R. Daniel Eidensohn, author of the recently published Da'as Torah, posted the following excerpt to Avodah of an article by Dr. Marvin Fox, "Nahmanides on the Study of Aggadot" in The Journal of Jewish Studies 40:1 (1989), pp. 98-100 link, I'm adding paragraph breaks based on conjecture):
Septimus makes the argument even stronger by virtue of the fact that he cites in his notes numerous sources which sustain the point that has just been made. His views are summarized with caution and clarity, when he says, "I would venture to say that anyone who reads Nahmanides' commentary will find ample evidence that he did not accept the absolute authority of all aggadah." What makes this entire discussion puzzling is the fact that anyone who would take the trouble even to read casually in Nahmanides' work, particularly, as Septimus points out, his commentary on the Torah, would see immediately how strange it is to assert as a truth beyond all question that his actual view was that we are obligated to believe all aggadot. There is hardly a page in that commentary where Ramban does not reject openly a midrash or a talmudic aggadah.
This approach to aggadah does not imply, as some of the commentators on this subject supposed, irreverence or impiety. Neither does it imply a lack of appreciation for the importance and value of aggadah. It simply indicates an understanding that, in contrast with our relation to halakhah, we have here the option, nay, the need, to be selective. We have a revealing parallel in the way that Ramban relates to the commentaries of his two great predecessors, Rashi and Ibn-Ezra, and to the works of Maimonides. He is unsparing in criticism when it is called for in his judgement, even to the point of seemingly unrestrained acerbity. Yet it is immediately evident that he has the highest regard for these earlier commentators, learns a great deal from them, often expresses agreement and appreciation, while allowing himself the right to react to their work with careful selectivity. When he thinks they are relatively superficial or when he thinks they are wrong, he says so without hesitation. This does not diminish his respect for them or his recognition of their importance or that of other scholars and commentators who preceded him. No one who has studied the works of Nahmanides can doubt for one moment that he held Maimonides in very high regard. Yet, in one of the most extreme cases of opposition to the views of Maimonides, he attacks his teaching with remarkable lack of restraint, concluding his attack with the statement that the account which is given by Maimonides "contradicts Scripture so that it is forbidden even to hear these words, to say nothing of believing them." Even so, there is no possible question about the deep regard in which Ramban held Rambam. Yet Ramban is totally confident that he knows how to be selective and which criteria to apply in particular cases.
This same attitude is evident throughout his treatment of aggadot. Respect and appreciation do not imply that one is obligated to abandon all independent judgement. The interpreters who find it difficult to believe that Nahmanides could have meant what he said in his report of the disputation have not paid close enough attention to his statement in its context. Ramban is saying that we do not have to believe in the truth or correctness of any given midrash. This does not mean that he approaches the whole of rabbinic aggadah with an initial attitude of disbelief, irreverence or outright rejection, but that we are permitted, even mandated, to exercise our intelligence and our learning in order to determine when to accept and when to reject a particular midrash.
Those who are shocked by Nahmanides' stance have paid insufficient attention to the nature and status of aggadic literature within the complex of the Judaic canon. They are equally lacking in an understanding of the style and mentality of one of the greatest Jews of the Middle Ages. What is particularly revealing is the extent to which these interpreters are dominated by their preconceptions of the nature of Jewish orthodoxy. Since, as is clear from their writings, they identify orthodoxy with literalist fundamentalism, they reach the understandable conclusion that Ramban could not both be a voice of orthodoxy and question the authority of "canonized" aggadic texts. Had they taken the trouble to look at the sources with an unprejudiced eye, they could easily enough have discovered that very early authorities had already expressed doubts and antagonisms toward aggadah in general. In the view of R. Zeira, the very unstructured character of aggadah makes it dangerous. He asserts that one cannot readily extract sound doctrine from it since it tends to turn things upside down. He thus advises his son to have nothing to do with the study of aggadah, but to devote himself exclusively to halakhah. In another well-known passage it is stated that a divine curse and other dire consequences await those who occupy themselves with aggadah. These statements do not represent, by any means, the only opinion expressed on this subject in the rabbinic sources, but it is certainly not a view which is unique or totally idiosyncratic. Why then should Ramban be considered to have uttered heretical statements when he simply propounded the far milder position that we are not obligated to believe every aggadah?