On the control/raising ambiguity with aspectual verbs: a structural account

Abstract

In what follows, I first briefly review Perlmutter (1968, 1970), in which it is argued that aspectual verbs are ambiguous between control and raising. I suggest that while the argument for the raising analysis is solid, the arguments supporting the control analysis of aspectual verbs are less so. As an alternative hypothesis to consider, I introduce the structural ambiguity hypothesis. In Section 3, I review three recent analyses of control and raising. Although there are important differences among them, they all share the basic assumption that the control/raising distinction is due to differences in selectional restrictions that the lexical items impose. Under such an assumption, the lexical ambiguity hypothesis is the only available option. In Section 4, I present evidence for the structural ambiguity hypothesis from studies concerning aspectual verbs in languages from four distinct families, German (Wurmbrand 2001), Japanese (Fukuda 2006), Romance languages (Cinque 2003), and Basque (Arregi Molina-Azaola 2004). These data strongly suggest that across languages aspectual verbs can appear in two different syntactic positions, either below or above vP, or the projection with which an external argument is introduced (Kratzer 1994, 1996, Chomsky 1995). Given these findings, I argue that it is the aspectual verbs' position with respect to vP which creates the control/raising ambiguity. When an aspectual verb appears in a position that is lower than vP, an external argument takes scope over the aspectual verb. Thus, it is interpreted as control. When an aspectual verb appears in a position that is higher than vP, on the other hand, it is the aspectual verb that takes scope over an entire vP, including the external argument. Thus, it is interpreted as raising. In section 5, I extend the scope of this study to include a discussion of want-type verbs in Indonesian, as analyzed in Polinsky & Potsdam (2006). Polinsky & Potsdam argue that the Indonesian want-type verbs must be raising in at least certain cases where they allow a rather peculiar interpretation. Although they assume that there are also control counterparts of the want-type verbs, I argue that applying the proposed analysis to the want-type verbs does away with the need for stipulating two distinct lexical entries for these verbs. Section 6 concludes the paper.

On the Verbal Complements of Aspectual Verbs

“…Note that a tendency can be observed with regard to the voice of the aspectual verb: for example, the verb pauō (active voice) [that belongs to the ifrst group of aspectuals] can take an in ifnitive or a participle as its complement; while pauomai (mediopassive voice) appears to select a participle as complement. This distribution accords with Fukuda (2007), who has shown that aspectual verbs in the passive should be more natural with aspectuals in L-Asp (+participials) 12) We have not included ekleipō in our indicative list, given it seems that most of the uses of ekleipō do not involve any of the basic aspectual meanings we are dealing with (start-continue-stop).…”

“…Sevdali (2007) who argues (a) that all in ifnitives in Ancient Greek are CPs, and (b) that there is a split between two types of CPs, normal C*Ps (strong phase) that have a contrastive focus feature and Control in ifnitives that do not have this feature and consequently do not have availability of accusative for the in ifnitival subject. 19) That is similar to the analysis of gerundives of Modern English as more limited than in ifnitives with respect to a range of possible interpretations (Bolinger 1968, Freed 1979, Brinton 1988, Fukuda 2007. In English, states are natural with in ifnitives but often unnatural with gerundives.…”

“…A functional head analysis has been proposed for aspectual verbs in German (Wurmbrand 2001), Romance languages (Cinque 2003), Basque (Arregi & Molina-Azaola 2004), Japanese (Fukuda 2006), and Modern English (Fukuda 2007): the main evidence for most of these languages is that long passive, i.e. passivisation of an embedded object with the passive morpheme appearing only on the matrix/aspectual verb, is the only option with aspectual verbs in L-Asp, since the passive morpheme is above the position of the aspectual verb, while the regular embedded passive is the only option with an aspectual verb in H-Asp, since the passive morpheme is below the position of the aspectual verb.…”

“…We will show (based on other morpho-syntactic evidence as well) that this dichotomy in Ancient Greek, Old (and Modern) English corresponds to a structural ambiguity between the alternative syntactic positions that aspectual verbs can occupy (H-Asp(ect) position [+ in ifnitive (to-in ifnitive) or participle (bare in ifnitive/gerundive) as complement] or L-Asp(ect) position [+ participle (bare in ifnitive/gerundive) as complement]). Travis (1991), Borer (1994Borer ( , 1998Borer ( , 2005, and Fukuda (2007) have shown that aspectual verbs are functional heads that encode aspectual information about events, while Fukuda (2007) has argued that Aspect heads can occur in two positions.…”

Crossed control revisited; The structure and interpretations of “want” and so on + passive verb in Malay/Indonesian

“…In (29), for example, the individual who is careful and who is under an obligation can be either John or the doctor. Indeed, some previous authors have ascribed crossed reading to the auxiliary status of mau and other words meaning "want" (Isamu Shoho 1995;Shin Fukuda 2007;Simon Musgrave 2001;Hirobumi R. Sato 2004). For mau, the authoritative dictionary Kamus Besar Bahasa Indonesia (KBBI) lists its intention/desire meaning as adverbia, which includes auxiliaries as well as adverbs.…”