ABSTRACT
Until the early 1960s, the computer simulations used in morphogenesis problems had developed in an inchoate and divergent manner, based on issues that were essentially specific to professional mathematicians. It is important, therefore, to first reconstruct the factors that enabled these simulations to evolve towards greater biological real-ism. The simulations I will consider are of three types: geometry and probability-based, logic-based and pluriformalized. We will see how their implementation broke not only with the traditional uses of the computer as a calculator of models, but also with the pragmatic epistemologies of these models. These simulations, which were successfully adopted by biologists, were the first to establish closer links with work in the field, but they would not be the last. This series of initial intersections with the empirical opened the way to an era of convergences with many different dimensions. In this chapter, in particular, I will show that certain biologists (such as Dan Cohen and Jack B. Fisher) were ultimately able to make good use of simulation once they managed to accentuate its ability to produce a representation on a geometrical level – albeit at the cost of reducing their ability to take the temporal heterogeneity of plant-growth rules into consideration.
