This article on the epistemology of computational models stems from an analysis of the Gaia hypothesis (GH). It begins with James Kirchner’s criticisms of the central computational model of GH: Daisyworld. Among other things, the model has been criticized for being too abstract, describing fictional entities (fictive daisies on an imaginary planet) and trying to answer counterfactual (what-if) questions (how would a planet look like if life had no influence on it?). For these reasons the model has been considered (...) not testable and therefore not legitimate in science, and in any case not very interesting since it explores non actual issues. This criticism implicitly assumes that science should only be involved in the making of models that are ‘‘actual’’ (by opposition to what-if) and ‘‘specific’’ (by opposition to abstract). I challenge both of these criticisms in this article. First by showing that although the testability—understood as the comparison of model output with empirical data—is an important procedure for explanatory models, there are plenty of models that are not testable. The fact that these are not testable (in this restricted sense) has nothing to do with their being ‘‘abstract’’ or ‘‘what-if’’ but with their being predictive models. Secondly, I argue that ‘‘abstract’’ and ‘‘what-if’’ models aim at (respectable) epistemic purposes distinct from those pursued by ‘‘actual and specific models’’. Abstract models are used to propose how-possibly explanation or to pursue theorizing. What-if models are used to attribute causal or explanatory power to a variable of interest. The fact that they aim at different epistemic goals entails that it may not be accurate to consider the choice between different kinds of model as a ‘‘strategy‘‘. (shrink)

Foreword -- Preface -- Introductory -- What is Gaia? -- Exploring Daisyworld -- The Archean -- The middle ages -- Modern times -- The contemporary environment -- The second home -- God and Gaia -- Gaia since 1988 -- Epilog -- References -- Further Reading -- Index.

In this article, I will discuss possible differences between ecosystems and organisms on the basis of their intrinsic complexity. As the concept of complexity still remains highly debated, I propose here a practical and original way to measure the complexity of an ecosystem or an organism. For this purpose, I suggest using the concept of logical depth (LD) in a specific manner, in order to take into account the difficulty as well as the time needed to generate the studied object. (...) I illustrate this method with fully controlled Daisyworld simulations (i.e., simulations based on the Gaia hypothesis) that have often been proposed to mimic living systems. The method consists of the following sequential stages: (1) identification of the shortest program able to numerically model the studied system (also called the Kolmogorov–Solomonoff complexity); (2) running the program, once if there are no stochastic components in the system, several times if stochastic components are there; and (3) computing the time needed to generate the system with LD complexity. This measure is supposed to estimate the system complexity. It appears in this study that LD estimations fit well with the intuition we have of complex systems, with higher complexities being found for more realistic Daisyworlds. With such a method of capturing the time needed to build a system, we expect to detect in future studies any quantified differences between complex ecosystems and organisms. (shrink)

This article investigates artificial life image-making in relation to and as constituent of the moving image, specifically artificial life visualized in three-dimensional computer-generated space . Of particular interest in this examination is the view or `window', from the virtual camera, into the artificial life computational model or `world' , and how it organizes a dense field of expectations. Analogous to looking through a telescope or microscope, the view into the artificial life world is monocular and often fixed in the world; (...) in this regime we look `at' computer models, which are often referred to as agents, `creatures', `organisms' or `cyberbeasts'. This tactic of looking through the instrumentality of science, the arts of reality, is parallel to looking through André Bazin's `long take' in cinema and documentary film-making in which we look `at' an unmediated view of reality; in other words, in looking `at' an image of artificial life we look `through' a non-intrinsic regime of seeing. This investigation into the interpretative regime provides only a partial account of the intercultural traffic between artificial life and the moving image. I further this discussion by arguing that artificial life screen-based works are a contemporary extension and reworking of the nature film in that they provide accounts of the `natural' world that are familiar and similar to those of their filmic and cinematic predecessors. I contend that artificial life visualizations and the stories that often accompany the images are the latest manifestation of this photographic and filmic tradition. To illustrate my point I discuss the scientific project Daisyworld, the associated narrative that frames the research, and the researchers' attempt to mathematically model an `imaginary planet [with] a very simple biosphere'. This sets the scene to explore artificial life in relation to Disney wilderness and nature films. (shrink)

This article on the epistemology of computational models stems from an analysis of the Gaïa hypothesis. It begins with James Kirchner’s criticisms of the central computational model of GH: Daisyworld. Among other things, the model has been criticized for being too abstract, describing fictional entities and trying to answer counterfactual questions. For these reasons the model has been considered not testable and therefore not legitimate in science, and in any case not very interesting since it explores non actual issues. (...) This criticism implicitly assumes that science should only be involved in the making of models that are “actual” and “specific”. I challenge both of these criticisms in this article. First by showing that although the testability—understood as the comparison of model output with empirical data—is an important procedure for explanatory models, there are plenty of models that are not testable. The fact that these are not testable has nothing to do with their being “abstract” or “what-if” but with their being predictive models. Secondly, I argue that “abstract” and “what-if” models aim at epistemic purposes distinct from those pursued by “actual and specific models”. Abstract models are used to propose how-possibly explanation or to pursue theorizing. What-if models are used to attribute causal or explanatory power to a variable of interest. The fact that they aim at different epistemic goals entails that it may not be accurate to consider the choice between different kinds of model as a “strategy“. (shrink)