Bracha b'mitzvat tefillin on tefilla shel rosh

Question

The bracha that appears in all modern siddurim that I'm aware of for the tefilla shel rosh (whether normally or in case of interruption between the two tefillin) ends with the phrase al mitsvat tefillin (על מצות תפילין). This is also what appears in our printings of the Babylonian Talmud at Brachot 60b and Menaḥot 36a.

However, the manuscripts of Menaḥot that we have uniformly have the text b'mitsvat tefillin (במצות תפילין). See also Dikdukei Sofrim, where it's also noted that b'mitsvat is the text of Halachot Gedolot (Hilchot Tefillin), Rif (Tefillin 8a in manuscripts), HaEshkol (Vol 1 p228 ed. Albeck), HaIttur (Hilchot Tefillin Ḥelek 8), and Or Zarua (Siman 582). On top of that list, I found that Rosh (Pesachim 1:10), Rashba (Hameyuḥas leHaRamban 202), Kol Bo (146:95), Halachot Ketsuvot, and Geonic Responsa have b'mitsvat, and Rosh and Kol Bo explicitly prefer it over al mitsvat.

My question: What is the history of b'mitsvat since the end of the Rishonic era? Is there any trace remaining? Has anyone dealt (historically or halachically) with b'mitsvat vs. al mitsvat since the Rishonic era? Has anyone from any era discussed this explicitly and come down on the side of al mitsvat? (I'm not interested in a preference for al mitsvat by simply omitting b'mitsvat, eg. Rambam in Mishne Torah; I'm interested in explicit discussions of the preference.)

Answer 1

Did some searching for you...

Note in Oz Vehadar:

Kol Bo:

Footnote 543:

Birkei Yosef:

Answer 2

Professor R. Naftali Wieder addressed this issue in two places (that I know of): once in his article פרקים בתולדות התפילה והברכות (footnote 48; p162 in התגבשות נוסח התפילה במזרח ובמערב volume 1) and in his article חמשה נושאים בתחום התפילה (section 2; p184 in התגבשות volume 1). In the first, he points out all the early sources that contain במצות תפילין: besides the ones in the OP, he also lists Rav Natronai Gaon, Halachot Gedolot, Rav Amram Gaon (also here), R. Shlomo of Sijilmasa, Sefer HaMakhria' (in quoting Rabbeinu Shlomo ben HaYatom in the name of Rabbeinu Gershom), Rokeach (when quoting R. Yehudah HeHasid), and R. Yehudah ben R. Kalonymous of Speyer.

He expresses surprise at the fact that, even with the plethora of sources and its apparent widespread use, it has completely disappeared completely from current use, and not because of a great argument on the behalf of על מצות. He only found one early source that explains why it should be על מצות (but not against במצות תפילין), which is Orḥot Ḥayyim (#16): ויש אומרים שמברכין עליהן על מצות תפילין, וסימן: ושמן על ראשך אל יחסר, שדרשו ז״ל אלו תפילין של ראש היא ב„על“, ואית דאמרי במצות תפילין. He suggests two reasons why על מצות won out: (1) the usual form of brachot is with ל or על, and not with ב, and (2) the phrase צוה ב isn't in Tanakh or the Mishna.

In his second article, he quotes R. Asher ben Ḥayyim of Montsant (grandstudent of the Rashba) in his Sefer HaPardes (starting כמה אני מצטער), where he writes strongly against those who say על מצות instead of במצות. In brief, R. Asher writes that all the Catalonian gemarot have במצות, and all his colleagues agree with him that it should be במצות. He writes that על מצות is a mistake, and that the fact that Rambam has על מצות in his Mishneh Torah is a copyist's faulty "correction" (although none of the Frankel, Qafiḥ, or Rabinowitz editions mention anything except על מצות), although he admits that he can't defend Rabbeinu Tam (who wrote explicitly that it should be על מצות) using that explanation. He also disagrees with על מצות on the basis of the rules for choosing between על and ל (where it seems that ב comes down on the side of ל, perhaps with the implied לקדשנו במצות תפילין, as suggested by R. Louis Ginsburg).

Finally (in footnote 15), R. Wieder brings the Ḥida (Shiyurei Bracha on OḤ 25) who quotes R. Asher ben Hayyim and answers R. Asher's objection to Rabbeinu Tam, and adds that every rule has its exceptions, and we shouldn't nitpick and change a bracha because of a rule. He says that since the Rambam, who was very picky about his nusaḥ, chose על מצות, we should keep it.