Objective conjugations in Eskaleut and Uralic

Languages in several eastern branches of the Uralic family have “objective” verbal paradigms, so called because they encode information on both the subject and the direct object. This feature has traditionally been compared, particularly from the genetic standpoint, with the Eskaleut transitive conjugation. The justification for this comparison is nevertheless weakened by current views on the nature and functioning of the objective conjugation in Proto-Uralic. In particular, it is now assumed that pronominal endings were always directly attached to the verb root. In this paper, I use data from Mansi (traditionally classified as Ob-Ugrian) to show that the Uralic objective conjugation must have resembled, and still resembles today, the Eskaleut transitive conjugation more than is currently allowed, at least in structural terms.