AbstractThe contribution clarifies from an epistemological and hermeneutical perspective some of the indispensable assumptions which are generally made during hypothetical deductive and hypothetical inductive inferences in text interpretations, but which are rarely reflected as such. The modes of inference of the natural sciences often serve as a model for, e. g., a ›hermeneutics of nature‹, even though these modes of inference cannot be directly compared to those of the humanities (1). In order to understand the conditions for making such a comparison, we consider the problem of interpretation conceived as the problem of the arbitrariness of interpretations. This problem, which has the three components of question, of evaluation, and of knowledge, can be solved by a concept of interpretation in which the arbitrariness is avoided or limited by a methodology of text interpretation (2). Meeting the general demands of a theory/methodology of interpretation requires us to explicitly specify the underlying concepts of meaning and interpretation (3). When interpretation is understood as relating a text to a context, the concept of meaning fixes a primary context for this text-context relation. The concept of interpretation associated to the concept of meaning distinguishes further contexts, possibly in a hierarchical order. Relations between different interpretations can then be analyzed as relations between the concepts of meaning and interpretation. Furthermore, disputes about the sense and relevance of questions can then be understood analytically as well as historically. We illustrate this with the example of the question whether Don Quixote was a Marrano. This will show that even with fixed concepts of meaning and interpretation no a priori criteria for the definitive closure of the interpretation of a work of literature can be given (4). Inferences of interpretation presuppose relations of meaning, by which we understand a relation between the signifier and the signified. For poetic texts the manifold dimensions of the differences between semantic, exemplifying, and analogical relations of meaning have to be taken into account, which may, with respect to a given text, be mutually in competition or in support. We illustrate this with the example of Hugo von Hofmannsthal’s Chandos letter. Exemplifications and analogies often enlarge the spectrum of meaning of poetic texts (5). Among the requirements of text interpretations we also count general background assumptions. In addition to heuristic assumptions there are presumptions. While hypothetical assumptions do not have to be justified a priori and while assertions do require proof, for a presumption the burden of proof lies with the critic. The status of a presumption is determined by the epistemic situation in which the interpretation is carried out (6). When the validity of two hypotheses of interpretation is to be compared, the hypotheses have to be evaluated. The evaluation of the plausibility of an interpretation can refer to either the initial hypotheses or the results of the interpretation. We observe that, just like in the natural sciences, evaluative decisions are underdetermined. In this context we also briefly comment on some assumptions concerning the ›inference to the best explanation‹ (7). Finally, we deal with the often raised suspicion of circularity of hermeneutic arguments. This suspicion is frequently declared to be the central difference between inferences in the natural sciences as opposed to inferences in the humanities. Historically, the hermeneutic circle and its inevitability is an invention of modernity. It can be shown systematically that not every circular inference needs to be vicious. Rather, whether it is vicious or not is rather determined by the epistemic situation. As long as the implicit knowledge that guides an argument is not fully explored, the circular or non-circular character of an argument cannot be judged (8).
Read full abstract