MARK WASHOFSKY

such"is the way of nature." Moreover, Bacharach writes, Rashi accepts this reasoning: although the fetus is not a legal person, it is only because of the threat to the mother's life that the abortion is permitted.?® Thus, in the case before him, Bacharach cannot agree to the woman's request.

Trani and Bacharach, in other words, differ on more than merely the answer to the question: when is abortion permitted? They differ fundamentally over the identification of precedents: precisely which sources are to be consulted in the consideration of the issue? Trani does not take the rodef analogy into account, and the warrant for abortion accordingly extends to instances other than mortal danger. Bacharach, while recognizing that the Talmud does not call the fetus a rodef in a situation of difficult childbirth, sees Maimonides ' ruling as important enough to resolve it(leyashvo) with the Talmudic passage(Sanhedrin 72b) that evidently contradicts it. The upshot is that the Talmudic text is reinterpreted; in Bacharach's rendition, it permits abortion only in cases where the fetus, prior to its emergence from the womb, endangers the mother's life.

This analysis suggests the fluidity of the Jewish legal discussion on abortion, at least in its early stages. In the seventeenth century, it was not at all obvious that the fetus could be aborted only when it threatened the mother's life. It was not at all obvious that Rambam 's interpretation of the Talmudic materials was correct. Nor, as Trani 's responsum shows, did it seem obligatory to take his position into account in reaching an halakhic conclusion. Jewish law was hardly univocal on the subject of the warrant for abortion.

How did univocality come to take the place of fluidity? How did Rambam 's restrictive stance become the halakhic consensus? The answer, I believe, lies in halakhic method, the techniques of legal reasoning which recent authorities have employed in order to determine the"correct" legal conclusion from among the possible alternatives. These techniques, as I have indicated, are excessively formal and do not reflect legal discussion at its best. But, as the following three examples of halakhic"method" demonstrate, the

47