CRLC Workshop 2006


The How and Why of Syntactic Relations
Randy LaPolla
(La Trobe University)

This paper starts with a discussion of the basic nature of communication as ostension and inference, and language structures as constraints on the inference part of communication, then moves on to how and why syntactic relations (which are seen as individual constraints on the inference of reference and role identification) have grammaticalized in some languages through repetition of a particular pattern of (co)-reference often enough for the inference of (co)reference to become conventionalized, and what this means to those languages in terms of clause structure and to the speakers of the language in terms of communication, and also discusses languages where there has not been grammaticalization of syntactic relations such as those in English, or where there has been conventionalization of a different set of constraints on referent identification, and what this means for those languages and their speakers.

Back to programme

Yoking phonology in the service of syntactic reconstruction:
the case from Tangkic
Erich Round
(Yale University)

In the 1996 grammar of Kayardild, Evans reconstructs a detailed picture of the history of Tangkic syntax. The reconstruction focuses mostly on the grammatical functions attached to morphological markers. Here I show that ongoing investigations into the historical phonology of Tangkic can begin to shed light on another aspect of grammar, namely word order.
Within a larger research project on the (phonological) history of the Tangkic system of nominal morphology, a number of sound changes can be reconstructed as having occurred at domain boundaries, particularly deletions at right edges. From this we can derive certain insights into which word classes did or did not appear at these right edges.
In one instance, so-called ‘truncation’ is reconstructed for early proto-Tangkic as having been a process which was obligatory phrase finally, and optional word finally. This system is altered in all daughter languages, and in one particularly telling way in Kayardild. Amongst the languages for which we have extensive documentation, Kayardild is distinguished in having retained proto-Tangkic truncation mainly within idiosyncratic allomorphs of morphological suffixes. Moreover, in just a few such allomorphs, one finds a modern alternation between forms in final /-rpa/ versus /-ra/. This can be linked straightforwardly to the postulated original obligatory truncation of /-rpa/ > [-r] in phrase final position. When the truncation rule changed so as to only delete final -a, new allomorphs in /-ra/ were inferred from surface [-r], but only for morphemes which originally occurred phrase finally. This hypothesis gains direct support, for example, from compounds in /cuŋarpa-X/ versus /X-cuŋara/, both containing */cuŋarpa/ ‘big’. Pursuing this line of argumentation, we also infer (1) that ‘prior’ subordinate clauses preceded their main clauses; (2) that the order in NPs was for ascriptive (adjectival) nominals to precede substantives; (3) that Evans correctly derives Kayardild /-arpa/ nominalisations as coming from now headless, erstwhile pre-head relative clauses. Finding (2) is further corroborated by the evidence from another, somewhat earlier, phonological change involving loss of word-final nasals in phrase internal positions (i.e., before a following consonant in connected speech). Here, although the evidence is not as clear cut, we can infer that predicative nominals tended to come phrase finally, suggesting a SUBJ V NPRED or V SUBJ NPRED order.
Some concluding remarks will be offered on the prospects for similar evidence to be employed in the syntactic reconstruction of other Australian languages: the outlook is positive, given the reliance here on diachronic changes which are expected, on phonological grounds, to be found relatively commonly in Australia.

Back to programme

Syntactic variation in Shetland Dialect

Dianne Jonas
(Yale University)

The topic of this talk is the verb syntax of Shetland Dialect, a variety of Scots spoken in Shetland. Shetland Dialect retains verb movement operations of various types: main verb movement in interrogatives and imperatives, and verb movement to the left of negation for a subset of main verbs. A discussion of these facts is presented together with an overview of auxiliary use and agreement facts in expletive constructions.

Back to programme

Directional Verbs as Kickoffs in Grammaticalization Processes:
Cantonese Verbal Particles as an Illustrative Example
Winnie Chor
(University of Sydney)

Cantonese is noted for its rich inventory of grammatical particles and it is this distinctive nature and pervasive use of particles that sets Cantonese apart from other Chinese dialects. While theories of grammaticalization have been flourishing in the last few decades (Heine, Claudi & Hunnemeyer 1991; Traugott & Heine 1991; Hopper & Traugott 1993; etc.), it was not until the late 1980’s that Chinese scholars started to discuss the phenomenon more systematically within a theoretical framework (Sun 1994; Shen 1994 & 1998; etc). In comparison, relatively little has been done on Cantonese from the perspective of grammaticalization. The aim of this paper is to report on some recent observations by the author in this direction (Chor 2004; Luke & Chor 2005; Chor 2005). Adopting a broadly functional approach, this paper attempts to show how verbs with a very particular directional/spatial meaning can get turned over time into a particle with very interesting, diverse, non-directional idiosyncratic properties. These include: the grammaticalization of faan1 (back/return) from a directional verb to a tone-of-voice particle; the extension of gwo3 (over) from giving directional meanings to the suggestion of experience and repetition; and, the development of hei2 (up) from a directional verb to a temporal particle, signifying “completion”. As suggested by cross-linguistic evidence, the use of the directional/spatial domain as a source domain for metaphor is common universally, and plays an important role in grammaticalization. The observations made in this paper are primarily based on a synchronic corpus consisting of about eight hours of genuine, colloquial, unscripted and spontaneous speech, and, records of historical materials (such as old Cantonese dictionaries, textbooks for Cantonese learners, documents written by missionaries, and, transcriptions of Cantonese films in Hong Kong from the 1940’s to the 1990’s). It is hoped that this study can provide insights to account for how and why meanings of directional particles in Cantonese are enriched and to shed light on grammaticalization processes in Cantonese generally.

Chor, Winnie 2004. A Semantic and Pragmatic Study of Verbal Particles in Cantonese. MPhil thesis. Hong Kong: University of Hong Kong.
Chor, Winnie 2005. “Two Completive Particles in Cantonese: ZO & HEI”. Paper presented at the 10th International Conference on Yue Dialect. Dec 12-14.
Heine, B., Claudi, U. and Hunnemeyer, F. 1991. Grammaticalization: A Conceptual Framework. Chicago: University of Chicago Press.
Hopper, Paul J., & Traugott Elizabeth C. 1993. Grammaticalization. Cambridge: Cambridge University Press.
Luke, Kang Kwong & Chor, Winnie 2005. “The Grammaticalization of faan1: from Verb to Expressive Particle”. Seminar presentation. Chinese University of Hong Kong.
Shen, Jia-xuan 1994. “Grammaticalization – an Overview”. In Foreign Language Teaching and Research, 4: 17-24.
Shen, Jia-xuan 1998. “Mechanism of Grammaticalization”. In Contemporary Linguistics, 3: 41-46.
Sun, Chao-fen 1996. Word-Order Change and Grammaticalization in the History of Chinese. Stanford, California: Stanford University Press.
Traugott, Elizabeth C. & Heine, Bernd 1991. Approaches to Grammaticalization. Amsterdam and Philadelphia: John Benjamins.

Back to programme

Does Grammaticalization Exist in Medieval Chinese?

Priscilla Ngan Yuk Han

(Centre de Recherches Linguistiques sur l’Asie Orientale)

Grammaticalization is a human linguistic cognitive process. Its existence in different languages is proved by numerous studies since the last century. The process is considered to be unidirectional. However, in recent years, its unidirectionality is reappraised. Besides, two terms denoting the reversal of grammaticalization are put forward: lexicalization and degrammaticalization. Much controversy regarding these two terms has been made until now but the conclusion about the “legitimate term” representing the reversal of grammaticalization is not yet drawn.
For this reason, we searched for examples of lexicalization in Chinese, a non-flexional language, in order to see if it represents really the reversal of grammaticalization. Here are two examples:

1. gong zhu tong hao (Classical Chinese)

make-public it/them-to too like

Make it/them public to (people who) like it too.

2. ni beng gen ta shuo (Modern Chinese)

you don’t to him tell

Don’t tell (it) to him.

In the above two examples, “zhu” is the contraction of “zhi (it/them) + yu (to)” and “beng” the contraction of “bu (not) + yong (necessary)”. These two lexemes are typical examples of lexicalization and they do not result from the reversal of grammaticalization of any syntactic entity.
These two examples show that lexicalization has no relation with the reversal of grammaticalization in Chinese. So the only term left is degrammaticalization. Since Chinese is a non-flexional language, its degrammaticalizational mechanisms are reanalyse, metaphorical shift and pragmatic inferencing.
Among the different periods of Chinese syntax, we chose the medieval one because it is the first period of the history of Chinese syntax in which the grammaticalization of the second verbs in “verb1 + verb2” constructions is proved. We are interested in finding examples of the reversal of grammaticalization of this period so as to see if degrammaticalization exists in Chinese and if grammaticalization coexists with its parallel reverse phenomenon.
We will undergo a thorough examination of two works representing the language used in medieval China: Luoyang qielan je, the history of the temples in Luoyang and Bai yu jing, Sutra of the 100 Parables. Then we will present the results in statistics after analysing the examples. Finally, we will draw the conclusion.

Back to programme

The diachronic evolution of directional complements in Mandarin
James McElvenny
(University of Sydney)

Directional complement constructions are a family of verb complement constructions in Modern Mandarin and other modern Chinese dialects whose typical function is to indicate a direction of motion associated with the action or event expressed by a verb. For example, in the sentence ta zouchulai 'He walks out hither', the verb zou 'walk' expresses the action of walking and the directional complement chulai indicates that the action is performed going from inside to outside and in the direction of the speaker. Directional complements can also have a variety of abstract senses that are derived from their basic directional semantics through metaphor. For example, the same complement can be used with non-motion verbs to indicate that something becomes known as a result of the action described by the verb. This is perhaps an instance of the metaphor 'to be outside is to be known.' In the phrase kànchulai 'tell from looking', the verb kàn expresses the action of looking and chulai indicates that something becomes known as a result of the act of looking.
In my paper, I trace the diachronic evolution of the directional complement constructions from their earliest forms as independent verbs in serial verb constructions in pre-Qin times up to their present state in Modern Mandarin. I identify the various formal and functional properties of the constructions at each stage in the history of the language and show how these properties change from one stage to another. I also investigate the factors that condition these changes.
My research is based on a corpus of vernacular texts that cover each period in the development of the constructions from pre-Qin times up to the present. In discussing the grammaticalisation of the directional complements from independent verbs, I treat the forms at each stage of the grammaticalisation process and their associated semantic and pragmatic functions as ‘constructions', as that term is understood in the theory of construction grammar (Kay, Fillmore and O'Connor 1999), rather than as the products of generative rules. In my analysis of the evolution of the semantics of the constructions, especially the extended metaphorical senses of many of the constructions, I draw on the theory of cognitive semantics.
KAY, PAUL AND CHARLES J. FILLMORE. 1999. Grammatical constructions and linguistics generalizations: the What's X doing Y? construction. Language, 75.1-33

Back to programme

From Ethic Datives to Number Markers in Murriny Patha
Joe Blythe
(University of Sydney)

Murriny Patha, a polysynthetic language from Northern Australia, makes a four way number distinction: singular, dual, paucal and plural; as well as a sibling versus non-sibling distinction for dual and paucal participants. The distinction is made through the absence or presence of dual and paucal non-sibling number marking morphemes that combine with subject and object-marking pronominals. Different combinations distinguish different groups of participants. In this presentation I wish to propose that two members of a no-longer productive series of ethic dative bound pronouns were reanalysed as dual number markers, thus enabling a gender distinction to be made in the duals. So doing, I’ll lay out some of the pathway Murriny Patha may have gone down in order to move from a system which had minimal gender marking to a system with greatly enhanced gender marking.

Back to programme

Interactions between pragmatics and syntax in the development of Australian pronominal clitics

Ilana Mushin and Jane Simpson

(University of Queensland and Sydney University)

Many Australian languages have both bound and free pronouns, either as verbal prefixes or as enclitics. Enclitic pronouns vary in their attachment possibilities, but are largely found prosodically attached either to the initial constituent (ie. in second position) or to the predicate. In this paper we present cross-linguistic evidence of the range of pragmatic and syntactic pressures that lead to the development of different types of pronominal enclitics.
Capell (1956) has proposed a general path of development of bound pronouns in Australian languages, claiming that they begin their lives as free pronouns and become prosodically bound and reduced over time. At such time as these forms are fully bound, new free forms emerge to replace the old system of free pronouns.
While this scenario accounts for the formal process of grammatical change, it does not explain why free pronouns should find themselves in positions to develop into enclitics. Free and clitic pronouns not only differ in their formal properties, but they also serve different grammatical and discourse functions. Clitic pronouns represent the person, number and sometimes noun class features of arguments. These forms are usually obligatory in the languages which have them. It is debateable whether such forms might be analysed as the arguments themselves, or as agreement markers to external arguments (e.g. Jelinek 1984, Austin & Bresnan 1996). Free pronouns are usually said to act like nominals, with respect to case-marking and distribution. In languages with bound pronouns, they are usually not obligatory and may be used in discourse in contexts of emphasis or contrast.
Here we examine the distribution of free pronouns in the discourse of languages which have no bound pronouns (Jiwarli), languages which have free and bound pronouns (Warlpiri, Kalkatungu, Nyamal), languages which have primarily bound pronouns (Warumungu) and languages in which free pronouns appear to be following a path to becoming bound pronouns (Garrwa).
We show that in some of these languages, the distribution of free pronouns differs from that of nominals. In languages which have only free pronouns these have two distinct information structure properties - as continuing topics, and as emphatic contrastive elements. These properties are associated with different syntactic positions, in particular, the positions for continuing topic are very often post-verbal or following the contrastive information. These preferences then lead on the one hand to post-verbal bound pronouns and on the other to second position clitics.
Austin, Peter and Bresnan, Joan. 1996. Non-configurationality in Australian Aboriginal languages. Natural Language and Linguistic Theory 14:2:215-268.
Capell, Arthur. 1956. A new approach to Australian linguistics: handbook of Australian languages, part 1. Sydney: University of Sydney.
Jelinek, Eloise. 1984. Empty categories, case, and configurationality. Natural Language and Linguistic Theory 2:1:39-76.

Back to programme

Clause merger and the evolution of the Iwaidja/Mawng reciprocal construction

Nicholas Evans, Ruth Singer and Bruce Birch

(University of Melbourne)

Iwaidja and Mawng (non-Pama-Nyungan, Northern Territory) have a cross-linguistically unusual construction for encoding reciprocity: to express 'they two saw each other', one says the equivalent of 'he-her-saw and'.
Historically, this appears to derive from a truncated biclausal construction from which the second verb has been ellipsed - something like 'he-her-saw and she-him-saw'.
Synchronically, however, there is clear evidence that there is a single clause. Firstly, other material from the first clause can be placed outside the 'in turn' pronoun, such as theme objects of ditransitives, or other arguments of three-place constructions: 'he-her-gave and food' for 'they gave each other food', and ' and lice' for 'they searched each other for lice'. Secondly, the 'in turn' pronoun, for some subject/object combinations, is not always the one which was not the subject of the first clause, showing that a single constructional clause-level template is determining the choice of bound and free pronouns.
In this talk we reconstruct the historical changes which led to this highly unusual case of clausal merger, drawing on comparative morphosyntactic evidence but also on evidence from prosody which shows that particular phonological reductions of the 'and PRONOUN.IN.TURN' occur in the reciprocal construction but not in true contrastive subject uses.

Back to programme

Substantival Adjectives in the History of English and the Nature of Syntactic Change

Cynthia L. Allen

(Australian National University)

Present Day English (PDE) differs from most other Germanic languages in generally not permitting attributive adjectives to be used in what is traditionally referred to as a ‘substantival use.’ In PDE, we cannot use the phrase a blind to refer to a single blind person, but in Old English (OE), it was possible to use se blinda ‘the blind man’ to refer to an individual, as is also the case in other modern Germanic languages generally. Adjectives can be used substantivally in some very restricted situations, as when the blind is used generically to refer to blind people.

Since the other Germanic languages retain some inflection on adjectives, while English lost all inflection on adjectives during the Middle English (ME) period, it is natural to attempt to link the loss of inflection with the requirement that attributive adjectives must be followed by a nominal head. This is the traditional view, expressed, for example, by Brunner (1960-62: 74), Fischer (1992: 222), and Rissanen (1997: 99-100). By this view, no claim is made that substantival adjectives became impossible with the loss of inflection, only that the loss of inflections probably played an important role in the development of a new system, in which a prop-word such as one was commonly used, and the expression of a head eventually became obligatory. However, a stronger and more deterministic link between adjectival morphology and the grammaticality of substantival adjectives has been assumed within the recent generative tradition; specifically, Kester (1996) proposes that a pro head is licensed by inflectional features of adjectives. Using the results of a corpus study, I will show that some predictions that Kester’s proposal makes are falsified by the data. The similarities in the behaviour of adjectives in ‘inflection rich’ and ‘inflection impoverished’ ME dialects are difficult to reconcile with the idea that adjectival morphology is essential for the non-expression of a head N. I also argue that a performance-based models such as those proposed by Bresnan, Dingare, and Manning (2001), Hawkins (Hawkins 2004 ), etc. to account for how ‘soft constraints’ or preferences can become grammatical rules is a more promising approach to this particular sort of syntactic change than the ‘competing grammars’ approach widely adopted in generative theory (Kroch 2001).

Back to programme

Adding to the class of modals: Ought it be considered grammaticization?

Barbara Kelly

(University of Melbourne)

This paper investigates whether ought is grammaticizing toward full modal status in two dialects of English. It examines the current spoken use of ought and traces use changes. The development of a class of modals from a class of full verbs occurred along a cline of grammaticization from the original source meaning as an independent verb to current auxiliary uses. At present ought is considered to be an auxiliary verb and it has been classed as a quasi-modal (Harris 1986, Krug 2000, Myhill 1997, Nordlinger and Traugott 1997, Palmer 2001). Beyond outlining current uses of ought this paper illustrates that the class of modals is not a closed set as suggested by Bybee (1995) and Bybee, Perkins and Pagliuca (1994). It shows that new modals can develop despite not following the same path to grammaticization as existing modals.
Data for this study is from three databases of spoken American and British English. These databases were investigated to examine frequency and types of use of ought in spoken English discourse.
When we examine use of ought in Present Day English we find evidence that ought is being used consistently in two different ways in British English, one of which is only minimally evident in American English. This is illustrated in Table 1, which indicates the breakdown of overall frequency of use of ought across the database.


British English

American English


ought + to 

13 (46%)

8 (89%)


ought + infinitive

15 (54%)

1 (11%)



28 (100%)

9 (100%)


Table 1. Uses of ‘ought’ across British and American English databases
Table 1 above indicates that for American English uses, ought with a directly following infinitive –to is the most frequent type of use found, e.g. You ought to hear that guy play while for British English ought with no following infinitive is the most frequent form used, e.g. You ought not see him undressed. American English uses of ought+to totalled 89%, while in the British English corpus ought+to accounted for 46% of uses. This investigation of spoken British and American English indicates that in British English, ought is following the same morphosyntactic historical path toward full modal status as other verbs have taken, such as should, can, and must.

Back to programme

How to get from have to have got

Heidi Quinn

(University of Canterbury)

This paper looks the syntactic properties of stative HAVE in two corpora in the Origins of New Zealand English (ONZE) archives, and proposes that the form HAVE got is the by-product of the categorial reanalysis of stative HAVE from a functional to a lexical head. Data from the Canterbury Corpus and additional empirical surveys suggest that in current NZE, DO+have is common in negative declaratives (1) and tag questions (2), even though HAVE got is clearly favoured in corresponding positive present tense sentences with a stative interpretation (1)-(2). In the past tense, we consistently find had in positive declaratives (3), and DO+have in negatives (4) and questions (5).

(1) and we don’t actually have any normal light bulbs but we’ve got the long ones (fyn01-10a, female, nonprofessional, born 1981)

(2) They’ve got quite a big house, don’t they.

(3) we had a Zodiac car Mum’s still got her Zodiac car (fyn95-13, female, nonprofessional, born 1965)

(4) cos they didn’t have an answer phone (fyn98-1, female, nonprofessional, born 1978)

(5) what did they have? [training options] (fon94-25c, female, nonprofessional, born 1940)

Instances of HAVE got do already occur in the earliest subcorpus of the ONZE archives (the Mobile Unit corpus), but the majority of speakers in this corpus appear to favour HAVE without got in positive present tense utterances (6). Questions and sentential negation with stative HAVE are rare in the Mobile Unit corpus, but the negatives in (6)-(7) and the question in (8) suggest that, for at least some speakers of early NZE, stative HAVE used to have auxiliary-like properties both in the present tense and in the simple past.

(6) Aunt Izy here has one of his medals no . hasn’t (mu-41b, male, born 1871)

(7) oh they hadn’t the quantities (mu-1c, female, born 1894)

(8) eight hundred was our population and what have we now? (mu-1c, female, born 1894)

I propose that the increasing popularity of HAVE got in positive present tense utterances and DO+have in negatives and questions is due to a categorial reanalysis of stative HAVE: In early varieties of (NZ)English, stative HAVE was a functional head, whereas in current NZE it is a lexical verb. In the proposed analysis, the additional got is the pronounced copy of a stative HAVE that has undergone short movement to a functional head projected in the simple present but not in the simple past.

Back to programme

Vowel Grade Harmony in Syntactic Change

Lawrence A. Reid

(University of Hawai’i)

The concept of vowel harmony has long been recognized as a factor interfering with the normal processes of sound change in language. Similarly, the concept of vowel grade by which morphological features are signaled by change in the quality of the vowel of a given form has long been recognized. The purpose of this paper is to demonstrate that vowel harmony not only interferes with the normal processes of sound change within a word, but can operate across structural boundaries in a sentence to create vowel sequences that mark the same (or similar) morphological features or “grades”. Noun phrases in many Philippine languages are commonly described as being introduced by “phrase markers” which specify certain syntactic and semantic features of the noun phrase they introduce. There are typically unstressed clitic forms having a CV or CVC shape. However the quality of the vowel varies from language to language, thus in Ivatan, the forms which introduce common noun phrases all have an /u/ vowel, while those which introduce personal noun phrases all have an /i/ vowel, while in Tagalog the forms which introduce common noun phrases all have an /a/ vowel, while those which introduce personal noun phrases all have an /i/ vowel, like Ivatan. Recognizing that the similarity in vowel quality of “phrase markers” in a language is commonly the result of vowel grade harmony, and is not necessarily the result of regular phonological change, provides an explanation for the multiple irregularities that are found in attempting to reconstruct the protoforms of the “phrase markers”.

Back to programme

Development of Applicative Constructions in Austronesian Languages

Ritsuko Kikusawa

(National Museum of Ethnology, Osaka, Japan)

In this paper, I will discuss morpho-syntactic changes that took place in Austronesian languages, focusing on applicative sentence structures and their development.

Austronesian languages spoken in Indonesia, Madagascar and Oceania are often described as having an applicative sentence structure as part of their system. The applicative sentence commonly shows the same sentence structure as the typical transitive sentence, but with different verb morphology. It has been claimed that the two (applicative and typical) transitive structures developed by converging several transitive structures that are reconstructible for Proto-Extra Formosan (Kikusawa 2003). Because verbs occurring in each transitive structure in Proto-Extra Formosan are considered to have carried formatives (affixes) that differed depending on the structure, a claim is made here that their reflex-structures (structures that developed from a certain proto-structure) and cognate-structures (structures that have developed from the same proto-structure observed in sister languages) must be identifiable by tracing cognate formatives occurring on verbs.

Based on this idea, integrating the results of morphological comparison of applicative verb forms, including those in Oceanic languages by Evans (2003) with the comparison of sentence structures, and following the method applied in the reconstruction of applicative sentence structure for Proto-Mayan (Mora-Marín 2003), I will illustrate the scenario as to how the earlier multiple transitive system developed into the two transitive systems in each language (or, language group). Special attention is paid to the methodology for determining reflex-structures in the daughters of a proto-language, and cognate-structures among sister languages.

Evans, Bethwyn. 2003. A study of valency-changing devices in Proto-Oceanic. Pacific Linguistics 539. Canberra: Pacific Linguistics.
Kikusawa, Ritsuko. 2003. The development of some Indonesian pronominal systems. In Current Issues in Linguistics Theory. Historical linguistics 2001: selected papers from the 15th International Conference on Historical Linguistics, Melbourne, 13-17 August 2001, vol. 237, edited by Barry J. Blake, Kate Burridge and Jo Taylor, 237-268. Amsterdam: John Benjamins.
Mora-Marín, David. 2003. Historical reconstruction of Mayan applicative and antidative constructions. International Journal of American Linguistics 69(2):186-228.

Back to programme

On the gradual nature of syntactic change: support from split-ergativity in Maori

Anna Pucilowski

(University of Canterbury)

Harris and Campbell (1995: 49, 101) suggest that syntactic change by extension is a gradual but systematic process, applying to one word, or a group of words at a time. This paper argues that the present-day status of the Maori case system provides
evidence that the extension of an intransitive clause pattern to transitive verbs in the Eastern Polynesian group of languages proceeded gradually, from the least transitive to the most transitive verbs. It is generally assumed that Proto-Polynesian was an ergative language, like modern Tongan. In the Eastern Polynesian subgroup of Proto-Polynesian, an intransitive pattern with an absolutive subject plus oblique, exemplified in (1) from Tongan, was extended to all transitive verbs, originally as an imperfective construction (Clark 1973: 600). The original transitive pattern, as in (2), remained as a perfective construction. The resulting imperfective-perfective contrast was subsequently reanalysed as active-passive, as in an accusative system.

(1) Na’e kai a e tamasi’i i he ika
TAM eat ABS the child DO the fish
the boy ate some of the fish/partook of the fish’

(2) Na’e kaii a e ika e he tamasi’i
TAM eat.Cia ABS the fish ERG the child
the boy ate the fish’ (Clark 1973: 49)

I propose that the perfective to passive reanalysis did not take place in Maori because the Maori equivalent of (2) does not code perfectivity alone. Rather Maori is a split-ergative language, where the most transitive clauses are ergatively marked, while the less transitive clauses follow an accusative pattern. Following Hopper & Thompson (1980), I assume that transitivity is determined by a variety of factors, and I will show that, in Maori, the important factors are the affectedness of O, the number of
participants, the dynamism/punctuality of the verb, and the aspect of the clause. The evidence from Maori indicates that the extension of the intransitive pattern in Eastern Polynesian was a gradual process, that first applied to the least transitive clauses, and only spread to the most transitive clauses after Måori had split from the other Eastern Polynesian languages.

Clark, R 1973. Transitivity and case in Eastern Oceanic languages. Oceanic Linguistics, 12. 559-605.
Harris, A & L. Campbell. 1995. Historical syntax in cross-linguistic perspective. Cambridge: Cambridge University Press
Hopper, P & S. Thompson 1980. Transitivity in grammar and discourse. Language. 56. 251-299

Back to programme

Argentinian River Plate Spanish: The Preterit and the Present Perfect

Celeste Rodríguez Louro

(University of Melbourne)

This work investigates the use of the Preterit and the Present Perfect (PP) in Argentinian River Plate Spanish (ARPS). The evidence offered indicates that in contemporary ARPS the Preterit prevails over the PP and expresses canonical perfective meanings as well as most notions of temporal relevance encoded through the PP in other Spanish dialects. The PP is reduced to encoding canonical resultative and experiential meanings and to performing an innovative emotional function. This study examines email messages, 20th century sociolinguistic interviews in Buenos Aires, and letters composed by criollos (natives of pre-independent Argentina) in the 16th to 18th centuries. All data were coded for Past verbal forms (N=1238). The Preterit is widespread throughout whereas the PP represents 13% of the colonial data, 15% of the 20th century data, and only 2.04% of contemporary email data. Findings in this study challenge Comrie’s (1976) claim that in Romance the PP is forcing out the Simple Past. Further, they question grammaticalization proposals for an anterior to perfective cline cross-linguistically including Romance (Bybee, Perkins, and Pagliuca 1994) and different varieties of Spanish (Schwenter 1994.)

Back to programme

An activity of the

This Web Site was created by Pascale Jacq, maintained by the Administrator:
This document last modified: 11th November, 2003
Copyright © 2003 Centre for Research on Language Change, ANU.