The article argues that there isn't just one way to define scientific rationality, but rather multiple ways based on different criteria. This raises the question of which criteria are better than others. The author suggests that when we try to justify a certain idea of rationality, we end up using another form of rationality to do so. This implies that the underlying rationality we use to justify our ideas is not so different from the ideas themselves.