Date: Wed, 8 Jun 2005 10:25:31 -0400 From: Keira Ballantyne < firstname.lastname@example.org > Subject: Deixis and Demonstratives in Oceanic Languages
EDITOR: Senft, Gunter TITLE: Deixis and Demonstratives in Oceanic Languages PUBLISHER: Pacific Linguistics YEAR: 2004
Keira Gebbie Ballantyne, Department of Linguistics, University of Hawai'i at Manoa
This edited volume contains an introduction, short descriptions of the deictic systems of seven Oceanic languages, plus Malcolm Ross's final chapter which partially reconstructs aspects of the Proto Oceanic deictic system and considers some common diachronic changes to that system. The book is likely to be of interest not only to scholars of Oceanic and Austronesian languages, but to typologists interested in the scope of deictic systems across languages.
Senft's introductory chapter serves to outline the notion of deixis. He briefly catalogs concepts important to the study of deictic systems and the linguistic resources pertinent to deixis, and summarizes the various papers in the anthology.
Malcolm Ross's first contribution to this volume, "Aspects of deixis in Takia", begins with a brief introduction to pertinent features of the structure of this papuanised member of the Bel subgroup, spoken on the island of Karkar and by small populations elsewhere in the Madang province of Papua New Guinea. The paper goes on to describe a three-way contrast preserved across six series of demonstrative, locative and manner forms.
Perhaps the most intriguing aspect of Ross's description is the semantic contrast which pervades this system. Takia contrasts speaker proximal, distal, and a third series, the "a-series", which Ross argues was historically a hearer proximal but has lost its deictic function in the modern language and is used as a "pragmatic-definite" marker. The set of circumstances under which a-series forms may refer include when the referent is discourse old, when it is visible, and when the referent is new but identifiable by means of modification. Apparently referents which are part of general or encyclopedic knowledge may not take a-series demonstratives; I note in passing that this appears to be an exception to theories of accessibility in which referents identifiable by means of the NP alone are generally less accessible than those which are already in long-term memory (e.g. Gundel Hedberg & Zacharski 1993).
The discussion of directional and positional verbs includes an interesting set of observations on the effects of papuanisation on chains of such items. In canonical Oceanic languages, when a transitive verb of motion is followed by a second verb, the object of the first is coreferential with the subject of the second. Unlike typical Oceanic systems, Takia also contains patterns in which the subjects are shared. The paper also notes that Takia lacks simple indigenous compass-point terms, using Tok Pisin borrowings for "north" and "south", and circumlocutions ("from where the sun rises/sets") for "east" and "west".
Anna Margetts' contribution focuses on the semantic and syntactic features of the deictic system of Saliba, a Papuan Tip language spoken on Saliba and adjacent islands as well as on the nearby mainland in the Milne Bay province of Papua New Guinea. Saliba preserves a tripartite speaker-based contrast across four form classes. Margetts notes an asymmetry between the speaker proximal and hearer proximal forms, finding that a greater area is covered by the former, and that under certain circumstances the speaker proximal form can intrude into the addressee's personal space (a particularly nice example shows a speaker proximal form accompanied by a touching gesture used to refer to part of the addressee's body).
The Saliba deictic paradigm is comprised of four formal classes. The free emphatic forms may be pronouns or adnominals, and are often accompanied by the clitics. Formally they are bimorphemic, being composed of the speaker proximal clitic form plus a deictic base -- Margetts argues that the speaker proximal form contributes to their emphatic nature. The clause final demonstrative class members instantiate the monomorphemic base form, can only be pronominal and occur for the most part in non-verbal clauses. The set of place adverbs (base + i) express stative location or the goal of motion.
The fourth form class has two distinct functions. In the first, this set of forms are determiner enclitics, and this paradigmatic class encompasses not only the tripartite deictic contrast but also a dedicated givenness marker which is not deictic in nature. Margetts further notes that the distal enclitic is extending its semantic domain and appears to be coming to mark specificity. The second function borne by this form class is that of the demonstrative particles, which are distinguished from the determiner clitics by bearing stress and by appearing in non-verbal constructions.
Senft's contribution describes selected aspects of the deictic system of Kilivila, a Papuan Tip language spoken in the Trobriand Islands in the Milne Bay province of Papua New Guinea. Kilivila is remarkable for its sizable set of noun classifiers (Senft's term of choice is "classificatory particles"). Kilivila demonstratives distinguish proximal, medial and distal terms, which may be divided into two classes. The first of these classes requires some sort of ostensive gesture for felicitous use. The second set is built around classificatory particles and has a particularly transparently iconic morphology with the medial containing an extra morpheme and the distal containing both that extra morpheme and an elongated vowel (rendered orthographically as a triplet).
Senft uses the "tabletop space" model developed by Pederson & Wilkins (1996) as an elicitation tool for investigating the way in which the tripartite system is applied to objects at various spatial distances. Given the resources of their deictic system, speakers make use of reference to proximity of interlocutors, demonstratives, and adverbials such as "in the middle" to distinguish between multiple items spread out upon a table.
This piece also briefly touches on the use of deictics in the vertical dimension, anaphoric uses of deictic demonstratives, locatives derived from body-part terms and ways of indicating direction in Kilivila.
The first of the two Polynesian languages considered in this volume is Næss' account of Pileni, a Polynesian outlier spoken within the Reef Island group in the Solomon Islands as well in some small communities on Santa Cruz. Deixis in Pileni is speaker-based with a tripartite system of speaker proximal, hearer proximal and distal. Pileni is remarkable for its seven postverbal directional particles, three of which relate to the speaker-based deictic dimensions, three of which describe vertical motion and a final particle which has the general meaning "away". Deictic demonstratives in this language are found on nouns, on verbs, and within relative clauses, and Næss suggests that the latter forms may be moving toward becoming grammaticalized as relative pronouns. A highlight of this paper is the exploration of the discourse functions of deictic demonstratives, with Næss showing that the contrast between the hearer proximal and the distal maps on to anaphoric distance. The speaker proximal is reportedly reserved for spatial deixis in Pileni discourse.
Næss argues that there are certain facts which suggest that the Pileni system is moving from speaker-based contrast to a distance-based system. Evidence for this contention includes the fact that the hearer-proximal form is used more frequently, is fixed in a number of idiomatic expressions, and appears to be the default or unmarked demonstrative for anaphors in discourse. Although these facts point to a clear asymmetry between the speaker proximal and the addressee proximal (of the type observed in the Margetts paper), it is not clear to me from the present work why these facts about the Pileni hearer proximal mean that it is moving toward being reanalyzed with a medial function.
Bril's contribution to the volume considers the system of deixis in Nêlêmwa, a language of the Kanak family spoken on the northwestern coast of New Caledonia.
Bril describes a system composed of three deictic markers, three anaphoric markers and five directionals, all of which suffix to various classificatory material (including a small set of nominal roots, number markers, a locative root, a temporal root and some kinds of pronouns) -- the resulting compound is a determiner. A particularly fascinating aspect of this language is the observation that determiners are prenominal when the NP referent is new and postnominal when it is given. Bril notes that in certain situations of emphasis, a noun may take both a pre- and a post-nominal determiner. In the current paper, the exegesis of this system is necessarily brief, but I look forward to more detailed analysis of this in future work.
In addition to attaching to classificatory material in order to form determiners, Nêlêmwa deictic suffixes participate in the formation of independent and presentative pronouns, locative and temporal adverbs, and the predicate expressing similarity.
Bril additionally examines the metaphorical extension of spatial and directional deictics. The directionals expressing "up" and "down" are mapped aspectually to completion and ongoing actions, respectively; and in addition to the familiar alignment of up/inland and down/oceanward, up/down map to the cardinal directions -- down is oceanwards, north and west; up is inland, south and east.
The paper concludes by briefly touching on the rich use of spatial and directional morphosyntax in Nêlêmwa narrative, which Bril interprets as emerging from the fact that the TMA system in this language contains few distinctions.
Ozanne-Rivierre's paper concerns deixis in Iaai, a language of the Loyalty Islands subgroup of New Caledonian, spoken on Uvea. Like Kilivila and Nêlêmwa, Iaai also attaches deictic markers to classificatory material. The set of roots to which these markers attach is, however, far smaller in Iaai, consisting of two static markers (specified vs. unspecified) and two dynamic markers (source and goal). These compound forms are then available to modify nouns or verb phrases.
Again we find the system of spatial deixis mapping to the temporal domain; in this case, the speaker and hearer proximal terms refer to proximal time, while the distal form indicates the distant future. Past in Iaai is conceived as being "down".
Ozanne-Rivierre also analyzes the alignment of relative and absolute directional terms. This paper presents a particularly clear analysis of the effects of topography and geography in an Oceanic locative-directional system. The island of Uvea curves around a lagoon to the east and has high cliffs on its convex west coast. Ozanne-Rivierre points out that alignments within the deictic system are motivated by the particular geographical specificities of the Uvea landscape, giving rise to two sets of polysemies: down/ west/ oceanward and up/ east/ inland. She further points out that the forms indicating east and west can be used in a relative, local context to mean up or down without evoking the cardinal directions.
Mosel's analysis of demonstratives in Samoan draws on perhaps the broadest range of data in this selection, incorporating data from elicitation and field observation, traditional oral and written narratives, modern short stories and email correspondence. Her analysis and description covers the syntax and semantics of the deictic systems of demonstratives, local nouns, directional particles, deictic verbs and temporal adverbial particles, with a heavy emphasis on the complex system of demonstratives. The Samoan demonstrative system shows contrasts along the parameters of participant orientation, spatial orientation, a formal/informal contrast, and contrasts of singularity, plurality and specificity.
The syntax of local nouns, deictic verbs (which are of the form "be/do/say/think like this/that"), pronominal and adnominal demonstratives and temporal adverbials is described and analyzed. In accounting for the semantics of these forms, Mosel separates situational from non-situational uses, and in the latter category investigates discourse deixis and reference tracking in narrative data. She finds that it is the social rather than the distance-based parameter which maps to non-situational uses of these deictics. The form which expresses an item close to or in the possession of the speaker is exploited for cataphoric discourse deixis and for the tracking of prominent thematic participants. Mosel interprets this as expressing that the speaker has some information in his or her possession that will be imparted to the addressee. Conversely, the form used for items in the possession of or close to the hearer is mostly found in anaphoric contexts, to express information that the hearer already possesses.
In the final contribution to this volume, Ross reconstructs a four-way semantic contrast in the Proto Oceanic (POc) demonstrative system -- a tripartite speaker-based paradigm plus an additional fourth term reserved for anaphoric use. Although the way in which languages carve up the semantic space of the demonstrative system has remained fairly stable over time, reconstructing the forms of these items proves to be an intransigent problem.
Rather than reconstructing a single paradigm for POc demonstratives, Ross's analysis posits multiple sets. He hypothesizes a rapid dispersion of POc speakers contemporaneous with the loss of the Proto Malayo-Polynesian (PMP) voice system. The sets of reconstructed demonstratives derive from PMP demonstrative bases and from the neutralized PMP voice markers -- the contention here is that as speakers dispersed, they took different subsets of this disintegrating voice system and co-opted the morphology for use in the determiner system.
Far more amenable to analysis are the local nouns found throughout the Oceanic family. Two classes -- or rather, as Ross is careful to argue, two constructions -- of local nouns can be reconstructed for POc, free local nouns and inalienably possessed local nouns. The free nouns for the most part have both a locational meaning and a common noun meaning. When the inalienably possessed nouns have a common noun counterpart, it refers to a body part (a well described phenomenon throughout the language family). Ross also points out a robust relational local construction in which an adposition is combined with a local noun to indicate a location with respect to some entity. He notes that in some languages the local noun has however been reanalyzed as an adposition (although the examples in this section leave me unsure whether Ross interprets all of the examples where the adposition is lost as examples in which the local noun has been reanalyzed in this fashion, or whether there is some intermediate step).
Ross finishes by examining systems of directionals in Oceanic, concluding that POc had a series of directional verbs which often occurred as final verbs in serial constructions, and that these were reanalyzed as directional clitics in a number of daughter languages. Some languages further shift the category of these items by analyzing them as the second element of a compound verb.
The organization of a set of papers like this in a single volume will be of particular utility to scholars who are trying to come to grips with describing deixis in an Oceanic language -- each of the seven chapters dealing with the system of a particular language could very well stand as a chapter on deixis in a comprehensive grammar.
The sample of languages considered is rather skewed toward the western edge of the Oceanic region, and the inclusion of two Polynesian (Pileni, Samoan) and two Papuan Tip (Saliba, Kilivila) languages in the sample of seven leaves some subgroups unrepresented (in particular, a description of a Micronesian language would have been a welcome addition). Given, however, the vast number of languages in the family, and the relatively small fraction for which complex contextually sensitive data are available, the skewing of the sample is more a reflection of the availability of data than of editorial oversight.
A particularly intriguing observation which emerges again and again is the asymmetry between speaker proximal and hearer proximal forms in speaker-based deictic systems (e.g. remarks by Ross on the pragmatic-definite function of the hearer proximal in Takia, by Margetts on the unmarkedness of the speaker proximal in Saliba and by Næss on the similar function of the hearer proximal in Pileni). No broad theoretical analysis is offered for this pattern, although Mosel's analysis of the interaction between social and spatial information is insightful and I suspect similar analyses might account for some of the puzzling data. Also absent is any data showing distinctions between items introduced by the speaker and those introduced by the addressee in dialogue -- such data has shed light on asymmetries between proximal and distal forms in languages like English (see e.g. Strauss 2002).
Finally, a particularly strong point of the volume is the inclusion of a great deal of data in context. Not only are single sentence examples presented with ample explication of their discourse context, a number of the papers include fairly extensive paragraph-length examples of deictics in context, and two of the contributions have appendixed texts (Bril on Nêlêmwa, Ozanne-Rivierre on Iaai).
All in all, the volume represents a significant contribution to the field of Oceanic linguistics. Despite the fact that it does not answer all of the questions that it brings up, its coherent articulation of both the breadth of and the gaps in our current knowledge is sure to inform and stimulate the reader.
Gundel, Jeanette K., Nancy Hedberg & Ron Zacharski. 1993. Cognitive Status and the Form of Referring Expressions in Discourse. Language. 69(2):274-307.
Pederson, Eric & David Wilkins. 1996. A cross-linguistic questionnaire on 'demonstratives'. Manual for the 1996 Field Season. Cognitive Anthropology Research Group. Nijmegen. 1-13.
Strauss, Susan. 2002. This, that and it in spoken American English: a demonstrative system of gradient focus. Language Sciences. 24: 131-52.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Keira Gebbie Ballantyne recently finished her Ph.D. at the University of Hawai'i at Manoa. Her doctoral work investigated reference (including the demonstrative system) and tense-mood-aspect in narrative and non-narrative discourse in Yapese, an Oceanic language spoken in Micronesia. Her research interests also include typological and comparative approaches to textual structure, corpora of underdocumented languages, and cognitive and functional approaches to language.