I remain amazed by how many people are willing to take China’s official statistics at face value. Yet news story after news story refers to China’s supposed high literacy rate.

If you know any Chinese characters, try to see how many of the following items you can pronounce. (But even if you don’t know any Chinese characters, please keep reading.) The pronunciation needn’t be in Mandarin if you speak another Sinitic language. Moreover, if you aren’t sure how to pronounce some characters but know the meaning of the word nonetheless, give yourself full credit for that item anyway. Characters following a slash are, of course, “simplified” forms.

Scroll down for the answers and more information.

For reference, I have added the frequency of the characters used. Once past the 3,000 or so most frequently used characters, however, figures for frequency of use are difficult to come by and relatively unreliable because these characters are relatively infrequent. Of course, this doesn’t mean these can be ignored completely, because they do still occur and, at present, Chinese orthography doesn’t allow for the insertion of Hanyu Pinyin into a string of characters the way furigana or other non-kanji scripts can be used in Japanese.

If your score fell short of 10, perhaps you’d like to know that the median for PRC university graduates was 6.

These were used in a test of literacy in the PRC that was part of a 1996 “stratified national probability sample” of some 6,000 adults ages 20-69. Care was taken in the selection of those interviewed, so that “for all practical purposes, we have representative national samples of China’s rural and urban populations,” according to Donald J. Treiman, who gives the results of this study in The Growth and Determinants of Literacy in China. For more on this study, which was a monumental undertaking, see Treiman’s Life Histories and Social Change in Contemporary China: Provisional Codebook. UCLA has made available much of the data for the study.

The selection of words, alas, was not particularly good, especially the choice of so many items from Literary Sinitic (classical Chinese). At least three of the final four words should have been tossed out in favor of more examples within the 2,000 or 3,000 most commonly used characters. Nevertheless, the data can be used to provide hints of the true extent of illiteracy in China.

In 1996 China’s adult literacy rate (15+) was about 85 percent, according to Beijing. (The age range for literacy in China is not always clear. Sometimes it refers to all adults. Sometimes it doesn’t include the elderly, whose rate of illiteracy is much higher than those born more recently. Sometimes it excludes everyone born before the founding of the PRC on October 1, 1949. And sometimes other limits are used.) The threshold for literacy was recognition of 1,500 characters for a rural inhabitant, and 2,000 characters for a “worker or staff member employed by an enterprise or institution or any urban resident.” (One country, two literacy thresholds?) (As most students of Mandarin could note, 1,500 characters isn’t going to provide anything resembling full literacy. Far too many characters in texts will be unknown, and far too little of a native speaker’s vocabulary will be unwritable with just 1,500 characters — in stark contrast to literacy through Pinyin, which would be easier to obtain and far more complete.) Moreover, literates were to be able to “read popular magazines and essays, to keep simple accounts, and to write simple essays.” Yet at the same time some one-fifth of China’s adult population could not recognize even such common and simple words as yiwan as written in extremely common and relatively simple Chinese characters (一万). Moreover, the characters in 姓名 (xìngmíng) and 糧食 (liángshi) are also well within the 1,500 most frequently used characters and should thus be known by all literate Chinese. The cumulative figure for those unable to identify all the characters given within the 1,500 minimum (for rural inhabitants) is 24 percent (see table below). That speaks of a literacy rate no greater than 76 percent, which is considerably less than the 85 percent the government was claiming.

A couple more factors need to be considered. First, Treiman’s study took roughly equal samples from China’s rural and urban populations (3,087 urban residents and 3,003 rural residents). But in 1996 about 75 percent of China’s population lived in rural areas, where literacy tends to be significantly lower than in the cities:

Relative to those who at age 14 had rural hukou status and resided in a village, those with urban hukou status residing in cities would be expected, on average, to be eight percentile points higher on the literacy scale. That is, the difference between the two extreme residential circumstances for otherwise similar people is the equivalent of about 1.6 years of schooling. (Treiman, p. 9)

Thus, because the relatively literate urban population is overrepresented, the literacy figure needs to be adjusted down from the 76 percent given earlier. (Sorry, I’m not much good at the math of adjusting sampling rates, so I’ll give a rough figure.) So now it’s at, say, 72 percent, which would give an illiteracy rate about twice as high as China was claiming (and which I think still underestimates the difference between “literacy” in the cities and the countryside). But the picture is still more bleak.

Another factor that cannot be overlooked is that real literacy, even by China’s own limited definition, requires the ability to write, not just read. Remembering how to write Chinese characters accurately, however, is much more difficult than the already difficult task of being able to recognize at least 1,500 of them passively. With this in mind, even doubling the illiteracy rate would not be extreme, I believe. This would yield an actual literacy rate below 50 percent.

Although this method leaves much to be desired, I believe its results better represent reality than official figures.

Literacy has been measured in China primarily according to the quantity of characters recognized (known) by an individual, normally 1,500 characters for rural dwellers and 2,000 characters for urban residents and rural leaders. These measures are not verified directly during a national census. Rather, survey teams note educational attainment and check illiteracy-eradication certificates. County level education departments or work units (danwei) are responsible for assessing through surveys or tests the literacy of and awarding literacy certificates to individuals who have not completed the fourth grade of six-year primary school, the third grade of five-year primary school, or an intensive primary school. — China Country Study, n. 5

Thus, the completion of as little as three years of primary school is enough to get someone listed automatically as literate, regardless of their actual literacy. Although that might be good enough to serve as a measure of basic literacy in a language that uses an alphabet, it isn’t when dealing with Chinese characters, which not only take many years to learn but also require a great deal of reinforcement through practice lest the learner lapse back into illiteracy. Other people are listed as being literate based on possession of an illiteracy-eradication certificate. These certificates, however, are awarded by authorities at the county level or at a person’s danwei; inflation of figures at the local or danwei levels, however, is common; the reasons for this can be summed up as “Individuals worry about punishment, officials worry about performance assessment, and enterprises worry about additional charges.”

(For an excellent look at how state planning and the use of statistics tend to become perverted under certain systems, see Dictatorship, State Planning, and Social Theory in the German Democratic Republic, by Peter C. Caldwell of Rice University. No, this doesn’t have anything to do with literacy or China, but many aspects of socialist planning in the former East Germany were the same as in the PRC.)

Before I close this unusually long post, I’d like to return for a moment to the characters in the literacy quiz. Note the approximate number of strokes in the various characters. Having only a few strokes doesn’t necessarily make a Chinese character “easy” to know. 彳and 亍 have but three stokes each, while 糧 and 食 have a total of 27. Yet more than 50 times as many people could identify the latter pair than the former one. The so-called simplification of Chinese characters did not, and could not, make Chinese characters simple to know or use.

A few words on the China Country Study cited above. This uses official (i.e., inflated and otherwise inaccurate) figures from the PRC. But it covers a wide enough range to be quite useful. It also has a very good bibliography of English sources. But all those pages about literacy — this is a long report — and not even a mention of how damn much trouble Chinese characters are. And essentially nothing about pinyin, either. Very strange.

Here’s its table of contents:

1. Introduction: A Snapshot of Literacy and Illiteracy in China
2. Literacy and Illiteracy in the Chinese Context: Historical and Contemporary Patterns of Literacy Provision
   • A Chronology of Literacy Policy, Definitions and Practice: 1905-2005
     • 1905-1949: Literacy for Saving, Securing, and Strengthening China
     • 1949-1976: Language Reform, Literacy for Collectivization and Production, and the Unequal Expansion of Schooling
     • 1978-1988: Literacy and the Modernization Decade: “Blocking, Eradicating, and Raising”
     • 1988-2005: Literacy for and Assimilation of the Margins
3. Minority Nationalities, Languages, and Literacy
4. Remaining Barriers to Literacy for All
5. Trends in Literacy and Illiteracy Across Regional and Rural-Urban Divides and Across Gender, Ethnicity, Income, and Disability
   • Literacy and Gender
   • Literacy and national minority populations
   • Literacy and disabled populations
   • A Rough Check on the Taken-for-Granted Mathematics of Chinese Literacy
6. Conclusion: Future Outlook and Challenges for Literacy in China
7. Bibliography

If you’d like references other than in the study above, Barend ter Haar has compiled an annotated bibliography on literacy, writing and education in Chinese culture.

And, finally, John DeFrancis has some important things to say on this topic in The Chinese Language: Fact and Fantasy, especially the chapter “The Successfulness Myth.”