ONlwSG

urrachdag f. [ˈuɍəxkaɡ̊], [ˈuɍaxkaɡ̊], [ˈuɍɔxkaɡ̊], gen. urrachdaig -[æɡ̊ʲ], -[ɛɡ̊ʲ], ‘thole-pin’ is derived by Henderson (1910, 149) from ON urga f. ‘the end of a rope’ (NO). Henderson says that ‘the Gaelic form urrachdaig [sic] is a double diminutive: urr-ag-ag, contracted to urcag, and that the plural occurs in SG na h-urrchdagan maide 

With SG maide ‘wood, timber’ in genitive position.

[sic] “the thole-pins”.’ 

De Vries (1962) cites Henderson, and McDonald (2009, 425) considers the loan likely, although he suggests that ON urga via ur(ǝ)ga [sic] yields urrac(h)a + g (from the Gaelic suffix -ag), but this is improbable as ON g in urga represents fricative [ɣ]. Oftedal (1962a, 119) is extremely doubtful of a connection.

There are several issues with this analysis: urrachdaig is either a misprint for urrachdag or a (perhaps normalised) dative form; a form *urragag would be expected, via syncope, to yield *urrgag *[ˈuɍ͡uɡ̊aɡ̊] or *urgag *[ˈuɾ͡uɡ̊aɡ̊], not urcag ?[ˈuɾ͡uʰkaɡ̊] (see below), although syncope of the clear vowel in -ag (originally a long vowel: EG -óc) is unlikely; and the consonant cluster in urrchdag is otherwise unattested and the form is probably in error for urrachdag — at any rate, ON urga would be formally expected to yield SG *urgha *[ˈuɾ͡uɣə] or *urrgha *[ˈuɍ͡uɣə] (Cox 2022, 273–76). In Henderson’s defence, his explanation may have been obfuscated by an attempt to incorporate and/or account for variant forms; there are:

A. SG uracag, urracag
A 1. SG uracag [ˈuɾəʰkaɡ̊]
Attested in the plural uracagan in Dwelly 1911.

A 2. SG urracag [ˈuɍəʰkaɡ̊]
In MacKenzie 1841, 376; MacLennan 1925; AFB^˄: /uRəxgag/; An Stòr-Dàta 1993.

B. SG urachdag, urrachdag
B 1. SG urachdag [ˈuɾəxkaɡ̊], [uɾaxkaɡ̊], [uɾɔxkaɡ̊]
In Armstrong 1825; (pl.) urachdagan in Dwelly 1911: for uracagan (A 1).

B 2. SG urrachdag [ˈuɍəxkaɡ̊], [uɍaxkaɡ̊], [uɍɔxkaɡ̊]
(i) urrachdag (Armstrong 1825: see urachdag; Faclan bhon t-Sluagh^˄: Coll);
(ii) urrachdaig (Henderson 1910, 149), ?in error for or a normalised dative form of urrachdag (i).

C. SG urcag ?[ˈuɾ͡uʰkaɡ̊]
In MacBain 1911: N. Lochabar (?= Nether Lochabar); MacLennan 1925.

Most sources give the sense of urrachdag etc. as ‘thole-pin’, except for MacLennan (1925), who for urcag gives ‘thole-pin’ but for urracag gives ‘thole-pin, timber-head’ – timber-head referring to ‘the head or end of any timber, specifically such an end rising above the deck and serving as a bollard; (rarely) blockhead, idiot’ (≈OED^˄). Urrachdag etc. conceivably derives from Scots (Shetland) urek (Foula) [urək], (otherwise generally) [ūrək] ‘a very small creature or thing; a little child; a tiny fish; a stunted creature (fish or young animal, occasionally of a person); a small part (portion) of something’ (Jakobsen 1928) — with the addition of the Gaelic suffix -ag. As well as the dialectal variation indicated by (A 1, B 1) r ~ (A 2, B 2) rr, there is variation in the degree of preaspiration indicated by (A) ur(r)acag ~ (B) ur(r)achdag. The phonetic value of (C) urcag is probably [ˈuɾ͡uʰkaɡ̊], possibly [ˈuɾ͡uxkaɡ̊]. Cf. similar variation in SG currac ~ currachd ‘cap, woman’s headdress’, and the derivative curracag (curraiceag) 

E.g. AFB^˄. The Easter Ross pronunciation /kuruɡɑɡ/ (cf. Watson 2022, 333) would seem to indicate its having been borrowed from a full-form svarabhakti dialect area, cf. fn 6, below (pers. comm. Professor Seòsamh Watson).

~ currachdag 

E.g. Forbes 1905, 25.

~ curcag 

E.g. Dwelly 1911: see curacag.

~ currcag 

E.g. Oftedal 1956, 142: /kuʀùkaɡ/ [i.e. [ˈkʰuɍ͡uʰkaɡ̊]], Lewis.

‘lapwing’.

MacBain (1911) derives SG currachd (in the sense ‘cap’) from Scots curch /kʌrtʃ/ ‘kerchief’, and curcag (in the sense ‘sandpiper’) from EG cuirrech ‘marsh, fen’ (eDIL^˄), but one might compare EG cuirce ‘?crest, tuft, headdress’ (eDIL^˄). Both MacBain (1896) and Dwelly (1911) derive SG cuircinn ‘a kind of woman’s headdress’ from Scots curch, but, given EG cuirce, above, MacBain (1911) is less sure.

^,

A development of SG urrachdag from EG airrecht ‘act of approaching, seeking out, making for’ (eDIL^˄) + the suffix -óc, while phonetically plausible (cf. EG ailaḋ, dat. ailaiḋ ‘tomb’ > Ir. ula ‘idem’, SG ulaidh ‘treasure’ (O’Rahilly 1929, 66)), would seem semantically implausible.