The Georgian word for “silk” is “აბრეშუმი” (abrešumi), which the Georgians acknowledge is from the Persian “ابریشم”. The Georgian pronunciation implies that in Persian it used to be “abrêshum”, which would mean it ought to be “abrishom” today, but it is usually not: The most common pronunciation is “abrisham”, although in Tajikistan the official pronunciation is “абрешим” (abrêshim), which would be “abrishem” in Tehran.

Nişanyan claims that this word derives from the Avestan “upa rēshma”, “upa” being cognate to “به” (be) in modern Persian (“bi” in Kurmanji, “ба/ba” in Tajikistan), “rēshma” to Persian “رشته” (Kurdish “rêze”).

If that’s the case, however, it doesn’t explain why Kurmanji speakers refer to silk as “hevrîşim”. “Hupa” doesn’t appear to be a form of the Avestan prefix in question, although “Etymological Dictionary of the Iranian Verb” agrees with the etymology that Nişanyan gives. It could be that this is similar to the case of the phantom “h” in the Kurmanji “hesp”, from the disputed PIE root *hewkos/*ekwos (the former possibility helps explain the breathy voice in the Greek word which became (through Latin, where it has an “h” due to the Greek “breathy voice”, which also appears on all Greeks words with initial “r”*, hence the “rh” in “rhythm”) “hippo”. But if this word really did always start with an “h”, how come this sound was only preseved among Kurds, especially considering the cognate root in Kurdish doesn’t even start with an “h”? Further questions are raised by the sporadic appearance of the word “هبریشم” (four results on my google search) AND the fact that as with “hesp”, Sorani speakers do not follow the Kurmanji lead with the word-initial “h” in “hevrîşim” (“ئه‌سپ” and “ئاوریشم”, respectively, note that most of the vowels there are identical to the Kurmanji accent, however the initial vowel after the missing “h” in “ئاوریشم” would be Kurmanji “a”, not “e”).

*Similar to how word final “r” is always devoiced in the Turkish of Turkey.