Abstract
For a physical theory to be useful, we should be able to extract from it quantitative predictions for physical observables. With most theories, this process of practical application requires some approximation method that is both tractable and adequately accurate. The lack of a useful approximation scheme can seriously impede progress in a field of research. Molecular quantum mechanics and critical phenomena, for example, have both suffered from this problem during periods of their development. The physics of strong interactions is in such a period today. Naturally, our inability to extract useful predictions from a promising theory is a strong incentive to develop novel approximation techniques.