ONlwSG

langa f., 

Armstrong 1825: lang; M^cAlpine 1832; Dwelly 1911; MacLennan 1925; Dieckhoff 1932; masculine in Armstrong 1825: langach; AFB^˄.

gen. langa, 

But gen. langann, dat. langainn (McDonald 1972: South Uist).

‘ling, Molva molva’ is generally derived from ON langa f. ‘idem’; 

MacBain 1911: comparing Scots laing [leg. ling] and Eng. ling; Bugge 1912, 305: after MacBain; MacLennan 1925; Christiansen 1938, 4, 14: Lewis; Borgstrøm 1937, 79, 112, 230: Barra; 1940, 27: Lewis, 173: Barra; 1941, 36: Skye, 92: Ross-shire; Oftedal 1956, 115: Lewis; 1983; de Vries 1962: after Christiansen; Ó Baoill 1994, 171: after Bugge and MacBain; Stewart 2004, 410: after MacLennan; McDonald 2009, 380.

the phonetic development is regular.

Cf. SGDS, Item 836: teanga.

Both SG langa 

Shaw 1780; Mac Farlan 1795; M^cAlpine 1832; Forbes 1905, 44, 370; Dwelly 1911; MacLennan 1925; McDonald 1972: South Uist; Faclan bhon t-Sluagh^˄: Harris, Islay (further, see below); Garvie 1999, 84; An Stòr-Dàta 1993.

and, less commonly, lang are found.

Lhuyd 1700, 137 X.30: Argyllshire; Armstrong 1825: †lang, after Shaw, who in fact gives langa and long (see below).

The former is rendered /Laŋgə/ phonemically in AFB^˄, cf. (Benbecula) [ʟɑ̣ŋġə] (LASID IV, Item 1167, p. 239), and (Barra) [Lɑŋgɔ̆] (Borgstrøm 1937, 79, 112, 230, and 1940, 173).

Cf. also (Ross-shire) [ʟɑŋgə] (LASID IV, Item 1167, p. 271), but this may be an introduced pronunciation for this dialect (see below).

In Argyllshire, there is a tendency for the stressed vowel to be fronted and for the nasal to be lost, sometimes with a glottal stop before the plosive: (North Argyll) [ɫɑ̈̃ŋɡə] (Faclan bhon t-Sluagh^˄), (Mull) [lɑ̟̃ʔɡə] (ibid.), (Kintyre) [ʟɑ̣’kə] (LASID IV, Item 1167, p. 218), [ʟɛg̊ə] (Holmer 1962, 28), and (Islay) [ṭɛkə] [l˚ɛkə] (Holmer 1938, 182).

For the initial dental, cf. SGDS, Items 549–55: làmh, Points 53–56.

Original ng is realised as a velar nasal in Skye: [laŋə] (Holmer 1938, 182), [Lɑŋɔ] (Borgstrøm 1941, 36), as a velar fricative in Lewis: [Lɑ̃ɣə] (Borgstrøm 1940, 27), /ʟɑ̃ɣə/ (Oftedal 1956, 115), [lɑ̃ɣ̃ə] (LASID IV, Item 1167, p. 238), Ross-shire: [Lɑ̃ɣə] (Borgstrøm 1941, 92), [Lɑ̃ɣə] (Wentworth 2003, s.v. ling), and Sutherland: [ʟɑ̃ɣə] (LASID IV, Item 1167, p. 280) and, with apocope, [ɫɑ̃ɣ] (Faclan bhon t-Sluagh^˄); but vocalised in East Sutherland: (Embo) /ɫãũ/, (Brora and Golspie) /ɫɔ̃ũ/ (Dorian 1978, 155), with diphthongisation following apocope before loss of (final) ng.

Dorian (ibid.) also records a diminutive form /ɫɔ̃ũag/ for Brora and Golspie.

Apocope is also found in (Glengarry) [La(ng)g] (Dieckhoff 1932, s.v. langa).

Like SG langa, Ir. langa is generally derived from ON langa.

Meyer 1891, 461; Craigie 1894, 163; MacBain 1911; Bugge 1912, 305; Marstrander 1915a, 61, 93, 159 (so also de Vries 1962); Greene 1976, 79; 1978, 121: citing only the Irish form; Ó Baoill 1994, 171: after Bugge and MacBain; McDonald 2009, 380; Ó Muirithe 2010; and Kelly 2023, 297. Note that de Vries (1962) cites Old Irish and Schulze-Thulin (1996, 106), Mac Mathúna (2001, 76), Ó Muirithe (2010) and Robert Farren (2014, 43) Middle Irish, but the earliest attestation of the Irish word is 1684 (Hardiman 1846, 289.9).

George Farren (1946, 13) lists a number of forms according to locality; these are compared below with those in LASID: 

Dinneen (1947) gives langa, also lang (lamhang), long, longa.

Reflexes of Ir. langa

Notes: (a) Idem (Holmer 1938, 208); (b) Cf. Mx lhing, from Eng. ling (Ó Baoill 1994, 171); (c) Also /Lɑŋgə/ (Quiggin 1906, 106 §302); [langə] [?leg. [laŋgə]] (Marstrander 1915a, 159); (d) ?With a palatal initial under the influence of Eng. ling; (e) Cf. (Blasket Islands) [lauŋgə] (Marstrander 1915a, 159).

long, longa
Scharff (1916, 113) opines that Ir. longa ‘seems to be derived from the Norse längan’, by which is perhaps meant långan, the definite form of Swed. långa ‘ling’, cf. longe, a variant of Norw. lange ‘idem’ (cf. Scots (Shetland) longa-fish, a tabu name for ling (Jakobsen 1928, s.v. longa)). However, LASID’s transcriptions suggest that the forms long and longa are frequently unjustified spellings of Ir. lang and langa, and the former can probably be taken as developments of the latter, with a ~ o variation.

Shaw’s (1780) Scottish Gaelic dictionary also lists long (so also Armstrong 1825, and most probably Forbes 1905, 44, 370, and Dwelly 1911, who writes †lung ?in error). However, in spite of East Sutherland’s /ɫɔ̃ũ/, Shaw’s ‘long “the fish called ling” ’ seems likely to have been adopted from O’Brien’s (1768) Irish dictionary’s identical ‘long “the fish called ling” ’.

langach
In addition to †lang and long, Armstrong’s (≈1825) Scottish Gaelic dictionary lists langach (so also Dwelly 1911), but this is unattested elsewhere in Scotland and may have arisen from a copying error: he gives ‘†lang (Irish idem) “a kind of fish (after Shaw 1780)” ’, immediately followed by ‘langach adj. “slim, slender” ’ and ‘langach m. (Irish idem) “ling” ’, and, below, ‘long “the fish called ling (after Shaw)” ’, for which compare O’Reilly’s (1817) Irish dictionary, in which langa ‘a sort of fish’ is followed only by ‘langach “slim, slender” ’ and, below, ‘long “the fish called ling” ’.

While SG langach may be a ghost word, Ir. langach is noted by Ó Dónaill (1977) as a variant of langa, cf. George Farren’s Donegal langach (Downings) and lanach [sic] (Gortahork), and LASID’s Item 1167, (in Donegal) Points 71, [Vol. IV] p. 59: [ʟɑŋɑh] (Downings/Rosgill), 74, p. 92: [ʟɑ̣ŋ̣ɑ̣] (Gortahork), 78, p. 111: [ʟɑŋɑx] (Rannafast: ‘mullet’), and 79, p. 119: [ʟɑɴɑ̣x] (Aranmore: ‘mullet’, contrast [l´ɪŋ] ‘ling’ (ibid.)), (in Mayo) Point 53, [Vol. III] p. 253: [Lɑnɑx] ‘mullet’ (Curraun Peninsula), and (in Galway) Points 37, p. 116: [lɑŋgəx] (Carnmore) and 46, p. 194: [lɑːnəx] ‘?mackerel’ (Cárna). However, while langach in the sense ‘ling’ is a plausible formation in itself (i.e. from Ir. langa + a substantival use of the suffix -ach), there seems to be a degree of conflation with Ir. lannach ‘mullet’ (?from Ir. lann ‘scale’ + -ach, ?on account of the relatively large scales around its head and gills 

British Sea Fishing.

).

longan
George Farren’s Donegal variant longan may be the result of confusion with Ir. longan (otherwise langán, longán) in the sense ‘spent fish; what remains of a potato when seed sets are cut from it’ (Dinneen 1947), ‘discarded portion of seed potato’ (Ó Dónaill 1977).

Cf. Ulster-Scots langens (pl.) ‘the parts of cut potatoes of no use for seed’ < Ir. (pl.) langáin (Macafee 1996).

Robert Farren (2014, 34) takes langán to be a derivative of Ir. langa + the nominally diminutive suffix -án (here -an), but it is perhaps more likely to be a loan-blend from Scots ‡lank ‘a lean creature, specifically a spent herring’, with the sense ‘spent fish’ in Irish extended to ‘spent seed potato’.