesthetic choice.  Mill doesn't like this one little bit.  But I think Comte is right.

Dislike of introspective psychology

Interestingly Comte is the father of behaviourism since he claims that you can't have a positive psychology; You can however have a study of moral and intellectual processes as a branch of physiology.

Hence he's the father of a materialist theory of mind.  He unfortunately wants to reuse the word phrenology here, while throwing away the pseudo-science of phrenology as it

existed in Comte's time, to describe what we would now call neuro-biology.  Despite the poor word choice, he again is proved right, though, again, Mill seems displeased.

This wonderfully expressed Mill-on-Comte critique foreshadows twentieth century psychology's great arguments between its famous schools - phenomenological, behaviourist,

cognitive behavioural, neuro-biological.