ONlwSG

gearrbhall m., gen. gearrbhuill, 

This is a reconstructed form, based on AFB^˄’s own gearra-bhuill; the genitive is otherwise unattested.

‘great auk, Penguinus impennis (now extinct)’. Martin (1698, 100) cites the term gairfowl and Macaulay (1764, 156) garefowl, which Swann (1913, 93) takes to be from ON geirfugl m. ‘idem’, so also Sommerfelt (1952a, 230; 1952b, 375) and McDonald (2009, 357). Forbes (1905, 30, 234) cites Ice. geyr(-)fugl (sic) and Henderson (1910, 129 fn 3) Ice. geyrfugl (sic) and geirfugl.

While Lockwood early on sees gairfowl, garefowl as an anglicisaton of a Scottish Gaelic reflex of ON geirfugl, he later concludes that the second element of a Scottish Gaelic form gearrbhall would be unlikely to develop into Eng. fowl, instead proposing that ON geirfugl yields SG gearrbhall via an unattested Orkney ‘Norn’ intermediary, and that consequently ‘gairfowl, garefowl can be interpreted as essentially Orkney Norn, with gair-, gare- regularly corresponding to ON geir- (Marwick 1929, xlii) and -fowl as the expected anglicisation of -fugl’ (≈Lockwood 1978, 396).

ON geirfugl 

From ON geir, the stem form of geirr m. ‘spear; point of an anvil’ + fugl m. ‘bird’.

would formally yield EG *[ˈɡʲeːɾ ̩fuːɫ̪] > SG *gèarful *[ˈɡ̊ʲeːɾ ̩fuɫ̪] or *[ˈɡ̊ʲeːɾ ̩fəɫ̪], but a likely association with bird names in SG gearr/geàrr/gearra- and perhaps a folk etymological connection with SG ball ‘spot, mark’ might almost inevitably see a further development to gearrbhall: the bird had a white spot by each eye. Open compounds such as SG gearra-breac ‘black guillemot’ and gearra-gort(a) ‘quail’ consist of an initial weakly stressed generic element + a fully stressed adjectival element.

Here with gearra, a compound form of SG geàrr, EG gerr (possibly a nominal form of the adjective geàrr (gerr) ‘short, ?mincing, ?quick’), + breac adj. ‘speckled’ and the genitive of gort ‘field’, respectively. EG gerr initially yields SG gearr, with its short vowel + long consonant developing in most dialects into gèarr, geàrr, with a long vowel + short consonant (cf. SGDS Item 470: geàrr ‘to cut’), e.g. geàrr-chlamhan ‘buzzard’ (cf. AFB^˄), but the short vowel is retained in stressed epenthesis (svarabhakti) environments, e.g. gearrbhonn ‘the outer sole of a shoe’ (cf. Dieckhoff 1932), and when followed by a vowel, e.g. gearrachlamhan (cf. Wentworth 2003, s.v. kite). Note that there would be a tendency to shorten the long vowel of forms such as geàrr-chlamhan by virtue of its weakly stressed position.

The stress pattern of ON geirfugl, however, accords with the stress pattern of closed compounds in Gaelic, in which a fully stressed adjectival element is followed by a lenited, weakly stressed generic element, hence notionally ˈgearrˌbhall, but probably ˈgearrbhall. The form gearrbhall occurs in two St Kilda songs (≈CG IV, 110, V, 46): M’ eudail thusa, mo lur ’s mo shealgair | Thug thu ’n-dè dhomh ’n sùl ’s an gearrbhall (‘You are my darling, my love and my hunter, | You gave me yesterday the gannet and the great auk’) and Bheir thu ’m fulmair ’s bheir thu ’n gearrbhall, | ’S bheir thu ’n sgarbh à calg an rubha (‘You’ll bring the fulmar and you’ll bring the great auk | And you’ll bring the cormorant from the tip of the headland’), with the rhymes shealgair : gearrbhall and gearrbhall : sgarbh indicating a nominal *[ˈɡ̊ʲɑɍ͡ɑvəɫ̪].

Although *[ˈɡ̊ʲɑɍ͡ɑvɑɫ̪] would seem to be a likely outcome given the folk etymological connection with SG ball and the consequent spelling.

A spelling gearrabhall, on the other hand, suggests a nominal *[ˈɡʲɑɍəvəɫ̪], perhaps on the model of gearra-breac etc.

Alexander Carmichael (collector of CG) gives gearrabhul in Harvie-Brown and Buckley 1888, 158. Fergusson (1886, 82) has gearbhal; Forbes (1905, 30) gearrbhall, gearrbhuil, gearbhal; CG II, 284, gearr-bhall and gearra-bhall; Dwelly (1911) and AFB^˄ gearra-bhall; Cunningham (1990, 135) garra-ball, in error; MacFhearghuis (1995, 59) and Garvie (1999, 63) gearra-bhall.

Pace Lockwood, it seems quite plausible that Martin’s gairfowl and Macaulay’s garefowl are partially folk etymologically driven scotticisations of SG gearrbhall, with its second element equated with Scots foul, fowl or Eng. fowl ‘bird’; indeed, Macaulay goes further in suggesting garefowl was a corruption of rarefowl, an explanation perhaps promoted by the bird’s relative scarcity: ‘The St Kildians do not receive an annual visit from this strange bird ... It keeps at a distance from them, they know not where, for a course of years’ (Macaulay ibid., 157).

Harris and Murray (1989, 27) state that the great auk bred in St Kilda ‘in the 17th century’. The last St Kilda bird is said to have been caught on Stac an Àrmainn ‘the stack of the hero’ (for the specific, see SG àrmann) NA151064 in 1840 (Harvie-Brown and Buckley 1888, 159), although MacKenzie (1911, 39) implies the bird was already extinct there by that year, and Wiglesworth (1903, 37–38) dates the killing on Stac an Àrmainn to between 1830–1835.

Scots gair-, gare-, however, suggests *[ɡeːr]-, from early SG *[ˈɡ̊ʲeːɾ]-.

The folk etymological explanation that gearrbhall derives from SG geàrr ‘short’ and ball ‘spot’ is seen in Alexander Carmichael’s translation ‘the squat spotted one’ (CG II, 284; so also Forbes), although gearrbhallach, with nominal use of the adjectival suffix -ach, might be expected (cf. Black 2006^˄). This may have influenced Carmichael's spellings gearr-bhall and gearra-bhall, and appears to be perseverated in AFB^˄’s phonemic transcription /gʲaRəvaL/, which suggests a reduced form of (lenited) ball /bɑuʟ/.

In summary, SG gearrbhall (now generally gearrabhall) appears to be a loan-shift (with complete morphemic substitution) from ON geirfugl, while Scots gairfowl, garefowl appears to be a loan-blend (with part phonemic substitution and part morphemic importation) from *gèarful, an earlier Scottish Gaelic reflex of the Old Norse word.