THE QUICHUA AND OTHER AMERICAN LANGUAGES AGGLUTINATIVE, LIKE THE TURANIAN AND IBERIAN LANGUAGES - COMPARISON OF DECLENSIONS IN NOUNS AND PRONOUNS - ULTIMATE IDENTITY OP THE GEORGIAN AND BASQUE PRONOUNS FOR THE FIRST AND SECOND PERSONS - CONCLUSION FROM THE WHOLE EVIDENCE. In support of the inferences to be derived from numerals, personal pronouns, and words of familiar use, the three branches of evidence which have occupied the two preced- ing chapters, it may now appear requisite that something should be added in reference to that most important test of linguistic affinity, grammatical construction. But, in the present case, this is needless for me to do, seeing that the work has already been satisfactorily done by others entitled to speak with authority upon such a question. It may, there- fore, be sufficient, if I merely give two quotations in corro- boration of my opinion, that the American languages, of which the Quichua is one, ar6 to be ranked with the Tura- nian and the Iberian, whether the type of the Iberian be sought in the Caucasus or in the Pyrenees. It is true that, in the first quotation which I shall adduce, it is intimated, in opposition to my own views, that there is not " the re- motest indication of a common origin" for the languages of America and the agglutinative languages of the Old World: but, if the numerals and words that I have compared to* gether render it necessary to modify or abandon such a statement, and to admit that there is good reason, not to say strong reason, for the belief in such a common origin, then the argument to be derived in favour of that belief from the agreement of grammatical structure may be fairly pressed, and will give completeness to the whole demon- stration.
142 PERUVIA SCYTHICA. My first quotation, then, is from Markham's Quichua Grammar and Dictionary, p. 60 : - "It will have been seen that the most remarkable feature of the Quichua language is the power of constructing words by means of affixes or particles joined to the root, and thus providing for the formation of innumerable words, and even whole phrases. Du Ponceau called this mode of composition polysynthesis ; and, as most of the American languages possess this faculty, he named them poly synthetic lan- guages. Dr. Max Miiller, however, considers this pecu- liarity to be nothing more than agglutination, as distin- guished from the inflexion of the Aryan and Semitic lan- guages. He, therefore, groups the American with the other languages which he calls agglutinative (a name given by William von Humboldt) in Asia and Africa, not because there is the remotest indication of a common origin, but from the absence of any organic differences of grammatical structure." My next quotation is from Max Muller's Lectures on the Science of Language, vol. i, p. 320 (ed. 1), and bears espe- cially on the resemblance between the American and the Iberian languages : - " The Bulgaric branch" of the Finnish stock " comprises the Tcheremissians and Mordvinians, scattered in discon- nected colonies along the Volga, and surrounded by Russian and Tataric dialects. Both languages are extremely artifi- cial in their grammar, and allow an accumulation of prono- minal affixes at the end of verbs, surpassed only by the Bask, the Caucasian, and those American dialects which have been called Poly synthetic." To these statements I subjoin in illustration a comparison of declensions, choosing the word ' house' as the word to be declined. The analogy between the Quichua and any mem- ber of the Turanian group seems here as well marked as the analogy which unites that particular Turanian language
PERUVIA SCYTHICA.
143
to the rest of the same group : indeed, it might be said that the Quichua comes as near to the Hungarian or Lap- ponic as either of those two members of the same Turanian family, the Finnish, do to one another. They will, there- fore, immediately follow the Quichua in the comparison below : -
Singular.
Plural.
Quichua.
Nom. huasi
huasi-kuna Gen. huasi-p
huasi-kuna-p Dat. huasi-pak
huasi-kuna-p ak Ace. huasi-kta
huasi-kuna-kta
Hungarian. hdz
hdz-ak hdz-nak
hdz-ak-nak hdz-nak
hdz-ak-nak hdz-at Lapponic, hdz-ak-at * wieso
wieso-h wieso-n
wieso-i wieso-i
wieso-it wieso-b
weiso-it
Japanese, i iye
iye no
iye ni
iye wo
The Japanese for i house' is also written ihe, and is analogous to the Chinese uh, 'house', as well as to such Brazilian words for ' house' as the Curetu uee, the Omagua uca (also in Peru and Ecuador), the Tupi oca (whence oca $ui, 'at home'), and the Mundrucu deka (whence deka ihto, 'at home'). See ante, p. 61. We now pass to the Iberian: -
144 PERU VIA SCYTHICA.
Basque.
eche (or iche)
eche-ren 1
eche-ri
eche
Georgian • chokhfi
chokh-ehi chokhisa
chokh-ebisa chokhsa
chokh-ebsa chokhi
cliokh-ebi
In the Tibetan, and in other Bhot languages, an m ter- mination is affixed to the base of the previous words, which produces for ' house' the Tibetan khyim, and such words as the Nepalese khim and dhim,, and the Gyarung dhem; which may be compared with South American words for c house' like the Botocudo kjiemm and tiemm, the Guayacuru dimi, and the Coropo scheh-me {ante, p. 61). The Tibetan khyvm is thus declined : - Tibetan. khyim khyim-rnamas khyim-kyi khyim-rnamas-kyi khyim-la khyim-rnamas-la khyim khyim-rnamas Turkish. ev ev'ler ev-in ev'ler-in ev-eh ev-ler-eh ev-i ev-ler-i Next to these Turanian languages may come the most Turanian of the languages classed as Aryan : - 1 The r is merely euphonic here, and in the dative. ■ * Cottage', or ' hut'.
PERUVIA SCYTHICA. 145 OSSETIC. ehazar chazar-thd chazar-iiy chazar-th-iiy chazar-dn chazar-th-dn m m chazar-iiy chazar-th-iiy So also we have in Ossetic, from the root arf, € deep' (apparently allied to the Armenian unrap, 'a very deep ditch'), the nominative arfade, € depth', with the accusative arfadiiy or arfadiy. Arfadiy may be compared with the Lycian accusative, Srafazeya, * tomb'; a word occurring in ' the following well-known bilingual inscription, in company with prinafatu, tedSeme, and lade, which have been analysed and compared above (pp. 60 to 76) : - iwieya erafazeya mdte prinafatu to fJivrjfjLa roBe enovqaaTo sedereya pe 4 . . . neu tedeeme urppe etle euwe %t&apio$ HapfievovTO? vlos eavrcoi se* lade euwe se tedSeme p . . . e . . . Wye* icai T7]c yvvauci tcai viau Hufiuikrji In placing the case-suffix after, instead of before, the plural suffix, the Ossetic is more like a Turanian than an Aryan language. But the German hdus-er-n, as compared with the Latin dom-ib-us, presents the same anomaly. The Ossetic case-suffixes are like those of the Tuschi, who dwell in their vicinity : - Tuscm. Singular. Plural. Norn. Dal, ' God\ DaUi Gen. Dal-e Dal-a Dat. Dal-en Dal-i-n Ace. Dal Dal-i This Tuschi word for ' God', Dal, I have compared else- where with the diel, dil, * sun', of the Albanians, who may be u
146 PERUVIA SCYTHICA. considered as the representatives of the ancient Illyrians. In the Toscan dialect of the Albanian, kyen, ' a dog*, is thns declined in the indefinite and definite forms : - Toscan.
If now, from the declensions of the Caucasian Tnschi, in whom we may recognise the Tusci of Ptolemy (lib. v, cap. 9), and also from those of the Albanian or Illyrian Toscans, both immediately preceding, we are next led to examine those of the ancient Italian Tuscans, or Etruscans, we shall find in them the regular marks of Aryan declension. At least this would be the case, if we may judge from the following inflections of what we may assume to be proper names in the Perugian Inscription; a monument which, however, is not bilingual, so that the inferences as to cases are not absolutely certain : - Etruscan. Velthina Afuna Velthinas Afuna 8, Afunes Velthinam Afunam Similarly we have, in a Phrygian inscription, materes and materan, = Latin matris and matrem, = Sanskrit mataras and mataram. 1 The Etruscan and Ossetic are Aryan lan- guages crossed with Iberian, as English is a German lan- guage crossed with French. But, in Etruscan, it is nume- rals instead of declensions that show the effects of Iberian intermixture. The converse is the case in Ossetic, where 1 For this Phrygian inscription, as interpreted from the Armenian language, see my Asiatic Affinities of the Old Italians, p. 71.
PKRUVIA SCYTHICA. 147 all the numerals of the decade are regularly Aryan, with the exception of far-ast, ' nine'; and even that contains the Ossetic form of the Aryan ' eight', ast. The 8 termination of the Aryan genitive bears some like- ness to the termination of the Georgian- genitive and dative, -sa. But this sa seems rather the dative than the genitive suffix, for shen 3 ' thou', is thus declined in Georgian : - Nom. shen Gen. shen-i Dat. shensa or shen-da The plural suffix of the Quichua, as will have been seen above, is kuna; and it is well compared by Lopez (p. 98) with a Tibetan plural suffix, leun, a word signifying likewise ' all 'in Tibetan. Lopez also compares the same Quichua plural suffix, kuna, with the Sanskrit fjuna, € quality', tri- guncby ' triple 9 , etc. : but the previous explanation from the Tibetan is manifestly preferable. And this Tibetan word, kun, s all, whole, every', which is at the same time a plural suffix in that language, as kuna is in Quichua, may again be employed to explain another Quichua suffix, nkuna, imply- ing ' every': thus from huata, 'year', is formed huata-nkuna, 'every year' (Nodal, Qramdtica Quichua, p. -210). For pronouns, the Tibetan uses cog as a plural suffix; and here there seems another analogy between the Quichua and the Bhot languages : -
Quichua. Tibetan. Lohorong (Nepal). ' I', noka na ka
nya ka-na 'We', noka-ndik iia-cag ka-6ika iioka-yku nya-ngo ka-di Thou*, ham khyod khe hana 'Ye', kam-cik khyod-6ag khe-ngo hana-di
148 PKRUVIA SCYTHICA. In these pronominal forms, the Bhot members of the Turanian group seem to approach nearer to the Qnichua than Iberian languages do. Of these Iberian languages, which, with the American and the Turanian, I consider to compose the great Scythian family of tongues, I here take as specimens the Basque and the Georgian, giving in each case the common and the agglutinative forms. These last will be recognised by the presence of a hyphen, which, in the actual languages, would be replaced by a verbal root. Thus, 'thou art' is in Basque sera, and in Georgian khar : 'ye are 9 is in Basque zerate, and in Georgian khavth. 'Thou' is, therefore, represented here in Basque by 3-, and in Georgian by kh- ; and 'ye', in Basque by z-te, and in Georgian by kh-th. The Ossetic plural suffix, -tha {ante, p. 145); may be compared with the Basque -te and the Georgian -th, which convert 'thou* into 'ye*.
QUICHUA. Basque. Georgian. ' I\ noka ni nik me
n- m-
-t V- i Wq\ riokandik gen- 6cen
gu gv-
guk v-th Thou*, kam zen~ z- zu zuk hik shen
hi kh- 9' ' Yo\ kamdik zen-te thchven
peru via scythica. 149 Quichua. Basque. Georgian. 'Ye\ zuek z-te kh-th g-th By combining together the Georgian forms for 'we* and ' y e '> given above, the full form for r we' would come out as dventh ; and for ' ye', as thohventh. The Georgian pronouns of the first and second persons may therefore, as it seems, have been originally constructed thus. There would have been two bases for the first person singular, me and dven, from which last would be derived the present genitive of me, ' V, namely, dem-i, which implies dem, ' I\ The two Sanskrit bases, ma and aham, may be compared with the Georgian bases, me and dven. From the form me (=Basque nij comes the Georgian prefix, m- (= Basque n-), as from the form dven (=Lesgi den, € V) would come the Georgian prefix, v-, with the Basque suffix, -t. The Georgian plural suffix for pronouns would be -th, thus giving dventh, 'we*, as the plural of dven, e V; and from dventh, € we', may be derived the present Georgian form for ' we*, dven (with the Basque prefix, gen-), as well as the Georgian prefix, gv- (=» Basque gu, r we'), and the Georgian prefix-suffix, v-th. This Georgian dven, 'we', which I have inferred to mean formerly 'P, with the actual genitive, dem-i, 'me-V, is like the Chinese din, dam, " anciently used for I ; now the royal we" (Williams' Tonic Dictionary). The second person singular in Georgian seems to have been originally formed from the first person singular, dven, by prefixing th. Its full form would thus be thdven (cf. Sanskrit tvam), from which can be deduced the present Georgian form, shen (with the Basque prefix, zen-), and the Georgian prefixes, g- and Ich- (with the Basque forms, hi and zu, and the prefix, z-). Finally, the second person plural in Georgian would be formed, as in the case of the first person, by the addition of
150 PfiEDVU SCYTHICA. -tli ; and its full form would therefore be thdoenth, which contains the present Georgian form thchven (with the Basque prefix-suffix, zen-tej, and the Georgian prefix-suffixes, g-th and kh-th (with the Basque prefix-suffix, z-te). I dwell on these analogies between the Basque and the Georgian, because, if both languages are to be referred to one stock, the Iberian, we may be able to see at once how probable it is that the Pre- Aryan population of the countries which i form the northern shores of the Mediterranean Sea was i Iberian also. Conclusion. Haying now brought the enquiry to an end, as far as the Quichua is concerned, it remains for me to point out the results which may be derived from the linguistic compari- sons that have been instituted, in answering the questions propounded at the outset of our investigations (p. 5). With regard to the first of those questions : - " By what route did the Americans pass from the Old World into the New ?" - nothing more than has been said need be said ; as there is no good reason for deviating from the most common opinion, that the Americans as a body passed into the New World from North Eastern Asia. The second question was : - " To which of the races of the Old World are the Ameri- cans most nearly allied ?" Six races may be mentioned in the world : - 1. The African or Ethiopian, now occupying the southern half, or rather the south-western half, of Africa. 2. The Semitic or Syro- Arabian, occupying (when taken in conjunction with the Sub- Semitic, or Libyan) the northern or north-eastern half of Africa, with Arabia, Syria, Meso- potamia, and Chaldaaa. 3. The Aryan. 4. The Iberian, or Caucasian.
PERUVIA SCYTHICA. 151 5. The Turanian. 6. The American. The nearest affinities of the American race would be with the Turanians and the Iberians, and all three races might either be thrown into one under the name of € Turanian,' or else, which may appear preferable, we might adopt the name of ' Scythian* for the whole race collectively, and employ the names of ' Iberian,' ' Turanian,' and € American/ for its three subdivisions. In doing this, we should be carrying out and extending the idea of Bask, who divided as follows, about half a century ago, the race which he called ' Scythian' : Tartar (or Turkish). North Asiatic (Finnish and Samoyed). Mongol-Tungusian . North American. The inhabitants of South Eastern Asia have since been added to this class, but the claim of the Americans to be admitted into it has hardly been recognised, nor even, universally, that of the Caucasians. Our previous six races are now reduced to four : - 1. African. 2. Semitic and Sub-Semitic. 3. Aryan. 4. Scythian (Iberians, Turanians, and Americans). We have next to consider to which of the first three races the Scythian race is most nearly allied. This is a more difficult question to decide ; but it seems to me that the Scythians are, at any rate, less closely related to the Semites and Sub- Semites than they are to the Aryans and the Africans ; and that, as far as their vocabulary is concerned, their connection is rather with the Africans than with the Aryans. Admixture alone is quite insufficient to explain the coincidences between African and Scythian words ; nor is the idea of an affinity between Turanians and
152 PERU VI A SCTTHICA. Africans, as well as one between Turanians and Americans, at all a new one : - "An extension of the Turanian family to these two con- tinents (Africa and America) has been hinted at by several scholars. The Greenland language has been pointed out as showing a transition of Turanian into American dialects, and the researches of physical science have clearly indicated the islands east of Siberia, as the only bridge on which the seeds of Asia could have been carried to the New World. As to African dialects, all is still conjecture, except this, that, besides the Semitic type of some African languages, such as the Galla, spoken north of the Equator, there is another grammatical character impressed on other idioms, as, for instance, the Hottentot, which, by its mechanical perfection and artificial complication, invites a comparison with the grammatical system of the descendants of Tur." 1 The degrees of relationship which unite the four great races of the globe may be conveniently represented by the following scheme, where each race would have its nearest kindred on both sides, and its remotest kindred opposite:- Aryans. Semites \ r Iberians, and > Scythians X Turanians. Sub-Semites J \ Americans. Africans. As the Scythian languages seem most attached to the African languages by their vocabulary, but to the Aryans by their numerals, it becomes ^probable that the Scythians continued in contact with the Aryans after the Africans had disappeared from that quarter of the world where both Scythians and Aryans remained. And this brings us to our third question : - 1 Max Miiller in Bunsen's Philosophy of Universal History (1864), vol i, p. 484.
PERUV1A SCYTHICA. 153 " In what part of the Old World was, most probably, the primitive seat of the ancestors of the Americans as a dis- tinct race ?" We have already inferred that the Aryans and the Scy- thians remained for a long time in communication together during the early ages of mankind : and therefore, as the Aryans have been traced np to a single tribe, or even house- hold, in Mount Imaus, we should be inclined to conjecture that the primitive seat of the Scythian race was in the same neighbourhood. The Yarkand country, on the east side of Mount Imaus, might be selected as a likely position for that primitive seat: and, as the Scythians spread abroad from such a centre as this, a time may have come when the little Aryan family was completely enclosed in its mountain- home by the three nations into which the Scythians became divided; the Iberians occupying the country between Imaus and the Caspian ; and the Great Desert of Gobi, which so nearly cuts Central and Eastern Asia into two halves, separating the Turanians on the south from the Americans on the north. About this time, the Semitic nations may have been on the Euphrates, while the rest of the Old World, America being as yet unpeopled, was thinly over- spread with scattered Africans; the African region then comprising Africa, Europe, Siberia, and the maritime coun- tries of Eastern and Southern Asia. In Siberia and Northern Europe, however, any African population that may have existed must always have been extremely sparse, and would soon have been lost among the advancing Scythians. Of this Scythian stock, the Iberian branch, stretching out its arms towards the west, would take pos- session of Asia Minor and Europe, while Northern Asia would fall, in like manner, into the hands of the Americans ; and while the Turanians, still on the south side of the Desert of Gobi, would be led from Tibet, by pursuing the courses of the great rivers, to the occupation of India and x
154 PEBDVIA SCTTHICA. China and South Eastern Asia. The next change would probably be, that the Americans would pass over into the New World from Northern Asia, being succeeded in Siberia and Tartary by a Turanian population, which may either have followed their steps from the common home of the Turanians and Americans on the east of Imaus, or else have taken a northerly course from some point or points in the enlarged Turanian area between Imans and the Pacific. When the Aryans, having multiplied in Imaus, at length emerged from its valleys, the Old World may have been thus peopled:- Africa, south of the Desert of Sahara, may have been occupied by the African or Ethiopian race, while the valley of the Nile and the coasts of the Mediterranean belonged to the Libyans or Sub-Semites, the Semites themselves possessing Arabia, Syria, and Assyria. The North of Europe, with Northern, Central, Eastern, and South Eastern Asia, including India (and with, perhaps, the addition of Iran or Persia), would have been Turanian ; and the remaining and greater part of Europe, with Asia Minor, Armenia, and the Caucasus (and, perhaps, Iran), would have been Iberian. But, when the Aryans issued from Imaus, the advancing hordes of Celts, taking their route into Europe on the north of the Caspian and Euxine, and followed by Germans and Slavonians, would have con- fined the Turanian Fins to the extreme North of Europe, and the Iberians to the South of the same continent, where they held their ground, mingled with Aboriginal Africans, as Iberians, Ligyes, Tuscans, and Pelasgians. Finally, at a later period than that of the Celtic immigration into Europe, the Thracians, a branch of the Southern Aryans, having detached themselves from the Iranians, and deprived the Iberians of Armenia, where a Thracian language is still in use, pushed themselves eventually as far westward as Italy, if not farther, expelling, subduing, or incorporating with themselves, as they advanced, the previous Iberian popula-
PERUVIA SCYTHICA. 155 tion. Yet in Asia Minor, under the protection of Mount Taurus, the Lycians still survived as a relic of the former Scythian possessors of the country, as the Basques still sur- vive in the Pyrenees to represent the ancient Spanish Iberians. In Dacia, again, though the old Dacian language was Thracian, and akin to the Armenian, a previous Iberian occupancy seems, nevertheless, indicated by the termination of Dacian town-names, -dava, = Georgian daba, € village'. 1 And, lastly, in Etruria, although the ruling language there is of Thracian origin, yet the important foreign element in the Etruscan vocabulary, including all, or very nearly all, the numerals, which are to be considered in the next chapter, still marks the co-existence of Ibero- Africans among the Thracians, not only, it may be, in Etruria itself, but likewise all along the route by which the Thracians advanced from the Caspian. Thus the Tuscans of the West became Aryanised through the instrumentality of the Thracians, and were succeeded by Thraco-Tuscans in Etruria, nearly as Gelto-Ligyes and Celt-Iberians arose in Gaul and Spain under Celtic influence, and as the mixed race and language of the Ossetes has arisen in the Caucasus among the de- scendants of the ancient Tuscans, Ligyes, and Iberians of the East. Here, in the ' mountain of languages', as it is called, the Old Iberian still exists in many different forms ; but nowhere else, except in the Pyrenees, is a true Iberian language to be heard. What the Celfciberian may have been, we do not know, but may suppose that it was Iberian rather than Aryan, and mixed with African. Our fourth and last question was : - " In what part of America are the nearest kindred of the Quichuas to be sought : or, in other words, who appear to stand to them in a relationship at all resembling that 1 See ante, p. 72. So, also, the Dacian for * tongue 1 , Zd\\a, appears Scythian, and = Fatagonian del, Turkish del, tel, Australian tale, African dali y etc. (ante, p. 93).
156 PERUVIA SCYTHICA. which the Germans bear to the English, or the Welsh to the Irish ?" As far as this question can be answered, it has been so above (p. 50) in favour of the territory called New Mexico or Arizona ; a country in which evidences of a former civi- lisation, analogous to that of the ancient Peruvians, continue to be found while I write. 1 In North America, it is the languages of California, New Mexico, and Texas, which have offered most parallels to the Quichua ; and it is possible to identify all the ten Quichua numerals with those of the Tesuque, one of the Pueblo nations in New Mexico (ante, p. 50). 1 A recent discovery of such antiquities is mentioned in a paragraph in the Times of the 8th of June, 1874, headed : - Ancient Ruins in Arizona.