Belfast english and standard english dialect variation and parameter setting by alison henry

Similarly, a number of accounts have been proposed for why main verbs in English except be and have never raise out out of VP: for example, Pollock 1989 suggests that they cannot assign

STANDARD ENGLISH

Richard Kayne, General Editor

Principles and Parameters of Syntactic Saturation

Edited by Adriana Belletti and Luigi Rizzi

Discourse Configurational Languages

Edited by Katalin E Kiss

Clause Structure and Language Change

Edited by Adrian Battye and Ian Roberts

Belfast English and Standard English: Dialect Variation and Parameter Setting

Alison Henry

BELFAST ENGLISH AND STANDARD ENGLISH

Dialect Variation and

Oxford New York Toronto

Delhi Bombay Calcutta Madras Karachi

Kuala Lumpur Singapore Hong Kong Tokyo Nairobi Dar es Salaam Cape Town

Melbourne Auckland Madrid

and associated companies in

Berlin Ibadan Copyright © 1995 by Alison Henry

Published by Oxford University Press, Inc.

200 Madison Avenue, New York, New York 10016 Oxford is a registered trademark of Oxford University Press, Inc All rights reserved No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of Oxford University Press Library of Congress Cataloging-in-Publication Data

Henry, Alison.

Belfast English and standard English : dialect variation and

parameter setting / Alison Henry.

p cm.—(Oxford studies in comparative syntax)

Includes bibliographical references and index,

ISBN 0-19-508291-5 ISBN 0-19-508292-3 (pbk.)

1 English language—Dialects—Northern Ireland—Belfast.

2 Belfast (Northern Ireland)—Social conditions.

on acid-free paper

One of the goals of linguistic theory is to establish the extent to whichall languages are similar, and the boundaries within which they may dif-fer This book is a study of how the syntax of a non-standard dialect ofEnglish differs from standard English The analysis presented is under-taken within the Principles and Parameters framework, but it is hopedthat it will also be of interest to those working in other frameworks, orinterested simply in a description of the grammar of Belfast English,something which has not hitherto been available With this in mind,each chapter begins with a description of the differences betweenBelfast English and standard English in relation to the structure beingconsidered, before going on to consider their analysis

This is, to the best of my knowledge, one of the few wide-rangingstudies of a non-standard dialect of English that has been undertakenwithin the Principles and Parameters framework; indeed, even article-length studies are very few This is at first sight very surprising, giventhe vast amount of attention that has been devoted to standard English,and the potential contribution to the theory of studies of closely relatedvarieties of language, as evidenced by the large quantity of fruitfulresearch on the Romance languages, and on the Scandinavian lan-guages A rich source of information has thus been largely unavailable

to linguists hitherto; indeed, not only have treatments within the ples and Parameters framework been unavailable, but, because of thelow status generally accorded to non-standard dialects, even descriptiveaccounts have not been compiled It is possible to find out somethingabout a standard language like English by looking at descriptive or ped-agogical grammars; but such grammars have not in general been written

Princi-for non-standard varieties, which have often been regarded simply asdegenerate versions of the standard (see Milroy & Milroy 1991).Although this book discusses a particular variety of English, that spo-ken in Belfast, the capital of Northern Ireland, many of the features dis-cussed occur in other varieties Thus a number of them occur inHiberno-English in general—for example, inversion in embedded ques-tions—and others are also found in some North American dialects, for

example for to, which has also been documented in the Ozarks and the

Ottawa Valley, possibly as a result of emigration It should thus be ofinterest to those working on other non-standard or regional varieties ofEnglish

Earlier versions of Chapters 2 and 3 of this book were presented aspapers at the Linguistics Association of Great Britain Conference, and I

am grateful to audiences there for much useful feedback Chapter 4 isadapted from a paper which appeared in Natural Language and Linguis-tic Theory (Henry 1992)

I am very grateful to Jim McCloskey, Nigel Duffield, Sten Vikner,Fritz Beukema, Marcel den Dikken, David Pesetsky, and Angelika vanHout for helpful comments and discussion on the topics considered inthe book

5 Inversion in Embedded Questions 105

6 Subject Contact Relatives 124

7 Conclusion 136

Notes 759

References 142

Index 147

STANDARD ENGLISH

1 Introduction

The Study of Dialect Variation

One of the major goals of linguistic theory is to establish what is versal in human language, and what are the limits on linguistic varia-tion, that is, in how far and in what ways the grammars of speakers maydiffer from one another

uni-One way of approaching this question is to look at languages whichare superficially very different from one another, and to find out to whatextent there are true deep-seated differences between the languages, andhow these differences can be accounted for A complementary approach

is to look at closely related languages or dialects and to examine thedegree of variation that is possible between grammars which are inmany other ways similar The latter approach has been used produc-tively in relation to Romance and Scandinavian languages and dialects,but there has been comparatively little work in this framework ondialects of English, where most research has been on the standard vari-ety of the language In this book, we consider how a non-standard vari-ety of English, Belfast English, differs from the standard language.The study of dialect variation presents a particular challenge to aview of language which sees the ways in which languages may vary asbeing highly restricted The Principles and Parameters approach togrammar views human language as essentially invariant, with the possi-ble ways in which grammars may vary from one another being limited

to the setting of a small number of innately specified parameters alongwhich language may vary Dialects at first sight do not seem to varyfrom one another in precisely the way this view would predict

3

Although some researchers have hoped that studying dialect variationwill enable us to find dialects which differ in a single parameter, thissearch has not been easily fulfilled, and it will not be fulfilled in thisstudy Rather, we will show that Belfast English differs from standardEnglish in a number of ways, but these are not all derivable from a sin-gle difference in parameter setting.

Moreover, within Belfast English, indeed within a single constructionsuch as the imperative, there are, as shown in Chapter 3, a number ofdifferent grammars possible

What we will show, however, is that the different possible grammarsall result from parameter setting differences permitted by UniversalGrammar That is, there is more variability in grammars between stan-dard and dialect, and even within dialect speakers, than one mightexpect But the differences all clearly reflect possible choices of para-meter setting, rather than, for example, the presence or absence of lan-guage-particular rules

What is noticeable about the differences between Belfast English andstandard English is that, in many cases, they derive from different char-acterisations of particular lexical elements, rather than of a functionalcategory as a whole Thus, for example, Belfast English infinitives dif-

fer from standard English infinitives in that for can be a clitic in the

for-mer but not in the latter Imperatives differ in that the imperative pheme, which appears in C, can be strong in Belfast English, forcingmovement of the verb to C The effect of this difference is that inimperatives, the main verb can raise to C as in the V2 languages; but,because this difference is tied to the property of a specific item, theimperative morpheme, rather than being a property of C holdingthroughout the language ("The V-feature of C is strong"), the superfi-cial difference between the dialects is not very great: overall the wordorder patterns are the same, except in imperatives

mor-The co-existence of different grammars in a speech community raisesinteresting questions about how language acquisition proceeds; childrenwill usually have input from more than one adult, and the adults whomthey hear will often have grammars with different parameter settings Itcan be observed that where this is the case, children do not necessarilydevelop a grammar which covers all of the data in the input to whichthey have been exposed Rather, they appear to select the parameter set-ting which is compatible with the majority of, but not necessarily all,the data in the input This shows that language learning is strongly

determined by innate factors; learners do not add language-particularrules which would enable them to develop grammars which could coverall the data, nor do they necessarily select the parameter setting whichwould cover the entire range of possibilities exemplified in the input.Rather, from the small range of possible grammars permitted by UG,they select the one which best fits the data For a detailed discussion ofthis process of selection, see Chapter 3, where we discuss Belfast Eng-lish imperatives.

Apart from imperatives, Belfast English differs from standard lish in a number of other interesting ways, which will also be discussed

Eng-in detail Eng-in subsequent chapters

Subject-verb agreement is optional:

(1) The eggs is/are cracked.

(2) The machines works/work well.

Inversion is possible in embedded questions, whereas it is restricted

to matrix questions in standard English

(3) I wonder did they go.

(4) She asked had anybody called.

For to is possible before infinitives:

(5) They seem for to be late.

(6) I want them for to win.

Relative clauses where the subject is relativised occur without overtrelative pronouns

(7) We had a window looked out on that side.

(8) There's a woman in our street went to Spain last year.

and certain finite subordinate clauses occur with null subjects, in what isotherwise a non-pro-drop language

(9) They were lucky got away.

(10) You were as well took the job when you were offered it.

The analysis of these constructions is relevant to a number of currentissues within syntactic theory in general and English syntax in particular

A number of characteristics of English which have been claimed toderive either from universal principles or from parameter settings hold-ing for English seem to differ in Belfast English Thus, it has been

claimed (by Rizzi & Roberts 1989 and Vikner 1991, among others) that

the unavailability of inversion in embedded questions in English derives

from the Wh-criterion (Rizzi 1991; for a discussion of this feature, seeChapter 5) Although the Wh-criterion seems otherwise to apply inBelfast English just as in standard English, inversion is freely available

in embedded questions in Belfast English, calling into question thisanalysis unless some other factor can be brought in to account for itsavailability Similarly, a number of accounts have been proposed for

why main verbs in English (except be and have) never raise out out of

VP: for example, Pollock (1989) suggests that they cannot assign their0-roles from higher positions because AGR is weak in English; but asshown in Chapter 3, Belfast English imperatives seem to show suchmovement, and any account of why verb raising is not generally possi-ble must therefore take account of the fact that it does happen in thiscase Examining such structures in detail will allow us to see whetherthe analyses proposed for standard English are correct, but additionalfactors in Belfast English mean that the facts look different, or whether

in fact the standard English analyses do not hold up when faced withdata from another dialect

Before we go on to look in detail at the differences between BelfastEnglish and standard English (and within Belfast English), it will beuseful to set the scene by considering some background issues Thus,although the study is concerned with the grammars of contemporaryspeakers of Belfast English rather than the historical development ofthat variety, it will be useful to consider briefly the historical and geo-graphical background of this variety This is done in the next section

We also need to outline briefly the theoretical framework in which thestudy is undertaken; this will take up a later section, titled "Principlesand Parameters Theory." The final section of this chapter considers thespecial methodological issues and problems which arise in working onnon-standard dialects

Belfast English: Some Background Information

Belfast English is the variety spoken in and around Belfast, the capital

of Northern Ireland Belfast is a major commercial and industrial centrewith a population of some half a million people

English speakers in Belfast are largely monolingual, and there is no

community of native Irish speakers in the area, although there is a smallbut growing number of Irish-medium schools for children whose par-ents wish them to be educated in Irish For most Belfast English speak-ers Irish is a subject learned at secondary school if at all; there are fewbilingual speakers, and thus any influence from Irish almost certainlyderives from historical, rather than contemporary, contact between thetwo languages.

The English spoken in this area of Ireland descends largely from thatintroduced by the plantations of Ireland, when English and Scottish set-tlers came to Ireland in the seventeenth century, bringing their languagewith them The local population at that time was Irish-speaking, andindeed quite a few of the settlers learned Irish The use of Irish in theBelfast area had however died out by the end of the nineteenth century.The plantation took place on an extremely large scale; the censusreturns for 1658/9 show that of a total population of 31,221 in Antrimand Down, the counties which border Belfast, 13,614 were of English

or Scottish descent Although many of the settlers learned Irish, theintroduction of such a large number of English speakers, who held theeconomic and political power, marked the beginning of the decline inthe use of Irish in this area (Patterson 1880)

Rural Ulster speech is generally considered to consist of two maindialects, Ulster Scots, which is closely related to Scottish English, in theNorth-east, and the Central or Mid-Ulster dialect, which shows moreinfluence of Irish Belfast, as the major city in Northern Ireland, hasattracted such an influx of population in search of work from manyother areas that its dialect cannot be clearly defined as belonging toeither of these two groups

In relation to standard English and some other varieties, Belfast lish can be considered to be a conservative one (it is indeed part of thelocal folklore that we speak English as it was spoken in Shakespeareantimes) Thus, some of the constructions discussed here were used in ear-lier forms of standard English, and they have been retained only inBelfast English and other conservative dialects today For example the

Eng-use of for to with infinitives was found in earlier stages of English thermore, the co-occurrence of wh-elements with that and inversion in

Fur-imperatives are also found in earlier stages of standard English

One of the interesting characteristics of Belfast English is that,although Belfast is known to be in many ways a divided society, withoften little contact between the Protestant and Catholic communities,

Belfast English is not distinguished, either phonologically or cally, along religious lines All of the constructions discussed in thisbook are used by both communities, and where there is any distinction

grammati-in usage, it is between workgrammati-ing and middle-class speakers or older andyounger speakers, rather than along religious lines It is simply not pos-sible to tell to which community persons belong by how they speakEnglish Belfast English is thus very much something which the com-munities have in common, something which tends not to be noticedbecause of two factors: First, the fact that the allegiance of one commu-nity to England (and standard English), or sometimes to the rural UlsterScots dialect with its clear Scottish roots, and the other to Ireland (andIrish) means that the local variety of speech is championed by no one.Second, and perhaps more important, this is a variety of English whichhas little status and which is not officially recognised Schools, bothProtestant and Catholic, devote a great deal of time to the teaching of

"correct" (=standard) English, and the ability to use standard syntax isconsidered to be a mark of education; conversely the use of local syntax

is considered a badge of the lack of education Many people consider itquite legitimate to discriminate against users of local syntax in employ-ment Milroy & Milroy (1991) quote the following letter from a localnewspaper as typical The issue in question is the use of non-standardpast tense forms; many verbs have different past tense and past partici-ple forms from those found in standard English For example, the past

tense of see is seen and the past participle saw, whereas the opposite is

the case in standard English (for details see Finlay 1988)

For many years I have been disgusted with the bad grammar used by school-leavers and teachers too sometimes, but recently on the lunch- time news, when a secretary, who had just started work with a firm, was interviewed her first words were: "I looked up and seen two men" etc It's unbelievable to think, with so many people out of work, that she could get such a job ("Have went," Saintfield, N Ireland)

Although there are now a few government-funded Irish-mediumschools in Belfast, there is no right to be educated in the local dialect ofEnglish; all education is based on standard British English Childrenwho use the structures discussed in this book in school work, with acouple of exceptions that have acquired the status of a local standard,will simply be marked wrong Things appear to be changing slowly,with the requirement in the new Common Curriculum that children

learn something about local dialects, and a local dialect dictionary isbeing compiled; but there is a long way to go before Belfast English isseen as anything other than a deviation from the standard Hence, thedisbelief with which most local people greeted the news that I was writ-ing a book about the grammar of Belfast English; in their view, it hadn't

got any

This book, then, is about a dialect which is widely spoken but whichhas no official status; thus, studies of its syntax have been few Therehas been some excellent sociolinguistic work on Belfast English (seeMilroy 1980, Milroy 1981), but much of it has focussed on phonology

It is hoped that, apart from the analyses it discusses, this book willmake available a description of the syntactic characteristics of BelfastEnglish which will be of value also to readers outside the focus of Prin-ciples and Parameters theory

Principles and Parameters Theory

This book is written within the framework of Principles and Parameterstheory, in particular its latest version, the Minimalist program of Chom-sky (1992) This theory views language as largely invariant, with thedifferences between languages deriving from choices between a smallnumber of innately determined parameter settings

It is beyond the scope of this introduction to provide a wide-rangingaccount of that theory, and the reader is referred to works such as Rad-ford (1988) and Haegeman (1991) for comparatively introductoryaccounts, or to Chomsky (1986, 1992) for more technical statements ofthe nature of the theory However, it is worthwhile to draw attentionhere to some relevant aspects of the framework which will be assumed

in this book, particularly insofar as they represent differences betweenthe Minimalist approach and earlier approaches with which some read-ers may be more familiar

The only levels of representation recognised within the Minimalistapproach are "interface levels," where the linguistic system interfaceswith other systems: these are the level of Phonological Form, where itinterfaces with the pronunciation system, or at Logical Form, where itinterfaces with other cognitive systems Thus, there is no level of D-structure or S-structure as assumed in earlier models

Chomsky (1992) argues that all languages are similar at Logical

Form; the differences between languages result from the fact that someprocesses apply earlier in some languages than others; those processeswhich apply early are overtly visible, while those which apply later arenot.

While earlier models envisaged words being introduced from the icon in their base form, and moving in the course of the derivation toacquire Case-marking or to pick up affixes, the Minimalist model envis-ages words being inserted from the lexicon in their fully specified,inflected form, and moving for checking to affixes; once checked, theaffixes delete What determines whether movement for checkingapplies early or late is whether the affix in question is weak or strong Ifstrong, it is visible at PF if undeleted, and it will be identified as anunattached affix at that level, causing the derivation to crash; therefore,where a functional element is strong, it triggers overt movement, before

lex-what is called spell-out If it is weak, movement does not occur until

LF

The reason that movement is delayed until LF if possible is the crastinate Principle: movement occurs as late as possible, LF movementbeing in some sense "less costly" than overt movement and thereforepreferred

Pro-Another important principle with a "least effort" flavour within malism is that movement only occurs if forced; there is no optionalmovement

Mini-An important difference between the Minimalist approach and otherframeworks which will be relevant for our study here relates to Case.Earlier approaches incorporated a Case filter operating at surface struc-ture, which excluded structures in which an overt Noun phrase had notCase At S-structure, in order to have Case, NPs had to be in a Caseposition, or be part of a chain of which one element was in such a posi-tion

While the Minimalist approach requires Case to be checked, this can

be done either in the overt syntax or later Since NPs are inserted fromthe lexicon with all their inflectional properties, the fact that an NP isovertly Case-marked does not mean it has to be in a Case position orpart of a Case-marked chain It can raise to check Case at LF, subjectsraising to SPEC/AGRS and objects to SPEC/AGR0 As we shall see, thisseems to work well for Belfast English where in imperatives, subjectscan be in a non-Case-marked position at S-structure, something whichwould have presented a problem in earlier approaches to syntax

In accordance with the Minimalist proposals, the phrase structureconfiguration we will be assuming in our discussions is as shown in(11), with a range of functional projections above the verb phrase.

Since the Minimalist approach is relatively new, some areas, forexample that of infinitives, have not received extensive treatment in thenew framework In discussing these areas we will also discuss possibleanalyses in an earlier framework

The Minimalist approach presents a particularly restrictive approach

to cross-linguistic variation: language varieties can only differ in tion to the strength or weakness of the morphological properties offunctional elements; it will thus be interesting to see whether such arestrictive theory can accommodate the range of variation foundbetween dialects

rela-Before we go on to consider this possibility, however, we need tolook at some particular issues involved in the study of non-standard lan-guages.

Studying Non-Standard Language Varieties

A number of problems arise in working with non-standard dialects,which do not manifest themselves, at least to the same extent, in study-ing standard language, and it is worth noting them here

First, there are problems in relation to obtaining judgements on tences from native speakers Speakers of Belfast English are aware thatmany of the things they say are regarded as ungrammatical in the pre-scriptive sense When asked whether a sentence is grammatical or not,their first reaction is to say whether they think it would be correct instandard English It is of course possible to overcome this in part byexplaining that what one is interested in is what people actually say inthe local variety of speech, not in what is commonly regarded as "cor-rect" English There is nevertheless a strong tendency towards what Ihave termed "negative overreporting" (Henry 1992), that is, indicatingthat structures are ungrammatical when in fact the speaker actually usesthem; it is thus important to check judgements against naturally occur-ring data where possible, particularly to ensure that structures said to beungrammatical have not in fact been so judged simply because they are

sen-non-standard An example of this occurred when I was studying for to

infinitives; I asked a native speaker whether it was possible to have

sen-tences where for to occurs directly after a verb like want; the speaker

said that sentences like:

(12) I want for to go

were ungrammatical But only a few minutes later, he remarked

(13) 1 want for to be helpful

and this was not just a performance error; when we went back to theoriginal examples, he agreed that he would use them, although, headded, "Of course they wouldn't be right."

It is in my view particularly important to make sure that the data areaccurate in studying non-standard dialects, because once data andanalyses are available, they tend to be discussed and reanalysed without

the original data necessarily being able to be checked Such checking isparticularly difficult in the case of non-standard dialects, of which fewlinguists are native speakers If I make claims about the grammaticality

of sentences in standard English, or French, or Japanese, there will bemany linguists who can check those judgements against their own intu-itions; but if I make a claim about Belfast English, this is not the case Ihave therefore tried to be particularly careful in establishing what thedata are, and I have where possible checked grammaticality judgementsagainst naturally occurring data, by which I mean spoken utterances Ihave not used any data from literary texts; although a number of lin-guists have used texts in looking at Hiberno-English (see for exampleDoherty 1993, Duffield 1993) Although such sources have the advan-tage of being easily verifiable, I regard them as not necessarily reflect-ing the actual usage of native speakers As Milroy & Milroy (1991)point out, writers attempting to write in a dialect of which they are nottruly native speakers may not know exactly what the rules of the dialectare, and the language they produce may in fact be ungrammatical Anexample from Belfast English quoted by Milroy & Milroy is the follow-

ing, from a television script based on Gerald Seymour's novel Harry's Game, where a Northern Irish speaker says:

(14) He's a hard man, but so is you Billy.

In Belfast English, as we will see in some detail in Chapter 2, it is ble under certain circumstances for the third person singular form of the

possi-verb to occur with all subjects, a process known as singular concord It

is clearly this process that the writer is trying to reflect here, but in factthe sentence produced is ungrammatical in Belfast English Non-agree-ing third person singular verbs never occur in inverted structures, nor dothey occur with simple personal pronouns

(15) a The girls is late.

b *They is late.

c *Is the girls late?

so that in the sentence used, it is ungrammatical to use is; the agreeing form are must occur.

(16) He's a hard man, but so are you Billy.

Thus I have avoided literary data sources

A second point, not unrelated to the unreliability of literary sources,

relates to identifying native speakers of the variety concerned, andensuring that one is obtaining judgements from speakers who actuallyuse the structure in question Most speakers from Belfast will haveheard people using the structures discussed in this book; but not all ofthem will use every one of those structures themselves Thus for exam-

ple the usage of for to seems to be restricted to middle-aged and older

speakers, and many younger speakers will not use it Similarly, the lack

of subject-verb agreement discussed above tends not to occur in thespeech of educated middle-class speakers However, such speakersoften nevertheless think they have sufficient passive competence fromhearing other people use these structures to say when they are used.Often, of course, they are right, but they can also be quite wrong Thuswhile speakers who do not themselves use non-agreeing verbs are usu-ally aware that agreement is obligatory with pronouns, many do notrealise that, for speakers in whose grammar this construction actuallyoccurs, it is also ungrammatical with inversion It is thus very important

to be sure that judgements come from speakers who actually use thestructure in question

A final problem relates to the co-occurrence of features in the mar of individual speakers If one is interested simply in documentingthe linguistic structures used in Belfast, one can be content with estab-lishing which structures are used by anyone in the area But if one isinterested in defining the nature of linguistic competence, it becomesimportant to know whether structures can, or must, co-occur in thegrammar of a single speaker This is particularly important in relation toparameter setting; one must be sure that structures claimed to be theresult of a single parameter setting in fact necessarily co-occur in thegrammars of speakers For example, as noted above, subject-verbagreement sometimes fails to take place in Belfast English; in Chapter

gram-2, we argue that this is because SPEC/TP is available in Belfast English

as a checking position for the subject; in Chapter 5, we consider sion in embedded clauses, which as we point out might conceivably beexplained along similar lines, since one proposal made in relation toother languages with similar phenomena is that the subject may be inthis position We in fact reject this analysis independently on linguisticgrounds, but we note also that even if it were possible to make theanalysis work, we would still have to take into account the fact thatmany speakers who use embedded inversion never use non-agreeingverbs; there might of course be a separate explanation for this, but nev-

inver-ertheless it is important to note whether features necessarily co-occur inspeakers' grammars; otherwise, one may make misleading analyses interms of parameter settings.

It is thus necessary to proceed with some care in studying a dard dialect With that caution in mind, let us go on to look at some ofthe ways in which Belfast English differs from standard English

Subject-Verb Agreement

There is a noticeable difference in subject-verb agreement patternsbetween standard English and Belfast English

In standard English, the -s ending occurs only with third person

sin-gular subjects; in Belfast English, a plural subject NP may occur with a

verb showing the -s ending, a phenomenon pointed out in a number of

sociolinguistic studies, and known as "singular concord" (Policansky

1976, Milroy 1981, Finlay 1988)

(1) a These cars go/goes very fast,

b The eggs are/is cracked.

It is not simply the case that the plural endings have a single form, 0,

in standard English but two variants, 0 and -s in Belfast English, -s is not in free variation with 0, for the occurrence of -s with plural subjects

is restricted in a number of ways Thus as Milroy (1981) notes, sonal pronouns cannot normally have singular concord

per-(2) a *They goes very fast.

b *They is cracked.

Singular concord is also generally impossible with inversion

(3) *Is the eggs cracked?

It will be argued here that "singular concord" verbs are in fact pletely unmarked for agreement; in sentences with singular concord,AGRS is weak in both V- and N-features, so that the verb does not raiseabove Tense before spell-out, and the subject (which we take followingKitagawa [1986], Koopman & Spottiche [1988], Kuroda [1988J, and

com-16

many others to originate VP-internally) need not raise to SPEC/AGRSP,but rather moves only as far as SPEC/Tense.

The chapter is organised as follows In the first section, we outlinethe facts of singular concord and show that it is indeed a syntactic phe-nomenon; we argue that singular concord verbs are in fact completelyunmarked for agreement Later, under the section titled "Pronouns andSingular Concord," we discuss the impossibility of singular concordwith certain pronouns The final section, "Inversion and Singular Con-cord," examines the interaction of singular concord with I-to-C move-ment

What Is Singular Concord?

We begin by outlining the facts of singular concord; next, we establishthat it is a syntactic, as distinct from semantic or pragmatic, phenome-non; we go on to show that singular concord verbs, which appear toshow third person singular agreement, in fact are unmarked for agree-ment We conclude by showing that singular concord is incompatiblewith nominative Case and argue that this is because the subject is not inSPEC/AGRSP, but in the SPEC/Tense P position

The Facts of Singular Concord

We begin here by outlining the facts of singular concord, since this struction has not previously been described in the syntactic literature.Singular concord is always optional; that is, it is always possible tohave the plural form of the verb with a plural subject As pointed out byPolicansky (1976), it would therefore be more correct to use the term

con-"variable concord"; However, it should be noted that the variabilityonly exists for plural subjects; thus, while it is possible to use a singularverb when the subject is plural, as in (1) above, it is not possible to use aplural verb with a singular subject

(4) *This car go very fast.

(5) *The egg are cracked.

Singular concord is available for most speakers in all tenses of theverb which marks agreement; thus in addition to the present, it appears

with the verb be in the past tense (Be is of course the only verb to show

agreement in the past tense in English)

(6) The students was late.

And it also occurs in the "historic present." This is a form of the verbwhich is used in storytelling contexts, and it differs from the simple pre-sent in using -s on the end of the first person singular as well as the thirdperson

(7) The girls goes and tells them

However, while all users of singular concord can use it in the present

tense, for some it is unavailable in the past tense of be and/or the

his-toric present

As we noted above, singular concord is impossible if the subject is asimple personal pronoun However, there are certain circumstanceswhen pronominal subjects are possible Thus for example pronounswhich are part of a co-ordination can have a singular verb, provided thatthey are not nominative

(8) Us and them is always arguing

(9) Him and me goes there every week

(10) Her and her mother works there

(11) *We and they is always arguing

(12) *He and I goes there every week

(13) *She and her mother works there

Demonstratives allow singular concord Note that them, rather than

those, is the distal demonstrative in Belfast English.

(14) These is cracked

(15) Them is no good

Moreover, Belfast English has an additional set of plural personal

pronouns not found in standard English, usuns, yousuns, and themuns;

these allow singular concord

(16) Usuns was late

(17) Themuns has no idea

Thus it is not the case that all pronouns require agreement; only a

subset of them does, the series of simple personal pronouns; we, they, and youse, which is the second person plural pronoun in Belfast Eng-

lish, all require agreement

As can be seen from the previous examples, singular concord is in

general possible both with raising verbs and with verbs which remain in

VP However, it is impossible with raising verbs if an adverb intervenesbetween the subject and verb; whereas (18a) is grammatical, as in stan-dard English, (18b) is not

(18) a The children really are late.

b *The children really is late.

Thus, it seems to be the case that the adverb position between the ject and the topmost projection of INFL which exists in English isunavailable in singular concord

sub-There is not a general adjacency requirement between subject andverb in singular concord, however; an intervening adverb is fine withverbs which remain in the VP

(19) The children really likes pizza.

(20) These books probably costs a lot.

We noted above that agreement is obligatory where subject-auxiliaryinversion has taken place Thus the following are ungrammatical

(21) *Is the students here?

(22) *Has the children arrived yet?

Summarising, then, in Belfast English a verb with a third person gular ending can occur with a plural subject NP, provided the subject isnot a simple personal pronoun, the verb is not inverted, and, if the verb

sin-is one which rasin-ises, nothing intervenes between the subject and theverb

Singular Concord as a Syntactic Phenomenon

It is important to show at the outset that singular concord is indeed asyntactic phenomenon For it is well known that subject-verb agreementcan be influenced by non-syntactic factors; thus, in standard BritishEnglish, a syntactically singular NP can have plural agreement if itrefers to a group

(23) The government is/are planning to resign.

(24) The committee has/have agreed on this.

It might be thought that singular concord is the converse of this; that is,

that a syntactically plural NP identifying a single "group" might have asingular verb However, this is not the case Singular concord occurswith all types of NP except pronouns, and does not favour any particu-lar type of NP Moreover, and more clearly, singular concord structuresdiffer from the collective plural examples in that they do not behavelike sentences with normal agreement patterns Thus, as noted above,inversion is impossible This is not the case with the collective exam-ples, which permit inversion.

(25) Are the government planning to resign?

(26) Have the committee agreed on this?

(27) *Is the members planning to resign?

(28) *Has the office-bearers agreed on this?

In addition, the collective examples permit an adverb to intervenebetween the subject and verb with raising verbs, whereas as notedabove, this is impossible with singular concord

(29) The government really are planning to resign

(30) The committee probably have agreed on this

(31) *The members really is planning to resign

(32) *The office-bearers probably has agreed on this

Thus, there are distinct differences between singular concord casesand those where group nouns are concerned In the latter case, there donot seem to be structural differences between these and sentences show-ing normal agreement, whereas in the singular concord cases, there aredefinite structural differences

"Singular Concord" as Lack of Agreement Marking

One of the superficially surprising aspects of singular concord is that itinvolves the use of a form with an overt inflection, where a base form ofthe verb would suffice; rather than appearing in its base form, as in (33),the verb is in fact gaining an ending, as in (34)

(33) The children shout all the time.

(34) The children shouts all the time.

We will begin by showing that so-called singular concord does notinvolve merely substituting the appropriate singular verb for a plural

one; rather, it involves complete lack of agreement marking agreement would therefore be a more accurate term, but we will con-tinue to refer to it by its traditional name of singular concord.)

(Non-We noted above that co-ordinated pronouns may have singular verbs;now if singular concord simply involved substituting the appropriatesingular verb for the plural, we would expect a first person plural sub-

ject to have a first person singular verb We can test for this with be, the

only verb in English to have a first person singular which differs fromthe first person plural What we find is that the first person singularnever occurs with a plural subject; apart from the first person pluralform, the only form we find is the third person singular

(35) *John and me am going.

(36) John and me is going.

(37) *Me and you am supposed to go.

(38) Me and you is supposed to go.

That these are indeed first person plural subjects is seen from their responding reflexive forms

cor-(39) John and me kicks ourselves.

(40) Me and you is supposed to help ourselves.

Similarly, -uns pronouns and pronouns which occur as part of a larger

NP always have the third person singular verb form

(41) *Usuns am happy.

(42) Usuns is happy.

(43) *Us students am very hardworking.

(44) Us students is very hardworking.

This indicates that singular concord is not simply agreement in whichplurality is disregarded; rather, where there is not full agreement theverb is always in the form normally found with the third person singu-lar, which appears to be a kind of default agreement

The lack of agreement raises the question of whether the subject is infact in SPEC/AGRSP in this structure Mohammad (1989), discussing asimilar but not identical agreement phenomenon in Arabic, suggeststhat lack of agreement occurs when the subject does not raise to INFL,remaining rather in VP; thus, it is unable to copy its features to AGR,and there is no agreement His analysis suggests that the subject posi-tion is filled by a null expletive; this will not work for Belfast English,

for two reasons First, there is otherwise no evidence of null expletives

in Belfast English; just as in standard English only overt expletives arepossible

(45) a * NP [e] is a book on the table.

b There is a book on the table.

(46) a * NP [e] is surprising that John won.

b It is surprising that John won.

Second, there is no word order difference between the sentences withagreement and those without In Arabic, which is a VSO language,agreement is obligatory in SVO sentences whereas non-agreementoccurs in VSO sentences, as (47a and b) (from Mohammad 1989) indi-cate

(48) *Is the eggs cracked.

(49) *Has the students arrived.

Chomsky (1992), commenting on these facts in Arabic, suggests thefollowing analysis within the Minimalist program: The NP feature ofAGRS can be strong or weak in Arabic, strength or weakness correlatingwith agreement or non-agreement Thus, a verb showing agreement has

a strong NP-feature and triggers obligatory subject-raising in the tax, whereas one without agreement has a weak NP feature, meaningthat the subject may not raise before spell-out This analysis has thesame problems as Mohammad's in relation to Belfast English; here, theword order is identical in both agreement and non-agreement sentences,showing that the subject has raised to a position above the verb in bothcases; we cannot therefore straightforwardly propose an analysis underwhich non-agreeing verbs do not trigger subject raising

syn-Thus, the details of neither Mohammad's nor Chomsky's analysiscan be wholly adopted for Belfast English But there is some evidencethat it is similar to the Arabic case in that the subject is not inSPEC/AGRSP An important piece of evidence in this respect relates toCase; singular concord is incompatible with nominative Case-marking.Now, it is widely accepted that nominative Case is assigned or checkedunder Spec/head agreement between the subject NP in SPEC/AGRSPand AGRS (see for example Chomsky [1989,1992], Mahajan [1990]) If

we can show that nominative Case does not occur in singular concord,then this may be evidence that the subject is not in (and cannot get by

LF into) a position in which nominative Case is checked; of course, thesubject will have to receive Case-checking in some other way, or thederivation will crash because it will contain an NP whose Case has notbeen checked; we return later to look at how this occurs For themoment, let us consider the data relating to nominative Case-marking.Since singular concord is impossible with simple personal pronouns,

we cannot use these to test for nominative Case-marking However, wecan again use pronouns which are part of a larger NP

Co-ordinate pronouns can occur with singular concord, as notedabove In general, co-ordinated pronouns in Belfast English may occureither in their strong form (which is morphologically identical to theaccusative, and does not vary for Case), or bear the Case assigned thewhole NP of which they are part (For a discussion of Case-marking inEnglish co-ordinates, within a different framework from the one weadopt here, see Parker, Riley, & Meyer 1988) Thus in Belfast English,and indeed most varieties of colloquial English, both (50a) and (50b)are possible

(50) a He and I are going.

b Him and me are going.

Some dialects of English appear to allow strong forms identical to thenominative

(51) They went with he and I.

(52) John helped you and I.

but such forms are impossible in Belfast English Co-ordinated nouns always occur with either the Case assigned to the whole NP ofwhich they are part, or in the strong form, which is identical to theaccusative

pro-In Belfast English, non-agreement is possible only if the pronounsare not nominative.

(53) a *He and I is going.

b Him and me is going.

(54) a *You and they is going.

b You and them is going.

Although there is considerable variability in agreement with nate NPs in English in general (see Cowart 1991), and one patternfound is agreement with one of the conjuncts, it is clearly not the case in(54b) that the verb is agreeing with one of the coordinated NPs; for both

co-ordi-you and them should trigger are rather than is.

In structures such as the following also, both the nominative and thestrong form may occur in colloquial English

(55) a We students are going

b Us students are going

Again, in Belfast English, singular concord is impossible with the inative form

nom-(56) a *We students is going

b Us students is going

It might be objected here that what we are detecting in examples like(56a), and also in co-ordinate examples like (53a) and (54a), is notungrammaticality as such but rather a sociolinguistic mismatch—withsingular concord being a non-standard feature, and thus soundingstrange when placed together with the formal, prestige form involvingco-ordinate pronouns in the nominative case It is, however, possible torule this out by testing the effect of similar mismatches; thus for exam-ple the use of non-standard past tenses is highly stigmatised, much more

so than singular concord But the use of co-ordinated nominative nouns with such verb forms is judged to be grammatical by nativespeakers

pro-(57) He and I seen them.

(58) We students done the work.

On the other hand, the combination of nominative pronouns with gular concord is judged to be completely ungrammatical

sin-Assuming that in the structures in question pronouns may freelyeither appear in their strong form or receive the Case assigned to the

larger NP of which they are part, it seems that singular concord isincompatible with the nominative Case.

Thus, our analysis of singular concord must exclude the possibility ofnominative Case being checked in this construction As we notedabove, nominative Case is often considered to be a property ofSpec/head agreement in AGRSP (Chomsky 1989, 1992; Mahajan 1990);thus, it is not surprising to find that where agreement is absent, so isnominative Case-marking, and this provides evidence that the subject isnot in these cases in SPEC/AGRSP, where it would have to have nomi-native Case

This is not to say that nominative Case-marking is universally patible with the subject occurring outside SPEC/AGRSP In examples(47a and b) from Arabic (repeated below), the NP bears nominativeCase regardless of whether it is in SPEC/AGR,, as in (a), or not, as in(b), at spell-out Nominative Case must of course be checked inSPEC/AGRSP at some stage in the derivation, but this need not bebefore spell-out; the subject can raise to SPEC/AGRS at LF, and checkCase at this level

Case-A piece of evidence which may help ascertain the position of the ject can be found in adverb placement, which as mentioned above dif-fers between sentences with agreement and those without As Pollock(1989) notes, in addition to the preverbal adverb position found in bothEnglish and French, there appears to be an additional position betweenthe subject and the highest head of IP available in English, so that anadverb may appear between the subject and a verb which raises, as in(59a) and (60a) Now it is unclear exactly what this position is, or whyadverbs should be restricted to occurring in this or a small number ofother positions However, the important point in relation to our presentconcerns is that the adverb position between the subject and a raisedverb is not available in singular concord

sub-(59) a The eggs really are cracked.

b *The eggs really is cracked.

(60) a The girls probably have left.

b *The girls probably has left.

This again indicates that the subject is not in the same position as innon-singular concord cases

Note that the unavailability of this adverb position in singular cord is in fact an argument in favour of the existence of such a position;Baker (1991) argues that there is a single adverb position and that insentences where apparently an adverb intervenes between a subject andraised verb, this is because in fact the verb has not raised Thus forBaker the structure of (59) and (60) are identical to that of (61), wherethe verb has not raised; but note that in Belfast English, the adverb posi-tion between the subject and verbs which remain in VP is available insingular concord structures, indicating that a different structure isinvolved

con-(61) a The girls probably like coffee.

b The girls probably likes coffee.

Thus, we have noted that in singular concord, the subject cannot haveovert nominative Case and the verb is not marked for agreement.Assuming that the subject originates in VP, it seems that the subject hasnot risen as high as SPEC/AGRSP; the question then arises of where thesubject is at spell-out Apart from the difference just noted, raisingverbs (in the sense of Pollock [1989]) manifest exactly the same pattern

of placement relative to negatives and adverbs with singular concord asthey do in sentences with agreement, indicating that the verb has raisedoutofVP

(62) The eggs is not cracked.

(63) The eggs is probably cracked.

Chomsky (1992) argues that the verbs be and have raise for checking

in the syntax, because they are too weak semantically to be visible formovement at LF Since verbs are marked for tense in the singular con-

cord construction, be must be at least as high as the Tense position in

the overt syntax; otherwise, since movement is unavailable at LF, theV-feature of tense would remain unchecked at LF, and the derivationwould crash Given the subject-verb order, the subject must occupy anode higher than Tense The obvious candidate here is SPEC/Tense P

A strong argument in favour of this is the dependence of the ability of singular concord on the choice of tense for some speakers As

avail-we noted above, although most speakers have singular concord in alltenses, for some it is restricted to present tense, or restricted to presentand past and excluded in the historic present Thus while most users ofsingular concord find (64), (65) and (66) all grammatical, for a substan-tial proportion, only (64) and (65) are grammatical, and for a smallergroup only examples in the present tense like (64) are possible

(64) The teachers is busy.

(65) The teachers was busy.

(66) The teachers goes and tells the Principal about it.

(historic present)

Thus, the availability of singular concord is for some speakers dent on the content of Tense, suggesting that it is Tense which isresponsible for Case-checking, and that for these speakers only certaintenses can check Case

depen-Note that this means that the ability to assign or check Case must be aproperty, not just of functional elements (e.g., Tense) but of particularinstantiations of those elements (e.g., the [+pres] Tense morpheme).Although this increases the number of grammars potentially available,

it is in line with views which see the differences between languages asrestricted to the properties of individual morphological elements in thelexicon, rather than some abstract global properties of the morphology

in general; and it is going to be necessary in any case to handle the fact

that for example certain complementisers (e.g., for in English: see

Chapter 5) assign/check Case, whereas others do not

An argument for the availability of Case-marking/checking inSPEC/Tense in Belfast English comes from the behaviour of negativepolarity items in that dialect In Belfast English, for some speakers, neg-ative polarity items can occur in subject position in matrix clauses

(67) Anybody wouldn't be able to do that.

(68) I was surprised that anybody didn't go.

Such sentences are ungrammatical in standard English, a fact generallyattributed to a universal requirement that NPIs be strictly c-commanded

by a negative operator at S-structure (Linebarger 1987, Laka 1990);since the subject position is not c-commanded by NEG in matrixclauses in English, negative polarity items cannot occur there Thequestion arises as to how Belfast English manages to escape this

requirement Duffield (1993) argues that the licensing condition on NPIitems can be reformulated to apply to LF, and is as follows:

(69) Case Condition:

For a Negative Polarity Item to be properly licensed, at least one Case-marked member of the NPI chain must be c-commanded by a negative operator.

Now clearly in general in NP chains, only the highest member of thechain is Case-marked However, Duffield points out that if, in BelfastEnglish, there are two potential Case-marking positions for the subject

NP, SPEC/Tense and SPEC/AGRs, then there can be a Case-marked trace

of the subject in SPEC/TP Duffield argues that in a sentence like: (70)

(70) Any country couldn't stand it.

the negative element c-commands a subject trace in SPEC/TP, as in(71),

(71) [AGRPAny country i [AGR S couldn't]

T0 t

j] [VP t; [v' stand it ]]]]

and that this subject trace is Case-marked There is at first sight a lem with allowing Case-marked positions in NP chains if one is tomaintain the least-effort principle: it may not be clear why an NP wouldraise if its Case has already been checked It may be that the conditionshould be re-formulated to refer to potential Case-marking positionsrather than actual traces Alternatively, it may be that the requirementthat the head of an NP-chain be Case-marked is not strictly correct An

prob-NP may be able to check Case in one position (e.g., SPEC/Tense) andthen move to another position to check agreement, so that the highestposition of the NP-chain is not the one which is Case-marked in fact.However, there is another way of interpreting this, without reference

to subject traces Suppose that, as Duffield argues, the negative elementleft-adjoins to TP at LF Now in Belfast English, if no further raisingtakes place at LF, then at LF the negative element will be in the follow-ing configuration:

(72) [ TP Neg[ TP Any country [ T could [ VP stand it]]]]

If the subject has already checked Case, and the (non-agreeing) verbdoes not require to check agreement, then there is nothing to force LF-raising; the subject and verb can remain in Tense P, and presumably

AGRS will delete since it will not be required for semantic tion, leaving the LF representation as in (72).

interpreta-One way to distinguish between these two analyses would be to sider what happens when there is a plural NPI subject If as we haveargued the subject in singular concord is in SPEC/Tense P, then itshould be impossible to have subject-verb agreement with an NPI sub-ject That is, a sentence like (73a) should be impossible, whereas thecorresponding sentence without agreement should be possible

con-(73) a *Any animals aren't coming

b Any animals isn't coming

This seems to be the case; NPIs are only possible with non-agreeingverbs, indicating that it is not the potential availability of Case-marking

in a certain position, but rather the actual presence of a Case-markedelement in the structure, that is the important issue

If an approach along these lines is right, then it would be, as Duffieldpoints out, an important step towards reformulating the conditions onNPI licensing as LF, rather than S-structure, conditions, for it will beobserved that at S-structure, the negative element is below the negative

polarity item; for example in the following sentences, any is above n 't and not, which are presumably in Tense (with did) and in NEG, respec-

tively

(74) a Any student didn't apply for the job

b Any student did not apply for the job

Under the Minimalist program, S-structure does not exist as a rate level and all conditions are interface conditions, applying at LF or

sepa-PF The availability of negative polarity items in subject position inBelfast English seems to indicate that this is the case; at S-structure theNPI in subject position is not c-commanded by the negative element, aswould be required by the S-structure condition on negative polarityitems However, if the condition is an LF condition, and negatives raise

to adjoin to TP, then we have a natural way to account for the ity of NPIs in subject position in standard English, and their impossibil-ity in standard English NPIs are possible in Belfast English because thesubject can be in SPEC/TP at LF

possibil-That there is a link between singular concord and NPI-licensing insubject position seems clear from the the fact that the two phenomenaseem to go together in speakers' grammars; those speakers who allow