Language Typology and Syntactic Description, Volume I: Clause Structure, Second edition
Participial adjuncts are clauses introduced by the particle
Download 1.59 Mb. Pdf ko'rish
|
Lgg Typology, Synt Description v. I - Clause structure
Participial adjuncts are clauses introduced by the particle nang, which express action simultaneous with that of the main clause, with the subject suppressed, but understood as coreferential to an actor or object argument of the main clause, regardless of focus. Coreference with a dative or prepositionally marked argument is not allowed: 210 Avery D. Andrews (124) a. B[in]isita ni Juan ang hari nang nag-iisa [perf]-visit(op) act Juan piv king adv ap.imperf-one ‘Juan visited the king alone [either Juan or the king is alone]’ b. B[um]ista si Juan sa hari nang nag-iisa [ap.perf]-visit piv Juan dat king adv ap.imperf-one ‘Juan visited the king alone [only Juan is alone]’ c. H[in]uli ng polis ang mgananakaw nang pumapasok perf -catch(op) act police piv thief adv ap.imperf :enter sa banko dat bank ‘The police caught a/the thief entering the bank [either thief or police are entering]’ d. Nang-huli ng mgananakaw nang polis nang pumapasok ap.perf -catch obj thief piv police adv av.imperf:enter sa banko dat bank ‘The police caught the thief entering the bank [either thief or police are entering]’ The non-ambiguity of (b) shows the difference beween the (non-core) dative, which can’t be understood as the subject of the nang construction, and the arguments marked by ng and ang in the other examples, which can be. This, together with Kroeger’s other tests, establishes a core–oblique divide with ng phrases on the core side, regardless of whether they are objects of ap verbs or the agents of op ones. So we have a situation where actor and object are core arguments regard- less of whether they are pivot or not, and where furthermore the actor is a privileged target for complement subject ellipsis, and also outranks other argu- ments on a hierarchy relevant for reflexivization (an actor can reflexivize any- thing, and nothing can reflexivize it). The concept of core vs oblique arguments seems supported, but one of the core arguments has subject-like properties regardless of whether or not it is the pivot. In terms of the ideas introduced at the beginning of the section, the actor will be the a-subject, the pivot the p-subject. Therefore in ap sentences the a-subject and the p-subject are the same np, which is also the a/s (giving a sentence structure similar to that found in nominative–accusative languages), but in op sentences the p is p-subject while the a is the a-subject, giving a grammatical structure similar to what is found in syntactically ergative languages. This analysis thus provides the useful properties of the passive and antipassive analyses without suffering from their drawbacks. The major functions of the noun phrase 211 Splitting the subject grammatical relation into a-subject and p-subject there- fore helps to elucidate the Philippine type as well as syntactic ergativity, both mixed and unmixed. The difference between these types of languages and more familiar languages such as English is that in the latter there is only one subject- like grammatical relation rather than two, with the sole subject-like relation tending to have the typical properties of both a-subject and p-subject (one could think of both kinds as existing, but always being the same np). 4.3 The universal status of a- and p-subjects We have now seen that some languages have a ‘full’ subject combining the properties of a- and p-subjects, whereas others split them into two distinct grammatical relations. A further question is whether these two kinds of subject are always found, whether individually or combined. The answer appears to be that p-subjects are clearly not universal, while the issue is rather doubtful for a-subjects. The languages without p-subjects would be languages such as Warlpiri, which lack passive or antipassive rules that alter the semantic role of the np in a recog- nizable pivot position (subject for nominative–accusative languages, absolutive for languages with ergative syntax). Although formally inclined linguists have tended to neglect the different significance of a putative subject relation in lan- guages with and without a passive rule, it has been discussed at some length in Van Valin (1981), Foley and Van Valin (1984), and Van Valin and LaPolla (1997:265–6); see also Foley chapter 8). It is hard to avoid the conclusion that if a language lacks any rules altering the semantic role of an np in ‘subject’ position, the significance of that position in the functioning of the language must be different from that of a similar position in a language that has such rules. Presence vs absence of a p-subject provides a straightforward account of the difference, and has been what has been proposed since Foley and Van Valin (1984), under various terminologies. P-subjects, for example, are frequently preferred, or required, to be definite, but, clearly, no such requirement can plausibly exist in a language without passives, where traditionally recognized subjects would be a-subjects. What about absence of a-subject? One possible case is Dyirbal, where there is no clear and compelling evidence for grouping s and a together, but only s and p. So Dyirbal might be a language with a p-subject, but no a-subject, and the same would be true of other ‘pure syntactic ergative’ languages, if these exist. But, after more than twenty-five years, Dyirbal is still the only reported case of a pure ergative language that has withstood scrutiny. Furthermore the data on Dyirbal is limited, and there is little prospect of getting additional data relevant to the question of whether or not it has a-subjects. |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling