BOOK REVIEWS
237
the path of English speakers learning Mandarin. A pronunciation problem exists because most English speakers do just what Huang directs them to do, i.e., make voiced-voiceless contrasts for these initial sounds instead of the voiceless aspirate-inaspirate contrasts made by the native speakers. If such mistakes are understood by the native speaker it is only because they are adapting to the pronunciation errors and not because the errors are unimportant. Such adaptation is of course automatic and easy if this is the only mistake a beginner makes. However, any beginning student can be expected to add mispronunciations of tone and vowel to the problem of the initials and it is clear that the pronunciation errors can easily multiply beyond the ability of the native speaker to compensate for them. In the long run every area of mispronunciation is equally important and the language instructor cannot afford to let students reinforce any weaknesses.*
It is ironic that many older works were more accurate than Huang in this respect. Wade-Giles p', t', etc. represent a clear attempt to demonstrate that these contrasts were NOT voiced versus voiceless. Huang misses the point altogether when (pp. xx-xxi) he says "voiceless plosives should be written as P, T, K, and the voiced ones simply as B, D, G." The Yale system uses such a symbolization but obviously the symbols themselves are totally unimportant, purely a matter of agreement among the users. The critical point is how these symbols are related to pronunciation in articulatory terms. Here Huang misdirects the student, while the textbooks using the Yale romanization write b, d, g, but clearly describe them as representing voiceless consonants.
Huang continues to mislead his readers by equating English and Chinese sounds when they are not equal. On page xxvi he equates Mandarin yi with the English vowel in see. This is phonetically inaccurate and touches on another common area of mispronunciation in Chinese. The English vowel as in see normally has an off-glide [-iy] with a shift in position of articulation. The Chinese vowel in yi has no offglide and, after the initial consonant or semiconsonant, is pronounced as a pure vowel [-i]. A similar situation exists
*Y. R. Chao suggests (Mandarin Primer, p. 21) that, failing to master voiceless initial stops and affricates, the student may "as a last resort" pronounce them as their voiced English counterparts. He then notes, "The use of voiced consonants.... is not absolutely correct and will give a strong foreign accent."