the luhmann challenge!!!

I challange you, Luhmannites. In 150 words or less, can you give me an example of an empirical phenomenon or puzzle that was clarified, explained, or resolved using autopoeisis or any other of Luhmann’s concepts? For example, will I understand the rise of capitalism in England better? Why African Americans have lower levels of education than Whites? Increased political polarization in America? Seriously, dudes, throw me a bone. Name me one empirical question that I can better answer because I spent the weekend reading Social Systems.* I dare you. I double dog dare you.

* And yes, I also checked out some of his “empirical” works like The Reality of the Mass Media. Still feel burned.

UPDATE: I’m changing the terms of our agreement.  Every 60 minutes that pass without a clear empirical application of autopoiesis, one of these fellows will “accidentally” find themselves underneath an annotated copy of Social Systems.


Written by fabiorojas

February 10, 2010 at 12:22 am

36 Responses

Subscribe to comments with RSS.

  1. I think you need to reread the first paragraph of “Instead of Preface to the English Edition” to Social Systems.



    February 10, 2010 at 12:47 am

  2. The thing is, Luhmann wrote a lot of full-length essays on specific issues (power, civil society, art, love, mass media, religion, the education system, law, science, economics, organizations and risk, to name a few). So these materials must be checked. I’m not sure how many are available in English though.

    The Spanish translations are excellent if you are interested.



    February 10, 2010 at 1:38 am

  3. Guillermo, I’ve read some of these essays, and frankly, it just seems to be puffery. For example, what was so original about Luhmann’s analysis of the media? I’ve got a copy right next to me. Tell on what page there is a passage that lays our the great insight. Give me the page number. Please. I am not joking.



    February 10, 2010 at 2:43 am

  4. Fabio: Maybe an image will help clarify things (well, not on the empirical side) — autopoiesis and other reflexive/mystical/über concepts such as structuration are often clarified and represented by Escher’s drawing hands image (if I see it in one more presentation…):



    February 10, 2010 at 4:35 am

  5. TF, you just doomed another itty bitty kitty to a twelve hour recitation, in German, of “Consequences for Epistemology,” the final chapter of Social Systems. Hope you’re happy now.



    February 10, 2010 at 4:39 am

  6. I don’t have the book with me, and I am definitely not a Luhmann fan, but there is some okay stuff in it about how journalistic discourse and practice get auto-referential. Of course he was not the first to make this point….

    Luhmann is okay for how art, law, media, etc. get auto-referential (art is increasingly about art, law is increasingly about law etc.) and get more and more autonomous and differentiated. He is of course kind of right and kind of wrong. He tends to make the same argument for all social systems, though. His writing is also not very good.


    Ari Adut

    February 10, 2010 at 5:28 am

  7. Ari! Thank you god you chimed in, I just caught a stray tabby… Like you, I don’t think I’m blown away by the self-reference –> more differentiation observation, but at least it’s an empirical phenomenon that could be observed and tested. (oops! I did positivism again!)



    February 10, 2010 at 5:37 am

  8. I am not one of the greatest fans of Luhmann. Below are some refs that I skimmed up and copy&pasted from my thesis. So sorry for babbling, but perhaps it helps…

    Maturana and Valera who original proposed autopoiesis strongly objected to applying autopoiesis to social systems, too (Maturana and Varela, 1979, pp. 80-81). They regard social systems as being constituted by the interactions of their members. Social systems arise due to the structural coupling of “real” autopoietic systems (humans), the components of social systems. A social system influences its members, as part of their environment, and is in turn influenced by the interactions between its members. Members can participate in more than one social system at the same time and each social system is constituted by a distinct set of interactions.

    Consequently, Maturana and Varela (1979, pp. 80-81) propose the use of the term “autonomy” instead of autopoiesis as the proper term for referring to the identity-preserving capability of social systems; autonomous systems strive to maintain their identity by subordinating all changes to the maintenance of their own organization as a given set of relations. Therefore Whitaker (1995) proposes (1) to follow Maturana and Varela (1979) and Varela (1981) by switching from claiming that “enterprise” (social system) autopoiesis exists to a claim that these processes can be explained by autonomy and (2) to leave this esoteric debate in the ivory towers and work with the non-controversial aspects of the theory.

    In my opinion Luhmann does not give us anything new that was not there before. Autopoietic systems are not determined by the environment: they are related to the environment, but their operations are totally closed. There is no deterministic input-output-relationship between the system and its environment; therefore there is no possibility of controlling an autopoietic system completely from the outside. This is, of course, a basic postulate of Cybernetics (self-organization). However, Luhmann does not specify the operators and mechanisms of these self-organizing processes in social systems, and is criticized for only providing a very abstract and impoverished view of social interactions which pays no attention to how the individual human actors interact in order to establish requisite variety (Mingers, 2002, pp. 289-294). Therefore I am still not convinced that autopoiesis as a theory of self-producing/-organizing systems can be transferred to social systems “per se”, and I am not alone in this (Hejl, 1984, pp. 72-75, Kay, 2001, Kickert, 1993, Mingers, 2002); but I do believe that we could use the concept of autopoietic systems (humans) on the micro-level for coming up with something useful on the group- or macro-level (processes of self-organization). But wether we really need Luhmann for that or just stick to less abstract basics…. :-).


    HEJL, P. M. (1984) Towards a Theory of Social Systems. Self-Organization and Self-Maintenance, Self-Reference and Syn-Reference. IN H. ULRICH, G. J. B. P. (Ed.) Self-Organization and Management of Social Systems. New York et al.

    KAY, R. (2001) Are Organizations Autopoietic? A Call for New Debate. Systems Research and Behavioral Science, 18, 461-477.

    KICKERT, W. J. M. (1993) Autopoiesis and the Science of (Public) Administration: Essence, Sense and Nonsense. Organization Studies, 14, 261-278.

    MATURANA, H. R. & VARELA, F. J. (1979) Autopoiesis and Cognition. The Realization of the Living, Boston, MA, USA, Kluwer.

    MATURANA, H. R. & VARELA, F. J. (1987) The Tree of Knowledge., Boston, MA, USA.

    MINGERS, J. (1995) Self-Producing Systems. Implications and Applications of Autopoieses, New York, NY, USA, Springer.

    MINGERS, J. (2002) Can social systems be autopoietic? Assessing Luhmann’s social theory. Sociological Review, 50, 278-299.

    ROBB, F. F. (1989) The application of autopoiesis to social organizations – a comment on John Mingers’ “An Introduction to Autopoiesis: Implications and Applications”. Systemic Practice and Action Research, 2, 343-348.

    SEIDL, D. & BECKER, K. H. (Eds.) (2005) Niklas Luhmann and Organization Studies, Abingdon, UK et al., Liber & Copenhagen Business School Press.

    VARELA, F. G., MATURANA, H. R. & URIBE, R. (1974) Autopoiesis: The organization of living systems, its characterization and a model. BioSystems, 5, 187-196.

    VARELA, F. J. (1981) Autonomy and Autopoiesis. IN ROTH, G. & SCHWEGLER, H. (Eds.) Self-organizing Systems. An Interdisciplinary Approach. Frankfurt am Main, Germany et al.

    WHITAKER, R. (1995) Self-Organzation, Autopoiesis, and Enterprises. ACM SIGOIS Illuminations.


    Christoph Rosenkranz

    February 10, 2010 at 7:49 am

  9. And this is of course no example for answering the challenge because I am with Fabio on that ;-).


    Christoph Rosenkranz

    February 10, 2010 at 8:00 am

  10. Well, I guess asking what you can find out better according to these specific questions is the wrong approach. With systems theory, you cannot implicitly give better answers to *all* questions. Rather, you either put other questions (focus shift) or you have a different view on empirical phenomenons (btw, what are you thinking of, when you consider the opposite of “empirical” phenomenons?). There are indeed some empirical studies (although I know only about German studies) that combine systems theory (as theoretical framework, or meta-theory) and qualitative, reconstructive methods, thus the empirical data gives an example on how, e.g. structural coupling is “working in practice” – I mainly refer to research from Werner Vogd, who analysed decision-finding-processes in hospitals from a systems theoretical point of view.

    If you take the concept of autopoiesis seriously, you’ll have several conclusions. E.g. assumptions on causality. “Communication” (social systems) does not determine the social systems but keeps up a state of uncertainty whose ascertainment is achieved during the process of communication – which, again, creates new alternatives on how to proceed and thereby maintains the uncertainty. Effects in social systems cannot be explicitly assigned to specific causes (e.g. the motives of underlying actions). And then you have the aspect of Eigen-Values (von Foerster): A quantitative approach knows (or better: believes to know, but actually doesn’t know anything) about an empirical phenomenon a priori. It’s only a question of good operationalization of variables. From a system theoretical point of view, you would be less pretentious, saying that a system creates its own history, description etc. which cannot be understood in a comprehensive way by an outside observer (complexity).

    In my opinion, there’s a lot of empirical potential using systems theory, but we haven’t really startet trying it out yet… Unfortunately, due to the lack of English translations, you cannot entirely follow the further development of Luhmann’s systems theory and its application in empirical research…



    February 10, 2010 at 9:22 am

  11. Btw, not explicitly the Luhmann’s systems theory, but a kind of further development in the direction of George Spencer Brown’s form theory is achieved by Dirk Baecker. You find some English papers here: and his blog here

    Best wishes



    February 10, 2010 at 9:25 am

  12. Your question is clearly dependent on a system of observation that cannot observe the system of sociological explanation. Therefore it is impossible to answer your question in any way that would be intelligible to you. Sorry.



    February 10, 2010 at 10:19 am

  13. There are several applications of autopoiesis in organization practice in a recent volume by Magalhaes and Sanchez (2009).
    Among these applications is a simulation study that provides an explanation to the somewhat counterintuitive and so far unexplained phenomenon that organizations are able to learn and create knowledge in spite of high personnel turnover or even severe lay-offs.

    The literature on knowledge management more or less suggests that losing staff is like losing memory, whereas Luhmann (and works beyond) points to the autopoiesis of organizations as the produce and reproduce themselves through communication. In effect, it clearly differentiates between organizational knowledge and individual knowledge in a way only systems thinking can.

    Magalhaes, R. & R. Sanchez (Eds.) (2009). Autopoiesis in Organization Theory and Practice. Emerald, London.



    February 10, 2010 at 11:03 am

  14. Thanks Steffen!


    Christoph Rosenkranz

    February 10, 2010 at 11:35 am

  15. regarding ecolocical problems:

    political system can only use coding in power/no-power, likewise the economy system as income/no-income. so both are doomed to not understand the other.

    one way to crack up this is for example to introduce “pay a lot of money if you harm ecology” after “uh, fucked up ecology is a thread to our power”.

    or the paradox of technology: it introduced a lot of ecological problems that only can be solved through use of technology…

    and so forth..

    pretty easy to understand if you read “ecological communication” (I don’t know the translated title of “Ökologische Kommunikation”)



    February 10, 2010 at 11:49 am

  16. political system can only use coding in power/no-power, likewise the economy system as income/no-income. so both are doomed to not understand the other.

    This seems implausible.

    On the other hand, I am reluctant to intervene in Fabio’s trolling of the German intelligenstia.



    February 10, 2010 at 12:59 pm

  17. Talking (writing) about functional systems needs some more space than just some words… In short, you can say that functional systems like science, economy, politics, religion etc. “sort” the communication according to certain codes (which are dual, with a positive and negative side). This has to be understood as a very abstract level of social systems. Empirically, you never find a communication which is only about one single aspect (e.g. in hospitals you not just have a “medical” communication, but hospitals also have to deal with legal issues, finances etc). Thus, organizations appear as “meeting-point” (Luhmann) of functional systems, as organizations have to deal with the different requirements from politics, economic, law etc.



    February 10, 2010 at 1:14 pm

  18. There are lots of possibilities to answer your question, I’ll try three of them:

    1. “Did you ever try not to think?”

    This meets your “150 words or less” challenge – but should not be taken too serious. My point is: Our consciousness produces thoughts out of thoughts in a recursive process, autonomically rather than autarchically. It can (and needs to) be irritated from time to time.
    Furthermore it requires specific activity of the brain (one shouldn’t equate consciousness with the brain, by the way! It represents an emergent order) – Luhmann calls this structural coupling (“strukturelle Kopplung”): Human consciousnesses and brains are separate systems, whereas the first ones are portrayed as autopoietic systems. Thoughts are the basic elements of psychic systems – and as long as a thought is followed by another one the system is operating (death and zen buddhism need to examined at some other time, perhaps). Let’s call answer 1 the “fun answer”.

    2. For empirical input you might have a look here: or here – but generally it seems hard to find english texts on Luhmann’s theory.
    Let’s call this the “academic”, “scientific” or “factual answer”.

    3. And number three? Let’s call it the “Luhmannite answer”. Communication is regulated by understanding. Understanding means differentiation between information and message (Luhmann defines communication as the unity of information, message and understanding). The recipient actualizes contingent dimensions of meaning (“Sinndimensionen”), e.g. the – social (the “fun answer”),
    – temporal (let’s wait and have a look at some more answers to your challenge) or
    – factual dimension (like in answer number 2, the “academic” reference; Steffen did the same above).

    Luhmanns theory allows to step back and observe observations. In general that’s a very clarifying feature for me. I hope you got my point, I’m kind of unfamiliar with English conversion about Niklas Luhmann. If you didn’t: feel free to ask again.

    Cheers, Sebastian

    post scriptum: The question whether these concepts are “new” (or Luhmann “gave us something new”) or not is unrewarding. What’s really *new* – especially in academic debates? ;-)

    post post scriptum: Empirical studies often fail to enregister themselves (and therefore can’t see, that they don’t see, what they can’t see) – but that’s another point of another discussion.



    February 10, 2010 at 3:44 pm

  19. One example, Luhman is able to explain Why Society Describes Itself as Postmodern



    February 10, 2010 at 3:55 pm

  20. I feel terrible for the rapidly thinning feline population of the world given the performance here.



    February 10, 2010 at 4:11 pm

  21. @Kieran: your reluctance of intervening trolling seems implausible to me, too.

    systems operate on basic, dual codes as they ‘translate’ the ‘irritations’ received from ‘outside’ the system. what’s so implausible about this?

    those codes differ from system to system – they have to differ as it is the internal mode of sorting communication (more precise: mode of observation)

    now, again: the binary code of the economic system is ‘money:no money’. the code of the political system is ‘power:no power’. as they don’t ‘speak’ the same language they can’t understand the other.

    now, think about ecological problems: how to tell the economic system to act responsible? you have to find something that is expressable in their language, in their coding system: “money:no money”. and who can and should express such? the political system – but they only ‘understand’ things if translated in their mode of perception, their coding: ‘power:no power’

    seems pretty plausible – if plausibility is a “Wahrheitskriterium” (something like ‘criteria of truth’) for you ;-)



    February 10, 2010 at 8:18 pm

  22. I find it interesting that Luhmann attracts more fervent interest among English-speaking scholars working in domains in which semantic traditions, dogmatics, systems of codification, and the like are important (thus: literature and literary criticism, philosophy, aesthetics, religion, law, and so on), which suggests an affinity between his work and certain kinds of qualitative research, I guess. Not surprising, given the influence of Koselleck and the historical semantics tradition on his work. As someone who studied political theory before coming to sociology, I liked a number of his conceptual pieces, like the short essay on ‘Familiarity, Confidence, and Trust’ – not necessarily groundbreaking, but nevertheless provocative enough to have seen it cited in more concrete empirical works on trust, etc.

    In response to some of the comments above, I don’t think his arguments about the binary coding schemata in functional subsystems will ever be convincing to those not fully committed to a Luhmannian perspective (which I’m not), though I agree that it’s one take on long-standing controversies in differentiation theory. His overall theory of modern society as functionally differentiated is a perspective, perhaps not convincing but nevertheless retaining a certain coherence and originality that I think is worth considering, for those interested in and focused on that problematic.

    I don’t find autopoiesis particularly convincing (or clear), but coupled to the thesis of functional differentiation it led him to hypothesize that systems come under increasing temporal pressures, since they have to reproduce themselves out of their own reserves. For instance, science (as a system of communication) comes to depend increasingly on publications to guide its own reproduction, but the more publications there are, the more there is to observe, which requires time. Various mechanisms – star systems, doctrinal wars, etc. – arise to provide more time by allowing vast quantities of research to be discounted as irrelevant, etc. A similar phenomenon can be found in the political system, where ideological posturing becomes a functional substitute for temporizing and delaying decisions to the point where choices are ‘forced.’



    February 10, 2010 at 10:05 pm

  23. That’s 3 or 2 : 0 for Luhmann then. Start reading “Social Systems” again instead of throwing it at cute kittens. :-)



    February 10, 2010 at 10:27 pm

  24. I think anon just axiomatically deduced the impossibility of lobbying.



    February 11, 2010 at 2:49 am

  25. @Omar:

    No would seriously deny the possibililty of lobbying (and lobbyism is no evil, btw). When it comes to bribery, for example, Luhmann call this a lack of “Interdependenzunterbrechungen” (roughly interdependence interruptions, compare the German text here: ). Nevertheless: The systems operate according to their code. Money can’t buy truth. Or love.



    February 11, 2010 at 5:41 am

  26. Oh! Apparently one word is missing in the first sentence of my second comment: “No one would seriously deny…” – that’s what I wanted to write. Sorry!



    February 11, 2010 at 6:19 am

  27. John Padgett’s has been working on what he calls “autocatalysis”. He recently presented the first chapter of a book he is working on with Woody Powell.

    When Padgett was asked about the comparison to Luhmann’s autopoiesis, he was a little glib, so I don’t want to put him on the hook. But he acknowledged some common origins (including I think Maturana and Varela mentioned above).

    He said (in short) said that a big difference is that Luhmann doesn’t include the math, so it’s not “fertile” in the same way as far as teasing out what it really means and applying it (exactly what rubbed Fabio wrong).

    The new Padgett/Powell book, which brings together specific empirical studies that illuminate the concept along with computational elaborations, may be a very satisfying but circuitous answer to the question in the post (more than 150 words for sure). If you have an interest in big-picture but empirical social structure questions, you ought to check it out.

    Read the First Chapter



    February 11, 2010 at 3:07 pm

  28. […] that is hard to follow and even harder to translate, indeed. No wonder poses the Luhmann Challenge to give “an example of an empirical phenomenon or puzzle that was clarified, explained, or […]


  29. I’m not Luhmanian enough to believe that social systems exist without the “psychological systems” or the “biological systems” we tend to name human beings. And I tend to think that systems are implying a coherence which is not found that way in the empirical world.
    However, one theoretical figure I think is worth to be defended precisely because it explains lots of empirical phenomenons: The notion that if you make an observation, you have to draw a distinction and thereby reduce and categorize the world outside. How could we, for example, imagine us falling in love if we could not draw a distinction between desireable and undesireable character traits? How could we decide what is quality if we don’t draw a distinction between the worthy and the unworthy? Luhmann insisted that there is no ontology “out there” which we just have to discover but that it is communication and interaction which construct these categories. And essentially, when we come down to these human beings, it is striking that however well-meaning society and their institutions are, they can’t really tell you what to like, what to dislike, what to find interesting and what to find boring, what to care about and what not, with whom to emphatize etc. Of course, you can try hard to follow outside guidiance and adopt a perspective but it won’t be your perspective without a process of translating for yourself what is relevand, expressed in your own language (same with learning a theory!) because..well, Luhmann would say because your programs and codes are not the same as those of the environment. So in a sense, he would like to remind us about the possibility of the non-identitcal, I think.
    However I agree that with system theory alone, we don’t solve any practical problems or understand the reasons of state failure in Haiti.


    Thorsten Kogge

    February 14, 2010 at 2:59 am

  30. “I’m not Luhmanian enough to believe that social systems exist without the “psychological systems” or the “biological systems” we tend to name human beings.”

    This is what Luhmann called (according to Spencer Brown) “conditioned co-production”. A system has always to be observed by its relation to its environment – no system can exist without environment. And Luhmann always said, social systems need certain conditions to emerge: one is, of course, some kind of presence of “consciousness” or “bodies”, or humans etc. If it would be hotter than 100 degrees outside, the environmental conditions are not optimal for social systems to emerge – since no human being could live there and irritate the communication.

    It seems to me that one great misunderstandings in reading Luhmann comes from trying to find empirical equivalents in a way like: “How can I count or measure systems?” Systems theory rather offers a heuristic for observing phenomenons to see the “structural background” of why things happen as they happen – without determining the observed object a priori(!). Systems theory also means functional(ly) analysis: We can observe an empirical phenomenon, e.g. behaviours of organizations that seem absolutely irrational (see Brunson: irrational organization), but for systems theory, irrationality or paradoxes are no problem: They serve a certain function. And the first step is to find the related problem that is solved by this “dysfunction”, the second step could be the search for functional(ly) equivalents, which may solve the problem better.

    So: “However I agree that with system theory alone, we don’t solve any practical problems or understand the reasons of state failure in Haiti.” – I would disagree to this sentence…



    February 18, 2010 at 5:00 pm

  31. M & V and others were, IMHO, mistaken when they picked on individual humans as the elements of possibly autopoietic social systems. Mingers correctly pointed out that in no way could humans be said to meet the AP criteria. However it has been argued that the AP of social systems derives from the language/concepts that humans employ in social interchanges, that these are the self-regenerating elements and that this provides a handy tool for describing social behaviour.
    Of course the question of whether or not AP is ‘better’ than any other model at reflecting ‘reality’ is a silly one because of the limitations of induction …
    The point about the autopoietic model is not that it deals with causal relations but that it helps to describe some kinds of social phenomena. AP seems to be a handy way of describing how people aggregate their ideas and form common world-views on which they act. AP easily describes ’tides of enthusiasms’, mass hysteria, crowd behaviour, superstitious and religious movements etc. but, perhaps, more interestingly, it helps to describe corporate loyalty, the phenomena of group-think and patriotism and the growth and decay of arcane social systems such as the law, accounting, medicine, the craft guilds, professional bodies, that contrive to exclude the uninitiated and privilege the acolytes. Like many other descriptive languages, it provides a handy package to use when making ostensive definitions, but it does have its limitations – especially when trying to determine where the boundary of an AP system might lie and what might be entrained in its conversations and what might not. AP systems generate their own boundaries for their own benefit, not that of the observer.
    Autopoiesis is a strictly ‘scientific’ term – it’s about description, NOT about ‘explanation’. Hypotheses non fingo! Leave explanation to metaphysics!



    February 18, 2010 at 9:46 pm

  32. […] 5. Kieran and I can live without Luhmann. […]


  33. I have been a bit puzzled by “application” of Luhman in Organization Studies and Journal of Management Studies. It seems its ok to say that some things don’t communicate with one other because Luhman said so. When people start stating things like ‘people don’t communicate, communications communicate’ then it is highly likely that there is no argument there just words.
    The answers to Fabio here seem to highlight rather trivial phenomena that could be explained without a colossal, confusingly written book. The empirical insights are hardly exciting. Davidson’s notion seems to apply very well to Luhman: “it is hard to improve intelligibility while retaining the excitement.”

    I am not certain whether we should then approach Luhman as a social theorist rather than a sociologist, like we read Giddens and Habermas? I know my Giddens pretty well and his social theory on structure-agency duality, rationality as a reflexive process, notion of identity politics, etc. solve no explicit empirical problems. They provide what I find useful/intersting perspectives to approach empirical domains.



    December 22, 2010 at 7:15 am

  34. […] – el cual es, en este sentido, un autor primariamente local (¿podría eso explicar en algo esto otro?). Es interesante observar también que la presencia de Parsons en estos textos es bastante pareja […]


  35. i too am not up to the specific challenge posed regarding the applied utility of a luhmannian framework to empirical research (a great challenge btw!), but i did come across this recent exegesis of luhmann by harrison white whose own work on ‘meaning’ being generated out of identities seeking control is also very much indebted to the cybernetic turn in information theory . . .



    January 3, 2011 at 12:16 am

Comments are closed.

%d bloggers like this: