Variable Grammars: Verbal Agreement in Northern Dialects of English (Linguistische Arbeiten) [1 ed.] 3484304960, 9783484304963

Citation preview

Linguistische Arbeiten

496

Herausgegeben von Hans Altmann, Peter Blumenthal, Klaus von Heusinger, Ingo Plag, Beatrice Primus und Richard Wiese

Lukas Pietsch

Variable Grammars: Verbal Agreement in Northern Dialects of English

Max Niemeyer Verlag Tübingen 2005

Bibliografische Information der Deutschen Bibliothek Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.ddb.de abrufbar. ISBN 3-484-30496-0

ISSN 0344-6727

© Max Niemeyer Verlag GmbH, Tübingen 2005 http: //www. niemeyer. de Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Printed in Germany. Gedruckt auf alterungsbeständigem Papier. Druck und Einband: Digital PS Druck AG, Birkach

Contents

List of maps List of figures List of tables List of abbreviations Acknowledgments 1.

VII VII VII IX X

Introduction 1.1. Data and method 1.2. The Northern Subject Rule: descriptive problems 1.2.1. Type of subject, position of subject 1.2.2. Type of verb 1.2.3. Other patterns of agreement variation 1.2.4. The 'markedness paradox'

1 4 5 7 12 13 14

2. Theoretical accounts of verbal agreement variation 2.1. Proposed formal analyses of the NSR 2.1.1. Henry (1995) 2.1.2. Corrigan (1997) 2.1.3. Börjars and Chapman (1998) 2.1.4. Hudson (1999) 2.2. Theoretical approaches to variation 2.2.1. Competence and performance 2.2.2. Individual grammar and community grammar 2.2.3. Variable-rules theory 2.2.4. Stochastic Optimality Theory 2.3. A usage-based approach

16 16 16 21 25 27 28 29 32 35 36 38

3.

History and origins of the NSR 3.1. The development of the NSR from Middle English 3.2. The Origins: Old English affix reduction and analogical spread 3.3. The Scandinavian hypothesis 3.4. The Celtic hypothesis

45 45 50 57 60

4.

Verbal agreement in the SED 4.1. The data 4.2. An overview 4.3. Non-NSR agreement patterns 4.3.1. West Midlands verbal -n forms 4.3.2. Verbal-Ä with / 4.3.3. Verbal -5 with thou 4.3.4. Levelling of was/were 4.3.5. Verbal -5 with we, you, and they 4.4. The Northern Subject Rule in the SED

63 64 69 71 71 74 76 76 80 82

5.

Verbal agreement in the NITCS 5.1. The data 5.2. Pronominal subjects 5.3. Full NP subjects

93 93 97 103

VI 5.3.1. 5.3.2. 5.3.3.

5.4.

5.5. 6.

Regional Variation Social and apparent-time Variation Linguistic conditioning factors: syntactic constellation 5.3.3.1. Inverted clauses 5.3.3.2. Relative clauses 5.3.3.3. Subject-verb distance 5.3.4. Linguistic conditioning factors: types of subject NPs 5.3.4.1. Demonstrative pronouns 5.3.4.2. Interrogative pronouns 5.3.4.3. Indefinite quantifying phrases 5.3.4.4. The indefinite pronoun ones 5.3.4.5. Lexical items: times, days, things 5.3.4.6. Other conditioning factors 5.3.5. Conclusions: the conditioning of NSR effects Existential there 5.4.1. Regional variation 5.4.2. Social and apparent-time variation 5.4.3. Linguistic conditioning factors 5.4.3.1. Negation 5.4.3.2. Clause structure Conclusions: agreement in Northern Irish English

Verbal agreement in FRED 6.1. The data 6.2. Pronominal subjects: present tense verbs 6.3. Pronominal subjects: was/were levelling 6.4. Existentials and -r levelling 6.5. Full NP subjects

103 105 109 110 110 113 114 115 116 117 118 118 119 121 123 125 127 130 130 132 133 135 135 137 138 140 145

7. Conclusions

151

Appendixes

155

A. Varbrul multivariate analysis methods

155

1. Maximum likelihood regression analysis 2. Significance testing with Varbrul 3. Dependency and interaction between factor groups 4. Clustering speakers B. SED data lists C. NITCS data tables D. FRED data tables

155 156 157 159 160 190 196

References

199

Index

213

VII

List of maps

Map Map Map Map Map Map Map Map Map Map Map Map Map Map Map Map Map Map

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

The NSR in Late Middle English (after LALME I: 467, IV: 1 lOff) Verbal -n in Late Middle English (after LALME I: 467, IV: 11 Off) Important subject-verb agreement isoglosses in the SED Verbal -n forms in the SED Verbal -s forms with / in the SED Verbal agreement forms with thou in the SED Was/were neutralization in the SED The was/weren 't negation effect in the SED Verbal -s with I/you/we/they in the SED Plural verbal -s in the SED (1) Plural verbal -s in the SED (2) Religious communities in the NITCS NITCS locations and traditional dialect areas NITCS Locations clustered for NSR usage NITCS locations clustered for -s/-r frequencies in existentials FRED locations Division of SED locations into seven areas Fieldworkers in the NITCS

46 47 70 72 75 77 79 81 83 85 87 96 97 104 126 136 190 192

List of figures

Figure Figure Figure Figure Figure

2.1. 2.2. 2.3. 2.4. 3.1.

A construction inventory for Standard English agreement A construction inventory for NSR agreement A construction inventory for variable NSR agreement A partial construction inventory for structured variation in NSR agreement Development of agreement paradigms in Old and Middle English

41 41 42 43 52

List of tables

Table Table Table Table Table Table Table Table Table

2.1. 2.2. 4.1. 4.2. 4.3. 5.1. 5.2. 5.3. 5.4.

Distribution of verb forms in Irish and English Plural verbal -s in relatives and non-relatives Partial was (S) and were (R) paradigms Present tense verbal -s with NP subjects in the SED, by area Present tense verbal -s with NP and pronoun subjects in the SED, by area Verbal -5 with pronominal subjects in the NITCS Plural verbal -s in the NITCS, by area and aspect Plural verbal -s in the NITCS, by area and verb type Plural verbal -s in the NITCS, by area and age group

23 35 78 89 90 100 105 106 106

VIII Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table

5.5. 5.6. 5.7. 5.8. 5.9. 5.10. 5.11. 5.12. 5.14. 5.15. 5.16. 5.17. 5.17. 5.18. 6.1. 6.2. 6.3. 6.4. 6.5. 6.6. 6.7.

Plural verbal -s in the NITCS, by area and sex Plural verbal -s in the NITCS, by area, sex, and age group Plural verbal -s in the NITCS, by area and religious community Plural verbal -s in the NITCS, by area, religious community, and age group Plural verbal - i in relative clauses in the NITCS, by relative type Plural verbal -s in the NITCS, by subject and clause type Partial Varbrul model for the NSR in the NITCS Existential -r in present tense and past tense in the NITCS, by area Existential -r in the NITCS, by area Existential -r in the NITCS, by area and age group Existential -r in the NITCS, by area, sex and age group Existential -r forms in the NITCS, by age group and community Existential -r in the NITCS, by number and polarity Existential -r in the NITCS, by number and clause type Was with pronoun subjects in FRED Nonstandard -r forms in FRED Varbrul model for existentials in FRED (speakers A) Frequencies of were in FRED (speakers B) Varbrul models for singular were in FRED (informants B-D) Effects of the BE constraint on the NSR Varbrul model for NSR in FRED

107 107 108 108 Ill 117 122 127 127 128 129 130 131 132 139 141 142 143 144 147 148

SED token listings B.l. Verbal-« B.2. Generalized'm B.3. lis B.4. Thou is B.5. Thou bist/be/bin B.6. Thou are/art B.7. Thou does B.8. You be B.9. Plural was B. 10. Singular were Β. 11. Use of -s with I, other verbs Β. 12. Use of -s with thou, other verbs B.13. Use of -s with we, other verbs Β. 14. Use of -s with you, other verbs Β. 15. Use of -s with they, other verbs B.l6. Verbal -s in NSR contexts Β. 17. Non-use of verbal -s in NSR contexts

161 162 163 163 164 165 166 167 167 168 169 170 171 171 172 173 188

Corpus data tables B.l. B.2. B.4. C.l.

NITCS locations NITCS informants Existential -r in the NITCS FRED informants

191 192 195 196

IX

List of abbreviations

Agr s CLAE BNC FRED FW F. W. GLA Infi IP LA Ε LALME NITCS NP NPI OT SED SED-BM Spec SV Τ TRS TRS-D VP

Subject-Agreement (functional head in Principles-and-Parameters syntax) Computer-Developed Linguistic Atlas of England (Viereck 1991/1997) British National Corpus Freiburg Corpus of English Dialects Fieldworker Factor weight ( Varbrul) Gradual Learning Algorithm Inflection (functional head in Principles-and-Parameters syntax) Inflection Phrase (functional projection in Principles-and-Parameters syntax) Linguistic Atlas of England (Orton/Sanderson/Widdowson 1978) Linguistic Atlas of Late Middle English (Mclntosh/Samuels/Bensik 1986) Northern Irish Transcriped Corpus of Speech (Kirk 1991) Noun Phrase Negative Polarity Item Optimality Theory Survey of English Dialects Survey of English Dialects-Basic Material (Orton et al„ ed. 1962-1971) Specifier (position in Principles-and-Parameters syntax) Subject-Verb (clausal constituent order) Tense (functional head in Principles-and-Parameters syntax) Tape-Recorded Survey of Hiberno-English Speech Tape-Recorded Survey of Hiberno-English Speech-Digitized (Hickey 2002) Verb Phrase

English and Scottish counties (pre-1974) (First column: FRED, second column: SED abbreviations) ANS BAN CAM CHS CUL DBY DFS DUR FIF HEF HUN INV KCD LAN LEI LIN LKS

(C) (Ch) (Cu) (Db) (Du) (He) (Hu)

(La) (Lei) (L)

Angus Banffshire Cambridgeshire Cheshire Cumberland Derbyshire Dumfriesshire Durham Fife Herefordshire Huntingdonshire Inverness-shire Kincardineshire Lancashire Leicestershire Lincolnshire Lanarkshire

NBL NFK NTH NTT PEE PER ROC SAL SEL STS SUT WAR WES WLN WOR YKS

(Nb)

(NO (Nth) (Nt)

(Sa) (St) (Wa) (We) (Wo)

(Y)

Northumberland Norfolk Northamptonshire Nottinghamshire Peebleshire Perthshire Ross and Cromarty Shropshire Selkirkshire Staffordshire Sutherland Warwickshire Westmorland West Lothian Worcestershire Yorkshire

Acknowledgments

This book is the revised version of my doctoral thesis, which was submitted to the University of Freiburg in January 2003 under the title Subject- Verb Agreement in Northern Dialects of English. The research for this study was supported, during its first phase, by a doctoral grant from the Studienstiftung des Deutschen Volkes, and later, as part of the research project on "English Dialect Grammar in a Typological Perspective" at the University of Freiburg, by a project grant from the Deutsche Forschungsgemeinschaft. I thank both organizations for their generous support. Being undertaken by a non-native scholar working abroad, this study could not have been written without those who gave their assistance in making empirical data accessible. My thanks are due to John Kirk for providing access to the NITCS; to him and Raymond Hickey for their efforts at recovering the lost treasures of the Tape-Recorded Survey of Hiberno-English Speech; and to Clive Upton for assistance with the unpublished material of the SED. Very special thanks go to those individuals and organizations who generously contributed speech recordings, often the fruits of long years of their own labour, to be used in the new Freiburg Corpus of English Dialects. Most crucial for the present work was the contribution by Roger Leitch, from the wealth of his personal collection of oral history interviews made in Scotland. Other material comes from Jane Petrie (Highlands County Archives), the Ambleside Oral History Group, Falkirk Archives, Northumberland Record Office, Middlesbrough County Archives, Pennine Heritage at Hebden Bridge, and the Oral History Group at Huyton (Merseyside). The student assistants at Freiburg University were of invaluable help in sharing the work of transcribing and proofreading the corpus recordings with me. Among them, Allison Felmy, Tara McCartney, Mhairi Pooler, and Ricky Walla made great contributions to the database. Thanks are also due to Anthea Gupta, Clive Upton, Esther Asprey and Kate Wallace (Leeds) for special assistance with transcription problems. I am indebted to Kevin McCafferty, Sali Tagliamonte, Peter Auer, Hildegard Tristram, Ingo Plag, Sally Thomason, and Elena Lieven, who gave valuable feedback and advice during the preparation of this work. The same goes for the participants of the 3rd UK Language Variation Conference in York, July 2001, of the Workshop on Dialect Syntax in the Westgermania in Freiburg, November 2001, and of the 39th Linguistics Colloquium in Amsterdam, September 2004, where I had the opportunity of presenting parts of this work; and for the many colleagues at the Research Centre on Multilingualism at Hamburg University with whom I have had the pleasure to work and share ideas during the time that led up to the final publication. Most special thanks are due to Bernd Kortmann for his constant support during all phases of this work, to Christian Mair for stimulating discussions; and to my old Freiburg colleagues, Lieselotte Anderwald, Manfred Krug, Nuria Hernandez y Siebold, Benedikt Szmrecsanyi, and especially Susanne Wagner and Tanja Herrmann, who were the greatest team to work with. And, of course, to Effi, without whom all this would not have been possible at all. September 2004

Lukas Pietsch

1. Introduction

This study deals with change and variation in verbal agreement in the dialects of northern England, Scotland, and the north of Ireland. What unites these northern varieties of English is a common grammatical feature inherited from northern Middle English: they observe verbal agreement with personal pronoun subjects directly adjacent to their verbs (e.g. he sings vs. they sing), but allow for an invariant, neutralized verbal -s form everywhere else, most notably with full noun phrase subjects (birds sings) or in the second of two coordinated verbs (they sing and dances). This phenomenon has been known in the literature as the "Northern Subject Rule", henceforth abbreviated as NSR. Crucially, all the modern varieties in question display reflexes of this alternation in a variable form. While the agreement rule is largely categorical in the case of the pronoun subjects, including the neartotal prohibition of verbal -s after they, the use of non-standard verbal -s elsewhere is subject to competition with the standard forms (birds sing, they sing and dance). Hence, much of the present work will be concerned with the factors that condition inherent linguistic variation in this domain. The morphosyntax of verbal agreement in English non-standard varieties is a field that has received some more attention in past research than many other phenomena of grammatical, especially syntactic, variation. In many areas of genuinely syntactic variation, serious research traditions have only recently begun to be established and have had to face considerable practical and methodological difficulties (see for instance Milroy 1987: 143ff., Kortmann 1999, 2002; Glaser 1997; Barbiers/Cornips 2002). Verbal agreement, being primarily a morphological phenomenon - though syntactically conditioned - has proved somewhat easier to approach. This is certainly due to the fact that it is a relatively highfrequency phenomenon, whose occurrences in texts can fairly easily be identified and counted. Also, it has long been recognized that verbal agreement displays some fairly distinctive regional patterns in different varieties of English. It has therefore been given some treatment in traditional dialectology, at least occasionally (e.g. Murray 1873; Wright 1892, 1905; Orton et al„ ed. 1962-1971). The NSR, in particular, is routinely mentioned in dialect descriptions, though often only in a summary fashion (Beal 1993, 1997, 2004; Miller/Brown 1982; Miller 1989, 1993, 2004; Macafee 1983, 1994; Harris 1993; Ihalainen 1994; Robinson 1997, Filppula 1999, Shorrocks 1999). Some of the major historical patterns of verbal agreement have also been documented for older varieties (e.g. Macafee 1993, Mclntosh/Samuels/Bensik 1986). Recently, verbal agreement variation has become a stock-in-trade of corpus-based studies in variation and change. Most of these more recent studies have been inspired by either historical or sociolinguistic research questions, or a combination of both. Some of them have been based on diachronic corpora (e.g. MeurmanSolin 1992, Kytö 1993, Montgomery 1994, Ogura/Wang 1996), while most used synchronic local corpora from individual speech communities, either historical (e.g. Bailey/Ross 1988, Bailey/Maynor/Cukor-Avila 1989, Montgomery, Fuller/DeMarse 1993; Schendl 1996, 2000, Wright 2002, McCafferty 2003) or contemporary (e.g. Tagliamonte 1998, Smith/Tagliamonte 1998, Godfrey/Tagliamonte 1999, Britain 2002, Schreier 2002, Peitsara 2002).

2 The present study intends to contribute to this line of research in two ways. With respect to empirical description, it will take a predominantly geographical-comparative perspective. Being based on dialectal speech recordings sampled over relatively large areas, it can attempt to trace the distribution of related variation patterns across space (for similar approaches see Ihalainen 1991a, Klemola 1996, Anderwald 2002). This geographical comparison is all the more interesting because in many areas the NSR co-exists and competes with other non-standard patterns of agreement variation, some of them of more restricted local range and some of wider, supra-regional validity. The focus of the investigation will be to identify commonalities and differences in the linguistic conditioning of inherent, quantifiable variation across related varieties. Among other things, it will be shown that over and above the defining constraints constituting the NSR itself- there exist a number of recurrent types of constraints which condition agreement variation. They can be detected statistically in the form of probabilistic effects, and they can best be described in terms of prototypical syntactic environments which idiomatically favour a certain morphological realization of the verb over the other. The second contribution lies in an attempt to forge a closer link between this descriptive, empirical work on the one hand and recent grammatical theorizing on the other. Most of the existing studies of verbal agreement have treated the phenomenon of quantifiable variation mainly under its societal (in Chomskyan terms: "E-language") aspects. On the other side, those studies of dialect grammar that have been inspired by theoretical concerns about individual linguistic competence ("I-language") have usually concentrated on 'variation' in the sense of 'differences between lects', and have rarely addressed the issue of variability within lects (see papers in Black/Motapayane 1996 and Barbiers/Comips/van der Kleij 2002; for a notable recent exception see Henry 2002). There are few specific analyses regarding a formal theoretical characterization of NSR-related verbal agreement phenomena (Henry 1995, Corrigan 1997, Börjars/Chapman 1998, Hudson 1999), and none of them addresses inherent variability. In this study I will therefore propose a fresh attempt to relate the empirical observations regarding variable grammatical performance to the theoretical question of how variation is anchored in grammatical competence. In doing so, I will seek explanations in a theoretical framework inspired by functionalist and cognitive research traditions (Bybee 1985, Bybee/Hopper 2001, Langacker 1987; Croft 1995, 2001) Before proceeding, some introductory explanations regarding terminology are necessary. There are a large number of competing names for the phenomenon that stands at the centre of this study. Among them, "Northern Subject Rule", originally introduced by Ihalainen (1994), is the term that has found most currency in recent years (e.g. Godfrey/Tagliamonte 1999; Klemola 2000; McCafferty 2003, 2004; Chambers 2004). Both because of its implicit emphasis on the historical continuity associated with its northern origins, and because of its descriptive neutrality as to the various sub-patterns involved, this term will be adopted throughout the present study. Among the other names that have been in use for the same phenomenon is that of the "northern present-tense rule", introduced by Montgomery (1994) and also employed by other writers on Scots (Robinson 1997, King 1997). This term is descriptively not quite adequate for the present-day dialects, because in many of them the effects of the rule are seen first and foremost not so much in present tense verbs but in the alternation between was and were. More neutral and comprehensive terms, lacking the explicit reference to the northern origins, are the "personal pronoun rule" (Mcintosh 1989), or the "NP/PRO constraint" (Bailey/Maynor/Cukor-Avila 1989). It may be noted that the

3 latter term, in emphasizing the categorical distinction between pronouns and non-pronoun subjects, ignores the second vital ingredient of the rule, that of adjacency. A couple of other descriptive terms such as "singular concord" (Milroy 1981, Policansky 1982, Henry 1995), "nonconcord" (Corrigan 1997, Filppula 1999) or "nonconcordance" (Kallen 1991) refer only to the one side of the alternation, the non-standard use of verbal -s, but not to the alternation pattern (birds sings versus they sing) as such. As for the geographical scope of the present study, it was implicitly defined above by reference to the historical and geographical continuity of the NSR. The book deals with those varieties of British English that show direct reflexes of this northern pattern in their verbal agreement systems. For the present purposes, this definition includes Scots (notwithstanding any considerations regarding its status as separate from English). New-world Englishes had to remain out of the scope of this study for practical reasons, although many of them have also inherited some of the relevant features (cf. for instance Montgomery 1988, 1997b, Tagliamonte 2002). However, the additional inclusion of Ulster, as an early colonist variety, seemed justified because of its particularly close historical links with mainland northern varieties, especially Scots, and because of the attested great vitality of the NSR in the modern Ulster dialects. In grammatical statements of verbal agreement behaviour I will employ the term "agreement verb" to refer to all verbal forms which have the morphological potential of displaying a person-number contrast. This includes all finite present tense verb forms except for modals and subjunctives, and in addition the past tense forms was and were. The term "verbal -s" is used to include the irregular forms is and was, unless otherwise specified. The distinction between "pronominal" and "non-pronominal" subjects plays a crucial role in the definition of the Northern Subject Rule. In this context, "pronoun" and "pronominal", where not otherwise qualified, refer only to the closed set of simple personal pronouns: I, thou, he, she, it, we, you, they, and their dialectal reflexes. Strong pronouns (such as me, us), demonstratives (them, those), or indefinite pronouns and numerals (some, any, one) are classed together with all other, lexical noun phrases under "full NP subjects", for lack of a better term. In using this term, I do not wish to make any claim as to whether simple pronouns should be analysed as technically of the same category as other phrases in the framework of some formal syntactic model (for instance as DPs, NPs and so forth). Nevertheless, I will argue that the pronouns, by virtue of their high discourse frequency and closer affinity to the verb position, are likely to be processed and represented differently from other noun phrases. The difference between the simple personal pronouns and the other pronominal elements is of crucial descriptive importance. It will be shown that many of the non-personal pronouns show a behaviour exactly opposite to that of the simple personal pronouns, in that they strongly favour verbal -s where the latter prohibit it. This book is organized as follows. In the remaining sections of this introductory chapter, I will first give an overview of the data sources and quantitative evaluation methods to be used in the empirical parts of the study. This will be followed by a descriptive survey defining some of the core properties of the NSR and related agreement phenomena, and discussing some of the empirical problems connected with these definitions. This will serve to set the scene for the rest of the study, which is organized in five main chapters. The first two of them provide theoretical and historical background discussion, while the remaining three present empirical findings from three different sources of data.

4 Chapter 2 discusses theoretical approaches to the NSR in particular, and to probabilistic inherent variability in general. It first contains a critical evaluation of four proposed formal analyses of NSR systems. This is followed by a survey of the theoretical discussion about probabilistic variation in grammar, ranging from the variable-rules approach developed during the 1960s and 70s, to recent proposals of probabilistic extensions to Optimality Theory. Chapter 2.3 then gives a tentative sketch of a possible characterization of the NSR system within the usage-based framework advocated here. Chapter 3 is in some sense parenthetical within the rest of the study, presenting an excursus into the history and origins of the NSR. It serves to demonstrate how the theoretical approach developed in the previous chapter can account in a natural way not only for the synchronic existence of the NSR but also for its diachronic development. At the same time, it provides a discussion of an important alternative explanation recently proposed, which links the origins of the NSR to a language contact effect with a Celtic substrate (Klemola 2000). Of the following three data-oriented chapters, Chapter 4 provides a survey of the role of the NSR in the verbal agreement systems of the traditional (mid-20th century) rural dialects of Northern England. Chapter 5 investigates verbal agreement in a body of corpus data from Northern Ireland, and Chapter 6 integrates both sets of results into a larger picture, complementing them with more recent data from different parts of the northern British Isles. Chapter 7 presents a brief conclusion summing up the findings.

1.1.

Data and m e t h o d

The present study employs qualitative and quantitative methods of analysis in investigating corpora of (mostly) spontaneous, naturally occurring spoken language data. In doing this, its geographical-comparative perspective causes an obvious practical problem. It is virtually impossible for a single researcher, or even for a team of researchers, to collect original data representing such a large geographical area through their own fieldwork. In a study that needs to be based on spoken language data of vernacular varieties, the researcher therefore has to resort to re-using existing data sources wherever possible, and some of the methodological standards established in, for instance, current sociolinguistic work - regarding consistency of sampling and similar criteria - must by necessity be relaxed to some degree for this purpose. Three groups of data sources were used for the present study. The reasons for their choice will be outlined here; further discussion of the nature and characteristics of each can be found in the introductory sections of each of the three major data-oriented chapters later (Sections 4.1, 5.1, and 6.1). The only research projects that have in the past produced linguistic data sets of dialectal speech covering large areas with a sufficient amount of regional representativity (i.e. internal consistency of sampling, sufficient geographical density of data points, etc.) are the great dialect atlas surveys, such as the Survey of English Dialects (SED). These surveys have traditionally not focussed on morphosyntax, and their data does not (or not primarily) come in the form of natural speech recordings. Nevertheless, with the necessary amount of

5 methodological caution, data from the SED can be (and has been) used for the study of subject-verb agreement. SED data is the basis of the discussion of traditional dialects of northern England in Chapter 4. The same use could unfortunately not be made of the Linguistic Survey of Scotland (Mather/Speitel 1975-1986) and of the Survey of AngloWelsh Dialects. (Parry 1977-1979). Excellent material is available for Northern Ireland, in the form of the Tape-Recorded Survey of Hiberno-English Speech (TRS) and its electronic offspring, the Northern Ireland Transcribed Corpus of Speech (NITCS). This material, which represents rural speech recorded during the 1970s, has provided the single richest data source on verbal agreement variation within the present study and forms the basis of an extensive quantitative analysis in Chapter 5. At the University of Freiburg, a new corpus representing dialectal British English speech has been under construction since 2000: the Freiburg Corpus of English Dialects (FRED). It contains transcribed speech samples collected from a variety of sound recording sources, representing for the most part informants who speak conservative, markedly regional varieties, most of them born at around the turn of the last century and recorded between the 1970s and 1990s. A subset selected from a preliminary version of this collection forms the basis of Chapter 6. Taken together, these three sources provide a reasonably dense and consistent amount of material for comparative work. In the case of England, the data provides for a real-time component through the possibility of comparison between the SED and FRED. In the case of Northern Ireland, there is instead an apparent-time component, inherent in the design of the TRS. As a consequence, the material can be used to investigate both regional and diachronic variation, despite some obvious limitations regarding representativity. Some further data sources which could not be used for the present study include the British National Corpus (BNC) - as Anderwald (2002) has shown, its coverage of dialectal speech is rather poor - and the Helsinki Dialect Corpus, which was unavailable for technical reasons. The corpus data of FRED and the NITCS were analysed employing the multivariate analysis techniques known under the name of Varbrul, widely familiar as a tool for corpus analyses of this type in the sociolinguistic literature. For the purposes of the geographical investigations of the NITCS, a somewhat innovative extension of this statistical technique had to be devised. This technique is described in Appendix A, which also contains a brief introduction to standard Varbrul practices, intended for readers unfamiliar with the method.

1.2.

The Northern Subject Rule: descriptive problems

The constant core feature of the present-tense agreement system that has been characteristic of the northern dialects since the Middle English period can most concisely be described as follows: (1)

The Northern Subject Rule (A): Every agreement verb takes the -s form, except when it is directly adjacent to one of the personal pronouns I, we, you or they as its subject.

However, this is only a somewhat idealized statement, describing a hypothetical, pure northern agreement system. Indeed, varieties that come reasonably close to this are attested,

6 for instance in Older Scots and northern Middle English (Murray 1873: 211; Montgomery 1994; Mclntosh/Samuels/Bensik 1986). In these older varieties, verbal -s in fact occurred with near-categorical regularity in almost all environments where this rule licensed it. In Modern English varieties, however, the state of affairs seems to be always a variable one. It has therefore been customary in the literature since Montgomery (1994) to describe the variation patterns observed in terms of two separate constraints, the first of them most often called the Type-of-Subject Constraint, the second variously Position-of-Subject Constraint, Sequence Constraint, Proximity-to-Subject Constraint or similarly. (2)

The Northern Subject Rule (B): a. All third singular subjects (and, where preserved, the old second singular thou) always take verbal -s. b. Type-of-Subject Constraint: All other subjects except the personal pronouns /, we, you, they (and, where it exists, youse) take verbal -s variably. c. Position-of-Subject Constraint: Non-adjacency of subject and verb favours verbal -.s.

In a sense, the variable state of affairs in these modern varieties can be described as a state of competition between two conflicting grammatical systems, each of which consists only of categorical rules: the 'ideal' NSR system described as in (1), and the Standard English agreement system, which can be stated as in (3): (3)

The Standard English Agreement Rule: All third singular subjects take verbal -s, all other subjects do not.

The environments where variation exists according to rule (2) are just those where the two ideal systems are in conflict, most notably those with nonpronominal plural subjects. Third person singular subjects on the one hand, and the constructions involving adjacent I, we, you, and they on the other hand, take the same forms in both systems and therefore show little or no variation in the modern varieties. The notion of two competing rule systems, each of which is in itself categorical, so that observed variation can be thought of as a question of choice between the two "grammars", is a convenient descriptive abstraction and will be used as such throughout the present work. However, this ought not to be taken to mean that speakers' linguistic competence should necessarily be thought of as comprised of two distinct, independently represented rule systems, so that observed variation would be an outcome of essentially an act of codeswitching. To the contrary, I shall argue that, since the choice of forms is itself heavily conditioned by grammatical factors, the competition process that leads to variable output can best be thought of as an integral part of one unified but heterogeneous system of grammatical representation and processing, and that it is therefore a matter of competition not between grammars, but between constructions (or "variable rules") within one grammar. Before proceeding to a more detailed discussion of possible theoretical accounts of the phenomenon of the NSR, and of probabilistic grammatical variation in general, a few more descriptive observations on the phenomenon to be analysed are in order.

7 1.2.1. Type of subject, position of subject The definition of the NSR given under (2) above has left the exact relation between the component rules (2b) and (2c) deliberately unspecified. In principle, the Position-of-Subject Constraint may apply to all kinds of subjects. This means, on the one hand, that the prohibition of verbal -s with I, we, you and they may be overridden if the verb and the pronoun are not adjacent. With respect to the other environments, on the other hand, it means that there is commonly a quantitative effect further increasing the likelihood of verbal -s in varieties where it is variable. Both effects have indeed repeatedly been observed (e.g. Montgomery 1994, Robinson 1997: 127) and will also again be found in the present study, especially in the data from Northern Ireland. However, both aspects of the Position-of-Subject Constraint are empirically not unproblematic, and the exact nature of the rule is difficult to define. While some of its effects are indeed distinctly observable as salient characteristics of the northern dialects, some others appear rather to be cross-dialectal universals that can be observed even in varieties not directly affected by the Northern Subject Rule. First and foremost, it is important to note that non-adjacency of subjects occurs in a range of rather different syntactic configurations. Formal syntactic theories must be reluctant to ascribe agreement behaviour to conditions of pure linear surface order, as such a condition is not so easily captured in standard tree-structural accounts of syntactic constellations. But it is clearly conceivable that it is not linear order per se that causes the variable agreement behaviour, and that the behaviour of verbal -s in each of the different nonadjacency scenarios might be quite different. The descriptive generalization contained in the traditional statements of the Position-of-Subject Constraints would then be oversimplifying and misleading. It is therefore vital to check empirically whether all syntactic conditions of non-adjacency really favour verbal -s in the same way. As for pronoun subjects, obviously, direct adjacency of pronoun and verb is the canonical and by far most frequent constellation in Modern English, and this seems to have been the case at least since the Middle English period. A second important type of subject-verb configuration is that involving a postverbal subject. In Modern English - setting aside the special case of the notional subjects after existential there, which will be dealt with later, and some very minor other exceptional patterns - subject-verb inversion occurs only with the operator verbs BE, HA VE, and DO, and under the condition of direct adjacency. However, as is well known, it was much more freely possible also with main verbs in older stages. Postverbal placement of subjects has quite different effects with respect to pronouns and full NP subjects. As far as I am aware, postverbal pronoun subjects have always taken the same non-affixed verb forms as the preverbal ones, in all NSR varieties on record, and the formulations involving "direct adjacency" (and not "direct precedence") found in most of the descriptive literature correctly reflect this fact. On the other hand, inversion involving full NP subjects is apparently among the factors that favours verbal -s. The classic case that is most often cited in the literature to illustrate the Position-ofSubject Constraint is the one where a subject pronoun is followed by two conjoined verb phrases. Here, the second verb is obviously non-adjacent to the subject, and a typical effect of the NSR is indeed for this second verb to take -s, as in they sing and dances. This pattern is fairly regular in older varieties, and still found in some of the modern dialects (e.g. the Northern Irish varieties represented in the NITCS; on the other hand, Shorrocks 1999: 112

8 does not confirm its existence in modern dialects of south Lancashire). Of all the positionof-subject effects discussed here, this is probably the one that is most distinctively characteristic of the traditional dialects of the north. A second important type of non-adjacency occurs in subject relative clauses, i.e. clauses in which the relativized element is the subject of the embedded clause. In Standard English, the verb of such a clause agrees with the head noun, which is outside the relative clause and usually separated from its verb by at least the relative element. (As a matter of terminological convenience, I will use the term 'subject' to refer to the clause-external head NP, not the relativizer itself.) Modern northern dialects often also have zero subject relatives, i.e. where the relativizer is omitted. Here the head noun and the verb can be adjacent on the surface, but they are arguably still divided structurally by a clause boundary. As will be shown in the empirical parts of this study, subject relatives of either type are among the environments that most strongly favour verbal -s, apparently in all of the NSR dialects studied here. This suggests that the factor of "non-adjacency", if at all relevant to this case, may have to be described on a somewhat abstract structural level rather than in terms of mere surface wordorder. It should be noted that a relative clause constellation may occasionally license verbal -s even when the 'subject' head NP is one of the personal pronouns which otherwise block it categorically. This is true even though cases where this rule can potentially apply are not particularly frequent. Relative clauses with pronominal heads appear most typically in cleft constructions. Like other relative clauses, cleft constructions are an environment very much favourable to the use of verbal -5. But, as the examples in (4) show, many modem dialects avoid the use of the simple personal pronouns as heads of such relatives, employing instead strong or demonstrative pronouns, often identical to object case pronouns (compare e.g. Henry 1995: 24; Shorrocks 1999: 78ff.). With these strong pronouns, verbal -s is licensed by the Type-of-Subject Constraint anyway, and its appearance need therefore not be imputed to the Position-of-Subject Constraint as such. (4)

a. It's me that goes b. It's you that goes c. It's them that goes

It must be noted, however, that this restriction did not always apply in older stages of the language. Example (5) below, from one of the earliest Middle English texts that display the NSR, shows verbal -s in relatives both after the plural demonstrative pronoun pa and after the simple personal pronoun pai ( = they). Therefore, within a descriptive account of the NSR that intends to cover all its historical forms, the constellation of personal pronoun and subject relative must be taken into account as a genuine case in which the Position-ofSubject Constraint can potentially override the Type-of-Subject Constraint. (5)

pa [>at heyes [>am, J)ai sal be mekid; & pai |Mt mekes [jam sal be heyed. [Kock (1902: 11), quoted from Helsinki Corpus of English Texts],

A further case in point is that of conjunct subject noun phrases including pronouns, as in (6): (6)

You and me is going

9 In cases like this, just as with other conjunct subject noun phrases, a tendency not to observe agreement and to treat the conjoined phrase as singular irrespective of the personnumber features of the two components seems to be near-universal in English (Visser 1963: 80), and not restricted to the north. Again, many modern dialects have the tendency not to use the simple personal pronouns in these positions in the first place, but to replace them with strong pronoun forms such as me. Compare, however, the Early Modern English example in (7), quoted after Visser (ibid.): (7)

1 and my company was arrested ij days at Dunckyrke.

It shows that simple personal pronouns in these types of environments were not systematically excluded in all stages of the language. The examples given so far, where the conjoined NP is conceptually clearly plural, ought to be distinguished from others where two conjoined singular NPs can take a singular verb form even in Standard English because they are conceptually interpreted as a singular, as in (8) (cf. Quirk et al. 1985: §§10.37ff„ 13.59). Structures of this type will remain out of consideration in the present study. (8)

With all the turmoil and trouble that 's all over the north of Ireland [NITCS L9.3:JM23]

The empirically most problematic case of non-adjacency is found in those constellations where a pronoun subject is divided from its verb only by an adverbial or other similar element. My formulation of the Position-of-Subject Constraint given above follows those usually found in the descriptive literature, implying that these cases allow for the Positionof-Subject Constraint to override the Type-of-Subject Constraint just as in the other types of environments discussed so far (cf. Montgomery 1994: 89). However, the existence of verbal -s in these environments seems to be a rather elusive phenomenon in many of the modern dialects. In the data analysed later in this study, it is decidedly rare. Only occasional examples such as (9a) were found in the Northern Irish data of the NITCS (Chapter 5.2). As for older varieties, Murray (1873: 21 If.) attests structures such as (9b) in 19th century spoken Scots. (9)

a. You never was up that Fivemiletown Road? [NITCS L24.3:CM115] b. Huz tweae quheyles gangs theare

What makes the phenomenon so problematic is the seemingly rather more common appearance of such structures in the dialects of Yorkshire and Lancashire (lOa-b). (10) a. 1 never says it [SED Y9: III.1.16] b. I oft pulls his leg [SED Y16: 1.2.17]

The problem here is that these examples may be due not to the NSR proper, but rather to quite another rule competing and overlapping with it. In parts of Northern England, nonstandard verbal -s is reported to be used as a marker of habitual aspect independently of the NSR, that is, even with adjacent pronoun subjects (Shorrocks 1999: 112, 116f.). There is some evidence that a similar condition favouring verbal -s in clauses of habitual or durative semantics may be operative in parts of Northern Ireland too (Chapter 5.3 below, cf. also Montgomery/Fuller 1996: 221). Now, many of the adverbs that can typically intervene

10 between subjects and verbs in Modern English are of just the kind that denote temporal meanings of habitualness: always, often, sometimes, never being among the most frequent ones. It is therefore hard if not impossible to distinguish whether the use of verbal -s in such examples is triggered by the syntactic non-adjacency of the subject, or by the collocation with these particular adverbs, or by the temporal semantics of the clause as such. The relation between these factors is one question that the present study will not be able to resolve, and more research clearly needs to be done on this.1 It should also not be forgotten that in Modern English there are different classes of adverbials whose positioning behaviour between the subject and verb differs slightly but significantly. Under certain conditions adverbs tend to be placed before the first auxiliary of the verb phrase, hence always causing non-adjacency between subject and inflected verb, while under other conditions the preferred position is after any first auxiliary or after the copula, hence causing non-adjacency only when the verb phrase consists of only a main verb (Quirk et al. 1985: §8.18). There is some evidence suggesting that even these different classes of constructions may behave differently with regard to agreement (cf. Section 2.1.1 on page 16). Another related case, which also deserves some separate consideration, is that of nonadjacency being caused by intervening quantifiers such as all or both, as in we both go(es). The surface distributional behaviour of these items in present-day English is quite similar to that of adverbs (cf. Quirk et al. 1985: §3.28), and some data suggest that they share with adverbs the property of triggering a Position-of-Subject effect that can potentially override the Type-of-Subject Constraint: (11) a. they both has one [SED Dbl: IV.12.2] b. they both was yoked onto it [NITCS L19:JB83] c. we all hes mater to thank God [Older Scots, Montgomery 1994: 89] d. yet they all professes that they are not able to give [Older Scots, Montgomery 1997b: 136]

It is descriptively not inconceivable to simply lump these constructions together with the corresponding adverb constructions (which is what I will do throughout the empirical descriptions in the later chapters, if for no other reason than lack of sufficient data to effectively differentiate the two). However, current formal syntactic discussions of English usually treat them as structurally rather different. In recent generative work these items are commonly discussed as "floating quantifiers", i.e. elements generated in an underlying subject position within the VP, which are then left behind in one of the intermediate Spec positions through which the subject raises on its way to its surface Spec-IP position (Haegemann/Gueron 1999: 230). To the best of my knowledge, any theoretical consequences that the existence of these constructions would have for a formal analysis of the NSR have not until now been discussed in the literature. 1

Tentatively, it might be suggested that the two may actually be causally/historically related. I am not aware of any historical explanation of the origins of the non-NSR conforming verbal -s occurrences in Northern English, nor of the habitual constraint which conditions them according to Shorrocks (1999: 116). It could very well be the case that it is an extension of the NSR pattern. Since the pattern pronoun + adverb + verb, with verbal -5 originally triggered by the NSR, would so often have coincided with habitual semantics, the verbal -s could have been reanalysed in these collocations as a marker of the temporal semantics, and its use extended from there into the pronoun-adjacent environments where it would not originally have been licensed by the NSR.

11 There are some further, minor types of configurations where subject pronouns are not adjacent to their verbs: for instance, intervening parenthetical material, pauses, or asyndetically conjoined verb phrases. The latter resemble true subjectless sentences in which there is ellipsis of the subject pronoun, as in (12), and where verbal -5 is also observed: (12) They take, eh, the 1-, füll load with them. Lifts ones and takes them. [NITCS L24.3:CM70]

Apart from elliptical sentences, which are of course fairly common in spoken language, another much more restricted idiomatic type of subjectless clause occurred in first person singular epistolary formulae in formal letter writing in Early Modern English and Older Scots, as in (13), quoted after Meurman-Solin (1993: 248f.; cf. also Montgomery 1994: 89): (13) and nocht hauing farder for this present, bot to wish yiow all happiness, restis your lordships maist affectionat to serue yiow.

Of course, in older stages of the language, especially in Middle English, yet other types of syntactic configurations occurred where verbs were not adjacent to their pronominal subjects (e.g. intervening object pronouns, the negation particle ne, etc.) To my knowledge there has not been any detailed study of their effects on subject-verb agreement; and since the older stages will not form the object of the empirical investigation in the present study, they will not be further discussed here. However, given the obvious historical continuity of the NSR phenomenon from Middle English into the modem vernaculars, it would certainly be desirable for a full descriptive and explanative account of the NSR to cover it in all its historical and geographical extension, and hence to be able in principle to account for these cases too. Summing up, it may be said that whereas the Type-of-Subject Constraint has been a stable and fairly unmistakeable feature of northern dialects, the Position-of-Subject Constraint may be regarded as the composite effect of several different patterns, of varying degrees of regularity, only some of which are specifically characteristic of the northern dialects whereas others are shared with many varieties elsewhere. It is only in those older dialects which were least affected by standardizing influences from the south that the Position-ofSubject Constraint can be regarded as a single, unified, tightly integrated feature of a consistent grammatical system. As for the second aspect of the Position-of-Subject Constraint, its role in quantitatively favouring verbal -s with full NP subjects according to their position (i.e. in cases where it does not conflict with the Type-of-Subject Constraint), the main empirical problem is the definition of 'distance'. The most straightforward formal definition would of course be based on the occurrence or non-occurrence of words or constituents between the right edge of the subject noun phrase and the finite verb. However, it has been observed that besides such intervening constituents external to the subject NP, the occurrence of internal postmodification within the subject NP itself can have a quite similar effect of favouring subsequent verbal -s. The effects of this "Heaviness-of-Subject Constraint" (cf. Bailey et al. 1989), which may be described either as a sub-feature of the Position-of-Subject Constraint or as an independent constraint complementing it, will also be confirmed in the corpus material analysed in the present study. It seems possible that the two effects can be reduced to a single principle, where 'distance' is defined not (or not exclusively) in terms of adjacency of nodes in a tree structure but rather in terms of surface distance between the verb and the head of the subject NP.

12

1.2.2. Type of verb Having discussed some of the issues regarding the definition of the Position-of-Subject Constraint, we can turn to another area that is descriptively problematic about the NSR in general: the question of what verbs it applies to. The problem here lies mainly in the forms of BE. As is well known, BE is morphologically exceptional in Standard English insofar as it preserves more person-number contrasts than other verbs do: three as opposed to two in the present tense (or four as opposed to three if one also counts the old second person singular), and a contrast of two forms as opposed to a single form in the past tense. Moreover, even the contrast of the two forms of BE in the past tense patterns differently from the contrast of the two forms of other verbs in the present tense, given that was is used not only in the third singular but also in the first.2 The NSR historically originated in the lexical verbs and only later spread also to BE. What is important to note here is that in the lexical verbs, the non-standard plural -s forms are not, as one might expect, an innovation in the NSR system, and not a takeover from the singular, but rather a retention of an earlier, genuine plural affix. In contrast to this, the -s forms of BE (is and was) are indeed purely singular forms in origin. However, all Modern English dialects that are affected by the NSR have transferred the NSR pattern analogically to BE, so that is and was are now also used in the plural. In the most straightforward form of this system, the forms of BE after pronouns are used just as in Standard English, while in all environments where -5 is licensed with other verbs, analogical is and was can be found too. This seems to be the dominant system in most of Scotland and Ulster. In Northern England, however, different systems are found; paradigms of BE even after pronoun subjects differ from the Standard (e.g. North Yorkshire I is, or Southwest Yorkshire he were) and variable agreement is partly conditioned by factors independent of the NSR, such as negation in the case of was and were. (cf. Chapter 4.3). Montgomery (1994), following Murray (1873), gives a timeline for the development of BE agreement as opposed to lexical verb agreement in Scots. He shows that the analogy never affected BE totally. Whereas the NSR system was observed near-categorically for lexical verbs until the beginnings of anglicization in the mid 17th century, the use of is and was in the plural never surpassed a rate of approximately one third of all tokens, and was also the first to decline after the 17th century. On the other hand, some modern dialects appear to have taken a reverse path in the latter respect: here, dialect levelling seems to have affected the NSR system for lexical verbs more strongly, whereas patterns of was/were variation partly reflecting the older NSR system are still frequent (e.g. Tagliamonte 1998). In the present study it will be shown that a similar effect, a tendency to preserve the NSR pattern more strongly for was (and possibly for is) than for other verbs, can be seen in many parts of the survey area.

2

For the purpose of describing verbal paradigms it is convenient to treat you as the plural form throughout, in accordance with its historical origin. In that perspective, the was/were contrast is a true singular vs. plural opposition.

13 1.2.3. Other patterns of agreement variation In many of the varieties studied, the NSR system co-exists with other patterns of nonstandard agreement variation that are partly independent of it. Two of them, the Northern English habitual -s, and Northern English BE variation, have already been mentioned. Another is the usage reported from many modern varieties involving verbal -s in narratives in the so-called historic present. Attestations in northern varieties originally belonging to the NSR area can be found in Shorrocks (1999) for the area of Bolton (Lancashire), Henry (1995) for Belfast, Macafee (1983) and Hagan (2002) for Glasgow. This pattern is also common in other parts of Britain. Another area of variation, which however will not receive special attention in the present work, is the variation between singular and plural verb forms which exists even in Standard English as a result of conceptually different interpretations of certain subject NPs, for instance with group nouns (the police, the government). This phenomenon is arguably of quite a different kind than the NSR, as it is apparently governed semantically, not syntactically. Yet another type of variation that is arguably unrelated to the NSR proper is processing-induced non-agreement, of the kind that has been called number shifting, number attraction, or proximity agreement. (14) a. A good knowledge of English, Russian, and French are required for this position [Quirk et al. 1985: §10:35] b. The people who Clark think are in the garden [Kimball/Aissen 1971]

See Visser (1963: 62-96) for a historical survey of some such phenomena; Quirk et al. (1985: §10:35-45) for a descriptive account of them in present-day Standard English; Kayne (1995) for a generative analysis of structures like (14b). Probably the most prominent and widespread agreement variation phenomenon occurs after existential there, and this does require some special consideration in the course of this work. As is well known, existential clauses in formal Standard English require the verb to agree in number not with the syntactic dummy subject there, but with the so-called notional subject, the NP that follows it. This constitutes a systemic anomaly, as the morphosyntactic properties of subjects are divided between two constituents. The dummy subject there acts as the subject of the clause as far as word order is concerned, but the following NP acts as the subject in so far as it controls agreement. Many varieties of English share the tendency to level out this irregularity, by allowing invariant singular verb forms after there. Thus the notional subject is finally deprived of the last remnants of its subject status. Historically this can be seen as part of a long-term trend of grammaticalization of the there construction, in which there has gradually changed its status from being originally a deictic adverb to being a subject (cf. Breivik/Swan 2000; pace Van Gelderen 1997: 88-109). The phenomenon of non-agreement in existentials has sometimes been conflated with that of NSR effects in other plural environments in statistical analyses (Policansky 1981, Kallen 1991), and sometimes it has implicitly been used even as another indicator characteristic of the Scottish/Northern origins of a variety (e.g. Montgomery 1997b). However, this form of non-agreement has been a structural option in all forms of English since the Old English period (Visser 1963: 62; Breivik/Swan 2000). Moreover, it is also a feature almost universally present in colloquial forms of present-day English (see Rupp fc. for a

14 recent survey of both variationist and theoretical studies on the issue). Particularly with respect to the Northern Irish data analysed in Chapter 5,1 will argue that (non-)agreement in existentials and (non-)agreement according to the NSR are two phenomena which are best kept separate, because they have been following quite distinct paths of historical change, they have different sociolinguistic status in the dialects in question (cf. also Wilson/Henry 1999: 12), and their distribution is governed by different sets of linguistic constraints.

1.2.4. The 'markedness paradox' After this global descriptive survey of the NSR and related phenomena, it will be useful to return briefly to one of its more general characteristic properties, which has repeatedly led to a certain amount of confusion and contradictory approaches in the literature. I will call this feature of the NSR the markedness paradox. It is a common practice to describe forms within a morphological paradigm in terms of their being marked or unmarked. As will be demonstrated later, all existing attempts at formal analyses of NSR systems, in one form or other, implicitly operate with this concept, even when they do not explicitly use the term. However, in the case of the NSR, which of the two forms (-s, - 0 ) is the marked and which is the unmarked member of the paradigm? In Standard English, the -s form unambiguously encodes one particular person-number combination, the third singular, whereas the - 0 form acts as the default for all other persons and numbers. It is therefore fairly unproblematic to characterize the -s suffix as an agreement morpheme in the normal sense, and hence the -s form as the marked member of the paradigm. This is a straightforward analysis, even though in a cross-linguistic comparison the system may be a rather uncommon one - the third singular usually being the most likely form for zero marking within agreement paradigms (Bybee 1985). However, in the NSR system matters are rather different. In pronominal environments the system works just as in the Standard, encoding the same grammatical information with the same means. But in all other environments it is the generalized presence of the -s form that leads to a complete neutralization of all person-number agreement contrasts. It can therefore be argued that the -5 form, though formally carrying the overt agreement morpheme, acts as a functionally featureless form devoid of person-number information. The -s morpheme can then be characterized as a mere tense-mood marker. In contrast to this, the presence of the formally featureless - 0 form, where it occurs, serves to uphold agreement oppositions, particularly that between singular and plural in the third person. The - 0 form is therefore usually regarded as the genuine 'plural' form, and hence as the one that functionally does carry genuine person-number agreement features. Given this somewhat paradoxical distribution of formal and functional markedness in the modern NSR system, it must be added that quite a different state of affairs held in its late Old English predecessor system. Here, at one stage, it must be assumed that the functional status of these forms was still exactly the reverse. Diachronically one may therefore speak of a process of 'markedness reversal'. As was mentioned earlier, the -s affix in the plural is historically not a takeover from the singular, as might be expected judging from the modern systems. Rather, the -s is etymologically just as genuine a reflex of a real person-number agreement marker in the plural (northern OE -ad/-as) as in the singular (northern OE

15

-ed/-es). The - 0 forms, by contrast, are originally a product of agreement loss, apparently through phonological erosion of the agreement morphology. They took on their new functional load as carriers of agreement information only when the two formerly distinct endings -ed/-es and -ad/-as happened to fall together. Thus the neutralized forms were reanalysed as default 'singular' forms, and the new affixless forms were reanalysed as 'plural' agreement forms. These historical developments will be discussed more fully in Chapter 3. The markedness paradox will be encountered repeatedly as a problem that has haunted previous attempts at descriptive and explanative accounts of the NSR. Several such accounts of a synchronic-theoretical nature will be discussed in the next chapter, and similar problems with one diachronic proposal will be discussed in Chapter 3.4.

2. Theoretical accounts of verbal agreement variation

There have been several attempts to describe the NSR within formal theories of syntactic competence, mostly in frameworks of the generative tradition. However, all of these proposals have difficulties in addressing the full range of distributional facts of the NSR. In the present study I will employ a different theoretical approach, derived from functionalist and usage-based strands of research. This chapter will serve to motivate this choice, and to demonstrate some of its consequences for the synchronic description of the NSR, for possible diachronic explanations of its development, and for the description of the quantifiable, probabilistic variation observed in its use. The chapter will be organized as follows. Section 2.1 presents four different formal analyses of NSR systems, within different formalist theoretical frameworks. Section 2.2 discusses theoretical approaches to the description of probabilistic variation in general. Finally, Section 2.3 presents a short sketch of a possible usage-based approach that unites a natural description of the NSR system with an account of its variability.

2.1.

P r o p o s e d f o r m a l a n a l y s e s of t h e N S R

Besides some very brief and inconclusive discussion in Roberts (1993) and Van Gelderen (1997), I am aware of three more elaborate proposals of formal analyses of the NSR phenomenon: Henry (1995), Corrigan (1997), and Börjars/Chapman (1998). All three follow the tradition of generative grammar, in so far as they try to deduce the observed behaviour of a construction from the postulated properties of a small set of atomic, universal elements, and their configuration in a universal system of possible tree structures. A somewhat different approach to the NSR, also within a rigidly formalized framework but with a radically different stance regarding linguistic universals, has been outlined in a brief sketch in Hudson (1999). However, all four of these formal proposals are burdened with problems on the level of descriptive adequacy, and none of them even attempts to account for the most crucial feature of the reflexes of the NSR in the modern dialects: its inherent and structured variability. The four proposals will be reviewed here in turn.

2.1.1.

Henry (1995)

Henry (1995) is a description of non-standard patterns in the syntax of contemporary Belfast English, based on a late Principles-and-Parameters version of generative grammar. As such, it is an example of the type of variation studies that has come to be known as 'microparametric syntax' in generative work (cf. papers in Black/Motapayane 1996 and Barbiers/Cornips/van der Kleij 2002). Henry's data are based on acceptability judgments elicited in interviews with Belfast informants. While Henry devotes extensive discussion to proposed parametrical links between the phenomenon of optional plural verbal -s and some

17 other non-standard grammatical phenomena in this dialect, she makes no attempt at integrating her analysis of the narrowly local variety of Belfast with a historical or larger geographical account, with respect to the continuity of the phenomena in question across a wider range of varieties. As to her methodology, it must be noted as unfortunate that Henry fails to give any detailed account of the number and choice of informants, her methods of sampling, or the set of test sentences used. Neither does she give any detailed quantitative or qualitative account of how her informants reacted to each test sentence. Henry's description of the NSR effects ("singular concord", in her terminology) differs in some respects from those found elsewhere in the dialectological literature, and it is sometimes difficult to judge whether these apparent differences reflect genuine divergence between dialects or whether they are merely due to diverging research methodologies. First, Henry does not deal with the Position-of-Subject Constraint as usually stated in the descriptive literature. Neither relative clauses (structures of the type they that goes), nor the 'classical' northern pattern they sing and dances are mentioned in her analysis; nor does she say anything about the effect of intervening quantifiers or adverbs with pronominal subjects, as in they both goes, they often goes (the latter structures admittedly being rare in Northern Irish English but by no means non-existent, at least in the traditional dialects surrounding Belfast). The most striking descriptive discrepancy between Henry's account and those found of related varieties in the literature elsewhere (and also the findings of the present study) regards subject-verb inversion. Henry states that plural verbal -s is strictly prohibited under inversion in Belfast English (1995: 16ff.). Her example sentences are of the type in (15). (15) *does the children sing?

This constraint is not confirmed in the data of the present study (see page 86 and 110 below). Sentences with subject-verb inversion are regularly attested in conservative varieties of the northern type both in Ulster and in Britain, and they were even found to be among the environments most favourable to verbal -s. It may, however, be the case that wh- questions are somewhat more frequent among these than yes-no questions. While the overall textual frequency of potential application environments of this type in the corpora is quite low - the bulk of all question clauses actually produced contain either pronoun subjects, or modal non-agreement verbs, and thus fall outside the scope of the rule - the relative frequency of verbal -5 in the few tokens produced is extremely high. In this respect, the present study agrees with findings by Harris (1993: 156), who, speaking for Ireland in general, sees inverted clauses as even "the most favourable context" for verbal -s. Börjars/Chapman (1998: 86) remark on the clash in the findings of both authors, and are inclined to attribute them to genuine differences between varieties. However, given the poor documentation, one cannot exclude the possibility that the apparent difference may also be an artefact of divergences between informants' actual discourse behaviour and their grammatical intuitions targeted by acceptability questions (cf. also Filppula 1999: 152). A second interesting point on which Henry's account differs from those in the dialectological literature is an effect that seems to act in exactly the opposite direction of what is elsewhere described as the Position-of-Subject Constraint. Henry (1995: 19,26) reports that with adverbs intervening between a full-NP subject and an operator verb ("raising verb" in her terminology, that is, a verb that raises to INFL before 'spell-out'), verbal -s is reported not to be licensed. This means that contrary to the common pattern observed elsewhere,

18 non-adjacency of subject and verb in this case would favour the standard agreement pattern. Thus: (1)

a. The children really are late. b. The children is late. c. *The children really is late.

Henry explains that "it seems to be the case that the adverb position between the subject and the topmost projection of INFL which exists in English is unavailable in singular concord" (1995: 19). She does not state explicitly whether the same kind of environment has any influence on agreement with personal-pronoun subjects (they really !isPare late), but from her statement of the rule it would seem that a fortiori the -s forms should be ruled out in that case too. In this instance, the data of the present study suggest that Henry's observation may in fact agree with a tendency observable in other related varieties. The adverb position Henry is speaking of here is the one which Quirk et al. (1985: §8.18) define as "initial-medial": the subtype of medial position adverbials that go between the subject and the first 'operator' verb if one exists - instead of after the first operator, as most medial adverbials do. Quirk et al. note that this position in Standard English tends to be reserved for particular semantic-pragmatic environments, among them those where the adverb is to be excluded from the scope of a following negation, and those where the following verb carries prosodic focus. The latter type, according to Quirk et al., is particularly common with the verb BE, as in the examples in (2): (2)

a. Well, you never WERE fond of work! b. That's where she usually IS at this time of day.

The sentences Henry uses appear to be examples of the latter type. Given the special pragmatic and prosodic properties of this type of sentence, it does not become entirely clear from Henry's discussion whether her informants' reluctance to accept verbal -s in these positions really was caused by some property of the phrase structure as such, or by other factors such as stress conditions on the verb form, or even the mere rareness of the construction type. A comparison of her observations with some data from the NITCS confirms the possibility that verbal -s in some of these environments may indeed be rare or avoided. The NITCS records several instances of adverb plus operator verb, and those that most closely correspond to the type Henry describes almost all have the non-.j form (3): (3)

a. People just don't have time [L23.3:MK45] b. And those horses never were home from February until November [L7.2:TM80] c. Although houses certainly are being pulled down [L27.2:NI8] d. Some st-, a few still do [L32.2:MM26]

Counterexamples are given in (4): however, one of them contains a self-correction from the -s form to the non-i form (4a), in another the adverb may be interpreted as a postmodifier of the subject NP rather than as an adverbial (4b), and the same might arguably be the case for the third example too (4c). (4)

a. The night that the rehearsals actually was held, were held, that, mmh... [L27.2:NI6] b. The menus then was usually... Well, 1 had done the menus. [L4.3:ML12]

19 c. You got used to walking on the stones. The roads even wasn't as good then either. Not at all. [L36.3:C053]

Another finding that also seems to fit in with Henry's observation is that a primary verb which arguably is not in operator position, namely HA VE wherever it has full verb status, clearly does allow verbal -s with preceding adverbs, as is shown by the examples from Northern Irish and English NSR varieties, from the NITCS and SED, in (5). Unfortunately, the set of examples given by Henry is too limited to allow for any far-reaching structural conclusions. (5)

a. Most people now has these big hay sheds [NITCS: L21.2:JM21] b. Some people always has a bit of bacon [SED: Lei2: V.7.4] c. 1 oftens has to work [SED: Y11: 1.3.1] d. They both has one [SED: Dbl: IV.12.2]

As far as the Type-of-Subject Constraint is concerned, Henry states that 'singular concord' (i.e. non-standard use of verbal -i) is only possible with subjects that are not openly marked as nominative (1995: 23f.). This generalization covers the prohibition of verbal -s with the simple personal pronouns I, we, and they. The pronoun you must be treated as an exception, as it does not have an overt case distinction between subject and object case but nevertheless patterns together with the other three with respect to the NSR. Henry takes as a strong piece of evidence for her claim that, according to her observations, the prohibition of verbal -s with these items does not apply only to the standalone pronouns but also to coordinated NPs involving any of them. The strong forms of the pronouns, which are identical to the accusative forms, are not subject to such a constraint. Co-ordinated NPs involving these can freely take 'singular concord': (6)

a. You and 1 are going b. You and me are going c. *You and I is going d. You and me is going

Henry reports that her informants found sentences of type (6c) "completely ungrammatical" to a degree which, as she argues, excludes the explanation that the ungrammaticality judgment could be due only to a "sociolinguistic mismatch - with singular concord being a nonstandard feature, and thus sounding strange when placed together with the formal, prestige form involving co-ordinate pronouns in the nominative case" (1995: 24). Consequently, Henry makes the 'nominative constraint' the focus of her analysis. In the Principles-and-Parameters framework adopted by Henry, agreement phenomena are assumed universally to involve 'checking' in a specifier-head configuration between the two elements involved. Therefore, the question of agreement or non-agreement logically leads to the question "whether the subject is in fact in SPEC/AGR s P in this structure" (1995:21). Henry discusses, and rejects, an analysis according to which the subject in 'singular concord' sentences (those with non-standard -s) would remain in VP, while in those sentences that take standard agreement it would raise to SPEC/AGR s P. Her own proposal involves a different set of nodes in the tree: while verbs and subjects of standard agreement clauses raise to AGR S and SPEC/AGR s P respectively, just as in Standard English, verbs and sub-

20 jects of 'singular concord' clauses raise only to one node lower in the tree, Τ ('tense') and SPEC/TP respectively. This reflects the intuition that the generalized verbal -5 of the 'singular concord' clauses is a mere tense marker, and that only tense features but no agreement features are involved in this position. Following common assumptions in generative work on English, the movement of the verb is supposed to happen overtly (at 'spell-out') only with the 'raising verbs' BE and HAVE, while all other verbs undergo the raising operation only at the level of LF. As for the subject, the Τ node is also assumed to be unable to assign nominative case to the subject in its specifier position, as this is an exclusive property of AGR S . However, Henry stipulates that Τ in this dialect has the property of being able to assign 'non-nominative' or 'default' case. Therefore, all subjects except those morphologically marked for nominative can appear in this position, and this explains the 'nominative constraint'. The question of why and when the elements are forced to raise to either of the two target positions is handled, following the preferred mode of thought in the Minimalist paradigm, by assuming that features are either 'weak' or 'strong'. In this case, it is stipulated that the NP features of AGRs can optionally be either weak or strong in the dialect, whereas in Standard English they are always strong. A central empirical argument Henry adduces for her analysis (1995: 29) is the behaviour of negative-polarity items (NPIs) in this dialect, which she claims to be parametrically linked to NSR behaviour. She finds that non-standard structures such as (7) are possible in Belfast English: (7)

a. Anybody wouldn't be able to do it. b. I was surprised that anybody didn't go.

However, such negative-polarity subjects can only occur if there also is 'singular concord': (8)

a. Any friends isn't coming. b. *Any friends aren't coming.

Following the generative tradition, Henry explains the licensing of negative-polarity items in terms of their being c-commanded by a raised negative operator on the level of LF. Her point is that for theory-internal reasons raising of the negative element above the negativepolarity subject is easily conceivable if the subject is in SPEC/TP, but not so easily if it is in SPEC/AGRsP. Again Henry seems to exclude the possibility that the co-occurrence restriction between singular concord and negative-polarity subjects could be due simply to stylistic mismatches of elements from different registers: That there is a link between singular concord and NPI licensing in subject position seems clear from the the [sic] fact that the two phenomena seem to go together in speakers' grammars; those speakers who allow singular concord also permit NPIs in subject position, and conversely nonusers of singular concord find NPIs in that position strongly ungrammatical. This is a clear case where careful examination of dialects or sub-dialects can help to show whether proposed connections are real or not, and where it is important to check what co-occurrence constraints there are on dialect features. (1995: 29f.)

Henry's proposal is probably the most comprehensive and empirically best argued analysis of NSR-related phenomena in a formal framework so far. Nevertheless, it must be noted that it does not provide a unified analysis of NSR systems. The presumed parametrical link

21 between the NSR and the NPI behaviour is just such a case where more "careful examination of dialects or sub-dialects" would have been a welcome complementation. The problem is that NSR phenomena are also common in varieties which, to all appearances, do not agree with Belfast English with respect to the NPIs. A similar and even stronger limitation applies to Henry's account of the 'nominative constraint'. Even if her rule is descriptively adequate for the Type-of-Subject Constraint data found in the particular variety she studies, there does not seem to be a straightforward way how it should carry over to varieties where the Position-of-Subject Constraint played a more prominent role. Such varieties, after all, do allow overtly nominative pronouns to co-occur with verbal -s, if only in non-adjacent positions. Given the central role of the 'nominative constraint' in Henry's analysis, it is difficult to see how the analysis could be adapted in such a way as to account for those other varieties. This, of course, does not logically invalidate Henry's analysis as such. It might in principle very well be the case that present-day Belfast English differs fundamentally from NSR systems in other (or older) dialects. In that case, Henry's analysis of the former would stand but explanations along rather different lines would have to be sought for the latter. However, an argument in the other direction holds. Any explanation that could account for the 'classic' older NSR systems with their stronger Position-of-Subject effects (such as Older Scots or late northern Middle English) would very likely be able to account also for the more restricted range of NSR phenomena in Belfast English in a unified way. Such a universal explanation would then certainly be preferable to the parochial one Henry proposes for Belfast English.

2.1.2.

Corrigan (1997)

It was shown above that a central idea in Henry's proposal is that the collocations of pronouns and affixless verb forms represent real agreement whereas the constructions with the -s form do not. This is entailed by the assumption that the former but not the latter involve checking of identically marked features of subject and verb in a specifier-head constellation. The two proposals to be discussed next, Corrigan (1997) and Börjars/Chapman (1998), both imply the exact opposite. Both of them interpret the co-occurrence of pronouns and affixless verbs as a kind of 'anti-agreement' (in the terms of Corrigan 1997: 203, quoting Roberts 1997: 109), that is, the avoidance of double marking of the same feature. In this perspective, then, it is the -s forms which represent the real agreement morphology, and the - 0 forms that represent the lack of such. Obviously, the existence of these two diametrically opposed approaches is a consequence of the 'markedness paradox' of the NSR as described earlier (cf. page 14). None of the authors involved seems to be aware of the possibility of the opposite view. Corrigan (1997) is a description of the English of South Armagh, written mostly from a historical and comparative perspective and based on recorded dialect material from the area. In theoretical terms it is couched in a Principles-and-Parameters approach similar to Henry's. In her chapter dealing with subject-verb agreement, Corrigan presents data regarding the distribution of was and were, showing that the traditional NSR system is preserved intact for these forms. (As for the other verbs, her corpus contains too few present tense forms to allow for a detailed analysis.) In her theoretical discussion she gives an overview

22 of the historical development and a comparison with other non-standard varieties in which variable agreement is found, but also adds an extensive discussion of similar patterns in Irish. Her thesis is that the preservation of the Type-of-Subject Constraint in Armagh English can be seen as an effect of "convergence" (1997: 219) between the Northern English/Scots superstrate and the Irish substrate, whereas some other aspects of the NSR, notably some conditions regarding the Position-of-Subject Constraint, can be understood as functionally motivated dialect universals. Both these suggestions are certainly valuable. As for the Irish hypothesis, one might prefer a term such as 'conservation' instead of 'convergence' (cf. also Filppula 1999: 158f.); after all, Corrigan is not claiming that anything substantial in the NSR system actually changed (i.e. 'converged' towards Irish) because of the contact. But the idea that a partial isomorphism between the English and the Irish systems might have helped to keep the NSR more robustly alive in Northern Irish English than elsewhere is prima facie plausible. The same is true for the idea that there may be functionally motivated universal tendencies for agreement to be affected under conditions such as non-adjacency or heaviness of subject. Unfortunately, in her formal analysis of the phenomena in question Corrigan's discussion displays some confusion, both in her presentation of the historical development in older stages of English and in her formal comparison with Irish. Having demonstrated convincingly that a tendency to treat full NP subjects differently from pronoun subjects has been common in many varieties of English, Corrigan draws the conclusion that the rule which makes affixless verb forms obligatory after pronouns such as / and they should in principle also hold for third singular pronouns (1997: 195). But in this regard, it seems that her interest in tendencies universal to all varieties of English leads her to pass rather too lightly over the clear geographical divisions that existed in Middle English and in the traditional dialects of Modern English. Her main evidence on this point is a study of the Cely letters (Bailey et al. 1989), an Early Modern English corpus consisting of the writings of a 15th century London-based merchant family. This study indeed finds that all three forms ( - 0 , -s and -th) were used variably in all environments including the third singular, and that the presence of full NP subjects favoured the affixed forms in all of them. However, the existence of affixless forms in the third singular in these idiolects is a feature that probably has little to do with the northern system in terms of historical developments. In the situation of dialect mixture characteristic of the London variety during the process of incipient standardization in Early Modern English, it is most likely a feature derived from East Anglia varieties. (Stein 1987: 428; for a critical re-appraisal of the Bailey et al. study, drawing attention to the fact that data from different members of the Cely family may mirror rather heterogeneous geographical sources, see also Montgomery/Fuller/DeMarse 1994: 354.) Corrigan places so much confidence in one example quoted from the Cely letters study (he lyke the passe! wyll) that she repeatedly includes the alternative form, he lykes, in lists of allegedly 'ungrammatical', asterisked forms (1997: 201, 203, 204). This, however, is a rather serious misreading of her source: Bailey et al. (1989: 290) state that each of three different verb forms, -th, -s, and - 0 , occurred with almost equal frequency with third singular pronouns in their corpus; thus there seem to be no grounds for saying that the -s form was ungrammatical in that environment. While it is certainly an interesting fact that a quantitative condition in some way parallel to the northern Type-of-Subject Constraint may have been active in the third singular in a non-northern variety, I am not aware of any documented case where this effect ever became categorical with respect to the third person sin-

23 Table 2.1. Distribution of verb forms in Irish and English Irish

Ν SR English

Std. English

3rd Ps. Sg.

pronoun singular NP

unmarked unmarked

-s -s

-s -i

Others

plural NP relatives zero subjects pronouns

unmarked unmarked marked unmarked

-s -s -s -0

-0 -0 -0 -0

gular in a way comparable to the other persons and numbers in the northern system. To confuse matters further, Corrigan then goes on to call upon Murray (1873: 211) as a witness for her wide statement of the rule, according to which pronoun subjects "have a propensity to mark adjacent verbs with { - 0 } even in the third person singular" (1997: 195) - only to find out a few pages later that her misreading of Murray's rule does not match the facts reported by the same author (1997: 205).' Despite these confusions, Corrigan does in the end argue on the basis of correct data that one particular formal analysis of the NSR system - treating the -5 suffixes as "incorporated pronouns" - is not tenable for the northern system, as it leads to wrong predictions regarding the third singular (1997: 204). Corrigan's comparison between the Irish and the northern English system again runs into problems due to the markedness paradox on the English side. Northern English has one affixed but (arguably) functionally personless verb form (-s), and one affixless but (arguably) functionally marked form (-0). The relevant constructions in Irish, on the other hand, involve a contrast between one unmarked, third singular verb form, and distinctive marked forms for other person-number combinations. In order to draw any comparison between the two systems, one first needs to make up one's mind which English form is to be compared with which of the Irish ones. Should the affixed English third singular be treated as equivalent of the unmarked Irish third singular, or of the marked Irish forms in the other persons and numbers? In Irish, plural noun phrase subjects, as well as relative clauses after first and second person subjects, take the same unmarked verb form as the third singular (cf. Table 2.1). Irish thus displays non-agreement after nonpronominal subjects in a way rather similar to northern English - if one regards the two 'third singular' verb forms as equivalent (as does, for 1

Corrigan's misunderstanding of Murray is hard to explain, as it is in clear contradiction to his formulation which she quotes verbatim on the same page: "aa leyke, wey leyke, yoo leyke, thay leyke, are used only when the verb is accompanied by its proper pronoun". In the later passage (1997: 205) she passes over the observed mismatch speculating that the alleged rule in the wide form she has ascribed to Murray "may well be correct for nineteenth century Scots" - but Murray is quite explicit in that he includes Older Scots in his description too. Corrigan then goes on to state her own version of the rule for Older Scots, but this time makes it too narrow: she thinks that the Type-of-Subject Constraint in Older Scots was valid only for "first person singular and third person plural" - but it did indeed also hold for the first and second Plural, as not only Murray but also Macafee (1993) and Montgomery (1994) clearly show.

24 instance, Filppula 1999: 158f.). Seen this way, one finds basically the same parallel in Irish which also holds in Welsh. 2 On the other hand, Irish has two different functional equivalents of the English verb-plus-pronoun construction: in some parts of the verbal paradigm Irish employs an analytic construction, again using the unmarked verb form combined with an overt pronoun. In other parts it uses marked verb forms inflected for person and number, but no overt pronoun. Pronouns and person-number inflections are thus in complementary distribution, and overt pronouns (differently from those in Welsh) act in the same way as full noun phrases. This distribution of zero subjects and pronouns provides a possible second - though somewhat more elusive - parallel with English, based on the opposite pairing of forms: this parallel works only if one equates the English -5 forms with the Irish inflected forms. It also requires that some construction in English involving the -s form be identified as a zero-subject construction. It can then be said that both Irish and northern English have unmarked verb forms with pronouns and marked verb forms with zero subjects. Note, however, that this analogy then does not apply to NP subjects; neither does it cover the case of the third singular. Corrigan, without apparently being aware of the alternative, opts for this second approach. This decision forces her to ignore the much more obvious parallel of the non-agreement in noun-phrases. She actually states that the Irish system in those aspects rather "resembles the Modern English Standard [sic] paradigm in which analytic [i.e. affixless] forms are used for all persons except the third singular" (1997: 199). As for the zero-subject cases in English, which she needs for her comparison, she finds them in the conjoined-VP construction (they sing and dances), assuming that the second VP in these constructions carries an empty subject element (1997: 200). The rest of her chapter is concerned with reviewing various formal accounts of agreement in Irish and other languages, only to find that each of them, on the basis of the analogy she has chosen, collides with the behaviour of either the noun phrase subjects or the third singular in the NSR system. It is also not obvious how her analysis involving empty elements would carry over to the intervening-adverb scenario (they often sings), of whose existence in older varieties she is aware (1997: 197). Corrigan also admits that the analysis treating the English -s forms as equivalents of the Irish inflected forms is functionally problematic because of the fact that the English -s form does not overtly code any actual person-number contrast (1997: 200). This finally leads her to the conclusion of characterizing the NSR system in earlier varieties primarily in terms of a functional, processing feature: the -s forms "acted as the default for all persons and numbers in discourse contexts where identification of the non-adjacent subject required greater than usual processing" (ibid.). Again, this is a functionally plausible idea but rather lacks the formal explicitness which the theoretical framework she professes to follow would require. Corrigan finally quotes Henry's proposal involving the 'nominative constraint', which she finds "offers an elegant analysis" (1997: 225), but without relating it to her own earlier formal discussion, and without apparently being aware of the implicit contradictory claims regarding the markedness relation between the two English verb forms.

2

The parallel with Welsh has led to the hypothesis, not to be confused with Corrigan's "convergence" proposal, of an English-Celtic language-contact effect on the earliest emergence of the NSR in Old English (Harap 1976, Klemola 2000). For a discussion, see Section 3.4 on page 60.

25 2.1.3. Börjars and Chapman (1998) Börjars/Chapman (1998) is an analysis of general properties of NSR dialects in the framework of Lexical-Functional Grammar (LFG). The authors draw upon the intuition, first developed in Chapman (1998), that pronoun subjects and adjacent verbs stand in a closer syntactic relation with each other than other subjects and verbs, and that together they form an integrated unit in syntactic processing - in the words of Chapman (1998: 39): "a syntactic unit which is interrupted if additional information is added in the form of a second pronoun, for example, or some sort of modifier between the pronoun and verb". This proposal, then, differs from the preceding two in explicitly focussing on the effects of the Position-ofSubject Constraint, as illustrated in (9). (9)

a. They often talks b. They talk

When seeking an explanation along formal lines for the contrast in (9), one must resort to either of two principal strategies: either posit that the two instances of they in (9a) and (9b) are in two different structural positions in the syntactic tree (which would not be the case in the corresponding sentences in Standard English), or posit that the two instances of they are somehow not the same thing. The analysis proposed by Börjars and Chapman combines both these options. As for the tree positions, they choose an analysis essentially opposite to Henry's: whereas Henry assumes the subjects of the -s verbs to be structurally closer to the (original) position of the verb in the tree configuration (the Τ node instead of the Agrs node), Börjars/Chapman ascribe a closer position to the subjects of the affixless verb forms. As for categorial status, they assume the existence of two partly homonymous sets of pronouns with different roles: those which co-occur with an -s ending on the verb: he, she, it, 11, youi, we/, theyi; and those that occur without an -s ending: I2, you2, we2, they2. The central point of their proposal is that this latter set has the categorial status of verb inflections, more exactly: "pronominal inflections" - they are really not part of the syntax but part of the verb morphology (1998: 76, 83). In what follows I will use the convention introduced by the authors of writing these constructions with hyphens to indicate their presumed bound status: they2=talk. Börjars/Chapman go on to specify the necessary properties of the two sets of pronouns in terms of feature settings in the formal mechanism of LFG. These properties entail the three most important aspects of their behaviour: that the second set must always be adjacent to the verb, that it cannot co-occur with another competing inflectional affix on the other side of the verb stem - namely, the -s affix - and that neither of the two sets can co-occur with each other or with another overt subject. In sentences with 'pronominal inflection', the affixal pronoun is generated under the V node, and these sentences therefore lack a subject in the canonical subject position, the sister of the VP. These sentences are therefore, technically speaking, pro-drop structures. Börjars/Chapman seem to assume, along the lines of the generative tradition, that a canonical subject position is nevertheless generated and that it contains an empty element. At least, this seems to follow from their discussion of the effects of their analysis on the so-called COMP- trace effect (1998: 86ff). Thus the following grammatical structures are generated:

26 (10) a. [MP theyi] [ V p often [ v talks]] b. [ N p e ] [ V p [v they 2 =talk]]

Structure (11), on the other hand, containing both the -s ending and a pronominal inflection they, is correctly predicted to be ungrammatical: (11) *[Np e ] [ V p [ v they 2 =talks]]

There are various problems with this analysis. First, as pointed out by Shorrocks (1999: 112), the assumption of bound (affixal or clitic) status intuitively conflicts with a central property usually connected with bound elements: phonological dependence. While it may be true that preverbal pronouns in connected speech often tend to be realized as phonologically reduced forms (Börjars/Chapman 1998: 97), these pronouns can nevertheless also take contrastive stress, and will then still occur with the non-inflected verb form in NSR dialects. A second problem concerns the descriptive adequacy of the proposal with respect to the variability of the system. It seems somewhat unfortunate in this respect that the authors base their analysis - as they admit themselves - on rather sketchy and incomplete data, collected from some cursory descriptions of agreement variation in the dialectological literature. This creates a particularly serious problem because the Position-of-Subject Constraint features so prominently in their discussion. As mentioned earlier, the effects of the Position-of-Subject Constraint in most modern dialects are rather marginal. Börjars/Chapman's proposal predicts, incorrectly, that the presence of -s after intervening adverbs should be categorical, i.e. that they often talk should actually be ungrammatical which it certainly is not, in any of the modern varieties I have looked at. Even if one concedes that the authors' goal is only to describe an idealized system and that their generalizations "may hold only for a subset of subjects or a subset of verbs in each dialect" (1998: 75), this constitutes a weakness of the proposal that cannot easily be glossed over. Moreover, problems may remain even within the proposed formal mechanism itself. To name but the most important ones: first, I fail to see how the model, in and of itself, accounts for even as much as the ungrammaticality of *they talks. Börjars/Chapman analyse this surface string as in (11) above, repeated here as (12b), which the model correctly predicts to be ungrammatical. However, one could just as well construe the same surface string as the structure in (12c), which does not seem to violate any of the principles posited by the authors. For all I can see, there is nothing in the proposal that would block this structure. (12) a. *they talks b. *[NP e ] [Vp [V they2=talks]] c. (?) [NP theyi] [vp [v talks]] The second problem stems from the, by now well-known, markedness paradox inherent in the NSR system. As stated earlier, Börjars/Chapman's proposal is based on the assumption that the affixless verb forms are unmarked in terms of agreement features. This may work for lexical verbs, but it hardly works for the forms of BE. Forms like am and are following the adjacent pronouns cannot easily be regarded as non-inflected in the same way as e.g. talk can. This is particularly obvious with I=am, where identical person-number information is clearly encoded twice, in just the way the LFG mechanism is designed to preclude.

27 Finally, the proposal clearly needs to be complemented with a mechanism to deal with inversion - possibly yet another set of 'pronominal inflections' that would attach to the right instead of the left of the verb stem. Börjars/Chapman (1999: 86) briefly discuss this problem but make no definite statements, because they they do not actually have appropriate empirical information on the behaviour of verbs in this situation. Their model is based on the guess that inverted pronouns should trigger the non-standard -s form, which unfortunately happens not to be the case. There is even a misunderstanding in Börjars and Chapman's review of what little literature they have available on this point: both Henry (1995: 16) and Harris (1993: 156), in the passages quoted, are talking about inversion between the verb and a full N P subject, a situation that is not really relevant to the question at hand, and happens to behave quite differently from that of verb-pronoun inversion in the dialects involved. 3

2.1.4.

Hudson (1999)

The latest proposal for a formal analysis of NSR systems is only a cursory, tentative sketch within a larger discussion of the Standard English agreement system, found in Hudson (1999: 204). Hudson puts forward an analysis within his own brand of formal dependency grammar, called Word Grammar (cf. Hudson 2001). As his contribution is not centrally concerned with the NSR dialects, it is not surprising that the descriptive information on which his account of them is based is even sketchier than in the case of Börjars/Chapman the only authority cited is the short description of NSR effects in Harris (1993). Nevertheless, his proposal offers an interesting move beyond those discussed up to now, as it avoids the methodological imperative of having to deduce observed grammatical behaviour from a small set of postulated universal elements. In Word Grammar, features and syntactic categories can be language-specific and can be freely assigned by ad-hoc rules. Words and grammatical entities form a taxonomic hierarchy, in which all elements can either inherit default properties from their superordinate nodes or override these by specific rules. To account for some special cases in the agreement behaviour of subject noun phrases in Standard English, Hudson introduces a novel ad-hoc feature of "agreement-number" that exists side by side with the normal morphological "number" feature. Both are linked through the default rule of identity, but may differ from each other in special cases (e.g., the noun family may have singular "number" but, by way of exception, plural "agreementnumber"). A similar mechanism of exceptional feature assignments specified for individual words is used to account for the exceptional behaviour of the pronouns I and you, and of the verb BE. In this framework, the formal description of the Type-of-Subject Constraint becomes surprisingly simple: in an NSR dialect, all nouns by default have either singular

3

There may be yet other formal problems regarding the possibilities of adverb placement. On the assumption that the presumed 'pro-drop' structures have an empty but structurally existing subject slot, and the usual tree positions for inserting adverbials around it, it seems to me that the proposal would wrongly predict the possibility of structures such as *[ NP e] [Vp often [ v they=talk]] or "•[fortunately [ N P e] [ V p never [ v they=fail]]].

28

"agreement-number" features or no "agreement-number" at all, and only the pronouns I, you, we, they have an exceptional plural "agreement-number".4 The main problem with this proposal is that it does not provide for a way to deal with the Position-of-Subject Constraint, of which Hudson seems not to be aware. Below, I will argue that a theory which makes very similar use of language-specific entities, taxonomic hierarchies, default inheritance, and non-separation between grammar and lexicon, but which regards the 'construction' rather than the 'word' as the central taxonomic unit for which grammatical rules are specified, provides a more natural account of both the Typeof-Subject and the Position-of-Subject Constraint. (About the close relationship between such a construction grammar approach and Word Grammar see also Holmes/Hudson 2001.) Summing up, each of the four formal proposals, by Henry, Corrigan, Börjars/Chapman, and Hudson, seems to capture valuable intuitions about the NSR but falls short of providing a credible and unified account of the whole phenomenon. Henry's analysis rightly draws attention to the interesting fact that the set of pronouns involved in the Type-of-Subject Constraint is partly identical with the set of items that has preserved the morphological contrast between subject and object case in English. Corrigan offers interesting thoughts about functional and processing motivations for the dialect-universal tendencies involved in the Position-of-Subject Constraint. Börjars/Chapman draw upon the plausible intuition that the syntactic structural link between pronoun subjects and adjacent verbs is somehow closer than that between other subjects and verbs and that they together form an integrated unit in syntactic processing. Hudson emphasizes the fact that the NSR system is made up of arbitrary exceptional properties associated not so much with abstract configurational properties of syntactic structures but rather with the pronouns as specific individual items. My own approach to the phenomenon, discussed in Section 2.3 below, will combine such elements and basic ideas especially from the latter two proposals. In doing this, it will try to avoid some of the descriptive narrowness of these formalist approaches, and additionally it will attempt to integrate the element of quantifiable, probabilistic variation into the account of the NSR.

2.2.

Theoretical approaches to variation

The formal analyses discussed up to now all follow the generative tradition insofar as they make statements only about categorical grammaticality judgments. Variability is catered for in the form of 'optionality' if at all: statements that license two alternative forms, but remain quantitatively unspecified. This is not to say that authors in the generative tradition necessarily ignore the relevance of the quantitative distribution of variation in discourse. For instance, Wilson/Henry (1998) are in full agreement with authors such as Kroch (1994, 2001) or Pintzuk/Tsoulas/Warner (2001) on the point that quantitative patterns of frequency are crucial for the explanation of language acquisition and language change. However, 4

It is part of Hudson's exceptional-feature mechanism that the agreement behaviour of / is paradoxically subsumed under the plurals. To account for its special behaviour with the forms of BE, Hudson uses another ad-hoc feature, "agreement-person", which only applies to I.

29 within the generative literature, the mechanisms that determine the actual outcome of the quantitative distribution in performance have traditionally been located somewhere outside of'grammar' proper. Henry (2002) is a rare exception within this research tradition, explicitly acknowledging the necessity of accounting for quantitative variability as part of 'core syntax'. Some recent proposals such as those of Henry (2002), Bresnan (2000), and (with some limitations) Yang (2003) can be seen as attempts at finally bridging the longestablished gap between variationist studies and syntactic theorizing. In the present study, I too will adopt an approach that describes quantifiable patterns of probabilistic variation as an integral part of a speaker's linguistic knowledge, closely integrating the statement of grammatical rules with statements regarding the probability of their appearing in discourse. The next sections will serve to motivate this choice and give a survey of existing theoretical proposals in this direction.

2.2.1. Competence and performance To integrate statements of probability into a description of grammatical knowledge is not a new idea. However, a comparison of older and much more recent contributions to the issue (for instance, Bickerton 1971, Berdan 1975, Sankoff/Labov 1979, Newmeyer 1998, 2003, Aissen/Bresnan 2002, Henry 2002) leaves one with the impression that the same old arguments are still being exchanged and that the discussion of the last thirty years has, on the whole, not moved much. I would like to suggest, first of all, that the concept of probability as part of grammatical knowledge does not, in principle, imply a denial of the logical distinction between competence and performance, as Newmeyer (2002b: 98) has claimed. At most, it implies a departure from the hypothesis of the autonomy of competence from performance, in the sense debated between Croft (1995) and Newmeyer (1998). A description of a speaker's 'competence' is, if anything, a description of those mental structures that determine his or her potential performance. If such a description may contain statements about the licensing of two alternative forms of a linguistic variable (i.e. optionality), then there is no a priori reason why it cannot also contain statements about the grammar's propensity to prefer the one form over the other by some quantifiable amount. Some theoretical objections that have been raised against the concept of probability as part of competence can be shown to stem from fundamental misunderstandings of the idea of randomness (see for instance Bickerton 1971 and the refutation in Sankoff/Labov 1979, as well as similar arguments in Aissen/Bresnan 2002). Decisive arguments for the conceptual and empirical necessity of probabilistic statements within grammar were summed up succinctly already by Berdan (1975), building on arguments in Wolfram (1973). Berdan argues against the by then orthodox view "that language behaviour is either categorical or outside the scope of grammatical description" (1975: 22). He goes on to demonstrate the difference between inherent variation on the one hand and code-switching or style-switching on the other: switching refers to quantitative shifts in the behaviour of several variables at once, sustained over an extended stretch of time in discourse, for instance along a formal-informal dimension; whereas inherent varia-

30 tion manifests itself as random variation that can occur independently for different variables and at any point in time. He sums up the arguments (1975: 24): Variation with respect to different features of the grammar is independent, it is not code switching. [...] Patterns of variation are not idiosyncratic; they can be observed across generations. Children do not only acquire the form of a rule, they acquire the probability of its use. Patterns of variation are observed to co-occur with other linguistic entities; they are language and dialect specific, not universal.

An essentially similar argumentation regarding one of these points, the cross-generation learning of frequency distributions, has recently been adopted again by Henry (2002), who gives examples of NSR-related verbal agreement variation across generations in Belfast English. Whereas Henry (2002) has acknowledged the empirical necessity of 'variable rules' as part of grammar, Newmeyer (1998: 56f.; 2003) is representative of the traditional generative division, excluding variation patterns from the concept of competence. In his view, the task of competence is only to define the range of possible structures available to the speaker, and it is then a matter of performance to choose between them: "speakers have to know what their options are before deciding (however unconsciously) which of them to use" (2000: 57). However, his arguments do not do full justice to the fact that variation is not only inherent but also structured, and therefore they do not easily apply to the variation phenomena to be discussed in the present study. In all the examples of variation that Newmeyer discusses in the passage quoted, the choice between two forms is demonstrably determined by either pragmatic or by processing factors, as in (13a) versus (13b), or (14a) versus (14b). (13) a. John hit the ball. b. It was the ball that John hit. (14) a. That Mary will be elected is certain, b. It is certain that Mary will be elected.

These factors can plausibly be seen as universally available, functionally grounded mechanisms of performance. It must be conceded to Newmeyer that they are thus of a different kind than those language-specific regularities that provide the various structural options in the first place. (This is true regardless of whether or not one might nevertheless wish to subsume them under 'linguistic knowledge' in some sense different from that adopted by Newmeyer.) However, in the case of verbal agreement variation, such a distinction is hard to maintain. There is no way in which the sentences in (15) are pragmatically or semantically different from each other in the way those in (13-14) are: (15) a. They was going b. They were going

Nor is there any universal functional explanation why was should be pragmatically more appropriate or easier to parse or to process in (16b-c) than in (16a): (16) a. Many was going b. Lots of them was going c. Them ones was going

31 Yet, systematic, quantifiable differences in frequencies of verbal -s between clauses such as these are consistently found in corpus data. Quantitative empirical studies of subject-verb agreement variation in English dialects or in historical varieties (for representative examples see Kytö 1993, Tagliamonte 1998 and Godfrey/Tagliamonte 1999, together with the data in the present study), suggest that the choice of agreement morphology in varieties of English may be sensitive to environmental dimensions of at least the following different types: phonological environment; syntactic constellation between subject and verb; heaviness of the subject NP; the lexical item of the verb; the auxiliary function of the verb; the occurrence of specific lexical items or determiners within the subject NP; the occurrence of negation; and possibly the temporal-aspectual semantics of the clause. Kroch (1994: 183), speaking of partly similar variation phenomena, acknowledges the existence of such structured, environmentally graded distributions, but ascribes them to factors which "are not well-understood but must involve psycholinguistic and information processing preferences". In a similar vein as Newmeyer, he concludes from this that their existence is "orthogonal" to the existence of the different forms itself. However, while it is certainly true that some of the factors mentioned above may indeed be indirectly linked to cognitive processing explanations, others seem to be just as language-specific (or dialectspecific) and as arbitrary as the existence of the structures as such. The fact that, for instance, clauses whose subject contains the item them are a preferred environment for the use of verbal -s (as opposed to they, which behaves in the opposite way) cannot be predicted from any more general principle but must be learnt as such. Usage preferences of this type are therefore most naturally described as parts of the linguistic knowledge of speakers, just as the structures themselves are. Moreover, the competition that takes place between the alternative structures (as in 15a-b) is to such a high degree sensitive to factors that are themselves clearly within the realm of 'grammar' that the competition can most naturally be described only as part of that grammar. Even though this competition is instantiated, by logical necessity, only in the time and space of 'performance', it cannot be described independently from the properties of the 'competence' that determine it. Kroch (1994: 185) also implies that the perceived theoretical problem of accounting for morphosyntactic variation can be alleviated if one assumes that variables arise through dialect and language contact and compete in usage until one or the other form wins out. Due to their sociolinguistic origins, the two forms often appear in different registers, styles, or social dialects; but they can only coexist stably in the speech community if they differentiate in meaning, thereby ceasing to be doublets.

While this statement is certainly valid as a description of the origins and historical role of many variables, it says nothing about the way the variable is represented in speakers' grammars. It is therefore hard to see how this argument should serve to explain away the facts of variation so that they would fall outside syntactic theory, as Kroch seems to intend. (For further critiques of Kroch's position, cf. also Hudson 1997, and Bresnan/Deo 2001: 37ff.).

32 2.2.2.

Individual grammar and community grammar

It is a fact commonly accepted in most variationist research that quantifiable patterns of variation can be found both in the individual output of speakers and, by consequence, in the collective output of a speech community. However, the straightforward assumption that the variation patterns found across the community directly mirror those in the individuals is empirically not unproblematic (cf. Mufwene 1994), and a possible objection against the method chosen in the present study must be countered at this point. This is crucial in the present case, because the phenomenon investigated is not of sufficiently high discourse frequency as to make it easy to collect amounts of data sufficient for reliably describing the effects of all variables within the production of individuals. Reliable, quantitatively significant results can often only be reached by counting tokens across a corpus consisting of the pooled data from many individuals, and it must be tested - as far as possible - to what degree the assumptions underlying this method are justified. One possible problem regarding the relationship between the individual and the community grammar was pointed out already by Bickerton (1971). He argues against the assumption that statistical grading effects observed across the speech community necessarily mirror probabilistic constraints in the individual. Linguistic constraints that govern variation might alternatively be of an 'implicational' instead of a probabilistic nature (cf. Rickford 2002). The distribution of a linguistic variable across an ordered set of linguistic environment types may be such that each individual has one out of a set of different possible grammars, each of which imposes a categorical choice rather than a probabilistic weighting on each environment type. The set of environments may form an ordered hierarchy, such that every possible grammar that allows a given form in a given environment will also allow it in all environments further down the hierarchy. Thus, a state of ordered variation, where some environments seem to probabilistically favour and some seem to probabilistically disfavour a given form, may well be observable across the collective output of the speech community, but this may in effect be only an epiphenomenon of pooling the output of the individuals, each of whom in reality shows categorical rather than variable behaviour. Bickerton illustrates this with data from the highly variable linguistic systems typical of post-creole continua, suggesting that categorical or at least near-categorical behaviour across ordered hierarchies of environment types on the individual level may be more common than assumed by Labov and others. He also suggests that a tendency for variables to display a distribution close to either 0% or 100% on the idiolectal level may be indicative of rapid change in progress. Shifts from one categorical rule to a different categorical rule may, for each speaker and for each environment type, pass through a period of variation, but that period of variation is likely to involve the typical S-shaped curve: "a rule-change is slow to get under way, shoots forward relatively rapidly to a point near completion, and then again takes some time to complete entirely" (1971: 489). Any state where a variable has a quantitative distribution closer to 50%, for any individual and environment type, would then tend to be relatively brief and unstable. While Bickerton's model of implicational constraints is certainly a valid possibility, other research has shown many times that quantifiable constraint effects on the individual level do occur, as do relatively stable diachronic states where quantitative patterns of varia-

33 tion are passed on from one generation to the next with little or no change (Berdan 1975, Hudson 1997, Henry 2002). To illustrate this point, I will here present some data from the corpus study of Northern Irish English speech which forms the topic of Chapter 5. As will be discussed in more detail in that chapter, there is strong statistical evidence for the following conditioning effect on NSR-related use of verbal -s (see page 110): subject relative clauses (i.e. those relatives in which the relativized element is the subject, and hence verbal agreement is governed by the clause-external head NP) have a significantly higher likelihood of displaying the nonstandard verbal -s forms than other, canonical clauses with clause-internal subjects. This is one of the conditioning factors that figure prominently in the multivariate analysis of variation patterns in that dialect, as discussed more fully in Chapter 5. The question of whether this constraint is of the implicational or of the truly probabilistic type meets with two obvious methodological problems. First, given the relatively low frequency of the phenomenon in question, it is difficult to find sufficient amounts of data from individuals. I have chosen the relative clause effect for this demonstration because it is quantitatively better attested than most other effects of this kind discussed in this study. Second, asking questions of categoricity of a corpus is always problematic, because a corpus does not provide negative evidence. A corpus can only confirm that forms are possible; it can never prove which forms are impossible. At best, the non-attestation of a form in a corpus will prove that it is unlikely to occur, but it requires at least a very large set of counterexamples until one can conclude with reasonable confidence that a certain form is anything close to non-existent. It is easy to find speakers in a corpus who have, say, verbal -s in zero out of twelve, or in twelve out of twelve cases in a given type of environment - but that is hardly enough to predict with any confidence that the same speaker could not have produced the opposite form the next moment. 5 In the present case, all that can be done is to look, with a lot of caution, into the recorded production of some of the best-represented individuals within the corpus, especially those who do display variable behaviour in both environments. Informant L9.3:JM is an educated retired farmer from northern Tyrone, living in an area where NSR usage is quite common in the dialect, but he himself is not a particularly heavy verbal -s user. He produced a total of 5 instances of verbal -s out of 23 potential opportuni5

It must be stressed at this point that alternative methods such as acceptability tests are likely to suffer from problems no less severe than this one. When it comes to fine-grained differences between idiolects co-existing within the same speech community, acceptability testing presupposes a degree of reliability of informants' introspection that cannot realistically be expected. The informants would need to be able to distinguish reliably between what they themselves would or would not produce, and what they hear or do not hear from others. An informant would have to be able to state with confidence, for instance, that he himself never produces a certain form in a certain environment, even when the same structure is frequently used by others in the same speech community, and the informant actually uses the form himself too in only slightly different environments. In a matter as fine-grained as the present question, this is likely to be beyond the introspective capacities of most if not all informants. The dangers of both over- and under-reporting would very likely invalidate most results. Acceptability tests and similar tasks may very well give valid evidence for what types of constructions are more or less common (for some details of methodology see Glaser 1997, Barbiers/Cornips 2002), but as for real categoricity in people's individual grammars, they are not likely to provide any more reliable evidence than the corpus findings do.

34 ties. Among these are 3 out of 6 tokens (50%) in relative clauses, and only 2 out of 17 tokens (12%) in other clauses. Clearly, the relative clause effect plays some role in his speech; nevertheless, both in relative clauses and elsewhere his production is clearly variable. Neither can the observed variation be explained by code switching between two distinct varieties, each with its own (presumably categorical) rules. Tokens with and without verbal -s follow too close to each other within the discourse. For instance, within the span of two utterances, the informant produces one relative clause without, and one non-relative clause with verbal -s (17a-b). Later, he produces two immediately following repetitions of an identical sentence, once with verbal -s and once without (17c). (17) a. Look at the growths of villages that were only a church, and a pub, and a, twenty years ago, are now developed into towns. [L9.3;JM8] b. The local people's to blame, because they've to... [L9.3:JM 10] c. Eh, the improvements that's there, in my time, in my, to my knowing, the improvememts that are there, eh, would compare with any other area in the north of Ireland. [L9.3.JM51]

Informant L22.2:JH is a forestry worker, aged in his mid-forties, who lives in western Fermanagh. He has plural verbal -5 in 14 out of 20 tokens. Among these are 5 out of 6 (83%) in relative clauses, and 9 out of 14 (64%) in other clauses. Again, the conditioning effect seems to be active, and it might be the case that for this informant the preference for verbal -s in relatives even verges on the near-categorical. However, the one exceptional non-5 form in a relative occurs again quite near to other sentences in which he does use verbal -s in less favouring environments. Moreover, this one exception (18) happens to contain the verbal form were, although another cross-cutting probabilistic conditioning factor (discussed on page 104) predicts that the likelihood of the -s form should be particularly high with this particular verb. This indicates again that all the relevant constraints in the idiolect of this informant allow for possible violations. (18) Well, all the men that were used to work with farmers turned out that day, you know. [L22.2:JH66]

Informant L17.3:LD is a retired domestic servant aged in her late sixties, and we will encounter her again as one among the heaviest NSR users in the corpus. Like the first informant, L9.3.JM, she lives in northern Tyrone, and she speaks with quite a broad Mid-Ulster accent. She is the only informant who shows more than just isolated use of verbal -s after non-adjacent I, you, and we (cf. page 101), and she also has plural verbal -5 in 19 out of 25 cases of plural NP subjects. Among these are 7 out of 9 tokens (78%) in relative clauses, and 12 out of 16 (75%) elsewhere. Her preference for verbal -s is so strong in all contexts that the relative clause effect does not seem to make much of an additional difference. In any case, despite her heavy preference for verbal -s, her grammar clearly allows for some residual variation in either type of environment. Table 2.2 summarizes the counts for these and some additional selected informants. Some of them have 0% of verbal -s everywhere, some have 100% everywhere. For those in between, the proportions in relative clauses are almost always equal or higher than those in non-relatives. There are also some informants who have either 0% or 100% of recorded verbal -s in one type of environment but variation in the other. However, the number of such cases is not larger than what would be expectable on grounds of random effects alone,

35 and the differences in observed proportions are not large enough to warrant a claim that these informants have grammars qualitatively different from the rest. As the total token counts for most individuals are much lower than the ones shown in the table, a rigorous statistical testing of these claims is not possible. However, the picture seems clearly compatible with the view that the observed distribution of verbal -s is governed by probabilistic mechanisms on the level of the individual speaker. Table 2.2. Plural verbal -s in relatives and non-relatives for selected NITCS informants Informant L23.3:MK L32.2:MM LI 1.2:TS L16.2:IC L13.2:IP L9.3:JM L7.2:TM L17.3:LD L36.2:JO L2.2:JM L16.3:PT L10.2:BC

2.2.3.

Relatives 0/4 0% 0/6 0% 0/6 0% 2/13 15% 2/6 33% 3/7 43% 2/3 66% 7/9 78% 8/8 100% 4/4 100% 6/6 100% 2/2 100%

Non-Relatives 0/21 0% 0/12 0% 3/30 10% 0/12 0% 0/16 0% 2/16 13% 20% 3/15 12/16 75% 4/8 50% 9/12 75% 4/4 100% 9/9 100%

Total 0/25 0% 0/18 0% 3/36 8% 2/25 8% 2/22 9% 5/23 22% 5/18 28% 19/25 75% 12/16 75% 13/16 81% 10/10 100% 11/11 100%

Variable-rules theory

Theoretical attempts at integrating probabilistic variation into an account of linguistic competence were first proposed as extensions to an otherwise generative-transformationalist view during the 1960s. Sankoff/Labov (1979: 215) quote a number of early proposals such as Klein (1965), Grenander (1967), Kherts (1968), Horning (1969), Suppes (1970), and Soule (1974). However, the most influential proposal in this respect was clearly the concept of 'variable rules', developed in the framework of the sociolinguistic studies of W. Labov and others associated with that research tradition (see the seminal programmatic statement in Weinreich/Labov/Herzog 1968). A mathematical model suitable for the description of variable grammars according to the variable-rules approach was developed and refined during the 1970s by D. Sankoff (for an introduction to the model cf. Sankoff 1988; for its mathematical justification in more detail cf. Sankoff/Labov 1979; for its application cf. papers in Sankoff 1978; for more recent state-of-the-art surveys and introductions to the method, see Bayley 2002 and Paolillo 2002). Early statements of variable-rules methodology were still couched within a formal terminology that marked it as an extension to transformationalist grammar. However, as the idea of integrating probabilistic variation into grammatical description was not taken up by mainstream generative research, and as the theoretical interests of the formalist and the sociolinguistic research programmes were felt to diverge radically from each other (cf. Sankoff 1998, Wilson/Henry 1998), this conceptual link was gradually given up in its practical application.

36 Independently of the theoretical perspective, variable-rules descriptions have remained, if only as a seemingly theory-neutral heuristic analysis tool, a methodological mainstay of much variationist research. Variable-rules descriptions are applicable "wherever a choice can be perceived as having been made in the course of linguistic performance" (Sankoff 1998: 151). Statistical methods developed for them are implemented and widely available today in the Varbrul software packages (e.g. Pintzuk 1988, Robinson 2002). They can be used as a convenient multivariate analysis tool wherever two or more groups of intersecting factors are perceived as jointly determining the choice of a dependent variable, no matter of what nature and theoretical status these factors are. Varbrul will also be used as a statistical tool for the analysis of corpus data in the present study (for a brief introduction to its application see Appendix A on page 155.)

2.2.4.

Stochastic Optimality Theory

A new attempt at integrating probabilistic variation into a formal theory of grammatical competence has recently been made within the framework of Optimality Theory (henceforth OT). Optimality Theory (Prince/Smolensky 1993) is a formal framework for describing the interaction between linguistic universals and language-specific properties of grammars. In this approach, universals are assumed to be represented in individual grammars in the form of constraints, which penalize certain linguistic structures but allow others. Constraints can compete with each other and may be violated. Language-specific settings are implemented by different rankings of the universal constraints. Out of any set of possible linguistic structures that compete for the realization of a given underlying content, the one structure that violates the least highly ranked constraints wins out. Hence, the rank order of constraints in this framework can be thought of as fulfilling a similar role as parameter settings in a Principles-and-Parameters based approach. While OT was originally designed for handling phonological problems, it has repeatedly been invoked also in studies of morphosyntax too. Haspelmath (1999) and Aissen (1999, 2000) have proposed treatments of syntactic typology in an OT perspective. Bresnan/Deo (2001) have brought the OT approach into the study of English dialect grammar, demonstrating how the different morphological paradigms of BE in traditional English dialects can be modelled through OT constraint rankings. Some versions of OT, including that invoked by Bresnan/Deo, also contain a probabilistic component. In Stochastic Optimality Theory (Boersma/Hayes 2001) constraint rankings are modelled mathematically not as discrete ordinals, but as points on a continuous scale, with each constraint being assigned an arbitrary value accompanied by an arbitrary range of numerical variability. On each subsequent activation of the grammar (during production) the weights of each constraint are assumed to vary slightly according to stochastic principles. This means that if the weighting of two conflicting constraints are sufficiently close to each other, the outcome will be sometimes in favour of the one and sometimes in favour of the other. In principle, quantitative distributions of a variable can be modelled in such a framework in a way rather similar to a variable-rules model. Like the latter approach it has been criticized, from a more orthodox Chomskyan perspective, for undermining the competence-performance dichotomy (Newmeyer 2002b: 98).

37 Stochastic OT also comes with a numerical model of acquisition, the 'Gradual Learning Algorithm' (GLA), which is designed to explain how an appropriate constraint ranking can be determined by a learner in such a way as to replicate the frequency distributions found in the learning input (Boersma/Hayes 2001; see Boersma/Weenink 2000 and Hayes/Tesar/ Zuraw 2000 for computer implementations). This attempt of formally tackling the learnability issue for a probabilistic grammar is certainly a welcome contribution to the theory, as it provides an attractive account both of individual acquisition and of the diachronic, cross-generation proliferation of variable grammars. Nevertheless, it is doubtful if the algorithm, in the form invoked by Bresnan/Deo, can be regarded as psychologically realistic. This is because it presupposes that the learner have a priori access not only to the surface forms to be learnt but also to semantic feature specifications associated with each of them but these are in reality themselves part of what has to be learnt. This leads to evidently absurd situations in the particular case studied by Bresnan/Deo. Their analysis of English dialectal BE paradigms depends crucially on the concept of semantically underspecified entries. For instance, the form is may, in some dialects, represent the full semantic feature structure [3Sg], in others only [Sg], and it may be semantically empty [ ] in yet others. Taken at face value, the learning model would then require that learners be able to tell a hypothetical form w[3Sg] from a homophonous «[Sg] or «[ ] when they encounter one in the learning data. Moreover, when for instance the learning data contains zs[ ] and the learner's intermediate grammar would produce zs[3Sg], the learner is supposed to notice the difference and adjust the grammar accordingly. If the OT algorithm is to become a psychologically realistic model of morphosyntactic learning, then it would seem that it needs at least to be refined in such a way that the two learning processes - inferring semantic feature structures for lexical items, and setting constraint weights - can proceed together by way of mutual trade-off. Apart from this, the main empirical problem of Stochastic OT seems to be how to map the multiple cross-cutting factor effects observed in real-world variation data to a (preferably small) set of universal constraints. Existing studies in the Stochastic OT approach have mostly dealt with variation only along the dimensions of very few independent variables. I am not aware of a Stochastic OT treatment of a set of data as complex as the multivariate tableaux commonly encountered in variationist studies. In purely mathematical terms, it stands to reason that many types of distributions that can be described by a variable-rules multivariate model, or even more, can also be captured by a Stochastic OT model in principle. However, for the latter to become more than just a notational variant of the former, it is necessary that every constraint posited within the OT model be identified and motivated as universal. How could a multivariate variable-rules analysis be 'translated' into a Stochastic OT account? The answer is straightforward in the case of variables of inter-speaker variation, such as age, sex, etc.: different speakers can be assumed to have different settings for the same constraints. In the case of style or text-type related variables, such as formality, a solution is sketched out in Bresnan/Deo (2001: 37f.): a given constraint may be promoted or demoted by some arbitrary constant under certain situational circumstances. The problem is more difficult for those remaining, purely linguistic context factors that govern inherent variation. Here, for any factor observed in the data, there are in principle three alternative explanations: either posit a new (universal!) constraint which, through its competition with the other constraints, causes the favouring or disfavouring effect; or posit some

38 language-specific mechanism (for instance, triggered by some property of a lexical item) that promotes or demotes an existing constraint; or delegate the effect to some performance mechanism 'outside grammar proper', such as processing ease. The first of these alternatives may entail an undesirable proliferation of new 'constraints' of doubtful universality. The second alternative weakens the universalist elegance of the model as a whole, and the third is tantamount to giving up the O T model as an explanative approach for those aspects of variation. To give yet another example of an apparently arbitrary, unmotivated variation effect: it has repeatedly been reported for various dialects of English that aspectual semantics may be a factor influencing subject-verb agreement behaviour (e.g. Godfrey/Tagliamonte 1999: 105f.), and some evidence to that effect will also be presented in the present study for Northern Irish English. Godfrey and Tagliamonte state that sentences of type (19b) were significantly less likely to display verbal -s in their Devon corpus than sentences of type (19a). This effect can hardly be explained by cognitive processing mechanisms outside grammar proper. But it is just as difficult to accommodate in terms of universal O T constraints. Comparing sentences such as (19a-b), it is difficult to conceive of any possible universal constraint regarding syntactic structure that would be violated by the one but not by the other: (19) a. I does few jobs (habitual) b. As long as I gets there tomorrow, I don't care (punctual) Clearly, a constraint of the form "avoid verbal agreement contrasts on habitual verbs" is not a suitable candidate for an innate property of Universal Grammar..

2.3.

A usage-based approach

In order to account for the variation data in the present study, a theoretical approach is required which accepts inherent variation as a natural component of linguistic knowledge. In addition, I will adopt the view (recently stated most forcefully by Croft 2001) that grammatical structure need not be described in terms of a set of atomic, universal elements (such as 'features' on lexical and functional 'heads' in tree configurations) in the way it is usually assumed in most formal syntactic theorizing, and that, as a consequence, grammatical 'explanations' need not consist in deducing the observed behaviour of grammatical constructions from the hypothesized properties of such atomic elements. Instead, the 'construction' itself, as a holistic schema, can be regarded as the central and thoroughly language-specific unit of grammatical description. In assuming these views, the present study takes its place within what has come to be called the "usage-based framework", a research tradition inspired by such strands of research as Cognitive Grammar (Langacker 1987, 1990), typologically informed functionalist research in language universals (Croft 1990, 2000, 2001), grammaticalization studies (Hopper/ Thompson 1993), and usage-based approaches to morphology (Bybee 1985). The mode of explanation employed in this research tradition crucially depends on the idea that grammars are not stable f r o m the completion of language acquisition during child-

39 hood, but can continuously change, being influenced by the language encountered throughout the lifetime of a speaker. A central role in this process is attributed to the relative frequencies with which different linguistic structures are encountered in discourse (Krug 2003). Different quantitative exposure to structures in discourse is directly mirrored by quantitatively different representation of these structures in linguistic knowledge. This assumption crucially diverges from the idea common to most theorizing in the generative tradition, which assumes that core grammar consists of purely qualitative representations, which either allow or disallow a certain linguistic structure. The most important cognitive mechanism invoked in modelling frequency effects on mental representation is "entrenchment", strengthening of representations through their activation in use (Langacker 1987: 59f.). A massive body of evidence has been collected in the study of morphology and of the diachronic behaviour of high-frequency collocations, especially during processes of grammaticalization, showing that frequency effects can be of crucial importance for the explanation of many processes of language change, and hence also of the synchronic states resulting from it (for recent studies, cf. various papers in Bybee/Hopper eds. 2001). Similarly, frequency distributions in the input can directly be mirrored in children's language acquisition patterns (Theakston/Lieven/Pine/Rowland 2005). As a theoretical framework for modelling syntactic knowledge represented in such a way, many proponents of a usage-based framework have appealed to construction grammar in recent years (Langacker 1987, 1993; Barlow/Kemmer 1994; Croft 2001; Tomasello 2003). In a construction grammar approach, properties of grammatical constructions, on all levels of abstraction ranging from individual words trough structural idioms to the most general principles of word order, are assumed to be all represented in essentially the same way. Constructions form a multiple taxonomy, ranging from highly schematic to highly specific, where the more specific constructions can inherit but also override the grammatical behaviour specified in the more general schemata of which they are instantiations (Kemmer/Israel 1994: 166; Croft 2001: 25ff). The behaviour of constructions at all levels can be thought of as non-universal and arbitrary, which means that there is no need to search for 'explanations' in the sense of the highly abstract generalizations common in the microparametric syntax of the Chomskyan tradition. The conceptual simplicity of this approach lies in the non-separation of grammar and lexicon. In classic generative approaches, as well as in frameworks such as LFG or HPSG, all idiosyncratic, non-universal behaviour of constructions is supposed to be rooted in the lexicon, i.e. in some feature structures specified for individual items - single words or, in the case of recent Principles-and-Parameters models, so-called functional heads in the tree structure. The task of the analyst is then to describe a way how these lexicon-specific feature structures of the atomic components can entail the observed behaviour of the construction as a whole, by making their contribution to the complex, universal calculus mechanism defined in the formal machinery of phrase structure, derivational movement and so forth. In the construction grammar approach, idiosyncratic behaviour can be specified directly for the construction as such. Moreover, in the cognitive version of construction grammar assumed here, these idiosyncratic properties can include quantitative effects, caused by quantitatively different strength of entrenchment of each construction. Applied to subject-verb agreement in English, such a construction-based approach can easily describe, for instance, the combination they + verb as a low-level pattern that instantiates a higher-level, more general schema subject + verb, and which, while sharing some or

40 most of its properties with this more general construction by default, may also partly differ from it in some arbitrary way, for instance with respect to its agreement behaviour. The existence of specific, lexically filled construction schemata reflects the fact that these specific units are acquired first in language acquisition, and that children build up higher-level abstract construction schemata over them only later during acquisition (Tomasello 2003). Moreover, I will follow the assumption suggested by Langacker (1987: 41 Iff.) and Croft (2001: 27f.) that entrenchment of high-frequency combinations may result in their retaining unit status even in adult grammar and that they may continue to exist as independently represented schemata side by side with the more abstract schemata they instantiate. This means that some information in syntax may in fact be represented redundantly, in a way similar to what is often assumed for inflectional morphology (Bybee 1985, 1995). I shall also assume that different construction schemata can compete with each other during production of an utterance, and that variability in a speaker's production can be explained by this competition (Bates/MacWhinney 1987). The probabilistic results of this competition provide "the basis for a natural account of the sort of structured variability traditionally modelled by variable rules" (Kemmer/Israel 1994: 165). The more heavily entrenched a constructional schema is in memory, the higher is its probability of being selected as the relevant categorization unit for the production of a specific usage event (cf. Kemmer/Israel 1994: 170f.). A frequency and usage-based approach of this type can help to account both for the complex state of variation observed in the present-day dialects, and for the historical processes that led to the emergence of the Northern Subject Rule in the first place. At the very heart of the northern agreement pattern are constructions whose principal distinguishing property is their high discourse frequency: combinations of personal pronouns and finite verbs. Pronouns have, by their very nature, a much higher discourse (token) frequency than any lexical noun has. The processing of recurrent collocations of adjacent pronoun subjects and their verbs involves a much higher degree of entrenchment and routinization than that of combinations containing lexical noun subjects (Theakston/Lieven fc. and Theakston/Lieven/Pine/Rowland 2005). This can account in a straightforward way for historical tendencies that lead to morphological reduction and irregularity associated with these constructions. As will be explained in section 3.1, it was exactly such a process that seems to have provided one of the main triggers for the emergence of the Northern Subject Rule in early Middle English: the phonological reduction of the verbal inflexional suffixes in the highfrequency environment with adjacent clitic pronouns. In this perspective, the emergence of the Northern Subject Rule can be seen as a case of morphosyntactic irregularity associated with a high-frequency construction, induced by mental routinization. I hypothesize that the verb-pronoun combinations involved in the Northern Subject Rule have unit status in the grammars of English speakers, i.e. that they are represented as partly separate construction schemata. Such independent representations may exist not only for specific collocations of individual pronouns and individual verbs, such as they + were, but also, at an intermediate level, for more schematic categorizations consisting of a specific pronoun and a free slot for any verb (plus the agreement morphology associated with the construction). Such a construction, e.g. they + verb, may share all or most of its properties with the more general construction subject + verb, which it instantiates, but may also override the behaviour specified for the latter in idiosyncratic ways.

41

(see Figure 2.1). From the perspective of a maximally parsimonious grammar, a minimum of three construction schemata, i.e. nodes in the taxonomic hierarchy of stored grammatical representations, need to be posited (ignoring, for the moment, the special cases of BE and that of the non-agreement verb forms). One is a maximally abstract schema specifying the relationship of subject + verb in general. It has two daughter nodes, one for third person singular subjects, specifying the use of the verbal -s morphology, and one for all other combinations, specifying the use of the base forms of the verbs. In a usage-based account, it may be further assumed that some additional, subordinate nodes may exist for more specific, frequent combinations, involving for instance pronoun subjects and/or high-frequency verb forms (shown in dotted lines in Figure 2.1). These may be stored and processed as separate schemata, having attained unit status through their entrenchment as high-frequency combinations, but they are not associated with any observably different formal behaviour and therefore formally redundant. Figure 2.2. A construction inventory for NSR agreement

42 Figure 2.3. A construction inventory for variable NSR agreement

Compare this system with that of an NSR dialect, considering first the case of a hypothetical, 'pure' system in which the NSR is categorical throughout (Figure 2.2). Here, the structure of the inventory is somewhat different. There must be a set of nodes representing the exceptional behaviour of the pronouns I, we, you and they (which take the - 0 forms), in opposition with a more abstract schema representing the use of the default verbal -s form for all the rest. Again, there might also be further, more specific nodes which represent other high-frequency combinations but which are formally redundant. In particular, it seems quite likely that those pronoun combinations that take the verbal -s forms (he/she/it + V-s) may also be stored and processed as independent units just like the other pronouns, owing to their high frequency of occurrence. We now come to the representations needed in the case of a hybrid dialect involving variation between standard-conforming and NSR-type agreement. I assume that each of the formal options available to a speaker of such a dialect is represented by a construction schema of its own, and that these schemata can compete with each other in production. Figure 2.3 shows a partial network structure of the construction inventory needed for such a system. In this system, the schemata NP V-s (of the NSR system) and NP P | V - 0 (of the standard system) will compete with each other. Any usage event of a particular subject and a particular verb will be processed as instantiating either of these two, whichever is more salient when the construction is being processed. Depending on which of the two is more deeply entrenched, either the one or the other will have the higher likelihood of winning out. Structured variation effects conditioned by further context factors, of the type described by variable-rules models, can be integrated in a system of this type by again assuming further, subordinate nodes in the network. In a construction grammar approach, separate representation as a unit in the structured inventory of a speaker's knowledge is commonly ascribed to "[a]ny construction with unique, idiosyncratic morphological, syntactic, lexical, semantic, pragmatic, or discourse-functional properties" (Croft 2001: 25; emphasis in original). In order to capture a situation of structured, inherent variation, it will be necessary to add a further factor to this list: any type of construction that is associated with arbitrary, quantifiable differences with respect to a linguistic variable must also have unit status and

43

be separately represented. In this way, nodes in the construction network may correspond to what would be modelled as "factors" in a variable-rule model. Figure 2.4 shows a partial network model of such a system, modelling one type of frequency effect repeatedly found in corpus data (as reported in chapters 4—6): likelihood of -s use varying between different individual verbs. Both the competing nodes from Figure 2.3 (NP V-s and NP P | V - 0 ) are here shown as having a number of daughter nodes. The subnodes shown represent schemata associated with individual verb forms. Each of the schemata represents either a standard-conformant or an NSR-conformant use. Each of them may be thought of as characterized by its individual degree of strength of entrenchment (symbolized in the graph by boxes of different thickness). During production, there will be a pairwise competition between relevant nodes. In this way, a situation can be modelled in which, for instance, plural use of is as opposed to are is relatively more likely than plural use of goes as opposed to go, just as it is found in the corpus data from many northern dialects. A usage and schema based approach such as this can also account in a natural way for the existence of the two principal types of constraints on subject-verb agreement that have been reported for many varieties of English, not only those in the direct historicalgeographical lineage of the NSR systems: constraints sensitive to the type of subject and to the position of the subject. Various (possibly non-categorical) forms of type-of-subject constraints may arise as combinations of verbs and personal pronouns, owing to their high discourse frequency, are entrenched separately and hence form extensions to the more abstract construction schemata for the subject-verb construction. Position-of-subject constraints can be seen as a corollary of the fact that both the more specific and the more general schema may compete with each other in the production of a usage event. I hypothesize that the more specific schema in memory - a gestalt consisting of a particular pronoun and a verb - will be more salient, and hence more likely to be activated as the relevant categorizing unit, if the utterance that is being formed involves a direct collocation of the pronoun and the verb. If both items are not adjacent, the construction less closely matches the gestalt prototype of the stored schema. It will then be less likely to be categorized as an instance of that specific schema, and by default the more abstract schema will be more likely to win out as the relevant categorizing unit.

44 It may be conjectured that effects of these types are more likely to emerge in a situation where a verbal agreement paradigm has previously undergone heavy erosion and reduction. As will be shown in the next chapter, such a state of affairs held just at the point in time in which the NSR first emerged in northern early Middle English. The present tense verbal system had been reduced, by a series of changes, to only one formal opposition between -(e) and -(e)s, so that only the first singular was reliably distinguished from all the rest of the paradigm. If language learners are confronted with such a heavily eroded - typologically and functionally implausible - agreement system, then a distinction such as that between pronominal and non-pronominal subjects may become cognitively more salient in processing than the person-number distinction, and a re-structuring of the agreement system along these dimensions may be the long-term consequence. It may also be argued that a similar state of affairs has again been reached, in a different form, in Modern Standard English, where only the third singular is distinguished from all the rest of the paradigm (and where, in addition, the contrast between was and were patterns according to yet a different rule, contrasting first or third singular as against the rest). This may be a reason why variation phenomena resembling the NSR have emerged, independently as it appears, in some modern varieties outside the original NSR area (see page 49 below). In the following chapter, which will review what is known about the historical origins and development of the NSR, I will attempt to show how this usage-based framework can offer a plausible model for the development that led to the split of the original common Old English agreement system into the NSR system of agreement with pronouns and nonagreement with other subjects. The same model will also provide a framework for describing the mechanisms that constrain verbal variation in the empirical corpus data analysed in Chapters 4—6.

3. History and origins of the NSR

The Northern Subject Rule is attested in the north of England since the Middle English period. Its developments since that time can be traced through texts of northern Middle English, Older Scots, and some documents of non-standard forms of Modern English. Unfortunately, however, its ultimate origins and earlier developments lie in the dark. The NSR seems to have emerged in early Middle English, a period from which no written documentation of the relevant dialects is extant. In the latest surviving northern Old English documents from before this gap, dating from the mid-10th century, no evidence of the NSR can yet be found. However, certain changes that can be regarded as a necessary corollary and prerequisite of its later appearance are well underway. In the earliest reliable northern Middle English documents from after the gap, dating from the 14th century, the NSR system is already more or less fully developed. The present chapter will first review the literature on the attested developments of the NSR from Middle English into the Modern English varieties. It will then take a step back and enter into a discussion of possible reconstructions of the processes that led to its first emergence. Some of this discussion will by necessity be speculative. I will critically review reconstruction proposals by Berndt (1956) and Klemola (2000) that have found some echo in the literature elsewhere, and in doing so I will also review some hypotheses about the role of language contact as an explanative factor in the emergence of the NSR. Against these hypotheses, I will sketch out a model of development that makes reference primarily to language-internal functional factors, and which is in principle independent of exterior, contact-induced causes.

3.1.

The development of the N S R from Middle English

The NSR is attested in more or less its modern form across all of northern England since late Middle English (Mustanoja 1960: 48If.). Its geographical distribution has most thoroughly been researched in connection with the Linguistic Atlas of Late Middle English (henceforth LALME; Mcintosh, Samuels/Bensik 1986; cf. also Mcintosh 1988, Samuels 1988, Laing 1988). (1)

Major Middle English verb paradigms

North Midlands South

3Sg. -es -eth -eth

3P1. NP -es -e(n) -eth

3 PI. pronoun -0 -e(n) -eth

The northern system was one of three major dialectal agreement patterns in Middle English (1). The north (including much of the northern Midlands) had -s in the second and third singular, and the -s/-0 alternation according to the NSR elsewhere. The rest of the Mid-

46 Map 1. The NSR in Late Middle English (after LALME 1: 467, IV: 1 lOff)

0

50

ι

100km

lands had -th in the third singular, and -en (later reduced to -e/-0) everywhere in the plural. The south had -th both in the third singular and in the plural. Map 1 shows the extent of the area where -5 was used in the third plural according to LALME. The southern boundary is more or less identical with the isogloss separating third singular -s from -th, forming the well-known Chester-Wash line. (A boundary running slightly further north was assumed by Moore/Meech/Whitehall 1935; cf. Laing 1988: 96.) As will be shown later on the basis of the SED data, this southern boundary of the NSR seems to have remained quite stable all the way into the traditional dialects of the mid-20th century. LALME does not give detailed accounts or statistics regarding the distribution across different syntactic environments, or any information about the first and second person in the documents studied. The only information given per document is whether -s forms

47 Map 2. Verbal -η in Late Middle English (after LALME I: 467, IV: 1 lOff).

were used regularly or rarely (LALME I: 298). However, it is assumed that the Type-ofSubject Constraint tended to be categorical, in the sense that adjacent pronoun subjects never took -s forms (LALME I: 554). Any residual variation in this respect, if it existed, was judged insignificant by the editors and has not been documented. Laing (1978: 244ff.) and Mcintosh (1988) draw attention to several types of intermediate systems that occurred along the southern boundaries of the NSR area. In all of them, an alternation between two forms was apparently governed by the same NSR constraints as in the northern system proper, but involved different morphological material. In much of the northern Midlands, i.e. the southern part of the NSR area, the opposition was not always between -s and - 0 , but commonly also between -5 and -n. These -n forms were like those typical of the Midlands dialects proper, to which they were geographically adjacent. Map 2

48 shows the total distribution of plural -n forms according to LALME. Its overlap with the area of the -s plurals represents the area of the intermediate systems. In some documents, a second similar pattern of alternation is found not between -s and - 0 , but between -th and - 0 . Mcintosh (1988) locates this system in a small area in the southeast Midlands, just south of the verbal -s isogloss, and geographically separated from the larger conservative area further south where -th was still the dominant plural ending throughout. Mcintosh argues that these -th plurals are not direct reflexes of the original Old English -(i)ad forms, but rather an innovation, an analogical re-introduction on the basis of the third singular forms and the neighbouring northern -s forms (cf. also Schendl 1996: 149, Laing 1978: 246). The precise nature of all these intermediate systems seems not yet well understood, and more research is clearly needed in this area. (2)

Major and intermediate Middle English verbal paradigms:

North N.Midi (1)1 N. Midi (2) N. Midi (3) Midi. South

3Sg.

3P1. NP

3PI. pronoun

-es -es

-es -es -e(n) -eth -e(n) -eth

-(e) -e(n)

-es/-eth -eth -eth -eth

-(e)

-e(n) -e(n) -eth

Whereas the emphasis in much relevant research on the Middle English varieties of England, in the context of the LALME project, has been on the dialect-geographical distributions, research on Scottish varieties has been mostly concerned with the language-internal details of the conditioning of the NSR. Besides the seminal discussion in Murray (1873), several recent studies of Older and Middle Scots (Meurman-Solin 1992, reprinted in 1993; Macafee 1993; and Montgomery 1994; see also King 1997) have given a fairly comprehensive picture of the traditional Scottish agreement system. Both the Position-of-Subject Constraint and the Type-of-Subject Constraint acted more or less categorically with lexical verbs in Scots, from the earliest attestations in the 14th century until the process of anglicization from the 17th century onwards. This included not only verbal -s in the third plural, but also quite regular usage of verbal -s with non-adjacent I, we, and you/ye. The old second singular pronoun thou, as in northern England, regularly took -s (more rarely also -st) affixes. Was/were and is/are came to be used in a pattern isomorphic with the NSR, with is and was analogically extended to non-third singular contexts, but only variably (Montgomery 1994). 1

The descriptive accounts in Laing (1978) and Mcintosh (1988) - both authors involved with the LALME project - do not agree as to how the variation patterned in cases where -5, -n and - 0 cooccurred. Laing (1978: 244) states for the south of Lincolnshire that -en endings were "current [...] beside -s" but "very uncommon when the verb immediately follows a personal pronoun, even in texts which have -n regularly in other positions." Mcintosh (1988: 117), with reference to the earlier study by Laing and without openly expressing disagreement, nevertheless states that "in the south of the N[orthern] area" the plural affixes were "often -en" instead of - 0 with adjacent pronoun subjects, and contrasted with regular -es after full NP subjects. While the first statement describes a dominant opposition of the type they sing: birds singen, the second description seems to imply a dominant opposition of they singen : birds sings.

49 In Early Modern English, plural verbal -th and -s temporarily appeared in patterns resembling the NSR in the emerging standard language, even outside the northern dialect zone. There are several older studies (Poliert 1881, von Staden 1903, Knecht 1911, Bambas 1947) dealing with -th and -s in the early literary language. More recent sociolinguistically oriented studies can be found in Bailey/Maynor/Cukor-Avila (1989), Bailey/Ross (1988), Montgomery/Fuller/DeMarse (1993), Schendl (1996, 2000), and Wright (2002). Schendl (1996) interprets the phenomenon as an analogical re-introduction of the affixes into plural environments on the model of the northern system. This would have been triggered by dialect contact through immigration from the north. Schendl demonstrates on the basis of demographical documentation that there was a massive influx of population from the north into London, especially during the 16th century (see also Nevalainen/Raumolin-Brunberg 2000a). The temporary rise and subsequent fall of plural -s in these areas coincides with the change of -th to -s in the third singular, which can also plausibly be seen as an effect of northern dialect contact (Stein 1987, Kytö 1993, OguraAVang 1996, Nevalainen/RaumolinBrunberg 2000b). Research in agreement variation in historical vernacular forms of Modern English in Britain has been fostered in the literature especially by an interest in the historical roots of overseas varieties. Montgomery (1988, 1992, 1997a, 1997b), Montgomery/Gregg (1997), Corrigan (1997), Filppula (1999), and most recently McCafferty (2003, 2004) discuss the linguistic influence of Scottish and English varieties on Ireland, especially Ulster. Another important line of research, which reaches beyond the scope of the present study, is concerned with the emergence of Afro-American Vernacular English (AAVE) (e.g. Schneider 1983; Bailey/Maynor 1985a; Bailey/Ross 1988; Poplack/Tagliamonte 1989; Montgomery/Fuller/DeMarse 1993; Montgomery/Fuller 1996 Wolfram 2000). Important studies of modern overseas varieties include Feagin (1979) on Alabama; Bailey/Maynor (1985b) on Texas; Eisikovits (1991) on Sydney; Schilling-Estes/Wolfram (1994) on North Carolina; and most recently Schreier (2002) on the isolated South Atlantic island community of Tristan da Cunha. A recurrent finding of all these studies is that subject-verb agreement variation patterns can be observed that resemble to some degree the historical NSR pattern of northern Britain. Comparative surveys and discussions of these studies can be found most recently in Tagliamonte (2002) and Chambers (2004). A question raised repeatedly in this recent research is to what degree the recurrent patterns found in geographically distant varieties reflect functional universals common to all English vernaculars, or to what degree each of the modern varieties can be explained on the basis of its direct historical heritage from the old British varieties. While for instance Corrigan (1997) and Chambers (2004) tend towards a universalist view, authors such as Montgomery and Tagliamonte have advocated a perspective that gives more weight to dialect-particular historical lineage. For instance, Montgomery/Fuller/DeMarse (1993: 337) stress the hypothesis that early AAVE is based "not on a mixture of dialects from the British Isles, but on a single variety", namely the Northern (Scottish-Irish) one. While for some modern overseas varieties, and perhaps also for the emerging Londonbased standard in Early Modern English, it can reasonably be argued that the observed NSR-like structures are direct historical reflexes of contact with originally northern varieties, there are some cases where such a connection is implausible. This is especially true for dialects in southwest England. Southwest England is the second major area in Britain beside the north where non-third singular verbal -s forms are a feature of traditional dialects.

50 These -s forms are ultimately reflexes of the southern Middle English -th forms. Unlike in the north, they traditionally occurred with all types of plural subjects, pronominal and nonpronominal. Nevertheless, some studies of present-day south-western varieties have found quantitative (though non-categorical) constraints, resembling the NSR, to be active in agreement variation (Godfrey/Tagliamonte 1999; Peitsara 2002): verbal -s is used at a significantly higher rate with full NP subjects than with pronouns. The apparently independent rise of NSR-like structures in these varieties strengthens the case for some dialect-universal mechanism which favours such structures - a "vernacular primitive" to use the terms of Chambers (2004). This mechanism may indeed be found in the cognitive principles of processing and entrenchment outlined above in Chapter 2.3: owing to their higher discourse frequency, adjacent pronouns will tend to be entrenched and represented independently from the more general subject-verb construction schemata, and may then easily become associated with some morphosyntactic behaviour differing from that of other subjects. Besides the inheritance of verbal -5 in the north and southwest, there is a third variation pattern that has an even wider geographical spread: levelling of the was/were contrast, most often in favour of plural was. This occurs also in many areas outside the traditional NSR area, which otherwise do not have plural verbal -s in lexical verbs (see for instance Schreier 2002, Britain 2002). Here too, patterns involving type-of-subject or position-of-subject effects have been found (e.g. Cheshire 1982). Within the core geographical area of the NSR, was/were variation tends to be aligned with the variation patterns in the other verbs, but there are also border areas where was/were variation follows idiosyncratic rules that cross-cut with those of the NSR. A prime example is the use of singular were in Yorkshire and Lancashire, but also in parts of Scotland and Ulster. Details regarding the modern dialects will be discussed in Chapters 4-6. About the historical developments of these special patterns, very little is known.

3.2.

The Origins: Old English affix reduction and analogical spread

I now turn to the question of the reconstructed earliest origins of the NSR. The first, crucial point to be repeated here is that the verbal -i forms in the northern system, in those environments where they do not match modern Standard English, are not a northern innovation. They were not, as one might think, an intrusion into these positions from the third person singular. Rather (for the lexical verbs, at least) these -5 forms are a conservative retention. Etymologically, verbal plural -s is no less genuine a reflex of an original agreement marker (northern OE plural -ad > -as > -5) than third person singular -s (northern OE -ed > -es > -s). What is new in northern system is really the introduction of the affixless forms in the pronominal environments. Seen in this light, the evolution of the NSR can be seen as part of the general drift of affix loss that affected all the Germanic languages. It may appear somewhat unexpected that it should have been the northern dialects that eventually preserved more of the traditional affix system than others, which eventually came to lose the plural affixes completely. After all, it is widely agreed that the trend towards affix loss was, on the whole, particularly vigorous in the north (possibly owing to a situation of intensive language contact with Scandinavian). But then, the apparent conserva-

51 tism of the northern dialects in this particular respect only applies to a last formal remnant of the affix system, and not to its actual function of agreement marking. Paradoxically, in keeping the affixed verb forms, northern English eventually preserved more of the phonological substance of the old agreement system than the standard did, but only at the price of having in effect less agreement. This paradoxical outcome may be interpreted as the result of a conspiracy of two independent developments originating in different dialects. One of them was the weakening and subsequent neutralization of a set of previously distinct but phonologically similar affixes (-ed/-ad/-iad/-is > -s). This development originated in the north and was well advanced by late Old English. The other was the innovation of affixless, so-called syncopated forms, at first only in a certain restricted set of syntactic environments adjacent to pronouns. This development was apparently headed by the southern dialects and only began to reach the north at some time during late Old English. At this point, the previous neutralization process had already brought the older, fuller agreement system to the verge of breakdown in that dialect, having obliterated almost all distinguishing contrasts in the verbal paradigm. The new affixless forms, the output of the second innovation, were therefore apt to be reinterpreted and pressed into service as carriers of a new agreement contrast. In the end, this led to a stable new system with an effective split of the once uniform morphological paradigm of subject-verb agreement into two distinct paradigms, valid for two complementary sets of environments: agreement after pronouns, and non-agreement after other subjects. This paradigm split can be seen as a major structural innovation in the history of English, comparable perhaps to the common Germanic split of the adjective inflection into the strong and weak paradigms (cf. Ferguson 1996: 183). In what follows, I will outline the hypothetical developments of the two different lines of innovation in more detail. A schematic overview of the development can also be found in Figure 3.1 on page 52. At the outset, the northern dialects of Old English shared the common Old English agreement system (Brunner/Sievers 1965: §352-378). The three persons in the plural had already been syncretized in pre-Old English and were -ad or -iad in the present indicative of lexical verbs, -en or -un in other paradigms. The singular forms in the present indicative were originally the common Germanic forms -u, -is, -id (in the case of the strong verbs), with the vowels undergoing a process of weakening to -e, -is, -ed during Old English. The endings in the weak classes were similar. Despite the vowel weakening, the plurals remained distinct from the third singulars for some time (strong verbs -ad versus -ed; weak verbs -iad versus -ed or -ad). In the course of the Old English period, the second singular added a -t affix after the -s, apparently owing to a reanalysis of a cliticized subject pronoun du. During the transition to Middle English, the three major dialect areas developed in different ways. The Midlands dialects replaced the plural -ad endings with -en, apparently through analogy with the past and subjunctive paradigms. Through later phonological reduction of the endings, these -en forms developed directly into the modern affixless forms of Standard English. The south retained the -d forms throughout, and replaced them with universal generalized -s in the modern dialects (Ihalainen 1994). The north was generally more progressive than the other dialects in the overall drift of affix reduction and loss. For instance, the phonological change whereby final -n was generally lost in all affixes was active in Northumbrian already during the Old English period,

52 Figure 3.1. Development of agreement paradigms in Old and Middle English

North

sing-e sing-es sing-es sing-es sing-es sing-es

sing Spread sing-s sing-s sing-s sing we sing-s sing j e sing-s sing öai

>

of -s to 1 Sg. we sing j e sing öai sing

Affix reduction sing-e sing-es(t) sing-eö sing-aö sing-aö sing-aö

sing-s I sing sing-s sing-s sing-s we sing sing-s you sing smg-s they sing

Northern Subject Rule

Spread of a-ffixless forms

Middle English

South

sing-e sing-est sing-eö sing-aö sing-aö sing-aö

Midi. repl. -d > sing we single

J Syncopation

ι

sing sing-est sing-eö sing-en sing-en sing-en

Modern English sing sing-est w sing-s sing sing sing

whereas other dialects followed only later during Middle English. Also, the weakening of vowel contrasts in unstressed affixes occurred earlier in the north than elsewhere (Berndt 1956). In the verbal paradigms, the north replaced the -d forms, both in the plural and in the third singular, with -s. This happened at some time during the late Old English period. The older northern documents, such as Caedmon's Hymn, the Leiden Riddle, and - with one notable exception - the runic inscriptions, all have the old -ό forms. 2 A later group of Northumbrian documents from the mid-10th century, the Lindisfarne Gospels, the Rushworth(2) Gloss, and the Durham Ritual, are witness to a stage where -ό and -s were used variably. The variation in these documents has repeatedly been the object of quantitative analysis (most thoroughly by Berndt 1956, but see also Holmqvist 1922, Ross 1934, Blakeley 1949 and most recently Stein 1986). These studies suggest that variation was conditioned, among other things, by phonetic environment factors, but also already by a slight tendency to treat pronoun subjects differently from full NP subjects; this in some way foreshadows the later Type-ofSubject Constraint. 2

The first, exceptional -s form occurs in the runic inscription on a cross in Urswick, North Lancashire, in the phrase gebidces per saulce. (Cf. Ross 1934: 68; Holmqvist 1922: 2, Page 1959: 402.) There is no certainty about the dating of this inscription. Estimates range from between 750 and 850 on the one hand (Page 1973: 144) to the early 10th century on the other (Collingwood 1927: 198; cf. also Collingwood 1911, Page 1959).

53 During the same time when -ö changed to -s, the vowels in the two endings also lost their contrasts. The new -st ending in the second singular was only a temporary intrusion in the late Northumbrian documents and was then again replaced by -s. The -e ending in the first singular became mute. Taken together, these changes meant that by the time of Middle English the present tense paradigms of lexical verbs contained only two distinct forms, - 0 and -(e)s. A hypothetical Late OE or Early ME stage can be posited in which all forms except the first singular had been neutralized in -5. (In the present subjunctive, neutralization had gone even further, due to the loss of final -n, so that only one form - 0 was left.) At about this time, the second of the two conspiring developments mentioned above became active. By the time of the first northern Middle English documents, only the second and third singular had retained the -s forms throughout, while new - 0 forms had spread into the plural and had already become near-categorical in pronominal environments. The most likely source for these new -s forms can be found in a development that apparently started a few centuries earlier, but in the south. Affixless forms conditioned by their syntactic position existed in one type of environment even in earlier southern (i.e. standard) Old English. In this variety, the original plural -ad and -en affixes were often syncopated before the postverbal plural subject pronouns we and je. In being sensitive to a pronominal subject and its position with respect to the verb, this rule already bears more than a superficial resemblance to the NSR. Brunner/Sievers (1965: §360) suggest that this reduction first occurred in the subjunctive. Here it would be easy enough to explain as an early application of the overall trend of deleting final -n. It would then have spread by analogy into the indicative. The affix reduction seems to have been triggered in a clitic environment. It has been commonplace in recent research on Old English syntax to assume that subject pronouns, both pre- and postverbal, acted as clitics (Pintzuk 1996, Kroch/Taylor 1997, Haeberli 2001). Within the usage-based approach pursued here, the tendency for cliticization itself finds a natural interpretation as a sign of a frequency-based routinization process, and the phonological reduction as a concomitant effect of this. These so-called syncopated forms remained restricted to the first and second person plural in standard Old English. Although this pattern seems to have had some effect still in southern and Midlands Middle English, it left no lasting mark on the language in those areas. In the late Middle English emerging standard variety, based as it was mostly on the Midlands dialects with their -n paradigms, the formal contrast temporarily introduced by the innovative affixless forms was neutralized again as soon as the -n was lost in the remaining environments too. Despite the evident similarity between this pattern of syntactically conditioned alternation and the NSR, an immediate historical connection between the two is not entirely unproblematic, because apparently the reductions with we and je occurred first and most frequently in southern, not northern Old English. In the three surviving Old Northumbrian texts which document the ongoing neutralization of the -s and -d forms, only isolated instances of the syncopated forms in the indicative are found. Berndt (1956: 213ff.) enumerates seven such forms in the first and second plural indicative from Lindisfame and Rushworth^), and nine from the Durham Ritual, plus 15 forms of plural imperatives (some of them with zero subjects) in all three texts. In the subjunctive plural the reduction of -en to -e is already the rule before we and je, and is also spreading into other subject-verb constellations in the Old Northumbrian texts, whereas the usage in the past tense paradigms differs between texts and between groups of verbs. The present indicative plural of the copula aren

54 is also often reduced to are in the same positions (Berndt 1956: 299ff.). 3 That the syncopated Old English forms were the direct source of the later affixless forms of the NSR was assumed explicitly by Rodeffer (1903: 33-36), and implicitly already by Murray (1873: 213). However, Berndt (1956: 216) argues strongly against this connection, and ever since, to the best of my knowledge, the question has not been taken up again. Schendl (1996: 157), citing Rodeffer and Berndt, apparently as being the latest studies on the topic, remains indecisive. As Berndt's decided verdict may have played some role in silencing the debate and blocking further exploration along this line, his argumentation deserves some scrutiny at this point. Berndt pleads for a model according to which the old syncopations were first eliminated completely, together with the -d forms, in favour of a complete unification of the indicative forms in -es. Affixless forms were then re-introduced in a new, unrelated development, leading to the NSR. His objections are partly of an empirical, partly of a theoretical nature. The empirical argument is that, of the three late Northumbrian documents, the Durham Ritual has more syncopated forms than the others, but is otherwise more conservative with respect to the d > s and other changes. Indeed, the syncopated forms represent only approximately 1% (7 out of 604) in the Lindisfarne and Rushwort 2 texts, but 10% (9 out of 88) in the Ritual, according to his figures. This comparison is enough for Berndt to claim that the welje syncopations in late Northumbrian must also be a conservative feature; they must have been on their way out, not on their way in. The theoretical argument is based in Berndt's conception of functional motivation in language change. Berndt sees language change as driven by "purposeful tendencies" ("Zweckstreben"), an eternal inner drive for economy and expressivity (1956: 11 et passim). He locates this drive not so much in the purposeful behaviour of individual speakers as many modern pragmatics-driven theories of language change would do (see e.g. Keller 1990, Hopper/Thompson 1993, Croft 2000) - but seems to conceive of it as a teleological trend immanent in a language system as a whole. In this sense, the main thesis proposed throughout his book is that the reduction and syncretism of the agreement affixes in Old English, as a move towards economy, was triggered because they had become functionally redundant after the rise of subject pronouns as the routine way of expressing person reference. It was therefore the 'goal' of the change to unify the whole present indicative paradigm in just one form, -(e)s, which then acted as a mere tense-mood marker. In contrast to this, a spread of the affixless forms would have led to a systematic syncretism not only within the indicative paradigm but also between the indicative and subjunctive. But this formal distinction, as opposed to the person-number distinction, was at that time still a functional necessity. It later suffered the same fate as the person-number system, being made functionally redundant by a new analytical construction type. This happened when the new system of analytical modal verb construction in later stages of English provided an alternative means to express modality. The rise of the modal verb constructions, Berndt concludes, was to become the causative factor for a second wave of affix reduction, which led to the NSR in early Middle English (1956: 21 Of.). As in late Old English this new alter3

There are also a handful of textually dubious forms (apparent scribal mistakes etc.), and 45 reduced ambiguous forms in positions other than before we and je. which translate Latin indicative forms but in the context of the Old English text could be explained as optatives. Of the latter, 28 are in positions that would warrant the reduced form in the later Middle English NSR system (Berndt 1956: 215).

55 native was not yet in place, the mood distinction was still carrying its functional load. Hence, a systematic syncretism between indicative and subjunctive would still have been "at cross-purpose" ("zweckwidrig", 1956: 211) with the primary teleological goal of unifying the present indicative paradigm. Berndt, in short, finds it hard to accept that two changes whose 'goals' are partly in conflict with each other could be active at the same time. This is why he refuses to accept any historical continuity between the original syncopated forms and the later process that would lead to the NSR, and believes that between these two stages the syncopated forms must temporarily have dropped out of use. I would like to argue that Berndt's argument does not stand up to scrutiny. First, current theories of language change should have no problem in principle with the idea of partly conflicting trends in linguistic change going on at the same time. Since they see the locus of language change in the pragmatics-driven behaviour of the individual speaker, the kind of economy invoked by these theories is a matter of local, not global optimization. Second, Berndt's argument is a logical non sequitur. Even if one grants that the process which led to the original emergence of the Old English syncopated forms and the process which led to the rise of the later NSR were governed by different functional causes, and that these functional causes in turn were incompatible with the cause that governed the intervening process of the ό > s change, it does still not logically follow from this that the syncopated forms could not have survived (if only marginally) and guaranteed some historical continuity between the two phenomena. The syncopated forms, as the output of the earlier change, could very well have provided the input for the later one, even if the two changes were different processes and even if a third conflicting change had intervened in the meantime. As regards the textual findings, there is a straightforward solution to Berndt's difficulty if one sees the difference between the Durham and the Lindisfarne/Rushworth texts as a reflection not only of a diachronic development but also as a sign of geographical and/or stylistic variation. The Lindisfarne and Rushworth texts are generally agreed to be more advanced in the specifically northern innovations, but the syncopations were originally a southern innovation. It is therefore not necessary to conclude from the evidence that the use of the syncopations was declining in the north. It may just as well have been in a stage of its first spreading into that area, and hence increasing. In that case, the Durham text, being more strongly influenced by southern dialects than then other two, would simply be more strongly reflecting the advance of a southern feature which is not yet reflected with the same strength in the more purely northern dialect of the other two texts. (3)

Hypothetical northern late OE / early ME verb paradigm Indicative

Subjunctive

preverbal pronoun subj.

postverbal pronoun subj.

non-adjacent or Ν Ρ subj.

Sg

1 2 3

ic binde du bindes he bindes

binde bindes bindes

ic du he

... binde ... bindes ... bindes

PI

1 2 3

we bindes j e bindes dai bindes

binde binde bindes

we 3e öai

bindes bindes bindes

ic öu he

binde binde binde

we binde ye binde öai binde

56 All in all, whatever the development of the Old English syncopated verb forms may have been up to the time of late Old Northumbrian, it seems safe to say that at the time when the development towards the NSR pattern started, the syncopated forms with enclitic we and were available at least as optional forms in the dialect and could have served as a trigger for further analogical spread. A hypothetical system of northern latest (post-Lindisfarne) Old English or earliest Middle English, assuming the existence of the we/ye syncopations, is shown in (3). A natural extension of the we/je syncopations, based on syntactic and semantic similarity, would obviously be to the third plural. The problem here is maybe not so much to explain why this extension happened, but rather why it did not happen earlier and also in the South - given that identity of forms across all persons of the plural was the rule everywhere in the Old English verbal system. Stein (1986: 645f.) draws attention to the fact that the third plural tended to be rather more conservative than the first and second plural also in some other processes of change in Old English, and that it would be typologically natural for the first and second persons to form a class separate from the third person on grounds of semantic similarity (quoting cross-linguistic evidence from, among others, Benveniste 1966, Moravcsik 1978). But still, why did the northern dialects come to behave differently from the southern ones? One straightforward part of the explanation may be related to the fact that the pronouns themselves were different. The third plural pronoun form was originally hlo/heo, and remained so in the southern dialects well into Middle English. But the northern dialects were replacing hlo/heo with the Scandinavian dai at about that time. The difference in the phonological shapes of the two pronouns - the one starting with the phonetically highly salient consonant / δ / , the other with the much less salient / h / - may well have played a role here: in the junction between the verb and the clitic, a cluster / s 3 / (as in northern bindes dai) would arguably be somewhat more likely to drop the verbal / s / than a cluster / ö h / (as in southern bindad hie) would be to drop the /Ö/. Thus, the greater phonological naturalness of this reduction, together with the analogical pressure from binde we and binde ye, may account for the fact that reduced verb forms spread into the third plural in the northern dialects. Parallel to the spread from the first and second plural to the third plural, it must also be assumed that the new reduced forms spread from the postverbal subject environments to the environments with preverbal pronoun subjects. Which of the two changes occurred earlier is of no importance. What is important is that the reduced forms spread to the environments with pronoun subjects but stopped short of those with non-adjacent pronouns or NP subjects. No definite explanation can be given for this, but again, assuming that verb-pronoun combinations tend to be strongly entrenched and may possess separate unit status in grammatical representation, owing to their high discourse frequency, helps to explain why these combinations can easily come to behave differently from other, less frequent, subject-verb constellations. The functional motivation for the quick and thorough adoption of the affixless forms seems to have been that they were co-opted to serve as genuine agreement forms, reintroducing and stabilizing a formal contrast that was no longer signalled reliably by the older set of affixes. To complete the full development of the NSR, two additional changes were necessary. First, the first person singular - which had possessed no consonantal affixes from the outset - introduced new and unetymological -s forms in positions not adjacent to the subject. This obviously happened by way of analogy with the other persons and

57 numbers, as the (old) zero ending of the first singular became treated as equivalent with the (newer) zero endings in the plural (cf. Holmqvist 1922: 49). Then, also by way of analogy, the usage of is/are and of was/were was partly aligned with that of the other verbs, as was and is gradually began to be used in plural environments in the same way as the -s forms of other verbs (Montgomery 1994). At this point, the 'ideal' Northern Subject Rule system as described in definition (1) on page 5 above had been reached.

3.3.

The Scandinavian hypothesis

Given the cloud of uncertainty that surrounds the beginnings of the NSR, it is perhaps not surprising that researchers have taken recourse to language contact as a hypothetical causative factor in its development. Indeed, two quite different languages and two quite different historical periods have been invoked in this context: Celtic (Brythonic) during the time of the Anglo-Saxon settlement and expansion; but also Scandinavian (Old Norse) during the time of the Danelaw. The latter contact situation has been discussed not so much with regard to the NSR proper but with regard to the concomitant d > s change. But since this change can plausibly be seen as a necessary prerequisite to the emergence of the NSR, the Scandinavian contact may be seen as having played an at least indirect causal role in its respect. Whereas the idea of Scandinavian influence is appealing for its historical timing and its obvious fit to dialect-geographical data, the Celtic substrate idea owes its attractiveness to certain apparent typological parallels between the NSR and Celtic agreement patterns. In reviewing both hypotheses, I shall argue that the Scandinavian contact offers a possible partial explanation for some of the prerequisite changes that led the way to the NSR, thus qualifying as at least a plausible indirect causal factor. The Celtic hypothesis, on the other hand, meets with serious difficulties regarding the temporal and logical order of the various developments involved. Since it was first proposed (Keller 1925), the idea of a Scandinavian influence on English verbal -s has often been met with considerable scepticism (for a survey of the older literature see Brunner 1962: 177), but it seems to have gained ground again in recent decades in the light of current language contact theories (Bailey/Marold 1977: 45, Thomason/Kaufman 1988, Stein 1986; for a recent sceptical view see Ferguson 1996: 178). The basic argument as proposed first by Keller was that Old Norse would have offered a model of analogy for the spread of -s from the second singular to the third singular. Old Norse had syncretized the second singular and third singular a short time earlier, with both forms ending in fYU - a sound developed from earlier Germanic /z/, which at the time of the Viking settlements would have been realized as a voiced apico-palatal fricative [3] (Noreen 1923: §530, Haugen 1984: 197). Before c. 1100 this sound merged with Μ in Old Norse, but at the time of contact it would likely still have had a sibilant quality perceived as similar to /s~z/ by English speakers. The identity of this s-like ending across the second and third singular in Old Norse would have been an obvious basis for an analogical spread of -s from the second singular to the third singular in English. There are several difficulties with this argument. One is that at the time when -s appeared in the third singular in Northumbrian, its supposed source, the second singular, was

58 itself moving away from -s in an apparently unrelated development, the change from -s to -st. At least this is the picture deriving from the surviving Northumbrian documents (cf. Brunner/Sievers 1965: §356). Given the non-survival of -st in later northern English, one might counter this argument with the hypothesis that their presence in the documents may possibly be only an intrusion of an orthographical practice taken over from the south and not necessarily reflecting the local dialect. A more serious problem is that the Scandinavian parallel does not provide an explanation for the spread of -s also to the plural. Old Norse still had separate distinctive endings for all three persons in the plural and would have offered no model for syncretism in this domain. Most authors agree that if Scandinavian influence was responsible for the -d to -s change, one would at least expect the new -s forms to appear earlier and more frequently in the third singular than in the plural. However, in the surviving Northumbrian documents, the exact opposite is the case. According to all quantitative evaluations of the Lindisfarne and related texts (for references see p. 52 above) the -s forms are significantly more frequent in the plural than in the third singular. All of these studies except Berndt (1956) - who comes to different results owing to different ways of weighting the evidence from the three existing texts (cf. Stein 1986: 639f.) - also agree that within the plural the second person stands out as having the most -s forms. This leads to the hypothesis of a spread first from the second singular to the second plural, based on semantic closeness of these categories, and then into the other plural forms. (Incidentally, the earliest isolated recorded form of verbal -s, in the Urswick inscription, also happens to be a plural imperative, a form that is usually identical to the plural indicative in all forms of Old English.) Samuels (1985: 111), the chief proponent of the Scandinavian hypothesis in the more recent literature, dismisses the Lindisfarne evidence regarding a priority of plural over singular, on the grounds that it was written near Durham, outside of what he identifies geographically as the "focal" area of the "Great Scandinavian Belt", and that the origins of the -s forms must be sought a good deal earlier, and within that focal area. The Lindisfarne data can therefore be expected to reflect only the idiolect of its scribe, into whose own local variety the -s forms had presumably begun to spread only at a later stage, and possibly in a different order. But this is begging the question: Samuels wants the -s forms to derive from Old Norse, so, for him, any text not originating in the direct vicinity of Old Norse cannot reflect the true origins of the -s forms. A stronger argument (ibid.) for the historical priority of the third singular -s over the plural -s, and hence for a direct Norse influence, is that the singular -s had spread across a slightly larger area in Middle English than the plural -v (cf. Stein 1986 and dot maps in LALME I.) Another, not uninteresting, possible Scandinavian factor that could have played a role in favouring the -s endings, according to Samuels, is the existence of the innovative MedioPassive paradigm in Old Norse, derived from fusion of the verb with the reflexive pronouns mik/sik (cf. Noreen 1923: §542f.). Samuels explains (1985: 112): Here, therefore, was a radical difference of conjugation between the two languages, and, in situations which assumed some degree of mutual intelligibility, this feature was bound to be misinterpreted by one side or the other. An obvious source for such misinterpretations would exist where Old English intransitive verbs corresponded to Old Norse reflexives, e.g. OE hie sittad 'they sit down', ON peir setjask; a probable compromise form would be hie sittas, as in late Old Northumbrian.

59 The debate about d/s and Scandinavian, which started in the earlier 20th century, was largely concerned with traditional concepts such as the Neogrammarian dichotomy of analogy versus sound-change as explanatory factors in language change. It is in the different research context of modern language contact studies that several authors besides Samuels have more recently taken up the issue again. According to this approach, it is not so much specific similarities or analogies that are assumed to have triggered a specific change. Rather, it is widely seen as a general condition of contact situations that morphological systems and their phonology tend to become simplified in the direction of markedness reduction. Thus, ö > s is seen as a typical leniting phonological change (cf. Lutz 1991) which would be rather natural to occur in such a situation. Bailey/Marold (1977: 45), in the context of their well-known Creole hypothesis, cite the neutralization of the verb forms in -es an example of the "general Creole tendency to simplify morphology". Stein (1986: 648), more reluctantly, sums up: Just how strong the effect of language contact was, and to what extent something approaching real creolization has to be assumed is a matter of debate, but it seems reasonable to suggest that the Scandinavian language contact has helped along or reinforced already independently existing drifts or developments, and has given them their specific Northern twist.

Ferguson (1996: 178) finds the contact idea likewise "tempting to hypothesise", since on considerations of markedness and phonological naturalness the d> s change would be "typical of transfer in second language acquisition or bilingualism". However, he finally dismisses it as unlikely, primarily because Old Norse itself still had -d affixes elsewhere in its morphological system. The interest in the idea of phonological markedness typical of this line of research, drawing on the evidence collected from contemporary variation studies, also lends new weight to findings about the phonotactic conditioning of the change. Blakeley had shown as early as 1949 that the distribution of the two ending types in the late Northumbrian texts is heavily constrained by the stem-final consonants preceding it (cf. Stein 1986: 648). While a Scandinavian influence on the -es affix neutralization is functionally plausible but still controversial in its historical details, none of the authors quoted here has placed the Scandinavian influence in direct connection with the evolution of the NSR proper. Scandinavian offers no direct model for a split of agreement constructions according to type or position of subject; nor would a split such as this generally be considered a simplification typical of contact situations. This split, which in some sense adds complexity to the morphosyntactic system instead of reducing it, still stands out within the overall development, similar to how Ferguson (1996: 189) characterizes the innovation of the Germanic strongweak contrast in the adjectives: as "the occasional clear instance of an innovation that goes contrary to the main drift and persists along with it for long periods of time." Nevertheless, if one accepts the idea Scandinavian influence on the development of verbal -s, in whatever form, then one may indeed plausibly see the NSR as an indirect long-range effect of the Scandinavian influence.

60 3.4.

The Celtic hypothesis

Whereas the Scandinavian hypothesis finds obvious support in the historical timing and space of the events, and in the undoubted strong presence of other Scandinavian effects on English that date from the same historical situation (Thomason/Kaufman 1988), the case for Celtic presents more historical difficulties. It seems to have been a consensus in research up until recently that very little, if any, substratal influence from Brythonic on English is discernible in England (e.g. Crystal 1995: 8). The ousting of Brythonic by Anglo-Saxon is thought to have been the effect either of wholesale physical replacement of populations, or of a swift and almost total language shift of the conquered Britons to the language of the conquerors. The conspicuous lack of Celtic loanwords in English is taken as proof that the Anglo-Saxon conquest was not followed by any profound Celtic substrate influence. Some authors have recently attempted to revise this received picture (for a survey of the literature, see Filppula/Klemola/Pitkänen 2002, Tristram 1999). Klemola (2000), the main proponent of the Celtic substrate hypothesis regarding the NSR, relies on historical evidence collected by Higham, who in a series of publications (1992, 1993, 1994, 1995, 2001) has proposed the thesis that there was "massive ethnic continuity from late Roman Britain to Anglo-Saxon England" (1992: 209), and that speakers of Brythonic continued to live side by side with Anglo-Saxon speakers for a relatively long time, especially in the north of England. Tristram (1999: 12) quotes archaeological evidence from Laing/Laing (1990) and Härke (1995a, 1995b, 1998, 1999) that points into a similar direction, suggesting that Anglo-Saxons were only a rather small aristocratic minority of the total population after the conquest. Filppula/Klemola/Pitkänen (2002: 5) sum up that by now a "widespread consensus among historians and archaeologists" has evolved which must lead linguists to reckon with the possibility of relatively peaceful and long-standing co-existence between the two populations. They believe that "the time is ripe for a critical reassessment" of the question of Celtic influence in the development of the English language (2002: 26). The argument about the lack of loanwords is countered by referring to the language contact theory of Thomason/Kaufman (1988; cf. also Sankoff 2002), which would lead to the expectation that in a substrate to superstrate shift situation, phonological and structural rather than lexical features are likely to be transferred from the substrate to the superstrate language first. Several authors have recently proposed instances of innovative phenomena in Old and Middle English that have parallels in the Celtic languages and could therefore be taken as examples of such features (Poussa 1990, Hickey 1995, Vennemann 2001, 2002a, 2002b, as well as other papers in Filppula/Klemola/Pitkänen eds. 2002; see also Tristram 1997a, 1997b, 1999). One of these phenomena is the NSR. The idea of a link between the NSR and the Celtic languages first appears in the literature in a brief remark by Hamp (1976: 73), who finds that the NSR "looks for all the world like" an effect of substrate influence, and who specifically locates its source in the Brythonic varieties of Cumbria and Strathclyde. His idea is quoted by Macafee/0 Baoill (1997: 279), and elaborated into a stronger but still tentative proposal by Klemola (2000). His argumentation in turn is accepted by Tristram (1999) and Vennemann (2001). Anglo-Saxon settlement had progressed up to a line reaching diagonally from the Severn to the Humber, as well as into areas along the east coast of Yorkshire and Northumberland,

61 by c. AD 600 (Jones 1978: 60, Jackson 1953: passim). West Yorkshire, Lancashire, Westmorland, Cumberland, Durham and Northumberland were settled by Anglo-Saxons only from the mid 7th century onwards. According to an optimistic estimate, a bilingual stage in these northern areas could have lasted "until the late 8th century, perhaps even later in parts, especially in Cumberland" (Klemola 2000: 340f.). In a similar vein, Tristram (1999: 16) gives a time span of "more than six generations" in which pockets of Brythonic speakers could have persisted in the northwest. This would still place the time of EnglishBrythonic contact in that area well before the English-Scandinavian contact, which started in the late 9th century. As Klemola recognizes, this timeline leads to an obvious problem: the effects of the NSR surfaced in the written record only several centuries after the supposed time of contact. Klemola (2000: 341) counters this with the observation that innovations quite often take a lot of time spreading from the spoken language into the written register, and that the same is true for some Scandinavian-induced changes (cf. also Tristram 1999: 27ff. with a similar argument). But the timing problem is more severe than this, because it is not just a question of the absolute dating of the NSR alone, but also a problem with regard to the logical and temporal order between the NSR and other innovations, especially the d > s neutralization and the innovation of the affixless forms. The Celtic hypothesis rests on the fact that the distribution of English 'singular' forms (bindes) and 'plural' forms (binde), from Middle English onwards, displays a parallel to the distribution of singular and plural forms in Welsh (and its reconstructed parent, Brythonic). Simplifying somewhat, the Brythonic agreement system uses synthetic verb forms, fully inflected for person and number, whenever the clause has either no overt subject NP or only a weak personal ('echo') pronoun. In all other cases, notably with plural subject NPs, an unmarked third person verb form is used. (For the syntax of verbal agreement in Old Welsh cf. Lewis/Pedersen 1961: §343, 398f., 433ff; Evans 1971; for a general typological comparison between Welsh and English see Tristram 1999.) This parallel, however, is only valid under the condition that bindes is indeed a singular, and binde is indeed a plural. This is less obviously the case than it may seem, given the 'markedness paradox' discussed so often before. Given what was discussed in the preceding sections, the English forms simply did not yet have the semantic values required for the parallel to work, at the time when English-Celtic contact was in its critical stage. The form bindes, according to all available evidence, was not yet in existence, and its various direct predecessors, the -is, -ed, -ad and -iad forms, were a mixed bunch of both singulars and plurals. The form binde was likewise as yet non-existent in its present function. That the bindes-binde contrast should express a singular-plural number opposition can only reasonably be claimed after the process of markedness reversal that came about through the neutralization of the -is, -ed, -ad and -iad forms in -es, and the spread of the new .s-less forms into the pronominal plural environments - that is, apparently, after the time of the late Old Northumbrian documents. In short, the necessary conditions under which a Celtic analogy could possibly have led to the outcome observed in Middle English were given only after the Viking period, not before. If a tendency to directly transfer the Brythonic agreement rules had been active at an earlier stage in the history of Old English, before the -es neutralization, we should expect, if anything, an outcome opposite to that observed in Middle English. If speakers had been introducing non-agreement into plural noun phrase environments, then it should have been

62 the noun phrase environments where a change would have manifested itself first - we would expect a new, neutralized form (possibly taken over from the third singular, or affixless) to be introduced in the noun phrase environments, and the original, genuine inflected plural forms to be retained in the pronominal environments. It is easy to see that such a system - if it should ever have existed - could not then possibly evolve into the NSR system as it is attested later. And, in contrast with the prediction just made, it was consistently the pronominal environments that were more progressive than the noun phrase environments in all the changes actually observed. First, the pronoun environments were apparently quicker to accept the δ > s change (Berndt 1956); and subsequently it was again the pronoun environments that first replaced the new -s forms with the even newer affixless forms, not the other way round. To defend the Celtic hypothesis, there are therefore in principle three logical possibilities: 1.) claim a later date for the Celtic contact, following or overlapping the Scandinavian one; 2.) claim an earlier date for the vowel neutralization, the δ > s change, and the availability of the affixless 'plural' forms; 3.) claim that early Celtic contact had left some other, hidden, mark on the English language system, which became active and indirectly caused the NSR only after an incubation period of several centuries. For hypothesis one, it would be necessary for Brythonic speakers to have persisted in northern England longer - much longer - than suggested by even the most optimistic estimates, until at least well into the Scandinavian period. Whereas for Scotland a long presence of Celtic (including the Brythonic branch) is trivially true, there seems to be little evidence for this regarding northern England. The dialect-geographical evidence also provides no arguments for the possibility of an origin of the NSR in Scotland and a later spread southwards into England. Hypothesis two would force us to assume that the neutralizing changes and the introduction of the new affixless forms happened earlier - much earlier in the spoken language than seen in the surviving documents. Without exception, all documents (both the older ones that still have the old endings, and the younger ones that apparently show the neutralization in progress) must then be very far away from reflecting the spoken language of their respective times. Exploring possible scenarios for hypothesis three can, at the present moment, be little more than pure speculation, and is beyond the scope of the present work. I would not wish to deny the possibility that early Celtic-English contact could have brought about some change in English which later, indirectly, contributed to causing the rise of the NSR. However, that the evolving NSR system then came to be actually isomorphic with the Brythonic one would still be a chance effect. To sum up, it seems the most likely that the developments that led to the rise of the core regularity of the NSR can be explained primarily through language-internal factors, by recourse to such time-honoured concepts of historical linguistics as syncretism, affix reduction, and analogy. All of these find a natural interpretation in a usage-based view of grammatical representation (cf. Bybee 1985). While language contact with either Brythonic or Scandinavian may or may not have played some indirect role in these developments, language-internal dialect contact between the north and south seems to have been a necessary factor, regarding first the spread of the affixless forms into the north and later the spread of the -s forms back into the south. The processes of dialect contact, levelling and competition between various cross-cutting patterns that ensued are the ultimate source of the complex patterns of variation found in the modern dialects, which will be presented and analysed in the following three chapters.

4. Verbal agreement in the SED

This chapter deals with verbal agreement variation in the traditional dialects of the English North and Northern Midlands as reflected in the Survey of English Dialects (SED). Verbal -s following the pattern of the Northern Subject Rule can be found in the SED data in a large area covering all of northern England and much of the Midlands. In large parts of these areas the pattern defined by the NSR is modified by, or competes with, other local non-standard patterns of agreement. Some of these apply to the morphological paradigms of BE, in either the present or the past tense; some apply to the agreement forms taken by certain pronouns (such as thou)\ and some (such as Midlands verbal -ri) affect the agreement paradigms of verbs in general. The following sections will first describe these different non-NSR agreement phenomena, as a background for the subsequent discussion of how the NSR itself interacts with them. Special attention will be given to the phenomenon of only partly isomorphic agreement patterns of lexical verbs on the one hand and the forms of BE on the other hand. It will be shown that forms such as is or was may often, within one and the same dialect, be sensitive both to rules that treat them as verbal -s forms along with all others (such as the NSR), and to other rules that affect only them individually and may partly override the former (for instance, local patterns licensing forms such as / is, he were, or they was). It will also be shown that in those areas where the NSR competes with the standard agreement pattern under conditions of dialect levelling or attrition, plural verbal -5 tends to become restricted to certain favouring environments. Very much the same favouring environments will be encountered again in the more recent corpus data in Chapters 5-6: most of all, subject relative clauses, as well as clauses with certain indefinite or demonstrative pronoun subjects. Finally, it will be shown that when these special clause types - especially as they are documented in the original, unpublished fieldworker notebooks - are taken into account, the geographical area affected by the NSR turns out to be larger than indicated by previous studies based only on the published material of the SED. The non-NSR agreement patterns that are to be dealt with here are in principle well known and are often summarily mentioned with reference to the SED in the literature (e.g. Wakelin 1977, 1984, Trudgill 1990, Ihalainen 1994). Much of the relevant SED data has also long been easily available in the form of atlas maps. Nevertheless, it will be useful to include a description of these phenomena in the sections below as a background, before proceeding to the analysis which regards the NSR proper. The focus of this analysis will be on how the NSR interacts with the other competing patterns in different areas. The resulting findings concerning favouring and disfavouring effects on verbal -s in different syntactic environments can then be used as a background for the corpus-based study of some more recent varieties spoken in the same geographical area (Northern England) as well as elsewhere (Scotland and Northern Ireland).

64 4.1.

The data

Data from the SED can, for several obvious reasons, not be analysed with the same quantitative methods and in direct comparison with corpus data such as that from the NITCS or FRED. For one thing, the SED data consists of lists of isolated elicited tokens, not exhaustive records of all forms produced in natural speech,1 and hence does not lend itself to a quantitative assessment of intra-speaker variation. Second, as is well known, the fieldwork for the SED was conducted during the 1950s, and hence reaches back a real-time span of almost a generation before the NITCS, and even more in comparison to FRED. Nevertheless, the SED data provides valuable evidence for the geographical distribution of morphosyntactic variables in the traditional dialects, and, if interpreted with the necessary caution, also for some of their quantitative aspects (for an example of a quantifying approach to its use, cf. for instance Ihalainen 1991a). It has also repeatedly been used for studies of subject-verb agreement. Some of the data relating to verbal morphology has long been accessible in the form of maps in either of the two SED-based atlas publications - the Linguistic Atlas of England (Orton/SandersonAViddowson 1978, henceforth LAE), and the ComputerDeveloped Linguistic Atlas of England (Viereck 1991/1997, henceforth CLAE). Some references to subject-verb agreement data from the SED can also be found in Trudgill (1990), Ihalainen (1991a, 1994), Klemola (1996, 2000), Bresnan/Deo (2001), Wright (2002), and Britain (2002). However, none of these studies has fully exhausted the possibilities of the SED with regard to a study of the NSR and related phenomena. The most detailed treatment so far is that by Klemola (1996: 54-60), who analyses the geographical distribution of NSR as a side-issue to a much more detailed investigation of verbal -s in the Southwest. His study is based on the evaluation of the Basic Material for twelve SED items. The empirical findings regarding the NSR have been re-published in similar form in Klemola (2000). The SED fieldworkers recorded, apart from the answers for the systematic elicitation items, a large number of additional utterances produced spontaneously by the informants during the interviews, whenever they felt these utterances illustrated interesting dialect features. These recorded tokens are known as "incidental material" in the SED. In cases where the features illustrated by these utterances were also the topic of one of the systematic questionnaire items, incidental material evidence was often included in the answer lists for the respective headings in the published "Basic Material" of the SED (Orton et al. 19621971, henceforth SED-BM). However, the coverage of the incidental material in SED-BM, especially in the case of morphosyntactic variation, is far from complete. This means that a wealth of additional data is still enclosed in the original hand-written fieldworker notebooks and has never become available in publications. This is mostly due to the fact that many morphosyntactic phenomena were not specifically covered by any questionnaire item. In the case of subject-verb agreement, and especially of NSR plural verbal -s, the gaps in the published material turn out to be substantial, and for this reason the geographical picture that has so far been derivable from SED-BM alone is in some respects distorted. For in1

From many SED locations, tape recordings of some stretches of spontaneous dialogue are also extant. At the time of writing, this material had been transcribed and was being prepared for publication as an electronic corpus by J. Klemola. Unfortunately this valuable material did not become available in time for being included in the present study.

65 stance, there is one questionnaire item (III. 10.7, bulls bellow) under which the editors have included incidental material tokens for plural verbal -5. This list in itself is reasonably complete, but unfortunately it covers only tokens that match the elicited grammatical context in a rather narrow sense: namely, subject NPs headed by lexical nouns in canonical SV clauses. Some of the most interesting grammatical environments for the NSR are therefore missing: relative clauses, verbs after plural demonstrative and indefinite pronouns, to name but a few. Many incidental material tokens that illustrated verbal -s in these environments are either scattered across a large number of other headings in SED-BM,2 or have up to now not been accessible at all. To remedy this situation, it was necessary to return to a study of the original fieldworker notebooks archived at the Special Collection at Leeds University Library. This investigation was conducted for 139 of the 311 SED locations, covering the northern half of England roughly down to the Chester-Wash line and somewhat beyond. All incidental material tokens relating to subject-verb agreement were excerpted from the notebooks, converted from the original phonetic to an orthographic transcription, and later collated with the data found under various headings in the printed Basic Material. The resulting token lists, which form the basis for the analyses in the following sections, can be found in Appendix Β beginning on page 160. They include approximately one thousand tokens ofNSR-related verbal -5 and another thousand tokens representing various other phenomena. Although this sheer amount of material may appear like a sufficient quantitative basis even for a study using corpus analysis techniques, a good deal of caution about its representativity is required in the analysis. The utterances recorded are obviously only a small fraction out of the many hundred hours of conversation that took place between the fieldworkers and the informants, and the selection of this material must therefore be highly subjective. The fieldworkers only noted down such utterances as drew their attention because they felt some salient dialectal feature was represented in it. There is hardly any indication about how often informants did not use these forms; for instance, how regularly they used Standard English forms instead. Likewise, it may well be the case that certain linguistic variables had less salience for the fieldworkers' attention than others and may therefore be underrepresented. Finally, so-called fieldworker isoglosses - systematic differences in the recorded data due to different data collecting practices of individual fieldworkers - are a serious problem in the geographical analysis (cf. also Klemola 1996: 39).3 2

3

The richest single source of such tokens is the collection of relative clauses for questionnaire item IX.9.5 (I know a man who will do it for you.) Note that the sentences actually elicited for this item, being singular, are irrelevant with respect to the NSR. More than half of the locations covered in this study were done by one fieldworker, S. Ellis. He was responsible for all the fieldwork in the four northernmost counties, as well as in Cheshire, Derbyshire, Nottinghamshire and Lincolnshire. The fieldwork in most locations in Lancashire and Yorkshire was done by P. Wright, and the rest of the Midlands was divided between A. Playford, D. Sykes and P. Gibson. The fieldworker problem most directly affects the bare amount of incidental material collected, as can be ascertained from even a cursory glance at the questionnaires. On average, Ellis tended to record only about half as many tokens per location than Wright did in the same geographical area. This clearly relativizes Orton's statement that all fieldworkers "made a point of collecting as much of this material as was feasible in the situation" (Orton et al. 1962: 17). It is, however, true that all fieldworkers were clearly aware of the dialectal importance of verbal agreement variation in general, and of the NSR in particular, and it seems safe to assume that

66 The material excerpted from the fieldworker notebooks was used to complement a systematic evaluation of the data printed under a variety of headings in SED-BM, comprising in turn both elicited and incidental tokens. Most of the SED items in question are those dealing with the morphology of the primary verbs BE, HA VE and DO. Some others deal with the agreement morphology of lexical verbs after various pronouns. Only two questions are specific to NSR contexts, i.e. lexical verbs after full noun phrase subjects. A number of other questions were primarily aimed not at morphosyntactic but at various lexical, phonological or idiomatic targets, but nevertheless provide material that contains tokens of present-tense subject-verb combinations, which can be included in the analysis. As for most other questions eliciting verbs as answers, the material unfortunately does not regularly include subject forms and is therefore unusable. The richest set of elicited data is provided in the SED for the forms of BE. The following questionnaire items elicit present tense forms in various persons and numbers, and in various syntactic environments (negated and non-negated, stressed and unstressed, tags, questions etc.): VIII.9.5 IX.7.1 IX.7.2 IX.7.3

1X.7.4 IX.7.5

IX.7.7 IX.7.9 IX.7.10

(We drink water when...) we are thirsty. Convert for: I am, they are. (Charted in LA Ε Μ16, 18.) (To find out whether you 're right, you ask quite simply:...) Am 1 right? (Charted in LA Ε M28, and CLAE I, M48. (To find out whether I had a wife, you'd ask me:...) Are you married? Convert for: are they (Charted in LA Ε M29, 30.) (Ifyou saw me wheeling a pram and then gathered from our conversation that I was not married, you might ask me in some surprise:...) Aren't you married? Convert for: aren't they. (Charted in LA Ε M27.) (Of a man who has just won a thousand pounds, you would say: Isn't he lucky? And if it was you, you'd say of yourself:...) Aren't I lucky? (Charted in LA Ε M26. (You can say: We 're all right here, aren't we? Now, speaking of that man over there, you can say: He's all right there,...) isn't he? Convert for: aren't I, aren 'tyou, aren't they. (Charted in LAE M12, 14, 15.) (Which of you is English here? For yourself, you could answer...) I am. Convert for: you are, they are. (Charted in LAE Μ1, 2, 4; and CLAE I, M49, 50.) (If I say: You people aren't English, you can contradict and say: Oh yes...) I am. Convert for: you are, we are. (Charted in LAE M5, 6, 8.) (If I said to you: You're drunk, you would answer: Oh, no,...) I'm not. Convert for: they aren't. (Charted in LAE M9, 11.)

The coverage of past tense was/were is less complete. Three questions eliciting ten forms are found in the questionnaire: VIII.9.5B (We drank water because...) we were thirsty. Convert for: 1 was, she was, they were. (Charted in LAE M20-23.)

they all followed a policy of noting down tokens of NSR usage when they could. Nevertheless, a fieldworker isogloss may be distorting the quantitative distribution of NSR just in its most interesting boundary areas. In some of the maps presented in this chapter, it was therefore deemed necessary to use different weighting criteria for tokens recorded by Stanley Ellis and those recorded by the rest, especially by Peter Wright.

67 IX.7.8 IX.7.6

(We'd all buy lots of things if...) we were rich. Convert for: I were, she were. (Charted in CLAEI M53-55, LAE M24-25.) (I was late,...) wasn 7I? Convert for: weren 'tyou, weren't they.

The following are the SED items dealing specifically with the inflectional forms of the other primary verbs, DO and HA VE: IX.5.1 IX.4.10 VIII.3.7 IX.5.4 IX.6.1 IX.6.2 IX.6.3 IX.6.4

(You don't care for things like that, but...) I do. Convert for: we do. (If it means nothing to you whether tomorrow it's wet or fine, you say:...) I don 7 care. (In talking over with an old friend the happy times you had together long ago, you'd ask him from time to time,...) Do you remember? (Ifyou want to know how much rent Jack pays for his house, you ask him: Jack, how much rent...) do you pay? " (Charted in LAE M36.) (Have you got a match?) Yes, I have. (Charted in LAE M40.) (I have a match, but...) he hasn 7. Convert for: I haven 7, we haven 7. (Charted in LAE M42,44.) (1 can 7 tell you what Smith's house is like, because so far...) I haven 7 seen it. (You say to a friend: Shall I give you one of these pups? But he answers: No thanks...) we have got one. (Charted in LAE M45

The following questions elicit lexical verbs. However, all of them were targeted not at the verbal agreement morphology but at lexical, phonological or other information. This means that it is not guaranteed that all forms cited in SED-BM were uttered in exactly the syntactic environments suggested by the question. For instance, in question IX. 1.9 below, it cannot be assumed that all tokens o f creep were produced without a resumptive pronoun subject adjacent to it, so a study of the effects o f the Position-of-Subject Constraint is out of the question. 4 VI.5.11 IX. 1.9 VI.3.2 VI.4.2 V.2.12 VI.2.8

(When I have an apple,...) I eat it. (To please the children, I often go down on hands and knees) and creep. (What do you do with your eyes?) - You see. (What do you do with your ears?) - You hear. (How do you see in this room when it gets dark?) We put the light on. (When two little girls get cross with each other, what do they often do?) They pull each other's hair.

In contrast to this group, the next three items below were designed to focus on the verbal forms themselves, and therefore provide more reliable material for a study of subject-verb agreement with pronoun subjects. However, some caution is still in order: the elicitation routines were formulated in such a way that the subject they, in each case, was provided by the fieldworker, and it was not recorded whether or not the informant actually repeated the The listings in the Basic Material for VI.5.11 and IX. 1.9 also include numerous tokens of the verbs eat and creep excerpted from the incidental material. Klemola (1996: 52 and 2000: 332) uses these, assuming that they are all examples of verbs with pronominal subjects. But the fieldworker notebooks show that this is by no means always the case. Even where such tokens are labelled for their person/number environment, it is often not clear whether they come from pronominal or full noun phrase subject environments. Another SED item which Klemola includes but which turns out to provide no disambiguated subject-verb pair tokens at all is the following: V.7.7

Of you see that your gravy is too thin, what do you do to it?) Thicken it.

68 subject in his utterance or only completed the clause where the fieldworker hat left off. In a situation where verb forms may be sensitive to the type and position of the subject, an utterance that contains such an unnatural speaker shift right between subject and verb can certainly not be guaranteed to provide the most natural form. VIII.5.1 VIII.5.2 IV.6.2

(What do good people do on Sunday? They...) go to church. (Charted in CLAE I, M38). 5 (But some lazy people like to read the Sunday papers, and so they...) stay at home. (Some people have a shed and a wire-netting run at the bottom of their garden in which they...) keep hens. (Charted in CLAE I, M28).

The only questions in the SED that were specifically designed to elicit subject-verb agreement behaviour with full NP subjects and lexical verbs are the following two: VIII.7.5 III.10.7

(We ordinary people buy the things we need, but...) burglars steal them. (Charted in CLAE I, M40). Bulls bellow, horses whinny... (Charted in CLAE I, M27).

In both cases, in order to control for the syntactic environment, these questions were preceded by others which elicited only the word or words in question, for lexical or phonological reasons. In the case of question VIII.7.5 the preceding prompt was: "What do burglars do? They break into houses and...". The informant was then invited to repeat the whole sentence including the subject noun, using the prompt: "So you can say: We ordinary people buy the things we need, but...". Unfortunately, the fieldworkers did not always document whether they succeeded in eliciting the exact syntactic pattern of bare noun subject and adjacent indicative verb, rendering only the verb form itself. It is also far from obvious in many cases whether the answers to the preceding question can be used as evidence for structures of the they sing and dances type, as the phrasing would seem to suggest. The situation is similar with question III. 10.7. Here, the preceding context was a series of questions after a variety of animal sounds (III. 10.2-6). The Basic Material for these questions is itself not very informative for the present purpose, because the syntactic context for each verb was usually not documented. 6 The informant was then asked to repeat the whole series of animal names and the verbs for their sounds in one go: "Now let's have it all together. You tell me once again: ..." Unfortunately, the fieldworkers again often failed to document the complete utterances and recorded only the verb forms, or only provided a summary comment stating if the informants used -s or not. Finally, the questionnaire contained a number of questions eliciting idioms or other verbal expressions, which gave rise to various expressions with inflected forms of verbs and various types of subjects. VIII.2.8 IX.9.4

5

6

(On meeting a friend in the street and inquiring about his health, you would say to him:...) How are you? (A little boy comes up and talks to you in the street, and you are not sure you know him, so you say: Tell me,...) who are your parents?

Note that incidental material from the northern locations, including numerous non-standard agreement tokens, is printed in an appendix in the SED-BM volume dealing with the East Midlands. Nevertheless, two of these items (III.10.3A/B horses whinnies, horses neighs) have been charted for their evidence of verbal -s in CLAE M27 and M28.

69 VI.5.8 VIII.7.9 VI.14.2 III. 1.12 111.5.2 V. 7.21

Have you got toothache? (To get rid of someone quickly who has been a bit of a nuisance, you'd say:...) Off you go. (Your wife, when for the first time she wears a nice new hat that goes well, likes to hear you say: My word!...) It suits you. " Her bones have come off J The horses are baiting.8 The old potatoes are finished.

The target of the following item was the phrase which one rather than the following verb form, but the answers occasionally include verb forms such as do you, doest thou, art thou,

etc. In the West Midlands volume these forms have been included in SED-BM; for the other areas they have to be looked up in the fieldworker notebooks. VII.8.18 (Ifyou offered a boy the choice of six apples, you'd ask him:...) Which one will you have?

4.2.

An overview

The geographical situation found in the SED can be summed up as follows (cf. Map 3). There is a central northern area, covering the three northern counties 9 of Cumberland, Westmorland, Durham, and the southern half of Northumberland, which have a fairly consistent, homogeneous subject-verb agreement system. Its most important features are: preservation of the thou versus you distinction, with thou always taking verbal -s forms; generalized use of is as the singular present tense form of BE in all three persons; preservation of the was versus were number distinction as in the standard; hence parallel paradigms of BE in both tenses (I/thou/he

is, we/you/they

are, I/thou/he

was,

we/you/they

were); application of the NSR to all verbs, including BE. This bundle of features, which quite closely conserves the northern Middle English system, can conveniently be called the Central Northern 10 agreement system. In the northern half of Northumberland, we find more or less the same features, but with less consistency, BE paradigms more similar to the standard, and a lack of thou. This area can conveniently be labelled the "Upper North". 7

8

9

10

The question for this item was: "When a cow shows signs of giving birth, you say...". According to the SED editors, this and similar expressions ("her slacks drop ", "her strings have gone "etc.) refer to "the anatomical changes observable in the pelvic region of the cow shortly before parturition." The question dealt with expressions for feeding a horse while resting from work in the field. The cases that are of interest here are those where the informants provided expressions meaning 'the horses are eating' rather than 'to feed the horse'. Following SED practice, all reference to English and Scottish counties in the following chapters and maps is in terms of the traditional (pre-1974) county divisions. Designations such as "Central North", "Lower North" etc. are intentionally reminiscent of the well-known, partly homonymous labels introduced by Trudgill (1990) for the "modern dialect areas", but are not equivalent to them. As used throughout this chapter, these terms designate areas defined exclusively in terms of the subject-verb agreement data in the SED.

70

systems can be found. Verbal -s conforming to the NSR competes with inflectional forms in -n. Moreover, there is also a strong tendency to neutralize the was/were distinction in favour of generalized were forms. The second singular form of BE is generally thou are or

71

thou art. These three features together are characteristics of an area between Derbyshire and southern Lancashire, also reaching into the southwestern border areas of Yorkshire, and may be labelled the "Northwest Midlands" agreement system. To the north of it, one finds a gradual shift from the Northwest Midlands features towards the Central Northern features. It can be visualized as a bundle of successive isoglosses stretching through Yorkshire and northern Lancashire. These isoglosses centre around a line from Morecambe Bay to the mouth of the Humber. This line, also called the Humber-Lune-Ribble line, has long been known as one of the most important and oldest boundaries in the traditional dialects of English. It is usually seen as related to the boundaries of the Anglo-Saxon kingdom of Northumbria (Trudgill 1990: 34, Ihalainen 1994: 219), but has also been explained as reflecting the distribution of Scandinavian influence (Samuels 1988). For the present purposes, the area spanned by the whole isogloss bundle, i.e. most of Yorkshire and northern Lancashire, can conveniently be called the "Lower North". Further southwest into the Midlands, more variation in the paradigm of BE is found, involving present indicative forms such as be, bist, and bin. Also, the area preserving inflectional -« overlaps in the south with an area that has generalized verbal -s independent of the NSR. The East Midlands differ from the West Midlands mainly in that present-tense paradigms of BE are more similar to the standard, and second singular thou is generally absent. The two areas are fairly sharply divided by a line that runs between the counties of Derbyshire and Staffordshire in the west, and Nottinghamshire and Leicestershire in the east. However, one feature that both the west and the east Midlands share is the tendency to neutralize the was versus were number distinction, in the one or other direction. NSR-conforming plural verbal -s with lexical noun subjects, as exemplified in the SED items burglars steals them and bulls bellows, is found everywhere north of a line running from Merseyside to the Wash (cf. Ihalainen 1994: 211, Klemola 1996: 59, 2000: 334," CLAE I M40, Viereck/Viereck/Ramisch 2002: 84). This burglars steals isogloss, as it may conveniently be called, cuts right through both the West and East Midlands with their otherwise very different systems. The following sections will give a more detailed account of the different agreement phenomena and the ways they interact with each other.

4.3.

N o n - N S R agreement patterns

4.3.1. Northwest Midlands verbal -n forms In one compact relic area in the northwest Midlands, overlapping with the NSR area, verbal -n suffixes can still be found in the SED. They are a continuation of the Middle English plural -n paradigms typical of the Midlands (cf. Map 2 on page 47 above for their much 11

Note that in Klemola (2000: 334f.) the captions for two of the relevant maps, one dealing with pronouns and one with full NP subjects, are apparently exchanged by mistake. Cf. the correctly labelled maps in Klemola (1996: 56, 59.)

72

Map 4. Verbal -η forms in the SED

50

100km present-tense bin Δ generalized am/'m Ο verbal -n (other verbs)

Northumberland '4·

Durham 4» 5·

Cumberland,

Westmorland

2· 13· 14· 19 ,

18· "

23>

Yorkshire 2 0

24

2

J Lancashire 32·

33·

.34· LJeibystmc i f Q pheshiri

{2·

oO

Norfolk

, I· iunt sh. 3

ι

1

·

Warwicksh4.SNorth'sh· -ambridgesh./

Herefordshire

WorcestershV v

7

76 found in the north. 15 This is part of the area that has generalized plural verbal -5 irrespective of the NSR, as will be shown below (cf. also Klemola 1996: 50ff.). Appendixes B.3-11 starting on page 163 give a detailed list of tokens.

4.3.3.

Verbal -s with thou

The old second person singular pronoun thou is found preserved in the SED data in a large coherent area covering all of the North, with the exception of the East Midland counties, as well as of northern Northumberland, where it is found less frequently and where ye/you is generally used instead (cf. Trudgill 1990: 86, Beal 2004.). The pragmatic and sociolinguistic function of the thou-you contrast is not of primary interest for the present purposes. 16 Neither do the different morphological realizations of the thou form (thou, thee, [öa], [ta] etc.) need to be discussed here. What is important is that the thou forms, where they exist, almost invariably command agreement with -s forms (rarely also -st forms) of the verb, irrespective of the Position-of-Subject Constraint. As in the original northern Middle English system, they do not fall under the NSR but rather align themselves with the third singular. Variation occurs only in the paradigms of BE. As for the past tense, the variation between thou was and thou were will be dealt with together with that in the other persons and numbers in the next section below. As for the present tense, three fairly distinct regional types can be distinguished (Map 6). In the Central North one regularly finds thou is, alongside I is and he/she/it is. Further southwest, in southern Lancashire, southwestern Yorkshire, Derbyshire, Cheshire, and northern Staffordshire, the forms are thou are or thou art, in inversion sometimes contracted to art'. Going yet further south into the West Midlands, one finds forms such as thou bis(t), as well as other forms with the verbal stem be in which the plural forms have been generalized to the second singular (thou be, thou bin). Token lists can be found in Appendixes B.4—12 beginning on page 163.

4.3.4.

Levelling of was/were

In the Central North, the usage of indicative was and were is, as a rule, parallel to that of is and are. The singular of all three persons (including thou) is invariably was, while the plural (including you) mostly has were but allows for was in accordance with the NSR. This is mainly true for the four northern counties of Cumberland, Northumberland, Westmorland and Durham. The SED data suggest a near-categorical validity of the number contrast with adjacent pronoun subjects in this area, with only one exceptional token of we was in location Du4, and two tokens of I were in Nb3 and Nb4. (However, compare the attestation of plural was levelling in more recent Tyneside and Northumberland dialects, according to Beal 1993: 194). Even plural was conforming to the NSR is rather scarce in the SED data from these northern locations, with only one token recorded (2).

15

16

It should be noted that the original fieldworker notebooks were not consulted for these locations, so the picture may be less complete here than elsewhere. For data relevant to this question, see SED-BM under question IX.7.2.

77 Map 6. Verbal agreement forms with thou in the SED

50

100km Ο thou is

•O'

thou bist Φ thou are/art Δ thou be/been

Northumberland.,

fO /

Cumberland?"})

JT%

O

W

l O

j

P

Ö

L o

J ^

Ο thou -s (other verbs)

\ 3

^

ί Ο - ΐ Α ' ^ Ό

Q o

_ o

^

^

„ l l ( m

-

U O 19 ίο/φ

^ © © O •)Lancashir^N „rfgX)

Ii3 (φ

Yorkshire*

; © "(($/)

Ό ο

^

^,(Ο)

(φ

2·

3 Γφ)

32ΗΜ33(^. /

ι 4*

'·

— Ja Detbyshire 1 10

3 ·

1

Westmorland ,„4· ί^ν/β· 7* s· i*13

iA ^ ^

9·

10·

4* N

I I *

iq>6v Yorkshire" , Ο

" Δ

i f ® Lancashire 12^

1·

2&

Derbyshire

I B

1

• 5·

··

(5· \ a

*

, δ

· *iottinghamih.

kC

7A

Lincolnshirt 12A

j5 ·

e·

Wjorcestersh/^-

, Δ

3 A

9]

r ^ j .

2O ^O'^Vw

Herelprdsnir|Q

2

ιη/*\

Cheshire^/'

1

Norfolk

«V'

^Hunt'slT

ambridgesh./

80 (3)

They was very cheap, clay pipes was [Ch2: 1.1.3]

Certain tendencies can be discerned within this general state of variation. Generalized were forms are more common in the negated (tag) contexts than in the affirmative clauses (although the exact opposite distribution also appears in some token sets, for instance in northern Lancashire). The data gives no direct indication as to whether the difference is due primarily to a negation versus non-negation contrast or to a tag versus non-tag contrast. However, recurrent results from variation studies in many areas of England suggest the former. Tendencies for was/weren't levelling has been reported, for instance, by Cheshire 1982, Tagliamonte 1998, Anderwald 2002, and Britain 2002. This negation effect on was/were levelling is discernible in data from all the area surveyed, with the exception of the Central North and those parts of the NW Midlands and Lower North where were levelling is predominant in all environments. Map 8 on page 81 shows where the was-weren 't contrast was found in the relevant SED questionnaire items, either consistently in all elicited forms or only as a tendency in most of them. It indicates that to the north of the central generalized were area, negation tends to affect only the singular forms, with were remaining near-categorical in the plural, except where was is licensed by the NSR. The existence of systems of this kind is consistent with findings by Tagliamonte (1998), who reports the combined effects of singular were generalization, the negation effect, and the NSR (though with no NSR usage preserved in the lexical verbs) in the more recent local variety of York. To the south of the generalized were area, the tendency is for both the singular and the plural to pattern according to the negation constraint. A second effect which has repeatedly been reported (Tagliamonte 1998, Chambers 2004) is that you was is more common than we was and they was. Unfortunately, the SED data does not lend itself to an investigation of this effect, because second person environments were not well covered by the questionnaire. Only one question (IX.7.6) involved a second person address, and this was in a negated tag in the singular. No item in the questionnaire targeted the opposition between semantic singular and plural in the second person. You forms are therefore reported in IX.7.6 only from those speakers who happened not to use thou in the singular. In five of the Midlands locations, fieldworkers recorded wasn 't you as opposed to weren't they (LI, L10, He3, He7, Wa3), but in four locations they also recorded the opposite distribution, weren 'tyou as opposed to wasn't they (L6, StlO, Nth3, Nth4).

4.3.5. Verbal -s with we, you, and they Apart from the variability in the two BE paradigms and in the second singular, as discussed so far, non-standard verbal -5 in lexical verbs also occurs occasionally with plural pronominal subjects, contrary to the Position-of-Subject Constraint. The geographical distribution is very similar for all three plural pronouns, as well as for / (see section 4.3.2 above). The distribution of verbal -s with all four pronouns is therefore shown together in Map 9 (cf. also the similar map in Klemola 1996: 56 and 2000: 335). Four distinct regions can be distinguished in this map. They are basically the same as those shown in the map for / (Map 5 on page 75 above). In the Southwest Midlands, i.e. south from Herefordshire and Worcestershire, generalized verbal -s is the rule in all persons and numbers and was quite regularly

81 Map 8. The was/weren 7 negation effect in the SED

50

100km singular was-weren 7 (as a tcndcncv) •

(consistently)

Northumberland., plural was-weren 7

Δ

(as a tcndcncv)

Δ

(consistently)

Durham 4* 5·

Cumberland

Westmorland 11.4®

ι 8

φ

22- 23·

i 9

Q Yorkshire

24»

I i · Lancashire 12· 32· 34*

33· 6

Cheshire

?Δ

Derbyshire 4* 5» 4 » 1/ __ ^ottinghamih."

¥

Λ

δ

-,Δ Lincolaehjre coljwhjri

Shropshire

Norfolk

iunt's'h. " s Q ß ( j j ^ ·

North'sh. 4· _ambridgesh./

' Ό Herel ordsmre .

' ' - Τ ·v,L. WorWStersh/

)

m\ 5 *

82 recorded in the various morphological elicitation items in the SED.18 This area extends further south throughout much of the Southwest of England (CLAE I: M38, II: M28). In the Lower North, especially in Yorkshire and Lancashire, verbal -s was recorded sporadically, mainly in the incidental material and much less often in the elicited items. These two areas are separated from each other by a broad gap which extends from Cheshire into the East Midlands. Verbal -s with pronominal subjects in these areas seems to be quite rare or absent. 19 Another area where verbal -5 with plural pronouns and / is found much more rarely is the Central and Upper North. 20 These data suggest that the Type-of-Subject Constraint is not equally strong throughout the area affected by the NSR. Both in the Central and Upper North and in the East Midlands it seems to be near-categorical, whereas in between, especially in southern Yorkshire and Lancashire, it can be overridden by other patterns (cf. Wright 1905: 296). These patterns are probably the same as those that have been reported elsewhere in the literature (see especially Shorrocks 1999): the narrative use in the historic present, and the use for habitual semantics. As was pointed out earlier (page 9), tokens such as I often goes present a difficulty in determining whether the verbal is triggered primarily by the Position-of-Subject Constraint of the NSR, or by the habitual semantics expressed by the adverb, as described by Shorrocks (1999: 112). Many of the tokens in the SED are indeed found in environments with such adverbs and could be licensed by either rule. However, there are also a large number of tokens with adjacent pronouns. A tentative quantitative analysis of their distribution will be undertaken in the next section. As for the habitual semantics, nothing much can be said here about these tokens, given the absence of discourse context in the SED data.

4.4.

The Northern Subject Rule in the SED

Having outlined the distribution of non-NSR subject-verb agreement phenomena, we can now turn to the question of NSR related verbal -s proper. The picture of the geographical distribution of the NSR in the SED that has prevailed in the literature up to now has been based on the data listed in SED-BM under the headings III.10.7 (bulls bellows) and VIII.7.5 (burglars steals) (cf. CLAE II: M27,1: M40; Klemola 1996: 59, 2000: 334; Viereck/Vier18

19

20

Note that the fieldworker notebooks were not evaluated for many of these locations, which means that the counts recorded in the map are likely to be under-stated in comparison to the northern locations. Klemola (1996: 52 and 2000: 332) lists some more tokens of verbal -s in this area, especially in Staffordshire and Lincolnshire. However, this is a result of misreading the data for one SED item, V.7.7 (to thicken it). The incidental material printed in SED-BM under this heading was selected for having post-verbal object it rather than for having specific types of subject, and the subjects were not documented. Contrary to Klemola's silent assumption, not every token of verbal -5 in this list can be assumed to come from a plural pronoun context. Cf. the list of verifiable tokens in Appendix B . l l , 13, 14, and 15. The token recorded for location Dul actually has subject us, not we. Us being a strong pronoun in origin, the -s form may also be regarded as licensed by the Type-of-Subject Constraint in this instance, just like with them. The SED does not provide appropriate data to decide on the exact status of us with regard to the NSR.

83

Map 9. Verbal -s with I/you/we/they in the SED

84 eck/Ramisch 2002: 84). Map 10 shows this distribution. Similar to Klemola's map, it indicates all locations where plural verbal -s was recorded with verbs other than BE and with full clause-internal NP subjects headed by lexical nouns. It shows one area using verbal -s in the southwest, which is more or less identical to the one where generalized verbal -s is also common with pronouns, hence part of the southwestern non-NSR area; and a second, quite sharply delimited area covering all the north down to a Chester-Wash line. The delimitation of the latter is the burglars steals isogloss shown also in Map 3 on page 70 above. Map 10 shows the burglars steals isogloss crossing the East Midlands slightly further south than in Klemola's map. 21 But what is even more important, incidental material evidence for plural verbal -s occurs sporadically yet further south in the East Midlands, but only in certain contexts that were not covered by headings III. 10.7 and VIII.7.5 in the SED: subject relative clauses, interrogative clauses, plural demonstrative and indefinite pronoun subjects. Some additional tokens with the verb is (also not represented in the map) also occur south of the main burglars steals line. An overview of the quantitative distribution of all NSR tokens per location, based on the evaluation of the full set of data from the SED fieldworker notebooks including these contexts, is shown in Map 11,22 It indicates a core area of frequent NSR usage in Yorkshire and Lincolnshire. Surprisingly perhaps, the Central and Upper North are somewhat less well represented, although at least a couple of tokens of verbal -5 are attested in every single location. Towards the Southeast Midlands there is not a sharp boundary but a transitional belt of less and less frequent NSR usage. The northernmost locations in this border zone that have no verbal -s recorded at all are found in a belt stretching from Cheshire to Cambridgeshire. Two things are noteworthy here. First, with the transitional zone included, the NSR at the time of the SED survey apparently still covered precisely the same area as it did in Late Middle English. Geographically it must have been remarkably stable for at least six centuries, as can be seen from a comparison with Map 1, showing the LALME evidence, on page 46 above. Second, the linguistic environments that seem to favour the retention of verbal -s in this transitional zone - and which unfortunately were not elicited in the SED - are almost exactly the same environments that will also be found as favouring factors in the Northern 21

22

Klemola excludes locations Lei2, Lei6, and Rl. These differences in detail are due to certain inconsistencies in the selection of incidental material tokens. This is a problem especially for the locations near the borderline of the NSR area. For instance, Klemola omits the token from Lei6: "th'ashes makes good stuff", although it is printed under III. 10.7. On the other hand, the editors of the Basic Material failed to include the token from Lei2: "where the spokes goes in". In addition, the Basic Material is inconsistent about clause types such as relative clauses, e.g. including the token from location Lei7: "there's two or three comes up at five o'clock", but omitting many structurally similar ones, for instance that from location Wa2: "there's a good many uses them". The present map is based on the full token list as drawn from the fieldworker notebooks (Appendix B.16). Note also that Klemola counts maximally one token per location and SED heading, even where several incidental material tokens are printed under the same heading. Owing to the different data collection habits of some fieldworkers, it was necessary to use different weighting criteria for tokens recorded by Stanley Ellis and those recorded by the rest (see page 65, footnote 3.) Locations whose fieldworker notebooks were not consulted are excluded in this map.

85 Map 10. Plural verbal -s in the SED (I) (lexical noun subjects, lexical verbs, canonical SV clauses only)

50

100km Ο 1 token 2 or more tokens slorthumberlani

m eC)\

(o Ό

Durham 4 * 5 *

Cumberlan^^

%

norland

~ IU

• 0 , s 0

„ o in Λ

i9 A

Yorkshire0^

φ > 2 # 23* 1 T V Lancashiri • 12 3 2 β 33|

οι

Derbyshire Cheshire • Φ * · , Wforces aTcestersR i0 Herefprdsnire.. 4 Jo

3

Norfolk

isestershifg8

y •

Lincolnshire 12*

Ο

Y^·

" ^fiunt'sk

arwicksh,.SNorth'sh· _ambridgesh./

86 Irish data discussed below in Chapter 5.3: subject relative clauses, inverted clauses, certain demonstrative pronoun subjects such as them, and possibly some of the primary verbs. This also fits in with observations reported by Wright 23 as early as 1892: in the dialect of Windhill in the West Riding of Yorkshire, plural verbal -s had largely become restricted to relatives and to HA VE and BE, and it occurred in other environments only rarely as a relic form (1892: 156). Wright also includes several examples with the subject them among lists of typical occurrences: (4)

(5)

a. Them's the men that does their work best b. Them men's been very good to me c. Us that's done so much for him d. Me that 's so poorly e. The coals isn7 done yet a./Vedone b. They're at it again c. Them men do their work very well

Wright's location is situated well north of the burglars steals isogloss in the SED. Although the SED data does attest the continued presence of verbal -s in that same area even outside the special favouring environments, more than half a century after Wright, his observation suggests the existence of a strong favouring tendency, and as such it fits in well with the observation in the SED that verbal -s was preserved longer, and further in the south, in just these favouring environments. The same rule that appeared as a strong quantitative constraint within the one local variety investigated by Wright, manifests itself as an implicational constraint when comparing different varieties along the north-south cline. Shown below are typical SED attestations of relative clauses, including presentational relative clauses after existential there, most typically with zero relativizer and with nonagreement in the preceding existential construction (7), as well as after cleft it's (8). (6)

a. You don't see many has holes now [La7: V.3.3] b. Them what's got a few [Lei5: VII.7.9] c. Hedges that hasn Ί been done [Lei9: IV.2.5] d. The ones that goes across was braces [R2: IV.3.7/III.10.7] e. It kills the thorns as grows round it [Nth2: IV. 10.7] f. I know several signs as is pretty sure [Nth4: VII.6.7] g. People what was used to it [Hul: IV.4.4]

(7)

a. b. c. d. e.

(8)

We're both right, aren't we, it's t' others 'at's wrong [Y16: X.4.4]

There's a lot of the people now as doesn7 talk like they used to [L9: VI. 14.11 ] There's any amount takes cattle on [Nb4: iii.3.8] There's not so many fills a ten-quart tin [Cu3: VIII. 1.24/IX.9.5] There 's a lot of people kills 'em [L15: IV.5.4/III.10.7] There 's two or three comes up at five o'clock in the morning [Lei7: IV.8.3/III. 10.7]

Plural verbal -s in inversion is exemplified by tokens such as those in (9). Note that there are several instances where 'em, a weak form of demonstrative them, is used as a tag subject just like a simple pronoun they, i.e. apparently without emphasizing or deictic force. It 23 The examples from Wright (1892) are quoted after Chapman (1998: 38f.).

87

Map 11. Plural verbal -, La4 (3[3]>, La5 (3), La6 (2[4]), La7 (2), La8 (1[2])( La9 (2), La 10 (2[1]), L a l l ( l ) , L a l 2 ( 2 ) , Lal3 (2), Lal4 (1), Y1 (3), Y2 (2), Y3 (3), Y4 (1), Y5 (3), Y6 (3), Y7 (3[1]), Y8 (2[ 1 ]), Y9 (3), Y10 (2), Y l l (1), Y12 (3), Y13 (2), Y14 (1[1]), Y15 (2), Y16 (2[1]>, Y17 (2), Y18 (2), Y19 (1), Y20 (1), Y21 (2), Y22 (2), Y23 (2[2]), Y24 (2), Y25 (2), Y26 (2), Y27 (3), Y29 (2[1]), Y30 (2[1]), Y31 (2[1]), Y32 (2[1]), Y33 (2), Y34 (1), Chi (3), Ch3 (2), Ch4 ([1]), Ch5 (2[1]), Dbl (1), Db2 (1), Db3 (3), Db4 (3), Db5 (2), Db6 (3[2]), Db7 ([1]), LI (1), L2 (1), Sa5 (3), Sa7 (1), Sa8 (1), Sa9 (1), SalO (1), Sal 1 (1), Stl (2), St2 (1), St3 (3), Hel (1), He2 ([1]), He3 ([1]), He4 (1), He5 (1), He6 (1), Wo2 (1), Wo3

(2), Wo4 (1), Wo5 (1), Wo6 (2), Wa7 (1). Nb9 Cu2

VI.14.2 VI.2.5

Cu4

V.5.9

Cu4 We3 We4 We4 La2

VI.14.2 VI.14.2 VI. 14.2 1.9.10 VII.6.18 IX.8.8

La3

III.3.2

La3 La4 La4 La4 La6 La6

VII.3.16 VI.2.7 VI.3.2 VI.3.2 VII.6.19 1.5.3 1.9.2

La6

II.5.3

VI.14.2

thou seems it thou has a tousy head ([5u: tz]) where tha (they?) was cheese-making, (no fw. tr.) tha looks well in that tha looks a gay sight in't. thou knows Thou knows ([du na:z]) who has tha had (=who have you had) [wa czta] has tha watered thy cows? ([ezta] when tha likes ([5a laiks]) tha grows a beard tha sees summat tha often pulls me up tha puts it over tha head spell it as tha likes it. ([azta laiks]) by gum, tha's some good

171

La6

III.5.2

La8 La8 LalO Y7 Y8 Y14 Y16 Y23

VI. 14.2 VI.14.2 VI. 14.2 VI.3.2 VI.3.2 VI.14.2 VI. 14.2 VI.3.2

VI.14.2 VI. 14.2 VI. 14.2 VI.3.2 VI.3.2 VI.14.2 VI. 14.2 VI.3.2

headed barley (=you have) ([öaz]) if tha going a long journey wi' 'em, tha takes the baiting wi' tha tha suits that tha looks gradely well tha looks well in it thou sees tha sees tha looks a gay sight in it thou looks well in it tha sees

Y23 Y29 Y30 Y31 Y32 Ch4 Ch5 Db6 Db6 Db7 He2 He3

VI.5.11 VI.5.11 VI.5.11 VI.5.11 IX.9.6

VI.5.11 VI.5.11 VI.5.11 VI.5.11 IX.9.6 IX.6.4 1X.6.4 IX.7.2 IX.7.2 VIII.3.7 VIII.3.7 VIII.3.7 VIII.3.7 IX.6.4 VI.3.2 VI.3.2

tha eats tha eats tha eats tha eats that's the chap tha knows thee 's got thee s got hast tha got a missus? canst thee remember? canst' remember? thee's got tha sees

He4 He4 He4 He6 Wo4 Wo4 Wo5 Wo6

V.2.12 IX.5.1 V.2.12 V.2.12 VI.3.2 V.2.12 V.2.12

V.2.12 IX.5.1 IX.5.4 V.2.12 V.2.12 VI.3.2 V.2.12 V.2.12

we puts the we does does we? we puts the we puts the we sees we puts the we puts the

Y17 Y32

III. 12.9 VI.3.5

III.12.9

He3 He7 Wo2 Wo3 Wo6

VI.3.2 VI.3.2

VI.3.2 VI.3.2 IX.5.4 VI.3.2 VI.3.2

you renders 'em i' t' oven. why doesn't you look where tha art going? you sees you sees does you? you sees you sees

B.13. Use of -s with we, other verbs Verbal -s with we or us (including the present-tense forms of the primary verbs is/has/does) was recorded 13 times in the following locations. Dul LalO Hel He2 He3

V.l.7 V.7.23 V.2.12 V.2.12 V.2.12

V.2.12 V.2.12 V.2.12

us calls us nips we puts we puts we puts

it 'em the light on the light on the light on

light on

light on light on light on light on

B.14. Use of -s with you, other verbs Verbal -s with you was elicited 6 times in question VIII.7.9, and recorded 17 times (including is/has/does) in other contexts [numbers in square brackets], in the following locations: Cu4 ([1]), Cu5 ([I]), La9 ([1]), Y7 ([1]), Y12 ([3]), Y14 ([3]), Y17 ([2]), Y32 ([1]), L4 ([1]), He3 (1[1]), He4 (1), He7 ([1]), Wol (1), Wo2 ([!]), Wo3 ([1]), Wo4 (1), Wo5 (1), Wo6 (1[1]). Cu4 Cu5

II. 1.5 III.5.4

La9 Y7

II.9.5 V.8.13

Y12

III.12.9

Y12

VI.8.13

Y14

II.7.5

you goes up t' next. you sticks it nose = you stick its nose you pikes it H.9.5 I I I . 12.9 you is welcome at home ([yaraz]) you cuts it up and renders it you is welcome at home as you're here ([j3raz]...|ja:na]) you take that there up and shakes it out;... you turns it

VI.3.2 VI.3.2

172

Β. 15. Use of -5 with they, other verbs Verbal -s with they and present-tense verbs was recorded 94 times, in the following locations: Cu4 (2), Cu5 (1), Du4 (1), Du5 (1), La6 (4), Y3 (3), Y6 (2), Y7 (1), Y9 (2), Y l l (1), Y12 (1), Y14 (1), Y17 (1), Y18 (1), Y19 (1), Y21 (2), Y27 (1), Y29 (1), Y30 (2), Ch4 (1), Dbl (1), Db5 (1), Db6 (1), Ntl (1), L12 (1), Sa9 (2), SalO (2), Sal 1 (1), St2 (1), Lei2 (1), LeilO (1), R2 (1), Hel (1), He2 (1), He3 (3), He4 (1), He7 (6), Wol (1), Wo2 (1), Wo3 (2), Wo4 (7), Wo5 (3), Wo6 (7), Wo7 (12), Wa2 (3), Wa7 (1), C2 (1). Cu4 Cu4 Cu5 Du4 Du5 La6 La6 La6 La6 Y3 Y3 Y6 Y6 Y7 Y9 Y9 Yll YI2 Y14 Y17 Y18 Y19 Y21 Y21 Y27 Y29 Y30 Y30 Ch4 Dbl Db5 Db6 Ntl L12 Sa9 Sa9 SalO

III.7.9

when they (tha?) does that ([3a diz]) VIII.5.1 they pumps it up VIII.5.1 they reads the papers VIII.5.1 they lets 'em in IV.6.2 IV.6.2 they keeps hens VI.2.8 VI.2.8 They lugs one another's hair VI.6.5 they gets in the gullet. VIII.5.1 VIII.5.1 they goes to church IX.4.16 doesn 7 they V1I.2.8 they travels miles [CONV] they maybe wants VII.8.21 III.10.7, farmers goes out and they VIII.5.1 drives 'em off VIII.5.1 they whinnies VIII.5.1 VIII.5.1 they calls 'em IX.5.1 they does they is IX.7.7 IX.7.7 VIII.5.1 VIII.5.1 they goes to t' church III. 11.8 III.10.7 they saves it VIII.5.2 VIII.5.2 they stops at home they laps theirselves IV.5.5 VIII.5.1 VIII.5.1 they goes to t' church VIII.7.1 they gets a holiday VIII.5.1 they starts II.l.l VIII.5.1 they leaves it a bit if they's no tred there'll be IV.6.9 no chicken ([3az no:]) V.5.6 VIII.5.1 they hangs it up and makes it into cheese VIII.5.1 they puts 'em i ' t ' corner II.1.5 o ' t ' field IV.2.4 VIII.5.1 they puts it length-way VIII.5.1 they opens VIII.5.2 VIII.5.2 they stops at home II.6.6 VIII.5.1 they carts it home they tears one another's VI.2.8 VI.2.8 hair VII. 1.13 VIII.5.1 they gallops about VIII.5.1 they doesn 7 do they keeps the poultry IV.6.2 IV.6.2 VIII.5.1 they draws VIII.5.1 they scratches

SalO Sail St2 Lei2 LeilO R2 Hel He2 He3 He3 He3 He4 He7 He7 He7 He7 He7 He7 Wol Wo2 Wo3 Wo3 Wo4 Wo4 Wo4 Wo4 Wo4 Wo4 Wo4 Wo5 Wo5 Wo5 Wo6 Wo6 Wo6 Wo6 Wo6 Wo6 Wo6 Wo7 Wo7 Wo7 Wo7 Wo7 Wo7 Wo7 Wo7 Wo7 Wo7 Wo7 Wa2 Wa2 Wa2 Wa7 C2

IV.6.2

IV.6.2 VIII.5.1 VIII.5.1

they keeps fowls they thinks they wants they has him now (['az m]) V.7.6 VIII.5.1 they isn 7 VIII.5.1 they calls 'em morphs IV.6.2 IV.6.2 they keeps fowls VIII.5.1 VIII.5.1 they goes to church VIII.5.1 VIII.5.1 they goes to church VIII.5.2 VIII.5.2 they stops at home they keeps fowls IV.6.2 IV.6.2 they keeps the fowls IV.6.2 IV.6.2 VIII.5.1 they inoculates VIII.5.1 they ricks VIII.5.1 they knows VIII.5.1 they pulls it and sells it VIII.5.1 they seems VIII.9.5 they's (=they is) they does IX.5.1 they pulls each other's hair VI.2.8 VI.2.8 VIII.5.1 they works VIII.5.1 they gets VIII.5.2 VIII.5.2 they stops at home VIII.5.1 they says VIII.5.1 they burns VIII.5.1 they carries on VIII.5.1 they dies VIII.5.1 they begins IV.6.2 IV.6.2 they keeps fowls IV.6.2 they keeps fowls IV.6.2 VIII.5.1 they drags VIII.5.1 they knows IV.6.2 IV.6.2 they keeps fowls VIII.5.1 VIII.5.1 they goes to church VIII.5.1 they fights VIII.5.1 they sells VIII.5.1 they sheds VIII.5.1 they calls VIII.5.2 VIII.5.2 they stops at home VIII.5.1 VIII.5.1 they goes to church VIII.5.2 VIII.5.2 they stops at home VIII.5.1 they builds VIII.5.1 they lets VIII.5.1 they revels VIII.5.1 they calls VIII.5.1 they uses VIII.5.1 they wonders VIII.5.1 they both coos VIII.5.1 they reckons VIII.5.1 they sits about VIII.5.1 they has VIII.5.1 they gets VIII.5.1 they lives VIII.5.1 they miaows VIII.5.1 they haves [sic!] some

173

Β. 16. Verbal -5 in NSR contexts Tokens of verbal -s with full plural noun phrase subjects were elicited 146 times in questions III.10.7 and VIII.7.5, in the following locations: Nbl (1), Nb4 (2), Nb5 (1), Nb8 (1), Nb9 (1), Cu2 (1), Cu3 (2), Cu4 (2), Cu5 (1), Cu6 (1), Dul (1), Du2 (27), Du3 (2), Du4 (1), Du5 (3), Du6 (2), Wei (1), We2 (1), We3 (1), We4 (2), Lai (1), La2 (1), La3 (1), La4 (1), La6 (2), La7 (1), La8 (1), Lai 1 (2), Lal2 (2), Lal3 (1), Y2 (2), Y3 (2), Y5 (1), Y6 (1), Y8 (1), Y9 (2), Y10 (3), Y12 (1), Y13 (2), Y15 (1), Y17 (4), Y19 (1), Y21 (1), Y22 (1), Y24 (3), Y26 (1), Y28 (2), Y30 (1), Y31 (1), Chi (2), Ch6 (2), Dbl (1), Db2 (2), Db3 (1), Db4 (1), Db5 (1), Nt3 (2), Nt4 (1), LI (2), L2 (1), L3 (2), L4 (5), L5 (2), L6 (1), L7 (1), L9 (3), L10 (3), LI 1 (1), L13 (1), L14 (6), Sa2 (1), Sa3 (1), Sa5 (1), Sa6 (8), Sa7 (6), Sa8 (1), Sa9 (3), SalO (1), Sal 1 (1), Hel (1), He2 (1), He3 (1), He5 (1), He6 (1), He7 (1), Wa5 (1), Wa7 (1). Apart from this, 984 tokens of verbal -s with full noun phrase subjects or non-adjacent pronoun subjects were recorded in other contexts. As for the pronoun tokens, this list contains those that have been interpreted as effects of the Position-of-Subject Constraint, and hence as examples of NSR, because the informants otherwise used non-s forms with these subjects fairly consistently, or because a non-s form was found with the adjacent pronoun in the immediate context (as for instance in "they peel 'em and boils 'em"). Otherwise tokens of verbal -s with pronominal subjects have been included in the various lists above, illustrating non-NSR variation phenomena. The NSR tokens were recorded in the following locations: Nbl (5), Nb2 (5), Nb3 (3), Nb4 (8), Nb5 (4), Nb6 (2), Nb7 (4), Nb8 (9), Nb9 (1), Cul (3), Cu2 (3), Cu3 (2), Cu4 (5), Cu5 (7), Cu6 (2), Dul (3), Du2 (3), Du3 (2), Du4 (9), Du5 (7), Du6 (5), Wei (1), We2 (4), We3 (5), We4 (9), Lai (5), La2 (7), La3 (3), La4 (7), La6 (29), La7 (35), La8 (2), La9 (1), La 10 (2), Lai 1 (4), La 12 (3), Lal3 (4), Lal4 (1), Y1 (7), Y2 (23), Y3 (13), Y4 (12), Y5 (4), Y6 (20), Y7 (16), Y8 (10), Y9 (23), Y10 (17), Y11 (22), Y12 (8), Y13 (4), Y14 (14), Y15 (5), Y16 (10), Y17 (31), Y18 (16), Y19 (3), Y20 (14), Y21 (12), Y22 (13), Y23 (1), Y24 (23), Y25 (27), Y26 (21), Y27 (30), Y28 (29), Y29 (16), Y30 (15), Y31 (6), Y32 (16), Y33 (15), Y34 (4), Chi (2), Ch2 (15), Ch4 (5), Ch5 (3), Ch6 (1), Dbl (3), Db2 (1), Db4 (1), Db5 (3), Db7 (2), Ntl (4), Nt2 (7), Nt3 (9), Nt4 (4), LI (5), L2 (5), L3 (5), L4 (7), L5 (6), L6 (8), L7 (10), L8 (19), L9 (10), L10 (12), L l l (17), L12 (8), L13 (12), L14 (11), L15 (1), Sal (2), Sa2 (1), Sa4 (2), Sa6 (3), Sa8 (1), Sa9 (1), St2 (1), St4 (1), Leil (7), Lei2 (6), Lei4 (1), Lei5 (1), Lei6 (4), Lei7 (1), Lei9 (5), Lei 10 (4), R1 (6), R2 (3), He7 (7), Wo4 (1), Wo5 (1), Wo7 (4), Wa2 (1), Wa3 (1), Wa7 (1), Nth2 (1), Nth4 (1), Nth5 (1), Hul (1), Nf2 (1), Nf7 (2).

The list below contains all tokens, including both the elicited and the incidental material. Nbl

III.8.1

Nbl Nbl Nbl

III.10.5 III.10.7 IV. 10.3

Nbl Nbl Nb2

VI.2.8 VI.17.7 1.1.6

Nb2 Nb2 Nb2

III. 12.4 III. 12.5 IV. 14.13

Nb2 Nb3

IX.7.2 III.1.12

Nb3 Nb3

V.7.18 IX.7.2

Nb4

1.2.2

Nb4

1.3.17

Nb4

1.5.2

Nb4 Nb4 Nb4

II.4.6 H.9.11 III.3.8

Nb4 Nb4 Nb4 Nb4 Nb5 Nb5 Nb5 Nb5 Nb5

III. 10.6 III. 10.7 IV.7.5 VIII.7.5 III.4.1 II1.10.1 III.10.7 IV.9.8 V.l.11

Nb6

IV .9.2

Nb6 Nb7 Nb7 Nb7 Nb7 Nb8 Nb8

VI. 14.13 II.2.9 III.2.8 IV.3.3 V.7.21 1.5.1 1.5.2

the sheep's onto the brick III. 10.7 dogs howls III. 10.7 sheep bleats III.10.7 the pigs is fond of them III. 10.7 the nannies has them some wears mits he could see how things was gonna gang ([waz gan a gain]) some pickles't some doesnae cure it III.10.7 trousers is the long straight ones IX.7.2 is them two married? III. 1.12 her bones has comen off. III. 10.7 some folk does who's them? ([hwe:z 0i:m]) III.10.7 the straw and hay is in the heck some puts a halter onto them some has an open bridle some calls 'em some teds it there's any amount takes cattle on some on 'em squawks III.10.7 sheep bleats people shoots 'em III. 10.7 VIII.7.5 burglars steals them some has a clarty hole III. 10.7 only the tups kens. III. 10.7 sheep bleats some calls them newts some is hung on the crooks. slugs bides on the ground VIII.5.1 (APP)them does the cows doesn Ί eat it these is split III. 10.7 where the crooks is on. V.7.21 the old taties is done some does some gans wi' open blinders.

174 Nb8 Nb8

1.7.8 II.4.5

Nb8 Nb8 Nb8

II.7.6 III.1.11 III. 10.7

III.10.7

Nb8 Nb8

VI.7.5 VII.2.2

III. 10.7 III. 10.7

Nb8 Nb9 Nb9 Cul Cul

VII.14.16111.10.7 II.9.6 II.9.3 III. 10.7 III.10.7 IV.9.6 VI.7.1 III.10.7

Cul Cu2

VII.6.17 III.10.7 II.4.3

Cu2 Cu2

III. 10.7 V.7.21

III. 10.7

Cu2 Cu3

IX.7.2 III. 10.7

IX.7.2 III.10.7

Cu3 Cu3

III.13.11 VIII. 1.24 IX.9.5

Cu4 Cu4

VIII.7.5 VIII.7.5 1.3.18 III.10.7

Cu4 Cu4

II.5.2 III.1.12

Cu4 Cu4 Cu4 Cu5

III.1.12 III.10.7 IX.4.1 III.9.4

III.1.12 III.10.7 III.10.7

Cu5 Cu5 Cu5 Cu5 Cu5

III.10.7 IV.6.13 V.7.21 VI.5.8 VII.6.17

III. 10.7 III.10.7 V.7.21 VI.5.8 III.10.7

Cu5 Cu5 Cu6 Cu6 Cu6 Dul Dul

VIII.6.3 VIII.7.5 III.10.7 V.7.21 VI.10.6 III.10.7 V.7.21

III.10.7 VIII.7.5 III.10.7 V.7.21 III. 10.7 III.10.7 V.7.21

Dul Dul

VI.3.10 VI. 14.1

III.10.7 III.10.7

III.1.12

some talks about some calls them clamps some people has nets some does sheep bleats (Fw.: "Some plurals are inflected.") these things happens rabbits kittles (=bring forth young) the fanners wears 'em A lot mows Sheep blears The boys catches them Mi hands is all hacked. ([mi hanz iz a: hakit]) The roads is gay clarty some doesn't snag them horses whinnies these is t' last of t' old taties is them two married? bulls croons, cows croons some does there's not so many fills a ten-quart tin burglars steals them all the byres is fothered off top ([o:t baiaz iz foöad Df tDp]) some doesn 7. t' bones is η't slipped plenty. t' bones is slipped cows heals shops keeps all kinds a lot of people 'at's never seen it cats mews the hens is in t' moult. t' old taties is done is thy teeth warking? thi shoes is all puddle today. the bairns gets holiday. burglars steals them bulls croons t' old ones is done your fingers swells up. sheep maes the old taties has gonen back my lugs is cold old women wears bonnets

Du2 Du2

III.10.7 IV. 10.6

Du2

V.7.17

Du2

VII.3.7

Du2 Du3 Du3 Du3

VIII.7.5 III. 10.7 III. 13.15 IV. 12.2

Du3 Du4 Du4

VIII.7.5 1.5.2 1.7.16

Du4 Du4

II.4.5 II.9.10

Du4

III.1.12

Du4 Du4

III.4.10 III.10.7

Du4 Du4

V.2.6 IX.8.6

Du4

CONV.

Du5

1.1.10

Du5 Du5

III. 1-12 III. 10.7

Du5 Du5

III.12.5 IV.7.7

Du5 Du5 Du5 Du5 Du6 Du6 Du6

V.7.21 VI.14.10 VII.1.10 VIII.7.5 III.2.6 III.10.7 IV. 12.5

Du6 Du6 Du6 Du6 C.P. CP

VI. 13.4 VIII.6.3 VIII.7.5 IX.7.7 V.3.3 VI.6.1

III.10.7 III.10.7

bulls bellows wood-pigeons feeds on them what s these at the top? it's getting very backendish now: the leaves has flown. VIII.7.5 burglars steals them III. 10.7 bulls croons them's rabbit-trimlings III.10.7 birds rests on (i.e. boughs) VIII.7.5 burglars steals them some calls 'em some calls 'em threelegs III.10.7 after the taties is took some does them theirselves III.1.12 the bones drops when she's close on calving a bonny lot doesn't III.10.7 sheep bleats (Fw.: "plurals are often inflected with s in ordinary conversation") some does these is mi father's boots, they're not mine. VIII.5.1(APP)they gang and never speaks when the beast gangs out (=when the cattle go out) her bones has dropped III.1.12 III.10.7 bulls blows, cats miaows some does when them gets hold of you V.7.21 t' old ones is done where's my gallowses? some calls it a dozen. VIII.7.5 burglars steals them some calls 'em III.10.7 sheep blares III. 10.7 a lot of people lops 'em III.10.7 birds gapes for worms t' bairns gets holiday III. 10.7 VIII.7.5 burglars steals them them is IX.7.7 them's the oldest thing them's things I's never had a great love for

175 Wv. Wv.

VI. 11.3 VIII.4.9

Wei Wei We2 We2

1.3.1 III.10.7 III. 10.7 IV. 1.3

III. 10.7 III. 10.7 III.10.7

We2

IV.9.4

III. 10.7

We2

V.7.21

We2 We3

VI.4.6 1.10.6

III. 10.7 IX.9.5

We3 We3

III.1.12 III.10.6

III.1.12 III. 10.6

We3 We3

IV.8.4 VI. 14.17

We3 VIII.7.5 VIII.7.5 We4 III.1.12 III. 1.12 We4 I1I.8.1 We4 III. 10.7 III. 10.7 We4 IV.9.8 We4 V.4.4 We4 V.7.21 V.7.21 We4

We4

VI.14.17

VII.7.6

III.10.7

We4 VIII.6.3 We4 VIII.7.5 We4 IX. 1.9 Lai III.1.12 Lai V.7.21 Lai VIII.7.5 Lai IX.5.6

VIII.6.3 VIII.7.5 IX. 1.9 III.1.12 V.7.21 VIII.7.5

Lai Lai La2

CONV. CONV. II.4.4

III. 10.7 III. 10.7 III.10.7

La2 La2 La2 La2 La2 La2 La2 La3 La3 La3 La3

III.1.12 III. 1.12 III.10.7 III.10.7 III.12.5 IV.3.6 V.7.21 V.7.21 VII.6.1 IX.7.5 IX.7.5 III.1.12 III.1.12 III. 10.5 III. 10.7 IV.6.2 VIII.7.5 VIII.7.5

all them warts is ulcers tinkers... goes around the country wagoners uses them bulls croons bulls beats t' next lot that comes ([kuz]) t' baims goes here for them. these is t' last o ' t ' taties pigs has snouts. there's some has them up. t' strings has gone it's nobbut bairns 'at yowls. (fleas) 'at bites you. ...that holds 'em around = that hold them around here burglars steals 'em. t' slacks is gone pigs has snouts bulls bawls them's t' same ones. them s hot. these is t' last o ' t ' old taties Where's my yorks at? (=where are my yorks?) beast is selled (=cattle are sold) children has holiday burglars steals 'em the youngsters creeps the slacks has gone t' old ones is done burglars steals 'em both on 'em stands here great snows comes doors hangs the tops goes to the cows. her slacks drops sheep bleats some puts it in a tub some goes that way t' old ones is done some does isn 7 'em? t' slacks is gone dogs barks some keeps 'em burglars steals 'em

La4

II.3.6

La4

III.1.12

III. 1.12

La4 La4 La4

III.10.7 IV.8.11 IV. 9.6

III. 10.7

La4 La4 La4 La6

IV. 11.2 V.7.21 IX.7.2 1.5.12

III.10.7 V.7.21 IX.7.2

La6

1.7.3

III. 10.7

La6

II.1.7

La6

II.2.1

La6

II.2.11

La6

II.4.1

III. 10.7

La6

II.4.3

III. 10.7

La6

11.9.4

La6

III.4.3

III. 10.7

La6 La6

111.5.2 III.8.5

III.10.7

La6 La6 La6 La6 La6 La6

III.10.7 III. 10.7 III. 10.7 IV.2.8 IV. 7.2 IV.8.4

III.10.7 III.10.7 III.10.7 III.10.7

La6 La6

IV.8.13 IV.9.8

La6

IV.9.11

La6 La6 La6 La6

IV. 10.7 VII.5.11 VII.6.14 III. 10.7 VII.8.10 IX.9.5

III.10.7

some does wi' both and some wi' one. her slacks is going down bulls bawls is 'em? (=are they) doesn 7 'em? (=don't they) some folk does th' old ones is done is 'em married? what a deal of people uses (=what a lot of people use) them kists was on purpose to put provand I thought them was peas. them doesn't grow much a lot of folk grows 'em i ' t ' dark. far apart and bigger your plants grows (=the farther apart, the bigger your plants grow) farmers carts 'em and gives ([giz]) 'em to the cows them follows t' machine abound two or three yard they gone out of date, is horses. the horses is baiting they're not worth bringing up, isn't little pigs ([Sa'nat...]) bulls bells dogs kills 'em these vets cuts 'em off birds lives on them the old ones feeds 'em them big ones as makes grubs them s i ' t ' dry cops ducks goes up the dykes and eats 'em all. sin t' docks was openend a lot of baskets is made them as wasn Ί roads is slippy There's many a lot axes for a pipe of 'bacco.

176 La6

VIII.5.1

La6 La6

VIII.9.3 VIII.9.5

La6 La6 La7

IX.2.10 IX.6.3 1.1.8

La7 La7

1.3.11 II.2.3

La7

II.7.8

La7

H.9.3

La7

III. 1.3

La7

III.4.7

III. 10.7

La7 La7

III.5.2 III.7.8

III.5.2

La7

III. 11.9

La7

IV. 1.8

La7

IV.5.2

La7

IV.7.8

III. 10.7

IV.5.2, III.10.7 III. 10.7

La7 La7

IV.9.3 IV.9.3

III. 10.7 III. 10.7

La7

IV. 10.3

III.10.7

La7

IV.10.12 III.10.7

La7 La7 La7

IV.10.13 III.10.7 IV.12.3 III. 10.7 V.l.11 III. 10.7

La7

V.3.3

La7

V.3.8

La7

V.6.5

IX.9.5

La7

V.7.21

V.7.21

La7

V.10.2

La7

VI.4.20

IX.9.5

some folk goes to t' church; they goes to t' church. little pigs has big ears, they peel 'em and boils 'em any on 'em as is near buses comes through them's the folks as said it. the tanks is still there them's the worst to get shoot on you'll get odd uns on land as is let run wild some opens corners out wi' a scythe them's a fine lot of beast his stones hasn t dropped the horses is baiting some has stocks and some taken 'em on their back on t' floor our own is (=our own are), no context given these as goes out into t' Wyre (=a river) the horses is baiting it wings is webs (=its wings are webs, said of a bat) chickens likes 'em. gorging-grubs (=slugs) is eating 'em off them things lads picks up off t' wood, some folk dyes easter eggs some folk comes here them t'others is leaves door studs, what t' doors is hung to. you don't see many has holes now t h e m ' s old-fashioned, ain't they? there's lots of people bakes their own these is t' last o' th' old praters (=potatoes) I can't thread mi needle, mi eyes is not so good. there isn't many folk wears garters now.

La7

VI.5.7

La7

VI.5.10

La7

VI.7.9

La7

VI. 10.6

La7 La7 La7

VII.6.14 III. 10.7 VII.6.15 VIII.7.5 VIII.7.5

La7 La7

VIII.7.6 IX. 11.5

La7

CONV.

La8

II.7.3

III. 10.7

La8 La8 La9

III.10.7 V.2.13 III.7.6

III.10.7

LalO V.7.21 La 10 VI. 12.3 L a l l 1.1.6 Lall

1.3.4

Lall Lall

III. 10.7 V.7.21

Lall

VI.5.11

V.7.21 III. 10.7

III.10.7 V.7.21

L a l l VIII.7.5 VIII.7.5 La 12 III.10.7 III. 10.7 La 12 V.8.4 La)2 Lal2 La 12 Lal3

V.7.21 V.7.21 VI.3.1 III. 10.7 VIII.7.5 VIII.7.5 1.9.1 III.10.7

Lal3 III. 1.11

III. 10.7

La 13 La 13 Lal3 Lal4 Y1

III.2.3 III. 10.7 V.7.21 VI.7.10 II.5.3

III. 10.7 III. 10.7 V.7.21

Y1

III.3.3

III. 10.7

Y1

IV.2.8

III. 10.7

IX.9.5

eye-teeth, some calls them. it's thy gums as is aching anybody as bites their nails, they're badtempered. thi feet gels too big for thi shoen when tha has them there. the roads is very slippy them as shapes they break into houses and steals dogs buries bones lads as is rougher than yourselves. when potatoes was cheap spoutings is round a house ([spa:tinz...a:z]) bulls bellows some calls 'em some clips on a stock (=some clip on a stock) th' old praters is done a good deal was beast ([biaz]) is in meadow. some soles 'em up wi' a rope bulls bawks th' old praters is gone done some o ' t ' childer comes burglars steals 'em dogs barks as is soft (=which are soft) (no more context) all t' praters is done his eyen's coming out burglars steals 'em. the spokes fastens on felloes. odd uns does come that way odd uns parts with it bulls bawks the old ones is done some calls 'em there's some hasn't awns on t' bairns has played all sorts w i n ' t ([wint] = with it) t' old hares gets

177

Yl

V.l.8

Y1 Yl

V.7.21 V.7.21 V1II.6.3 III.10.7

Yl

IX. 1.5

Y2

1.1.2

Y2

1.5.3

Y2

1.7.16

Y2

II.9.11

Y2 Y2

III.2.8 III.3.5

III. 10.7

Y2 Y2 Y2

III.5.2 III. 10.7 III. 11.9

III.10.7 III. 10.7 III. 10.7

Y2 Y2 Y2

III. 13.7 III.10.7 111.13.13 V.l.12 III. 10.7

Y2 Y2

V.7.21 V.8.10

Y2 Y2

VI.5.10 VI.7.3

Y2 Y2

IX.9.5

V.7.21 III. 10.7

VI.7.13 VI. 10.1

Y2 Y2 Y2 Y2

VI. 11.4 III.10.7 VII.2.8 VII.5.11 III. 10.7 VIII.3.5

Y2 Y2 Y2 Y3 Y3 Y3 Y3

VIII.7.5 VIII.7.5 VIII.7.7 IX.8.7 1.10.5 III.4.9 III.4.10 III.6.7

Y3 Y3

III.10.7 V.3.10

Y3 Y3 Y3

V.6.9 VI.7.1 VI.7.15

III. 10.7

III. 10.7 IX.9.5

through them some of these doors is same old taties is done the children has a holiday I can picture people 'at has said it. there's plenty of farmers keeps they'll not know what these is those here men 'at's in now hay-tethers is gone out of date them's t' hoofs, my daughters pulls me up t' horses is baiting bulls bellows kennels generally gels 'em cats does and all rabbits does stable doors turns on crooks and door-bands the old taties is done some folks doesn Ί use it. my gums is sore my hands is naught but chops, and they're terribly sore, these is peas my feet's bad (but homophonous: my foot's bad, [fiats]) his hands is very horny them fronts lets down clocks is coming they're real hard gossips, is them, burglars steals 'em both means t' same, these is this cow's legs them what has them bones what sticks out them s th' hoofs, they're rough mutton, is tups, bulls beats them's done away with. some on 'em says what my hands is frozen when there's four comes

Y3 Y3 Y3

VI. 14.8 VII.4.9 VII.6.2

Y3 Y3 Y3

VIII.7.5 VIII.7.5 VIII.8.2 CONV. III.10.7

Y4

1.7.8

III. 10.7

Y4 Y4 Y4

1.11.5 III.5.2 V.l.9

III. 10.7 III.5.2

Y4

V.6.1

III. 10.7

Y4 Y4

VI.2.9 VI.7.4

Y4

VI.13.14 III.10.7

Y4

VI. 14.6

Y4

VII.3.18 III. 10.7

Y4

CONV.

Y5 Y5

1.9.7 III.8.9

Y5 Y5 Y5

III.10.4 III. 10.7 V.7.21

III. 10.7 V.7.21

Y6

1.2.4

III. 10.7

Y6

1.3.14

III.10.7

Y6

1.11.5

Y6 Y6

III. 10.7 VI.8.1

Y6

VI. 14.18 III. 10.7

Y6

VII.8.21 III.10.7

Y6

IX.4.8

Y6

IX.4.8

III. 10.7

Y6

IX.7.11

IX.9.5

Y6

IX. 10.5

Y6

IX. 10.5

III. 10.7 III. 10.7

III. 10.7

III.10.7

III.10.7

which is them? the bells rings at night clouds is low, it's very dull burglars steals 'em them six has been off 't wurzels J liking well ([t wuzalz liakan], =the wurzels look well) farmers makes their own seaves is for thatching t' horses is baiting them's corned up ([komd]) the fine sharps is taken out of t' meal some on 'em rives it my lugs is getten frostbeaten the old calves dies wi' that bairns has jackets, t' old men has coats till his eyen was quite red (=his eyes, [i:n waz]) dockings is the worst grass. them's it inside folk hasn Ί had much to do wi' pigs some does do bulls croons them's t' finish-up o' t' old taties. when the tups is brought off hoof (?) them kind of besoms is gone out them 's nae good, will throw it up. bulls croons animals 'at puts their tongues out loughrams is things 'at are knitten farmers goes out and they drives 'em off others does what they shouldn't some children does that there's a lot of houses has one. them s farther off and yon's farther off. thur's near to me

178

Y6 Υ6 Υ6 Υ6 Υ6 Υ6 Υ6 Υ6

Υ6 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ7 Υ8 Υ8 Υ8 Υ8 Υ8 Υ8 Υ8 Υ8 Υ8 Υ8

([3orz]) (thur = these) all them houses is bluebottles goes for III. 10.7 that CONV. III. 10.7 cows is kept CONV. III. 10.7 farmers sells it CONV. some smits 'em on t' cheek CONV. them 'at hasn Ί CONV. these tups has... CONV. III. 10.7 you know what rattens and mowdies is ([ratanz an maodiz]) CONV. III. 10.7 them calves's had dry feed III.1.12 III. 1.12 her strings is dropped III.5.2 III.5.2 the horses is baiting III.10.8 III. 10.7 the cobblers puts 'em on III.10.8 III.10.7 the cows eats it III. 11.2 them's draft yows IV.6.19 III. 10.7 't hens is fair floutered (=frightened) to death IV.10.2 IX. 10.4 them s them V.7.9 t' taties is terrible wallow V.7.15 VIII.5.1(APP)them wants a stiffer soil V.10.1 III. 10.7 the things u r n knit (=knitted) VI. 11.4 they're segs, is them. VI.13.3 my legs aches a bit VII.3.7 III.10.7 backend starts after the cows is tied up VIII. 1.7 III. 10.7 mi father and mi grandfather wasn Ί IX.3.9 has thi taties corned up yet IX.7.2 is both you women wed? IX.7.7 both on us is English IX.7.9 my wife's folks has lived there III.10.7 1.9.6 spokes is all the same. IH.10.7 I doubt my days is II.4.6 numbered now III.1.12 her tits is waxy III.1.12 III.1.12 her strings is going III.10.7 III. 10.7 bulls beats III.11.5 III.10.7 long things that cats generally gets some on ' e m p u t s 'em III.12.6 on a table IV. 10.8 III.10.7 squirrels picks 'em for the winter's keep VI.5.8 VI.5.8 does thy teeth wark? VI. 10.1 there's lots on 'em has one foot now CONV. CONV.

Y8 Y9 Y9

CONV. 1.1.3 1.5.10

III. 10.7

Y9

1.7.9

Y9

1.9.3

Y9

II.4.7

Y9

II.6.7

II.6.7

Y9 Y9 Y9 Y9

II.9.3 III.5.2 HI.10.7 IV.8.3

III.5.2 IH.10.7 III.10.7

Y9

IV.9.3

Y9

IV.9.9

Y9

V.1.9

Y9

V.4.5

Y9 Y9

V.5.1 V.5.2

Y9

VI.8.6

IX.9.5

Y9

VI. 11.4

HI.10.7

Y9

VI.14.13 III.10.7

Y9

VII.5.3

Y9

VIII.3.7

Y9 Y10

VIII.7.5 VIII.7.5 1.1.8

Y10

1.1.9

Y10

1.5.1

Y10

1.10.6

Y10

II.4.2

III.10.7

Y10 Y10 Y10

III.3.7 III. 10.5 III. 10.7

HI.10.7 IH.10.7

III.10.7

beast eats it funds is over low these is t' trappings 'at gans over the loins garden gripes is flat, farmers' gripes is round. a winter sled 'at baims plays wi'. hazels is the best stack-stubs them 'at's on the wagon loadens it a lot of folk doesn't horses is baiting bulls bellows I'll bet those kids has crums (=lice) in their hair them snails is eating all mi cabbages as long as our names isn Ί there gates has snecks on and all there's a hole here, where t' ashes drops folks is different some puts milk i ' t ' cellar. there's few mothers feeds a child some farm lads gets 'em athout ([aöu:t] = without) breechings goes round the arse them's good enough to see all sorts of jobs as isn Ί my own burglars steals 'em some do and some doesn Ί if there's happen one or two out 'at comes to your engine these goes over this here swingle-tree. them goes on when they're leading hay or straw. the orchard, where t' apple trees and t' pear trees grows some on 'em does that t' old yows is blaring cows bellows, cats

179

ΥΙΟ ΥΙΟ

IV.6.18 IV.11.3

ΥΙΟ ΥΙΟ

V.5.10 VI.5.IO

ΥΙΟ

VIH.1.1

ΥΙΟ ΥΙΟ

VIII. 1.8 VIII.7.7 III.10.7

ΥΙΟ

VIII.8.14

ΥΙΟ ΥΙΟ

CONV. CONV.

III. 10.7 ΙΧ.9.5

ΥΠ ΥΠ

1.1.11 1.7.16

III. 10.7 III.10.7

ΥΠ

1.8.2

ΥΠ

Χ.5.3

Υ11 ΥΠ ΥΠ ΥΠ

Χ. 11.5 Χ.15.5 II.5.1 Ι1Ι.2.3

ΥΠ ΥΠ ΥΠ ΥΠ

III.2.8 ΙΙΙ.5.2 ΙΙΙ.7.11 ΙΙΙ.8.1

ΥΠ ΥΠ ΥΠ Yll Yll

III.8.7 III. 12.3 III.10.7 III. 12.3 III.10.7 III.13.5 VII.4.11 III. 10.7

Yll

VII.8.5

Yll

VIII.4.4

III. 10.7

III.10.7

ΙΙΙ.5.2

III.10.7

Yll Yll

VIII.4.8 IX.7.3 IX.7.2

Yll

CONV.

Y12 Y12

III. 1.12 III. 8.1

III. 1.12

miaows, bulls bellows what s them? same as these as grows... (=these that grow) some calls it beastlings my gums is sore ("older", [sea]) vs.: my gums are sore ("modern", [sua]) some on 'em has up to twelve on 'em yes, some folk does them lads is chucking stones both thi hands and thi face is darted up wi' treacle nits is t' eggs there's none uses them now. Dutch barns is run out my boots isn 't theat (=water-proof) old-fashioned ploughs, is these. isn Ί both on 'em getting wed soon? clouds is low these is mi own them 's swads is mi glasses there, missus? them's the hoofs horses is baiting these is bars they're rum things, is the pigs. and t' others is boars and t' others is hams farmers doesn Ί some has kiddies gets Christmas boxes some folks thinks they're rich if they have a hundred some isn't much at weals ([wi:lz] = isn't much good at) hooping iron onto t' wheels a lot 'at gives over isn't both on 'em married they're cheap, is kittens her slacks is dropping they're a bit queer, is

Y12 Y12 Y12

IV.6.10 IV.6.15 VI.14.8

III.10.7 III.10.7

Y12 Y12 Y12 Y12 Y13

VIII.7.5 IX.7.2 IX.7.5 IX.7.7 III.7.8

VIII.7.5 IX.7.2 IX.7.5 IX.7.7

Y13 Y13

III. 10.7 IV.1.11

III. 10.7

Y13 Y13

VI. 14.16 IX.10.4 VIII.7.5 VIII.7.5

Y13

IX.2.9

Y14 Y14

1.7.3 1.7.9

Y14

1.7.12

Y14

1.9.4

Y14

1.10.6

Y14 Y14

1.10.7 X.2.12

Y14

II.2.7

Y14

III. 11.9

Y14

V.7.3

Y14

V.7.15

Y14

VII.6.5

Y14

VIII.4.7

Y14

VIII.4.8

Y15 Y15 Y15

III.1.12 III.1.12 IV.8.6 VIII.4.8 III. 10.7

Y15 Y15 Y16

VIII.7.5 VIII.7.5 CONV. III.10.7 1.9.2

Y16

X.l.ll

t' eggs is chipping mi hens is moulting them 'at does wear 'em ([at doz wiar]) burglars steals 'em is 'em married? isn Ί 'em? 'em is some clips 'em on t' ground and some on a stock bulls croons you chaps is in f o r ' t . ([jii: tfaps iz]) them s out of date (burglars) steals (them) some skellers w i ' t ' sun (=some [doors] warp in the sunlight) some sews 'em up some has four (=some [forks] have four [prongs]) they're all shafts, is them. these hasn Ί been ago so long in this country them as goes back and forwards ... is these. sheep 'at's been buried a fortnight them doesn 7 come up while about April some doesn t get up while dinner-time is them porridge ready? his eyen's as red as shoops ([iz i:nz az]) a lot of people goes by that t' old times is better nor t' new ones, them people has to work to get it her strings is gone these is bad enough odd uns works most o' t' Sunday burglars steals 'em thi feet is mucky now them's oldfashioned there isn't many 'at doesn't know

180 Y16

Y16 Y16 Y16 Y16 Y16

Y16 Y16 Y17 Y17 Y17

Y17

Y17 Y17 Y17 Y17 Y17

Y17 Y17

Y17 Y17

Y17 Y17 Y17 Y17 Y17

Y17

X.4.4

we're both right, aren't we, it's t' others 'at's wrong X.12.2 III. 10.7 poultry loses their feathers X.12.4 folks 'at's found inside III.7.1 III.7.1 sucks (Fw.: "3 pr. pi.", no further context) III.7.11 them s the things VII.2.12 them 'at has most pudding will have most meat VIII.4.12 chaps 'at gangs from farm to farm VIII.7.9 them houses is pulled down 1.3.17 VIII.5.1(APP)they go in and cuts 'em down II.6.3 six sheaves makes a stook II.8.2 they throw that down and picks another sheaf up II.8.2 you stack it up and carls it up to where you want it. II.9.2 they get done, does meadows. II.9.10 some uses a strick III.1.5 there is some sucks themselves III. 10.7 they roar, does a bull, and snorts. III. 10.7 III.10.7 sheep baes, lambs maes, cats mews, dogs barks III.12-5 we always kills 'em and cures 'em wi' salt. IV.7.8 III. 10.7 partridge comes in in September (=partridges come) IV.8.10 IX.9.5 there's lots of childer calls 'em that. them t'others has been IV.9.3 agate (=have started) eating the cabbage them's t' tongs for V.3.7 lifting coal on t' fire. V.3.10 them s t' bellowses. V.7.12 them s pea swaths (=pea pods) V.7.18 them t'others is runner-beans V.7.20 you put middling of water into it and pulls it down V.8.3 them 'at doesn Ί eat meat

Y17

V.10.7

Y17

VI. 12.4

Y17

VI.13.3

III.10.7

Y17

VI.14.1

III.10.7

Y17

VI. 14.11

Y17 Y17

VI. 14.25 VII.6.14 III. 10.7

Y17

VII.7.2

Y17 Y17

VII.7.10 IX.7.2

Y17 Y17 Y18

IX.8.8 CONV. 1.3.16

Y18

1.7.15

1.7.14

Y18

1.8.3

III.10.7

Y18

1.9.12

Y18

X.5.2

YI8

II.5.3

Y18

II.6.8

III. 10.7

Y18 Y18 Y18

III.3.6 IV. 10.8 VI.13.5

III.10.7 III.10.7

Y18 Y18

VI.14.22 VI.14.22

Y18 Y18

VIII.5.5 VIII.5.10 III.10.7

Y18

CONV.

Y18

CONV.

Y19 Y19 Y19 Y19 Y20 Y20 Y20

V.11.2 III. 10.7 V.7.21 V.7.21 VI.13.5 III. 10.7 VIII.7.5 VIII.7.5 1.13.2 II.3.5 11.5.2

IX.9.5

III. 10.7

III. 10.7

there's a few takes their knitting with 'em they're very nasty, is mumps mi guts aches this morning it's what persons wears when they're in mourning at a funeral a deal of folks calls 'em these is made t' roads is right slaip ([sle(a)p] = slippery) farthings hasn 't been much i-go t' others is pound notes them 'at hasn't had schooling them t'others is snecks them 's Gargrave folks 'at cow hind feet stands on (=that cow's...) they vary, does steesteps what t' horses pulls it with (=by which the horses pull it) they've recently corned, has them. there's many a lot of them comes to me they're rum things, is barley-horns tines isn 7 so long, neither, ([tainz]) some seals t' paps up hazel-nuts grows horses sweats and all (=too) them 'at has no top they're laced boots, is these of mine. them is tombstones if ever t' joiners comes here t' end of t' gate, where t' crooks is on chickens hatches out o ' t ' eggs women doesn Ί wear t' old ones is done horses sweats burglars steals 'em them's thistles sheep sees them's good heads

181 Y20 Y20

IV.6.2 IV.6.8

Y20

IV.9.3

Y20

IV. 10.9

Y20

V.7.12

Y20 Y20

Vl.2.7 VI.5.9

Y20 Y20

VI.7.7 VII.3.17

Y20

VII.4.5

Y20

CONV.

Y21

1.1.8

Y21 Y21

1.5.9 II.2.8

Y21

III.1.16

Y21 Y21 Y21

III.5.2 III. 10.5 IV.3.4

Y21

VI.4.1

Y21

VI.8.3

Y21 Y21 Y21

VI. 14.25 VIII.7.5 VIII.7.5

Y22 Y22

1.3.13 1.6.4

Y22

II.2.6

Y22

II.3.1

Y22

III.8.1

Y22 Y22

III.10.7 III. 10.7

Y22 Y22

III. 11.2 III.11.2

some keeps ducks them 'at has some fowls they're a nuisance, is them things IX.9.5 we have some grows i' t' village them s a fine lot of peas these is sideboards them s to chew the meat with III.10.7 my hands is terrible any amount of farm fellows is paid at Saturday. III.10.7 shops was open while ten III. 10.7 them things is handed down III.10.7 a place where t' cattle roams these is t' breechings they others grows ([3iar oöaz]) them beast has corned up. III.5.2 us horses is baiting III. 10.7 sheep bleats the hingings (=hinges) dons in (=fit on) them 's us ears (=these are our ears) there is childer 'at has... some has leather laces burglars steals 'em VIII.7.5 III.10.7 some folk buys the coals, but some folk steals 'em these is muck-forks IX.9.5 there's a lot uses tractors docken and t' thistle is t' worst, but we haven't many. you go fair down t' middle and pulls one each way. III.10.7 some folks brings pigs up with a bottle III. 10.7 cats mows chaps 'at's i ' t ' band likes to come III. 10.7 all cows is graded. VIII.5.1 (APP)they feel over 'em and weighs 'em, does butchers.

Y22 Y22 Y22

IV. 1.4 IV.6.10 VI.7.1

Y22

VI. 14.14 IX.9.5

Y23

VI. 11.2

Y24

1.5.12

IX.9.5

Y24

III.10.7

III.10.7

Y24 Y24

III.10.7 IV.3.10

III. 10.7 III.10.7

Y24

IV.4.1

Y24

IV.6.4

Y24

IV.9.2

Y24

V.3.10

Y24 Y24

V.7.21 VI.1.5

V.7.21 III. 10.7

Y24

VI.13.3

III.10.7

Y24

VII.5.12

Y24 Y24

VII.6.2 VII.8.21 IX.9.5

Y24

VIII.4.7 III.10.7

Y24 Y24 Y24 Y24

VIII.6.3 III.10.7 VI11.6.3 VIII.7.5 VIII.7.5 VIII.7.5

Y24

IX.2.13

Y24

IX.4.I3

Y24

IX.4.17

Y24

IX.5.6

III.10.7

Y25

V.7.21

V.7.21

Y25

1.1.11

III. 10.7

Y25 Y25

1.5.2 1.9.4

III.10.7

III.10.7

III.10.7

half on 'em's dead. eggs « j u s t hatching my hands is gotten right chapped w i ' t ' frost. there's none so many doesn 't some wrinkles a lot sooner nor others there isn't many wants whips. bulls beal, horses whinnies fellies goes round t' horse's feet pogs into't ground (=sink into the ground) them 'at falls generally have a maggot in they're dear, is eggs, now. leaves is beginning to drop these is t' bellowses and them's t' tongs. t' old taties is all gone it's where t' barracks is my guts aches this morning plenty takes their dinners with 'em these is clouds. there's not as many comes as what used to do. farmers has to work on Sundays. bairns gets holiday most likes to laik burglars steals 'em burglers goes and steals they won't know what taties is. most on 'em goes and takes one. they're daft 'at makes butter. maybe t' kids has done it these is all t' old taties we have left ([ciiaz ist]) things these farm-lads doesn Ί know nought about know that them is them's inside o ' t ' iron

182

Y25

X.4.3

Y25

X.15.6

III 10.7

Y25 Y25

II.9.1 II.9.10

Y25 Y25 Y25

HI.8.1 IV.5.7 IV.6.22

Y25

IV.8.4

Y25

IV. 11.2

Y25 Y25

1V.12.3 VI.2.8

Y25 Y25

VI.7.3 VI. 13.4

Y2S Y25

VI.14.13 VII.6.11

Y25 Y25 Y25

VII.6.17 III.10.7 VII.7.9 VII.7.9

Y25 Y25

VIII.2.8 VIII.2.8 CONV.

Y25

CONV.

Y25

CONV.

Y26

1.8.3

III.10.7

Y26 Y26

III.5.2 III.7.9

III.5.2

Y26

III. 10.4

Y26

IV.7.6

III. 10.7

Y26

IV.8.3

III. 10.7

Y26

IV.8.6

Y26

IV.9.2

Y26

IV. 11.8

Y26

V.l.6

Y26

VI. 1.3

III.10.7 III.10.7

III.10.7

III.10.7

i' t' wheel both on 'em hasn Ί been very well. these legs is belonging that cow (=are that cow's) some says grass stock likes it (=cattle like it) funniest things, is pigs. stoats does sheep does it more than cows they always crawl upwards, does lice. they talk to you, does grass(??) them's leaves them's sideboards what was cut off my hands is chapped they're very affectionate, is pigs. them s breeches them's ice-candles (=icicles) your boots is white some calls it that some on 'em calls 'em that yet. how s things? implements is seasonable things some s wider (=some are...) there's a lot doesn Ί know that where t' horses is fastened on to. the horses is baiting dogs has driven 'em out cows is bleating out again them big owls catches vermin pig lice crosses t' pig back every twenty-four hours flies has started astirring. leather-jackets lives on cabbage these is peas, aren't they ([o:nt 3a]) eaves is where the roof starts them s fruzzings.

Y26 Y26

VI. 11.5 VI. 11.8

Y26

VI 1.2.11

Y26 Y26

VIII.6.3 VIII.6.4

Y26 Y26

VI II.7.5 VIII.7.5 IX. 1.9

Y26

IX.1.11

Y26

IX.9.4

Y26 Y27 Y27 Y27

IX.10.4 1.1.9 1.3.7 1.5.12

IX.9.5

Y27

1.9.6

III. 10.7

Y27 Y27 Y27

X.2.16 II.4.6 II.8.5

III.10.7

Y27 Y27

II.9.8 III.5.1

Y27 Y27

III. 11.6 III.10.7 III. 11.12 III. 10.7

Y27

IV.6.13

Y27 Y27

1V.6.18 IV.9.3

Y27

IV.10.3

III. 10.7

Y27

V.l.3

III.10.7

Y27

V.l.16

Y27

V.6.3

Y27

V.7.15

Y27

VI.5.11

Y27 Y27

VI.7.5 VI.10.10

Y27 Y27

VI. 13.3 VI. 14.11

III. 10.7

III.10.7

blebs raises up carbunkels has generally three or four holes in 'em them i' t' tax office has no work to do childergeto- holiday there's lasses 'at takes to it burglars steals some creeps and some lurches (said of babies) things starts a-going round you do find odd uns what's woven. them 'at deals some calls it some on 'em's iron. there's very few uses whips. where all t' spokes fits in. some folks calls it them 's big uns. some on 'em's laid to blow t' chaff off. them s them. has them horses been served? (=fed) some people does some people calls 'em that we ('re) not getting eggs, all t' hens is i ' t ' moult. them 'at's profitable some calls 'em cocktails. oak-apples grows on oak-trees. cow-hull, where t' beast goes in (=cattle go in) scavengers comes round and takes 'em away them 'at gets t' last share gets t' best t' pea-pullers is off by. (=are going past) there's a lot o f ' e m 'at's died out these is your knuckles. some s six foot and some s five foot your legs is bad I think them's t'

183

Y27

VIM.17

Y27

VII.6.14

Y27

VIII. 1.2

Y27

VIII.5.14

Y28 Y28 Y28 Y28

1.1.9 1.3.15 1.9.9 1.11.8

III. 10.7 III. 10.7

Y28 Y28 Y28 Y28

II.5.1 III.5.2 III.5.2 III.9.4

Y28

III. 10.7

III.10.7

Y28

IV. 1.7

III. 10.7

Y28

IV.3.10

III.10.7

Y28

IV.3.18

III.10.7

Y28

IV.4.1

III.10.7

Y28 Y28 Y28

IV.5.8 IV.5.11 IV.6.1

III. 10.7

Y28

IV.7.2

III.10.7

Y28

IV. 10.2

Y28

IV. 11.4

Y28 Y28

V.7.12 V.7.21

V.7.21

Y28

VI. 1.3

III.10.7

Y28 Y28 Y28 Y28

VI.1.7 III. 10.7 VI.2.7 VI.6.16 VI. 14.20

Y28

VIII.5.8

III. 10.7

Y28

VIII.6.6

III. 10.7

Y28

VIII.8.19

IX.9.5

wannest there's more has one (=there're more people who have one) the roads is slaip (=slippery) them that was on that ship 'at went down them s mostly for soldiers' graves. pigs runs in some people does them's felves where's them come fra? (=where have they come from) half o n ' e m was apt. horses was baiting horses is baiting when two horses gels together bulls beals, cows blares roads is full of potholes t' horse feet sinks in t' ground them cows and things is making a bonny track all them apples is tumbling your hands smells some on 'em does they're ferocious things, is badgers them birds is fit for flying a lot of folks makes 'em there's some down here grows is taties fit? taties is done, this is last on 'em. your hands gets chapped lots of people says it some on 'em calls 'em some does them that doesn 7 know some spots is soil, some is clay, and you can hardly dig it. them what geese tears up most on 'em is

Y29

1.8.4

Y29

1.9.6

Y29

III.3.7

Y29

III.5.2

Y29 Y29 Y29 Y29 Y29 Y29 Y29 Y29 Y29 Y29 Y29 Y30

Y30

Y30 Y30 Y30 Y30 Y30 Y30 Y30 Y30 Y30 Y30 Y30 Y31

Y31 Y31

III.10.7

these here small whipple-trees is hooked on to the end of this rock-tree these here 'at goes around them 'at doesn 7 pay him the horses is baiting

III. 10.7, III.5.2 III.10.2 III.10.7 women wears 'em III.10.3 III.10.7 cows gets 'em III. 11.4 VIII.5.1(APP)they rope 'em and then pulls 'em in. childer calls 'em III.10.7 IV.9.6 IV. 11.1 childer'i off blagging (=blackberrying) V.6.10 muffins is round ones and loafs is oblong. V.8.10 some always sups i ' t ' bait. VI. 11.5 them's sore and all if you burst 'em VI.14.2 farmers has ear-caps VI.14.10 them s the galowses VII.4.6 mills is running now II.7.3 t' housings (?) (=eaves) is dropping again II.8.5 them ' a t ' s nearest t' back-door, them's t' best oats([wDts]), and them 'at's nearest t' chaff is t' light oats. III.1.5 small ones 'at keeps a two-three milkers III.6.2 III.10.7 his father and mother lives there III. 10.7 III.10.7 bulls roars IV.3.9 them's foot-marks of a horse IV. 10.6 yon turnips is sprouting IV.10.12 them's young bracken, aren't they? IV. 11.8 them's peas V.3.10 these is t' bellowses V.7.17 kidney-beans, them's t' worst to get ready V.11.3 plenty folk does IX.3.10 III.10.7 gardeners always puts 'em i ' t ' sun III.2.10 VIII.5.1(APP)they run at one another and brods (=prick) one another. III.3.3 they reckon them's t' best III. 11.6 some calls it pluck

184 Y31

IV. 12.2

Y31

VI.5.9

Y31

VII.5.12 III. 10.7

Y31 Y32 Y32 Y32

VIII.7.5 VI1I.7.5 1.1.10 III. 10.7 1.3.8 1.7.8

Y32 Y32

1.10.5 III.5.3

III.10.7

Y32

III.9.2

III.10.7

Y32 Y32

IV.3.11 VI.4.1

Y32 Y32

VI.5.11 VI.7.3

Y32

VI.13.15 III. 10.7

Y32

VIII. 1.8 III.10.7

Y32 Y32

VIII. 1.10 III.10.7 VIII.3.5 III. 10.7

Y32 Y33

VIII.8.6 VIII. 1.24, III. 10.7 1.3.18

Y33

II.2.8

Y33

II.4.4

III. 10.7

Y33 Y33

11.4.6 III.1.16

III. 10.7

Y33

III.4.7

Y33 Y33 Y33 Y33 Y33 Y33

IV.6.2 V.7.2I VI.4.1 VI.7.9 VI.9.4 VI. 13.20

Y33

VI. 14.6

Y33 Y33

VI. 14.9 III. 10.7 VII.3.14 III.10.7

Y34 Y34 Y34

IV. 12.2 III. 10.7 V.7.21 V.7.21 VIII.5.11

III. 10.7

III. 10.7 III.10.7

V.7.21 III.10.7 III.10.7 III.10.7

boughs is branches, and then there's twigs and leaves my gums has swelled right rats has taken my snap ([az 'te:sn ma: snap]) burglars steals 'em fowls runs in it. these was like that there is some 'at does a bit these fits on farm-horses generally lives on bran and oats pigs eats another, cows eats another footpaths 'at runs... mi tabs is (has?) ([az]) cold childer uses it my hands is cracked all over two heads is better nor one these three is t' grandkids. pens is same as folks women does spend a lot of time gossiping how many missuses does obey? the first forty year's t' worst they're terrible things, them is taties comes after fallows my taties is all big uns there's some on 'em wants help they're bad, them is, to do ought with. them's fine potatoes t' old taties is done my tabs does ache mi corns is bad doctors uses that mi teeth does nack ([nak]) men 'at's brushing hedges painters uses smocks when the pubs turns out branches is t' biggest t' old taties is finished. men as carries 'em

Y34

VIII.6.1

Chi Chi Chi

III.10.7 III. 10.7 III.10.7 III. 10.7 IV.02.01

Chi Ch2

V.4.6 1.1.3

Ch2

III.5.4

Ch2 Ch2 Ch2 Ch2

IV.2.11 IV.4.5 IV.6.8 IV.7.6

Ch2 Ch2 Ch2 Ch2

IV. 11.2 IV. 11.5 IV. 11.6 VI. 1.6

Ch2 Ch2

VI.5.16 VI.11.3

Ch2 Ch4

CONV. III. 12.5

Ch4

IV.6.11

Ch4

1V.6.19

Ch4 Ch4 Ch5 Ch5 Ch5 Ch6

IV.7.8 VI.7.2 1.1.4 VI.6.1 VI.7.1 III. 10.7

Ch6 Dbl Dbl

VIII.6.3 III.10.7 III. 10.7 III. 10.7 IV.2.4 IX.5.1

Dbl Db2 Db2

IV. 12.2 III.10.7 V.7.21

Db2 Db3 Db4 Db4 Db5 Db5 Db5 Db5 Db6

VIII.7.5 III. 10.7 VII.8.19 VIII.7.5 III.1.12 IV.3.10 IV.4.1 VIII.7.5

III.10.7

III.10.7 III. 10.7

IX.7.5

III.10.7

III.10.7 V.7.21 VIII.7.5 III.10.7 III.10.7 VIII.7.5 III.1.12 III. 10.7 III. 10.7 VIII.7.5 IX. 5.4

t' kids is coming out of school cats miaows bulls bellows big ditches as divides [relative clause or noun?] some calls it so they was very cheap, clay pipes was. them pigeons was there beasts makes them some calls it some doesn't these doesn't, these stands childer gets 'em cattle likes it birds has 'em some calls it headache, some calls it neuralgia in t' head ([njurradja]) some does beast has 'em on their tits (i.e. warts) nits comes butchers has special places them always chips where the beak is. some has very big ones isn't 'em? a lot does some calls 'em ricks some walks a lot some hands is deep cut sheep bleats, cats miaows children gets holiday cows bowls some does and some doesna they both has one sheep bleats these is t' last o ' t ' old ones burglars steals 'em bulls bellows the buggers does burglars steals 'em her bones is going where t' bars fits in his feet sinks in burglars steals 'em what's 'em call it? (=what does them)

185 Db7 Db7

1.3.2 IV.11.1

Ntl Ntl

III.1.12 IV.7.3

III.1.12

Ntl Ntl Nt2 Nt2

VI.7.4 III. 1.12 III.1.12 IV.2.4

III. 10.7

Nt2

IV.9.9

III. 10.7

Nt2

V.5.2

Nt2 Nt2

V.7.18 III.10.7 VIII.2.4

Nt3

1.6.4

Nt3 Nt3 Nt3 Nt3

III.1.12 III. 10.7 IV.2.1 IV.6.10

III.1.12

III.1.12 III.10.7

Nt3 Nt3 Nt3 Nt3 Nt3

III. 10.7 VI.5.1 VII.6.17 III.10.7 VIII. 1.1 VIII.7.5 VIII.7.5 IX. 1.9

Nt3

IX.8.6

Nt4 Nt4

III.1.12 III.6.8

III.1.12 IX.5.1

Nt4 Nt4 Nt4 LI LI

III.10.4 VIII.7.5 CONV. III.1.12 III. 10.7

III.10.7 VIII.7.5 III. 10.7 III. 1.12 III.10.7

LI

VI. 11.1

LI

VI. 11.6

LI

VI. 17.17

LI

V1II.3.5 III. 10.7

L2

1.6.5

L2

11.3.4

L2

III.8.4

III.10.7

them 'atparts 'em I donna know what them is her gristles goes woodpigeons as builds in the trees your hands is hot her bones is going her bones is going they cut 'em down low and lays 'em children comes and catches 'em all the bits as comes off a cow's bag kiddies calls it is your relations coming? they 've gotten wom while some on 'em's through her gristles has gone bulls blows some on 'em has they've spretched(?) and the chickens is hatching out my lips is cracked your boots is mucky them as comes in burglars steals 'em and some never creeps at all (i.e. babies) these is mi father's boots her muscles is going, they take some watching, some does. bulls Mothers burglars steals 'em bracken and fems is her gristles is dropped cocks crows, and if hens crows they kill 'em some on 'em has frettles some on 'em comes to a head them i Eastoft leggings when two on 'em's laid on horses is going out of date them s two ploughhorses some ί smaller than

L2 L2

IV.3.11 IV.6.15

IX.5.1

L2

V1H.7.5 VIII.7.5

L3

1.9.2

L3

III.1.12

L3

III.6.2

L3

III. 10.7

L3 L3 L3 L4

1V.6.20 VI.3.2 VHI.7.5 1.3.18

L4 L4

II.2.7 III. 10.7

L4

IV. 1.2

L4

IV. 12.3

L4

VIII. 1.2 III. 10.7

L4 L4 L5 L5 L5 L5

VIII.7.5 VIII.7.5 CONV. III. 10.7 1.6.1 III. 1.12 III.1.12 III.10.7 III. 10.7 IV.6.10

L5

IV. 11.6

L5 L5 L5 L6

VI.7.3 VIII.7.1 III. 10.7 VIII.7.5 VIII.7.5 1.9.5

L6 L6 L6

III.1.12 III. 10.7 III. 13.1

III.1.12 III. 10.7 III.10.7

L6 L6

IV.8.11 IV. 12.2

III. 10.7

L6

V.7.21

V.7.21

III.1.12

III.10.7

VIII.7.5

III. 10.7

III. 10.7

others some does plenty on 'em comes over here (burglars) steals (them) drayers(?) has sides to fit on a wagon when her bones has gone sheep's one of the stupidest things there is. (Fw: "plural verbs in

η

some sticks 'em some calls it slab-stone burglars steals 'em they lead bulls wi' staffs, puts a band on their homs, puts it undemien (=undemeath) their hind leg, and tie it to the tail. some folks smokes it. bulls blows, horses whinnies, cows blows, sheep blears the rigtails stops the water from getting in them that hangs over anywhere things wasn Ί the same then burglars steals 'em I thought tadpoles was four horses is a team her bones has gone bulls beats so many eggs has spretched them there what the bairns is gathering them is cracks, kids is slithering again burglars steals 'em there isn't many loads com now her bones is dropping (bulls) beals kell-scraps is the nicest scraps you can eat battletwigs eats plums, branches is the biggest, the boughs is the smaller. the old taties is done

186 L6

VII.6.6

L7

1.3.2

L7

1.3.10

L7

III.2.9

L7

III.5.5

IX.5.1

L7

III.10.7

III.10.7

L7

III. 11.2

L7

IV.5.7

III.10.7

L7 L7 L7 L8

IV. 11.4 V.8.9 CONV. II.5.2

III. 10.7

L8

III.1.12

L8 L8

IV.2.1 IV.6.5

L8

IV.6.22

III. 10.7 III.10.7

III. 10.7, VI.2.8

L8 L8

V.3.1 V.3.4

III.10.7 III. 10.7

L8

VI.13.11 III. 10.7

L8 L8 L8

VII.5.2 III.10.7 VII.6.14 III.10.7 VIII.5.1 VIII.5.1

L8 L8

CONV. CONV.

L8 L8 L8 L8

CONV. CONV. CONV. CONV.

L9

1.10.1

L9

III.10.7

L9

IV. 1.3

L9

IV.5.12

L9

IV.6.22

III. 10.7

III. 10.7 III. 10.7 III. 10.7

III. 10.7

I know some 'at 's been putten up some has, and some has iron. some goes into the crews. there's black cows has no horns, you know, they want brushing and cleaning, horses does. (Fw: "inflected plurals in z") them that's the best, they're penning 'em. old hens is nasty when they get broody, yew-trees is. some says mashing, two heads is peas and beans doesn Ί come in ears, they come in pods, the cows has eaten it up again.

L9

IV.10.8

III.10.7

L9

V.l.12

III.10.7

L9

V.l.14

L9

VI.7.3

L9

VI.14.11

L9

VI.14.17 III. 10.7

L9 L10

VIII.7.5 III.4.6

L10

IH.5.4

L10 L10

III. 10.7 III.10.7

III.10.7 III. 10.7

some calls 'em dykes, some on ' e m s redified (sic! = made red) two on you pulls, and them what gets the longest has the wish some folks does when the cinders tumbles out some folk says he's a boozer.

L10

IV.5.5

III. 10.7

L10

IV.6.22

L10

IV.9.11

L10

V.7.23

sommer fallows is. the roads is very slaip they're as good as them as goes to church. folks makes a mistake is them people there today? some has more nerve the lanes is the horses is the t'others what fruits (=that give fruit) them s what they call the rungs bulls blows, cows beats several on 'em calls it a ford there's maybe odd people as goes people as is a bit superstitious

L10 L10 L10

V.8.9 V. 10.6 VI. 11.3

L10 L10 LI 1

VII.6.14 VIII.7.5 VIII.7.5 1.7.14 III.10.7

LI 1

II.7.1

LI LI LI LI LI

II.9.16 III.3.2 III.5.7 III.10.7 IV.4.2

III. 10.7 III.10.7

LI 1

V.l.9

III.10.7

LI 1

V.7.6

LI 1

VII.3.9

1 1 1 1 1

III. 10.7

VIII.7.5

III.10.7

VIII.5.1

catkins comes afore the leaf. aye, a lot on 'em does today some on 'em has none at all. mi hands is chapped badly there's a lot of the people now as doesn Ί talk like they used to do. the lasses was always yucking the zips some folks steals 'em. some on 'em sooner goes over the time when you get into Nottinghamshire, they mix it up and calls 'em servers, (bulls) bellows lambs makes a bit different noise the gypsies eats 'em as they are, rolls 'em up in clay ([tki]) and cooks 'em over a fire, two on you takes hold on it kippers is dried herrings some calls 'em head and tailing 'em ([teilinam])(?) some calls it brewing some calls 'em spools them bairns has a lot of warts on 'em. the roads is very slaip burglars steals 'em some ladders as is made with spindles some has square stacks and some has round ones some gets afire some doesn't sit at all some on 'em's that fat bulls bellows same as where my roses is they're old-fashioned things, is snecks. the men what's gone from here have died young men them ί the sort wi' a

187

LI 1 LI 1

VII.5.19 III.10.7 VIII.5.6 VIII.5.1

LI 1

IX.7.2

LI 1

CONV.

LU

CONV.

III.10.7

LI 1

CONV.

III.10.7

LI 1 L12

CONV. II.4.5

III. 10.7

L12 L12

II.6.4 III.12.1

L12 L12 L12 L12 L12 L13

IV.2.7 V.7.21 CONV. CONV. CONV. III.1.13

L13

III.3.5

L13

III.9.4

L13

IV.3.1

L13 L13

IV.6.6 IV. 12.2

VIII.5.1 V.7.21 III. 10.7 III. 10.7 III. 10.7

III. 10.7

L13

IV. 12.5

L13

VI. 10.6

L13

VI.13.1

L13

VI. 14.23

L13 L13 L13 L14

VIII.7.5 VIII.7.5 CONV. III.10.7 CONV. III. 10.7 1.7.2 III. 10.7

L14

1.8.5

(VI.2.8)

L14

II.7.7

III.10.7

L14

III.3.4

L14

III.3.9

III. 10.7

III. 10.7

bit of swath on 'em pot-holes was bad them all lies between (=they all lie in between) I wonder if them two's married some has grazing seeds the wheels on a morphrey is the last lot 'at I did was ferrets is some's better than others some calls 'em stooks some on 'em has wheels on. them's out of date the old taties is done the fat uns is plenty calls most on 'em whistles some calls it th' afterbirth some as has ([az az]) real good uns some Λ coarser than others some on 'em never shuts the gates at all some on 'em J red the boughs and the branches is the same thing some says lop but I'd say slat some people's subject to 'em ([subdjekt t,am]) some cows has ([az]) always got somewhat the matter only them as goes to work wears boots (burglars) steals it trays is wages was bags is smaller than what the sacks is. what the horses pulls with most on 'em is not wood most on 'em has one (['εζ won]) my ducks is in it

L14

III.10.7

III.10.7

L14

IV.5.5

L14

IV.6.12

III.10.7

L14

IV.8.3

VIII.5.1

L14 L14 L15

IV. 10.3 CONV. IV.5.4

III.10.7 III. 10.7 III.10.7

Sal Sal Sa2 Sa2 Sa3 Sa4 Sa4 Sa5 Sa6

1.3.18 V.7.21 VI.7.2 VIII.7.5 VIII.7.5 III. 1.12 V.7.21 VIII.7.5 III.10.7

Sa6 Sa6 Sa6 Sa6 Sa7 Sa8

III.10.7 III.10.7 V.7.21 V.7.21 VIII.7.5 VIII.7.5 VIII.7.5 VIII.7.5 V.7.21 V.7.21

Sa8 Sa9 Sa9 SalO Sail Sa7

VIII.7.5 V.7.21 VIII.7.5 VIII.7.5 VIII.7.5 III.10.7

VIII.7.5 V.7.21 VIII.7.5 VIII.7.5 VIII.7.5 III. 10.7

Sa9

III.10.7

III.10.7

St2 St4

III.1.16 V.l.16

Leil

1.1.5

Leil Leil

III. 1.8 IH.3.4

IX.5.1

Leil Leil

VII.3.8 IX. 1.7

IX.5.1

Leil

IX.4.12

V.7.21 VIII.7.5 VIII.7.5 III.1.12 V.7.21 VIII.7.5 III. 10.7

bulls blows, horses whinnies, sheep blows, cows moos, dogs barks, pigs squeals hedgehogs sucks the cow's tits both has one (['bo0 az 'won]) theen is (['öran iz]) (=these are) oaktrees bears 'em the hollows is there's a lot of people kills 'em some calls 'em lofts the taties is finished my hands is chapped burglars steals 'em burglars steals 'em her bones is dropped the old taties is gone burglars steals 'em bulls roars, horses neighs, horses whinnies, cows moos, cows bawls, sheep baes, cats miaows things goes things has growed the taties is all finished burglars steals 'em burglars steals 'em the old taties is finished burglars steals 'em the taties is finished burglars steals 'em burglars steals 'em burglars steals 'em bulls bellows, horses neighs, cows lows, sheep bleats, cats miaows dogs barks, pigs squeals them makes best mist we got some as goes out some calls 'em pigsties some on 'em does some on 'em reckons it in't some does good thackers as thacks the houses them's old housen (sic, n-plural!)

188 Leil Lei2 Lei2 Lei2

II.2.3 1.8.1 1.9.6 1.9.7

Lei2 Lei2

III.13.4 V.7.4

Lei2

VII.2.2

Lei4 Lei5 Lei6 Lei6

VII.7.9 VII.7.9 1.1.3 II.2.9

Lei6

III.12.5

Lei7

IV.8.3

Lei9 Lei9

II.2.9 III.3.8

Lei9

IV.2.5

Lei9

V.l.8

III.10.7

Lei9

VII.5.2

VIII.5.1

III. 10.7 VIII.5.1

III.10.7

III.10.7

III. 10.7

VIII.5.1

Lei 10 III.1.16 Lei 10 IV.2.7 Lei 10 IV.6.2

III. 10.7 IV.6.2

LeilO V.l.17 R1 Rl

1.7.4 1.7.15

IX.5.1

R1 Rl Rl

1.9.9 IV.5.2 IV.8.1

VIII.5.1

long ones is them's th' handles them's the spokes where the spokes goes in some does some people always has a bit of bacon. these is the front of these docks is them what's got a few folks has been(?) th' ashes makes good stuff some's wood, some's lead there's two or three comes up at five o'clock in the morning. some calls it soot-fog. them's gisters in the top field. hedges that hasn Ί been done. our doors is on hooks, more like a gate. them 's oum (=they are ours). some calls 'em steer, some bullocks lights is a good many keeps hens these chalises(?) as was all lost some does some calls 'em sties, some spindles them '.v fellies them 's shrew-mices two or three folks here has got one

Rl

V.9.10

R2

IV.3.7

III. 10.7

R2

VII.4.8

III. 10.7

Hel He2 He3 He5 He6 He7 He7 He7 He7 He7 He7 He7 He7

VIII.7.5 VIII.7.5 VIII.7.5 VIII.7.5 VIII.7.5 III.10.7

VIII.7.5 VIII.7.5 VIII.7.5 VIII.7.5 VIII.7.5 III.10.7 III. 10.7 III.10.7 III.10.7 III. 10.7 III.10.7 III.10.7 IX.9.5

Wo4 Wo5 Wo7 Wo7 Wo7 Wo7 Wa2

III.10.7 III.10.7 III.10.7 III.10.7 III.10.7 III. 10.7 IX.9.5

Wa3 Wa5 III.10.7 Wa7 III.10.7 Wa7 Nth2 IV. 10.7

III. 10.7 III. 10.7 III.10.7 III. 10.7

Nth4 VII.6.7 Nth5 IX.9.2 Hul IV.4.4

III. 10.7

Nf2

11.09.10

Nf7 Nf7

V.7.2 VIII.7.6 IX.5.1

Nbl

III. 10.3

Nb2

III. 10.7

Nb2

IV.6.3

some uses a jug and gets it broke the ones that goes across was braces all the children comes home burglars steals 'em burglars steals 'em burglars steals 'em burglars steals 'em burglars steals 'em bulls roars horses pulls the bones goes ducks goes arms gets ladies wears quists comes two people (who) speaks the same birds carries spokes goes bakers works eggs begins feathers grows some mothers yields there's a good many uses 'em. leggings comes outside cats miaows bulls Marts peas looks it kills the thorns as grows round it I know several signs as is pretty sure the houses was thatch people what was used to it some on 'em was made of mortar th' old spuds is done some on 'em does here

Β. 17. Non-use of verbal -s in NSR contexts Verbal base forms in potential NSR contexts were excerpted in 59 cases from the fieldworker notebooks of the following locations: Nbl (1), Nb2 (5), Nb3 (1), Nb5 (1), Nb7 (1), La3 (1), Y11 (2), Y16 (2), Y17 (4), Y18 (1), Y21 (1), Y22 (5), Y24 (1), Y25 (1), Y26 (1), Y28 (I), Y30 (1), Y31 (2), Y34 (1), Db2 (1), L4 (2), L8 (5), L9 (1), L10 (1), L12 (1), L13 (1), L14 (1), Lei4 (1), Nfl (5),Nf2 (7).

they generally come wi' a whistle (Fw: "none of the plurals have inflected endings, I tried repeatedly. Certain incid. material shows the presence of such endings") the hens are on the perks

189 a lot divn Ί milk them them that have health and strength Nb2 VI. 1.3 some call them chaps some folk call them Nb2 VI.4.1 Nb3 IX.7.7 they both are. I'll bet his bairns get some Nb5 V.7.14 and all. you whiles gel them Nb7 V.7.5 the animals as give you La3 1.2.1 wool LalO III. 10.7 III.10.7 cats miaow (Fw: "does not inflect verb in plural") Lal4 III.10.7 (Fw: "would not give any plurals with inflection") Yll II.5.1 beans have no swads VII.5.12 there is some men lake Yll their dinners with them them are peas, are they? Y16 V.7.13 ([3em a:]) cows and sheep hooze Y16 VI.8.3 (=cough where t' horses are ([a.r]) Y17 1.4.1 Y17 VIII.6.5 teachers are just going home now our lads were a-going Y17 [CONV] astray ([war a goin]) Y17 [CONV] these here stones are more stones nor dust they never were ([wa.r]) Y18 1.9.4 Y2) 1.3.7 beasts' heads were sealed (=tied) together ([wa si:ld]) Y22 IV.6.7 hens 'at haven Ί been trodden Y22 IV. 10.5 little baims grow under t' rose-bushes Y22 IV.12.5 lads go chumping Y22 V.2.4 some on 'em sleep i ' t ' attic Y22 IX.2.1 folks laik at Saturday Y24 V.7.16 shallots don Ί get nought as big Y25 I..01.4 if them glasses tumble on t' floor, they'll smash to smithereens. Y26 1.2.1 when t' horses were i-go ([i'g°a]) Y28 VI.7.12 some on 'em aren 't Y30 IV.8.13 these here come through that Y31 1.7.4 these were com-kists Y31 VII.4.11 fair near (['ηεηφ where t' old folks were (['wo]) Y34 IV. 11.7 birds eat 'em Db2 III.7.4 them as were bringing 'em Db5 III.10.7 (Fw: "no inflected forms") L4 crooks are in the post, IV.3.5 vartiwells are on the gateNb2 Nb2

VII.8.19 VIII.4.8

L8 L8

VII.2.1 VIII.2.1

L8

VIII.3.5

L8 L8

VIII.7.8 [CONV]

L9

VI. 1.4

L10 L12

II.9.17 III.10.7

L13

VI. 14.24

L14

VI. 1.4

Lei4

III.2.6

Nfl Nfl

III.1.8 1V.6.15

Nfl Nfl

IV. 10.7 V.2.6

Nfl

VII.6.18

Nf2

1.5.4

Nf2

1.8.1

Nf2

1.9.17

Nf2

II.2.6

Nf2

VI.5.1

Nf2

VII.4.6

Nf2

IX.3.8

head. some folks call it string our neighbours are very good yon people have been gossiping for an hour foxes do that too. the bearers are starting tomorrow some folks about here say they've got the palsy. when the beast eat it the bulls bell when they're in the grass-field the roads were all rough stone some people say they got palsy they're dangerous cattle, them are\ some old boys say wild geese come off the marshes (present tense) osiers, some call 'em lots of people say pantyladder march winds blow the silt into the roads some are made of iron and some of wood our ploughs what we use here are different some people leave it for seed some lands are festered with it some people call it the mug lot of people work Sundays and all some say grab it

190 Map 17. Division of SED locations into seven areas for purposes of Chapter 4.4

50

100km

Central/Upper North Cu5· Du3· Cu4 *w c ,. Wc2 · Wc3 " Υ7· , Υ5· Y12*YI3*Y|4· y „
uthwest Midlands1'"1* Sill Sail) Nlh2· Sail· SI6·

Sl7·

1^7» Hcl* Wo4" [ Hc2 · Hc3* Wo lk4*

Wu5· Wa6* Southeast Midlands

191 C. NITCS data tables

References to NITCS informants within the text are given as in the example: "L17.3:LD"; where "17" is the NITCS location number; "3" is the index of the interview within that location, which usually also identifies the age group of the informant; and "LD" is an individual identifier, usually consisting of the name initials. References to speech examples additionally contain the consecutive number of the utterance within the interview, as in "L17.3:LD83". The numbering of the locations within the NITCS is not identical with the larger numbering scheme of the TRS, shown in the second column in the following table.

C. 1. NITCS locations NITCS Location LI L2 L3 L4 L5 L6 L7 L8 L9 L10 LI 1 L12 L13 L14 L15 L16 L17 L18 L19 L20 L21 L22 L23 L24 L25 L26 L27 L28 L29 L30 L31 L32 L33 L34 L35 L36 L37 L38

TRS Location A12 A13 A14 A19 A20 A21 A22 A28 A29 A30 A31 A32 A33 A34 A39 A40 A41 A42 A44 A45 A46 A49 A50 A51 A52 A53 A54 A55 A56 A57 A59 A60 A61 A63 A64 A65 A66 A73

Name Dervock Armoy Glendun Shan tallow Dungiven Garvagh Glarryford Clady Plumbridge Cranagh Desertmartin Toomebridge Doagh Ballycarry Scraghey Omagh Carrickmore Stewartstown Crumlin Ballystockart Dunover Braade Irvinestown Clabby Ballygawley Benburb Lurgan Kinallen Derryboy Ballygalget Belcoo Kinawley Lisnaskea Madden Lurganmore Cabragh Ballynoe Attical

County Antrim Antrim Antrim Londonderry Londonderry Londonderry Antrim Tyrone Tyrone Tyrone Londonderry Antrim Antrim Antrim Tyrone Tyrone Tyrone Tyrone Antrim Down Down Fermanagh Fermanagh Fermanagh Tyrone Tyrone Armagh Down Down Down Fermanagh Fermanagh Fermanagh Armagh Down Down Down Down

Word Count 4,201 7,401 1,704 3,755 5,918 4,469 9,527 10,216 5,665 7,074 10,808 9,306 5,850 4,600 3,752 8,903 9,996 4,247 7,118 2,549 10,438 9,104 10,299 4,994 9,263 4,927 7,397 426 3,987 1,405 10,658 7,046 8,613 4,060 2,427 7,556 450 600

192

C.2. N1TCS informants NITCS Loc. LI LI L2 L2 L2 L3 L3 L3 L4 L4 L4 L5 L5 L5

TRS Loc. A12 A12 A13 A13 A13 A14 A14 A14 A19 A19 A19 A20 A20 A20

Interview

ID

LI.IB L1.1G L2.1 L2.2 L2.3 L3.2 L3.3 L3.3 L4.1 L4.2 L4.3 L5.1 L5.2 L5.3

BC MK WS JM AM MM BO MO CL PL ML KF TD RF

Age group 1 1 1 2 3 2 3 3 1 2 3 1 2 3

Sex

Rel.

FW

Μ F Μ Μ Μ F Μ F F F F F F F

C C Ρ C C C C C C C C C C C

JR JR JR ED JR JR JR JR CG CG CG LC LC LC

Plural -s tokens 0/16 0/19 0/0 13/16 3/7 1/1 3/3 0/0 0/0 7/10 2/3 0/2 0/4 0/2

Word count 1,991 2,272 791 5,339 1,540 381 1,220 125 638 2,702 657 868 2,342 612

193 NITCS Loc. L5 L6 L6 L6 L6 L7 L7 L7 L7 L7 L7 L8 L8 L8 L8 L9 L9 L9 L10 L10 L10 LI 1 LI 1 LI 1 L12 LI2 L12 L12 L13 L13 L13 L13 L13 L13 L14 LI4 L15 L15 L15 L15 L16 L16 L16 L17 L17 L17 L17 L18 L18 L18 L18 L18 L19 L19 L19 L19

TRS Loc. A20 A21 A21 A21 A21 A22 A22 A22 A22 A22 A22 A28 A28 A28 A28 A29 A29 A29 A30 A30 A30 A31 A31 A31 A3 2 A3 2 A3 2 A3 2 A3 3 A3 3 A3 3 A3 3 A33 A3 3 A34 A34 A39 A3 9 A3 9 A39 A40 A40 A40 A41 A41 A41 A41 A42 A42 A42 A42 A42 A44 A44 A44 A44

Interview

ID

L5.3 L6.1 L6.2 L6.2 L6.3 L7.1 L7.2 L7.2 L7.3 L7.3 L7.3 L8.1 L8.1 L8.2 L8.3 L9.1 L9.2 L9.3 L10.1 LI 0.2 L10.3 LI 1.1 LI 1.2 LI 1.3 L12.1 L12.2 L12.3 L12.3 L13.1 L13.2 L13.3 L13.3 L13.3 L13.3 L14.1 L14.2 L15.1 L15.2 L15.3 L15.3 L16.1 L16.2 L16.3 LI 7.1 LI 7.2 L17.2 L17.3 L18.1 L18.1 LI 8.2 L18.3 L18.3 L19.1 L19.2 L19.3 L19.3

TF CR ANON MM AH AM ANON TM 2ANON ANON WG ANON CM PM WC CA JW JM SM BC JM PG TS JM JH HC ANON LM IM IP 2ANON 3ANON ANON NR KS SG MM BM ANON OM EM IC PT NK ANON HR LD ANON JA ΗΝ ANON FC DF WC ANON JB

Age group 3 1

Sex

Rel.

FW

Μ Μ

?

?

C C ?

2 3 1

C C Ρ

?

F Μ F 7

2

Μ

? ?

? ?

3

Μ

Ρ 7 Ρ Ρ

?

?

?

1 2 3 1 2 3 1 2 3 1 2 3 1 2

Μ Μ Μ F Μ Μ F Μ Μ Μ Μ Μ Μ Μ

?

?

3 1 2 7

F Μ Μ 7 7 7 Μ Μ Μ F Μ 7

C C C C C C C C C C C C C C 7 C Ρ Ρ 7 7 7

LC CG CG CG CG LC LC LC LC LC LC CG CG CG CG CG CG CG CG CG CG CG CG CG MNR TL MNR MNR JR JR JR JR JR JR JR AP CG CG CG CG CG CG CG CG CG CG CG CG CG CG CG CG JR JR JR JR

? ? 3 1 2 1 2

? 3 1 2 3 1

?

? Ρ Ρ C C C C Ρ C C C

?

Μ F F Μ F ?

2 3

Μ F

C Ρ

?

?

?

1 2 7

F F 7

C Ρ 7

3 1 2 3 3

Μ Μ Μ F Μ

C Ρ Ρ 7 ?

?

Plural-s tokens 1/6 0/2 0/0 0/12 8/8 0/3 0/0 5/18 0/0 0/0 1/2 0/0 0/9 1/7 13/15 0/6 6/7 5/23 2/5 11/11 8/11 0/3 3/36 9/14 0/4 6/6 0/0 5/13 0/1 2/22 0/0 0/0 0/2 0/3 0/2 0/10 0/0 1/1 0/0 7/9 0/3 2/25 10/10 1/2 0/0 1/13 19/25 0/0 0/5 0/15 0/0 5/6 2/10 10/12 1/2 5/5

Word count 2,273 648 17 2,473 1,531 3,061 3 4,210 19 604 1,891 5 2,720 3,328 4,557 724 2,582 2,572 1,035 4,015 2,250 1,973 4,573 4,809 1,782 3,429 4 4,320 236 3,683 2 1 25 2,152 1,790 2,931 353 1,182 27 2,332 1,017 5,405 2,843 2,818 10 2,782 4,788 3 689 1,258 17 2,445 2,813 2,536 174 1,870

194 NITCS Loc. L20 L20 L21 L21 L21 L21 L22 L22 L22 L22 L23 L23 L23 L24 L24 L24 L25 L25 L25 L26 L26 L26 L27 L27 L27 L28 L28 L28 L29 L29 L29 L29 L30 L30 L30 L31 L31 L31 L31 L32 L32 L32 L32 L32 L32 L33 L33 L33 L33 L33 L34 L34 L34 L35 L36 L36

TRS Loc. A45 A45 A46 A46 A46 A46 A49 A49 A49 A49 A50 A50 A50 A51 A51 A51 A52 A52 A52 A53 A53 A53 A54 A54 A54 A55 A55 A55 A56 A56 A56 A56 A57 A57 A57 A59 A59 A59 A59 A60 A60 A60 A60 A60 A60 A61 A61 A61 A61 A61 A63 A63 A63 A64 A65 A65

Interview

ID

L20.2 L20.3 L21.2 L21.2 L21.3 L2I.3 L22.1 L22.2 L22.3 L22.3 L23.1 L23.2 L23.3 L24.1 L24.2 L24.3 L25.1 L25.2 L25.3 L26.1 L26.2 L26.3 L27.1 L27.2 L27.3 L28.1 L28.2 L28.3 L29.1B L29.1G L29.2 L29.3 L30.1 L30.2 L30.3 L31.1 L31.2 L31.2 L31.3 L32.1 L32.1 L32.2 L32.2 L32.3 L32.3 L33.1 L33.2 L33.3 L33.3 L33.3 L34.1 L34.3 L34.3 L35.3 L36.1 L36.2

HS MS ANON JM DA JA MC JH ANON JO KO PM MK. EM AW CM SM MM PH JM BM TK GB NI SB CS MM GG AW EP TM FD KW JP PB AT DM RM JR AN KM ANON MM ANON MM OH EM 2ANON ANON MM JS ANON PH MM NG JO

Sex

Rel.

FW

Μ F F Μ Μ Μ Μ Μ F Μ F Μ F F F Μ Μ F F F Μ Μ Μ F Μ F

Ρ Ρ Ρ Ρ Ρ Ρ C C C C C C C C Ρ C C C C C C C Ρ Ρ Ρ

?

Ρ

F Μ F Μ Μ F Μ Μ F Μ F Μ F F Μ F Μ F Μ F

? ?

C C C C C C C C C C C C

?

?

?

3 3 1

F Μ Μ

C C C

AP AP AP AP AP AP CG CG CG CG CG CG CG CG CG CG CG CG CG CG CG CG MVB MVB MVB GBA GBA GBA GBA GBA GBA GBA GBA GBA GBA CG CG CG CG CG CG CG CG CG CG CG CG CG CG CG AM AM AM AP TL TL

Age group 2 3 2 2 3 3 I 2 3 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 1 2 3 1 2 3 1 2 2 3 1 1 1 2 3 3 1 2

?

Ρ Ρ Ρ ? C

?

?

?

?

3 3 1 2

Μ Μ F Μ

7 C C C

Plural-s tokens 2/5 5/11 0/0 4/7 0/27 8/12 0/2 14/20 1/2 7/9 0/18 2/20 0/26 0/0 0/8 10/12 0/5 3/9 11/11 1/16 0/4 6/7 1/4 1/15 3/7 0/0 0/0 0/0 0/2 0/0 0/9 3/25 0/0 0/0 2/4 0/7 5/24 0/0 4/5 0/1 0/2 0/0 0/18 5/5 1/13 0/5 3/6 0/0 1/3 10/120 0/3 0/0 2/9 3/11 1/4 12/16

Word count 757 1,876 51 921 7,354 2,470 2,005 3,999 471 2,949 3,971 3,167 3,604 789 1,803 2,605 2,095 3,569 3,758 2,391 1,078 1,646 1,528 3,621 2,318 134 79 232 213 155 758 2,982 101 694 674 2,090 4,982 2 3,886 69 635 31 2,435 824 3,430 2,024 1,241 13 127 5,581 1,771 12 2,430 2,538 2,624 3,886

195 NITCS Loc. L36 L37 L38 L38 L38 Total

TRS Loc. A65 A66 A73 A73 A73

Interview

ID

L36.3 L37.3 L38.1 L38.3 L38.3

CO PO LG ANON MH

Age group 3 3 1 3 3

Sex

Rel.

FW

F Μ Μ Μ F

C

TL GBA GBA GBA GBA

? ? ? ?

Plural-s tokens 4/5 0/1 0/0 0/0 2/4 316/938

Word count 1,392 463 70 47 512 239,001

C.3. Existential -r in the NITCS cross-tabulated for three social and three linguistic factor groups.

Area North

Northwest

East

Southwest

SouthEast

Total

Age 3 2 2 1 1 3 3 2 2 1 1 3 2 1 1 3 3 2 2 1 1 3 3 2 2 1 1

Sex m m f m f m f m f m f m m m f m f m f m f m f m f m f

Past Tense (were-was) Non-Negated Negated PI PI Sg Sg 16/21 3/3 0/0 14/23 1/1 0/0 4/8 4/4 0/0 1/1 0/0 0/0 0/0 0/0 0/0 0/0 0/0 1/1 0/4 1/5 9/47 19/42 20/33 9/11 2/8 0/0 4/9 5/10 7/40 7/21 5/9 3/5 1/4 0/24 4/8 0/2 0/0 6/11 2/3 1/1 0/0 0/0 1/5 0/1 0/2 1/1 1/13 7/9 0/3 0/1 0/18 1/3 0/0 0/0 0/4 4/6 0/0 0/0 0/0 0/0 2/20 2/12 6/60 8/45 2/27 0/10 0/41 0/22 1/15 0/3 2/26 1/12 0/10 0/4 0/22 4/15 0/0 0/0 0/6 0/2 0/0 0/9 0/0 0/3 0/4 0/1 0/13 0/2 0/2 0/3 0/6 1/12 0/4 0/1 0/15 1/8 0/1 0/0 0/3 2/2 0/0 0/0 0/7 0/1 0/0 0/0 0/4 0/3 38/146 17/57 54/418 87/260

Present Tense ('re- 's) Negated Non-Negated PI PI Sg Sg 4/4 0/0 4/14 8/13 3/4 3/3 11/17 8/28 0/0 0/0 0/2 0/0 0/0 0/0 1/5 0/6 0/1 0/0 0/20 0/11 13/41 18/31 15/23 8/8 2/3 0/0 0/10 0/5 4/15 3/4 6/53 20/39 0/2 2/3 0/32 8/13 1/1 0/0 4/20 3/10 0/1 0/0 1/25 3/28 2/7 7/9 2/61 12/19 1/6 2/20 1/10 3/5 0/0 0/0 0/9 2/4 0/0 0/0 0/2 0/0 1/9 1/27 4/6 0/18 1/10 0/2 0/41 1/11 0/11 1/3 0/33 4/25 0/10 6/24 0/0 0/22 0/4 0/2 0/18 0/43 0/2 0/0 0/43 3/32 0/0 0/14 0/3 1/8 0/3 0/0 0/7 0/8 1/7 0/3 0/18 0/12 0/0 0/0 0/3 2/4 0/0 0/2 1/15 0/15 0/0 0/0 0/15 0/6 35/126 31/50 42/623 104/387

Total

49/78 34/65 1/3 1/11 2/42 111/236 13/45 55/186 15/88 17/46 5/60 32/121 8/66 6/23 0/2 24/197 4/164 9/128 10/107 0/75 3/89 1/45 1/41 2/68 4/13 1/40 0/28 408/2067

196 D . FRED data tables

References to FRED informants are given in the text as in the example "WES4:HL", where the first three letters indicate the county of origin (see the list of county abbreviations on page IX), the number identifies the corpus text, and the following letters the informant (usually by name initials). The TRS-D recordings, not formally part of FRED but included in the same analysis in Chapter 6, are listed at the end. They are identified by their TRS location and interview number. See also Map 16 on page 136 for the geographical distribution of the texts.

D. 1. FRED informants Location ANSI ANSI ANS2 BAN1 DFS1 DFS1 DUR1 DUR2 DUR3 FIF1 INV1 KCD1 KCD1 KRS1 LAN5 LKS1 NBL1 NBL6 NBL6 NBL7 PEE1 PEE2 PERI PER2 PER3 PER4 ROC1 SEL1 SEL2 SEL4 SUT1 SUT2 SUT3 SUT4 WES3 WES4 WES6 WES8 WLN1 WLN2

Inf. ID AN JT BS JS AN WH ML NB EL PS WR ES RS WS EG GH WL GS JB AR JP MT SS TJ AB HK AM JB JM AN DS GM GM TR AQ AY BL BR TW MO

Born 1918 1919 1914 1910 c. 1910-20 1908 1897 1914 1952 1924 1917 1906 1917 c. 1900 c.1900 c. 1890-1900 1891 1952 1905 1902 1913 1913 1901 1902 1926 1931 1914 1922 1901 1929 1898 1905 1901 1908 1927

Sex Μ Μ Μ F F Μ Μ Μ F Μ Μ F Μ Μ Μ Μ Μ Μ F Μ Μ F Μ Μ Μ Μ F Μ Μ Μ Μ Μ Μ Μ Μ Μ F Μ Μ F

NSR Tokens 0/0 18/19 3/6 7/16 0/0 6/17 9/23 2/16 13/21 2/11 11/19 5/7 10/15 0/7 26/60 3/3 1/2 2/3 3/9 2/7 1/8 9/25 1/2 0/12 19/21 0/24 1/22 7/14 2/8 0/5 0/0 4/18 0/27 7/10 9/10 2/4 0/7 13/26 6/13 2/4

Word Count 11 5,640 1,676 5,349 15 9,563 9,304 8,318 9,098 3,471 2,000 1,700 3,471 5,561 28,263 3,585 1,000 2,000 1,500 1,168 2,131 12,160 3,672 2,259 8,168 4,740 5,807 4,304 2,919 1,870 173 3,807 3,124 3,398 5,033 8,035 3,377 11,759 1,754 1,630

197 Location WLN4/5 WLN6/7 YKS1 YKS2 YK.S3 YKS4 YKS5 YKS6 YKS7 YKS8 YKS9 YKS9 YKS13

Inf. ID MD ES WF JV SL EH GR HS ML MW MG WG AT

Bom 1908 1931 1910 1899 1912 c.1910 c.1910 1910 c.1900 c. 1900 c. 1930 c.1905 1899

TRS-D recordings integrated with the FRED recordings: TRS18.3 AN c.1900 TRS25.3 AN c.1900 TRS37.2 AN c. 1930 TRS58.3 AN c.1900 TRS7.2 AN1 c. 1930 TRS7.2 AN2 c.1930 TRS72.2 AN c.1930 Total

Sex F Μ Μ Μ Μ F Μ Μ Μ F F Μ Μ

F F F Μ F Μ Μ

NSR Tokens 2/16 4/9 5/13 15/23 11/14 4/8 6/8 1/7 1/22 0/42 0/1 13/15 1/6

Word Count 8,131 6,412 10,153 9,695 12,070 1,988 3,262 7,372 11,377 9,809 431 5,434 3,576

22/22 1/21 1/28 0/0 0/5 0/3 5/6 298/790

2,516 1,475 2,911 1,352 84 3,121 2,484 296,466

References

Adams, G. Brendan (1973): "Language in Ulster, 1820-1850." - In: Ulster Folklife 19: 50-55. Aissen, Judith (1999): "Markedness and subject choice in Optimality Theory." - In: Natural Language and Linguistic Theory 17: 673-711. - (2000): "Differential object marking: iconicity vs. economy." http://ling.ucsc.edu/~aissen. - , Joan Bresnan (2002): "Optimality and functionality: objections and refutations." - In: Natural Language and Linguistic Theory 20: 43-80. Anderson, Stephen R. (1974): "On dis-agreement rules." - In: Linguistic Inquiry 5: 445-451. - (1985): "Inflectional morphology." - In: Timothy Shopen (ed.): Language typology and syntactic description. Vol. 3: Grammatical categories and the lexicon, 150-210. Cambridge: Cambridge University Press. Anderwald, Lieselotte (2001): "Was/were variation in non-standard British English today." - In: English World-Wide 22: 1-21. - (2002): Negation in non-standard British English: gaps, regularizations and asymmetries. - London: Routledge. Bailey, Charles, Karl Maroldt (1977): "The French lineage of English."- In: Meisel (ed.), 21-53. Bailey, Guy (1989): "Sociolinguistic constraints on language change and the evolution of are in Early Modern English." - In: Joseph B. Trahem, (ed.): Standardizing English: Essays in the history of language change, 158-171. Knoxville: University of Tennessee. - , Natalie Maynor (1985a): "The present tense of be in southern black folk speech." - In: American Speech 60: 195-213. - , Natalie Maynor (1985b): "The present tense of be in white folk speech of the southern United States." - In: English World-Wide 6: 199-216. - , Natalie Maynor, Patricia Cukor-Avila (1989): "Variation in subject-verb concord in Early Modern English." - In: Language Variation and Change 1: 285-300. - , Garry Ross (1988): "The shape of the superstrate: morphosyntactic features of Ship English." In: English World-Wide 9: 193-212. Bambas, Rudolph C. (1947): "Verb forms in -s and -th in Early Modern English prose." - In: Journal of English and Germanic Philology 46: 183-187. Barbiers, Sjef, Leonie Cornips (2002): "Introduction." - In: Barbiers, Comips van der Kleij (eds.). Barbiers, Sjef, Leonie Comips, Susanne van der Kleij (eds.) (2002): Syntactic microvariation. http .//www. meertens. nl/books/synmic/. Barlow, Michael, Suzanne Kemmer (1994): "A schema-based approach to grammatical description." - In: Lima, Susan, Roberta Corrigan, Gregory Iverson (eds.): The reality of linguistic rules, 19-42. Amsterdam: Benjamins. Barry, Michael V. (1981a): "The southern boundaries of Northern Hiberno-English speech." - In: Barry (ed.), 52-95. - (1981b): "The methodology of the tape-recorded survey of Hiberno-English speech." - In: Barry (ed.), 18-46. - (ed.) (1981): Aspects of the English dialects in Ireland. - Belfast: Queen's University, Institute of Irish Studies. Bates, Elizabeth, Brian MacWhinney (1987): "Competition, variation, and language learning." - In: Brian MacWhinney (ed.): Mechanisms of language acquisition, 157-139. Hillsdale, NJ: Erlbaum. Bayley, Robert (2002): "The quantitative paradigm." - In: Chambers/Trudgill/Schilling-Estes (eds.), 117-141.

200 Beal, Joan (1993): "The grammar of Tyneside and Northumbrian English." - In: Milroy/Milroy (eds.), 187-213. - (1997): "Syntax and morphology." - In: Jones (ed.), 335-377. - (2004): "The morphology and syntax of English dialects in the North of England." - In: Kortmann/Upton (eds.). Beals, {Catherine et al. (eds.) Π 994): Papers from the 30th regional meeting of the Chicago Linguistics Society. Chicago: Chicago Linguistics Society. Benveniste, Emile (1966): Problemes de linguistique general. - Paris: Gallimard. Berdan, Robert (1975): "The necessity of variable rules." - In: Fasold/Shuy (eds.), 11-26. Bergs, Alexander T. (1999): Modern Scots (Languages of the World Materials). - München, Newcastle: Lincom Europa. Berndt, Rolf (1956): Form und Funktion des Verbums im nördlichen Spätaltenglischen: Eine Untersuchung der grammatischen Formen und ihrer syntaktischen Beziehungsbedeutungen in der großen sprachlichen Umbruchsperiode. - Halle: Niemeyer. - (1982): A history of the English language. - Leipzig: VEB Verlag Enzyklopädie. Bickerton, Derek (1971): "Inherent variability and variable rules." - In: Foundations of Language 7: 457-492. Bjorge, Anne-Kari (1989): " Ί poor man is singing': Subject-verb concord in 18th-century English grammar." - In: Leiv-Egil Breivik et al. (eds.): Essays on English language in honour of Bertil Sundby, 47-65. Oslo: Novus. Black, James R., Virginia Motapanyane (eds.) (1996): Microparametric syntax and dialect variation. Amsterdam: Benjamins. Blakeley, Leslie (1950): "The Lindisfame s/d problem." - In: Studia Neophilologica 22: 15-47. Boersma, Paul (2000): "Learning a grammar in Functional Phonology." - In: Dekkers/van der Leeuw/van de Weijer (eds.), 465-523. - , Joost Dekkers, Jeroen van de Weijer (2000): "Introduction: Optimality Theory: phonology, syntax, and acquisition." - In: Dekkers/van der Leeuw/van de Weijer (eds.), 1—46. - , Bruce Hayes (2001): "Empirical Tests of the Gradual Learning Algorithm." - In: Linguistic Inquiry 32: 45-86. - , David Weenink (2000): "Praat computer program." http://www.fon.hum.uva.nl/praat/. Börjars, Kersti, Carol Chapman (1998): "Agreement and pro-drop in some dialects of English." - In: Linguistics 36: 71-98. Breivik, Leiv E. (1981): "On the interpretation of existential there." - In: Language 57: 1-25. - , Toril Swan (2000): "The desemanticisation of existential there." - In: Christiane Dalton-Puffer, Nikolaus Ritt (eds.): Words: structure, meaning, function, 19-34. Berlin: Mouton de Gruyter. Bresnan, Joan (2000): "Optimal syntax." - In: Dekkers/van der Leeuw/van de Weijer (eds.), 334—385. - (2001): "Explaining morphosyntactic competition." - In: Mark Baltin, Chris Collins (eds.): Handbook of contemporary syntactic theory, 11-44. Oxford: Blackwell. - , Ashwin Deo (2001): "Grammatical constraints on variation: 'Be' in the Survey of English Dialects and (Stochastic) Optimality Theory." http://www-lfg.stanford.edu/bresnan/be-final.pdf - , Shipra Dingare, Christopher Manning (2001): "Soft constraints mirror hard constraints: voice and person in English and Lummi." http://csIipublications.stanford.edU/LFG/6/lfg01-toc.html. Britain, David (2002): "Diffusion, levelling, simplification, and reallocation in past tense be in the English Fens." - In: Journal of Sociolinguistics 6: 16-43. Brunner, Karl (1962): Die englische Sprache. Vol. 2. - Tübingen: Niemeyer. - , Eduard Sievers ( 3 1965): Altenglische Grammatik. - Tübingen: Niemeyer. Bybee, Joan (1985): Morphology: α study of the relation between form and meaning. - Amsterdam: Benjamins. - (1995): "Regular morphology and the lexicon." - In: Language and Cognitive Processing 10: 425^155.

201 -

(2001): "Frequency effects on French liaison." - In: Bybee/Hopper (eds.), 337-360. (2003): "Mechanisms of change in grammaticalization: the role of frequency." - In: Brian Joseph, Richard Janda (eds.): Handbook of historical linguistics, 602-623. London: Blackwell. - , Paul Hopper (2001): "Introduction." - In: Bybee/Hopper (eds.), 1-26. - , Paul Hopper (eds.) (2001): Frequency and the emergence of linguistic structure. - Amsterdam: Benjamins. - .Joanne Scheibmann (1999): "The effect of usage on degrees of constituency: the reduction of don't in English." - In: Linguistics 37: 575-596. - , Sandra Thompson (2000): "Three frequency effects in syntax." - In: Berkeley Linguistics Society 23: 378-388. Chambers, Jack K. (2004): "Dynamic typology and vernacular universals." - In: Bernd Kortmann (ed.): Dialectology meets typology. Berlin: Mouton de Gruyter. - , Peter Trudgill, Ν. Schilling-Estes (eds.) (2002): Handbook of language variation and change. Oxford: Blackwell. Chapman, Carol (1998): "A subject-verb agreement hierarchy: evidence from analogical change in modem English dialects." - In: Richard Hogg, L. van Bergen (eds.): Current Issues in Linguistic Theory: Historical Linguistics 1995. Vol. 2, 35-44. - Amsterdam: Benjamins. Cheshire, Jenny L. (1982): Variation in an English dialect: A sociolinguistic study. - Cambridge: Cambridge University Press. - , Viv Edwards, Pamela Whittle (1989): "Urban British dialect grammar: the question of dialect levelling." - In: English World-Wide 10: 185-225. Clark, Brady Z. (2003): "On stochastic grammar." - http://www.stanford.edu/~bzack/papers/ stotgrammars .pdf. Collingwood, William G. (1911): "A rune-inscribed Anglian cross-shaft at Urswick church." - In: Trans. Cumberland Westmorland Antiq. Archaeol. Soc., ser. 2, 11: 462-488. - (1927): Northumbrian crosses of the pre-Norman age. - London: Faber Gwyer. Corrigan, Karen P. (1990): "Northern Hiberno-English: the state of the art." - In: Irish University Review 20: 91-119. - (1997): "The syntax of South Armagh English in its socio-historical perspective." - PhD thesis, University College Dublin. Croft, William (1990): Typology and universals. - Cambridge: Cambridge University Press. - (1995): "Autonomy and functionalist linguistics." - In: Language 71: 490-532. - (2000): Explaining language change: an evolutionary approach. - London: Longman. - (2001): Radical construction grammar. - Oxford: Oxford University Press. - (fc.): "Construction grammar." - In: Dirk Geeraerts, Hubert Cuyckens (eds): Handbook of Cognitive Linguistics. Berlin: Mouton de Gruyter. Crystal, David (1995): Cambridge encyclopedia of the English language. - Cambridge: Cambridge University Press. Dekkers, Joost, Frank van der Leeuw, Jeroen van de Weijer (eds.) (2000): Optimality theory: phonology, syntax, and acquisition. - Oxford: Oxford University Press. Dingare, Shipra (2001): "The effect of feature hierarchies on frequencies of passivization in English." -http://roa.rutgers.edu/files/467-0901/. Eisikovits, Edina (1991): "Variation in subject-verb agreement in Inner Sydney English." - In: Jenny Cheshire (ed.): English around the world: sociolinguistic perspectives, 235-255. Cambridge: Cambridge University Press. Ellis, Michael (1994): "Literary dialect as linguistic evidence: Subject-verb concord in 19th-century Southern literature." - In: American Speech 69: 128-144. Evans, D. Simon (1971): "Concord in Middle Welsh." - In: Studia Celtica 6: 42-56. Feagin, Crawford (1979): Variation and change in Alabama English. Georgetown: Georgetown University Press.

202 Ferguson, Charles A. (1996): "Variation and drift: Loss of agreement in Germanic." - In: Gregory Guy et al. (eds.): Towards a social science oj language: papers in honor of William Labov. Vol. 1: Variation and change in language and society, 173-198. Amsterdam: Benjamins. Filppula, Markku (1994): From Anglo-Irish to Hiberno-English: divergence and convergence in the Irish dialects of English. - Stuttgart: Franz Steiner. - (1997): "The grammar of the Irish Dialects of English: Some problems and controversies." - In: Tristram (ed.), 225-244. - (1999): The grammar of Irish English: language in Hibernian style. - London: Routledge. - , Juhani Klemola, Heli Pitkänen (2002): "English and Celtic in Contact: the state of the art in research." - In: Filppula/Klemola/Pitkänen (eds.). - , Juhani Klemola, Heli Pitkänen (eds.) (2002): The Celtic Roots of English. - Joensuu: University of Joensuu. Garcia-Bermejo Giner, Maria, Michael Montgomery (1997): "British regional English in the nineteenth century: the evidence from emigrant letters." - In: Alan R. Thomas (ed.): Issues and methods in dialectology, 167-183. Bangor: University of Wales. Gillies, William (1993): "Scottish Gaelic." - In: Martin J. Ball (ed.): The Celtic Languages, 145-227. London: Routledge. Glaser, Elvira (1997): "Dialektsyntax: eine Forschungsaufgabe." - In: Schweizerdeutsches Wörterbuch: Bericht über das Jahr 1996. Zürich: Rotkreuz. 11-31. Godfrey, Elizabeth, Sali Tagliamonte (1999): "Another piece for the verbal -s story: evidence from Devon in southwest England." - In: Language Variation and Change 11: 87-121. Görlach, Manfred (2002): A textual history of Scots. - Heidelberg: Winter Gregg, Robert J. (1972): "The Scotch-Irish dialect boundaries in Ulster." - In: Martin F. Wakelin (ed.): Patterns in the folk speech of the British Isles, 109-139. London: Athlone Press. Grenander, Ulf (1967): "Syntax-controlled probabilities." - Technical report, Brown University, Division of Applied Mathematics. Guy, Gregory (1988): "Advanced Varbrul analysis." - In: K. Ferrara, B. Brown, K. Walters, J. Baugh (eds.): Linguistic change and contact, 124—136. Austin: University of Texas Press. - (1997): "Violable is variable: OT and linguistic variation." - In: Language Variation and Change 9: 333-324. Haeberli, Eric (2001): "Adjuncts and the syntax of subjects in Old and Middle English." - In: Pintzuk/Tsoulas/Wamer (eds.), 109-131. Haegemann, Liliane, Jacqueline Gueron (1999): English grammar: a generative perspective. - Oxford: Blackwell. Hagan, Anette (2002): Urban Scots dialect writing. - Berlin: Lang. Hamp, Eric P. (1976): "Miscellanea Celtica." - In: Studia Celtica 10/11: 54-73. Härke, Heinrich (1995a): "Finding Britons in Anglo-Saxon graves." - In: British Archaeology 10: 7. - (1995b): "Immigrants and natives: a provisional model of Anglo-Saxon ethnogenesis." - In: Early Medieval Europe 4: 1-18. - (1998): "Briten und Angelsachsen im nachrömischen England: zum Nachweis der einheimischen Bevölkerung in den angelsächsischen Landnahmegebieten." - In: Hans-Jürgen Häßler (ed.): Die Wanderung der Angeln nach England: 46. Internationales Sachsensymposium, Schleswig 1995, 87-119. Oldenburg: Isensee. - (1999): "Sächsische Ethnizität und archäologische Deutung im frühmittelalterlichen England." In: Hans-Jürgen Häßler (ed.): Sachsen und Franken in Westfalen. Paderborner Kolloquium 1997, 109-122. Oldenburg: Isensee. Harris, John (1984): "Syntactic variation and dialect divergence." - In: Journal of Linguistics 20: 303-327. - (1991a): "Ireland." - In: Jenny Cheshire (ed.): English around the world: Sociolinguistic perspectives, 37-50. Cambridge: Cambridge University Press.

203 -

(1991b): "Conservatism versus substratal transfer in Irish English." - In: Trudgill/Chambers (eds.), 191-214. - (1993): "The grammar of Irish English." - In: Milroy/Milroy (eds.), 139-186. Haspelmath, Martin (1999): "Optimality and diachronic adaptation." - In: Zeitschrift fur Sprachwissenschaft 18: 180-205. - (1984): Die skandinavischen Sprachen: eine Einführung in ihre Geschichte. [= German translation of The Scandinavian languages (1976)] - Hamburg: Helmut Buske. Hawkins, John A. (1996): A performance theory of order and constituency. - Cambridge: Cambridge University Press. Hayes, Bruce P. (2000): "Gradient well-formedness in Optimality Theory." - In: Dekkers/van der Leeuw/van de Weijer (eds.), 88-120. - , Bruce Tesar, Kie Zuraw (2000): "OTSoft software package." http://www. linguistics, ucla. edu/people/hayes/otsoft/. Hazen, Kirk (1996): "Dialect affinity and subject verb concord: the Appalachian-Outer Banks connection." - In: The SECOL Review: Southeastern Conference on Linguistics 20: 25-53. Henry, Alison (1994): "Singular concord in Belfast English." - In: Belfast Working Papers on Language and Linguistics 12: 134—176. - (1995): Belfast English and Standard English: Dialectal variation and parameter setting. - Oxford: Oxford University Press. - (1997): "The syntax of Belfast English." - In: Kallen (ed.), 89-108. - (2002): "Variation and syntactic theory." - In: Chambers/Trudgill/Schilling-Estes (eds.), 267-282. Hickey, Raymond (1995): "Early contact and parallels between English and Celtic." - In: Vienna English Working Papers 4: 87-119. - (2002): Tape-Recorded Survey of Hiberno-English Speech: digitized sound-files. CD-ROM. Higham, Nicholas (1986): The Northern Counties to AD 1000. - London: Longman. - (1992): Rome, Britain and the Anglo-Saxons. - London: Seaby. - (1993): The Kingdom of Northumbria AD 350-1100. - Phoenix Mill: Alan Sutton. - (1994): The English conquest: Gildas and Britain in the fifth century. - Manchester: Manchester University Press. - (1995): An English empire: Bede and the early Anglo-Saxon kings. - Manchester: Manchester University Press. - (2002): "The Anglo-Saxon/British interface: history and ideology." In: Filppula/Klemola/Pitkänen (eds.). Holmes, Jasper, Richard Hudson (2001): "Constructions in Word Grammar." http://www.phon. ucl.ac. uk/home/dick/construct. htm. Holmqvist, Erik (1922): On the history of the English present inflections, particulary -th and -s. Heidelberg: Winter. Hopper, Paul (1987): "Emergent grammar." - In: Berkeley Linguistics Society 13: 139-157. - , Sandra Thompson (1993): Grammaticalization. - Cambridge: Cambridge University Press. Horning, James J. (1969): "A study of grammatical inference." - Stanford University, Computer Science Department. Hudson, Richard (1997): "Inherent variability and linguistic theory." - In: Cognitive Linguistics 8: 73-108. - (1999): "Subject-verb agreement in English." - In: English Language and Linguistics 3: 173-207. - (2001): "Word Grammar." http://www.phon.ucl.ac.uk/home/dick/hcl.htm. Ihalainen, Ossi (1985): "Synchronic variation and linguistic change: evidence from British English dialects." - In: Roger Eaton, Olga Fischer, Willem Koopman, Frederike van der Leek (eds.): Papers from the 4th International Conference on English Historical Linguistics, Amsterdam, 61-72. Amsterdam: Benjamins.

204 -

(1986): "An inquiry into the nature of mixed grammars: two cases of grammatical variation in dialectal British English." - In: Kastovsky/Szwedek (eds.), 372-379. - (1991): "On grammatical diffusion in Somerset folk speech." - In: Trudgill/Chambers (eds.), 104— 119. - (1994): "The dialects of England since 1776." - In: Robert Burchfield (ed.): The Cambridge History of the English language. Vol. 5: English in Britain and overseas: origins and developments, 197-276. Cambridge: Cambridge University Press. Jackson, Kenneth (1953): Language and history in early Britain. - Edinburgh: Edinburgh University Press. Jones, Charles (ed.) (1997): The Edinburgh history of the Scots language. - Edinburgh: University Press. Jones, Glanville R. J. (1978): "Celts, Saxons and Scandinavians." - In: Robert A. Dodgshon, Robin A. Butlin (eds.): An historical geography of England and Wales, 57-80. London: Academic Press. Kallen, Jeffrey (1986): "Linguistic fundamentals for Hiberno-English syntax." - PhD thesis, Trinity College. Dublin. - (1991): "Intra-language transfer and plural subject concord in Irish and Appalachian English." In: Teanga: The Yearbook of Applied Linguistics 11: 20-34. - (ed.) (1997): Focus on Ireland. - Amsterdam: Benjamins. Kastovsky, Dieter, Alexander Szwedek (eds.) (1986): Linguistics across historical and geographical boundaries. Vol. 1: Linguistic theory and historical linguistics. - Berlin: Mouton de Gruyter. Kayne, Richard S. (1989): "Notes on English agreement." - In: CIEFL Bulletin I: 41-67. - (1995): "Agreement in three English dialects." - In: Hubert Haider, Susan Olson, Sten Vikner (eds.): Studies in comparative Germanic syntax, 159-165. Dordrecht: Kluwer. Keller, Rudi (1990): Sprachwandel: von der unsichtbaren Hand in der Sprache. - Tübingen: Francke. Keller, Wolfgang (1925): "Skandinavischer Einfluss in der englischen Flexion." - In: W. Keller (ed.): Probleme der englischen Sprache und Kultur: Festschrift Johannes Hoops, 80-87. Heidelberg: Winter. Kemner, Suzanne, Michael Israel (1994): "Variation and the usage-based model." - In: Beals et al. (eds.), 165-179. Kherts, Μ. M. (1968): "Entropy of languages generated by automated or context-free grammars with a single-valued derivation." - In: Nauchno-Tekhnicheskaia Informatsia 2: 29-34. Kimball, John, Judith Aissen (1971): "I think, you think, he think." - Linguistic inquiry 2: 242-246. King, Anne (1997): "The inflectional morphology of Older Scots." - In: Jones (ed.), 156-181. King, Ruth (1994): "Subject-verb agreement in Newfoundland French." - In: Language Variation and Change 6: 239-253. Kirk, John M. (1991): Northern Ireland Transcribed Corpus of Speech. Vol. 1 (with S. West and S. Gibson): Textfile. - Colchester: Economic and Social Research Council Data Archive, University of Essex, Colchester. - (1992): "The Northern Ireland Transcribed Corpus of Speech." - In: Gerhard Leitner (ed.): New directions in English language corpora: methodology, results, software developments, 63-73. Berlin, New York: Mouton de Gruyter. - (1997): "Ulster English: The state of the art." - In: Tristram (ed.), 135-179. - (1998): "Ulster Scots: realities and myths." - In: Ulster Folklife 44: 1-25. - , Georgina Millar (1998): "Verbal aspect in the Scots and English of Ulster." - In: Scottish Language 17: 82-107. Klein, Sheldon (1965): "Control of style with a generative grammar." - In: Language 41: 619-631. Klein, Wolfgang (1974): Variation in der Sprache: ein Verfahren zu ihrer Beschreibung. - Kronberg: Scriptor. Klemola, Juhani (1996): "Non-standard periphrastic do: a study in variation and change." - PhD thesis, University of Essex.

205 -

(1999): "The Survey of English Dialects on CD-ROM: The spoken corpus recorded in England 1948-1961."- Presentation given at the Freiburg-Helsinki Workshop on Dialect Corpora and Dialect Syntax, Freiburg, May 1999. - (2000): "The origins of the Northern Subject Rule: A case of early contact?." - In: Tristram (ed.), 329-346. Knecht, Jakob (1911): Die Kongruenz zwischen Subjekt und Prädikat und die 3. Person Pluralis Präsentis auf -s im Elisabethanischen Englisch. - Heidelberg: Winter. Kock, Emst A. (ed.) (1902): The Benedictine Rule: three Middle-English versions of the Rule of St. Benet and two contemporary rituals for the ordination of nuns. - London: Early English Text Society. Kortmann, Bernd (1999): "Typology and dialectology." - In: Caron, Bernard (ed.): Proceedings of the 16th International Congress of Linguists, Paris 1997. CD-ROM. Amsterdam: Elsevier Science. - (2002): "New prospects for the study of dialect syntax: impetus from syntactic theory and language typology." - In: Barbiers/Kornips/van der Kleij (eds.). - , Clive Upton (eds.) (2004): Varieties of English: The British Isles. - Berlin: Mouton de Gruyter. Kroch, Anthony S. (1989): "Reflexes of grammar in patterns of language change." - In: Language Variation and Change 1: 199-244. - (1994): "Morphosyntactic variation." - In: Beals etal. (eds.), 180-201. - (2001): "Syntactic change." - In: Mark Baltin, Chris Collins (eds.): Handbook of contemporary syntactic theory, 699-729. Oxford: Blackwell. - , Ann Taylor (1997): "Verb movement in Old and Middle English: Dialect variation and language contact." - In: Ans van Kemenade, Nigel Vincent (eds.): Parameters of morpho-syntactic change, 297-325. Cambridge: Cambridge University Press. Krug, Manfred (2000): Emerging English modals: A corpus-based study of grammaticalization. Berlin: Mouton de Gruyter. - (2001): "Frequency, iconicity, categorization: Evidence from emerging modals." - In: Bybee/Hopper (eds.), 309-336. - (2003): "Frequency as a determinant in grammatical variation and change." - In: Günter Rohdenburg, Britta Mondorf (eds.): Determinants of grammatical variation, 7-67. Berlin: Mouton de Gruyter. Kytö, Merja (1993): "Third-person present singular verb inflection in early British and American English."- In: Language Variation and Change 5: 113-139. Labov, William (1994): Priciples of linguistic change. Volume I: Internal factors. - Oxford: Blackwell. Laing, Lloyd R., Jennifer Laing (1990): Celtic Britain and Ireland: The myth of the dark ages. Dublin: Irish Academic Press. Laing, Margaret (1978): Studies in the dialect material of medieval Lincolnshire. - PhD thesis, University of Edinburgh. - (1988): "Dialectal analysis and linguistically composite texts in Middle English." - In: Speculum 63:83-103. Lambrecht, Knud (1988): "There was a farmer had a dog: Syntactic amalgams revisited." - In: Berkeley Linguistics Society 14: 319-339. Langacker, Ronald W. (1987): Foundations of Cognitive Grammar. Vol. 1: Theoretical prerequisites. - Stanford: Stanford University Press. - (1990): Concept, image, and symbol: The cognitive basis of grammar. - Berlin: Mouton de Gruyter. - (1991): Foundations of Cognitive Grammar. Vol. 2: Descriptive application. - Stanford: Stanford University Press.

206 Lewis, Henry, Holger Pedersen (1937, 2 1961): A Concise Comparative Celtic Grammar. - Göttingen: Vanderhoeck & Ruprecht. Lightfoot, David (2002): "Introduction." - In: David Lightfoot (ed.): Syntactic effects of morphological change, 1-22. Oxford: Oxford University Press. Lutz, Angelika (1991): Phonotaktisch gesteuerte Konsonantenveränderungen in der Geschichte des Englischen. - Tübingen: Niemeyer. - (1992): "Lexical and morphological consequences of phonotactic change in the history of English." - In: Rissanen/Nevalainen/Taavitsainen (eds.), 156-166. Macafee, Caroline (1983): Glasgow. - Amsterdam: Benjamins. - (1988): "Some studies in the Glasgow vernacular." - PhD thesis, University of Glasgow. - (1993): "A short grammar of Older Scots." - In: Scottish Language 11/12: 10-36. - (1994): Traditional dialect in the modern world: a Glasgow case study. - Frankfurt: Lang. - (1996): A Concise Ulster Dictionary. - Oxford: Oxford University Press. - (1997): "Ongoing change in Modern Scots: the social dimension." - In: Jones (ed.), 514—550. - , Colm 0 Baoill (1997): "Why Scots is not a Celtic English." - In: Tristram (ed.), 245-286. Macaulay, Ronald K. S. (1991): Locating dialect in discourse: the language of honest men and bonny lasses in Ayr. - New York: Oxford UP. McCafferty, Kevin (1998a): "Shared accents, divided speech community? Change in Northern Ireland English." - In: Language Variation and Change 10: 97-121. - (1998b): "Barriers to change: Ethnic division and phonological innovation in Northern HibernoEnglish." - In: English World-Wide 19: 7-35. - (1999): "(London)Derry: Between Ulster and local speech - class, ethnicity and language change." - In: Paul Foulkes, Gerard Docherty (eds.): Urban voices: Accent studies in the British Isles, 246264. London: Arnold. - (2003): "The Northern Subject Rule in Ulster: How Scots, how English?" - In: Language Variation and Change 15: 105-139. - (2004): "'[T]hunder storms is verry dangese in this countrey they come in less than a minnits notice ...': The Northern Subject Rule in Southern Irish English." - In: English World-Wide 25: 51-79. Mac Eoin, Gearoid (1993): "Irish." - In: Martin J. Ball (ed.): The Celtic Languages, 101-144. London: Routledge. Mcintosh, Angus (1988): "Present indicative plural forms in the later Middle English of the North Midlands." - In: Angus Mcintosh et al. (eds.): Middle English dialectology: essays on some principles and problems, 116-122. Aberdeen: Aberdeen University Press. - , Michael L. Samuels, Michael Bensik (1986): A linguistic atlas of late mediaeval English. 4 vols. - Aberdeen: Aberdeen University Press. MacWhinney, Brian (2001): "Emergentist approaches to language." - In: Bybee/Hopper (eds.), 449470. Mather, James Y., Hans-Henning Speitel (1975-1986): The linguistic atlas of Scotland: Scots section. 3 vols. - London: Croom Helm. Meechan, Marjory, Michele Foley (1994): "On resolving disagreement: linguistic theory and variation - there's bridges." - In: Language Variation and Change 6: 63-85. Meisel, Jürgen Μ. (ed.) (1977): Langues en contact - Pidgins - Creoles - Languages in contact. Tübingen: Narr. Meurman-Solin, Anneli (1992): "On the morphology of verbs in Middle Scots." - In: Rissanen/Nevalainen/Taavitsainen (eds.), 611-623. Meurman-Solin, Anneli (1993): Variation and change in early Scottish prose: studies based on the Helsinki Corpus of Older Scots. - Helsinki: Suomalainen Tiedeakatemia. Miller, Jim E. (1993): "The grammar of Scottish English." - In: Milroy/Milroy, 99-138. - (2004): "The morphology and syntax of Scottish English." - In: Kortmann/Upton (eds.).

207 - , Keith Brown (1982): "Aspects of Scottish English syntax." - In: English World-Wide 3: 3-17. Milroy, James, Lesley Milroy (eds.) (1993): Real English: The grammar of English dialects in the British Isles. - London: Longman. Milroy, Lesley (1987): Investigating and understanding natural language. - Oxford: Blackwell. Moessner, Lilo (1988): "Is the subject-verb concord in Scots a number concord?" - In: Richard Matthews, Joachim Schmole-Rostovsky (eds.): Papers in language and mediaeval studies presented to Alfred Schöpf, 177-191. Frankfurt: Lang. - (1997): "The syntax of Older Scots." - In: Jones (ed.), 112-155. Montgomery, Michael (1988): The roots of Appalachian English. - Clevedon: Multilingual Matters. - (1991): "The anglicization of Scots in seventeenth-century Ulster." - In: Studies in Scottish Literature 26: 50-64. - (1992): "The linguistic value of Ulster emigrant letters." - In: Ulster Folklife 41: 26-41. - (1994): "The evolution of verb concord in Scots." - In: Alexander Fenton, Donald McDonald (eds.): Studies in Scots and Gaelic: Proceedings of the Third International Conferences on the Languages of Scotland, 81-95. Edinburgh: Canongate. - (1997a): "A tale of two Georges: Regional variation in 18th century Hiberno-English." - In: Kallen (ed.), 227-254. - (1997b): "Making transatlantic connections between varieties of English: The case of plural verbal -s" - In: Journal of English Linguistics 25: 122-141. - , Janet M. Fuller (1996): "What was verbal - i in 19th-century African American English?" - In: Edgar Schneider (ed.): Focus on the USA, 211-230. Amsterdam: Benjamins. - , Janet M. Fuller, Sharon DeMarse (1993): "'The Black men has wives and sweet hearts (and third person plural -s) jest like the white men': Evidence for verbal -s from written documents on 19thcentury African American speech." - In: Language Variation and Change 5: 335-357. - , Robert Gregg (1997): "The Scots language in Ulster." - In: Jones (ed.), 569-622. Moore, Samuel, Stanford B. Meech, Harold Whitehall (1935): "Middle English dialect characteristics and dialect boundaries." - In: Essays and Studies in English and Comparative Literature 13: 1-60. Moravcsik, Edith A. (1978): "Agreement." - In: Joseph Greenberg (ed.): Universals of Human Language. Vol. 4: Syntax, 331-374. Palo Alto: Stanford University Press. Morris-Jones, John (1913): A Welsh Grammar, historical and comparative: phonology and accidence. - Oxford: Clarendon. Mufwene, Salikoko S. (1994): "Theoretical linguistics and variation analysis: strange bedfellows?" In: Beals etal. (eds.), 202-217. Murray, James A. H. (1873): The dialect of the southern counties of Scotland. - London: Philological Society. Mustanoja, Tauno F. (1960): A Middle English syntax. - Helsinki: Societe Neophilologique. Nevalainen, Terttu, Helena Raumolin-Brunberg (2000a): "The changing role of London on the linguistic map of Tudor and Stuart England." - In: Dieter Kastovsky, Arthur Mettinger (eds.): The history of English in a social context, 279-337. Berlin: Mouton de Gruyter. - (2000b): "The third-person singular -(e)s and -(e)th revisited: The morphophonemic hypothesis." In: Christiane Dalton-Puffer, Nikolaus Ritt (eds.): Words: Structure, meaning, function. Berlin: Mouton de Gruyter. Newmeyer, Frederick J. (1998): Language form and language function. - Cambridge/Ms.: MIT Press. - (2002a): "Optimality and functionality: A critique of functionally-based Optimality Syntax." - In: Natural Language and Linguistic Theory 20: 1-42. - (2002b): "A rejoinder to Bresnan and Aissen." - In: Natural Language and Linguistic Theory 20: 81-95. - (2003): "Grammar is grammar and usage is usage." - In: Language 79: 682-707. Noreen, Adolf ( 4 1923): Altisländische und altnorwegische Grammatik (Laut- und Flexionslehre) unter Berücksichtigung des Urnordischen. - Halle: Niemeyer.

208 Ogura, Mieko, William S. Wang (1996): "Snowball effect in lexical diffusion: The development of-5 in the third person singular present indicative in English." - In: Derek Britton (ed.): English historical linguistics 1994: Papers from the 8th International Conference on English Historical Linguistics, Edinburgh 1994, 119-142. Amsterdam: Benjamins. 0 Dochartaigh, Cathair (1992): "The Irish Language." - In: Donald Macaulay (ed.): The Celtic Languages, 11-99. Cambridge: Cambridge University Press. Ο hÜrdail, Roibeärd (1997): "Hiberno-English: Historical background and synchronic features and variation." - In: Tristram (ed.), 123-134. Orton, Harold, et al. (ed.) (1962-1971): Survey of English Dialects. (B): The basic material. 4 vols. Leeds: E. J. Arnolds. Page, Raymond I. (1959): "Language and dating in OE inscriptions." - In: Anglia 77: 385-406. - (1973): An introduction to English runes. - London: Methuen. Paolillo, John C. (2002): Analyzing linguistic variation: Statistical models and methods. Stanford: CSL1 Publications. Parry, David (1977-1979): Survey of Anglo-Welsh Dialects. 2 vols. - Swansea: University College. Peitsara, Kirsti (2002): "Verbal -s in Devonshire: The Helsinki Dialect Corpus evidence." - In: Helena Raumolin-Brunberg (ed.), Variation past and present: VARIENG studies on English for Terttu Nevalainen, 211-230. Helsinki: Societe Neophilologique. Percy, Carol (1991): "Variation between -(e)th and -(e)s spellings of the third person singular present indicative: Captain James Cook's 'Endeavour' journal 1768-1771." - In: Neuphilologische Mitteilungen 92: 351-358. Petyt, Keith M. (1985): Dialect and accent in industrial West Yorkshire. - Amsterdam: Benjamins. Pietsch, Lukas (2001): "Subject-verb agreement in Northern Ireland: Variation across space and apparent time." - Paper held at the 3rd UK Language Variation Conference, York, July 2001. Pietsch, Lukas (2005): "'Some do and some doesn't': Verbal concord variation in the north of the British Isles." - In: Bernd Kortmann, Tanja Herrmann, Lukas Pietsch, Susanne Wagner (eds.): A comparative grammar of English dialects: Agreement, gender, relative clauses. Berlin: Mouton de Gruyter. Pintzuk, Susan (1988): "Varbrul, version 2.3." Software package and documentation. ftp://ftp. cis. upenn. edu/pub/ldc/misc-sw/varbrul. tar.z. - (1993): "Verb seconding in Old English: Verb movement to Infi." - In: The Linguistic Review 10: 5-35. - (1995): "Variation and change in Old English clause structure." - In: Language Variation and Change 7: 229-260. - (1996): "Cliticization in Old English." - In: Aaron Halpern, Arnold Zwicky (eds.): Approaching second: Second position clitics and related phenomena, 375—409. Standford: CSLI Publications. - (1998): "Syntactic change via grammatical competition: Evidence from Old English." - In: Paolo Ramat, Elisa Roma (eds.): Sintassi storica, 111-125. Rome: Bulzoni. - , George Tsoulas, Anthony Warner (2001): "Syntactic change: Theory and method." - In: Pintzuk/Tsoulas/Wamer (eds.), 1-24. - , George Tsoulas, Anthony Warner (eds.) (2001): Diachronic syntax: Models and mechanisms. Oxford: Oxford University Press. Policansky, Linda (1982): "Grammatical variation in Belfast English." - In: Belfast Working Papers on Language and Linguistics 6: 37-66. Poliert, Karl (1881): "Die 3. Person Pluralis auf -s bei Shakespeare." - PhD thesis, University of Marburg. Pollner, Clausdirk (1985): "Linguistic fieldwork in a Scottish New Town." - In: Manfred Görlach (ed.): Focus on Scotland, 57-68. Amsterdam: Benjamins. Poplack, Shana (2001): "Variability, frequency, and productivity in the irrealis domain of French." In: Bybee/Hopper (eds.), 405-430.

209 -

, Sali Tagliamonte (1989): "There's no tense like the present: Verbal -s inflection in early Black English." - In: Language Variation and Change 1: 47-84. Poussa, Patricia (1990): "A contact-universal origin for periphrastic DO, with special consideration of OE-Celtic contact." - In: Sylvia Adamson et al. (eds.): Papers from the 5th International Conference on English Historical Linguistics, 407-434. Amsterdam: Benjamins. Prince, Alan, Paul Smolensky (1993): Optimality theory: Constraint interaction in generative grammar. - Piscateway, NJ: Rutgers University Center for Cognitive Science. Pyles, Thomas, John Algeo (1993): The origins and development of the English Language. - Orlando: Harcourt Brace. Quirk, Randolph G., Sidney Greenbaum, Geoffrey Leech, Jan Svartvik (1985): A comprehensive grammar of the English language. - London: Longman. Rickford, John R. (2002): "Implicational scales." - In: Chambers/Trudgill/Schilling-Estes (eds.), 142-167. Rissanen, Ossi I., Terttu Nevalainen, Irma Taavitsainen (eds.) (1992): History of Englishes: New methods and interpretations in historical linguistics. - Berlin: Mouton de Gruyter. Roberts, Ian (1993): Verbs and diachronic syntax: a comparative history of English and French. Dordrecht: Klüver. - (1997): Comparative syntax. - London: Arnold. Robinson, John (2002): "GoldVarb 2001, v. 1.0.3." Software package and documentation. http://www.york.ac.uk/depts/lang. Robinson, Philip (1984): The plantation of Ulster: British settlement in an Irish landscape, 16001670. - Dublin: Gill and Macmillan. - (1989): "The Scots language in seventeenth-century Ulster." - In: Ulster Folklife 35: 86-99. - (1997): Ulster Scots: A grammar of the traditional written and spoken language. - Belfast: Ullans Press. Rodeffer, John D. (1903): The inflection of the English present plural indicative with special reference to the northern dialect. - Baltimore: John Murphy. Rohdenburg, Günter, Julia Schlüter (2000): "Determinanten grammatischer Variation im Früh- und Spätneuenglischen." - In: Sprachwissenschaft 25: 443-496. Ross, Alan S. (1934): "The origin of the j-endings of the Present Indicative in English." - In: Journal of English and Germanic Philology 33: 68-73. Rousseau, Pascale, David Sankoff (1978a): "Advances in variable rule methodology." - In: Sankoff (ed.), 57-70 Rousseau, Pascale, David Sankoff (1978b): "A solution to the problem of grouping speakers." - In: Sankoff (ed.), 97-118. Rupp, Laura (fc.): "Constraints on non-standard -s in expletive there sentences: A generativevariationist perspective." - In: English Language and Linguistics. Sabban, Annette (1982): Gälisch-englischer Sprachkontakt. - Heidelberg: Groos. - (1985): "On the variability of Hebridean English syntax: The verbal group." - In: Manfred Görlach (ed.): Focus on Scotland, 125-143. Amsterdam: Benjamins. Samek-Lodovici, Vieri (2001): "Agreement impoverishment under subject inversion: A crosslinguistic analysis." http://ruccs.rutgers.edu/roa.html. Samuels, Michael L. (1985): "The great Scandinavian belt." - In: Roger Eaton, Olga Fischer, Willem Koopman, Frederike van der Leek (eds.) (1985): Papers from the 4th International Conference on English Historical Linguistics, Amsterdam, 269-281. Amsterdam: Benjamins. - (1988): "The great Scandinavian belt." - In: Angus Mcintosh et al. (eds.): Middle English dialectology: essays on some principles and problems, 106-115. Aberdeen: Aberdeen University Press. Sankoff, David (ed.) (1978): Linguistic variation: Models and methods. - London: Academic Press. - (1988): "Variable rules." - In: Ulrich Ammon (ed.): Soziolinguistik, 984-997. Berlin: de Gruyter.

210 -

(1998): "Sociolinguistics and syntactic variation." - In: Frederick Newmeyer (ed.): Linguistics: The Cambridge Survey. Vol. 4, 140-161. Cambridge: Cambridge University Press. - , William Labov (1979): "On the uses of variable rules." - In: Language in Society 8: 189-222. Sankoff, Gillian (2002): "Linguistic outcomes of language contact." In: Chambers/Trudgill/Schilling-Estes (eds.), 636-668. Schendl, Herbert (1996): "The 3rd plural present indicative in Early Modern English: Variation and linguistic contact." - In: Derek Britton (ed.): English historical linguistics 1994: Papers from the 8th International Conference on English Historical Linguistics, Edinburgh 1994, 143-160. Amsterdam: Benjamins. - (2000): "The third person present plural in Shakespeare's first folio: A case of interaction of morphology and syntax?" - In: Christiane Dalton-Puffer, Nikolaus Ritt (eds.): Words: Structure, meaning, function, 263-276. Berlin: Mouton de Gruyter. Schilling-Estes, Natalie, Walt Wolfram (1994): "Convergent explanation and alternative regularization patterns: Were/weren't levelling in a vernacular English variety." - In: Language Variation and Change 6: 273-302. Schneider, Edgar (1983): "The origin of the verbal -s in Black English." - In: American Speech 58: 99-113. - (2002): "Investigating variation and change in written documents." - In: Chambers, Trudgill, Schilling-Estes (eds.), 67-96. Schreier, Daniel (2002): "Past be in Tristan da Cunha: The rise and fall of categoricality in language change." - In: American Speech 77: 70-99. Schrijver, Peter (2002): "Phonological evidence for early medieval close contacts between Celtic and West Germanic." - In: Filppula/Klemola/Pitkänen (eds.). Shorrocks, Graham (1999): A grammar of the dialect of the Bolton area. Vol. 2: Morphology and syntax. - Berlin: Lang. Smith, C. A. (1896): "Shakespeare's present indicative -s endings with plural subjects."- In: Publications of the Modern Language Association 11: 362-376. Smith, Jennifer, Sali Tagliamonte (1998): '"We was all thegither, I think we were all thegither': Was regularization in Buckie English." - In: World Englishes 17: 105-126. Soule, Stephen (1974): "Entropies of probabilistic grammars." - In: Information Control 25: 57-74. von Staden, Wilhelm (1903): "Entwicklung der Präsens-Indikativ-Endungen im Englischen." - PhD thesis, University of Rostock. Stein, Dieter (1986): "Old English Northumbrian verb inflection revisited." - In: Kastovsky/Szwedek (eds.), 637-650. - (1987): "On the crossroads of philology, linguistics, and semiotics: Notes on the replacement of th by ä in the third person singular in English." - In: English Studies 68: 406-432. Stoelke, Hans (1916): Die Inkongruenz zwischen Subjekt und Prädikat im Englischen und in den verwandten Sprachen. - Heidelberg: Winter. Strauss, Dietrich, Horst W. Drescher (eds.) (1986): Scottish language and literature: Medieval and renaissance. - Frankfurt: Peter Lang. Suppes, Patrick (1970): "Probabilistic grammars for natural languages." - In: Synthese 22: 95-116. Tagliamonte, Sali (1998): "Was/were variation across the generations: View from the city of York." In: Language Variation and Change 10: 153-191. - (2002): "Comparative sociolinguistics." - In: Chambers/Trudgill/Schilling-Estes (eds.), 729-763. - , Jennifer Smith (2000): "Old was, new ecology: Viewing English through the sociolinguistic filter." - In: Shana Poplack (ed.): The English history of African American speech, 141-171. Oxford: Blackwell. Tesar, Bruce, Paul Smolensky (2000): "The learnability of Optimality Theory: An algorithm and some basic complexity results." http://ruccs.rutgers.edu/roa.html.

211 Theakston, Anna L., Elena V. Lieven (fc.): "The acquisition of auxiliaries be and have: An elicitation study." - In: Journal of Child Language. Theakston, Anna L., Elena V. Lieven, Julian M. Pine, Caroline F. Rowland (2005): "The acquisition of auxiliary syntax: be and have." - In: Cognitive Linguistics 16. Thomason, Sarah G., Terrence Kaufman (1988): Language contact, creolization and genetic linguistics. - Berkeley: University of California Press. Thumeysen, Rudolf (1980): A grammar of Old Irish. [Revised and enlarged edition, transl. by D. A. Binchy and Osborn Bergin, of Handbuch des Altirischen (1909)]. - Dublin: Institute for Advanced Studies. Todd, Loreto (1984): "By their tongue divided: Towards an analysis of speech communities in Northern Ireland." - In: English World-Wide 5: 159-180. Tomasello, Michael (2003): Constructing a language: A usage-based theory of language acquisition. Cambridge/Ms.: Harvard University Press. Tristram, Hildegard (1997): "Introduction." - In: Tristram (ed.), 1-26. - (ed.) (1997): The Celtic Englishes. - Heidelberg: Winter. - (1999): How Celtic is Standard English? - St. Petersburg: Nauka. - (ed.) (2000): The Celtic Englishes II. Heidelberg: Winter. - (2002): "Attrition of inflexions in Old English and Old Welsh." - In: Filppula/Klemola/Pitkänen (eds.). Trudgill, Peter (1990): The dialects of England. - Oxford: Blackwell. Trudgill, Peter ( 4 2000): Sociolinguistics: An introduction to language and society. - Harmondsworth: Penguin. - , Jack K. Chambers (eds.) (1991): Dialects of English: Studies in grammatical variation. - London: Longman. Upton, Clive, David Parry, John Widdowson (1994): Survey of English Dialects: The dictionary and grammar. - London: Routledge. Van Gelderen, Elly (1997): Verbal agreement and the grammar behind its 'breakdown': Minimalist feature checking. Tübingen: Niemeyer. Vennemann, Theo (2001): "Atlantis Semitica: Structural contact features in Celtic and English." - In: Laurel Brinton (ed.): Historical Linguistics 1999: Selected Papers from the 14th International Conference on Historical Linguistics. Amsterdam: Benjamins. 351-369. - (2002a) "Semitic > Celtic > English: The transitivity of language contact." - In: Filppula/Klemola/ Pitkänen (eds.). - (2002b): "On the rise of 'Celtic' syntax in Middle English." - In: Peter J. Lucas, Angela M. Lucas (eds.): Middle English from tongue to text: Selected papers from the Third International Conference on Middle English: Language and Text, Dublin 1999. Bern: Peter Lang. Viereck, Wolfgang (1991/1997): The Computer-Developed Linguistic Atlas of England. 2 vols. Tübingen: Niemeyer. Visser, Frederic. Th. (1963): An historical syntax of the English language. Vol. 1. - Leiden: E. J. Brill. Wakelin, Martyn F. (1977): English dialects: An introduction. - London: Athlone Press. - (1984): "Rural dialects in England." - In: Peter Trudgill (ed.): Language in the British Isles, 7 0 93. Cambridge: Cambridge University Press. Weinreich, Uriel, William Labov, Marvin I. Herzog (1968): "Empirical foundations for a theory of language change." - In: Winfried P. Lehmann, Y. Malkiel (eds.): Directions for historical linguistics. A symposium, 95-188. Austin: University of Texas Press. Wilson, James (1915): Lowland Scotch, as spoken in the lower Strathearn district of Perthshire. London: Oxford University Press. Wilson, John, Alison Henry (1998): "Parameter setting within a socially realistic linguistics." - In: Language in Society 27: 1-21.

212 Wolfram, Walt (1973): "On what basis variable rules?" - In: Charles Bailey, Roger W. Shuy (eds.): New ways of analyzing variation in English, 1-12. Washington DC: Georgetown University Press. Wolfram, Walt (2000): "Issues in reconstructing earlier African-American English." - In: World Englishes 19: 39-58. Woods, Anthony, Paul Fletcher, Arthur Hughes (1986): Statistics in language studies. - Cambridge: Cambridge University Press. Wright, Joseph (1892): A grammar of the dialect of Windhill in the West Riding of Yorkshire. - London: Kegan Paul, Trench, Trübner and Co. - (1905): The English Dialect Grammar. - Oxford: Henry Frowde. Wright, Laura (2002): "Third person plural present tense markers in London prisoners' depositions, 1562-1623." - In: American Speech 77: 242-263. Yang, Charles D. (2003): Knowledge and learning in natural language. - Oxford: Oxford University Press.

Index

acquisition 39,41,42 Adams, G. Brendan 109, 111 adverbials (between subject and verb) 10, 18, 129 Afro-American Vernacular English 51 age see apparent time agreement passim in Principles-and-Parameters theory 20-21 agreement verb 3 Aissen, Judith 14,31,38 Alabama 51 analogy 60, 61 Anderwald, Lieselotte 90, 146 Antrim see Chapter 5, passim apparent time 121,143 Armagh 23 see also Chapter 5, passim aspect 10,40,93, 120, 124 Ayrshire 140 Bailey, Charles 61,62 Bailey, Guy 23,51 Bambas, Rudolph C. 51 Barlow, Michael 41 BarTy, Michael 108,110,119 Bates, Elizabeth 42 Bayley, Robert 37, 173 be application of Northern Subject Rule 12-13, 51,60 contracted are in existential there clauses 139-41, 157-62 dialectal forms 38, 74-90, 154, 180 see also was/were Beal, Joan 84 Belfast 14, 17-22, 121 see also Northern Ireland Benveniste, Emile 59 Berdan, Robert 31,37-38 Bemdt, Rolf 47, 54, 56-58, 61, 65 Bickerton, Derek 31,34 Blakeley, Leslie 55,63 ΒNC see British National Corpus Boersma, Paul 38^10 Bolton 14,80 Börjars, Kersti 18,26-28,29 Breivik, Leiv E. 14

Bresnan, Joan 30, 31, 33, 38^10, 68 Britain, David 52, 68, 90 British National Corpus (BNC) 5, 147 Brunner, Karl 56 Buckie 106 Bybee, Joan 15,41,42 Caedmon's Hymn 55 Cambridgeshire 96 Cely letters 23 Central North (dialect area) 75, 81-84, 156 Chambers, Jack K. 52, 90, 106, 157 Chapman, Carol 18,26-28 Cheshire (county) 78-81, 96 Cheshire, Jenny 52,90, 147 Chester-Wash Line 48, 77, 96 cleft constructions 127-28 cliticization 27, 54, 56 code switching 31, 35, 169 Cognitive Grammar 40-46, 169 collective nouns 14,135 competence vs. performance 31-33,39, 169 competing grammars 7, 33 Computer-Developed Linguistic Atlas of England (CLAE) 68 see also Survey of English Dialects (SED) construction grammar 29,40-46, 105, 139, 169 convergence 23 co-ordination in subject NP 20, 136 of verb phrases 8, 18, 25, 104, 116, 129, 155 Corrigan, Karen 22-26, 29, 52, 111,139 Croft, William 31,40-^6,57 Cumberland 64, 74 Cumbria 156 Deo, Ashwin 38^Φ0 Derbyshire 77,78-81 Derry see Londonderry dialect attrition 67, 124, 163, 164 dialect contact 66, 67, 99 do (dialectal forms) 71,80 Donegal 151 Down (county) see Chapter 5, passim Dumfriesshire 156 Durham (city) 62 Durham (county) 74, 156

214 Durham Ritual 55, 56-58, 61, 63 Early Modern English 9, 11, 23, 51 East Midlands (dialect area) 77 Eastern Ross 155 economy 57 Eisikovits, Edina 51, 126 ellipsis (of subject) 129,149 Ellis, Stanley 70 entrenchment 40-46 existential there clauses 14, 106, 139^*9, 15762,215 annex relative clauses after there 127-28 Feagin, Crawford 51 Ferguson, Charles A. 5 4 , 6 1 , 6 3 Fermanagh see Chapter 5, passim fieldworker isoglosses 69, 109 Filppula, Markku 51, 63, 149 floating quantifiers 11 FRED see Freiburg Corpus of English Dialects free variation 31, 169 Freiburg Corpus of English Dialects (FRED) 5 list of informants 216 see also Chapter 6, passim frequency 30, 40-46, 56 Gaeltacht 158 Gibson, Peter H. 70 Glasgow 14 Görlach, Manfred 165 grammaticality judgments 20, 30 grammaticalization 41 Gregg, Robert J. 51, 118, 142, 149 Grenander, Ulf 37 habitual 10,40,93 see also aspect Haeberli, Eric 56 Hagan, Anette 14 Hamp, Eric P. 64 Härke, Heinrich 64 Harris, John 113,146 Haspelmath, Martin 38 have dialectal forms 71, 80 sensitivity to Northern Subject Rule 120 Hawkins, John A. 138 Hayes, Bruce 38-40 Hazen, Kirk 126 Heaviness-of-Subject Constraint 12,23, 129, 137, 167 see also Position-of-Subject Constraint

Helsinki Dialect Corpus 5 Henry, Alison 14, 15,17-22, 29, 30, 31, 113, 121, 134 Herzog, Marvin I. 37 Hiberno-English 51, 110, 149 Hickey, Raymond 64, 108, 152 Higham, Nicholas 64 historic present 13, 93, 113-14, 154 Holmqvist, Erik 55 Hopper, Paul 41,57 Horning, James 37 Hudson, Richard 29-30, 33 Humber-Lune-Ribble Line 77 Ihalainen, Ossi 2, 54, 67, 68 Inverness 155 inversion 8, 96, 100, 125-26, 165 in existential there clauses 148 in Yorkshire tag clauses 102,166 says/(idiom) 113,155 Ireland 51, 149 see also Northern Ireland, Ulster Irish 23, 24-26 Israel, Michael 41,42 Kallen, Jeffrey 14 Kaufman, Terrence 6 1 , 6 3 , 6 4 Kayne, Richard 14 Keller, Rudi 57 Keller, Wolfgang 60-63 Kemmer, Suzanne 41, 42 Kimball, John 14 Kincardineshire 156 King, Anne 51 Kirk, John M. I l l Klein, Sheldon 37 Klemola, Juhani 4, 47, 63-66, 68, 72, 91, 96 Knecht, Jakob 51 Kortmann, Bemd 1 Kroch, Anthony S. 30, 33, 56 Krug, Manfred 41 Kytö, Merja 32,51 Labov, William 31,37-38 Laing, Jennifer 64 Laing, Lloyd R. 64 Laing, Margaret 50 LALME see Linguistic Atlas of Late Middle English Lancashire 10, 14, 52, 77, 78-81, 156 Langacker, Ronald W. 40-46 language contact with Irish 24-26, 111

215 with Old Norse 53, 60-63, 77 with Welsh/Brythonic 25, 63-66 Leicestershire 77 Leiden Riddle 55 Leitrim 151 Lexical-Functional Grammar (LFG) 26-28, 41 LFG see Lexical-Functional Grammar Lieven, Elena V. 41,42 Lincolnshire 87 Lindisfarne Gospels 55, 56-58, 61, 63 Linguistic Atlas of England (LAE) 68 see also Survey of English Dialects Linguistic Atlas of Late Middle English (ULME) 47-50,62,80 Linguistic Survey of Scotland 5 Londonderry (county) 158 see also Chapter 5, passim Lothian 156 Louth 152 Lower North (dialect area) 77,90-93, 156, 160 Macafee, Caroline 14, 50, 64 Macaulay, Ronald K. S. 126, 140 MacWhinney, Brian 42 markedness 15-16, 23, 24, 28, 62, 65, 147 Maroldt, Karl 61,62 McCafferty, Kevin 51, 124, 146, 149 Mcintosh, Angus 6, 47-66 Meurman-Solin, Anneli 50 Middle English 9, 12, 42, 46, 47-66, 80 Midlands Middle English verbal paradigms 48, 50, 54, 56 Miller, Jim 140,164 Milroy, Leslie 1 Montgomery, Michael 2, 6, 10, 13, 14, 50, 51, 60 Moore, Samuel 49 Moravcsik, Edith A. 59 Mufwene, Salikoko S. 33 Murray, James 6, 10, 13, 50, 57 -n (verbal affix) 47-50, 54, 76, 78-81, 87, 180 negation 90, 136, 146, 157, 159 Nevalainen, Terttu 51 Newmeyer, Frederick 31,39 Ni Ghallchoir, Caitriona 109 NITCS see Northern Ireland Transcribed Corpus of Speech Nominative Constraint 20-21 see also Type-of-Subject Constraint Nonconcord(ance) 3

see also Northern Subject Rule; nonagreement (in existential clauses) North Carolina 51 Northern Ireland religious communities 109, 123,145 see also Chapter 5, passim Northern Ireland Transcribed Corpus of Speech (NITCS) 5, 19 list of informants 211 see also Chapter 5, passim Northern Present-Tense Rule 2 see also Northern Subject Rule Northern Subject Rule passim alternative names 2 definition 6 geographical extension 48, 67, 77, 93-99 Northumberland 74, 156 Northumbria 77 Northumbrian see Old English Northwest Midlands (dialect area) 77, 78-81 Nottinghamshire 77 NP/PRO Constraint 3 see also Northern Subject Rule NSR see Northern Subject Rule number shifting/attraction 14 0 Baoill, Colm 64 Ogura, Mieko 51 Old English 47-66 Old Norse 53,60-63 ones (indefinite pronoun) 133 Optimality Theory (OT) 38^10 optional ity 30,31 Orton, Harold 68, 70 Paolillo, John C. 37, 173 Peitsara, Kirsti 52 performance see competence Personal Pronoun Rule 3 see also Northern Subject Rule Perthshire 140 phonological conditioning factors 136, 169 Pine, Julian M. 41 Pintzuk, Susan 30, 56 Playford, Averil 70 Pluraliatantum

135

Policanski, Linda 14 Poliert, Karl 51 Position-of-Subject Constraint 6, 7-12, 45 with NP subjects 129-30, 137, 167 with pronoun subjects 18, 26-28, 29, 104, 114-15, 155

216 postmodification (of subject noun phrases) 12 see also Heaviness-of-Subject Constraint Poussa, Patricia 64 pragmatic constraints 32 Prince, Alan 38 Principles-and-Parameters theory 17-22, 41 processing constraints 26,32, 139 pronouns 3 complex plural pronouns (themuns, youseuns) 134 demonstrative pronouns 3, 33, 67, 96, 131— 32 indefinite pronouns 3, 67, 96, 132, 133 interrogative pronouns 132 nominative marking 20-21 personal pronouns 42,80, 100, 112-17 see Type-of-Subject Constraint strong pronouns 9, 20, 100 proximity agreement 14 Proximity-to-Subject Constraint see Positionof-Subject Constraint relative clauses 8, 18, 67, 96, 99, 126-28, 133, 167 annex clauses in existentials/clefts 127-28 with existential there 148 zero relatives 126-28 Roberts, lan 17 Robinson, Philip 110, 113, 124 Rodeffer, John D. 56-58 Ross, Alan S. 55 Rousseau, Pascale 173,177 Rowland, Caroline F. 41 rune inscriptions 55, 61 Rupp, Laura 14 Rushworth Gloss 5 5 , 6 1 , 6 3 -s (verbal affix) passim distribution in Middle English 47-50, 62 in southwestern England 54, 82, 90-93 its origins 53-56, 65 Samuels, Michael L. 60-63 Sankoff, David 31,37-38, 173, 177 Sahkoff, Gillian 64 Schendl, Herbert 51,57 Schilling-Estes, Natalie 51,147 Schreier, Daniel 51 Scotland 13, 52,66, 106, 140 see also Chapter 6, passim Scots 3, 10 Older Scots 6, 11, 13, 50 Ulster Scots 110,124,162

see also Chapter 5, passim SED see Survey of English Dialects sex (sociolinguistic variable) 122, 143 Shorroeks, Graham 10, 13, 80, 87, 93 Shropshire 78-81 Sievers, Eduard 56 singular concord 3, 18 see also Northern Subject Rule Smith, Jennifer 106, 126, 147, 157 Smolensky, Paul 38 Soule, Stephen 37 southwest England 40, 52 Southwest Midlands (dialect area) 90-93 Staden, Wilhelm von 51 Staffordshire 77,78-81 Stein, Dieter 5 1 , 5 5 , 5 9 , 6 1 , 6 2 Strathclyde 64 Suppes, Patrick 37 Survey of Anglo- Welsh Dialects 5 Survey of English Dialects (SED) 5 basic material 68, 179 fieldworker notebooks 69, 179 incidental material 68, 80, 179 questionnaire items III.1.12 74 III. 10.2-6 73 111.10.3 73 III. 10.7 69,73,93-99, 193 III.5.2 74 IV.6.2 73 IX. 1.9 72 1X.4.10 71 IX.5.1 71 IX.5.4 71, 185 IX.6.1 71 IX.6.2 71 IX.6.3 71 IX.6.4 72 IX.7.1 70, 81, 180, 182 IX.7.10 71, 182 IX.7.2 70, 84, 180, 183, 184, 186 IX.7.3 70, 183, 184 IX.7.4 71, 182 IX.7.5 71, 182, 183, 184, 186 IX.7.6 71,87,90, 187, 190 IX.7.7 71, 81, 180, 182, 183, 184, 186 IX.7.8 71,87 IX.7.9 71, 81, 180, 182, 183, 184, 186 IX.9.4 73 IX.9.5 69,71

217 V.2.12 72, 190 V.7.21 74 V.7.7 72 VI. 14.2 74 Vl.2.8 72 VI.3.2 72 VI.4.2 72 VI.5.11 72 VI.5.8 74, 190 VII.8.18 74 VIII.2.8 73 VIII.3.7 71 VIII.5.1 72 VIII.5.2 73 VIII.7.5 73,93-99, 193 VIII.7.9 74, 190 VIII.9.5 70, 87, 180, 182, 187 see also Chapter 4, Appendix B, passim Sutherland 155 Sydney 51 Sykes, Donald R. 70 tag clauses 90, 140 Yorkshire type 102,166 Tagliamonte, Sali 13, 32, 40, 52, 90, 106, 126, 147, 157, 161, 164 tags clauses with existential there 148 Tape-Recorded Survey of Hiberno-English Speech (TRS) 5, 108-10, 151,216 Teesside 156 Texas 51 -th (verbal affix) 47-66, 54, 60-63, 65 thae (they) (demonstrative pronoun) 131 Theakston, Anna L. 41, 42 them (demonstrative pronoun) 100, 131, 167 themuns (them ones) 134 there see existential there clauses these 131 they 9, 59 used as demonstrative see thae Thomason, Sarah G. 6 1 , 6 3 , 6 4 Thompson, Sandra 41, 57 those 131 thou 51,74,84, 155, 182-86, 189-90 times 134 Tomasello, Michael 41,42 topic-focus structure 138

Tristan da Cunha 51 Tristram, Hildegard 63 TRS see Tape-Recorded Survey of HibernoEnglish Speech Trudgill, Peter 67, 68, 75, 81, 145 Tyneside 84 Type-of-Subject Constraint 6, 7-12, 20-21, 45 Tyrone 158 see Chapter 5,passim Ulster 3, 13,51,52, 155 see also Northern Ireland Upper North (dialect area) 75,156 Urswick Cross (runic inscription) 61 Van Gelderen, Elly 14,17 Varbrul 5, 38, 173-78 clustering speakers 177 dependency (between factor groups) 175 factor contrast test 175 interaction (between factor groups) 176 see also Chapters 5-6, passim speaker clustering method 118-20 variable rules 37-38,42,67, 169 see also Varbrul Vennemann, Theo 64 Viereck, Wolfgang 68 Wakelin, Martyn F. 67 was/were in existential there clauses 139—41, 157-62 in NSR pattern 23, 105-6, 120, 124 with pronouns 52, 71, 74, 84-90, 117, 15557 weans (wee ones) 134 Weinreich, Uriel 37 Welsh 25,63-66 West Midlands (dialect area) 82 Westmorland 74 Wilson, James 140 Wolfram, Walt 51, 147 Word Grammar 29-30 Wright, Joseph 93,99, 104 Wright, Laura 51,68 Wright, Peter 70 Yang, Charles D. 30 York 90, 106, 161, 164 Yorkshire 10, 13, 52, 77, 99, 102, 105, 156, 166 youse 112 youseuns (youse ones) 134