Gra cha in eng wor

Free ebooks ==> www.ebook777.comVolume 67 Grammatical Change in English World-Wide Edited by Peter Collins SCL focuses on the use of corpora throughout language study, the development of

Trang 1

Free ebooks ==> www.ebook777.com

Grammatical Change in English World-Wide

www.ebook777.com

Trang 2

Volume 67

Grammatical Change in English World-Wide

Edited by Peter Collins

SCL focuses on the use of corpora throughout language study, the development

of a quantitative approach to linguistics, the design and use of new tools for processing language texts, and the theoretical implications of a data-rich discipline

For an overview of all books published in this series, please see

The Tuscan Word Centre/

The University of Siena

Trang 3

Grammatical Change

in English World-Wide

Edited by

Peter Collins

University of New South Wales, Australia

John Benjamins Publishing Company

Amsterdam / Philadelphia

www.ebook777.com

Trang 4

8TM

Cover design: Françoise Berserik

Cover illustration from original painting Random Order

by Lorenzo Pezzatini, Florence, 1996.

The paper used in this publication meets the minimum requirements of

the American National Standard for Information Sciences – Permanence

of Paper for Printed Library Materials, ansi z39.48-1984.

Trang 5

Table of contents

Peter Collins

part 1 Inner Circle Englishes

Diachronic variation in the grammar of Australian English: Corpus-based

Cross-variety diachronic drifts and ephemeral regional contrasts: An analysis

of modality in the extended Brown family of corpora and what it can tell

The evolution of epistemic marking in West Australian English 205

Celeste Rodríguez Louro

May and might in nineteenth century Irish English and English English 221

Marije Van Hattum

The present perfect and the preterite in Australian English: A diachronic

Xinyue Yao

www.ebook777.com

Trang 6

 Grammatical Change in English World-Wide

part 2 Outer Circle Englishes

Recent diachronic change in the progressive in Philippine English 271

Peter Collins

Linguistic change in a multilingual setting: A case study of quotatives

Julia Davydova

Patterns of regularisation in British, American and Indian English: A closer

look at irregular verbs with t/ed variation 335

Bernard De Clerck & Klaar Vanopstal

An apparent time study of the progressive in Nigerian English 373

Robert Fuchs & Ulrike Gut

American influence on written Caribbean English: A diachronic analysis

of newspaper reportage in the Bahamas and in Trinidad and Tobago 389

Stephanie Hackert & Dagmar Deuber

Cultural keywords in context: A pilot study of linguistic acculturation

Joybrato Mukherjee & Tobias Bernaisch

Recent quantitative changes in the use of modals and quasi-modals

in the Hong Kong, British and American printed press: Exploring the potential

of Factiva® for the diachronic investigation of World Englishes 437

Dirk Noël & Johan Van der Auwera

The development of an extended time period meaning of the progressive

Bertus Van Rooy & Caroline Piotrowska

Trang 7

doi 10.1075 /scl.67.01col

Introduction

Peter Collins

University of New South Wales, Australia

Until fairly recently we have had to rely on unsystematic and impressionistic sources for information on grammatical change in contemporary English, by contrast with sound change in progress, which has been subject to a good deal of sociolinguistic research As corpus-based studies have begun to gather momentum, there are signs that real progress is being made The most significant research to date is that reported

in Leech et al.’s landmark 2009 volume, Change in Contemporary English In this

inno-vative contribution to the long tradition of research on the historical development

of the grammar of English, the authors demonstrate the capacity of a corpus-based approach to quantify recent changes in a range of grammatical categories, including the modal auxiliaries, progressive, subjunctive, passive, genitive and relative clauses,

in British and American English At the same time they explore the role played in this process by a range of linguistic factors (such as grammaticalisation), discourse-level factors (such as colloquialisation) and socio-historical factors (such as Americanisa-tion and prescriptivism) A more recent volume whose focus is also on current, rela-tively short-term, change in English grammar – more specifically the verb phrase – is Aarts et al (2013) While Leech et al.’s work is based on the ‘Brown family’ of corpora, contributors to Aarts et al avail themselves of a wide range of corpora, and there is a notable concern with questions of methodology

The focus in both of these collections, which demonstrate the power of corpus guistic techniques to provide valuable quantitative insights into changes in the English language, is squarely upon on the British and American ‘supervarieties’ of English The investigation of postcolonial varieties of English from a diachronic rather than synchronic linguistic perspective has, however, been largely neglected The existence

lin-of this gap in the World Englishes research paradigm was recognised by Noël, Van der

Auwera & Van Rooy, in their capacity as editors a recent special issue of the Journal of

English Linguistics (Volume 42, 2014, “Diachronic Approaches to Modality in World

Englishes”) The papers in this issue seek to illuminate distinctive grammatical terns in selected postcolonial varieties using concepts and methods from historical linguistics, eschewing the hitherto more common approach involving synchronic comparisons between postcolonial Englishes and the parent variety, accompanied

pat-www.ebook777.com

Trang 8

in which English is the first language for the majority of the population and the guage in which almost all public and private interaction is conducted (Part 1), and on the other hand ‘Outer Circle’ varieties, in which English is usually a second language learnt in school, despite its status as an official language (Part 2) Part 1 contains five chapters on ‘antipodean’ southern hemisphere Englishes (by Collins, Peters, Rodriguez Louro, and Yao on Australian English, and by Hundt on both Australian and New Zealand English); two on Irish English (by Kirk, and Van Hattum); and two on Cana-dian English (by D’Arcy, and Meyer) One paper, by Mair, explores the implications for other Englishes of changes that have occurred in British and American English The chapters in Part 2 represent the following regions: South-East Asia ( Collins’s on Philippine English, and Noël & Van der Auwera’s on Hong Kong English); South Asia (Davydova’s and De Clerck & Vanopstal’s, both on Indian English; and Mukherjee & Bernaisch’s on Indian, Pakistani and Sri Lankan Englishes); the Caribbean (Hackert & Deuber’s on Bahamian and Trinidad/Tobagan Englishes); and Africa (Fuchs & Gut’s

lan-on Nigerian English, and Van Rooy & Piotrowska’s lan-on Black South African English).Four overarching research questions were identified as considerations for con-tributors to bear in mind in preparing their papers, as follows:

1 How do the diachronic tendencies observed in a particular variety differ from those of the parent variety (British English for all the postcolonial Englishes bar Philippine English, whose parent is American English)?

2 What are the possible causes of the diachronic tendencies observed? These may include, for example, the evolutionary status of a variety in Schneider’s (2007) dynamic model, the characteristic style orientation of a variety (has English become rooted in informal registers or is it a formal choice?), English teaching traditions and learner strategies in institutionalised L2 varieties, prescriptivism (as manifested in the pursuit of codification and in the presence of a complaint tradition), and internal changes in registers (such as the increasing use of direct speech or free indirect speech in fiction)

3 Do you observe different rates of change in the same direction from one variety

to another, or in different directions? Do the observed changes converge with, or diverge from, or run in parallel with, those in the parent variety? Are the changes regionally specific, found in a particular variety but not attested in others)? How

Trang 9

(Meyer), and do-support (D’Arcy; Hundt), a smaller number are concerned with morphology (De Clerck & Vanopstal on -t/-ed variation in verbs; Peters on adverb

suffixation), and with topics on the periphery of grammar (Mukherjee & Bernaisch’s paper on cultural key words is concerned with the lexis-grammar interface; Rodriguez Louro’s on epistemic markers with the pragmatics-grammar interface; Davydova’s on quotatives with the discourse-grammar interface) The majority of chapters focus on

a single variable, but two explore the ‘bigger picture’ afforded by investigation of a set of variables (Collins on Australian English; and Hackert & Deuber on Caribbean Englishes)

A major challenge for contributors was the paucity of resources suitable for the torical study of postcolonial Englishes Within the World Englishes paradigm the most well-known and widely used resource is the International Corpus of English (ICE) collection While, strictly speaking, the chronologically parallel ICE corpora are ame-nable only to synchronic comparisons, they have been used as the basis for indirect, apparent time, comparisons by various linguists (for discussion of the apparent-time construct see Labov 1994: 43–72) Some previous ICE-based studies have extrapo-lated findings for ongoing change from differences between speech and writing, based

his-on the assumptihis-on that changes tend to be more advanced in spoken than in written texts (e.g Collins 2009; Van der Auwera, Noël & de Wit 2012) A further possibility – exploited in Fuchs & Gut’s chapter on the progressive in Nigerian English – is the use of synchronic corpora for apparent time studies that compare speakers of different age groups based on the assumption that changes will be more advanced in the usage

of younger than older speakers Yet another strategy is to identify changes in ent time via comparisons of postcolonial varieties and their ‘parent’ variety, based on the assumption that extent of divergence will be an indicator of advancement (Mair & Winkle 2012; Mukherjee & Bernaisch in this volume)

appar-The problem of a short supply of corpora suitable for the real time historical study of postcolonial Englishes is addressed in various ways by the contributors to this volume

Some use the strategy employed in Leech et al (2009) of using parallel or parallel (sub-)corpora representing differing time points Mair in fact introduces the latest member of the American Brown-family (the 1930s ‘Before-Brown’ corpus) in

near-www.ebook777.com

Trang 10

Other contributors use historical corpora in which texts are sampled over a period

of time at regular intervals The studies by Collins (Part 1), Hundt, Peters, Van Hattum and Yao all make use of Clemens Fritz’s facetiously named Corpus of Oz Early English (COOEE), compiled for his doctorate on the origins of Australian English (see Fritz 2007) COOEE comprises four macro-genres, fourteen text categories, and covers the period from 1788 to 1900 Yao and Collins’s chapters also use a recently-compiled multigeneric corpus of 20th century Australian English (AusCorp), comprising news, fiction, and scientific texts organised in ten year periods In order to draw comparisons with earlier British and American English, Collins, Yao and Hundt draw data from ARCHER (A Representative Corpus of Historical English Registers), version 3.2, a multigeneric corpus with texts divided into 50-year periods from 1600 (for British English) and 1750 (for American English) till the end of the 20th century Hundt’s New Zealand data are derived from the 19th and 20th century Corpus of Early New Zealand English (CENZE), which was designed to be as similar in its design to ARCHER as the availability of texts would allow The diachronic dimension of Meyer’s study derives from his use of two generically-matched corpora: a 19th century corpus (comprising texts from the new Corpus of Early Nineteenth-Century Ontario Newspaper English, and various non-fiction and fiction texts) and from the multigeneric Strathy Corpus of Canadian English (which comprises over 50 million words of texts produced from the 1920s to the present day) Van Hattum uses a self-compiled corpus of historical Irish English and English English trial proceedings and personal letters, taken from a vari-ety of sources including CORIECOR, the Old Bailey Corpus, COOEE and ARCHER.The historical corpora described thus far are all multigeneric Some contributors use monogeneric corpora comprising newspapers collected across a set of time points

D’Arcy’s study is based on a set of issues of a Canadian newspaper, the British Colonist,

from 1858 to 1935 Noël & Van der Auwera’s database comprises issues of Hong Kong’s

South China Morning Post and several major American and British newspapers, at

three data points (1990, 2000, and 2010) Hackert & Deuber’s study is based on press

Trang 11

Introduction 5

data from two countries (the Bahamas, and Trinidad and Tobago), at two time points (1968 and 2002–2012, “pre-” and “post-independence” respectively) Van Rooy & Piotrowska use a database comprising mainly 20th century (1884–2012) newspapers, (with a smaller amount of fiction) Also monogeneric is Kirk’s primary data-source, CORIECOR

Of the other types of corpus data used, brief mention may be made of the tions of transcribed spoken material used in two chapters Rodriguez-Louro uses a collection of oral histories housed in the State Library of Western Australia, recorded from speakers born between 1874 and 1983 and grouped into four categories: 1874–

collec-1889, 1922–1933, 1951–1958, and 1964–1983 Davydova draws on data obtained via sociolinguistic interviews from a multilingual community in the south of New Delhi, collected between 2007 and 2011, a part of the Hamburg Corpus of Non-Native Vari-eties of English Finally there is one chapter – the only one – which makes systematic use of a web-based corpus: De Clerck & Vanopstal use data from the Indian compo-nent of the GloWbE corpus, along with a disparate array of other corpora

The first chapter in Part 1, by Peter Collins, explores developments in ten phosyntactic variables in Australian English over the past two centuries (-t/-ed past verb forms, ’s-genitives, the mandative subjunctive and were-subjunctive, concord with collective nouns, light verbs, non-finite complementation with help and pre-

mor-vent, do-support, and be-passives) Data derived from the news and fiction sections

of two historical corpora, COOEE and AusCorp, are compared with those senting British and American English from ARCHER Australian grammatical pat-terns are found to be mostly – in all but two cases – more advanced than those of its British colonial parent, this divergence reflecting Australia’s increasing indepen-dence from British linguistic norms At the same time Australian usage is shifting towards that of American English – the new centre of gravity of grammatical change

repre-in English world-wide – which emerges as the most advanced variety on eight of the ten variables

Alexandra D’Arcy investigates the expression of stative possession by have and

(have) got in Canadian English (specifically that of Victoria, British Columbia) British

and Southern hemisphere varieties – notably Australian and New Zealand English –

are known to have been shifting towards have got, more rapidly than North can varieties, where got has long been subject to strong prescriptive censure Despite

Ameri-the position of Victoria as Canada’s ‘most British city’, it follows Ameri-the North American

pattern of resistance towards have got According to D’Arcy the explanation is to be

found not merely in “sociohistorical timing and contexts”, but also in the interaction

of language internal-forces: the expansion of do-support for stative have has exerted

an inhibiting effect on the spread of the innovative form have got in North American

dialects, but not in British dialects, where the shift from auxiliary to full verb status has been less wide-ranging

www.ebook777.com

Trang 12

6 Peter Collins

Marianne Hundt confirms that the variability of do-support in 19th century

British English, particularly with verbs of the so-called ‘know-group’ (Ellegärd 1953),

was also a feature of the Antipodean colonial Englishes in New Zealand and Australia

In the second half of the 19th century, Hundt notes, the Antipodean varieties develop

in parallel with their colonial parent and with their more established American colonial sibling, albeit with American English slightly ahead of the other varieties in

post-the regularisation of do-support Most strongly resistant to this trend has been post-the lexical verb have, which Hundt shows – confirming the results of previous studies – to

have been more resistant in the British (and Antipodean) varieties than in American English

John Kirk investigates changes in the frequency and uses of the progressive in Irish

English since the late 18th century He finds the progressive to be highly frequent in Irish English, but does not attribute this to increases in such basic uses as the progres-sive passive and ‘special uses’ such as the interpretive and futurate progressives, in view

of their comparable frequencies in his British data While such uses are shared in mon with the majority of Englishes world-wide, Kirk describes the progressive in Irish English as “Janus-like”, with further features transferred from Irish (e.g progressives

com-with auxiliary do as in Don’t be worrying, and ‘extended-now’ progressives as in How

long are you living here?) Kirk makes the interesting suggestion, which would certainly

warrant further investigation, that Irish English might have influenced the ment of the progressive in varieties of English world-wide, through emigration in the 19th century to larger British cities and to the new world

develop-While Christian Mair’s chapter differs from the others in its concern with the

‘supervarieties’ of World English, its relevance to the volume is grounded in Mair’s tention that the more reliable the information we have about the two global reference varieties in the 20th century, the more reliable they will be as benchmarks for the study

con-of the many New Englishes which have developed distinctive endonormative prcon-ofiles only in the 20th century Mair’s study follows in the footsteps of research on modality based on the “Brown family” of corpora (LOB, FLOB, Brown and Frown), and more specifically studies such as Leech & Smith (2009) and Leech (2013) which extend the diachronic coverage of earlier studies based on the British branch of the family Mair’s study doubles the three-decade time depth of the American branch, with data from the

‘Before-Brown’ corpus complementing that from Brown and Frown The modals are found to have remained entrenched in written American English, despite a significant

decline for must, may and shall, while the semi-modals have either remained stable or increased in frequency (with the single exception of be to).

Matthias Meyer focuses on passive clauses containing ‘ditransitive’ verbs such as give, sell and teach in Canadian English: ‘first passives’ (e.g He was given the letter),

‘second passives’ (e.g The letter was given him) and ‘prepositional passives’ (e.g The

letter was given to him), in 19th century and present-day Canadian English The main

Trang 13

Introduction 7

diachronic development he notes is a shift from the dominance of prepositional sives in the 19th century data to that of first passives in the present day, the survival of both types being supported by their functional complementarity Meanwhile second passives have become even rarer than they were in the 19th century, as in American English

pas-Pam Peters investigates five adverbs which have ‘dual’ (‘zero’ and -ly) forms, such

as bad/badly, high/highly, and slow/slowly, in Australian and British English of the 19th

and 20th centuries Peters observes a decline in the use of the zero forms over the course of the 20th century in Australian English, albeit one milder than has occurred

in British English, suggesting that the divergence between the varieties could be uted to either colonial lag or emerging republican independence She also notes that free variation between dual adverb pairs is more commonly found in Australian Eng-lish than in British English, where zero adverbs are associated with a more limited set

attrib-of verbs

Celeste Rodriguez Louro focuses primarily on the expression I think, and its

gram-maticalisation in Australian English from a stance marker taking a clausal complement

to an ‘epistemic/evidential parenthetical’ used to express opinions and mitigate tive judgements Multivariate analysis of Rodriguez Louro’s oral history data reveals differences between speakers born in the period 1964–1983 and those born earlier (for

nega-example a preference by the latter for I think in initial position, but for medial/-final position by the former) which suggests that grammaticalisation of think

clause-was essentially a late 20th century phenomenon Another finding of the study is that

guess only entered the Australian English system of epistemic/evidential verbs in the

early 20th century, quite possibly as an import from American English

Marije Van Hattum’s study suggests that developments with may and might in

19th century Irish English and English English run largely in parallel In objective

possibility contexts might is found to have been restricted to either past or remote contexts, whereas may was used only in non-past and non-remote contexts In subjective possibility contexts, however, might increasingly lost its ability to signal past time reference, and along with may was increasingly used with the perfect construction to

express propositions about past time situations The only period showing a cally significant difference between the two varieties was the early 19th century, where

statisti-might was used more frequently in non-past, non-remote contexts in Irish English

Van Hattum suggests that this finding may be attributable to Irish influence (the past

form b’ fhéidir ‘perhaps’ being preferred over the present form is féidir in non-past,

non-remote contexts)

Xinyue Yao compares the developmental patterns associated with the present

per-fect and the preterite in Australian English with those in British and American English

In the two reference varieties the present perfect has been losing ground to the erite since the 18th century, attributable in large part to a functional shift which has

pret-www.ebook777.com

Trang 14

8 Peter Collins

seen the present perfect become associated more with ‘extended-now’ contexts than resultative contexts American English has been leading the way, not merely in the rate of frequency decline of the present perfect (relative to the preterite), but also in its increasing limitation to temporally specified, negative, and other contexts Yao finds the ratio of present perfects to preterites to have been relatively stable in Australian English, displaying a British-like conservatism that suggests a retention of patterns found in earlier stages of the language

The first chapter in Part 2, by Peter Collins, seeks to shed light on the

‘evolution-ary’ status of Philippine English via comparisons of the progressive in 1960s and 1990s data According to Schneider (2007: 141) the endonormative consolidation of Philippine English is little more than incipient, a view opposed by Borlongan (2011), who argues that it has in fact reached an advanced level Collins’s findings suggest that the jury must remain out on this debate Using findings reported by Leech et al (2009) for the two ‘supervarieties’ – British English and American English (the colonial ‘par-ent’ of Philippine English) – as a benchmark for comparison, Collins presents results that in some cases support Schneider’s position (for example Philippine English fol-lows American English in its distaste for progressive passives, and in the relative popu-larity of progressives with stative lexical verbs), and in other cases support Borlongan’s position (for example Philippine usage diverges from American, and British, in its dispreference for present progressives and contracted progressives)

Julia Davydova explores recent diachronic developments in the quotative marking

of Indian English The study investigates three major components of the IndE

quota-tive system: (i) conservaquota-tive mainstream forms such as say and think; (ii) global vative variants such as be like; and (iii) local innovations such as okay (fine) The most

inno-conspicuous developments noted are a sharp decrease in the frequency of verbs of

reporting, contrasting with a strong rise in that of be like and okay (fine) Looking more

closely, Davydova finds sociolinguistic proliferation, with older mainstream variants preferred by mesolectal male speakers and innovative variants by acrolectal female speakers, a finding noted to be consonant with that of other non-Western societies in which women have been found to be less conservative than men

In their chapter Bernard De Clerck & Klaar Vanopstal seek to determine if the trend in British and American English towards the adoption of -ed endings over -t

endings in the conjugation of irregular verbs is attested in Indian English as well They find that Indian English uses a set of hybrid features, some similar to British English and some to American English Ultimately they question the relevance to their results

of Inner Circle dependent concepts such colonial lag or colonial innovation, ring to interpret the processes of change they identify in Indian English as driven by variety-independent, localised forces In attempting to account for internal variation they note a tendency, albeit mild, for vowel change to be a factor in the retention of

prefer t forms.

Trang 15

Introduction 9

Robert Fuchs & Ulrike Gut compare the usage of speakers of three different age

groups in ICE-Nigeria: ‘older’ (50 years and older), ‘middle-aged’ (30–49 years), and

‘younger’ (18–29 years) Using a multivariate logistic regression analysis, they find that while there is no significant difference in the frequency of progressive use between younger and middle-aged speakers, both groups have a significantly higher frequency than older speakers The conclusion that is inferred from this, that the progressive is

on the increase in Nigerian English, as it is generally in English world-wide (compare Collins, Van Rooy & Piotrowska in this volume) Also found to be significant factors

in changing progressive use are ethnicity (with the rate of use by Yoruba speakers nificantly higher than that of Igbo speakers) and text category (a higher frequency occurring in more persuasive texts and a lower one in more formal ones) Finally, the frequency of extended uses of the progressive with verbs referring to habitual durative activities and stative verbs is found to be stable across age groups in Nigerian English

sig-Stephanie Hackert & Dagmar Deuber compare developments in four

grammati-cal features – contractions of negatives and verb forms, the be-passive, relative that

vs which, and pseudotitles – in newspaper reportage in the Bahamas and in Trinidad

and Tobago, over the past half-century or so The results are interpreted in the light of several broad factors Americanisation appears to have been a mildly influential fac-

tor, with changes in the Caribbean data (e.g increase in contractions, relative that, and pseudotitles, decline in be-passives) in the same direction as those in American usage,

albeit of considerably less magnitude There is also mild evidence for colloquialisation (with contraction rates), and densification (with pseudotitles) One clear difference with American (and British) newspaper language is the retention by Caribbean jour-nalists of a distinct “flavour of formality”

Joybrato Mukherjee & Tobias Bernaisch explore lexicogrammatical routines in

three South-Asian Englishes (Indian, Pakistani, and Sri Lankan) More specifically,

they examine the collocation between the ‘cultural keywords’ government, religion and

terror and the verbs that follow them, in the relevant components of the South Asian

Varieties of English (SAVE) corpus The data is synchronic, but understood to have diachronic ramifications: the greater the divergence between a particular South Asian

variety and its British colonial parent (as measured by a ‘diversity/unity (d/u) ratio’

proposed by the authors), the greater its diachronic advancement is assumed to be Of

the three keywords studied, government is found to have a high degree of shared verbal collocates, religion a lower degree, with terror in between.

The context for Dirk Noël & Johan Van der Auwera’s study is the exchange between

Millar (2009) and Leech (2011, 2013) concerning frequency changes in modality Millar’s finding of a frequency increase in the modals in the American news publica-

tion Time Magazine between 1923 and 2006, appeared to undermine previous findings

by Leech and others (e.g Leech 2003; Leech et al 2009) of a decrease in the modals in American and British writing in general In response Leech claimed that his original

www.ebook777.com

Trang 16

in terms of the contrast between his use of ‘representative’ corpora and Millar’s use of

a genre-specific corpus Noël & Van der Auwera, who argue the descriptive benefits

of studies such as Millar’s, deriving from the role played by genre in language change, restrict their study to newspapers, comparing modal and quasi-modal frequencies in major Hong Kong, American and British newspapers between 1990 and 2010 The British and Hong Kong press are found to pattern similarly, with the modals in decline and the quasi-modals on the rise, whereas in the American press both modal catego-ries are increasing in frequency, the quasi-modals particularly strongly

Bertus Van Rooy & Caroline Piotrowska show that Black South African English

exhibits the rising trend for the progressive attested in varieties of English world-wide, but they are unable to determine with certainty whether it results from the influence

of White South African English, or whether it has been internally fuelled As in 20th century British and American English (see Leech et al 2009: 142) there is little evi-dence, despite frequency increases, for changes in the functions and uses of the pro-gressive The capacity of the progressive in Black South African English to combine more readily than that in the native varieties with stative and achievement verbs, and

to express an extended time period, have been constants in the variety rather than changing features Van Rooy & Piotrowska argue for the likelihood of transfer from substrate (Bantu) languages

The papers in this volume show that the imaginative use of both available pora and of newly-prepared purpose-built corpora can provide fresh insights which promise to address the ‘diachronic gap’ in the World Englishes paradigm It is to be hoped that the book will provide a stimulus for more studies in this relatively new field

cor-of enquiry Finally, I would like to thank the contributors for agreeing to participate

in this project, and to record my gratitude to Xinyue Yao for her gracious help in the preparation of the manuscript for publication

References

Aarts, Bas, Close, Joanne, Leech, Geoffrey & Wallis, Sean (eds) 2013 The Verb Phrase in English:

Investigating Recent Language Change with Corpora Cambridge: CUP.

DOI: 10.1017/CBO9781139060998

Borlongan, Ariane 2011 Relocating Philippine English in Schneider’s Dynamic Model Paper presented at the 17th International Association of World Englishes Conference, Melbourne, Australia.

Collins, Peter 2009 Modals and quasi-modals in world Englishes World Englishes 28: 281–292.

DOI: 10.1111/j.1467-971X.2009.01593.x

Trang 17

Introduction 11

Ellegärd, Alvar 1953 The Auxiliary ‘do’: The Establishment and Regulation of Its Use in English

Stockholm: Almqvist & Wiksell.

Fritz, Clemens 2007 From English in Australia to Australian English: 1788–1900 Frankfurt:

Peter Lang.

Kachru, Braj 1985 Standards, codification, and sociolinguistic realism: The English language

in the outer circle In English in the World, Randolph Quirk & Henry Widdowson (eds),

11–30 Cambridge: CUP.

Labov, William 1994 Principles of Language Change, Vol 1: Internal Factors Oxford: Blackwell.

Leech, Geoffrey 2003 Modality on the move: The English modal auxiliaries 1961–1992 In

Modality in Contemporary English, Roberta Facchinetti, Manfred Krug & Frank Palmer

(eds), 223–240 Berlin: Mouton de Gruyter.

Leech, Geoffrey 2011 The modals ARE declining: Reply to Neil Millar [2009] International

Journal of Corpus Linguistics 16: 547–564 DOI: 10.1075/ijcl.16.4.05lee

Leech, Geoffrey, 2013 Where have all the modals gone? On the declining frequency of modal

auxiliaries in American and British English In English Modality: Core, Periphery and

Evidentiality, Juana I Marín-Arrese, Marta Carretero, Jorge Arús Hita & Johan Van der

Auwera (eds), 95–115 Berlin: Mouton de Gruyter.

Leech, Geoffrey & Smith, Nicholas 2009 Change and constancy in linguistic change: How

grammatical usage in written English evolved in the period 1931–1991 In Corpus

Linguis-tics: Refinements and Reassessments, Antoinette Renouf & Andrew Kehoe (eds), 173–200

Amsterdam: Rodopi.

Leech, Geoffrey, Hundt, Marianne, Mair, Christian & Smith, Nicholas 2009 Change in

Contem-porary English: A Grammatical Study Cambridge: CUP DOI: 10.1017/CBO9780511642210

Mair, Christian & Winkle, Claudia 2012 Change from to-infinitive to bare infinitive in ficational cleft sentences: Data from World Englishes In Mapping Unity and Diversity

speci-World-Wide: Corpus-Based Studies of New Englishes [Varieties of English around the World

G43], Marianne Hundt & Ulrike Gut (eds), 243–262 Amsterdam: John Benjamins.

DOI: 10.1075/veaw.g43.10mai

Millar, Neil 2009 Modal verbs in Time: Frequency changes 1923–2006 International Journal of

Corpus Linguistics 14: 191–220 DOI: 10.1075/ijcl.14.2.03mil

Schneider, Edgar 2007 Postcolonial English: Varieties Around the World Cambridge: CUP.

DOI: 10.1017/CBO9780511618901

Van der Auwera, Johan, Noël, Dirk & De Wit, Astrid 2012 The diverging need (to)’s of Asian Englishes In Mapping Unity and Diversity World-Wide: Corpus-Based Studies of New

Englishes, [Varieties of English around the World G43], Marianne Hundt & Ulrike Gut

(eds), 55–76 Amsterdam: John Benjamins DOI: 10.1075/veaw.g43.03van

www.ebook777.com

Trang 18

Trang 19

part 1

Inner Circle Englishes

www.ebook777.com

Trang 20

Trang 21

doi 10.1075 /scl.67.02col

Diachronic variation in the grammar

of Australian English

Corpus-based explorations

Peter Collins

University of New South Wales

This chapter uses data extracted from two recently compiled historical corpora

of Australian English, the Corpus of Oz Early English (19th century) and

AusCorp (20th century) to examine developments in ten morphosyntactic variables over the past two centuries At the same time comparisons are drawn with British and American English, using data from ARCHER (A Representative Corpus of Historical English Registers) While Australian usage is found to have diverged from that of its British colonial parent, reflecting increasing independence from British linguistic norms, it has shifted towards that of

American English – the new centre of gravity of grammatical change in English world-wide – which emerges as the most advanced variety on eight of the ten variables

Keywords: grammar; Australian English; diachronic; corpus; variation

1 Introduction

Scholarly interest in the characteristic phonological and lexical features of Australian English (AusE) can be traced back about half a century, to the publication of such landmark studies as Mitchell and Delbridge (1965) and Ramson (1966) Work on grammatical features has been slower to gain momentum, these being generally per-ceived to differ less significantly from those of other national varieties, a reflection of the truism that “accent divides, and syntax unites” (Mair 2007: 97) The chapters in Peters et al.’s (2009) volume, with their comparative focus on the grammar of AusE, provide a number of avenues for interpretation that are potentially relevant to the findings of the present study In some chapters similarities noted between patterns of usage in contemporary Australian and British usage might simply reflect coinciding preferences, but alternatively may suggest the persistence of an allegiance in Australia

to the grammatical norms of the colonial ‘parent’ variety For example Smith (2009)

www.ebook777.com

Trang 22

(2009) finds the tendency for the connective adverbs however, thus and therefore

to be used in ‘run-on sentences’ (introducing a second main clause within a single orthographic sentence) to be stronger in AusE than in BrE and AmE The findings

of other chapters are compatible with an interpretation of Australian subservience to AmE norms For example, in Peters’ (2009a) chapter the frequency of the mandative subjunctive in AusE is found to outstrip that in BrE, and at the same time noted to approximate those recorded for AmE in other studies

One limitation of these and other grammatical studies is their reliance on chronic data, one which restricts their capacity to identify historical developments leading up to the present day, and to make plausible predictions about future trends This problem is in turn a by-product of the unavailability of suitable data for the investigation of diachronic grammatical changes in AusE The smallness of the tem-poral gap between the two currently available contemporary representative corpora

syn-of AusE, the Australian Corpus syn-of English (ACE) and the Australian component

of the International Corpus of English (ICE-AUS), makes them unsuitable for chronic study as a pair: the texts for ACE, a Brown-family corpus, were sampled

dia-in 1986, while those for ICE-AUS were sampled dia-in the early 1990s Fortunately, recently assembled diachronic corpora of AusE now cover the period 1788–2000: see Section 3 below

This study examines 19th and 20th century developments in a number of matical variables, in AusE, BrE and AmE The variables were selected on the grounds

gram-of their susceptibility to historical change in Late Modern English and are ously discussed in recent studies such as Leech et al (2009), Rohdenburg & Schlüter (2009), and Peters et al (2009) They include not only (lexico-)grammatical variables

(do-support, light verbs, be-passives, and non-finite complementation), but also phological (irregular past forms, and ’s-genitives), and some that lie on the borderline

mor-between morphology and syntax (subjunctives, and concord with collective nouns) There is also, unfortunately but inevitably, a spread of frequencies from comparatively

large (for example be-passives with 4109 raw tokens), through middling (for example

collective nouns with 227 tokens), to very infrequent (for example mandative tives with 21)

subjunc-The structure of the remainder of this paper is as follows Section 2 surveys ous studies of grammatical change in AusE Section 3 provides details of the corpora and methodology used in the study Sections 4–6 present the findings for the morpho-logical, morphosyntactic and syntactic variables respectively Section 5 is devoted to concluding remarks

Trang 23

previ-Free ebooks ==> www.ebook777.comDiachronic variation in the grammar of Australian English 17

2 The study of grammatical change in AusE

The corpus-based diachronic study of AusE grammar is very much in its infancy

In some studies apparent time implications have been drawn from comparisons of spoken and written genres (e.g Collins 2009), and in others from comparisons of age-graded responses in questionnaire data (e.g Elsness 2009) Fritz’s (2007) pioneering corpus-based examination of various morphosyntactic features focuses on the 19th century, with some comparisons made on the basis of frequencies drawn from more recent corpora

Two recent studies (Collins & Yao 2014; Collins 2014) have explored changes in selected categories of the verb phrase in Australian English, using the same paral-lel corpus material for the 19th and 20th centuries that is used in the present study (differing only in that these two studies are limited to fiction) Collins and Yao’s find-ings, for the categories they examined, suggest that while AusE has been evolving

in the same general direction as the two longer-established reference varieties over this period, it has generally been lagging behind them (in its frequency increase for the progressive and most of the quasi-modals, and in its decline in the proportion

of present perfects to preterites) One motivation for the present study was to test whether such a pattern of diachronic development, popularly referred to as ‘colo-nial lag’ (the alleged conservatism of postcolonial varieties), was in evidence across

a wider range of grammatical variables in AusE Following suggestions made by Hundt (2009a), in her investigation of differential grammatical change in BrE and AmE, I shall adopt a more fine-grained typology of diachronic patterns in which the notions of colonial lag and its converse, ‘colonial innovation’, are supplemented by those of ‘revival’, ‘survival’, ‘divergence’, ‘parallel change’, and ‘overtake’.1 For instance the 20th century American-led rise of the mandative subjunctive can be considered

as a postcolonial ‘revival’, in view of the steady decline that the previously- productive mandative had suffered from Early Modern English until the 19th century (see Övergaard 1995; Hundt 2009a: 30–31) By contrast frequencies adduced by Leech et

al (2009: 64) suggest that the were-subjunctive in hypothetical conditionals, a

rem-nant of the Old English inflectional subjunctive paradigm, is a ‘survival’ that is more strongly endorsed in late 20th century AmE than BrE ‘Divergence’ is in evidence

in a lexico-grammatical variable such as prevent NP (from) Ving where the

from-less variant, though attested in both British and American usage in the 18th and 19th centuries, has been retained in present-day BrE but according to Leech et al (2009: 193–194) has virtually disappeared from AmE The increasing use of singular

1 Hundt (2009a: 33) proposes the colourful term ‘kick-down development’, for the last of these.

www.ebook777.com

Trang 24

18 Peter Collins

concord with collective nouns since the 19th century in BrE and AmE is suggested

by Hundt (2009a: 33) as a possible case of (American-led) ‘parallel development’ Finally, the ‘overtake’ pattern is exemplified by the quasi-modals, whose trajectories are initially more conservative in AmE than BrE, but run parallel in the 19th century until AmE overtakes BrE in the 20th century (Krug 2000)

While the role of Australia’s colonial parent in shaping AusE grammar deserves systematic attention, the more recent influence of AmE must also be taken into account Having emancipated itself from the hitherto prestigious variety spoken

in the motherland, AmE has become the centre of gravity of much grammatical change in English world-wide A consequence of this change is that a number

of general tendencies that have been observed to be particularly associated with American usage (Rohdenburg & Schlüter 2009: 5–6) have variably penetrated BrE, AusE, and other varieties These include the tendency for AmE to be tolerant of and inclined towards features characteristic of spoken colloquial usage (e.g quasi-modals), another for it to have a preference for explicit, analytic, marking of gram-

matical functions (e.g. do-support), and another for it to dispense with function

words that are semantically redundant and grammatically omissible (e.g infinitival complements)

bare-3 Corpora and methodology

This section presents information about the three corpora used in the present study, and about the methodology used for extracting relevant data and using it to anal-yse the nature and extent of grammatical change The study examines a set of gram-matical variables in two genres (news and fiction) across three regional varieties of English (AusE, BrE, and AmE), over the 19th and 20th centuries (subdivided into four half-centuries)

3.1 The corpora

The three corpora are all multigeneric, but for the sake of generic consistency the study was limited to the two genres – news and fiction – that were common to all three corpora Between them, news and fiction cover much of the variation associ-ated with written English, and are directed at a non-specialist readership News, whose concern is with the reporting and interpretation of events, is an innovative,

‘agile’ genre (Hundt & Mair 1999) that is highly receptive to change Fiction, whose concern is with constructing imaginary worlds, has a reputation for being more informal than both news and the ‘uptight’ genre of learned writing, particularly

Trang 25

Free ebooks ==> www.ebook777.comDiachronic variation in the grammar of Australian English 19

fictional texts in which the dialogic recreation of speech looms large (see further Biber et al 1999)

a COOEE: Compiled by Clemens Fritz, COOEE (Corpus of Oz Early English) comprises texts written between 1788 and 1900, from more than a hundred dif-ferent sources, including books, letters, diaries, proclamations, and newspapers (for corpus information see Fritz 2004, 2007) In order to match time periods

in the other corpora, the study was restricted to texts produced in the period 1800–1899 The genre of fiction is readily available in COOEE, represented

by the category ‘narratives (novels and short stories)’ However the genre of news lacks a direct counterpart, so a selection was made of 96 news reports from the COOEE category ‘newspapers and broadsides’ The remaining 57 texts

in this category are excluded insofar as they represent literary prose or press editorials

b AusCorp: AusE data for the 1900–99 period was taken from AusCorp, a corpus

of 20th century AusE recently compiled by Xinyue Yao and the present author, comprising news, fiction and scientific text samples organised in ten year peri-ods The news section contains around 170,000 words evenly distributed across

the decades, and sourced from major Australian newspapers such as the Sydney

Morning Herald and The Australian The fiction section also comprises around

170,000 words, with some bias towards the late 20th century Text-extracts are

taken from two sources: anthologies of Australian literature such as The

Macqua-rie PEN Anthology of Australian Literature (Jose 2009); and the Digital Archive of Colonial Australian Popular Fiction, an online collection of Australian fiction pro-

duced during the period spanning the 19th century and early 20th century (see http://www.apfa.esrc.unimelb.edu.au/) For each writer, only about 2000 words were selected

c ARCHER: Data for BrE and AmE was drawn from ARCHER (A tive Corpus of Historical English Registers), version 3.2.2 ARCHER 3.2 is a multi-genre historical corpus of BrE and AmE, with texts divided into 50-year periods starting from 1600 (for BrE) and 1750 (for AmE) till the end of the 20th century

Representa-Table 1 below summarises the corpus categories used in the present study, from COOEE for AusE 1800–1899, AusCorp for AusE 1900–1999, and ARCHER for BrE and AmE

2 The corpus was accessed at Northern Arizona University in November 2012.

www.ebook777.com

Trang 26

20 Peter Collins

Table 1 Word counts for the fiction and news subcorpora of COOEE, AusCorp

and ARCHER used in the study*

deverbal disyllabic adjective learned from the set of preterite and past participial forms

of learn) In other cases the number of irrelevant tokens was high (as for the light verbs have and take, for which it was necessary to remove all instances in which the

following indefinite NP complement was not derived morphologically from a verb, as

in have a husband) In some cases it was not possible to run exhaustive searches: for

the mandative the most common suasive expressions governing subordinate clauses containing subjunctive verb forms were targeted, and the list of 35 collective nouns

in Hundt (2009b) was used to investigate concord with collectives For be-passives

the subcorpora were POS-tagged to identify past participles, and then a search string with regular expressions was applied to accommodate items that may occur between the past participle (for example, adverbials, noun phrases and negators) and forms of

be Post-editing was not feasible in view of the vast number of tokens, so an estimated

error rate of 5–10% has to be taken into consideration in assessing the results

An important task in corpus-based studies of grammatical change is to mine the most appropriate methodology for measuring changes in frequency A commonly used and quite straightforward method of measurement involves nor-malised frequency counts, usually to tokens per million words (pmw) This is the main method adopted by Leech et al (2009), where it is readily justified by their use of the one- million word nuclear Brown-family corpora, whose parallel status

deter-is guaranteed by “the employment of a strict sampling frame, down to one-to-one mapping of equivalent 2,000-word samples” (Smith & Leech 2013: 71) A variant of the normalised frequency method involves changing the baseline frequency from

Trang 27

words to a grammatical category such as VPs, in order to compensate for the tially skewing effects of generically- and temporally-determined variations in the

poten-VP density of texts (as used for example by Smitterberg 2005 in his study of the

pro-gressive) This measure is used for be-passives in the present study (see Section 6.4

it isolates a linguistic variable, thereby reducing the possibility that changes could

be due to a variety of factors For example, application of a proportional approach

to change in the frequencies of bare-negation versus do-supported negation (see

Section 6.3 below) enables us to avoid the question of whether changes are due merely to those in the frequency of negative clauses Again, the question of whether the rise of singular concord is due simply to a rise in the frequency of collec-tive nouns is bypassed if we focus on the alternation between singular and plural concord

While determining the variants of a linguistic variable may be a straightforward

matter (for example, the complementation of help by either a bare- or to-infinitive),

it can also be subjective and imprecise For example, as light verbs, have and take clearly alternate in combination with certain objects (such as have/take a look and

have/take a sip), but not others (for example have/*take a quarrel, but *have/take a step) In other cases the problem is not merely one of determining alternation, but

furthermore the impracticality of inspecting a vast number of tokens For example, in

the case of ’s- genitives a proportional analysis was not applied, because the difficulty

of delimiting a category of ‘of-genitives’ would have necessitated testing for the stitutability of every token by an ’s-genitive: see further Section 4.2 below).

sub-In the present study pmw frequencies are presented in the figures that represent proportional information, even though proportions could equally well be calculated

on the basis of the raw frequencies The reason for this is to maintain consistency with

the two non-proportional pmw-based analyses (of ’s-genitives and be-passives) All

the pmw frequencies presented represent averages of those for each of the two genres investigated, news and fiction

3 While it is a simple matter to calculate the number of VPs in a syntactically parsed corpus,

a close approximation to the number of finite VPs can also be derived from a POS-tagged corpus by counting the number of finite verbs.

www.ebook777.com

Trang 28

22 Peter Collins

4 Morphological variables

In this section we shall consider two morphological variables: the past tense and past

participial forms of irregular verbs, and ’s-genitives.

4.1 Regularisation of irregular past tense and past participle forms

Findings by Hundt (2009a: 24–25) suggest that until the early 19th century it was BrE that led the way from AmE in the re-regularisation of irregular past tense and

past participle forms of verbs such as burn, learn, dwell, spell, smell, dream, kneel,

lean, leap, spill and spoil Subsequent post-colonial developments saw AmE overtake

BrE and remain in the vanguard of change ever since Hundt’s (1998a: 32) age figures for the regular forms, 96.7% for Brown and 64.8% for LOB, indicate the extent of the gap between the reference varieties in the 1960s The strong American preference for the regular forms is possibly attributable to the putative general regu-larising tendency in AmE that was referred to in Section 2 above, the conservatism of BrE perhaps to “an avoidance strategy treating the regular forms as a morphological Americanism” (Hundt 2009a: 25), or to prescriptive endorsement of the older irregu-lar forms (Peters 2009b: 14)

percent-The present study was restricted to just two verbs, burn and learn, the only verbs

with appreciable frequency.4 The findings presented in Figure 1 generally confirm the claim above that AmE overtakes BrE in the 19th century and maintains a strong preference for the regular forms thereafter The findings for AusE contradict those of Peters (1994, 2009b), who claims contemporary AusE to be even more conservative than BrE.5 In the present study, AusE swings from colonial conservatism in the 19th century to advancement in the 20th when, as in the case of a number of other vari-ables, it assumes a position mid-way between the more innovative AmE and the more conservative BrE

4 This restriction might have skewed the frequencies by comparison with those of studies based on a larger set of verbs, given the differing responses of verbs to regularisation Figures

presented in Hundt (1998a: 32) show burnt to be about twice as popular as burned in ACE and WCNZE (Wellington Corpus of New Zealand English), but learned to be about twice as popular as learnt.

5 Based on age-graded evidence in elicited questionnaire data, Peters (2009b) suggests that irregular forms may even be making a comeback in contemporary AusE amongst younger speakers.

Trang 29

Figure 1 -ed vs -t forms of burn and learn in AusE, BrE and AmE (frequencies pmw)*

* In this and subsequent proportional figures the four fifty-year periods 1800–1849, 1850–1899,

1900–1949 and 1950–1999 are represented notationally as ‘19/1’, ‘19/2’, ‘20/1’ and ‘20/2’ respectively.

4.2 ’s-genitives6

The Old English genitive inflection, according to historical accounts presented in Rosenbach (2002) and Altenberg (1982), underwent a decline in Middle English

As the alternative of-genitive option began to increase from late ME, the ’s-genitive

became semantically restricted (to animate nouns) Since Early Modern English,

how-ever, the ’s-genitive has undergone a revival, its expansion bolstered by its use as a clitic

in ‘group-genitives’, as in (1), where it has scope over each other, and by a weakening of

its restriction to animate nouns, as in (2)

(1) if you are involved in each other’s work all the time

(Sydney Morning Herald, 7 July 1986)

(2) the last of the sun’s rays streamed down in sloping planks

(Thiele, The Sun on the Stubble, 1961)

Figure 2 presents the findings for historical change with the ’s-genitive in the present

study I have eschewed the proportional approach used widely throughout the present

study, in view of the challenges that are presented to the task of identifying of-genitives

(i.e the subset of the vast set of of-phrases in the corpora that may be considered to

be semantically and formally interchangeable with ’s-genitives) by the multiplicity of

grammatical and non-grammatical factors determining selection of one or the other

construction (see Rosenbach 2002; Hinrichs & Szmrecsanyi 2007).7

6 The search routine targeted all tokens of ’s and s’, and irrelevant tokens (mainly

contrac-tions as in there’s) were manually discarded.

7 Leech et al (2009: 224–225), likewise observing the impracticality of an exhaustive

anal-ysis, present a proportional comparison of ’s-genitives and of-genitives based on a 2% sample

from their corpora, conceding its status as a mere approximation.

www.ebook777.com

Trang 30

24 Peter Collins

The findings comply with those of other studies in recording a 20th century

increase in the ’s-genitive that has advanced further in AmE than in BrE (Leech et

al 2009: 223; Rosenbach 2002: 3), but which has been more sluggish in AusE (Hundt

1998a: 45–46) In Figure 2, frequencies of ’s-genitives for the second half of the 20th

century show AmE to be marginally ahead of BrE, and the latter ahead of AusE thermore, between the first half of the 19th century and the second half of the 20th, AmE enjoys an increase of +145.1%, leading the way over BrE (+89.9%) and AusE (+28.8%)

con-tive as in if need be, God save the Queen and lest he be, is highly fossilised and restricted

(and for this reason not included in the present study) The other two, both of which have a viable frequency in contemporary English and which I focus upon in this study, are the ‘mandative subjunctive’ (in subordinate clauses following ‘suasive’ expressions

Trang 31

such as require, important, recommendation, etc.), and the were-subjunctive in

condi-tional and concessive clauses

In all three constructions there remain, in contemporary English, merely remnants

of the fully fledged inflectional subjunctive paradigm of Old English The remaining verb forms are as follows: the ‘plain’ or ‘bare’ form of lexical verbs with a third person

singular subject as exemplified in (3) below; the verb be in its ‘plain’ or ‘bare’ form as

in (4); and the past subjunctive were with a first- or third-person singular subject as in

(5) In all four cases the subjunctive form is distinguishable from a possible indicative

form: gives in (3), is in (4), and was in (5).

(3) Spanish officials said Spain would demand that the United States give up

its use of the Torrejon Air Base outside Madrid (ARCHER 1975atl2.n8a) (4) Mr Walpole was told by the council clerk that advertising the intention to

borrow would enable residents to ask that a poll be taken

(Advocate, 09 June 1954)

(5) Amused by this game, they began to treat me with a sort of teasing

condescension, as if I were a toy, almost a mascot

(Anderson, Tirra Lirra by the River, 1978)

After the Old English period the subjunctive went into decline, but the 20th century saw a revival of the mandative subjunctive, one in which AmE is leading the way over BrE (Johansson & Norheim 1988; Övergaard 1995; Hundt 1998b; Schneider 2011;

Peters 2009a; Leech et al 2009: 51–70) By contrast the ‘conditional’ were-subjunctive,

described by Quirk et al (1985: 158) as ‘something of a fossil’, has been observed to be

in decline, and more so in BrE than AmE according to Leech et al (2009: 64)

5.1.1 The mandative subjunctive

The mandative subjunctive has periphrastic modal alternatives That with should,

exemplified in (6) below, is its most direct rival in that there is no appreciable semantic

difference between them (substitution of be for should be in (6), might subtly alter

the associative meaning of the sentence – its style or level of formality – but would

not affect the primary sense) In addition to the 67 instances of periphrastic should in the results, there were 12 cases with other modals (might, would, must and shall) An example with must is provided in (7):

(6) In the Assembly Mr J C Ross, who presented a petition from the Public

Service Association, moved that counsel should be heard at the Bar of the

House in reference to the Salaries Bill

(Sydney Morning Herald, 03 June 1936)

(7) the Russians have stipulated that France’s and Britain’s nuclear weapons

must be included in SALT III (ARCHER 1959man1.n8b)

www.ebook777.com

Trang 32

26 Peter Collins

The corpora were interrogated for mandative subjunctives via the suasive expressions found in the matrix clause I used the same items as Leech et al (2009: 53), in order

to facilitate comparison with their results.8 The study revealed a certain amount of

lexical conditioning in the occurrence of the mandative subjunctive: move proved

to be by far the most common suasive expression with support also for ask, demand and suggest.

The frequencies for the mandative subjunctive and periphrastic alternatives in the second half of the 20th century, as presented in Figure 3, generally support Leech

et al.’s (2009: 54–55) finding that the mandative subjunctive varies from a position of dominance over modal periphrasis in AmE, to being a relatively infrequent variant to

it in BrE That this contrast is of longer historical reach is suggested by the average centage of 62.5% that the mandative subjunctive enjoys over the two centuries in AmE, compared with only 23.7% in BrE AusE arguably lies mid-way between the two refer-ence varieties While the Australian penchant for the mandative across the 200-year period (29.3%) is only marginally higher than that of BrE, it has risen steadily over the period from virtual non-existence (4.0%) in the 19th century to just over half (52.0%)

per-in the 20th century This fper-indper-ing for AusE is compatible with Schneider’s (2011: 169) Brown corpus-family based finding that AusE is positioned mid-way between AmE and BrE It is also supported by the results of Peters’ (2009: 132) study of the mandative subjunctive in the spoken components of the ICE corpora, which did not include AmE but found AusE to be more innovative in its deployment of the mandative subjunctive than its colonial parent, BrE In the revival of the mandative subjunctive then, AusE – like a number of postcolonial Englishes (see Hundt 1998a, Peters 2009a) – seems to

be following the lead of AmE (which has maintained a preference for the subjunctive

over should since the latter half of the 19th century) and to be eschewing the more

conservative behaviour of British speakers (who have maintained a dispreference for the mandative over the same period).9

8 Leech et al.’s list comprised seventeen suasive verbs (advise, ask, beg, demand, desire, direct,

insist, move, order, propose, recommend, request, require, stipulate, suggest, urge and wish), plus

eight related nouns (demand, desire, proposal, recommendation, request, requirement,

sugges-tion and wish), and five adjectives (anxious, essential, important, necessary and sufficient): see

Trang 33

Figure 3 The mandative subjunctive vs modal periphrasis in AusE, BrE and AmE

(frequencies pmw)

5.1.2 The were-subjunctive in hypothetical conditional and concessive clauses

In this section I consider the alternation between subjunctive were and indicative was,

in subordinate clauses expressing a hypothetical condition (introduced by if, as if, even

if, and as though).The alternation is only possible in the first person singular, as in (5) above, and third person singular as in (8):

(8) If such a treatment were meted out to residents of the capital, it would in all

probability precipitate a riot (The Townsville Baile Bulletin, 23 June 1954) Subjunctive were alternates with indicative was, as in (9), the latter tending to be asso-

ciated with a greater degree of informality

(9) Gawd, she thought, this should be a fabulous feeling – cruising with a

beau – if only a girl wasn’t afraid of dying (Winton, Cloud Street, 1991) The percentages in Figure 4 suggest that the decline of were in AmE did not gain

momentum until the 20th century, when its percentage drops to a level (57.5%) similar to that of 20th century BrE (58.1%) In the 20th century the AmE percentage remains relatively constant, but the construction undergoes a decline in BrE Leech

et al (2009: 68–69) posit several possible explanations for the conservative American

situation: American susceptibility to prescriptive support for subjunctive were and proscription of indicative were; support from the strongly performing mandative sub-

junctive; and a “complicated pattern of post-colonial revival” (though the almost static situation in 20th century AmE is perhaps indicative of survival rather than revival) AusE, like AmE, exhibits colonial conservatism It begins with a very high endorse-

ment of were (83.4%) similar to that in AmE, which subsequently drops but increases

www.ebook777.com

Trang 34

28 Peter Collins

again to 63.7% in the second half of the 20th century, conservatively stronger than the final percentages in BrE and AmE It is difficult to know whether or not Australian usage with this variable might be being influenced by American conservatism

Figure 4 Subjunctive were versus indicative was in AusE, BrE and AmE (frequencies pmw)

5.2 Concord with collective nouns10

Concord with collective nouns is realised in the selection of singular or plural verbs (and pronouns) Examples (10) and (11) feature both verbal and pronominal concord, singular in the former and plural in the latter

(10) The government having declared Melbourne an open city was continuing its

deliberations on the ‘terms’

(Eldershaw, Tomorrow and Tomorrow and Tomorrow, 1947)

(11) The party were furnished with two fowling pieces, and after strolling about

the fields and shore, trying their skill on the sea-gulls, crows, sparrows

(ARCHER BrE 1822eva1.n5b)

Historically, singular concord is the older pattern, but it began to be challenged by plural forms which, having begun to appear as early as 1000, peaked in the17th and

10 Searches were run for instances of the 35 nouns used by Hundt (2009b: 211), which she

in turn took from Quirk et al (1985: 316): army, association, audience, board, cast, clan, class,

club, college, commission, committee, community, company, corporation, council, couple, crew, crowd, department, family, federation, gang, generation, government, group, institute, majority, ministry, minority, opposition, party, population, staff, team, university Of these, five had no

tokens (cast, clan, class, institute, university) in the present corpus data.

Trang 35

18th centuries and then declined in the 19th century (Levin 2001: 36) The current stronger preference for singular concord in AmE than in BrE and other varieties, as noted by Fritz (2007: 217), Hundt (2009b: 208) and others, is regarded by Marckwardt (1958: 77) as simply representing the perpetuation of the older pattern, with AmE lagging behind BrE in embracing plural concord However the increasing popularity

of singular concord is more plausibly interpreted as an American-led revival, rather than a survival, of the older latent pattern Figure 5 shows that over the past two cen-turies the most conservative increase in singular concord has occurred in BrE, which

is overtaken firstly by AmE in the second half of the 19th century and then by AusE in the first half of the 20th century The findings do not support Fritz’s (2007: 221) claim, based on a comparison of his COOEE findings with those of Dekeyser (1975: 58–64) for BrE, that in the 19th century AusE was more advanced than BrE in its use of singular concord.11 It can be inferred from Figure 5 that AmE and AusE, having both transitioned from colonial conservatism to advancement, throughout the 20th century maintain a larger gap between singular and plural agreement (9.1:1 in AmE, 4.3:1 in AusE) than that in BrE (3.1:1) As for most of the variables in this study AmE emerges as the most advanced variety, BrE as the most conservative, with AusE in between

169

309

646 744

128

359 187

Figure 5 Singular concord vs plural concord with collective nouns in AusE, BrE and AmE

(frequencies pmw)

11 Inspection of the frequencies in Fritz’s (2007: 222) Table 4.50 reveals a number of tion errors (AusE should be 51.5% rather than 58.2%, and BrE should be 61.1% rather than 42.3%) The corrected figures thus confirm the validity of the claim made here that, in the 19th century, BrE was more advanced than AusE.

calcula-www.ebook777.com

Trang 36

30 Peter Collins

6 Syntactic variables

In this section we shall examine light verbs, non-finite complements, do-support, and

be-passives The first two of these are arguably lexicogrammatical, rather than purely

syntactic, influenced as they are by collocational restrictions For example the light

verb have, but not take, may be used with the deverbal noun talk; help can take an infinitival but not gerundial complement, as in Mary helped to paint/*painting the fence (by contrast with stop for example, as in Mary stopped *to paint/painting the fence).

6.1 Light verbs

‘Light verbs’ are so-called on the grounds that “their contribution to the meaning

of the predication is relatively small in comparison with that of their complements”

(Huddleston & Pullum 2002: 290) The two light verbs investigated in this study, have and take, are exemplified in (12) and (13) respectively.

(12) No, they must have a try, to satisfy themselves; if that trial be successful,

there they will stick (COOEE 3–014)

(13) Sometimes the old mate would stay over Sunday, and in the forenoon or

after dinner he and father would take a walk amongst the deserted shafts of

Sapling Gully or along Quartz Ridge (COOEE 4–334)

Unfortunately there is a lack of agreement in the literature on how the light verb struction should be defined In most cases it is specified that the object NP with which the light verb enters into construction must be indefinite, in order to exclude examples

con-such as have mercy, take pity, and have the drink A more contentious issue, however,

is whether or not the complement must be headed by a deverbal noun formed via version (as required for example by Leech et al 2009: 173, and Smith 2009: 141) The vast majority were certainly of this type, but I also allowed instances – which Leech

con-et al and Smith would presumably have disallowed – of the type have a suspicion,

where the noun is derived via suffixation rather than conversion

According to Hiltunen (1999) and Claridge (2000), light verb constructions did not begin to appear until Early Modern English, though Leech et al (2009: 171) claim that they are attested as early as Old English They began to increase substantially in frequency in Late Modern English (Bailey 1996; Brinton 1996), one factor most likely precipitating their spread being the increasing productivity of conversion as a word

formation process in Modern English Until around 1800 take was more common than

have, but during the 19th century, as have increasingly acquired dynamic uses, it began

to overtake take (Leech et al 2009: 171).

The most frequently noted aspect of regional variation with light verbs is the

con-trast between BrE, whose speakers are more inclined to use have than take, and AmE,

Trang 37

whose speakers are more inclined to use take than have (e.g Algeo 1995) This ence can be attributed to the failure of have to acquire as much dynamism in AmE as

differ-it did in BrE (Trudgill et al 2002: 11), and is plausibly interpreted as an instance of colonial conservatism by Hundt (2009a: 33) and Rohdenburg and Schlüter (2009: 399)

A further sign of the more advanced status of BrE is its more frequent use, by contrast with AmE, of light verbs generally (Leech et al 2009: 172)

The findings of the present study, presented in Figure 6, provide a measure of port for these claims The non-proportional form of presentation in this figure reflects

sup-the fact that not all tokens of sup-the light verbs have and take in sup-the data are truly variable BrE has a higher frequency of have and take together (170 tokens pmw across the four

half-centuries) than AmE (136), with AusE in between (147) Furthermore the alleged

AmE predilection for take is in evidence in the 20th century, in which the take vs have ratio is 1.5:1, by contrast with the BrE preference – albeit milder – for have over take (1.1:1) The results for AusE reveal a level of innovation (with a have:take ratio in the

20th century of 1.7:1) that is even stronger than that shown by BrE Ratios aside, the results certainly support claims by Wierzbicka (1982) and Kearns (2002) that AusE

frequencies for light have are similar to those for BrE and unlike those for AmE (in

the 20th century: BrE 125 pmw, AmE 52 pmw, AusE 105 pmw) On this variable BrE emerges as uncharacteristically innovative and AmE conservative, with AusE aligned more closely with the former)

51 151

Figure 6 Light verb have versus take in AusE, BrE and AmE (frequencies pmw)

6.2 Non-finite complementation with help and prevent

The spread of infinitival subordinate clauses at the expense of finite clauses has been both long term (dating from Old English) and systematic in the history of English

www.ebook777.com

Trang 38

32 Peter Collins

(see Rohdenburg 1995) In this section I target several specific constructions

involv-ing infinitival clauses functioninvolv-ing as complement to the matrix verbs help and

pre-vent The gerund is by contrast a more recent development, originating in the 17th

century, and not achieving ascendancy in its competition with infinitival

comple-ments (in for example start Ving vs start (to) V) until the 20th century (see Leech

et al 2009: 185).12

6.2.1 Help (NP) (to) V

Help, whether intransitive as in (14) and (15), or transitive as in (16) and (17), may take

either a bare infinitive as in (14) and (16), or a to-infinitive as in (15) and (17) (14) Each can destroy Each helps destroy my parents; each helps them lay waste

about them (Porter, The Watcher on the Cast-Iron Balcony, 1963)

(15) But, having faith in the engineer, the structure should last for several

generations, which should help to pay for it

(The Singleton Argus, 2 June 1954) (16) “They help us prepare free meals for the poor and clean the streets,” he said

(The Daily Mirror, 23 June 1986) (17) Kelvin Hansen, 15, who had been helping his father to nail down a loosened

sheet of iron roofing, was thrown to the footpath by a gust of wind

(The Sydney Morning Herald, 23 June 1975)

This variability has been attributed to structural, semantic and stylistic factors (e.g

stylistically, the to variant is preferred in more formal styles, the bare variant in more

informal styles) Leech et al (2009: 189) note two parallel developments in late 20th century BrE and AmE: both varieties are witnessing a mild decline in the frequency of

the to-infinitive, and a sharp rise in the frequency of the bare infinitive, with AmE more

advanced than BrE in both complementary trends According to Mair (2009: 275), the

rise of the bare infinitive with help points to “an auxiliarisation/grammaticalisation

process in which this verb is becoming more like a semi-auxiliary of causation than a lexical verb expressing the notion of assistance” Furthermore the fact that the popular-ity of the bare infinitive is greater in speech, and increasingly so, suggests that a fur-ther factor in its rise is colloquialisation (Leech et al 2009: 190–192; Mair 2009: 273)

Figure 7 confirms the leading role of AmE in the complementary rise of help V and fall of help to V, developments that are not apparent in the British data until the second

half of the 20th century AusE lies in between the innovative American pattern and the

12 The initial search included not only help and prevent, but also start and stop

Unfortu-nately the number of tokens for the latter was too small to make further analysis viable.

Trang 39

conservative British, with a level of support for the bare infinitive in the 20th century that is not far behind that in AmE

48 199

158

12 0

65

149

307 112

47 54

Figure 7 Help (NP) V vs help (NP) to V in AusE, BrE and AmE (frequencies pmw)

6.2.2 Prevent NP (from) Ving

According to Leech et al (2009: 193), while prevent NP from Ving is available in all varieties of English, there is a from-less version (prevent NP Ving) that was attested in

both AmE and BrE in the 18th and 19th centuries However the varieties have diverged,

with the from-less version all but disappearing from AmE in the 20th century, and its

continuing robustness in BrE assuring its status as a ‘Briticism’ today The two structions are exemplified in (18) and (19)

(18) a black nor’-easter would keep Victoria awake all night; a southerly buster

would prevent her from eating a bite of breakfast (Farmer, Melpo, 1982) (19) Senator Pearce may say that this provision was intended to prevent the

smaller States combining to obstruct legislation

(Sydney Morning Herald, 09 June 1914) The distribution of prevent in the present study, as represented in Figure 8, largely sup- ports the historical picture presented above: the from-less version is spurned in AmE,

by contrast with BrE, where its percentage has risen from 45.7% in the first half of the 19th century to 72.7% in the second half of the 20th century Meanwhile AusE diverges

from BrE, its steadily decreasing endorsement of the from-less version – possibly

influ-enced by the American dispreference for this variant – contrasting with the continuing British endorsement of it

www.ebook777.com

Trang 40

34 Peter Collins

11

63 37 34

13 93

181

26 25 81

23 35

prevent NP Ving prevent NP from Ving

Figure 8 Prevent NP Ving vs prevent NP from Ving in AusE, BrE and AmE (frequencies pmw)

In Present-Day English lexical verbs require do-support in negative (and

interroga-tive) contexts In fact, according to Ellegård (1953), there were but a handful of verbs

that continued to allow bare negation beyond the end of the 18th century, namely care,

know, doubt, fear, say, and think Of these, only the first three yielded more than two

instances of bare negation in the present study Examples follow:

(20) even though she cared not one straw about them (COOEE 2–060)

(21) He knows not even whence nor how this is (ARCHER 1839long f5a)

(22) [she] withdrew to her own room: where, we doubt not, she was followed by

the rebukes of her conscience (ARCHER 1822sedg.f5a)The search routine yielded three further verbs that were represented by more than two

tokens each – have, need and dare – as exemplified below:

(23) He had not mental capacity enough to comprehend (COOEE 3–193)

(24) Death is a serious thing, as I am sure I need not remind you, Mr Trefusis.

(ARCHER 1887shaw.f6b)

13 The search items were not and its contracted form n’t.

Định dạng
Số trang	494
Dung lượng	4,31 MB

Gra cha in eng wor

Multivariate analyses and diachronic comparisons

The Bahamas and Trinidad and Tobago: Sociohistorical and