There is no universally accepted system for spelling the Mongolian words and terms of the age of empire and after and, in some cases, there is even a lack of agreement upon the proper forms themselves. A large part of the problem is due to the fact that we have so few documents in Mongolian from the period in question and must rely on reconstructions made from spellings of the words in Islamic or other foreign sources. They often have long manuscript traditions resulting in many erroneous forms transmitted to the present day. These reconstructions are particularly problematic when Chinese transcriptions are our only source since, although, thanks to early Chinese use of printing, the evidence is more reliable, Chinese (q.v.) is too simple phonologically to represent the sounds of many Mongolian, Turkic, and other foreign words found in the sources of the time. Most Mongolian and, particularly, Turkic finals simply do not exist in Chinese and representing the complex diphthongs characteristic of most Central Asian Turkic languages of the past and of the present is quite beyond Chinese. There is the added problem of having to take into consideration the specific Chinese pronunciation of the era (so-called Old Mandarin) and possible dialectical variants (the speech of Peking, barely in the Old Mandarin zone in the 13th century, for example) that we may not understand today.
Thus, recovering Mongolian words and terms from Chinese sources is an uncertain enterprise at best, and only possible at all due to our ability to make comparison between various versions of what is apparently the same word, the drafting of modern forms to understand the past, and the existence of a few texts in Chinese intended for the serious study of the Mongolian by Ming Dynasty (q.v.) translators, for example, and utilizing a more scientific transcription system with a better representation of finals and diphthongs. The longest and most complete of these reproduces the entire text of the Secret History of the Mongols (q.v.), which has been largely reconstructed from this Ming version. Yet even this text has many problems due in large part to the fact that it is possible to misread forms in Uighur-script Mongolian (see Mongolian Script) because of the ambiguity of the script itself, and the Chinese transcribers of the Secret History, who had a now-lost Uighur-script version of the text before them, have done just this.
In any case, the existence of such fairly reliable texts is the exception rather than the rule. They are also exceptional in that they seem to represent a reasonably standard usage of the form of Mongolian considered the most important standard language of the time, Middle Mongolian (q.v.). This is the language in which Ming Dynasty informants recited the Uighur-script original text, which reflects other linguistic values entirely. More typical are forms that do not appear to reflect any rules or even standard usage. Some are outright strange. While much of this may reflect transmission problems or a simple mishearing of native Mongolian words by nonnatives, there is also the issue of other dialects than standard Middle Mongolian. There is also the possibility that the Mongolian spoken environment of the time may have been heavily Turkicized and the forms that have come down to us are mixed, linguistically, as a consequence.
All of this is to say that while some reconstructed forms are more reliable than others, and many are even likely to be entirely accurate, the potential for error and misunderstanding is great, all the more so because the detailed linguistic studies of the various court and other environments of the Mongol era are largely lacking, particularly for China. Users of this dictionary are thus warned about overly exact reconstructions where the available evidence simply does not allow it. Many are highly artificial and arbitrary, even if widely quoted in the literature out of lack of anything better in most cases.
My own approach is probably artificial and arbitrary too, but I have at least striven to be consistent in using a single authority in most cases. This is Igor de Rachewiltz’s Index to the Secret History of the Mongols, Bloomington, Indiana: Indiana University (Indiana University Publications, Uralic and Altaic Series, 121), 1972. The Secret History of the Mongols is not only our single largest treasury of Mongolian words and phrases from the era of Mongolian empire and its successor states, but the de Rachewiltz text is highly reliable in part because it has been computer-checked for consistency. In his Index he employs a modification of the transcription system originally employed by Paul Pelliot in his reconstruction of the text of the Secret History of the Mongols, but with fewer diacritics, making his spellings easily reproducible with even a minimal character set. Mongolian words occurring in the Index and used in this Historical Dictionary, with a few exceptions, mostly where another form has become established in the literature, are spelled as they occur there except that I have made a choice between alternative forms in some cases (using the form more common in the text). Other Mongolian words and terms of the period, those found in the texts but not existing in the Secret History of the Mongols, I have endeavored to spell in the same way, but not to the extent of making possibly arbitrary reconstructions. Note too that some Mongolian forms given below are Classical Mongolian forms, as they are spelled in the Uighur script, e.g., qaγan, not qa’an, or are the current standard Khalkha forms, e.g., khaan. For the convenience of the more specialized reader, Appendix B provides an index of all Middle Mongolian words and expressions used in the text as well as a few other forms.
All Chinese terms used below are transliterated in accordance with the Pinyin system, except that I have not changed bibliographical entries using the old Wade-Giles system. Chinese characters are universally provided for the first occurrence of a Chinese word, term, or name in a given entry or section, except for the names of dynasties for which see the table below on page xxvii. No attempt has been made to provide Old Mandarin forms, which are not that different from modern in any case. This is generally the convention in the field as well.
The Hepburn system is used for Japanese with macrons. Russian words are transliterated in accordance with the system employed in reports published by the Foreign Broadcast Information Service (except for a few names whose forms are well established in the literature), and the same practice is followed for any modern Turkic forms now written in the Cyrillic script. For older forms I follow G. Doerfer, Türkische und mongolische Elemente im Neupersischen, four volumes, Wiesbaden: Franz Steiner Verlag, 1963–1975, or some other indicated authority, if there is good reason to do so. In each case these forms have full diacritics. Persian and Arabic forms are spelled as in the Cambridge History of Iran with full diacritics.
Exceptions to the above rules include many forms that are firmly a part of popular usage, or whose spelling is firmly a part of some source tradition. Thus I write Mongol, and not Mongqol, except in phrases (yeke mongqol ulus, “the Great Mongol Patrimony,” i.e., the Mongol Empire), Bukhara instead of Bukhārā, Baghdad instead of Baghdād (but Baghdād-qatun), Ghazan, instead of an unfamiliar Qasan, or Ghazan, although I have resisted Genghis Khan and Ulan Bator, writing, respectively, Cinggis-qan and Ulaanbaatar. When well-established forms exist in the text of Marco Polo I have at least cross-referenced those that are the most well known. In all of this I have firmly followed the maxim of using forms that will be usable by the nonspecialist but that will still be reasonably solid.