congresos y reuniones científicas
Optimality Studies: Models, Theories and Idealizations
Workshop; First Workshop on Models and Idealizations in Science; 2016
Optimality studies have a long tradition in the history of biology. For instance, it was common both in Aristotelian biology and Natural theology to explain the presence of a particular trait by appealing to the fact that it is was optimally designed to satisfy its télos or its function (see Blanco (2008); Henry (2013)). Nevertheless, it is only in the 1960´s and 70´s that the contemporary approach to this kind of studies develops, via the so-called "optimality models". Nowadays, these kinds of studies are usual, and sometimes central, in many parts of biology. Optimality models are typically said to have three components or "parts". These are (a) A trait-type that can assume a certain range of values; (b) A "currency" or variable being optimized; and (c) A set of constraints, of possibly different kinds (physical/mechanical, environmental, developmental, etc.). The example we are going to look at comes from the field of OFT, and was proposed by T. W. Schoener (1974). This model studies the optimal diet that a predator should follow (understood as the kinds of prey that it should include in it). That is, element (a), the trait-type, is the diet of the predator, while its possible values are all the different combinations between prey types available in the environment. For instance, if in a given hunting territory there are three different kinds of prey (let´s say, triangles, circles and squares, noted as T, C, S), then, the possible diets will be {T}, {C}, {S}, {T, C}, {T, S}, {C, S}, {T, C, S}. What Schoener chose as an optimization criterion or "currency" is the energetic efficiency of each diet (measured as average calories per unit time obtained with it). The constraints of the model include energetic ones (such as the average energy spent on searching, pursuing, handling and digesting prey, and the energy obtained from the prey included in the diet) and time-based ones (time spent searching and handling for prey in the diet). Some of those, in turn, can be thought of as defined notions; for instance, the average time spent searching for prey in a diet depends on the number of prey included in the diet (having more items in the diet diminishes search time), the abundance of said prey, the velocity at which the predator moves or searches its environment, etc.The trade-off being modeled here is the following. Prey items can be "ranked" according to the net energy (total energy they give minus the energy spent handling them) over the average pursuit time they provide. That is, the highest ranking prey should always be included in the diet because, once found, it returns the highest amount energy for the time spent pursuing it. Including only highly ranked prey in the diet will make the average energy gain from each unit of time spent in pursuit very high. However, if those high-quality prey types are not very abundant, it will also make searching time and energy high. All of this can be represented mathematically by the following equation: Therefore, a predator might be better off including lower quality prey in its diet, thus reducing average energetic gains from the time spent in pursuit, but increasing the frequency with which it encounters prey. From this equation and these sorts of considerations, Schoener derives a mathematical optimum. We can conceive the predator as starting with a diet consisting only of the highest ranked prey. It then adds, one by one, lower quality prey in its diet, until the diminishes in time and energy spent on searching stop compensating for the diminishes in average energy gained per hunt and the increases in average pursuit time.There are two different ways in which the nature of optimality models is studied: by their role in evolutionary studies, and by its role in functional biolgy. In many scientific papers, optimality models appear to be conceptually related to natural selection somehow. In both in the philosophical and biological literature, there have been strong disagreements about what the precise nature of this relationship is, and, correspondingly, about the legitimacy of the optimality approach as a whole. For example, some accusations have been made that optimality models presuppose adaptationism, and are therefore unfalsifiable and useless (Gould & Lewontin 1979; Gray 1987; Pierce and Ollason 1987). On the other hand, it has been argued recently that optimality models reflect only selective dynamics?not evolutionary dynamics as a whole?(Potochnik 2009), or even no dynamics at all, by giving equilibrium explanations of where selection will take a population in the long run (Rice 2012), or by merely allowing one to identify environmental factors that could be relevant to selective processes (Bolduc and Cézilly 2012). On a seemingly different note, philosophers who defend the propensity interpretation of fitness argue that optimality models give (at least sometimes) ways of determining fitness differences that are independent of actual reproductive success (Beatty 1980; Brandon and Beatty 1984; Millstein 2014).On the other hand, some authors consider that optimality models play a role?independent of evolutionary studies?in functional biology. This is the case, for example, of Wouters (1999), who believes that optimality explanations are a kind of ?explanation by design?, which are, in the view of the author, functional non-evolutionary and non-causal explanations. In this work we will discuss the role and nature of optimality models by showing the way in which both approaches complement each other. That is, we show, on the one hand, that (at least some) optimality explanations are a kind of functional explanation; and on the other hand, that they can and do perform certain roles within evolutionary biology, but only in the way that functional studies usually do.We will approach both matters through the explication of the fundamental concepts which figure in these models, through the recognition of the structure of the explanations they help to provide, and through the study of the relation between such concepts and explanations with the standard evolutionary and functional ones.The discussion about the precise nature of optimality models has some interesting and more consequences regarding models in biology, and in science more generally. First, it illustrates that the use of the word "model" (when used in opposition to "theory", or to denote the lack of such, and not generally to refer to every scientific construct) to describe a particular scientific construct, can be mistaken. The source of this errors may be (as it is in the present case) a lack of metatheoretical clarity about the nature of this construct and its links with other constructs from the same or related disciplines. In the case examined here, we show that optimality explanations can be treated as instances or specifications of some standard theoretical explanation-patterns. Second, another reason for the use of the word "model" (in opposition to "theory") sometimes has to do with the relatively high degree of idealization. While we recognize that idealizations are omnipresent in science (and optimality explanations are not exempt from this), we argue that, in the present case, the idealized character of these models is greatly exaggerated or misconstrued. That is, many of?what are thought to be?idealizations of the models themselves are not so, but only of their application in evolutionary contexts. Lastly, and more generally, the arguments presented in this work go against the (extended) idea that explanations in biology have a peculiar nature or status (because they do not appeal to laws, or are a priori, more highly idealized, etc.). In the present case, after careful explication of the terms that figure in these explanations and reconstruction of the explanatory patterns in which these terms figure, the peculiar appearance disappears, and what is left looks a great deal like other kinds of scientific constructs taken from other disciplines.BibliographyBeatty, J. (1980). Optimal-Design Models and the Strategy of Model Building in Evolutionary Biology. Philosophy of Science, 47(4), 532?561.Blanco, D. (2008). La naturaleza de las adaptaciones en la teología natural británica: análisis historiográfico y consecuencias metateóricas. Ludus Vitalis, XVI(30), 3?26.Bolduc, J.-S., & Cézilly, F. (2012). Optimality modelling in the real world. Biology and Philosophy, 27(6), 851?869.Brandon, R., & Beatty, J. (1984). The Propensity Interpretation of ?Fitness?--No interpretation Is No Substitute. Philosophy of Science, 51(2), 342?347.Gould, S. J., & Lewontin, R. C. (1979). The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme. Proceedings of the Royal Society of London, 205, 581?598.Gray, R. D. (1987). Faith and Foraging: A Critique of the ?Paradigm Argument from Design.? In A. C. Kamil, J. R. Krebs, & H. R. Pulliam (Eds.), Foraging Behavior (pp. 69?140). Springer US. Retrieved from, D. (2013). Optimality and Teleology in Aristotle?s Natural Science. In B. Inwood (Ed.), Oxford Studies in Ancient Philosophy.Millstein, R. L. (2014). Probability in Biology: The Case of Fitness. In A. Hájek & C. R. Hitchcock (Eds.), The Oxford HandbookPierce, G. J., & Ollason, J. G. (1987). Eight Reasons Why Optimal Foraging Theory Is a Complete Waste of Time. Oikos, 49(1), 111?118., A. (2009). Optimality modeling in a suboptimal world. Biology and Philosophy, 24(2), 183?197.Rice, C. (2012). Optimality Explanations: A Plea for an Alternative Approach. Biology and Philosophy, 27(5), 685?703.Schoener, T. W. (1974). The Compression Hypothesis and Temporal Resource Partitioning. Proceedings of the National Academy of Sciences of the United States of America, 71(10), 4169?4172.Wouters, A. G. (1999). Explanation Without a Cause. The Netherlands: ZENO.