Etymology

ME gres (as opposed to the more common ME gras, PDE grass) has often been explained as showing the vocalism of an ON *gres, cp. OIcel -gresi (as in e.g. illgresi ‘weed’), OSw græs, Sw gräs, ODan græs (thus TGD, Jordan-Crook §§32 Rem. 3, 34 Rem. 1, McGee 334, 517, 519,  Bj. (DP 24 n.1), Kluge 1901: 938 (citing Brate 67); Luick §382.2 and Serjeantson 83 regard the vocalism as distinctive of loan from OEN); this is apparently the reflex of a PGmc *grasja-, a collective formation on the PGmc *grasa- which gives OE gærs, græs (etc.) and Go gras, OIcel gras, OFris gers, gres, OS, OHG gras. On the other hand there are a number of forms spelt with <e> in OE texts, esp. with metathesis (notably the common late Nhb gers) but also without (see Sievers-Brunner §84 Anm. 3 and DOE), and thus an Angl. gres may have come about by native processes.  There is also the possibility of a ME change /a/ > /e/ before /s/ (see kest (v.)) (Morsbach §87).

PGmc Ancestor

*grasa- 

Proposed ON Etymon (OIcel representative)

-gresi (e.g. illgresi ‘weed’)
(ONP gras (sb.))

Other Scandinavian Reflexes

 ODan græs, OSw græs, Sw gräs

OE Cognate

OE gærs, græs (etc.) incl. Nhb gers, gres 'grass'

Phonological and morphological markers

Summary category

CC1ac

Attestation

Gres- is attested in ME from Orrm onwards and especially characteristic of N and E texts, but not exclusively (e.g. Chaucer TC 2.515), and often spelt <gres> when it occurs as a loan in medieval Latin.  

Occurrences in the Gersum Corpus

Gaw 235, 527, 2181; Pe 10, 31, 245; Cl 1028; WA 338

Bibliography

MED gras (n.) , OED grass (n.1) , HTOED ,  Dance gres, Bj. 30n.1, 211, 286, DP 24 (n.1), Falk-Torp græs, Hellquist gräs, Nielsen græs, de Vries gras, -gresi, Mag. gras, -gresi, Bj-L. gras, Orel *ʒrasan, Kroonen *grasa-, AEW græs, gærs, DOE gærs, græs, EPNE gærs, græs, gres