DRAFT – not
for circulation or citation!!
Bridie
Andrews
History of Science,
"A man should keep his little brain-attic
stocked with all the furniture that he is likely to use, and the rest he can
put away in the lumber-room of his library, where he can get it if he wants it.”
(Sherlock Holmes, in
Conan Doyle’s The Five
Orange Pips, 1892)
This paper is an attempt to explore the overlooked role of reference
works, and dictionaries in particular, in the transmission of knowledge and
culture. I first came across this topic
when I noticed that the first dictionary in Chinese devoted to Chinese medical
knowledge was published as late as 1921: Xie Guan 謝觀 (Liheng)’s Encyclopaedic dictionary of Chinese medicine (中國醫學大辭典). On examination of the context for the
publication of this medical dictionary, I discovered that there was something
of a publishing revolution in early twentieth century
In this paper, I argue that the early twentieth-century understanding of
what it meant to be “scientific” has
shaped the modern Chinese intellectual world.
Tracing the influence of this idealized “science” is
not difficult, since the historical actors of this period were obsessed with it
as a yardstick of social and cultural development. By following this rhetoric into the design of
reference books and other aids to acquiring knowledge, we will be afforded a
glimpse of how the pursuit of “science” has
changed the whole shape and meaning of Chinese culture.
“Science” is of course a complex
term with a range of meanings, both in English and in Chinese. I am not so unwise as to attempt a definition of it
here. Instead, I wish only to indicate a couple of
key characteristics of “science” in
the early twentieth century that are relevant to this story.
1.
First and foremost, “scientific”
activities were identified by being limited to only observable cause and
effect. This meant that the social and ritual elements of many practices were
superfluous and even considered harmful as “superstition”. Examples of this in medicine would include
the religious ritual which usually accompanied variolation
and, later, vaccination against smallpox.
As Chang Chia-feng has observed, worship of
the goddess of smallpox, dou zhen niang niang 痘疹娘娘
was often an integral part of Chinese families’ management of their
children’s vaccination, even when the vaccination was carried out by
someone trained in western medicine. Since the
efficacy of such rituals
was not empirically observable in terms of cause and effect, the practices
could not be scientific. Of course, this only reinforced a long-standing
prejudice among many Chinese literati against “superstitious practices”,
and a pre-existing preference for empiricism among educated Chinese tended also
to reinforce this particular understanding of “scientific”. This meaning of “science” is
also evident in Western writings, as for example when nineteenth-century homeopaths claimed that
their remedies, based on Hahnemann’s direct observations
of the effects of minute doses of drugs, were more scientific than the heroic
interventions of orthodox medicine such as purging, vomiting, and bloodletting.
2.
The word “Science”, in the form of the
Chinese term kexue
科學, entered
So early twentieth century meanings of “science”
included new, specialized and differently compartmentalized knowledge from
abroad, and a strict limiting of explanations to only those where the link
between observed cause and observed effect could be explained without recourse
to invisible beings or non-natural sources of power.
We should note that this was before the philosophical ideal of modern science first formulated
in the inter-war years by Karl Popper, and criticized from a sociological
perspective in the early post-war period by Thomas Kuhn and others. Before the end of the Second World War, the
ideal of science did not necessarily consist of a process of observation – conjecture – hypotheses –
experiments – refutations – better hypotheses –
scientific laws; nor did it consciously acknowledge the importance of “thought
collectives” and training in scientific paradigms. A comparative study of just what the ideal
of science and the label “scientific”
meant in Euro-America and in other parts of the world around the turn of the
twentieth century would be useful here.
For now, we should at least note that the meanings of “science”
have definitely changed.
Two aspects of social history are vital to this story of the changing
meaning of cultural knowledge in
1.
Printing
Early Protestant missionaries in Serampore,
Technologies invented by missionaries for the purpose of proselytizing
in Chinese were thus essential to the development of an indigenous printing industry
capable of producing large quantities of cheap print in much smaller typefaces
than before. As movable type replaced
xylography, a whole cottage industry that often depended on female home-based
labor for relief-carving the plaques, was slowly eradicated. Other social aspects of this industrializing
process include the fact that traditional methods were better suited for
producing ephemeral literature (pamphlets, flyers), especially in the early
nineteenth century, when most missionary printing was illegal. After the Treaties of Nanjing
and
2.
Educational reform
A decree of 1906 stated explicitly that henceforward, “education
in China was not merely to discover men of talent, but to educate the whole
nation and to inculcate loyalty to the throne, respect for Confucius, the
awakening of the people to a sense of their national responsibility, the
promotion of a military spirit, and the creation of a practical and realistic
sense” (Purcell, 67).
As we now know, the awakening nationalism and sense of military
inferiority convinced many that only a new government could save
After the establishment of the Republic of China in 1912, the first
Education Minister Cai Yuanpei 蔡元培
(1867-1940) ordered that the Chinese classics were no longer to be taught at all. This left seven
faculties at the college level: arts, science, law, commerce, medicine,
agriculture and applied science (eg
engineering). However, the political
instabilities of the ‘warlord period’ in
modern Chinese history meant that enforcement of the new curriculum was patchy
at best, and that much education was still carried out in the many small,
privately-run traditional schools. Women’s
education continued to lag far behind that of men, so that statistics for women’s higher education from 1924 show that women
comprised only 2.54 of the student population in government universities and
colleges, while they made up 14 per cent of the student body in missionary
colleges. (Cleverly, 45-8; Purcell, 70)
Nonetheless, in this
era of education reform, with the civil service examinations gone and the
Classics on
the wane, there was a pressing need for
new curricula, new textbooks, and teachers educated in the new subjects of
instruction.
The Commercial Press was established in
In the year of the founding of the Republic of
China, 1912, the Commercial Press' successes in the provision of textbooks for
the new school system suffered a setback when the editor of two such series of
textbooks, Lufei Kui
陸費逵 (1886-1941) resigned to set up his own
publishing company, the China Book Trading Co. 中華書局, taking the series with him. Lufei was a noted educator:
for a while he edited a journal known in English as the Educational Review, and the new education system promulgated by Cai Yuanpei in 1912 was partly
designed by him (Who’s who in China, 1936: 182). The two houses
were to be fierce rivals until the start of the next Sino-Japanese War in 1937
(Wang, 1973: 69). Between them, they dominated the
market for the new textbooks and reference works of the twentieth century.
Imperial
During the
'Hundred Days' Reform of 1898, Liang Qichao perceived the absence of a national
spoken language as one of the causes of
By the end of the Qing dynasty in 1911, about a dozen different systems for
rendering the pronunciation of Chinese in an accessible way had been devised by
various Chinese scholars. One consequence of any application of a
phonetic writing system for Chinese is that communication would be in the
vernacular, or something very close to it. If
writing was to be a means by which anyone could communicate, that writing would have to be in the language of ordinary people. Literacy would no longer be inextricably
linked to the canonical texts. So the concern for increasing
literacy rates in
This change has immense
consequences. As many first-hand
accounts of growing up in
With the discarding of the canonical
texts as the basis for education, and the flood of new translations and
transliterations coming into
Exacerbating these frustrations was the view, widely held among
foreigners and influential among many Chinese, that the Chinese language lacks
a formalizable grammatical structure. In the mid-nineteenth century, missionary
educator Joseph Edkins had argued that Chinese was a
primitive language and likely to retain traces of the mythical “Ursprache”, the language of all humanity before
From its publication in 1716 until after the fall of
the Qing dynasty in 1911, the standard Chinese
language dictionary was the Kangxi zidian 康熙字典. This
dictionary, like most of its predecessors, indicated the pronunciation of each
character with two further characters, the first of which indicated the initial
sound, and the second of which gave the final and the accompanying tone (the fanqie 反切 system). However, these pronunciations had been copied
from Tang dynasty (618-906 AD) readings, and often
deviated greatly from any modern dialect of Chinese. The Kangxi dictionary therefore also gave a further character
as an indication of pronunciation. This
new clue to pronunciation was homophonous with the dictionary entry in the
When western missionaries arrived in
In 1912, the editors of The Commercial Press made the first Chinese attempt to replace the Kangxi dictionary with a modern
reference tool when they published the New character dictionary 新字典 under the editorship of Lu Erkui 陸爾奎 (1862-1935). This dictionary included many of
the new characters coined by missionaries and Japanese to translate scientific
terms, and also some characters used to indicate dialect words, which had been
ignored by the purely literary dictionares of the
past. The fact that it specifically
aimed to replace the Kangxi dictionary is indicated by a preface
from Cai Yuanpei, who
listed the inadequacies of the Kangxi dictionary
for modern scholars and praised the New
character dictionary for correcting many of them. The New character dictionary was relatively
small, however, with only about 40,000 character
entries (fewer than the Kangxi dictionary, which had 47,000), 1000
pages, and with no attempt to render the pronunciation of characters in any of
the new phonetic ways; it still used the archaic fanqie 反切 system.
Lufei Kui, who had left the
Commercial Press in 1912 to set up the China Book Trading Company 中華書局, was also competing with his old employer in the
production of a new dictionary. He, too,
aimed to publish a replacement to the Kangxi dictionary,
and in 1915 the China Book Trading Company published the Chinese encyclopaedic character dictionary 中華大字典. Its
title page calligraphy was executed by the famous woman calligrapher, Wu Zhiying, and the following pages boasted dedications by both the President, Yuan Shikai,
and the Vice-president, Li Yuanhong. Next came eight prefaces by various literary
luminaries (including Liang Qichao) and the editors of the
project, Lufei Kui and Ouyang Tuancun. Lufei Kui's preface listed
several failings of the old Kangxi dictionary, including
its lack of detail, non-inclusion of common expressions, and difficult
organization. He then described how he
had marvelled at the convenient organisation
of Japanese and English bilingual dictionaries, and decided to apply their
principles to Chinese.
For example, the Character
dictionary attempted to provide a standardized pronunciation system - again
a version of fanqie,
but one based on a Song dynasty work that had recorded
the pronunciation of a record number of characters - over 53,000 of them. The Dictionary
innovated further by numbering each meaning of each character, and giving a
single example of the usage of each meaning.
It also provided a new, supplementary index of characters ordered by the
number of strokes without reference to their radicals, particularly useful for
those characters for which the radical is not obvious. Perhaps most significantly, it was the first
Chinese-edited dictionary to include a few
common two-character words ('colocations') as subentries under the
first character of the word, enabling
the dictionary to be used as a 'word' dictionary 辭典 to a certain extent.
The significance of this change is that in contrast to the literary language, the Chinese vernacular was very rich in such
two-character words, and increasingly so with the accelerating adoption of
neologisms - particularly technical terms - from Japan.
In the same year, 1915,
the rival Commercial Press published its second attempt at a
modern dictionary, the Word-source 辭源. In contrast
to the Character dictionary, this had
no prefaces or dedications. In place of
a preface, the chief editor Lu Erkui wrote a short
introduction. Clearly feeling that
The Ciyuan dictionary was a break with the past in that
words composed of two or more characters were given separate entries to
themselves, and it also innovated in the inclusion of biographical entries, explanations of some idiomatic phrases (chengyu 成語) and
encyclopaedia-style explanations rather than bald definitions, especially of new
scientific terms and concepts.
Both these major dictionary projects that reached
publication in 1915 were intended to replace the old Kangxi dictionary. However, neither of them embraced the
contemporary movements either to provide an accessible phonetic rendering of
Chinese characters or to bring the written language significantly closer to a
vernacular. Indeed, the first preface of
the Chinese encyclopaedic
character dictionary was written by an energetic opponent of the vernacularisation of written Chinese, Lin Shu 林紓. (Lin is
well-known for his translations of foreign novels into elegant literary Chinese.) Nonetheless, both attempted to fill the
perceived need – and evident
market niche – for guides to
contemporary usage of the written language, as opposed to the guides to archaic
usage and rhymes represented by previous scholarly dictionaries.
[Note: show slides of the plates from da zidian]
[Add Cihai, 1936, encyclopedic dictionary, with illustrations throughout the
text, and appendices to indicate pronunciation according to three different
systems; the names and pronunciation of the chemical elements; a comparative
table of international weights and measures; and a list of almost 12,000
foreign words with their Chinese translations cross-referenced in the back. ]
Wang Yunwu 王雲五 (1888-1979) is perhaps best known to westerners as the inventor of the
four-corner system for looking up Chinese characters in reference books. Since
he was also the chief editor of the Commercial Press from 1922, he was able to have this
system reproduced in a great many dictionaries and other works. Every dictionary
produced by the Commercial Press after this time had at least an index of
characters by Wang’s method, and in
many works, the body of the work was also arranged this way, with supplementary
indexes by traditional radicals and stroke order. Less
well known are his efforts to create a library classification system that would
meet
Wang’s four-corner
system considers each character’s structure, and assigns a numeral value to each of the strokes closest
to the corners of the imaginary square in which the character is written. The upper and lower left and right strokes
are independent of the stroke order, so that very little education is necessary
beyond the rules of the system itself, in order to determine the four- or
five-digit number of each character. The
characters may then be listed in an index, or even arranged in a reference
work, in numerical order.
By adding a fifth (supplementary) digit, referring to the number of
strokes of the first type to occur in the character (ie
at the upper left), to characters when several ended up with the same
four-digit number, Wang was able to claim that 89.9 per cent of all Chinese
characters would be found in groups of five or fewer characters under a single
number. [slide] This compares
very favorably to the huge numbers of characters found under each of the 214
radicals of the Kangxi system, and the dozens found even when
only characters with a certain number of strokes after the radical are grouped
together. (As an example, Wang cites the
205 characters of eight strokes under the 草 radical, and 204 with
nine strokes under the same radical in the Kangxi Dictionary.
Wang explained the
rationale behind his system like this:
The
arrangement of words or characters – a condition which never occurs in countries
where the phonetic system of writing is in vogue – becomes a pressing problem in
Wang considers the possibility of
arranging characters phonetically, but dismisses this solution on the grounds
a. That “a
phonetic system is applicable only to a unified national tongue, which can
hardly be expected of
To solve these problems, Wang devised
eight principles of an ideal lexicographic system, the first four of which are:
1.
That every person can understand and use
it;
2.
That every character can be found or
located rapidly;
3.
That every character can be found or
located directly, without resort to any other index;
4.
That there is absolutely no necessity to
understand the order of strokes in a character; ... (Wong, 1926:9)
Wang’s system, therefore, was designed to be used by students even at a very
elementary level of literacy. He was
inspired, he says, by consideration of the Chinese telegraphic code. In this, every Chinese character is
represented by an arbitrarily-assigned four-figure number. Looking up the appropriate number is as
time-consuming as looking up an unknown character in the Kangxi Dictionary, so that the professional translators employed to
render telegraphic messages into code have had to memorize the numbers of
several thousand characters. By
contrast, the translation process from code back into characters is
incomparably faster: the operator merely searches in a single numeral index,
and scarcely needs to be literate in order to render the telegraphic message
accurately back into character form.
Wang’s twin concerns with the efficiency of conducting business in Chinese,
and with the accessibility of his system to the barely literate and to students
studying on their own, nicely mirror the cultural concerns of the day.
Ding’s case illustrates the fact that this desire to
access information directly was part and parcel of the whole challenge to
traditional authority that pervades the history of the period and goes well
beyond the “May Fourth” movement of 1919 and
thereafter.
Stephen Owen highlights the difference in Mencius’
thought between wen 文,
the literary language, and ci 辭, the statement.
For Chinese thinkers since Mencius, he writes,
“wen – the
patterned, literary word and the written word – is not a figuration or deformation of
plain language, but the full and final form of language. Although it is most open to misunderstanding,
wen is also the entelechy of
language.” (Owen, 25)
As is well known, this regard for the literary language as the
depository of culture and the epitome of fully developed thought became a
casualty of the Chinese enthusiasm for science in the early twentieth century.
The appearance of dictionaries and reference works document the changing
mechanisms of the transmission of Chinese culture in the early 20th
century. When these are taken into
account, it becomes easy to see why so many people opposed both the vernacular
style and the development of a phonetic script.
They feared, with good reason, the loss of the shared literary tradition
upon which Chinese élite culture was based.
Indeed, now that we are blessed with hindsight, even the rather conservative measures
implemented in the PRC since 1949 to facilitate widespread literacy - the
simplification of some characters and the use of the vernacular in most kinds of printed
literature - have very quickly produced a generation of scholars who have
trouble reading unsimplified characters – that is, any book published before the
mid-1950s – and who struggle to understand classical
Chinese.
To return to the dictionary that started me off on this course, what is
perhaps most striking about Xie Guan’s
Encyclopedic Dictionary of Chinese
Medicine is the indication it gives us that even specialists could no
longer be expected or required to memorize large amounts of classical
texts. In spite of the fact that Xie deliberately restricted the entries in his dictionary
to only terms occurring in China’s own medical literature
(the dictionary included no entries about western medicine at all) the very
fact of its existence shows that students were moving away from the idea that
competence resided principally in mastery of the classical texts. They were moving towards becoming the kinds
of scholars we are today: highly skilled at knowing where to find information,
but no longer the repositories of it.
Bailey, Paul J. Reform
the People. Changing Attitudes Towards Popular Education in Early Twentieth-Century
Cai, Yuanpei
et al 蔡元培等. Wan Qing
sanshiwu nian lai zhi Zhongguo
jiaoyu (晚清三十五年來之中國教育
Education in
Chen, Bingtiao 陳炳迢 Cishu Gaiyao (辭書概要 Essential Aspects of
Dictionaries).
Chen, Jiang
陳江 et al (comp. & ed.): Shangwu yinshuguan da shi ji (Major events in
the history of the Commercial Press 商務印書館大事記)
Chiang, Yee. A
Chinese Childhood.
Cleverley, John. The Schooling of
De
Francis, John. Nationalism and language reform in China.
Ding,
Fubao 丁福保. Foxue xiao
cidian (佛學小辭典 A Short
Dictionary of Buddhism). Shanghai: Yixue shuju, 1919.
Drège,
Jean-Pierre. "Les Aventures De La Typographie Et Les Missionnaires
Protestants En Chine Au Xixe Siècle."
Journal asiatique
280 no.3-4 (1992): 279-305.
Fan, Hongye 樊洪業 ."Cong "gezhi" dao
"kexue" (From the ‘investigation
of things’ to
‘science’ “從‘格致’到‘科學‘”)，Journal of dialectics of
nature 自然辯辯證 法通訊
vol. 10 no.3
(1988): 39-50.
Fung, Yulan.
"Why
Gao, Hanqing 高翰卿. Ben guan chuang ye
shi (本館創嶪史
'History of the founding of the Press') in Commercial Press, The, ed. Shangwu yinshuguan
jiushiwu nian (商務印書館九十五年
95 years of the Commercial Press).
Nakayama, Shigeru. "Translation
of 'Science' to 'Kagaku, Kexue 科學'.".Presentation
to the Third International Conference of the History of Chinese Science,
Norman,
Jerry. Chinese.
Owen, Stephen.
Pepper, Suzanne. Radicalism
and Education Reform in 20th-Century
Purcell, Victor. Problems of Chinese Education.
Wang, Li 王力. (Modern Chinese Grammar). 2 vols.
Wang, Yunwu 王雲五. Shangwu yinshuguan
yu xin jiaoyu
nianpu (商務印書館与新教育年譜
Annual chronicle of the
Commercial Press and modern education [in
Wang, Yunwu 王雲五. Xin Muluxue De Yi Jiaoluo (新目錄學的一角落 One
Angle on the New Methods of Cataloguing).
Wong, Su-Ling, and E.H. Cressy. Daughter
of Confucius.
Wong, Y.W. (Wang Yunwu 王雲五). Wong's System of
Chinese Lexicography: The Four-Corner Numeral System in Arranging Chinese
Characters. Shanghai: Commercial Press, 1926.
Xie, Guan 謝觀, ed. Zhongguo yixue da cidian
(中國醫學大辭典 Encyclopaedic Dictionary of Chinese Medicine).Four
vols.
Zheng, Zizhan 鄭子展. Lufei Bohong
xiansheng nianpu (陸費伯鴻先生年譜
Annual chronicle of [the life of] Lufei Bohong [Kui]).
* I would like to thank
Peter Buck, who patiently listened and suggested improvements to many of the ideas presented here; and to Lesley Wei-chung Ma for invaluable research assistance.
[1] When, as a graduate
student, I joined Michael Loewe’s undergraduate
classes in Classical Chinese, and had to read chapters from Wang Chong’s Lunheng, or Dong Zhongshu’s Chunqiu fanlu, I was sometimes surprised that the traditional
commentators glossed words that seemed (even to me) as though they ought to be fairly elementary. When
we remember that even the traditional type of dictionary was a fairly uncommon
item in traditional
[2] To this day there is a greater proportion of Esperanto speakers in
[3] This is the origin of
that rule against splitting your infinitives, by the way. Since they cannot be split in Latin, they
ought not to be split in English.
[4] Examples of ways in which
modern Chinese has become ‘Europeanized’
include the much more frequent use of particles (extra characters) to indicate
tense, plurals, or parts of speech. It
is also much less frequent to have to infer the subject of a sentence from the
context since it is usual modern practice to state this explicitly.
[5] This observation is
confirmed in the first paragraph of the “Editorial principles” (bianji da gang 編輯大綱) preface to the first
edition of the Cihai
辭海 dictionary of 1936: [about early word
books and rhyme manuals]: 大抵可供行文獺祭之用，而不可以供讀書明理之用也。”In
general, they are useful for writing like ‘otters who sacrifice
[fish]’, but are no use for reading books and explaining
them.” (‘Otters sacrifice [fish]’
comes from an allusion in the Book of
Rites to the otter behavior of leaving partially-eaten fish on river banks,
from which it came to mean superfluity).