1. The Harrowing of Hell (see W. Gillies, in Celtica 13 (1980), 32-55. There is a third copy in the earlier (fourteenth-century) manuscript Rawlinson B 486.); 2. Etymology of étáil (Support for the author’s suggestion that étáil is a secondary form of the vn of ad-cota; see B. Ó Cuív, in Celtica 13 (1980), pp. 125-145 (esp. 142 ff.).).
OIr. verb *tucaid / *tocaid derives from a causative verb meaning ‘to cause to become solid’ > ‘to destine’, containing the root *tenk- (‘to congeal’). Tocad (‘fortune, chance’) derives from a -to-particle of this verb.
1. An unnecessary emendation (vs. W. Stokes’s emendation of snidhfed to sn[a]idh[m]fed, in RC 12 (1891), pp. 306-08 (§14), reproduced in E. A. Gray, Cath Maige Tuired (1982); also discussion of the theme of womenfolk’s intervention in political affairs); 2. OIr. -tochus (Interprets MS toc̄sa (§67) as tochus, prototonic fut. 1 sg. of do-cing ‘steps, strides forward; advances, comes’); 3. A redactorial intrusion [in §69].
gaibid, meaning ‘provides for’, in phrases of structure: gaibid + object (‘a fixed day’) + idirect object. Discusses use in LU version of Táin bó Flidais (see LU ll. 1631-32).
Discusses is ed mod, is ing, is ar éigin ‘scarcely, hardly’, and is obair ‘it is hard, difficult’. Suggests that ModIr. fhóbair, (fh)obair ‘almost’ represents a confusion of impersonal verb fóbair and use of obair replacing earlier mod.
vs. E. P. Hamp, Journal of Indo-European Studies 1 (1973), p. 321; concludes that OIr. -op(a)ir does not contain a PIE preverb of the shape *op, and that OIr. -op(a)ir cannot be equated with Lat. offero.
Argues that OIr. -bria (subjunctive to bronnaid) is a suppletive form connected to unattested *brenaid, which can be reconstructed with the aid of Indo-Iranian from a PIE nasal present *bhri-n-eH-ti.
Calls attention to the variant paradigm in fua- attested in Modern Irish for the past tense of téigh, and argues that this, along with the paradigm in chá- (cf. bardic cháidh and ScG chaidh), results from the blending of synonymous ro-fáith and do-chuaidh.
ad Gospel of Thomas, q. 21 (as. ed. by J. Carney 1964 [BILL 2778]). Rejects previous emendations and interprets MS domais as the 3rd sg. deut. pret. rel. of do-maisi ‘concocts’.
Lat. lodix glossed with Ir. sléic (means ‘pumice’, possibly related to slíachtaid ‘smoothes’), ruamnae (earlier form of rúamna ‘colouring matter, redish colour’), diol (‘fillet, diadem’): all exx. of ornamentum muliebre. Also suggests Ir. slíogadh ‘smoothing, polishing’ derives from ON slíkja ‘to smoothe’, although slíocadh forms may have been influenced by Engl. slick ‘to slick, polish, smooth’.
Argues that boíthi replaced a morphologically ambiguous 3rd sg. pret. of the substantive verb *boí (< InsC *bowe + en), which may survive as the anomalous 3rd sg. pret. of the copula boí found in some passages of Amra Choluim Cille: §25 boe seim sáth ‘there was to him slender food’ (as ed. by W. Stokesin RC 20.168 [Best¹, p. 237]).
Discusses other PIE parallels of ‘Act of Truth’ and cognate verbal expressions of the following four expressions: 1. Is tre ḟír flathemon ‘it is through the ruler’s truth’; 2. mortlithi (mórslóg no) márlóchet di doínib dingbatar ‘plagues, (a great host, or) great lightnings are warded off men’; 3. gáu ḟlathemon ‘ruler’s falsehood’; 4. ní[n]-aurdallat dána (support for emendation to ní-n-aurdallat anai ‘let not riches blind him’ (see F. Kelly, AM §31); 5. to- aidble éisc i sruthaib -snáither (emends to to- aidbli éisc i sruthaib -snáither ‘with abundance of fish it is swum in streams’, taking to-snáither to be an impersonal passive rather than 2nd sg. deponent (see F. Kelly, AM §20).
Repr. in Watkins selected writings II, pp. 626-643.
Addresses the semantic difficulties in O. Bergin’s hypothesis (in Ériu 12.227 [Varia 21: Old Irish téit]) that OIr. 3rd sg. téit/-tét derives from a root aorist *ten-t(i).
1. On the Old Irish dative singular in *+mi;
2. ‘Yes’ and ‘no’: (a) taccu; (b) tó ‘yes’ (DIL s.v. 1 to); 3. Ir. uirghe f.; 4. tene, ten masc. > fem. ‘fire’;
5. teng, ting ‘tongue’; 6. for·érig. Cf. E. P. Hamp, in Ériu 53 (2003), pp. 185-186.
Argues that OIr. (h)uisse is not related to L iūstus ‘j{u}st’ but that it originated as a past participle in the factitive (or causative) verbal paradigm of the root *Hi̯eudh-, possibly related to imperative forms uind-se, uinn-si ‘look, behold, etc.'.
Argues that do·uccai derives from an old causative *h2/3ōnḱ-ye/o- to the PIE ‘Narten’ present *h2/3ḗnḱ-ti, *h2/3énḱ-n̥ti, and that do·ratai is an old primary verb (< *deh3- ‘give’) absorbed into the weak a-class.
Presents evidence from Audacht Morainn (as ed. by A. Ahlqvist 1984) for the survival of the middle voice in two Old Irish verbal forms, teclannathar (< do·eclainn) and méthathar (< méthaid).
Studies the alternation between con- and cond- in forms of con-ric (some analysable as ro-icc preceded by relative con- or the conjunction con), and argues that the -d- in these forms is not a pronominal but the result of anaptyxis.