Standard statistical measures of strength of association, although pioneered by Pearson deliberately to be acausal, nowadays are routinely used to measure causal efficacy. But their acausal origins have left them ill suited to this latter purpose. I distinguish between two different conceptions of causal efficacy, and argue that: (1) Both conceptions can be useful; (2) The statistical measures only attempt to capture the first of them; (3) They are not fully successful even at this; (4) An alternative definition based more squarely on causal thinking not only captures the second conception, but also can capture the first one better too.

Publications in contemporary science journals often include figures like graphs, diagrams, photographs, and MRIs, which are presented as support for the hypothesis the author is defending. As a first step to explaining how figures contribute to confirmation, I present an account of visual representation and use examples to show how the visual format is involved in the support those figures provide the authors’ conclusions. I then show that attempts to explain what figures contribute to scientific arguments without analyzing them as visual representations (for example by translating them into linguistic representations) will not succeed.

For nearly as long as the word ‘correlation’ has been part of statistical parlance, students have been warned that correlation does not prove causation, and that only experimental studies, e.g., randomized clinical trials, can establish the existence of a causal relationship. Over the last few decades, somewhat of a consensus has emerged between statisticians, computer scientists, and philosophers on how to represent causal claims and connect them to probabilistic relations. One strand of this work studies the conditions under which evidence accumulated from non-experimental (observational) studies can be used to infer a causal relationship. In this paper, I compare the typical conditions required to infer that one variable is a direct cause of another in observational and experimental studies. I argue that they are essentially the same.

Critics of functional explanations in social science maintain that such explanations are illegitimate unless a mechanism is specified. Others argue that mechanisms are not necessary for causal inference and that functional explanations are a type of causal claim that raise no special difficulties for testing. I show that there is indeed a special problem that confronts testing functional explanations resulting from their connection to second-order causal claims. I explain how mechanisms can resolve this difficulty, but argue that this does not provide support for methodological individualism since it is not necessary that the mechanisms be described in terms of individual interactions.

Jaegwon Kim has argued that if psychological kinds are multiply realizable then no single psychological theory can describe regularities ranging over psychological states. Instead, psychology must be fractured, with human psychology covering states realized in the human way, martian psychology covering states realized in the martian way, and so on. I show that even if one accepts the principles that motivate Kim's argument, his conclusion does not follow. I then offer a dilemma that forces Kim to concede the possibility of a unified psychology. I close with a discussion of what, according to Jerry Fodor, is “really bugging” Kim.

The aim of this paper is to introduce the instrumental variables technique to the discussion about causal inference in econometrics. I show that it may lead to causally incorrect conclusions unless some fairly strong causal background assumptions are made, assumptions which are usually left implicit by econometricians. These assumptions are very similar to, albeit not identical with, James Woodward's definition of an ‘intervention’. I discuss similarities and differences of the two points of view and argue that—understood as a practical method of causal inference—the set presented here is superior.

I consider the following questions, central to recent disagreements between Longino and Kitcher: Is it constitutive of making judgments of the cognitive acceptability of theories that they are made under certain social relations that embody specific social values that have been cultivated among investigators (Longino)? Or is making them (sound ones) just a consequence of social interactions that occur under these relations (Kitcher)? While generally endorsing the latter view, I make a distinction, not made by the philosophers under discussion, between sound acceptance and endorsement of a theory, and argue that Longino's view applies to endorsement.

If feminists argue for the irreducibility of the social dimensions of science, then they ought to embrace the idea that feminist and non-feminist scientists are not in collaboration, but in fact defend different interests. Instead, however, contemporary feminist science studies literature argues that feminist research improves particular, existing scientific enterprises, both epistemically (truer claims) and politically (more democratic methodologies and applications). I argue that the concepts of empirical success and democracy at work in this literature from Longino (1994) and Harding (1996), to Longino (2002), Gilbert and Rader (2001), and Keller (2001) are not sufficiently critical, and fail to do justice to the truly revolutionary work done by feminist scientists. I offer the beginnings of an epistemology of dissensus (as opposed to consensus), using the work of Haraway (1978), Lyotard (1984, 1998), and Ziarek (2001). How would such an epistemology relate to feminist discussions of the possibility of democratic, responsible knowledge?

Several feminist philosophers of science have tried to open up the possibility that feminist ethical or political commitments could play a positive role in good science by appealing to the Duhem-Quine thesis and underdetermination of theories by observation. I examine several different interpretations of the claim that feminist values could play a legitimate role in theory justification and show that none of them follow from a logical gap between theory and observation. Finally, I sketch an alternative approach for defending the possibility that feminist political commitments could play a legitimate role in science.

A number of recent biologists have used multilevel selection theory to help explain the major transitions in evolution. I argue that in doing so, they have shifted from a ‘synchronic’ to a ‘diachronic’ formulation of the levels of selection question. The implications of this shift in perspective are explored, in relation to an ambiguity in the meaning of multilevel selection. Though the ambiguity is well-known, it has never before been discussed in the context of the major transitions.

Samir Okasha argues that clade selection is an incoherent concept, because the relation that constitutes clades is such that it renders parent-offspring (reproduction) relations between clades impossible. He reasons that since clades cannot reproduce, it is not coherent to speak of natural selection operating at the clade level. We argue, however, that when species-level lineages and clade-level lineages are treated consistently according to standard cladist commitments, clade reproduction is indeed possible and clade selection is coherent if certain conditions obtain. Despite clade selection's logical coherence, however, we share some of Okasha's pessimism. Whether or not clades are a unit of selection is ultimately a question of empirical support and theoretical import.

Evolutionary convergence is often appealed to in support of claims about multiple realization. The idea is that convergence shows that the same function can be realized by different kinds of structures. I argue here that the nature of convergence is more complicated than it might appear at first look. Broad claims about convergence are made by biologists during general discussions of the mechanisms of evolution. In their specialized work, though, biologists are often more limited in the claims they make. I will examine a standard example to show how claims about convergence can be oversimplified.

This paper examines the empirical aspect of the debates over group selection. I argue that specific narrow readings of the notion of ‘group’ result in a deficient experimentation of the group selection process. Criteria for empirical testability are presented and used to reexamine two well-known experiments of group selection. I argue that the former holds a narrow image of ‘group’ that does not distinguish group selection from selection at other levels; while the latter holds a multifaceted image of ‘group’ that offers a way to achieve substantive progress in the debate. However, most supporters and opponent of group selection maintain a narrow and restricted image of ‘group’, which implicitly undermines informative details regarding population dynamics and population structure. These details may improve our understanding of the evolution of altruism and the evolution of sociality. To conclude, adopting a wide-ranging meaning of ‘group’ may empirically advance the group selection debates.

Background theories in science are used both to prove and to disprove that theory choice is underdetermined by data. The alleged proof appeals to the fact that experiments to decide between theories typically require auxiliary assumptions from other theories. If this generates a kind of underdetermination, it shows that standards of scientific inference are fallible and must be appropriately contextualized. The alleged disproof appeals to the possibility of suitable background theories to show that no theory choice can be timelessly or noncontextually underdetermined: Foreground theories might be distinguished against different backgrounds. Philosophers have often replied to such a disproof by focussing their attention not on theories but on Total Sciences. If empirically equivalent Total Sciences were at stake, then there would be no background against which they could be differentiated. I offer several reasons to think that Total Science is a philosophers’ fiction. No respectable underdetermination can be based on it.

Against the well-known objection that in the history of science there are many theories that are successful but false, Psillos offers a three-pronged defense of scientific realism as the best explanation for the success of science. Focusing on these, I criticize Psillos’ defense, arguing that each prong is weakened when we recognize that according to realist rebuttals of the underdetermination argument and versions of empiricism, realists are committed to accounting for the explanatory success of theories, not their mere empirical adequacy or instrumental reliability. I conclude by indicating how ‘explanationist’ realism might be recast to accommodate my arguments.

The Pessimistic Induction from falsity of past theories forms a perennial argument against scientific realism. This paper considers and rebuts two recent arguments—due to Lewis (2001) and Lange (2002)—to the conclusion that the Pessimistic Induction (in its best known form) is fallacious. It re-establishes the dignity of the Pessimistic Induction by calling to mind the basic objective of the argument, and hence restores the propriety of the realist program of responding to PMI by undermining one or another of its premises.

Two strong arguments have been given in favor of the claim that no selection process can play a role in explaining adaptations. According to the first argument, selection is a negative force; it may explain why the eliminated individuals are eliminated, but it does not explain why the ones that survived (or their offspring) have the traits they have. The second argument points out that the explanandum and the explanans are phenomena at different levels: selection is a population-level phenomenon, whereas adaptation occurs on the individual level. Thus, selection can explain why individuals in a certain population have a certain trait, but it cannot explain why a certain individual has this trait. After pointing out that both arguments ignore the significance of the limitation of environmental resources, I will construe a positive argument for the claim that cumulative selection processes can, indeed, play a role in explaining adaptations.

Evolutionary processes such as natural selection and random drift are commonly regarded as causes of population-level change. We respond to a recent challenge that drift and selection are best understood as statistical trends, not causes. Our reply appeals to manipulation as a strategy for uncovering causal relations: if an appropriately controlled manipulation of variable A results in a systematic change in variable B, then A is a cause of B. We argue that selection and drift can be manipulated to produce different kinds of population-level change. They should therefore be regarded as causes.

Evolutionary theorists have encountered difficulty in explaining how altruistic behavior can evolve. I argue that these theorists have made this task more difficult than it needs to be by focusing their efforts on explaining how nature could select for a strong type of altruism that has powerful selection forces working against it. I argue that switching the focus to a weaker type of altruism renders the project of explaining how altruism can evolve significantly less difficult. I offer a model of weak altruism that can avoid many of the difficulties that evolutionary accounts of altruism have traditionally faced.

In this paper, it is argued that selection and drift might be distinct. This contradicts recent arguments by Brandon (2005) and Matthen and Ariew (2002) that such a distinction “violates sound probabilistic thinking” (Matthen and Ariew 2002, 62). While their arguments might be valid under certain assumptions, they overlook a possible way to make sense of the distinction. Whether selection and drift can be thought of as causally distinct, I argue, depends on the source of probabilities in natural selection. In particular, if the probabilities used in defining fitness values are at least partly a result of abstracting from or ignoring certain features of the environment, then selection and drift might in fact be causally distinct.