Perhaps the clearest aspect of the mitsvah is that it originally applied to the writing, or correcting, of an entire Torah. While the Torah refers to writing hashirah hazot, which might only mean the song recorded in Haazinu, everyone agrees that one has to write the entire Torah in order to write Haazinu properly. The gemara is also fairly clear that the mitsvah applies anew in each generation, so that one may not rely on a sefer Torah inherited from one’s ancestors. A significant body of opinion does allow buying a sefer Torah as a less-preferred fulfillment of the mitsvah (starting with Rashi and including the Vilna Gaon; surprisingly, however, Rema in Yoreh Deah 270 rules against this view). All agree that correcting an invalid sefer Torah qualifies as writing it.

The aharonim question the distinction between inheriting and purchasing a sefer, since in both cases the person is not writing it himself. Minhat Hinukh suggests that a person buying a sefer at least expends the effort and monetary cost of the purchase, whereas the heir does not involve himself in the mitsvah at all. This leads to other interesting questions, such as whether receiving a sefer as a gift would qualify or not, but we will leave those for another forum.

A more central issue around this mitsvah was raised by Rosh and affects several halakhic issues. Rosh pointed out that in the time of the gemara, people not only read from a sefer Torah in public, but used it for their private study of Torah. In that context, he claimed, Hashem obligated people to write their own sefer Torah, so that they could always have one available to use for purposes of study. Nowadays, when people no longer study from the actual sefer, the mitsvah is no longer to write a Torah, but rather to write (or purchase) the books one would use in study (including Humash, Mishnah, Gemara, and commentaries).

Taken at its face value, Rosh seems to be saying that the mitsvah was always an adjunct of the mitsvah to study Torah; as people’s practices around Torah study changed, the way in which we were supposed to fulfill the mitsvah changed as well. Rambam seems to share this assumption, since he excludes women from the obligation. He could have nevertheless thought that the obligation depended on the actual writing of a sefer despite its being connected to the mitsvah of Talmud Torah, but at the least his perspective on women (which he does not explain at all) seems to agree with Rosh's assumption.

Several aharonim read Rosh as adopting the position we mentioned above, most notably the Perishah in his commentary on the Tur. Others, including Beit Yosef, cannot accept that Rosh would have assumed such a radical change in the obligatory performance of a mitsvah. Instead, they suggest that he meant only to articulate an added aspect of the mitsvah, but that the original obligation to write a sefer Torah itself remained in full force.

Sha'agat Aryeh (no. 35) is so sure that the mitsvah of writing a sefer Torah is independent of the mitsvah of Talmud Torah that he cannot accept Rambam's view that women are exempt, nor can he accept Perishah's reading of the Rosh. Instead, he suggests that the mitsvah is completely Rabbinic in the post-Mishnaic period, since the Talmud candidly admits that we have lost the definite tradition of certain spellings (haserot and yeterot, words that are written with or without a vav or a yud on certain occasions) of the Torah. Since we cannot be sure that we are writing sifrei Torah exactly as Moshe wrote the one he gave the Jewish people, Sha'agat Aryeh says, the mitsvah de-oraita was no longer fulfillable anyway.

Having lost the mitsvah de-oraita, Hazal instituted a rabbinic version of the same commandment, to insure that the Torah not be lost from the Jewish people. This version, however, was connected to Torah study, so that women might not be included and, as methods of study changed, the mitsvah could as well. The crucial point that led Sha'agat Aryeh to his view was his rejection of Rosh's (and perhaps Rambam's) assumption that this mitsvah depends upon or is an adjunct to, the mitsvah of Torah study.

Hatam Sofer 254 similarly rejects the connection to study. He notes a recurring debate in the gemara (centrally in Baba Metsia 115) as to whether darshinan ta`ama de-kra or not, whether we allow ourselves to shape our fulfillment of certain mitsvot according to our independent understanding of the Torah's goal in commanding them. Since thegemara clearly rejects the view that allows doing so, Hatam Sofer writes that Rosh must only have been making his comments in order to explain how that point of view would currently understand the mitsvah. We, who do not shape our performance of mitsvot in that way, could not do so.

Here, too, Hatam Sofer's point, while well taken, contradicts the general reading of Rosh, since all the other aharonim agreed that Rosh was writing halakhah le-ma`aseh. Too, in this case (it seems to me) darshinan ta`ama de-kra should not arise, since the Torah itself gives the reason that Rosh used to shape his view of the mitsvah (which the gemara itself accepted as justifying basing the halakhot of a mitsvah on its underlying reason). The Torah says to write the shira, teach it to the Jewish people, and place it in their mouths, clearly indicating that the writing is a part of the teaching and the placing in their mouths. Rosh's connecting the mitsvah of writing to the rest of the verse should therefore not have any problems of darshinan ta`ama de-kra.

Which leaves us, in my meager understanding, with the view among rishonim that this mitsvah is connected to that of Talmud Torah, as evidenced by Rambam's assumption that women are exempt and by Rosh's view that the mitsvah changed as the methods of Torah study changed. Whether we accept Rosh's view in practice is a matter of debate among aharonim, some of whom instead assumed that the mitsvah is simply to write a sefer, without any connection to study at all. Shabbat Shalom.