This article focuses on the amount of rationality we can expect from the outcomes of democratic procedures. I will examine two traditions within democratic theory that posit competing and opposing conceptions of participation. These are: (A) theories of democratic discourse, and (B) disequilibrium theories of social choice. The stated arguments seem to be mutually exclusive and theorists who work within one field tend to ignore the conclusions reached in the other. What has not been noticed in the literature to date is that both theories present almost identical assumptions concerning the requirements for a fair procedure; they both demand equal access to debate, the absence of a powerful agenda setter, unrestrained access to raise and object to amendments and so on. Despite these similarities, one argument suggests that the procedure of democratic discourse will yield morally legitimate outcomes, while the other suggests that it has a tendency to produce chaos and disequilibrium. At heart, the difference between the theories is that supporters of discourse argue that the more democratic the system is, the better, while social choice theorists suggest that more democracy leads to instability and arbitrary results. Both cannot be correct and a resolution to the puzzle would seem crucial, especially considering the current support afforded both discourse theory and social-choice theory in the discipline.