Résumé
Industrial ecology (IE) is a new research field, mainly developed by engineers, which aims at saving materials and energy. Based on the analysis of industrial metabolism, it promotes symbiotic relations between actors (industries) on different territories. Originally designed to optimize flows of productive activities (on business parks for example), it has been extended to other situations in different contexts (i.e. urban heating networks). Different experiments such as the Kalundborg symbiosis are now well documented. Though very heterogeneous, all these experiments share a common methodology which lays on the study of the industrial metabolism. The approach of IE is based on the idea that the natural metabolism (the biological systems) can be reproduced on the scale of human activities. Graedel (1996) has thus proposed a representation of a completely symbiotic system in which the only external energy is solar. From the perspective of social sciences, this representation can be described as eco-centric. Different models are under discussion. Debates have emerged with the assessment of different experiments and failures were explained by the non inclusion of the «human factor» (Mirata, 2005) into the experiments. More explicitly, two different perspectives about IE emerged from the debate between Allenby (2006) and Ehrenfeld (2007). Allenby, pictures IE as a spontaneous process, based on opportunities and market based exchanges, whereas Ehrenfeld defends a vision of IE as a socially constructed process, where transactions are embedded in the social sphere - using the terminology of Granovetter (1985). In the first case, and even if not explicitly mentioned, human exchanges are compared to natural exchanges (natural systems) and environmental constraints are, at best, taken into account by imitating natural systems. It therefore appears appropriate to speak about an eco-centric vision (Hess, 2010). The second approach focuses instead on the social dimension, but fails to explicit the links with the natural environment. This explains why Hess (2010) considers the reference to ecosystems is metaphoric in this approach, and goes along with an anthropocentric point of view. This paper aims to reconsider the debate between the two conceptions of IE. That is, firstly, to clarify the different epistemological positions. And secondly, we aim to go beyond the traditional debate in order to promote a real interdisciplinary perspective, notably between economics and social sciences on the one hand, and more (so called) applied sciences (ecology, genetics), on the other. To do so, we start by reviewing each position. We can contrast two ways of representing relationships between disciplines, which reflects two implicit epistemological positions. This first is a strategy ofjuxtapositionwhich consists in accepting the results of each other without questioning the status of the research object. Each side mobilizes its own tools, that is to say, its own principles of action and regulation (ie rationality and the market on the one hand, evolution and natural selection of the other). As a consequence, opposed or irreconcilable epistemological point of views can coexist easily, without apparent contradictions. Another way consists in embedding the two opposed positions into a wider theoretical corpus. When doing so, the opposed positions becomes a special case of this general corpus. Again, two positions can be identified. The first, named thebusiness model position, aims at reducing human action to a general objective : profitability. Survival (literally) therefore requires cost saving and synergies are instruments associated with cost saving. Natural systems are subordinated to human goals, nature is to serve productive purposes. In contrast, the other vision is completely at the opposite. This model is based on the assumption of subsidiarity of human actions which are under laws that manage biological systems. As a result, the real (true) survival model in the medium / long run imposes that human activity must be managed within this framework. However, natural systems are generally regarded as having no purpose other than their survival and reproduction, whereas human systems are reflexive and therefore able to design goals, criteria and modalities of action. One crucial points of this debate, lays on the question ofintentionalitywhich distinguishes natural systems from human systems. This communication aims to evaluate each model by showing its assumptions, characteristics and mechanisms, to highlight their normativity. We show that the so called vision supported by biologists and geneticists, is more complex and mobilizes mechanisms that are more subtle than those proposed by critics. The functioning of biological systems, with operates at different levels of complexity is difficult to grasp. Thus, the mechanisms of natural selection are often assimilated to the survival of the fittest (strongest), which does not seem to give room to chance. This leads to the development of different strategies from the same starting point. Meanwhile, the market metaphor itself is normative, and has its own weak points, perhaps better known by the social sciences community. There are very restrictive assumptions about the actors' behavior. Regulation encounters market failures, inefficiencies and problems of fairness. There are also difficulties in integrating environment and externalities. Finally, the communication will attempt to propose criteria to be discussed in order to build common properties, to qualify action and systems (objectives, criteria for action, context). A middle option might be to change the basis for discussion, namely to turn towards the notion ofeco-situation rather than the notion ofeco-centrism. The criteria of complexity as a departing point is preferred to the criteria of opposed conceptions. Human systems are under various constraints, conservation, functional redundancy, diversity, they have to cope with disturbances, inefficiencies in the functioning of organizations ... Much can be learn for evolution principles of ecosystems and some can be applied to human sustainability (or sustainable development) maybe reconciling apparently irreconcilable approaches.