4.1. Greek

4.1.1. Greek evolution

Evolution of Disintegrating Indo-European to Proto-Greek, probably through a Graeco-Armenian stage, include the following phonetic changes (Bubenik 2017):

·       Final laryngeal evolution (see §II.2. Laryngeal evolution).

·       Devoicing of voiced aspirates: *bh → *ph; *dh → *th; *dh → *th. Plain voiced and voiceless stops do not change.

·       Dissimilation of labiovelars to plain velars adjacent to *u, e.g.  Gk. ouk.

·       LPIE *s → PGk. *h except adjacent to itself, a voiceless stop, or when final.

·       Vocalisation of syllabic resonants with epenthetic vowels, unstable still in the common period, e.g. Gk. karterós/kraterós.

·       Preservation of *; *- is lost, although in initial position it is found weakened to *h- or strengthened to *d-.

·       Palatalisation, affrication, and depalatalisation and merging with cluster *ts: *t → *t̓ → *t̓ś → *ts; *k- → *k̓- → *t̓ś- → *ts. The voiced counterparts were further simplified *d, *g → *dz → *z.

Morphological features include (García Ramón 2017):

·       The eight LPIE cases can be reconstructed for Proto-Greek before the syncretism of later dialects.

·       A postposition added to the accusative ending, *-de, can also be reconstructed as a “directive” found in ancient dialects.

·       Thematic nominative plural in *-i.

·       Preservation of pronouns and demonstratives, with certain innovations.

·       Fairly conservative verbal system, including dual, opposition active/middle and passive (in aorist and future stems), three aspectual stems, threefold opposition of tense, four moods (plus the inherited injunctive, still alive in Mycenaean).

4.1.2. Contacts with Pre-Greek sources

Proto-Greek is supposed to have entered the Greek peninsula after ca. mid–3rd millennium BC. The cultural and genetic steppe-related impact is relatively small compared to that of Bell Beakers (see §3.2.1. North-West Indo-European community), and this is reflected in the heavy inheritance of pre-Greek forms. Place names of Archaic Greece show a mixture of Greek and non-Greek forms.

The following are examples of non-Greek names (many of which show Pre-Greek morpheme *-ān-), many found already in Mycenaean texts: Kórintho-, Tírunth-, Erúmantho-; Mukḗnā-, Athā́nā-, Halikarnāssó-, Knōssó-; Thbai (Myc. te-qa-de = Thēgwans-de ‘to Thebes’), Krēt-, etc. and names of mountains such as Ólumpo-, Parnassó-, Díktā-, etc. The combination of Greek and non-Greek place names is proof that Greeks found central places such as Athā́nā- and Kórintho- already inhabited and named, but smaller places could still be given names in their own language, such as those found in Mycenaean texts: Pleurṓn- ‘side of an area’, from pleura ‘rib, side’; Marathṓn- ‘rich in fennel’, from márath()o- ‘fennel’; Selinóont- ‘rich in celery’; Hrío- (<*sri-o-) ‘peak’; Hél-es- (<*sél-es-) ‘swamp’; Leũk-tr-o- ‘lookout’, from léusse (<*leuk-e-); and Plataiái ‘plain’.

To the same substratum belong culture words such as asámintho- ‘bath-tub’, and kupárisso- ‘cypress’, both found in Mycenaean and Homeric Greek.

Dubious is the origin of words that are analysable as Indo-European, but which may have a non-Indo-European origin, such as Mycenaean terms with obscure etymologies anakt- ‘king’, and its later replacement gwasileu- (originally a local clerk); or forms with difficult morphological interpretations, such as atástha-lo- ‘overconfident, carefree’, khróno- ‘time’, thálassa ‘sea’.

The replacement of the most common noun for wealth and livestock, *pekū, into Gk. próbaton (close to próbasis) shows that it continues the Graeco-Aryan tradition of cattle and sheep-goat herding economy, but replaces the old root with an innovative economic term for moveable property (Benveniste 1969).

4.1.3. Anatolian and Semitic contacts

It has been argued in the past that an Anatolian language may be behind the Pre-Greek substrate, at least of place names in -sso-, -tto-, and -ntho-, believed to correspond to Anatolian place names in -ssa and -anda[xiii]. While the nature of the substrate language is difficult to ascertain, it is clear that Mycenaeans had contact with contemporaneous Anatolians (Hajnal 2018):

·       Mycenaean had presence in Milet ca. 1450-1100 BC, and this is identified with Millaa(n)da in Hittite texts (identical with Greek Milet). Therefore, the western Aḫḫia(a) described by Hittites must be identified with a mainland Greek empire, and in Mycenaean with the state name *Akhaiu̯ia, associated with the ethnic name Ἀχαιοί (Akhaiu̯oi) the leading class in the palaces of Knossos, Khania, Pylos, Mycenae, Tiryns, and Thebes, and the name by which Homeric Greeks designated themselves in the Trojan war.

·       In Mycenaean, women from Milet, Milātiai, or Halicarnassus, prisoners of war, are referred to as from Asia, identical to the Hittite toponym Aššua.

·       Names from Knossos contain pi-a-, which seem to correspond to frequent Luwian names with a first verbal component pi-oº- ‘give’.

The Late Bronze Age contacts between Greek and Anatolian dialects can also be seen in borrowings on lexical, as well as on phonological, morphological, and syntactic levels. However, unlike the heavy influence of the Pre-Greek substrate, these close contacts did not leave substantial traces in Mycenaean or in Anatolian languages.

Semitic loanwords may be found behind Mycenaean khrū- ‘gold’ (noun and adjective), kúmīno- ‘caraway’, sā́sama ‘sesame’ (Attic-Ionic sḗsamo-), khitṓn- ‘undergarment worn on the body’, etc.

4.1.4. Schleicher’s fable in Proto-Greek

óu̯is kwoi-kwe

óu̯is ās lā́nos ne es

kwons dédorke;

ton gwarún u̯ókhon u̯ókhentã,

ton makrón phóron,

ton mrətón ōkã phérontã.

óu̯is kwoihi eikwe:

kərdā ákhnutoi moi,

anérã u̯idóntei kwons agóntã.”

kwoi u̯eikwont: “klúthi óu̯i!

kərdā ákhnutoi ãsmí u̯idóntei,

anḗr, denspóthis, ói̯on lā́non

su̯oi kwhermón u̯éstrãn kwoéei,

ói̯on-kwe u̯lā́nos ne ésti.”

toi kluu̯ṓs óu̯is agróm phégwoto.



·       Post-Mycenaean *hippos < **ékwos < LPIE *hékos is striking because of the aspiration, the double -pp-, and the -i-. Raising of -e- to -i- in labial contexts is normal. Archaic and Classical -pp- most likely are due to Proto-Archaic gemination **-kwkw-, to maintain the original prosodic structure of the etymo[xiv] (for both developments see §4.10. Lusitanian). The aspiration remains unexplained, because initial laryngeals had been lost already before the Proto-Greek period. It may be of expressive origin, or it was contaminated by another word.

·       For the pronoun Gk. ammí<*smí, an intermediate stage with -s- can be seen in the oldest stage (Bičanová and Blažek 2014).