Um diesen Artikel korrekt lesen zu können, müssen Sie zuerst die beiden True-Type-Schriftsätze Aaron und Amairgin (Autorin: Melanie Malzahn) herunterladen und auf Ihrem Computer installieren.

Notes on Celtic Nominal Morphology

1. Celtiberian -unei, Luguei 1

von David Stifter, Universität Wien

Abstract: The infinitives in -unei and the form Luguei suggest that in Celtiberian the dative singulars of proterokinetic stems were reshaped and took on the appearance of hystero- or amphikinetic stems.

Basically two reasonable approaches exist to explain the Celtiberian infinitival suffix -unei attested four times on the first bronze tablet from Botorrita (taunei [K.1.1,A-2], tizaunei [K.1.1,A-2], u]ertaunei [K.1.1,A-2], a]mbitinkounei [K.1.1,A-6]). Both build on the assumption that the forms in question are dative singulars of neutral verbal abstracts, from heteroclitic wer/n-stems on the one hand (most recently GARCÍA RAMÓN 1994, 52 f. and 1997, 66 f.; MEID 1995, 151 ff. and 163; ZIEGLER 1997, 634; RUBIO ORECILLA 1999, 623), from men-stems with the sound change Vmn > Vwn on the other hand (see, inter alios, EICHNER 1989, 35; DE BERNARDO-STEMPEL 1994, 295; SCHMIDT 1999, 315 with literature; undecided STÜBER 1998, 78 f.; without discussion JORDÁN CÓLERA 1998, 83 and 96). A third etymology (< *-wonei [sic], UNTERMANN 1997, 412) is implausible for phonological and morphological reasons:  PIE o > u in Celtic only in final syllables, otherwise > a, and suffixal o is only to be found in nominatives in PIE. VILLAR’s views (-unei = /uney/, 1997, 939) do not become clear to me.

Both strategies face the morphological difficulty that in the dative singular of a proterokinetic paradigm (to which both neutral wer/n-stems and neutral men-stems belong, cp. STÜBER 1998, 45 ff.) full grade of the suffix is expected (SCHINDLER 1975a, 9 and SCHINDLER 1975b, 263), in the present case *-weney, resp. *-meney.

I want to point out a further such case. In the Celtiberian rock inscription of Peñalba de Villastar, written in Latin letters, two times the form Luguei [K.3.3,3 and K.3.3,6] is found. This is generally interpreted as the dative singular of the divine name Lugus as one of the recipients of the dedicatory inscription. The first form could perhaps be read as Lugues (UNTERMANN 1997, 626), but ESKA 1990, in a syntactical analysis of the text, makes a dative singular plausible for both instances.

Lugus is a Celtic u-stem (for the attestations see below). The regular u-declension of Celtic apparently goes back to the corresponding proterokinetic inflexion of PIE. This replaced other accentual paradigms even where they were inherited into Celtic (cp. MCCONE 1994, 114 ff.; see NUSSBAUM 1998, 150 f. for relic forms of other inflectional types, recognisable by unusual root ablaut).

Evidence for the Celtic inflexion of u-stems can be found in the following forms:

Nom. sg.:  Gaul. DagolitouV ‘having good celebrations (?)’ [G-271], Esus [L-14] a God's name, trigaranus ‘having three cranes’ [L-14] (these last two words are likely to be u-stems in view of the o-stem nom. sg. Cernunnos and – admittedly ambiguous – taruos ‘bull’ on the same inscription), molatus ‘praise’ [Lezoux 4],  OIr. cruth ‘shape, form’ < PIE *-us.

Gen. sg.:  Ogam [AM]B[I]CATOS ‘having battle around himself’ [Mac. 500] (cp. Gaul. Ambigatus for *Ambicatus?), VERGOSO [Mac. 121] ‘manly vigour’ (cp. McMANUS 1991, 105 and 116), OIr. crotho < Common Celtic (CC) *-ows < PIE *-éws (or perhaps from PIE * -ows, cp. JASANOFF 1990, 183 2  and KLINGENSCHMITT 1992, 118 f.).

Dat. sg.:  Gaul. taranoou ‘to God Thunder (= Taranus)’ [G-27], gussou ‘to violence (?)’ [Lezoux 7] < Proto-Italo-Celtic *–e/ow < PIE *-éwey (with analogical loss of *-ey after the haplological loss in the i-stems *-ey < *-éyey, cp. KLINGENSCHMITT 1992, 105 f. and 119); OIr. cruthL < CC *-u either < instr. PIE *-uh, a relic of the acrostatic inflexion, or under influence of the o-stem dative in *-u < PIE *-oy, -od, -oh.

Acc. sg.:  OIr. cruthN < PIE *-um.

Abl. sg.:  Celtib. karauez ‘from Carauis’ [A.66] is very doutbful as a u-stem (VILLAR 1995, 13; but a different opinion in VILLAR 1997, 923).

Nom. pl.:  OIr. crothae, MCymr. kadeu, Cymr. cadau ‘battles’ < OCymr. -ou, OCorn. gemmou ‘gems’, OBret. airmaou ‘battle fields’ < CC *-owes < PIE *-éwes or perhaps from amphikinetic *-ówes; the ending of OIr. crothai has been taken over from the i-stems CC *-eyes, the ending of OIr. crotha probably only reflects the late Old Irish and Middle Irish tendency towards loss of distinctive vowel quality in final syllables.

Gen. pl.: perhaps Celtib. [e]dnoum ‘?’ [K.3.7] (= /ednom/?, cp. SCHMOLL 1959, 43) < CC *-owo/um < PIE *-ém; the ending of OIr. crothaeN has been taken over from the i-stems CC *-iyom.

Dat. pl.: OIr. crothaib is ambiguous and may go back to CC *-owbi(s), *-owobi(s) (with addition of thematic o to the stem) or *-obi(s) (with replacement of u by thematic o). The ending is that of the instrumental.

Acc. pl.: perhaps Celtib. matus ‘good (?); bear (?), period of time (?)’ [K.1.1.,A-2], see UNTERMANN 1997, 515 for literature on the possible meanings and the stem class (possibly o-stem); OIr. cruthuH < CC *-us < PIE *-uns.

Many more u-stems are attested as first members of compounds in Gaulish (the translations of the following Gaulish names are only tentative), among them:
*bitu- ‘world’ in Bituriges ‘kings of the world’ etc., cp. OIr. bith, Cymr. byd, OCorn. and OBret. bit, Corn. bys, Bret. bed ‘id.’.

*bratu- either ‘judgement’ (to OIr. bráth, Cymr. brawdd, Corn. bres ‘id.’, MBret. breut ‘debate, plea’ < *qrHtus, thus LEIA B-80 and MCCONE 1991, 11 f.  or < [rHtu-, thus LAMBERT 1996, 93) or ‘gratefulness’ (tu-abstract *qrHtus with semantical connection to Lat. grates, Osc. gen. sg. brateis ‘thanks, gratuity’) in Bratuspantium ‘place where vows are spoken’ (LAMBERT 1996, 93) and perhaps in formulaic bratoudekantem/n (the last interpreted as two separete words, the first as instr. sg. of an o-stem adjective ‘gratefully’ < *qrHtoh by SZEMERÉNYI 1974, 252 ff. and 1991, 306).

*litu- ‘ceremony’ or ‘fury, anger’ in Litugena/us ‘engendered by a celebration; born from fury’, Litumara/us ‘great in celebrations; great in fury’ etc., cp. OIr. líth, Bret. lid ‘celebration’, or Cymr. llid ‘fury’.

*magu- ‘boy, servant’ in Magurix ‘king of servants’ etc., cp. also Gaul. Magunia/us, OIr. mug ‘slave, servant’, Cymr. meudwy ‘servant of God’, MCorn. maw ‘boy, servant’, MBret. mao ‘happy, joyful’.

*medu- ‘mead’ in Medugenos ‘engendered by mead’, Medoureix ‘king of mead’ etc., cp. Celtib. mezukenos [K.1.3,I-4, -46, -60, II-4, -12, -21, III-11, -29, IV-9 Botorrita III], OIr. mid, Cymr. medd, OCorn. med, Bret. mez ‘id.’.

*taranu- ‘(God) Thunder’ in Taranucnos ‘Thunder’s son’ etc., cp. also Gaul. dat. sg. taranoou besides o-stem tarano- and i-stem tarani-, cp. OIr. o-stem torann, Cymr. taran, OCorn. and OBret. taran ‘id.’.

*webru- ‘amber’ in Vebrumara ‘great in amber’ etc., cp. also Gaul. ouebroukkoV, Cymr. gwefr ‘id.’.

For further evidence see BILLY 1993 sub uocibus.

The Celtic attestations of the divine name Lugus do not diverge from this pattern:

Nom. sg.:  Gaul. LougouV [G-159], OIr. Lug (the OIr. evidence is from DIL sub uocibus Lug and Lugnasad ‘Lug’s wedding’) < CC *lugus (MCymr. Llew, Lleu in Math fab Mathonwy, the fourth branch of the Mabinogi, influenced by llew ‘lion’ < Lat. leo and lleu ‘light, bright’?).

Gen. sg.:  OIr. Loga < CC *lugows.

Dat. sg.:  OIr. LugL < CC *lugu.

Acc. sg.:  OIr. LugN < CC *lugum.

Nom. pl.:  Gaul. Lugoues [CIL XIII 5078, Avenches, Switzerland] < CC *lugowes.

Dat. pl.:  Celtib. latinised Lugouibus [CIL II 2818, Osma, Soria], Celtiberian weak stem *lugow- with Latin ending -ibus, perhaps for original Celtib. athematic *lugowbos (attested in Galicia as Lucoubu 3 UNTERMANN 1997, 403; Lucoubus 3 in REVUE ARCHÉOLOGIQUE 50 (1957), 322) or for Celtib. *lugobos (for thematic vowel ù in front of the ending -bos cp. o-stem dat. pl. akainakubos [K.1.1,A-9 Botorrita I], arekoratikubos [K.6.1,1 Luzaga], beskuauzuetikubos [K.30.1], loukaiteitubos [K.0.7], r-stem dat. pl. [ma]trubos ‘to the mothers‘ [CIL II 2848, Muro de Agreda, Soria], cp. VILLAR 1997, 922) or for *lugowebos (perhaps attested in Spain as Lucouebus 3, REVUE ARCHÉOLOGIQUE 19 (1912), 457; for thematic vowel e in front of the ending -bos cp. perhaps consonant stem dat. pl. tikerzeboz [K.6.1,3 Luzaga]; VILLAR 1997, 918 however does not recognise this form as a dative plural).

The bare stem *lugu- is attested as first compound member for example in Continental Celtic and British place names like Lugudunum ‘Lugus’ town’ (for a critical discussion of the meaning see MAIER 1996, 128 ff.), in the OIr. name of a festival day Lugnasad ‘Lug’s wedding’, and in the personal names OIr. Lugaid, gen. Luigdech, Ogam Irish e.g. Lugudeccas [Mac. 263]  ‘worshipper of Lug’ < *lugudek- (cp. perhaps also the Gaul. personal name Lucudeca [CIL XIII 5926 Rep. Bourbonne-les-Bains]), Cymr. Llywelyn < *lugubelinos and Gaul. Lugurix [L-4] ‘king of warriors’ (I will not go into a discussion of the various meanings of lugu-, cp. for this MAIER 1996).

The only form which does not square with this fundamentally proterokinetic pattern is the Celtiberian dat. sg. Luguei. This may conceivably stand for /lugwey/ or /luguwey/. Unlike VILLAR 1997, 923 I do not see any reason why we should see a phonological reality /lugowey/ behind this spelling. Neither the Gaul. dat. sg. ending *-ow, nor the OIr. dat. sg. LugL can go back to CC †lugwey (> Gaul. †lugwe, OIr. †luigb) or †luguwey (> Gaul. †lugowe, OIr. †logae or Gaul. †luguwe, OIr. †lugae (?); for the development of CC uw > ow before a vowel other than i cp. McCONE 1996, 55, for a different opinion cp. JASANOFF 1988, 301). Therefore this form has to be a Celtiberian innovation.

If Luguei represents /luguwey/ the only place where the u could be got from would be the stem lugu- of the nom. and acc. sg.  (I owe this suggestion to Martin PETERS). In the case of paradigmatic levelling in favour of the ‘strong’ stem the spread of this stem variant to the whole paradigm should be expected. The dat. pl. Lugouibus and Lucoubu without a stem lugu- speak against this possibility. Therefore I take Luguei to stand for /lugwey/ with a zero grade suffix comparable to that of –unei.

In view of Luguei and the infinitives in -unei I set up the rule, that in Celtiberian paradigms that continue the PIE proterokinetic inflection in the dative singular the inherited full grade variant of the suffix was replaced by the apparently hysterokinetic or amphikinetic zero grade variant.

According to the teaching of Jochem SCHINDLER proterokinetic stems had a descriptively ‘hystero-’ or ‘amphikinetic’ instrumental singular (SCHINDLER with reference to Vedic instr. deviyá ‘goddess’ << diwiHé/óhapud HOLLIFIELD 1980, 45 and in class; similarly NUSSBAUM 1998, 154187 with reference to Vedic instr.  paraƒvá from paraƒú- ‘axe’). Martin PETERS has expanded this theory to the directional case, thus deriving the sigmatic aorist infinitives of Greek of the type deixai from a neuter s-stem deykes- with the directional ending *-ay and zero grade in the suffix.

Hence it can be assumed that the starting point for the spread of the hysterokinetic or amphikinetic suffix variant to the dative singular of originally proterokinetic stems in Celtiberian lay in the instrumental and directional singular (personal information from Martin PETERS). Mutual influence among these cases is even more likely in view of the fact that dative and instrumental show a tendency towards syncretism in Celtic (cp. for example above the OIr. dat. sg. and pl.).

As to the etymology of the Celtiberian infinitival suffix -unei, I derive it from *-mney, on grounds of the frequency of neutral men-stems among the Insular Celtic verbal nouns and the absence of heteroclitic wer/n-stems in Celtic with the one exception of the isolated OIr. arbar, arbae ‘corn’ on the one hand and on grounds of the sound change Vmn > Vwn, well attested in all of Continental Celtic and British, on the other hand (cp. for example Celtib. gentilic name kounesikum [K.1.1,B-1 Botorrita I] ‘neighbours’ < *kom-ness-iko-, Noric personal name Counertus ‘having great strength’ (LOCHNER VON HÜTTENBACH 1989, 63) < *kom-nert-o-, Lepont. ualaunal (LEJEUNE 1971, 407), Gaul. *wellawnos, OCymr.  °guallaun etc. ‘ruler’ < CC *we/alla-mn-o-, cp. LAMBERT 1996, 94 f.; more about this question in an article to appear).


[A.1] = UNTERMANN 1975

BILLY, Pierre-Henri
1993 Thesaurus Linguae Gallicae, Hildesheim - Zürich - New York

1994 Das indogermanische m(V)no-Verbaladjektiv im Keltischen, in: ed. Roland BIELMEIER et al., Indogermanica et Caucasica. Festschrift für Karl Horst Schmidt zum 65. Geburtstag, Berlin - New York, 281-305

DIL = ed. QUIN, Ernest Gordon
Dictionary of the Irish Language. Based Mainly on Old and Middle Irish Materials. Compact Edition, Dublin 1990

1989 Damals und heute: Probleme der Erschließung des Altkeltischen zu Zeußens Zeit und in der Gegenwart, in: ed. Bernhard FORSSMAN, Erlanger Gedenkfeier für Johann Kaspar Zeuß, Erlangen, 9-56

ESKA, Joseph
1990 Syntactic notes on the great inscription of Peñalba de Villastar, BBCS 37, 104-108

1980 Inscriptions gauloises sur céramique et l’exemple d’une inscription de La Graufesenque et d’une autre de Lezoux, ÉC 17, 111-144

[G-1] = LEJEUNE 1985

1994 Correspondencias y ecuables en la reconstrucción indoeuropea: chipr. to-we-na-i /dowenai/, véd. daváne, IE *d]-wén-, in: ed. Luis M. MACÍA APARICIO et al., Quid ultra faciam? Trabajos de griego, latín e indoeuropeo en conmemoración de los 25 años de la Universidad Autónoma de Madrid, Madrid, 45-56
1997 Infinitive im Indogermanischen? Zur Typologie der Infinitivbildungen und zu ihrer Entwicklung in den älteren indogermanischen Sprachen, InL 20 [1998], 45-69

HOLLIFIELD, Patrick Henry
1980 The Phonological Development of Final Syllables in Germanic. Part 1, Sprache 26, 19-53

1988 Old Irish boí ‘was’, in: ed. Mohammad Ali JAZAYERY et al., Languages and Cultures. Studies in Honor of Edgar C. Polomé, Berlin - New York – Amsterdam, 299-308
1990 The Origin of the Celtic Comparative Type Oir. tressa, MW trech ‘stronger’, Sprache 34/1 [1988-1990], 171-190

1998 Introducción al Celtibérico, Zaragoza

[K.1.1.] = UNTERMANN 1997

1992 Die lateinische Nominalflexion, in: ed. Oswald PANAGL et al., Latein und Indogermanisch. Akten des Kolloquiums der Indogermanischen Gesellschaft, Salzburg, 23.-26. September 1986, Innsbruck, 89-135

LAMBERT, Pierre-Yves
1994 La Langue Gauloise. Description linguistique, commentaire d’inscriptions choisies, Paris
1996 Notes gauloises, in: ed. Wolfgang MEID et al., Die grösseren altkeltischen Sprachdenkmäler. Akten des Kolloquiums Innsbruck, 29. April-3. Mai 1993, Innsbruck, 86-106

[L-1] = LEJEUNE 1988

LEIA = ed. VENDRYES, Joseph (BACHELLERY, Édouard and LAMBERT, Pierre-Yves)
 Lexique étymologique de l’irlandais ancien, Dublin - Paris 1959-

1971 Lepontica, ÉC 12, 357-500
1985 Recueil des Inscriptions Gauloises, Vol. I. Textes gallo-grecs, Paris
1988 Recueil des Inscriptions Gauloises, Vol. II fasc. 1. Textes gallo-étrusques, Textes gallo-latins sur pierre, Paris

[Lezoux 1] = FLEURIOT 1980, 128

1989 Die römerzeitlichen Personennamen der Steiermark. Herkunft und Auswertung, Graz

[Mac. 1] = MACALISTER 1945

MACALISTER, Robert Alexander Stewart
1945 Corpus Inscriptionum Insularum Celticarum, Vol. I, Dublin

MAIER, Bernhard
1996 Is Lug to be Identified with Mercury?, Ériu 47, 127-135

1991 The Indo-European Origins of the Old Irish Nasal Presents, Subjunctives and Futures, Innsbruck
1994 An tSean-Ghaeilge agus a Réamhstair, in: ed. Kim MCCONE et al., Stair na Gaeilge in Ómós do P[h]ádraig Ó F[h]iannachta, Má Nuad, 61-219

McMANUS, Damian
1991 A Guide to Ogam, Maynooth

MEID, Wolfgang
1995 Das Verbum im Keltiberischen, in: ed. Heinrich HETTRICH et al., Verba et Structurae. Festschrift für Klaus Strunk zum 65. Geburtstag, Innsbruck, 135-167

NUSSBAUM, Alan Jerry
1998 Two Studies in Greek and Homeric Linguistics, Göttingen

1999 El verbo celtibérico y el verbo protocéltico: problemas y perspectivas, in: ed. Francisco VILLAR et al., Pueblos, lenguas y escrituras en la Hispania prerromana. Actas del VII coloquio sobre lenguas y culturas paleohispánicas. Zaragoza, 12 a 15 de Marzo de 1997, Salamanca, 619-634

1975a L’Apophonie des thèmes indo-européens en -r/n, BSL 70, 1-10
1975b Zum Ablaut der neutralen s-Stämme des Indogermanischen, in: ed. Helmut RIX, Flexion und Wortbildung, Akten der V. Fachtagung der Indogermanischen Gesellschaft, Wiesbaden, 259-267
1977 Notizen zum Sieversschen Gesetz, Sprache 23, 56-65

SCHMIDT, Karl Horst
1999 Review of: UNTERMANN 1997, IF 104, 310-317

1959 Die Sprachen der vorkeltischen Indogermanen Hispaniens und das Keltiberische, Wiesbaden

1998 The Historical Morphology of n-Stems in Celtic, Maynooth

1974 A Gaulish dedicatory formula, KZ 88, 246-286
1991 Review of: Patrizia DE BERNARDO STEMPEL, Die Vertretung der indogermanischen liquiden und nasalen Sonanten im Keltischen, Innsbruck 1987, ZCPh 44, 299-312

1975 Monumenta Linguarum Hispanicarum, Bd. I. Die Münzlegenden, Wiesbaden
1997 Monumenta Linguarum Hispanicarum, Bd. IV. Die tartessischen, keltiberischen und lusitanischen Inschriften, Wiesbaden

VILLAR, Francisco
1995 A new Interpretation of Celtiberian Grammar, Innsbruck
1997 The Celtiberian Language, ZCPh 49/50, 898-949

1997 Zur Syntax und Entwicklung des Verbalnomens in den keltischen Sprachen, in: ed. Emilio CRESPO et al., Berthold Delbrück y la sintaxis indoeuropea hoy. Actas del Coloquio de la Indogermanische Gesellschaft. Madrid, 21–24 de Septiembre de 1994, Madrid – Wiesbaden, 631-644

My work on this article was made possible by a doctoral scholarship of the Österreichische Akademie der Wissenschaften. I want to thank especially Martin Peters for many invaluable suggestions and his advice through all stages of this work. For remaining errors and inaccuracies I alone am to blame.

2  I do not subscribe to JASANOFF’s view that the i-stem gen. sg. ending –o of Old Irish goes back to -*oy Apart from the phonological difficulties (there is no other evidence that PIE and CC oy developed into anything else than i in absolute auslaut and stayed oy elsewhere in Old Irish) there is nothing implausible about mutual influence among i- and u-stems paradigms, despite JASANOFF’s claim „that the i- and u-declensions had little if anything in common at the pre-Ogam linguistic state“ (1990, 183). On the contrary, already in Common Celtic or even Proto-Italo-Celtic the dat. sg. *–e/owey of the u-stems was remodelled to –e/ow in analogy to *-ey < *-eyey of the i-stems (cp. above sub dat. sg.). Furthermore in Old Irish the u-stems took over the ending of the i-stem gen. pl. –eN< *-iyom and the ending of the nom. pl. –i < *-eyes as an alternative to the inherited ending –e < *-owes.

3 The frequent spelling with c suggest that perhaps we are dealing with a different word *luku- here, perhaps related to the name of the Galician town and province Lugo < Lucus. This wouldn't, of course, change anything of the u-stem character of the word.