Somewhat provocatively for the historian, in this article I pose a question concerning the extent to which it is acceptable to look for a reflection of the real world in literary sources that are primarily fictional. I base myself mainly on a method which seeks realities immanent in individual works, which accentuate the autonomy of a work of art and thwart the possibility of interpreting these texts mimetically. The potential information encoded in the work can be classified into three groups. With implicit communications there is a greater risk that its mimetic interpretation will be over-speculative. The interpretations have therefore to be confirmed or refuted by a method based on the internal immanent reality. My starting point is that there is a fundamental difference between the real/natural world and the fictional world. The literary work is regulated as an aesthetic sign by a set of rules. These rules need not be equivalent to those known and valid in the real world. In a polemic with Alfred Thomas about the reflection of medieval rulers' power in the chivalrous romances, I argue by means of an analysis of the two visits of Duke Ernest to a city in Cyprus. In my first point I demonstrate that the only thing the author of a work can fully bring under control is meaning and function. By contrast, as far as the manner in which the meaning or function of the text is communicated, the author is to a large extent constrained by an autonomous literary system, which influences not only the selection of individual elements, but also the manner of their composition. In my second point I demonstrate, with the help of a specific example, that an unmediated mimetic interpretation of the text is impossible. I demonstrate that the passivity of several figures in the chivalrous romance, which Thomas explains mimetically, is merely apparent, and manifests no connection with real-life circumstances.