Language Typology and Syntactic Description, Volume I: Clause Structure, Second edition
participant, which, if present must be in a [
Download 1.59 Mb. Pdf ko'rish
|
Lgg Typology, Synt Description v. I - Clause structure
participant, which, if present must be in a [ +oblique] case like dative. The antipassive version of (90c) is (91) (- ŋ a(y) is the derivational verbal suffix for antipassive): (91) bayi ya a (bagun dugumbil-gu) balgal- a- u DET.ABS man. ABS DET.DAT woman- DAT hit- ANTIPASS-TNS [ + A ] [ – A ] – oblique + pivot + oblique [ ] ‘The man hit the woman’ The pivot system of Dyirbal is ergative–absolutive (ergative: [ −oblique], [ −pivot]; absolutive: [−oblique], [+pivot]), as opposed to English, which A typology of information packaging 397 is nominative–accusative (nominative: [ −oblique], [+pivot]; accusative: [ −oblique], [−pivot]). This claim can be substantiated for Dyirbal in the same way it was for English: by showing a number of grammatical constructions for which the [ −a] pivot is the controller or target. A number of these constructions have English analogues; consider, for example: (92) controlled np in nonfinite purposive clauses ↓ ↓ (a) balan d ugumbil ba ŋ gul ya a- ŋ gu balga-n [ badi-gu] [ −a] [ +a] [ −a] det .abs woman.abs det.erg man-erg hit-tns fall-purp ‘The man hit the woman, (causing her) to fall’ ↓ ↓ (b) *balan d ugumbil ba ŋ gul ya a- ŋ gu wawu-n [ [ −a] [ +a] [ +a] det .abs woman.abs det.erg man-erg fetch-tns balan nayinba walmbil-i [ −a] det .abs girl.abs get up-purp ‘The man fetched the woman to get the girls up’ ↓ ↓ (c) balan d ugumbil ba ŋ gul ya a- ŋ gu wawu-n [ [ −a] [ +a] [ +a] det .abs woman.abs det.erg man-erg fetch-tns bagun nayinba-gu walmbil- ŋ ay-gu [ −a] det .dat girl-dat get up-antipass-purp ‘The man fetched the woman to get the girls up’ Dixon (1972, 1994) Remember the pivot for Dyirbal is the absolutive np, the [ −a] argument of a transitive verb or the sole [ ±a] argument of an intransitive verb. The pivot is the target for control in these nonfinite purposive clauses. Example (92a) is fine because the controlled np is the [ −a] sole argument of an unaccusative intransitive verb bad i- ‘fall’, which, if overt, would be in the absolutive case (see (90b)). Example (92b), on the other hand, is ungrammatical because the targeted np for control is the [ +a] argument of a transitive verb walmbil- ‘get up’, which would appear, if overt, in the [ −pivot] ergative case (note that the English translation of (92b) is perfectly grammatical because English, of course, has the opposite choice for pivot from Dyirbal, namely the [ +a] argument of a transitive verb). For the meaning of (92b) to be expressed grammatically, an 398 William A. Foley antipassive must be used, in which the [ +a] argument is presented as the pivot of a derived intransitive unergative verb. This is found in (92c): the transitive verb root walmbil- ‘get up’ is now marked with the detransitivizing antipassive suffix - ŋ a(y). Its core [ +a] argument, if overt, would have to be marked with absolutive [ +pivot] (see (91)). As pivot, it can now be a target for control so that (92c) is well formed. Languages with an ergative–absolutive alignment for pivot choices, i.e. the pivot is the [ −a] participant of a transitive verb in contrast to the [+a] choice for languages with a nominative–accusative alignment, were once thought to be rare. But intensive study since the mid-1970s has unearthed more examples. Besides Dyirbal and some other languages of far north Queensland in Australia (see Dixon (1977a, 1980, 1981, 1994)), other languages of this type include Eskimo (Manning (1996)), some Mayan languages, especially those of the Mamean subgroup like Mam (England (1983a, 1983b, 1988)) and perhaps a number of Western Austronesian languages like Chamorro (Cooreman (1987, 1988)). An interesting problem confronted by all languages with this typology is the class of ditransitive verbs, with their multiple [ −a] participants. There can only be one [ +pivot] np per clause. Ditransitive verbs present no problems for nominative–accusatively aligned [ +pivot] languages like English because the pivot is the [ +a] participant, of which there is only one: (93) Egbert gave Hortense a python [ + A ] [ – A ] [ – A ] – oblique + pivot ⎡ ⎣ ⎤ ⎦ – oblique – pivot ⎡ ⎣ ⎤ ⎦ – oblique – pivot ⎡ ⎣ ⎤ ⎦ But true ditransitive verbs should be impossible or at least highly shunned in ergative–absolutively aligned [ +pivot] languages, because it would potentially permit two pivots which is impossible, so that structures like the following should be ungrammatical: (94) *Egbert gave Hortense a python [ + A ] [ – A ] [ – A ] – oblique – pivot – oblique + pivot – oblique + pivot * A typology of information packaging 399 In language after language of this typology, this prediction is borne out: there are no true ditransitive verbs; there is only one [ −a] argument so that ‘give’ is a transitive verb only: (95) Dyirbal (a) balam mira ba ŋ gul ya a- ŋ gu [ −a] [ +a] det .abs bean.abs det.erg man-erg wuga-n bagun d ugumbil-gu [ +oblique] give-tns det.dat woman-dat ‘The man gave beans to the woman’ (b) balan d ugumbil ba ŋ gul ya a- ŋ gu [ −a] [ +a] det .abs woman.abs det.erg man-erg wuga-n ba ŋ gum mi a -d u [ +oblique] give-tns det.instr bean-instr ‘The man gave the woman (with) beans’ Dixon (1972) Dyirbal Download 1.59 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling