É. Apor (ed.): Codex Cumanicus. Ed. by Géza Kuun with a Prolegomena to the Codex Cumanicus by Lajos Ligeti. (Budapest Oriental Reprints, Ser. B 1.)
L. Ligeti: Prolegomena to the Codex Cumanicus
30 L. I-IGETI Both views have their advocates, and both have ample arguments for and against. 3 9 When interpreting the Persian material of the Codex Cumanieus, both the errors of the copier (just as in the Coman part) and those arising from G. Kuun's edition must be taken into account. In this context, the latter are negligible, and even Teza ignored them in his review of the publication. Later editors could easily disregard Kuun's erroneous readings as the facsimile edition of the Codex was then available to them. Kuun's edition, however, eontains a section worthy of our attention. This is the Persian word-list (pp. 308 — 322), where the Persian words in Latin script are followed by the Arabic-Persian written forms of the classical Persian language, which Kuun intended as an explanation. Every so often they prove the priority of Kuun's correct interpretations. It is also noteworthy that relying upon Vullers' dictionary Kuun quoted the «vulgar» forms of Lit. Pers. in addition to the classical forms (e.g. ayn «ferrum» (besides ahin), aou «aqua» besidesaja).These data illustrate that the Persian material of the Codex was based on the living, and not on the standard written language. Variants coexisted, which reflect either inconstant features in a single dialect or differences in pronounciation of the several native informants. We must also be aware of the simple literals of mediaeval manuscripts also present in the Persian data of the Codex. Monchi-zadeh (M) and Bodrogligeti (B) pointed this out, too. One of the most common mistakes is the mixing of t and c. Such are largos for cargos «hare», t alamari for calamari «interpreter», noghitc for noghitt «pea», due for dut «smoke». Similar errors of the copier are: batran for barran «rain», aut for art «flour», etc. 3 9 According to G. Györfiy, Aulour du Codex Cumanieus, pp. 121 — 122, the Persian-C'oman words were dictated by a single interpreter with a knowledge of both languages. He points out that 363 equivalents of the 414 Latin entries are missing in both languages. From among the non-parallel lacunae, 13 Persian equivalents and 38 Coman ones are missing. If only the figures are considered, one might conclude that the interpreter had a better command of Persian than of Coman. However, Györfiy also stiesses that the Persian and Coman words frequently overlap. He selected seven concordant names of birds from the chapter «Nomina volucrum» (p. 109), which do not even follow each other. Such a comparison is misleading, since there are altogether 24 bird names (Kuun's 23 is a mistake), 20 of them to be found in Persian, and 23 in Coman (one Latin entry has neither a Persian nor a Coman equivalent; this was left out by Kuun). If we take the whole material into consideration, we find that only 8 words are identical in Persian and Coman, and only three of the 8 are of Turkish origin. This was not what Györffy aimed at. His objective was to prove that the Persian —Coman part could not have been written gradually testified to by the large number (which is not so large, alter all) of identical words in the two languages. His assertion is also weakened by the fact that tire list includes the international loan words of falconry. I am afraid this issue cannot be definitely settled on the basis of the extant manuscript, as it is a copy itself. Under such circumstances, naturally, the opposite of Györffy's hypothesis cannot be proved either.
