Is it Rashi who says the krei should really have been לה?

Summary: Explaining one of Rashi's comments, and considering whether Rashi really said it.

Post: In parashat Behar, a pasuk states:

30. But if it is not redeemed by the end of a complete year, then that house which is in the city that has a wall, shall remain permanently [the property] of the one who purchased it throughout his generations. It will not leave [his possession] in the Jubilee.

ל. וְאִם לֹא יִגָּאֵל עַד מְלֹאת לוֹ שָׁנָה תְמִימָה וְקָם הַבַּיִת אֲשֶׁר בָּעִיר אֲשֶׁר [לא] לוֹ חֹמָה לַצְּמִיתֻת לַקֹּנֶה אֹתוֹ לְדֹרֹתָיו לֹא יֵצֵא בַּיֹּבֵל:

Rashi explains {my translation}:

		וקם הבית וגו' לצמיתת: יצא מכחו של מוכר ועומד בכחו של קונה:
asher lo` chomah -- we read it lo {to it}. Razal say that [this is the law] even though it does not have it now, since it had it beforehand. And ir, 'city', is feminine, such that it should have required {a krei of} לה, lah. But, since it needed to write the [consonantal] text of לא within, they established לו is the masoret, this falling upon that {since both may be pronounced identically}.		אשר לא חמה: לו קרינן, אמרו רז"ל אף על פי שאין לו עכשיו, הואיל והיתה לו קודם לכן. ועיר נקבה היא והוצרך לכתוב לה, אלא מתוך שצריך לכתוב לא בפנים, תקנו לו במסורת, זה נופל על זה:
		לא יצא ביבל: אמר רב ספרא אם פגע בו יובל בתוך שנתו לא יצא:

Shadal approves of this Rashi. He writes:

אשר לא חומה : יפה כתב רש " י ( בקצת ספרים ) שהתקינו לקרוא " לו " להיותו קרוב במבטא למילת " לא " שבכתיב , אך לפי האמת היה ראוי לקרוא לה ( בלמ " ד קמוצה ובמפיק בה " א ) חוזר לעיר .

Indeed. Note the reference to it occurring in קצת ספרים. Apparently, not every girsa of Rashi has this. In fact, here is one manuscript, from Rome, 1470, that lacks it:

Thus, it has note of the krei and the derasha on this phrase found in the Sifra (and the gemara in Megillah, etc.). But there is no note of the feminine nature of the word ir, such that we should really expect lah.

I will note that the Sifra and gemaras do not explicitly relate the derasha to the ketiv / krei switchoff. But Rashi certainly appears to do so.

How about a girsa of Rashi which does have it? Munich, 1233:

However, if you compare to other Rashi texts, you will see that this dibur hamatchil is out of order. Asher lo chomah is in the middle, yet Rashi comments on it in this manuscript last. Add the general tendency of this particular early manuscript to add derashot from elsewhere in Chazal and its own peshat interpretations, and I would take the fact that it is out of place as an added indication of modification from the original text of Rashi.

Yosef Daas often discusses girsology in Rashi, so it is a good idea to check him out on something like this:

Thus, he modifies the manuscript text, inserting the word הואיל. This seems right. He also writes that afterwards is the dibbur hamatchil of לא יצא ביובל. This is another correction, since the manuscript put it out of order, as I discussed above.

The Samaritans adopt the krei as their ketiv:

Yet they don't change it to לה. They thus adopted the krei, which made more sense to them, and emended their consonantal text to match, but did not dwell upon it enough to realize the 'mistake' that 'Rashi' points out.

Aharon the Karaite says that the krei is the ikkar:

It strikes me as somewhat funny that the Karaites should prefer the krei, which is an Oral tradition, over the written tradition of the text. Though I suppose that one could deem it appropriate that they consider the Mikra as primary to the Masores.

This is not the same as their typical rejection of derashot Chazal, though. They could readily say that this is no derasha, but rather an irregularity in the spelling, while the sound is the same.

I have more to say on a related topic, though I don't know if I will get to it. The topic: what is the nature of the derasha in Sifra? Is it based on the krei vs. the ketiv, or something else? What might the two other derashot on the same phrase reveal about this? Do I agree with the Yerushalmi which makes the two competing derashot about the krei vs. ketiv? Or will I try to make it a simple derasha about לו coupled with the irregular אשר, as opposed to the earlier phrasing as עיר חומה? Also, is the consonantal text לא or לא, and Rashi vs. Tosafot on this, and variants texts listed in Vetus Testamentum which support each. And a claim I made way back in my year in Israel, that there actually is a difference in pronunciation between לו and לא, since one is a cholam malei and the other is a cholam chaser, such that it makes sense to speak of krei and ketiv here. As you can see, this would be quite a handful. So, I'll leave off here, perhaps to return. But I'll refer you to this rather nice devar Torah and analysis of parts of the inyan by Ephraim Stulberg.