Decreasing and Rearranging Alternations in Hindi

This paper reports on the ongoing work in creating the Hungarian data-base for the PaVeDa project, with special attention to a selected array of lan-guage specific issues that may also have relevance beyond Hungarian. Hun-garian has relatively free word order, but its morphology is rich. Valency al-ternations are typically coded on the verb, and non-subject arguments receive designated case markers. This rich morphology opens a window onto a world of fine-grained variation in argument coding that may be gently disguised in languages with less articulated morphology.