Chinese Personal Names and Titles:

Problems in Cataloging and Retrieval

Shuk-fong Lau
V i c k y Wang

ABSTRACT. The increasing availability of Chinese language ma-

terials and other materials of Chinese authorship in North American
libraries are posing major complications for cataloging which lead to
retrieval problems, as exemplified by Chinese personal names and
titles in this study. These complications include: the intricate nature
of the Chinese script along with the extensive use of the traditional
and simplified Chinese characters, and the application of variant ro-
manizatlon schemes. includine Pinvin. Proaosed solutions to im-
prove access to c h i n k langugge miterials i h u d e : providing more
access ooints in Pinvin form. establishine standards for international
practick in romani& ~ h i n d s epersonal~amesand in publishing the
order of the famil and given names of all authors, and linking cata-
loging authority &es wlth OPACs.

Before 1957 there w e r e no standardized rules for cataloging Chi-

nese language materials in North American libraries.' Descriptive
cataloging w a s carried out in different w a y s b y individual libraries.
In v i e w o f the need for standardization, in 1953, joint efforts were
m'ade by the American Library Association (ALA) and the Library
o f Congress (LC) t o revise the A. L.A. Cataloging Rules for Author
and Title Entries,' the Rules forDescriptive Cataloging in the Li-
brary of Congress,Qnd A Manual of Romanization, Capitalization,

Shuk-fong tau, MLIS, is Referencefinformation Retrieval Librarian, Refer-

ence ~ e ~ a r t m e ;nd
n t Vicky Wang, MLS, is Cataloger, Cataloging Department,
Memphis State University Libraries. Memphis, TN 38152.
The authors wish to express their gratitude to Ann Denton, John Evans and
Betsy Park for their constructive comments on this article.
Cataloging & Classification Quarterly, Vol. 13(2) 1991
O 1991 by The Haworth Press, Inc. All rights reserved. 45
46 CATALOGING & CLASSIFICATION QUARTERLY

Punctuation, and Word Division. This cooperation eventually led to

the publication of the Preliminary Rules and Manual for Cataloging
Chinese, Japanese, and Korean Materials in 1957.4 The revised
rules worked under the principle that they could, and should, be
applied to all language materials. Many North American and Asian
libraries with Chinese language collections adopted and began fol-
lowing the revised rules. Chinese names, when used as the main or
added entry, were entered in romanized form using the Wade-Giles
system. This trend of conformity was consolidated with the publica-
tion of the Anglo-American Cataloging Rules (AACR)I in 1967 and
the revision of 1978 (AACR2).6
Even though much has been done in the past three decades to
provide greater access to Chinese language materials through ro-
manization (Wade-Giles), application of standard cataloging rules
(AACR2), and inputting Chinese characters in machine readable
format, e-g., by use of the Online Computer Library Center's
(OCLC) Chinese, Japanese, and Korean languages (CJK) software,
numerous problems and difficulties still persist in cataloging Chi-
nese language materials. This article addresses some of these prob-
lems and proposes some solutions. Particular attention is focused on
the problems of transliterating Chinese characters and establishing
entries for personal authors and titles.

NATURE OF CHINESE CHARACTERS

The immediate problem in cataloging Chinese language materials

in North American libraries is the language barrier. Traditional Chi-
nese characters are divided into two classes: wen, simple figures,
and tzu, compound letters. They are subdivided into six categories,
forming the so-called "Six Scripts":'

1. hsiang-hsing, pictograms. These characters are mainly picto-

rial symbols of certain objects, e.g., 'jen' (man) represents a
person standing on two legs.B
2. chih-shih, indicative symbols. These characters are also picto-
rial but suggest the meaning of logical or abstract terms, e.g.,
'ming' (bright or clear) is a combination of two words, 'jih'
 Shuk-fang Lau and Vicky Wung 47

(sun) and 'yueh' (moon), which have in common the attribute

of luminosity."
3. hsing-sheng, phonograms. About seventy-five percent of all
Chinese characters are of this type.'These characters are com-
posed of two elements: the semantic element called a radical
and a phonetic element indicating the pronunciation of the
word. An example is 'chan,' where the left-hand element
gives the meaning of water, and the right-hand element gives
the sound of 'chan.' When combined, it means to moisten,
and is pronounced as 'chan."'
4. hui-i, ideograms. These characters are formed by associating
two or more symbols for objects to those of ideas, e.g., 'fa'
(invade or attack) denotes a lance or spear held by a man."
5. chuan-chu, associative characters. These characters are
formed by extending, deriving, modifying or distorting the
original meaning of certain concrete characters into new
meaning. For example 'wang' is a picture of a fishing-net
which, by extension it becomes a general expression to catch.'"
6 . chia-chieh, borrowed characters. These characters are used in
a sense that is not their own. They are formed as a result either
of error, being used for another character, or by conversion,
designating an object which has no special character. An ex-
ample is 'shuo,' which in modem usage means to speak, but in
the ancient times it meant to rejoice, now written as 'yueh.""
Table I is a summary of the examples of the Six Scripts of Chi-
nese characters given above.

TRADITIONAL AND SIMPLIFIED

CHINESE CHARACTERS
Chinese characters are arranged in dictionaries according to the
radicals of which they are composed or with which they are tradi-
tionally associated. .Some radicals can stand by themselves as a
word, but some have to combine either with the phonetic element
which gives the sound or with the primitive which does not give the
48 CATALOGING & CL4SSIFICATlON QUARTERLY

TABLB I. Examples of the Six scripts of Chinese characters

Categories g X A W P L E S
of the
Six scripts Romanization Weaning Chinese

hsiang-hsing man

hsing-sheng chan moisten

hui-i fa invade/attack

chuan-chu vang fishing-net

chia-chieh shuo speak(modern usage) ( '@ )

rejoice(ancient usage)

Yueh rejoice(ancient usage) ( )

sound of the character. There are 214 radicals arranged in modern

dictionaries according to the number of pen strokes used in writing
them.ls
The intricate and complex nature of the Chinese characters cer-
tainly proves to be challenging for the catalogers of Chinese lan-
guage materials. The learning and understanding of the language is
further complicated by the existence of several writing styles, often
expressed in calligraphic form. Though the general principles of
construction of the Chinese characters have remained relatively un-
 Shuk-fong Lau and Vicky Wang 49

changed, the variation in number, shape, and position of strokes in

writing has resulted in the development of different styles from time
to time."
The earliest form of Chinese writing, which can be traced back to
the Sheng dynasty at about 18th century B.C., is called 'Chia-Ku-
Wen,' or the Shell-and-Bone Script. These characters are engraved
on tortoiseshell or animal bones. The majority of these characters
are found to be pictograms, followed by ideograms and phono-
grams." In the Chow dynasty (12th century B.C.), the 'Ku-Wen' or
the Ancient Script, was developed; then followed by a new style
called 'Ta-Chuan' or the Great Seal, which was widely used till the
Chin dynasty (246-207 B.C.). Later, a new style, called 'Hsiao-
Chuan' or the Small Seal, was adopted. From the Chin dynasty to
the Han dynasty there gradually developed 'Li-Shu' or the Official
Script, which is also known as 'Tso-Shu' which serves as a means
to quicker and easier writing. The 'Pa-Fen' Style was also devel-
oped during the Han period, which is closely related to 'Li-Shu.'
After Han, three new styles, which involve variations in number,
shape and position of strokes, were adopted. They are: 'Kai-Shu,'
the Regular Style; 'Hsing-Shu' the Running Style; and 'Tsao-Shu,'
the Grass Style.'BThese three styles are still widely used by contem-
porary Chinese calligraphers.
The Regular Style is adopted as the book script in publishing
Chinese language materials in many Chinese communities like Tai-
wan and Hong Kong. However, the Mainland Chinese government
advocates a simplified version of Chinese characters that usually
requires fewer strokes to write a word. Currently all Chinese publi-
cations from Mainland China use the simplified characters. In order
to catalogue a Chinese publication, ideally speaking, the cataloger
must know both scripts, a task requiring a lengthy learning process.

ROMANIZATZON
In addition to the problem of the intricate nature of the Chinese
characters, there is also the problem of transliteration or romaniza-
tion. Basically, transliteration is the use of one alphabet to represent
the scripts of another language, so that users can get access to vari-
ous language publications unrestricted by his or her knowledge of
50 CATALOGING & CL4SSIFICATION QUARTERLY

that particular language. There are numerous romanization schemes

used by different institutions to transliterate publications in Chinese
to English, e.g., Wade-Giles, Pinyin, Yale University, Chu-yin
Tzu-mu, and Gwoyeu Romatzhy etc."
The most prevalent transliteration scheme, adopted by LC from
near the beginning of the century and then by most of the Western
world, is the Wade-Giles system. It was first developed in 1867 by
the British diplomat, Thomas Francis Wade and was subsequently
modified by Herbert A. Giles in his Chinese-English Dictionay in
1912." In 1958, however, the Pinyin scheme was adopted for publi-
cation by the Mainland Chinese government. It employs the 26 let-
ters of the Roman alphabet, 4 digraphs (zh, ch, sh & ng), and 4 tone
marks ('high,' 'high-rising,' 'falling-and-rising' and 'falling'). It is
now the official romanization scheme used in Mainland China to
serve as "a teaching aid in studying the characters ... and in
fostering the standard spoken language known historically as Man-
darin.""
Though there are some computer programs which can convert
from Wade-Giles to Pinyin and vice versa,22the problem of revers-
ibility exists no matter which system is adopted. Ideally, any "sci-
entific" romanization should be reversible, i.e., the original script
can always be reconstituted from its romanized form. With Chinese
characters this is difficult to achieve, for many characters with the
same sound have different meanings. Further, a Chinese character
can be an unique word by itself, but when combined with other
characters, a new character with the same or a new meaning can be
formed. Its pronunciation may be identical or different from the
original.
The diversity and multiplicity of dialects spoken throughout
China adds to the problem of transliteration. When romanizing, the
cataloger has to look for contextual clues to find the exact character
in order to deduce the correct meaning. This is extremely difficult
to accomplish if the cataloger lacks the knowledge of the dialect.

PERSONAL AUTHORSHIP AND TITLE

AS MAIN O R ADDED ENTRIES
Taking into account these two fundamental difficulties of master-
ing the Chinese language, i.e., the intricate nature of the Chinese
 Shuk-fong Lau and Vicky Wang 51

characters and the use of variant romanization schemes, the follow-

ing is a discussion of establishing personal authorship and title as
the main or added entries in cataloging Chinese language materials.
The choice of a proper entry for a work will affect the accuracy of
the classification as well as the cataloging. However, selecting the
main or added entries from a Chinese publication is not a simple
task. Some of the difficulties encountered are discussed in the fol-
lowing sections.

Personal Name Entries

Dificulty in distinguishing between the last and the first name:

One of the most significant differences between Western and Chi-
nese names is the order of the last and the first names. Generally
speaking, the Chinese surname precedes a given name of one or two
character(s). AACR2 Rule 22.1D2 requires the cataloger to "retain
hyphens between given names if they are used by the bearer of the
name" and "include hyphens in romanized names if they are pre-
scribed by the romanization system adopted by the cataloging
agency."
The problem of identifying the last and the first names of a Chi-
nese author is not serious if the text is published in Chinese and the
cataloger knows the language. However, it becomes a problem
when the work is a translated one, with the author's name transliter-
ated in English on the title page.
At present, there is no standard practice by authors, translators,
or publishers in presenting the order of the non-Western last and the
first names. Publishers who are issuing contemporary translations
of works of early Chinese scholars may or may not follow the prac-
tice of the Wade-Giles scheme and the rules of AACR2. Contempo-
rary Chinese authors who write in English may transliterate their
own names into English, accordhg to their spoken dialects, result-
ing in a form that is very different from the Wade-Giles convention.
They may or may not follow the Western practice of presenting
their given names first followed by their surnames. Some scholars
from Mainland China prefer to publish their transliterated names in
the Chinese order while some overseas Chinese scholars publish
their names in the Western order.
To avoid such unnecessary confusion, AACR2 Rule 22.4B2 re-
52 CATALOGING & CLASSIFICATION QUARTERLY

quires that if the first element of a person's name is a surname, a

comma be placed after the surname to distinguish it from given
names on the title page. However, when no comma is used, it be-
comes difficult for catalogers to seek a correct access point for the
catalog. Confusion and errors also arise because common Chinese
given names are often spelled the same as surnames, a useful exam-
ple is the case of Cheng Liu, the author of Soils and Foundation.
Because Cheng and Liu are popular Chinese surnames, it becomes
difficult, if not impossible, to distinguish between the last and the
first names. Similar confusion occurs in the case of two different
authors whose names appear in identical form on the title.pages of
two different publications. The author, Shen Fu, whose surname
was "Shen," noted for Chapter from a Floating Life, is a famous
scholar in Chinese literature. His name is transliterated the same as
the artist, Shen Fu, whose surname is actually "Fu." In this partic-
ular case, it would be easy for a cataloger to mistake them as the
same person. From the above examples it is evident that if LC's
Name Authority File is not consulted, the choice of name entries
can be a puzzle to catalogers and library users alike. Sometimes a
Chinese author may have a two-character compound surname but a
single character given name. When this occurs, the author's sur-
name may be hyphenated; and if it is published in the Chinese order
of surname first, e.g., Ou-yang Hsiu, the hyphenated surname may
be misunderstood as a two-character given name. In a more extreme
case, a Chinese author may have a hyphenated two-character sur-
name followed by a hyphenated two-character given name, e.g.,
Chu-ko Ching-yun. This may cause catalogers to spend a tremen-
dous amount of time distinguishing and verifying the family and
given names, if they are not familiar with the Chinese compound
surnames.
Difficuly in choosing between different names used by the same
author: In ancient times, it was the general practice for a Chinese
person to use different names in various places, ages, and occa-
sions. The types of names given and used were: "milk name"
(nai ming) during childhood; "school name" (shu ming; hsueh
ming) from the first period of education; "courtesy name" or
"style" (tzu) upon reaching adulthood; "literary name" (hao);
"studio name" (pieh hao); "official name" (kuan ming) by earning
an academic degree or official rank; "pen name" (pi ming); "post-
 Shuk-fong Lau and Vicky Wang 53

humous name" (shih ming) and "temple name" or "dynastic title"

(miao hao) for emperor^.'^ In contemporary times, a Chinese author
may add a "non-Chinese name" (yang ming) when communicating
with non-Chinese speakers who experience difficulty pronouncing
his or her Chinese given name.
Regarding the choice of name, AACR2 Rule 22.1 requires one to
choose "the name by which he or she is commonly known." In
case a person has more than one name, Rule 22.2A requires one to
"choose the name by which the person is clearly most commonly
known, if there is one." An examination of LC's Name Authority
File reveals that the established heading for the famous Chinese
poet of the Sung dynasty, Su Shih, does not conform to the above
rules. It is well known among the Chinese that Su Shih is more
commonly known and revered by the use of his literary name or one
of his pseudonyms, Su T~ng-p'o.~The similar dilemma also occurs
in the choice of name for Tao Chien, who is more commonly
known among the Chinese as Tao Yuan-ming. Besides, if using
AACR2 Rule 22.2C2, one should establish a person's dominant
pseudonym as the main entry. For instance, LC now accepts the
better known Mark Twain as the main entry for Samuel Langhorne
Clemens, the famous American author who wrote under the Twain
pseudonym after 1863.x However, it does not apply the same rule
to the Chinese scholars, such as the cases of Su Tung-p'o and Tao
Yuan-ming.
Another example is Shang Ming-hsuan's Sun Chung-shan Chuan
(Biography of Sun Yat-sen). Although the established entry, Sun
Yat-sen, is well known in the Western countries, "Sun Chung-
shan" is the name used and preferred by the Chinese from Taiwan,
Mainland China, Hong Kong, and overseas. It seems that LC's
choice is based on the fact that the name is clearly most commonly
known by Westerners instead of being the majority preference of
the Chinese. Even though LC's Name Authority File has solved
some difficult name authority problems for Western authors, more
needs to be done for non-Western authors.
Dificulty in identifying an author whose name has been roman-
ized in different schemes: With the end of the Cultural Revolution
and the adoption of the Open Door Policy in China in the late
1970s, the Western world was exposed to the Pinyin scheme more
so than at any previous time. As Western newspapers and journals
54 CATALOGING & CLASSIFICATION QUARTERLY

began to use the Pinyin system, e.g., the Mew York Times in 1979,
readers began to see Beijing and Mao Zedong rather than Peking
and Mao T s e - t ~ n g . ~ ~
The Western cataloging world faces the dilemma of whether to
continue to use the Wade-Giles system or to change to the Pinyin
system. Although LC has announced its decision to tentatively
adopt the Pinyin system for cataloging of Chinese language publica-
tions on June 29, 1979,27the decision was immediately dropped a
year later after opposition from major university libraries in the
United state^.'^ While the "controversy over romanization" ended
with LC's decision,29the essence of the problem still has not been
solved.
AACR2 Rule 22.3C requires the cataloger to "choose the form
of name that has become well established in English-reference
..
sources . If variant English language forms are found, choose the
form that occurs most frequently." For instance, when cataloging a
work written by the current Chinese leader known in Pinyin style as
Deng Xiaoping Wenxuan, catalogers have to conform to the tradi-
tional practice of establishing the author's main entry under "Teng
Hsiao-ping" even though "Deng Xiaoping" has occurred most fre-
quently in English publications and references.
On the other hand, AACRZ Rules 22.1A and 22.1B suggest that
the basis of the name heading be determined by the most commonly
known name from the chief sources of information. It is known that
for more than three decades Mainland China has printed the Pinyin
form of an author's name in publications. In this case, catalogers
following AACR2 Rule 22.3C strictly should adopt the Pinyin form
of name as the main heading, but following the established practice
obviates that choice. Nevertheless, the inconsistency of rule appli-
cation does happen in the choice of name for main entry. An exam-
ple is that the name Wang Zhongshu has been established in LC's
Name Authority File as main entry, even though this is a romanized
name in Pinyin form.
Another factor causing different romanized forms of names is the
multiplicity of dialects spoken by the Chinese in different regions.
Prior to 1949, the government of the Republic of China established
a national spoken language called "Kuo-yu," later known as Man-
darin, based on the dialect spoken in Peking. Since then Mandarin
has become the most popular spoken dialect in Mainland China and
 Shuk-fang Lau and Vicky Wang 55

Taiwan. However, Cantonese and Fukienese are two popular dia-

lects in the southern part of China. People from these southern re-
gions were the very first ones to emigrate to Southeast Asia and
North America. Eventually, these overseas Chinese transliterated
their Chinese names in English according to the sound of their indi-
vidual dialect.
While the distinction between the English transliteration from
Mandarin, Cantonese, or any other dialect is neither obvious nor
seems important to the gen&al library patrons, it is of concern to
catalogers. For instance, the name of the late President of the Re-
public of China, "Chiang Kai-Shek," as established in LC's Name
Authority File, was actually transliterated according to the Canton-
ese dialect. In Mandarin it is transliterated as "Chiang Chieh-
shih." Thus, when the title page of a publication has a transliterated
name from Chinese, the cataloger immediately faces the dual prob-
lems of identifying the dialect and the romanization scheme. An-
other example is Wong Yoon-wah who wrote Wang Jun-hua Tzu
Hsuan Chi. The established main entry for the author's name is also
not romanized in Wade-Giles form. In another case, Chan Wing-
tsit, who translated Reflections on Things at Hand, has given both
the Chinese and transliterated (in Cantonese form) names on the
title page of his work. His own transliterated name is chosen as the
main entry, even though in Wade-Giles it should be transliterated as
Chen Jung-chieh. Thus, when the title page of a publication has a
transliterated name from Chinese, LC's practice is to use the one
that the author uses even though it is not in Wade-Giles form.
As the same Chinese family name can be romanized in different
ways based on the dominant dialects spoken in differcnt regions, for
instance, Mandarin in China and Taiwan, Cantonese in Hong
Kong, Fukienese in Malaysia and Singapore, catalogers experience
difficulty in distinguishing who's who in Chinese." This could eas-
ily lead to the possibility that several different authors are repre-
sented by one romanized name. This kind of ambiguity is chiefly
caused by the multiplicity of dialects spoken in various Chinese
communities.' Table I1 gives some examples of the variations of
Chinese surnames romanized in different Asian regions to illustrate
this chaotic situation:
From Table 11, it tells that the same Chinese surname, such as
Chen can be romanized as Chan in Hong Kong, Tan or Ding in Ma-
56 CATALOGING & CLASSIFICATION QUARTERLY
TABLE 11. Romanised Chinese Burnames in various Asian Regions

aomani~edNames in Various Asian Regions

Chinese
BUnXmeS China Taivan ~ o n gxong nalaysia & eingapore

Chen Chen Chan Tan, Chan, Ding

Lin Lin Lam Lilo
Li Lee Lee, Li Lee, ~e
LIu Liu Lau Low, Liew

wu wu Ng Ng, Wu, 00, Hoo, Goh

Zhang Chang Cheung Chong, Teoh
ZhU ChU Chu Choo
wang wang Wong WonQ, Ong
Huang Huang, Hwang Wong Wong

laysia or Singapore, often based on the dominant dialect spoken

there.
Another problem is that some Chinese scholars are reluctant to
convert their dialect based romanized names into the Wade-Giles
form, as it may become "another surname." The pride of family
names reflected in the emphasis on the "correct" way "our name"
is spelled, is also shared by some contemporary Western scholars."
Title Entries
Problems encountered by catalogers in selecting title entries from
variant forms of romanization for Chinese language materials are
just as complicated as are those of personal name entries. Some of
the problems in original cataloging of establishing title entries are
discussed below.
The immediate problem of publications in Chinese with no trans-
 Shuk-fong Lou and Vicky Wang 57

lated or transliterated data on the title page is to identify whether the

traditional or simplified script has been used. Occasionally, a paral-
lel title either in translated or romanized form, may be found along
with the Chinese title on the title page, the colophon page, the book
cover, or elsewhere in the publication. It would be an easy task to
make a parallel title entry if a translated title in English or a roman-
ized title under Wade-Giles is given. Having a good command of
Chinese, the cataloger certainly has no problem in transliterating
Chinese according to Wade-Giles.
With publications in English, it is relatively easy for the cataloger
to procure the necessary bibliographic data from the title page or the
cover. Yet, given only a transliterated title that is romanized under
neither Wade-Giles nor Pinyin, the cataloger must determine which
romanization scheme has been used. An appropriate example is the
volume of Chinese classic poetry, whose title has been romanized
as She King or ShiKing by different scholars. Neither transliteration
conforms to the Wade-Giles or the Pinyin schemes. The second
word, "King," when romanized in the Wade-Giles form would be
"Ching," but in the Pinyin form it should be "Jing." Faced with
the situation where the form of romanization can not beeasily iden-
tified and where other sources are unavailable, catalogers are hard
pressed to make a successful conversion.
It is safe to assume that publications of the last thirty years from
Mainland China have employed the Pinyin system. Catalogers who
wish to follow the AACR2 rules and convert from Pinyin to Wade-
Giles will observe some problems. Under the Wade-Giles scheme
each Chinese character is romanized as a single word and multi-
character personal or geographical names are hyphenated.'= How-
ever, under the Pinyin scheme, a single character may or may not
be romanized as a single word. Some characters may be grouped
together as a single compound word without hyphenation. The fol-
lowing are two examples:
PINYIN WADE-GILES
1) Bing Ju Fen Xi
-
- Ping Chu Fen Hsi
2) Putonghua Zhengyin Shouce = Pu Tung Hua Cheng Yin Shou Tse

The first example is a straightforward conversion from one char-

acter to one word. The second title in Pinyin contains both two-
58 CATALOClNC & CLASSIFlCATlON QUARTERLY

character and three-character words. Under the Pinyin scheme, the

first word, "Putonghua" actually contains three separate charac-
ters, romanized as "Pu Tung Hua" under the Wade-Giles scheme.
"Zhengyin" is a two-character word for "Cheng Yin," and
"Shouce" is a two-character word for "Shou Tse." Under such
circumstances, it is extremely difficult for catalogers to do a suc-
cessful conversion unless they know both schemes well.
While LC has established English uniform titles to improve ac-
cess to some classical Greek works such as Homer's Odyssey and
Aristophanes' Birds, providing uniform titles for Chinese publica-
tions in library catalogs is also indispensable for users to gain access
to the desired record. A good example is the Chinese classic novel,
Hung Lou Meng, also known as Shih Tou Chi, and variously trans-
lated as The Dream of the Red Chamber, The Dream of the Red
Mansion, The Stoly of the Stone, etc. The author's name, roman-
ized in Wade-Giles as "Tsao Hsueh-chin," and in Pinyin as "Cao
Xueqin," has been printed on title pages differently by each transla-
tor. If the cataloger does not know all these variant translated or
transliterated titles and names, it will be difficult to establish a uni-
form title entry to link them together. This eventually will lead to
retrieval problems in an Online Public Access Catalogs (OPACs)
environment.

RETRIEVAL PROBLEMS
OCLC's CJK software offers member libraries access to Chinese
script records. Member libraries can directly interact with OCLC's
online system through the M300 workstations." Currently, most
libraries in North America do not have separate Chinese databases
to store and retrieve their Chinese language materials nor do they
have a system that can accommodate inputting and retrieving both
in English and Chinese. Nevertheless, much has been done in this
direction by many vendors and librarians in Asia." In most cases,
Chinese collections in North American libraries are entered in ro-
manized form according to the Wade-Giles scheme. To retrieve
Chinese language materials from the catalog, the patron faces the
same problem as the cataloger does. To many users, any romaniza-
 Shuk-fang Lou and Vicky Wang 59

tion system is merely a repository of meaningless phonetic words.

Users who know Chinese well may not be familiar with the roman-
ization scheme being used and would encounter difficulties in re-
trieving Chinese language materials.
Even though Wade-Giles has been well established and many
scholars are familiar with its use, Pinyin is becoming more domi-
nant in the contemporary world. Western writers; editors, reporters
and columnists generally look more favorably on Pinyin:" As a
result, the newer generation of scholars, researchers, or students are
exposed more to Pinyin than Wade-Giles. The mass media in North
America, withits emphasis on timeliness and recency, are more
dynamic and prepared to adapt to the changing world than the cata-
logers in research libraries. Catalogers must consider the hundreds
or thousands of records that would be affected by the adoption of
new standards. Even though it may seem desirable to convert from
Wade-Giles to Pinyin, the obvious concerns are the cost, the accu-
racy, and the time required for massive conversions. More research
is needed to determine the feasibility of such a change.
In addition, there arises the question of who would really benefit
from the conversion. Except patrons from Mainland China who
know Pinyin, most users do not have any formal training in either
Wade-Giles or Pinyin. In 1987188 there were 62,480 foreign stu-
dents of Chinese origin studying in institutions of higher education
in the United States. Among these students, 25,170 were from
Mainland China, 10,650 from Hong Kong and 26,660 from Tai-
wan." It is natural for students from Mainland China to retrieve any
Chinese language materials from library catalogs by using the Pin-
yin scheme. If the catalog does not utilize Pinyin or provide any
cross-references to the established entries in Wade-Gi'les, the user
will definitely have a frustrating,experience. The same dilemma is
also faced by the younger generation of scholars in North America
who have more exposure to Pinyin than Wade-Giles. However, as
for students from Hong Kong and Taiwan, even though Chinese is
their first language, it is likely that most of them are unfamiliar with
either of the schemes. Hence, it is debatable who would benefit
most from the conversion- the catalogers, the patrons, or neither!
 CATALOGING & CLASSlFICATION QUARTERLY

SOL UTZONS
During the past decade more and more libraries closed their card
catalogs and turned to OPACs that can provide more access points
and flexibility in retrieving bibliographic records. The more sophis-
ticated of the OPACs allow for global changes of headings through
"search and replace" features. Such a capability reduces greatly
the onerous burden of updating terminology. However if the system
does not provide flexible keyword searches for any romanized
schemes, it may still be difficult for patrons to retrieve Chinese
language materials unless they know the conventions of the Wade-
Giles romanization scheme.
More Access Points b y Provision
of Entries in Pinyin
The most important need to improve the access of the Chinese
language materials is the provision of more access points through
additional entries of variant names, e.g., tracing Pinyin translitera-
tions of author and title entries. The two actions required for imple-
menting this idea are the revision and updating of AACRZ rules for
cataloging Chinese language materials and a decision by LC to add
Pinyin transliteration to its Name Authority File. Both would defi-
nitely save the time and effort of catalogers of Chinese language
materials, and would in turn facilitate success in Chinese name and
title searches by users of the OPACs.
Consensus in Romanizing and Presenting
Last and First Names
If authors would arrive at a consensus in using only one roman-
ization scheme in romanizing their names, and in presenting the
order of their last and the first names, readers and catalogers would
be well served. These goals can best be accomplished by the Na-
tional Information Standards Organization (NISO), which has long
been developing and promoting technical standards for information
science, libraries, and related publishing practices under the um-
brella of the American National Standard Institute (ANSI). By
standardizing these two practices, NISO can set a yardstick, and
 Shuk-fong Lau and Vicky Wang 61

eventually an international standard, for libraries and the publishing

world to follow.

Authority Control

An alternative for libraries wishing to help patrons perform suc-

cessful comprehensive searches is to provide authority control in
the OPAC. Authority control is a series of intellectual decision
making processes aimed at "maintaining consistency in headings in
a bibliographic file through reference to an authority file."" With
the availability of the bibliographic utilities like OCLC, RLIN, and
WLN, bibliographic data and resources can now be shared. LC's
Name Authority File, used by OCLC's participating member li-
braries, serves to provide accurate and consistent headings, to con-
trol access points from variant and related headings, as well as to
update and maintain authority records. After gaining online access
to this file, many libraries simply do away with their own authority
files.
Even though authority work has been considered essential for
years by cataloging departments, some system vendors provide no
authority control in their OPAC. The lack of such an important
element limits the possibility of the success of a search. If the ro-
manized name of the author or title does not retrieve any records,
users may conclude the library does not have that particular work,
not realizing that the form of the name entered simply does not
match the name used in the catalog. Users may be unsure in decid-
ing when to end an apparently unsuccessful search. Even if the user
knows that the author has pen names, there is no guarantee that the
relevant publications will be retrieved unless all possible names
have been input by the cataloger. This is further complicated if both
the title and the author's name are transliterated.
Despite the fact that OPACs can provide rapid and flexible re-
trieval through truncated, keyword and Boolean searches that by-
pass the control of subject headings and corporate names, authority
control is indispensable in OPACs. In an OPAC with authority con-
trol, the success of a search will be enhanced with the provision of
"SEE" and "SEE ALSO" references which link all the related and
variant names.
62 CATALOGING & CLASSIFICATION QUARTERLY

Cooperation among library personnel can greatly improve a cata-

log's search capabilities. It is the reference librarian who reflects
user needs in information retrieval to the system designer or pro-
grammer who determines the framework and the function of the
OPAC. It is the cataloger who, after consulting various authority
files and applying relevant cataloging and classification codes, de-
termines the content of the catalog. It is in such collaborative efforts
where authority control finds its finest expression.
Hence, to make an effective use of the online catalog, one must
link the authority work of the cataloging department to the public
display of information in the OPAC. The smoother the linkage and
more transparent it is at the level of user interface, the more en-
hanced the search process will be." Once the linkage is established,
the cataloger has to maintain the authority file accurately and in a
timely manner. Authority work is expensive, and if it is only avail-
able at the cataloging level, it fails to achieve its full potential.
However, if it is also at work in the OPAC, its cost is justified as it
facilitates the success of searches. The satisfaction of users' infor-
mation needs is the mission of all types of libraries. Hence, to pro-
vide truly effective service in information retrieval, authority con-
trol should be provided in OPACs.

National and Transnational Cooperation

Beginning in April 1988, LC/NACO (National Coordinated Cat-

aloging Operations) has launched a pilot project, the National Coor-
dinated Cataloging Program. Under this Program, selected research
libraries adhere to agreed-upon practices and policies for creating
national level bibliographic records on Western European language
materials. LC's database serves as its core to receive, maintain and
distribute contributed records nationally and internationally." It is
anticipated that such practice would be extended to non-Western
language materials, including Chinese publications. In this single
national database, each access point of each bibliographic record
would be in compliance with a verified authority record, thus elimi-
nating confusions and inconsistency in cataloging different lan-
guage materials, leading to the establishment of a national uniform
standard. By then LC could serve as a model for other Asian coun-
 Shuk-Jong Lau and Vicky Wang 63

tries like Taiwan, China and Hong Kong to follow. With the rapid
technological advances in computers and telecommunication net-
works, a trans-national effort may become feasible to work toward
a universal catalog!4o
With the increasing availability of Chinese language materials in
North American libraries, the growing influence of the Pinyin ro-
manization system, and the growing number of Chinese students in
North America, there arises the need to revise the contemporary
cataloging practice. It is time for LC, NISO and librarians to give
more serious thought and take actions to establish standards in cata-
loging Chinese language materials so that improved retrieval ser-
vice could be available in the near future.

REFERENCES
1. James Anderson, "Cataloging and Classification of Chinese Language
Library Materials," in Cataloging and Classification of Non- Western Material:
Concerns, Issues and Practices ed. by Mohammed M. Aman, (Phoenix, Ariz.:
Oryx, 1980), p. 93-129.
2. Clara Beetle ed., A.L.A. Cataloging Rules for Author and Title Entries
(Chicago: American Library Association, 1949).
3. Rulesfor Descriptive Cataloging in the Library oJCongress (Washington,
DC: Library of Congress, 1949).
4. Hsu Lee-Hsia Ting, "Problems of Cataloging Chinese Author and Title
entries in American Libraries," Library Quarrerly 36, no. 1 (January 1966): 1-13.
5. Anglo-American Cataloging Rules, North American text (Chicago: Amer-
ican Library Association, 1967).
6. Anglo-American Cataloging Rules, 2nd ed. (Chicago: American Library
Association, 1978).
7. L. Wieger, Chinese Characters: Their Origin, Etymology, History, Clas-
sification and Signification. A Thorough Study from Chinese Documents (New
York: Paragon Book Reprint & Dover Pub., 1965).
8. Tsuen-Hsuin Tsien, Written on Bamboo and Silk: the Beginnings of Chi-
nese Books and Inscriptiom (Chicago: University of Chicago Press, 1962), p. 25.
9. Yee Chiang, Chinese Calligraphy: an Introduction to Its Aesthetic and
Technique, 3rd ed. (Cambridge, Mass.: Haward University Press, 1973), p. 27.
10. The New Encyclopedia Britannica, 15th ed. (Chicago: Encyclopedia Bri-
tannica, 1986), S.V. "Chinese Writing System", p. 238.
11. Wieger, Chinese Characters, p. 10.
12. Chiang, Chinese Calligraphy, p. 11.
13. Richard Newnham, About Chinese (Middlesex, England: Penguin Books,
1971), p. 36.
64 CATALOGING & CLASSIFICATION QUARTERLY

14. Wieger, Chinese Characters, p. 11.

15. The New Encyclopedia Britannica, 15th ed., s.v. "Chinese Writing Sys-
tem," p. 238.
16. Tsien, Written on Bamboo and Silk, p. 182.
17. Ibid., p. 24.
18. Chiang, Chinese Calligraphy, p. 41-42.
19. John J. Deeney, comp., SIyle Manualand Transliteration Tablesfor Man-
darin (Taipei, Taiwan: Tamkang College, 1973), p. 18-31.
20. H. A. Giles, A Chinese-English Dictionary, 2d ed. (Shanghai: Kelly &
Walsh, 1912).
21. "Times Due to Reverse its Chinese Spelling," New York Times, Sunday,
4 Feb. 1979, p. 10:6.
22. Kwoon-choh Yue. "Computer Conversion from Wade-Giles Romaniza-
tion to Pinyin or from Pinyin to Wade-Giles," Hong Kong Library Association 5
(1980): 67-72.
23. Anderson, "Cataloging and Classification," p. 102-103.
24. Burton Watson, trans., Su Tung-P'o: Selectionsfrom a Sung Dynasty Poet
(New York and London: Columbia University Press, 1965), p. 4-5.
25. Patrick H. Kellough, "Name Authority Work and Problem Solving: The
Value of the LC Name Authority File," Technicalities 8, no. 6 (June 1988): 3-5.
26. "Times Revises Style of Chinese Spelling," New York Times, Sunday, 4
March 1979, p. 10:l.
27. Library of Congress Information Bulletin No. 38 (June 29, 1979): 239-
240. and No. 39 (May 2, 1980): 140.
28. "Controversy Arises on Chinese Spelling," New York Times, Sunday. 30
Sept. 1979, p. 5:l.
29. Beatrice Ohta, "The Library of Congress and Chinese Romanization,"
Journal of Library and Infomation Science 6 (October 1980): 151-153.
30. Ting, "Problems of Cataloging," p. 3-4.
31. Charles Clement. Ottilia Koel and Mary Lou Miller, "Where is the Solu-
tion; Where is the Problem?" RQ 24 (Spring 1985): 282-284.
32. Cataloging Sewice Bulletin No. 118 (Summer 1976): 35-36.
33. David Y. Hu, "OCLC and Its CJK Software," Journal of Educational
Media & Library Sciences 23, no. 1 (Autumn 1985): 63-75.
34. Grace Cheng, "The Design of a ChineseiEnglish Library System: a Hong
Kong Librarian's Perspective," Journal of Librarianship 21 (October 1989):225-
245.
- .
35. Olive Holmes, "The Pinyin Syndrome," Scholarly Publishing 11, no. 3
(April 1980): 221-227.
36. National Center for Education Statistics, Digest of Education Statistics
1989 (25th ed., Washington, DC: U.S. Department of Education, Office of Edu-
cational Research and Improvement, 1989), p. 392.
37. Arlene G. Taylor, "Authority Files in Online Catalogs: an Investigation
of Their Value," Cataloging & Classification Quarterly 4, no. 3 (Spring 1984):
1-17.
 Shuk-fong Lau and Vicky Wang 65

38. Larry Auld, "Authority Control: an Eighty-year Review," Library Re-

sources and Technical Services 26, no. 4 (October,December 1982): 319-330.
39. Henriette D. Avram and Beacher Wiggins, "The National Coordinated
Cataloging Program," Library Resources and Technical Services 32, no. 2 (April
1988): 111-115.
40. Mary Piggot, A Topography of Cataloguing: Showing the Most Impoflant
Landmarks, Communications and Perilous Places (London: Library Association,
1988), p. 11.

for faculfy/professionals wilh journal subscription recommendation

authority for their insfitutionallibrary. . .
If vou have read a reorlnt or oholocoov of lhis article, would
ydu like lo make suie that your librah also subscribes to
lhis iournal? If vou have the authoritv lo recommend sub-
scriptions to your library, we will send you a free sample
copy for review with your librarian. Just fill out the form below-and make
sure that you type or wrlte out clearly both the name of the journel end
your own name and addrear.
( ) Yes, please send me a complimentary sample copy of this journal:

(please wrile in complele journal title here-do no1 leave blank)

Iwill show this journal to our institulional or agency library for a possible
subscription.
The name of my institutionallagencylibrary is:

-
-

INSTITUTION:
ADDRESS:
CITY: STATE: ZIP:
Return to: Sample Copy Department, The Haworth Press, Inc.,
10 Alice Street. Bingharnton, NY 13904-1580