This was an argument against vitalism. Vitalism was a basically metaphysical doctrine arguing on the level of ontology. It claimed that living processes contain a nonphysical force or principle that make them what they are, which can never be explained by the physical sciences. Therefore, in Mayr's text, the assumption of the genetic program was also a metaphysical argument, working on the level of ontology. It explained how is it possible that living beings come into existence, and their way of existing in the world. In terms of theory building, the idea of the genetic program replaced intentionality in nature and the vitalist non-physical forces. It was therefore negative metaphysics that Mayr was arguing for. With the physically plausible assumption of a genetic program it was no longer necessary to believe in nonphysical forces and intentions to explain the apparent functional organization and goal-directed behavior of organisms. Jacob and Monod [15] were comparatively more constructive in their approach. They saw how such a genetic program could be conceptualized: as a sequence of regulatory steps whereby genes are regulated by the products of other genes.
But was this discursive move to the programs containing genome just a step towards dropping unnecessary metaphysical ballast and therefore in itself ontologically or metaphysically ‘innocent’ or neutral? I do not think so. In order to see this, we need to advance another 45 years in the history of biology and look at the ideas of contemporary systems biology. Kunihiko Kaneko, a Japanese biologist and influential theorist of complex molecular systems, summarizes the current approach to explaining the emergence of relatively stable and regular developmental pathways of organisms in a very simple way as follows:
‘Thus the situation is one of mutual influence, not unidirectional causation. Hence, although the genes can be thought of as in some sense controlling such processes, in fact it is not true that an understanding of the genes alone is sufficient for their complete description. For example, even if we were somehow able to obtain the DNA of a dinosaur, unless we also knew the initial conditions of the cellular composition that allow their proper expression of genes, we would not be able to create a Jurassic Park. The conclusion we reach from these considerations is that… we should be studying models of interactive dynamics. Then, we should inquire whether, within such dynamics, the asymmetric relation between two molecules is generated so that one plays a more controlling role and therefore can be regarded as the bearer of genetic information’ [17, p 20].
If we compare this idea that Kaneko is outlining, referring to vast experimental evidence and to mathematical models, with the image inherent in the ‘genetic program’, several important differences become obvious. The relation between different molecules and processes in the cell are seen as mutual influence, instead of a unidirectional causality contained in sequences of linear if-then events. Parts interact in many ways, loops of causal influence going forward and backward, branching in many directions and making the system as a whole relatively open or relatively closed. The division of roles within such a system is not a precondition, but must itself be explained as a result of the interactive dynamics of the system. Therefore, the apparent specialization of DNA as the bearer of genetic information, and the many very important roles singular genes can play in the development of the organism, are products of the interrelation of the parts of the system and their dynamics. This is the second striking difference to the genetic program approach. There, it was thought that a causal program is a precursor of development in the shape of the sequential composition of the DNA polymer. Thirdly, a regularity of developmental events that could be described as something like a program is to be located on the level of those interactive dynamics, not on the level of one component of the system. In terms of the distinction between genotype and phenotype, the assumption of systems biology is that the program (if anybody still wants to talk of programs) is a phenotypic regularity. What behaves regularly in foreseeable and reproducible sequences of events is the whole organism within an environment, not one isolated molecule.2
Molecular genetics, particularly in the context of developmental biology and genomics, has contributed to this enlargement of the picture. A wide variety of mechanisms that enable the cell to use DNA sequences in different ways have been discovered. The active RNA molecules (still suggestively called ‘messenger’ RNA) are compiled in much more complicated ways, and most of the RNA molecules (MicroRNAs) are no templates for proteins at all. DNA sequences that code for proteins (containing the genes in a classical sense) and some proteins are multifunctional. Which effects will be realized depends on a multitude of other factors, and sometimes on spatial information as well, i.e. on the place within the cell or a multicellular network. Some genes overlap, some genes can be spliced in multiple ways, depending on the situation, and the resulting RNA variants will lead to different proteins that are all related to the same DNA stretch. Sometimes the cell uses fractions of one and sometimes fractions of the other of the two single DNA strands as a template for producing a functional RNA. There are also switches to alternative reading frames, i.e. the shifting of the three-letter code by one, resulting in different sequence information. There is sometimes even post-transcriptional editing of mRNA, i.e. the introduction of changes in the sequence of an mRNA molecule after its composition, which also leads to a different amino acid sequence of the protein being built from it [14, 19-22]. Genes are multifunctional [23], and therefore they can no longer be considered as independent factors in a chain of events. But this is precisely what the idea of the ‘genetic program’ suggested.
Taking these and other phenomena into account, Eva Neumann-Held [24] has reconsidered the very concept of the gene from a systems perspective. If we still want to call what explains the biosynthesis of a particular type of protein in a cell a ‘gene’, we can no longer say that one stretch of DNA is responsible. It is rather a range of factors, interacting with DNA and with each other, and processes sometimes transgressing the boundaries of the cells and the body, that are actually contributing. This set of contributing factors includes DNA, but also much else. Neumann-Held bases her reflection on a groundbreaking book by Susan Oyama from 1985 that has the title ‘The Ontogeny of Information’ [25] and has inspired many authors to new formulations under the umbrella term of a ‘developmental systems approach’ [26, 27]. The key idea was a new attempt to theorize development. Previously, we thought development was basically a result of two different information resources, one internal and genetic, the other external, i.e. social, environmental, or cultural. The divide between these two information resources has materialized in the ‘nature or nurture’ debate in developmental psychology. One school emphasized the genetic contributions (sociobiology, evolutionary psychology, etc.), the other more the social and cultural factors. Oyama's point was that development is better seen as an interaction of both,