Berthold Delbrück and the syntax of cases. An analysis of the case ending -φι in Homer

In this article, we investigate Delbrück’s analysis of the cases by discussing the case ending -φι in Homer. He himself treated the issue on three occasions and stated that this case form remained insufficiently and unsatisfactorily explained. Initially, he argued that the original meaning was the instrumental-comitative one, expanded with locative and ablative functions and the original number was the plural. Later, however, he assumed that the ending could also be used for the genitive and the dative. As Mycenaean had not been discovered at his time and he could therefore not have included it, we focus in our analysis solely on -φι in Homer and only refer to Mycenaean sparingly. First, we provide an overview of the scholarship on Homer, distinguishing between the scholars writing before the decipherment (including Delbrück) and the ones after it. In a second step, we determine our corpus by discussing some passages with uncertain transmission (αὐτόφι versus αὐτόθι and Ἰλιόφι κλυτὰ τείχεα). We then provide the figures and proceed to the actual analysis. We start with number, animacy and concreteness, then discuss the case usages – locative, instrumental, object marking – and the use of the forms with prepositions. After that, we take a closer look at the distribution of the forms and their co-occurrence with genitive and dative forms in the same sentence and the same syntagma. At the end of the article, we analyse the instances for which more than one interpretation is possible. Our analysis shows that the suffix is numerus-indifferent and that instances with an unambiguous plural are relatively uncommon, that the suffix appears almost exclusively with inanimate entities, is used predominantly with concrete elements, has often instrumental and locative meaning, but can also be used as an ablative and is very common with prepositions (more than half of the instances). Our findings differ from what Delbrück himself noted in two respects, namely the lack of animacy and the lack of plural meaning, and from the data in Mycenaean in three respects, namely the mostly singular meaning of the suffix (in Mycenaean, the suffix is almost exclusively used in the plural), its common use with prepositions and with the ablative and locative-directive functions (these two uses are very rare in Mycenaean). They also make the interpretation of the ending as a simple oblique case marker or a simple poetic tool less likely and shed new light on the original function and the origin of the ending and its use in Proto-Indo-European (PIE) and other Indo-European languages, but that discussion cannot be performed in the current article.