Model Thinking in the Life Sciences: Complexity in the Making Tudor M. Baetu, Ann-Sophie Barwich, Daniel Brooks, Sébastien Dutreuil & Pierre-Luc Germain Biological Theory ISSN 1555-5542 Biol Theory DOI 10.1007/s13752-013-0100-8 1 23 Your article is protected by copyright and all rights are held exclusively by Konrad Lorenz Institute for Evolution and Cognition Research. This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self- archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication. 1 23 Author's personal copy Biol Theory DOI 10.1007/s13752-013-0100-8 MEETING REPORT Model Thinking in the Life Sciences: Complexity in the Making Second European Advanced Seminar in the Philosophy of the Life Sciences, ‘‘In Vivo, ex Vivo, in Vitro, in Silico: Models in the life sciences’’ Hermance, Switzerland, 10–14 September 2012. Tudor M. Baetu • Ann-Sophie Barwich • Daniel Brooks • Se´bastien Dutreuil • Pierre-Luc Germain Received: 12 February 2013 / Accepted: 13 February 2013 Ó Konrad Lorenz Institute for Evolution and Cognition Research 2013 The European Advanced Seminars in the Philosophy of the Gene`ve) and Dr. Philippe Huneman (IHPST), and was hos- Life Sciences (EASPLS) are biennial symposia organized by ted by the Brocher Foundation in Hermance, on the shores of members from six leading European institutions in the phi- Lake Geneva. losophy of the life sciences–the ESRC Centre for Genomics The seminar’s overarching theme focused on the notions of in Society, University of Exeter (Exeter); the European ‘‘model,’’ ‘‘modeling,’’ and ‘‘simulation,’’ with an emphasis School for Molecular Medicine (SEMM, Milan); the Institut on the philosophical implications of the pervasiveness and d’Histoire de la Me´dicine et de la Sante´ (Geneva); the Institut striking diversity of models used in the life sciences. Several d’Histoire et de Philosophie des Sciences et des Techniques, key philosophical issues were covered, including the Paris 1–Panthe´on-Sorbonne (IHPST, Paris); the Konrad explanatory value of models, the epistemic values that models Lorenz Institute for Evolution and Cognition Research should satisfy, the simplifications and idealizations required to (Altenberg); and the Max Planck Institute for the History of build a model, and the connections between models and other Science (Berlin). The second edition of the EASPLS was key concepts such as ‘‘theory,’’ ‘‘representation,’’ and ‘‘fic- coordinated by Prof. Bernardino Fantini (Universite´ de tion.’’These issues were addressed in relation to a wide variety of models, ranging from material model organisms to abstract mathematical models and computer simulations, and a wide spectrum of disciplines, ranging from chemistry to ecology. & T. M. Baetu ( ) The presenters incorporated a diverse set of professional Konrad Lorenz Institute for Evolution and Cognition Research, Altenberg, Austria backgrounds, bringing together perspectives from philosophy e-mail: [email protected] and history of the life sciences, and from working scientists. Indeed, one of the explicit goals of this meeting was to stim- A.-S. Barwich ulate a dialogue between scholars and researchers whose areas Egenis, University of Exeter, Exeter, UK e-mail: [email protected] of expertise concern different modeling practices, and dif- ferent areas of the life sciences, as well as philosophically D. Brooks complementary issues, approaches, and perspectives. This Abteilung Philosophie, Universita¨t Bielefeld, diversity served the session discussions well, reflecting the Bielefeld, Germany e-mail: [email protected] multifaceted nature of the use of models in the life sciences. The opening lectures of the seminar focused on a number S. Dutreuil of issues concerning the use of living systems as models in Institut d’Histoire et de Philosophie des Sciences et des the life sciences. The speakers and commentators widely Techniques (IHPST), Universite´ Paris 1-Panthe´on-Sorbonne, Paris, France agreed that the dissemination of terms like ‘‘model organ- e-mail: [email protected] ism’’ or ‘‘model system’’ during the 20th century has given rise to difficulties in keeping track of their various uses, and P.-L. Germain cross-referencing the term with related ones (such as European School of Molecular Medicine (SEMM) & University of Milan, Milan, Italy experimental system, animal model, etc.). For one thing, the e-mail: [email protected] ubiquity of the term itself demands that researchers be more 123 Author's personal copy T. M. Baetu et al. sensitive to what they take the organisms to be ‘‘models of’’ functions outside of theorizing. They may serve explana- in the first place. It is often assumed that the principle of tory interests as well in furthering our ability to manipulate common descent is sufficient to conclude that some simi- or control a phenomenon in an experimental setting, or larities between model organisms and other organisms that even provide understanding of a phenomenon with or are the target of investigation will be found. Applied too without a theoretical framework (Nina Atanasova). strongly, however, this assumption may reintroduce ele- Subsequent talks highlighted very interesting points ments of typological thinking when generalizing these sim- regarding the relationships between models and theories, ilarities from one species to another, or from one species to a on the one hand, and models and empirical knowledge, on higher taxa (Alessandro Minelli). At the same time, seeing the other. In some cases (e.g., metabolic control analysis), certain ‘‘types’’ repeated across taxonomic kinds is what models are initially developed as instruments for theory allows us to generalize results collected from one species of construction, but with time they acquire a partial inde- model organisms to other groups of organisms, as such pendence. The acquisition of autonomy cannot be antici- similarities indicate stable macroevolutionary features that pated, but can be diagnosed retrospectively by looking at scientists wish to comment on in the first place (Jan Baedke). the divergence between the developmental trajectories of Along these lines it was suggested that the use of model models and theories (Josephine Donaghy). In other cases, organisms serves as a general basis for a kind of analogical models remain tightly intertwined with theories: models reasoning by which to extrapolate scientific findings, though are built on the basis of theoretical knowledge and con- the term’s usage in itself tends to remain ambiguous con- strain it in return. For example, computational models were cerning the specific features at work in this reasoning (Jean used in evolutionary theories of sociality to constrain the Gayon). This ambiguity can be balanced, in part, by the conditions under which one can think that sociality will stringent demands placed on the choice of model organism, emerge (Silvia de Monte). Conversely, theoretical concepts which must fulfill increasingly fine-grained criteria before such as ‘‘evolvability’’ and ‘‘facilitated variation’’ play an they are considered exemplary for a particular biological important role in developing new approaches that may phenomenon (Tarquin Holmes). enable open-ended evolution to be genuinely modeled One important topic discussed during this session cen- (Andreea Esanu). Equally important, the logical formal- tered on the epistemic functions that organism-based ization of current empirical knowledge (e.g., molecular models may serve in scientific reasoning. In particular, reactions in a cell) can help us both better organize this attention was given to explicating the purposes that knowledge and generate new predictions (Giovanni Boni- researchers may ascribe to a model’s ability to represent a olo). Simple logical tools may help us grasp biological phenomenon of interest. In biology, modeling can often be complexity–an interesting by-product of this would be the continuous with experimentation, serving, among other standardization of biological data–although prediction may things, to help uncover causal dependencies in the system be hindered by biological complexity (Dan Nicholson). or systems of interest. This gives such organism-based The nature of the relationship between models and the real ‘‘experimental models’’ a distinctly concrete basis in con- world was covered as well. Some models investigate trast to the abstract, computation- and mathematics-based relations between properties that are abstracted from a model systems of physics and chemistry, and other areas of specific target in the world. In other cases, however, we biology (Marcel Weber). Yet such ‘‘experimental models’’ study models not because they bear similarities with bio- also deserve the full philosophical status ascribed to tra- logical systems, but because the models themselves are ditional scientific models, since both material and abstract interesting objects of study (Philippe Huneman). In such models nonetheless represent features of the system they cases, the simulation does not ‘‘simulate’’ any process other are modeling by virtue of the purposes attached to them by than itself (Fridolin Gross). Finally, some numerical sim- researchers. Weber argued this point by showing that the ulations specifically