Publishing history:v1.0
v1.0: 29/04/26
rùm  
While the spelling rum accommodates a greater range of pronunciations, the spelling rùm is adopted here as it is in common use.
Cf. /Ruːm/ (AFB˄), /Rɔ̃ũm/ (Oftedal 1956, 93: Lewis).
Cf. Shaw 1780; Mac Farlan 1795; Armstrong 1825; HSS 1828; McAlpine 1832; MacEachen 1842; Dwelly 1911; AFB˄.
McAlpine (1832) notes the origin of SG rùm as ‘Teut[onic]’.
E.g. Campbell in McDonald 1972, s.v. fo. NB AFB˄ gives an extended sense of the phrase in thig fo rùm ‘go to (your own) bed, turn in!’
So also McDonald (2009, 396–97).
Thomson loosely (but incorrectly) translates ON rùm [sic] as ‘rowing bench’, so also McDonald (2009, 396–97). Cf. SG fo lobhtaidh ‘below deck’ ([An] Dughalach 1829, 192), s.v. lobhta.
For SG rùm, in addition to ‘room: space; room: chamber’, Dwelly (1911) gives the sense ‘floor’. This is from Armstrong (1825), who draws on Lhuyd’s (≈1707), if not directly on O’Clery’s (1643), Irish dictionary. The latter gives ‘rúm .i. eadarfholamh, no urlár[:] rúm na raṫa .i. urlar na ratha etc.’ [‘rúm, i.e. an intervening space or floor: the rúm of the boat, i.e. the floor of the boat etc. (for the sense ‘boat’, see eDIL˄)’].
Lhuyd’s (1707) own entry reads ‘†rúm, eadarfholamh, nó urlár: rúm na ratha, i.e. urlár na ratha, a floor. Q. antræ, crann-rum, ruma no urlaire idir chrannaibh Pl[unkett]’. Lhuyd rightly questions his interpretation of Plunkett (≈1662), whose own (Latin-Irish) dictionary reads ‘antræ, rectius ancræ [a valley or gorge] “crannrúma, rúma m.c. no urláir idir chrannaibh”, for which compare Thomas’s (≈1593) Latin-English dictionary’s ‘antræ “spaces between trees: also valleyes”. Rectius ancrae.’
Stokes (ibid.) also asks whether MG rúm might be a genuine Celtic word, but the question is crossed out in ink in Marstrander’s copy of Stokes 1909˄.
MG rúm appears as rumh and ruma in In Cath Catharda (Stokes 1909, 216.2814 and 2838), hence Schulze-Thulin (1996, 107) lists ON rúm /ruːm/ as yielding MG rúm(h) /ruːμ/. However, mh is no doubt in error for m [m] /m/. Stokes (p. 541) describes rúma as a ‘sister form’ of rúm: the context is (p. 216) [r]otuitsetar іаrum cenn i cenn 7 taeb re taeib i ruma na rat[h]a iarsin, and, while Stokes translates (p. 217) ‘[t]hey tumbled head against head and side against side on the floor of the raft’, we should probably translate ‘they then thereupon fell together into the rooms of the vessel’, with accusative plural 
As suggested by Marstrander in his copy of Stokes 1909.
eDIL˄ cites two further instances of MG rúm. Firstly, the Book of Leinster’s version of Togail Troi (‘the destruction of Troy’) contains the phrase ruchtgal na rúm ica ruth (TCD MS 1339, p. 236 a 16; Best and O’Brien 2006, 1100.32229), the context of which is a run of mainly alliterative phrases descriptive of a storm at sea, for example briscurnach na m-bárc ica m-brissiud; sessgal na n-ethar cá n-imram; ruchtgal na rúm ica ruth; trethan na tét icá timsugud, which Stokes (1881, [35], 100, 176) translates as ‘the crashing of the barques a-breaking; the sound of the seats 
I.e. from sess ‘rowing bench’ + gal (gol) ‘wailing etc.’ (ibid., 178).
Cf. sisarnach ‘hissing etc.’.
Cf. eDIL˄, s.v. ruth ‘the act of overthrowing, casting down, breaking?’.
With metaphorical use of trethan ‘(stormy) sea’ (cf. eDIL˄, s.v.).
MG rúm, and consequently usage in SG fo rùm ‘below deck’, may indeed derive from ON rúm in the sense ‘one of the spaces between a ship’s ribs accommodating a pair of oarsmen’, although broadly the same sense occurs in both Eng. room (OED˄, s.v., III.7: ‘a space or compartment lying between the timbers of a ship’s frame, the thwarts of a boat, etc.’) and Scots room (SND˄, s.v., I.n.5: ‘the compartment or space between the thwarts in a boat’). It seems more likely, however, that Ir. rúm and SG rùm in the senses ‘room: space; room: chamber’ go back to Eng. and/or Scots room or their antecedents.
‘The term rúm is used in Donegal corresponding to Eng. room “chamber”, but not for “space (to be filled)”. It is strongly perceived as an English loan-word. An informant of mine from South-West Donegal corrected it to seamra while recounting a story.’ (pers. comm. Seòsamh Watson).
SG cùm ‘to keep’ (< MG congḃaiḋ ‘contains, maintains etc.’ (eDIL˄, s.v. con-gaib)), spelt with a lengthmark, distinguishes it from the homophonous SG cum ‘to fashion, shape etc.’ (< MG cummaiḋ ‘fashions, shapes etc.’), spelt without.
Table: SG rùm and cùm
| A Location | B SG rùm | C SGDS Item 286 chum ‘kept’ | |||
| (i) | (a) | 1. | Benbecula | [ruːm], [ruːmə] (LASID IV, 230, Item 184; 233, Item 517) | Pt 22 [xVːm] |
| 2. | Duirinish | [ruːm] (LASID IV, 265, Item 184; 267, Item 517) | Pt 103 [xVːm] | ||
| 3. | Gairloch | [ʀũːm], gen. [ʀũmə] (Wentworth 2003a, s.v. room) | Pts 121–122 [xVːm] | ||
| 4. | Easter Ross | /ruːm/, dim. /ruːmɑn/ (Watson 2022, 276) | Pts 152–161 [xVːm] | ||
| (b) | 5. | Lewis | /Rɔ̃ũm/ (Oftedal 1956, 93) [R[ȯu̇]m] (Borgstrøm 1940, 46) a [ʀuːm] (LASID IV, 253, Item 517) | Pts 1–9 [xVVm] | |
| (ii) | 6. | North Argyll | [rũ̜m] (Faclan bhon t-Sluagh˄) | Pts 60–71 [xVm], [xVmˑ] or [xVmː] | |
| 7. | Islay | a [ruːm] (Holmer 1938, 205) | Pts 53–56 [xVmˑ] (also cum [kʿVm] in Holmer 1938, 149) | ||
| 8. | Blair Atholl b | (EPG) ‘apartment’: (MS) /rumː/, pl. /rumːxˊɩn/, and ‘space, scope’: (MS) a /ruːm/, (JM) /rumː/ (Ó Murchú 1989, 392–93) (WPG) ‘apartment’: /rumː/, pl. /rumːxˊɩn/, and ‘space’: /rumː/ (Ó Murchú 2021, 308–09) | Pts 191–193 [xVmː] |
Notes: (a) The following form contrasts with that of Column C; (b) Blair Atholl: ‘EPG’ and ‘WPG’ = East and West Perthshire Gaelic, after Ó Murchú 1989 and 2021 (‘MS’ and ‘JM’ are the initials of informants). Despite the east-west designations, both sets of data originate from Blair Atholl Parish, and in this instance coincide with SGDS’s Points 191–193.
Lengthening develops in stressed syllables ending in -m in Scottish Gaelic, but normally not when followed by a vowel. On the one hand, therefore, the genitive [ʀũmə] (Table, row 4) suggests that the original radical form had a short vowel; on the other hand, the diminutive /ruːmɑn/ (row 4), the plural /rumːxˊɩn/ (row 8), as well as the radical [ʀuːm] for Lewis (row 5) and [ruːm] for Islay (row 7) (which would otherwise be expected to contain a diphthong and short vowel, respectively) indicate that the original radical form had a long vowel. The word may have been borrowed more than once, of course, as indicated by the forms /rumː/ ‘apartment’ ~ /ruːm/ ‘space, scope’ (row 8); cf. also the long-short stressed-vowel variation apparent under Derivatives.
Derivatives: SG rumail (Wentworth 2003a, s.v. roomy: rumoil) and rùmail (HSS 1828; MacLennan 1925), with the adjectival suffix -ail; rumaich (HSS) and rùmaich (MacEachen 1842), with the verbal suffix -aich; and rùmalachd (McAlpine 1832; MacLennan), with the abstract noun suffix -achd. Faclan bhon t-Sluagh˄ lists ròramach, ròromach ‘extravagant’ for Islay, noting a tentative derivation from SG ro- + rùm, a suggestion possibly drawn from McAlpine’s (1832, s.v. roram) dictionary; cf. Robertson’s (1899, 263) ròramach ‘profusely hospitable’ for Arran. While the intensive prefix ro- (or the noun rò ‘excess’) may well lie behind these forms, a compound with SG rùm seems unlikely.
In summary, then, SG rùm may have a number of points of origin: Eng. room and MScots roum with their long vowels, and Scots room with its short vowel, while the phrase fo rùm ‘below deck’ may go back to ON rúm, although the possibility that it may in fact go back to English and/or Scots usage cannot be excluded. Analogy has no doubt also played its part.