Archives and Rare Books

The draft was brought to the Ricci Institute, then located at the University of San Francisco, from its homologue in Taipei, where it had apparently been sitting for a number of years. 

The dictionary draft was not produced there, however, but in Taichung, some 110 miles (175 km) south of Taipei on Taiwan’s western coast. Its origins are even older, and can arguably be traced back to 1940s Beijing (then known as Peiping). The draft constitutes one part of one of the largest projects of Chinese foreign language lexicography undertaken in the twentieth century. In this essay, I will introduce the project that produced this collection and suggest an approach for studying it.

The project that produced this multilingual dictionary draft does not appear to ever have had a formal name. In publications from the 1950s and later, it is simply referred to as the “Polyglot dictionary,” Dictionnaire chinois polyglotte, and, with more gravitas, Novum glossarium sinicum plurium linguarum.” I have chosen to refer to it as the Taichung polyglot in this essay.

At the Jesuit language school at Chabanel Hall—established in the alley of Shihu hutong 石虎衚衕 in 1937 (Meynard 2006: 98-101)—Zsámár “began collecting slips of paper with Chinese characters and their Hungarian equivalents,” since there was no Chinese-Hungarian dictionary available at the time. “When the slips of paper filled several shoe boxes, the idea of a dictionary took hold,” the New York Times later reported. (In 1953, “all the news that’s fit to print” apparently included Chinese lexicography.)

The Communist victory in the Chinese civil war in 1949 first circumscribed and eventually proscribed missionary work in China. Foreign missionaries were expelled in the 1950s. As the tide was turning, Zsámár made his way to the Portuguese colony of Macao, where he reconnected with a French missionary of similar lexicographic bent.

André Deltour, S.J. (Du Ande 杜安德, 1903–1983) and Henri Pattyn, S.J. (Ba Zhiyong 巴志永, 1902–1985), like Zsámár, had been engaged in various lexicographical projects during the Japanese occupation of Peiping. Pattyn, notably, had worked on a thesaurus and other editorial projects that were all “suspended because of lack of funds” at war’s end (Wright 1947: 360). Deltour, however, was able to bring “eight large trunks, six containing books and two containing manuscripts” from Peiping to Macao (Grand dictionnaire Ricci 2001: vol. 1, vi).In 1949, Zsámár and Deltour teamed up, and as the Portuguese outpost filled with missionaries fleeing the Chinese mainland, others joined them. The Jesuit residence had in 1951 become “practically a house of research with eight Fathers and two Brothers exclusively dedicated to the compilation of a monumental polyglot Chinese dictionary” (Mateos 1995: 107). Chinese-Hungarian, Chinese-French, Chinese-Spanish, and Chinese-Latin dictionaries were to be compiled concurrently on the basis of a shared Chinese lemmata list.

In July 1952, when it looked as if the PRC might take control over Macao, the project was moved to Taiwan, where the missionary lexicographers felt safe under the protection of the United States Navy. The move was carried out in September of that year (Motte 1953: 9). The Jesuits established themselves in a part of a property belonging to “an old Formosan family” on the outskirts of Taichung. It appears to have been a rustic setting, with an old banyan tree shading the gate. In Taichung, a Chinese-English dictionary was added to the other four, giving the project its final form.

In 1953, Zsámár retired to Macao and was replaced as leader of the editorial work in Taichung by Yves Raguin, S.J. (Gan Yifeng 甘易逢, 1912-1998), one of the old Shanghai Jesuits. Around twenty individuals worked on the project, including a number of local “assistants” who were not priests.

In retellings of the editorial work, these scholars are referred to as “Chinese,” and it is not clear whether this meant refugees from the mainland or members of the local Hokkien or Hakka communities. Raguin mentioned some of them by name: Zhang Yi 張毅, Zhang Yuanbo 張遠博, Zhang Keming 張克明 (who worked on the project for 38 years), Liu Pengjiu (Raguin does not give the Chinese characters), and “Mr. Xue,” who died suddenly sometime in the 1950s. The new English section was headed by Thomas Carroll, S.J. (He Duomo 賀多默, 1909–1964), who had a PhD in East Asian studies. The Jesuit superior, Juan Antonio Goyoaga Uriarte, S.J. (Gao Yugang 高欲剛, 1917–1998), a veteran of the Spanish Jesuit mission to Wuhu, Anhui, was in charge of the project but was not involved in the editorial work (Raguin 2000:56-57). 

Manuscript dictionaries of Chinese and European languages were produced already by members of the old Jesuit mission (sixteenth to eighteenth century). Most of them remained unpublished. In the nineteenth century, Protestant missionaries entered the field as well. After the Jesuits of the new mission started to arrive in China in 1842, more dictionaries were compiled and published.

One of the individuals involved in the Taichung polyglot, Joseph Motte, S.J. (Mu Qimeng 穆啟蒙, 1906–1990), was cognizant of the contributions of earlier works such as Séraphin Couvreur’s (Gu Saifen 顧賽芬, 1835–1919) Dictionnaire classique de la langue chinoise (Classical dictionary of the Chinese language; 1890, later revised), Augustin Debesse’s (Mei Ruzhou 梅汝舟, 1851–1928) Petit dictionnaire chinois-français (Small Chinese-French dictionary; 1901, later reprinted), Charles Taranzano’s (Tao Deming 陶德明, 1866–1942) Vocabulaire français-chinois des sciences mathématiques, physiques et naturelles (French-Chinese vocabulary of the mathematical, physical, and natural sciences; 1914, third ed. under a slightly different title in 1931), the lexicographical component of Léon Wieger’s (Dai Suiliang 戴遂良, 1865–1933) pedagogical works and the supplementary Néologies (1919 and 1925, see Bernard 1927), and Robert Henry Mathews’s (Ma Shouzhen 馬守真, 1877–1970) Chinese-English dictionary (1931, rev. ed. 1943; PL1455.M34 1972). Yet all of these dictionaries had shortcomings that motivated the compilation of a new work, according to Motte. Couvreur’s dictionary was limited to the classical language, Debesse’s was not comprehensive, Taranzano’s was limited to the vocabulary of a specific domain, and Wieger was out of date. Mathews was good, but Motte and his colleagues had the ambition of compiling something “both more precise and more complete, especially with regards to the modern language.” The goal of the Taichung team was to produce a work for use by “the China missionaries as well as Chinese who want to improve their knowledge of European languages” (Motte 1953: 9).

Another (anonymous) report written around the same time described the intention behind the work in a little more detail. The new dictionaries were not intended to replace dedicated dictionaries for classical Chinese, but would cover modern technical vocabulary fairly comprehensively, in order to serve as a general-purpose dictionary for everyday use. In other words, it would be a kind of multilingual Larousse for Chinese. This was a reference to the encyclopedic French dictionary of the nineteenth century, which in histories of lexicography is often contrasted with the Littré, which was more narrowly focused on language rather than facts. In the polyglot dictionary, a “gentleman [will] find, in condensed form, all of the vocabulary that he might encounter in conversation or reading.” The dictionary should thus contain everything necessary to understand the great classics of Chinese literature, but especially the vocabulary of the modern language in both its spoken and written forms. The missionary lexicographers vowed to include vernacularisms that more prudish writers might avoid. Coverage of all the dialects would not be feasible, but should be included to some extent, with northern Mandarin given preference (“Notice” 1953: 565).

The team proceeded by first establishing a kind of inventory of the Chinese lexicon. In this process, they relied most heavily on Chinese monolingual dictionaries rather than bilingual missionary lexicography. 

 Rhyme books were arranged according to the pronunciation (rhyme) and character books were arranged according to the graphic form of characters. Dictionaries of words and phrases certainly existed (many bilingual dictionaries were of this type, for example, as was the imperial poetic thesaurus Peiwen yunfu 佩文韻府), but the most widely used rhyme books and character books were dictionaries of single Chinese characters. Rhyme books were very popular in the late imperial period, to a large extent because they were useful when writing poetry. New lexicographical genres emerged in the early twentieth century, and it was books of these new kinds that the Taichung team relied on in their work.

The team in Taichung had at their disposal around two hundred Chinese dictionaries (Motte 1953: 12), but the lemmata list was established on the basis of a much reduced set of three: Guoyu cidian 國語辭典 (Dictionary of the national language; 1937), Ciyuan 辭源 (Source of words; 1915–1939), and Cihai 辭海 (Sea of words; 1936). Zsámár used Guoyu cidian to create a collection of index cards with Chinese lemmata while he was still in Peiping. This book, on which work had begun in 1928, represented “a crucial step in the modernization of lexicography in China.” The work “for the first time combines essential features of modern Chinese lexicography: sound-based arrangement, inclusion of single and multiple character expressions, coverage of the modern lexicon, and indication of standard pronunciation” (Klöter 2019: 325).

The sound-based arrangement in Guoyu cidian was more akin to European-style alphabetical order than to imperial-era rhyme books, as the headwords in it were arranged according to the pronunciation of their first elements. The Taichung project was similar in that it used alphabetical arrangement based on the Romanization of the Mandarin pronunciation of the Chinese words. Given this arrangement, the right Romanization was essential, and Guoyu cidian was the most authoritative lexicographical resource in this regard.

Ciyuan and Cihai, by contrast, were arranged in the manner of late imperial “character books.” Their primary contribution was as milestones in the development of dictionaries of polysyllabic words as opposed to single characters. Ciyuan “focuses on the development of the language approximately up to the Opium War (c. 1840), but it also contains some lexical innovations of the second half of the nineteenth century.” By contrast, Cihai “covers ancient and modern words, including scientific and technical terms as well as personal and geographical names” (Klöter 2019: 322). The influence of this wide coverage is clearly discernable in the Taichung project as described by its participants. 

The workflow in Taichung

To draft their dictionary, the Jesuit lexicographers first established a preliminary list of headwords by cutting up copies of the most important dictionaries at their disposal and distributing the words onto index cards. Yet not only dictionaries were used to compile the wordlist, but also all the words included in Chinese high-school textbooks. As for the Catholic vocabulary so important for missionary work, it reflected current usage rather than the preferences of the compilers (“Notice” 1953: 565).

After a first culling of this material, the Jesuit lexicographers had about 15,000 different Chinese characters in combinations that amounted to c. 165,000 words and phrases. (These figures vary somewhat between reports on the project, which probably reflects the different stages that it went through.) Each card contained an identifying number, the Chinese headword, its pronunciation, the various definitions and translations (drawn from seven or eight dictionaries), sometimes accompanied by example sentences (Raguin 1956: 262; 1995: 7).

The team was at work from 8am to noon and then again from 2:30pm to 6pm. During these hours, “everyone remain[ed] at their post, like workers at their station along an assembly line.” Consulting the local collaborators—who were knowledgeable in both their own language and in at least one European language—the team members made further changes to the lemmata list. “Central desk” (Bureau central) created the cards with headwords, which were then sent to the French team in boxes containing up to one thousand cards each (Raguin 1956: 262). French had been the dominant language within the Jesuit mission to China in the late Qing and Republican period, and its status was apparently maintained in Taichung. The editors working on the translation into the other European languages used the French translation as their basis, while also carrying out a kind of peer review of the work of the French group (“Notice” 1953: 566). “Battles that could go on for several days” sometimes resulted from this work of mutual criticism, with the local Taiwanese and Chinese assistants often placed in the uncomfortable position of arbiters (Raguin 1995: 7).

At this stage, in 1954, the team appears to have already been working on the basis of typed sheets as opposed to index cards, so additions and deletions were made by means of scissors and glue. Five sets of typescripts, one for each language, were produced and bound into identical notebooks. By means of a spirit duplicator (polycopieuse), extra copies were made and sent to the Philippines, France, and the United States (Raguin 1995: 7-8). This duplicated American set is not the origin of the Ricci Institute materials, which, as mentioned, come from a set that remained in Taiwan and only much later was moved to San Francisco and then Boston.

The dictionary continued to be revised in the late 1950s. Morohashi Tetsuji’s 諸橋轍次 (1883–1982) monumental Dai Kan-Wa jiten 大漢和辞典 (Unabridged Chinese-Japanese dictionary), published in 1955–59 (and later revised), was a key reference in this process. Goyoaga, unable to secure third-party funding for the project, wanted the lexicographers to quickly bring it to a close, but Raguin insisted that it needed several more years. In the early 1960s, when Raguin spent most of his time teaching in south Vietnam, the polyglot was put on the back burner, with only a skeleton crew remaining in Taichung. Carroll died unexpectedly while visiting Hong Kong in 1964, which was an important loss for the project. In the same year, however, Raguin returned permanently from Vietnam and the French section of the dictionary was brought to what he considered to be a publishable state (Raguin 2000: 57-58).

It was decided to publish an abridged Chinese-French dictionary first. This task was the express focus of the Ricci Institute that was established in Taipei in 1966. Among the news bulletins and newspaper clippings related to the China and Taiwan Jesuits that Francis Rouleau, S.J. (Hu Tianlong 胡天龍, 1900–1984) collected (now held by the Library), two announcements of the Taipei Ricci Institute’s recent establishment are found, both of them highlighting the dictionary as its most urgent project.

The abridged Chinese-French dictionary was indeed published, as was an abridged Spanish-Chinese dictionary. A Chinese-Hungarian dictionary based on Zsámár’s contribution to the project was envisioned but never appeared. In the mid-1980s, work recommenced on the unabridged Chinese-French dictionary, which was completed as a collaborative research project in France rather than Taiwan. In 2001, it was published as the Grand dictionnaire Ricci de la langue chinoise, which differs substantially from the 1950s drafts.

Additions and corrections have been carried out in manuscript. Our set contains primarily Chinese-English material, with some being Chinese-French and some Chinese-Latin. Some of it is dated. When we moved it into boxes in the spring of 2025, we found volumes dated 1956 and 1958. Several volumes in addition contains the names of the lexicographer who worked on them. They include Gerald Pope, S.J. (Bao Jinghua 寶靖華, 1908–1980), James Enda Thornton (Tao Yagu 陶雅谷, 1910–1993, a photocopy of whose two-page typescript autobiography is also held at the Library as part of the Rouleau-Malatesta archive), and Richard B. Meagher (Mi Xixian 彌希賢, 1909–1989).

Research possibilities

Our set holds great potential as a source for the history of Chinese lexicography. Not only are different stages in the project’s development discernable from the manuscript annotations, but the timing of the production of the Taichung polyglot is also particularly intriguing. The Jesuit project in fact coincided with another great abortive undertaking of Chinese-foreign lexicography.

This other project was housed at the Harvard-Yenching Institute in Cambridge, Mass. As it turns out, Raguin had some knowledge of this project some eight years before he took charge of the Taichung polyglot. In 1946, Raguin arrived at Harvard to study toward a PhD, but only stayed for two years before events in China compelled him to leave for Shanghai without having finished his degree. While in Cambridge, however, he learned that the director of the Institute was looking for funding for a Chinese-English dictionary project that had already yielded a dataset of a full sixteen (Raguin writes seventeen) Chinese dictionaries cut up and pasted onto index cards (Raguin 1995: 3-4).

William Hung (Hong Ye 洪業, 1893–1980) and Nieh Ch’ung-ch’i (Nie Chongqi 聂崇岐, 1903–1962) had initiated this project in 1936 in Peiping, which was then the location of the American-funded Yenching University. The influential Chinese-American linguist Yuen Ren Chao (Zhao Yuanren 趙元任, 1892–1982), who, I believe, was at Harvard during the war years, was put in charge of the project in 1942. The following year, Chao published a specimen of what the dictionary might look like. Chao left the project in 1946, but the bibliographer Achilles Fang (Fang Zhitong 方志浵, 1910–1995) joined it, as did, for two years, the historical linguist Li Fang-kuei (Li Fanggui 李方桂, 1902–1987). Hung took charge again in 1948. Several other American sinologists contributed to the project, including the Mongolist Francis Woodman Cleaves (1911–1995). A 68-page specimen of the dictionary appeared in book form in 1953. The project aspired to create an exhaustive etymological dictionary for Chinese, in which definitions would be supported by quotations from primary sources (Harvard-Yenching Institute 1953; Chen 1955). The two projects—one in Macao and Taiwan, one in Massachusetts—advanced simultaneously, but they were very different. In an article on the Taichung polyglot project written in September 1955, Raguin even explicitly contrasted the Jesuits’ approach to that of the Harvard-Yenching group (Raguin 1956: 264).

I do not know what happened to the Harvard-Yenching project after 1953. Is the archive left behind by this project still held somewhere in the bowels of the Harvard-Yenching Library? I do not know, but since the Ricci Institute is a mere bus ride away from Harvard, a more enterprising scholar might in the future consider a joint investigation of the Taichung polyglot and the Chinese-English dictionary project.

By Mårten Söderblom Saarela, Special Collections Librarian.

With thanks to Anthony E. Clark, Joseph W. Ho, Albert Hoffstädt, Mark Mir, Yuzhou Bai, and the Boston College interlibrary loan staff.

Note: The black-and-white photos in this essay were taken by Frederic J. Foley (Fu Liangpu 傅良圃, b. 1917). They were scanned from contact prints.

September 19, 2025. Updated February 2, 2026.

References

Bernard, Henri. 1927. “Bibliographie méthodique des œuvres du père Léon Wieger.” T’oung Pao. 2nd ser. 25, no. 3/4: 333-45.

Chen, Shih-hsiang. 1955. Review of Chinese-English Dictionary Project, Fascicle 39.0.1: Preliminary Print, by the Harvard-Yenching Institute. The Far Eastern Quarterly, 14.3: 395-402.

“Five Dictionaries Due Soon in China: Priest Back from Orient Tells of Monumental Project of Scholars in Formosa.” 1953. The New York Times, November 8: 115.

Grand dictionnaire Ricci de la langue chinoise. 2001. Compiled by the Institut Ricci (Paris) and the Taipei Ricci Institute. 7 vols. Paris & Taipei: Desclée de Brouwer & Institut Ricci. PL1459.F8 G83 2001.

Harvard-Yenching Institute. 1953. Chinese-English Dictionary Project, Fascicle 39.0.1: Preliminary Print. Cambridge, Mass.: Harvard University Press.

[Jesuits in China. China general. 1923-1967]. Archival collection consisting of printed, typescript, and manuscript materials in folders. Held at the Library of the Ricci Institute, Boston College, with the call-number BX3743.J47 1923.

Klöter, Henning. 2019. “China from c. 1700.” In The Cambridge World History of Lexicography, 317-339. Edited by John Considine. Cambridge: Cambridge University Press.

Mateos, Fernando. 1995. China Jesuits in East Asia: Starting from Zero, 1949 1957. Taipei: s.n. BV3415.2.M47 1995.

Meynard, Thierry. 2006. Following the Footsteps of the Jesuits in Beijing: A Guide to Sites of Jesuit Work and Influence in Beijing. St. Louis: Institute of Jesuit Sources. DS795.7.M497 2006.

Motte, Joseph. 1953. “Un travail de Bénédictin: le Dictionnaire Chinois-Polyglotte.” Chine Madagascar, no. 38: 9-14.

“Notice sur une Série de Dictionnaires Chinois.” 1953. China Missionary Bulletin 5, no. 6: 565-66.

Raguin, Yves. 1956. “Une grande entreprise lexicographique: la collection de dictionnaires chinois des pères jésuites de Taichung (Formose).” Études: revue fondée en 1856 par des Pères de la Compagnie de Jésus, January: 261-67.

Raguin, Yves. 1995. “L’aventure du Grand Dictionnaire Ricci de la Langue Chinoise, la fondation des trois Instituts Ricci et de l’Association internationale Ricci pour les études chinoises.” Paper presented at the VIIIe Colloque International de Sinologie de Chantilly, 3-6 September.

Raguin, Yves. 2000. “The History of the ‘Grand dictionnaire Ricci de la langue chinoise’.” The Ricci Bulletin, no. 3: 53-64.

Wright, Arthur F. 1947. “Sinology in Peiping 1941–1945.” Harvard Journal of Asiatic Studies 9, no. 3/4: 315-372.