ABSTRACT
What do educators normally do when they want to educate pupils in some subject, or discipline, or department of life'? Since 'education' implies not just any change (or even 'development') in the pupil's thought and action, but a change by learning, towards more understanding, truth, knowledge or rationality, they have first to agree on how-in general terms-pupils ought to think about the area. They take it, for instance, that the pupils who want to learn science should not look in crystal balls, or slavishly follow the authority of Aristotle, but use the procedure and reasons incorporated in (to use a brief phrase) 'scientific method'. In history, they are not encouraged to give reasons such as 'An angel did it' or 'Whatever the facts, it must have been the wicked Jews plotting', but rather reasons that emerge from a consideration of certain kinds of (historical) evidence: documents, private diaries, and so on. In mathematics, it is not the shape of the numbers which is to count, but their value. And so on.
