The Case Differential: Syntagmatic Versus Paradigmatic Case – Its Status In Synchrony And Diachrony


Werner Abraham and Elisabeth Leiss
Department of German Linguistics
University of Munich
Schellingstrasse 3/RG
80779 Munich


In today’s leading discussions on case and its licensing mechanics there is only one approach: the syntagmatic type of case assignment. In contrast, the present article describes patterns of paradigmatic case assignment. Paradigmatic case assignment means that two different m(orphological)-cases fill one slot licensed for the same theta role, whereas under the strategy of syntagmatic case assignment there is just one case form assigned for one theta property. By paradigmatic case, we mean that the same case-assigner/licensor may opt for more than one m-case, depending on a number of factors among which aspect, nominal reference and animacy. Illustrations will be provided from modern languages, mainly Germanic and some Slavic, as well as their older states. See the following case pair (Acc vs. Gen) from Middle High German (c. 1100) er azdaz brot.accvn tranc dazvo. /eîns wazzers.gendaz er […] vant‘he ate (up) the bread and drank some water that he found‘. While drinking up the water in the bucket is impossible (forbidding telic austrinken‘out-drink’), thus justifying the partitive genitive eîns wazzers, daz brot‘the (piece of) bread’ expressed by the accusative denotes a bounded, telic event (permitting telic aufessen‘up-eat’). The deeper reason for paradigmatic case is that the alternates encode covert differences of reference and discourse status in languages, or their diachronic states, where the respective functional equivalents (such as (in)definite articles) have not (yet) emerged or are amiss. We argue that such paradigmatic bare case doubling occurs only when the triggering force is an active and powerful aspect system as in Modern Russian and Polish as well as in early states of Germanic (Gothic, Old High German). There are different kinds of paradigmatic case assignment, among which the differential object marking/DOM) in various modern languages. They may be seen as a fossilized trace of this earlier, far more dominant, case-encoding strategy. In the present discussion we focus on what we label ‘one theta role two bare case forms’ (OTTC).