Niklas Luhmann’s apples-to-oranges category error impairs his theory of society
I have read and reread Niklas Luhmann’s books for several years, initially in puzzlement, later with excitement about the possibilities offered, and finally with disappointment that is nonetheless tempered by gratitude for all that can be learned from his expertly-conceived high-risk journey to imaginary societies that probably never will nor can exist. Discovery, after all, is the essence of social science. Even in a well-intentioned, ambitious, systematic and arduous voyage across the seas which at the end discovers nothing of substance, some smaller, wonderful and indispensable discoveries will have been made along the way. I will try to balance the good against the bad. In retrospect I now see why Luhmann cannot be considered of equal stature to Weber and Parsons in the pantheon of social theorists, though the conceptual lineages between these three are clear to see. It was for the concepts and that lineage that I first read Luhmann, and this was worth while. Furthermore, since Luhmann is now the world’s most celebrated system theorist of contemporary society, he must be read (and criticized) by anyone who aspires to do systems theory.
The writings of Luhmann and Hayek, though entirely different in method, share in common their championing of the idea of self-generated societal order. Both are weakened by depending on a single assumption that is probably mistaken, and which is certainly irreconcilable with my thesis. To sum this up, the core problem in Hayek’s approach is that he separates rules of society from rules of government, when in reality there is at least one rule joining society to government. The core problem in Luhmann’s approach, I will show, is his failure to separate subsystems of society from subsystems of the state, which in reality are neither conceptually nor functionally equivalent.
My criticism of Luhmann’s theory of society and social systems (segments of which will eventually appear on this site) is undertaken from a Weberian standpoint, and in defence of methodological individualism and state-centred social order. I will suggest post-Weberian means whereby systems theory can amalgamate a theory of society’s evolution with a theory of social action, and a state system at the heart of the social system.
From this perspective, Luhmann’s theory has three main analytical components: the subsystems of society, the binary codes of subsystems, and a conceptualization of society as constituted by self-referring communications between the coded subsystems. In political terms, the element of interest is Luhmann’s emphasis on a decentred society of operationally-closed, self-reproducing and self-regulating systems with irrelevant human agency and ineffective volitional governance.
The pertinent features of Luhmann’s theory can be outlined through a critique focused on the intangibility of society’s communications, the functional non-equivalence of society’s subsystems, and logical inconsistencies in the conceptualizations of social order, the state, and organizations.
However, I want strongly to emphasise that Luhmann’s writings offer insights and innovations that can be fruitfully adapted. His conceptual inconsistency, methodological imprecision, and psychic detachment from reality can be surmounted in a new conception of volitional governance. Some of his concepts will be gyrated for this purpose. They are indispensable, just as Weber’s concepts were indispensable for Parsons, and as Parsons’s concepts were indispensable for Luhmann.
The critique must begin, though, with the idea of ‘subsystem’. The empirical referents of Luhmann’s theory of self-generated order are the subsystems of contemporary society. What are these subsystems? Potentially they are infinitely varied, a veritable melange. The ones Luhmann most often discusses are politics, law, economy, education, sport, love, art, religion, health, and science. Seemingly he persuaded himself that each of them has not only a function by which it distinguishes itself from other subsystems, but also “incorrigible operational autonomy” to reproduce and organise itself.
It should also be noted that Luhmann’s preferred term for self-generation, self-production or self-reproduction is ‘autopoiesis’, which, had it been employed literally rather than by analogy with the biological form of autopoiesis (which has weak scientific foundations), should not have presented great difficulties, since its origin lies very simply in the Greek words for self, maker, and create.
Modern society, we are urged to infer from Luhmann’s worldview, is not a society by virtue of its territorial boundary, its mode of production, its governance structures, or shared value and culture systems. Nor is society defined by interactions between individuals. Instead, society is the process by which the subsystems are built and “functionally differentiated” from one another.
The magnitude and daring of Luhmann’s attestation here deserves to be dwelt upon — “modern society is to be understood as a functionally differentiated society”, and society’s “subsystems assume universal competence for their specific function”. This theoretical claim provides the foundation for the conceptual withering away of the state system in Luhmann’s schema of society. Society looks after itself because every one of its subsystems is almost entirely and more or less equally competent for its own function insofar as that function is feasible and socially significant.
From a Weberian viewpoint, Luhmann’s schema of society’s subsystems poses a theoretical impossibility, because one of them — law — belongs within the state, and another — politics — is action oriented to exerting pressure upon, within or between organisations of the state. Politics is itself not a system with clearly demarcated interacting parts. The action of politics in modern society has its home within subsystems of the state whose purposes are policy, representation, and legislation. Luhmann must have seen the difficulty, since he devotes much time and space to depicting politics and law as specially and constitutionally “coupled” subsystems of society.
The first failure of Luhmann’s conceptualisation of subsystems, then, is that although they do perform differentiated functions within social systems, and although it is useful in some contexts to distinguish them in his terms, empirically they are not functionally equivalent system concepts. Law and politics are not functionally and systemically equivalent to science, art, or health care.
The collapse of Luhmann’s logic on this point can be traced right back to the inconstancy of his initial evolutionary schema of society from ancient to modern times, which is set out as a “catalog of forms” through which society constructs and reconstructs its own “unity” — 1) segmentary differentiation 2) centre-periphery differentiation 3) stratificatory differentiation, and 4) functional differentiation. Luhmann is careful not to refer to governance in this context, though governance is implicit in words he uses to describe empirical components of the various differentiations, such as ‘stronghold’, ‘fortress’, and ‘nobility’. Each of the differentiations at least implies forms of societal governance, or actions to establish or maintain social order. But this is not Luhmann’s concern.
Luhmann’s terminology of evolutionary differentiation would be more persuasive if the break from status or stratificatory differentiation to functional differentiation were conceived in conceptually consistent terms as a transition from hierarchically stratified, state-elite control to equally divided controls over functional state domains.
As it stands, however, Luhmann’s differentiation schema is really a more conventional sociological classification of social subsystem evolutions according to ethnicity, or residence and community, or control of territory and resources, or lineage and status, or special functions. It does not account for the emergence of separation of powers in the modern era, which are equivalent but differentiated powers to perform specialized functions inside states.
The inconstancy arises because Luhmann’s stratificatory differentiation “between nobility and commons” gives way — in the crucial transition to functional differentiation — to a centreless society. Some element of centralized control or power was implied by segmentary, peripheral and status differentiations. In Luhmann’s schema of the transition to modern society, central control recedes in the face of differentiations not of controls or powers but rather of trivial differentiations between social functions of economy, education, sport, love, art, religion, health, and science.
In his most cogent sociological statement about why modern society lacks a centre, Luhmann insists that transitions to functional differentiation require a renouncement of “all macrosocietal criteria” for comprehending relations between social subsystems. The scientific claim, which is only inferred, is that once modern society emerges any equivalence of a sociological ideal type for conceptualizing each of the fundamental differentiations — i.e. social forms of central control in segments, centres, hierarchies — disappears. The categories need no longer be equivalent.
If we are already inclined to doubt the empirical-theoretical claim that society’s centre evaporates, our doubts are magnified to incredulity upon seeing the insufficient credibility of the concepts here employed. The great masters of conceptual precision (Leibniz, Weber) would not approve. There is no utility an evolutionary schema in which the final stage bears no equivalence to earlier stages.
A second criticism follows from the first logical inconsistency, and is similarly rooted in a lack of adequate equivalence between the major concepts. As I see them, whatever functions belong in categories of ‘law’ or ‘politics’ are special because they limit or influence the functions performed by all the other social subsystems. They are not ordinarily social, as Luhmann’s schema has it.
To the extent law or politics are ‘systems’ they must be conceptualized as state subsystems because they factually determine conditions of existence for the economy, education, sport, love, art, religion, health, and science. The partial exception is the economy which possess a unique ‘operational autonomy’ by virtue of market processes that truly are beyond the control of central directives. Not a single social subsystem has commensurate reverse influence on law or politics.
On the contrary, the chain of causation, in so far as it is concrete and tangible in the form of laws, regulations, and authoritative decisions runs entirely in the opposite direction — from politics or law to art and science. That is a reality of modern society, and it is hard to pretend otherwise.
Luhmann’s vision of modern society relies on a second major claim, namely that each of these functionally differentiated societal subsystems — with the possible exception of religion — develops a unique central “binary code” to communicate and apply its own positive and negative preferences internally, and in order that its special function can be observed by other subsystems.
The binary codes for ‘political’ systems include government-opposition, governing-governed, and legitimate-illegitimate power. The code for science is true-false or truth-untruth. Art is coded by what is art and not art, and beautiful-ugly or coherent-incoherent. The code for health is health-illness. Economic scarcity is expressed as a more complex coded distinction of property-nonproperty, with secondary money coding of payment-nonpayment or having-not having, and a possible further law-economy coding for the enforceability of property rights and contracts.
Luhmann’s concept of functional coding for law — lawful-unlawful or legal-illegal — is more persuasive than the other codes because of the realism of “operational coupling” or “structural coupling” for connecting law with politics and the economy. However, the general schema of subsystem coding is susceptible to criticism on the grounds that all these codes are no more than truisms.
My criticism of Luhmann’s subsystem-specific codes is two-fold. Firstly, it is difficult to date them historically for the purpose of defining evolutionary differentiation within or between societies. Secondly, coding does not validate Luhmann’s central assumptions about the existence of a centreless society, i.e. one that reproduces itself without human agency and state ordering. In Luhmann’s schema it is not possible to conceive the existence of a central code or state code.
My objective has been to show that Luhmann was guilty of a magnificent category error akin to the apple-to-orange comparison. The subsystem concept mistake which I outlined above undermines his whole schema at the outset. Nevertheless, in this otherwise fruitless voyage we have discovered a thing of great value — the binary code as the distinguishing mark and rule of society systems.
Sources all written by Niklas Luhmann:
Social Systems, Stanford University Press 1995
Theory of Society: Volume 1, Stanford University Press 2012
Theory of Society: Volume 2, Stanford University Press 2013
Law as a Social System, Oxford University Press 2004
Political Theory in the Welfare State, Walter de Gruyter 1990
Risk: A Sociological Theory, Walter de Gruyter 1994