4.6. Celtic

4.6.1. Celtic evolution

Certain common phonological features from Celtic include:

·       Newly arisen sequences of velar + labial glide merged with labiovelars: cf. *eko- Cel. *ekwo- behind Gaul. epo-, OIr. ech ‘horse’.

·       Labiovelars:

o   Delabialise in various contexts *Kw → *K: before *, before *n, before *u.

o   In the other cases, *gw → *b.

o   Shortly after this change, *gwh becomes *gw (see following point).

·       No opposition *T – *Th. Probably:

o   Voiced aspirates lost their aspiration: *Dh → *D.

o   Later intervocalic voiced obstruents were “lenited”, becoming the corresponding voiced fricatives: *Dh → *Đ.

·       Important changes are seen with *p weakened to a voiceless bilabial fricative *φ:

o   *φ *x before *s and *t; *φ → *βbefore liquids.

o   After *s, it survives as an allophone (maybe *b).

o   In all other contexts *φ **h → *Ø. At this stage, preceding liquids geminated: *RH → *RR.

·       Merge of *k →*kw.

·       Dental + velar sequences are metathesised: e.g. *dhghom- → *gdon-.

·       Complex developments of clusters of three or more consonants with a sibilant in the middle.

·       Resonants resolved into sequences of resonant + vowel, or vowel + resonants, usually *CR̥C → *CaRC, but e.g. *Ri before non-continuants.

·       Clusters of labial segment + *n showed complex developments.

·       Few changes to vocalism, including reduction of long vowels to three, *ā, *ī, *ū (e.g. *o *ū). Similar to Italic or Greek, Celtic keeps the distinction between short *a and *o.

Important morphological developments include:

·       Generalisation of stem *so- for the demonstrative pronoun.

·       Aorist and perfect merge into a single preterite category.

·       Passive of preterite formed from a separate stem, based on verbal adjctive in *-to- (used originally in the copula as a periphrastic passive similar to Lat. factus est ‘was made’).

·       Present participle and infinitives lost.

·       Different augments, widespread in *ro- < *pro-, and *no-.

·       Rise of present stem classes through the loss of various intervocalic consonants belonging to either the root or the suffix.

The first references to Celtic peoples are found in connection with the Greek settlement in Massalia by Hekataios of Milet (indirectly attested in the Ora maritima of Festus Rufus Avienus in the 4th century). Celts are also described later in the works of Herodotus (5th c. BC), Aristotle (4th c. BC), Polybios (2nd c. BC), Poseidonius (1st c. BC) and Caesar’s De bello Gallico (1st c. BC).

The first Celtic inscriptions are ca. 6th–1st c. BC in Lugano alphabet in Lepontic, a p-Celtic dialect (either early Gaulish or an independent Celtic branch); Gaulish, written in the Greek alphabet in Southern France from the 3rd c. BC, but also Cisalpine Gaulish in the Piedmont, in a variant of the North Etruscan alphabet, dated to ca. 1st c. BC; and Celtiberian in the Celtiberian variant of the north-eastern Iberian script in the 2nd-1st c. BC (Vath 2017).

The three broadly described Celtic groups with different SVO syntax allow us to infer an older emergence of a Common Celtic language probably in the centuries around the turn of the 2nd–1st millennium BC. The expansion of Hallstatt and La Tène have been traditionally associated with the expansion of Celtic peoples.

4.6.2. Proto-Celtic–Early Proto-Fennic contacts

Proof of Early Proto-Fennic (EPF) adstrate in Proto-Celtic include certain terms concentrated in the field of hyonyms, i.e. words designating pigs and boars (Hyllested 2016):

·       Cel. *brokko ‘badger’, also in Gmc. *brakka- ‘(scent) hound, dog used for hunting’, could be traced to EPF *mäkrä ‘badger’, with Cel. *br- coming from an older *mr- and stress shift with vowel loss in the first sillable. The word may be ultimately of Altaic origin (and thus connected to the contacts of the Seima-Turbino network), given the more recent similar loans into Russian or Hungarian.

·       Cel. *lub-ia- (cf. OIr. f. luib ‘herb, plant’), also Gmc. *luba- n. ‘herb, potion’ (cf. Goth. lubja-leisei ‘witchcraft’, ON lyf ‘medicine, healing herb’, OE lyb ‘medicine, drug, potion’). A concept associated with magic, their similarity is often interpreted as a loan from one language to the other. However, they may have been adopted from Uralic *luppo, cf. Fi. luppo ‘lichen’.

·       Cel. *mokku- ‘swine’ (cf. OIr. mucc, W moch, Bret moc’h, Gaul. Mocccus, the name of a pig divinity) < EPF *emä ‘womb; mother (also of animal) + suffix *-kko, cf. Finn. emokki, Est. emak, Votic emakko, etc. Instead of being a NWIE loan or later wanderword, Kroonen proposes that MLG mocke f. ‘breeding sow’ and MLG MDu. mocke f. ‘sow’ are later loanwords directly from Gaulish, since they are confined to the Franconian part of the Germanic area (Kroonen 2013).

·       Cel. *sukko- ‘sow’ (cf. W hwch, OBret. hoch, Corn. hoch, OIr. socc-) looks similar to the common Late PIE hyonym *suH-s, but the reconstruction of a laryngeal for Proto-Celtic is not possible, and the West Germanic forms in -g- come not from an older *k, but from a regular velar development (Verschärfung) of hiatus or -- between two high vowels if at least one of them is u (Kroonen 2013). Lat. sūcula comes from diminutive ending -cula. Therefore, NWIE *seuk- did not exist. On the other hand, Ltv. cūka, Lith. dial. č ‘pig’ may point to a common origin in EPF *či/uka, cf. Kar. čugu, Finn. sika (compare with N. Saami dial. sohki, Inari Saami šahe). Supporting this, Latvian compound mežacūka ‘wild boar’ (lit. ‘forest pig’) has the same etymological source as Finn. metsäsika ‘badger’ (metsä ‘forest’ is in turn a borrowing from Baltic *meda-).

·       Cel. *turko ‘wild boar’ (W. twrch, OBret. torch, OIr. torc) < EPF *tora ‘tusk of a wild boar’ + denominal suffix *-kko, cf. Kar. torakko, torikko ‘tusk of a wild boar’. The root is also found e.g. in Finn. tora- ‘struggle, fight, battle’, N. Saami doarro-.

These direct contacts between Early Proto-Fennic with Proto-Celtic must come then most likely from direct contacts with the Tumulus culture (ca. 1600-1200 BC) and the succeeding Urnfield culture (ca. 1300-750 BC) with cultures from east Europe. The importance of pigs in the Baltic during the Iron Age is attested by Tacitus in his ethnographic work Germania, from around AD 98, in a commentary on the Aestii, a Northeast European tribe:

They worship the mother of the gods: as an emblem of that superstition they wear the figures of wild boars: this boar takes the place of arms or of any human protection, and guarantees to the votary of the goddess a mind at rest even in the midst of foes. (Transl. Hutton 1914).

4.6.3. Schleicher’s fable in Proto-Celtic

ou̯is ekwūs-kwe

ou̯is i̯āi ulanā ne est

ekwūns dorke;

som trummom u̯egnom dukontam,

som magi̯os baskim,

som gdoni̯om rinnom berantam.

ou̯is ekwobos sekwet:

“kridi̯os agetor mu,

ēdonteiirom ekwūns rēdontam.”

ekwūs sekwont: “roklinu, ou̯i!

kridi̯os agetor nos u̯ēdontbos:

iros, tigernos, ou̯i̯om ulanām

su̯ebei tepesmim linnām kwrinouti.

ou̯i̯om-kwe u̯lānom ne esti.”

om sosim klust ou̯is magosam tekwt.