▪ RMN_1 ‘…’ ↗… ▪ RMN_2 ‘…’ ↗…♦ Semantic value spectrum in ClassAr (acc. to BAH2008): ‘to mend, to patch up; animal halter; to decay, decayed remnants, rotten and decayed bones.’ – The word rummān is classified by the philologists under this root and also under root RMː(RMM).
▪ Jeffery1938, 144-45: »The generally accepted opinion among the Muslim authorities is that it is a form fuʕlān from ↗√RMː (RMM) (cf. Rāghib, Mufradāt, 203), but some had considerable doubts about it as we see from LA, xv, 1 48; and Jawharī, sub voc. – Guidi, Della Sede, 582, noted it as a loan-word in Ar, and Fraenkel, Fremdw, 142, suggested that it was derived from the Syr rūmanā, the Ar form being built on the analogy of tuffāḥ. As the Eth [Gz] rōmān and the Phlv ideogram rōramnā or romanā,1
are of Aram origin we may assume the same for Ar rummān, but the ultimate origin of the word is still uncertain.2
It occurs in Hbr as רמון, in Aram רימונא and רומנא, as well as Mandaean רומאנא,3
but appears to be non-Semitic.4
Horovitz, Paradies, 9, thinks that if it is true that the pomegranate is a native of Socotra we may have to look in that direction for the origin of the word. It is, of course, possible that it is a pre-Semitic word taken over by the Semites. (See Laufer, Sino-Iranica, 285.)« ▪ Orel&Stolbova1994#2122: From Sem *rimān‑ ‘granate’ < AfrAs *riman‑ ‘fruit’. If Eg rrm.t ‘fruit’ (NK) really is cognate, then it would show both assimilation of liquida and metathesis.