Română(articol de popularizare) English abstract
κάγα: an important Dacian word in Tomitan inscriptions1
The next lines will briefly present you a discovery I made in 1989: that of the first getic word found in epigraphic material, fortunately also an important one.
The first document I rely upon is the inscription nr. 128 from the IInd volume of the series "Inscriptiones Daciae et Scythiae Minoris Antiquae", containing the inscriptions found at Tomis (today Constanţa at the Black Sea), a volume edited and published by our regretted specialist in Tomitan inscriptions, Iorgu Stoian. The text of this bilingual - Latin and Greek - document was read by its editors as follows:
Ti(berius) Claudius Mu-
casius v(otum) s(olvit) l(ibens) m(erito)
Ηερώϊι (sic!) ΚΑΤΑ Τι(βέριος)
The first lines of the text, those in Latin, easy to read at the discovery of the stone but almost illisible today because of the poor conditions it was preserved, raise no problem of interpretation: "Sacrum to Heros. Tiberius Claudius Mucasius respected his vote, gladly and on merit". Let us see what the Greek part says: "To Heros KATA. Tiberios Claudios Mukasios consacrated as promised." If, as we saw, the Latin variant is fairly clear, in the Greek one we find instead this KATA, unsatisfactory explained by any of its succesive editors. The first of them, Grigore Tocilescu, who also discovered the monument in 1895, trying to find a plausible meaning for the sequence KATA, took it as an adjective of Heros and interpreted it as an abbreviation of κατα(χθόνιος) which he considered to be an attribute or a hypostasis of the Thracian god. The lack of any other more consistent explanation of this word made his interpretation finally accepted in the litterature. It has nevertheless at least two weak points which draw our attention: 1. it supposes a considerable and unexpected difference between the Latin and the Greek variants of the text and 2. it proposes an attribute twice unique (and I insist to enounce here an important principle, which I shouldl formulate as Always beware of hapax!): it would be the only time that καταχθόνιος would be abbreviated κατα and the only time that the epichoric god Hērōs would be called "of the underworld". All these difficulties could have been avoided if both Tocilescu, but especially the later editors, would have taken a closer look at the stone, where - as you can see by yourself in the images below - the lapicida wrote not ΚΑΤΑ , but ΚΑΓΑ!
Fig. 1 Inscription nr.128 - photo and drawing
Click on the image to enlarge
It is true that this new lecture would not have inlightened, right the way and by itself, the meaning of the text, but would have at least avoid to let go the inexistent κατα(χθόνιος), an epithete that Hērōs, of course, never bore. If we leave apart, for the moment, the etymologic discussion and consider this word not an adjective of Heros but an equivalent of the Latin sacrum, then all the difficulties mentioned above disappear: sacrum to Heros. etc. would suddenly become identical with kaga to Heros. etc. Though already more plausible, the unusual of this interpretation and the deontological prudence would still oblige us to admit other possible explanations, for example an error of the carver, and to subsequently consider the term as doubtful. But, fortunately again, here comes a second inscription to insure the validity of this new found word: it is nr. 36 of the same volume, the left half of a marble plate holding the following inscription:
Δήμητρος ἄρχο[ ντες… ]
Ποσειδώνιον Νουμηνίου Καλλίμαχον Κα[… Διο -]
νύσιον Διοσκουρίδου ΚΑΓΑ Ηρωί[δου ἐτίμησαν ]
|Under Demeter's priesthood, the archo[nts…
Poseidonios of Noumenios, Kallimachos of Ka[llimachos?,… Dio]
nysios of Dioskurides, Kaga Ērōi[…
as you can see in these pictures
Fig. 2 Inscription nr.36 - photo and drawing
Click on the image to enlarge
The string of names in the accusative determined Iorgu Stoian to take this text as a honorific dedication, erected by the arhonts in the honour of some citizens of the town. Following this logic, Kaga had to be itself a first name and the following Ērōi[..] the beginning of the patronym, restored by the editor, exempli gratia, Hērōi[dou]. If we look carefully this precise spot on the photography, we could notice that after Hērōi we don't find, as expected, the rest of a letter, but, more probably, a blank. Anyway, the lecture ΚΑΓΑ is for the second time assured and it obviously cannot be a coincidence that its neighbour is, again, Hērōi. In my eyes, these reasons were enough to convince me that this word indeed existed and was still used in Tomis towards the beginning of the IInd century. Its meaning was, very likely, the same as that of the Latin sacrum. Once the job of the epigraphist done, the linguist must take control and explain where could this word come from.
Kaga cannot be, for sure, neither Latin, nor Greek. It doesn't look scytho-sarmatian either, because there is no such word in any of the iranic languages we know. One could say that the ending -aga is much like the ossetic adjectival suffix -äg, but it is not the case here, since in our word the sequence ag must be part of the root. There is therefore no reason why we should not consider this word autochtonous in Moesia, quite on the contrary. At the beginning of the first century AD, the unwillingly Tomitan poet Ovidius complained that at his time in Tomis were only spoken Greek and Getic. This is, of course, an exageration which we can understand and forgive, but it is definitely not an invention of the sensitive Roman poet. The inscriptions found in this left pontic metropol confirm that, at that time, Roman speakers were indeed rare. This is the age of the "mixellenic Tomis". One hundred years after, when our inscriptions dates from, the situation changed considerably. In 46 AD, under the emporor Claudius, this region, until then part of the Odrysian kingdom, known as Ripa Thraciae, became the Roman province Moesia. It was the greatest crosspoint in the history of the peoples living here, because since that moment they changed the old Greek cultural domination with the new Roman one. Moesia will be from now on the spinebone of the Romanity in the Haemus paeninsula. The status of Tiberius Claudius Mucasius is an illustrative example of the ethnic and linguistic reality of these regions at the beginning of the IInd century AD. He was a Roman citizen, as his tria nomina prove. His cognomen, very likely pronounced Mucazius (because the panthracian form of this name was Mukazis), shows us that he comes from the traditional layer of population in Tomis, that one shown by Ovidius: a mixture of Greeks and Geto-Moesians, maybe also of Thracians (as those Bessi or Lae consistentes, from the surrounding villages). Another indice of his Moesian origin is the very fact that he worships Heros, whose sanctuaries and dedications are most frequent in this province, mainly around eastern towns such as Marcianopolis or Odessus. Therefore we must expect kaga to be a Geto-Moesian word.
There is nevertheless another decissive argument that comes to solve for good both its linguistic origin and its meaning. I speak about a well known text of Strabo's Geography which mentions the oronym Kōgaionon, a text whose information was not well enough interpreted until now. Here you are the Geographer's words, relating about the place with cave in the mountains where Zamolxis had retired (7.3.5 sqq.): "…καὶ τὸ ὄρος ὑπελήφθη ἱερόν , καὶ προσαγορεύουσιν οὕτως‧ ὄνομα δ᾿αὐτῷ Κωγαίονον ὁμώνυμον τῷ παραρρέοντι ποταμῷ." As we see, Strabon says, as clear as possible, that the name Κωγαίονον is in fact the adjective κωγαίονον and that it means ἱερόν. The ressemblence in form and the semantic identity of κωγαίονον(= ἱερόν) and καγα(=sacrum) are enough reasons to conclude that both forms belong to a single Geto-Moesian root.
Accepting our hypothesis, prof. Dan Sluşanschi (at the Faculty of Classical Languages of the University of Bucharest) made an etymological study upon this word. He proposes, as a primary etymon of this root, the PIE *∂3kw - > okw "eye". From here came thereafter the derivated *(∂3)kweg(h) "obvious", inherited in Daco-Moesian both as *kōgayō, a verb of intensive aspect having *kōgayonom as a participle, and *kogos, its corresponding noun which became, according to the already known phonetic laws of this language, *kaga(s). As for their meanings, prof. Sluşanschi concludes that if "καγα = sacrum, then Κωγαίονον seems to be = consecratum". From the initial meaning "to show (himself), to manifest (himselfself)" this word naturally got the meaning "to indicate" and then "to say". This very stage was kept in Slavic languages in (s)kazat'(sya), from the same root as kaga. The next step, the transition from "to show, to say" towards "to consecrate" seems to have taken place in Daco-Moesian only, but other IE words, which suffered the same evolution, legitimate it. For example IE *deik "to show" (preserved as such in gr. δείκνυμι , got. táíknjan etc.) became in Latin both dico, -ere "to say" and dico,-are "to consacrate"; from the same root the sanskrit has diśati "to show", but also dikşati "to consecrate, to sanctify". But, returning to professor Sluşanschi's article, though his authority imposes his explanation as the most probable (provided that kaga indeed comes from an IE *kwag-), we must say that it might also come from other IE roots by some not yet known Thracian phonetic mutations; this is why I shall propose two other possible explanations in a future work.
At the end I should like to draw your attention upon some methodological issues. The discovery I just presented to you also demonstrates that ancient sources could still hide important information, despite the pessimism of some researchers in this regard. In order to bring it to daylight, Thracologists must review these sources themselves and must never accept others' conclusions without verifying them. Assuming the risic to repeat myself, I shall underline, once again, that the "eye" of a researcher experienced in TDM linguistical materials could recognize potential autochtonous words or names there where other classicists would say "hic sunt leones".
1. The present work is a newer and more detailed remake of the article "Kaga and Kogaionon - the Issue" by Sorin Olteanu (see also "Kaga and Kogaionon - Philological and Linguistical Analysis" by prof. Dan Sluşanschi), in "Traco-Dacica", 1989, tom X, p. 216 and 219, both in Roumanian, with short English abstracts. The conclusions I reached in 1989 are today accepted by Roumanian historiography, cf. "The History of the Roumanians" (Alexandru Vulpe), vol.I, 2001, p.435.
2. Today I am asking myself more and more if this καγα isn't rather an old loan from gr. ἅγιος, or the end of the evolution of another stem. One could hypothesize that it may come, for example, from a prefixed *sag- 'holy, saint' (the etymon of *ἁγιο-); cf. in this respect κινούβοιλα which, if comparations work, is akin to the lit. šunobuõlas. There are a lot of examples of an alternance between aspiration and velar consonant, like in Carpi/Ἅρπιοι, Daci/Dahae, Hebrus/Cebrus a.s.o. It might be bound also to the name Καιγεισος which can be very well explained from *kaygo-(which is *kagyo-[*kag-yo] with metathesis of iota), if this name is not only a variant of Κειάγεισος, which is certainly another dish.