The use of informational terms is widespread in molecular and developmental biology. The usage dates back to Weismann. In both protein synthesis and in later development, genes are symbols, in that there is no necessary connection between their form (sequence) and their effects. The sequence of a gene has been determined, by past natural selection, because of the effects it produces. In biology, the use of informational terms implies intentionality, in that both the form of the signal, and the response to it, have evolved by selection. Where an engineer sees design, a biologist sees natural selection.

In many texts on evolution the reader will find a characteristic depiction of inheritance and evolution, one showing the generations of an evolving population linked only by a causal flow from genotype to genotype. On this view, the genotype of each organism in this population plays a dual role as both the motor of individual development and as the sole causal channel across the generations (cf., e.g., Maynard Smith 1993, Fig. 8). This picture is known to be literally false. In many species, parents exert direct causal influence on their offspring, and some of those influences cause the offspring to resemble the parent. For example, a butterfly that lays her eggs on the same plant host on which she hatched thereby exerts a causal influence on her offspring, and one apt to cause them to resemble her more than they would, had she chosen a plant of a different species. None of this is at all controversial, but it poses a puzzle. If the “Weissmanian” conception is literally false, why is it seen as a perspicacious representation of evolution?

Maynard Smith is right that one of the most striking features of contemporary biology is the ever-increasing prominence of the concept of information, along with related concepts like representation, programming, and coding. Maynard Smith is also right that this is surely a phenomenon which philosophers of science should examine closely. We should try to understand exactly what sorts of theoretical commitment (if any) are made when biological systems are described in these terms, and what connection there is between semantic descriptions in biology and in other domains.

Maynard Smith notes that he provides a natural history and not a philosophical analysis of the use of concepts of information in contemporary biology. Just a natural history, however rich, would do little to resolve the ongoing controversy about the role of these concepts in biology. None of the disputants deny that the biological use of these concepts is pervasive. The dispute is about whether these concepts—and the framework in which they are embedded—continue to be of explanatory value in contemporary biology. Fortunately, Maynard Smith does much more than provide a natural history: his contribution is also a sustained attempt to justify many (though not all) of the uses of information in biology.

Inevitably, much of this reply will concern points on which I disagree with one or other of the commentators. I should start, therefore, by saying that I am delighted to find that their thinking is in many ways similar to my own, and that a serious attempt is under way to clarify the concept of information in biology.

This paper considers the implications for the relational-substantival debate of observations of parity nonconservation in weak interactions, a much neglected topic. It is argued that ‘geometric proofs’ of absolute space, first proposed by Kant (1768), fail, but that parity violating laws allow ‘mechanical proofs’, like Newton's laws. Parity violating laws are explained and arguments analogous to those of Newton's Scholium are constructed to show that they require absolute spacetime structure—namely, an orientation—as Newtonian mechanics requires affine structure. Finally, it is considered how standard relationist responses to Newton's argument might respond to the new challenge of parity nonconservation.

Biological knowledge does not fit the image of science that philosophers have developed. Many argue that biology has no laws. Here I criticize standard normative accounts of law and defend an alternative, pragmatic approach. I argue that a multidimensional conceptual framework should replace the standard dichotomous law/accident distinction in order to display important differences in the kinds of causal structure found in nature and the corresponding scientific representations of those structures. To this end I explore the dimensions of stability, strength, and degree of abstraction that characterize the variety of scientific knowledge claims found in biology and other sciences.

I develop an account of predictive similarity that allows even Antirealists who accept a correspondence conception of truth to answer the Realist demand (recently given sophisticated reformulations by Musgrave and Leplin) to explain the success of particular scientific theories by appeal to some intrinsic feature of those theories (notwithstanding the failure of past efforts by van Fraassen, Fine, and Laudan). I conclude by arguing that we have no reason to find truth a better (i.e., more plausible) explanation of a theory's success than predictive similarity, even of its success in making novel predictions.

The physical and/or intrinsic connection approach to causation has become prominent in the recent literature, with Salmon, Dowe, Menzies, and Armstrong among its leading proponents. I show that there is a type of causation, causation by disconnection, with no physical or intrinsic connection between cause and effect. Only Hume-style conditions approaches and hybrid conditions-connections approaches appear to be able to handle causation by disconnection. Some Hume-style, extrinsic, absence-relating, necessary and/or sufficient condition component of the causal relation proves to be needed.

Terence Horgan and John Tienson claim that folk psychological laws are different in kind from basic physical laws in at least two ways: first, physical laws do not possess the kind of ceteris paribus qualifications possessed by folk psychological laws, which means the two types of laws have different logical forms; and second, applied physical laws are best thought of as being about an idealized World and folk psychological laws about the actual world. I argue that Horgan and Tienson have not made a persuasive case for either of the preceding views.

Despite the results of David Lewis, Peter Gärdenfors, and others, showing that imaging and classical conditionalization coincide only in the most trivial probabilistic models of belief revision, it turns out that imaging on a proposition A can always be described via Popper function conditionalization on a proposition that entails A. This result generalizes to any method of belief revision meeting certain minimal requirements. The proof is illustrated by an application of imaging in the context of the Monty Hall Problem.