Some Implications of Niklas Luhmann's Equivalence-Functionalism
ソウル大学で開催された、東大とソウル大の社会学研究室の合同シンポジウムで報告しました。以下は、その報告原稿の全文です。（PDFはこちら）
I read the following paper at the graduate student session of Seul National University - University of Tokyo Joint Sociology Forum (3. Nov. 2006, at Seoul National University).
Some Implications of Niklas Luhmann's Equivalence-Functionalism
Introduction
Niklas Luhmann is known today as a system theorist in sociology. Indeed he is one of the greatest and most famous sociologists of the 20th century and wrote many books and articles on social systems theory. His contributions to (social) systems theory and, as subfields of this, theories of interaction, organization, society (including economy, politics, law, art, religion, mass media, etc.), the general theory of communication and action, theories of communication media (power, love, truth, money, etc.) and so on, are so broad and complicated that nobody has succeeded evaluate them in their totality.
This difficulty of evaluation occurs also from his own saying that his theory is circularly constructed and not axiomatically or linearly. When a theory is circularly constructed, a critic or a evaluator cannot know where he should begin to criticize or evaluate the theory. Of course it is important that he wrote too many books to read in some years (it is said that he wrote 20,000 pages in total! ― 70 books and nearly 400 articles, and what is more for us (for me?), almost all of them are written in German language).
In these situations, it is necessary for us (who are interested in Luhmann’s theory and sociological theory in general) to investigate whether his theory actually has some circular construction, and if so, what kind of circularity it is. If circularity is only in an appearance and there is some basis on which his theory is linearly constructed, we could focus on that basis and use it as a starting-point of evaluation. Even if circularity is genuine one, it is still possible to investigate the basis of this circularity itself, and if such a basis be discovered, to use it as a starting-point. When the circularity is genuine and groundless, there is no reason for us to read Luhmann any more. In any case, what is required is not an uncritical introduction or application, but a critical investigation.
Based on these considerations, we distinguish here between systems theory and functional method. In this paper, it is not our concern whether Luhmann’s systems theory is circularly constructed or not. For in any case, in our view, his attempt of theory construction itself is to be thought as based on a functional method. Indeed in the earliest period of his academic career, it could be said, his main aim was to free functionalism from systems theory. To clarify the significance of this liberation, we need first to take a general view of the relation of functionalism to systems theory before Luhmann. And then, we can see, in comparison to this, how Luhmann liberated functionalism from systems theory and how he gave functional analysis a new meaning.
Functionalism before Luhmann ― Two Types
It could be generally said that functionalism in sociology is the position that requires that the concept of function is necessary and plays some essential role in sociology. But according to the way in which functional concept plays this role, that is, according to the aim of using the concept, it is possible to distinguish some different types of sociological functionalism. In this paper, we distinguish following three types. In this section, we concentrate two types among them.
(1) Functionalism as a Substantive Theory or Explanation-Oriented Functionalism. This is the type of functionalism that uses the concept of function as describing a property or a principle of the social world. In this position, it is said that, in the society (or more generally, in the social system), there is a property which could be named function, and that any theory which claims that there is no such property is wrong because of making such a claim. In other words, this type of functionalism introduces a variable of “function” into a system of variables which describe the social world (claiming that the variable of function is irreducible to any other variables in the system). In this position, the aim of using the concept is the explanation of social phenomena. Its typical form of proposition is “There is A, because it fulfills the function F to B.” In this proposition, A is an explained variable, and F (to B) is an explaining variable. The condition on which this proposition is right is that the social world is so constituted. The examples of this position are found among the social anthropologists like Radcliffe-Brown [1952].
Since 1950s, especially in the United States, the writers who criticize functionalism began to deny that functionalism is a special position irreducible to any other position. This is a criticism upon the special status of the concept of function which is irreducible to any other explaining concept, particularly causal concepts. Critics said that functional explanation is not a special method of explanation which stands side by side with causal explanation, but a special expression or a special stage of development of causal explanation. The core of their criticism is the thesis that what is called function is an abbreviation of more complex causal relations and can be reduced to these relations, and that, unless all of these ‘true’ explaining factor be found, the explanations using the concept of function would be insufficient. So Bredemeier [1955] required that usual functional explanations should be complemented by motivation analyses (for him, motive is the only immediate cause of any human action). Or, on a similar but more general ground, Kingsley Davis [1959] said that functionalism is another name of sociology and social anthropology in general. After all, their criticism was a denial of function as a special property, and therefore, a denial of the specialness of functional method of explanation.
(2) Functionalism as a Method of Theory Construction. This is Parsons’ position (Parsons [1951] etc.). He thought that the ideal and final state of theory development consists in a system of equations which contain all the relevant factors as variables. Since his theory is the action theory, this system of equations constitutes the action system. But in the present conditions of sociological knowledge, it is impossible to determine the way in which each (variation of a) variable influences all the other variables and to construct a complete theoretical system. To cope with this difficulty, Parsons proposed a “structural-functional” way of theory construction.
Because of our limited possibilities of observation in comparison to natural sciences, many items which constitute the social world remain invariable and constant through all the observations. We cannot conceive these items as variables, we can only conceive them as constants. Parsons names the set of these constants as structure, and the items which can be set as variables as processes. And he introduces the concept of function to express the relation of these processes to structure.
All the details must be omitted here, but what we need to see is the fact that Parsons’ structural-functionalist terminology expresses a method of theory construction under some limited and difficult conditions, and its validity is independent from any particular state of the social world. Therefore, this kind of functionalism is not substantive, but formal. And at the ideal and final stage of theory development (Parsons calls this stage a dynamic theory), functionalism with the use of the concept of function should disappear (1).
- (1) "As dynamic knowledge is extended the independent explanatory significance of structural categories evaporates" (Parsons [1951: 21]).
Not only critics of functionalism, but also functionalists themselves were not clearly aware of this difference. But this is an essential one, based on the difference of position in relation to the explaining force of function. Substantive functionalism thinks that it has such a force. Parsonian functionalism thinks that it doesn’t (2).
- (2) In Parsonian (structural) functionalism it is a structure that has such a force. Moreover, this is a tentative and not a genuine explaining force.
But by logical analyses of the form of (substantive) functionalist explanation performed by Hempel [1959] and others, it proved that it is insufficient as a logical explanation. They pointed out that even if a social system has a set of prerequisites which is necessary for its existence, and if this social system exists now, it then follows only that there exists some item which fulfils one or more of the prerequisites. So even if we know that an item fulfils that prerequisite(s), that is, this item fulfils some necessary function, it is impossible to deduce the necessity of existence of this item from this knowledge alone. So this kind of functional explanation is always insufficient.
| Functionalism as | ||
| a substantive theory | a method of theory construction | |
| Purpose | Explanation | Explanation |
| Form of theory | Systems of variables | Systems of variables |
| Position of function | One of the explaining variables | A tentative relation of variables to constants |
| Difficulty | Insufficient as a logical explanation | Not functionalism in the true sense |
Luhmann's Equivalence-Functionalism: Its Implications in Three Directions
In his debut article "The Concept of Function in Administrative Science” (1958), Luhmann proposed a new definition of “function.” According to him, “function of function” is “to regulate the interchange of possibilities in reference to a presupposed perspective,” and “based on this interpretation, the concept of function can be defined for our purpose as the relationship of x (a so-called variable) to y, when y serves as a point of view for statement of equivalence of x (that is, as a point of view of variation of the variable x)” (Luhmann [1958: 99]). In some later articles (“Function and Causality” (Luhmann [1962]), “Functional Method and System Theory” (Luhmann [1964]), etc.), he speaks also of “to compare possibilities” instead of “to regulate interchange of possibilities” in the same sense. He calls this type of functionalism “equivalence-functionalism.”
As we have seen, Hempel and others have criticized (substantive) functionalism on the basis of the possibility of existence of functional equivalents. Luhmann accepts their criticisms as valid, and uses this possibility as the defining property of the concept of function. Our task is to investigate the question: What meaning does this redefinition have?
An important point is that Luhmann gave up any explanation as a goal of functional analysis. In his view, an only legitimate goal of functional analysis is comparison between functional equivalents. By this reorientation, the task and theoretical premises became lighter than before. Explanations are only possible in particular cases, when the general theory which includes those cases as parts has been more or less completed. So explanation-oriented functionalists have tried first of all to construct a functional theory of social world and have been denied the possibility of such a theory construction. Contrary to this, comparison itself requires no such premiseful theory. While explanation-oriented functional analysis depends on the plausibility of the theory, comparison-oriented one does not.
This reconstruction of functional analysis has important implications in three directions: methodological, theoretical and practical.
First, methodologically, functional analysis of this kind can be performed in the very simple framework. In principle, we need only a problem and some possibilities which solve the problem. The set of solution possibilities need not be determined, because the discovery of alternative solutions is itself an scientific insight. Unlike the case of explanation, comparison does not require the theory which can make comparison plausible. Of course, for more fruitful analysis, it may be required to combine several comparisons. For this purpose, a theory that unites several problems is needed. And Luhmann’s systems theory is of this kind, that is, a problem-providing type (3).
- (3) This problem-providing character can be seen even at the later (that is "autopoietic") period of his theory development. "In our remarks so far we have given precedence to formulating problems and carefully avoided a structural determinations of the theory itself. We have not presented any "models" to avoid the appearance of a theory that is already on its way to determining structures. We have limited ourselves to enriching an understanding of the problems confronting systems theory. This is a consequence of the concept of self-referential systems. At the same time, it secures points of departure for functional analysis" (Luhmann [1984: 83] = [1995: 52], italics original).
Finally, from the practical point of view, this type of functional analysis opens another form of utilizability of the fruit of social-scientific research. When explanation-oriented, practically usable sociological knowledge often takes the form of prediction, because explanation and prediction have the logically same form of deduction from some premises. When we want to take a critical attitude to the status quo of our society, since the deductive process is supposed to be necessary one, if the present state of affairs is evaluated as bad, then the premises or principles of society which bring about this social state must be substituted with another premises or principles (for example, “capitalism must be substituted with communism”). This substitution on the level of principles is more or less difficult in its feasibility and its great (possibly harmful) influence to another part of society. But the comparison-oriented functional analysis proposes alternative possibilities under the present premises or principles. Substitution based on this kind of knowledge is superior to above one in its feasibility and less influence to another part. It brings more realistic possibilities of rationalization.
Remaining Tasks
These considerations above are a brief sketch. And the approach which we take here opens a lot of questions which should be kept in mind when we read Luhmann’s works. In this concluding section, we enumerate some of these questions as follows: What is the philosophical or anthropological ground of equivalence-functionalism? What kind and degree of influence does equivalence-functionalism have to Luhmann’s attempt to construct his own systems theory (which has at least superficially circular construction). Why did Luhmann prefer an attempt to build a universal (and in a sense very speculative) theory to an attempt to propose practically fruitful insights based on equivalence-functionalist analysis? What was the problem of existing theories from the functionalist perspective? and so on.
References
- Bredemeier, Harry C., 1955, “The Methodology of Functionalism,” American Sociological Review 20, pp. 173-180
- Davis, Kingsley, 1959, “The Myth of Functional Analysis as a Special Method inn Sociology and Social Anthropology,” American Sociological Review 24, pp. 757-772
- Hempel, Carl G., 1959, “The Logic of Functional Analysis,” in: Llewellyn Gross (ed.), Symposium on Sociological Theory, Row, Peterson, pp. 271-307
- Luhmann, Niklas, 1958, “Der Funktionsbegriff in der Verwaltungswissenschaft,” Verwaltungsarchiv 49, pp. 97-105
- Luhmann, Niklas, 1962, “Funktion und Kausalität,” Kölner Zeitschrift für Soziologie und Sozialpsychologie 14, pp. 617-644
- Luhmann, Niklas, 1964, “Funktionale Methode und Systemtheorie,” Soziale Welt 15, pp. 1-25
- Luhmann, Niklas, 1984, Soziale Systeme: Grundriß einer allgemeinen Theorie, Suhrkamp
- Luhmann, Niklas, 1995, Social Systems, translated by John Bednarz Jr. with Dirk Baecker, Stanford University Press
- Parsons, Talcott, 1951, The Social System, The Free Press
- Radcliffe-Brown, A. R., 1952, Structure and Function in Primitive Society, The Free Press
