The effects of generalist and specialist scientists on scientific discovery

Thumbnail Image
Issue Date
Journal Title
Journal ISSN
Volume Title
The current scientific education leads up to become a scientific specialists. This means the generalist scientist is in the minority and their contribution to science may not be as large as it could be. In this thesis the effects of different distributions of these two scientist types is measured through agent-based modelling. In this model the scientists observe the situations and behaviours of a cognitive system represented by a finite state transducer of which they try to describe the internal structure. Depending on how many correct predictions their theories make, the agents become satisfied. This satisfaction is the main measurement of progress within the simulations. The differences between the two modelled scientist types lies in the way they generate their theories and in how much of the previously created theories they can see. Besides the progress in satisfaction the agents and their theories are measured on accuracy and complexity — measured in size of the finite-state transducer. During the implementation process a lot of concessions had to be made and because of this some of the observed effects are only artefacts of the implementation. In the end, the generalist and specialist agents each showed advantages and disadvantages on the speed of scientific discovery and the agents are far from perfect. The effect each scientist has is therefore not defining for the quality of science and it is up to preferences of the individual scientist to determine their path in science.
Faculteit der Sociale Wetenschappen