38. Cases with Initial and Intervocalic Labial Consonants
A. Sauvageot starts an analysis of the series of words and parallels found in other authors' publications and he increases them with his own parallels. The first three series (vide supra, p. 123), containing eighty-five cases, are closely connected with the problem of aspiration and bilabialization in Tungus. Since this problem in Tungus has nothing to do with the loss of the initial bilabial tenuis, the whole series might be dismissed from the analysis; but I will not do it, for parallels are brought forth to prove the common origin of words, a problem of certain interest by itself. Yet if we leave out of the discussion the problem of the loss of this consonant in all other languages except the Tungus language and temporarily agree with the creators of this hypothesis, then in the Tungus language, even the pra-Tungus language, these words might be received already without the initial consonant and afterwards pass through the process of aspiration and bilabialization [105]. Therefore I shall compare these series too.
After a minute analysis of the first eighty-five cases I have found that there are nineteen cases of Tungus parallels lacking; in six cases there are parallels only from Tungus and Samoyed, so that they have little value as evidence for the Ural-Altaic hypothesis; in twenty-three cases it was impossible to agree with the validity of the Tungus parallels; in twenty-four cases the parallels are present, but for various reasons are found to be invalid; and in thirteen cases the parallels are found to be worthy of further analysis and use.
So that, in so far as the Tungus language is concerned, in the problem of the Ural-Altaic hypothesis the outcome of the work done is not great. However, the cases rejected, perhaps, in some instances, may be saved when the parallels are completed from Tungus and other materials. In the cases to be discussed there are some met with which have been discussed twice under different numbers, so that the actual number of stems is below eighty-four.
The thirteen cases here analysed may be regarded as actually common words, but what their origin is and why they are common need special investigation which cannot be successful in all cases. Yet, when the common character of the words is established, I give some additional remarks when it seems to me desirable.
Case 2. G. Ramstedt's stem *pYjY, *pYYsYY, which means «the nest.» In fact, the Manchu feje has corresponding words in the Northern Tungus; e.g., uyi, ujin (Bir.), which in Tungus might be derived from the stem VjV or VgV; but whether uyi and feje is the same stem or not cannot be asserted, for uyi may be explained in a different manner. Yet the difficulty of this case is also that several other Tungus dialects possess other stems; namely, as shown below: (1) ur (Ur., Castr.) and or (Mank.) are borrowed from Buriat (Ur., Castr.); (2) omuk, omunin, xamun (Neg., Sch.), umuk (Ner.), omokon, omok (Tungus quoted by P. P. Schmidt), onto (Goldi, W. Grube); (3) (several other words found in more than one dialect, also «nest» of special description.) Omuk cannot be connected with feje nor pYjY, but it is the most numerous in Tungus. After all, can we assert that the Tungus language ought to be connected through Manchu feje with other languages? On the other hand, the Uralian languages have, instead of j, different sounds, as s, z,t,d, etc., so that the comparison is confined to the first initial p, which itself is a hypothesis.
Case 6. The stem *pYrY expresses the idea of «wheel,» «circle,» «round,» «turn,» etc [106]. Forombi (Manchu Writ.)—«to turn,» is found, as well as a series of words like for on—«the turn,» etc., forgon—«the rotation,» «cycle,» etc., and various Northern Tungus words, as orol (Ner., Bir.), orul, arul (Bir.), orolko (Ner.)—«to turn oneself»; oroli (Khin.) (Ur., Castr), horoli (Ur., Castr.)—«the circle.» This may be regarded as an established fact, but the stem is oro, and not poro, so that what is actually compared is VrV or r. However, the Mongol stem urb has a different meaning [107].
Case 9. The stem includes the initial labial consonant and r with a broad meaning «das obere,» «die hohe,» «kopf,» «spitze,» «ende» in Samoyed and there are compared foron (Manchu Writ.)—«the top,» e.g., «the top of the head,» also «the top of a mountain» (P. P. Schmidt), etc. This is connected with horon, oron, of various Northern Tungus dialects, namely, oron (Bir., Kum., Khin., RTM, Ner., Barg., Mank.), horon (RTM) (Ur., Castr.), oro (Khin.) also Poro (Olcha, Sch.) compared by all authors with Mongol oroi and horoi (P. Pelliot). Thus the stem is oro. However, further steps are made and Mongol oroi is compared with orgil, etc., of the same meaning. Moreover, in Tungus there is another stem, namely, ojo (Manchu Writ.) (Bir., Kum., RTM, Ner.) (Mank. Castr.), pojo (Olcha, Sch.), which has its parallels in Bargut ojjo (Rudnev). The case is not simple if we do not regard oro and ojo as one and the same stem. Such a possibility does exist, but the equation must be proved. A. Sauvageot concludes his parallels by a remark,—«L'absence de toute etymologie fgr. sure ne permet pas de decider si les mots considers remontent a une origine ouralo-altaique, d'autant plus qu'en turk, il ne nous a pas ete possible d'etablir une equation equivalente» (op. cit., p. 14). As compared with other conclusions, this one is unusually wise.
Case 20. The stem *pYjY (Uralian) and *fYlY (Mongol and Tungus) means «trembling poplar.» The Tungus words which may be compared are fulxa (Manchu Writ.), ulan (Ner.), uloka (Bir.), huldan (Neg., Sch.), hulugdan (RTM), xulu (Oroc., Sch.), polo (Goldi, W. Grube), oleson (Ur., Castr.). The latter may be compared with the Mongol series — ubaso (Buriat, Castr.), uljasun, ulisun, ulasu, etc. (Mongol). Since -xa, -n, -dan, -gdan, also -so, -sun, -son are probably suffixes, the stem is ulV, or even VIV in Tungus and Mongol. The question as to what the relationship between pYjY and VlV is answered in a positive sense by A. Sauvageot in the reason of the two hypotheses; namely, the original one of the loss of p and the new one of l = j. Indeed, as to the last one in principle there is nothing impossible in it, for, as is well known, j may appear in any combination as one of the moments of palatalization. A. Sauvageot gives another similar instance in Case 21, where the Uralian j is supposed to correspond to the Tungus d, which is not also a fact but a hypothesis [108].
Case 32. The stem is *pYy and means «souffler.» In the Uralian languages y alternates with b, v, w, u, and zero. Amongst the Tungus parallels are found fuka (Manchu Writ.) and fukanambi (Manchu Writ.), which semantically seem to be connected with the idea of «round,» «circle » and as, perhaps a secondary meaning, «bubble,» also perhaps «protrude» and not «to Mow.» The Northern Tungus is also figured in the form uvu, already discussed by me [109]. A. Sauvageot has omitted the inconvenient parallel brought forth by P. P. Schmidt — fulgjembi (Manchu Writ.) —«to blow,» connected by me with another stem pul, even perhaps ul. The Tungus words have the intervocalic v, which does not alternate with y and l if the stem ul is also a Tungus stem, so that here we again have two hypotheses — the loss of p and v. But the Turk words do not seem to be convincing to A. Sauvageot. On this occasion he expresses his opinion as to the grammarians who have a tendency to overestimate the importance of onomatopoetic origin of words. This time, however, to every one's mind comes the idea of such an origin of many European words. This idea does not seem to be so repugnant, as it is to A. Sauvageot, that the parallels from the so-called Ural-Altaic languages come to the same line.
Case 46. The stem is supposed to be jVpV and means «wet» followed by a series of other equivalents like «macerate,» «to wash,» «le lavoir,» «marshy place,» etc. Since, as shown by A. Sauvageot, j~v and j~l, also j~d, the parallels are easily found. From Tungus are given only two parallels, namely, cekpa (Goldi, which needs to be checked up, for it is not found in other Tungus dialects) and deptom (Ur. Castr.), which is borrowed from Mongol, so that this case could be easily dismissed. However, in Tungus there is a stem which would not be omitted by A. Sauvageot, so I will use it too. It is ulapkun (Bir., Kum., Khin.) (W. T. C. Schiefner), utapkun (Ner., Khin.), ulapcha (as part. perf. from ulap) (Ner., Bir., Kum., Khin.). The serious objection is the initial u, but it may be easily dismissed under some pretext and under similar circumstances..In fact, the case is not so simple as that. We have ula (Ang., Samog. Tit.) — «to make wet»; ulikcha (Turn.) — «wet»; and ulgambi (Manchu Writ.), wulgambe (Manchu Sp.) — «to wet,» «to soak.» In the first series containing p, we have it as a suffix of transitive «verbal» forms, which in some dialects is not needed, for the «transitiveness» is understood without a special suffix [110]. Yet -kun, -la, and probably the Manchu -ga are suffixes too. So the stem is ulV and it has nothing to do with jVpV. The stem deptI dismiss as one recently borrowed from Mongol. The Goldi word cekpa must be first checked up [111].
Case 48. The stem is *kYp with the meaning «l'ecorce,» «la croute,» and afterwards «la peau.» A. Sauvageot has lost the good opportunity of comparing Tungus kiwa, etc., already discussed by me on the occasion of Case I (vide supra, p. 126) and which in all Tungus dialects and Gilak means «birch-bark.» However, it looks like a local phenomenon naturally correlated with the geographical area of betula alba.
Case 54. The stem is *kYp, with the meaning -«plat» and with the further extension of «handflaeche,» «fussblatt,» «niedrig,» etc. The Tungus hapse (Olcha), hapse (Goldi), about which one must have a certain caution, for A is a mere aspiration, are also found; they are translated (W. Grube) — «flach ansteigend.» It is not mentioned by P. P. Schmidt. In several Tungus dialects we have afsa, hafsa, etc. — «the low standing box» (cf. Mongol apsa). There is another word in Tungus which comes near to kYp and the idea of «plat»; namely, kaptaka (Bir.) — «the low flat box,» but kaptaka is made of wooden, planks which are in the Tungus dialects kaptasun (Bir., Khin.) (Khin. is from Bargut), kaptayon (Bir., cf. Dahur kaptayon), kaptayin (RTM from Yakut xaptasyn), kaptagun (Neg., Sch.), etc., which is not perhaps Tungus but probably Mongol, where it receives a great variety of derivatives from the stem xabta (cf. also Yakut derivatives), while in Tungus they are limited. However, there is one more stem in Tungus which might serve A. Sauvageot; namely, kapaxun (Manchu Writ.) — «flat,» e.g., «nose»; kaparambi (Manchu Writ.) — «to become flat»; also kapahi (Orochi, Sch.) kapasu (Goldi, Sch.) — «the plank,» «the board.» With these additions it may be recognized that we have the stem kVp; but whether the Tungus stem is of Tungus origin or Mongol, and whether kapa, met with in Manchu, Goldi, and Orochi, is the same stem as kapta or not, are questions to be answered by further investigations. However, the Tungus dialects possess other stems for «flat.» In this case, as in other similar cases, the parallelism of different stems with the same meaning is an extremely complex matter.
Case 58. The stem is *tYp, which means «clou.» In Tungus a stem tipk, which originally meant «the nail,» and probably «wooden nail» is met with. So we have tipka (Bir., Khin.) — «to fasten» (e.g., the cover of a box with the nails); tipkasun (Khin.), tipkon (Neg., Sch.), typkon (Goldi, Sch.), tipa (Olcha, Sch.) — «the nail»; tipkocin (Lam.) — «the wooden nail» (for fastening the tent) [112]. Besides this series there is another series, still more numerous than the former one, of words designating «nail,» but it should be connected with Mongol. These are: kadasin (Ner.), kadasun (Khin.), kadahun (Mank., Ur., Castr.), corresponding to kadahan (Buriat, Castr.), kadasun (Xalxa, Podg), xadayan (Buriat, Podg.), connected with gada, kada || kadayasun (Mongol, Rudnev), whence kada (Manchu Writ.) — «the nail»; kadambi (Manchu Writ.) — «to fasten with the nail.» There is one more word in RTM which is interesting — toyohol (RTM)— «wooden nails» (for fastening the board of the birch-bark canoe) connected with toyoso (Yakut, Pek.) — «the nail,» «the wooden nail,» «stake,» etc. These three series of words are interesting, for all of them originally meant «the wooden nail,» «stake,» «pile,» etc. All three are now met with in Tungus. A. Sauvageot has compared his series with Mongol tab —«tete d'un clou.» A. D. Rudnev gives nearly the same meaning tap (Gorlos) || tab. However, various meanings of tap may exist; e.g., with the semantic emphasis on «head» and not «nail.» Anyhow, in Tungus tipk k cannot be dismissed and the Mongol vowel is a and not i.
Case 65. The stem is *pYrY || bYrY, with a wide meaning; namely, «gravier,» «sable,» «argile,» «poussiere, «suillure,» «cendre,» etc. The Tungus parallels shown are buraki (Manchu Writ.) — «the dust»; berten (Manchu Writ.) — «the stain,» [113] none of which have corresponding words in other Tungus dialects. A. Sauvageot also gives buru — «flint.» In fact, buru (Ur. Castr.) (W. Okhot.) (Barg., Ner.), bur (Bir., Kum.) — «flint,» perhaps, in general, «stone which may be used for production of sparks,» but its connexion with «poussiere,» etc., is rather artificial. Buraki being isolated in Tungus finds, perhaps, its best etymology in Mongol. In Tungus the words for the above-indicated semantic groups («poussiere,» etc.) have nothing to do with *pYrY and bYrY.
Case 66. The stem is *pYr (and bYr), which is given a number of semantic equivalents, e.g., «storm,» «steam,» «fog,» «rain,» «snow,» etc. There are good parallels from Manchu Writ. e.g., burga, burgan — «dunst,» «dampf,» [114] and burgashambi — «to drift»; e.g., «dust,» «clouds,» «smoke,» etc., and some other words of the same stem. However, in other Tungus dialects I do not know it. A. Sauvageot points out that the words for the designation of these phenomena were subject to borrowing from one group to another. For those who want to find «common words» with European languages, one may quote a long series of Russian words like «bur'a,» «buran,» «purga,» «uragan,» etc., which are of various origin, also «bourrasque,» «brouillard,» «hurricane,» and many others. It is evident that here the question is more complex than the simple case of «common words.»
Case 72. The stem is *pYng and *bYk, which means «excroissance,» «noeud.» As to the Tungus words quoted, we have different initial words; namely, in Manchu Writ, stem bukt with the meaning of «curved,» «bent,» «hunch,» etc., with other derivatives like mukch,, well known in several other Tungus dialects, and «the wart,» «spunk» — boyono (Kum., Khin.), boyoto (Bir.), moyo (Bir.); while bokoto (RTM), (Neg., Sch.), bogoto (Neg., Sch.), boxto (Goldi), bokto (Ner.), etc. — «the cone» (of a coniferous tree). Yet there is a series of words, like bukachan (Bir., Kum.)—«a small island,» bukachan (Ner.) — «a small hill,» «hillock,» «heap»; bukchan (Lam.) — «a small island»; in all these words we have thus the idea of something protruding, coming out. It ought to be noted that the «spunk,» «cone,» and even «hillock» in their translations into the local (Siberian) Russian spoken by the natives is usually rendered as shishka. I do admit that such a misunderstanding was possible in many a case. One may close one's eyes to the phonetic differences and semantic discrepancies only when one leaves oneself to the charm of finding common words. They do, indeed, need careful and accurate analysis.
Case 83. The hypothetic Tungus stem *bYlk, lk of which is admitted to be Samoyed l, means «gosier.» In Tungus the usual form is bilga, bilya, whence Goldi bilzha. The Manchu Writ. bilxa (bilya of Manchu Sp.). However, in Samoyed the stem is fal.
Case 84. The stems are *bYtY, bYUtYU, with the meaning «conduire,» «mener.» The case is said to be «le plus singulier.» The Tungus parallels are fude (Manchu Writ.) (let us remark that they are accepted by G. Ramstedt and P. Pelliot), which have a very special meaning connected with the customs of meeting persons, sending presents, dowry, etc., also expedition of official documents, etc., all of which, with a certain effort, may be brought to the idea of «conduire.» Another word is udd (Ur. Castr.) — «to escort,» already connected by M. A. CastrSn with Buriat udenep, and compared by P. Pelliot with hiida (mediaev. Mongol). The Manchu word seems to be a Mongol word used for ceremonial and office occasions which is frequently met with in Manchu. Whether the Mongol stem udV or ud may be brought to the hypothetic stem bYtY or not will be shown by further investigations in Mongol; but before it is shown, the stem must be considered as ud(V) and not bYtY. The parallels from Turk are lacking. It may also be noted that specialists in the Uralian languages have already connected this stem with the Indo-European *wedh. Theoretically speaking, such instances must be rather frequent.
* * *
The analysis of the above-mentioned cases with the initial and intervocalic labial consonants has shown that out of thirteen cases which could be analysed as instances where the possibility, both from the phonetic and semantic points of view, of common origin or at least common character may be suspected. However, in Cases 2, 6, 9, 20, 32, and 84 the hypothesis of the loss of the initial p greatly reduces the validity of parallels from Tungus. Some other hypotheses as to the alternation of consonants still more reduce the validity of these parallels. In some cases, as, for instance, Cases 6 and 9, the semantic extension is so broad that the value of parallels becomes very doubtful. Practically speaking, in all cases except Case 84, there are at least two hypotheses brought forth. However, even under these conditions A. Sauvageot could not find parallels for all «Ural-Altaic» languages; e. g., in Case 9 the Turk and Finn parallels are lacking, in Case 32 the Mongol parallels are lacking, and in Case 84 the Turk parallels are lacking.
If we agree with A. Sauvageot as to the validity of his hypotheses and ignore the fact of data in some cases, then we may have the following series:
Case 2, Manchu word feje — «the nest»
Case 6, Tungus oro — «the idea of circle»
Case 9, Tungus oro — «the top,» «summit»; Turk and Finn parallels lacking
Case 20, Tungus ula — the name of a tree Case 32, probably onomatopoetic; Mongol parallels lacking Case 84, Mongol uda; Tungus parallels lacking
Cases 46, 48, 54, 65, 72, 58, and 83 are free from the hypothesis of the loss of the initial p, but hypothetically established alternations are met with in Cases 46, 72, and 83; the semantic hypotheses are met with in Cases 46, 65, and 66; in the following cases the parallels are lacking: 72 (Turk), 58, and 83 (where only Tungus and Samoyed are found); in Cases 48 and 54 the Tungus is given by me. If we agree with all the hypotheses proposed by A. Sauvageot, we shall have:
Case 46, Mongol dep — «to wet»
Case 48, Tungus kiva — «birch-bark»
Case 54, uncertain Tungus kapta — «the plank»
Case 65, uncertain Manchu buraki — «dust»
Case 66, Manchu burga, consideied by A. Sauvageot as a loanword
Case 72, Tungus boko — «the idea of protruding»
Case 58, Tungus tipk — «the (wooden) nail»; only Samoyed parallels
Case 83, Tungus bilg — «the oesophagus»; only Samoyed (doubtful) parallels
So if we agree with all A. Sauvageot's hypotheses, but exclude all cases which are not Tungus and met with only «like Tungus» words (stems) and all cases where there are no parallels in other languages, we shall receive a short list of five stems for the following words and ideas, — «circle,» «top,» «name of a tree,» «birch-bark,» and «protruding» which may be regarded as «common» with the Ural-Altaic languages. The conclusion which may be drawn from these facts, in so far as the Tungus language is concerned, is evident, on which I could drop the analysis, but since these eighty-five cases represent only a part of all the cases I shall here analyse other cases too.
105. Indeed, such a hypothesis is not needed if we do not postulate the common origin of the Ural-Altaic languages.
106. This case has already been rejected (vide Sec. 36). As a matter of fact, it contains more than one stem to be discussed and for this reason I have dissected it into two parts, one of which has been rejected and another here analysed as good for comparison.
107. It means «to change oneself,» etc., and hurba (P. Pelliot) —«retourner,» «renverser.» Semantically, it may be compared with ubal (Manchu Writ.), obal, obol (Kum. Bir.)—«to transform oneself.» The latter is perhaps found in Buriat (Podg.) in an «aspirated» form xubila, borrowed in Ner. as kuvil (go), also known with the stabilized glottal k in Mongol and Turk. In this modification it is found in Manchu Writ, kubulimbi by the side of the bilabialized form u ubalamb'e of Manchu Sp. with a slightly different shade of meaning. This case is interesting, for it shows the traces of diffusion of the idea of transformation practised by the shamans, amongst the Tungus. In this connexion it may be pointed cut that tlie Tungus of Manchuria have obal, etc., from Manchu Writ, ubal, and the Tungus of Tran&oaikalia have huvil from Buriat — xubila.
108. The question is about fodo, fodoxo (Manchu Writ.) = vda, udu, etc. (Mongol dial.) — «tire willow,» «salix,» which, in so far as I know, is not known in the Northern Tungus dialects using different words for the same plant. In Manchu, fodoxo refers to a specified kind of salix unknown in the territory of the Northern Tungus. Furthermore, A. Sauvageot agrees with the parallels given by P. Pelliot; namely, Mongol and Euriat hicasun, isaghan, etc.—«the willow,» through Tuik usak, etc.— «the poplar.» Leaving aside the question whether ud and ish, ich are of the same origin, as well as the question of semantic generalization (the willow and poplar are certainly plants), I want to point out that A. Sauvageot admits the possibility of d=sh=c, so that the series of alternations is extended ad infinitum s, z, h, x, k, etc. What remains as a fact is d in Zyrian and Votiak and Mongol, for Manchu fodoxo seems to be a non-Tungus stem and the initial p is a hypothesis.
109. Cf. «Notes on the Bilabialization and Aspiration in the Tungus Languages.»
110. The suffix of the transitive form of verbs is va (e, i,) (V), which changes into a tenuis occlusive bilabial through the assimilation c. It is a very common suffix, indeed.
111. In cekpa if it it a Tungus correctly recorded word; -pa may also be a suffix.
112. The Goldi parallels from W. Grube's work, as tukpa, tukpe, perhaps need some correcting, as seems to be done by P. P. Schmidt.
113. There ought also to be added borton, bortonombi—«dirt,'' to dirt which probably must not be connected with buraki, but perhaps compared with borzhin (Mongol, Rudnev) — «dirt,» «dust.»
114. This word is lacking in I. Zaxarov's dictionary. The translation does not seem to be absolutely exact; cf. burgashambi. This word looks like that well known in Asia for «snowstrom» and «dust storm.»