Massei, The Arei Miklat

Our parsha talks of the cities of the Leviim, including the Arei Miklat.  It would be natural to assume that since the purpose of the Arei Miklat is to provide refuge for people who had accidentally killed someone from the victim's vengeful relatives, that a fortified, walled city would be optimal.  In fact, however, the simple meaning of several gemaros  (Arachin 33b and Makos 10a) is that an Ir Miklat can not have a wall (אלא בתי ערי חומה ללוים מי אית להו.)  None of the 48 cities of the Leviim can have a wall.  If a Knaani city that had a wall happened to be apportioned by lot to the Leviim, the wall would have to be knocked down (Arachin 33b, למיסתרינהו קיימי.)

The Rambam does not pasken like the plain meaning of these Gemaros.  The Rambam (8 Rotzeiach 8) only excludes cities that are very large or very small.  He does not exclude walled cities.  The Aruch Laner (Makkos 10a dh אלא עיירות בינוניות) explains that the Rambam understands the Gemara's exclusion of walled cities only because walled cities were almost always very large, so having a wall is an indicator of a problem, not a problem in itself.

But according to those rishonim that understand the Gemara as excluding walled cities, where would such a rule come from? There is no passuk that says anything about it, so there has to be a rationale.  What is it?

There are two svaros: the Gevuras Ari/Liflagos Reuven technical explanation, and the Mishna Lemelech/Radvaz practical explanation.

The Gevuras Ari says that the problem is that if a refugee became a Metzora, he would have to leave the Machaneh Yisrael.  This is only true in a walled city.  Many people do not realize this; everyone knows a metzora is sent outside the camp, as Miriam was in the Midbar.  But it is clear that this only applies in walled cities (Mishna Keilim 1:7 and Rashi Megilla 10b.  Tosfos in Arachin 32b suggests a Gzeira Shava Moshav/Moshavo as the source.)   If, under any circumstance, a refugee would be forced to leave a city, that city cannot be an Ir Miklat, because the passuk says אשר נס שמה- שם תהא דירתו.  Rav Bengis in the Liflagos Reuven says a similar thing; one of the dinim of Ir Miklat is that if the refugee dies before the Kohen Gadol dies, he has to be buried in that city.  But it is assur to bury the dead in a walled city, as that Mishna in Keilim says, and so the rules of Ir Miklat could not be fulfilled.

The Mishna Lemelech (on the Rambam there) and the Radvaz (teshuvos 2:681) say that a walled city is always a commercial center with many visitors, and because a large city cannot be an Ir Miklat because the vengeful relative could mix with the anonymous crowd (like internet commenters) and kill the refugee, a walled city cannot be an Ir Miklat.

A contemporary of mine in NIRC  (Moshe Nachum Sochaczewski) published an article in 1989  (see the second page) saying that the machlokes Rashi and Rambam whether a walled city can be an Ir Miklat can be explained with the Gevuras Ari.  He brings that the Ambuha D'sifri in Massei questions the Gevuras Ari's assumption that the metzora would be forced out of the city.  The Rambam (3 Bias Mikdash 8) says that the din that the Metzora leave the city is only a mitzvas asei (בדד ישב;) there is no lav if he stays in the city. So if faced with a conflict between the din Ir Miklat that he stay in the city and the din of Metzora that he leave the city, he would stay; an assei is not docheh an assei, and you would say sheiv al taaseh.  Rav Sochachewsky said that with this we can explain the machlokes: It so happens that unlike the Rambam, Rashi (Pesachim 67a) holds that a metzora that doesn't leave is over on a lav and an assei (בדד ישב and ולא יטמאו את מחניהם.)  So according to Rashi, the refugee would have a conflict between the Mitzvas assei to stay in the Ir Miklat, and a lav and an assei to leave because he's a metzora.  The lav plus assei would overcome the assei, and he would have to leave the city.  According to Rashi, he would have to leave.  According to the Rambam, as the Ambuha Desifri noted, it would be assei/assei, and he should stay put.   This is why Rashi holds a walled city is passul for an Ir Miklat, and the Rambam holds that it is kasher.

Before getting to the main discussion, I have several minor comments.

1.  According to the Riva, the idea of אין עשה דוחה לא תעשה ועשה is because although the asei is docheh the lav, you remain with asei against asei, and we say sheiv al taaseh (as the Imrei Moshe explains in siman 14.) Applying the Riva to the case of a metzora in the Ir Miklat, we ought to say that the asei to stay inside is docheh the lo saaseh of staying inside, and you remain with asei against asei; but in that case, we would say sheiv al taaseh, and he should stay in the city.  True, Tosfos in Kiddushin 34 holds that the logic of אין עשה דוחה לא תעשה ועשה is that the asei strengthens the lav, so even the lav is not nidcheh, but the Riva and Tosfos in Chulin 141 hold that the asei if docheh the lav and you're left with asei/asei, so his teretz wouldn't work.

     Additionally, Rabbi Sochaczewski's idea that according to the Rambam it would be assei/assei and the rotzei'ach would stay in the city is only true if he was in the city before he became a metzora.  If, on the other hand, he was a metzora before he entered the Ir Miklat, the assei/assei would require that he stay out of the city- another sheiv al ta'aseh.  So even according to the Rambam that should make the city unfit to be an Ir Miklat.

As an indication of what a great lamdan Rav Sochaczewski is, I faxed a copy of this to his law office in Baltimore, and evidently he considered it beneath himself to acknowledge receipt, and certainly not to comment upon it. I don't mean this sarcastically chalilah; my chinuch in how Bnei Torah should relate to each other is the law of predators in the jungle.

2.   Another point- the Gemara and Rashi in Megilla 10a indicates that to have the dinim of a walled city regarding sending out a metzora requires an act of kiddush.  It's not automatic; Beis Din, or someone, would have to be mekadesh the city as a mukaf choma.  As the Chazon Ish (OC end of 153) points out, the Gemara in Erchin says that if a walled city fell to the Leviim they would have to knock down the wall.  According to the Gevuras Ari/Liflagos Reuven, that wouldn't be true.  They could keep the wall, just don't be mekadeish the city.  The Chazon Ish answers that Beis Din was mekadeish all the walled cities, and only afterwards did they distribute cities to the Leviim with a goral.  This is a good teretz only if the kiddush is just kedushas peh (Turei Even Megilla 10 and Tosfos Arachin 32b,) because that can be done wholesale.  If an act is required (Rashi Arachin 32b, but see Tosfos there regarding a stirah in Rashi,) it's not mistavra that they did the kiddush on every ir choma before the Goral.

**********************

After I said this to my shiur, someone (Banny Singer) came over to me and asked, how can it be that Arei Miklat and Arei HaLeviim cannot be walled cities, when in the parsha of Arei HaLeviim it says

ומגרשי הערים אשר תתנו ללוים מקיר העיר וחוצה אלף אמה סביב ?

I couldn't tell him that קיר doesn't mean a wall, because Onkelos translates קיר as כותל.  The Targum "Yonasan" has a girsa issue- one girsa is חזור משור קרתא ולבר, like Onkelos, but the other is חזור מקרתא ולבר without the שור.  But you can't rely on our Targum Yonasan, doubly so when there are girsa issues.  

One might answer that while it says קיר/כותל/שור, not every city with a wall has the dinim of a walled city.  For example, in Megilla 5b the Gemara says that it's not called a walled city if the houses are next to each other and comprise the city wall.  It has to be a wall built separate from the houses.

אשר לו חומה ולא שור איגר.

Rashi-

חומה ולא שור איגר - בבתי ערי חומה כתיב עיר חומה ולא עיר שאין לה חומה בפני עצמה אלא מוקפת בתים סמוכות זו לזו וחומות חיצונות של בתים נעשות חומה לעיר והיינו שור איגר שגגותיה חומותיה גג מתרגמינן איגר.

Similarly, if the wall was built after the houses were built, it doesn't have a din or kedusha of a walled city.

But it's a terrible dochak to limit the words of the passuk to these odd cases.  Furthermore, as the Mishna LeMelech I cited above says, it's not mistavra that a city that doesn't have the din of "Ir Choma" regarding metzora and megilla would not have a din of Ir Choma regarding miklat, at least according to the Mishna Lemelech/Radvaz practical svara I mentioned above.  (According to the Gvuras Ari/Liflagos Reuven technical svara, the definition of Ir Choma, of course, would be identical for both.  But that wouldn't explain why the Torah would davka mention a wall.)

More fundamentally, one could say that this is a problem endemic to the sugya, because the whole concept that אינו נחלט לסוף שנה in the cities of the Leviim only applies to Ir Choma, which is why the Gemara in Arachin has to find a case where the city had a wall, the Levi sold the house, then they knocked down the wall, as was necessary.  So if the Torah gives cases where there's a wall, although it needs to be knocked down, maybe the same is true here.  But I don't like that teretz.  We're talking here about measuring the migrash, and talking about walls does nothing but create a distraction.

I'd like to think that קיר is the wall of a house, here meaning the outermost house that defines the end of the city (or at least the point where the karpaf of the city begins,) while the wall of a city is a Choma.  But I can't prove it. 

Eli sent in this response, slightly edited:

This (question is) asked by several contemporaries (R. Avraham Shapira, R. Steinman, R. Ch. Kaniewski), but none provide a convincing answer.... I then posted the (question) in Otzar Hachochma, and one of the participants referred me to Rasag who translates קיר העיר  as חאיט
While I have no idea what חאיט is, the (writer) referred me to an article discussing the structure of the Karaite neighborhood in Y-M (9th-10th century).
http://www.ybz.kotar.co.il/KotarApp/Viewer.aspx?nBookID=23677466#13.1696.6.fitwidth
There (p.11) he brings the Kataite scholar Yefet ben Ali (1-2 generations after Rasag, quoted often by Radak and Ibn Ezra) who also translates קיר העיר as חאיט, but also adds the clarifying comment 'whether it's a wall or something else'
So, if we accept Yefet as משיח לפי תומו, the term חאיט  which Rasag (and Yefet) use to translate קיר העיר, is not necesarily a wall.

Eli was kind enough to send me images of Rav Kanievsky's and Rav Shapiro's words.  (Rav Steinman's remark doesn't advance the discussion.)  Both of them say that the word Kir must refer to the walls of the outermost houses.

Rav Shapiro, מנחת אברהם א page 123.

Rav Kanievsky, נחל איתן page 108.

These mefarshim are indeed saying that the word קיר as used here means the wall of a house, here meaning the outermost house that defines the end of the city (or at least the point where the karpaf of the city begins.)

Eli also cited an article that discusses Rav Sadia Gaon's unusual word.  Here is the section that addresses that word.

Earlier, I suggested that the word always means something other than a city wall, while a proper wall of a city is a Choma.  I had no proof of the distinction, other than the obscure word used by Rav Saadia Gaon.  Since then, I came across a something written by a friend's son- Harav Aryeh Leib Keller of the Dirshu chabura in Lakewood, an extraordinary young talmid chacham.  He brings Tosfos Ri'd in his peirush on Eicha (2:8), where he says

 חל וחומה: חו­מה היא גבוה, ו״חיל״ היא קיר נמוך בצדו לחוזק

We see here a clear distinction between חו­מה and חו­מה ,קיר being a tall enclosure, while קיר refers to a short wall.

He points out that while Onkelos translates מקיר העיר to mean "מכותל דקרתא,", he always translates חומה as "שור," as in Vayikra 25:29, where he translates עיר חומה as  "קרתא מקפא שור," and, there in 25:31, as "מקף שורין."  As I mentioned above, however, the Targum "Yonasan" does say שור here, but that does not weaken the proof from Onkelos.

He also brings the Chazon Ish (OC 110:29) as follows:

 ואמנם אם כל העיר מוקפת מחיצות נותנין עיבור חוץ לחומה לר״מ דנותנין קרפף לעיר אחת אע״ג דיש אויר בין הבתים לחומה כן מבואר בע״ה ובמ״מ פכ״ח ה״ה, ולכאורה האויר שבין חומה לבתים דינו כחצר ואי חצר אינו נמדד גס אויר

 החומה אינו נמדד מן העיר אבל י״ל דאויר שכל בני העיר משתמשין כו טדיף וכמש״כ ריטב״א בנחל שיש כו דקה דמודדין

 משפת השני של הנחל, וכתב ריטב״א דאע״ג דאין מתעבר אלא דירה ומקום דירה מ״מ נחל עדיף שכל בני העיר משתמשין ואינו דבר מסוים נוח הוא ליבטל להעיד וה״נ י״ל אויר שבתוך החומה, מיהו התם י״ל לרכנן קיימינן ויהבינן קרפף וכמש״כ לעיל, אכל הכא יהבינן קרפף חוץ לחומה נר״מ, כמש״כ המ״מ,

וקרוב הדבר לומר דדוקא חומה המגינה על העיר ולפיכך תוך החומה מקרי עיקר העיר אבל מחיצה במסיפס בעלמא לא, ולענין בתי ערי חומה ודאי דוקא חומה והא דמספקא לי׳ מגילה ה׳ ב׳ בטברי׳ אי משוס מגליא או משום דלא מיגנא ודוקא לענין קריאה המגילה אבל לענין בתי ערי חומה ממעטינן טברי מ״מ בלא מגליא לחוד לא סגי אלא תרתי בעינן לא מגליא ומגני, וכדכתיב דברים ג׳ ה׳ כל אלה ערים בצורות חומה וגו׳ לבד מערי הפרזי, אלמא דבעינן דמיגני

 וכיון דהגגה מלתא היא י״ל דהא דמודדין ממומחה למיהב עיבור ע׳ אמה חוץ לחומה הוא דוקא בחומה מקפת ומגינה.

The value of the Chazon Ish for this discussion is that despite the Gemara in Arachin that states unequivocally that the Arei HaLeviim cannot be "Walled Cities," it could be that not every perimeter barrier would automatically make a city a "Walled City."  The Chazon Ish has given us a factor in the legal definition of "Walled City," (completely unprecedented in Chazal,) such that a city might have a wall, a קיר העיר, but it does not make it a "Walled City," and therefore it can be a city of the Leviim.  It stands to reason, of course, that a wall that was not built for defense, that was built only to define a perimeter or impede casual trespass or funnel traffic, would be much smaller and less substantial.  This is, then, perfectly consistent with the Tosfos Ri'd's distinction between קיר and חומה, and it also explains Rav Saadia Gaon's choice of the unusual word חאיט.