Libraries in North America use a version of the pinyin romanization as modified by the Library of Congress. In terms of phonetic representation, this system is essentially identical to Han'yu pinyin fangan, the approved standard in the People's Republic of China. For example, zhong is always zhong; no permutation such as zhung occurs. There is no departure from the expected representation of sounds and library users can refer to any standard pinyin romanization chart when constructing a search in the catalog.
That covers the sound system. However, there are some important variations in word division practice that need to be taken into consideration. In the majority of cases, a search in pinyin formulated according to the Wade-Giles hyphenation conventions will produce the desired result. Unfortunately, there are some major deviations from and modifications to this principle that will cause searching difficulties and we hope the following notes will help library users to minimize failed searches in the online catalog.
Polysyllabic pinyin is the default standard in scholarly work and news sources in North America and Europe. It is the usual practice in China as seen in government documents, dictionaries, and text books for learning Chinese. Polysyllabic pinyin means that characters are joined together to form "words" or meaningful semantic units, in linguistic terms. This is sometimes referred to as word division. The book title, 中国历史概要 would be romanized in polysyllabic pinyin as: Zhongguo lishi gaiyao (3 words).
The pinyin romanization system as adopted by the Library of Congress is based on monosyllabic pinyin (with some notable and far-reaching exceptions which are discussed below). This means that romanized Chinese text is not grouped together in "words." LC decided to use monosyllabic pinyin because of the difficulties involved in writing a conversion program that could reliably change monosyllabic Wade-Giles into polysyllabic pinyin. Consequently, the book title mentioned above will appear in library catalogs as: Zhongguo li shi gai yao (5 words). All characters in the title, except those for Zhongguo, are romanized separately.
Although monosyllabic pinyin is the rule, there are a set of documented exceptions which affect searching in the online catalog. A summary of the major exceptions is given below:
The Library of Congress, as an agency of the federal government, is obliged to use the form of geographic names approved by the Board of Geographic Names (BGN). The BGN follows Chinese romanization practice - which is polysyllabic pinyin. Therefore, the BGN authorized form for most Chinese geographic proper names in catalog records will be in polysyllabic pinyin:
Certain terms for administrative divisions are joined together when used as part of a proper name. When used generically (not part of a proper name), they are romanized separately:
A list of joined administrative terms is given below:
For the time being, the BGN continues to mandate Wade-Giles form for Taiwan place names. When a Taiwan place name appears as a corporate author or as a subject entry in the catalog it will be in Wade-Giles form. As part of a book title or part of a corporate name it will be romanized in pinyin. See the following example involving the city of Gaoxiong (Kao-hsiung):
This policy results in subdivisions of the city (as defined by cataloging rules) being in pinyin, while the city itself is in Wade-Giles. Consider the following example for the Kao-hsiung City Council:
Problems with Hyphens
Syllables of general non-specific geographic terms are not joined. However, due to variant word division practices in Wade-Giles records, machine conversion has resulted in two pinyin forms in the catalog; joined and separate. In the following examples, both joined and separated forms can be found in the UW catalog as well as OCLC Worldcat.
In other cases, the machine has simply converted what it encountered in the record, based on the presence or absence of a hyphen:
"Di" and "De"
It was standard practice to uniformly romanize the frequently used character 的 as "ti" when it was used as a possessive particle even though it was usually pronounced as Wade-Giles "te".
The conversion software cannot tell when the syllable "ti" should be romanized as "de" since "ti" occurs very often when the character should indeed be pronounced as Wade-Giles "ti" (as in Chung-kuo ti li [geography of China]). Thus the software has uniformly converted the syllable "ti" to "di", resulting in incorrect conversion in almost every instance involving the possessive particle:
T'ai-wan jen ti wen hua kuan (臺灣人的文化觀) became Taiwan ren di wen hua guan, instead of: Taiwan ren de wen hua guan.
We will gradually change the incorrect conversion of "ti" to "di" when used as a possessive particle in catalog records. However, this will take time and in the meantime we advise users to do two searches - one with "ti" and one with "di".
Unexpected Phonetic Conversion Errors
Absent or improperly used diacritics can cause incorrect conversion:
Unexplained Phonetic Quirks
Partial Conversion
Errors such as those listed above will be corrected as they are encountered in the catalog. As one can imagine, this will be a long process.
Conclusion
A draft clarification of Library of Congress romanization guidelines is currently under discussion and should be finalized in the spring of 2002. The section dealing with geographic names may have some small revisions, but most likely will remain as it now stands. Consequently, we have described the important points of the sections relating to geographic names on this page.
Some major revisions are expected in the section dealing with pen names, courtesy names, terms of address, and titles for royalty. For this reason we have not summarized romanization practices in these areas even though some occurances can be found in the catalog. We feel it is better to wait until final revisons are released rather than present more paragraphs of confusing rules and examples - some or all of which may change in the near future. Persons interested in reading the romanization rules governing pen names and other variations too minor or infrequent to list here may consult the full Clarification document directly at Library of Congress Pinyin Conversion page.