the language of the macedonians (lecture handout)

E2 — ALEXANDER’S LEGACY: GREEK AS A WORLD LANGUAGE New Light On… Lecture – The Language of the Macedonians (MJCS)

1

Language and Dialect in Ancient Macedonia: General Issues

THE MACEDONIAN KINGDOM CA. 336 BC (Map via Wikimedia Commons)

• Evidence for Language and Dialect in the Macedonian Kingdom:

- Direct Evidence (i.e. archaeological remains) o Coin Legends o Epigraphic Documents

- Indirect Evidence (i.e. via medieval MSS traditions) o Mentions of literary authors remarking on the language o ~140 glosses and material in the Lexicon of Hesychius o Mythological, personal, and local onomastics recorded in literary authors

• Two Important Facts: 1. All surviving linguistic evidence is written in Greek 2. The Greek attested already shows a great deal of dialectal diversity within the area controlled

by the Macedonian kingdom: o West Ionic in Euboean colonies o Insular Ionic in Cycladic colonies o Attic in former member states of the Delian League o Corinthian at Potidaea


2

o Attic-Ionic koiné in Macedonian official inscriptions Regarding the ‘Greekness’ of the Macedonians and their Language(s)

• Evidence from literary sources is divided on the original ethnic and linguistic identity of the Ancient Macedonians:

o Hesiod fr. 7: Eponymous ancestor Μακεδών born to Zeus and one of Deucalion’s daughters.

o Herodotus (5.22, 8.137-139) strongly states a belief that Macedonians were Greeks o Thucydides (2.80.5-7, 2.81.6, 4.124.1) possibly considers Macedonians to be βάρβαροι. o Aristotle (1324b) includes Macedonians in a list of ἔθνη with despotic practices (ἐν

τοῖς ἔθνεσι πᾶσι τοῖς δυναµένοις πλεονεκτεῖν), but makes no reference to language. o Demosthenes (3rd Olynth. 24, 3rd Philip. 31, False Embassy 308) claimed Philip II was

non-Greek o Isocrates (Philip 108, 154) claimed Philip was Greek, but ruled over non-Greeks

• Demosthenes and Isocrates may have had political reasons to attack or attribute Hellenicity to Philip II because of the political situation at the time.

• Later historical sources mention relate at points the use of speech ‘in Macedonian’ (µακεδονιστί < µακεδονίζειν) at particular points of emotional excitement:

o Diodorus Siculus (90-30 BC): 17.101.2 o Plutarch (45-120 AD): Alexander 51.6, Eumenes 14.5 o Quintus Curtius (ca. 1st c. AD): 6.9.34-36; 6.11.4 relates Macedonian as being

unintelligible. o On the other hand, Livy (BC 57 – 17 AD) 31.29.15 reports that Macedonians, Aetolians,

and Acharnanians all spoke the same language. • Similar verbs in Greek exist meaning ‘to speak in dialect’ (e.g. Βοιωτιάζειν, Δωρίζειν), just as ‘to

speak in language’ (e.g. Περσίζειν, Καρίζειν) with identical in word-formation. The Issue over the Macedonian Spellings < Β Δ Γ > for Greek < Φ Θ Χ >

• Much debate over Macedonian dialect/language has centred around the interpretation of these spellings

• Greek < φ θ χ > / pʰ tʰ kʰ / from PIE voiced-aspirated plosives *bʰ *dʰ *gʰ (and labiovelar *gʷʰ) • Some PNs in Macedonian and Attic:

o Maced. Βερενίκα ~ Att. Φερενίκη (*bʰer-) o Maced. Βάλακρος ~ Att. Φάλακρος o Maced. Βουλοµάγα ~ Att. Φυλοµάχη (*bʰh₂u-sl-) o Maced. Κεβαλῖνος ~ Att. Κεφαλῖνος (*gʰebʰ-l-)

• Some Macedonian glosses in Hesychius: o ἀδῆ· οὐρανός. Μακεδόνες (αἰθήρ < *h₂eidʰ-, cf. Lat. aedes) o δώραξ· σπλὴν ὑπὸ Μακεδόνων. (θώραξ) o δανῶν· κακοποιῶν. Κτείνων (*θανόω = θανατόω; Maced. δάνος = θάνατος Plut. Moralia

2.22c) o γόλα (γόδα ms.)· ἔντερα (possib. Att. χολή ‘gall’, Hom. χολάδες ‘guts’) o βροῦτες· ὀφρῦες. Μακεδόνες


3

Hypothesis 1: Macedonian is not Greek (Non-Greek Hypothesis)

• The devoicing of *bʰ *dʰ *gʰ *gʷʰ to Proto-Greek *pʰ *tʰ *kʰ and *kʷʰ is one of the most fundamental early isoglosses of Proto-Greek vis-à-vis Proto-Indo-European. It follows that if the spellings < Β Δ Γ > for < Φ Θ Χ > are to be interpreted as strictly a deaspiration of the voiced-aspirated series, then Macedonian cannot be classified as a Greek dialect.

Hypothesis 2: Borrowings/Interference Hypothesis

• Macedonian words and onomastics (attested in Greek) are borrowed from an (otherwise undocumented) Indo-European sub- or adstrate language related to Thracian or Phrygian where the sound change *bʰ *dʰ *gʰ > b d g takes place.

• Herod. 7.73 relates a tribe of Phrygians called Βρύγες that did not travel to Asia Minor: ὡς Μακεδόνες λέγουσι, ἐκαλέοντο Βρίγες χρόνον ὅσον Εὐρωπήιοι ἐόντες σύνοικοι ἦσαν Μακεδόσι, µεταβάντες δὲ ἐς τὴν Ἀσίην ἅµα τῇ χώρῃ καὶ τὸ οὔνοµα µετέβαλον ἐς Φρύγας. “As the Macedonians say, these Phrygians were called Briges as long as they dwelt in Europe, where they were neighbors of the Macedonians; but when they changed their home to Asia, they changed their name also and were called Phrygians.”

o Cf. Neo-Phrygian written in Greek script exhibits the sound-shift *bʰ > b in αββερετ and αββερετορ derived from PIE *bʰer-.

• Under such a hypothesis there would be two or more languages present in the Macedonian kingdom: Local Greek dialects and at least one non-Greek Indo-European language as a linguistic adstrate, accounting for the characteristic “non-Greek” features such as the reflexes of the voiced-aspirates

• The borrowings/interference hypothesis is one plausible way of looking at the evidence, but since we have no direct remains of the hypothesized Thraco-Phrygian language in the Macedonian homeland, this is, at least in part, a hypothetical solution.

Hypothesis 3: The Greek Hypothesis

• Argues that the spellings < Β Δ Γ > for < Φ Θ Χ > is the result of a secondary voicing lenition of original Greek *pʰ *tʰ *kʰ to [ v ð γ ] (via intermediary fricativization / f θ x /).

• The Greek Hypothesis attempts to explain Macedonian entirely as a dialect of Greek, eliminating the need to postulate a ‘Thraco-Phrygian’ substrate language advocated by the Borrowings-Interference Hypothesis

• Arguments in favour of this hypothesis o Ancient Sources that state that the Macedonians were Greek and spoke a dialect

similar to that of Aetolia and Epirus o Macedonian Glosses can be interpreted as Greek with distinctive phonetics. o Most Macedonians (from the earliest recorded historical period) had Greek names or

names interpretable as Greek, e.g. Φίλιππος, Ἀλέξανδρος, Περδίκκας (πέρδιξ?), Ἀµύντας (ἀµύνω), etc.

o Some short texts attest a variety of Greek close to Doric.


4

o HATZOPOULOS and PANAYOTOU claim to have found features attributable to a Macedonian substrate in the inscriptions in Attic or the Internationalized ‘Great Attic’ koiné.

The Question of the PIE Plosives in Macedonian

• The Greek hypothesis holds that the interpretation by contrast holds that the interpretation of the voiced-aspirated series as voiced stops is a misconception.

o NB: It is assumed on the basis of γενέσται (Pella tablet), the voiceless-aspirates had passed through an intermediary stage as fricatives (as probably in Northwest Greek i.e. *-stʰ- > -st(ʰ)- where < Φ Θ Χ > = / f θ x /; Cf. MÉNDEZ DOSUNA 1985:333-94)

• This hypothesis, built up from earlier hypotheses by HATZIDAKIS in the late 19th century and more recently by BABINOTIS (1992), has been elaborated most fully by HATZOPOULOS (2007).

o Simple deaspiration of the voiced-aspirated stops does not fully account for all the attested data:

o In certain contexts *p *t and *k also appear to undergo voicing in contexts V_V, R_V, and V_R, found in some epigraphic documents from Macedonia:

§ ἐν βλαγάνοις Ἀρτέµιδι Δειγαίᾳ (= ἐν βλάχανοις Ἀρτέµιδι Δικαίᾳ) § Βάλαγρος ~ Φάλακρος § Βορδῖνος ~ Πορτῖνος § Δρεβελαος ~ Τρεφέλεως

o Certain words cannot be explained by assuming simple deaspiration of voiced-aspirates: The case of the gloss κεβαλή for κεφαλή < PIE *gʰebʰal- (cf. OHG gebal ‘skull’)

§ Βέροια: Πόλις Μακεδόνων, ἣν φασὶν ἀπὸ Φέρητος τινὸς κτισθεῖσαν, Φέροια, καὶ κατὰ Μακεδόνας, Βέροια, τροπῇ τοῦ φ εἰς β, ὡς Φερενίκη, Βερενίκη, ἡ γυνὴ τοῦ πατρὸς Πτολεµαίου· τοῦ φ τραπέντος εἰς β. Καὶ τὴν κεφαλὴν, κεβαλὴν λέγουσι. (Etymologicum Magnum 195.36-40, cf. LSJ s.v. κεβλη)

§ Also PN Κεβαλῖνος = Att. Κεφαλῖνος § If a simple deaspiration of voiced aspirates was what happened we would

expect †Γεβαλῖνος

• To account for the voicing of the voiced stops as well as the etymological voiced-aspirated series the Greek hypothesis postulates a secondary sound shift in Macedonian:

PIE General Greek Macedonian*

*p *t *k /p t k / < Π Τ Κ > /p t k / < Π Τ Κ > ~ [ b d g ] < Β Δ Γ >

*bʰ *dʰ *gʰ /pʰ tʰ kʰ / (~ / f θ x / in some dialects) < Φ Θ Χ > /f θ x / < Φ Θ Χ > ~ [ v ð γ ] < Β Δ Γ >

*b *d *g /b d g / (~ [ v ð γ ] in some contexts) < Β Δ Γ > /b d g / < Β Δ Γ > ~ [ v ð γ ] < Β Δ Γ >

*contexts for allophonic voicing/fricativization lenition: V_V, R_V, V_R (Schematization after MÉNDEZ DOSUNA 2012)

• Typologically this is not entirely an unusual allophonic distribution and similar conditioning and secondary waves of voicing are found in Modern Spanish dialects, for example.

• Thus under this explanation problematic gloss κεβαλά and PN Κεβαλῖνος can be explained: o PIE *gʰebʰ-l- > Proto-Greek *kʰepʰ-l- > Post-Proto-Greek *kepʰal- > Maced. κεβαλ-


5

• The charge that this only occurs in onomastics is refuted from lexical glosses δώραξ ~ θώραξ, etc.

• Word-initial voicing must be generalized through sandhi phenomena: o ἁ Φερενίκα > ἁ Βερενίκα, τᾷ Φερενίκα > τᾷ Βερενίκα, ὦ Φερενίκα > ὦ Βερενίκα (etc.)

• Additionally the Pella tablet (below) may show features consistent with this analysis.

NB: The Greek Hypothesis also deduces evidence from Modern Greek dialects of Upper Pieria collected and evaluated by TZITZILIS (2008), this is beyond the immediate scope of the lecture to discuss, but cf. MÉNDEZ DOSUNA (2012:138-139).

Final Thoughts:

• Only the Greek Hypothesis can explain data like κεβαλά = κεφαλή (and etymologically derived derivatives). The operation of Grassmann’s Law guarantees -β- must be secondary from original *pʰ in this example.

• Further direct data for the voicing lenition is frustratingly scarce. The chronology of the proposed Macedonian sound-shift is unclear. If the Pella text is Macedonian, it is only attested through the problematic ΔΑΓΙΝΑ = δαπινά(?) (and nowhere else). Frustratingly the feature we want to see is only found in an unclear part of the inscription.

• One important point: There is no good reason not to suppose that the Greek Hypothesis and the Borrowings/Interference Hypothesis need be mutually exclusive.

• It may be possible that the Macedonian elites were ethnically ‘Greek’ in origin and originally speaking a dialect related to Northwest Doric, but ruled over a number of Non-Greek peoples. In bilingual situations we should not exclude the possibility that some Macedonian glosses or references to ‘Macedonian’ in literature may refer to either a local variety of Greek, or a non-Greek language of subject peoples.


6

Two Possible Macedonian Texts:

1) SEG 43:434 — Makedonia (Bottiaia) — Pella — ca. 380-350 BC. VOUTIRAS (1992/1993:43-48). Cf.

DUBOIS (1995:190-197), VOUTIRAS (1998).

[Θετί]µας καὶ Διονυσοφῶντος τὸ τέλος καὶ τὸν γάµον καταγράφω καὶ τᾶν ἀλλᾶν πασᾶν γυ- [ναικ]ῶν καὶ χηρᾶν καὶ παρθένων, µάλιστα δὲ Θετίµας, καὶ παρκαττίθεµαι Μάκρωνι καὶ [τοῖς] δαίµοσι· καὶ ὁπόκα ἐγὼ ταῦτα διελέξαιµι καὶ ἀναγνοίην πάλειν ἀνορόξασα [τόκα] γᾶµαι Διονυσοφῶντα, πρότερον δὲ µή· µὴ γὰρ λάβοι ἄλλαν γυναῖκα ἀλλ’ ἢ ἐµέ, [ἐµὲ δ]ὲ συνκαταγηρᾶσαι Διονυσοφῶντι καὶ µηδεµίαν ἄλλαν· ἱκέτις ὑµῶ<ν> γίνο- (5) [µαι· ․․․]αν οἰκτίρετε δαίµονες φίλ[ο]ι, ΔΑΓΙΝΑΓΑΡΙΜΕ φίλων πάντων καὶ ἐρήµα· ἀλλὰ [ταῦτ]α φυλάσσετε ἐµὶν ὅπως µὴ γίνηται ταυτα καὶ κακὰ κακῶς Θετίµα ἀπόληται· [․․․․]․ΑΛ[ 5 / 6 ]․ΥΝΜ․․ΕΣΠΛΗΝ ἐµός, ἐµὲ δὲ εὐδαίµονα καὶ µακαρίαν γενέσται· [— — —]ΤΟ․[— — —]․․․Ε․Ε․ΩΑ․․ΜΕΤΕ[— — —]

l.6 ΔΑΓΙΝΑΓΑΡΙΜΕ: DUBOIS δαπινὰ γάρ ἰµε

Translation (after VOUTIRAS 1998:15-16):

Of Thetima and Dionysophon the ritual wedding and the marriage I bind by a written spell, as well as (the marriage) of all other women (to him), both widows and maidens, but above all of Thetima; and I entrust (this spell) to Macron and to the daimones. And were I ever to unfold and read these words again after digging (the tablet) up, only then should Dionysophon marry, not before; may he indeed not take another woman than myself, but let me alone grow old by the side of Dionysophon and no one else. I implore you: have pity for [Phila ?], dear daimones, [for I am indeed bereft ?] of all my dear ones and abandoned. But please keep this (piece of writing) for my sake so that these events do not happen and wretched Thetima perishes miserably. [- - -] but let me become happy and blessed. [- - -]

Dialectal Features of the Pella Text:

1. /ᾱ/ for Att.-Ion. Η a. Θετίµᾱς (2) b. γᾶµαι (4) c. ἄλλᾱν (4,6)

2. -ᾶν as 1st decl G.pl. a. τᾶν ἀλλᾶν πασᾶν (1)


7

b. χρᾶν (2) 3. Apocope of παρά (before any consonant) and of κατά (before dental stops)

a. παρκαττίθεµαι (2) 4. D.sg. ἐµίν (= Att.-Ion. ἐµοί) 5. Adverbs in –κα

a. ὁπόκα (3) 6. Hyphaeresis Θεο- > Θετίµᾱς (2), Θετίµᾱ (7) 7. Optative in temporal clause referring to the future

a. ὁπόκα ... διελ<ί>ξαιµι καὶ ἀναγνοίην (3) 8. στ for σθ

a. γενέσται (8) 1-5 common to all Doric 1 not diagnostic (archaism) 2-3 innovations (but not exclusively Doric) 4-5 exclusive to Doric and Boeotian 6 sometimes Boeotian, Delphian, Coan, Megarian 7-8 exclusively Northwest Doric (i.e. Phokian, Lokrian) Two important features of the Macedonian text:

1. 1st Compensatory Lengthening (e.g. 2.pl. pron. *usme- Pella ὑµῶ<ν> vs. Aeol. ὔµµων) completed (unlike neighbouring Thessalian)

2. Athematic D.pl. δαίµοσι not Thessalian (exp. δαιµόνεσσι) nor NW Koiná δαιµόνοις (cf. Phokian, Lokrian πάντοις etc.)

3. We may add an additional feature if Dubois’ conjecture δαπινὰ γάρ ἰµε ‘for I am indeed bereft’? = Att. ταπεινὰ γάρ εἰµι showing the Maced. tendency to voice voiceless stops in certain contexts.

a. The meaning of the restoration makes sense in context, however it is unfortunate that the feature we want to see is in the conjectural reading.

Summary of the importance of the Pella text:

• The Greek hypothesis is first formulated on the evidence of the glosses and onomastics known from the tradition of indirect transmission, the Pella text gives us the first potential direct epigraphic attestation of a Macedonian Greek dialect.

• The features of the Pella text do not match up to any single previously known Greek dialect precisely as they are, but appears to be closest to Northwest Greek.

• It may provide evidence that the Macedonians did originally speak a dialect of Greek, similar but not identical to known Northwest Doric dialects (i.e. Phokian, Lokrian, Aitolian, etc.).


8

2) DAKARIS, VOKOTOPOULOU, & CHRISTIDIS (forthcoming) no. 2493A, Dodona Museum 871; cf. MÉNDEZ

DOSUNA (2007:283-4)

Ζεῦ καὶ Διώνα, ἦ ἔσσονται παῖ- δες ἐκ τᾶς γυναικὸς Κεβαλίωι τᾶς νῦν ἔχει κ[α]ὶ ζώσοντι;

Translation (after MÉNDEZ DOSUNA 2012:144):

Zeus and Diona, shall Kebalios have children from the wife he has now, and shall they survive? Dialectal features of the Dodona Text: Text cannot be Attic-Ionic on account of:

1. Retention of original *ā in Διώνᾱ, τᾶς 2. Retention of geminate -σσ- in ἔσσονται 3. Article as relative in τᾶς νῦν ἔχει

Text cannot be Boeotian on account of:

4. Spellings ΑΙ in ἔσσονται, γυναικός, κ[α]ί, and ΕΙ in ἔχει (Boeot. ἔσσοντη, γυνηκός, κή, ἔχι) 5. Earlier *di, *gʷi > <Ζ> in Ζεῦ, ζώσοντι (Boeot. Δεῦ, δώσονθι) 6. ἐξ > ἐκ before a consonant in ἐκ τᾶς (Boeot. ἐς τᾶς) 7. Unaspirated stops in third plural endings ἔσσονται, ζώσοντι (Boeot. –νθη, –νθι) 8. Thematic dative singular –ωι in Κεβαλίωι (Boeot. –οι)

Text cannot be Thessalian on account of:

9. Spellings <Η> and <Ω> in ἦ, Διώνα, Κεβαλίωι, ζώσοντι (Thess. should normally have <ΕΙ>, <ΟΥ>) 10. ἐξ > ἐκ before a consonant in ἐκ τᾶς (Thess. ἐς τᾶς) 11. Thematic dative singular –ωι in Κεβαλίωι (Thess. –ου or –οι) 12. Unaspirated stops in third plural endings ἔσσονται, ζώσοντι (Thess. –νθαι, –νθι)

Further Reading: BABINOTIS, George. 1992. “The Question of the Mediae in Ancient Macedonian Greek Reconsidered” In:

Historical Philology: Greek, Latin, Romance, ed. by B. BROGYANYI & R. LIPP, 29-40. Amsterdam: John Benjamins.

*BRIXHE, Claude. 1997. “Un ‘nouveau’ champ de la dialectologie grecque: le macédonien” AION: Annali del Dipartimento di Studi del Mondo Classico e del Mediterraneo Antico - Sezione linguistica 19:41-71.

BRIXHE, Claude & Anna PANAYOTOU. 1988. “L'Atticisation de la macedoine: l'une des sources de la koiné” Verbum 11:245-60.

*CRESPO, Emilio. 2012. “Languages and Dialects in Ancient Macedon” In: In GIANNAKIS (2012) pp.121-31.


9

DAKARIS, S., I. VOKOTOPOULOU, & A. Ph. CHRISTIDIS. (forthcoming). Τα µολύβδινα χρηστήρια πινάκια της Δωδώνης των ανασκαφών Δηµητρίου Ευαγγελίδη. Athens.

*DUBOIS, Laurent. 1995. “Une tabelette de malédiction de Pella: S'agit-il du premier texte macédonien?” Revue des Études Grecques 108:190-97.

*GIANNAKIS, Georgios, K. (ed.) 2012. Αρχαία Μακεδονία: Γλώσσα, ιστορία, πολιτισµός. Thessaloniki: Centre for the Greek Language. [Contains articles by CRESPO, MÉNDEZ DOSUNA in Greek, English, French, and German] Freely available online at: http://ancdialects.greeklanguage.gr/sites/default/files/studies/_makedonia_tomos.pdf

HAMMOND, N. G. L. 1994. “Literary Evidence for Macedonian Speech” Historia 43:131-42. HATZOPOULOS, Miltiades. 1998. “Épigraphie et philologie : récentes découvertes épigraphiques et gloses

macédoniennes d'Hésychius” Comptes-rendus des séances de l'année... - Académie des inscriptions et belles-lettres 142:1189-218.

*HATZOPOULOS, Miltiades. 2007. “La position dialectal du macédonien à la lumière des découvertes épigraphiques récentes” In: Die altgriechischen Dialekte: Wesen und Werden - Akten des Kolloquiums Freie Universität Berlin 19.–22. September 2001, ed. by I. HAJNAL, 157-76. Innsbruck: Institut für Sprachen und Literaturen der Universität Innsbruck.

MÉNDEZ DOSUNA, Julián. 1985. Los Dialectos Dorios del Noroeste: Gramática y estudio dialectal. Salamanca: Ediciones Universidad de Salamanca.

MÉNDEZ DOSUNA, Julián. 2007. “Παρατηρήσεις στις νέες µαντειακές πινακίδες της Δωδώνης.” Μελέτες για την ελληνική γλώσσα [Studies in Greek Linguistics] 27:277-85.

*MÉNDEZ DOSUNA, Julián. 2012. “Ancient Macedonian as a Greek Dialect: A Critical Survey on Recent Work” In GIANNAKIS (2012) pp.133-45.

*MORPURGO DAVIES, Anna. 1987. “The Greek Notion of Dialect” Verbum 10:7-28. *PANAYOTOU, Anna. 2007. “The position of the Macedonian dialect” In: A History of Ancient Greek: From

the Beginnings to Late Antiquity, ed. by A.F. CHRISTIDIS, 433-43. Cambridge: Cambridge University Press.

SCARBOROUGH, Matthew J. C. 2013. Review of GIANNAKIS, Giorgios (ed.) 2012. Αρχαία Μακεδονία: Γλώσσα, ιστορία, πολιτισµός. (Thessaloniki: Centre for the Greek Language). In Bryn Mawr Classical Review. http://bmcr.brynmawr.edu/2013/2013-08-21.html. Bryn Mawr: Bryn Mawr College.

SOWA, Wojciech. 2006. “Die makedonischen Glossen in antiken Quellen” Studia Etymologica Cracoviensia 11:115-33.

TZITZILIS, Christos. 2008. “Τα αρχαία µακεδονικά στοιχεία στο ιδίωµα της ορεινής Πιερίας και οι σχέσεις της µακεδονικής µε τη θεσσαλική και της βορειοδυτικές δωρικές διαλέκτους” In: Θερµή και Φώς / Licht und Wärme, ed. by M. THEODOROPOULOU, 225-44. Thessaloniki: Centre for the Greek Language.

VOUTIRAS, Emmanuel. 1993. “Ένας διαλεκτικός κατάδεσµος από την Πέλλα.” Ελληνική Διαλεκτολογία 3:43-48. VOUTIRAS, Emmanuel. 1998. Διονυσοφῶντος γάµοι: Marital Life and Magic in Fourth Century Pella.

Amsterdam: Gieben.