With regard to the question what is to be tran$mitted in teaching two aspects have to be considered: a) the contents of the specialty; and b) the problem-solving methods with regard to the specialty. On the first aspect an overwhelming amoUnt of book$ and articles has bee~ written in medicine. The second 8Spect is usually only mentioned in passing. In my opinion, ~his subject has thusfar been greatly undervalued. When one is unable to trace the - problem-solving processes how can anyone determine the efficscy, the effectivity, and the efficiency of this process, or value~ the outcome. To state it in Magerien terms: "If you do not know where to g;o, you may very well end up somewhere else- and not even know it." How physicians solve clinical problems is the main object of this research. The investigator studied and modeled two of the eldest and famous ways of pJroblei!Jlrsolving: the deductive and the inductive strategy9 with the modern probability reasoning viewed as an extension of the latter strategy. All 68 physicians who participated in this investigation used the inductive strategy for the -usually four - presented patient-problems. Within the inductive strategy three variants could be distinguished. The consequences of this finding are far-reaching. As the inductive strategy does not include a logical hierarchy of argumentationsteps, retracing of the process is impossible. (This aspect is also relate~ to our opinions about experience-knowledge end teaching)A As the hypothesis generation is prior to the acquisition of infot~tion, this latter aspect can only be viewed in the light of the former~ and thus limited to a small number of domains. As the hypothesis generation is - partly - unrelated to the total available amount of information, the decision making (chopsing the ultimate diagnostic hypothesis) will usually follow implicit~ personal trends and standards, e.g. satisfying minimal eicpectations (Satisficing Theory, Simon} or risk-avoiding prospects (Prospect Theory, Kahnemann & Tversky). It suggests a highly personal character of diagnostics and/or the therapeutic management, which is contradictory to general accessibility of medical knowledge and medical teaching. This feature of personal concepts easily links up with Polanyi's theory of "PeJrsonal" or 91Tacit Knowledge" as contrasted to "Objective Knowledge" (Popper) 9 which has general accessibility and validity. During the investigation this as~oct came forth. The framework of the investigation (patient simulation) end the special definitions and coding of illness elements (symptoms, signs9 tests) all~wed for comparing similar conceptions (diagnoses, treatments) ~ong the participants. These comparisons confirm Polanyi's theory and the concepts of inductive reasoning. Mutual comparibility of diagnoses seems hardly possible when analysing these conceptions into their basic elements (symptoms etc.). This aspect touches upon one of the main cornerstones of clinical ~edicine. When the starting positions have not been unequivocally defined treatment, .,;ie-Jed as the intervention in the natural course of a disease, can only lead t4:!1 uncertain outcomes. The lack of · standardized :medical definitions and a tmiform, unambiguous taxonomy inhibits the application of a formalised, normative decision theorry for clinical medicine. Future planning aims at a reconsideration of medical terminology~ medical taxonomy and medical problem-solving methods by means of clustering the basic elements of clinical medicine.