A plausible interpretation of the experimental evidence is that energy confinement in tokamaks is governed by two separate considerations: (1) the need for resistive MHD kink-stability, which limits the permissible range of current profiles-and therefore normally also the range of temperature profiles; and (2) the presence of strongly anomalous microscopic energy transport near the plasma edge, which calibrates the amplitude of the global temperature profile, thus determining the energy confinement time tau E. The present analysis begins by identifying a hypothetical model of tokamak confinement that is designed to take into account the conflict between Te(r)-profile shapes arising from microscopic transport and J(r)-profile shapes required for gross stability. On the basis of this model, a number of hypothetical lines of advance are developed. Some TFTR experiments that may point the way to a particularly attractive type of tokamak reactor regime are discussed.