This dissertation is concerned with the “research-practice gap,” the notion that the development of scientific knowledge in education has been relatively ineffective in addressing the practical problems of schooling. Research and Practice have been differentiated on the basis of actors, practices, expertise, organizational contexts, interests, and more; such are the apparent factors driving the Research-Practice Gap. Yet, many education researchers, policymakers, and practitioners are hard at work in attempts to bridge the two.In imagining the bridging of the Research-Practice Gap as a precursor to the closure of the Achievement Gap, Research is figured as the external source of a pipeline of salutatory scientific interventions, which, provided appropriate translational work, can flow effectively into Practice. If proven effective, these interventions can be “scaled up” as a significant support of racial equality in the US. An investigation of the Research-Practice nexus brings up questions that are crucial not only to the operation of US education, but also the organization of social life in general, and the organization of a democratic society in particular. Can people meaningfully and effectively work together across difference? Can a nation professedly built on diversity be meaningfully United? This work is an investigation of communication, cooperation, and trust in the project of making American education scientific, itself a project of making America. Central to this dual scientific and political project are trust and objectivity. Trust and objectivity are complementary modes of organizing social relations: Where a lack of trust threatens to sever lines of cooperation, objectivity enters to tie them back together. Historians of science have remarked that objectivity is distinctive in its “exclusion of judgment,” as a “struggle against subjectivity” (Porter 1995, p.ix). As such, the governance of American education has largely turned to objectivity as a means of making decisions without a decider. As Porter puts it, “in science, as in political and administrative affairs, objectivity names a set of strategies for dealing with distance and distrust” (ibid.). As a set of methods for eliminating individual subjectivity, objectivity becomes a means of rendering unity out of diversity, of identifying signals amidst noise. In this vein, objectivity is asked to not only produce education research as a diverse but unified scientific community, but also to produce US education as a diverse but democratic system of schooling. I put forward three key distrusts which are managed in-and-through the 21st century American project of governing education through science. The first distrust is a bipartisan distrust of politics in education. This distrust of politics drives legislative appeals to science as a source of objectivity, as exemplified in the 2001 No Child Left Behind Act’s recourse to and codification of “scientifically based research.” It also sets the stage for the latter two distrusts, which are the focus of this dissertation. In finding it necessary to define Scientifically Based Research, NCLB points to the second distrust: governmental and academic distrust of educational research. This line of distrust takes the Research-Practice Gap to be the fault of Research—insufficiently objective; incoherent; easily captured by the political, marketable, and fashionable. It is met by another line of distrust which holds the Research-Practice Gap to be the fault of Practice: researchers’ distrust of practitioners’ ability to faithfully implement the solutions offered by Research. This work is concerned with describing education researchers’ management of these two latter distrusts by way of two strategies of objectivity: reliability and fidelity. Through demonstrations of reliability, researchers constitute themselves as a scientific community, as Research, authorized to study and govern the work of Practice. In the name of fidelity, researchers turn the work of Practice into “implementation,” permitting the capture of variation against a Research-given standard, and again capturing Practice within the jurisdiction of Research. In an ethnographic account of the real-time activities of producing fidelity and reliability in a randomized controlled evaluation of an educational intervention, I use a linguistic anthropological analysis to describe the role of such interactions in maintaining the organization of the Research-Practice nexus, extending to the maintenance of its troublesome Gap. In concluding, I expand on the implications of my analysis in suggesting that a scientifically democratic system of education must be a democratically scientific system of education.