|
|
|
|
China's Transition, by Andrew J. Nathan
Anyone who works in the field of area studies knows from experience that cultures are different. Indeed, the effort to understand the distinctiveness of cultures in comparative perspective is a central undertaking of the modern humanities and social sciences, not only in Asian studies but also in studies of other parts of the world. But works on the subject seldom discuss the conceptual and methodological issues involved. What do we mean by culture in the context of comparative statements? How can a culture's distinctiveness be conceptualized? What is required to demonstrate that such distinctiveness exists, what it consists of, and what influence it has on the performance of societies? In the case of Chinese studies, how far have we come in establishing that Chinese culture is distinctive, in what ways, and with what consequences?
It is helpful to discuss these issues in terms of two bodies of literature with different ways of conceptualizing culture and its distinctiveness, although I intend to blur the distinction at the end. 1 Following Ying-shih Yü, I will label the two approaches hermeneutic and positivistic. 2 I do not argue that one of the approaches is better than the other; each achieves goals that the other does not. The real problem is lack of clarity about the different logical statuses of the kinds of findings that typically emerge from the two approaches. This can lead to problems when insights are transposed from the hermeneutic approach into positivistic language or vice versa.
Although I will suggest two meanings of distinctiveness, both involve identifying a culture's differences from another culture or cultures along one or more dimensions. In analyzing differences, two operations are involved: abstraction (of a characteristic to be compared) and comparison. These moves are often made in order to take a third step: to use the differences in culture to explain a difference in some societal outcome, such as economic development or democratization.
In the course of these operations, knowledge about a single culture is sometimes restated as knowledge about two cultures, hermeneutic knowledge as positivistic knowledge, insight about what makes a society unique as comparison of differences among cultures. A reverse error is made when comparative differences among cultures are treated as absolute differences that place each society beyond comparison. In either of these cases, we learn something about what Chinese culture is like, but gain unreliable information about its distinctiveness. In this review essay, I will argue that we can enjoy the potential benefits of living in two methodological worlds, but only if we avoid pitfalls in negotiating back and forth between the two.
Hermeneutic Approaches to Distinctiveness
The hermeneutic approach views culture as a historically shaped, socially shared set of symbols, concepts, and ways of organizing them. The major concern of hermeneutic works is interpretation. Hermeneutic methods vary, but have in common the attempt to elucidate meaning in a text or a text-analogue (such as a pattern of belief or belief-revealing behavior) by paying attention to the text's context and its inner structure. This family of methods seeks to understand culture by exploring its "pattern of meanings," in the phrase of Clifford Geertz, or by sympathetically entering what Benjamin Schwartz calls its "world of thought" to explore the "problematiques" of its thinkers and discover their "shared cultural assumptions." 3 The approach predominates in all but two of the books reviewed here, reflecting the fact that the positivistic approach remains relatively underdeveloped in Chinese cultural studies for a variety of reasons.
A common theme of hermeneutic works is cultural identity, which is the chief subject of two works under review here. Wang Gungwu's book, a collection of sixteen previously published lectures and essays, describes "the Chineseness of China" from many angles, as a civilization, a place, a society, and a people. 4 The book contains chapters on Tang geopolitics, historiography in the imperial and communist periods, China's relations with its neighbors, the roles of Chinese intellectuals, and Mao Zedong. Wang builds up a picture of Chinese identity as ever-changing but rooted in an awareness of its own past. He concludes that "the Chineseness of China" is whatever went on among Chinese, so that "there is nothing absolute about being Chinese" (p. 266).
Much of the material in "The Living Tree" likewise addresses the question of what it means to be Chinese. 5 In his introductory essay, Tu Wei-ming argues that the qualifying factor is participation in the "symbolic universe" of Chinese civilization, even if the participant is living in the diaspora or is ethnically non-Chinese. The nine essays range from Mark Elvin's evocation of "the inner world of 1830" to meditations on the roles of the intellectuals and the overseas Chinese communities in what one author calls "the construction of Chinese identity." Chinese identity proves to be so elusive that the participants conclude that "the only relevant criterion of identity is the self-identity perceived by a person" (Hsu Cho-yun in end matter, no page). In Myron Cohen's words, where once there was a common Chinese culture, today the identity of "being Chinese is no longer buttressed by a firm sense of cultural participation in something Chinese," so that Chineseness has become "as much a quest as a condition" (p. 133).
If cultural identity is a subjective psychological affiliation that members of a society can accept, reject, or change, one can validly make only certain kinds of comparative statements about it. One can say, as Tu Wei-ming does, that identity has been more of a problem for Chinese than for Indian intellectuals (p. 2), or that Hong Kong people's Chinese identity is stronger than Singapore people's (p. 11). But one cannot lay out objectively what Chineseness consists of and say who is more Chinese than whom. For example, we should not read literally Wang Gungwu's statement that "The most Chinese thing about Mao Tse-tung was his poetry and his loyalty to its traditional forms" (p. 261), as if Wang were suggesting that a person who considered himself Chinese but was less loyal to traditional poetry than Mao would actually be less Chinese. For, as Wang says a few pages later, "an American-educated engineer who subscribed to Jeffersonian democracy and loyally served the Canton Government under Sun Yat-sen [was not] any less Chinese than Mao" (pp. 265-66). Like Tu and his contributors, Wang builds a mosaic of ways of being Chinese rather than a list of Chinese characteristics.
The hermeneutic approach also can be used to compare cultures objectively, as it is in two other books under review. Benjamin Schwartz's book consists of two lectures given at the University of Arizona in 1982. 6 In the first, in order to explore the "qualitative differences between cultures, which endure over time" (p. 3), he compares the civilization of ancient China to those of ancient India, the Middle East, Greece, and the Jewish world. As distinctive features of Chinese civilization, he identifies, among others, ancestor worship, the "religious quality of Chinese familial life," the theme of universal kingship, the familial model of the Chinese socio-political order, the unity of ruling and teaching, the "primacy" and "sacred" quality of the political order, the "faith that the good order had been realized in the past," and a sense of total order. The second lecture explores continuity and change in Chinese culture from the ancient period to the modern.
The other, Lucian Pye's book, is an effort to identify attributes of Chinese political culture that help explain why China has had a harder time with political development than some other countries. 7 He put his classic work back into print, with two new chapters, with the conviction that "the special importance of political culture for understanding China . . . lies in the ways in which China is unique at both the collective and the individual levels" (p. ix). To understand political culture, he uses an approach that he describes in a later work as "interpretive political culture studies." 8 As he explained in Spirit 's original preface (not reprinted in the new edition), in this method "our concern is to describe the constellation of sentiment and attitudes that we feel must have existed for the Chinese political system to have developed as it has. We are not concerned with questions about the actual distribution of attitudes and feelings throughout the Chinese population" (orig. ed., p. viii).
As with the works on identity, comparative statements in these works should be read in light of their conceptualization of culture. Although their approaches differ in many ways, Schwartz and Pye have in common that they conceptualize culture as a complex of attributes defined so specifically that they are not found in the same form anywhere else. When Pye speaks about the management of aggression as a central problem in Chinese political culture, he does not mean that aggression is unique to China, but that in China the problem is "particularly acute . . . because the impact of the modern world . . . disrupted the elaborate mechanisms by which the drives of aggression had been traditionally repressed" (pp. 33-34). When Schwartz singles out as distinctive "the religious quality of Chinese family life," he does not mean that in no other ancient civilization did family life have much of a religious quality at all, but that no other ancient civilization had a religious quality of family life quite like China's.
Similarly, when Schwartz argues that ancestor worship was a distinctive characteristic of ancient Chinese society, the statement should be understood in the following sense, as no doubt it was intended. Religious worship exists in all societies. Ancestor worship is a form of religious worship that exists in many societies. Specifically Chinese-style ancestor worship existed only in China. The finding of China's distinctiveness is a function of the level of specificity at which Schwartz has conceptualized his object of comparison, Chinese-style ancestor worship. By the same token, one could go on to say that ancestor worship in the Qing was different from ancestor worship in the Ming, and ancestor worship in Guangdong during the Qing different from ancestor worship in Hebei during the Qing, and so on.
Comparison rests on a prior operation of classification. Objects can properly be compared only with objects belonging to the same class, except to make the trivial point that they belong to different classes. If that class is populated in one society and empty in others, then the object is properly said to be unique, i.e., "distinctive" in the first of two senses of that word that I want to identify. But this finding is dependent on the level of generality at which the class is defined. A sugar doughnut is different from a jelly doughnut, but both are doughnuts, and doughnuts are cakes, and cakes are sweets. Although each object is unique, it can also be conceptualized as belonging to a class within which it is not unique.
Uniqueness is thus a function of location on what Giovanni Sartori has called the "ladder of abstraction." Moving up the ladder is a process of simplification, reducing the number of definitional attributes. Moving down the ladder is a process of complication, rendering the object more complex and hence more specific. 9 For a given object, at a given level on the ladder, a finding of uniqueness is preordained. In this sense it is true by definition, not in the sense that it is obvious to the ignorant but in the sense that it is a consequence of conceptualization. In effect, such finding of uniqueness is not a finding about a society's culture in comparative perspective, but a finding about one's way of defining the cultural complex one has selected for comparison.
We already knew that two cultures were not identical when we identified them as two cultures rather than one. A true-by-definition finding of uniqueness tells us in more detail how the two cultures are not the same. This is often worthwhile. But it may not take us as far as we often want to go. For a variety of reasons (discussed further below) we may also need to ask how much the two cultures differ, how they compare rather than how they are unique; in other words, how they are "distinctive" in the second sense of the word. In this second sense, unlike the first, the attribute exists in both cultures. We want to know about the degree of difference between the two in the attribute's quantity (extent, degree, etc.), quality (intensity, functional importance, etc.), distribution (sector of society, geographical location), or relationship (pattern of association with other attributes). From the level of abstraction at which one makes a finding of uniqueness, one has to move to a higher level of abstraction to make such findings about difference.
To anticipate a point developed further below, it would be fallacious either to restate a finding of uniqueness as a finding of difference, or to restate a finding of difference as a finding of uniqueness. To say Chinese aggression expresses itself in unique ways is not to say the Chinese are more aggressive than others; to say the Chinese are more aggressive than others is not to say they are unique in possessing the psychological drive of aggression.
Perhaps because so many of the contrasts noted in hermeneutic works are true by virtue of conceptualization, they are seldom given close specification as to time, place, and social group. Schwartz warns that culture is not unitary or changeless, that the existence of "dominant" and "persistent orientations" does not rule out the existence of alternative views, that there are class differences and differences between high and popular cultures. Yet he, along with others using this approach, seldom states explicitly which class or group of people he believes held which attitude at what time. 10 In addition, many writers in this tradition treat similarities across time as continuities, without attention to the alternative possibilities that they are parallel developments or not even really similar, and without asking how continuity was created. 11 Authors often discuss institutions as if they too were culture, setting up a tautological relationship between beliefs and behaviors, in which the behaviors are both the evidence for the attitudes and their purported consequence.
In many hermeneutic works the comparison culture--the object of contrast--is not seriously analyzed and is reduced to a stereotype. Wang Gungwu's chapter on "Power, Rights, and Duties in Chinese History" is an exception; it pauses to give nuanced, if brief, definition to the Western versions of the concepts with which he contrasts Chinese ideas. An even more fully developed example of such an exception is Rodney Taylor's careful comparison of the ideas of the Confucian sage and the Christian saint, in which he uses careful hermeneutic analysis to show that, although the two concepts are different in specifics, at a higher level of abstraction they have enough in common to support the view that Confucianism has an important "religious dimension." 12 By developing both sides of a comparison, such works use hermeneutic methods to achieve findings about difference rather than findings of uniqueness.
Positivist Approaches to Distinctiveness
Positivist approaches to cultural distinctiveness produce not better but different answers, having their own limitations. Contrary to caricatures by some critics, the kind of positivism influential today is not the version that flourished through the early 1960s, which sought natural science-like laws of history and behavior based on the comparison of epistemologically unproblematic, objective facts. It is a diminished positivism that requires only that any proposition be precisely specified (for example, with respect to time frame, social actors, geographical location, and indicators of amount or degree), that it be stated in a form that is potentially disprovable by reference to empirical evidence (the falsifiability criterion), and that it be treated as unproven until empirically proven. In order to meet these requirements, positivists conceptualize culture as a distribution among a population of specifiable and identifiable attitudes, values, and beliefs.
For cultural comparison, positivist approaches have advantages and disadvantages. The differences between the two approaches can be illustrated with examples from some recent studies that use survey research, a popular positivist method that has recently come into use in Chinese studies, to investigate some themes that are also treated in the hermeneutic literature.
For example, one of the most common hermeneutic findings about the distinctiveness of Chinese culture is that it lays greater stress than other cultures on the use of guanxi (personal connections), the Chinese term for certain kinds of particularistic ties between pairs of people. In the Tu volume, Ambrose King argues that "to know and practice guanxi is part of learned behavior--of being Chinese" (p. 79). Fei contrasts Western societies based on organizations with the Chinese pattern, in which "our social relationships spread out gradually, from individual to individual, resulting in an accumulation of personal connections [which] form a network" (p. 70). 13 Survey research in both the United States and China included the following question: "From time to time, most people discuss important matters with other people. Looking back over the last six months, who are the people with whom you discussed matters important to you?" By collecting data on the backgrounds of these people, researchers were able to compile information on social networks in the two societies. Among other findings, Chinese persons' networks are twice as large as Americans'; Chinese are more likely to associate with people of the same age and educational level; and Chinese are less likely than Americans to mention kin as important associates. In explaining these differences, the researchers referred to the importance of the work unit in Chinese society, which dictates that most Chinese have constant, intimate, and important relations with colleagues at work. 14 The findings tend to support the notion that guanxi are more important in China than in America, but also reveal that these ties are more instrumental than personal in nature. 15 Although China-U.S. differences exist, the comparison also demonstrates that networks built out of personal relationships are not unique to China. 16
Another theme treated in both hermeneutic and positivisitic studies is personalistic factionalism, or the use of guanxi in politics. Pye, for example, considers guanxi to be "the key ingredient of [Chinese] factions" (p. 212). In a 1985 Hong Kong survey by S. K. Lau and H. C. Kuan, respondents were asked to name effective ways to influence government policy. One-point-eight percent spoke of making use of relatives who are officials; 18.1 percent mentioned writing letters to officials or meeting them in person; 25.7 percent mentioned mobilizing and organizing the people affected; 14.7 percent mentioned exercising influence through political parties and other political organizations; 3.8 percent said they would demonstrate and protest; 1.8 percent gave other methods, and the rest said there was no way or did not answer. This can be compared with data from the 1959-60 Civic Culture surveys in five countries in which citizens were asked what they would do to try to influence their local government. None of the respondents mentioned working through relatives who were officials; the percentages in the U.S., U.K., Germany, Italy, and Mexico who said they would organize an informal group of friends and neighbors ranged from a low of 7 percent to a high of 56 percent depending on the country; the percentages who spoke of directly contacting elected officials or the press ranged from 12 percent to 45 percent; and a range of other items was mentioned. 17 The findings show that the cultural options for political action are distinctive in each of the six places, yet the range of possible approaches to political influence is basically similar. The distribution does not support the idea that the Chinese in Hong Kong are partial to the use of guanxi in politics to a striking degree.
A third theme treated in both kinds of literature is intolerance. Pye argues that Chinese political culture demands obeisance to whatever ideology is espoused by the current rulers (pp. 13-16). In a 1990 national survey in mainland China, 18 the following situation was posed: "There are some people whose ideology is problematic; for example, they sympathize with the Gang of Four." The respondent was then asked whether he or she thought that such people should be allowed to express their views in a public meeting, as a teacher in college, and by publishing articles or books. This can be compared with a question about "people who want to overthrow the government by revolution" asked by the International Social Survey Program in 1985 in six countries. 19 Fewer than one-fifth of the Chinese respondents were willing to allow sympathizers of a deviant viewpoint to express their views in a meeting, as compared with two-fifths to three-quarters of the populations in the ISSP survey. Much of the Chinese "intolerance edge" was accounted for by differing educational levels in the comparison countries. Among respondents with some college education, over 50 percent of Chinese respondents would tolerate deviant speech at a meeting, compared with anywhere from 59 percent to 84 percent of college-educated respondents in the other countries. In short, Chinese were less tolerant than residents of six other countries, but the differences among Chinese and others were less substantial among people of similar educational levels.
In each of these three examples, the fit between hermeneutic generalization and positivist test is less than complete. The social networks measured in the Tianjin survey are a pale reflection of the rich concept of guanxi presented by Fei Xiaotong or Ambrose King. The idea of influencing government policy hardly describes the wealth of purposes for which factions are supposed to operate, and intolerance for dissident speech only scratches the surface of what Pye means by the demand for consensus. This is partly because the survey literature is still underdeveloped. For example, it might be better to test Pye's idea about guanxi -based factions with mainland data, but for now only Hong Kong data are available. The question about sympathizers of the Gang of Four may not be the best way to measure intolerance, so different measures need to be tried in future surveys.
Even when these inconveniences are overcome, however, positivist tests of hermeneutic insights will continue to display a certain meagerness compared to the original insights. In order to have validity (the ability of an operational measure to tap the attribute it is supposed to measure) and reliability (the measure's ability to measure that attribute accurately), a questionnaire item must be limited to a single carefully defined variable, not a complex of attributes. The insight being tested loses in complexity what it gains in specificity. Of course, additional questions can be asked to gather data on more variables making up the complex, but often the original insight is either too elaborate or too vague to be adequately represented by any practical number of measurable variables.
Moreover, in survey research across cultures, care must be taken to provide an equivalent stimulus in each setting. 20 A valid comparative survey must ask questions that can be meaningfully asked in all the settings the research is taking place. To ask Americans about guanxi , a term they do not understand, would be as meaningless as asking Chinese about their participation in school board elections. Whatever is truly unique, such as a linguistically distinctive way of referring to a cultural attribute or the name of a particular institution, has to be squeezed out of a question before it can be asked in more than one society. Most of the hypotheses stated in the hermeneutic literature are stated in forms that are inherently untestable across cultures, and have to be abstracted to be tested.
Here, from another angle of approach, we reencounter an earlier conclusion: A culture's uniqueness or nonuniqueness is not a characteristic of the culture itself, but of the way its attributes are conceptualized. As an attribute is abstracted to be measured cross-culturally, it loses the uniqueness it possessed when it was located lower on the ladder of abstraction, where it was described in a more complex, specific form. It now becomes by definition an attribute that all cultures possess, differing only in degree.
This does not mean that the positivistic approach is incapable of finding uniqueness among cultures. But the uniqueness it can discover consists not in the possession or nonpossession of a particular cultural attribute or cultural complex, but in the pattern of distribution of an attribute among the population, or in its pattern of association with other (cultural or noncultural) attributes. For example, in China the expectation of fair treatment from government authorities is strongly affected by respondents' educational levels. In this respect China is similar to five other countries where the same question was asked. Where China is unique is in the shape of the curve: only in China do those with college educations have a lower level of such expectation than those with no formal education. 21
Yet after all the caveats have been entered about positivist findings, they still remain robust enough to raise doubts about the distinctiveness of Chinese culture. In the examples we have been using--guanxi , factions, and intolerance--the question the positivist findings raise is not whether these phenomena are important in China, but whether they are either more important in China than elsewhere, or important in different ways. The positivist literature, limited though it is, leaves doubt about whether there is an empirical basis for any strong claim of Chinese distinctiveness in these three areas, other than to say that the Chinese on the average have somewhat larger guanxi networks or are somewhat more intolerant than members of certain other cultures at certain points in time. The hermeneutic literature makes bolder claims, but usually offers little clarity about what it is claiming and little empirical evidence to back up its claims.
Survey research is not the only positivist approach, but other works within the same tradition lead to similar doubts about Chinese culture's distinctiveness. For example, Richard Wilson's Compliance Ideologies makes positivistic use of comparative history in the sense that Wilson uses historical data to test clearly specified propositions. 22 He starts with an innovative definition of political culture as a dominant ideology justifying compliance with a society's institutional system (p. 19). He conceives of two general types of compliance ideologies, which he calls contractual and positional, the first stressing compliance based on individual rights and obligations, the second stressing compliance based on one's place in society (p. 89). He abstracts the sets of ideas that would characterize each of these two types of political culture at each of four historic stages in the life of a society (p. 91). The resulting eight categories offer a grid within which societies can be placed and compared. In this conceptual scheme no culture is unique: each box in the grid potentially contains more than one case. However, the scheme does highlight important differences among societies. Wilson looks at monographic studies, especially on the United States and China, for evidence that each displays the characteristics suitable to its type. The methodological difference between a work like Wilson's, which I here classify as positivistic, and the work by Taylor, which I classified above as hermeneutic, is not sharp. Both use a typology and both look at evidence. Wilson's typology is more formally constructed and the evidence he uses depends less than Taylor's on the interpretation of states of mind. As I argue in the conclusion, ultimately the most successful works combine insightful interpretation and careful empiricism. The boundary between the two approaches is made to be crossed.
Although less formally conceptualized, Fei Xiaotong's book serves as an example of comparative anthropology used in a positivistic way. Originally published as a series of magazine essays in the 1940s, the book aims to describe the rural Chinese society of the day to Chinese urbanites who were out of touch with rural society. Much of it consists of an elementary description of rural life, some of it touching and true, some repetitive and simplistic. In light of recent research on the commercialization of late traditional Chinese rural society, Fei's picture of a timeless, changeless, tradition-bound rural society (p. 57) seems exaggerated; perhaps this reflects the fact that he conducted his fieldwork in the 1930s, partly in remote Guangxi and Yunnan, while during the war his students did their fieldwork in Yunnan. 23
Whether Fei's argument is right or wrong, however, it is positivistic in form. Citing Toennies and Durkheim, and influenced by Redfield, Fei argues from two universalistic ideal types (confusingly and inconsistently referred to, respectively, as "rural" or "Chinese" and "modern" or "Western") (pp. 41, 61ff.). The former type is conditioned by poverty, physical immobility, a simple division of labor, and clustered dwellings; the latter type by the opposite characteristics. Each set of social conditions generates corresponding cultural features, such as different types of personal relationships, different senses of time, and different modes of exchange. In other words, like Wilson, Fei presents a typology of societies and shows where China fits. The attributes of Chinese culture, while unique in their particulars, are characteristic of societies of the relevant type.
Uniqueness, Difference, and Hypothesis-Testing
One of the most important and common uses of cultural comparison is to use culture as an explanation for certain societal outcomes, among them economic development versus stagnation and democracy versus authoritarianism. Pye, for example, says in another work, "Briefly put, my thesis is that political power is extraordinarily sensitive to cultural nuances, and that, therefore, cultural variations are decisive in determining the course of political development." 24
Logic says that a causative proposition cannot be proved with a by-definition finding of uniqueness. To test a hypothesis about the effect of culture on a social outcome, it is necessary to define cultural attributes in a way that is cross-culturally valid in principle. 25 Weber would have argued fallaciously if he had tried to prove that the notion of Original Sin was essential to capitalism because capitalism developed only where this idea existed. His theory of the Protestant Ethic was convincing because he abstracted from the Protestant mentality the idea of an acquisitive rationality that he said was crucial to capitalism, and which might have existed elsewhere but apparently did not. Only because Weber built his argument at this level of abstraction, was it possible for Thomas Metzger later to argue that neo-Confucianism contained a concretely different but abstractly similar kind of transformative tension. 26 If we find that a cultural attribute that is supposed to affect a particular outcome is absent in one place and uniquely present in the other, the logic of the investigation is flawed if this finding is an artifact of conceptualization.
Comparative propositions in the hermeneutic literature are often stated at the wrong level of abstraction for hypothesis-testing. For example, Pye proposes that in China, "more than in most countries, politics revolves around clashes of ideas and sentiments that have to be played out in the context of exaggerated notions of authority, on the one hand, and straitjacket controls on dissent, on the other" (p. ix). How could this hypothesis be tested? In a later book, Pye argues that "cultural factors dominate public life in China more than in just about any other country." 27 What might this mean in operational terms?
In statements like these (which are not limited to Pye's work), an author takes an interpretive insight into a unique complex of attributes in Chinese culture and restates it in the guise of a comparative statement about the different magnitude, extent, or influence of a cultural attribute in two or more different cultures. Such a transposition turns what may have been a valuable insight into the web of meanings within a culture into an untestable proposition about two or more cultures. This is what I meant earlier by saying that it is fallacious to restate a hermeneutic finding of uniqueness as if it were a positivistic finding of difference. When the goal is to frame comparative propositions that are meaningful for testing the kinds of explanatory hypotheses that the literature contains, we cannot transpose findings of uniqueness from the hermeneutic approach into findings of difference stated in positivistic language.
Equally, it is fallacious to restate a finding of difference as a finding of uniqueness. This is the transposition conducted by Gary Hamilton and Wang Zheng in their translators' introduction to Fei Xiaotong's book. As noted earlier, Fei places rural China into a category of societies in which relationships are predominantly parochial and personalistic. He labels this cultural complex chaxugeju , or network structure, as distinct from the organization-based structure that he believes to be typical of modern Western society. In effect, Fei is referring to the familiar concept of guanxi .
While Fei states that Western and Chinese society both contain mixed patterns, and that the difference is that in Chinese society the network pattern is "predominant" or "more important" (pp. 80-81, although he offers no guidance as to what this means), Hamilton and Wang turn this comparative statement into a statement about China's uniqueness, proposing that "Chinese social structure is unlike Western social structure." And although noting that Fei himself does not claim the patterns he describes are unique to China, Hamilton and Wang argue further that "there is sufficient evidence to indicate that the network patterns described by Fei are distinctively Chinese," even in comparison to other Asian societies. In the large literature on guanxi , Hamilton and Wang may be the first to make such far-reaching claims for its importance: they see it as the root of differences between China and the West in the concept of the self, sex roles, and family functioning (pp. 25-33).
Such statements might be a case of overlooking nuance in order to stress a main point, yet they promote a major misunderstanding. They place us in the presence of a homogeneous, classless, stateless China, which is contrasted with an equally caricatured "West" that stretches, presumably, from Scandinavia to Sicily and relies entirely on organizations without networks. Hamilton and Wang have reified Fei's already over-simple typology of societies, essentialized his concept of culture, and restated his insight into what makes China different from some other societies as a claim that China is unique.
The Fallacy of Conceptual Relativism
Hamilton and Wang do not merely offer the factual (if false) proposition that "Chinese social structure is unlike Western social structure" (p. 32). They also make a meta-theoretical claim: "There are . . . no universal social patterns and no universally valid principles by which all societies are held together" (p. 16), and therefore "Western concepts did not work well" in studying China (p. 12). "What passes in the West for general social theory," they conclude, "is often, in fact, local knowledge" (p. 34). If this claim is acceptable, the problems I have addressed in this essay disappear. Cultural comparison will consist of using different concepts to describe different cultures. We will see cultures as different not in the degree to which conceptually abstracted attributes are present, nor even in the possession of unique complexes of attributes, but in the very concepts needed to grasp them.
In making this claim, Hamilton and Wang part company with their translatee. As I pointed out earlier, Fei uses universal categories to explain to his Chinese readers what is distinctive about the rural Chinese society of their day. In a Chinese book for Chinese readers, he of course uses Chinese terms. But this is a choice of language, not a methodological position. His method is functionalist like that of his teacher Malinowski, and hence dependent on universal concepts. Fei's intellectual biographer David Arkush states that one of the main achievements of this book was to introduce "to a wide Chinese public the basic categories and ideas of Western social science." 28
Hamilton and Wang, however, use Fei to revive the old debate about whether there can be culturally universalistic concepts in social science. The use of culturally "local" concepts as tools for analysis may have its attractions as a strategy for single-country hermeneutic insight (although I think they are limited), but it is self-defeating as a strategy for cross-cultural comparison because it deprives us of concepts that can mark off for comparison comparable entities in different societies. 29 The insistence on the ineffability or incommensurability of cultures would solve the question of distinctiveness by fiat while preventing actual comparative investigation of cultures.
To say this is not to claim that there is such a thing as a floating vantage point above culture. 30 The vantage point that comparison requires is no doubt culturally rooted, and members of different cultures may adopt different vantage points. Chinese social scientists, for example, may choose to compare American and Chinese society using concepts derived from Chinese tradition or from Marxism. But in any case, they will need to use a single set of concepts to grasp comparable entities within the two societies.
Hamilton and Wang themselves use, as they must, the universal concepts of "class," "state," "network," "family," "the self," and so on, to express their views as to what China has and the West does not and vice versa. 31 Indeed, "Fei's sociology" could hardly "demand that we in the West rethink ourselves" (p. 34) if all social knowledge is only local knowledge. Their emicism thus contains its own negation. It leads to the statements that "in China the state does not exist as an organization" (p. 29) and that "China should be considered not a class-based but a network-based society" (p. 33). As empirical claims, such statements are false; as conceptual claims, fallacious. A culturally particularist social science will never be able to go beyond the limited and in many senses trivial finding that all cultures are at some level by definition unique. From this position, it cannot engage in explanations based on culture.
Conclusion
While we have learned a great deal about Chinese culture, successful comparative work remains rare in both the hermeneutic and positivist traditions. One reason is practical: good comparison requires extended conceptual and empirical study of both entities being compared, which is time-consuming and difficult. The second reason is methodological: the failure to gain the benefits of both approaches by learning how to combine them in valid ways in comparative studies.
Hermeneutic insights that are useful for comparative studies are those that anchor their propositions in specific persons, times, places, and groups, that deal conscientiously with evidence that might disprove their arguments, and that avoid true-by-definition claims of distinctiveness by paying attention to the ladder of abstraction. Good positivist work makes use of hermeneutic insights both to develop hypotheses, concepts, and measures and to interpret findings. Interpretation is an essential phase of the positivist method: no findings are self-evidently meaningful, as shown in the examples from survey research discussed above.
Is Chinese culture distinctive? Although anyone who studies it must be convinced that it is, we have far to go to state clearly how it is distinctive and to prove it empirically.
Note *: This article is a contribution to the Forum on Universalism and Relativism in Asian Studies initiated in Journal of Asian Studies 50.1 (February 1991): 29-83. For comments, the author wishes to thank Roger Ames, Lili Armstrong, Myron L. Cohen, Michael Gasster, Charles W. Hayford, Daniel W.Y. Kwok, Margot Landman, Jeremy Paltiel, Randall P. Peerenboom, Andrew G. Walder, John R. Watt, and other participants in a seminar at the Univesity of Hawaii at Manoa on January 20, 1993, and in the Modern China Seminar, Columbia University, on May 13, 1993, as well as two anonymous reviewers for the JAS . Back.
Note 1: am concerned with culture as a pattern of values, attitudes, beliefs, and affects, not as a pattern of behavior. The anthropologists' view of culture as including both values and behavior has its uses. But when one wants to use culture as an explanation for behavior, one must define culture as a pattern of mental attitudes separate from the pattern of behaviors that such attitudes are thought to explain. Back.
Note 2: "Clio's New Cultural Turn and the Rediscovery of Tradition in Asia," keynote address by Ying-Shih Yü at the Twelfth Conference of the International Association of Historians of Asia, University of Hong Kong, June 24-28, 1991, published as a pamphlet by that group. The distinction does not entirely correspond to that between humanistic and social scientific approaches, because many practicing social scientists use hermeneutic methods and there is occasional use of positivistic methods in humanistic research. Nor do the hermeneutic and positivistic approaches exhaust the list of approaches available in social sciences and humanities; Rabinow and Sullivan, for example, also refer to structuralist and neo-Marxist positions: Paul Rabinow and William M. Sullivan, "The Interpretive Turn: Emergence of an Approach," in Rabinow and Sullivan, eds., Interpretive Social Science: A Reader (Berkeley, Calif.: University of California Press, 1979), p. 1. The present discussion, however, has neither the need nor the space to complicate the problem. Back.
Note 3: Geertz, The Interpretation of Cultures (New York: Basic Books, 1973), p. 89; Benjamin I. Schwartz, The World of Thought in Ancient China , Cambridge: Harvard University Press, 1985, pp. 3, 413. Schwartz rightly points to important differences between his approach and Geertz's, but these are not relevant here. Back.
Note 4: Wang Gungwu, The Chineseness of China: Selected Essays (Hong Kong: Oxford University Press, 1991). Back.
Note 5: Tu Wei-ming, ed, "The Living Tree: The Changing Meaning of Being Chinese Today," special issue of Daedalus , Spring 1991; published under the same title by Stanford University Press, 1994. Back.
Note 6: Benjamin Schwartz, China's Cultural Values , Occasional Paper No. 18, Center for Asian Studies, Arizona State University (Tempe, Arizona: Center for Asian Studies, 1985). Schwartz developed some of the themes of his 1982 lectures in The World of Thought in Ancient China . I have chosen to review the shorter and earlier of the two books because it makes more explicit comparative statements, and because it has had influence among social scientists seeking concise statements about what makes Chinese culture distinctive. I do not criticize what Schwartz says, but attempt to point out how he can be misread. Back.
Note 7: Lucian W. Pye, The Spirit of Chinese Politics , New Edition (Cambridge: Harvard University Press, 1992). Back.
Note 8: The Mandarin and the Cadre: China's Political Cultures (Ann Arbor: Center for Chinese Studies, University of Michigan, 1988), p. 8. Back.
Note 9: Giovanni Sartori, ed. Social Science Concepts: A Systematic Analysis (Beverly Hills, CA: Sage Publications, 1984), pp. 44-46; Sartori, "Concept Misformation in Comparative Politics," American Political Science Review 64:4 (December 1970), pp. 1033-53. Back.
Note 10: "Among the wise," states Pye in The Mandarin and the Cadre, "it is unnecessary, indeed somewhat insulting, to clutter up analysis with the obvious qualification that such collectivities are not homogeneous entities" (p. 28). Back.
Note 11: See Arthur L. Stinchecombe, Constructing Social Theories (New York: Harcourt, Brace, 1968), pp. 101-129. Thanks to Andrew Walder for the citation. Back.
Note 12: Rodney L. Taylor, The Religious Dimensions of Confucianism (Albany: State University of New York Press, 1990), ch. 3. Back.
Note 13: Fei Xiaotong, From the Soil: The Foundations of Chinese Society, A Translation of Fei Xiaotong's Xiangtu Zhongguo ,, trans. Gary G. Hamilton and Wang Zheng (Berkeley: University of California Press, 1992). Back.
Note 14: The comparison is limited to urban populations of the two societies. Ruan Danqing [elsewhere Danching], Lu Zhou, Peter M. Blau, and Andrew G. Walder, "A Preliminary Analysis of the Social Network of Residents in Tianjin with a Comparison to Social Networks in America," Social Sciences in China 11.3 (September 1990), pp. 68-89. Back.
Note 15: This point is developed in Ruan Danching, "Interpersonal Networks and Workplace Controls in Urban China," Australian Journal of Chinese Studies 29 (January 1993), pp. 89-105. Back.
Note 16: One finding of the research project described here is that "The similarity between the microstructures of interpersonal relations in the P.R.C. and the U.S. is impressive, considering the differences in culture and tradition." From Peter M. Blau, Danching Ruan, and Monika Ardelt, "Interpersonal Choice and Networks in China," Social Forces 69:4 (June 1991), p. 1049. Back.
Note 17: Lau Siu-kai and Kuan Hsin-chi, The Ethos of the Hong Kong Chinese (Hong Kong: The Chinese University Press, 1988), p. 101; Gabriel A. Almond and Sidney Verba, The Civic Culture , abridged ed. (Boston: Little, Brown, 1965), p. 148. Although Pye may not consider Hong Kong very Chinese culturally, Lau and Kuan make extensive use of Pye's arguments in interpreting their findings. In any case, the point here is to illustrate the types of findings that survey-based comparisons produce, and only secondarily to suggest a test for Pye's hypothesis. Back.
Note 18: Andrew J. Nathan and Tianjian Shi, "Cultural Requisites for Democracy in China: Findings From a Survey," Daedalus 122:2 (Spring 1993), pp.. 95-123; reprinted here as chapter 11. The present essay elaborates points originally offered in the last few pages of the Daedalus article. Back.
Note 19: The ISSP is a continuing program of cross-national collaboration conducted in Australia, Germany, the United States, Great Britain, Austria, and Italy. See International Social Survey Programme, Role of Government--1985 Codebook ZA-NO . 1490, Ann Arbor: ICPSR, University of Michigan. The stimulus posed in the 1985 ISSP survey was, "There are some people whose views are considered extreme by the majority. Consider people who want to overthrow the government by revolution." It was followed by the same three questions. Back.
Note 20: See Frederick W. Frey, "Cross-Cultural Survey Research in Political Science," in Robert T. Holt and John E. Turner, The Methodology of Comparative Research (New York: Free Press, 1970), pp. 173-294; Sidney Verba, Norman H. Nie and Jae-on Kim, Participation and Political Equality: A Seven Nation Comparison (Cambridge: Cambridge University Press, 1978), pp. 32-40. Back.
Note 21: Nathan and Shi in Daedalus , pp. 108-110. Back.
Note 22: Richard W. Wilson, Compliance Ideologies: Rethinking Political Culture (Cambridge: Cambridge University Press, 1992). Back.
Note 23: R. David Arkush, Fei Xiaotong interpretingand Sociology in Revolutionary China (Cambridge, MA: Council on East Asian Studies, 1981), pp. 57-103. After 1949, the people living in the Fei team's Yunnan field site were classified as non-Han minority Bais; see David Yen-ho Wu in Tu, p. 169. This gives an ironic twist to the effort to use Fei's insights to prove the distinctiveness of Chinese culture. My thanks to Charles W. Hayford for suggesting this point. Back.
Note 24: Lucian W. Pye, with Mary W. Pye, Asian Power and Politics: The Cultural Dimensions of Authority (Cambridge: Harvard University Press, 1985), p. vii. Back.
Note 25: There is debate about how strong an explanation culture gives of any societal outcome. Pye's analogies of culture to music (in Asian Power and Politics , p. 20) or to grammar (in Mandarin and Cadre , p. 9) seem to me about right. And how much difference has a country's musical tradition or grammar made to its modernization or democratization? The alternative to a cultural explanation is usually an institutional or structural one. See, for example, David J. Elkins and Richard E. B. Simeon, "A Cause in Search of Its Effect, or What Does Political Culture Explain?" Comparative Politic s 11 (January 1979), pp. 127-45; Bruce J. Dickson, "What Explains Chinese Political Behavior? The Debate Over Structure and Culture," Comparative Politics 25.1 (October 1992), pp. 103-18. Pye, however, considers this debate "pointless"; Mandarin , pp. 20-22. Back.
Note 26: Metzger, Escape From Predicament (New York: Columbia University Press, 1977). Back.
Note 27: Mandarin and Cadre , p. 30. Back.
Note 28: Arkush, Fei , p. 144. Back.
Note 29: Although the difference in the prevalence and functions of networks versus organizations in China and the West remains largely an empirical mystery, the two societies are demonstrably different in the self-image that each has about its respective reliance on networks and organizations. For this statement, the Fei and Hamilton essays themselves serve as evidence, because each embodies its own society's self-perception of the difference. Although the difference in self-perception does not prove that there is a difference in social functioning, it reasonably generates the hypothesis that there is. But one cannot test this hypothesis by using separate conceptual categories for separate countries. Back.
Note 30: Schwartz persuasively rebuts such a claim in The World of Thought , pp. 3-7. Back.
Note 31: Ambrose King seems to place himself in the same contradiction when he argues both that guanxi is a uniquely Chinese phenomenon and that for this very reason the word should be incorporated into modern social science as an analytic term. King in Tu, ed., p. 68. Back.
Email this citation