Reviewed by Mikael Thompson, Ulaanbaatar, Mongolia

Ordos is the dialect of Mongolian spoken in the Ordos Loop of the Yellow River (containing parts of Inner Mongolia, Ningxia, Shaanxi, and Gansu). This book, published in 1903, was the first major study of Ordos. A young French translator of Chinese in Beijing, Georges Soulié, who later produced several works in Chinese studies (including the first major Western study of Chinese acupuncture), studied Ordos under the direction of Monsignor Alfons Bermijn, after most of the language materials collected by his mission had been destroyed in the Boxer Rebellion (iii–vi). Ironically, however, the institutions that allowed him to produce this book eventually rendered it obsolete. The ablest Western Mongolist of the twentieth century, Father Antoine Mostaert, began his work in Ordos three years later, and his publications are the starting point for any study of Ordos.

This book was written after several months of study of written Mongolian, with Ordos pronunciations given by a non-native speaker, and is prey to all of the possibilities of error that this would suggest. In the chapter on pronunciation and orthography (9–18), the written forms are given along with a French-based transcription of the pronunciation; some pronunciations given are literary readings, not Ordos forms. Now, written Mongolian has a highly ambiguous script, and Ordos has undergone extensive sound change that the orthography does not always conceal, and S’s far too brief treatment is marred by extensive errors: ambiguities in written Mongolian show up in the pronunciation (e.g. gem ‘flaw’ and kem ‘limit’, spelled identically in written Mongolian but misspelled here as ‹gim›, are both given as gem), and the written forms are poorly edited and often crudely misspelled (e.g. *‹čiil› for ‹jil› ‘year’, *‹sibke› ‘dung’ for ‹sibege› ‘rampart’); worse, in several examples the pronunciation of a misspelled written form has been given in place of the actual Ordos (e.g. *ölčii for ölǰii ‘good fortune’). The book’s failings are symbolized by one pair of words meant to show how velar letters are used to distinguish the vowel harmony of words: ǰarlig ‘decree’ and ǰerlig ‘wild’ are identically misspelled as ‹yrliγ› with the back-vowel velar (13).

The grammatical sections are paradigm-based and discuss irrelevant European categories like gender at too great a length. The facts of vowel harmony are included implicitly but not properly emphasized; the discussion of the verb is a resume of the French verb with the corresponding Ordos forms inserted, without the benefits of a treatment based on Ordos agglutinative morphology. However, the forms given do appear to be accurate enough.

This book cannot be recommended to anyone besides Mongolists interested in the history of their field. Sadly, despite S’s admirable enthusiasm for his task and his deserved fame as a pioneering translator of Chinese and Japanese classics, this book is riddled with inaccuracies and should not have been reprinted.