HistLing Seminars,
Centre for Research on Language Change, ANU

Series 1, April-July, 2003

Wednesdays, 4pm. Coombs Seminar Room C, Coombs Building, Australian National University


30th April, 2003:

Evershed Amuzu

(School of Language Studies, Arts, ANU)

On Beyond Surface Grammatical Structures:
A language-production based grammatical analysis of mixed constituents in Ewe-English Codeswitching

(A Ph.D mid-term review seminar)

The truism in the labels 'Language Contact' and 'Languages in Contact' is that they do not refer to physical encounters of languages. Rather, they point to the fact that languages meet in the minds of the speakers who are/were exposed to them via bilingualism or some other kind of language contact (i.e. LC) situation. This metaphor of languages meeting in the minds of speakers invites one to anticipate finding in the LC literature a common assumption that the starting point of quests to understand the nature of LC phenomena is the determination of the nature of cognitive processes and constraints that guide the speakers to produce the bilingual grammatical structures that get labelled as one or the other kind of LC phenomena (e.g. Codeswitching/Codemixing, Borrowing, Language Intertwining, Language/Structure Change, Child Bilingual-Language Acquisition, Inter-language, Pidgin/Creole, etc).

Unfortunately, not much of the LC literature touches on the issue of cognitive processes in discussions of LC structures. Majority of LC linguists seem content with describing LC surface structure configurations exclusively. The avoidance of cognitive processes as integral aspects of the objects of LC research is due to widespread scepticism about the verifiability/Scientificity of hypotheses on abstract / mental processes. But the avoidance has led to a proliferation of often writer-specific and LC-specific explanations for the nature of LC data, a trend that is suggestive of the problem with the skipping of what should be the legitimate starting point in LC investigations.

It is against this background that we discuss the nature of structures of mixed constituents used by Ewe-English bilinguals with a theoretical framework that perceives LC phenomena as structural ramifications of some universal cognitive (i.e. language production) processes and abstract grammatical constraints. Specifically, we will go beyond the surface structures of various types of Ewe-English constituents to invoke the framework‚s hypotheses about how such surface structures have been derived from abstract grammatical structures. The framework is associated with Carol Myers-Scotton and it consists of the Matrix Language Frame model and crucially the more recently proposed Abstract Level model and 4-M model - i.e. model stipulating four universal kinds of morphemes (see Myers-Scotton 2002 for her latest exposition of the framework).

7th May , 2003:

Cynthia Allen

(School of Language Studies, Arts, ANU)

Our this fascinating construction:
investigating the interaction of determiners and possessives in the history of English

In Present Day English, it is not possible to combine a possessive and a determiner before a possessed noun, both *That our fascinating construction (henceforth 'Det Poss') and *our that fascinating construction ('Poss Det') are impossible. If we want to combine a demonstrative determiner and a possessive, we must use the 'Double Genitive' construction; e.g. That fascinating construction of ours. However, both the Det Poss and the Poss Det constructions are found in Old English (OE) texts, while the double genitive construction is not. While the Poss Det construction apparently disappeared by 1100 and has been ignored for the most part by research into diachronic English syntax, the Det Poss construction, which is also found in Early Modern English (EModE) texts, has frequently been assumed to be the forerunner of the Double Genitive construction.

This paper presents some results of a systematic investigation into the history of possessives in English and their interaction with determiners. I will argue that the absence of the Det Poss construction from the texts for more than three centuries is no mere data gap and that the usual assumption that this construction disappeared as the Double Genitive became ascendant does not hold up to close scrutiny. I will also demonstrate that the Det Poss and Poss Det constructions in OE were not simply interchangeable variants; there were several systematic differences between the constructions. The Det Poss and Poss Det constructions and their relationship to later English syntax can only be understood in the context of a systematic examination of their syntax as well as their discourse functions in OE, and I will offer an analyses of the OE NP/DP which accounts for the syntax of both constructions. I will also make some observations on their use in discourse and the relationship of the Det Poss construction to Latin models.

The investigation also raises some interesting methodological questions which need to be addressed in any text-based study of diachronic syntax.

14th May , 2003:

Paul Sidwell

(Department of Linguistics, Research School for Pacific and Asian Studies, ANU)

Looking for Su': the adventures, surprises and rewards
of finding a language when you don't really know where it is, or who speaks it

Until the 1990s the West Bahnaric languages were effectively unknown in the linguistic literature there was hardly more than some confused language names and a few short and sometimes inaccurate word lists. Mapping, describing, and classifying the languages has been, and continues to be something of an adventure. In the last 5 years unknown languages have been discovered, others lost and rediscovered, and a new and more complete picture of the linguistic situation in southern Laos is now emerging. In this seminar I will share some of the adventures in carrying out this research, and the challenges of dealing with the data to arrive at a coherent account of the disposition, and historical origins of this language family.

21st May , 2003:

Andrew Pawley

(Department of Linguistics, Research School for Pacific and Asian Studies, ANU)

Assignment of animate gender to non-human referents in English comparative and historical notes

A number of varieties of English use he and she for inanimate referents: trees, turnips, axes, houses, roads, rivers, rain, jobs and situations. My aim in this paper is to throw some light on the history of such patterns of usage, which I call ‘animation’. Animation is not to be confused with personification, where a non-human referent is endowed with a range of human qualities. The evidence indicating that at least two different kinds of non-standard gender assignment patterns go back several centuries in English dialects. One system is (or was formerly) used in Wessex English, centred in southwest England in Somerset, Dorset, Devon and certain neighbouring counties. This system is also found in Newfoundland, which received many settlers from the Wessex area. A second system of gender assignment is best documented in what I call Australia and New Zealand Vernbacular English, but I believe it is present as well in the USA and Canada, and probably in South Africa. It is this second system that I will be mainly concerned with here.

28th May, 2003:

Harold Koch & Pascale Jacq

(School of Language Studies, Faculty of Arts, ANU)

Towards the reconstruction of Pama-Nyungan verb inflection

In comparative Australian linguistics, the comparison of verb has figured in arguments for the unity of an Australian language family or phylum (Capell 1956, Dixon 1980). Meanwhile the existence of a large family called Pama-Nyungan (PN) within the Australian languages has been widely accepted since the 1960s. Detailed reconstruction of verb inflection within the presumed PN group has been largely ignored, except by Dixon 1980 and Alpher 1991. Their results and methods will be assessed. The aims and methods of a large-scale project comparing and reconstructing PN verbal etymologies and inflections, begun in 2002, will be described. The structure of the data-base compiled by Pascale Jacq (and in 2002 by Michael Dunn) will be described. Preliminary results from the comparison of monosyllabic verb roots will be presented with discussions of their implications for the reconstruction of Pama-Nyungan.

4th June, 2003:

Malcolm Ross

(Department of Linguistics, Research School for Pacific and Asian Studies, ANU)

Possession in Oceanic

The work reported in this seminar represents part of an ongoing project on the diachronic fate of grammatical constructions. A construction is regarded as a form--meaning pairing (as it is in the various current versions of Construction Grammar). In certain contact situations, a language's organisation of meaning is remodelled on the basis of its speakers' other language, and this in turn leads to a reorganisation of parts of the grammar. In situations where contact plays no role, however, the organisation of meaning can be remarkably durable, even when changes in form occur. Thus in Oceanic Austronesian languages a difference between two locative constructions has been maintained over 3500 years, despite changes in various Oceanic languages in the forms of these constructions, as I showed in my CRLC seminar in 2002.

Here I report on the diachronic fate of Oceanic possessive constructions. Proto Oceanic distinguished between inalienable and alienable possession, expressing them with different morphosyntax. Within the alienable possession construction there was a possessive classifier slot, which is usually assumed to have been filled by one of three contrasting classifiers: *ka- 'food', *mwa- 'drink', *na- 'general'. I will show, however, that the accepted reconstruction of Proto Oceanic possession is probably too simple. The (meaning) distinctions carried by the classifiers have again proven durable over 3500 years in some languages. But in other languages possessive constructions do show some changes in the organisation of meaning. One focus of this paper is to track these reorganisations and to seek out the principles which underlie them.

Various Oceanic languages also display changes in the structures used in these constructions. A second focus of this paper is an examination of cases where there has been change in structure, reconstructing the changes that have taken place and seeking to understand how/why they have occurred.

18th June, 2003:

Helen McLagan

(Master of Linguistics Student, School of Language Studies, Faculty of Arts, ANU)

“Of White Body and Fair Face Men”: placement of adnominal genitive phrases in Old English.

Conflicting claims have been made about changes that took place in the structure of noun phrases within the Old English period, but there is as yet no empirical base for evaluating these claims. Before it can be authoritatively stated how NP structure changed from Early OE to Late OE or from Late OE to Middle English, there need to be systematic investigations of the syntax of NPs for these periods.

This synchronic study concerns pre- and postplacement of adnominal genitive phrases in ∆lfric’s second series of Catholic Homilies, written during the last decade of the 10th century, and, using an electronic corpus, provides some of the data essential to an understanding of how genitive constructions operated in OE. Findings by Allen (1998), Rosenbach (2000, 2002), Altenberg (1982) and Anschutz (1997) for ME and Modern English show that such factors as meaning of the head, type of possessive relationship, and animacy and topicality status of the genitive phrase can be significant for placement.

In the present investigation, rules of placement are found to operate within a complex network of tendencies, but despite the numerous interactions in play, several distinct rules do emerge (eg. demonstrative pronouns with quantifiers are always postposed), and, beyond that, weaker tendencies are generally found to be overridden by stronger ones (eg. proper noun genitive phrases are generally preposed; pre-head adjectives generally trigger postposition). Possibility of selection therefore existed, and was exploited in the homilies most often for what appear to be stylistic reasons.

 16th July, 2003:

Angela Terrill

(MPI for Psycholinguistics
Nijmegen, The Netherlands)

Testing Punctuated Equlibrium in the Solomon Islands

The notion of punctuated equilibrium as a way of understanding language contact and change has become popular and well known since the appearance of R.M. W. Dixon's 'The rise and Fall of Languages'. The model was developed to explain aspects of the Australian linguistic situation, but is also intended to apply to other language situations around the world. However, the applicability of punctuated equilibrium as a way of understanding language change in any other regions of the world has not been systematically tested. This paper looks at an area of long-term contact in the central Solomon Islands, with a view to testing the Dixon model.

In many ways the Solomon Islands are a perfect laboratory for such a study: there are many Oceanic languages and a scattering of Papuan languages, all of which have been in close contact over a long period, enabling us to disentangle genetic inheritance from diffusion, which is something often not possible in the Australian context. The study finds that while the scenarios of punctuation and equilibrium can be a useful way to understand events of the past, for this area of the Pacific predictions made on the basis of the punctuated equilibrium model do not hold up. Instead, observed patterns of linguistic change can better be understood in terms of interaction-based accounts according to known patterns of social behaviour in Melanesia, as well as geographical accounts which predict trajectories of change based on human movement through particular types of geographical terrain.

An activity of the

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