JAVANESE AND PROBLEMS IN THE ANALYSIS OF ADVERSATIVE PASSIVE

Adversative passive is one of the means that languages use to code that an event may have detrimental effect on someone. The adversative passive differs from the standard passive in that the speaker perceives an event as unpleasant or unfortunate. The adversity semantic effect is normally encoded with an adversative passive affix attached to the verb.  Javanese has such coding with (1) prefix ke- and (2) the circumfix ke--an. However, Javanese adversative passive is not always associated with adversity.  In fact, an event described by Javanese adversative passive may have neutral or pleasant consequences. This proves to be problematic for the current frameworks on adversative passives such as Kubo’s (1992) and Pylkkänen’s (2002) because their frameworks assume that an adversative passive carries an adversative semantic property encoded in the malefactive head or with a passive morphology. Moreover, the subject of the ‘adversative passive’ in Javanese does not have to possess an object because the passive can have a reading in which the passive subject held an object belonging to someone else while experiencing a situation related to the object. This also poses a problem for Pylkkänen’s (2002) because she bases her adversative passive analysis on the possessor raising theory which requires a possesive relation between the theme and the affected argument. I argue that Javanese ‘adversative passive’  is best described as a combination of the prefix ke- and suffix –an (the circumfix ke--an) with the prefix  ke- carrying the accidental semantics property and the suffix –an as an applicative suffix adding an affected argument to the construction.