In this article the outlines are sketched of an empirical science of literature as close as possible to the model of the natural sciences. This raises the question of what the standards of an empirical science in the strictest sense should generally be. Practical relevance of its results soon turns up as the fundamental condition for an explanatory empirical science, if the ideology of nearing an empirical truth is no longer accepted and a mere pragmatic justification rejected as its insufficient substitute. As a consequence the logical priority of scientific explanation over scientific description is considered irrefutable. The so-called "sciences of understanding" - among them a "science of literature" - are confronted with a special difficulty, if they want to become explanatory empirical sciences. Their practical problems to be solved are related to linguistically or semiotically coded facts. To empirically describe these problematic facts causes this difficulty, which the author tries to overcome with recourse to ideas from Wittgenstein. And the practical relevance, which may be achieved with the help of nomological explanations, here doesn't consist in predictions for better practical production of means, direct or indirect, for survival, but in the provision of rational understanding for better practical teaching or criticism. This practical relevance is shown in the explanation of valuing acts of readers of The Golden Bowl by Henry James. The meaning of a literary text is no longer taken as a possible object of scientific empirical explanation, nor its rational justification as a scientific goal that might be practically relevant.