Berachot 51-52: Why say the Halacha is like Bet Hillel?

On Berachot 51b-52a:

Our Rabbis taught: The points of difference between Beth Shammai and Beth Hillel in relation to a meal are as follows: Beth Shammai say that the blessing is first said over the [sanctity of] the day and then over the wine, because it is on account of the day that the wine is used, and [moreover] the day has already become holy²¹  before the wine has been brought. Beth Hillel say that a blessing is said over the wine first and then over the day, because the wine provides the occasion for saying a benediction.²²  Another explanation is that the blessing over wine is said regularly²³  while the blessing of the day is said only at infrequent intervals, and that which comes regularly always has precedence over that which comes infrequently. The halachah is as laid down by Beth Hillel. 

That ends the brayata. Then, the gemara analyzes this brayta.

'And the halachah is as stated by Beth Hillel'. This is self-evident, for the Bath Kol²⁴  went forth [and proclaimed so]!²⁵  If you like I can reply that this statement was made before the Bath Kol [had issued forth], and if you like I can say that it was made after the Bath Kol and that it represents the view of R. Joshua, who said that we pay no attention to a Bath Kol.¹

When the parallel Yerushalmi cites the brayta, the last statement that the halacha is like Bet Hillel is absent. But look to the Tosefta, and you will find it there. In Berachot 5:25:

ארבעה דברים שבין בית שמאי ובית הלל בסעודה בית שמאי אומרים מברך על היום ואחר כך מברך על היין שהיום גורם ליין שיבא וכבר קדש היום ועדיין לא בא ובית הלל אומרים מברך על היין ואח"כ מברך על היום שהיין גורם לקדושה שתאמר דבר אחר ברכת היין תדיר וברכת היום אינו תדיר תדיר ושאינו תדיר תדיר קודם והלכה כב"ה.

A major difference here is that the Tosefta leads with the statement that there are four points of difference between Bet Shammai and Bet Hillel. The statement, at the end of the lengthy discussion of point #1, is meant to apply comprehensively to all four. Because we might have an alternative tradition that we pasken like Bet Shammai in one of them.

Indeed, consider point #4, and what the gemara later on Berachot 52a states, regarding the dispute between Bet Shammai and Bet Hillel, where Bet Hillel say one washes first and Bet Shammai say one sweeps first:

R. Jose b. Hanina said in the name of R. Huna: In all this chapter the halachah is as stated by Beth Hillel, save in this point where it is as stated by Beth Shammai. R. Oshaia, however, reverses the teaching⁹  and in this point also the halachah follows Beth Hillel.

We typically assume "reverses the teaching" means that the positions of Bet Hillel and Bet Shamai are transposed. An alternative (albeit perhaps awkward) understanding is that he reverses the statement that the halacha is like Bet Shammai here. Rabbi Oshaya and Rabbi Chiya authored braytot, and the Tosefta.

The Tosefta, meanwhile, does not reverse the positions in this last case.

____________________

Two more points, in passing.

First, the character of the Tosefta here seems to be one of a proto-gemara. Recall that Rav is cited in Tosefta, and Rabbi Chiyya and Rabbi Oshaya, the compilers of Tosefta, bridged the Tanna and Amora period as well. So we see a statement that the halacha is like Bet Hillel; and other such statements of who the halacha is like. And we see alternate suggestions of Bet Hillel and Bet Shamai's reasoning.

The gemara, naturally, treats this proto-gemara like any other brayta. And therefore it analyzes why Bet Hillel would say a davar achar, to give an alternate explanation. Meanwhile, it was not necessarily Bet Hillel who says davar achar but the proto-gemara. And similarly, the gemara analyzes why the brayta would say halacha ke-X.

See also this, from Wikipedia on Tosefta:

The traditional view is that the Tosefta should be dated to a period concurrent with or shortly after the redaction of the Mishnah. This view pre-supposes that the Tosefta was produced in order to record variant material not included in the Mishnah.

Modern scholarship can be roughly divided into two camps. Some, such as Jacob N. Epsteintheorize that the Tosefta as we have it developed from a proto-Tosefta recension which formed much of the basis for later Amoraic debate. Others, such as Hanokh Albeck, theorize that the Tosefta is a later compendium of several baraitot collections which were in use during the Amoraic period.

More recent scholarship, such as that of Yaakov Elman, concludes that since the Tosefta, as we know it, must be dated linguistically as an example of Middle Hebrew 1, it was most likely compiled in early Amoraic times from oral transmission of baraitot.,^[2] Professor Shamma Friedman, has found that the Tosefta draws on relatively early Tannaitic source material and that parts of the Tosefta predate the Mishnah.^[3]

Second, in terms of the gemara on 52b that says:

Another explanation is, so that the meal should follow immediately the washing of the hands.

This does not have to feed into the three teikef sugyas. Tosefta doesn't utilize the term teikef:

[דבר אחר אין נטילת ידים אלא סמוך לסעודה]

(and indeed it seems that some sources don't have the statement). Yerushalmi (58a) as well does not utilize the term teikef:

מה טעמהון דב"ה לעולם אחורי הכוס טמאין דבר אחר אין נטילת ידים אלא סמוך לברכה.