Warum die Aussage »Text T ist unzuverlässig erzählt« nicht immer interpretationsabhängig ist. Zwei Argumente

Abstract

Abstract This essay asks whether the attribution of unreliability to the narrator of a literary text is always dependent upon interpretation. The bulk of narratological research answers with »yes«. Yet the content of the term »interpretation-dependent« is understood in radically different ways. As a minimal consensus, it is commonly accepted that the attribution of unreliability cannot be described as »interpretation-neutral«, in the way that, for instance, the statement »The narrator in text T is a homodiegetic narrator« is interpretation-neutral. Following a few preliminary explanatory remarks on terminology, I propose two arguments for why this majority opinion is false. I argue that the statement »Text T is unreliably narrated« is not always interpretation-dependent. Within the framework of the first argument, I attempt to show that the criterion of »interpretation neutrality« depends upon some meta-theoretical assumptions. If one assumes that basic linguistic characteristics are valid independent of their interpretation and argues that a sentence such as »Call me Ishmael« establishes a homodiegetic narrator because the word »me« signals that he belongs to the narrated story, then one implicitly excludes as inadequate certain idiosyncratic theories of meaning that would ascribe a different meaning to »me«. That is not problematic in and of itself. But it shows that there are conditions of adequacy for theories of meaning that are fundamentally negotiable. And the set of statements which can be attributed the attribute of being »interpretation-neutral« can vary depending upon how these conditions of adequacy are defined. In a corresponding adaptation of the conditions of adequacy for theories of meaning and interpretation, it is therefore inherently possible that even statements about the reliability of a narrator could be granted the status of being interpretation-neutral. The second argument focuses on the praxis of interpretation. I seek to reconstruct how exactly the qualification of a narrator as homodiegetic (an attribute that is usually considered as interpretation-neutral) and as unreliable (an attribute that is usually not considered as interpretation-neutral) can come about in a process of interpretation. There appear to be cases in which criteria commonly cited to qualify a statement as an interpretation-neutral description of a text are also applicable for the attribution of narrative unreliability. Such cases are literary texts like Agatha Christie’s The Murder of Roger Ackroyd or Ambrose Bierce’s An Occurrence at Owl Creek Bridge, in which the unreliability of the narrator is apparent. The knowledge that the narrators in these texts at least temporarily withhold facts relevant to the plot, tell lies, make mistakes, hallucinate, etc. can just as much be attained on the basis of an unreflective understanding of the linguistic meanings of words as can the knowledge that the narrators are part of the stories they tell. If one wishes to not relinquish the interpretation-neutral status of statements about the ontology of a narrator (the qualification, that is, of a narrator as homo- or heterodiegetic) to a relativism that includes linguistic interpretations, then one is forced in principle to also retain the status of interpretation neutrality for statements about the reliability of a narrator. Both arguments lead me to conclude that the universal quantification that all determinations of the reliability of a narrator are dependent upon interpretation is false. I propose that we limit ourselves to more modest existential quantifications and that we do not attribute the attributes »interpretation-dependent« and »interpretation-neutral« to entire literary categories or types of statements in general, but rather to individual statements. Moreover, I give a short and tentative definition of the criterion »interpretation neutrality« that follows from these considerations.