Note: This site has permanently moved to You will be automatically redirected in 5 seconds. Please bookmark the new permanent site.

Please note: This is a preliminary draft prepared for the conference "Rethinking Science and Civilization: The Ideologies, Disciplines, and Rhetorics of World History," Stanford University, May 21-23, 1999. Not for citation or circulation beyond the purposes of this conference.

To appear in Tokens of Exchange: The Problem of Translation in Global Circulations, edited by Lydia H. Liu. Durham, N.C.: Duke University Press (in press); an earlier version is being published as "Translating Worlds: Incommensurability and Problems of Existence in Seventeenth-Century China," Positions 7, no. 1 (in press). Chinese characters have been removed.

Translating the Untranslatable:
From Copula to Incommensurable Worlds[1]

Roger Hart
Program in History and Philosophy of Science, Stanford University

By adopting certain naive presuppositions, studies of the asserted problems encountered in translations across languages have often reached dramatic conclusions about the fundamental differences between civilizations. These presuppositions are naive in that they circumvent many of the questions that should properly confront historical inquiry, adopting instead simple formulas. For example, on what level of social organization should historical explanation concentrate--what are the significant units of society in analyses of historical change? Instead of determining the complex networks of alliances that dynamically constitute groupings within societies, in such studies the boundaries are already given--drawn along lines of languages, or, more often, systems of languages that mark the purported divides between civilizations. What are the fracture lines in societies underlying antagonisms and conflict? Instead of analyzing complicated divisions along the dimensions of class, gender, status, allegiances, or competing schools of thought, all such differences are collapsed into a unity predetermined by the sharing of a single language (the same, that is, once all historical, regional, educational, and status differences are effaced). What kinds of relationships should historical analysis elucidate? With civilizations as the given units of analysis, such studies are typically content with assertions of similarities and differences. What is the relationship between thought and society? Instead of historicizing the role of ideologies, self-fashioned identities, and performative utterances in the formation of social groupings, individuals are instead reduced to representatives or bearers of entire civilizations. How does one understand thought through the transcriptions preserved in historical documents? Instead of explaining the dissemination of copies, commentaries, and interpretations of texts in their cultural context, such studies fix an original against which the correspondence of the translation can be compared. And what is the relationship between thought and language? Too often such studies implicitly presuppose a correspondence between words and concepts. After such a series of simplifying reductions, the conventional conclusions about civilizations are an almost inevitable result.

Rather than critiquing in a general fashion the aporias that inhere in claims made about civilizations in studies of translations, this essay will illustrate these aporias through the analysis of selected studies. To accomplish this, I return to one of the most intensely researched examples of translations across civilizations--the Jesuit missionaries and their translations of European religious and scientific treatises in China in the seventeenth-century. Admittedly, much of the historical literature on this episode hardly merits critique; I have chosen two exemplary studies of these translations that represent the best scholarship on the subject. I follow a tradition of applying historical research to philosophical problems, similar perhaps to what Pierre Bourdieu calls "fieldwork in philosophy."[2] I first outline the claims, presented in these two studies, of linguistic and conceptual incommensurability between seventeeth-century China and the West, claims that are based on the asserted difficulties of translating the copula and the concept of existence. I then turn to the theories of incommensurability that underwrite these studies, along with several related philosophical theories: Emile Benveniste's analysis of the copula to be, Jacques Derrida's critique of Benveniste, W. V. O. Quine's arguments on the indeterminacy of translations, and Donald Davidson's criticisms of assertions of conceptual schemes. Finally, as an alternative to incommensurability, I present an analysis of the translations by the Jesuits and the Chinese converts in cultural context.

China, the West, and the Incommensurability that Divides

Imagining China and the West to be two central actors in historical drama, writers since the eighteenth century have sought symbols that were to distinguish the two.[3] Terms such as modernity, science, and capitalism headed the list of mutually incongruous candidates invoked to portray stark differences: China was identified often by mere absence (for example, of science, capitalism, or modernity) or else designated by pejoratives (for example, practical, intuitionistic, or despotic).[4] Anthropomorphized through the assignment of personality traits (pride, xenophobia, conservatism, and fear), China itself became the subject of a praise-and-blame historiography of civilizations. In the period following World War II, Fairbankian historiography decried the lack of agency attributed to China, offering redress by assigning to China a limited capacity to respond to the West.[5] Joseph Needham proposed to restore for China its pride, correcting its slighting by making it an equal contributor among the tributaries that flowed into the river of modern science; his "grand titration" was to redistribute credit for scientific discoveries among civilizations.[6] Joseph Levenson projected Liang Qichao's thought onto a "mind of modern China" and psychologized China's historicist reaction against Western value.[7] These postwar approaches, then, focused on the ways that China had either responded to, contributed to, or rationalized away the West; the West remained for these writers conceptualized as essentially universal. Studies that emerged in the late 1960s and 1970s retained the China/West dichotomy but inverted the earlier triumphalist accounts of Western universalism, critiquing the exploitation, domination, and violence wrought by imperialism.[8] During the 1980s, a "China-centered approach" was articulated as an alternative that, however, for a field institutionalized under the rubric of "area studies," too often meant little more than a return to Sinocentrism, with its attendant claims of particularism and Chinese uniqueness.[9] What these later approaches share with their predecessors, then, is a continued credulity toward an essential divide between China and the West.[10]

In this context, studies of the "first encounter" of these two great civilizations have acquired a particular urgency. Interpretive approaches have often been limited to two alternative models--conflict, opposition, and misunderstanding, or synthesis, accommodation, and dialogue.[11] But in recent years, relativism--again formulated within the context of an assumed plausibility of a divide separating China and the West--has become yet another important approach.[12] Theories of linguistic and conceptual incommensurability often underwrite this relativism, providing for relativism perhaps its most rigorous formulation. These claims of relativism and incommensurability have played an important role in encouraging the analysis of Chinese sources and viewpoints by positing a special Chinese worldview protected from pretentious dismissal by a historiography mired in universalism. Yet they have done so at the cost of further reifying China and the West, and further radicalizing the purported divide that separates them. To elucidate the role played by claims about translation in theories of incommensurability, this section will examine two important recent historical interpretations of this encounter: Jacques Gernet's China and the Christian Impact[13] and a related analysis of the translation of Euclid's Elements by Jean-Claude Martzloff in his History of Chinese Mathematics.[14]

In probably the most sophisticated study of the Jesuits in China during the seventeenth century, Gernet's China and the Christian Impact adopts incommensurability between Western and Chinese concepts as the philosophical framework that is to explain the history of the translation and introduction of Christianity.[15] Against previous studies of the introduction of Christianity into China that had been based primarily on Western sources, Gernet proposes as a new approach the study of the "Chinese reactions to this religion." Previous approaches were often universalistic, assuming that "one implicit psychology--our own--valid for all periods and all societies is enough to explain everything." Gernet asserts that for the missionaries, the rejection of Christianity "could only be for reasons that reflected poorly on the Chinese." Later interpreters similarly have "a tendency to see the enemies of Christianity as xenophobic conservatives" while praising converts as open minded. This thesis, Gernet asserts, "is contradicted by the facts."[16]

Gernet's defense of the Chinese rejection of Christianity is based on a claim of the fundamental incommensurability of languages and the associated Chinese and Western worldviews: "The missionaries, just like the Chinese literate elite, were the unconscious bearers of a whole civilisation. The reason why they so often came up against difficulties of translation is that different languages express, through different logics, different visions of the world and man."[17] Gernet outlines this theoretical framework in the final sections of his concluding chapter. He offers several examples of the difficulties in bridging "mental frameworks": for example, "In trying to assimilate the Chinese Heaven and the Sovereign on High to the God of the Bible, the Jesuits were attempting to bring together concepts which were irreconcilable."[18] He discovers radical differences between Chinese and Western thought: "The Chinese tendency was to deny any opposition between the self and the world, the mind and the body, the divine and the cosmic.... For Chinese thought never had separated the sensible from the rational, never had imagined any `spiritual substance distinct from the material,' never had conceived of the existence of a world of eternal truths separated from this world of appearances and transitory realities."[19] These differences (although still often conceptualized by Gernet as absences) are adduced as evidence that demonstrates the "radical originality" of China: "Ultimately, what the Chinese criticisms of Christian ideas bring into question are the mental categories and types of opposition which have played a fundamental role in Western thought ever since the Greeks: being and becoming, the intelligible and the sensible, the spiritual and the corporeal. Does all this not mean that Chinese thought is quite simply of a different type, with its own particular articulations and its own radical originality?"[20]

The philosophical framework of conceptual incommensurability that Gernet employs in this work is based on the linguistic theory of Benveniste: "Benveniste writes: `We can only grasp thought that has already been fitted into the framework of a language. . . What it is possible to say delimits and organises what it is possible to think. Language provides the fundamental configuration of the properties that the mind recognises things to possess.'"[21] More specifically, Gernet asserts that the two fundamental differences between Chinese and Western languages are categories of thought that derive from language and the concept of existence:[22] "Benveniste's analysis illuminates two characteristics of Greek--and, more generally, Western--thought, both of which are closely related to the structure of Greek and Latin: one is the existence of categories the obvious and necessary nature of which stems from the use to which the language is unconsciously put. The other is the fundamental importance of the concept of being in Western philosophical and religious thought."[23] As I will argue below, Gernet's examples--the translation of Christian terms--present special philosophical problems. So before exploring these, I will examine the translation of Euclid's Elements as a more concrete but related example for the comparison of Chinese and Greek thought and language.

The translation of Euclid's Elements into China by Xu Guangqi (1562-1633) and Matteo Ricci (1552-1610) in 1607 would seem ideal for an examination of linguistic incommensurability, given the extant historical documents. Jean-Claude Martzloff, perhaps the most eminent Western historian of Chinese mathematics, has written extensively on the translation.[24] He adopts Gernet's incommensurability in his explanation of the history of the translation, arguing that the Chinese had failed to comprehend the deductive structure of the Elements precisely because of linguistic incommensurability. Martzloff argues that the central problem was the difficulty of translating the copula, because of its absence in classical Chinese:

In addition to the terminology, the even more formidable problem of the difference between the Chinese syntax and that of European languages had to be faced. The main difficulty was the absence of the verb "to be" in classical Chinese. The translators were unable to find better substitutes for it than demonstratives or transitive verbs such as you, wu and wei.... But often, the verb "to be" disappeared altogether, as in the following case:

[The] circle: [a] shape situated on flat ground (ping di) [sic] within [a] limit. [The] straight strings (xian) constructed from [the] limit to [the] centre: all equal.[25]

Martzloff then offers for comparison Clavius's original:

"Circulus, est figura plana sub una linea comprehensa, quae peripheria appelatur, ad quam ab uno puncto eorum, quae intra figuram sunt posita, cadentes omnes rectae linae, inter se sunt aequales."[26]

Martzloff then links the copula to questions of existence, asserting that "one might think that this type of phenomenon contributed to a masking of the conception, according to which geometric objects possess inherent properties, the existence or non-existence of which is objectifiable."[27] Although Martzloff apparently borrows this framework from Gernet's China and the Christian Impact and Benveniste,[28] in his argument he cites primarily A. C. Graham's "`Being' in Western Philosophy" as asserting that neither you, wu, nor wei are equivalent to the copula.[29]

In addition to Gernet's and Martzloff's assertions based on Benveniste, a wide variety of arguments on the relation of language to thought have been presented in historical studies of China. Peter Boodberg suggests that "the great semantic complexity of tao may have predetermined the rich system of associations surrounding Tao in its metaphysical and literary career."[30] Alfred Bloom notoriously asserts that the lack of counterfactuals and universals in the Chinese language inhibited the ability of the Chinese to think theoretically.[31] Many authors have presented claims that the Chinese language inhibited the development of science.[32] Until recently, such studies have rarely critically analyzed any of the details of the theories they cite;[33] the following section, then, will examine Benveniste's claims about the copula.

The Philosophy, Language, and Translation of Existence

Benveniste's central thesis is that language and thought are coextensive, interdependent, and indispensable to each other. "Linguistic form is not only the condition for transmissibility," Benveniste asserts, "but first of all the condition for the realization of thought";[34] the structure of language "gives its form to the content of thought."[35] Benveniste examines Aristotle's categories of thought to assess whether we have "any means to recognize in thought such characteristics as would belong to it alone and owe nothing to linguistic expression."[36] He concludes that Aristotle's categories were simply the fundamental categories of the language in which Aristotle thought--"the ten categories can... be transcribed in linguistic terms."[37] "Unconsciously," Benveniste argues, Aristotle "took as a criterion the empirical necessity of a distinct expression for each of his predications.... It is what one can say which delimits and organizes what one can think."[38]

In "The Supplement of Copula: Philosophy before Linguistics," as an example of the paradoxes in claims that language governs thought, Jacques Derrida critiques Benveniste's assertion that the Greek language determined Aristotle's categories.[39] It is Benveniste's own writings, Derrida asserts, that offer a "counterproof" against the assertion that being is nothing more than a category linguistically determined by the copula to be: Benveniste himself asserts that there is a meaning of the philosophical category to be beyond that expressed in grammar. For Benveniste argues, Derrida asserts, that (1) "the function of `the copula' or `the grammatical mark of equivalence' is absolutely distinct from the full-fledged use of the verb to be" in the sense of existence; and (2) "in all languages, a certain supplementary function is available to offset the lexical `absence' of the verb `to be,'" used grammatically as a mark of equivalence.[40]

Derrida then links this conflation of these two uses of to be--grammatical and lexical--to the history of Western metaphysics.[41] It is the "full-fledged" use that Heidegger wishes to recover when he suggests that being has become both compromised and effaced: "`Being' remains barely a sound to us, a threadbare appellation. If nothing is left to us, we must seek at least to grasp this last vestige of a possession."[42] This nostalgia for a return to the use of to be as existence is echoed, Derrida asserts, by Benveniste: "It must have had a definite lexical meaning before falling--at the end of a long historical development--to the rank of `copula.'... We must restore its full force and its authentic function to the verb `to be' in order to measure the distance between a nominal assertion and an assertion with `to be.'"[43] The copula thus transcends the grammatical categories of any particular language: in some languages it is denoted by only a lexical absence; on the other hand, the "full fledged" notion of to be cannot be a category determined by language if it is still to be possible to return from the effaced use of being to its "full force" and "authentic function."

Derrida's critique of Benveniste exemplifies the aporias that Derrida suggests inhere in assertions that philosophic discourse is governed by the constraints of language. For the oppositions of linguistics--"natural language/formal language, language system/speech act, insofar as they are productions of philosophical discourse, belong to the field they are supposed to organize."[44] Derrida thus inverts Benveniste's claim, asserting that "philosophy is not only before linguistics in the way that one can be faced with a new science, outlook, or object; it is also before linguistics in the sense of preceding, providing it with all its concepts."[45]

Among the claims for radical differences between languages, the absence of the copula has seemed to be both the most concrete and the most significant, for the copula has seemed to be the most plausibly connected with philosophically important consequences. Derrida's criticisms point to two fundamental problems in Gernet's and Martzloff's applications of the theories of Benveniste: (1) although the existence of the copula seemed to mark an important difference between the Indo-European and Chinese languages, Benveniste's claim is, rather, that the copula exists in all languages, and (2) the correlation between the copula and the metaphysics of existence results from the conflation of two separate uses of the same lexical term.[46]

If the linguistics of the copula does not demonstrate that radical differences exist between languages--much less between the thought expressed in those languages--the general question remains: To what extent can differences in thought be shown to result from differences in language? Derrida offered one answer: the concepts from linguistics that are to provide the basis for comparison are themselves constituted by the philosophy they purport to analyze. Quine offers a different critique: to compare systems of thought, we must first have solved the problem of translation. Against the views of Ernst Cassirer, Edward Sapir, and Benjamin Whorf that differences in language lead to fundamental differences in thought, Quine objects that we cannot, in principle, provide translation rigorous enough to assess such grand philosophical theses. He concludes that it is not that "certain philosophical propositions are affirmed in the one culture and denied in the other. What is really involved is difficulty or indeterminacy of correlation. It is just that there is less basis of comparison--less sense in saying what is good translation and what is bad--the farther we get away from sentences with visibly direct conditioning to nonverbal stimuli and the farther we get off home ground."[47]

If Quine's argument, based on the indeterminacy of correlation, concludes that it is impossible to rigorously compare differing conceptual schemes, Davidson, in "On the Very Idea of a Conceptual Scheme," warns that the notion of a conceptual scheme is itself ultimately unintelligible.[48] Davidson's argument is presented against the conceptual relativism of Quine, Whorf, Thomas Kuhn, and Paul Feyerabend.[49] Davidson notes the following paradox: the demonstration that two conceptual schemes are incommensurable requires the solution of the purported incommensurability in a frame of reference that incorporates both; assertions of conceptual schemes are always framed in a language which purports to explain that which cannot be explained.[50] The differences, he notes, "are not so extreme but that the changes and the contrasts can be explained and described using the equipment of a single language."[51] Davidson thus argues that "we cannot make sense of total failure" of translation, whether based on a plurality of imagined worlds or incommensurable systems of concepts employed to describe the same world.[52] Davidson concludes that "no sense can be made of the idea that the conceptual resources of different languages differ dramatically."[53]

Davidson's central argument is against the assertion of conceptual schemes by showing that they are unintelligible--that is, their very formulation is paradoxically circular. But in the course of his argument, Davidson also offers the following deflationary aside: "Instead of living in different worlds, Kuhn's scientists may, like those who need Webster's dictionary, be only words apart."[54] If Davidson is right that assertions of incommensurable conceptual schemes cannot be formulated coherently, then the question remaining for us to answer becomes, How are claims of different worlds constructed from differences in words? It is precisely because of the impossibility, suggested by Davidson, of describing radically different conceptual worlds that such claims must defer by analogy to other radical differences. Thus, for example, Kuhn explains the purported incommensurability of scientific paradigms through analogies to differing linguistic taxonomies.[55] Feyerabend explains incommensurability between "cosmologies" through analogies in forms of art.[56] Whorf explains differing linguistic systems through examples from science.[57] Quine explains differing conceptual schemes through analogies in physics.[58] And Martzloff and Gernet explain the incommensurability of thought through analogies to linguistic differences, via Benveniste's copula.

How are incommensurate worlds, then, created from words? The process of translation would at first seem an unlikely tool, since the putative goal of translation is to establish equivalences between two languages. However, it is this assumption that provides for these claims a crucial resource: the purported impossibility of finding equivalent words then itself serves as a sign of radically different worlds.[59]

The differences that appear in Martzloff's examples are not between Clavius's Latin version and Ricci's Chinese translation--no unmediated comparison is possible--but instead between the untranslated Latin, the translation that Martzloff provides of Ricci's Chinese into English, and ordinary expressions of English. Clavius's Latin represents an uncorrupted original by remaining untranslated and dehistoricized: effaced are the problems of translation from Greek to Latin to French and English, the complex history of the translation and editions of this text,[60] and in particular Clavius's redaction that altered and deleted much of the structure of the proofs.[61] Martzloff's translations convey the radical otherness of classical Chinese by employing techniques of defamiliarization similar to those used elsewhere to demonstrate (again, in English) the purported awkwardness of Chinese monosyllabism: "King speak: Sage! not far thousand mile and come; also will have use gain me realm, hey?"[62] Indeed, Martzloff argues against "more elegant, more grammatical" renderings, stating that "English grammaticality tends to obliterate the structure of the Chinese and the connotation of the specialised terms."[63]

The differences Martzloff presents are the artifacts of these choices he makes in his translation. First, he insists on an extreme literalism in his selections of possible equivalents, for example "straight strings" for xian and "flat ground" for pingdi, with the latter marked by sic to emphasize the inappropriateness of what was, after all, his own choice. He marks articles with brackets. However, Martzloff's most jarring technique is his omission of the copula in English, a language in which the copula is denoted lexically.[64] Although Martzloff cites Graham as asserting the lack of the equivalent of the copula in classical Chinese, Martzloff notes that "we shall not retain [Graham's] English translations, since these translations introduce numerous elements which do not exist at all in classical Chinese (for example, the verb `to be')."[65] Martzloff, however, then supplements English with a nonlexical symbol, the colon, to denote the absent copula: "since ordinary words are not sufficient, we also use a punctuation mark" (in his earlier translations in French, Martzloff employed a question mark paired with an exclamation mark).[66] It is then senseless to correct Martzloff's translation. While ordinary translation seeks to establish equivalences, Martzloff seeks to convey radical differences; but following these principles, an English translation of Clavius's Latin would be rendered equally bizarre, and thus Martzloff must leave the Latin untranslated. The extremes to which Martzloff takes the translation are necessary to evoke the linguistic differences that are to serve as an analogy for radical differences in thought.

If Martzloff's claimed linguistic incommensurability relating to existence is an artifact of his jarring omission of the copula in insistently preserving this "lexical absence" of classical Chinese in modern English and French that denote the copula lexically, is there any evidence to support claims of a radical incommensurability on the conceptual level? As noted above, Gernet's claims of conceptual incommensurability were based on his assertion, following Benveniste, of differences between China and the West in the fundamental "concept of being" and "the existence of categories" that unconsciously stem from the use of language;[67] Gernet adopts the claims of abstract differences in the concepts of existence and categories as a philosophical theory within which to frame his description of historical events. The critiques by Derrida, Quine, and Davidson suggested general philosophical problems with these claims; here instead I will seek historical explanations. That is, we must return to the debates on existence and categories not abstracted from but, rather, resituated within their historical context, and not as philosophy explaining history, but as philosophy inseparable from the history it was to explain. To do so, I will reexamine one of the central texts analyzed by Gernet, Matteo Ricci's Tianzhu shi yi (ca. 1596), for evidence of the problem of the translation into Chinese of notions of existence and categories.[68]

The problem that confronted Ricci in the Tianzhu shi yi, it turns out, is not one of an impossibility of expressing the philosophical concept of existence in the abstract but, rather, debates about existence with specific referents--in particular, spirits and God. Chapter 4 begins with the following summary of the previous chapter:

Chinese scholar: Yesterday after I took my leave and reviewed your distinguished instruction, sure enough [I] understood that it all is true. I do not know why the deluded scholars of my country should accept as the orthodox truth denials of the existence of spirits.

Western scholar: I have comprehensively examined the ancient classics of your esteemed country. Without exception [these texts] take sacrifices to the spirits as momentous occasions for the Son of Heaven and the feudal lords; thus [they] revered these spirits as being above them and all around them. How then could it possibly be that there is no such thing and thus in this [they] acted deceitfully! [69]

In the Tianzhu shi yi, there is no shortage of ways to express of spirits the predicate of existence: "Tang['s soul] continued to exist without dissipating"; "souls of the deceased exist eternally without extinction"; "the human soul does not dissipate after death".[70] Nor is there any shortage of debates on questions of existence of spirits. In Tianzhu shi yi, the Chinese scholar summarizes contemporary Chinese positions on the existence of spiritual beings as follows:[71]

Chinese scholar: Among contemporaries that discuss the spirits, each has his own viewpoint. Some state that nowhere in the world are there things such as spirits. Others state if one believes in them then they exist; if one does not believe in them then they do not exist. Others state that to assert they exist is incorrect; to assert that they do not exist is also incorrect; only to assert that they both exist and do not exist is to attain it [the correct viewpoint]![72]

The reply of the Western scholar proceeds from the claim that "all affairs and things that exist indeed do exist, [those that] do not exist indeed do not exist", and proceeds to the conclusion that spirits exist. Ricci's dispute with commonplace Chinese views is thus not that the spirits of ancestor worship do not exist; on the contrary, they exist eternally. Rather, there is a more important spirit that the Chinese must worship if they seek fortune and salvation. His defense of the existence of spirits is based on his own translation of an enigmatic phrase attributed to Confucius, interpreted through his knowledge of God:

Thus Confucius states: "Respect the spirits, and distance them."[73] Happiness, fortune, and forgiveness of sin are not within the powers of the spirits, but up to the Lord of Heaven alone. Yet the trendy curry favor [with the spirits], desiring to receive this [happiness, fortune, and forgiveness] from them; but it is not the true way to obtain this. The meaning of "distance them" is the same as "if one sins against Heaven, there is no one to pray to."[74] How is it possible that "distance them" can be explained as "they do not exist," snaring Confucius in the deceit that spirits do not exist?[75]

What is superstitious about Chinese ancestor worship, then, is not belief in these spirits themselves but seeking favor by worshiping the spirits of ancestors rather than salvation by worshipping the Lord of Heaven. Ricci's answer indicates that the problem is not a lack of belief but, rather, that the Chinese concept of God has been effaced. Contrary to Gernet's claims, the translation of European treatises into Chinese provides no convincing evidence to demonstrate difficulties in expressing the concept of existence, whether of tangible or intangible objects. In historical context, these problems were not about the impossibility of translating the abstractions of modern philosophy into Chinese; they were debates about the existence of spirits and, in particular, the primal noun (God) and the primal verb (Being).

Similarly, differences in categories of thought--whether based on Benveniste's view that language dictates categories through an unconscious process or Kuhn's view that incommensurability results from mismatched taxonomies--do not provide a framework outside history available only to modern commentators. Instead, the claim of differences in taxonomies was itself a strategy in Jesuit propaganda. Ricci argues,

In dividing things into categories, the learned men of your noble country state: some [things] attain form, such as metal and stone; some in addition attain the energy of life and grow, such as grass and trees; some also attain the senses, such as birds and beasts; and some are more refined and attain consciousness and intelligence, such as man.

Our learned men of the West seem to have made even more detailed [categories], as can be seen from the following chart. Only its classes of accidents are most numerous and difficult to list completely, and therefore they are only summarized, emphasizing their nine major classes.[76]

Linguistic differences and conceptual differences do exist--as they do within civilizations, cultures, and subcultures--but this is very different from a theory of radical conceptual incommensurability split along a purported China/West divide. As an explanatory framework, conceptual incommensurability is at best inflationary: Does the rejection of Ricci's claims asserting the existence of spirits really require for its explanation a theory of incommensurability asserting the impossibility of the translation of concepts of existence? Does contesting Ricci's God and the evidence that he adduces for Him from the Lun yu really require for its explanation a theory of two radically different philosophical worldviews? Does the disbelief in the soul really require for its explanation assertions of the incomprehension of Western concepts and scholastic philosophy? Worse, Gernet's thesis of incommensurability "is contradicted by the facts." The above examples suggest no insurmountable difficulties in expressing concepts of existence.

Gernet's claims themselves, like those of Martzloff, create the otherness of the world he purports to describe. These claims of differences derive their plausibility in the first place from the assumption of a Great Divide between China and the West. The complex similarities and dissimilarities between Jesuit doctrines and Legalism, Buddhism, Confucianism, Daoism, and popular religions are collapsed along a single radical China/West divide, which is then reinforced through the insistent, repeated assertion of difference: "The conclusion that emerges from the various texts which I have just cited is that Chinese conceptions are in every regard the opposite of those taught by the missionaries."[77] Counterexamples--the adoption of Western Studies [Xi xue ] by Chinese converts--are themselves appropriated to further reproduce this divide: "What appeared to fit these traditions--or rather what could be easily integrated--was accepted; the rest was unacceptable";[78] "Chinese who sympathized with Christianity praised it in terms of Chinese philosophy and attributed their own conceptions to it."[79] Through links forged by tracing linguistic continuities, this break between cultures then becomes the suprahistorical break between civilizations;[80] the rejection of Jesuit doctrines is inflated to represent the rejection of "the European vision of the world."[81]

The import for research on China of this barrier constructed by incommensurability was that--against dismissals common in the received historiography which narrated the inexorable triumph of Western universals--relativism protected the assertions of the Chinese protagonists. Relativism insisted not just on equivalence but on the impossibility in principle of any comparison. Elevated to the status of an equal, the claims and allegations of the Jesuits' Chinese opponents came to merit historical analysis. This, then, is the central contribution resulting from Gernet's adoption of relativism and incommensurability--his insistence on understanding the arguments of the Chinese opponents of the Jesuits and their converts through the analysis of Chinese primary materials. For this, China and the Christian Impact has been justly recognized as seminal.

Yet one consequence of this very relativism was that it effaced the social and political context it should have analyzed--it is important here is to see what this framework of incommensurability must, by construction, leave out. When social and historical context is shifted into the realm of discourse, rejections of Jesuit doctrines are explicable only by assertions of the impossibility of translation between two radically different worldviews; individuals are deprived of agency in deciding what is determined by these conceptual structures. Yet how can we be assured that the difficulty is one of translation if the existence of the referent of the term--whether spirits or God--is itself in question? How does one determine whether the notion of God was correctly translated? Without a grounding in a truth external to the text, the only criterion for assessing the translations is comparisons with explanations and glosses in other texts; in short, we have returned to doctrinal disputes within Christianity of which the Jesuits were themselves but one faction. And perhaps the most serious problem in Gernet's framework is the collapsing of the complex interactions between individuals and subcultures to two mutually exclusive poles, China and the West; the Chinese converts, belonging to neither, can only be minimized as transparent, passive translators.

The Contextual Turn

Historical studies framed within the linguistic turn have often displaced social context into the realm of discourse;[82] theories of incommensurability have been but one variant on this trend. One important alternative that re-places discourse in social context is suggested by Mario Biagioli's proposal of a "diachronic approach" to incommensurability by analyzing "its emergence in relation to the internal structure, external boundaries, and relative power or status of the socio-professional groups involved in the non-dialogue."[83] Biagioli's concern is to explain how phenomena that have been interpreted as linguistic incommensurability are in fact strategies adopted in social conflicts between groups that share the same language.[84] I take a related approach, showing how claims of purported difficulties in translations between different languages themselves served as resources in social conflicts. Claims about translation will not be conceptualized as disembodied philosophical theories that provide a framework external to history within which events are analyzed; instead, claims about translation will themselves be examined as in need of historical explanation. That is, instead of accepting claims made by historical protagonists about translation, whether its accuracy, difficulties, or impossibility, we must analyze, contextualize, and render historically intelligible these claims themselves. For example, in seventeenth century Europe, theories about the translation and accommodation of the spoken word of God to the languages of the illiterate masses provided both important resources and alibis in debates over heliocentric theories;[85] theories about translations of the Bible and the Book of Nature were central sites of contest in debates over the legitimacy of Galileo's claims.[86] How, then, did claims about translation provide for the Jesuits and the Chinese converts important strategies in their proselytism?

To answer this question, first we must examine translations, conceptualized not as an impossibility demonstrated by posing against translations objections from biblical hermeneutics but, rather, analyzed as historical events.[87] Perhaps the simplest approach to translation, and the one adopted by the Jesuits and their converts most frequently (except for the translation of important theological terms), was the creation of neologisms. In some cases, they resorted to loan-words--neologisms created by transliteration.[88]

The study of anima (translated, means soul or nature of the soul), within philosophia (translated, means the study of "investigating things and exhausting principles"), is the most beneficial and the most respected.[89]

Transliteration provided one possible translation for the term God. For example, in the first chapter of Tianzhu shi yi, titled "Lun Tianzhu shi zhi tian di wanwu, er zhuzai anyang zhi" [Showing that the Lord of Heaven created heaven, earth, and the myriad things, and controls and sustains them], Ricci argues that there must be a creator of the heavens, and "Thus this is the Lord of Heaven, the One our Western nations term Deus".[90] A transliteration for Deus was, in fact, the choice for the translation of the term `God' into Japanese. This historical possibility of creating neologisms by semantically neutral transliteration undermines theories of incommensurability that assert radical impossibilities based on nothing more than the absence of lexical terms.[91]

Another approach was loan translations--the creation of semantic neologisms by combining characters. This was the approach most often employed by the Jesuits and their converts. The following examples will serve as a very small sample: in theology, terms for omniscience (zhi zhi), omnipotence (zhi neng), and infinite goodness (zhi shan);[92] in Aristotelian philosophy, terms for cause (suoyiran), active cause (zuozhe), formal cause (mozhe), material cause (zhizhe), final cause (weizhe), the four types of causes--proximate, distant, universal, and special (respectively, jin, yuan, gong, si),[93] substance (zilizhe), and accident (yilaizhe);[94] and in Euclidean geometry, point (dian), line (xian), and surface (mian).[95] A very different strategy of translation adopted by the Jesuits and their converts was the selective omission of doctrines that would have subjected them to even harsher attacks: the Jesuits often omitted mention of the Trinity, revelation, and, with the exception of baptism, the sacraments;[96] they also frequently failed to mention the crucifixion of Jesus.[97]

For the translation of the most important terms in the most important subject, Christian theology, the Jesuits used semantic extension, borrowing and redefining terms from Buddhism and Confucianism. From Buddhism the Jesuits appropriated terms that had been employed as equivalents for Sanskrit: for Heaven, tiantang (Sanskrit devaloka, mansion of the gods); for Hell, diyu (Sanskrit naraka); for Devil, mogui (mo for the Sanskrit mâra); for angels, tianshen (Sanskrit deva; Protestant translators later employed tianshi); for soul, linghun.98 But it was from Confucianism that the Jesuits borrowed their most crucial terms. As Gernet notes, "The first missionaries were especially delighted to find in the Classics--the works venerated above all others among the literate elite--the term `Sovereign on High' (shangdi), invocations to Heaven and expressions such as `to serve Heaven' (shi tian), `to respect' or `fear Heaven' (jing tian, wei tian)."[99] For example, the overall project of the Jesuits was described by Xu Guangqi as "self-cultivation and serving Heaven" (xiu shen shi Tian).[100] And among the terms translated, the one of the utmost significance for the Jesuits was God. Instead of the transliteration of Deus, early translations made use of several choices: Tianzhu (a term which appears in Buddhist texts), Shangdi (a term that appears in several early Chinese texts, including the Li ji Shi jing, Shu jing, Mo zi, and Shi ji), Tiandi (a term appearing primarily in Buddhist texts but also in Zhan guo ce) and Shangzun.[101]

Different choices in translation provided the Jesuits and their converts with different opportunities. Phonemic and semantic neologisms necessitated lengthy explanations and commentaries; examples include treatises explaining the concept of anima and the soul.[102] But borrowings from Buddhism and Christianity provided important additional opportunities. The borrowing of terms from Buddhism was the result of an attempt by the Jesuits in the early years of the mission in China to represent themselves as similar to the Buddhists. But in later years, their use of Buddhist terms also provided the Jesuits with the claim that their doctrines corrected Buddhist distortions. For the Jesuits, Buddhist theories were doubly false: Buddhist doctrines were perversions of Indian beliefs that were no longer accepted in India, doctrines that had in fact originated in the false beliefs of Pythagoras.[103]

The Jesuits' strategy in China was focused--as in Europe--on gaining patronage with the monarch. The official Confucian orthodoxy provided crucial opportunities for this project and, conversely, for the elite literati-officials who collaborated with the Jesuits. For the Jesuits and the converts, the problem in the choice of the proper term for God was not a lack of possible equivalents but, rather, the opportunities offered by each that entailed complex strategic implications and consequences--social, political, philosophical, philological. (The conflict over these choices led to the Rites Controversy, with the use of Tian and Shangdi being forbidden by the pope in 1704.)[104] Claims made by the Jesuits and their converts about translation served as an important means of legitimation. As Gernet notes,

The missionaries also often resort to another idea: namely that part of the ancient Chinese tradition had disappeared in the Burning of the Books ordered by the first of the Qin emperors in 213 BC and that it was precisely that part that set out the thesis of an all-powerful, creator God, the existence of heaven and hell and the immortality of the soul; the teaching of the missionaries fortunately made it possible to complete what had been lost in the classical traditions of China. This is, indeed, pretty well the thesis put forward by Ricci in The True Meaning of the Master of Heaven, where he explains to a Chinese man of letters why it is that the Classics make no mention of paradise and hell.[105]

For the Jesuits and their converts, translation was not theorized as introducing new knowledge but, rather, as a recovery of knowledge that had been lost from the Chinese tradition.

The most important opportunity offered by translation was in the selection among possible equivalents for the term God: ambiguities in translation were a crucial resource to camouflage ambiguities in the loyalty of elite literati-official converts toward the Ming dynasty. These elite converts were central in the dissemination of Western Studies in China, not merely as the object of Jesuit proselytism strategies, but in translating Western Studies into the language and literary style of the elite, helping to legitimate it, and building their careers on its advocacy.

The most important among these converts was Xu Guangqi; his success in patronage exemplifies how problems of translation permitted the converts to produce documents that could be read by both the Jesuits and the Ming imperial court as expressions of loyalty and faith.[106] Xu had been trained in the Hanlin Academy to write memorials to the Ming court on issues ranging from taxes to water conservancy, from military proposals to astronomy; as Ray Huang notes, "because [the Ming] empire was created to be controlled from the center by documents; field experience or lack of it made very little difference."[107] In his memorials Xu repeatedly risked his career for the Jesuits, while at the same time fashioning himself as a statesman with novel practical solutions to Ming dynasty crises. Probably the most important example is Xu's explicit defense of the Jesuits in "Bian xue zhang shu" [Memorial on distinguishing learning]:[108]

Thus the learning of serving Heaven transmitted [by the Jesuits] can truly be used to supplement the moral influence of our sovereign, aid Confucianism and correct Buddhism. Thus in the West there are more than thirty neighboring kingdoms which implement this doctrine. For over a thousand years up to the present, the large and small help one another; the superior and the inferior live together in peace; borders require no defenses; dynasties exist without change; countries are entirely without cheats or liars; ever since antiquity there has been no lasciviousness or thieving; people do not pick up objects lost on the roads; and doors are not locked at night. And as for disturbances and rebellions--not only are they without such affairs and without such persons--there are not even words or written characters to denote such things.[109]

For Gernet, this passage is evidence of a Chinese "ancient mental framework" incommensurable with Western thought: "For Zhang Xingyao and Xu Guangqi, the ancient mental frameworks remain unchanged despite their conversions: orthodoxy must contribute towards the universal order and is recognisable by its beneficial moral and political effects."[110] Yet the content of this passage reflects not the workings of a mental framework but, rather, the intended audience; it is, after all, a memorial on political policy presented to the Ming imperial court. What Gernet ignores is evidence from other sources that can equally support the claim that Xu adopted the Christian faith: Xu's extant letters show, as Fang Hao notes, that "there are many places that express the sincerity of his religious beliefs."[111] In particular, Xu's eleventh extant letter to his family shows that not only did Xu have his father-in-law converted, Xu was also worried that his father-in-law had not been given absolution before his death.[112] Questions about Xu's beliefs, then, cannot be answered a priori on the basis of differences in language or worldview; instead, it is precisely these questions that were central matters of debate among the historical protagonists. If instead of theorizing translation as an impossibility we seek to understand the process by which translation did occur, this passage is representative of that process. Xu's novel solutions--military, moral, mathematical, and astronomical--appealed to a desperate Ming court that promoted him to one of the highest posts;[113] in turn, Xu's success helped legitimate the Western Studies that he advocated.[114] And in this passage, Xu's most dramatic evidence that this constructed world of the Western Other offered solutions to Ming dynasty crises was a lexical absence--the word for rebellion. Here again--as was the case with Martzloff and Gernet--fantastic claims about the radically different Other are betrayed by assertions of the absence of words.


Relativism and the incommensurability--linguistic or conceptual--on which it was based required for its initial formulation the assumption of a radical divide between two imagined communities, China and the West. Differences in languages then served as a natural symbol for this presumed divide--both by paralleling political boundaries and by providing compelling metaphors for the suprahistorical continuity of civilizations. This use of languages as emblems for civilizations could, however, with equal ease be used to support either universalistic or relativistic conclusions: differences in languages, through purported hierarchies constructed for languages (e.g., their precision and scientificity, or the development of alphabetization) have been linked to other civilization-defining teleologies (e.g., science, or capitalism);[115] on the other hand, differences in language have also provided important metaphors to construct relativism's radically different Other.

For the purposes of this latter relativism, features of languages were further essentialized and radicalized: the difficulties of translation came to represent impervious barriers; mutual intelligibility came to represent an essentialized, systemic unity; diachronic continuities came to represent suprahistorical self-identity. The enormous diversity of strains of thought collapsed into an essentialized China provided crucial alibis in the forging of continuities and discontinuities on which the claims of incommensurability depended. On one hand, this diversity provided a wealth of examples to demonstrate the opposition between China and the West; yet this same diversity could always provide examples to explain the acceptance of Western doctrines in China as nothing more than the acceptance of the Chinese tradition.

Historically contextualized, in their translations the Jesuits and their converts adopted for their own religious concepts terminology appropriated from Buddhism. They created neologisms from terms in the Confucian tradition. Their neologisms provided the opportunity for explanations and commentary. Jesuit doctrines were then claimed to be the recovery of the lost meanings of the Confucian classics destroyed in the Qin burning of the books. But most importantly, problems in translation served as an important patronage strategy of the converts. By introducing ambiguities in the translation of terms such as serving Heaven and Sovereign on High, the converts produced documents that could be read by both the Chinese court and the Jesuit missionaries as expressions of allegiance. Translation was thus not an obstacle to dialogue but a crucial resource; the Chinese converts were not transparent scribes but active agents manipulating these translations.


I have removed the Chinese from this version because of the problems of displaying Chinese in web browsers.

[1]I would like to thank Steve Angle, Mario Biagioli, Yomi Braester, Paul Cohen, Benjamin Elman, David Keightley, Ted Porter, Roddey Reid, Haun Saussy, and Wen-hsin Yeh for detailed criticisms. Two anonymous reviewers offered important suggestions. Versions of this paper have been presented at Berkeley, MIT, Stanford, the Association for Asian Studies, and the Mathematisches Forschungsinstitut Oberwolfach (Germany); I would like to thank all those who offered comments. Positions kindly published an earlier version of this paper. Finally, I owe special thanks to Lydia Liu for criticism, encouragement, and advice.

[2]See Pierre Bourdieu, "`Fieldwork in Philosophy,'" in In Other Words: Essays towards a Reflexive Sociology, 3-33.

[3]Two important early examples are Georg W. F. Hegel, The Philosophy of History; and Max Weber, The Religion of China: Confucianism and Taoism.

[4]For critical analyses of works asserting that China did not have science, see Nathan Sivin, "Why the Scientific Revolution Did Not Take Place in China--Or Didn't It?"; and Roger Hart, "On the Problem of Chinese Science."

[5]For example, see Ssu-yü Teng, John K. Fairbank, E-tu Zen Sun, Chaoying Fang, et al., China's Response to the West: A Documentary Survey, 1839-1923; and Edwin O. Reischauer, John K. Fairbank, and Albert M. Craig, A History of East Asian Civilization. For a critical analysis of Fairbank's and related works, see Paul A. Cohen, Discovering History in China: American Historical Writing on the Recent Chinese Past, chap. 1.

[6]Needham proposes a retrospective competition between the West and China, fixing dates of discovery through a "grand titration" that compares "the great civilizations against one another, to find out and give credit where credit is due, and so. . . to see why one combination could far excel in medieval times while another could catch up later on and bring modern science into existence" (Joseph Needham, The Grand Titration: Science and Society in East and West, 12).

[7]Joseph Levenson, Liang Ch'i-ch'ao and the Mind of Modern China; idem, Confucian China and Its Modern Fate. I am indebted to Wen-hsin Yeh for her suggestions on these and other points.

[8]For examples, see the early issues of Bulletin of Concerned Asian Scholars (1968- ), esp. James Peck "The Roots of Rhetoric: The Professional Ideology of America's China Watchers"; John K. Fairbank and James Peck, "An Exchange"; and Joseph Esherick, "Harvard on China: The Apologetics of Imperialism". For an analysis of these and other studies critical of imperialism, see Cohen, Discovering History in China, chap. 3.

[9]The China-centered approach is advocated in Cohen, Discovering History in China, chap. 4. It should be noted that Cohen explicitly asserts that China should not be treated in isolation (196). Cohen's China-centered approach, in fact, decenters China: the approach he advocates "disaggregates China `horizontally' into regions" (following G. William Skinner); it "disaggregates Chinese society `vertically' into a number of discreet levels"; and it encourages interdisciplinary theoretical analyses (186). For a criticism of Sinocentrism, see Philip Huang, "Theory and the Study of Modern Chinese History: Four Traps and a Question."

[10]An important exception to these examples is the work of Benjamin Schwartz. See Schwartz, "The Limits of `Tradition Versus Modernity' as Categories of Explanation"; idem, In Search of Wealth and Power: Yen Fu and the West, chap. 1.

[11] Important studies include Jonathan D. Spence, The Memory Palace of Matteo Ricci; Charles E. Ronan and Bonnie B. C. Oh, eds., East Meets West: The Jesuits in China, 1582-1773; John D. Young, East-West Synthesis: Matteo Ricci and Confucianism; David E. Mungello, Leibniz and Confucianism: The Search for Accord; and idem, Curious Land: Jesuit Accommodation and the Origins of Sinology. For important criticisms of the received historiography, see Lionel M. Jensen, Manufacturing Confucianism: Chinese Traditions and Universal Civilization.

[12] For example, see Journal of Asian Studies 50, no. 1 (February 1991), which is dedicated to the issue of relativism, and esp. David D. Buck's summary, "Forum on Universalism and Relativism in Asian Studies: Editor's Introduction," 29-34. For Philip Huang, "culturalism" and the resulting relativism is one of his "four traps" (Huang, "Theory and the Study of Modern Chinese History," 192-201). For collected essays on the debates on relativism, see Martin Hollis and Steven Lukes, eds., Rationality and Relativism, and Michael Krausz, ed., Relativism: Interpretation and Confrontation.

[13] Jacques Gernet, China and the Christian Impact: A Conflict of Cultures; originally published as Chine et christianisme: action et réaction (1982).

14 Jean-Claude Martzloff, A History of Chinese Mathematics; originally published as Histoire des mathématiques chinoises (1987).

[15]Gernet, China and the Christian Impact, and Spence, Memory Palace of Matteo Ricci, are generally recognized as the two seminal works describing this period. For two important critiques of Gernet's book, see Paul Cohen's review in Harvard Journal of Asiatic Studies, and Howard L. Goodman and Anthony Grafton, "Ricci, the Chinese, and the Toolkits of Textualists."

[16]Gernet, China and the Christian Impact, 1-2.

[17]Ibid., 2. For criticisms of assertions that individuals are the bearers of whole cultures, see Johannes Fabian, Time and the Other: How Anthropology Makes Its Object.

[18]Gernet, China and the Christian Impact, 193.

[19]Ibid., 201.

[20]Ibid., 208.

[21]Ibid., 240. Gernet cites Emile Benveniste, "Catégories de pensée et catégories de langue"; reprinted in Problèmes de linguistique générale.

22These two claims are in fact related, since Benveniste's central example of categories of thought is based on the copula.

[23]Gernet, China and the Christian Impact, 240. Gernet quotes Benveniste's assertion that the existence of the verb to be in Greek made possible the philosophical manipulation of the concept, with the result that the concept of being became central in Greek thought.

[24]In addition to Martzloff, History of Chinese Mathematics, 111-122, 273-277, 371-389, see idem, "Matteo Ricci's Mathematical Works and Their Influence"; idem, "Eléments de réflexion sur les réactions chinoises à la géometrie euclidienne à la fin du XVIIe siècle: le Jihe lunyue de Du Zhigeng vu principalement à partir de la préface de l'auteur et de deux notices bibliographiques rédigées par des lettrés illustres"; idem, "La géométrie euclidienne selon Mei Wending"; idem, "La compréhension chinoise des méthodes démonstratives euclidiennes au cours du XVIIe siècle et au début du XVIIIe." For a more detailed study of the translation that supersedes Martzloff's work, see Peter M. Engelfriet, "Euclid in China: A Survey of the Historical Background of the First Chinese Translation of Euclid's Elements (Jihe Yuanben, Beijing, 1607), an Analysis of the Translation, and a Study of Its Influence up to 1723."

[25]Martzloff, History of Chinese Mathematics, 116-18. Martzloff's French original: "Le cerle? Une forme située sur la terre plate (ping di) (sic) entre de la limite! Les `fils' (xian) droits construits de la limite au centre? Tous égaux!" (Histoire des mathématiques chinoises, 103). Sic is in the originals, both English and French.

[26]Martzloff, History of Chinese Mathematics, 118; Martzloff quotes from Clavius, Euclidis Elementorum, bk. 1, def. 7.

[27]Martzloff, History of Chinese Mathematics, 118. In the original French edition, Martzloff states directly that the central problem is of "the concept of existence": "Ce type de phénomène rendait pour le moins hasardeuse la transmission du concept d'existence si important en mathématiques (parallèles, constructions géométriques, raisonnements par l'absurde dans lesquels on prouve qu'un certain objet mathématique n'existe pas)" (Martzloff, Histoire des mathématiques chinoises, 103).

[28]It should be noted that a recent article by Martzloff instead asserts that it was a commensurability between the European and Chinese concepts of space and time that led to the rapid acceptance of Jesuit astronomy (67). Yet incommensurability thesis remains in Martzloff's claim that the Chinese and Europeans had "fundamentally different orientations" (82-83). Gernet, in the following article, argues that the commensurability that Martzloff argues for is only apparent. Martzloff, "Space and Time in Chinese Texts of Astronomy and Mathematical Astronomy in the Seventeenth and Eighteenth Centuries," and Gernet, "The Encounter between China and Europe."

[29]A. C. Graham, "`Being' in Western Philosophy compared with shih/fei yu/wu in Chinese Philosophy." Graham's conclusions are in fact in many ways the opposite of Martzloff's. Martzloff also cites Gilles Granger, La théorie aristotélicienne de la science.

[30]Peter A. Boodberg, "Philological Notes on Chapter One of the Lao Tzu," 601.

[31]Alfred Bloom, The Linguistic Shaping of Thought: A Study in the Impact of Language on Thinking in China and the West. For a critical review, see Kuang-ming Wu, "Counterfactuals, Universals, and Chinese Thinking--A Review of The Linguistic Shaping of Thought: A Study in the Impact of Language on Thinking in China and the West."

[32]For a recent summary--and one further example--of these arguments, see Derk Bodde, Chinese Thought, Society, and Science: The Intellectual and Social Background of Science and Technology in Pre-Modern China.

[33]Two important recent studies are Haun Saussy, The Problem of a Chinese Aesthetic, and Lydia H. Liu, Translingual Practice: Literature, National Culture, and Translated Modernity--China, 1900-1937. In their first chapters both works provide critical discussions of much of the literature on philosophy and translation.

[34]Emile Benveniste, Problems in General Linguistics, 56; originally published as Probèmes de linguistique générale (1966).

[35]Ibid., 53.

[36]Ibid., 56.

[37]Ibid., 60.

[38]Ibid., 61; emphasis in original.

[39]Jacques Derrida, "The Supplement of Copula: Philosophy before Linguistics," in Textual Strategies: Perspectives in Post-Structuralist Criticism.

[40]Ibid., 114. Cf. Boodberg's discussion: "(d)jwer (modern wei), a copula-like particle common in the language of the Shih and Shu. If this be the lost Chinese verb `to be'. . ." (Boodberg, "Chapter One of the Lao Tzu," 603).

[41]In this sense, grammatical peculiarities of the Indo-European languages have influenced the development of philosophy within those languages, by the conflation of to be and existence.

[42]Martin Heidegger, An Introduction to Metaphysics, 58-61; quoted in Derrida, "Supplement of Copula," 118, with Derrida's italics.

[43]Benveniste, Problems in General Linguistics, 138; quoted in Derrida, "Supplement of Copula," 119, with Derrida's italics.

[44]Derrida, "Supplement of Copula," 82.

[45]Ibid., 98.

[46]It should also be noted that elsewhere in his argument asserting that language determines categories of thought, Benveniste uses the example of science and the Chinese language to argue against linguistic incommensurability: "Chinese thought may well have invented categories as specific as the dao, the yin, and the yang; it is nonetheless able to assimilate the concepts of dialectical materialism or quantum mechanics without the structure of the Chinese language proving a hindrance. No type of language can by itself alone foster or hamper the activity of the mind. The advance of thought is linked much more closely to the capacities of men, to general conditions of culture, and to the organization of society than to the particular nature of a language. But the possibility of thought is linked to the faculty of speech, for language is a structure informed with signification, and to think is to manipulate the signs of language" (Benveniste, Problems in General Linguistics, 63-64).

[47]Willard V. O. Quine, "Meaning and Translation," in On Translation, 171-172.

[48]Donald Davidson, "On the Very Idea of a Conceptual Scheme," in Inquiries into Truth and Interpretation.

[49]Benjamin Whorf, "The Punctual and Segmentative Aspects of Verbs in Hopi," in Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf; Thomas S. Kuhn, The Structure of Scientific Revolutions; Paul Feyerabend, "Explanation, Reduction, and Empiricism," in Scientific Explanation, Space, and Time; Willard V. O. Quine, "Two Dogmas of Empiricism," in From a Logical Point of View. Davidson argues that Quine's conceptual schemes constitute yet a "third dogma." For Quine's response, see "On the Very Idea of a Third Dogma."

[50]This is precisely the contradiction found in Martzloff--his assertion of incommensurability between Chinese and Latin itself requires a solution to the problem of incommensurability in a firm ground for philosophical translations between Greek, Latin, English, Chinese, and French, a solution that could not have existed four hundred years ago at the time of the translation.

[51]Davidson, "On the Very Idea of a Conceptual Scheme," 184.

[52]Ibid., 185.

[53]Davidson, Inquiries into Truth and Interpretation, xviii. More precisely, Davidson states, "Our general method of interpretation forestalls the possibility of discovering that others have radically different intellectual equipment. But more important, it is argued that if we reject the idea of an uninterpreted source of evidence no room is left for a dualism of scheme and content. Without such a dualism we cannot make sense of conceptual relativism" (xviii).

[54]Davidson, "On the Very Idea of a Conceptual Scheme," 189.

[55]See Kuhn, Structure of Scientific Revolutions; for a later formulation, see idem, "Second Thoughts on Paradigms."

[56]Paul Feyerabend, in Against Method, 170-226, offers as an example of incommensurability two contrasting forms of Greek art that he identifies as cosmologies A and B. A is the "archaic style" in which a "paratactic aggregate" forms a "visual catalogue" through the placement of standard figures in varying symbolic positions--for example, death is the standard figure drawn horizontally (Feyerabend cites Emanuel Loewy, Die Naturwiedergabe in der älteren Griechischen Kunst). In B, art is arranged so that the underlying essence is grasped through representations that trigger illusions (e.g., two-dimensional drawings). Feyerabend then defines incommensurability: "A discovery, or a statement, or an attitude [is called] incommensurable with the cosmos (the theory, the framework) if it suspends some of its universal principles" (215).

[57]For example, Whorf asserts that the adoption of Western science entails the adoption of the Western system in its entirety: "That modern Chinese or Turkish scientists describe the world in the same terms as Western scientists means, of course, only that they have taken over bodily the entire Western system of rationalizations, not that they have corroborated that system from their native posts of observation" (Whorf, Language, Thought, and Reality, 214).

[58]Quine's famous analogies are from boundary-value problems in physics: "The totality of our so-called knowledge or beliefs. . . is a man-made fabric which impinges on experience only along the edges"; "total science is like a field of force whose boundary conditions are experience" (Quine, "Two Dogmas," 42, quoted in Davidson, "On the Very Idea of a Conceptual Scheme," 191).

[59]Arguably, it is the purported correlations of words with things that underwrites claims of deep ontological differences. Claims of radical differences are based on the assertion that one language fails to have a word for a particular phenomenon, not that the phenomenon cannot possibly be described.

[60]John E. Murdoch, "Editions of Euclid" in Dictionary of Scientific Biography.

[61]On Clavius's mathematics and science, see James M. Lattis, Between Copernicus and Galileo: Christoph Clavius and the Collapse of Ptolemaic Cosmology.

[62]The example is from Schleicher, quoted in Whitney, 331, quoted and criticized in J. R. Firth, "Linguistic Analysis and Translation," in Selected Papers of J. R. Firth 1952-59, 76. For analysis of this translation, see Haun Saussy's article in this volume; I would like to thank him for bringing this passage to my attention.

[63]Martzloff, History of Chinese Mathematics, 118.

[64]For criticism of Martzloff's translations, see the review of his Histoire des mathématiques chinoises by Catherine Jami in Historia Scientiarum.

[65]Martzloff, History of Chinese Mathematics, 273.

[66]Ibid., 274. For the French, see n. 24 above.

[67]Gernet, China and the Christian Impact, 240.

68Tianzhu shi yi, in Tianxue chu han (TXCH); translated into English as Matteo Ricci, The True Meaning of the Lord of Heaven (T'ien-chu Shih-i). Several of the examples that I analyze are in fact cited or translated in Gernet, China and the Christian Impact, and provide important evidence against the conclusions he draws in the final chapter.

[69]Tianzhu shi yi, 450; Ricci, Lord of Heaven, 34-36. In the following translations I have consulted Lancashire and Hu's translations, borrowing and making alterations where appropriate.

[70]Tianzhu shi yi, 451-452; Ricci, Lord of Heaven, 176-177. Just as the asserted absence of a word precisely equivalent to the English verb to be is hardly evidence for broader philosophical conclusions such as conceptual incommensurability (as I have argued above), so too the translations I offer in this article of various Chinese phrases into English terms such as existence, soul, and forms of the verb to be are not meant to suggest a transparency of translation or to imply the exact equivalence of Chinese and English words or concepts (I would like to thank Marta Hanson for her suggestions on this point). The (im)possibility of translation cannot demonstrate either radical incommensurability or exact correspondence; these are questions not of linguistics but of philosophical interpretation and intellectual history.

[71]For an important earlier example, see the chapter "Gui shen" [Ghosts and spirits] in Zhuzi yu lei [Conversations with Master Zhu, arranged topically].

[72]Tianzhu shi yi, 452; Ricci, Lord of Heaven, 178-179. This passage is also translated in Gernet, China and the Christian Impact, 68: "Some say that they definitely do not exist. Others say that they do exist if one believes in them and do not if one does not believe in them. Others say that it is just as false to say that they do exist as to say that they do not and that the truth is that they both do and at the same time do not exist" (emphasis mine). Gernet concludes that these debates were "hardly. . . to the missionaries' liking," but does not explain why this translation does not controvert his claims in his concluding section, "Language and Thought," and in particular his central assertion that "there was no word to denote existence in Chinese, nothing to convey the concept of being or essence, which in Greek is so conveniently expressed by the noun ousia or the neuter to on" (241).

[73]The entire passage states, "Fan Chi inquired about intelligence. Confucius stated: `Endeavor to make the people righteous, respect the spirits and distance them; this can be termed intelligence.' [Fan Chi also] inquired about benevolence. [Confucius] stated: `The benevolent face difficulties first and obtain afterwards; this can be called benevolence.'" (Lun yu [Confucian analects] 6.22, in Shisan jing zhu shu (SSJZS) [Thirteen classics with annotations and subannotations], 2:2479). Thus in the comparison Confucius appears to be exhorting Fan Chi to concentrate on the affairs of the living rather than spirits. See also Lun yu 11.12, SSJZS 2:2499.

[74]Lun yu 3.13, SSJZS 2:2467.

[75]Tianzhu shi yi, 1.468; Ricci, Lord of Heaven, 202-203.

[76]Tianzhu shi yi, 461; Ricci, Lord of Heaven, 190; Gernet, China and the Christian Impact, 243.

[77]Jacques Gernet, "Christian and Chinese Visions of the World in the Seventeenth Century," Chinese Science 4 (September 1980): 1-17.

[78]Ibid., 15.

[79]Ibid., 16.

[80]For example, Gernet notes, "The difference between the philosophical ideas inherited from Ancient Greece and those of the Chinese emerges clearly here" (Gernet, China and the Christian Impact, 210).

[81]From Gernet's concluding sentence to "Christian and Chinese Visions of the World," 17.

[82]The most sustained attempt to formulate such a theory is Michel Foucault, The Archaeology of Knowledge; originally published as L'archéologie du savoir (1969). For a critique, see Hubert L. Dreyfus and Paul Rabinow, Michel Foucault: Beyond Structuralism and Hermeneutics. The theoretical reformulation presented in Archaeology of Knowledge should be contrasted with Foucault's earlier historical works.

[83]Mario Biagioli, "The Anthropology of Incommensurability," 184.

[84]Incommensurability, Biagioli asserts, is often associated with instances of trespassing professional or disciplinary boundaries. He concludes against Kuhn that while the possibility of developing bilingualism is not logically wrong, the assumption is unwarranted in ignoring "the fundamental relation between social groups and cognitive activity" (ibid., 207).

[85]Robert S. Westman, "The Copernicans and the Churches," in God and Nature, ed. D. C. Lindberg and R. Numbers, esp. 90-91.

[86]Mario Biagioli, "Stress in the Book of Nature: Galileo's Realism and Its Supplements." Biagioli argues that "Galileo's so-called mathematical realism was, in fact, a form of scriptural fundamentalism" (2).

[87]For an important theoretical analysis of the problem of translation between Chinese and Western languages and a critical overview of theories on translation, see Liu, Translingual Practice, esp. chap. 1. These issues will not be further developed here.

[88]For an analysis of translation and a discussion of terminology, see Theodora Bynon, Historical Linguistics, chap. 6; see also John Lyons, Semantics.

[89]F. Sambiasi, "Ling yan li shao yin" [Preface to "A preliminary discussion of anima"], in TXCH, 2.1127. The phrases in parentheses are commentaries, distinguished from the original text by half-width characters.

[90]Tianzhu shi yi, 1.381; Ricci, Lord of Heaven, 70.

[91]For examples of the translation of Christian doctrines into European languages, see Bynon, Historical Linguistics, chap. 6.

[92]"Tianzhu shi yi yin", Tianzhu shi yi, 1.370; Ricci, Lord of Heaven, 62.

[93]Tianzhu shi yi, 1.390-391; Ricci, Lord of Heaven, 84-86.

[94]Tianzhu shi yi, 1.406; Ricci, Lord of Heaven, 108. Hu and Lancashire note that the current terms are ziliti and yifuti; see Ricci, Lord of Heaven, 108, n. 18.

[95]For a comprehensive table of Euclidean terms translated by the Jesuits, along with their modern equivalents, see Engelfriet, "Euclid in China," 164-167.

[96]Presumably, the most difficult of the sacraments to explain would have been the Eucharist--that bread and wine were in actuality the body and blood of Christ.

[97]Yang Guangxian later accused the Jesuits of frequently failing to mention the crucifixion, arguing that the Jesuits wished to deliberately conceal that the Lord they worshiped was nothing more than a criminal during the period of the Han dynasty.

98These examples are from Lancashire and Hu, introduction to Ricci, Lord of Heaven, 35-36.

[99]Gernet, China and the Christian Impact, 25.

[100]Xu Guangqi uses this term to characterize the Jesuits' mission in China in a memorial to the Ming court. The phrase Shi tian appears in the Li ji [Records of rites] (SSJZS, 2:1612); xiu shen appears in several passages from the Li ji, and most notably, shen xiu appears in Da xue. See Zhu Xi (1130-1200), Da xue zhang ju [The great learning, separated into chapters and sentences] in Si shu zhang ju ji zhu [Four books, separated into chapters and sentences, with collected annotations], 3.

101Tianzhu shi yi, 359, 359, 366 and 369, respectively. Other terms were also used for God in more specific contexts, such as Dayuan (Tianzhu shi yi, 1.399; Ricci, Lord of Heaven, 96). Lancashire and Hu note that Yuan was sometimes used by Ricci, along with Yuan, "to express `source' or God as Creator" (Ricci, Lord of Heaven, 96, n. 32). As noted by Lancashire and Hu, these terms were sometimes preceded by a blank space to denote respect; the terms Zhongguo and Zhonghua were also preceded with a blank space (Tianzhu shi yi, 367). To denote this, I have capitalized these terms in pinyin; however, it should be noted that the addition of a blank space in the Chinese text is not used uniformly throughout.

[102]F. Sambiasi and Xu Guangqi, "Ling yan li shao" [A preliminary discussion of anima], in TXCH.

[103]For examples of Jesuit claims of Buddhist distortions, see E. Zürcher, "The Jesuit Mission in Fujian in Late Ming Times: Levels of Response," in Development and Decline of Fukien Province in the 17th and 18th Centuries, ed. E. B. Vermeer, 418 n. 3.

[104]In 1704 Pope Clement XI banned the use of tian and shangdi as translations for God. Lancashire and Hu, introduction to Ricci, Lord of Heaven, 20.

[105]Gernet, China and the Christian Impact, 28.

[106]It should be noted that the specific mechanisms of patronage in China are in many ways different from those found in Europe. Perhaps the most important difference is the examination system and the official bureaucracy and the way that these structured patronage in the Ming court.

[107]Ray Huang, 1587, A Year of No Significance: The Ming Dynasty in Decline, 50.

[108]In Tianzhu jiao Dong chuan wenxian, xu bian [Documents on the Eastern transmission of Catholicism, second collection], 1:25-26. In his introduction to this collection, Fang Hao states that this reproduction is printed from a Ming edition preserved at the Vatican Library. Fang notes that there are many editions of this treatise, and many are incomplete. Fang offers several examples of editors that removed portions of the above quotation (3-4); the edition copied by Wang Zhongmin in XGQJ omits parts of this passage (431).

[109]This passage is translated in Gernet, China and the Christian Impact, 110; I have made minor alterations to his translation where appropriate.

[110]Ibid., 110-111.

[111]Fang Hao, Zhongguo Tianzhujiaoshi renwu zhuan [Biographies of Catholics in China], 103.

[112]XGQJ 2.492. For an English translation of these letters, see Gail King, "The Family Letters of Xu Guangqi."

[113]Xu's career followed the pattern of exile and returns to power typical of the period; with the ascendance of Sizong in 1628, Xu returned to power. In the last year of his life, 1633 (Chongwen 6), he reached one of the highest posts in government, the Grand Guardian of the Heir Apparent, and Grand Secretary of the Hall of Literary Profundity (Taizi taibao wenyuange daxueshi).

[114]This argument is presented in detail in my book manuscript "Proof, Propaganda and Patronage: The Dissemination of Western Studies in Seventeenth Century China."

[115]Important examples include Jack Goody, The Logic of Writing and the Organization of Society, and Bodde, Chinese Thought, Society, and Science.