Since ancient times, Greek has a great variety of forms. Each region, each city reveals its own language; each literary genre has a linguistic form consecrated by tradition, but treated differently by each author. These varieties fall into a number of main types, which are called dialects. As elsewhere, so also in Greece it is not possible to establish a clear division between dialect and dialect, between variety and dialectal variety. Thinking of linguistic differentiation as the result of successive divisions of a primitive nucleus and representing the variety of dialects in the form of a family tree can only be correct in the case that a fraction of the people, migrating, has lost contact with the others, i.e. when due to political or religious differences, two or more parties a linguistic community have become estranged from each other. But more often linguistic differences arise within a continuous territory. Each grammatical novelty spreads from a given point over a given area; evidently the limits of the different areas intersect; and since a dialect is defined by a sum of linguistic innovations, it is clear that each dialect can have one or more of them in common with others, so that all the dialects appear linked together as links in a chain. Furthermore, it can happen – and it has happened several times in Greece – that one people overlaps another: in this case the language that succumbs usually reacts on the winning language, which sometimes takes the appearance of a mixed language. The actual conditions are therefore very complex; to obscure them is added that, while we are well informed of some dialects, of others there are few or no documents. The following distribution tries to reconcile linguistic facts with historical and geographical data.
Doric: 1. Laconia (with Taranto and Eraclea). 2. Messenia. 3. Argolis (with Aegina). 4. Corinth (with Corcira and Syracuse) and Sicyon. 5. Megara (with Byzantium and Selinunte). 6. Crete. 7. Milo and Tera (with Cyrene). 8. Rhodes (with Gela and Agrigento). 9. The other Doric islands of the Aegean (Nanfio [Anáfē], Stampalia, Calimno, Cnido, Kos, Nisiro, Folegandro, Piscopi [Tḗlos]). 10. The Doric colonies in Sicily not remembered with the respective metropolis.
The dialect of Achaia and its colonies in Italy (Crotone, Metaponto, Posidonia, Sibari) resembles the Doric, but the insufficiency of the documents does not allow to determine its exact position: the archaic inscriptions are too scarce and the less ancient, more numerous, they no longer reflect the dialect of the region. With Acaic we usually remember the dialect of Kefalonia, Ithaca and Zakynthos (Zante), but the connection is not sure.
The dialect of Elis occupies an intermediate place between groups I and IV, from which, however, it is distinguished by some peculiarities of its own.
North-west group: This includes: 1. the dialects of Locride and Phocis, which show some affinity with the Doric ones of the Peloponnese; 2. those of Epirus, Acarnania, Aetolia, Phthiotis (southern Thessaly) and the Aenians, poorly documented.
Aeolian: Includes three divergent varieties: 1. Lesbos dialect (with Tenedo and the northern part of the Asian Aegean coast). 2. Northern Thessalian, distinguished in local varieties among which those of the Thessaliotid and the Pelasgiotis stand out. 3. Beotic. The original close kinship of these dialects, attested by a series of common characters, is obscured by the penetration of Western elements to a more limited extent in Thessalian, wider in Boeotic, which, while retaining a predominantly Aeolian aspect, could be said to be an intermediate dialect between the wind farms and those of the north-west.
Arcade-Ciprio: Common characteristics link these two geographically divided and distant dialects. They are the remains of the “Achaean” population that once stretched from the Peloponnese to Cyprus and was then overwhelmed by a subsequent immigration of Hellenic peoples.
Panfilio: Isolated from the rest of the Greek world and exposed to the influence of barbarian languages, this dialect had a very particular development. The documents, not very copious, reveal a strange mixture of discordant characters, which seems to confirm the memory of a variety of lineages, contained in the name Παμϕύλιοι. However, the affinities with ciprio and arcadius prevail, to which he connects them, as the third member of the Achaean group, the Meillet.
Ionian-Attic: It is distinguished first of all in Ionian and Attic. In the Ionian there are three varieties corresponding to three geographical areas: 1. Asiatic Ionia with the islands of Chios and Samos. 2. Cyclades (Delos; Naxos and Ceo; Paros with Thasos and Faro; Sifno). 3. Euboea (Chalcis with the Reggio and Cuma colonies; Eretria and Stira).
Macedonian is not taken into account in this enumeration, which some scholars (including the linguists O. Hoffmann and GN Hatzidakis) consider as a Greek dialect, whereas others keep it for a distinct language, perhaps akin to Illyrian (Hirt) or formed for the mixture of Greek and Illyrian or Thracian elements (Kretschmer), or declares (Meillet) that the quality, rather than the scarcity, of the material transmitted to us does not allow us to resolve the question (see fruit salad).
Some linguists (D. Pezzi, CD Buck) have tried to reduce the whole variety of Greek dialects to two basic types: Ionic and non-Ionic. Historical and linguistic reasons lead to a preference for a tripartite division. O. Hoffmann distinguishes: Doric in a broad sense, Achaean (subdivided into North-Achaean, that is Aeolian, and South-Achaean that is Arcadian-Cypriot) and Ionic. A. Thumb gathers the eight main dialects (we enumerate) into three large groups, calling respectively western, central and eastern the Doric in the broad sense, the Aeolian and Achaean and the Ionic-Attic, and putting aside the yacht as a mixture of ‘western and central elements.
The geographical qualification refers to the relative position of the groups in prehistoric times. The distribution of the dialects into three groups historically means that the immigration of the Hellenes occurred, as it were, in three bands, or in other words, that the three groups represent three successive ethnic strata. There is no doubt that in ancient Greece there was an overlap of different lineages; whether the layers were three, or more than three, cannot be said with certainty. A. Debrunner (op. Cit., P. 525) observes that the historical analogy makes a more complex stratification likely; and Meillet certainly distinguishes four groups (Ionic-Attic, Arcado-Ciprio, wind, west) as many representatives “poussées of envahissement” (op. cit., p. 79). The succession of layers is also a matter of discussion. Nobody disputes that the immigration of the Western bloodlines was the last (in this case the ancient tradition of a “Doric invasion and has a background of historical truth); as for the previous immigrations, opinions differ. P. Kretschmer (in Glotta, I, 1907, p. 9 ff. And at Gercke-Norden, Einleitung, 3rd ed., Leipzig 1923, I, p. 528) from the presence of oriental (Ionic) elements in the central (Achaean) dialects he inferred that the first Greek invaders were Ionic (or, more precisely speaking, that people of which the Ionians of history are a scion), and that the later Achaeans they drove them out of Greece (except Attica) in the way that the Dorians later drove the Achaeans out of the greater part of the continent and islands, or mingled with them. This doctrine has so far found more dissension than consensus. A. Thumb (Handbuch, Heidelberg 1909, p. 66) considers it a probable but unnecessary hypothesis.
The sources for the knowledge of the Greek dialects are the epigraphic texts, the literary texts and the information handed down to us by the ancient grammarians and lexicographers. The inscriptions reveal a large number of dialectal varieties, but their data are too often fragmentary. Relatively few inscriptions in dialect remain from the VII-V centuries BC. C. Those of the IV-II centuries, much more numerous, are more or less influenced by the koin ḗ Hellenistic which was already spreading throughout Greece; not a few could be said to be written in “common language” with dialect varnish. We have an extensive epigraphic documentation only of the attic, that is to say precisely of that dialect that even literature makes us know better than any other. Literary texts, in comparison with inscriptions, offer more copious material, but only those few dialects that were used in literature. The inscriptions give us valuable information on the phonetics and grammar of the various speeches, but their vocabulary is mostly restricted and the phraseology little varied. Literary texts give us in return a richer vocabulary and a more varied syntax.