The figure [neo-roman writers] wish to hold out for our admiration is described again and again. He is plain and plain-hearted; he is upright and ful… - Quentin Skinner

If this is nothing more than a stipulation about how we ought to use the term ‘ideological’, then perhaps it will do no harm. But if it is a proposal about how historians ought to set about the business of explaining beliefs, then it seems to me fatal for just the reasons I have sought to give. It refuses to recognise that one of the reasons why someone may hold a certain belief is that there is good evidence in favour of it, that it fits well with their other beliefs, and so on – in a word, that it is rational for them to hold it. If we refuse to speak in these terms, we deprive ourselves of an indispensable means of identifying the most appropriate lines of enquiry to follow in any given case.

Having gestured at the concept of rationality, I ought to stress that I intend nothing very grand or precise by that much abused term. When I speak of agents as having rational beliefs, I mean only that their beliefs (what they hold to be true) should be suitable beliefs for them to hold true in the circumstances in which they find themselves. A rational belief will thus be one that an agent has attained by some accredited process of reasoning.

I am convinced, in short, that the importance of truth for the kind of historical enquiries I am considering has been much exaggerated. I take this to be a product of the fact that so much of the meta-historical discussion has hinged around the analysis of scientific beliefs. In such cases the question of truth may perhaps be of some interest. But in most of the cases investigated by historians of ideas, the suggestion that we need to consider the truth of the beliefs under examination is, I think, likely to strike the historian as strange.