Reviewed by: The morphology of Chinese: A linguistic and cognitive approach by Jerome L. Packard Katia Chirkova The morphology of Chinese: A linguistic and cognitive approach. By Jerome L. Packard. Cambridge: Cambridge University Press, 2000. Pp. 335. What a word is and how to divide the lexicon into lexical categories has long been a topic of heated debate in Chinese linguistics in the West and in China. Due to its absolute lack of inflectional morphology, Chinese was long considered a language with no distinct word classes in the West (Karlgren 1949:68, Maspero 1934:35). In the Chinese philological tradition, on the other hand, the notion of lexical categories did not exist. Instead, all words were classified into full words (shící) with a concrete meaning and empty words (xūcí) with a broader grammatical meaning. The first scholar who seriously addressed the question of word definitions and lexical categories was Mǎ Jiànzhōng. In his grammar of Classical Chinese, Mǎshì wéntōng (1898), western categories of word classes were for the first time systematically applied to Chinese. After Mǎshì wéntōng, the discussion on the nature of words and word classes carried on for several decades with many prominent linguists such as Chén Chéngzé, Wáng Lì, Zhū Déxī, and Lǚ Shūxiāng contributing to the debate. It was obvious that in the absence of inflectional morphology, criteria for defining words and assigning word classes should be different from those used for most Indo-European languages. One solution was the principle of distribution, which found its most detailed analysis in the works of Zhū Déxī. A major turning point in the discussion was the methodical description of word classes in Chinese by Yuen Ren Chao in his Mandarin primer: An intensive course in spoken Chinese (1948) and Grammar of spoken Chinese (1968). Since then, however, despite the general agreement that Chinese has word classes, no concurrence of the standards for classification has been reached. Packard’s Morphology of Chinese (hereafter Morphology) comes as a welcome addition to this discussion. It serves as an important link between the Chinese linguistic tradition and late twentieth-century advances in western morphology. P proposes a new perspective on what a word is and gives a solution for the assignment of noun and verb classes to words in Chinese. Moreover, by analyzing word properties and word forming principles in Chinese, this study makes significant cross-linguistic generalizations about the structure of words and the functions of word components. Ch. 1 explains the importance of the study of Chinese words, discusses the crosslinguistic concept of word, and outlines the main issues discussed in the book. In Ch. 2, the reader is introduced to various existing criteria for defining words. Ch. 3 focuses on possible descriptions and the assignment of form classes for word components. P introduces the headedness principle which deduces the class identity of bisyllabic words from the right- or left-hand-side position of its constituents. Ch. 4 analyzes the range of complex noun and verb words viewed by component form class and morphological identity. In Ch. 5, P proposes an ‘X-bar analysis’ of Chinese words and also extends the proposed analysis to English data. Ch. 6 focuses on lexicalization as a source of new words and as a way of explaining the nature of the relationship between components and words. The same chapter also analyzes exceptions to the headedness principle as well as the relationshipbetween lexicalization and grammaticalization and between lexicalization and the formation of new words. Ch. 7 considers issues involving the relationshipbetween Chinese words, character orthography, and the lexicon. Ch. 8 concludes the book and summarizes the generalizations discovered. Morphology investigates a variety of highly interesting phenomena and raises numerous significant questions. Below I take up a number of questions which might still require further analysis. On p. 246 P states that ‘the occurrence of -zi, -tou and -r affixed nouns as constituents within complex nominals is generally disfavored in Mandarin’, for example, the word cuò ‘mistake’ can be affixed with the suffix -r while the constituent cuò ‘mistake’ of the word cuòwu ‘mistake’, that is *cuòrwu...