The Gemara in Shabbos (21b) records that one should preferably light Chanuka candles at the doorway of one’s home. Rashi (s.v. Mibachutz) writes that one lights in a chatzer (shared courtyard) but not in the reshus harabim, whereas, Tosfos (s.v. Mitzvah) writes that one lights at the entrance of one’s house only if it is adjacent to the reshus harabim, but if the house is inside a chatzer, one must light at the entrance of the chatzer adjacent to the reshus harabim.
At first glance, it is hard to understand motivatied Tosfos to say this. After all, the Gemara just says “Mitzvah lehanicha pesach beiso mibachutz,” implying that entrance of one’s house is always the appropriate place to light. Why does Tosfos understand that this is only true when the house is adjacent to reshus harabim? Tosafos bring two proofs to their position: Firstly, the Gemara on 23a says that if there is a chatzer with two doorways, one must light in both of those doorways. If Rashi is correct and one is obligated to light at the entrance of one’s home, why would the Gemara discuss a case where one if lighting at the entrance of their chatzer. Tosafos further point to a Gemara on 23b that says a ner (meaning, a container where on places oil, as Rashi explains there) with two openings (two wicks), may be used by two people to fulfill the mitzvah. Tosafos say that this must be referring to a case where there are multiple houses in a chatzer and therefore there are many people lighting in one place: the entrance to the chatzer adjacent to reshus harabim. According to Rashi, why would multiple people light in one place? They should all be lighting at the doorways of their own homes. [i]
Rashi addresses both this issues.[ii] Regarding the Gemara on 23a about a chatzer with two doorways, Rashi (s.v. Chatzer) explains that there Gemara is referring to a house which has two doors leading into the chatzer. Regarding the Gemara on 23b about one ner with multiple wicks, Rashi (s.v Shtei) explains that multiple people would want to be yotzei in the same location if a household is trying to fulfill the mitzvah according to the level of mehadrin, where one candle is light for each family member. The Maharsha points out the Tosafos does not understand the Gemara this way, because if the Gemara was limited to a case of people fulfilling the mitzvah at the mehadrin level the Gemara should have specified this. Furthermore, Tosafos couldn’t have said that it is a case of mehadrin min hamedhadrin, because they hold that mehadrin min hamedhadrin involves only one person lighting.
At the end of the day, however, it seems that Rashi’s position is difficult. Why does he explain the Gemara on 23a to refer to a house with two openings into a chatzer when the simple reading of the Gemara is in accordance with the view of Tosafos, that the chatzer has two openings into reshus harabim. Additionally, later in the same beraissa there Gemara quotes that if one lives in an attic, one should light in a window close[iii] to reshus harabim, implying that the lighting earlier in the beraissa was also done facing reshus harabim.
Perhaps we can suggest that Rashi and Tosafos argue over the correct definition of pirsumei nisa – does it mandate publicizing the miracle of Chanukah to the entire world, or just to one’s family. Tosafos assumes that pirsumei nisa’s intended audience is the general public, in which case the menorah must be placed near the reshus harabim. Rashi, however, things that both audiences are intended, and the halacha takes them both into account. One can’t light in their one home because that would lack in pirsumei nisa for the general public, but one can’t light at the entrance to one’s chatzer either, as that would lack in pirsumei nisa for one’s family. Therefore, one should light at the entrance of one’s home in order to accomplish both pirsumei nisa for one’s family and for, to a limited extent, the public. If this is case, it is understandable why Rashi understood the Gemara on 23a differently that the simple reading, and ignored a minor proof from Tosafos on 21b.
However, we still have to understand what led Rashi to assume that this definition of pirsumei nisa is correct. Perhaps his source was another Gemara on 21a which says that, at a basic level, one fulfills the mitzvah of Chanukah with one candle per household. A higher level, mehadrin, is fulfilled by lighting one candle for each member of the household, and the highest level, mehardin min hamehadrin, is fulfilled by changing the number of candles lit as the holiday progresses. Tosafos (s.v Vehamehadrin) and the Rambam (Hilchos Megillah Uchanukah 4:1) dispute the correct way to the fulfill mehardin min hamehadrin. Tosafos says that this level refers back to ner ish ubeiso, and each household would only light one menorah. They contend that if we say that each member of the household were to light depending on the night, then that would override the entire point of lighting in the this manner which is to make it clear to onlookers which night of the week it is. If, for example, there were six candles on the first night because there are six members of the, or there were three members of the house lighting three candles on second night, one would not be able to recognize that the number of candles lit correspond to the day of the holiday. The Rambam ignores this claim and contends that the Gemara builds each level on top of the other, with the fulfillment of mehardin min hamehadrin including the practice of mehadrin to light for each family member. I would contend that Rashi agrees with this approach, because it is the simpler way of reading the Gemara and Rashi doesn’t comment otherwise. How would the Rambam and Rashi respond to Tosafos’ problem? Perhaps they would say that we really don’t care about the onlookers and their perspective. The principle concern is for the members of the household who understand the number of candles they are seeing, not for the general public.[iv]
[i] The Rashba (s.v. Ner) brings another Gemara as a proof to Tosfos, which is the opinion of R’ Yehuda quoted later on 21b that a storekeeper who lights his Chanuka candles in the public domain is not considered to be negligible if a camel laden with tinder was caught aflame by his menorah (since it was within his right to place the candles outside), which implies that one would be lighting the candles in the Reshus Harabim, because the owner of the camel would have no claim if he were trespassing into the Chatzer. The Tosfos Harosh (s.v. Mitzvah) points out that this isn’t a valid proof, because it could just be that the store was located adjacent to the Reshus Harabim.
[ii] This is all assuming that Rashi holds that one would actually light at his house inside the courtyard, not like the Ritva (s.v. Mitzva) who says that Rashi really agrees with Tosfos and that Rashi is just saying that when one lights at the entrance to the courtyard adjacent to the Reshus Harabim, he should be place his menorah inside the courtyard. However, the Ritva’s understanding is difficult in light of the Rashi on 23a that we discuss.
[iii] Although, perhaps the language of Semucha would actually imply like Rashi that the window isn’t directly opening to Reshus Harabim but is the window of the house that is closest to the Reshus Harabim (one manuscript of the Meiri, in fact, has the language of Pesucha which implies that it is directly opening to the Reshus Harabim).
[iv] This might relate to the idea of Hadlakas Neiros Chanuka being a Chovas Habayis (meaning, that since it is a Chiyuv that is incumbent on the household, therefore it must be done in way that the household actually sees it) for more on this idea see Beikvei Hatzon (Siman 20). Additionally, the Gra (O.C 672:2) explains that the Rambam argues on Tosfos because he holds that the reason that we change the number of candles each night is based on Maalin Bekodesh (which would apply to each member of the household) as opposed to Tosfos who says that it has to be recognizable because it is based on the number of days. (The Gra and Beikvei Hatzon were pointed out to me R’ Dovid Willig.)
The Lamdan 