Follow up to this question about Bereishit 31:53.
A word from the "eloh" root is used 3 times in that pasuk. In the Artscroll, there is the word "chol" in the margins for 2 of them, which echoes Rashi's commentary. The marking of "Chol" apparently has to do with the inherent holiness of the name which might impact a sofer, or the rules regarding erasing the text (as per the Siftei Chachamim on the rashi on the pasuk). However there is some confusion regarding the status of the the third use, the "elohei" which precedes "avihem."
There are 2 statements about the pasuk (one in Bereishit Rabbah and one in Masechet Soferim) and Rashi according to sites like Sefaria (and, by the omission of any other attribution, my 2 sets of Mikra'ot Gedolot) quotes Bereishit Rabbah for all three instances but for the third cites the statement as made in Soferim.
He quotes Bereishit Rabbah which states
אֱלֹהֵי אַבְרָהָם, קֹדֶשׁ, וֵאלֹהֵי נָחוֹר, חֹל, אֱלֹהֵי אֲבִיהֶם, מַשְׁמָעוֹ קֹדֶשׁ וְחֹל (בראשית לא, נג).
but Rashi writes
אלהי אביהם: חול:
which seems to reflect the Soferim Quote
אלהי אברהם (בראשית ל״א:נ״ג) קודש ואלהי נחור (שם) חול אלהי אביהם (שם) חול
I know that the Minchat Shai deals in some way with the difference between the two understandings (and I'm not saying I fully understand what he is getting at) but I'm curious more why Rashi would cite one but quote the other.
This is all obviated by later publishers and publishings which assign the third instance to Masechet Soferim (I have yet to find a published Chumash which attributes all three to M"Soferim), but that raises a question as to why Rashi would use one source for the first two and a different source for the third.
Certainly, this is all moot if the attribution is just a collection of later (and sloppy or arbitrary) guesswork, but I'm hopeful that there is some deeper reason or logic at work here.