“So shall you also separate T’rumah” (Bamidbar 18:28). Although the main intent of the verse is to instruct the Levi’im that they also must give T’rumah to the Kohanim, the Talmud (Bava M’tziya 71b) says these words (especially the extra word “also”) teaches us that “sh’lucho shel adam k’moso,” something done by an authorized representative is considered as if it was done by the person he is representing. However, in the main “sugya” (in depth discussion) of “sh’lucho shel adam k’moso,” in Kiddushin (41a-b), the Talmud gives a reason why this concept being taught regarding T’rumah cannot be automatically applied elsewhere. Instead, the concept of “sh’lichus” (representation) is learned from a different source, to the extent then when it applying when separating T’rumah is also learned from those other source-texts. (The Talmud there says the extra verbiage in the T’rumah verse teaches us something else.) How can the Talmud in one place say that the source that “sh’lichus” works is from T’rumah, if, as part of the main discussion about the topic, it says that T’rumah can’t be the source?
This becomes more puzzling when we consider that Rashi (Bava M’tziya 71b) says the source that the concept of “sh’lichus” (applying all over) is learned from T’rumah is in Kiddushin! How can Rashi refer us there if that’s where the Talmud says that not only don’t we learn that “sh’lichus” works elsewhere from T’rumah, but we don’t even learn that “sh’lichus” works for T’rumah from T’rumah? Rashi reiterates that the source for all “sh’lichus” is T’rumah elsewhere as well (Shabbos 153b, Y‘vamos 113a, Bava M‘tziya 22a and 96a and Chulin 12a), as does Tosfos (Gittin 23b). Rambam (Hilchos T’rumah 4:1) also says that T’rumah is the source that “sh’lichus” works for separating T’rumah, and (in Hilchos Ishus 3:17) strongly implies that it is the source that “sh‘lichus” works elsewhere as well. Why is the source for “sh’lichus” constantly presented as being from our T’rumah verse if the Talmud explicitly says that it isn’t?
Rashbuh (in Kiddushin) is among the commentators who reinterpret the Talmud in Bava M’tziya, so that instead of meaning that the verse in T’rumah is the source for “sh’lichus” (all over), it is only the source for limiting when “sh’lichus” works (which is what the Talmud in Kiddushin says the extra T’rumah verbiage teaches us, at least according to the first opinion there). Nevertheless, this is difficult to read into the Talmud’s words in Bava M’tziya, and next to impossible (if not impossible) to read into Rashi’s words (or those of Tosfos or Rambam).
P’nay Y’hoshua (Gittin 23b) suggests that when it’s said that the source for all “sh’lichus” is T’rumah, it really means from a combination of T’rumah and Geirushin (a “tzad ha‘shaveh”). Aside from this not being what the Talmud in Kiddushin says (it says it is learned from a combination of Geirushin and Kodshim), Geirushin is not mentioned as a source with T’rumah by anyone else, most notably perhaps, not even by Rambam (in Hilchos Ishus) when that is the topic of discussion.
Rabbi Yosef Korkos, in his commentary on Rambam’s Hilchos T’rumos, makes two suggestions to explain how Rambam could say the T’rumah verse is the source that “sh‘lichus” works for T’rumah (rather than the “tzad ha‘shaveh” the Talmud presents as the source) . His second suggestion, based on certain cases where “sh’lichus” will not work for separating T’rumah, is that the extra verbiage is needed to teach us that not all cases of “sh’lichus” for T’rumah are being excluded, which means it is also teaching us that “sh’lichus” will work most of the time (i.e. when it’s notnone of those exceptions). While this is a valid approach to explain how Rambam can say it is the source for “sh’lichus” when separating T’rumah, it will not explain how this makes T’rumah the source for all “sh’lichus” if the Talmud says it can’t be extended to other situations. Rabbi Korkos’ first approach is that the verse’s extra verbiage teaches us not only certain limitations regarding who can be a “sh’liach,” but also that “sh’lichus” works when separating T’rumah. This approach has the same issue to overcome as his second approach, while also having to explain how the extra verbiage can teach us both things, that “sh’lichus” works when separating T’rumah and that certain people cannot be that “sh’liach.”
I would therefore take his second approach a step further, and re-explain the Talmudic discussion. The Talmud had asked why we need a verse to teach us that “sh’lichus” works when separating T’rumah if we would know this anyway from the combination of it working by Geirushin (serving divorce papers) and Kodshim (bringing offerings). Instead of the answer being “you’re right, it doesn’t teach us that; rather, it teaches us who can’t be a ‘sh’liach,” the answer is really “you’re right, we don’t need the extra verbiage to teach us that ’sh’lichus’ works when separating T’rumah, but it needs to be taught here for a different reason, so that ‘sh’lichus’ is included in the verse, thereby allowing us to make a comparison between the person who can be a ‘sh’liach’ and the person who makes him the ‘sh’liach.” Just as the Talmud’s answer according to Rabbi Shimon is that the extra verbiage (“also”) teaches us that we can appoint a “sh’liach” to separate T’rumah even though the word “you” excludes certain cases of separating T’rumah from “sh’lichus,” so too according to the Talmud’s first answer does the extra verbiage (“also”) teach us that “sh’lichus” works for separating T’rumah (even though we would know that anyway from Geirushin/Kodshim) so that the “you” can teach us a limitation in “sh’lichus.”
Once “sh’lichus” is also (pardon the pun) learned from T’rumah, the limitations that are also (pardon the reverse pun) learned from T’rumah apply to all situations of “sh’lichus,” making T’rumah a primary source for “sh’lichus” (and its limitations).