In 1959, scientist and novelist C.P. Snow labeled art and science as two conflicting cultures, unable to communicate. Over the last fifteen years the landscape has dramatically changed. Collaborative research projects between artists and scientists have been flourishing in the UK, putting Snow's thesis to the test. This has spawned a debate on the value of art-science research projects and how it can be assessed and validated. This dissertation attempts to contribute to this debate through the investigation of the value of art-science collaborations in terms of knowledge. To this purpose, a theoretical model of knowledge construction has been adopted to test the hypothesis that art and science collaborations can create knowledge. The model, for the first time in this research, has been applied to an art-science setting and the hypothesis has been tested against a case study, to which the knowledge-building theory has proved to be only partially applicable. Therefore an alternative model has been developed. Since this model has been devised on the basis of one case study, it needs to be validated by further research.