Pyrgi tablets

The Pyrgi tablets.

A structural comparison of Etruscan with the Kartvelian languages

by Dan Alexe

Etruscan, the language of the people that dominated central and northern Italy from prehistoric time until the rise of Rome, has hitherto not been entirely deciphered. The riddle posed by the nature of the tongue of the early masters of Rome remains a permanent irritant; the more so since the solution to this enigma would help shed a new light on the early history of Mediterranean civilisation.

The present approach focuses on a structural comparison of Etruscan with the South-Caucasian (Kartvelian) language family. We find a complete concordance with Kartvelian of the whole system of the attested Etruscan casual terminations, but also an identity of their usage, which is so unusual and complex as to exclude any explanation by coincidence. We also find a compelling number of core cognates and the possibility of applying a typological cross-verification: by using grammatical patterns from one of the two compared systems one is able to predict similar patterns in the other.[1]

”One structural feature in language predicts another, implies its presence, or limits its functional or distributional possibilities“.[i] The question whether there is a logical predictability between two grammatical systems that present similar traits can be answered only after a detailed investigation, primarily involving those similarities that go beyond the statistical possibility of a mere typological coincidence.

Here I will argue that Etruscan and the Kartvelian (South-Caucasian) language family (Georgian, Mingrelian, Laz, and Svan) are not only typologically similar, but possibly also genetically related. Although some correspondences and similarities between Etruscan and Kartvelian have been suggested by linguists such as Pauli [ii] and Thomsen[iii], one century ago, access to scientifically presented Svan material has been scarce until the recent decades. Secondly, comparisons have always been made in a piecemeal way, sometimes even by bringing in supposed similarities with the other two, unrelated Caucasian language families, the Adyghe and the Nakho-Daghestanian (both spoken on the Northern slopes of the Caucasus mountain range). A systematical, structural comparison of Etruscan and Kartvelian (especially Svan) has never been performed until now.

The fact that the two language systems (Etruscan and Kartvelian) present similarities, grammatical and lexical, that seem to suggest linguistic kinship would have huge implications for the study of the birth of European civilisation, since the Etruscans, the founders and rulers of Rome for centuries, passed on to the Romans an important part of their culture and skills.


The nominal morphology


When we compare the morphology and the grammatical structure of the Kartvelian languages, especially that of Svan, with what we know about the Etruscan, a series of remarkable correspondences become apparent. First of all, identical case terminations and the dominance in both systems of an overused ”dative“ case. For lack of a better nomenclature, this case has been labeled ”dative“ in Kartvelian linguistics, although its uses overlap with those of the accusative in other linguistic structures.

As has been shown[iv], Etruscan is a language with no grammatical genders or nominal classes and with no accusative construction. These two facts alone would not suffice to allow for the inclusion of Etruscan into the Kartvelian group, which also shares these traits. Nevertheless, a closer comparison shows that the similarities go much further. The system of declension in Etruscan and Svan (and Kartvelian at large) is identical in its main points, that is: the lack of specific terminations for the nominative (except for the Georgian recent –i) -combined with the absence of the notion of ”accusative“, the direct and indirect objects of a transitive verb being both indicated by the ”dative“ case-, down to the similarity of the terminations themselves.

Like in Svan, the Etruscan plural is formed by the addition of the particle –er, –ar. In Etruscan, as well as in Svan, there is a multiple noun declension and case terminations of the plural are identical to those in the singular, simply added by agglutination to the noun in the plural. In both languages, there is ablaut of the root of the declined noun in the oblique cases.

As has been well established within Etruscan studies, the declension of a well known term such as clan (son), known from countless funerary inscriptions, is:

………………SG                       PL

NOM          clan                      clen-ar

Gen          clan-s                   clen-ar-a-s

DAT          clen-s                   clinii-ar-a-s

Similarly, in Svan, a one-syllable noun such as xäm (pig) gives:

……………….SG                       PL

NOM          xäm                     xam-är

GEN           xäm–iš                 xam-är–iš

DAT           xäm-s                   xam-är-s

In Etruscan, as well as in all Kartvelian languages, the terminations of the genitive and of the dative are very similar, to the point of becoming almost identical, as in Georgian. They are, thus, in Georgian, Laz, Mingrelian and Svan :

………..G       LAZ  MG    SV

Gen. –is      –š      –š      –iš

Dat.   –s      –s      –s      –s

The similarity of the terminations for genitive and dative in Etruscan allows us to address the riddle of the inscription adorning a kylix discovered in Tarquinia (Testimonia Linguae Etruscae, 156): itun turuce venel atelinas tinas cliniiaras. The unanimously accepted translation is: ”This was offered by Venel Atelina to the Disokouroi“ (i.e. Tina’s sons, tinas cliniiaras, Tin, or Tina, being the supreme god) seems to carry the same termination for both genitive and dative. Tina (Jupiter) is in the genitive (Tina-s), while clenar/cliniiar (cliniiar-as) in the dative. The Etruscan Tinas cliniiaras (to Tina’s sons), functions exactly like in Old Georgian kacisa ʒesa, or kacis ʒes, ”to the son of man“, where kac-is(a) is in the genitive (from kaci, man), and ʒe-s(a) is in the dative (from ʒe, son).

This could pass for mere chance, were it not for the total concordance in the special uses of this case, which, for lack of a better term, has been traditionally called ”dative“. By applying the grammatical pattern of the Kartvelian languages, we immediately understand the peculiar Etruscan uses of the dative, which could not be explained until now.

Kartvelian languages have no accusative case, using in its place what we call the ”dative“. For Etruscan, no accusative has been postulated either (although it is thought that traces of an ”accusative“ case are still retained by the known personal pronouns). In Kartvelian, the direct object of a transitive verb is expressed by what we usually call the ”dative“. The Kartvelian dative is thus, as one linguist put it, a ”functionally heavily burdened“ case[v]. Thus, in Mingrelian :

muma arʒen-s cxen-s skua-s, the father is giving a horse to his child, where both the direct object, and the indirect object, cxen (horse) and skua (child), are in the dative. Or, in Svan : eǯa xo:te bo:pš-s diär-s – he (eǯa) cuts bread for the child. Or, to give an example in Svan, in the plural:

lamp’räl-s at’wra:lix, they light the lamps, where the termination of the dative is simply added to the plural: lamp’räl-s.

The Etruscan dative in the plural is formed in exactly the same way: by agglutinating the same termination -s, as in the singular, after the mark of the plural: cliniiar-a-s, to the sons.


Once the rule is understood, the morphology and syntax of Etruscan start to become comprehensible. The Kartvelian extensive use of the dative in -s allows us to understand the similar use and the ordering of the famous inscription on the Florence statue known as the ”Arringatore“ (on which the signs for ś, x, and s, s, are inverted, as was the custom of the Northern Etruscan cities):

auleśi meteliś ve(lus) vesial clenśi

cen flereś tece sanśl tenine tuθines χisvulicś

To Aule Metele, son of Vel and Vesi,

this statue (fler) was dedicated/put- sanśl tenine- by the city (tuθ).


The meaning of the above translation has been well established. However, it has never been understood why fler, statue, offering, sacrifice, should be in the dative case, flereś, since fler is the direct object of the active verb tece (put/dedicated). But, in the Kartvelian grammatical logic, fler/statue has to be in the dative. (cf. Georg. cers c‘ign-s, he writes a book, c‘ign-). Later on, we will also imply that Kartvelian grammatical logic explains also why the demonstrative article eca gives cen in the dative, so that the group eca fler becomes in the dative cen flereś (see infra: Svan nom. eǯa -dat. ečən).

Another peculiar utilisation of the dative in Kartvelian is in the designation of time and duration. Thus in Old Georgian, in the dative singular:

dγe-sa šabat-sa, on the day of Sabbath

In Mingrelian, with day and night in the dative singular:

sum dγa-s do sum ser-s , three days and three nights

The same occurs in Laz:

mažura ndγa-s, on the second day

In the same way, in Etruscan, on funeral inscriptions, the word designating the year, avil, is always in the dative singular, with –s: XX avil-s lupu, ”he died at the age of XX years“. The word avil (year), appearing on thousands of funerary inscriptions, was one of the first to be deciphered with certitude; however, the fact that it should take the termination –s of the dative has always puzzled researchers. In Kartvelian, the dative case is the one used for designations of time and duration. The identity of the termination of the dative, –s, combined with its two special uses, for the designation of the direct object of an active verb, and its temporal use, should exclude any explanation by chance.


Such a comparison of the two grammatical systems yields thus a whole series of similarities. Starting from both systems, I should say that from now on I will argue that by using the mechanism of the Kartvelian languages one can understand the formation of a certain number of vocables in Etruscan, and vice versa. By using the Kartvelian key, we discover that Etruscan possessed another case, undetected until now, which we can name, following the Kartvelian tradition, ”adverbial“. The Kartvelian adverbial case, through the adjunction to the root of –d, or –t , indicates a function, or a particular quality. The termination is –d, or –ad, in Georgian and Svan, –t in Mingrelian. Thus in OG:

mamak’ac-ad da dedak’ac-ad kmna igini, he created them man and woman.

man daswa igi mepe-d – he installed him as king.

Let us now take the extremely frequent Etruscan term zilaθ, or zilat, which has been known to designate the consul, the Roman duumvir, and the well identified numeral zal, ”two“, (z representing, as has already been proposed, the affricate [ts], corresponding to the Kartvelian c and c‘). The meaning of both terms, zal and zilaθ, has long been known, but the relation between the two was never established.

By applying the mechanism of the Kartvelian adverbial case in –d or –t, we see that Etruscan could form, from the numeral zal (2) the very frequent term zilaθ, designating each of the members of the dual, collegial, annual leadership of a given city, or of a function. We know that the Etruscans introduced the institution of duumviri at Rome. It was the second Tarquinius who created the office of the duumviri sacris faciundis [vi].

The Etruscan zilaθ, a very frequent terms in Etruscan epigraphy, seems thus obviously derived from the root of the numeral zal, two, the Latin duumvir being just a translation, an adaptation of zilaθ. The mechanism through which zal furnishes zilaθ can be explained only through Kartvelian. One can also bring a supplementary argument proving that the termination –θ is not organic to the root zal-: the fact that the name of the same function of duumvir was also rendered by zilaχ[vii], with the –χ termination of agents and ethnonyms, like in Rumaχ (Roman), and Velznaχ (a native of Volsinii.

On the other hand, the common root *zal, *zel, *zil is very well attested in Kartvelian (cal-, cel-, cil-). In verbs, it indicates the splitting in two: Sv. li-cel, to be broken in two (li– is the prefix of all Svan infinitives), transitive: li-cle, to break in two; Ge. mo-cil-eba, separation from someone; cal-k’uli, separated. The Kartvelian root cal– also designates the member of a couple of objects, or simply the half: cal-tvala, one-eyed; cal-k’uza, with one hump, dromedary; cal-rkiani, unicorn; cal-pexa, one-legged; cal-xela, with one arm. Starting from here: mo-cile = enemy, the one who is in front of me. Through the adverbial suffix –ad, *cal– produced also in Georgian calad, alone. Precisely, the formation of calad is exactly parallel to the one of the Etruscan zilat, which from this perspective appears clearly as the one whom the Romans were naming a “duumvir”, a member of a pair.

We have an exact parallel of this Etruscan-Kartvelian use of one single root in the sense of unity, duality and half in the way in which the Indo-European root *sem– produced on the one hand the Greek numerals ἕν and εἷς, one, and the Latin semel, once, and on the other hand the Greek ἥμι and the Latin semi, half.

Typological cross-verifications

If we accept that such a similarity of structural traits as we show here cannot be simply attributed to mere chance, we must also deduct that by using grammatical patterns from one of the two compared systems we should be able to predict similar patterns in the other. Such would be the obligatory use of the preposed personal pronoun in the conjugation. In the Svan conjugation, as in all other Kartvelian languages, the category of person and the direct, as well as the indirect, object of a transitive verb are preposed and form an integral part of the verb: mirdi, he raises me; ǯirdi (ǯ is is Svan the result of the palatalisation of an archaic g-), he raises you. M- and ǯin Svan (m- and g- in the other K. languages) representing the 1st and 2nd personal pronouns in the singular, and the forms being identical for direct and indirect objects, mirdi and ǯirdi can also translate: he raises for me and he raises for you.

Now, let us take two Etruscan dedicatory inscriptions :

— mi titasi cver menaχe (TLE 282, on a mirror, translated tentatively ”Tita gave me as a gift“)

— mi qutun lemauśnaś ranazu zinake (TLE 28, ”I am the kōthōn of Lemauśna; Ranazu made/offered (me?)“, this being the traditional interpretation)

It has always be thought that menake and zinake (elsewhere menaχe and zinaχe; the various spellings of the guttural is due to the difference in regional conventions of spelling) represent different verbs, although, given the profusion of other verb forms whose meaning has been tentatively rendered as ”to give“, or ”to offer“ one would be justified in trying to find an alternative explanation.

In our inscriptions, following the Kartvelian pattern, menaχe and zinake simply seem to reproduce the m– and ǯ– of a Kartvelian, m-ena-χe and z-ena-χe (-χe is the Etruscan termination of the past). zinake would thus mean ”ranazu made/gave (me) for you“ (ǯ-inake), while menake is the same verb with the preposed 1st person instead of the 2nd (m-enake). We would thus have a confirmation of the fact that the alphabet of the Etruscans was very ill adapted to the phonetics of their language and that z was reproducing the whole series of the palatals, as well as of the affricates. This is immediately obvious in some graphical choices, like the spelling of the numeral huθzar, where the combination of hut (probably 5) and śar (10) produced [č], spelled θz.

It thus follows that z reproduces indiscriminately the two series of affricates (c, c‘, Ʒ) and palato-alveolars (č, č‘, ǯ) of the Kartvelian languages. We have in zal (two) the voiceless affricate of Kartv. cal-, cel-, cil-, but in other Etruscan words z corresponds to the K. ejective (glottalized) affricate c‘, as in the root for ”book“ and ”to write“: ziχ (cf. Georg. c‘ig-ni = book identical with ”book“ in the epitaph of Lars Pulena: ancn ziχ neθśrac, correctly interpreted traditionally as ”this holy book“ (this-book-holy), neθśrac itself, ”holy“, being obviously a composed word that can be compared with the old Georgian net‘ar, as in the Biblical expression ”net’-ar-i Saaba“ = ”blessed Sabah“.

By using the Kartvelian key, we can even tentatively decrypt the etymology of some Etruscan terms. Thus, in the religious text of the Liber Linteus, written on a mummy wrapping preserved in a Zagreb museum, we have a succession of administrative entities in favour of which prayers are to be addressed to the gods. The formulae are clearly similar to those in the Umbrian language from the Tabulae Iguvinae. The succession of administrative entities in the Liber Linteus is: ”sacnicleri spureri meθlumeri“. The meanings of spura (= city, cf. the Etruscan neighbourhood of Suburra in Rome) and meθlum ( = territory) have long been identified. The logic of the succession of the invocations shows that sacnicla has to be something inferior, in descending order, from the territory to the city, and down the administrative scale. These sacnicleri that have to be blessed, coming down the ladder from territory to city, and lower, can only be the fields, the arable fields.

The Kartvelian grammatical mechanism would offer an immediate explanation for the formation of the word : qana or qona is the plough-able field in Svan and the sister K. languages. By using the very frequent K. sa– prefix denoting collectives, or entities, we obtain sa-qna (in Svan, the root vowel also usually disappears in composition, as in u-qna : unploughed), sa-qna designating the totality of the fields; this would be as logical as to see that the country of the Georgians is actually called in their own language sa-kartvel-o, or that house is sa-xl-i, from an unknown root *xel-.

The same procedure of creating abstracts or collectives from a syncopated root by the addition of the prefix sa– would offer, when reversed, a satisfactory etymology to the very Kartvelian name of the house: saxli, from the unattested root xel- : hel-, pl. hilar, is the name of the ”land“ in Etruscan, identified in the expression helu tesne rasne : terrae jus Etruriae, according to the Latin translation. Hilar was also used to designate the limes, the border of a territory, as we see on countless stone stelae. From hel, we get sa-hl-i (-i is just the termination of the nominative in Georgian), in the same way in which from K. qona we would get Etr. sa-qna.

(Incidentally, another Etr. word that we know to mean territory: ceχana –which we find, to quote just one instance, in the ”inscription of Velθur Partunus“: zilχ ceχaneri = ruler of the land– is identical with the old Georgian kweq’ana, as in: kmnna γmertman cay da kweq’anay / God mad the heaven and the earth.

The pronominal flexion

The most striking similarities centre on the known personal pronouns. The Etruscan personal pronouns for the 1st and 3rd persons singular are identical with their Kartvelian correspondents: mi, and eca (eca, it has been shown, is actually a demonstrative, which, as in many languages, was used as personal pronoun: he/she/it). Mi is present on hundreds of Etruscan objects in inscriptions of the kind: ”I belong to…“. Eca has been identified thanks to scenes like the one painted on a crate from Vulci, where we see Alceste, who descended into Hell to sacrifice herself in order to free her husband: eca ersce nac aχrum – flerθrce: “She (i.e. eca) went (ers-ce) to (nac) the Acheron (aχrum). She sacrificed herself”. The Etruscan pronouns mi and eca correspond exactly to the Svan mi and eǯa, as well as to the Georgian me and ege/igi. (Historically, the Svan ǯ derives in certain conditions from g; [ega] was certainly the actual pronunciation of the Etruscan eca, the Etruscan alphabet making no distinction between voiced and voiceless consonants.)

But the Etruscan and Svan pronouns are also identical in their flexion. Thus, in Svan, the dative of eǯa (Etruscan eca) is ečən, while in Etruscan eca becomes cen. All Etruscan demonstrative pronouns form the dative in the same way, e.g. ita > itun, etc. (again, this case was wrongly considered until now as an ”accusative“, a grammatical notion absent both in Etruscan and in Kartvelian). We have thus two identical personal pronouns out of three in the singular.

Coming back to the Liber Linteus, we also find what could be the possessive form of the Etruscan first person plural pronoun: nienaś = our, present more than a dozen times at the end of the enumeration of those entities for which prayers were offered: śacnicleri spureri meθlumeri enaś / ”(for) our land, city and country“. Enaś seems the possessive, or the genitive form, of a first person plural pronoun ni, in the same way in which in Svan niš-ge is the possessive of , we, (-ge is a mere suffix appended to all Svan possessive pronouns). Correspondences between personal pronouns are a sure sign of kinship between languages. As is well established in linguistics, personal pronouns are among the most resistant elements of any given language and they are never subject to borrowings.

In the plural, we can thus acquire the certitude of identifying the Etruscan pronoun for the first person, we: ni, which in Svan is na, , näy, according to the dialect. Ni adorns a small series of votive objects found across Etruria and it was considered until now to be an error of the engravers for the first person mi. This is highly unlikely, especially in such a simple and usual word. Pallotino‘s Testimonia Linguae Etruscae[viii] lists five instances of ni, and no other possible confusion between an m and an n is known. The survival of votive objects carrying such botched simple inscriptions is also unlikely, for obvious reasons of economic and social prestige: nobody would have offered or put in a tomb such valueless artefacts. Objects with inscriptions like ni larisa larecenas must have been part of a set, a series, each one of them announcing: ”We belong to Lar Larce“…

To continue with the typological cross-examination, if we know with a total certitude the pronouns mi (I) and eca (he/she/it) and have found a plausible ni (we), together with the possessive enaś, we absolutely have to identify a sin/sen or śin/śen for the missing 2nd person singular, because this is the form that the 2nd person singular pronoun has in the K. languages.

We actually find, in the columns V and IX of the Liber Linteus: śin eiser śic śeuc, where śin is clearly in apposition with the same formula used elsewhere in the text: eiser śic śeuc, without śin. In the same way, we find śin eiser faśeiś, or śin vinum flere neθunsl. These look like imperative sentences where śin is used alternately in an apposition. The second formula –vinum flere neθunsl– is very clear and it was always translated as: ”sacrifice wine (as in a libation) to Neptune“. śin is clearly optional and it could only be the personal pronoun singular: thou, identical with the one in the K. languages and perfectly completing the system.

Other lexical correspondences / Etruscan – Svan core cognates

At the lexical level, Etruscan and Kartvelian share a number of important isoglosses (we leave aside the personal pronouns already mentioned, and those words, like zilaθ whose formation was explained above). At the same time, in what concerns the lexical comparison, we have to keep in mind that the phonetics of Svan has been profoundly dislocated :

Father – Etr. apa, phon. [aba] – Kartv. baba

Mother Etr. ati, phon. [adi] – Kartv. di

In both instances, the voiced quality of the Etruscan stops is hidden by the writing, which made no distinction between voiced and voiceless. One could object that K. baba might as well be a loanword from Turkish, but it is present in all four K. languages, and Klimov[ix] accepts it as native, based on the fact that in Georgian, under the form babua, it designates the uncle. The root for mother, di-, still found in pure form in Svan, is reduplicated in the other languages: dida in Mg.-Laz, deda in Georgian.

Daughter, Son – Etr. seχ (daughter) – Mg. skua, Laz ski-ri, Sv. sge, ske (son)

The divergence in meaning is secundary: Klimov[x] has shown that both Kartvelian terms come from the root *-šw-, to beget, to give birth.

To write, book

ziχ-, to write, book, has been shown supra.

Moon, Etr. tivr :

In Old Georgian this is attested as m-tovare, m-tvare, where m– is just a phonetical excrescence without any etymological value (cf. G. mgel-i, wolf, Min. ger), the oldest attested form being thus tovare. (Matth. 24, 29: m-tovare-man ara gamosces nateli tvisi, the moon will not give its light). In Georgian and the other related languages this lexeme also exists in shorter forms, in G. twite, ttue, tue, in ming. tuta, Laz (m)tuta. Svan is here alone in having an aberrant form, došdul, došt’ul, about which all that can be said is that it is a diminutive (suffix –ul).

Sun, Etr. usil (phon. [uzì:l]):

The Kartvelian forms are: G. m-ze, Inguri dialect zej, Ming. b-ža, Laz – m-žu-a, m-žo-ra, svan miž/məž. This is an example of the well-known use in Etruscan of the same letters to represent the voiceless consonants, as well as the voiced and ejective phonemes. The letter s (s) in usil was obviously rendering the voiced spirant [z] (the Etruscan letter z (z) being used, as we have seen, exclusively for the affricates corresponding to the Kartvelian c and c‘). Usil appears to be a diminutive formed with the suffix –il. The root was *-z-, which we find in the kartvelian ze-, b-ža, m-žu-a, m-iž (the correspondence between a Georgian z and a Svan-Mg-Laz ž is a well attested fact in the history of Kartvelian languages).

Usil, the Etruscan name of the sun, is obviously a secondary formation, proceeding either from the adjunction of the diminutival suffix, –il, thus us-il, or with the help of the homophone suffix –il found in Georgian in words like dum-il-i, silence, or by a circumfix u-il.          One would thus obtain u-s-il, ”the shiny“, from a root *-s- ([-z-]), to shine, in the same way in which in Georgian, from the root –dg-, to stand, we have a-dg-il-i, place. (Klimov[xi] raises a similar hypothesis as to the form of the Kartvelian root meaning ”sun“, ”to shine“, which he tentatively reconstructs as *mze-, allowing that m– could also be just an excrescence, or the first part of a circular affix me applied to a root *-z– meaning ”to shine“). All things considered, everything points in the direction of usil being a diminutive. The reason for which i in usil kept its phonetic value, instead of closing into ə or a, as we would expect in Etruscan, which usually had the accent on the first syllable, is furnished by the Svan phonology. In this language, diminutives in –i:l move the accent of the word on the suffix itself, which has a long vowel, cf. Sv. dedber-ì:l (old woman, from the Georgian dedaberi). It appears thus that usil should be read [uzì:l].

Young, new, husrna:

This is a case in which we not only discover the exact correspondent of an important Etruscan adjective in Svan, but we also uncover another proof that the Etruscan writing system simply did not represent voiced consonants, although these were perfectly present and functional in the phonetic structure of the language. In this case, the Etruscan h (h) corresponds to a Svan [γ] (the modern greek γ, or to the Spanish jota), which shows that one and the same graphic sign, h (h), was representing both the voiceless spirant [x], and its voiced correspondent [γ], while [ž] was rendered simply by s (s, x), sometimes z (z), or other combinations. We see this on two mirrors engraved with figures of mythological characters representing young Mars. On both, the adolescent Mars is called Mariś husrnana, interpreted as “Mars juvenilis” already by Trombetti[xii]. The epithet is husrna with a reduplicating suffix: husr-na-na. In Svan, γwžur or na-γwžur, is the male adolescent. A brave man, the man in the sense of vir is called γwažmāre, pl. γwažär (māre is simply « man » in Svan, γwaž adding a sense of young force). A brave adolescent is called: maxe-γwaž, from maxe = new, young. The Etruscan husr-na is thus in a perfect symmetry with the Svan na-γwžur. Reciprocally, like in a mirror, the Svan. γwažmāre corresponds to the Etruscan mariś-husrnana. On other inscriptions, we have the term husiur clearly representing “sons”, male children. On those inscriptions, [ž] was obviously rendered by the group –si-, γwžur = husiur, instead of a simple s. The word husrna designating the adolescent, it follows thus that huslna vinum offered for libation in the religious ritual reproduced in the Liber Linteus can only mean “new wine”, with γwžul- for γwžur, which shows simply a dialectal particularity. Having to represent the sounds [γ] et [ž], the Etruscans have employed the letters at their disposal: h (H) and s (S). Mariś husrnana is thus to be read as Maris γwžrnana.

Herm, γerm – the notion of the «sacred»

The notion of the sacred can be identified principally through the quasi-bilingual golden tablets of Pyrgi, where, in the longest Etruscan inscription, heramasva, at the end of the 1st line and beginning of the 2nd, corresponds exactly to the Punic ‘ŠR QDŠ, the ”sacred place“ (which brings the Pyrgi tablets closer to the status of a bilingual than was thought until now). The Punic LRBT L’ŠTRT ‘ŠR QDŠ corresponds thus exactly to heramasva vatieχe unialastre „this sacred place of the divine Astarte“, Unial-Astre (Uni in Etruscan, the Latin Juno being of Etruscan origin) being in the dative: unialastres. Whatever the signification of vatieχe, heram-asva is equivalent to ‘ŠR QDŠ, «the sacred place».

In identifying herm– as the root designating the «sacred», we find a confirmation in the epitaph of the Tarquinia priest Lar Pulena, where herm– figures no less than 5 times, in the following expressions: caθas hermeri; caθas paχanac alumnaθe hermu; alumnaθ pul hermu huzrnatre (where huzr– renders again the Svan γwažär, with h corresponding to γ in two successive words, hermu and huzrnatre , in the same inscription); again hermu; and in the name of the month of h/e/rm/e/rier. It is obvious that such a profusion of the root herm– in a short text describing the cursus honorum of a sacerdote can only be related to God and the divine. The total absence of the root herm– from the texts containing precise names of gods, like the Piacenza bronze liver, is easily explained by the adjectival nature of the term: herm– wasn’t the name of a particular god, but represented the ”godhead“, a notion rendered in Kartvelian by γerm-/γertem.

The simple fact that the supreme god of the Etruscans, who, according to Latin testimonies, was named Vertumnus, or Voltumna, does not appear in any inscription should also have raised some questions. What we find in the inscriptions is only herm-. The explanation for this strange absence is that the word was spelled differently in Etruscan and in Latin and that the Latin V at the initial of some words borrowed from Etruscan hides in fact a very different phonetic reality. The structure of the theonym Vertumnus/Voltumna is clear: vertum-, or voltum-, is followed by the adjectival termination –na (cf. Etr. śuti-na, for the tomb, funerary, or husr-na, young). What was not clear until now was the nature of the initial V, in order to explain why Vertumnus was never found in any Etruscan inscription. The explanation is that the Latin initial V was trying to reproduce the voiced spirant [γ]. In the Kartvelian languages, the word for God is built on the root γertm -, or γmert-, one being the metathesis of the other. We have thus γmert(i) in Georgian, γermet in the Lower Bal dialect of Svan (γerbet in Upper Bal, with the ergative γertem), γoront(i) in Mingrelian. Since the phoneme γ did not exist in Latin, it was rendered by the Romans, in their written texts, by v (= w). Vertum-/Voltum– (-nus/-na being simply adjectival extensions) reproduces γertem/ γermet. In the native Etruscan writing, γ– was rendered by h-, hence the frequency of the root herm– in the Etruscan texts in which gods are mentioned. (Etr. herm– has traditionally been interpreted as referring to the Greek god Hermes. This is a totally gratuitous interpretations, passed over from a generation of scholars to another. There is no scriptural evidence of a Hermes in all the Etruscan iconography, and the god corresponding to the Greek Hermes is always designated by the national name: Turms.)

Once we understand this, we also realise that the Pyrgi tablets are closer to a bilingual than admitted hitherto.

Musician (particularly piper), subulo

Subulo is glossed as tibicen, flute player, but the Etruscan tibicines were also dancers. Most surviving representations show us the flute players in the act of dancing while playing. (Varro says about subulo : “ita dicunt tibicines Tusci”. From his unreliable authority, for two millennia, it was assumed that subulo was a piper… But when we look at the tomb painting and all other representations of flute players, we always see them in the act of dancing. ) Thus, the Svan sub– to dance, subi, he dances, can only remind us of the Etruscan/Latin subulo. We have in Svan : subi, he dances, süb-da, he has danced, an-sub-ni, he shall dance. The derivation from the root sub– by means of the prefix –ul, –il, poses no problem to arrive at sub-ul, the –o being simply the result of the latinisation of the word, like in lucum-o, the Latin name of the Etruscan kings. By extension, subulo took also in Latin the meaning of «debauched », which shows that the term was designating initially more than a precise function and that it was rather a generic term covering any participant in the Etruscan games that the Romans were considering so obscene.

Actor, istrio – (h)istrio

The Latins knew already that the initial h- was not organic, but simply graphic: the root –ster– means to play in Laz: i-ster-s : he/she plays.

The missing ergative case

When one reviews all the similitudes that we reveal here: the known pronouns, identity of the casual system, perfect superposition of the use of the various cases, down to the most improbable idiosyncrasies and lexical similitudes, one is forced to admit that we are confronted with two identical morpho-syntactic systems, a concordance which excludes the intervention of chance. Still, one major objection can arise: the apparent absence of the ergative (or narrative, as it is sometimes called) case in Etruscan. Languages built on the model of the Kartvelian can only be of the ergative type. (The ergative is the case of the subject of transitive verbs and sometimes of intransitive actives that stand in a form of the aorist).

At first sight, there seems to be no place in the Etruscan casual system for a possible ”ergative“ termination, but this is because until now linguists have not thought of Etruscan as of a functioning system, but as of an accumulation of isolated characteristics. When one tries to identify the missing ergative termination, one realises that, given the nature of the surviving texts, this case could only be appended to proper names (such and such has given, offered, made, etc.). We then see that a great number of names of donors, offerers and makers that appear in the inscriptions bear the termination –s, a circumstance which was never explained. In fact, the Svan uses (again!) the dative in -s to mark the actant subject of a verb in the past where in other languages we would expect a mark of the ergative.

Thus :

eǯi-s p‘lat‘uk xočo:na – he has rolled (it) up into a cloth,

where eǯi-s is the dative of eǯa – he, she, in the function of the subject of a transitive verb in the past.

bepšw xa:mne lezweb-s mäydär-s – the child gave bread to the hungry ones, where child -the actant subject- is in the dative where we would expect an ergative.

Kartvelian offers thus the only possible key for explaining the triple use of the termination –s in dedications such as itun turuce venel atelinas tinas cliniiaras, (”This was given by Venel Atelina to the sons of Tina/Jupiter, i.e. the Dioscouroi) where we have atelina-s (dat.), itun (dat.) tina-s (gen.) and cliniiara-s (dat.) exactly like we would have in Svan eǯi-s (xa:mne) lezweb-s mäydär-s – he gave bread to the hungry ones.

Etr. — atelina-s (turuce) itun tina-s cliniiara-s – Atelina gave to the sons of God

Sv. — eǯi-s (xa:mne) lezweb-s mäydär-s – he gave bread to the hungry ones.

Trubetzkoy[xiii] has also shown that in some languages it is the form of the genitive that coincides with that of the ergative. We know that in Old Georgian, the function of the genitive was more extended than today, and it was often found before other endings, like those of the adverbial: Parnavaz-is-ad, compared to the regular Parnavaz-ad. The ergative is also the only case that presents no common, or coherent terminations across the spectrum of the Kartvelian languages.

The most likely conclusion, in front of this schematically presented evidence, is that Etruscan and Kartvelian are genetically related. By comparison, a language possessing a first person pronoun with the form *eghwom and an accusative case with the termination –m, which termination, besides being applied to the direct object (libru-m) would also serve for expressing various adverbial notions, like direction (eo Roma-m) and time (nocte-m) would unhesitatingly –and rightly so- be accepted as Indo-European, even if most of its vocabulary would remain inaccessible and incomprehensible. We wouldn’t be dealing with random lexical coincidences, but with the repetition of one and the same complex and delicate morpho-syntactical system. A similar assumption should be made about the relationship between Etruscan and Kartvelian. Here also, we don’t only have a perfect concordance of the very forms of all the attested (in the case of Etruscan) casual terminations, but also an identity of their usage, which is so unusual and complex as to exclude any explanation by coincidence.

The identical first person pronoun mi, the complex functions of the overburdened dative case in –s, combined with the similarity of the termination of the dative case with the termination of the genitive case show again that coincidence should be totally excluded. To the similarities of the grammatical systems we can add a series of lexical correspondences –usil (sun), tivr (moon), husar (young), herma (sacred)-, and also the fact that the etymology of some important words (zal – zilaθ, two – duumvir) becomes clear only from the comparison of the two linguistic systems.

Klimov[xiv] has already suggested that Svan separated from the other Kartvelian languages at the beginning of the second millennium a. Ch. Etruscan is attested in writing in central Italy starting from the 7th century a. Ch. The lack of a marked dialectal dispersion in Etruria speaks, as Helmut Rix has pointed out[xv], ”for a relatively late spread of the language from a limited area“. An archaic inscription written in a language that is clearly what could be called an ”Etruscan dialect“ has been discovered at the end of the 19th century on the Greek island of Lemnos (cf. inter all. Bonfante[xvi]). The path is cleared.



Cf. also:

Sucking the victim‘s mother‘s teats – the Etruscans and the Caucasian vendetta…

[1] For the Kartvelian grammatical examples the main sources are The Indigenous Languages of the Caucasus, A. C. Harris, Ed. (Caravan Books, Delmar, New York, 1991), vol. I, Kartvelian Languages, but also G. Dumézil, Contes lazes, (Institut d’ethnologie, Paris, 1937) and G. Dumézil, Récits lazes en dialecte d’Arhavi : parler de Şenköy (Presses univ. de France, Paris, 1967), as well as my own knowledge of the languages. The Etruscan examples are taken strictly from already published collections such as CIE (Corpus Inscriptionum Etruscarum), TLE (Testimonia Linguae Etruscae), or from quoted authors. The transcription for the Kartvelian languages is the one used in the ILC. For Etruscan, it is the traditional translation, using the Latin alphabet with the addition of a few Greek letters.

[i]           J. Nichols, Linguistic Diversity in Space and Time (University of Chicago Press, 1999), p. 1.

[ii]           Pauli, C. E. Altitalische Forschungen. (J.A. Barth, 1885).

[iii]          V. Thomsen, Remarques sur la parenté de la langue étrusque, in Samlede Ahfandlinger Vol. II. (København, 1919)

[iv]          Cf. G. Bonfante, L. Bonfante, The Etruscan language: an introduction. (Manchester University Press; New York, 2nd ed., 2002).

[v]           H. Fähnrich, Old Georgian, in The Indigenous Languages of the Caucasus, A. C. Harris, Ed. (Caravan Books, Delmar, New York, 1991), vol. I, Kartvelian Languages, p. 188.

[vi]          Dumézil, G. La religion romaine archaïque : avec un appendice sur la religion des Étrusques (Payot, 1987), pp. 593-594.

[vii]         D. Briquel, Les Etrusques : peuple de la différence. (A. Colin, Paris, 1993), p 106.

[viii]         M. Pallottino, Testimonia linguae etruscae. (La nuova Italia, Firenze, 1968)

[ix]          Г. А. Климов, Этимологический словарь картвельских языков. (Москва, 1964), p. 47.

[x]           Г. А. Климов, op. cit. pp 139-140.

[xi]          Г. А. Климов, op. cit. p. 134.

[xii]         Trombetti, A. La lingua etrusca. Grammatica, testi con commento, saggi di traduzione interlineare, lessico. (1928), p. 180.

[xiii]         N. S. Trubetzkoy, Studies in General Linguistics and Language Structure. (Duke University Press, London, 2001), p. 80.

[xiv]         G. A. Klimov, Die kaukasischen Sprachen. (Buske, Hamburg, 1969), pp. 45-46.

[xv]         H. Rix, Etruscan, in The Ancient Languages of Europe, R. Woodard, Ed. (Cambridge University Press, 2008), p. 142.

[xvi]         G. Bonfante, L. Bonfante, The Etruscan language: an introduction. (Manchester University Press; New York, 2nd ed., 2002).