Professor Whitty has endorsed the consensus that research into education is empirical social science, distinguishing "educational research" which seeks directly to influence practice, and "education research" that has substantive value but no necessary practical application. The status of the science here is problematic. The positivist approach is incoherent and so supports neither option. Critical educational science is virtually policy-inert. The interpretive approach is empirically sound but, because of the value component in education, does not support education research either, or account for this component. A solution to the latter problem is sought in the debate between Carr and Hirst on the relationship between philosophy and education. This shows Carr making claims that rely on a conception of philosophy that he rejects, while Hirst insists on this conception, uses it to justify practical claims, but denies that this is possible. To achieve a practically relevant analysis of educational research, both need to include second-order, normative, conceptual enquiry into the philosophies that drive educational policy-making and partly regulate teaching methodology. Deweyan, first-order, "reflective practice" needs, then, to be supplemented with second-order reflection. Educational research is philosophy- not science-driven, and is value-led. Consequently, it has the status not of scientific discovery but of practical recommendation. (Contains 21 notes.)