Popper's philosophy has been often misrepresented in the debate about the scientific status of psychoanalysis by pseudo‐Popperians who use the idea of falsifiability without rejecting positivism and inductive logic. The debate itself is misconceived because it is based on the outdated positivist assumption that what is not science is not important. Partially submerged by this, however, is a very important debate about method. The issues are clearly focused by what is here called ‘Hill's problem’, that of finding a common framework in which the contributions of organic, psychodynamic and social psychiatry can be discussed and applied to clinical problem solving.Will's paper brings out many fundamental problems. This reply discusses misapprehended aspects of Popper's world view and introduces his plural interactionalist metaphysical research programme. It demonstrates the close similarity of Popper's view of science which Lorenz calls ‘hypothetical realism’ with Bhaskar's ‘transcendental realism’ which Will advocates for pyschoanalysis. The differences hinge on concepts of explanation and the methods of theory testing. It is here argued that Popper's view is better fitted for the task Will proposes. The resemblance of ‘hypothetical realism’ to the clinical tradition further demonstrates its applicability