Although the process of scientific discovery is not fully understood, there are two aspects of it that are important. One involves such activities as theorizing, gaining insights, speculating, using intuition, and extrapolating. This part can be called ‘extension’ (in the sense of adding to current ideas). The other aspect involves collecting data, noticing patterns, identifying processes, and exhaustively describing given phenomena. This can be called ‘consolidation’ (in the sense of filling in the gaps in what is known). While linguists are always engaged in both types of activities (with different focus during different eras), within generative work there has been special emphasis on ‘extension’, on formulating theory and using intuition; indeed, this emphasis has been one of the attractions of generative grammar. Complementary to this emphasis on extension there has been developing an increased interest in ‘consolidation’, in doing field work and conducting phonological experiments, for example. My purpose here is to discuss experimental phonology as a consolidative activity, and to show how the consolidative function of experiments puts special requirements on the way they are run. A secondary purpose is to place consolidative work within a moderately general framework.