We argue that Horgan's program for nominalizing science fails, because its translation of quantitative statements destroys the inferential structures of explanations, predictions and retrodictions of nonquantitative scientific facts.

I point out that conceptions of particles as mathematical, or quasi mathematical, entities have a longer history than Resnik notices. I argue that Resnik's attack on the distinction between mathematical and physical entities is not deep enough. The crucial problem for this distinction finds its locus in the numerical indeterminancy of elementary particles. This problem, traced by Heisenberg, emerges from the discovery of antimatter.