Shute gives a good clarification of Lonergan's often misunderstood term 'pure' desire to know: it is "intelligence unencumbered by the drag of the biases." (Shute, Lonergan's Discovery of the Science of Economics, Toronto 2010, 79 n 57) Thus he is not affirming that we have, at any time, a pure desire to know. It is rather the pure desire considered apart from the biases. It is a 'line of reference' or a first approximation, somewhat like the 'state of pure nature' in medieval theology.
Lonergan also talks of the pure theory of external human action ("Essay in Fundamental Sociology," LEER 17), of the 'pure cycle' and 'pure surplus income' in his economics, and of 'pure formulations' in his theory of interpretation (Insight CWL 3:602).