Coping With Strongly Coupled String Theory

String theory, if it describes nature, is probably strongly coupled. As a result, one might despair of making any statements about the theory. In the framework of a set of clearly spelled out assumptions, we show that this is not necessarily the case. Certain discrete gauge symmetries, combined with supersymmetry, tightly constrain the form of the effective action. Among our assumptions are that the true ground state can be obtained from some perturbative ground state by varying the coupling, and that the actual numerical value of the low energy field theoretic coupling ${g^2 \over 4\pi}$ is small. It follows that the low energy theory is approximately supersymmetric; corrections to the superpotential and gauge coupling function are small, while corrections to the Kahler potential are large; the spectrum of light particles is the same at strong as at weak coupling. We survey the phenomenological consequences of this viewpoint. We also note that the string axion can serve as QCD axion in this framework (modulo cosmological problems).