In this article I scrutinize the question whether clinical medicine, in order to be considered a hermeneutical enterprise, must be thought of as a reading of different "texts." Three different proposals for a definition of the concept of text in medicine, suggested by other hermeneuticians, are discussed. All three proposals are shown to be unsatisfying in various ways. Instead of attempting to find a fourth definition of the concept of text suitable to a hermeneutics of medicine, I then try to show that the assumption that one needs to operate with the concept of text in order to develop a hermeneutics of medicine is false. Clinical interpretation can be shown to essentially consist in a dialogical hermeneutics, the pattern of which can be found in the philosophy of Hans-Georg Gadamer. This kind of hermeneutics is not a methodology of text reading, but an ontological, phenomenological hermeneutics in which understanding is a necessary feature of the being-together of human beings in the world. This being-together in and through language takes on a peculiar form in the clinical encounter, since the medical meeting is typically characterized by an asymmetrical enstrangement and has a specific goal - health for the patient - absent in other forms of hermeneutics. Central issues of Gadamer's philosophy, e.g. "fusion of horizons," are shown to fit the structure of clinical practice.