He remained, for all his study, something believed to be unknown where he ought to have known and known where he ought to have not known. One kind of knowing was modelled on what he saw in others and he was quick to fill in the absence by triangulating back from responses he encountered. A deal of imagination was required here, of the kind that fills in the details in dreams from a few vague indications. In this process he was all too likely to put more work into the fleshing out of the world than that into himself, whom he could keep posed as naively as in a child's drawing. His mistake was to begin inquiry from doubt instead of from the neutral state of not knowing. Doubt, as Descartes had shown, leads from repudiation to affirmation, a sort of push-pull too easily anticipated and misapplied. Doubt, as a tool used to develop knowledge, was far less innocent than it seemed since it begins by taking up a dominant attitude towards that which it intends to question and thus fails to recognise it.