Skip to content. | Skip to navigation

Personal tools


Consider the world interacting

  • Logo CNRS
  • Logo ENSL
  • logo-enssib.jpg
You are here: Home / Agenda / Seminars / "Les engagements épistémiques des théories de la complexité »

"Les engagements épistémiques des théories de la complexité »

THE CONFERENCE WAS POSTPONED (update : 11th march 2020). Hopefully, a new date will be be possible in November, 2020. The two-day conference “Epistemic commitments of complexity theories” aims at gathering practitioners of the field, as well as researchers in philosophy, history, sociology, in order to address the axiological, ideological and political aspects of some of the most important complexity theories: complex systems sciences (Santa Fe Institute style), cybernetics and systemics (autopoiesis, Hayekian neoliberalism, Odum ecology, etc.)…
When Nov 19, 2020 09:00 AM to
Nov 20, 2020 06:00 PM
Where ENS de Lyon
Contact Name
Add event to calendar vCal

THE CONFERENCE IS POSTPONED (update : 11th march 2020)

Hopefully, a new date will be be possible in November, 2020. 

Several social and human science works have shown how science and knowledge are not merely a world of ideas. Scientific theories are embedded in different practices, techniques, ontologies, institutions, materiality, norms, and ideologies (Hackett et al., 2008). Scientific narratives, theories, models, and even equations are – despite their often supposed neutrality – carriers of normative and political viewpoints that must be explicated and democratically debated (Granjou & Arpin, 2015; Knorr-Cetina, 1982; Jasanoff, 2015; Vieille-Blanchard, 2007).

Following such an anthropological take on complexity theories, this workshop will address the normative, ideological and political underpinnings of complexity science, including questions such as:

  • What kind of ontology of the social world is implicit in complexity theories?

  • What performative and normative effects do such ontologies have on complexity specialists’ visions of politics?

    • What are complexity specialists’ commitments in the academic, social, economic, and political fields?

List of invited speakers

- Jean-Pierre Dupuy (
- Catharina Landstrom (
- Andrew Pickering ( - Thomas Turnbull (
- Michel Morvan (

Call for papers

Two slots are open for external speakers. Complexity practitioners are welcome. The conference will provide for transport and hosting for the accepted speakers.




Thursday 19 : From cybernetics to complexity institutes

10 - 10.30       Introduction on complexity history: Fabrizio Li Vigni
10.30 - 11.45  Andy Pickering : British cybernetics
11.45 - 12.30  contributed talk
12.30-14         Lunch
14 - 15:15       Thomas Turnbull: Second Cybernetics: the Science of Saving Energy
15.15 - 16       David Chavalarias : "De Cybersyn aux macroscopes sociaux : contributions de la recherche sur les systèmes complexes à une réflexivité sociale"
16 - 17.15       Jean-Pierre Dupuy : From cybernetics to complexity

Friday 20 : Complexity, simulations and social systems

10 - 10.45      Catharina Landstrom: modeling and expertise in social simulations
11.45 - 12.45  Michel Morvan : CosmoTech: from complexity research to business.
12.45 - 14       Lunch
14 - 15.15       From complexity theories to lean management, David Vallat (Sciences Po, Lyon)
15.15 - 16.15  Fabrizio Li Vigni : Santa Fe Institute as a neoliberal symptom
16.15 - 17       Final discussion




 Representing or intervening? When less is more in environmental modelling  - Catharina Landström

How much ‘real world’ complexity can you ignore? This is a question that all computer simulation modellers have to address. While computer models make it possible to represent complex systems there are limits to the number of parameters it makes sense to include in a model. Different modelling approaches have different affordances in this regard. The purpose of modelling is also important. Scientific modellers often take pride in the number of elements and relationships that the models they create can represent. However, modelling for environmental management requires a different perspective on complexity and model representation. To build models that provide actionable knowledge for environmental decision makers it is necessary to minimise the number of elements and processes included in the model. Decision makers are not interested in knowing as much as possible about the functioning of a system, they need to know what they can do, to which effects. Models for environmental management must be trustworthy and provide actionable knowledge. But, what is a modeller to do when all the decision makers involved do not operate in the same realities? In this presentation I will consider how the tension between representing complex realities and intervening in them is present in environmental modelling for decision making. Drawing on empirical examples from the field of water management in the UK I will discuss how complexity can be addressed in modelling for the management of socio-hydrological systems.

gordon pask and the cybernetic method -  andrew pickering

In 1955 Stafford Beer defined cybernetics as the science of 'exceedingly complex systems'—wicked systems we cannot understand a priori. Many of the systems we struggle with in the world today fall into this category, so it is interesting and important to think about how to engage with them. In 1958 Gordon Pask contrasted the cybernetic method appropriate to such systems with the scientific method. The cybernetician aims to 'maximise' interaction with an assemblage, while the scientist aims to 'minimise' it (hypothesis-testing). Unfortunately, Pask's discussion relates to his very ambitious (and unsuccessful) attempts to develop a structure of metal threads to replace the human management of a factory, which I find impossible to follow in any detail. This talk therefore focusses on a much simpler real-world example, the adaptive management of the Colorado River, and seeks to clarify, expand and possibly correct Pask's image of cybernetic method. Pask saw this method as a way of developing a 'language' appropriate to managing an exceedingly complex system, and I show that 'language' here has to be understood as a 'vocabulary' of patterned actions—a 'conversation' of actions, not words. Pask claimed that such conversations amount to building up a 'concept,' while I argue that they are not oriented to concept-formation but aim instead at a choreography of agencies (a multiplicity of river flows, sediment deposition, dam operations)—in a process that, following Heidegger, I have previously referred to as poiesis. My principal conclusion is, then, that we need to organise our dealings with complex systems in performative and choreographic (rather than scientific and representational) terms. I make a further post-Paskian distinction between science and cybernetics. If science aims at 'detachable' and dualist mechanisms (machines that dominate nature and function independently of us), the cybernetic method gears continuing human actions into nature (the timing of water flows from dams in relation to river-flooding, say). The cybernetic method thus aims at a different sort of end-state from conventional science. If time permits, I will discuss more examples, including psychiatry (also mentioned by Pask) and anti-psychiatry.