Etymology

The absence of palatalization of /k/ here is often attributed to some form of ON input, cp. OIcel bekkr 'bench' (cp. Dan bænk without consonant cluster assimiliation) < PGmc i-stem n. (gender varies in WGmc) *banki- (cp. Burg *banki, OE benc, OFris benk, bank (fem., masc.), OS bank (fem.), OHG banc (masc., fem.)), which is usually connected (see Orel, Kroonen) to a PGmc masc. n. *bankōn 'elevation' formed on the same root (cp. bakki ‘bank (of river, lake), shore; slope, escarpment; edge (etc.)’, LG bank 'bank, shore'). The N distribution of this form and its earliest attestion in Orrm (where as Bj. notes, the native form also occurs) may be taken as further circumstantial evidence in favour of interpreting the non-palatalized form as influenced or reinforced by ON, if not derived from it.

PGmc Ancestor

*ƀanki-

Proposed ON Etymon (OIcel representative)

bekkr 'bench'
(ONP bekkr (1) (sb))

Other Scandinavian Reflexes

Far bekkur, bonkur, Icel bekkur, Norw bekk, Dan bænk, Sw bänk 

OE Cognate

benc 'bench'

Phonological and morphological markers

[absence of palatalization of */k/] (possibly diagnostic)

Summary category

CC2ac

Attestation

First attested in Orrm (where the native variant also occurs) and then mostly from N texts subsequently in ME; N and Sc in modern dial (though EDD also cites one instance from Nhp.)

Occurrences in the Gersum Corpus

WA 625, 2927, 4236 etc.

Bibliography

MED benk (n.) , OED3 bink(n.) , HTOED , EDD bink (sb.), Bj. 145, de Vries bekkr, Mag. bekkur (1), Orel *bankiz, Kroonen *banki-, DOE benc, AEW benc