II. SHARED UNDERSTANDING OR MISUNDERSTANDING

§1. The Observable Moral Data

• A. The Complications of Shared Understanding
  • 1. The Equivocality of the Terminology
  • 2. The Semantics of Shared Understanding
• B. Social Fact and Social Meaning
  • 1. Émile Durkheim's Sociology of Morals
  • 2. Social Fact
  • 3. Between Fact and Meaning
    • a. Source
    • b. Relativity
    • c. Substance
    • d. Objectivity
• C. Cultural Tradition
  • 1. The Concept of Culture
  • 2. Twitch and Wink
  • 3. Sheep Raid
• D. The Interpretation of Shared Understandings

§2. The Objectivity of Social Meaning

• A. Simple Objectivity
• B. Complex Objectivity
  • 1. The Objectivity of Judgement
  • 2. The Objectivity of Social Construction

§1. The observable moral data

Walzer’s practical particularism is grounded on what he calls “the opinions of mankind” in Just and Unjust Wars.^{1Wars, p. 15.} He later elaborates this idea into “shared understandings” in Spheres of Justice. This is a decisive turn from the early project of modernity which endeavours to found morality on a universal rationality. His particularist methodology challenges the early modern aspiration to building a universal platform for humanity. Some scholars interpret Walzer’s move as a kind of communitarian critique of liberalism. This classification is correct so long as communitarianism is loosely understood as a concern for culture and community. However, a communitarian perspective is inadequate to interpret Walzer’s writings.^{2Walzer draws a line between the communitarians and himself in The Communitarian Critique of Liberalism, in Political Theory 18 (1990) 6-23.} His enterprise, I believe, is more ambitious: he actually tries to bridge the gap between scientific knowledge and traditional knowledge. Modernity created this cleavage, which was then institutionalized in the academy as the division between the traditional discipline of theology or philosophy and the modern discipline of social science. The old and the new disciplines, each relies on a different methodology, and each claims to have its own authority, have ever since gone their own way. Walzer’s interpretation of shared understanding, I will argue, offers a way to rejoin them.

A. The complications of shared understanding

The term “shared understanding” appears many times in Spheres of Justice, but nowhere can we find any definition of it. Perhaps, Walzer assumes that his reader would have no difficulty in grasping it. At first sight, shared understandings refer to some languages, knowledge, practices, and social institutions that are shared among the members of a particular society. As a kind of vague everyday language, the use of the term “shared understanding” is acceptable. Few people would doubt that a community speaks at least a common language, shares some common knowledge, and upholds certain common values. But when a vague idea is taken to represent a philosophical concept, it invites severe criticism. Indeed many critics find the idea of shared understanding fatal to the otherwise well-argued book.

One straightforward attack is to deny that there is any understanding shared at all. Ronald Dworkin, who considers shared understandings as social conventions and observes that there are debates on virtually every political matter in America, once said, “… the very fact that we debate about what justice requires, in particular cases, shows that we have no conventions of the necessary sort.”^{3R. Dworkin, To Each His Own, p. 4, c. 4.} Here Dworkin, and some other critics as well, interprets “shared” as “conformation.”^{4Conformation is how Joshua Cohen understands the word “shared.” He says, “A value is only said to be shared if its members conform to its requirement (J. Cohen, Review of M. Walzer, Spheres, p. 462).”} Since the citizens of the United States always disagree, there is no conformation, and thus no shared understanding. Even if there are shared conventions, the Walzerian ethic will still be nothing but a literary presentation of moral statistics. Such a moral anthropology is too parochial to be critical at all.^{5Cf. N. Daniels, Review of M. Walzer, Spheres.} These are serious accusations that demand unequivocal answers.

1. The equivocality of the terminology

The first appearance of “shared understandings” is found in the Preface of the Spheres.^{6Spheres, p. xiv.} In its context, the term suggests that our ideal of a free and just society is not a utopia located nowhere, but a practical possibility already embedded in our commonly held meanings of goods. It emphasizes the localness of justice in contrast to the universal conception of justice. Except for this assertion, there is no further explanation. Understandably, the preface only introduces the key concepts, and we would expect Walzer to elaborate on them in the body of the work. Some critics even express the opinion that the notion of shared understanding is so new and fundamental to Walzer’s theory of justice that nothing is adequate except a comprehensive theory of community. Contrary to their expectation, Walzer gives no definition in the Spheres, let alone a theory. It seems that Walzer treats the idea of shared understanding as a self-evident fact that needs no specific definition.

To further complicate the issue, Walzer uses, rather casually, an array of similar expressions, all of them ill-defined. To give an idea of what they are, I group them according to the substantives (page references are given in parentheses):

1. Understanding-expression

2. Meaning-expression

3. Conception-expression

4. Value-expression

A preliminary study of the expressions shows that Walzer has used at least four substantives: understanding, meaning, conception, and value. To qualify them, he uses four adjectives as well: shared, social, common, and cultural. The most frequently used substantive and adjective are “meaning” and “shared” respectively. The expressions on the whole convey the impression that the meaning of a good or a thing is shared among the members of a community. Superficially, these expressions seem to carry similar connotations. But, the problem is, what do they exactly refer to? Do they refer to the same reality? These are complex questions that Walzer has not clarified in the Spheres. In order to answer them, I will extract the meanings of the various expressions from the immediate contexts in which they stand.

2. The semantics of shared understanding

After a survey of the various expressions, we can draw out five inferences:

1. In most cases, “shared” and “social” are interchangeable, likewise for the terms “understanding,” “meaning” and “conception.”

No effort is made to distinguish “shared” from “social,” or to differentiate “understanding” from “meaning” or “conception.” In his discussion of social goods, Walzer writes, “Goods in the world have shared meanings because conception and creation are social processes.”^{7Spheres, p. 7.} From this statement, we can deduce that meanings are shared because their formation is a social process. When “social” is used to qualify “meanings,” it carries the double connotations of “shared” and “the reference to the collective effort.” As for the three nouns, no significant distinction can be made. Very often, the expressions are in the form of “shared understandings of goods,” “social meanings of goods,” or “shared conceptions of goods.” Sometimes, “social meanings” may stand on its own, but the context indicates that it refers to “social meanings of goods.” Yet, “shared understandings of goods” and “shared understandings,” in some instances, may have different meanings. “Shared understandings” may denote the totality of the moral world as the expressions “the world of meanings that we share” and “shared understandings of the members” indicate, whereas “shared understandings of goods” can be interpreted as a subset of “shared understandings.”

2. Shared understandings of goods, with respect to distributive justice, concern every aspect of society.

According to Walzer, society is primarily a community of distribution. He says, “Distributive justice is large idea. It draws the entire world of goods within the reach of philosophical reflection. Nothing can be omitted; no feature of our common life can escape scrutiny. Human society is a distributive community.”^{8Spheres, p. 3.} People come together, organize themselves into a particular community and create things—such as membership, power, honour, commodities, and feasts—for sharing, dividing, and exchanging. Different political arrangements enforce and different ideational systems justify different patterns of distributions. Therefore, all the shared understandings of a particular social world are relevant, especially the political arrangements and the moral ideals. Here, we find some ambiguity of Walzer’s use of “shared understandings” and “shared understandings of goods”. Logically, shared understandings contain all the social meanings of which shared understandings of goods is a subset. Though distributions are not the whole of communal life, they may be related to every aspect of it. In other words, the set of the shared understandings of goods is open. If we find any shared understanding relevant to distributive justice, it belongs to this set. Consequently, within the domain of distributive justice, shared understandings are the same as shared understandings of goods. It implies that any idea, though it may not relate directly to the meanings of goods, can be invoked as an argument to justify certain distributive principle.

3. Shared understandings are particular.

Different communities have by and large different shared understandings despite the fact that there may be similar or even overlapping shared meanings of goods in some communities. A question immediately arises: which shared understanding should concern us in distributive justice? Fortunately the question itself has hinted at the answer: “us” points to the solution. For all the peoples within “us,” their shared understandings are relevant. Walzer thinks that the best boundary of “us” is the political community.^{9Spheres, pp. 28-30.} Although the political community is not a self-contained distributive world, it comes closest to a world of common meanings: common language, common history, common culture. Moreover, politics creates its own commonality. Distributions are to a large extent authorized and regulated by political power. Hence a political community constitutes the basic setting of distributive justice.

4. Shared understandings of goods are examples of distributions.

Walzer’s shared understandings of goods often refer to historical cases. In the explanation of his methodology, he writes, “My examples are rough sketches, sometimes focused on the agents of distribution, sometimes on the procedures, sometimes on the criteria, sometimes on the use and the meaning of the things we share, divide, and exchange.”^{10Spheres, p. xiv.} These examples are contemporary and historical accounts of distributions. They are not confined to a particular community. Distributive stories of other political communities are cited in contrast to the American ones. These exotic examples are used to “suggest the force of the things themselves.”^{11Spheres, p. xiv.} But how much force does an example from a foreign society carry? How can we scale the weights of the shared understandings of different societies? These are difficult questions. If we read Walzer straightforwardly, only the contemporary shared understandings of goods of one’s country are obligatory; all the others are illustrations. Walzer, however, does not strictly observe this rule. He uses both historical and exotic examples to vindicate his case. If the contemporary opinion of a certain distribution does not support the ideal of complex equality, Walzer may present a historical or exotic story to counter the contemporary one. That is why some of the critics become confused. For instance, arguing that medical care is a welfare and thus should be distributed to every citizen, Walzer cites the historical accounts of ancient Athens, the medieval Jewish community, and the European principles of “cure of the souls” and “cure of the bodies” to oppose the dominant American practice of free exchange between doctor and patient.^{12Cf. Spheres, pp. 64-91.} Is Walzer inconsistent? Is he not conflating different sets of shared understandings of goods? These questions lead us to the “deeper understandings of social goods.”

5. The deeper or real understandings of social goods are the hidden or interpreted shared understandings of goods.

These two questions, in Walzer’s opinion, are crucial in determining the distributive arrangements of a society. The first question asks us to look back on our history, and the second directs us to reflect on the shared understandings. Once we know the answers, we should understand how to distribute justly here and now. It is important to note that the word “really” put within parentheses is the crux of Walzer’s interpretive method. Are there understandings which are not really shared? How can we know what understandings are really shared? Walzer does not elaborate this point in the immediate context. Another sentence in a later passage sheds more light on it:^{14Spheres, p. 26.}

This statement gives Walzer leeway to retreat from the tyranny of the majority, the dominant or popular opinions, the established practices and institutions, or perhaps even the restrictions of temporality and locality. Because of dominance and monopoly, existing distributions may have been the results of oppression, and the popular ideals may have been distorted to favour the ruling class. Walzer’s basic assumption is that we have to be faithful to the shared understandings if and as long as these understandings reflect the true and free expressions of mankind. There is no need indeed to respect a popular but oppressive idea imposed by the powerful. Since oppression is always present in human history, interpretation inevitably involves finding the deeper or true shared understandings of goods.

The semantic analysis above is only one of the two possible means to investigate the meaning of shared understandings. The other involves an inquiry of what foundation Walzer has chosen for the development of his theory. Weighing on the material we have discussed above, it is true that Walzer has not clearly defined or explained his fundamental concept of shared understanding, thus inviting severe criticism. Nevertheless, we can still grasp the general idea of shared understandings as the social meanings of things embedded in history. Such understandings often reside in historical cases which require interpretation if we are to uncover their moral significance. They are similar to the precedents in the legal system. Moral ideals, like laws, are devoid of pre-existent definitions, and we have to look for their meanings in historical precedents. The procedure, Walzer compares, is like a judge or a lawyer making legal decision from within the legal tradition.^{15Cf. M. Walzer & R. Dworkin, 'Spheres of Justice': An Exchange, in The New York Review of Books, 21 July 1983, pp. 43-46, pp. 43-44; Interpretation, pp. 18ff.} If the legal system can work properly without serious theoretical objection, why can’t a moral system function on the same basis? Indeed, Walzer assumes that it can. Thus it is unnecessary to define shared understanding in detail. Furthermore, the methodology used in the Spheres is not entirely novel. Walzer has employed more or less the same technique in the Wars.^{16Walzer himself declares that he uses the same methodology in the two books. In an interview, he says, “I am still mixing arguments from principle and arguments from historical cases in the way that I did in those books [Wars and Spheres].” See M. Walzer, M. Carleheden, R. Gabriëls, An Interview with Michael Walzer, in Theory, Culture and Society 14 (1997) 113-130, p. 113.} Since nobody has challenged him there, he would not expect it to be rejected here. Most probably, the negative feedback is caused not so much by the idea of shared understanding as by its assertive particularist expression.

Regardless of the intention of the critics, their queries cannot simply be brushed aside. What does shared understanding refer to? It seems that shared understanding is moral tradition. If so, why does Walzer invent a new name instead of using the well-known term “moral tradition”? Apparently, Walzer has two choices. Either he develops a theory of community as some critics suggest, or he connects the idea of shared understanding to some well-established concept like tradition, history, or culture. Walzer chooses the latter, but his analogy to law cannot satisfy his critics. Georgia Warnke comes to his aid and attempts to fill the gap by proposing that Walzer’s methodology is a kind of interpretation which can be understood by hermeneutics.^{17Cf. G. Warnke, Justice and Interpretation, Cambridge, MA, 1992, pp. 13-37.} She first establishes a parallel between literary interpretation and social interpretation, and then draws the conclusion that Walzer’s interpretation accords with the advanced hermeneutics, whereas his critics’ understanding of his interpretation is defective. Hermeneutics, of course, can be one possible way of explaining Walzer’s methodology. There is, I believe, an alternative more immediate to Walzer’s thinking, that is, to explain it in the light of anthropology. I am inclined to think that the anthropological approach is more appropriate since Walzer himself claims that his theory relies on “history and anthropology.”^{18Spheres, p. xviii.} In order to better understand the proposition of the “interpretation of shared understandings,” I shall give a rough sketch of the sociological/anthropological background by introducing Émile Durkheim’s notion of “social fact,” and Clifford Geertz’s “interpretation of cultures.”

B. Social fact and social meaning

1. Émile Durkheim’s sociology of morals

For a deeper appreciation of the concept of shared understanding, it will be illuminating to compare Walzer with Durkheim on this subject matter. A full investigation is obviously beyond the scope of my project. My aim has to be minimal: I attempt to show that Durkheim and Walzer are revolting against the ethical foundation founded by Kant, and that they are both on a journey towards a mature modern moral theory. There are certainly signs of continuity between them, but I do not plan to establish the link here. I intend only to draw out some salient features so as to cast the notion of shared understanding in a sociological background. Could shared understanding be the source of moral investigation from the sociological perspective? Could Walzer avoid the pitfalls stemming from Durkheim’s positivist standpoint, for which the latter has been severely criticized? These are the main questions that I try to address.

At first glance, Walzer and Durkheim do not have much in common. The two authors are almost separated by a century, and there is no direct link between them except that both are of Jewish origin. No apparent similarity can likewise be found in their works to affirm their correlation. But such superficial impression cannot withstand a deeper investigation. Walzer’s Spheres is an implicit critique of Rawls’s Theory of Justice. Rawls treats society as an assemblage of individuals, who are without intrinsic bondage either with each other or with their ancestors. Walzer judges this approach to be a kind of “bad sociology.” He criticizes Rawls for separating people from each other and from their history. Such an individualistic ethical theory cannot fully encompass our moral experience. Walzer has the ambition to work out an ethic which will be consonant with the findings of sociology. As a matter of fact, Rawls depends heavily on Kant. In his own account, Rawls says, “The theory that results is highly Kantian in nature.”^{19J. Rawls, A Theory, p. xviii.} Indeed, his notion of persons in the original position is an elaborate version of Kant’s categorical imperative. Walzer is not alone in the effort to refute the Kantian deontological ethic. Durkheim has done the same thing a century ago. In fact, Durkheim is one of the pioneers who tries to replace Kant’s individualistic approach to morality with another scientific theory, which Robert Hall calls the “sociology of morals.”^{20See Robert T. Hall’s Introduction, in E. Durkheim, Ethics and the Sociology of Morals, trans. R. T. Hall, Buffalo, NY, 1993, pp. 11-51. Hall has excellently demonstrated that Durkheim is a sociologist of morals, whose life-long endeavour is to establish a sociology of morals which, he thinks, forms the basis of philosophical ethics.}

In a way, Durkheim and Kant are related to Newton, not in the subjects they study, of course, but in the methodologies they use. Newton’s science, since the appearance of the Principia Mathematica, has become the model and the touchstone of modern sciences. He had been seen for some time as the lawgiver of the universe. His significance was captured by the English poet Alexander Pope in the proposed epitaph for our modern lawgiver:^{21A. Pope, quoted in I. Prigogine & I. Stengers, Order, p. 27.}

In the poem, Pope implicitly compares Newton to Moses. He implies that both Moses and Newton receive God’s revelation but in different domains. One gives us the law of human, and the other the law of nature.

Pope’s dichotomy represents a common understanding of Newton’s achievement. Actually Newton’s science has implications beyond the movement of physical bodies. In the ancient cosmology, the heavenly bodies are considered as vital to all forms of life on earth. These bodies transmit their energy to the earth, and exert a determinative influence on human beings. The heavenly bodies circulate the earth in a regular manner, and are in turn moved and controlled by some other angelic beings, as is believed. Newton’s theory proposes that the bodies are not moved by angels but are in fact acted upon by gravitational forces, and that their paths of movement can be accurately represented by mathematical formulae. Although Newton himself refuses to accept gravity as a material force, it can nonetheless be interpreted as thus, and from which we can further deduce a mechanistic world view. If the heavenly bodies, which were regarded as mystical in the past, can be controlled by a mechanistic law, could human beings be also governed by another mechanistic law?

Kant resists this temptation. Although he goes so far as to agree with the Newtonian view that human behaviour is circumscribed by a law, he refuses to accept that that law is mechanistic in nature. Kant endeavours to limit the application of the Newtonian science to the domain of the physical world. To him, human moral behaviour belongs to another territory and obeys another category of law.

Newton’s science may be summarized as consisting of three main characteristics: observability, comprehensibility, and controllability. It assumes that everything is observable through human senses, and that the collected data can be analysed by the intellect and be rendered into a set of comprehensive principles. As a result, the principles can in turn be used to manipulate the corresponding objects. These three elements constitute the foundation of any scientific methodology. Kant, however, argues that the moral behaviour of rational beings cannot be wholly circumscribed by this kind of methodology. According to some Western moral traditions, a moral act is first and foremost determined by the intention of the actor. No doubt an action is open to observation. But without knowing the intention of the actor, we cannot make a moral judgement. Take for example a policeman shot dead a robber in an exchange of fire. This action is observable and can be judged to be in total compliance with the requirements of law. It is lawful. But, is it moral? In most cases, it will be reasonable to assume that it is moral. Now, suppose the policeman had an overwhelming hostility toward robbers because his ex-partner was killed by a thug. To revenge for the tragic death of his colleague, he vowed to kill every robber insofar as the law permits. In this case, his intention is not peace-keeping but revenge. His act of killing a robber would become immoral even though it would still be lawful. Human intention is entirely internal and beyond the scrutiny of an outside observer. Thus, the rule of human action, unlike the movement of physical objects, cannot be accessed through any empirical study of human behaviour. In other words, human moral behaviour is determined by the will of individuals rather than by external forces. Following this line of thinking, the absolute good can only be a will that wills good. Kant translates this absolute good into the self-made law (autonomy) which says that an act is moral if and only if it can be generally accepted by every free-willed being.

In the view of Durkheim, the Kantian methodology is a regression to premodern thinking. While Kant accepts the comprehensibility and the controllability of human behaviour, he rejects the (complete) observability of moral acts. Kant’s argument is based on the assumptions that society consists of atomized individuals, and that each individual acts according to his or her free will. Durkheim observes that these assumptions do not constitute an accurate account of the real world. People act, for most of the time, not according to their free will, but to the norms already laid down by the society in which they live and interact with each other.

The overemphasis of individuality has a deeper root in Kant’s grounding of his epistemology on an a priori basis. Philosophers since Aristotle have discovered that some ideas, such as time, space, class, number, cause, and substance, are impossible to be reduced to other ideas. This set of fundamental ideas is called the categories of the understanding. Together they constitute the most basic framework which encompasses all thoughts. One could hardly think of anything without reference to time, space, or number. All other knowledge, including the law of physics, appears to be in a continuous flux, but the categories of the understanding remain constant. They appear to have a special relationship with human intelligence. The apriorists, among them Kant is one of the advocates, think that the categories of the understanding belong to the innate structure of human beings. All sensory data and experience are organized according to this structure. The empiricists disagree with the apriorists; they reject any a priori structure. They claim that the basic concepts are derived solely from experience. Consider an individual agent who repeatedly performs a certain action, which yields similar results and experience. After an indefinite number of times, the agent begins to be aware of the basic ideas involved in its experience. Finally, he articulates the ideas and uses them to perceive and apprehend other experiences. For instance, an ancient man threw a stone upward into the sky and saw it fell down onto the earth. Each time he repeated the same action, the stone fell back to the ground. During the process, he gained the experience that the stone sometimes fell faster and sometimes slower depending on the force he applied. The apriorists would say that the experience of fastness and slowness points to the innate idea of time. Whereas the empiricists counter by asserting that the experience of speed or time is but a relative experience dependent on the background of the moving stone and the position of the observer. His sense of speed is built up as a result of repeated experiences. For any person to understand the idea of time, he must repeat the same or similar action until the experience is firmly ingrained in his mind.

Durkheim opines that the apriorists and the empiricists have both missed the point.^{22Durkheim does not make a distinction between pure apriorists and the Kantian apriorists who claim that only the synthetic a priori statement is a priori. To simplify our discussion, I maintain the commonly accepted distinction here.} He follows the analysis of Kant and points out that the apriorists and the empiricists are speaking of “two sorts of knowledge, which are like the two opposite poles of the intelligence.”^{23E. Durkheim, The Elementary Forms of the Religious Life, trans. J. W. Swain, New York, NY, 1965, p. 27.} On one pole stand the categories of the understanding which are not attached to any particular object and thus are independent of every particular subject. They are universal and necessary for the understanding of other knowledge. On the other pole are the empirical data, which rely upon objects and the perception of subjects. They are essentially individualistic and subjective. The empiricists deny the first kind of knowledge and are ignorant of its formal and stable properties. These features are forces constraining the perception of an individual. If a person sets aside the categories of the understanding, he will meet with great difficulties and resistance, that is, he will lose sense of the society where he lives, and will be unable to communicate with other members. This framework does not merely depend upon us; it actually imposes itself on us. Kant rightly recognizes the two distinctions of knowledge. He realizes that the categories of the understanding are necessary to safeguard an objective standard which is independent of the individuals. He tries to explain these “pre-existed” ideas as the inborn structure of the human mind. For Durkheim, this explanation is a regression to the premodern mindset. He says, “For it is no explanation to say that it is inherent in the nature of the human intellect.” If we attribute something to instinct simply because we cannot explain it, this is the same as to say that we do not know where it comes from. A scientific mind should first exhaust all explanation derived from observable data before coming to speculative, metaphysical, or mythical explanation. Kantians resort too early to transcendental argument or inexplainable assumption. There is, in fact, a possibility of further scientific inquiry. The crucial question, Durkheim thinks, “is to know how it comes that experience is not sufficient unto itself, but presupposes certain conditions which are exterior and prior to it, and how it happens that these conditions are realized at the moment and in the manner that is desirable.”^{24E. Durkheim, The Elementary Forms, p. 27.} In this respect, Durkheim seeks his answer in society. He thinks that the categories of the understanding are nothing but collective representations. He says, “collective representations are the result of an immense co-operation, which stretches out not only into space but into time as well; to make them, a multitude of minds have associated, united and combined their ideas and sentiments; for them, long generations have accumulated their experience and their knowledge.” In other words, collective representations are the concentration of the intelligence of many generations. That is why they are “infinitely richer and complexer than that of an individual.”^{25E. Durkheim, The Elementary Forms, p. 29.} They are external and prior to individuals, and thus form the framework of the understanding for all the members of a society.

Evidence shows that Durkheim’s sociological explanation of the categories of the understanding is closer to the human learning process than that of Kant’s inherent structure of the mind. Anthropological research reveals that there exist different conceptions of time. For instance, Westerners see time as linear but the people in ancient China regard it as circular. The advance in the knowledge of navigation, physics, and mathematics has also changed our conception of space. The classical definition of a straight line is defined as the shortest distance between two points where the straightness is commonly understood as the edge of a standard ruler. As our power and experience in travelling expand, we find out that the idea of straightness is in fact a limited comprehension of reality. If we travel across the continents on the high seas or through the sky, this idea of straightness is completely inapplicable. Since the shape of the earth is a globe, a “straight-line” course for a ship may require it to dive into the sea and bore through the crust of the earth, and a “straight-line” path for an aeroplane will send it into the outer space. In order to accommodate to our expanded experience, we have to redefine straightness as an arc rather than the edge of a ruler. Yet we may still argue that while the conception of straightness expands, the a priori synthetic definition of a straight line holds, that is, a straight line is still the shortest distance between two points. At all events, the trueness of the assertion has something more to do with our determination to keep it as a higher abstraction of our experience than with the innate structure of our mind. Based on the axiom of a straight line, mathematicians have built a whole branch of a priori synthetic knowledge, called non-Euclidean geometry. But how many ordinary people understand abstract geometry? Few! Most people have not even heard of it. Even mathematicians themselves have had hard times in grasping such ideas and theorems. If these ideas are part of the innate structure of the mind, why are so many minds deprived of them? Children use a lot of time and energy to do drilling exercises before they could grasp abstract mathematical notions. This is a positive sign pointing to socialization rather than to instinct or innate structure.

What Durkheim suggests is that we should seek explanation of human behaviour in society itself. He thinks that society is an objective reality independent of individuals. Individual men and women gather together to form a society, but society as a complex whole transcends individuals. A society is an association that is far more complex than an aggregation of men and women. It is an entity in itself that opens to scientific investigation. In other words, it has objective data that are observable to an external observer. Durkheim calls these data “social facts.”

So far we have seen that both Durkheim and Walzer refute the Kantian individualistic approach to morals. Both plead that morals should be grounded on a social basis. While Durkheim articulates “social facts” to refer to the observable data of society, Walzer chooses the term “social meanings” which he frankly acknowledges to have it borrowed from anthropology.^{26M. Walzer, Objectivity and Social Meaning, in M. Nussbaum & A. Sen (eds.), The Quality of Life, Oxford, 1993, 165-177, p. 166.} But it is still unclear what kind of relationship “social meaning” stands with “social fact.” There is at least one attempt made by Jack Douglas to compare social fact with social meaning.^{27J. D. Douglas, The Social Meanings of Suicide, Princeton, NJ, 1967.} In his Social Meanings of Suicide, Douglas argues that it is impossible to speak about social facts of suicide consistently without touching upon social meanings. It is obvious to him that the study of suicidal actions as facts is unsatisfactory. Even Durkheim is unconsciously moving in the direction of treating suicide as meaningful act. If Douglas’s argument can be established, then Walzer’s use of social meaning may be a correction of Durkheim’s social facts. Let me now give a brief account of Durkheim’s social facts before proceeding to compare the concept with social meanings.

2. Social fact

According to Durkheim, a scientific study of human (social) behaviour can take place only if there is a distinct, objective subject matter. The thinking and sensual beings, for example, are the subject of psychology, and the human bodies that of biology. Sociology must also have its own source of observable data which are independent of individuals and non-reducible to either psychological or biological data. Durkheim strongly resists the reduction of social phenomena to psychological data, which he considers to be false. To him, it will be as false as to reduce psychological data to biological data, or to reduce biological data to chemical or physical data. In order to guard against reductionism, he insists that once a society comes into existence, it acquires a life of its own as evident and as real as an individual human being. The first task of a sociologist is to identify the body of data which express the moral and social reality, and which are susceptible to scientific manipulation. At the beginning of The Rules of Sociological Method, Durkheim affirms there exist such sociological data:^{28E. Durkheim, The Rules of Sociological Method, trans. S. A. Solovay & J. H. Mueller, New York, NY, 81964, p. 1.}

In his first major work The Division of Labour in Society (1893), Durkheim calls the social obligations and duties “moral facts.” Later in The Rules of Sociological Method (1895), he extends this preliminary notion to the more general idea of “social facts.” A social fact is a social responsibility that is encoded in law or in custom.

Contrary to the claim of Kant, Durkheim asserts that universality is irrelevant to social facts. Rather, the key to social facts lies in their collective property. He says, “A thought which we find in every individual consciousness, a movement repeated by all individuals, is not thereby a social fact.” Social facts are determined by the “collective aspects of the beliefs, tendencies, and practices of a group that characterize truly social phenomena.”^{29E. Durkheim, The Rules, pp. 6-7.} Apparently, Durkheim is trading universality for observable particular social data. His move is not a relapse to premodernity. Indeed, it can actually be justified within the reasoning of modernity.

The methodology of a study occupies a determinative role in modern epistemology. Not only is an acceptance or rejection of a scientific report based partly on the evaluation of its methodology, modernity actually enthrones scientific methodology as the foundation of knowledge. Strictly speaking, universality is not part of scientific methodology; it may be the end-product of a scientific research, or worse still, it may only be an aspiration of modernity. In sociological field, observable social facts are particular, that is, each society has its particular set of social data. An honest sociologist has to respect the fact that different societies have different social behavioural patterns, even if it contradicts his own belief. He should not disregard the particularity of social data and impose his universalist speculation on the data to arrive at a universal model of society. To alter the scientific methodology so as to suit one’s intended aim is the same as putting the cart before the horse. The sociological findings do not support the universalist aspiration. Sociologists have already uncovered dramatic deviance from society to society. We can at most say that common patterns of behaviour exist within a social group. In order to comply with the requirements of the scientific methodology, Durkheim has to confine the validity of social facts to a specific society.

The second characteristic of social facts is their objectivity. An objective social fact has two dimensions which Durkheim often confuses.^{30Giddens attributes the flaw of Durkheim to his inability to distinguish the difference between what a social actor faces and what a social observer sees. The first kind of objective facts is difficult to quantify (A. Giddens, Studies in Social and Political Theory, London, 1977, p. 292).} On the one hand, objectivity in relation to a social actor means that the demand, the force, or the coercion acting upon him is experienced as not from within but from without. The externality, without denying the possibility of its internalization, can readily be seen in the norms instituted by customs, morals, and laws. On the other hand, objectivity in relation to a scientific observer requires the rigorous application of a disciplinary methodology. He says:^{31E. Durkheim, The Division of Labor in Society, trans. W. D. Halls, New York, NY, 1997, p. xxix.}

To Durkheim, careful observation, accurate description, and classification, though fundamental, are inadequate in the study of social behaviour. One still has to look at social facts from the scientific perspective, that is, from the perspective that will render social facts unanimous to every (objective) social observer. Put differently, the description of a social fact should be established solely on the object itself. In Durkheim’s mind, social facts are “things” similar to physical objects.^{32Cf. E. Durkheim, The Rules, pp. 14-31.} They exist on their own and are external to human consciousness. Being strongly influenced by the positivism of his time, Durkheim believes that social facts can be quantified and measured with precision. Just as the sensation of heat and electric current can be gauged by thermometer and galvanometer, social facts can be represented by statistics.

Besides numerical data, Durkheim also recognizes other representations of social facts. In the passage just cited, he holds that social relation or social coherence is objectified in the laws of the juridical system. The primary rule for a social observer is to discard his subjective sensation and to look for external objective facts. The major sign of a social fact is its “constraint” or “coercion” exerted on individuals. Social facts are not merely external customs, norms, laws, or instructions which an individual has to consult before he acts. Rather, they are constraining forces that “impose themselves upon him, independent of his individual will.”^{33E. Durkheim, The Rules, p. 2.} Durkheim repeatedly stresses on the refractoriness of social facts to human will. Elsewhere he writes, “Indeed, the most important characteristic of a ‘thing’ is the impossibility of its modification by a simple effort of the will. Not that the thing is refractory to all modification, but a mere act of the will is insufficient to produce a change in it.”^{34E. Durkheim, The Rules, pp. 28-29.} It is interesting to note that Durkheim emphasizes the external coercive property of human behaviour in contrast to the Kantian accent on the actor’s will. Social facts will not easily bend to the will of individuals. They will by all means force themselves upon an individual until the social actor conforms to them. The constraint of social facts becomes apparent in the socialization process of education. Through it, society imposes “on the child ways of seeing, feeling, and acting which he could not have arrived at spontaneously.”^{35E. Durkheim, The Rules, p. 6.} It is a common practice for every society, organization, party, or family to fashion its members in its own image. Every society has its own set of social facts, which will be transmitted and revised through the process of socialization. It is almost impossible to imagine how a person brought up in society could evade being shaped by its social facts. If a person refuses to conform, he will be treated as insane and excluded from the community.

Having outlined the characteristics of social facts, let us now turn to the substance: what are social facts according to Durkheim? When he is pondering on the question about the best representation of “social solidarity” in his early work The Division, Durkheim singles out the laws in the juridical system as the only objective social facts for an investigation of social relationships. “Social life,” he writes, “wherever it becomes lasting, inevitably tends to assume a definite form and become organised. Law is nothing more than this very organisation in its most stable and precise form.”^{36E. Durkheim, The Division, p. 25.} Durkheim is not unaware of the fact that social relationships are represented by the morals in a broader scale, which actually form the basis of legal laws. But he entertains the ideas that some moral rules, though reflecting certain social relationships, are placed outside the scope of law, and that there are some other moral rules which even go against the underlying principles of law. For the former, the very fact that they are unable to attain a legal status reveals their secondary nature. Since they are not crucial to the sustenance and preservation of the social relationships, there is no need to elevate them to a legal standing which is backed up by the coercion of the state. For the latter, it is rare that moral rules would oppose the legal laws. If such were the case, then the moral rules would be “pathological” or “abnormal.” We are not told in the Division how to deal with abnormal or pathological moral rules. It seems that Durkheim treats them as rare cases that can be put aside. He simply assumes that “the law reproduces all those types [social relationships] that are essential, and it is about these alone that we need to know.”^{37E. Durkheim, The Division, p. 26.} In other words, the laws are the objective scientific data and at the same time fully represent the social relationships of that society.

Later Durkheim realizes that legal laws alone are insufficient for the investigation of the complex social phenomena. In The Rules, he amends the definition of social facts. Instead of listing the categories of social facts, Durkheim formulates a general definition which includes “every way of acting, fixed or not, capable of exercising on the individual an external constraint.”^{38E. Durkheim, The Rules, p. 13.} This is a lax statement focusing on constraining rules, which may include laws, morals, beliefs, customs, professional practices, and even fashions.

Durkheim is not a disinterested social observer. He creates a distance only to serve society better. His concern for society becomes conspicuous when he makes a distinction between “normal” and “pathological” social behaviour. This categorization is as important as the separation of the good from the evil in morality. We must know what is good or evil before we can promote the good and suppress the evil. Similarly, a sociologist must determine what is normal and what is abnormal before he can write a prescription for society. Durkheim sees the task of a sociologist not merely as mapping the contour of society. He is more like a doctor. The medical doctor diagnoses the physical condition of a patient, and prescribes a regimen to promote the patient’s overall well-being. Likewise, the social doctor has a duty of bringing society to a more stable state or mature stage.

We see an overlapping of duties between the emerging moral sociologists and the traditional ethicists who whip society toward a higher ideal. We are not sure if Durkheim intends to replace the old trade with the new profession. But he does criticize the ethicists’ approach to social behaviour as subjective and dogmatic. They presumptively set up a standard and use it to measure society. Every act which conforms to it is deemed to be desirable, and every act which deviates from it is condemned. Durkheim discerns that history shows another story which is much more complex. For example, polygamy is unacceptable to Christians and modern liberal society, but for the ancient Israelites, as well as many ancient nations, such marriage was regarded as necessary or even good. In the case of modern social system, the socialists accuse the capitalists of exploitation, and the capitalists charge the socialists with tyranny. Why do we have two antagonistic judgements regarding the same phenomenon? Why do some people accept polygamy, and some don’t? Why do some people accept capitalism, while some denounce it? The traditional ethicists will resolve this conflict by seeking out an absolute principle. One group may defend monogamy as the ultimate form of marriage or capitalism as the best way to organize society. The opponents may advocate the opposite value. The two sides compete with each other by propagating their own views. Sometimes this side wins; the next turn the other side carries the day. Contrary to the ethicists’ purpose, the outcome is interminable conflicts between different parties. Durkheim attributes this consequence to the one-sided perspective of the partisans in opposite camps. “The common flaw in these definitions,” he says, “is their premature attempt to grasp the essence of phenomena.”^{39E. Durkheim, The Rules, p. 54.} He opines that a moral sociologist should look at the social facts from a scientific perspective before passing his judgement. To explain a social phenomenon, he should use the principle extracted directly from the data themselves rather than to apply an extrinsic principle, whether it is derived from personal experience, or handed down from tradition, or borrowed from another field of science.

Durkheim uses crime to illustrate his point. Without doubt, Crime is everywhere regarded as evil. Society always discourages deviance from social norms. Instead it tries to eradicate some noncompliant acts by labelling them as crime and by imposing on them some social sanctions. Notwithstanding severe punishment, crime persists throughout history. There is no crime-free society in the world. Moreover, Durkheim notices that in contrast to our expectation, the rate of crime increases rather than decreases as society advances. This evidence compels us to rethink the role of crime. Perhaps, we should treat crime as a permanent and normal condition. Crime or deviation from social norm, Durkheim proposes, is necessary and beneficial to social health. For instances, Socrates, Gotama Buddha, and Jesus Christ have all been regarded as moral revolutionaries who have contributed to the well-being of humanity. But at their times, they not only violated the norms of their societies, they were actually looked upon as rebellious and dangerous men. Socrates and Jesus were both held as criminals. They were put on trial, convicted, and executed. Afterwards, people began to recognize the value of their teachings, which subsequently became the new norms of societies.

Another reason why crime is permanent and normal relates to the essence of society and the psychology of its members: ^{40E. Durkheim, The Rules, p. 67.}

Durkheim judges that it is the very nature of society to foster strong collective sentiments in order to maintain internal cohesion. Such collective sentiments are also the sources of criminality. A collective sentiment by definition is a sentiment imposed from without on every member of a society. Reasoning in the same way as Newton in his third law of motion—action and reaction are equal in magnitude but opposite in direction, Durkheim proposes that in reaction to the imposed collective sentiment, a member will generate an equally intense personal anti-sentiment.^{41This principle was known to the ancient people of China. They formulated it as yin yang. Confucius has suggested a different way that encourages people to do good without imposing a collective sentiment.} If the collective sentiment can successfully inhibit the personal anti-sentiment, a person will not commit the crime. But his anti-sentiment will seek other outlets and express in other forms of anti-social behaviour. Sometimes the collective sentiment does not work and some people may transgress the law and become criminals. Either way, crime cannot be eradicated. Of course, we can diminish the number of transgressions by making a more reasonable law or by providing better social service to alleviate personal anti-sentiment and frustration. But this cannot eliminate transgression altogether. Alternatively, we can abolish the collective sentiments by legalizing certain crimes such as divorce, homosexual marriage, abortion, and drug-consumption. This will automatically wipe out the crimes and the corresponding collective sentiments. Nevertheless, society has other collective sentiments which can never be thrown away for the mere sake of social cohesion and solidarity. Thus the types of crime may change, the number of transgressions may diminish, but crime still remains in society.

With this understanding in mind, Durkheim re-defines the normality of social facts as follows: ^{42E. Durkheim, The Rules, p. 64.}

Durkheim takes a relativist view. He thinks that morality is relative to culture as well as to time. Crime per se should not be considered as pathological. On the contrary, it is a normal social phenomenon for it exists in every society in every phase of its evolution. Cohabitation, polyandry, polygyny, monogamy, and sequential monogamy (one spouse at a time) do exist in different societies and in different periods. They are normal in their respective time and communities but abnormal if they appear at the wrong time or in the wrong place. For instance, polyandry and polygyny are abnormal in liberal society, while cohabitation, monogamy, and sequential monogamy are normal social facts. Even suicide is not to be considered as a universally recognized pathological behaviour.^{43Cf. E. Durkheim, Suicide. A Study in Sociology, trans. J. A. Spaulding and G. Simpson, London, 1970, pp. 208-216, 361-370.} In a certain society, at a particular time, where individualistic and egoistic trend is prevalent, suicide may be seen as a self-expression of egoism. People commit suicide because they can no longer balance the conflicting demands of their society, and they choose to follow the suicidal trend to end their lives. If the number of suicide is moderate, it should be considered as a normal behaviour. Society needs not intervene if the statistic does not reach an alarming level, which indicates that suicidal trend has become abnormal. What is normal or abnormal is relative to the social conditions; it cannot be judged by criteria outside the social facts.

Durkheim’s proposal for using scientific methodology to investigate social behaviour from a sociological perspective has its ground. Nowadays, most people accept that an explanation of social behaviour cannot be reduced to psychology. In addition, sociology has won over a sizeable number of policy-makers. It symbolizes a kind of authority, and becomes an indispensable reference in making social policy. Durkheim’s criticism of ethicists’ negligence of social facts has to be heeded. Nevertheless, his sociology of morals is not free of flaws. Since many critics have already voiced their criticisms, it does no harm not to repeat them here. But for our purposes, I will highlight those problems that are relevant to our discussion.

The aim of sociology, according to Durkheim, is not merely descriptive. Were it a social drawing, it would not, he says, “worth the labour of a single hour.”^{44E. Durkheim, The Division, p. xxvi.} Sociologists have the mission to maintain society in a normal state, or to transform it into a new equilibrium. Assuming a certain society is in a stable state, we can measure the average of a certain act or find out the majority opinion about it. We take this average or majority behaviour as the norm, which is in turn used to curb deviant social behaviour. If such method is strictly and effectively put into practice, it is not difficult to imagine that such society will become stagnant: it has neither the momentum to pursue a higher ideal nor the ability to adapt to the ever-changing environment and demands of its members. Such society will collapse before long. The problem and limitation of the science of morals are that either it clings to its descriptive objective and remains a kind of moral anthropology, forfeiting the ambition to regulate society, or it risks becoming the ideological machinery of the state apparatus, which is ordained to maintain the status quo at all cost. For most ethicists, neither of them is desirable. A second serious challenge to be raised against Durkheim is that his science of morals is inapplicable to society in a state of flux. Suppose a certain value is in dispute, and there are more or less equal number of supporters for and against it. How could we determine which one is normal and which one is pathological? It seems that the sociology of morals is helpless in the case of a moral dispute, and that the issue has to be settled ultimately by political bargaining or by philosophical argument.

Durkheim is well aware of these difficulties. Indeed, he is often confronted by his opponents. In a discussion with Gustave Belot about the transition from the scientific study of moral facts to the extraction of moral principles, Durkheim asked his fellow philosopher: “ How does the objective study of moral facts permit new ends to be determined which are different from those that the given morality assigns to conduct?” After Belot gave a not-so-satisfactory answer, Durkheim persisted: “How can the principle of a new moral orientation be created from even the complete systematization of moral data?” Durkheim does not put forward these questions to embarrass his friend. What he wants is a second opinion. “I am raising the objection which has often been made to me,” Durkheim said, “and I should be happy to see whether you are better able than I am to reply to it.”^{45E. Durkheim, A Discussion on Positive Morality. The Issue of Rationality in Ethics, in W. S. F. Pickering (ed.), Durkheim. Essays on Morals and Education, London, 1979, 52-64, pp. 57-58.} At the end, Durkheim parted from his friend disappointed. And his perplexity remained till the end of his day.

3. Between fact and meaning

Notwithstanding his confession of the problems in the sociology of morals, Durkheim insists that ethics must be founded on social facts. In his review of Lucien Lévy-Bruhl’s Ethics and Moral Science, Durkheim comments: “One will find in this work, analyzed and demonstrated with rare vigor, the very idea that is basic to everything we are doing here, namely that there is a positive science of moral acts, and that it is on this science that the moralists’ practical speculation must rely.”^{46E. Durkheim, Emile Durkheim. Contributions to L’Année Sociologique. Edited by Y. Nandan, New York, NY, 1980, p. 130.} Apparently Durkheim separates the study of morals into two branches. One is the new science of morals and the other the philosophy of morals (or ethics). The former employs scientific method to investigate morality, while the latter uses speculative method to appraise morality. One concerns the explanation of morality, and the other its evaluation.^{47Cf. E. Durkheim, Sociologie et philosophie, Paris, 1924, pp. 49-54.} According to Durkheim’s ideal, the sociology of morals precedes ethics and forms its basis, that is, the critical standards must be in some way related to the social data. Unfortunately, Durkheim is unable to bridge the gap between them. But neither could the theoretical moralists. Some philosophers do attempt to ground their critical principles in some moral data. For instance, Kant bases his theory on the common notion of duty, and Rawls applies the method of reflective equilibrium. It is not true that the moralists entirely neglect the social facts. Their common problem is that they consider only a limited set of moral data in their construction of their theories. A moral or political theory founded on scanty social facts is definitely not an adequate theory to comprehend and to evaluate complex moral behaviour. A better moral theory should formulate its critical moral criteria from a comprehensive set of observable data.

Walzer’s theory circumscribes a broad base of social data, and it also contains critical principles. In my judgement, Walzer has accomplished the unfinished task of Durkheim. Walzer’s success hinges on the paradigm shift from social facts to social meanings. He unties the knot that Durkheim has made out of his preoccupation with the objectivity of social facts. In order to assess Walzer’s achievement, we have to compare social fact with social meaning. Let me first recapitulate the main characteristics of social facts. They can be classified under the headings of source, particularity, objectivity, substance, and relativity. For particularity, both Durkheim and Walzer take society and its related ramifications as the artefacts of a particular group. Social fact and social meaning are thus particular per se. I assume that the reader would have no difficulty to recognize this similarity if he has read Walzer’s particularism in the preceding chapter. To spare the reader from boredom, I shall not repeat it here. For the rest of the characteristics, I will compare them with those of social meaning. The order of the subjects as laid out above will not be followed for reason that will become self-evident.

a. Source

Walzer’s basic moral tenet is that there exists a moral world in which we live, and by which our perception of reality is (not entirely but to a large extent) shaped and our behaviour constrained. The main task of an ethicist is not to seek out a transcendent moral law, nor to invent an advanced social principle that belongs to a higher evolution stage, but to interpret to one’s fellows the moral world they inherit. Walzer concretizes this idea early on, when he wrote the Just and Unjust Wars: ^{48Wars, p. 20.}

To Walzer, human moral behaviour is a learning process. We learn how to act in a socially acceptable way by studying those who precede us. We can hardly think of any present moral norm that is not in a certain way connected with the past. The learning process, however, is not merely mimicking. A mature actor has to comprehend his inherited moral world and analyse the present situation before he can figure out the best way to act. This moral deliberation is very complex. There does not exist a scientific perspective that will make for a unanimous decision. We may insist on making moral decision in a systematic way. Then several systematic perspectives may turn up. Opposing camps fight with each other, and each side claims that its resolution is the right answer. Different arguments are then put forward to fuel this endless debate. Some people, who come to be known as realists, perceive the indeterminacy of morality as a sign of power struggle. They think that there is no such thing as moral law in reality. Morality is the invention of the powerful to control the plebeians and to cover their political ambition. The realists want to unveil morality, and show us the real working of the world lest we would mis-calculate and fall into a disadvantageous position. To support their claim, realists commonly point to the extreme case of war. War, they contend, is only a matter of conquer or being conquered, a matter of killing or be killed, and there is no place for morality. Inter arma silent leges, thus says the maxim. Or, the realists can argue that morality changes from time to time and varies from place to place, and that it is futile in most cases to talk about morality because people can hardly understand each other. Against this background, Walzer insists that there is indeed a moral world, which precedes us and is external to us. It confronts us in our everyday life. The signs of its existence can readily be seen in everyday complaints and the responses to the complaints. We couch them in a common language, which is mutually comprehensible to us and to our opponents.

Later in Spheres of Justice, Walzer again invokes the same idea, this time in a modified and more articulate form. He says that an ideal of society already exists but is hidden in our world of shared understandings: ^{49Spheres, p. xiv.}

Modern political philosophers and sociologists have an overwhelming ambition to reform society. They first set up an ideal society, which is deemed to be superior to the present one, and use it as the standard to refashion the old social order. This practice is probably inspired by the idea of evolution. The life of organisms gives us a primitive concept of evolution. An organism grows from birth to death through various stages of development. Likewise, species develop from a simple cell into a multitude of complex organisms. When we borrow this developmental idea and apply it to society, then society can be viewed as a collective organic mechanism. It grows, develops, and passes through various stages of evolution. The philosophers and the sociologists think that they know the next stage of evolution or even the finality of society. Some lay out blueprints for a better social arrangement, and some design strategies to speed up the evolution process. Despite their relentless effort, Walzer thinks that it is only a fanciful utopia—an imaginary “no where.” A recognizable and workable good society can only be an existing one. If it does not exist, how do we know that it is good, or that we will be happy to live in there? We need to work out the ideal not because it does not exist, but because it is hidden in the moral morass. Our task is to unveil it. So, both Walzer and Durkheim assert that society contains the moral norms within itself.

b. Relativity

Particularism inevitably leads to relativism in the present world structure as we know it. They are, in fact, the two sides of the same coin. In order to introduce the idea of relativity to the science of morals in the Western world, where a strong universalist tradition prevails, Durkheim makes a distinction between moral ideals and moral facts. He does not deny the existence of universal values, but he separates them as moral ideals. The main concern of moralists is to find out universal moral standards and to use them to appraise the status quo. But social scientists have to put aside moral ideals, look for the social facts, and then extract principles from them. Such approach generates astonishing results and weird conclusions. Crime, for instance, is to be classified not as pathological, but as a normal persistent phenomenon. Even suicide should not be regarded as a universal pathological act because it may in fact be promoted as an expression of self-determination in the individualistic society. What is normal or pathological cannot be predetermined. It must be evaluated against the developmental history of society and the present stage of the society at stake.

Such kind of parallels can also be found in Walzer’s works. First of all, Walzer makes a distinction between good and justice. Good is traditionally conceived as a universal and ultimate idea. But Walzer thinks that it is possible, and indeed more practical, to think about morality without reference to this eternal idea. He refuses to discuss the idea of good because there is no sure way of knowing it. In the concluding chapter of the Spheres, he says, “I shall not attempt here to consider the question whether societies where goods are justly distributed are also good societies. Certainly, justice is better than tyranny; but whether one just society is better than another, I have no way of saying.”^{50Spheres, p. 312.}

To define justice, Walzer seeks its criteria empirically. After he has looked into our world, he concludes: “There are an infinite number of possible lives, shaped by an infinite number of possible cultures, religions, political arrangements, geographical conditions, and so on.” Communal life is contingent; its present shape depends on many factors. Since these factors will not be identical for two peoples, it is improper to use the ideal of one society to judge the other society. A more reasonable way to assess a society is to judge it from within and by its own ideal. Thus comes the oft-quoted definition of justice:^{51Spheres, p. 313.}

To be sure, most philosophers recognize that there are variations of justice. They admit that a just society can be implemented in an infinite number of ways but only within certain limits. For example, we have the American democratic system and the European democratic system. Some countries still retain their kings, queens, or nobility. Some have more female political representatives than the others. But these features are not regarded as crucial in evaluating the societies. So long as they follow a set of approved democratic procedures, they are all regarded as versions of democracy. Since democracy is the only justifiable political system in modernity, they are all versions of just society. While the political system of China has not come up to the Western democratic standard, it is not counted as a just society. Here, Walzer differs from the universalists by giving up predetermined values. Looking from an empirical point of view, social hierarchy exists in virtually every society, and inequality, contrary to what some progressive sociologists would like to have us believe, increases as society becomes more complex. Even in theory, organization works against equality. The more effectively a society is structured, the more unequal it has to be and it becomes. In the very beginning of the Spheres, Walzer writes, “Equality literally understood is an ideal ripe for betrayal.” And the reason seems obvious enough: “Committed men and women betray it, or seem to do so, as soon as they organize a movement for equality and distribute power, positions, and influence among themselves.”^{52Spheres, p. xi.} We have no way other than to recognize social hierarchy as a normal persistent phenomenon, and ask what level of inequality is acceptable to a given society, and what degree of equality is feasible. In every case, evaluation has to be made according to the social meanings shared by that particular society. To stress his point, Walzer pushes this argument to its extreme by saying that the Indian caste system could be justified. His argument offends the sentiment of most Westerners who find the caste system unthinkable. The caste system nonetheless has its internal meanings and justifications, which are, at least it seems to me, not so difficult to understand.

c. Substance

The major difference between Durkheim’s social fact and Walzer’s social meaning has two dimensions: the first is concerned with “fact” and “meaning,” and the second the understanding of what is social. Looking from Durkheim’s perspective, society is open to observation, and can be reported objectively as facts. A social scientist can extract principles from the facts, and use them in turn to explain the facts. This methodology apparently works effectively in natural sciences, but it encounters difficulty in its application to complex social phenomena, for human actions, unlike physical movements, carry meanings. It is inadequate just to observe the actions without taking into account their meanings. Therefore, the real aim of a social scientist, for Walzer, should be looking for meanings instead of facts.

Furthermore, Walzer’s criteria of “social” differ from those of Durkheim. In defining social facts, Durkheim lists two criteria. A fact is social if it is:^{53E. Durkheim, The Rules, p. 13.}

A social fact must satisfy either one or the other. Although social meaning is significantly different from social fact, Walzer also claims that social meaning is external to an individual and exerts coercion on the social actor. In the discussion of the social meanings of goods, he says, “Distributive agents are constrained by the goods they hold; one might almost say that goods distribute themselves among people.”^{54Spheres, p. 7.} Walzer wants to emphasize the point that goods exert coercion on the distributive agents because of their social meanings. Social goods are not neutral things that can freely be disposed of. They carry with them social meanings, which force the agents to comply. If an agent refuses to use a good according to its social meaning, he will encounter adverse repercussion. Both Durkheim and Walzer speak of coercion and externality. Nevertheless they differ at some delicate points.

There are at least three sources of coercion related to the social meanings of goods: the internal mechanisms of social goods, social sanctions, and the imperatives of moral ideals. They originate from the goods, the society, and the individuals respectively. Certain goods undeniably have intrinsic meanings, that is, by analysing the operation of the goods, we know how they should be used properly. Public honour, for example, is the recognition of the achievement of a person in a particular domain. Accordingly, it should be distributed to the person who deserves it. A woman swimmer who wins a competition fairly in the Olympic Games is awarded a gold medal. She is said to deserve it, and nobody can take away this honour from her. Suppose she leaves the medal in a taxi on her way to the airport, and somebody takes it and never returns it to her. She is still the owner of the title; the one who possesses the medal does not possess the title. Medicine is invented to cure disease. It should be distributed to the sick. It is a cruelty if it is administered to the healthy. Money, too, has certain internal mechanism. Valid banknote can only be issued by the authorized agency. If someone draws a paper money and uses it to buy food in a grocery, no matter how real it looks or how artistic it is, he will probably be sent away empty-handed. The intrinsic meanings of goods impose constraints on the agents. The agents cannot dispose of them in whatever way they like. The above three cases, though differ somewhat in their operation and strength, clearly show the limitations of social meanings on the social actors. We can reasonably assume that every good exerts its own constraints on the human agent who exploits it.

Walzer’s critics, however, are quick to point out that only honour and love can be said to have intrinsic meanings. Although it is nonsense to distribute medicine to those who do not need it, it is equally nonsense to say that medicine itself dictates to whom it should be distributed. A need does not automatically appropriate a right. Indeed society simply cannot answer every need even if we want it to. It is also true that most people will not accept fabricated money as the medium of exchange. But genuine money can effectively play the role as the universal medium of exchange. It has no restriction on what it should not buy. It can be used to buy political influence, commodities, beautiful women, servants, slaves, or anything that anyone wants to sell if it is not for the regulations imposed from without.

The critics are certainly correct, but yet they miss the point. Walzer, for most of the time, is not talking about this kind of intrinsic meanings. Rather, he is referring to the social meanings that society creates and imposes on goods. Intrinsic meanings are only the smaller part of social meanings. For the larger part of social meanings, their coercive force comes not from their internal operations but from society itself. Such force manifests itself in various forms and with different degrees of coercion. For instance, law is the most visible form of public sanction, which has been organized as the most systematic and regulatory coercion. Besides that, we have moral norms, rituals, customs, taboos, tacit consent, etc. Together, they mandate the socially acceptable behaviour, or better still, a particular mode of life and a collective identity. Any violation of them is regarded as an offence against the society, and will be met with public resistance. The purpose of public sanction is to bring deviant behaviour into compliance with the established social practices. In the case of medicine, it is true that need does not entail health care to be distributed to every patient, nor does need dictate it to be distributed equally. Medicine has, however, acquired a history of social meaning, and because of that, need is only one element of moral deliberation. As pointed out by Walzer, the cure of the souls (pastoral care) in the Middle Ages Europe was public, while the cure of the body (health care) was private.^{55Cf. Spheres, pp. 86-91.} Christians believed that eternal life was the most valuable thing. Thus, society had the foremost duty to organize its distribution so as to ensure that it was equally distributed to every member. In contrast, they ran medicine in the old Hippocratic tradition in which doctors only cared about the rich and the wealthy. Due to the advent of modernization, Europeans gradually view the soul as an abstract concept. The body becomes more real and more prominent. Consequently, the cure of the souls is transformed into the cure of the bodies. Medicine has since become public despite the fact that some doctors still cling to the Hippocratic tradition and resist outside regulation as much as possible. Nowadays, every government has to provide decent and equal health care to its citizens and residents. This principle can easily be confirmed in the Western European countries. For them, it would be outrageous to talk about privatization of health care. Likewise money has acquired social meanings, if mostly negative: it is illegal to buy public office; it is immoral to buy a wife or a husband, etc.

The third source of coercion comes from moral ideals, which exert strong influence on some people even in the absence of social sanction. Of course, moral ideals are also a kind of social meanings. (What I refer to as moral ideal is what Walzer designates as “hidden,” “deeper, ” or “real” social meaning.) Here, I intend to confine the term to refer to those ideals which are not yet widely recognized by the public but nevertheless are entities of the shared social meanings. I may immediately be accused of introducing contradiction into my statement. How can a meaning be shared but not widely recognized? If it is not accepted by the majority, can it still be said to be shared or social? This complication is possible and indeed exists in every society because social meaning is accumulative and thus historical. As a result, it is not solely dependent on the contemporary society. People may forget some social meanings of a good. Some may even distort the social meanings so as to benefit their own group. But because of the history of the social meanings, we can still recover the real meanings. Furthermore, “social meanings” treated as a complex whole has internal contradictions. It requires interpretation to resolve the conflicts and render the ideational system more consistent.

Let us take up health care again. The social fact about medicine in the United States is that the majority regards it as a marketable good. Perhaps government should provide a certain amount of health care to those who cannot afford it themselves. But the US citizens see no reason why they should not buy better health care if they could. Medicine at least is partly a commodity. Now, Walzer argues that the real meaning of medicine obliges it to be distributed equally to all citizens who need it. It implies that medicine as a marketable good is not the real meaning. Its deeper meaning is the cure of the bodies whose root is in the cure of the souls. If Americans consider their culture a continuation of that of the Western Europeans, they should admit that the cure of the bodies is their moral ideal. This is, it seems to me, logical enough. Nonetheless, the cure of the bodies is not widely accepted by the American public in the meantime. Walzer’s notion of medicine is only a moral ideal, which is hidden in the culture. If there is no legislation or public action, it would probably never be realized. In what sense is such moral ideal coercive?

Walzer is fully aware that a moral ideal has only internal moral urge. In his comment of the shabby welfare system of the United States, he writes:^{56Spheres, p. 84.}

On the one hand, Walzer agrees that there is nothing wrong in principle with not providing equal health care if this is the decision of the democratic citizens. But on the other, he does want to uphold the moral ideal by charging them with not living up to it. What kind of coercion, if any, has the moral ideal on the citizens? I think it can only exert some internal forces on those who accept it, such as bad conscience, a feeling of imperfection after its violation, or an urge to change the status quo. But this coercion comes by degrees. Different people may hold different attitudes. This is permissible since the moral ideal at stake has not yet been accepted by the public. As for those who doubt Walzer’s interpretation, there will be no sense of coercion or obligation at all. For the category of moral ideals, its coercion is not external and uniform, nor does it generally exist in society. Obviously, it does not fulfil either of Durkheim’s criteria. But it satisfies the second part of criterion (2), namely that it exists in its own right independent of its individual manifestations or denials.

d. Objectivity

Objectivity is, for Durkheim, the touchstone for sociology to be qualified as a science and not philosophy. Thus he emphasizes that social facts are things as real as material things. Like the physical world, they can be observed and subject to the same rigorous scientific investigation. And he believes that scientific laws about the social world can be discovered from the collected social facts. This optimism is one of the consequences of positivism that permeated the nineteenth and the early twentieth centuries. Durkheim is particularly influenced by Descartes’s philosophy. On his way to define social science, Durkheim wholeheartedly embraces Descartes’s programme of self-detachment. Concerning the prerequisite of a genuine social scientist, he says, “it is not enough [for a social scientist] carefully to observe, describe and classify them [collected social facts].” The ultimate test lies in his ability to, “in Descartes’ phrase, discover the perspective from which they become scientific.…” But how could we discover the scientific perspective? Descartes recommends us a programme of self-detachment. (A secular form of self-denial?) To discover the true nature of a thing, we must repeatedly admonish ourselves until we relinquish all our bias and prejudices, until our mind is as white as a printing paper ready to receive without distortion any image that happens to imprint on it. The scientific perspective, in fashionable language, is the impartial perspective. We see thing as it is, so to speak, not as we feel, think, or want it to be. Though impartiality is still popular in academic and political arenas, noticeably in the Anglo-American circle, its rhetoric is losing force. Many people begin to doubt its validity. Could we really empty ourselves towards nirvana and become nobody? Even if we could do this, how about the object of our investigation? Could society, the artefacts of human hands, be treated in the same way as a physical object? Could it be objectively observed? Could the facts thus extracted present themselves as governed by law?

Putting these questions in this way makes myself clear that the expected answer is “No.” My objections to Durkheim are that he uncritically accepts the positivist notion of objectivity, and that he mixes up fact and reality. Sociology as the study of social facts already contains, Giddens points out, “a confusion of more than purely terminological significance.” He reasons:^{57A. Giddens, Studies, p. 292.}

Durkheim’s lack of sensibility towards fact and reality can readily be seen in the first chapter of The Rules. When he introduces “social facts,” he writes, “each individual drinks, sleeps, eats, reasons….” These are, of course, social facts about people’s everyday life. But he goes straight to include social obligations—such as the roles of brother, husband and citizen, and social contracts—in the domain of social facts. These obligations and contracts are more than facts about society; they are themselves the constituent elements of society. Daily human activities can perhaps be described by simple statements, but social institutions can neither be fully described nor explained without taking into consideration the culture and value-system. We can, without great difficulty, find out whether two persons are brothers or not. But how can we observe the role as a brother? At a closer look, even the habitual activities of necessity, like eating and drinking, cannot be stated as facts. For example, the Jews have continuously built up a complex system of kosher food to distinguish which food can be eaten and which cannot, and to prescribe the proper manner of preparation and eating. We can give a detailed description of their practice, but without taking into account the Jewish history and culture, we will never understand what kosher means to them. And we would be puzzled as to why such people bother to keep so many food regulations.

This confusion of fact and reality is linked to the assumption that what the social observer observes is what the social actor experiences. In the study of society, it involves a society, the social actors, and at least one observer. An actor lives in the society and is confronted by the society. Durkheim assumes that the observer can observe the action of the actor, and knows exactly what the actor is doing without consulting the actor. Since the observer is detached from the society, and hence obtains a kind of objectivity, he is in a better position to understand the actor and the society. The validity of this proposition involves complex issues that we cannot deal with here. Let me simply state some limitations of a social observer. It is certainly not true that an observer can totally get rid of his bias, prejudices, values, and world view. An observer is equally not at an advantageous position if he does not know the culture and history of the society which he observes. It is also highly improbable that the collected data will reveal their underlying law or laws without the observer’s active effort to impose on them a paradigm which he finds encompassing.

Walzer’s idea of objectivity significantly differs from that of Durkheim.^{58Walzer has taken cognizance of the above-mentioned limitations and has devised measures to accommodate them. His reconstruction of objectivity merits a separate treatment on its own, which I will deal with fully in a later section.} The change from using social fact to social meaning indicates a fundamental shift in the view of objectivity. The new viewpoint recognizes the distinction between the social actor and the social observer. It denies the observer his attributed superior point of observation. It refuses to admit that the observer can extract law solely from social data without consulting a presumed paradigm or model. Nor has he the privilege to interpret the data according to the model of his own choice if his description is meant to be an accurate representation of the society. The social actor is continuously confronted and constrained by the society. The outside observer can hardly observe these forces. In other words, the social observer is not situated in a better position to understand or to explain what an action means to the actor within the social milieu. Most probably, he is in a less advantageous position since his detachment makes him insensitive to the acting social force. Stripped of his privilege, the observer has to interpret and explain social data from within the context of the society that he observes. Social data cannot be simply regarded as facts that everybody in a neutral position can pass the right judgement on. They carry meanings of their own, whose validity can only be judged within their respective social milieu.

The main difference between social fact and social meaning lies in their different views of objectivity. Durkheim sees society as containing observable facts which are regulated by scientific laws. After getting rid of one’s prejudices, the social scientist can accurately and truthfully formulate the laws that control human behaviour. Walzer probably does not believe that human behaviour can be summarized as a set of laws. He doubts if human behaviour can adequately be circumscribed either by laws or by principles. People have created society as a world of meanings. Their actions have to be understood from within the frameworks of meanings. There is no superior objective observational position. It is not the fact but the meaning that is crucial to the understanding of human behaviour.

The transition from fact to meaning bridges the cleavage between scientific theory and philosophical ideal. This dichotomy is the product of the debate between scientists and philosophers in the nineteenth century. The scientists insisted that they were objective in their methodology while the philosophers and the theologians were speculative. Only knowledge that came to us through objective methodology could be true or useful knowledge. Philosophical or theological knowledge was at best unverifiable, and thus inapplicable to the construction of modern society. In reality, both philosophers and scientists are searching for the hidden essence of an object. The real difference lies in their methodology. Philosophers tend to abstract it directly from the object through reasoning and the manipulation of language and concept. Scientists claim to acquire knowledge of the object by observing and experimenting with it in the absence of personal involvement. Walzer accepts the scientists’ criticism of the speculative methodology as unverifiable. But at the same time, he criticizes this kind of objective scientific view as untenable. The scientists are not aware of what they are really doing. No matter whether it is observation or experimentation, a scientist is actually experiencing the object through his own senses. To put it another way, he does not know the object as it is; he knows it only as he experiences it. Furthermore, he is not satisfied with the experience alone; he aims to make a theory out of it. Such act will no doubt attribute meaning to the object. Walzer’s search for meaning is pointing towards the dissolution of the distinction between scientific truth and philosophical truth. There are objects external to and independent of human existence, with which human beings are in constant interaction. One of the activities noticeable in the human mind is the comprehension of the external world by way of attributing meaning to it. And this process is a social one, that is, scientific research is conducted in and regulated by scientific communities. We may say that there is only one kind of social knowledge that can be assessed by reason empirically, which is social meaning. Since moral ideals are a kind of social meanings, the division between sociology of morals and philosophical ethics will be annulled.

C. Cultural tradition

Walzer’s approach to morals may also be compared with the anthropologists’ approach to cultures, in particular Clifford Geertz’s interpretation of cultures. I have to stress beforehand that though there are some outstanding similarities between them, I do not claim that Walzer derives his methodology from Geertz. As far as I know, there is one author, Michael Rustin, who makes a direct connection between Walzer and Geertz. In a footnote to his article Equality in Post-Modern Times, Rustin writes: “Walzer has in these matters [cultural analysis] been deeply influenced by the writings of his anthropological colleague at the Institute of Advanced Study, Clifford Geertz. The Spheres has as its basic method the ‘interpretation of cultures,’ in the form of elucidating the meaning of a whole variety of embedded social practices both of his own and of other societies.”^{59M. Rustin, Equality in Post-Modern Times, in Pluralism, 17-44, p. 20, n. 6.} Whether or not Walzer is “deeply influenced” by Geertz, I am not so sure. The fact that Walzer only generally mentions, among others, the name of Geertz in the Acknowledgments of the Spheres but not in the content or the footnote makes it difficult to assess to what extent he is influenced by Geertz. It is true that in the Spheres, Walzer is applying a methodology that bears striking resemblance to “the interpretation of cultures.” But one has to take into account that this methodology has appeared, albeit in a less developed form, in the Wars. Anyway, I am not so much concerned with who influences whom. My attempt is to understand “social meaning” from an anthropological perspective. Since Walzer has not comprehensively laid out the theoretical basis of his interpretation of morality, a study of Geertz’s anthropological methodology can yield some supplementary information.

1. The concept of culture

Social science, ever since its emergence, has quickly gained public acceptance. They are consequently institutionalized and incorporated into the structure of university as a distinct discipline. Sociology from its very beginning is designed as a tool to help shape the modern world, which is driven by the impulse of the endless accumulation of capital. As documented by Karl Marx, this capitalist world-system is full of internal contradictions. In order to mitigate social conflicts, various solutions have been suggested, and they are commonly classified in the lines of the right, the left, and the centre, or more tellingly as conservatism, socialism/Marxism, and liberalism. The conservatists strive to preserve the status quo and keep the entrepreneurs in check, while the radicals demand to transform society to the ideals of modernity in a short lapse of time. To balance both sides and at the same time to permit the capitalist principle to prevail, liberals ingeniously choose sociology as the handmaiden of modern society and use it to buffer the demands from the conservatists and the radicals. Sociology presents itself as an expert-system and employs the general strategy of criticizing conservative values as oppressive to true human nature and radical ideals as far-fetched. It claims its programme to be the middle way and the sure guide to social progress. In fact, it always aims to open the ways to business opportunities and technological innovations. Its so-called scientific perspective is overdetermined by these motives. Since sociology is geared to the liberal project, it is not surprising that sociologists encounter difficulties when they apply this methodology to study societies other than the advanced capitalist ones. As a result, two new branches come out to meet this challenge. For those nations which are non-Western and display strands of high culture, they are studied under the category of Oriental Studies. Since the histories, cultures, and social relationships of these nations are so complex that the researchers are not willing (or unable) to invest much time and resources into something that is not immediately useful, these studies by and large remain at the level of philology. As for the rest, they are identified as primitive societies and are studied under anthropology. Being free of the liberal mandates and not overwhelmed by the immense complexity of Oriental Studies, anthropology has the hope of giving due respect to the study of society.

I have stated briefly the relationships between sociology, Oriental studies, and anthropology in the purpose of pointing out the bias of the first, the limitation of the second, and the prospects of the third. Now we may proceed to the central subject concerning anthropologists. In a retrospect of what the anthropologists have done, Geertz writes that the whole discipline of anthropology evolves around the concept of culture.^{60C. Geertz, The Interpretation of Cultures, London, 1973 (repr., 1993), p. 4.} Its classic definition can be found in Edward Tylor’s Primitive Culture, in which he says: “Culture . . . is that complex whole which includes knowledge, belief, art, morals, law, custom, and any other capabilities and habits acquired by man as a member of society.”^{61E. B. Tylor, The Origins of Culture, [Primitive Culture, Vol. I], New York, NY, 1958, p. 1.} By dedicating himself to the study of human behaviour in its totality and to understand human action as the actor understands it, Tylor claims an approach different from sociology. Its accent is on the appreciation of human artefacts rather than on the manipulation of social behaviour, or in the words of sociologists, the improvement of society.

Culture defined as such is a promising idea that has inspired and guided many anthropologists in their study of society. It, however, cannot avoid the fate of falling into its self-created pitfall. Cultural study has for some time passed its prime of experimentation and settled down to the work of detailing and clarifying. An examination of the corpus of accumulated literature reveals that the idea of the most complex whole has reached the point of causing more confusion than giving illumination; it proves itself to be too ambiguous. In order to sustain anthropology as a scientific discipline, it is necessary to pin down the concept of culture. Anthropologists have performed their duty and bequeathed us an array of definitions. In an extensive survey done in 1952, Alfred Kroeber and Clyde Kluckhohn discovered more than 160 definitions of culture.^{62Cf. A. L. Kroeber & C. Kluckhohn, Culture. A Critical Review of Concepts and Definitions, New York, NY, 1963, pp. 81-141.} Kluckhohn himself has managed to narrow them down to eleven. To give an idea of it, I am going to quote a shortened list abridged by Geertz: “(1) the total way of life of a people; (2) the social legacy the individual acquires from his group; (3) a way of thinking, feeling, and believing; (4) an abstraction from behavior; (5) a theory on the part of the anthropologist about the way in which a group of people in fact behave; (6) a storehouse of pooled learning; (7) a set of standardized orientations to recurrent problems; (8) learned behavior; (9) a mechanism for the normative regulation of behavior; (10) a set of techniques for adjusting both to the external environment and to other men; (11) a precipitate of history.”^{63C. Geertz, The Interpretation, pp. 4-5.} Geertz comments that while Kluckhohn’s collection of definitions represents an improvement of Tylor’s most “complex whole,” it remains a kind of “pot-au-feu” theorizing about culture.^{64C. Geertz, The Interpretation, p. 4.} Obviously, such kind of “mixing ideas together” is undesirable in scientific research. If each anthropologist follows a particular definition throughout his study for a particular subject, anthropology could still be considered as consisting of different ways of studying culture. But if the anthropologist shifts freely from one definition to another in one project, we can anticipate how confusing it will become.

There are two major approaches taken by the anthropologists to clarify the concept of culture. The first approach follows the behaviourist perspective in viewing culture as consisting of behavioural events. To study culture is to construct the patterns of behavioural events for a particular community. This approach is congruent with the positivist idea of objectivity. In reaction to the behaviourists, the second group of anthropologists attribute culture to the psychological structure of individuals. One of their spokesmen, Ward Goodenough, says, “Culture [is located] in the minds and hearts of men.” He thus arrives at the definition: “A society’s culture consists of whatever it is one has to know or believe in order to operate in a manner acceptable to its members.”^{65W. Goodenough, quoted in C. Geertz, The Interpretation, p. 11.} From this point of view, the task of an anthropologist is to construct a system of rules of which if anyone follows closely, he (appearance aside) will be indistinguishable from a native.

The two above-mentioned solutions are, in Geertz’s opinion, unsatisfactory. In spite of their embarking on opposite directions, they have both uncritically employed, at a deeper level, the method of reduction to the simple. The behaviourists reduce social reality to observable patterns, while the psychologists reduce it to the structure of mind and formal rules. Human action can never be merely patterns of behaviour, for a human being has both body and mind. Many human actions are not simply reflexes. They are conscious actions intended by the actors to convey meanings and to achieve ends. Most often, a social action is repeated time and again not because people are conditioned to it but because they find in it meaning which is important to them. The second kind of reductionism, though refined and intricate, is still off the mark. It is a marriage between “extreme subjectivism” and “extreme formalism,” Geertz comments.^{66C. Geertz, The Interpretation, p. 11.} Since culture is a public creation, it cannot be solely located in the mind of individuals. And the rules of something cannot be taken as the whole thing itself. The rules of football are not a football game. Few football fans will be content just to attend an intensive course in football rules instead of watching a match played by their favourite team.

Reductionism is a powerful means to comprehend a reality which is so complex that it overwhelms human comprehension. We reduce reality to the level of our capacity, and in a second step, we reorganize our relation with it. This is how the modern world is constructed. If we give up reductionism, what alternative do we have? “Thick description of the complex” is the method that Geertz proposes. This idea, he frankly admits, does not originate from him. Many anthropologists have practiced it for many years, perhaps without fully aware of it themselves. Anthropologists, since the early time, have engaged in an expedition to the hinterland and brought back exotic stories of the aborigines. Their records are thick descriptions of people’s life in rich details of customs, manner, practice, etc. The most interesting point of their works, besides the strangeness of the stories, is that they try to keep their reports as accurate to the event and the local setting as possible, and to seek explanation from the viewpoint of the natives. In this way, they lay bare before us other worlds of meanings which are significantly different from ours. Without the help of the anthropological reports, we can by no means understand the aborigines’ behaviour. Thus Geertz concludes that anthropologists are looking neither merely for patterns of behaviour nor exotic behavioural laws but for local meanings of actions as understood by the natives. From this perspective, culture is best seen as a complex reality which is constructed by human beings in interactions with themselves and the environment. In so doing, they create complex networks of meanings. “Man,” in Geertz’s vivid metaphorical language, “is an animal suspended in webs of significance he himself has spun.” It follows that the most appropriate method to study it is not “an experimental science in search of law but an interpretive one in search of meaning.”^{67C. Geertz, The Interpretation, p. 5.}

2. Twitch and wink

To illustrate the idea of “thick description,” Geertz borrows the story of wink from Gilbert Ryle.^{68C. Geertz, The Interpretation, pp. 6-7.} Imagine two boys quickly contracting the eyelids of their right eyes. One is an involuntary twitch, but the other is an intended action signalling a conspiracy to a friend. For the actions alone, both movements are identical: each of the boys contracts the eyelids of his right eye once and sharply. We cannot tell the difference between them. In the Western culture, twitching may or may not carry a message. A twitch differs greatly from a wink as anybody who has mistaken the former for the latter would know. Twitch is an involuntary action with no significance, but wink is a body language that communicates a message of intimacy. The latter has to be conducted in a culturally defined manner by someone to a specific person for a special message with the intention to evade the notice of the rest of the company. Twitch is only a physical movement, while wink is double: it is a physical movement plus meaning. Though indistinguishable from an “objective” scientific observational point of view, a wink is as real as a twitch.

The story does not end here. A third boy intending to amuse his peers parodies the second boy’s wink.^{69Geertz says that the third boy parodies the first boy’s wink. Since the first boy does not wink, it would be more reasonable to say that the third boy parodies the second. See C. Geertz, The Interpretation, p. 6.} No doubt, he will do it emphatically so that his wink appears to be over-obvious, and perhaps it comes along with a grimace. The movement of contracting his right eyelids is the same as a twitch or a wink. Nevertheless, his is neither a twitch nor a wink but a parody. A parody is a ludicrous imitation which also carries a message, not that of conspiratorial signal but that of ridicule. This too, is performed according to a socially established code. Suppose our small comedian is really serious about his trick and practises it at home before a mirror. This time, he is neither twitching, nor winking, nor parodying, but rehearsing, though to a strict behaviourist, a rehearsal is the same as a twitch, a wink, or a parody. To make our story a little more complicated, suppose the second boy is actually not winking at someone but pretending to wink in order to mislead others into believing that there is a conspiracy going on, which in fact there is none. Thus, besides twitch, wink, parody, and rehearsal, we now have to add a new category, fake-wink, to our list. This hierarchy of meanings associated with the action of contracting the right eyelids is crucial to our perception, interpretation, and reproduction of the movement. Without it, we can hardly understand the action, or we may not even notice it at all.

I still remember a cultural shock of wink when I first set foot on Belgium several years ago. Among the few things a resident alien in Belgium has to do as soon as possible is to open a local bank account. So I went to the main street of the town where I lived, and looked for a bank. Finally I made up my mind, and I entered a bank. There I saw a handsome young man with square face and broad shoulders sitting behind the counter. Unfamiliar with the setting and at the same time searching for help, I stared at him, not quite certain if he was the right person to talk to. Conscious of somebody’s attention, the bank clerk lifted up his head. Our eyes met and lingered for a short moment. Suddenly the man quickly blinked his right eyelids. Perceiving that this was unmistakably a signal to me, I was immediately shocked by such unexpected action, for as far as I could recall, I had never received a wink in a public place, and especially by an unknown handsome and strong man. While I was pondering what this signal could possibly mean, I managed to conceal my astonishment, and proceeded to the usual banking routine. Now and then I am still wondering the exact meaning of this wink. It is true that the people in my part of the world also wink by contracting the eyelids of their eyes to convey a conspiracy. But somehow we regard this as an indecent act, and attach a strong disapproval to it. Parents will teach their children, if they happen to have picked up such habit, not to do it as soon as the children can reason. Devoid of a developed structure of meanings and lacking such social experience as I was, I was unsure of the exact meaning of the wink, though I could guess that it was a welcoming gesture. Now, with the aid of Ryle’s hierarchy of meanings, I guess that the possible meaning of the clerk’s blink is something between a wink and a fake-wink. Since the clerk and I have never met before, the signal cannot be a conspiracy. It too cannot be counted as a fake-wink without a company of cronies. The gesture is a fake-wink because there is no conspiracy whatsoever between us, and it is also not a fake-wink intending to mislead others. The closest option is that it fakes an intimate relationship, which does not exist between strangers. Hence, it is best to call it a greeting-wink.

A simple social action may convey different social meanings. Because of its complexities, we can give a thin description, or more appropriately, a thick description of it. For instance, a thin description of a rehearser is that he rapidly contracts his right eyelids. This description is indistinguishable from a twitcher, a winker, a parodist, or a fake-winker. Whereas a thick description of what he is doing would be that he is “practicing a burlesque of a friend faking a wink to deceive an innocent into thinking a conspiracy is in motion.”^{70G. Ryle, quoted in C. Geertz, The Interpretation of Cultures, p. 7.} Such thick description is possible because there exists, in Geertz’s own words, “a stratified hierarchy of meaningful structures in terms of which twitches, winks, fake-winks, parodies, rehearsals of parodies are produced, perceived, and interpreted, and without which they would not … in fact exist, no matter what anyone did or didn’t do with his eyelids.”^{71C. Geertz, The Interpretation, p. 7.}

3. Sheep raid

If Ryle’s story seems to be too artificial, Geertz provides us with a typical anthropological example. Since any form of reproduction will distort the significance of the record, I have no alternative but to quote in full this not-so-short narrative:^{72C. Geertz, The Interpretation, pp. 7-9.}