* * *

        The question of the origin of the Georgian language, of its affinity with other languages of the world—a question which later became the central problem around which he built up his subsequent research work—began to interest Nicholas Marr while still a pupil of the Kutais Gymnasium. It was during his university years, while studying Arabic under Professor V. P. Rosen, the Russian Orientalist, that Marr first observed a series of resemblances between Georgian and the Semitic languages. Further study convinced him that here, at last, was the right line of investigation. For two years he worked upon this subject and then published, in 1888, in the Georgian journal Iveria (No. 86), an article on " The Nature and Peculiarities of the Georgian Language.”
        Restrictions surrounding an academic life, however, hindered Marr from any immediate publication of progress made in this study, so that it was not until 1908 that The Fundamental Tables of the Grammar of the Ancient Georgian Language appeared, in which Marr claims to reveal the entire nature of the morphology of the Georgian speech, and particularly of the structure of Georgian verbs and their conjugation.
        But it was in his Introduction to this work, under the title “ Preliminary Information on the Affinity of the Georgian Language
[ 105]  
with the Semitic Languages ", and covering ten pages, that Marr’s so-called "Japhetic Theory" was propounded. In this article he enumerated concisely the characteristic traits common to both Georgian and the Semitic languages, such as the formation of words with the aid of prefixes; the peculiar distribution of vowels as elements for the formation of etymological categories, and of consonants as stems ; the triliteralness of stems; the affinity of the sounds themselves, etc. Yet, Marr added, " there is in Georgian much, and very much, which removes it from the Semitic. Georgian, therefore, cannot be said to be a Semitic language, it is only related to Semitic in a certain degree."
        Having established the relationship of Georgian and its affiliated languages with Semitic, Marr's next step was to find a name for his newly defined linguistic branch. As the Hamitic languages were then also supposed to be related to the Semitic, he decided to appropriate the name of the third remaining brother, Japhet, hence Japhetic became synonymous for Georgian and its related Caucasian languages. "For," said Marr, "the term Japhetic is not binding in any way. It is merely a conventional one which may be altered at will to cover any or all meanings according to results achieved in the process of research. Its meaning might even be completely altered.”
        In defining his Japhetic languages, Marr used the term "branch" and not “family”. In his opinion, neither Semitic nor Japhetic formed a linguistic family. Each was only an individualized branch, and only together did they, the Semitic branch and the Japhetic, form one linguistic family. For this parental family, Marr appropriated the name of Noah (father of Shem, Ham, and Japhet) and called it Noachian —a conventional term, also.
        Later on, though, Marr somewhat altered the degree of relationship which Georgian bore to Semitic. “The 'brotherly’ relationship as at first outlined between the Japhetic, Semitic, and Hamitic languages,” he said, "which made it permissible to call their common parental language Noachian, was found ... to be a doubtful one. It is now clear that Hamitic languages are further removed from the Semitic than these latter from the Japhetic.”
        Still further progress in the study of Japhetic languages demonstrated how inappropriate was the term "brotherly” as applied to the degree of relationship existing between even the Semitic and the Japhetic languages. But Marr, in spite of the fact that some of the claims proved to be groundless, always stoutly maintained his opinion of the relationship of these linguistic branches.
        In course of time, the sense conveyed by the term Japhetic was itself altered. The term in no way defined the biogenetic nature of a language, Marr said. A new materialistic approach to the problems of linguistics and the analysis of language as a specific ideological
[ 106] 
social superstructure (which he calls "a new teaching of language") led Marr to introduce into the classification of languages the term System, in place of “family", “tribe" (ethnos), “branch" (la souche, der Sprachstamm), etc. However, the name Japhetic was retained because of its analogy with the names Semitic and Hamitic.
        The first stage of development, the birth, as it were, of the Japhetic theory, was almost entirely taken up by the work of research into the nature of the inter-relationship of the Japhetic with the Semitic languages. Marr declared that the correct conception of the task of Japhetic linguistics at this stage was hindered, among other things, by the fact that, according to the then prevailing scientific interpretation, the circle of Caucasian languages, bound by ties of mutual relationship, consisted of only four living Trans-Caucasian languages, viz., Georgian, Megrelian, Chanian or Lazian, and Svanian. Even Abkhasian, it was thought at that time, did not belong to this circle— " Small as this circle of related languages was, its horizon was further shrunk by the predominance of the norms of only one historically cultural Georgian language."
        The second stage in the development of Japhetic linguistics began with the concentration of research within the circle of the Japhetic languages themselves. After a prolonged investigation into the ancient, medieval, and modern literary Georgian and Armenian languages, Marr set himself the task of determining the exact nature of the inter-relationships within the Caucasian linguistic group itself. With this aim in view, Marr undertook an intensive research into other, particularly the unwritten, languages of Caucasia, which resulted in a complete revaluation of the importance of the living and dead languages. While working in this direction, his analysis of the Megrelian (Iberian) and Chanian (Lazian) languages showed that these were fully developed languages and not mere dialects of the Georgian, as they had been previously considered. He established that these were independent languages of the same Caucasian branch, and that each of them showed characteristic peculiarities which made it possible to classify them as follows :—

         I. Sibilant languages :
        (a) Hissing (svistyashchie—dental-sibilant) characterized by s sound. (Georgian.)
        (b) “Hushing" (shipyashchie—palatal-sibilant) characterized by sh sound. (Megrelian and Chanian.)

        II. Spirant languages, characterized by h, k, etc., sounds with a series of their compound dental and affricate sounds.
        (a) Mixed with the hissing lang. (I, a), (Abkhasian).
        (b) Mixed with the “Hushing” lang. (I, b), (Svanian).

[107]
        III. Sonorous languages.

         "Extraordinarily complex and fine, but transparent correlations of the unwritten languages, correlations detected only by ear, were instrumental in elucidating the hybridized state of many Japhetic languages,” Marr wrote. "Theoretical interest in hybridized types was consequently intensified. In the first place, languages purely Japhetic were examined, such as Svanian, which proved to be a mixture of the language of palatal-sibilant or sh group, and of the language of the spirant branch (II, b).”
        As time went on, the process of hybridization was gradually detected in all the Japhetic languages. The so-called Armenian language proved of exceptional interest in this respect. It was found to consist properly of two languages: one, the ancient literary, and the other, the ancient unwritten, but still extant, Armenian language. Both languages were hybridized types of exceptional value. Marr demonstrated, in a whole series of works, that the Armenian people, their culture, and particularly their language, were most clearly bound up with the variegated, but essentially Caucasian milieu. The Armenian languages, it was concluded, were half Indo-European, half-Caucasian or Japhetic, and Marr explained why this was so in a series of articles under the common title : Yafeticheskie Elementy v Yazykakh Armenii, " Japhetic elements in the languages of Armenia.”
        Here, while on the subject of Marr's philological works which were devoted to the explanation of Armeno-Georgian, Georgio-Persian, Armenian-Syrian, Armeno-Greek, and Georgio-Greek, Armeno-Arabic and Georgio-Arabic literary interrelations, we must not forget his great magisterial dissertation, Sbornik Pritch Vardana, “ A collection of the Parables of Vardan,” which appeared in three volumes during the years 1894-1899 (Vardan was an Armenian moralist preacher of the twelfth-thirteenth century). V. B. Aptekar holds that this occupies first place among Marr’s works. In it the author explains precisely, side by side, the lexical influences and the mutual contributions of all cultural languages belonging to the medieval Hither-Asian world.
        “The problem of Armenology,” said Marr, was “to investigate the conditions under which the Armenian nationality was formed, and to define in historical sequence its independent, original
[ 108]  
manifestations, however insignificant that may seem, on that common background of the political and cultural life of Armenia which was created as a result of strong external influences.” To attain this, in his opinion, scientific prejudices must be renounced. "Armenology cannot found itself on only one philology in the narrowest sense of the word. Archaeology with its discipline and ethnology with its folklore should interest the armenologist as much as linguistics and literature.” Armenology being an independent science, would not establish itself on a firm basis until the student “equipped himself with real knowledge of local languages and real understanding of local antiquities.”
        The third stage of development of Japhetic linguistics began with the publication of the work Opredelenie Yazyka vtoroy Kategorii Akhemenidskikh Klinoobraznykh Nadpisey po dannym Yafeticheskago Yazykoznanija, "Definition of the language of the second category of the Achasmenian cuneiform inscriptions in accordance with the data of Japhetic linguistics.”
       Extensive research into Caucasian languages brought in its trail the necessity of studying written monuments that were chronologically more ancient, that is, the dead cuneiform languages. Marr began to study these, mainly Elamite, particularly the language of the second category of the Achasmenian inscriptions, and Khaldian, the language of Ancient Van. The former he established as a Japhetic language of the sonorous group, and the latter as both spirant and palatal-dental in structure, that is, of the type of Svanian, Megrelian, and Chanian. Marr examined also Sumerian.
"A most amazing and fundamental acquisition from these as yet unfinished investigations,” he said, ''is this, that although literature existed in at least one of them, namely, Elamite, long before the written Sumerian arose— and the antiquity of the Japhetic languages preserved in the cuneiforms is many millenia greater than any art of writing—nevertheless, the living Japhetic languages of Caucasia, including Georgian, represent, it seems, the most ancient type of Japhetic speech in perfect preservation.”
        The fourth stage began with the discovery in the Pamirs of a Japhetic language, namely, the Vershik, and with the definition of the European Japhetic languages, viz. the living Basque in the Pyrenees, and the dead Etruscan on the Apennine Peninsula. Deeper penetration into the study of chronologically more ancient written languages gave rise to the question : What relation does the written language bear to the living speech ? This, of course, necessitated concentration of research on the collection of materials and on the study of the unwritten languages of present-day Caucasia.
       It was only after such prolonged and persistent research into the
[ 109 ] 
ling uistic materials of the whole of Caucasia that Marr took up the problem of the relationship borne by Etruscan and Basque to the Caucasian languages, which relationship had already been indicated by western scientists. Theoretical analyses of these languages from printed manuals as well as from practical examination of the materials on the spot, established not only their affinity with the Caucasian, but also established them as Japhetic languages.
        The fifth stage of the development of Japhetidology began with the investigation of the genesis of the Japhetic languages and of the relation of the population of the Mediterranean to the population of Caucasia. The fact that the Japhetic family of peoples had in their great variety been steadfastly preserved only in Caucasia, naturally led one to think that the Japhetic peoples of the Mediterranean had emigrated from Hither Asia, and in particular from the regions around Ararat. The region of Ararat was considered as their point de depart on account of the predominance, in the toponymics and tribal names of this region, of the stem " ras " (rosh), which is found both in the geographical name Urartu-Ararat, and in the ethnic name Etruscan-Rasenian. This idea, however, was abandoned when independent investigations in the sphere of Mediterranean toponymics brought to light a series of Japhetic names scattered all over the countries adjacent to the sea and over the islands from Spain to Caucasia. They were found even further eastward, in Central Asia, where the contemporary language of the Pamirian Vershik— Bureshikians—was found to be " most closely connected with the North-Caucasian Japhetic languages." "The more I had the opportunity of studying materials of the Hindu-Kush group of languages," said Marr, “the more I became convinced that in it we had a new group of Japhetic languages, the Central Asiatic. With this discovery arose the necessity, not only of raising the question of the migration of Japhetic languages from Hither Asia into Central Asia, but of the revision of the question of the migration of North Caucasian Japhetic languages from southern Armenia to northern Caucasia."
        The study of the Vershik language stimulated Marr to research into the pre-Aryan languages of Iran.
        “Concentration of attention on the Japhetic languages and the bearers of its typological peculiarities… precised the Japhetidological methods, deepened the Japhetidological perspectives, and opened up the ethno-cultural strata of the history of Caucasia, which overthrew… our generally accepted idea of the influence of the external cultural-historical factors on the life of Caucasia, revealing in everything the primal predominance of the Japhetic norms and a great diapason of their influences, and the radius of the circle or volume of their reverse influence on the adjacent countries. We had to reject even the idea of what seemed to be the
[ 110] 
Iranian influence of Persia over Caucasia… of the Iranian origin of certain aspects of social life of the people of Caucasia, of the Iranian basis of epic stories, and of the Iranian basis of their beliefs and of the origin of the names of the pagan gods of Georgia, as well as of the whole of the corresponding terminology.”
        " With the widening of our knowledge of the Japhetic ethnic world, of the tribal strata and sub-strata of the population of Caucasia, Iranism proved to be either a mirage or a result of later adaptation of local, or generally of the Hither Asiatic, immemorially Japhetic, ethno-cultural aspects and forms, under the cloak of the later Iranian cultural-historical deposits. In this connection ... I was compelled not only to part with myself in the interpretation of a series of Georgian terms, but also completely to reject such work as Bogi yazycheskoy Gruzii po drevne-gruzinskim istochnikam, " Gods of pagan Georgia, in accordance with ancient Georgian sources,” with all its Iranian interpretations. The source of such an upheaval in scientific thought was again an accumulation of new facts, a more extensive and intensive registration of Japhetidological data. These new sources, as yet mainly if not exclusively linguistic, revealed first the elements, and then layers, of the converse influence in the languages not only of the neighbouring but also of peoples far away from Caucasia. The same sources revealed a different degree of relationship of the Japhetic languages with the Semitic… The same sources made necessary, which is most important… the revision of the classification of the Japhetic languages themselves. In consequence, a completely new picture is now being visualized of the interrelationship of the Japhetides not only with the Iranians of Persia and beyond the limits of Persia… but also with the people situate within the limits of the rise and development of Mediterranean culture.” (Japhetic Caucasia and Third Ethnic Element, pp. 11, 12.)
        In 1920 Marr began to analyse the Japhetic elements in Greek. The problem of the Greek language he declared to be indissolubly bound up with that of Pelasgian and Etruscan. His linguistic researches in this connection related not only to the problem of the relationship of Pelasgians to Greeks, but also to the formation of the Greek nation in general during the process of hybridization with the non-Aryo-European Ionian tribes.
        “In the light of Japhetidological data the hybridization could have taken place on the Caucasian, or near-by soil. The hybridization may have taken place in the very region of the Black Sea Coast which, in historical times, was known for the development of Greek colonial life… while in the prehistoric times of Greek national formation, Greece is connected with Caucasia by a rich selection of most ancient myths.”
        Under the circumstances, the search for a centre for the dispersion
[ 111]  
of the Japhetic race became pointless, and further extensive research led to the establishment of a third ethnic element in the creation of Mediterranean culture; this third ethnic element was a pre-Indo-European and pre-Semitic race, namely, Japhetic, which once extended all over the vast region of southern Europe and Hither Asia and later was submerged by Semitic and Indo-European waves which spared, however, the oases in the Pyrenees, in Caucasia, and in Central Asia in the region of the Pamirs.
        Such a formulation of the problem, which later was again somewhat changed, led to the investigation of possible Japhetic survivals in the languages of modern European populations, namely, in the Romaic and Germanic languages, which, indeed, were found to contain Japhetic elements to a considerable extent. This direction of research, which ever widened the Japhetic horizon, established beyond doubt the impossibility of giving this linguistic group a name of any geographical region, even that of Caucasia. For in Caucasia, itself, we have, though well preserved, yet an insignificant part of what was once a vast region populated by this third ethnic element. The appropriateness of the conventional biblical term Japhetic is thus fully confirmed.
        Each of the above-mentioned stages of development of Japhetic linguistics has its own problems requiring further research and clearer definition. One of many such problems is, for instance, the connection of Slav languages, in particular, of Russian, with Japhetic languages, which connection in Marr^Js opinion was effected through the Scythian language. “That the Scythians were originally Japhetides is clear from their name, ' Scythes ’ and from the national Scythian variety of the latter, viz. ‘Skolot ’”
        While working on this problem, Marr came across stories hitherto unknown from the history of ancient Russian and ancient Armenian literature. These he published under the title of Knizhnye legendy ob osnovanii Kuara v Armenii i Kieva na Rust (Skifskoe Predanie v Yafetidologicheskom Osveshchenii), “ Book legends on the foundation of Kuar in Armenia and of Kiev in Russia (The Scythian tradition in the Japhetidological light).” In this work he established connections between Japhetic Caucasia and Russia.
        From the foregoing it will be seen that Marr was constantly searching for this third ethnic element of Mediterranean culture in different parts of Asia, Europe, and even Africa, and he revealed it everywhere as forming the earliest racial cultural substratum. The fourth and fifth stages already mark the paleontological character of the research; with these stages, too, Japhetidology ceases to be a mere Caucasiological science.
        Transferring the centre of gravity of his linguistic research from
[ 112 ] 
the formal phonetico-morphological side to the inner sense of speech, that is, to Semantics, Marr noted connections between Japhetidological paleontology and the problem of the social origin of language.
        “The process of speech formation,” he said, " and its typological transformation was taking place, of course, before the Japhetides became a separate particular family, and the human race in its pre-Japhetic state requires another denomination. The fact is incontestable, however, that the Japhetic languages, more obviously than any other, have retained the accumulation of the remote stages of the development of human speech ; in them may be clearly traced the process of sedimentation of elements, even of whole layers of a language in its initial stage of development… Languages reflect the social state with its psychology of different epochs, including the most ancient. In the meanings of words, or Semantics, the abyss that separates prehistorical from historical times is revealed particularly clearly… The Japhetic languages show that originally ‘word’ was not conceived as something utterable, but as a means or weapon for conveying information. Japhetic articulate speech has preserved us such a conception of speech, among others, in the term 'to speak’ properly denoting ‘informing by mouth’ or ‘by face' (compare Georgian pir-utqvi 'wordless, animal', literally, one who cannot convey information by mouth or face). Human speech does not originate from some primordial single language, and the Ur-Sprache conceived in this sense is not a scientific proposition… Language is a creation of human society… Language is one of the social superstructures.
        …The development of language is determined (or defined) by the development of material productive relations, the dialectics of which it reflects… the material productive relations determine the dialectics through the thinking which arises simultaneously and is organically connected with it… The development of language does not proceed evolutionarily, but dialectically… The mutual reaction of language and thinking is the mutual reaction of form and content."
        It was this materialistic, ideological, and dialectic interpretation of linguistic problems that led Marr to introduce the term System in place of " family ", which fact is mentioned earlier in this article.
        Thus, according to Marr, a primitive articulate language, a unitribal language, does not, and could not have existed. It is only a product resulting from tribal economic interrelationship, and is a deposit left by this ever multi-tribal society.
        Who, then, were these primal tribes whose common, social efforts are responsible for the creation of articulate speech ? And how many tribes were there?
        Investigation into tribal and toponymic names had established
[ 113 ] 
a limited number of fundamental tribal terms in which very frequently occur the following stems : rosh, ber, sal, yon… which in their turn, by variation of the consonant element, expand to seven and more; or, on the other hand, contract to three : rosh, ber, sal, which by the process of hybridization have given rise to many ethnic names and languages.
        The usual method of establishing the basic meaning of the word is completely rejected by the Japhetic theory. According to it the word "sky" denoted, not only "sky” or "heavens", but also everything that the imaginationof pre-historic man connected with the sky, such as clouds, planets, stars, and even birds. After extensive research into this question, Marr came to the conclusion that the above-mentioned fundamental Japhetic tribal names with their derivatives lay at the basis of all the Semantic series, for these tribal names could be detected in the name of "heaven" in all its Japhetic aspects, viz., of the upper (the sky in its proper sense), the middle (the earth), and the lower (the infernal regions, also waters).
        In this way it became clear that each of the fundamental tribal names indicated was at the same time also the name of the heavens. But this name was also passed on to its derivatives, in particular to its derivatives from the world of animals. Thus, connected with the upper firmament are birds; with the lower, water and creatures of the infernal regions. Separate animals, too, and even metals appeared to bear the same radical or tribal name. Hence the task of Japhetidology was to establish which animals were connected with which tribe and to such animals to assign the term "totem". For instance, "horse/bird" was found to be the totem of the Urartians-Etruscans.
        Research in this direction was complicated by emergence of the fact that the basic Japhetic tribal names were detected in morphological particles; in the first place in the suffixes of the plural (tribe -- posterity --children --plurality). It was also revealed that the verbs "to speak", "to create", "to build", "to move", etc., were connected with tribal terms, so that peoples, as Meschchaninov puts it, rosh-ed, sal-ed, ber-ed, or yon-ed, in different meanings of speaking or acting. In other words, during the long process of development of human speech, the action had become bound up with the actor, that is, the action was only an external manifestation of the man himself, more exactly of the tribe itself. Likewise, the tribal names were detected in social terms, as "the tribe-subduer" and "the tribe subdued".
        Naturally, while working in this direction, Semantics held the first place in Marr's attention, and consequently he traced out a whole Semantic series, such as: Sky->top —> mountain -> head; Sky —> mountain -> temple -> house; Sky -> deity -> sovereign ; or Sky -> deity -> lord, master -> power -> ruler; hand -> own, possess -> power
[ 114 ] 
-> ruler, or Sky -> water -> birth -> to create -> to do, and hand -> to do, make -> create, etc., and with the accumulation of such data Japhetidology goes deeper and deeper into the process of the creation of speech. Thus Marr established that the primal Japhetic tribal names form the basis of the whole of human word-formation and word-derivation of a definite period, when sky held the foremost place in man’s conception of the world.
        Japhetic languages do not represent a racially individualized group, but a definite stage in the development of languages in general. This “stadium" under pressure of the changed conditions of social life, became transformed into Indo-European, the only survivals of this process of transformation being the Basque language in the Pyrenees and the Caucasian languages. Semitic and Hamitic are likewise branches of the languages of the Japhetic stadium, which circumstance explains the relationship of Georgian to Semitic, as Marr expounds in his “Tables".
        Finally, "among the many original explanations provided by the Japhetic theory, let us quote the case of the Grammatical Comparative degrees. The lower, that is, the positive degree, is explained as a deposit of the name of the lower social class; the middle or comparative degree as a deposit of the middle class; and the superlative as a deposit of the upper class. And as the social class or rank represents a survival of the tribes (“the tribe subduer and the tribe subdued”), therefore, in the corresponding degrees of comparison of the primitive adjectives “good", “bad", names were detected of the tribes who made up the people in whose language the degrees of comparison mentioned were used. In other words, most of the basic social terms in Caucasia are of ethnic origin, the result of particular tribal relations and of the transformation of tribes into social classes. Thus, the common or lower social class both in Georgia and in Armenia bears the name of the people known to the European antique world as Colchians, viz. Georgian glekhi ; Arm. grehik “a peasant"; of the same origin is also the old Georgian adjective and noun, glakhak <- mod. Geo. glakha "dirty, bad, a beggar”. A certain class of nobles (in Russian, dvoryane) was originally known in Georgia, and is still known among the mountain peoples of Caucasia, not excluding the Svans, under the ethnic name of a Japhetic tribe, viz., Varg//Marg, whence the Georgian adjective vargi, varga “useful, good”. The name of the Moskhians//Mosokhians, the local popular form of which is a-mysta, has been preserved by the Abkhasians as the title of a certain rank of their nobles, etc.[2]
        In Marr’s own words, his new teaching of language introduces two aspects: One concerns the affinity and interrelationship of
[ 115  ] 
t he indigenous languages of Caucasia, "the Japhetic languages,” found also outside Caucasia—an aspect of teaching which has a local importance for the regions where these Japhetic languages are spoken, that is, in Caucasia and Mesopotamia, in the Pyrenees Peninsula, and in the Pamirs; the other concerns the paleontology of human speech, namely, the general universal question of the origin of languages in general, of its different aspects, and, in connection with this, of the classification not only of Japhetic languages, but also of the languages of the entire world.