May 24, 2006

On Learning Mandarin in America

[Guest post by Victor Mair]

Two daughters of a couple whom Li-ching and I have known for many years both went to the same elite American university (one of the very best in the United States; extremely competitive to get into). The father and mother are mainland Chinese who were born in Taiwan, went to National Taiwan University, and attended graduate school in the United States. Both of the daughters were born in America, but the father and mother have always spoken Mandarin to them and required that they attend weekend schools for Mandarin instruction throughout their primary, junior high, and high school education.

The two girls, who are very diligent, intelligent, and obedient (TING1HUA4), probably recognized about 500 HAN4ZI4 before entering college and could speak at the intermediate level, but they could barely write anything and had difficulty reading even children’s books (admittedly, Chinese children’s books are infamous for not being written in a style appropriate for young readers). When they went to their elite university, they essentially had to start all over again at the beginning.

Both of the daughters went through the third-year level of Mandarin at the university. The elder graduated about five years ago, and her Mandarin now is probably at the same level she had achieved when she was in the fifth grade of elementary school – viz., barely functional for reading and writing, and moderately fluent for speaking and listening. Her younger sister is now in 3rd-year Mandarin at the university and is likely going to end up at about the same level in all four skills (reading, writing, listening, and speaking) as her JIE3JIE.

In fact, what prompted me to write this account at all is the sensational realization by MEI4MEI’s mother that she was writing her Mandarin compositions with a software program that permits one to type in English sentences and let the computer convert them into Mandarin! The mother discovered this situation when she read one of her daughter’s compositions and could scarcely believe how ungrammatical and unnatural it was. Such software is apparently in quite widespread use, not just in America, but even in Hong Kong, Singapore, and elsewhere.

Now, JIE3JIE and MEI4MEI are both highly intelligent and hard-working, but their Mandarin is, for all intents and purposes, useless when it comes to reading and writing, and minimal when it comes to speaking and listening – despite the fact that they have spent more than 15 years studying it in school and university (plus at home). And I’m sure that this tragic situation is by no means peculiar to these bright sisters, but is endemic among many (actually most) students across the country and, indeed, throughout the world. How can this be?

Well, as I knew intuitively when I began studying Mandarin in 1968 and have reiterated on countless occasions since then, it’s because there’s far too much emphasis on HAN4ZI4 from the very beginning. I believe that students should NOT be exposed to HANZI for **at least** the first year of instruction, and preferably not for the first two years. Only then should the characters be gradually introduced. Why? The answer is simple: students need to master the basics of the language (pronunciation, vocabulary, grammar, syntax, idioms, etc.) before they are required to memorize hundreds of HANZI. It is essential to internalize the patterns and structures of the language before spending endless hours vainly striving to master large numbers of HANZI. As a matter of fact, students who initially concentrate exclusively on learning pronunciation, grammar, syntax, etc. and are not burdened by having to memorize HANZI actually learn the HANZI more effectively and quickly later on when they do start to acquire them systematically **in relation to the language as a whole.** This has been proven in the ZHU4YIN1 SHI4ZI4, TI2QIAN2 DU2XIE3 (“recognize characters [through] phonetic annotation, speed up reading and writing”) experiment in the People’s Republic of China. It is also borne out by the experience of students who learn Cantonese, Taiwanese, and other Sinitic languages without having to be saddled with the HANZI. It is remarkable how swiftly such students can attain fluency when exposed to a full course of instruction and exercise.

The negative effects of having to learn HANZI during the first few years of instruction are numerous. In the first place, they consume limitless amounts of time that would better be spent on actual language learning. Secondly, they emphasize monosyllabic morphemes over multisyllabic words, which constitute the overwhelming proportion of the lexicon. Third, the HANZI themselves cannot be efficiently memorized in the absence of a prior mastery of the fundamentals of the language, thus a vicious cycle ensues, with neither language acquisition nor control of the script making any noticeable progress.

JIE3JIE’s and MEI4MEI’s mother is still earnestly seeking a way to improve the reading and writing skills of her daughters in a realistic manner. I have a simple solution: make available a large amount of quality literature with phonetic annotation for each character, preferably showing word division marked by spaces (FEN1CI2 LIAN2XIE3). This is comparable to KANJI texts with FURIGANA in Japanese.

Students should not be blamed for being poor learners when it is their teachers who employ outmoded, impractical methods.

[Guest post by Victor Mair.]

Posted by Mark Liberman at May 24, 2006 06:34 PM