Varieties of Civil Religion by Robert Bellah and Phillip E. Hammond
Robert N. Bellah is Ford Professor of Sociology and Comparative Studies at the University of California, Berkeley. Phillip E. Hammond is Professor of Religious Studies and Sociology, University of California, Santa Barbara. Copyright by Robert N. Bellah and Phillip E. Hammond.
Chapter 5: The Rudimentary Forms of Civil Religion
(Chapter 5 is a somewhat revised version of Religious Pluralism, Legal Development, and Societal Complexity: The Rudimentary Forms of Civil Religion, "Journal for the Scientific Study of Religion, 13 (June, 1974). pp. 177-189. I want to acknowledge the important research assistance of William A. Cole in developing this chapter.)
Not ever nation-state has a civi1 religion. Whatever the proclivity of enduring groups to project or represent themselves symbolically, not all succeed in doing so in transcendental terms. Only some make sacred their civil rituals; only some create theology out of their political myths. A several-faceted issue is thus raised: What distinguishes those nation-states with a civil religion from those without? Why might a civil religion develop in the first place? What needs, felt by whom, are met by a civil religion? And why are such needs not met by other (prior, noncivil) religions? To answer such questions is to advance our understanding of civil religion. At the very least the question of whether various ceremonies and ideologies should be regarded as rituals and theologies is clarified if a plausible reason can be given as to why they can be.
In this chapter I attempt only to illuminate some of the conditions that could give rise to civil religions, thus making plausible at least their existence. Briefly put, my argument is (1) the condition of religious pluralism prevents any one religion from being used by all people as a source of generalized meaning, but (2) people nevertheless need to invest their activity with meaning, especially when that activity brings together persons of diverse religious background. Therefore (3) a substitute meaning system is sought and, if found, the people whose activities have been facilitated by it will tend to exalt it.
I suggest a universalistic legal system can be such a substitute meaning system, though, as other chapters in this volume suggest, more than a universalistic legal system is surely involved.
Admittedly, the theory here is abstract. It suffers from the fault of so much of the social scientific study of religion, which in its efforts to show the sociological importance of religion has been forced to deal with global, far-reaching, and historical variables. In so doing, such study has frequently sacrificed the closely specified and carefully measured variables. This one is no exception. And yet I use quantitative data from a number of societies and may for that reason claim to avoid some of the pitfalls of mere theoretical analysis. Nevertheless the analysis makes several broad assumptions: regarding the time order of variables, regarding the meaning of "social complexity" (soon to be specified) found among present-day developing societies, and regarding the equating of contemporary nation-states with what in theory are "societies."
I previously identified a second caveat. The overarching theory states that a generalized legal system will be elevated to civil religious status insofar as it facilitates the interaction of religiously diverse people. In fact, however, the information (and thus the narrower theory examined here) pertains only to the first of the two implied assertions -- that a generalized legal system facilitates the interaction of religiously diverse people. The second assertion -- that such a legal system will then become sacrosanct -- remains unexamined by data in this chapter. It is in this sense that our understanding of civil religions will be advanced even if only a plausible reason for their existence can be given. With these two qualifications, then, I present and test the narrower of the two theories.
John Courtney Murray put the first part of this narrower theory as well as it can be put: Civil discourse or civil unity is complicated and laborious enough, but under conditions of religious pluralism it is more so because discussion of concrete affairs goes on in abstract terms -- in "realms of some theoretical generality" -- and pluralism creates different sets of these realms. Discourse becomes "incommensurable" and confused.1 Such confusion is easy to imagine since some taken-for-granteds get called into question. Would-be partners cease to be sure of each other’s commitments and thus trustworthiness. The relative strength of loyalties to kin, neighborhood, status group, and such are possibly challenged by one’s loyalty to one or another (or no) religious group. To the simple categories of believer and heretic get added the various ways to be a nonbeliever. Religious pluralism, to put it simply, complicates matters by ripping the once intact sacred canopy. Interaction, we might suppose, thereby is inhibited.
For this reason, and perhaps contrary to one’s initial guess, we can expect greater pluralism at the religious level to be associated with lesser amounts of social development (or "societal complexity"). At least we can expect no positive relationship between religious pluralism and societal complexity. This is the first proposition.
It is important to realize the meaning attached to "societal complexity" in this proposition, however. Since religious pluralism itself represents one kind of complexity, it should be clear that kind of complexity is excluded from this variable. The reference in the present complexity variable is rather to those "secular" activities that, if religious considerations can be ignored or overcome, may proceed unimpeded but into which religious considerations are likely to intrude unless special efforts are made. The type case is the market transaction wherein two parties can strike a bargain suitable to both if particularistic factors such as religion do not interfere. But the whole range of contacts generally regarded as "socioeconomic" also apply here. I recognize that the usual formulation is somewhat different, that is, as secular activities develop, such particularistic factors as religion diminish. In either view particularistic factors are held in abeyance to some degree; I am arguing that the condition of religious pluralism makes it more important for people to hold religious consideration in abeyance. What little difference exists between these two views revolves around whether religion’s diminution (secularization) occurs automatically with pluralization or instead will occur only if the "holding in abeyance" is helped along (as by a changing legal order).
In any event, though religious heterogeneity may very well appear alongside or as part of general social complexity, I expect a plurality of meaning systems or "sacred canopies" will inhibit development of those social interactions that typically cross religious lines, that is, secular or socioeconomic activities.
The second part of the theory deals with situations wherein secular activities have been or are in a position to be inhibited by religious pluralism. It states that the (threat of) impasse may be lessened by the existence of legal institutions that foster universalistic norms. In effect, this requires of the law some procedures to enable diverse parties to interact in spite of otherwise inhibiting characteristics. For example, two persons who are prohibited by religious scruples from transactions with each other may find it possible, through law, to use third parties (for example, the impersonal, universalistic procedures of a bank). Or, as another example, two groups who define themselves as "believers" and "heretics" or as "faithful" and "infidel" may nevertheless be convinced they will receive fair treatment in the courts and thus allow themselves to buy and sell, enter into contract, and so forth.
Max Weber’s distinction between substantive and formal rationality sets the terms for this part of the theory:
In general terms . . . the rationality of ecclesiastical hierarchies as well as of patrimonial sovereigns is substantive in character, so that their aim is not that of achieving the highest degree of formal juridical precision which would maximize the chances for the correct prediction of legal consequences. . . . The aim is rather to find a type of law which is most appropriate to the expediential and ethical goals of the authorities in question. . . . Yet in the course . . . of rationalization of legal thinking on the one hand and of the forms of social relationships on the other, the most diverse consequences could emerge from the non-juridical components of a legal doctrine of priestly make. One of these possible consequences was the separation of fas, the religious command, from jus, the established law for the settlement of such human conflicts which had no religious relevance. In this situation, it was possible for jus to pass through an independent course of development into a rational and formal legal system. . . . Juridical formalism enables the legal system to operate like a technically rational machine. Thus it guarantees to individuals and groups within the system a relative maximum of freedom, and greatly increases for them the possibility of predicting the legal consequences of their actions. Procedure becomes a specific type of pacified contest, bound to fixed and inviolable "rules of the game."2
Weber, as usual, is speaking here of ideal-typical forms of legal systems. Nevertheless, he sees the strategic importance played by the shift from particularistic to universalistic standards -- from what he calls substantive to formal rationality. Related to this shift is the obvious fact that religious particularism is not the only particularism. Thus Weber refers to "patrimonial" as well as "ecclesiastical" sovereigns.3 The proper interpretation is to recognize that whatever loyalties and commitments people may have -- to kin, guild, caste, and so on -- are likely, when institutionalized, to be expressed religiously. It is in this sense that the opening paragraphs of this chapter refer to meaning systems or sacred canopies as religion. In a totemistic, endogamous, clan society, for example, the "kin" loyalty and "religious" loyalty would be identical. But in any society kin loyalty is likely to find expression in religious language, rituals, and so forth.
The proposition in this second half of the theory can be stated as follows, therefore: The higher the level of religious pluralism in a society, the more will societal complexity depend upon, and thus be associated with, the presence of a universalistic legal system.
This second proposition appears almost to represent common knowledge. A rationalized, generalized, predictable legal system will foster social interaction and thus societal complexity.4 The claim here is more than this, however; it is that, while a rationalized legal system may facilitate societal complexity in any situation, it may be even more helpful for people who are otherwise religiously different. Here is the argument locating this chapter in the context outlined in the opening paragraphs: The legal order in some significant measure becomes a substitute for the religious order -- the order that supplies meaning -- and thus sets the stage for the emergence of civil religion. 5
Granted, the outcome may not be a sharply defined "civil religion" -- one that could be universally recognized as such -- but at the least it can be anticipated that some kind of "political religion" will be more likely to emerge in societies where legal structures take on meaning-bestowal qualities.6 Obviously such political religion can emerge in "totalitarian" as well as "democratic" societies, but in either setting it will be the law and not mere coercion that facilitates social development. This is Kozolchyk’s point regarding "fairness" in the law. If law is to facilitate societal complexity, it must instill trust, which means it cannot deviate far from the will of those governed. Nation-states of considerable as well as meager development are variously located along the pluralism axis. I shall show that regardless of political style, the societal complexity of religiously plural societies is helped along more by universalistic legal systems than is the complexity in religiously homogeneous societies.
Testing the Theory
In an ambitious project to assess the correctness of Talcott Parsons’ theory of evolutionary universals, Gary Buck accumulated masses of data for 115 contemporary nation-states from every part of the world.7 He developed elaborate indices (as of 1960 wherever possible) of the ten variables Parsons discussed: (1) communication, (2) kinship organization, (3) religion, (4) technology, (5) stratification, (6) cultural legitimation, (7) bureaucratic organization, (8) money and market complex, (9) generalized universalistic norms, and (10) democratic association.8 Information was taken from such sources as the United Nations Statistical Yearbook, the Yearbook of Labor Statistics, and UNESCO’s World Survey of Education. One index (kinship organization) was measured by only three indicators, but the others were more extensively measured, some by over a dozen separate pieces of information. I have borrowed five of Buck’s ten indices in order to measure two of the three variables.
Societal complexity refers to the degree of complexity in "secular" activities, of which the market transaction is the prime example. Perhaps socioeconomic development is a suitable alternative phrase. Four of Buck’s indices -- communication, technology, bureaucratic organization, and money and market complex -- seem to measure this variable very well. Altogether they entail fifty indicators (making their listing here too space consuming), but one indicator from each of the four indices will give an idea of their nature (see Table 2).
Table 2. Four Indices of Societal Complexity and Sample Indicators of Each
For the index of: An illustrative indicator is:
Communication Motor vehicles in use
While Buck found it necessary to keep these four indices separate -- if he was to test the sequential nature Parsons claimed for them -- I have no such requirement. Therefore since each measures a different but important aspect of "complexity," I merge all four indices. Each had a range of scale scores from one to seven, which means the composite index used here ranges from four to twenty-eight. For simplicity I have subtracted three from every unit, making the range from one to twenty-five. Countries were found at every score level. It is important to remember that all fifty indicators, despite their differences, refer to "secular" activities, that is to say, activities that can be engaged in by persons whether or not they "like" or "know" or "frequently see" each other. They measure how easily "strangers" are brought into interaction.
Buck also has an index he called, following Parsons’s terminology, "generalized universalistic norms." A legal system is generalized, Parsons says, to the degree it is "an integrated system of universalistic norms, applicable to the society as a whole rather than to a few functional or segmental sectors, highly generalized in terms of principles and standards, and relatively independent of both the religious agencies that legitimize the normative order of the society and vested interest groups in the operative sector, particularly in government."9 That is, a generalized legal system facilitates social relations. Such a process can be regarded as legal development (in the sense I have been using that concept here), and I therefore use the Buck index of generalized universalistic norms as its measure. He, in turn, built his index on eight pieces of information about each society contained in one of two sources.10
Seven of these pieces of information deal with the degree to which the law is used to "eliminate" opposition rather that to facilitate concerted action by disparate groups. Thus from Banks and Textor, Buck recorded the degree of freedom opposition groups enjoy, and from Feierabend and Feierabend he recorded (for the period 1948-1962) the amount of "repressive action against specific groups. "arrests of significant persons for political reasons," "arrests of insignificant persons," "significant changes of general laws," "politically motivated killing by the government," and the frequency with which the constitution is suspended and martial law is declared. The eighth indicator, from Banks and Textor, reflects a characterization (with the help of a legal scholar) of each nation’s legal system as ranging from "indigenous’’ (indicating low development) to "common law" (reflecting high development). While this final indicator can be criticized as being subjective (at least on the part of the legal scholar, if not on the part of those who appended the labels on each country’s legal system), it is worth pointing out that as only one of eight indicators, this one influences the index very little. Were it to be eliminated altogether, the relative scores on legal development in 115 nations would change hardly at all.
It may well be asked whether an index mainly of negative instances fairly measures the degree of something positively stated. Political reprisals and closing off opposition may reflect the absence of a widespread legal order, it might be argued, but is the nonappearance of such phenomena to be taken as indicating its presence? Two responses might be made. (1) The expedient one is that the Buck measure is available, and no preferable measure based on recorded data is known. (For example, information on the number of lawyers, law schools, and so forth in a society could be a substitute measure of legal development, but in that case the relationship with societal complexity would be well-nigh tautological.) (2) Moreover, I am not so much interested in the existence of law and lawyers as I am their effectiveness in subjecting persons’ behavior -- without coercion -- to legal rules. The record of legal repression seems, then, a reasonable measure of legal development, much repression indicating little legal development, and little repression more legal development.
The third index measures the degree of religious pluralism with information from various sources.11 Because a certain amount of arbitrariness was involved in this procedure, I want to be certain it is explicit. First, it became apparent that the highest levels of pluralism as measured by any available method are not found in those societies regarded today as very pluralistic (the Western industrialized democracies). To be sure, in the scoring scheme the United States would be quite plural -- with Catholics, Jews, and Protestants all in considerable number. But it is surpassed by a number of nations in Asia, Africa, and the Near East, where there exist not only Protestant and Roman Catholic or Christian and Jew but maybe Islam, Hindu, Animist, Shinto, and Buddhist as well. I decided therefore that though one could enumerate many sectarian expressions of Protestantism, the same information did not exist for Buddhism, Islam, Animism, and so forth; so I counted only the major groupings just listed (plus Eastern Orthodox and Confucianism), giving one point for the presence in any society of each of these ten religious categories.
Second, since even the most religiously uniform society is going to have a smattering of "deviants," it was decided that a religion must be represented by at least 2 percent of the population if it was to get counted. Without any firm evidence for selecting that cutoff point, I reasoned that clashing definitions of ultimate reality are likely to clash only when supported by at least a handful of opposing "believers." Moreover, it can be assumed the data are more imprecise on religious minorities; with a minimum of 2 percent the chances are good they are at least noticed.
The upshot, however, as already indicated, was that those societies generally regarded as most plural -- even to a point of the widespread exercise of religious freedom -- were categorized along with newly developing societies where tribal religions compete with the legacies of numerous missionaries. It was as if one were comparing a present-day African or Asian nation with the United States of 1825 (with Protestants and some Jews entrenched on the East Coast, Catholics entrenched in the Southwest, and "Animists" spread out in between). But in fact the measures of legal development and societal complexity are contemporary, thus making that comparison impossible or at least illegitimate.
The only solution seemed to be to eliminate those nations that had already "developed." Granted, this move has an arbitrariness about it, but the effect is to subject the theory to an immensely more stringent test. If societies like the United Kingdom, Canada, Sweden, Luxembourg, France, United States, and New Zealand are known in advance to be (1) highly complex, (2) highly developed legally, and (3) religiously plural, it is not much of a test to determine if legal development and societal complexity are strongly related in religiously plural societies. But what of Burma, Sudan, North Vietnam, Bulgaria, Cameroon, Trinidad, and Ceylon? It seemed reasonable to treat all these latter nations as developing rather than developed and to inquire about their legal and religious circumstances.
Consequently, eliminated are the 23 nations (of the original 115) that scored twenty or higher on the societal complexity index (these are all so-called developed nations). The remaining 92 nations ranged in scores on religious pluralism from one (low) to four (high).
Another obvious weakness characterizes the religious pluralism index; it counts only orthodox, traditional, universally recognized religions. These may, but probably do not, exhaust the variety of "meaning systems" in a society. Thus to take an example, Belgium is nominally all Roman Catholic; yet the Walloon-Flemish distinction is known to pervade all social behavior. India, to take a different kind of example, contains sizable adherents of several "religions," but the adherents of any one religion are themselves divided by language, ethnic, and regional loyalties of great depth. The pluralism index measures none of these religious (but not orthodox religious) differences, though it can be hoped that errors are distributed randomly. To the extent they are, they mute the corelationships that do show up.
It is also worth noting that, while these variables were measured with data from around 1960, the relationships between variables are not necessarily time-bound. Particular scores for particular nation-states change through time, but the tendency for one characteristic of a nation to associate with another characteristic changes only very slowly if at all. A pattern of relationships found at one time, therefore, is likely to be the pattern found at other times as well.
The first part of the theory states that religious pluralism and societal complexity are inversely related, the reasoning being that while religious pluralism is itself a kind of complexity, it has an inhibiting influence on those "secular" interactions necessary for the kind of development I have called societal complexity. Table 3 shows the findings on this first question.
A statistical fundamentalist would insist no guaranteed negative relationship exists in Table 3 because the number of cases is small and the differences are slight indeed. So be it. The argument is only that, contrary to idle speculation perhaps, religious pluralism is not just another aspect of social differentiation but is, if anything, inimical to societal complexity. The data seem to support this part of the theory.
Table 3. Religious Pluralism and Societal Complexity
Average Societal Complexity Scores 10.8 10.0 9.8 8.9
The second part of the theory is more difficult to test. It is not enough to show that legal development facilitates societal complexity, though Table 4 suggests it does.
Table 4. Legal Development and Societal Complexity
The less than perfect relationship reminds us societal complexity is an enormously involved phenomenon and therefore subject to a multitude of forces in addition to legal development. Had the relationship in Table 4 been reversed, I would have had to revise my thinking. But in a major way the finding of Table 4 is just what is expected: Nations with the least legal development are least complex; high legal development is associated with high complexity. Bear in mind, too, that putting back in the twenty-three nations already eliminated would, because they are both legally developed and societally complex, greatly sharpen the relationship in Table 4.
The evidence so far, then, supports the proposition that a universalistic legal system facilitates societal complexity. The theory states more than this, however. It states a universalistic legal system will be more facilitating of societal complexity the higher the level of religious pluralism. Operationally, this amounts to the prediction that with each increase in religious pluralism the association between societal complexity and legal development will also increase. Table 5 contains the evidence.
Table 5. Religious Pluralism, Legal Development, and Societal
It would appear that the theory finds considerable support (recognizing, of course, that mere association does not necessarily signify the causal path the theory claims). One possible criticism comes immediately to mind, however: Might the increasing association between legal development and societal complexity be artifactual? Might earlier colonial regimes, for example, have left both the makings of an industrial (complex) society and a legal system somewhat more developed than what they found? Table 6 suggests not. The four societies at once most plural, most complex, and most developed legally (Bulgaria, Malaysia, the Philippines, and Trinidad) are drawn from widely divergent areas of the world. They represent different ways to be religiously plural, but so also are their legal development scores arrived at in different ways. Common law traditions are found in the background of one or two, and modern-day colonialism existed in two or three, but neither attribute characterizes all four. In other words these four nations are not closely aligned historically, geographically, politically, legally, or religiously. Similar statements can be made of the other eleven groups of societies listed in Table 6. As with all research using static data, of course, such arguments do not "prove" the theory being advanced, but they do make it plausible. To reiterate, legal development does appear to make more of a difference in the complexity of religiously plural societies.
Implications for Civil Religion (Ted: 2 more pgs. of tables)
The question then arises as to what civil religious implications there may be in the role played by legal institutions in religiously
Low Societal complexity: Medium Societal Complexity: High Societal Complexity:
(scores of 1 -- 7) (scores of 8 -- 12) (scores of 13 -- 19)
Very High Religious
Pluralism Burma 4 Camaroun 2 Bulgaria 5
Medium High Religious
Pluralism Burundi 2 Albania 3 Hungary 3
Medium Low Religious
Pluralism Cambodia 5 Indonesia 2 Brazil 4
Very Low Religious
Pluralism Afghanistan 5 Bolivia 3 Algeria 2
___________________________________________________________ Turkey 3
136 Varieties of Civil Religion
plural society. It can be ventured that if people experience conflict, they attempt to resolve it. If the conflict itself includes a clash of "resolution systems" (meaning systems or "sacred canopies") -- if, in other words, A and B cannot agree because neither accepts the other’s basis for agreement -- then some other resolution will be sought. It is here that the law may be turned to, especially to the degree it is ‘ universalistic," thus overriding whatever parochial conditions have stood in the way. Insofar as the law responds satisfactorily -- its procedures respected, its orders obeyed, its sanctions upheld -- a legal "order" can be said to exist. But if in addition the agencies of this legal order use the language and imagery of purpose and destiny, if they not only resolve differences but also justify their resolutions, it is easy to see how something identifiable as civil religion could emerge. There would exist already a cadre of "clergy," a set of "rituals," places for "worship," and a number of directives for behavior. Add to these a "theology" -- an ideology of purpose and destiny or theodicy -- and a civil religion may be close behind. Whether Bulgaria. Malaysia, the Philippines, and Trinidad are further along in the development of civil religions than other societies in Table 6 I do not know. But it seems reasonable to suggest, on the basis of the theory here, that whatever level of civil religion they may possess at this time will, more so in their cases, be reflected by their legal systems. At least such a perspective directs us to further research into the issue of civil religion.
1. John Courtney Murray, We Hold These Truths (Garden City, N.Y.: Doubleday, 1964), pp. 27-29. See also Talcott Parsons, "Religion in a Modern Pluralistic Society,’ Review of Religious Research, 7 (1966), pp.125-146 Peter Berger, The Sacred Canopy (Garden City, N.Y.: Doubleday, 1967), pp. 48-49, 135 - 138.
2. Max Weber, On Law in Economy and Society, trans. Edward Shils and Max Rheinstein (New York: Simon and Schuster, 1967), pp. 225-227.
3. And Cowan speaks of "legal pluralism." "A second great complex of problems to which legal pluralism gives rise is that which has generally conic to be known under the heading of inter-personal conflict of laws, internal conflict of laws, or inter-gentile law. . . . How should one regulate disputes between members of different ethnic or religious groups living within the same political unit under different laws?" Denis V. Cowan, "African Legal Studies," in Hans W. Baade, ed.. African Law: New Law for New Nations (Dobbs Ferry. N.Y.: Oceana Publications, 1963), p. 18.
4. Compare with Kozolchyk’s statement: "The main function of a legal system in bringing about economic development is to instill trust in legal institutions." Boris Kozolchyk. "Commercial Law Recodification and Economic Development in Latin America," Lawyer of the Americas, 4 (1972) p. 2.
5. Thus Kozolchyk, ibid., whose research is based on Latin American societies, goes on, "And this trust results not only from the efficient operation of legal institutions, that is to say, when these institutions perform predictably and in the least costly manner, but also from their fairness." The point, of course, is that "fairness" automatically spills over into the religious sphere.
6. See David Apter, "Political Religion in New Nations," in Clifford Geertz, ed., Old Societies and New States (New York: Free Press, 1963).
7. Talcott Parsons, "Evolutionary Universals in Society," American Sociological Review, 29 (1964), pp. 339-357.
8. Gary Buck. A Quantitative Analysis of Modernization, Office of Research Analyses, Holloman Air Force Base, New Mexico, 1969. Buck found by scalogram analysis, incidentally, that the evolutionary sequence Parsons suggests (that is, the order in which I listed the ten variables) is strongly supported by the data on these 115 societies. I am indebted to Buck for his cooperation in this research.
9. Parsons. "Evolutionary Universals," p. 351.
10. Arthur S. Banks and R. B. Textor, A Cross-Polity Survey (Cambridge, Mass.: MIT Press, 1963); I. K. Feierabend and R. L. Feierabend, "Aggressive Behaviors Within Polities, 1948-1962: A Cross-National Study," Journal of Conflict Resolution, 10 (1966), pp. 249-271.
11. S. W. Coxill and Sir Kenneth Grabb, The World Christian Handbook (London: Lutterworth, 1968); Elisa Daggs, All Africa (New York: Hastings House Publications,1970); Deadline Data on World Affairs (Greenwich, Conn.: D.M.S., Inc.,1971); Facts on File (New York: Facts on File, Inc., 1971);John S. Mbiti, African Religions and Philosophy (New York: Praeger, 1969).