Since we are chiefly concerned with Tungus parallels, it will be useful to dwell here on the question of existing Tungus material. Some remarks are needed for showing what the character of the Tungus material is and how far it can be used. The Manchu language, which has attracted the attention of a great number of philologists and linguists, is the best-studied Tungus language. However, A. Sauvageot does not use all the material known — he uses H. G. von der Gabelentz's dictionary. His choice is not motivated. However, the existing Manchu dictionaries are not always reliable, and this is even so with I. Zaxarov's dictionary. The reason is very simple — all Manchurists used to study Manchu through the medium of the Chinese language, or with the help of Manchus who did not possess this language as their native one. Thus it is seen that the «meaning» of words was not usually received in the Manchu complex, but through the Chinese complex, so that these linguistic materials ought to be regarded as reflecting Chinese complexes which are translated into Russian (Zaxarov), German (Gabeientz), and French (Amiot) in terms corresponding to the Chinese complex, and thus they are greatly distorted. The whole complexes, as, for example, the terminology of social organization, the terminology of shamanism, etc., are represented very often in their Chinese adaptation. As a matter of fact, none of these authors could do it otherwise, for they did not deal with the Manchus who spoke Manchu and thought in Manchu terms. I can illustrate it with hundreds of examples. Any one who will make a new attempt at remodelling these dictionaries without studying the Manchu ethnographical complex will be in the same position. So when the terms refer to the Manchu specific conceptions, one cannot rely upon the dictionaries known [70].
Fortunately, we have good material for the restoration of the Nuichen language by W. Grube («Die Sprache und Schrift der Jucen»), who has transcribed Nuichen words from the Chinese transcriptions. The possible defects are evident — the Chinese complex has produced its effect upon the phonetic system and «meaning.» The same is true of A. Wylie's list of Nuichen words («Translation of the Ts'ing Wan K'e Mung, etc.»)
The other Tungus material is much poorer. W. Grube has made a heroic attempt at compiling a Goldi dictionary («Goldisch-Deutsches Worterverzeichniss») from various materials, both published and unpublished. The material is extremely heterogeneous. Besides the Goldi words, there are also included some words given in Manchu. Moreover, several dialects are also given. Since the material was not originally gathered by specialist-linguists, it is not always properly recorded from the phonetic point of view. Individual or occasional emphatic pronunciations are often mistaken for usual ones. In many instances the phonetic complex is greatly affected by being represented as reflected in the Russian phonetic system. The translations in W. Grube's dictionary have passed through two semantic complexes; namely, once through the Russian complex and the second time through the German. There are many instances of misunderstandings which have originated from this source. The translations into Russian even are not always exact when they are taken from the texts representing translations into Goldi from Russian (as shown in Protodiakonov's material); they are very often a mere adaptation of the nearest known words to the Russian complex. In many cases the translations of the original Goldi texts are entirely erroneous. W. Grube has very often shown parallels from other Tungus dialects, as well as Manchu. However, these parallels are very often defective, owing to the erroneous translations and erroneous transcriptions. The chief source of the Manchu comparative material was H. G. von der Gabelentz's dictionary. Moreover, the Goldi material covers only a small part of the existing lexic complex, so it is not at all complete. P. P. Schmidt, in his dictionaries, has used the Goldi material in a more or less corrected form. However, the corrections cannot always be regarded as sure records, for they are hypothetic in their nature. The Goldi language in its Olcha dialect has been studied by P.P. Schmidt, who has published a short dictionary. The using of the Goldi material requires, first of all, great caution and a thorough knowledge of published material and possible ways of distorting Tungus words.
Amongst the earlier publications, there are two publications of importance — a dictionary made by M. A. Castren on the Urulga and Mankova dialects of Transbaikalia. These small dictionaries were recorded a few decades prior to the extinction of the Urulga dialect, at which time it was already saturated with Buriat words used parallel with the Tungus ones or already incorporated into the Urulga dialect. The records of M. A. Castren are generally good, but in some instances the Tungus phonetic system is adjusted to the Buriat phonetic complex, as shown, for instance, in M. A. Castren's preference for a. etc. Another publication is that by A. Schiefner, who presented the material gathered by A. Czekanowsky. This dictionary is not very extensive and naturally suffers from its incidental character and uncertain translations.
Some other collections of Tungus words of different dialects exist, gathered amongst various Tungus groups by T. von Middendorff, R. Maack, A. O. Ivanovskii, Baron Maydell, and others. These collections hardly cover a few hundred words.
In recent years there was a series of four small Tungus dictionaries published by P. P. Schmidt resulting from his own records and those made by incidental travellers and also completed from corrected and revised records previously published. There are presented: (1) the above-mentioned Olcha dialect; (2) the Orochi (with at least two dialects); (3) the Samagir and Samar (one of the Goldi groups); and (4) the Negidal (four dialects, according to P. P. Schmidt) altogether containing about five thousand words. The words are not always free of the suffixes that are not found with the translations; the translations in some cases are not exact which depends chiefly on the persons who gathered the material. Generally speaking, these dictionaries present an important source for Tungus comparative studies. Two more valuable publications, by E. I. Titov («The Tungus-Russian Dictionary») and N. N. Poppe («Material for the Study of the Tungus Language. The Dialect of the Barguzin Tungus»), are known. E. I. Titov has published an extensive collection of words (in many cases illustrated with texts) of the Tungus of Transbaikalia and the Irkutsk Government. It contains about thirty-two hundred and fifty words of four or five dialects, unequally represented. Apart from the technical defects, such as transcription and wrong translations, the great defect of this dictionary is that the author himself was affected by a strong influence of the non-Tungus complex, so that many translations reflects his own complex and not the Tungus complex [71]. So this dictionary in some cases may involve into involuntary mistakes and thus it cannot be used before being carefully revised.
N. N. Poppe has published a collection of about twelve hundred words of the Nomad Tungus of Transbaikalia (Barguzin), which are well transcribed and usually well translated.
From the enumeration of the above-quoted publications one may see that the Tungus material is not numerous nor extensive. However, A. Sauvageot did not use all at his disposal, and in most cases confined himself to quoting parallels from W. Grube's Goldi dictionary.
Since I am going to quote various Tungus dialects not mentioned in the previous lines I must refer to my own and others' unpublished material. Besides about thirty thousand Northern Tungus «words» gathered by myself I have also some unpublished material from other persons and my extensive material regarding the Manchu spoken in the Aigun district. These data are under analysis and will be published, I hope, in due time. The enumeration of dialects and abbreviations are given at the end of this section.
The material published and unpublished which I have at my disposal is not yet sufficient for an exhaustive treatment of the Tungus lexic wealth. There are some dialects so little investigated that it is very hazardous to operate with them. Yet, as stated, the records are not always absolutely reliable. Is it not really premature to use such material for comparative purposes of great scientific responsibility?
In my work on the social organization of the Northern Tungus and others I have shown how ethnical units are formed amongst the Tungus. The distribution of dialects does not exactly correspond to that of the existing ethnical units (ethnoses). So, for instance, amongst the Birarchen, who are opposed to the Kumarchen, one may observe two dialects, one of which is so similar to that of the Kumarchen that it ought not to be regarded as a distinct dialect. Yet the dialect RTM does not differ very much from that spoken in the southern part of the Yakutsk Government. On the other hand, the Goldi language contains at least four dialects. There is no doubt that formerly the language spoken by the Kumarchen and the Khingan Tungus was the same; but the Khingan Tungus. under Mongol influence, have greatly changed their language, so that, compared with that of the Kumarcen, they regard it as their own and a different dialect. The Kumarchen agree with this distinction. The similarity between the Tungus dialects of certain regions, for instance, northern Transbaikalia and western Manchuria (except RTM), is so great that the groups speaking them can easily understand one another; but the difference, for instance, between the Barguzin dialect and the Khingan dialect is so great that the groups speaking them can hardly understand one another. The dissimilarity between the dialects is still greater when one compares the Anadyr dialects recorded by Baron Maydell (in A. Schiefner's publication) or the Lam. and Turn, dialects with the Transbaikalian Mankova dialect. Yet some Northern Tungus dialects have been so greatly influenced by other languages that, for instance, the Orochi and especially the Goldi, they preserve only traces of their Northern Tungus characters; yet we know that many Tungus groups have changed their languages and have adopted the Yakut, Russian, Mongol, Chinese, and other languages. In spite of this situation, by eliminating the various foreign elements, one may trace a large group of dialects back to their original linguistical complex, which appears to be that of the Northern Tungus, who, with rare exceptions, call themselves evenki; and another group of dialects may be traced back to the Southern Tungus as their original 1 nguistical complex. Yet the similarity between the Northern and the Southern groups is so great that after the elimination of the elements characteristic of these two groups, there remains a large number of common elements which may be regarded as pra-Tungus; i.e., as existing prior to the appearance of the Northern and Southern modifications. As to which elements have been lost by one of these two groups, and which elements have been appropriated (e.g., the system of suffixes) by them after the separation, is difficult to say, for we have no evidences as to the pra-Tungus language. The process in the formation of new dialects may now be observed amongst the Northern Tungus groups. This process at the present time is chiefly due to the various non-Tungus influences. If the group is isolated from other Tungus groups, and if it is in contact with some non-Tungus groups, it is very likely to be opposed to other groups under other and distinct influences. Under this condition, the formation of a new dialect is very likely, but it is not certain, for the language is sometimes preserved in spite of the essential changes in the cultural complex, as, for example, the economical system, «religion,» etc. This is the case, for instance, with the settled groups of Birarchen and the hunter-horse-breeder Kumarchen. Cases in which the complex is not changed, but the language is changed, are also known, as amongst the Mankova Tungus, who preserve their original language, and the Mongolized Tungus of Urulga, who now speak a «Mongol» dialect, but all of them preserve the same cultural complex. Thus, in the case of the Tungus the language and other ethnographical characteristics do not always correlate, as is also true of the anthropological types and various elements of culture. The latter may be more or less preserved, while the language may be lost altogether, and vice versa. The process of formation of dialects, ethnical units, cultural complexes, and anthropological amalgamations and crossed groups may proceed more or less independently one from another. I say more or less, for there is a certain, although very slight, correlation between the elements forming an ethnical complex.
For the convenience of my readers I shall now give a list of abbreviations which I shall use throughout the following pages:
A. The Northern Tungus Dialects According to the Present Author
Amur — the dialect of the Reindeer Tungus living in the eastern part of the Amur Government, also spoken by a Tungus group in Sakhalin Island
Barg. — the dialect of the Reindeer Tungus of Transbaikalia living in the Barguzin taiga
Bir. — the dialect of the Tungus of Manchuria living in the regions lying between the Amur, the Sungari, and the left tributaries of the Nonni (Biranchen)
Borz. — the dialect spoken by the Nomad Tungus living in south-eastern Transbaikalia, in the basin of the Borza
Khin. — the dialect of the Tungus of Manchuria (Mongolia) living in the Khingan Mountains, chiefly in Hulun Buir
Kum. — the dialect of the Tungus of Manchuria living in the basin of the Kumara, Panga, Albazixa, and Upper Nonni (Kumarchen)
Mank. —the dialect of the Nomad Tungus living in south-eastern Transbaikalia, in the Mankova region (volost)
Ner. — the dialect of the Reindeer Tungus (partly settled) living in the Nerchinsk taiga
RTM. — the dialect of the Reindeer Tungus of Manchuria living in the basin of the Bystraia and Albazixa, and also the Upper Kumara
B. The Northern Tungus Dialects, According to Unpublished Material Gathered by Other Investigators
Enis. — the dialects of the Enissy Tungus gathered by the missionaries, brought by I. P. Tolmac'ev, and put at the author's disposal by W. L. Kotwicz
Lam. — the dialect of the Reindeer Tungus of the Yakutsk Government included into Lamunxinskii rod (cf. S. K. Patkanov, «Essay on the Geographical and Statistical Distribution of the Tungus»), gathered and put at the author's disposal by P. V. Olenin
Turn. — the dialect of the group mentioned above included in the Tumunxanskii rod; see Lam.
C. The Northern Tungus Dialects Known from Other Publications
Ang. Tit. — the dialect of the Reindeer Tungus of the Angara region in northern Transbaikalia, according to E. I. Titov
Irk. Tit. — the dialect of the Tungus of Irkutsk Government, according to E. I. Titov
Kal. Tit. —the dialect of the Reindeer Tungus living in the basin of the Kalar and Kalakan rivers, according to E. I. Titov
Mank. Castr. — see Mank. above, according to M. A. Castren
Neg. Sch. — the dialect of the Negidals living in the basin of the Amgun, according to P. P. Schmidt
Nomad Barg. Pop. — the dialect of the Nomad Tungus of the Barguzin district, according to N. N. Poppe
Orochi — the dialect of the Orochi living in the Maritime Government, according to S. Leontovich, V. P. Margaritov, and P. P. Schmidt
Solon (Ivan.) — the dialect spoken by the Solons of Manchuria, according to A. O. Ivanovskii
D. The Southern Tungus Languages and Dialects
Manchu Sp. — the Manchu language as it is spoken by the Manchus of the Aigun district in Manchuria, partly in Peiping and in Sinkiang province, according to the author
Manchu — writ. according to I. Zaxarov and partly according to the author
Nuichen — according to W. Grube
Goldi (Grube) (Sch.) — according to W. Grube and P. P. Schmidt
E. The Non-Tungus Languages
Mongol (Podg.) (Rud.) — according to I. A. Podgorbunskil and A. D. Rudnev
Buriat (Podg.) (Rud.) — according to P. A. Podgorbunskil and A. D. Rudnev
Dahur (Ivan.) — according to A. O. Ivanovskii
Dahur — according to the author
Yakut (Pek.) — according to E. K. Pekarskil
Gilak — according to W. Grube
Various abbreviations as they are found in A. Sauvageot's work.
70. I have felt it myself fairly strong when I was translating Manchu epic poems, shamanistic and other «religious» texts, i.e. the material essentially Manchu. It is easier, of course, with the Manchu translations from Chinese, where the meaning of Manchu words may be found from the Chinese text, but these «meanings» will be Chinese and not Manchu. The chance of correcting these dictionaries is decreasing every year, for the Manchu language is becoming a «dead language,» even in the remote regions where it was spoken some twenty years ago. True, it is still preserved amongst some Northern Tungus and Dahurs, but one must remember when it passes through the Northern Tungus or Dahur complex, it is subject to modifications. So, practically, one must be very careful in using Manchu material.
71. I have already pointed out this curious case (cf. «Social Organization of the Northern Tungus»), illustrating in which degree the author's complex if uncontrolled by criticism may affect the outcome of his work. It may be stated without exaggeration that this dictionary may suffice to restore the author's complex. Of course, to do so one must operate with the Tungus words translated with specific meanings or interpretations.