The Nature of Irreality in the Past Domain

Evidence from Past Intentional Constructions in Australian Languages* This paper was written while I was a visiting research fellow at the linguistics department of the University of Melbourne. I would like to thank the department for its hospitality, as well as the Fund for Scientific Research–Flanders for its financial support. I am grateful to Rachel Nordlinger and two anonymous reviewers for very helpful comments on an earlier version of this paper, and to Bruce Birch, Alan Dench, Nick Evans, Ian Green, Shelly Harrison, Ekkehard König, Bill McGregor, Nele Nivelle, Marie-Eve Ritz and Adam Saulwick for general discussion or specific comments on some of the ideas presented here. I am of course responsible for any errors of fact or interpretation

Bibliographic Data

ID12059403
AuthorsJean‐christophe Verstraete (0000-0002-3576-9827, KU Leuven, corresponding author)
Year2006
Volume26
Issue1
Pages59-79
Publication date2006-03-16
Peer ReviewedYes
Open AccessNo
TypeARTICLE
VenueAustralian Journal of Linguistics (JOURNAL)
Journal identifiersISSN: 0726-8602 • E-ISSN: 1469-2996
PublisherTaylor & Francis (PUBLISHER • GB)
DOI10.1080/07268600500531636
OpenAlexW4214597446
LanguageEN
Citations received10
References cited18

This study investigates an unusual pattern of mood marking found in the past-tense counterparts of purposive and apprehensive complex sentences in some Australian languages. The secondary clause in these constructions, which will be called past intentional constructions, is typically marked for mood even though the event described is often known to have taken place, and in some cases the mood markers have counterfactual meanings even though the constructions as such do not have any counterfactual meaning. It is argued that past intentional constructions can be defined on the basis of a combination of two features: (1) mood marking in the secondary clause, which adduces a semantic feature of (un)desirability; and (2) a link of this mood to the agent of the main clause, which creates a relation of intended or avoided realization between the main clause and the secondary clause. This analysis can help both to motivate the presence of mood marking in these constructions, and to explain the apparently unmotivated use of mood markers with counterfactual meaning, which is shown to be a cancellable implicature rather than a basic meaning of the mood markers in question. In addition, the analysis also suggests that purposive, apprehensive and past intentional constructions constitute a formally and functionally intermediate category between constructions of speech and thought representation and constructions of adverbial modification. Notes This paper was written while I was a visiting research fellow at the linguistics department of the University of Melbourne. I would like to thank the department for its hospitality, as well as the Fund for Scientific Research–Flanders for its financial support. I am grateful to Rachel Nordlinger and two anonymous reviewers for very helpful comments on an earlier version of this paper, and to Bruce Birch, Alan Dench, Nick Evans, Ian Green, Shelly Harrison, Ekkehard König, Bill McGregor, Nele Nivelle, Marie-Eve Ritz and Adam Saulwick for general discussion or specific comments on some of the ideas presented here. I am of course responsible for any errors of fact or interpretation. 1The markers discussed in the text will be indicated in bold. The glosses in the examples are taken over from the sources, but were slightly normalized in a number of instances. The abbreviations used are A—transitive subject, ABL—ablative, ABS—absolutive, ACC—accusative, ALL—allative, APP—apprehensive, ASSERT—assertative, aug—augmented, CAUS—causative, CF—counterfactual, CLOC—complementizing locative, CM—conjugation marker, CONT—continuous, CT—contemporary, DAT—dative, Dl—dual, EMPH—emphatic, ERG—ergative, EXP—expectation, EXR—experiencer, FREQ—frequentative, FUT—future, G—goal, GEN—genitive, IMP—imperative, IMPF—imperfective, IMPL—implicated, INC—inclusive, INCR—verbal increment, IRR—irrealis, LOC—locative, m—masculine, min—minimal, MPROP—modal proprietive, NEG—negative, NF—non-future, NOM—nominative, NPP—non-past progressive, ns—non-singular, O—object, ORIG—origin, PASS—passive, PC—past completive, PERF—perfective, Pl—plural, PNCT—punctiliar, PCT—precontemporary, POSS—possessive, POT—potential, PRO—pronoun, PROP—proprietive, PRS—present, PST—past, PURP—purposive, R—realis, RDP—reduplication, REL—relative, RPST—recent past, RR—reflexive/reciprocal, S—intransitive subject, Sg—singular, SS—same subject, TEMP—temporal, uaug—unit augmented, UNM—unmarked. 2See Verstraete (Citation2004) for a more detailed analysis of these general tendencies, and the semantic relation between mood meaning and complex sentence meaning. 3In this case, the tense/mood marking is actually homophonous with infinitive marking, but McKay (Citation1975: 321–322) argues that pronominal prefixes in general play a disambiguating role because they indicate finiteness, distinguishing finite past counterfactual forms, which have pronominal prefixes, from non-finite infinitive ones, which do not have any. 4A further piece of evidence that may be relevant is the well-known fact that in a large number of Australian languages, the deontic mood marker found in purpose constructions appears to be related to a formative that marks dative case in the nominal domain [see the section on the bivalent suffix –ku in Dixon (Citation1976)]. For Arrernte, for instance, Wilkins (Citation1989: 238) analyses the purposive marker –tyeke, which is found in main clauses with deontic meanings as well as in purpose constructions [see examples (4) and (6) above], as deriving from a combination of a nominalizer –tye and a dative marker –ke. While this is not as direct a reflection of agent-binding as in the Ngiyambaa and Nyikina cases illustrated above, there are still good reasons to interpret the link with dative markers as a formal reflection of a semantic link with agent-binding. If we look at the range of functions of dative marking in the nominal domain, this typically includes agent-bound intentional meanings such as marking the goal of an activity, as in the Arrernte (Wilkins Citation1989) structure in (a) and the Wargamay (Dixon Citation1981) structure in (b) below. From this type of function in the nominal domain, it is of course only a small step to transfer intentionality to the clausal domain, replacing the nominal argument with a clausal one, which may be formally nominalized [as is the case in Arrernte in example (4)—see Wilkins (Citation1989: 238)] or not [as is the case in Wargamay in example (20)—see Dixon (Citation1981: 69)]. 5I am grateful to a reviewer for drawing my attention to this connection

Applied linguistics · Domain (mathematical analysis · Linguistics · Sociology · History · Language and cultural evolution · Linguistic Variation and Morphology · Mathematics · Philosophy · Syntax, Semantics, Linguistic Variation

Tools

Open DOISci-Hub