Viorel Pâslaru (University of Dayton)
I examine in this article the role of descriptions of recurrent mechanisms in explanation and prediction. I show that the new mechanistic philosophy is committed to the symmetry explanation-prediction. Contrary to this thesis, I examine the work of ecologist D. Tilman on the mechanism of competition and show that: descriptions of mechanisms for explanation are different from descriptions for predictions; descriptions of mechanisms for predictions allow novel predictions, while descriptions of mechanisms for explanation do not; and predictions are improved not by adding details to the descriptions of mechanisms for explanation, but to descriptions of mechanisms for predictions.
