¹ Edited by W. Bang, A. von Gabain, and G. Rachmati, in Sitzungsberichte der Akademie der Wissenschaften 20.90–192 (Berlin, 1934).

² From Iran. spur-xān ‘house of perfection’, see Gauthiot apud É. Chavannes and P. Pelliot, Traité Manichéen 132.

³ C. Brockelmann, Mitteltürkischer Wortschatz nach Maḥmūd al-Kāšγarī's Dīvān Luγāt at-Turk, Bibliotheca Orientalis Hungarica I, Budapest and Leipzig, 1928.

⁴ W. W. Radloff, Versuch eines Wörterbuches der Türk-Dialecte, 4 vols., St. Petersburg, 1893–1911; henceforth quoted as WB.

⁵ Ch. de Harlez, Manuel de la langue Mandchoue 211 f. (Paris 1884).

⁶ H. C. von der Gabelentz, ZDMG 3.2.178 (1864).

⁷ L.c. 89.

⁸ According to B. Karlgren, Grammata Serica 296 (Stockholm, 1940), No. 661 f–j, < Ancient Chinese, and Archaic Chinese c'i̯әm 'to lie down, sleep; ... back apartment of an ancestral temple'; i.e. this word was not borrowed into Turkic before Ancient Chinese final -m changed to -n in the Northern Chinese dialects, in other words, not before some time after the composition of the C'iä-Jün Dictionary (601 a.d.). The replacement of Chinese c'-, lacking in Altaic, by s- is regular.

⁹ Kāšγarī has an indistinctly written word which Brockelmann 186 transcribes as sut(?b?)-ran ‘something tall like a lighthouse’; it is said to belong to the language of the Γuzz (= Oγuz). This word is no doubt to be read (or written) subran, and represents the Γuzzic development of Ujγur suburγan ‘stūpa’ (likewise listed by Kāšγarī with the translation ‘pagan burial place’; see Brockelmann 184).

Ujγur, Kāšγarī suburγan and its Γuzzic equivalent subran have a good Turkic etymology: suburγan is the Ujγur and Čaγataj form suby- (or: subu-)-r-γan; while the Central-Asiatic languages show the old form, Γuzz shows omission of the γ initial of suffixes, and furthermore, disappearance of the middle syllable. The basis suby-/subu- is represented by several derivatives in the southwestern group (Oγuz group). WB IV lists Osman sübü < sübü-g ‘round and pointed; oblong and pointed’ (a reference to söbü and söbö is added; but either is lacking in its proper place); Qrym süwrü ‘stinging’ (WB gives a wrong translation); Osman sivri-, sivri ‘to be pointed; pointed, sharp’; Ałtaj, Teleut, Baraba sǖrü ‘pointed’ (WB IV.815). The forms süwrü, sivri, and sǖrü go back to *sibi-r-ig or *suby-r-yγ. The original meaning of the stem is ‘tall and pointed’. Thus, Ujγur suburγan, Γuzz subran, actually a loan-word from Iranian spur-xān ‘stūpa’, has apparently been subjected to popular etymology.

The word is also found in Mongol; there, suburγa(n) ‘tumulus, tour ou pyramide sépulcrale, pagode’ (Kovalevskij, Dictionnaire Mongol-Russe-Français, 1392) serves as a translation of Sanskrit stūpa and čaitja. In Mandžu the word is attested as subargan and subarxan (I. Zakharov, Połnyj Maňčžursko-Russkij Słovaŕ 638 f.; v.d. Gabelentz, l.c. 182). The word was borrowed from the Ujγurs by the Mongols and from them by the Mandžus.

¹⁰ S. A. Kozin, Sokrovennoje Skazanije Jüań Čao Bi Šy 1.591 (Moskva, Leningrad, 1941).

¹¹ See E. Berneker, Slavisches Etymologisches Wörterbuch 291.

¹² A. A. Šakhmatov, Očerk drevnejševo perioda istorii russkovo jazyka §§62, 71 note, 308–10, and N. Trubeckoj, in Sbornik Miletic 269–70 (Sofia, 1933).

¹³ Kovalevskij 1.449.

¹⁴ E. Berneker 398.

¹⁵ G. M. Vasilevič, Evenkijsko-Russkij dialektołogičeskij słovaŕ 80 (Leningrad, 1934).

¹⁶ Je. I. Titov, Tungussko-Russkij słovaŕ 123 (Irkutsk, 1926).

¹⁷ S. A. Kozin 1.595.

¹⁸ Ibid. 1.614.

¹⁹ Ibid. 1.616.

²⁰ Ibid. 1.591.

²¹ P. A. Mostaert, CICM., Textes Oraux Ordos, Monumenta Serica, I (Peip'ing, 1937).

²² G. M. Vasilevič 78.

²³ In an oral communication to me, Roman Jakobson maintained, as opposed to Berneker, that a derivation of gaj, haj, γaj from *gojiti, *gojъ (Berneker 319), is possible, considering gaj, haj, γaj as a lengthening degree of *gojъ (type krojiti-krajъ, etc.). Berneker declined to recognize a connection between *gojiti, *gojъ and gaj, haj because of the great semasiological discrepancies (l.c. 292). If Jakobson's derivation is correct—which remains to be proved—, we may here deal with one of the few hitherto known cases of genuine IE–Altaic relationship.

²⁴ A. Brückner, KZ 45.24 ff.; more recently, the new four-volume Tołkovyj Słovaŕ Russkovo Jazyka 4.843.

²⁵ WB IV.556.

²⁶ WB IV.1805.

²⁷ Menges, Archiv Orientální 11.7 ff.

²⁸ S. A. Kozin 609.

²⁹ Genuine Turkic is Kāšγarī's tünäk, i.e. tün-ä-k, known also from Ujγur (WB III.1549 f.); its original meaning is 'place where it is (dark like) night (tün)'. A semantic parallel is afforded by OCS. tъmъnića ‘prison’ from tъmъnyjъ ‘dark’.

³⁰ Brockelmann errs in his Kāšγarī edition (224) when he calls Radloff's opinion erroneous.

³¹ Lacking in I.I. Sreznevskij's Materials for an Old-Russian Dictionary; quoted without etymology in the new Tołkovyj Słovaŕ Russkovo Jazyka; see furthermore F. Miklosich, Türkische Elemente ..., Nachtrag II.35 (Vienna, 1888 ff.).

³² WB IV.636. According to Kāšγarī (Brockelmann 179) ‘lacquer, color’.

³³ Kovalevskij 1532 f.

³⁴ F. Steingass, Persian-English Dictionary 667.

³⁵ Ibid. 671.

³⁶ WB IV.591. Here, ụ designates an intermeditae vowel between o and u.

³⁷ WB IV.598. The u̿ designates an intermediate vowel between ö and ü.

³⁸ L.c. 676.

³⁹ Türkische Elemente ..., Nachtrag II.35.

⁴⁰ WB IV.519.

⁴¹ WB IV.518 ff.

⁴² Brockelmann 117.

⁴³ Ibid. 105.

⁴⁴ WB II.1100 ff.

⁴⁵ Kāšγarī, ed. Brockelmann, 105.

⁴⁶ WB II.1102 ff.

⁴⁷ Brockelmann 94.

⁴⁸ I. I. Sreznevskij, Mat. I. 1209. The word is not found in other Slavonic languages, with the exception of Polish, in which kirka is a borrowing from Russian. In Polish, it has the same meaning as in Russian, but is especially used in the sense of ‘część ostra mlotka mularskiego, do przecinania cegły’ (‘sharp edge of a mason's hammer [pick-hammer], for hewing bricks‘), Słownik Języka Polskiego, ułož. pod red. Jana Karłowicza, Adama Kryńskiego i Władysława Niedźwiedzkiego II.341 (Warszawa 1902).—Kirka is lacking in A. Brückner's Słownik Etymologiczny Języka Polskiego.

⁴⁹ Šakhmatov §§455–60.

⁵⁰ Brockelmann 105.

⁵¹ WB II.1107.

⁵² WB II.1365.

⁵³ WB II.1366.

⁵⁴ Kovalevskij 2515, 2552.

⁵⁵ L.c. 501.

⁵⁶ Vład. Daλ, Tołkovyj Słovaŕ Živovo Velikorusskovo Jazyka II.273, refers to Tatar and Persian origin. The statement in the new Tołkovyj Słovaŕ Russkovo Jazyka II.1358, referring to Pers. kärpič as the source, is incorrect since the Persian word has been borrowed from Turkic.