Game theory has played a critical role in elucidating the evolutionary origins of social behavior. Sober and Wilson (1999) model altruism as a prisoner's dilemma and claim that this model indicates that altruism arose from group selection pressures. Sober and Wilson also suggest that the prisoner's dilemma model can be used to characterize punishment; hence, punishment too originated from group selection pressures. However, empirical evidence suggests that a group selection model of the origins of altruistic punishment may be insufficient. I argue that examining dominance hierarchies and coalition formation in chimpanzee societies suggests that the origins of altruistic punishment may be best captured by individual selection models. I suggest that this shows the necessity of coupling of game-theoretic models with a conception of what our actual social structure may have been like to best model the origins of our own behavior.

The constitution of a collective judgment is analyzed from a contractarian point of view. The optimal collective judgment is defined as the one that maximizes the sum of the utility each member gets from the collective adoption of that judgment. It is argued that judgment aggregation is a different process from the aggregation of information and public deliberation. This entails that the adoption of a collective judgment should not make any rational member of the group change her individual opinion, and so the collective judgment can not have any kind of epistemic superiority over the individual ones.

Quartz (2002) argues that some recent findings about the evolution of the brain (Finlay and Darlington 1995) are inconsistent with evolutionary psychologists’ massive modularity hypothesis. In substance, Quartz contends that since the volume of the neocortex evolved in a concerted manner, natural selection did not act on neocortical systems independently of one another, which is a necessary condition for cognition to be massively modular. In this article, I argue that Quartz's argument fails to undermine the massive modularity hypothesis.

The spontaneous emergence of signaling has already been studied in terms of standard evolutionary dynamics of signaling games. Standard evolutionary dynamics is given by the replicator equations. Thus, it is not clear whether the results for standard evolutionary dynamics depend crucially on the functional form of the replicator equations. In this paper I show that the basic results for the replicator dynamics of signaling games carry over to a number of other evolutionary dynamics.

A common objection to adaptationist accounts of human emotions is that they ignore the influence of culture. If complex emotions like guilt, shame and romantic jealousy are largely culturally determined, how could they be biological adaptations? Dual inheritance models of gene/culture coevolution provide a potential answer to this question. If complex emotions are developmentally ‘scaffolded’ by norms that are transmitted from parent to offspring with reasonably high fidelity, then these emotions can evolve to promote individual reproductive interests. This paper draws on case studies of emotional development to illustrate how complex emotions satisfy these conditions. Many of the norms and parenting strategies influencing emotional development are absorbed during the early stages of life when a child is in primary contact with its parents and before the onset of complex cognition. These conditions make it likely that emotion-governing norms are transmitted vertically and with relatively little cognitive ‘contamination’.

Images come in many varieties, but for evidential purposes, photographs are privileged. Recent advances in neuroimaging provide us with a new type of image that is used as scientific evidence. Brain images are epistemically compelling, in part because they are liable to be viewed as akin to photographs of brain activity. Here I consider features of photography that underlie the evidential status we accord it, and argue that neuroimaging diverges from photography in ways that seriously undermine the photographic analogy. While neuroimaging remains an important source of scientific evidence, proper interpretation of brain images is much more complex than it appears.

As scientists begin to study increasingly complex questions, many have turned to computer simulation to assist in their inquiry. This methodology has been challenged by both analytic modelers and experimentalists. A primary objection of analytic modelers is that simulations are simply too complicated to perform model verification. From the experimentalist perspective it is that there is no means to demonstrate the reality of simulation. The aim of this paper is to consider objections from both of these perspectives, and to argue that a proper understanding and application of robustness analysis is able to resolve them.

Several quantitative techniques for choosing among data models are available. Among these are techniques based on algorithmic information theory, minimum description length theory, and the Akaike information criterion. All these techniques are designed to identify a single model of a data set as being the closest to the truth. I argue, using examples, that many data sets in science show multiple patterns, providing evidence for multiple phenomena. For any such data set, there is more than one data model that must be considered close to the truth. I conclude that, since the established techniques for choosing among data models are unequipped to handle these cases, they cannot be regarded as adequate.

It is a common assumption amongst theorists that the phenomenon of animal emotion supports the affect program theory of emotion. I argue that this assumption is mistaken by exploring two cases of animal emotion from studies in ethology: aggression in chimpanzees and fear in piping plovers. While the affect program theory fails to account for the cognitive complexity involved in each case, I do not argue for a cognitive theory of emotion. Instead, I suggest that paying attention to animal emotions helps the emotion theorist avoid the dichotomy between the extreme versions of the affect program theory and cognitive theories.

Ellis argues that certain essential properties of objects in the world not only determine the nature of these objects but also how they will behave in any situation. In this paper I will critique Ellis's essentialism from the perspective of the philosophy of chemistry, arguing that our current knowledge of chemistry in fact does not lend itself to essentialist interpretations and that this seriously undercuts Ellis's project. In particular I will criticize two key distinctions Ellis draws between internal vs. external properties and essential vs. accidental properties, showing that at the chemical level such distinctions are insupportable. If essential properties only exist at the level of sub-atomic physics, then Ellis's hopes that essentialism will provide a theoretical basis for a philosophy of chemistry are at best hopes for a very distant future, since the argument that chemical structure and dynamics can be explained at the quantum level derived is purely from analogy to much simpler systems than those chemists actually study. This suggests that we have very little scientific evidence that scientific essentialism is a viable ontology.

Two articles on the reduction of chemistry are examined. The first, by McLaughlin (1992), claims that chemistry is reduced to physics and that there is no evidence for emergence or for downward causation between the chemical and the physical level. In a more recent article, Le Poidevin (2005) maintains that his combinatorial approach provides grounding for the ontological reduction of chemistry, which also circumvents some limitations in the physicalist program.

Newton's methodology is significantly richer than the hypothetico-deductive model. It is informed by a richer ideal of empirical success that requires not just accurate prediction but also accurate measurement of parameters by the predicted phenomena. It accepts theory-mediated measurements and theoretical propositions as guides to research. All of these enrichments are exemplified in the classical response to Mercury's perihelion problem. Contrary to Kuhn, Newton's method endorses the radical transition from his theory to Einstein's. The richer themes of Newton's method are strikingly realized in a challenge to general relativity from a new problem posed by Mercury's perihelion.

A recent proposal by Norton (2003) to show that a simple Newtonian system can exhibit stochastic acausal behavior by giving rise to spontaneous movements of a mass on the dome of a certain shape is examined. We discuss the physical significance of an often overlooked and yet important Lipschitz condition the violation of which leads to the existence of anomalous nontrivial solutions in this and similar cases. We show that the Lipschitz condition is closely linked with the time reversibility of certain solutions in Newtonian mechanics and the failure to incorporate this condition within Newtonian mechanics may unsurprisingly lead to physically impossible solutions that have no serious metaphysical implications.

Mathematical idealizations are scientific representations that result from assumptions that are believed to be false, and where mathematics plays a crucial role. I propose a two stage account of how to rank mathematical idealizations that is largely inspired by the semantic view of scientific theories. The paper concludes by considering how this approach to idealization allows for a limited form of scientific realism.

In this paper we argue that the formalisms for decoherence originally devised to deal just with closed or open systems can be subsumed under a general conceptual framework, in such a way that they cooperate in the understanding of the same physical phenomenon. This new perspective dissolves certain conceptual difficulties of the einselection program but, at the same time, shows that the openness of the quantum system is not the essential ingredient for decoherence.

The literature on causal discovery has focused on interventions that involve randomly assigning values to a single variable. But such a randomized intervention is not the only possibility, nor is it always optimal. In some cases it is impossible or it would be unethical to perform such an intervention. We provide an account of ‘hard’ and ‘soft’ interventions and discuss what they can contribute to causal discovery. We also describe how the choice of the optimal intervention(s) depends heavily on the particular experimental setup and the assumptions that can be made.

Although scientific images have begun to receive significant attention from philosophers, one type of image has thus far been ignored: moving images. As techniques such as live cell imaging and videomicroscopy are becoming increasingly important in many areas of biology, however, this oversight needs to be corrected. Biologists often claim that there are relevant differences between video and static images. Most interesting is the idea that video images allow us to see causal relationships. By identifying the conditions that would be required for this to be true and showing that they are not satisfied at the micro level, I will show that videomicroscopy does not provide us with special access to causal information.

David Deutsch and others have suggested that the Many-Worlds Interpretation of quantum mechanics is the only interpretation capable of explaining the special efficiency quantum computers seem to enjoy over classical ones. I argue that this view is not tenable. Using a toy algorithm I show that the Many-Worlds Interpretation must crucially use the ontological status of the universal state vector to explain quantum computational efficiency, as opposed to the particular ontology of the MWI, that is, the computational histories of worlds. As such, any other interpretation that treats the state vector as representing real ontological features of a system can explain quantum speedup too.

It has been argued, partly from the lack of any widely accepted solution to the measurement problem, and partly from recent results from quantum information theory, that measurement in quantum theory is best treated as a black box. However, there is a crucial difference between ‘having no account of measurement’ and ‘having no solution to the measurement problem’. We know a lot about measurements. Taking into account this knowledge sheds light on quantum theory as a theory of information and computation. In particular, the scheme of ‘one-way quantnum computation’ takes on a new character in light of the role that reference frames play in actually carrying out any one-way quantum comptuation.

This paper argues that econometricians’ explicit adoption of identification conditions in structural equation modelling commits them to read the functional form of their equations in a strong, nonmathematical way. This content, which is implicitly attributed to the functional form of structural equations, is part of what makes equation structural. Unfortunately, econometricians are not explicit about the role functional form plays in signifying structural content. In order to remedy this, the second part of this paper presents an interpretation of the functional form based on Herbert Simon's definition of causal order. This begins to set out just what the functional form of structural equations represents.