Paper long abstract:

In an STS-informed empirical research I explore how scientific knowledge is produced in today's computational neuroscience. Therefore, my case in point for the computerized modeling of cerebral functions is the Blue Brain Project (and the emerged Human Brain Project), located in Lausanne (Switzerland). The project claims to be not an artificial intelligence (AI) project. At the same time it tries to reverse engineer the whole human brain in a supercomputer with a bottom-up approach.

Social scientists have described computer simulations as an epistemic practice with a new quality of artificiality and as a way of "experimenting with theories" (Küppers/ Lenhard). Simulations use the language of physics and mathematics to describe the epistemic object. According to Hans-Jörg Rheinberger, "nature itself" (in a technological and scientific sense) only becomes real as a model.

As in the "situated epistemology" (Kay 2001) of cybernetics and AI, todays neuroscience and cognitive science develop new ideas of how to meet the challenges of the complexity of the brain. My paper deals with the question, how the constellation of a brain simulation can be epistemically active, especially when it comes to new data driven approaches? Do computer simulations have special capabilities in an epistemic process to call them the third pillar of science next to theory and experiment? Or has science already reach the 4th paradigm, proposed by the computer science optimist Jim Gray?