A main implication of C.E. Ayres tool-combination principle is that the goal of technical progress is best served by a non-proprietary, open science public policy. Joseph Schumpeter claimed that new combinations are consequential only when they have been successfully commercialized. The capacity to privatize knowledge is, moreover, a powerful stimulus to innovation. This paper reexamines the Ayresian and Schumpeterian positions using evidence from the Bayh Dole experiment. The Bayh Dole Act, which gave universities title to inventions resulting from federally-sponsored research, created a laboratory wherein the trade-offs between diminution of the appropriable knowledge fund (due to patenting) and incentives to commercialization can be appraised.