ABSTRACT

On the other hand, according to Hawkes, 'for Goidelic itself, chronology might envisage something, not in the Hallstatt only, but even behind the Urnfields, back in the Bronze Age' (Hawkes 1987: 212), a statement that comes close to G. S. Mac Eoin's supposition: 'Is it legitimate ... to think in terms of a diffusion of Celtic to Britain and Ireland at that time before the growth of the chariot-riding aristocracies?' (Mac Eoin 1986: 172). From the linguistic point of view, there are grammatical and literary features which may indicate that proto-Celtic (PC) emerged in a rather early period. The grammatical features, that is, (a) the Gaulish future in ·sie-/·sio-; (b) the Old Irish (Olr.) equivalent of the Sanskrit desiderative formation; and (c) the Celtiberian (CI) inflected relative pronoun •ios, connect Proto-Celtic with eastern Indo-European languages, particularly with Indo-Iranian, Greek, Phrygian (and Slavic):

(I)a. Gaul. (Chamalieres) biss{et 'he will split' : *bheid-, *bhid-; pissfiu m{ 'I shall see' : OIr. ad-d'sees' < *kwis-e-t; !onr;naman toncsi{ontio 'who will swear the oath': Ved. kar-i-$yti-ti 'he will make', vak-$yti-ti 'he will speak, Avestan vax-syiz 'I shall speak', Lith. duo-siu 'I shall give', OCS bys~teje, bysQsteje < *bhu-sjo-nt-''to IlEJ..A.OV', Gk. KEiovn:c;' KOlllll~11(J6IlEVOt Hes. < *kej-hl-sjo-nt-= Ved. say-i-$yti-nt-;

b. SkI. dkir$ati: kar-'to make', krtti-, susrU$ate: sru-'to hear', sruM-; 0Ir. celid 'reveals': fut. 'dla < *cechliz-< *kikliz-< *kik/-h-se-I-so-;

c. CI (Botorrita) : iomui: IiSTas : TiTaS : sisonTi : somui (A7);

ias: osias : ueiTaTosue : Temeiue : ioPiSeTi : saum : TeCameTinas : TaTus (A8); ToCoiTei: iDS: ui (.) anTiomue: auseTi : aiaTimue: TeCameTam : TaTus (AlO)

(Schmidt 1986b: 173ff., 1988a: 245)

The fact that these grammatical features, while lacking in Italic, are shared by Celtic and eastern Indo-European languages, seems to point to early, pre-Italo-Celtic language contacts between Celtic and eastern IndoEuropean. On the other hand, the grammatical characteristics listed in (1) do not necessarily reflect the Proto-Indo-European model and therefore must be taken as common innovations rather than as common IndoEuropean inheritance.