ABSTRACT

The Greek language forms, by itself, a separate branch of the Indo-European family. It is one of the oldest attested Indo-European languages, being attested from c. 1400 BC in the Mycenaean Greek documents found on Crete (and from somewhat later, on the Greek mainland) written in the Linear B syllabary. Except for a break in attestation between the end of the Mycenaean empire (c. 1150 BC) and roughly 800 BC, a period sometimes referred to as the ‘Dark Ages’ of Greek culture, Greek presents a continuous record of attestation for the linguist, right up to the present day. Commonly called Greek in English, based on the term Graeci used by the Romans to

label all the Greeks (though originally the name may have properly applied only to a tribe in the north-west of Greece), the language is also referred to as Hellenic, from the Greek stem Εʽkkgm-*,1 used in the Iliad to refer to a Thessalian tribe but in Herodotus (and elsewhere) to designate the Greeks as a whole as opposed to barbarians; indeed, the Greeks themselves have generally referred to their language as e9kkgmijή, though contemporary Greeks also use the designation qxlaίija, an outgrowth of their connection historically with the Eastern Roman Empire based in Constantinople. Using a somewhat outmoded but nonetheless still widely cited putative dialect divi-

sion within Indo-European, one can classify Greek as a ‘centum’ language, for it shows a distinct set of reflexes for the Indo-European labio-velars, opposed to a single set of reflexes for the Indo-European palatals and velars combined; thus, Greek shows a root pqia-‘buy’ (cf. also Mycenaean qi-ri-ja-to ‘bought’ showing the labio-velar preserved as < q > ) from Proto-Indo-European *kwriH2-(cf. Sanskrit root krı-- ‘buy’), a noun jqέaς ‘meat’ from Proto-Indo-European *krewH2s (cf. Sanskrit kravis-‘raw flesh’), and a root jei-‘lie (down)’ from Proto-Indo-European *k

ei-(cf. Sanskrit root s´ı-- ‘lie’), in which the plain *k of the proto-language and the palatal *k

show a merger while the labio-velar *kw is kept distinct. Greek also shows some particular affinities with Armenian and Indo-Iranian, sharing with these branches, for example, the past-tense morpheme *e-(the ‘augment’), and the use of the negator *me-(Greek lή), and with Armenian alone the vocalisation of the Indo-European ‘laryngeal’ consonants in initial

a' Arm. where no other Indo-European language has precisely this form, or pqxjsός ‘anus’ = erastank’ ‘buttocks’). Moreover, Greek preserves the Indo-European vowel system (with long and short *a *e *i *o *u) more faithfully than any other language in the family. Differentiating Greek from the other members of the Indo-European family, though,

are several particular features. In morphology, Greek innovated a (past and future) passive marker -hg-and elaborated the infinitival system. With regard to phonology, Greek alone in Indo-European shows voiceless aspirates (in the ancient language) as the continuation of the Indo-European voiced aspirate consonants (e.g. ueq-‘carry’ from *bher-, cf. Sanskrit bhar-); in addition, Greek lenited Indo-European *s to h in many environments, ultimately losing it intervocalically (e.g. e9psά ‘seven’ from *septm

˚ , cf. Latin septem, or cέme-i ‘in, at, to a race, kind (dat. sg.)’ from *genes-i, cf.

Sanskrit janas-i ‘in the people (loc. sg.)’). Also, Greek deleted original word-final stops (e.g. lέki ‘honey’ from *melit, cf. Hittite milit ‘honey’). Moreover, although Common Greek preserved the Indo-European labiovelars as

such, to judge in part from their preservation in Mycenaean (cf. qi-ri-ja-to above), the ancient language is characterised by a number of complex dialectal developments with *kw, *gw and *gwh. Labial reflexes occur in some environments and in some dialects (e.g. pan-Greek interrogative stem po-from *kwo-, Aeolic (Boeotian) pέssaqeς ‘four’ from *kwetwr