Show Summary Details

Page of

Printed from Oxford Classical Dictionary. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 07 March 2021

Arcado-Cypriot dialectfree

  • Albio Cesare Cassio


In spite of being spoken in areas so far away from each other, Arcadian (written in the Greek alphabet) and Cypriot (written in a special syllabary) were two closely related ancient Greek dialects, hence the modern appellation “Arcado-Cypriot.” They descended directly from Mycenaean, some archaic vocabulary is unique to Homer, Mycenaean, and Arcado-Cypriot, and various other features set them apart from all the other Greek dialects.

Arcado-Cypriot is a modern name used for two ancient Greek dialects, Arcadian and Cypriot, which share in a number of peculiarities—both archaisms and innovations, the latter being of central importance for the reconstruction of an earlier Arcado-Cypriot unity (which was not recognised in antiquity). Obviously, beside numerous similarities, there are remarkable differences between Arcadian and Cypriot.

It is likely that both dialects, neither of which seems to have given rise to literary texts, descend directly from Mycenaean, hence the frequently used label of “Achaean” dialects. In archaic and classical times Arcadian was spoken by a population inhabiting central Peloponnese (see Arcadia), the Arcadians being surrounded by speakers of Doric dialects, while Cypriot was spoken by the Greek population of the island of Cyprus, on which Phoenician and some indigenous languages were also used (see Eteocypriot).1 The likeliest way to explain the remarkable similarities between these geographically very distant dialects is to admit that the Mycenaean ancestors of both populations lived in the Peloponnese in the late 2nd millennium bce and at some point were obliged to flee foreign invasions. Some of them took refuge in the mountainous central region of the Peloponnese (Arcadia), while others migrated to Cyprus (see dialects, greek, prehistory). It is hardly accidental that the places where most Cypriot Greek inscriptions were found are Marion and Paphos, in the westernmost part of the island.

Linguistic Features

Arcadian is known from glosses and alphabetic inscriptions (mainly official documents: temple regulations, decrees, etc.), the earliest of which go back to the beginning of the 5th century bce. With the passage of time, some archaic traits disappeared, and in Hellenistic times the dialect was more and more influenced by the surrounding Doric dialects and the koine. Cypriot Greek was written with a syllabary ultimately derived, like linear B, from Linear A; we now have more than 1,000 syllabic inscriptions (very short and of private nature with the exception of the Idalion bronze tablet and a handful of other documents; see Figure 1).2 The syllabary (of which a special variant is attested in Paphos) was in use mainly from the 8th to the late 4th century bce (but a very early inscription with the genitive of a personal name, o-pe-le-ta-u = φέλταυ‎ is dated to c. 1050–950 bce ).3 From the late 4th century bce, the Greek alphabetic system prevailed; however, the syllabary was still used (along with the alphabet) as late as the first part of the 2nd century bce, in a small number of dedicatory inscriptions at the sanctuary of the Nymph at Kafizin.4 We also have a high number of dialect glosses. In what follows, alphabetic forms belong to Arcadian and transcribed syllabograms to Cypriot. Note that many details of Cypriot phonology and morphology are obscured by the numerous ambiguities of the syllabic script.

Figure 1. Idalion bronze tablet.

Cabinet des médailles, Paris. Creative Commons License (CC BY-SA 3.0).

It is remarkable that some Greek nouns, verbs, and particles are rarely found outside Homer (and later poets) and Arcado-Cypriot, as in the case of δέ‎ (“and”) (= καί, κάς‎, ka-se elsewhere; Cypriot i-te = δέ‎); the Homeric epithet of Hermes, ριούνης‎ means “excellent runner” and must be linked to a unique root attested in Cypriot and Arcadian glosses (e.g., Hesych. o 1785 ονη· δερο. δράµε‎. ρκάδες‎ (in Arcadian, ονη‎ means “here, run”). Homeric οος‎ (“alone”) (< οϝος‎, cf. Avestan aēuua-, same meaning) is otherwise found only in Mycenaean o-wo-we (= οϝώϝης‎ “with a single ear [= handle]” and in Cypriot, dat. o-i-wo-i (“alone”).5 Apparently, Arcado-Cypriot continued to use some extremely archaic vocabulary that had become obsolete in the rest of the Greek-speaking world. Note, however, that new discoveries can change the picture: δέατοι‎ (“it seems”) is often found in Arcadian inscriptions, and δέατ'(ο)‎ “seemed” is found only once in Homer (Od. 6. 242). Although this is universally regarded as an Arcado-Cypriot feature in Homer, δέαται‎ has surfaced in a (Hellenistic!) inscription in Asiatic Aeolic.6po-to-li-se (= πτόλις‎ city) is amply attested in Cypriot, but only πόλις‎ in Arcadian inscriptions (yet Πτόλις‎, name of the citadel of Old Mantinea, is a plausible conjecture in Paus. 8. 8. 4). No explanation of initial πτ‎- is really convincing. Mycenaean has the personal name po-to-re-ma-ta = Πτολεµάτας; πτ‎- is also attested in Thessalian (assimilated ττολίαρχος‎, name of an official). Πτόλις‎ (“city”) and πτολίπορθος‎ (“destroyer of cities”) are frequently found in Homer.

The presence in Arcado-Cypriot of inherited phonology and morphology is remarkable, although in many cases, archaisms and innovations are inextricably intertwined. The retention of inherited [a:] (e.g. ‎̄µατα‎/a-ma-ta days < *Heh2m-; Homeric µατα‎) is attested in all dialect groups except Attic-Ionic; also in common with other dialects is the retention of [w], virtually in all positions, especially in the earliest inscriptions, e.g. ϝίσϝοις‎ (“equal”) dat. plur., κάταρϝος‎ (“cursed”), we-pi-ja (“words”) (= ϝέπια < ϝέπεα, ϝοικίας‎ houses). A notable feature of Arcado-Cypriot is a special development of the old labiovelars gw, kw, gwh (still alive and well in Mycenaean) before front vowels, a development apparently older than the one found in all the other dialect groups, in which these Indo-European phonemes are rendered with stops (e.g. πέντε/πέµπε‎ “five” < *penkwe, τλε/πήλυϊ‎ “far” < *kwēl-). As a matter of fact, before front vowels, Arcadian uses a special sign that, in all likelihood, was meant to render affricate sounds, while Cypriot has an [s]. In an inscription from Mantinea, 5th century bce, and in the newest one published by Carbon and Clackson, c. 500–450 bce , we find the outcome of the old labiovelars written with a special grapheme < ϟ‎ >, e.g. ϟις‎ = Attic τις‎ (“somebody”) (< k wi-), εϟε‎ = Att. ετε‎ (“or”) (< -k we), ϟελός‎ = Att. βελός‎ (“skewer”) (< *-gwel-; origin of ‎- unclear).7 In Arcadian glosses, < ζ >‎ appears as the outcome of gw- before [e] (Hesych. ε‎ 597 ζελεν·βαλεν‎ “he/she threw” and ζ‎106 ζέλλειν· βάλλειν‎ “to throw,” both < *gwelh1-); and in Cypriot τις‎ (“somebody”) is written si-se (= σις‎): the sibilant is confirmed by a gloss of Hesych. σ‎ 552 σί βόλε; ·τί θέλεις; Κύπριοι‎ (“the Cypriots”) say σί βόλε‎; for “what do you want?” (σί <kwíd).8 It is highly probable that the special Arcadian sign < ϟ >‎ found in the oldest inscriptions represented [dz] and [ts] (> [s] in Cypriot), and that these developments were chronologically earlier than the stops found in the other dialect groups.9 In later inscriptions, however, we find the familiar < τ‎ > < δ‎ > and < θ‎ >.

Various important developments are in common with other dialects, although variations and contradictions tend to be soft-pedalled in handbooks. It is often stated that in Arcadian, the outcome of the IE resonants [n̥], [m̥], [r̥] is not [a] and [ra]/[ar] as in Attic-Ionic and Doric, but [o], in agreement with Aeolic: Arcadian δέκο‎ (<-) (“ten”) = Att. δέκα‎, νϝότοι‎ (< *h1nu̯n̥t-) ϝέτει‎ (“in the ninth year”) (Att. νάτ‎), πανάγορσις‎ (< *-h2gr̥tis, probably -ᾱ́‎- with Wackernagel’s lengthening) (“festival”), Att. πανήγυρις‎. Yet in Arcado-Cypriot, [a] developments are not lacking: τριακάσιοι‎ (“The Τ‎hree Hundred”; -κασι- <‎ *-ḱm̥ti-), ϟεσϟάρο‎̄ν‎̣ “four” (gen. plur.) < IE *ku̯etu̯r̥-; Cypriot dat./loc. te-ka-to-i (before 500 bce )= δεκάτοι‎/-τ‎ “tenth” with -κατ‎- <ḱm̥t-), cf. Att. δεκάτ‎.10 In Arcado-Cypriot, oscillations between [a] and [o] as outcomes of vocalic liquids and nasals are far more frequent than in any other dialect.

In other cases, too, archaisms and innovations are inextricably intertwined, as in the use of ν‎ + acc. = Att. ες‎ + acc., e.g., ν δµον = εἰς δµον‎ (“to the people”), i-ta-ti-o-ne (= ν τν θιόν‎) = att. ες τν θεν‎ (“to the goddess”) (on ti-o see below).11 ν <ν‎ is an Arcado-Cypriot innovation, but ν‎ + acc. (also attested in other areas, Northwest Greek, Boeotian, Thessalian) is more archaic than ες‎ + acc., cf. Latin in patriam (“to the fatherland”). There is also a special development [e] > [i] between [m] and [n], µίνονσαι‎ (“staying”) (fem. plur., Att. µένουσαι‎), mi-no-ta-mo = Μινόδµος‎ (pers. name), att. Μενέδηµος‎. As a rule [e] becomes [i] before a back vowel in Cypriot (ti-o “god” = θιός < θεός‎) but not in Arcadian.12

Word-internal and final ti assibilate to si as in Mycenaean and Attic-Ionic: εκοσι‎ (same in Att., < -ḱm̥ti) (“twenty”), εδ‎' ν‎ ... φευρίσκωνσι‎ (-νσι <‎ -nti) (“if they find something more”), e-ke-so-si (= ξονσι‎), Att. ξουσι‎ (“they will have”). Also the modal particle ν‎ is the same as in Attic-Ionic.13 Hence Arcado-Cypriot κάς‎, ka-se = καί‎, probably < *kati, with assibilation and apocope.14 This is the same with πός‎, po-se < ποτί‎ (“toward”). In Cypriot, final and intervocalic [s] tended to become [h] with the passing of time, with various exceptions and chronological problems: po-ro-ne-o-i (“that they may consider”) (subj.) = φρονέωhι‎ (Attic φρονσι‎), to-po-e-ko-me-no-ne | po-se | to-ro-wo (“[land] that reaches the river”) (= τ(ν)ποχόµενον πς τ(ν)όFο(ν)‎ (= Att. τν προσεχόµενον πρς τνον‎).15 This is almost unknown to Arcadian, but note εδπς τι οκίαι µπόεστι κπος‎ (“if there is no garden adjacent to the house”) (= Att. εδπρός τοκίµπρόσεστι κπος‎), with πόεστι < πόσεστι‎.

Moreover Arcado-Cypriot simplifies -ss- to -s- precisely in the same instances as Att.- Ion. (e.g. µέσος‎ “middle” < IE *medhi̯o- etc.).16 But differently from the latter, Arcadian keeps a nasal before an [s] if it is the outcome of Greek-internal developments (sifflante forte, like -si <-ti/*-tsi), e.g. τος πολιτεύονσι‎ (Att. πολιτεύουσι‎) (“to those who are citizens”).

A remarkable archaism in common with Mycenaean but unattested in the other dialects is the -τοι‎ ending of the 3rd pers. sing. pres. middle, for example, [σ]ετοι‎ (“will be”) Att. σται‎; ν δέ τις ... φτοι‎ (“if one says”); ke-i-to-i (“she lies”) (Att. κεται‎): the -τοι‎ ending is older than the familiar -ται‎ one.17 The feminine participle of the verb “to be” appears with the old zero-grade of the suffix as ασα <ασσα <‎*h1sn̥tih2 vs. full grade οσα, οσα‎, οισα‎ etc. attested in other dialects again in agreement with Mycenaean (a-pe-a-sa “absent women” = Att. ποσαι‎).18 This archaism is also found in Messenian, and has recently surfaced also in (Hellenistic!) East Aeolic.19

In common with Thessalian, Boeotian, and Elean the subjunctive of the thematic forms shows the simple lengthening of the thematic vowel (εδ‎’ ν καταλλάσσε‎̄, probably “should he exceed”), not < -ηι >‎ (= -‎ in literary texts; either lengthening of -ει‎, or remodelling of -η‎ after -ει‎), familiar from Attic-Ionic.

Some Arcado-Cypriot active thematic infinitive forms show a remarkable affinity with their Ion.-Att. counterparts, e.g. -ην endings (φέρην‎ “bring,” with a long vowel more open than the one of Attic φέρειν‎, both < *-e-hen: in Arcado-Cypriot open vowels (<η>, <ω>‎) are the outcome of the contractions of *e-e and *o-o); yet we find short endings at Tegea (πέχεν‎ “remove” = Att. πέχειν‎; Doric influence); -εναι‎ is the ending of athematic actives (νθναι‎= Att. ναθεναι‎, both < *-the-enai), but the interpretation of Cypriot to-we-na-i (= δόϝεναι‎) is highly controversial.20

Other peculiarities are: Athematic inflection of verba vocalia: Arcadian δικήµενος‎ (= Att. δικούµενος‎) “suffering injustice,” ποίενσι‎ (Att. ποιοσι‎) “they do,” κύενσαν‎ (Att. κυοσαν‎) “pregnant woman”; Cypriot uncertain. In Arcadian ς‎ is the anteconsonantal form of ξ‎ (also found in other dialects), e.g. σδοτρες‎ = κδοτρες‎ “those who contract out”: there are “short” accus. plural of the thematic declension, in common with Thessalian and some Doric dialects, from -ονς‎ + consonant with elimination of [n], τςσγόνος‎ (“the descendants”) att. τοςκγόνους‎. In Arcadian the locative of the -o- stems is used for the dative singular, e.g. τι Θεσµοφόροι‎ (“to (Demeter) Thesmophoros”), but with oscillations (Cypriot unclear). The dat. plural is in -οις‎, not -οισι‎. There are special pronouns νί‎, νυ‎, Cypriot o-nu (“he”), gen. τννυ‎ (“of them”), νυ‎ probably being the enclitic particle found in Homer; το‎̃νυ‎ (“of him”) was found in an archaic Cretan inscription, SEG 27:631, probably a pre-Dorian remnant. The preposition νά‎ appears in Arcado-Cypriot as ν‎, also found in Thessalian and East Aeolic; ν‎ is often raised to ν‎, and “he/she dedicated” appears in Cyprus in three different shapes, a-ne-te-ke, o-ne-te-ke, u-ne-te- ke.

The -ευς‎ nouns (ερεύς‎ etc.) show -ής‎ in the nom., (φονε‎̄́ς‎ = Att. φονεύς‎ “killer,” ερής‎ = Att. ερεύς‎ “priest”) and -ήν‎ in the acc. (ερε‎̄́ ν‎ “priest” in Arcadian, not certain in Cypriot). If -ήν‎ is old, with -ēn < *-ēm by “Stang’s Law,” nomin. -ής‎ was backformed on the accusative.21

In Arcado-Cypriot, a final -o can appear as -υ‎ (pronounced [u], not [y]) in a number of cases: γένοντυ‎ = γένοντο‎ (“they became”), ke-no-i tu = γένοιτο‎ “may happen”), possibly πύ‎ = πό‎ (“from”) (not certain because Mycenaean a-pu seems to be an independent variant).22 The old (already Mycenaean -ο‎ gen. of the -ς‎ (Att. -ης‎) nouns appears in Arcado-Cypriot as -αυ‎, λεξιάδαυ‎ “(son) of Alexiadas,” Μιλτιάδαυ‎ “(son) of Miltiadas,” na-si-o-ta-u “(son) of Nasiotas,” o-pe-le-ta-u = φέλταυ‎ “of Opheltas.” At Tegea -αυ‎ was transferred to the fem. nouns (τςργωνίαυ‎ “of the contract,” τς ζαµίαυ‎ “of the penalty”).23 However the change -o > -υ‎ was far from systematic, e.g. δύϝο‎ “ten,” δύϝο‎ “two” (neither in Cypriot), ατό‎ “that,” (also Cypriot a-u-to). Probably the o > u change started early but in limited environments, and may have “acquired a wider distribution at a later stage, when the two dialects were already separated.”24

In Cyprus, a special thematic genitive written -o-ne is found, e.g. o-na-si-lo-ne (“of Onasilos”). It is probably a Cypriot innovation, with no convincing explanation. A special syntactical feature of Arcado-Cypriot, also interesting for its possible link with Mycenaean, is the use of the dative/locative instead of the genitive with prepositions meaning “from”: πτοῖ ἱερο‎ (Att. πτοῦ ἱερο‎) “from the shrine,” πτοῖ Ἀπόλλωνι‎ (Att. πτοῦ ᾽Απόλλωνος‎) “from Apollo,” e-xe-to-i wo- i-ko-i (= Att. κ τοοκου‎) “from the house”; note in Mycenaean pa-ro da-mo (da-mo probably representing dat. damōi or loc. damoi) “from the dāmos” (would have been παρά‎ + gen. in Att.-Ion).25

In sum, in spite of some features in common with Attic-Ionic and Aeolic, Arcado-Cypriot presents a number of lexical and morphological archaisms clearly linkable to Mycenaean, which, along with many other peculiar developments, set this dialect group remarkably apart from all the other ones.

Primary Texts

  • Egetmeyer, Markus. Le dialecte grec ancien de Chypre. Berlin: De Gruyter, 2010.
  • Masson, Olivier. Les Inscriptions chypriotes syllabiqes. Paris: Editions de Boccard, 1983.
  • Schwyzer, Eduard. Dialectorum graecarum exempla epigraphica potiora, Leipzig, Germany: Hirzel, 1923.


  • Buck, Carl Darling. The Greek Dialects, Chicago: University of Chicago Press, 1955.
  • Carbon, Jan-Mathieu, and James P. T. Clackson. “Arms and the Boy: On the New Festival Calendar from Arkadia,” Kernos: Revue internationale et pluridisciplinaire de religion grecque antique 29 (2016): 119–158.
  • Dubois, Laurent. Recherches sur le dialect arcadien, I: Grammaire. II: Corpus dialectal. III: Notes-Index-Bibliographie. Louvain-La Neuve, Belgium: Peeters, 1986.
  • Dunkel, George E. Lexikon der indogermanischen Partikeln und Pronominalstämme. Band 1: Einleitung, Terminologie, Lautgesetze, Adverbialendungen, Nominalsuffixe, Anhänge und Indices. Band 2: Lexikon. Heidelberg, Germany: Universitätsverlag Winter, 2014.
  • Lejune, Michel. Phonétique historique du mycénien et du grec ancien. Paris: Klincksieck, 1972.
  • Steele, Philippa M. Writing and Society in Ancient Cyprus. Cambridge, UK: Cambridge University Press, 2018.
  • Thumb, Albert. Handbuch der griechischen Dialekte. 2. erweiterte Auflage von A. Scherer, Heidelberg, Germany: C. Winter, 1959.