De l'un au multiple
| ,IV. Les aléas des textes
On Mengxi Bitan’s world of marginalities and “south-pointing needles”. Fragment translation vs. contextual translation
A propos de monde de la marginalité dans le Mengxi Bitan et des « aiguilles pointant vers le sud »
Résumé
Pour interpréter et traduire correctement la notice du Mengxi Bitan présentée par les historiens occidentaux des sciences comme un exposé de « la découverte de la déclinaison magnétique », il est nécessaire d’abord d’analyser la catégorie dans laquelle elle est insérée dans le texte même. Cette rubrique, Zazhi, qu’on a traduit par « notes diverses », signifie plutôt pour Shen Gua, l’auteur du Mengxi Bitan, « notes sur des faits marginaux ou anormaux » (littéralement non orthodoxes). Dans cette perspective, la notice en question concerne une curiosité naturelle (la déviation fréquente de l’aiguille vers l’est) et une technique pour obtenir la direction exacte du sud, nullement la « découverte » d’un principe fondamental. La traduction donnée ici diffère profondément de celle de Wylie, qui est citée. En outre, il n’y a aucune raison de considérer la description d’un procédé de correction comme une partie accessoire et sans importance de la note.
Cette utilisation, par Joseph Needham, de traductions fragmentaires pour démontrer des priorités chinoises – ici, la découverte par la Chine de la déclinaison magnétique avant l’Europe – illustre bien la manière dont les historiens des sciences « imposent » les taxonomies de la science moderne sur des textes dont la logique propre n’est pas prise en considération. Dans cette opération, l’intérêt réel de ces textes est complètement oblitéré : on ne fait pas l’histoire de la science chinoise telle qu’elle fut.
Cette imposition de taxonomies étrangères prend deux formes complémentaires : des récits grandioses de l’histoire de découvertes chinoises et des tableaux réorganisant des fragments d’origine et d’époques diverses. D.W. Fu présente ici une analyse critique des tableaux présentés par Wylie (1897) et Needham (1962) à l’appui de leurs interprétations de la découverte de la déclinaison magnétique en Chine.
Note de l’auteur
An early extended abstract of this paper was read as the keynote speech at the colloqium “Translation from Chinese into European Languages”, held at the École des hautes études en sciences sociales, Paris (June 1994). The title for that extended abstract was “Imposition of Taxonomy, or the Translation of ‘Scientific Fragments’ from Ancient Chinese Texts into Some European Languages in the Modem Era”. An early version of my discussion on the translations for Shen Gua’s famous jotting on “south-pointing needles” (zhinan zhen 
), in section II of this paper, was read and debated in the “Text Seminar” of the Needham Research Institute, Cambridge (April 1995). I am very grateful for the valuable and stimulating comments on my drafts on these two occasions. Many thanks also to Ling-fang Cheng, who helped to polish the English of this paper.
Texte intégral
- 1 . The interesting part of this “priority” is that usually the more we think of the “scientific valu (...)
1There is a peculiar tradition of translation in Chinese history of science, which we might call the tradition of fragment translation. Translations are often made by a highly selective process wherein fragments deemed “scientific” are selected from ancient texts, usually with no regard for the wider textual unity, context, and taxonomy in which those fragments are embedded. The fragment translations are then reorganized and reinterpreted in a new, Western taxonomy and narrative. Thus, the process of fragment translation plus the additional reorganization and reinterpretation constitute a social process of translation prevalent in the traditional field of Chinese history of science. In contrast to fragment translation, I propose a different kind of translation which we might call contextual translation. The basic unit of text for contextual translation should be a certain meaning unit, rather than a meaning fragment, from the original ancient text. A single jotting is naturally the basic kind of meaning unit in an ordinary historical text. The process of translating a text can begin only after we have become very familiar with the historical context of the text in question. The (modern) “scientific” value of the text has only a secondary place. The historical context of a text takes priority, in translation, over the modem scientific value of that text.1 Finally, the various results of the translation process should be organized and distributed strictly according to the historical taxonomy (historical “order of things”) in which the original texts are situated. The disciplinary boundaries of modem science have no place in contextual translation. Usually, we say serious translation creates deeper under-standing, but it is through contextual translation only that we can begin the process of understanding.
I
Mengxi Bitan3 contains 609 jottings with an internal conceptual structure of seventeen categories (men ). The position of the category of Zazhi is located almost at the end (the 16th) of a series of the seventeen categories: before the category “Deliberations on Materia Medica” (yaoyi ), and after the category “Wit and Satire” (jixue ). The position itself is not special, as in many other similar medieval Chinese jottings. But what is the meaning of Zazhi? Traditionally we understand it as “miscellaneous notes” – a group of notes very different from each other, or notes of mixed or various kinds. If we break the internal structure of the seventeen categories of Mengxi and randomly redistribute the 609 jottings, we can rightly call Mengxi a work of “miscellaneous notes”. If, on the other hand, we look into the similarity among the jottings within the category of Zazhi and compare this similarity with the internal similarities of other categories in Mengxi, we can hardly justify translating Zazhi as “miscellaneous notes”. In Chinese dictionaries, one important meaning of the character Za is certainly “mixture” or “miscellany”, which is frequently used throughout ancient Chinese texts. There is, however, another sense of this word, especially when used as zazhuan or in contrast with . This familiar contrast indicates a contrast in terms of “margins vs. the center”, “ille-gitimate vs. legitimate”, “residual vs. categorical” or “unorthodox vs. orthodox”. In this sense, the focus is not on a neutral or innocent “mixture”, but on a clearly unbalanced power relationship. In contrast with zheng , zhong and chun , the word za in this sense denotes an immense marginal space of residuals, things forgotten, illegitimate, unorthodox, and, in general, things hidden in shadow and darkness. Although Zazhi formally is a category at the end4 of Mengxi, it is actually a residual category, the category of marginality.Zazhi as “notes on marginalities and unorthodoxies”. First, there is the main theme of geographical (hence often political) marginality in Zazhi. The events and things described in the great majority of jottings of Zazhi happened and were located in the geographical margins of Northern Song. The places often mentioned (in the total of 71 jottings) such as
were usually marginal or actually barbarous foreign lands with respect to the political center.5 In general, things or events in Zazhi are from the margins, the outside, the “other” of the orthodox Northern Song. An interesting problem of this theme is that Zazhi contains jottings about some important cities in the traditional “south of the river” (Jiangnan ) or around the canal Bian . They are not clearly “marginal” in Northern Song territory, and a small number of jottings are about things which happened there. However, I find the things described there (old monuments, Tang customs, tombs) are all of a reminiscent nature,6 especially with respect to the long past high culture of Tang . Shen Gua did not really pay attention to Northern Song Jiangnan. If we furthermore compare the cultural significance of Jiangnan in Northern Song with that Jiangnan in Tang or in Southern Song, the answer is also clear. Northern Song Jiangnan was not very attractive. Therefore, the cultural (geographical) meaning of them in Zazhi can be considered as somewhat “marginal”.Mengxi, many “important” subjects were distributed in other, more noble categories, such as philological criticism (bianzheng ), music and mathematical harmonics (yuelü ), numerological regularities underlying phenomena (xiangshu ), literati affairs (renshi ), civil service (guanzheng ), and literature, calligraphy, painting and the like. Although some military and map-making techniques are discussed in Zazhi, they are basically concerned with how the power center can effec-tively deal with the troubles emerging in the margins. Topics in Zazhi, often of great interest to modern science, like “stone oil” (shiyou ), strong wind, astonishing barbarians, “jumping rabbits” (tiaotu ), or other curious birds, bugs, teas, vegetables, fish, snakes, stretches of water, plants, or the many strange/unheard of geological formations, needle deviations, odd diseases, folk techniques/medicine dealing with things from the outside world, were really topics culturally marginal to the center of Northern Song. With an astute and systematic gaze7 from the power center, Shen Gua nevertheless made excellent observation on them.human marginality in Zazhi can also be briefly discussed here.8 In general, the people and the stories told about them here are usually about marginal people in a marginal place. Although some anecdotes on literati were included in Zazhi, these were typically officers of lower rank, who were often former rebels, thieves, or barbarians that had been tamed and assimilated into the bureaucracy; or they were decadent or idiosyncratic literati,9 or people who refused to submit to the imperial order10 (), or others such as a greedy apprentice (), a mean servant (), an obscure local monk doctor (), a nameless alchemist who made a “water-pellet” (), “boat merchants” who knew the winds (), mystical geomancers and their needle secrets (), a good rebel loved by the local people (), an old prostitute who helped to defend Song territory (), or even a lost ghost who, as a man, failed in the official examinations but now, as a ghost, is eager to serve the emperor (). All these marginal people,11 though sometimes ridiculous and not approvable, were nonetheless tame and occasionally were of some service to the Song, and thus deserve mention or can serve as a lesson here. Sometimes they are simply the kind of small thieves who do things by mimicking cock-crowing and dog-barking ().12 On the other hand, Shen Gua’s marginal world in Zazhi teems with monstrous, blood-thirsty, lustful barbarians wandering outside, and penetrating into the territory of the orthodox Northern Song. These barbarian taies, however, usually end on a triumphant note when the Song army finally defeats the barbarous invasions, conquers and takes the foreign lands. Although the jottings in Zazhi concerning these barbarous affairs need further analysis, a general picture seems to emerge. While the marginal Song people were tame or could be tamed, the barbarous marginal people should be defeated, conquered, and then finally be tamed. Therefore, as I stated above, the category Zazhi refers to a marginal and unorthodox world dominated by a clearly unbalanced power relationship.
II
“has not only the undeniably earliest clear description of the magnetic needle compass in any language, but also a clear statement of the magnetic declination”. Needham also recalled, in a footnote on the same page: “I shall never forget the excitement which I experienced when I first read these words. If any one text stimulated the writing of this book more than any other, this was it.” A typical modem “stimulation” for writing Chinese history of science, perhaps. But behind the sinological history of jotting 437, there is a genealogy of sinologists and Chinese historians of science who tell (and debate) the stories of “the Chinese priority in discovering south-pointing needles and magnetic declination”: from Klaproth’s original story (Klaproth 1834), Wylie’s summary (1859, 1897) of Chinese records of magnetic declination, especially on the mysterious record in the 8th century by the monk Yixing , Edkins’ interpretation (Edkins 1877) of the structure of the geomantic compass () as knowledge of magnetic declination in its fossile form, to Hirth’s wonderful story of the “Origin of the Mariner’s Compass in China” in his Ancient History of China (1908), Moule’s definite separation (1924) of “south-pointing carriage” () from south-pointing needles, Mitchell’s challenge (Mitchell 1937) of the standard interpretation of jotting 437 in , via Wang Zhenduo’s establishment of the connection between the magnetic compass and the diviner’s board of the Han people (Wang Zhenduo 1948), Li Shuhua’s excellent Chinese summary (Li Shuhua 1953) of the current state of the art reclaiming the Chinese prior discovery on the mariner’s compass, to Need-ham’s grand summary (1962)13 of the current state of the field with important further studies on the relevance of geomantic literatures.
2Finally, here is the text of jotting 437:
3Compare Wylie’s translation (Wylie 1859):
Scientific people rub the point of a needle with load-stone, which imparts to it the property of pointing to the south, but it always declines slightly to the east, and is not direct south. When it floats on the surface of water, there is too much oscillation. The finger nail or the rim of a cup may be used to balance it on, when it will turn more freely; but being hard and smooth, it is liable to fall. The best plan is to suspend it by a silk fibre, thus – take a single-cocoon fibre of new silk, and attach it to the centre of the needle by a piece of wax the size of a mustard seed; then suspend it free from any current of air, then the needle will always point southward. Some are magnetized so as to point to the north. I have in my possession, both north and south pointing needles. The southern direction of the load-stone, like the Western direction of the cypress tree, is a principle hitherto unexplained.
- 16 . One significant difference between Wylie’s and Needham’s translations is that the term fangjia w (...)
4and Needham’s translation16 (1962):
Magicians rub the point of a needle with the lodestone; then it is able to point to the south. But it always inclines slightly to the east, and does not point directly at the south. (It may be made to) float on the surface of water, but it is then rather unsteady. It may be balanced on the finger-nail, or on the rim of a cup, where it can be made to turn more easily, but these supports being hard and smooth, it is liable to fall off. It is best to suspend it by a single cocoon fibre of new silk attached to the centre of the needle by a piece of wax the size of a mustard-seed – then, hanging in a windless place, it will always point to the south. Among such needles there are some which, after being rubbed, point to the north. I have needles of both kinds by me. The south-pointing property of the lodestone is like the habit of cypress-trees of always pointing to the west. No one can explain the principles of these things.
5with my own translation from a different perspective (italics are mine):
“its method is:”) to set the needle right into the southern direction. Historians of science used to take this contrivance as a way to resolve only the problems of its being “unsteady” (in water) or “liable to fall off” (of the finger-nail or the rim of a cup). They divided jotting 437 into two unrelated parts: (1) Shen’s “discovery” of magnetic declination (), and (2) his method for solving other minor problems. I disagree with this practice of dividing a jotting by way of what is or is not important to modem science,22 and I would like to stress the importance of reading the historical meaning of a jotting as a single, indivisible unit, very similar to the way Shen Gua himself treated it. As I indicated above in my translation and the detailed notes to it, we must bear in mind that Shen Gua had no modem conception of “magnetic declination”. Therefore, due to the crude apparatus used by Shen, the presumed maginalities in the category of Zazhi and the geographical variations of the actual magnetic field which effected his needles, it is probable that he noted this strange phenomenon from time to time and saw it more or less in terms of “frequency”, that is how often he saw or did not see clear deviations, at this place or that place, how much it deviated this time or that time, and so on. Obviously he did not know or see it as some kind of natural regularity,23 but, being an astute natural observer, Shen Gua had to notice it from time to time and thus had to consider it as an anomaly, not merely as an accident. This interpretation is not the same as Mitchell’s famous blunt comment in 1937, that “the deviation from the meridian is not regarded as a separate physical phenomenon but [as believed by Shen] as an accidental resuit due to imperfect support or suspension of the magnetised needle.”24 Although Mitchell saw clearly that the deviation must not be regarded as a separate phenomenon, his use of “imperfect support or suspension” as the definite cause to explain the often anomalous deviations only to the east (why not sometimes to the west?) is not convincing. True, Shen Gua wondered about possible causes such as “imperfect supports”; these no doubt interfere with the needle, but it is unlikely that he would consider them the definite causes of deviation to the east. As far as I can see, Shen must have felt the anomalous (not accidental) nature of this deviation to the east as something strange that he did not understand, but he provided a method (partly according to his common sense as hinted by Mitchell) anyway and corrected or greatly reduced the deviations, at least to his own satisfaction.25 Thus, if Shen found a problem in the frequency of “60%” of the cases of slight deviation to the east, then he might have felt self-satisfied with the much lower frequency of only “30%” of anomalous cases after applying his method to the needle (“Then the needle would most often point to the south!”).
- 17 . I translate the adverb chang as “usually” or “often”, not as “always”, as in almost all the old (...)
- 18 . is an ambiguous phrase. There are two ways to parse this phrase: . The first way corresponds to (...)
- 19 . I use this transitional “doubt” in parentheses to somewhat smooth the flow of Shen Gua’s sentence (...)
- 20 . The phrase is an important marker in many of the jottings in Mengxi, which usually points to the (...)
- 21 . It is important to appreciate the natural flow of writing to this point: . I wish to stress the t (...)
Geomancers rub the point of a needle with a lodestone; then it is able to point to the south. But it oftenlusually17 deviates slightly to the east and not always to the south [or: not show due south]18. (Could it be the following reasons which cause the problem of deviation?)19 When it floats on the surface of water, it is rather unsteady. The finger-nail or the rim of a cup may be used to balance it on, where it will turn more easily, but these supports being hard and smooth, it is liable to fall off. It is best to suspend it by a silk fibre. Its method is20: to suspend it by a single cocoon fibre of new silk attached to the centre of the needle by a piece of wax the size of a mustard-seed, hanging in a windless place. Then the needle would most often point to the south.21 Among such needles there are some which, after being rubbed, point to the north. I have needles of both kinds by me. The south-pointing property of the lodestone is like that of cypress-trees of pointing to the west. No one can understand these things.
6The related jotting 588, also concerning lodestone and needles, is situated in the category of Yaoyi. It may be noted that most ancient Chinese conceptions of “lodestone” (cishi
- 26 . This is an adapted translation from Needham’s (Needham 1962: 250), with an important difference, (...)
When the point of a needle is rubbed with the lodestone, then the sharp end often26 points south, but some needles point to the north. I suppose that the nature of the stones is different. Likewise, at the summer solstice the deer shed their horns, and at the winter solstice the elks do so. Since the south and the north are opposites, there must be a fundamental difference between them. This has not yet been investigated deeply enough.
- 27 . When it comes to Kou Zongshi’s Bencao yanyi of the Southern Song, perhaps the phenomenon of “dev (...)
7Here, Shen Gua is dealing with the two opposites “the south and the north” and their Chinese medical analogies with two other opposites “the summer and winter solstices”, but he does not mention at all his famous remark on “deviations to the east”, as he does in jotting 437. Implicitely, this significant absence has some implications: (1) that the anomalous “deviations” were not medically important to Shen, not to mention that they must have seemed somehow correctable to him through his “method” proposed in jotting 437, and (2) that a more theoretical explanation of these unstable, anomalous “deviations” in terms of medical theories/resources was not really necessary for Shen Gua.27
III
8In this section, I shall start considering a wider area of problems which, although still concentrating on the south-pointing needle, will briefly cover a history of sinological research on this topic. I will also reflect upon some sinological practices in the translation and interpretation of Chinese history of science. As I indicated before, we have a history of “south-pointing needles in China” by way of fragment translation — especially concerning the property of “magnetic declination”, when and where it was “discovered” in China, and how many years earlier than in Europe. Now let us get into some of its details.
9Firstly I quote here some interesting statements on translation by Needham, in a reflective mood, in some prefaces to his famous Science and Civilization in China.
Though all translators may be traitors, the duty of translation has at least the merit that it forces decisions about meanings, provisional though they may have to be (Needham 1959: xiv).
China to Europeans has been like the moon, always showing the same face – a myriad of peasant-farmers, a scattering of artists and recluses, an urban minority of scholars, mandarins and shopkeepers. Thus do civilizations acquire “stereotypes” of one another. Now, raised upon the wings of the space-ship of linguistic resource and riding the rocket of technical understanding (to use an Arabie trope), we intend to see what is on the other side of the dise, and to meet the physicists and engineers, the shipwrights and the metallurgists of China’s three-thousand-year-old culture. (Needham 1962: xxviii-xxix.)
10Very vividly pictured indeed. But, even as an historian of science most sympathetic to China, Needham is still willing to use the metaphor of the “moon” to characterize the cultural subject (China) in his own work. Actually, whether stereotypes such as this or other, truer faces of China, all these are faces of a European satellite. Needham thus inevitably held a Eurocentric viewpoint, with its entire conceptual apparatus and vehicles used to see the “physicists and engineers” on the other side of the three-thousand-year-old satellite of Europe.
11Secondly, there are two important characteristics of fragment translation. The first is the imposition of taxonomy. The taxonomy or the conceptual grid of modem Western science is often taken as the basis from which to approach ancient Chinese texts. Usually it is the imposed foreign taxonomy itself that dictates where, when, which fragments in the texts are translated, and even how. After the first step of fragment translation, all these translated fragments will be reorganized and placed into the proper boxes within the imposed foreign taxonomy. The process of reorganization is further mediated by a special kind of narrative in order to tell “a great story” of a certain Chinese discovery, usually in a form such as: some passages in some ancient text were found and after painstaking readings these show that a certain Chinese already discovered something which is actually a certain scientific fact, law, or theory (through a direct translation in terms of the Western modem science terminology).
- 30 . When later modem Chinese historians of science learned enough from these Western hunters, they be (...)
- 31 . This is a typical metaphor used by F. Hirth in the construction of his table (or “abstract of dat (...)
- 32 . In a process of contextual translation, this is precisely what I am trying to do: to recover or r (...)
12The second characteristic of fragment translation might be called looking for gold or treasure hunting. Usually the narrator is clearly a Western prospector30 in the (adventurous?) field of ancient Chinese texts. The story often begins with a clue, a legend or a surprise in which an important “scientific” secret might be buried or hidden. Some books or maps leading to that secret, seen or described by a certain eminent sinologist in the past, but apparently now “lost”, may also increase the attraction of the original legend and constitute further objects to “hunt down”.31 The narrator is not interested so much in the subjectivity of Chinese cultural taxonomy or the historical context of the clue (e.g., geomancy), as in the instrumental value of geomantic fragments for digging up evidence for magnetic declination in China. Through care-fully selected geomantic fragments, we may see Chinese “references” to magnetic declination, but the “meaning” of Chinese geomancy and the “context” of those geomantic fragments is not made any clearer,32 and so they can simply be dismissed in passing as “pseudo-sciences”. The hunt usually ends with a constructed table showing as evidence fragment translations collected from various times and places among the vast field of ancient Chinese texts. Regardless of the drastic cultural and historical differences by which those fragments are actually divided, “evidence” thus collected is treated as coherent and homogeneous. Together the fragments constitute the proof of the existence of “knowledge of magnetic declination” in China. Whereas tables of “declination observed” were enthusiastically constructed by A. Wylie (1897: 157), F. Hirth (1908: 135), Wang Zhenduo (1950: 212), and J. Needham (1962: 310), I shall now address the question of whether these are really historical tables or instead the “imaginings” of modern science’s own history? (For Wylie and Needham’s tables, see tables 1 and 2.)
Wylie 1897, III: 157.
2. Needham’s Table
Needham 1962: 310. The last lines of the table, not being relevant for the present discussion, have not been reproduced.
- 33 . I exclude the observation (“c. +720”) allegedly made by the famous monk Yi-xing, since there is n (...)
- 34 . We might ask, why “compass observations” by Western Jesuits in China should be considered “Chines (...)
13Let me give a detailed example about imaginings of this modem science of its own history. As an example, I take a table constructed by the best and most respected historian of Chinese science in the long tradition of, if I may say, fragment translation. Let us consider the underlying drastic differences in cultural and historical context of the many entries in Needham’s constructed table of Chinese compass observations (Needham 1962: 310) (fig. 2). The title of this table reads “Details on Chinese compass observations including magnetic declination, mainly before +1500”. This is a carefully chosen, interesting title. In the sixteen real entries33 (or observations) of this table, only the last six observations were made by people who knew anything about “magnetic declination”, and these were mostly European Jesuits. The remaining ten were made by people who can socially be divided into two groups: geomancers and literati. None of them knew anything of magnetic declination in the modem sense. Moreover, in most cases, these observers (geomancers, literati, Jesuits) did not communicate with each other, did not share the same conceptions of “deviations”, or live in the same historical period. In fact, many of them were puzzled and had expressed anxiety about these anomalous marginal events, and often did not consider these events as constant and regular phenomena. What sort of “table” is it then, that groups together these radically heterogeneous events and observations? In Needham’s title, “including”, I think, indicates a sub-set of all “Chinese compass observations”,34 which specifically mentioned or at least referred to magnetic declination. Thus, only an outside “other”, who has little historical interest in the Chinese meaning and context of these observations but who knows (from the Western scientific perspective) what these observations were “really about”, made the construction of this table possible. Following this tradition of table constructions by Wylie, Hirth, Wang, and Needham, treasure-hunting sinologists would continue the expansion of this very table, insert confirmed new entries, delete confirmed mistakes, and so on. But does this whole process help us understand more of the history of Chinese science (not the Chinese history of science)?
- 42 . Take the case of Chu Yong’s in the entry “c. +1230”, where Needham refers to the passage of “Dis (...)
14Finally, because of the constraints of space, I shall leave other controversial entries undiscussed.42 Let us now return to the entire table itself. Some moments ago, when we began considering this table, I asked what sort of “table” would group together these radically heterogeneous observations? Then I answered: only an outside “other”, who has little historical interest in the Chinese meaning and context of these observations, but knows (from the Western scientific perspective) what these observations were “really” about, could have constructed such a table. Whenever I think of this outside other who “knows”, a picture inevitably springs to my mind about the whole situation involved, and that is a picture of the Chinese tale about “a group of blind people trying to figure out what an elephant is like” (xiazi moxiang
Bibliographie
Les utilisateurs des institutions qui sont abonnées à un des programmes freemium d'OpenEdition peuvent télécharger les références bibliographiques pour lequelles Bilbo a trouvé un DOI.
Si vous souhaitez que votre institution souscrive à l'un des programmes freemium d'OpenEdition et bénéficie de ses services, écrivez à : contact@openedition.org
Bibliography
Cao Wanru 
1980. “Lun Shen Gua zai dituxue fangmian de gong-xian” 
, Kejishi wenji, 3: 81-83. Shanghai, Shanghai kexue jishu chubanshe.
Edkins, Joseph. 1877. “Chinese Names for Boats and Boat Gear; with Remarks on the Chinese Use of the Mariner’s Compass”, Journal of the North China Branch of the Royal Asiatic Society, 11: 123-142.
Fu Daiwie 
. 1993-94. “A Contextual and Taxonomic Study of the ‘Divine Marvels’ and ‘Strange Occurrences’ in the Mengxi Bitan”, Chinese Science, 11: 3-35.
– 1995a. “Higher Taxonomy and Higher Incommensurability”, Studies in History and Philosophy of Science, 26, 2: 273-294.
– 1995b. “The Needle Deviation (‘Magnetic Declination’): Looking for its Possible Influences in Some Early Medieval Texts of Geomancy.” (Lecture delivered at the College de France, Paris.)
“Guanshi dili zhimeng” 
, in Gujin tushu jicheng 
, Yishu dian 
, juan 655. Taipei, Ding-wen Publishing Company.
Guo Wenhua 
. 1996. “Shen Gua ‘yixue’ yanjiu ji ‘yaoxing’ chutan” 
. (Manuscript written in Daiwie Fu’s seminar on Mengxi Bitan at Tsing-Hua University.)
Hirth, Friedrich. 1908. Ancient History of China; to the End of the Chou Dynasty. New York, AMS Press.
Klaproth, Julius. 1834. Lettre à M. le Baron A. de Humboldt, sur l’invention de la boussole. Paris, Dondey-Dupré.
Kou Zongshi 
. 1115. Bencao yanyi 
(Southern Song Edition). Taibei, Yi-wen Publishing Company. (From Baibu congshu jicheng 
; reprinted Taibei, Shijie shuju, 1962.)
Lei Hsianglin 
and Daiwie Fu. 1993. “Language and Similarity in the Dream Brook – A Study of Prognostication, Divine Oddities, and Strange Events in Mengxi Bitan”, Tsing Hua Journal of Chinese Studies, 23, 1.
Li Shuhua 
. 1953. “Zhinan zhen de qiyuan” 
[Parts 1 and 2], Dalu zazhi 
, 7, 9: 1-7 and 7, 10: 1-10.
Si vous souhaitez que votre institution souscrive à l'un des programmes freemium d'OpenEdition et bénéficie de ses services, écrivez à : contact@openedition.org
Mitchell, A. Crichton. 1937. “Chapters in the History of Terrestrial Magnetism”, chapter II: “The Discovery of Declination”, Terrestrial Magnetism and Atmospheric Electricity, 42: 241-280.
DOI : 10.1029/TE051i003p00323
Si vous souhaitez que votre institution souscrive à l'un des programmes freemium d'OpenEdition et bénéficie de ses services, écrivez à : contact@openedition.org
Moule, A.C. 1924. “The Chinese South-Pointing Carriage”, T’oung Pao, 23: 83-98.
DOI : 10.1163/156853224X00103
Needham, Joseph. 1959. Science and Civilization in China [SCC], vol.3: Mathematics and the Sciences of the Heavens and the Earth. Cambridge, Cambridge University Press.
– 1962. Science and Civilization in China [SCC], vol.4, part I: Physics, section 26. Cambridge, Cambridge University Press.
Shen Gua 
. 1962. Mengxi Bitan jiaozheng 
. Shanghai, Zhonghua shuju. 2 vols. (Edited by Hu Daojing 
on the basis of the Northern Song standard edition of 1166, with the standard numerical ordering of the total 609 jottings.)
Wang Zhenduo 
. 1948/50. “Sinan zhinan zhen yu luopan jing” 
[part I] (The lodestone Spoon of the Han) and [part II] (The ‘Fish’ Compass, the Needle Compass, and the Early Work on Deviation), Zhongguo kaoguxue bao, 1948, 3: 119 ff. and 1950, 4: 185 ff.
Wylie, Arthur. 1859. “The Magnetic Compass in China”, North-China Herald, March 15, 1859. (Included in Wylie 1897, III: 155-157.)
– 1897. Chinese Researches. Shanghai/Taipei, Chengwen Publishing Company.
Notes
1 . The interesting part of this “priority” is that usually the more we think of the “scientific value” of a certain ancient passage, the more likely we are to mistranslate that passage.
2 . I have made a similar study on two other catagories of Mengxi Bitan, see Daiwie Fu 1993-1994. In a similar spirit, Lei Hsianglin 
and I published another paper entitled “Language and Similarity in the Dream Brook – A Study of Prognostication, Divine Oddities, and Strange Events in Mengxi Bitan” (Lei Hsianglin and Fu Dawei 1993). For a further theoretical elaboration of this key concept of “taxonomy” in connection with the late professor Kuhn’s ideas on taxonomy, see Daiwie Fu 1995a. In the appendix of my article in Chinese Science just cited (Daiwie Fu 1993-1994), I have argued that the 17 categories of Mengxi were really Shen Gua’s own taxonomy.
3 . The edition I used for this study is based on Hu Daojing’s 
critical edition of Mengxi Bitan 
(Mengxi Bitan jiaozheng), with the standard numerical ordering of 609 jottings in total.
4 . As mentioned above, Zazhi is not the final category of Mengxi; it is the 16th. However, in a careful study of the 17th category, Yaoyi, we find that this category is probably not in the conceptual structure of Mengxi but appended, as it were, from the outside. This idea comes from a manuscript “Shen Gua ‘Yixue’ yanjiu ji ‘Yaoxing’ chutan” 
by Guo Wenhua 
, written in my seminar of Mengxi Bitan in Tsing-Hua University (Guo Wenhua 1996). This finding provides strong evidence for us to say that Zazhi is conceptually a final, residual (dis-)category of Mengxi.
5 . In the total 71 jottings of Zazhi (including those in the Supplement Bu Bitan
), I can find only two jottings concerning events that happened in the capital of Northern Song. One is jotting 466, the other is jotting 572. The interesting thing about jotting 572 is that it involves the son Mo Zhan 
of the former king of the barbarous Xiqiang 
, thus this story is indirectly connected to the barbarous foreign lands. Jotting 466, although it concerns an anecdote about one Song emperor, is very unconventional or even unorthodox. As to some other possible “exceptions”, three jottings are about military exercises (475, 578, 579) and two about the (borders) maps made for the dynasty (472,575). They are all related to the issues of how the political center is to deal with the troubles in the marginal or foreign lands.
6 . In jotting 581, Shen admired Yangzhou 
in the Tang dynasty. In jotting 458, he talks about an interesting Tang custom and its legacy in Liangzhi 
. In jotting 438, he writes about an ancient tomb in Jinling 
. In jotting 431, he compares the traces of old canals in Jianghuai 
with those described in a Tang geography book Lainanlu 
. Jotting 429 again focuses on a Tang stone monument discovered in Suzhou 
of the Bian canal.
7 . For a more systematic discussion of Shen Gua’s “gaze” from the center, see my previous study on two other categories of Mengxi (Fu Daiwie 1993-94, esp. pp. 24-33).
8 . Some ideas here come from a manuscript by Zhang Zhaowei 
(1990).
9 . 
in jotting 460 can be a decadent person, whereas 
as commented by Shen Gua in jotting 465, can be an idiosyncratic person, who regretted all his life having once bought a book, which supposedly broke his discipline of self-restraint.
10 . In jotting 470, Shen Gua writes about those people in Jiangzhou 
who were loyal holdouts of a former local lord and refused to submit to the Song imperial order for three years. Later, after their defenses cracked, they were massacred by a Song general despite an imperial pardon, which came too late. This is the fate of these people, Shen sighs. Although the leader of these people, Hu Ze 
, is also mentioned in jotting 157 on “Literati Affairs”, the emphasis there is quite different. An almost opposite story is jotting 424 in Zazhi, where two local doctors want very much to become petty officials (
), but both suddenly die after the good news arrives.
11 . Generally speaking, “servants” (shiren 
) of the literati were very active in the category Zazhi. See, for example, jottings 436, 446, 458, 470.
12 . See jotting 478, about an old prostitute who helped to defend the Song from the barbarians Xiawei 
by openly shouting at the enemies the “dirty private affairs” of their mother-king who led the invasion.
13 . Klaproth 1834; Wylie 1897; Edkins 1877; 123 ff; Hirth 1908; Moule 1924: 83 ff; Wang Zhenduo 1950, 1953; Needham 1962.
14 . Although Shen Gua was well known as a polymath who specialized in many esoteric learnings, he talked surprisingly little about “geomancy” in Mengxi. Only a very small number of jottings are related to geomancy: jottings 145, 380, and 571.
15 . The making of this map was based on a set of fixed directions and distances between any two geographical points in question, similar to a bird flying directly from one place to another with a fixed direction and distance. This is different from the old method of map-making where the distance between two points is estimated by the length of the actual road from one place to another.
16 . One significant difference between Wylie’s and Needham’s translations is that the term fangjia 
was translated by Wylie as “scientific people”, but by Needham as “magicians”.
17 . I translate the adverb chang 
as “usually” or “often”, not as “always”, as in almost all the old translations (including the earlier French ones). But this 
does not necessarily mean high frequency or high probability of occurrences; it usually depends on the context. For example, we have no way of telling from the sentence alone 
how often or how many times he came in the last year. Thus, 
may mean something much less than “often”. I have checked quite a number of uses of 
as an adverb in Mengxi, and I found that almost all of them mean “usually”, “often” or even less, which is quite different from “always” – especially in the issues discussed in this paper. It is quite understandable why historians of science in the past liked to translate it as “always”. See further explanations in the following passages.
18 . 
is an ambiguous phrase. There are two ways to parse this phrase: 
. The first way corresponds to the first, italic translation, the second to the alternative translation in brackets. It is more natural to use the first one, where 
is used in contrast to 
as just discussed, and the meaning of this phrase is consistent with other passages in this jotting. Translated the second way, 
presumably means “due south”, but Shen Gua rarely used 
in this way. For example, in jotting 127, he did not use
, and in jotting 131, he used 
instead.
19 . I use this transitional “doubt” in parentheses to somewhat smooth the flow of Shen Gua’s sentences. It may reflect a hesitation, a pause, a change of tone on his part. It is also to emphasize the unity of this jotting and to avoid mutilating it into several unrelated fragments, like many translations did in the past.
20 . The phrase 
is an important marker in many of the jottings in Mengxi, which usually points to the key ideas Shen Gua wanted to express in the jotting in question. This marker also often indicates where the source or origin of that method is, and it is usually a book. Unfortunately, he did not indicate the source of “the method” in jotting 437.
21 . It is important to appreciate the natural flow of writing to this point: 
. I wish to stress the term 
, which signals a turning point on the way to resolving to Shen Gua’s satisfaction at least, all the problems (especially the “deviating slightly to the east”) raised before this point in this jotting. Again, I translate 
as “most often”, not “always”. Historians in the past saw this jotting from the perspective of modem science, thus they translated 
as “always” to observe the imperatives of magnetic declination. But Shen Gua had no such conceptions. Most likely he considered these deviations in terms of “frequency” – hence the terms “usually”, “often” and “not always”, given the crude apparatus he had at hand, the curiosity that was only marginal to his experience, and the geographical (and hence magnetic) variations by which his south-pointing needles were effected.
22 . Many make the taxonomy or conceptual grid of modem Western science the basis from which to approach ancient Chinese texts. We might call this practice an “imposition of taxonomy”, and I will focus on it in the third section of this paper.
23 . For the distinctions among “regularity”, “anomaly” and “accident”, it is easy to draw up a series of degrees or stages of epistemological status (from high to low) concerning certain units of “knowledge” or “phenomena”:
(i) knowledge and discovery (and its “normal science structure”, in the sense of Thomas Kuhn);
(ii) recurrent phenomena, a regularity, a prototype law, a building block for normal knowledge;
(iii) anomaly (acknowledged strange phenomena happen from time to time, which resist assimilation to normal knowledge and potentially challenge or conflict with the official knowledge);
(iv) aberration (strange phenomena which can be set aside or even forgotten);
(v) practical accidents (familiar phenomena which can easily be dismissed or explained as trivial).
24 . The italics are mine. See Mitchell 1937. As to Needham’s interesting arguments against Mitchell’s unconventional idea, see Needham’s footnotes (1962: 250d, 252c, 309c). In footnote p. 252c, Needham partially concedes Mitchell’s point by saying that what was in Shen Gua’s mind “is not certain”. But his strategy of using Kou Zongshi’s attempt to “explain the magnetic declination” (thus Kou presumably would recognize the phenomena to be explained as a “regularity”) in order to refute Mitchell’s point is, to me, not convincing. In the next section, I will make a very clear distinction between Shen’s and Kou’s situations; thus we cannot use Kou’s observation in the Southern Song to interpret Shen’s in the Northern Song. And the key issue here is the nature of Shen Gua’s world. In footnote p. 250d, Needham criticizes “the spirit in which some Europeans approach Chinese texts” (such as Mitchell’s), but fortunately I am not an European but a (Han) Taiwanese.
25 . Later, in section III, I will introduce a Southern Song reading (hence with no conception of “magnetic declination”) of Mengxi’s jotting 437 to confirm my interpretation here.
26 . This is an adapted translation from Needham’s (Needham 1962: 250), with an important difference, that 
(in 
) again is translated as “often”.
27 . When it comes to Kou Zongshi’s Bencao yanyi 
of the Southern Song, perhaps the phenomenon of “deviations” had become more regular and less anomalous. Kou did mention “deviations” in his medical discussions of the lodestone 
and also recorded a strange passage in which a Taoist medical concept of “five phases” (wuxing
) was evoked to explain them. More on this in section III of this paper.
28 . I agree with Needham (1962) that there is a westward-deviation of south-pointing needles in 9th- and 10th-century China, but for different reasons. Needham’s evidence for this (as abstracted in table 2) is too sketchy and unreliable. What I did is a serious reading of the Qingnangxu 
and Qingnang’aozhi 
, by the famous Tang geomancers Yang Yunsong 
and Zeng Qiuji 
, the Fuzhen fangqi 
diagram in Needham (Needham 1962: 304), and the Hushou jing 
(found in Dili xianpo ji 
). I found these to be consistent with some crucial passages in Chu Yong’s 
Quyi shuo 
(interpreted differently from Needham). Together, they all point to a very probable “westward-deviation” in 9th- and 10th-century China. In May 1995, I gave a talk on this subject at the Collège de France in Paris, entitled “The Needle Deviation (Magnetic Declination): Looking for its Possible Influences in Some Early Medieval Texts of Geomancy”. I would like to thank Dr. Martzloff, who made this talk possible.
29 . See 
edition, 
, juan 655 (
), 
: 6781-82, 6844.
30 . When later modem Chinese historians of science learned enough from these Western hunters, they became the modem “Chinese hunters”, equipped with the Western taxonomy of modem science to hunt down and at the same time defend their “national treasures”.
31 . This is a typical metaphor used by F. Hirth in the construction of his table (or “abstract of dates”) conceming the various issues of “south-pointing needles” in China (Hirth 1908: 135). The object of this “hunt down” for Hirth is the mysterious passage in a Chinese text concerning an observation of magnetic declination in the 8th century by the monk Yixing, a mysterious event first reported by the meticulous A.Wylie (1897: 157), who surprisingly did not provide any reference for the source of that report.
32 . In a process of contextual translation, this is precisely what I am trying to do: to recover or regain the meaning and context of some Tang geomantic texts which contain vague discussions of “needles”. Some preliminary ideas were presented in my talk, “The Needle Deviation” (1995b).
33 . I exclude the observation (“c. +720”) allegedly made by the famous monk Yi-xing, since there is no evidence for it. Although Needham (1962) did not uncover any evidence for Yang Yunsong’s 
discussion of westward-deviation (“c. +880”), I have found some. Hence, I include Yang’s observation.
34 . We might ask, why “compass observations” by Western Jesuits in China should be considered “Chinese”?
35 . See 
édition, 
, juan 655 (
) 
: 6781-82, 6844. Needham (Needham 1962: 302) translates as evidence a passage there (p. 6782), which is problematic. There is a puzzling passage on the same page: 
(p. 6782). A quite different passage is also in 
(p. 6844), “
”, this is clearly “eastern-deviation”. See 
edition, 
, juan 662.
36 . See 
edition, , juan 655 
: 6782. The commentator is talking about the Chinese calendar and run 
; the historical complexity involved here with regard to the history of Chinese geomancy/astronomy has yet to be studied.
37 . Based on a slightly different edition, Needham’s translation (1962: 251) contains the same problems (actually even clearer in this entry) as those in his translation of jotting 437 of Mengxi. I will not repeat them here.
38 . In 
, the entry 
includes the phrase: 
. The 
position in the 24-directions compass indicates deviation toward the east for one full position (a 15-degree westward declination), which is probably greater than at Shen Gua’s time (thirty to forty years earlier with a deviation “slightly to the east” – hence 5-10-degree westward declination, according to Needham’s estimation) and thus more noticeable and more familiar. Another very interesting change is that, although Kou quotes some passages from jotting 437 of Mengxi, he alters the key phrase 
to 
by dropping the adverb 
(“slightly”). This probably also means an increase of the magnitude of the deviation in Kou’s time. By the way, we should not forget that the adverb 
“often” is used again in this text.
39 . In 
, there are many quotations or paraphrases from Mengxi. In Mengxi’s category “Deliberations on Materia Medica”, for example, there are at least eight jottings which Kou had closely consulted: 490, 491, 493, 496, 498, 500, 502, 506. From these cases, we know Kou is very serious about quoting phrases. He was quite independent in judging the materials he consulted and was very organized in quoting or paraphrasing what he wanted. This shows that we should also take seriously the way Kou quotes Mengxi’s jotting 437.
40 . Shen Gua was well versed in the arts of the “five phases”, as amply demonstrated in his Mengxi, but he did not use it in jotting 437.
41 . Presumably, the object to be explained by a cosmographical theory in terms of the “incline” on a flat earth should be a much more “regular” phenomenon. But this is one hundred years after Shen Gua.
42 . Take the case of Chu Yong’s 
in the entry “c. +1230”, where Needham refers to the passage of “Distinguishing Needles” (bianzhen 
) (Needham 307-308). Since I have a quite different interpretation of this passage, I do not have the space here to elaborate on it. By the way, a version of the 
(“Canon of the Fox Head”) mentioned in this passage was found in 
. But the date of its completion is not clear. I suspect it actually referred to the rare “westward-deviation” in the tenth chapter Qiyuan pian 
.
Notes de fin
1 This allows 100 minutes to the degree.
Table des illustrations
© Éditions de la Maison des sciences de l’homme, 1999
Conditions d’utilisation : http://www.openedition.org/6540
