Evolution has resulted in organisms that develop. For a long time the processes of evolution and development have been theoretically and empirically largely independent. This was epitomized by the gene selectionist view of evolution, which rendered the process of development largely epiphenominal to the evolutionary business of evolving the genome. However, there is now a growing awareness that an understanding of development is essential to a complete understanding of evolution and vice versa. These sessions address the contributions of this 'developmental synthesis' and some difficulties it faces.

Organized by: Kelly Smith & Roger Sansom

Roger Sansom, University of North Carolina at Chapel Hill
"Adaptive Development and Trends in Evolution"

I shall begin with Lewontin's requirements of quasi-independence and continuity for adaptive evolution. I shall argue that quasi-independence is a type of continuity and investigate the importance of continuity to evolutionary theory and the selection process that favors it. I agree with Lewontin that continuity has significant general adaptive value.
Our interest in quasi-independence lead naturally to an interest in modularity, because modular developmental systems are quasi-independent. The general adaptive value of quasi-independence raises the interesting possibility of a general trend in evolution towards more modular developmental systems. But there are other ideas out there that seem to be at odds with this. For example, Wimsatt has argued that complex adaptive systems will have entrenched traits. In essence, their adaptationist argument for this relies on these traits losing their continuity. There is significant empirical support for entrenchment (it would explain the widespread presence of vestigal traits in early development) as well as some interesting counterexamples. I shall attempt to incorporate these ideas into one coherent picture of the adaptive influences on the evolution of development.

Kim Sterelny, Victoria University of Wellington
"Modularity, Symbiosis and Evolvability"

In this paper, I explore the connections between inheritance systems and modularity. In doing so, I assume rather than argue for a multiple inheritance model. Any mechanism whereby members of generation N influence their offspring in ways that tend to make those offspring resemble their parents is an inheritance mechanism. But not all inheritance mechanism are of equal importance in the evolution of biological diversity and complex adaptation. Genetic inheritance is only one of the systems that supports cross-generation similarity. But it is one of exceptional importance. It is ancient. It is (almost) universal. And it is highly evolvable. On the basis of this assumption, I argue:
(i) The vertical transmission of symbionts is a also bona fide inheritance mechanism. Moreover, it is an evolutionarily significant mechanism.
(ii) Though symbiotic transmission is an inheritance system, it is one with a very different character to the transmission of genes. Genetic inheritance is an information-based system, whereas symbiotic transmission is a sample-based system.
(iii) Evolution based on sample-based inheritance has different evolutionary dynamics than those based on information. Sample-based inheritance has less variance. But is more robust and more modular, than inheritance based on information.
(iv) Sample-based inheritance has the potential to support evolutionary innovation. It is highly evolvable, essentially because samples generate their phenotypic effects in ways that are relatively independent of other developmental resources.
In short, this paper defends a connection between sample-based inheritance, evolvability and modularity.

Chi-hua Chiu and Gunter P. Wagner, Yale University
"Are There Absolute Limits to Causal Explanations in Evo-Devo?"

Evolutionary developmental biology is gaining momentum and brings with it the conceptual and methodological tools that enable us to finally begin to address "how" questions, such as how the architecture of development produces variational tendencies in some structures, e.g. the limb. The greatest challenge to evo-devo research is establishing a causal link between observed genetic differences and morphological change. Here, we propose criteria that connect developmental and phylogenetic data to study claims of causation in developmental evolution. The most critical step in establishing causality, however, is experimentally recreating a phenotypic difference that arose in evolution. Recent models derived from the theory of phenotype landscapes predict that at least for some classes of evolutionary transitions, this approach is not feasible due to the inability to identify critical genetic changes in evolution. We will explain the model and discuss the possible epistemological implications.