מתני' מתנה שומר חנם להיות פטור משבועה והשואל להיות פטור מלשלם נושא שכר והשוכר להיות פטורין משבועה ומלשלם כל המתנה על מה שכתוב בתורה תנאו בטל וכל תנאי שיש מעשה בתחילתו תנאו בטל וכל שאפשר לו לקיימו בסופו והתנה עמו מתחילתו תנאו קיים:
In terms of the authorship of the Mishna at the top of 94a, the gemara was doing so well with attributing it to Rabbi Yehuda, only to veer off at the end and attribute it to Rabbi Meir (with a possibly far-reaching exception within the position of Rabbi Meir.) While the final objection, from the sefa of the sefa, is something that we need to come up with an answer for, it is much more convincing to say that this is Rabbi Yehuda. See also the Yerushalmi which concludes as a matter of course so, citing a brayta that allows masna al ma shekasuv baTorah for monetary stipulations, in general.
ושאני הכא דמעיקרא לא שעבד נפשיה
Indeed, this is a difference in kind. Marriage is marriage as defined by the Torah, carrying along all its benefits and obligations, monetarily or otherwise. If you want to buy into Biblical marriage, this is what it entails. But there are all sorts of monetary transactions or watchman obligations one could set up, and so one can set up a non-Biblical set of obligations…
פרק שמיני - השואל את הפרה
מתני' השואל ה"ג השואל את הפרה
An interesting Rashi establishing the correct girsa of the Mishna.
אם גנוב יגנב מעמו ישלם לבעליו
Throughout, the pasuk is quoted without the leading vav. So im instead of ve’im. This corresponds to the Samaritan text, rather that to the Masoretic text.
אם גנוב יגנב מעמו ישלם לבעליו
Derashos aside, the structure of the halachos makes logical sense and can be read, in peshat manner, into the pesuk. Afterwards, we can bring the midos shehaTorah nidreshes bahen to come to those conclusions within the framework of midrash halacha.
The setup is that “stolen or lost” form an equivalence class. What happens to one happens to the other. We first encounter this by the unpaid watchman (first paragraph), who swears and is exempt. Stolen is most explicit there, but in the intervening verse (al kol devar pesha), if we don’t take this as an out-of-context interjection (eruv parshiyos), mentions any loss (al kol aveida asher yomar ki hu zeh). So “stolen or lost” is an equivalence class.
When we reach the next tier, the paid watchman (second paragraph), we see that he pays for “stolen”, but only swears for the next level up, in the category of ones. So “lost” would come along as part of the equivalence class.
Meanwhile, the paid watchman swears and is exempt for a different equivalence class - dies of normal causes, broken, captured. So when we reach the next tier of borrower, and are told that he must pay for two members of that equivalence class, we understand that as a shorthand for every member of the class. The gemara sort of suggests this as a derivation, but it is pushed off, and we seek another derasha. Regardless, the structure of the halacha is as described above, even if the particular way of arriving at it via midrash halacha is different.
6) Dibra Torah
אלא למאן דאמר דברה תורה כלשון בני אדם כו'. אע"ג דלדידיה דרשינן בכל מקום לבד מהיכא דלא מיסתבר כדפי' בפ"ב (לעיל דף לא: ד"ה דברה) ה"נ לא מסתבר לרבות אבידה כיון דכתיב בלשון גניבה ואית לן למימר דברה תורה כלשון בני אדם:
That is, elsewhere, earlier in the masechta, we saw that even the one who says dibra Torah such that we don’t make a derasha will still make the derasha if needed and it is mistabar. So Tosafot explains how here it must not be mistaber, because the language of geneiva is not the same as loss. I would say it is mistaber since it was part of the same group for an oath to exempt by shomer chinam.
In truth, it is that other gemara earlier in Bava Metzia which is somewhat questionable, in attributing a bunch of duplicate language derashot to the side that says dibra Torah. Our local gemara does not strike me as problematic. It indeed may serve as a counterexample to the claims in that other gemara.