Weekly Seminars - First Semester 2006

Unless otherwise indicated, all seminars are on Fridays, beginning at 2.30 pm in Room 110, the Transient Building, Camperdown Campus.

They are followed by afternoon tea in the Tea-room.
All are welcome.


For further information: Jane Simpson, 9351-3655
E-mail: jhs AT mail.usyd.edu.au

Timetable

Abstracts

(arranged in alphabetical order by surname)

Cynthia L. Allen, Australian National University
Look at My One! Investigating Syntactic Change in Progress

The pro-nominal uses of the word one have been the subject of a great deal of discussion by
linguists, e.g. Rissanen (1997), Jackendoff (1977), Panagiotidis (2003). In formalist
accounts, one is usually treated as equivalent to e, an empty N. The use of one instead of e is
obligatory in some circumstances, optional in others. For many speakers, one cannot replace
e after a possessive expression such as John’s or my (where the possessive pronoun mine is
used instead), and there are many statements in the literature to the effect that one is
impossible or dubious in this position (the 2002 Cambridge Grammar of the English
Language finds my one etc. to be of ‘questionable acceptability, p. 1513). However, in an
unpublished 2005 honours thesis, Lucy Shorrock shows that expressions such as let’s take my
one are accepted (or preferred) by some speakers of Australian, British, and US English.
Shorrock finds that younger speakers in particular are adopting the POSS ONE construction
in what Zribi-Hertz (1997) terms the RELATIONSHIP meaning, e.g. She had no pencils so I lent
her mine. In contrast, speakers in her survey generally rejected the construction in the
POSSESSION meaning (the house is now mine).
In this paper, I explore the implications of Shorrock’s and my own empirical findings for
the study of (morpho)syntactic change. We have in the POSS ONE construction an unusual
opportunity to study the implementation of a grammatical change currently in progress using
both corpora (written and spoken) and native speaker intuitions. The POSS ONE construction is
an extension of the use of pro-nominal one in favour of a nominal expression without an N.
This extension seems natural and furthermore nudges English in the direction of languages
such as Spanish and Dutch, which require to differentiate between the RELATIONSHIP meaning
(Spanish el m'o, Dutch dat van mij ‘the one belonging to me’) from the POSSESSION meaning
(es m'o, het is van mij ‘it belongs to me’). However, it appears to be difficult to explain why
this extension should have started taking place just at this time, and not just in one variety of
English. It is difficult to find a ‘triggering’ factor for this extension; English speakers have
managed for centuries with unexpressed Ns in possessive phrases, and they appear still to be
an option for even the most enthusiastic of POSS ONE users.
There also appear to be no syntactic or morphological, factors which have a really marked
encouraging or dampening effect on the use of POSS ONE, other than its near- universal
rejection in the POSSESSION meaning, but some factors are associated with an overall higher
percentage of acceptance. The parallels between POSS ONE and the spread of overt N-heads
with attributive adjectives is interesting. The usual explanation for the fact that we now must
say the blind man/one where OE simply had se blinda is that the loss of morphology on
adjectives made expression of an N desirable. Formalisms which only allow for well-formed
or ill-formed structures and require a ‘trigger’ for every grammatical change do not deal well
with the implementation of such changes; for example, Kester 1996 posits that these N-less
expressions became completely impossible when adjectival morphology was lost. Such an
approach cannot account for the gradual nature of the change, as I demonstrate with the
results of a corpus-based study.
It seems instead that we must turn to performance-based models such as those proposed by
Bresnan, Dingare, and Manning (2001), Hawkins (2004) etc. to account for how ‘soft
constraints’ or preferences can become grammatical rules. More extensive studies of POSS
ONE and other new constructions are needed to throw light on the implementation of past
changes.

Monika Bednarek, Universities of Augsburg and Sydney
Local grammar and register variation

In this study I compare the language of ten British ‘popular’ and ‘quality’ newspapers, on the basis of a 70000 word comparable corpus, which comprises 100 texts from the ‘hard news’ register. The comparison focuses specifically on what is variously called evaluation, appraisal or stance – the linguistic expression of speaker opinion. This has mostly been discussed within systemic functional linguistics (APPRAISAL theory, e.g. Martin and White 2005) or in the context of English for Academic Purposes. In contrast, this paper applies Hunston and Sinclair’s (2000) corpus-based pattern approach to the analysis of evaluation. As these authors show, certain lexico-grammatical patterns are commonly associated with evaluation, and together make up what they call ‘a local grammar of evaluation’. An examples of such a pattern is: it + link verb + adjective group + clause (e.g. It seemed important to trust her judgement). Evaluative patterns of this kind will be examined in the corpus (focussing on adjective patterns), in order to find out if there are any differences in the patterns between the two types of newspapers (i.e. differences within the register of news reportage), and if so, what these differences are, and whether they confirm our stereotypes about the popular press: should the two types of news discourse be seen as different varieties of English in terms of a local grammar of evaluation?

Bronwen Dyson, University of Sydney
From syntax to morphology: a reinterpretation of developmental stages in ESL acquisition

This paper presents the findings of a longitudinal study of ESL acquisition which suggests that, contrary to the current view of incremental second language development (Pienemann, 1998, 2003), syntactic structures are acquired before morphological ones. To clarify the concept of stage, the paper first outlines three key components - a definition, research methodology and hypothesised explanations. Since the study tested the most recent explanation of the ESL stages, the paper focuses on Pienemann’s (1998, 2003) hypothesis in Processability Theory that stages can be defined in terms a hierarchy of skill-based processing procedures. These procedures, word (stage 1) > category (stage 2) > phrase (stage 3) > sentence (stage 4) > subordinate clause (stage 5), are accompanied by phrase structure rules which are not connected in a principled way to the procedures at stages 2 and 3. The findings are then presented for the grammatical acquisition over one academic year of two recently arrived adolescents (one male, one female) of Chinese language background. While these results indicate that there is evidence of ESL developmental stages and some support for the hierarchy, the predictions cannot satisfactorily explain the learners’ tendency to first acquire phrasal and clausal syntax and then gradually acquire grammatical morphology at stages 2 to 4. To explain this pattern, the paper proposes that stages are the product of an interaction between universal grammar and processing procedures.

John Giacon, ANU/University of Sydney
Linguistics and language revival - some thoughts and questions

A basic principle of language revival is that it serves the people whose language is being revived. At times academic linguistics can focus on the search for knowledge and analysis, but the focus in language revival is more on use of the language. There is a strong interplay of two groups, the language owners and the linguists. The background of both these groups has an impact on the process. Academic linguistics is often focussed on analysis as its endpoint, whereas language revival has language creation and language use as a major purpose.

I will describe some of the work that has been done in the Gamilaroi-Yuwaalaraay area, principles that have been adopted in doing this work, [for instance community decision making, language standardisation] and difficulties that arise. I will also raise questions including:
- what is the way to extract as much information as possible from historical GY sources
- to what extent can this be done by use of typology, beginning with closely related languages
- which areas of grammar should be given priority
- The influence of English and modernisation is very strong. In what areas should a change in traditional GY be welcomed or accepted? For instance kin terms have changed to reflect current social realities, pronunciation and rhythm have been substantially influenced by English, and there is a tendency is to have GY ‘tenses’ parallel English usage. These could be ‘legitimate language development’, ‘unavoidable changes’ or ‘avoidable and unnecessary changes in GY’.

Mark Harvey, University of Newcastle
The origin and development of subsection terminology

Subsection terminologies are widespread in north-central and north-western Australia. They are known to have spread in the post-colonial period. The principal work on the origin and development of subsections is McConvell 1985. McConvell proposes that subsection terminologies also spread extensively in pre-colonial times, and that the point of origin for subsections as a construct was among the Western Mirndi language varieties – Jaminjung, Ngaliwurru, and Nungali. As a result of research into the history of the Mirndi language family, I propose that while McConvell’s account captures important fundamentals, some revisions are required. Specifically, I propose the following.

1. The point of origin of subsection as a concept is distinct from the point of origin for the most widespread diffusion of a particular subsection terminology.
2. Subsections spread as a concept, independently of the spread of any particular terminology.
3. There is evidence for three distinct diffusions of particular types of subsection terminologies.
4. Subsections may be reconstructed for Proto-Mirndi

These revisions suggest that subsections as a concept are of some antiquity.

Reference:
McConvell, P. 1985. "The origin of subsections in northern Australia." Oceania 56:1-25

Matthew Toulmin, ANU
Reconstructing in a dialect continuum: sorting out chronology of changes in Kamta, Rajbanshi and North Bengali

Dialect continua, marked by non-discrete boundaries between speech communities, pose methodological problems for historical reconstruction. Innovations do not necessarily nest discretely within a lect or a subgroup of lects. Thus, given the right socio-historical conditions, a more widespread innovation may be more recent than a more localised innovation. Consequently, it may be impossible to reconstruct the relative chronology of innovations using only ‘system-internal’ or ‘a-social’ linguistic methodologies, including the comparative method.

In this paper I take the dialect geography of phonological changes (reconstructed using the comparative method) and interpret the linguistic geography using extant and hypothetical socio-historical scenarios. The socio-historical reconstruction supplements the comparative reconstruction by disambiguating the chronology of some innovations.

Michael Tyler, MARCS, UWS
How do adults and infants learn to segment words from continuous speech?

Words in continuous speech are rarely separated by acoustic gaps - our perception of gaps between words in our native language is an illusion. How do we achieve word segmentation in our native language and how do we learn to segment words in a second language? With adults, using "artificial languages" (Saffran et al., 1996, Science), my collaborators and I have been investigating the role of a) statistical probabilities between syllables, and; b) the rhythmic structure of the native language on word segmentation. I will report on these cross-linguistic experiments conducted in France, The Netherlands and Australia. Results of additional experiments on infants question the relevance of artificial languages and statistical probabilities to natural language acquisition.

Deanna Wong, Macquarie
Listening to Listeners: A corpus-based analysis of backchannels produced by Australian and New Zealand listeners

Whilst waiting for a turn to talk, we are expected to not interrupt the current speaker. But does ‘not interrupting’ mean we must remain silent? Backchannels, or the speech of listeners, occur simultaneously with the speaker’s talk, yet rarely interrupt it. In an analysis of conversations drawn from the Australian and New Zealand sub-corpora of the International Corpus of English, backchannels produced by listeners revealed an interesting mix of predictability and variation. This prompted both quantitative and qualitative analyses which revealed that listeners appear to exert conscious control over their backchannels and are capable of subtly shifting their support to signal their intention to assume the turn at talking.