Abstract

This study examines the debated linguistic relationship between Thracian and Illyrian, focusing on the proposed connection between the Thracian form MEZENAI and the Illyrian Menzanas, often linked to Albanian mëz (“foal”) and Romanian mânz. Through a critical reassessment of epigraphic evidence—particularly a Thracian inscription from the Arabaj tumulus—the paper challenges earlier interpretations by V. Georgiev, arguing that the reading connecting MEZENAI to a horse deity is unconvincing. A revised interpretation by S. Kaloyanov is presented as more systematic and plausible, undermining the supposed Thracian–Illyrian correspondence.

While rejecting the direct link between the Thracian and Illyrian forms, the study supports a broader Indo-European etymological network connecting Illyrian Menzanas, Albanian mëz, and Romanian mânz, alongside related forms in Latin, Celtic, and possibly Basque. The semantic consistency of “young animal/foal” across Balkan languages is highlighted as significant. Ultimately, the paper demonstrates how differing interpretations of fragmentary linguistic evidence can lead to fundamentally different conclusions.

Thracian–Illyrian linguistic parallels:
Thrac. MEZENAI ~ Illyr. Menzanas
Irina A. Kaluzhskaya (Moscow)

This paper addresses a rather difficult and controversial problem concerning the establishment of similarities between the linguistic remains of the Thracian and Illyrian languages. The essence is that conclusions about one or another correspondence often depend on the interpretation of the available material.

For this very reason, the preliminary analysis of linguistic units is so important. All of the above can be illustrated by a concrete case, where it becomes clearly visible how the interpretation of the initial facts influences the conclusion.

The Messapic epithet for Jupiter—Menzanas, attested by Festus in a passage describing the tribe of the Sallentini of southern Calabria (Sallentini, apud quos Menzanae Iovi dicatus [sc. equos] vivos coniciuntur in ignem – “The Sallentini, who during a sacrifice throw into the fire alive [i.e., horses]” – Festus f. 190, ed. M. Lindsay), was convincingly interpreted by Stier (KZ 1863, XI: 148). He successfully identified it with Albanian mëz (Gheg mâz) m. ‘foal’ and Romanian mânz m.

G. Meyer (1891: 276) accepted this identification, linking mëz with the Albanian verb mënd ‘to suck’.

V. Georgiev (1977: 111–114) discovered the Thracian equivalent of Illyrian Menzanas in the final fragment of a well-known Thracian inscription on a ring from the Arabaj tumulus (5th century BC) in the Duvanli region: HΥΣΙΗ…..ΔΕΛΕ ΜΕΖΕΝΑΙ. Since the inscription accompanies the depiction of a horseman, Georgiev considered them closely related and thought it included the name of the Horseman God.

Thus, the form MEZENAI was linked with Illyrian Menzanas, Albanian mëz, and Romanian mânz. Georgiev also assumed that the Thracian name Μεζαναία (Detschew 1957: 291) derives from the same root. According to the Bulgarian scholar, the fragment Μεζεναι means “the horseman” and appears to be the name of a well-known horse deity found in many reliefs in the eastern part of the Balkan Peninsula.

The entire inscription was interpreted by Georgiev as: “mighty (or swift, life-giving), help… protect… the Horseman.” The loss of the consonant v in the Thracian word was explained either as a result of dissimilation from the second v, or due to its position before ζ, similar to Albanian Lab dialect mëz. However, these phonetic assumptions cannot be proven by linguistic data.

First of all, it should be emphasized that the process of v-loss is not attested in the available Thracian material. The analogy with Albanian is not appropriate here, since the denasalization of ã is a relatively late dialectal process (Çabej 1959: 553). Moreover, Georgiev’s interpretation of the inscription itself is not as indisputable as it seemed to him.

More recently, S. Kaloyanov has proposed a new and, in our opinion, more convincing reading of the inscription (Kaloyanov 1988). The fact is that the signs on it are quite damaged (especially in the second fragment), therefore they can be interpreted in different ways. For example, the first publisher of the Arabaj inscription, B. Filov (Filov 1934: 129–131), read it as HΥΣ (or HΨΣ) AΕΛΕ ΜΕΖΕΝΑΙ.

Kaloyanov noted four silver vessels found in the neighboring Bashovo tumulus. All four contain the same inscription: ΔΑΔΑΛΕΜΕ. Based on this data, he proposed a new reading: HΥΣ ΔΑΔΑΛΕΜΕ ΖΗΝΑΙ.

In this way, it appears that the epigraphic material from the two neighboring and evidently related tumuli (it is not excluded that they belonged to relatives) contains the same graphic sequence: ΔΑΔΑΛΕΜΕ. The segment IH is well known from other Thracian inscriptions, for example from a phiale from Branichevo (Georgiev 1977: 137, 138), and may represent a numerical sign, perhaps a date or a year.

The complex ZHNAI is identical to Thracian personal names Ζηνᾶς, Ζηνίς, Ζηνιος (Detschew 1957: 184). Thus, despite the a priori nature of some of Kaloyanov’s assumptions, they appear to have a systematic character, which makes them more convincing than Georgiev’s hypothesis.

The conclusion that can be drawn from the above is that the connection of the Thracian sequence MEZENAI with Illyrian Menzanas, Albanian mëz, and Romanian mânz is impossible.

As for the origin of Illyrian Menzanas and its connection with Albanian mëz and Romanian mânz, it has already been mentioned that G. Meyer (1891: 276) accepted this correspondence and reconstructed Proto-Albanian *manza < *mandja- < *mondjo-, linking mëz (according to W. Tomaschek – BB 1885, IX, 101) with Albanian mënd ‘to suck’ < Indo-European mend- (Greek μαστός ‘breast’ < *μασ-τός < *mnd-to; Old High German manzon ‘teat’; Sanskrit mándati, mádati ‘to flow’). (Another view is given by Pokorny: 694–695).

G. Meyer classified Albanian mëz, Romanian mânz, as well as several similar forms from Romance languages (Italian manzo ‘bullock’, Sardinian Manzu ‘young bull’, Trentinian manza ‘young cow’, Gröden mans ‘bull’, mánža ‘heifer’, Bavarian manz, menz ‘barren cow’, Rhenish Minzekalb ‘heifer’) as “an Alpine Illyrian word” (Meyer 1891: 276).

Krahe (1955: 84–85, 103, 115) connected Latin mannus ‘a small horse of Gallic breed’ with an Illyrian source (< Illyr. manda-), including in this family Illyrian names such as Menda, Mandeta and place names in Calabria such as Manduria (Liv. 27, 15; Plin. 2, 102), Μανδύριον (Stef. Biz.), as well as Mandonia in Lucania and Μανδαραί in Macedonia.

This range can be expanded with Thracian names such as Μενδας in the Byzantium area, as well as toponyms Μένδη, Μέvδα, Mendaeum, Mendae—the city of the Sithones on the Pallene peninsula, and Μενδενίς—the name of the nymph who gave birth to the Pallene peninsula, whose daughter was Sithone (Detschew 1957: 293–294).

There is also a family of related Celtic forms: Middle Irish menn (< mendo-), mennan ‘young animal, calf, foal’, Gaelic mionnseach ‘kid’, Welsh mynnen ‘young’, Cornish min ‘kid’, Breton menn ‘young animal’, menn gaur ‘kid’ (Pokorny: 729), possibly also personal names Mendus (m.) and Menda (f.) (Holder 1961–62, II: 549). Of particular interest is Basque mando ‘mule’, which has been considered either as a pre-Indo-European inheritance (Hubschmid 1954: 28 ff.), or as a Celtic loan into Basque (Ernout-Meillet 1960: 558); the latter seems more acceptable.

It is worth noting that the phonetic shape of Albanian mëz allows the reconstruction (unlike G. Meyer) of a prototype with either the vowel o or e, cf. Albanian Tosk rërë, Gheg rânë ‘sand’ < Latin arena; Tosk qendër, Gheg qândër ‘center’ < Latin centrum; Tosk mënd, Gheg mẫnd ‘mind’ < Latin mentem, etc.

Thus, Albanian mëz may reflect the same prototype as Illyrian Menzanas and Celtic mendos. Another ablaut grade is represented in Latin mannus, Basque mando, and the Romance forms mentioned above. Unfortunately, we do not have more precise phonetic criteria to determine whether the Romanian word was borrowed from Albanian, or whether both derive from Illyrian, or all three from a common prototype. An indirect indication that the Balkan forms are closer to each other lies in their semantics, namely the fact that they denote a young horse (foal), in contrast to Celtic and Romance terms for young horned cattle.

All of the above appears to be a sufficient illustration of how the analysis and interpretation of initial facts can influence conclusions.

References

Irina A. Kaluzhskaya
“Thracian–Illyrian Language Parallels:
Thrac. MEZENAI ~ Illyr. Menzanas”
Institute for Slavic & Balkan Studies
Leninsky Prospekt 32-A, Moscow, Russia