The role of derivational history in aspect determination

Such verbs contain two or more prefixes and the imperfective suffix. In a given context they behave as either unambiguous perfective or unambiguous imperfective verbs, so their grammatical status is fully resolved. Now, this behavior fails to be revealed by the standard diagnostic tests used to determine the perfective or imperfective status of Russian verbs (see Section 3): under those tests, verbs in (1) are aligned with unambiguously imperfective verbs. Many studies also implicitly or explicitly deny the existence of affixed biaspectual verbs in Russian (e.g., recent theories by Ramchand 2004, Romanova 2004, Svenonius 2004b, Tatevosov 2008 and Tatevosov 2009; for more discussions about the size of the group of biaspectual verbs, see Gladney 1982, Čertkova and Čang 1998, Jászay 1999, Anderson 2002). However, it is clearly evident from a recent corpus-based study by Borik and Janssen (2012) that such verbs exist. They constitute an open class of lexical items with subgroups that follow productive patterns. The main focus of this paper is on the properties of verbs such as those in (1), which have so far remained largely unexplored, and at closer inspection pose challenges both to traditional and also to recent syntactic analyses of derivationally complex Russian verbs. The hypothesis defended in this paper is that a cetain class of biaspectual verbs sui generis must be acknowledged: namely, a class of verbs that cannot be neatly aligned with either imperfective or perfective verbs. A case in point is the class of derivationally complex verbs, as exemplified in (1). It will be argued that the biaspectual status of a given verb of this type is a consequence of the multiple derivational histories associated with it. Our approach not only pro-