Three hundred and sixty years after Descartes's Discours de la Methode we may be tempted to think that our science has by now reached a stage in which the "clear and distinct ideas" are no more an ideal, but a safe possession of any scientifically learned man. Descartes was right in questioning the whole received knowledge of his time, but - it is often assumed - science at the end of the XX century does not require any such radical treatment. One of the roots of this kind of complacency is the opinion that contemporary science progresses by absorbing smoothly and new discovery, so that the more recently accepted theory in any field contains, in some sense, all the worth the precedent theories could boast of. In this statement there are many concepts that need clarification, but we shall dwell only on the idea of the new theory as 'containing' the achievements of the past. The meaning of this is by no means clear, and on the whole historical record belies any naive interpretation...
