My dissertation articulates the main contours of a humanistic philosophy of social inquiry, appropriate to a “deliberative” or “dialogic” ideal of democracy. Such a conception makes certain demands on how we ought to conduct inquiry into the interests that influence and are shaped by decision-making. I specify what these demands are, and explore a range of practices that promise to help us better to meet them. Doing this helps to elucidate the manifold ways in which inadequate attention to the proper character of democratic theory as areflexive social practice can contribute to the construction of theories, and to the reproduction of practices, that bolster, rather than undermine, extant relations of domination. To try to avoid this potential, a reflexive approach openly recognizes and evaluates the influence of its practitioners’ values, interests, and cultural traditions – and not just at the level of “normative” theorizing but also at the levels of “empirical” description and explanation too. Such an approach highlights the need to include affected constituencies in the production and evaluation of information and knowledge pertinent to decision-making, a requirement that follows from the claim that decision-making should aim at “equal (deliberative) interest representation.” To elaborate and defend these views, I critically evaluate central features of American discourse surrounding U.S. “democracy promotion”, as well as U.S. support for authoritarianism, focusing particularly on Latin America. And I critically assess how specific struggles for meaningful participation have been historically investigated and portrayed. In doing so, I explore how certain forms of inquiry, and the reliance on certain conceptions of expertise, have sometimes served to obscure, and sometimes to bolster, illegal, undemocratic, and violent U.S. policies and interventions. Finally, I explore how certain practices could mitigate the potential for domination that inheres in these activities, considering how public opinion formation, legislation, and implementation might each come more closely to approximate the ideal of equal (deliberative) interest representation.