Benefactive, applicative, and possessor-raising constructions in Bimanese

Melanie Owens

Stanford University

Bimanese, like many languages, allows semantically 'non-core' arguments which are ordinarily expressed in oblique phrases to be realized as syntactic objects. This paper focuses on the realization of benefactive arguments as objects, as in (1), while (2) exemplifies the more typical situation where this non-core argument is instead realized as an oblique. Bold-faced in (1), but absent from (2), is a morpheme wea. From this pattern it may appear that wea accompanies, or even signifies, the expression of the benefactive argument as an object, yet (3) invalidates this conclusion by showing that wea may still be realized when the benefactive is realized as an oblique. The exact nature and function of wea is thus superficially unclear.

Wea also appears in possessor-raising constructions, such as (4), which similarly contain non-core arguments realized as syntactic objects. These constructions are characterized by a necessary relationship of inalienable possession between the 'raised' possessor and the base object, and by their propensity for malefactive readings. Yet a malefactive interpretation for the possessor is by no means entailed; the most that can be assumed is that these individuals are to some degree affected by the event described, by virtue of inalienably possessing the object. Since benefactive objects in double object constructions, such as (1), can similarly be described as being affected by the event described (albeit with the added information that the effect is beneficial), the morpheme wea is glossed as 'AFF[ ected]'. Wea has the constant semantic effect of signaling that there is some individual who is in some way affected by the event described, which I represent formally as a relationship between an event and an individual at argument structure, as in (5).

Although wea has a constant semantic effect, it does not have a constant syntactic effect. The factors which determine possible expression in double object form are in fact entirely independent of the instantiation ofwea. Wea does not enable a non-core argument to be realized as an object. Under my argument structure representation, this lack of syntactic effect arises from the requirement that the individual argument introduced by the AFFect relation must be fused (as in Alsina and Joshi, 1991) with another argument, resulting in a single syntactic realization, as in (6). Without anything to fuse with, this argument cannot be syntactically expressed, as in (7b) (although note from (7a) that the semantic implication of the existence of an affected individual does remain). But there is one situation in which the affected individual need not be fused, yet still may be syntactically expressed: when the base verb is intransitive, as in (8). These expressions, I argue, are correctly termed 'applicatives', in contrast to benefactive and possessor- raising constructions. I am thus arguing against the trend of collapsing these expressions into one category of 'applicative' that has been pervasive ever since Baker (1988). The data I present also directly counter the validity of Pylkkanen's (2000) distinction between high and low applicatives, which is representative of this tendency to over-generalize the notion of applicative.

 Examples

(1)        Nahu ndawi wea.ku Fero pangaha.

Ip make AFF.l Fero cake

'I made Fero a cake.'

(2)        Nahu ndawi.ku pangaha m'u Fero.

Ip make. 1 cake for Fero

'I made a cake for Fero.'

(3)        Nahu ndawi wea.ku pangaha m'u Fero.

Ip make AFF.l cake for Fero

'I made a cake for Fero.'

(4)        Nahu dompo wea.kujanga tuta.na.

1 p cut AFF.l chicken head.3

'I chopped chicken's head off.'/'I chopped its head off on the chicken.'

(5)        AFF < _event, _individual>

(6)        Formal representation of (1):

Argument Structure:             AFF < make-for <I, Fero, cake> _event,      Fero _individual>

             I    I

   I       I      I

Grammatical Relations:                                              SUBJ DBJ DBJ

(7)        a. Fero ka.raso                     wea.na pingga.

    Fero CAUS.clean AFF.3   plate

    'Fero washed the dishes (for someone).'

b. Argument Structure:             AFF < wash <I , dishes> _event, X _individual>

          I      I           I

    Grammatical Relations:                SUBJ DBJ            Ø

(8)        Nahu hari    wea..ku Fero.

Ip      laugh  AFF.l    Fero

'I laughed for Fero.'

References

Alsina, Alex, and Joshi, Smita. 1991. Parameters in Causative Constructions. CLS27: 1-15.

Baker, Mark. 1988. Incorporation: a Theory of Grammatical Function Changing. Chicago: The University of Chicago Press.

Pylkkanen, Liina. 2000. What Applicative Heads Apply to. University of Pennsylvania Working Papers in Linguistics, Volume 7.1. 197-210.