Irish English Volume 2: The Republic of Ireland 9781614511298, 9781614511687

Table of contents :
Preface
Acknowledgements
Map of Ireland
Content
List of abbreviations
1 Geography, demography, and cultural factors
2 Phonetics and phonology
3 Morphosyntax
4 Lexis and discourse features
5 History, including changes in progress
6 Survey of previous work
7 Sample texts
References
Index

Citation preview

Jeffrey L. Kallen Irish English Volume 2: The Republic of Ireland

Dialects of English

Editors Joan C. Beal Karen P. Corrigan Bernd Kortmann

Volume 9

Jeffrey L. Kallen

Irish English Volume 2: The Republic of Ireland

ISBN 978-1-61451-168-7 e-ISBN 978-1-61451-129-8 ISSN 2164-7445 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de. © 2013 Walter de Gruyter, Inc., Boston/Berlin Typesetting: PTP-Berlin Protago-TEX-Production GmbH, Berlin Printing and binding: Hubert & Co. GmbH & Co. KG, Göttingen ♾ Printed on acid-free paper Printed in Germany www.degruyter.com

Preface Anyone who is interested in the English language in Ireland is faced with a range of approaches to the subject. From the perspective of the history of English, Irish English has a special position as the oldest overseas variety of the language. Its early establishment has in turn influenced English in North America, Australia, New Zealand, and elsewhere. In a regional perspective, though, Irish English can also be seen within the spectrum of local dialects and links across Scotland, England, and Wales. Some elements of Irish English clearly reflect historical influences from varieties of British English and Scots. In other cases it appears more likely that developments in Ireland have been transported to parts of England and Scotland. Sometimes it may be impossible to ascertain a direction of flow, and it may be more realistic to think in terms of shared areal features. The long history of contact between the Irish language and English in Ireland has given rise to a further perspective on Irish English, within the field of contact linguistics. Here the insights from studies of language creolisation, code-switching, linguistic borrowing, and other contact effects put a different emphasis on what we can learn from the study of Irish English. Turning from the formal aspects of language to the life of language in society, we also see that Irish English offers insights as a language which has developed over centuries of language conflict that point to wider conflicts, contacts, and accommodations in the social and political world. In saying that Irish English is not just a “regional” variety of English but also a “national English”, we immediately raise questions as to the relationship between Irish English and concepts of nation, state, and community. As an element of Irish culture, Irish English has also functioned as the medium of expression in a wide variety of literature, whether written by speakers of Irish English such as Oscar Wilde and Samuel Beckett who have fashioned their literature largely outside of Ireland, or by writers such as John Millington Synge and Roddy Doyle, who have chosen different forms of Irish English as a focus for their artistic portrayals of speech in the local community. Though the scope for exploring all these aspects of Irish English in one volume must be limited, the treatment which follows is designed to touch on each of them. In so doing, we may be able to develop a three-dimensional view of Irish English, comprising aspects of form, culture, and use. This book is designed to follow Irish English, Volume 1 – Northern Ireland by Karen P. Corrigan (Corrigan 2010). The decision to divide Irish English into two areas based on the political division between Northern Ireland and the Republic of Ireland was not made lightly. The strong presence of the Scots element in Ulster makes an obvious case for treating Ulster English, and its connections with Ulster Scots, as a separate topic from English in the rest of Ireland. O’Rahilly

vi   

   Preface

(1932: 17) declared that “historically there were but two main dialects in Irish, a Northern and a Southern”, and details (pp. 161–191) both the historic influence of Scottish Gaelic on Ulster Irish and a lesser reverse flow from Ulster Irish into Scottish Gaelic. Such a division in Irish dialects could also motivate a parallel division in the treatment of Irish English on purely linguistic grounds. When we start to examine the evidence in linguistic detail, though, the fixing of a linguistic border becomes elusive. There are, for example, many distinctive features of Irish English which can be found throughout the entire island: a tendency to use nonvelarised or “clear” /l/ in all word positions, dentalisation of /t/ and /d/ before /r/ or /ər/, use of the after perfect (as in My friend’s car is after breaking down), and the use of discourse expressions such as don’t be talking are but a few such cases. It is often true, too, that features which are strongly associated with Ulster English, such as the use of aye ‘yes’ or wee ‘little’, are nevertheless variably to be found in other parts of Ireland, depending on the social networks and experience of speakers. Using a simple North–South linguistic division would also risk downplaying variation that exists within each area. As Corrigan (2010) makes clear, there are many linguistic divisions even within the six counties of Northern Ireland; if we were to add the traditional Ulster counties of Donegal, Cavan, and Monaghan to this mix – or to go one step further and include Louth, which O’Rahilly (1932: 18) includes in the zone of Ulster Irish  – the increased diversity of our data would require a more complex treatment. It would also be misleading to suggest homogeneity for any area we might wish to label as the ‘South’: local varieties of English spoken in Cork are markedly different from those in Galway, Dublin, or the midlands. Given, in short, the permeability of any North–South linguistic border which we might wish to suggest as a linguistic demarcation, and given the complexity of possible variation on either side of an idealised dividing line, the question of a North-South border on purely linguistic grounds is, I suggest, best thought of as a research agenda rather than as an organising principle for a volume of this kind. The political division between the Republic of Ireland and Northern Ireland, on the other hand, admits of no such ambiguities. The social life of language in each of these two jurisdictions is visibly different. For the most part, road, traffic, and other public signage in the Republic is consistently bilingual in Irish and English, while the same kinds of signs in Northern Ireland are almost invariably monolingual in English. Nearly every school child in the Republic studies Irish, and most will study it until they finish second-level education; provision and take-up of Irish in Northern Ireland is considerably less. Legal instruments, the civil service, and a wide range of public functions (from parking meters to tax forms) in the Republic work within a bilingual framework that contrasts strongly

Preface   

   vii

with most comparable functions in Northern Ireland. These differences are not accidental, but stem from historical and political developments, many of which are discussed in Chapter 1. These differences also affect our understanding of Irish English in its social context. When we talk about the role of English in society, or about people’s everyday experiences with the Irish language in relation to English, we describe, for most people, very different linguistic environments on each side of the political border. It is, I suggest, possible to discuss linguistic life in the Republic in a coherent way which reflects the foundation of the state on principles that include linguistic nationalism and which explain much of what happens in education and the public domain; these principles, in turn influence private language use in particular ways. It is equally possible to discuss linguistic regimes and relationships in Northern Ireland, but here the variables are different. As long as the political border exists, then, it remains as a well-defined, fixed point of reference. Not at the level of local dialect and linguistic features, but at the more abstract level of language and society, this point of reference forms what I take to be the preferable way of capturing what we can be sure of with regard to the North– South division in contemporary Irish English. One practical consequence of this division of labour is that less will be said here about the counties of the Ulster dialect zone which lie within the Republic than might be desirable. In broad terms, the phonology of this area shares much with the Ulster English phonology of Northern Ireland. Henry (1958: 154–155), for example, shows that the raising and fronting of /u/ in Co. Antrim, which yields forms such as [hyk] ‘hook’, [ˈs̱kʲyˑpɪd] ‘stupid’, and [hyː] ‘how’, is also characteristic of Monaghan [s̱kyˑl] ‘school’, [hyk] ‘hook’, and [s̱yːẕ] ‘shoes’. These forms contrast in Henry’s account with [ruːts] ‘roots’ and [ɡuːs] ‘goose’ in Co. Leitrim in Connacht. Yet there is also evidence to differentiate these more outlying counties from other parts of the Ulster dialect zone. Again comparing the /u/ vowel, Adams ([1950] 1986: 99–100) contrasts the use of [ü] in “standard east Ulster speech” with the south Donegal vowel transcribed as [ʊ]. Adams describes the Donegal vowel as having “more lip-rounding and a more retracted tongue position” than [ü], but not being equal to the [ʊ] of “southern English”. Barry (1981b: 87–95) presents maps based on dialect evidence which could be used to set the southern boundary of the Ulster dialect zone, but these maps raise many more questions which require more detailed investigation, particularly with regard to details within the border counties. Detailed discussion of the Ulster counties within the Republic, therefore, runs the risk of either duplicating a certain amount of description which is equally true in other parts of Ulster (for which we refer naturally to Corrigan 2010), or requiring a more detailed local description that would bring this book beyond its size limitations. Faced with this practical choice, the practice in

viii   

   Preface

this volume will be to make use of information from all counties of the Republic, but in the interests of space limitations, there will be less discussion of Ulster English within the Republic than there might otherwise be. Despite the importance which I would attach to the role of Irish in conditioning the social life of English in the Republic, I would also stress that this book is designed as a book about English. Though it would be impossible to treat Irish English comprehensively without discussing Irish – whether in its historical role as an influence on Irish English or in its contemporary role as a source for codeswitching and bilingual effects in language use  – the focus in this volume is intentionally on English itself, rather than on an extensive discussion of the possible effects of Irish on the development of Irish English. While it is something of a tradition in the study of Irish English to engage in debate on the “substratum” effect of transfer from Irish or the “superstratum” effect of British English, and while we will at times touch on matters of language contact and transfer, I consider this debate to be a matter for another day. My approach to contextualising features of Irish English will also at times lean towards other varieties of English, seeing Irish English in the context of world Englishes. Therefore in this volume where the discussion focuses on English without discussing comparable facts in Irish, there is no necessary implication that Irish would be irrelevant to a more detailed treatment: I simply keep to the principle that this is a book about English. Irish English material for this book comes from four main sources: (1) the published record for the linguistic description of Irish English, sometimes augmented by unpublished original theses and reports; (2) literary representations of Irish English, which have been used sparingly to illustrate – rather than provide the primary evidence for – particular uses; (3) linguistic corpora, most notably the International Corpus of English, Ireland component (ICE-Ireland), described in Kallen and Kirk (2008) and below, and (4) my own notes, which include transcripts of recordings as well as notes of incidental conversations, radio broadcasts, and the like; some of these notes have been reliably reported to me by others. Most of these sources are cited in conventional citation form; data from the last source will contain the note (JK) together with the provenance of the speaker where known.

Acknowledgements No work of this kind can be completed without the help of the many speakers of Irish English  – family members, neighbours, students, colleagues, friends, speakers on radio and television, and chance conversationalists of many kinds – who have contributed raw data and linguistic intuitions on which much of this book is based. I consider myself fortunate to have learned about the diversity of Irish English from them, and hope that they will recognise Irish English as they know it in this book. Formal thanks are extended to the Irish Times for permission to reproduce ‘The brother says there’s a black market in turps’ by Myles na gCopaleen, which is found as sample text 5 in this volume. Sample text 3, from The Real Charlotte by Somerville and Ross, also appears by kind permission of the copyright holders. Academics work in a network of professional relationships, and it would be difficult to thank all the people with whom I have discussed the materials and approaches which are developed here. In particular, though, I would like to acknowledge the help I have received from discussing matters of this kind with Hal Schiffman, Máirtín Ó Murchú, Lesley Milroy, Jim Milroy, Cathal Ó hÁinle, John Harris, Markku Filppula, Michael Montgomery, Salikoko Mufwene, Frans Hinskens, Peter Auer, Sandra Clarke, Alicia Beckford Wassink, Manfred Görlach, Edgar Schneider, Sali Tagliamonte, Alex Bergs, Goodith White, Michael Quigley, James Killen, Mark Hennessy, Peter Harbison, and Pat Hartigan. Particular thanks go to John Kirk, with whom I have worked on the ICE-Ireland corpus since the 1990s. Special thanks also go to Karen P. Corrigan, not just for her longstanding support of various aspects of this research, but for the valuable comments which she and an anonymous reviewer provided on a draft of this manuscript. Similar thanks go to Ciaran Brady, who also offered valued comments on an earlier version, and to Charlie Travis, who created the map for this volume. I have also been helped by the kind and patient advice which I have received from Emily Farrell and Lara Wysong at Mouton de Grutyer. On a personal note, I would also like to thank my parents, Arthur and Vivian Kallen, for their support and regular enquiries as to the state of ‘the book’. Thanks, too, go to my family in Dublin – Peigí Whelan-Cunningham, Mary da Hora, Rosie Whelan, Bill Whelan, and Esther Ní Dhonnacha – who have acted over the years as a ready-made panel of informants; Esther has also been a rigorous and insightful editor and critic. My deepest thanks, though, go to Margaret Mannion Kallen, who has been my help in every regard, and without whom this book would not have been written.

IRELAND Coleraine

DERRY

DONEGAL

ANTRIM

Frosses

TYRONE Tartaraghan

DOWN

Bragan

FERMANAGH

Sligo

SLIGO

MONAGHAN Kilclare

LEITRIM

MAYO Charlestown Balla

ROSCOMMON

Dundalk

Cavan

CAVAN

Nobber

GALWAY Galway

LOUTH River Boyne

Drogheda

MEATH

Emper

Athlone

Lambey Island

WESTMEATH Fingal

OFFALY

Loughrea

Carlingford

Rathdrummin

Kells

LONGFORD

Roscommon Glenamaddy

Tuam

ARMAGH

KILDARE

Howth Dalkey

Dublin Killakee

WICKLOW

Aran Islands

CLARE Ennis

Feakle

Limerick

Kilcommon

TIPPERARY

Athea

LIMERICK

Cashel

Kilkenny

Melleray

KERRY

CORK

Glencar

Ballyvourney Midleton Cork City Ballymakeery Kinsale Rosscarbery

Lismore

Carlow

CARLOW

KILKENNY Glenpipe Waterford

Smerwick

Wicklow Baltinglass

LAOIS

WATERFORD

WEXFORD

New Ross Wexford

Forth & Bargy Baronies

Kilmore Quay

Province Ulster Leinster Connacht Munster

Content Preface  v Acknowledgements  ix Map of Ireland  x List of abbreviations  xiv 1 Geography, demography, and cultural factors  1 1.1 Introducing Irish English  1 1.2 First contacts  2 1.2.1 The first settlement of English  10 1.2.2 Linguistic relations in medieval Ireland  13 1.3 Irish English in transition  18 1.4 Language shift and linguistic realignment  30 1.5 English and the Republic  38 1.5.1 English, Irish, and the nation  38 1.5.2 New directions on language and the nation  42 2 Phonetics and phonology  46 2.1 Typological overview  47 2.2 Consonants  49 2.2.1 The labial group  49 2.2.2 The coronal group  50 2.2.3 Palatals and velars  57 2.3 Vowels  59 2.3.1 Vowels and lexical incidence  60 2.3.2 Variation in the front vowels: MEAT and CLEAR  63 2.3.3 The pen/pin merger: /ɛ/ raising  65 2.3.4 Mid vowel monophthongs  66 2.4 Syllables  67 2.4.1 Complex syllable codas  67 2.4.2 Further epenthesis  68 2.5 Word-level stress patterns  68 3 Morphosyntax  70 3.1 Information structure  71 3.2 Subordination and coordination  77 3.3 The verb phrase  85 3.3.1 Progressive verb forms  85 3.3.2 Generic-habitual forms  90

xii   

   Content

3.3.3 Indicative be  93 3.3.4 Perfect aspect  94 3.3.5 Verbal negation  106 3.3.6 Question formation  110 3.4 Words and morphemes  111 3.4.1 Subject-verb agreement  112 3.4.2 Verbal morphology  114 3.4.3 Pronouns  118 3.4.4 Nouns and noun phrases  122 4 Lexis and discourse features  126 4.1 Lexical perspectives  127 4.1.1 Irish English and etymology  128 4.2 Irish and the lexicon  132 4.2.1 Lexical items and Irish  141 4.2.2 The dual lexicon  148 4.3 English, Scots, and other sources  152 4.3.1 Irish English usage: the example of kill  170 4.4 Prepositions  171 4.5 The dialect of Forth and Bargy  176 4.6 Discourse perspectives  179 4.6.1 Lexical and phrasal elements  181 4.6.2 Communicative style  203 5 History, including changes in progress  212 5.1 Early Irish English  212 5.2 Modern Irish English  220 5.3 Debating sources: a look at the perfect  224 5.4 Contemporary change in progress  229 6 Survey of previous work  234 6.1 General works: chronological overview  234 6.2 Historical, geographical, and cultural factors  241 6.3 Phonetics and phonology  243 6.4 Morphology and syntax  244 6.5 Lexis and discourse features  248 6.6 Historical change in Irish English  251 6.7 Uses of Irish English in literature  252 6.8 Irish English abroad  253

Content   

7

Sample texts  256

References  272 Website references  303 Index  304

   xiii

List of abbreviations In keeping with common practice, many dictionaries and glossaries which are frequently referred to here are denoted by short abbreviations. The following table lists the abbreviations along with their corresponding references. Abbreviation

Reference

BDI CUD DAI DARE DCS DHE DHS DIL Dineen DNE DSL EDD EDG FGB JPG LQ MED NPD OED

Byrne (2004) Macafee (1996) Ó Muirithe (1996a) Cassidy and Hall (1985–2012) Beecher (1991) Dolan ([1998] 2004) Partridge ([1961] 1972) Quin ([1913–1976] 1983) Dineen ([1927] 1934) Story, Kirwin, and Widdowson ([1982] 1990) Dictionary of the Scots Language Wright (1898–1905) Traynor (1953) Ó Dónaill (1977) Barnes (1867) ‘Lexical questionnaire’ in Kallen (1996, 1997) Middle English Dictionary Dalzell and Victor (2006) Oxford English Dictionary

1 Geography, demography, and cultural factors In some counties in Ireland, many of the poorest labourers and cottagers do not understand English, they speak only Irish, as in Wales there are vast numbers who speak only Welsh; but amongst those who do speak English we find fewer vulgarisms, than amongst the same rank of persons in England. The English which they speak is chiefly such as has been traditional in their families from the early settlers in the island. During the reign of Elizabeth and the reign of Shakespeare, numbers of English migrated to Ireland; and whoever attends to the phrases of the lower Irish may, at this day, hear many of the phrases and expressions used by Shakespeare. Their vocabulary has been preserved nearly in it’s [sic] pristine purity since that time, because they have not had intercourse with those counties in England, which have made for themselves a jargon, unlike to any language under Heaven.  – Edgeworth and Edgeworth ([1802] 1803: 151–2).

1.1 I ntroducing Irish English Richard Lovell Edgeworth  – scientist, engineer, educationalist, and essayist  – and his daughter Maria  – best known for her novel Castle Rackrent and other literary works – thus described the language environment in Ireland at the start of the 19th century. At this time, Irish was the dominant language for a majority of the ordinary population, and the Edgeworths could still plausibly associate the everyday use of English among members of the “lower” classes with those whose ancestors came to Ireland in Elizabethan times and (allowing for the Edgeworths’ poetic licence) subsequent settlements in the first half of the 17th century. Before the close of the century, however, the antiquarian, clergyman, and explorer Abraham Hume would argue (1858: 51) that a national version of English in Ireland, itself “not much older” than the 19th century, had become what he later termed “the recognised language of the country” (Hume 1877–78: 103). As the different viewpoints of the Edgeworths and Hume suggest, the 19th century was pivotal in the development of Irish English, especially in that part of Ireland which now constitutes the Republic. Yet in order to understand the position of Irish English today, it will be necessary both to look backward at the time before the 12th century, when speakers of Middle English first settled in Ireland, and to look towards the future of linguistic relations within the Republic in an era of increasing European unity and linguistic globalisation. In this perspective, we can see Irish English in several different ways: as the oldest of the “overseas” varieties of English, as a complex of geographically- and socially- defined variation with links both to the dialects of British English and Scots and to the Irish language, and as a distinctive national variety of English.

2   

   Geography, demography, and cultural factors

One enduring constant in this history is the linguistic division of labour known as diglossia. Classic diglossia as defined by Ferguson ([1959] 2003) refers to the distribution of two distinct language varieties within a single language, each variety being associated with a particular domain of usage. Those varieties (such as Classical Arabic or Katharevousa in Greek) which are associated with notions such as prestige, power, and formality  – usually acquired with formal education – are referred to as H (“high”) varieties, and those (such as colloquial Arabic or Greek Dhimotiki) which are less formal, often less prestigious, and associated with first language acquisition are referred to as L (“low”) varieties in this analysis. (It should be noted that Greek diglossia, which was one of Ferguson’s original paradigm cases, has changed considerably with the development of Standard Modern Greek following the official ending of diglossia in the language reform of 1976: Frangoudaki 1992 provides a review.) Though Ferguson’s (2003: 345) definition of diglossia explicitly excluded “the analogous situation where two distinct (related or unrelated) languages are used side by side throughout a speech community”, other analysts have developed the idea into what Fasold (1984: 53) calls “extended diglossia”, in which it is possible to examine any combination of socially stratified language varieties (languages, dialects, or other linguistic codes) which are separated by their domains of usage. In addition to this notion of extended diglossia, we will also make use of a distinction which has been developed by Fishman ([1967] 2003), who argues (pp. 362–3) that “bilingualism is essentially a characterization of individual linguistic behavior whereas diglossia is a characterization of linguistic organization at the socio-cultural level”. This account leads to a four-way combination of factors, by which, for example, a state may be diglossic without bilingualism, if two languages are spoken within its borders but the individuals within the state are not for the most part bilingual, or bilingual and diglossic when individuals are bilingual but the codes which they command are separated by domain of usage. As we will see in this chapter, there are many periods in Irish history in which diglossia without bilingualism is the predominant mode. The situation in the Republic today, however, is one in which government policy for the most part aims to create a society in which Irish-English bilingualism is widespread, but without diglossic stratification or separation of function for the two languages.

1.2 First contacts Taking a conventional approach and discounting the period before the 12th century will have an effect on how we view Irish English. Not only can the introduction of English appear as an abrupt and isolated event, almost accidental in



First contacts   

   3

view of the low official status of English in 12th century England, but we will fail to answer questions about the historical and linguistic context for the establishment of English: what other languages had been spoken in Ireland, by whom, and for what purposes – in short, how did English fit into the existing linguistic order in Ireland? An overview of this early time brings into focus the crucial position of Ireland in the sea lanes of western Europe. This position has brought with it lines of communication, commerce, political allegiance, and population movement, all of which have been crucial in the development of Irish society and thereby provide a context for the development of Irish English. We thus start by looking at what we know of the cultural and linguistic regime with which English was to make such a long-lasting and intimate encounter. Thousands of years of human habitation pass in Ireland before we have any indication as to what languages were spoken. The earliest human artefacts come from mesolithic settlements which date to at least 9,000 years before the present time (BP). We do not know where these first settlers came from: Mitchell and Ryan (1997: 116) point out similarities between Irish artefacts from this period and others found in Denmark, while Duffy (2000b: 10) suggests that the first settlers travelled from Britain. The earliest evidence of agriculture dates to the neolithic period roughly 6,000 years BP. This period shows the clearance of forest, though the most enduring evidence of human activity is in the construction of megalithic tombs. Sheridan (1994: 50–1) points out a range of archaeological finds which suggest contact at this time with communities in Scotland, the Orkney islands, Britain, and the Iberian peninsula. Stout and Stout (1997: 35) also offer evidence of shared features of passage tomb art in western Iberia, western France, and scattered parts of Britain. Bronze age civilisation came later, approximately 4,500  years BP, at which time too we find a form of pottery whose most distinctive element is a particular type of beaker. Mitchell and Ryan (1997: 193–4) point out that this pottery is found from “Poland to Iberia”, reflecting trade relations with other parts of Europe and “a fashion adopted by the cosmopolitan rich and powerful everywhere”. Irish metalworking from this period includes distinctive decorative goldwork, in which Cahill (1994) sees the possible reflection of Mediterranean and Scandinavian influences (see also O’Kelly 2005: 128–129). This metalwork provides a basis for import and export: Mitchell and Ryan (1997: 236) note the substantial import of amber from Jutland or the Baltic area during this time. Further evidence of cultural exchange is seen in the distribution of halberds, implements which include a blade mounted on a wooden shaft and attached by rivets. O’Kelly (2005: 124) points out the widespread distribution of the halberd “from southern Italy to Scandinavia”, and notes that approximately 40% of known halberds have been found in Ireland. This distribution suggests that the implement may have origi-

4   

   Geography, demography, and cultural factors

nated in Ireland and spread elsewhere, but this point is still open to debate. What is relevant for us is the evidence which these and other archaeological patterns give us of Ireland’s involvement, whether by trade, population movement, or a combination of the two, with the European culture of the time. Coming closer to historical time, the Iron Age civilisation known to archaeologists as the La Tène culture, so named from the discovery of archaeological remains at the La Tène site in Switzerland, is generally taken to have arrived in Ireland no less than 2,300 BP. We have no direct evidence of the language spoken by those who brought this civilisation to Ireland, but the La Tène culture which developed on the European mainland roughly two centuries earlier was, according to Raftery (1994: 107) “emphatically Celtic” and corresponds to that of the Galli or Galatae described by Roman commentators. Raftery (2005: 145) points out, though, that while La Tène civilisation was widespread on the European mainland, La Tène material is only sparsely distributed in Ireland, suggesting that “we can scarcely speak of a great, sweeping change of population” at this time. The limited nature of this material suggests instead a period of social stratification, in which, as Raftery (2005: 161) points out, “a significant part” of the La Tène material “reflects the trappings of an aristocratic élite” and in which “large sections of the contemporary population are unrepresented” by this archaeological evidence. Due to the connection between La Tène civilisation and attested evidence for continental Celtic language use, the advent of La Tène cultural artefacts is often taken to represent a plausible date for the first arrival of a Celtic language in Ireland. Since we know nothing of the languages spoken by civilisations predating La Tène culture in Ireland, this supposition has always been conjectural. Recent controversy has been generated by the “Celticization from the West” hypothesis, which Cunliffe and Koch (2010: 1) explain to hinge on the proposition that “Celtic probably evolved in the Atlantic Zone during the Bronze Age”: the “Atlantic zone” includes “Armorica and the north and west of the Iberian peninsula”. If the hypothesis is borne out by the linguistic, archaeological, and genetic evidence which is being used to test it, the rise of a Celtic language in Ireland would date from an earlier time and a different source than the La Tène account would have it. It follows by implication that any speakers of a Celtic language who came to Ireland in the La Tène expansion would have found another Celtic language already established on the island. Because of the time gaps between the archaeological evidence and the first evidence we have of writing in Irish, we do not know if the earliest stage of the language we know as primitive Irish was brought directly from the European mainland, or if it was forged in Ireland from contact between Celtic-speaking immigrants and the people they encountered on coming to Ireland.



First contacts   

   5

What we do know, though, is that relations between early Irish-speaking society and Roman civilisation provide subsequent points of linguistic and cultural contact. Iron Age archaeological items from north Africa, such as the skull of a Barbary ape found at Navan fort, Co. Armagh (the ape perhaps being brought as a gift), and an Egyptian gold bracelet found in Co. Derry, cited by Warner (1994: 112–3), are indicative of early contacts with the Mediterranean world. A frequently-quoted passage from Tacitus, a Roman historian who wrote in the late 1st century, gives a view of Ireland in which, as Stout and Stout (1997: 43) cite, “the interior parts are little known, but through commercial intercourse and the merchants there is better knowledge of the harbours and approaches”. As Warner (1994: 115) and Duffy (2000b: 15) show, the map of Ireland which can be reconstructed from the description given by the 2nd century Alexandrian geographer Claudius Ptolemaius, better known as Ptolemy, provides a very recognisable outline of the physical geography and population of Ireland, the details of which sometimes correspond well with known historical evidence. Though Ireland was never part of the Roman empire, we have archaeological evidence of Roman presence, and of trade with Roman civilisation, through Romanised Britain and directly with the European mainland. Warner (1976: 274), for example, interprets the cremation burial site in Stoneyford, Co. Kilkenny, which dates to ca. 2,100 BP, as evidence of “a strong and secure Roman community”. The 1st century Roman burial site on Lambay island near Dublin is attributed by Raftery (2005: 175) to “north Britons whose material culture was strongly influenced by Rome”. Mitchell and Ryan (1997: 246–247) point out that this burial site includes “brooches of Roman-British style and a local imitation of them”, and that on the nearby mainland at Drumanagh, Co. Dublin, artefacts and copper ingots of a Roman type suggest the manufacture of “high status objects” indicating the position of Drumanagh as “an important entrepôt” between Ireland and Roman Britain. Evidence of changes in fundamental features of Irish daily life are also suggested in Mitchell and Ryan’s view (1997: 248–9) that new methods of ploughing and dairy culture which came about as a result of Roman influence had a major impact on the development of Irish society around this time. Turning to language matters, it is from the 4th century that we find inscriptions in Irish using the ogam writing system. Writing in ogam is based on a system of lines and notches made along a central axis, and though it therefore does not visually resemble the Roman alphabet, its means of relating sound to writing follows what McManus (1991: 27) describes as “the classification of the letters of the Latin alphabet found in the works of the Latin grammarians”, modified to reflect distinctive features of Irish. Ogam inscriptions have been found in Ireland especially in Munster, and in Wales, Cornwall, and scattered parts of Scotland: for maps, see Ó Murchú (1985: 13), McManus (1991: 46, 48), Stout and Stout (1997:

6   

   Geography, demography, and cultural factors

44), and Edwards and Hourican (2005: 126–7). Commentators such as Jackson (1953), McManus (1991), and Russell (2005) note that it is not certain if ogam originated in Ireland and was brought to Wales, or if it originated in the Irish colonies of what Russell (2005: 415) calls the “Latinate milieu” of Wales in the late 4th or early 5th century. This indeterminacy is itself further indication of cultural links and networks of communication that encompassed Ireland and Britain in the late Roman period. The adoption of Christianity in the 5th century drew Ireland into new political and social relationships with other parts of Europe. A key date in this historical phase is the year 431, which saw the mission of Palladius “to the Irish believing in Christ”. This description of the mission presupposes an existing Christian community, well before the coming of St. Patrick later in the 5th century (see Ó Cróinín 1995: 14–23 for a review). The development of these relationships gave rise to a further role for Latin in Irish society. Picard (2003b: 47–55) gives evidence that the Irish attachment to Latin continued long after the collapse of the Roman empire in Britain, noting (p. 47) Bede’s comment in the History of the English People from the 660s that “many English youths, both from the nobility and from lesser classes, travelled to Ireland to get instruction”, and citing (p. 49) the early use in Irish education of the 7th-century Etymologies of Isidore of Seville as evidence that “Ireland was in touch with the rest of the Roman world and kept up with the latest scholarship”. Irish adoption of Christianity also brought about important changes in settlement patterns, with the growth of monastic or other ecclesiastical sites: Stout and Stout (1997: 512) document the wide range of establishments, while Duffy (2000b: 20–21) links the influence of “secularization, wealth and lay-patronage” during this time to the development of illuminated manuscripts and the arts of metal and stonework. Literary developments in Irish at this time point towards the elaboration of a system of diglossia. The number of texts in Old Irish is limited, and the major sources for Classical Old Irish, dating from the 8th and 9th centuries, consist of short passages and glosses on Latin religious and grammatical works: Thurneysen (1946: 4–11) gives a clear account. By the end of the 9th century, Classical Old Irish was developing into what is now referred to as Middle Irish, but it is frequently noted, as reviewed by Russell (2005: 412–414), that Middle Irish texts contain archaic language which suggests a composition date in the Old Irish period. As shown in Russell’s (2005: 440–450) review of variation in Old and Middle Irish texts, writers developed what Russell (2005: 443) terms “an elite register spoken by the nobility as well as poets, churchmen, and judges, and presumably also by those who aspired to high status”. We can also associate this period with the development of extended diglossia, by which Latin



First contacts   

   7

became a H domain language used for religion and formal education. As we will see in Chapter 4, the lexicon of Irish by this time started to adopt a wide range of Latin terms, some of which have in turn become elements of Irish English. The production of important Latin manuscripts in Ireland appears to have reached a high point in the 9th century, before coming to what Ó Cróinín (2005a: 404) describes as “a sad end” in the 12th: for details see also Lapidge and Sharpe (1985). Ireland’s connections with Scandinavia took a new turn at the close of the 8th century. The first known Viking raid took place in 795, on Lambay Island, a monastic site in Dublin Bay. Though reviews such as those of Ó Cróinín (1995: 260–5), Clarke, Ní Mhaonaigh, and Ó Floinn (1998), Byrne (2005), and Valante (2008) show considerable controversy as to the precise impact of the Viking presence on Irish society, some elements must be taken into account in the linguistic history. Particularly important is the Viking establishment of urban settlements in coastal areas: 9th century foundations in Dublin, Wicklow, Wexford, Waterford, Cork, and Limerick are all directly related to Norse settlements. Scandinavian links between England and Ireland also suggest the possibility of a linguistic conduit in the region. We may note, for example, archaeological evidence of the similarity between the 10th century layout of streets in Dublin, Waterford, Wexford, and Limerick and towns in Mercia and Wessex, which Bradley (1988b: 70) interprets as an indication that “the Scandinavians successfully transferred to Ireland the concept of the early tenth century town as seen in the south and west of England”. The settlements themselves did not bring about a very large change in the population, but they did form centres of wealth, power, and trade relations. Hiberno-Norse cities such as Dublin were involved in lively trade relations with the northwest of England, Northumbria, York, Chester, Bristol, ports of France, and further afield. Ó Corráin (1972: 107) notes that “English pennies circulated fairly widely in Ireland” during most of the 10th century; following the establishment of an Irish mint in 997, according to Wallace (2005: 838), Irish coins from this period have been found “in hoards as far away as Iceland, the Faeroes, Norway, Denmark, Sweden, Finland, Russia, the Baltic States, northern Germany, and even Rome”. Artefacts with English connections which have been found in Dublin from this period include not only hoards of Anglo-Saxon coins and a variety of personal objects, but a knife sheath with an Anglo-Saxon inscription and what Wallace (2008: 174), citing work by Elisabeth Okasha and John Bradley, describes as a “leather scrap with the first letters of the alphabet in Anglo-Saxon script”. Relations of this kind are bound to have an impact on language. We have little direct evidence of linguistic practice in the Viking cities of Ireland, though Johnson (2004: 84–85), who notes runic inscriptions on a variety of personal

8   

   Geography, demography, and cultural factors

objects that have been found in Dublin, states that the languages of Dublin in the Viking era included “Norse, Irish, Anglo-Saxon, church Latin and perhaps some French (through trading contacts with Rouen)”. We do know that Old Norse had a linguistic influence on Irish, notably in the addition of vocabulary related to trade and seafaring. We may cite Irish words such as margad (modern Irish margadh) ‘market, marketplace’ (derived by Greene 1976: 79 from Old Norse markaðr and described by DIL as a “late loan word” with sources in Old Norse and Latin mercatus); marg ‘a march, boundary’; fuindeóc (modern Irish fuinneog) ‘window’, derived by DIL from Old Norse vindauga; langa ‘ling’ (the fish Molva molva); trosc ‘cod’ (fish such as Gadus morhua), from Old Norse þorskr; pónair ‘beans’, for which Greene (1976: 79) cites Old Norse plural baunir; and bát (modern Irish bád) ‘boat’, for which DIL favours an etymology from Old Norse bátr. A number of place names in Ireland are also derived from Norse. Among the best known are Lambay itself (‘lamb island’), for which Ó  Muirithe (2010: xxv) cites a derivation from lamb-øy or lamba- øy; Wicklow (given by Room 1994: 127 as ‘vikings’ meadow’, from vikingr ‘viking’ and ló ‘meadow’); and Smerwick in Co. Kerry, derived from Old Norse Smjör-uík ‘butter bay’; see further Byrne’s (2005: 630–634) review of Oftedal’s (1976) early suggestions. In general, place names from Old Norse have entered English directly rather than via an intermediate incorporation into Irish. Oftedal (1976: 127–129), however, also points out that a further group of names (including the provinces of Ulster, Leinster, and Munster) are intermediate in mixing both Irish and Norse elements. The name Ulster, for example, combines the Irish tribal name Uladh with the Old Norse genitive -s and what Oftedal (p.  129) terms “appellative tír borrowed from the Irish tír ‘land’.” Oftedal also argued that it is Old Norse Ulaðstir which entered English in the form Ulster. How place names of Norse origin entered English independently of Irish is a matter of some concern for Irish English. Greene (1976: 81) deduced that “Norse must have survived as a spoken language in Dublin and some other settlements up to the time of the English invasion of 1169”, but Byrne (2005: 631) argues that many of these names entered the English language, “through trading contacts a century or more before the Anglo-Normans arrived in Ireland”. The late survival of Old Norse in Ireland and its use in contact with Britain are not mutually exclusive, but if Byrne’s account is correct, it implies a high enough degree of contact between speakers of English in Britain and speakers of Old Norse in the Scandinavian kingdoms of Ireland for the former group to incorporate Irish place names into their version of English. At the same time, Ó Corráin (2009: 69) argues that Irish elites of the 11th and 12th centuries also spoke Old Norse, and views literacy in Old Norse as “well established”. Language contact is implied in what is probably a late 10th century Irish text which lists, among “useless or



First contacts   

   9

pointless things”, the gíc-goc Gallgaidhel or ‘gic-goc of the Hiberno-Norse’ and the gib gab na gcennaigh, which Ó Corráin (2009: 65) translates as “the cant of the hucksters”. Ó Corráin’s discussion of work by Carl Marstrander suggests that, in these Irish texts, gíc-goc is a Norse phrase, while gib gab is derived from Anglo-Saxon. Citing literary evidence, Ó Corráin (2009: 72) pushes the date for the continued use of Old Norse “in what were coherent Hiberno-Norse communities” as far as the middle of the 13th century. How the details of contact involving Irish, Old Norse, and English would have worked at this time remains a subject for further research. The defeat of Viking forces at a battle in Clontarf, near Dublin, in 1014 weakened the Scandinavian position, though it did not bring about an end to Scandinavian influence. Historical identities established in the Viking era continued well past the time of Viking political power: even as late as 1263, according to Edwards and Hourican (2005: 38), “a number of Irish chiefs offered the crown of Ireland to King Haakan IV of Norway” as an unfulfilled part of the resistance to Anglo-Norman rule. Curtis (1919: 234) notes the case of Maurice MacOter in 1289 who petitioned that he and his clansmen be treated as English, rather than Irish, on the basis of their Scandinavian descent. Legal proceedings from 1295 (see Mills 1905: 59) show a dispute between William le Teynturer against Henry and John le Norreys. The latter parties complained that they should not have had to answer to le Teynturer since he was said to be hibernicus [Irish] and “of servile condition”. The reply from le Teynturer was that he was in fact Houstmannus (an ‘Ostman’, as the descendants of the Vikings were then known), having received, at the instigation of his mother, “the liberty of the Ostmen” in Limerick. The 11th century also gives many examples of what Richter (1985: 329) refers to as “horizontal loyalties” with the rest of Europe held by “small though significant groups” in Irish society. Among these we may consider the refuge taken in Wales by the Scandinavian king of Dublin, King Sitric, following his banishment from Dublin in 1036 (see Duffy 1997: 38), and the consecration of Irish bishops in England. The consecration in 1096 by the Archbishop of Canterbury of Máel Ísu Ua hAinmere as the first bishop of the Hiberno-Norse city of Waterford (see Flanagan 1989: 20) was particularly important; Duggan (2007: 121) notes that six Irish bishops were consecrated at Canterbury between 1074 and 1140. We may also note the activities of Harold and Leofwine, sons of Godwin (earl of Essex), who came via Bristol to Ireland in order to raise a fleet which subsequently attacked coastal areas in England in 1052. Harold’s own sons came to Ireland for help following his death during the Norman conquest of England in 1066 (see Richter 1985: 336 and Flanagan 1989: 59)

10   

   Geography, demography, and cultural factors

1.2.1 The first settlement of English Viewing Ireland in its position in the seaways of western Europe, we can understand the arrival in 1171 of the English king Henry II as part of a continuing story of contacts and allegiances. Richter (1985: 328) argued that “the coming of the English to Ireland from 1169 onwards was less of turning point than it is generally regarded”, while Martin (1987: 44) discusses the term “invasion” to describe this event by concluding that “there was no intrusion or intervention on such a scale, or of such a nature, as to merit it being described as ‘invasion’. Nor was there a conquest. That was not achieved by England until 1603”. Nevertheless, since it is this episode that provides the basis on which the English language developed in Ireland, we will consider it here in some detail. Both ecclesiastical and civil arguments could be advanced in accounting for the motivations underlying Henry’s interventions in Ireland. The Church in England had shown an interest in control over the Irish Church at least since the time of Lanfranc, who was appointed as the Archbishop of Canterbury in 1070. A proposal for the invasion of Ireland was discussed at the Council of Winchester in 1155, which was attended by important Church figures including the Archbishops of Canterbury and York; the proposal may have been fed, among other things, by opposition to the synod held in Kells in 1152, which had established a diocesan structure for the Irish church (discussed by Flanagan 1989: 38–42). The account of the Anglo-Norman settlement in Ireland written in the 12th century by Giraldus Cambrensis, known as Expugnatio Hibernica (see Scott and Martin 1978 for a recent edition), includes the text of a document of uncertain origin, now generally referred to as Laudabiliter (from the first words of the text, meaning ‘it is praiseworthy’). This document purports to give Papal authority for the intervention of Henry II in Ireland. Since there is no independent text of Laudabiliter that can be used to verify the Giraldus version, its status as genuine document or as a forgery has been a matter of historical debate. Detailed analysis by Duggan (2004: 140) concludes that the Giraldus text is best described as “an amalgam of accurate reporting and tendentious manipulation of the truth”. The main significance of this document, however, lies not in its a prior justification for action, but in its later use, as Duggan (2007: 156–8) explains, to “demonstrate the prophetic nature of Norman dominion both in Ireland and in Wales”. While ecclesiastical motivations for the decision of Henry II to intervene in Ireland may be unclear, there can be no doubt as to the importance of political and military factors. The turning point in a series of historical events is the request which the Irish king Diarmait Mac Murchada made to Henry II for assistance in regaining his position following a military defeat at the hands of Ruaidrí Ua Conchobair in 1166. The Scandinavian kings played a role, for it was, as Duffy



First contacts   

   11

(1997: 58) points out, the Ostmen in Dublin whose recognition of Ua Conchobair’s supremacy is mentioned in the Irish Annals of the Four Masters. Mac Murchada’s first point of contact in looking for help was Bristol, where, according to Martin (1987: 63–4), Diarmait had a “trusted friend” in Robert fitz Harding, the provost of Bristol. Mac Murchada and fitz Harding shared a common interest, since they had supported the claims of the Empress Matilda and her son Henry in opposition to the position of king Stephen in the battle over the succession to the English throne following the death of Henry I in 1135. Such entanglements help to understand what Hays and Jones (1990) refer to as the “Irish sea diplomacy” which Henry II developed to secure his interests in Wales and Scotland and bring them into line with his interests in France. In suggesting that Henry may have self-servingly viewed Diarmait’s request for troops as “an opportunity to siphon the discontented and bellicose Anglo-Norman knights away from Wales”, Hayes and Jones (1990: 297) bring to the fore the inter-connections of the various conflicts in Britain and the continent with which Ireland had become involved during Henry’s reign. Curtis (1919: 235) described the force of 300–400 soldiers allied to Dermot Mac Murchada who came in 1169 to Wexford as “a very motley crew”. Though this frequently-cited phrase accurately reflects the recruitment of soldiers whose backgrounds lay in various parts of Britain, France, and further afield, it would be misleading to take this description as a sign that the force was casual or poorlyorganised. In fact, the arrival of the forces put together by Mac Murchada took place over a period of nearly three years. Mac Murchada returned in 1167 with a force which Duffy (1997: 62) describes as a “small band of people of Flemish origin who had settled in Pembrokeshire”. An initial military force came in 1169, and later reinforcements arrived in 1170. This latter group included Richard fitz Gilbert de Claire, who had supported Stephen in opposition to Matilda and Henry, and who is better-known in later accounts as Strongbow. The subsequent power and potential independence of colonial forces in Ireland posed a problem for king Henry, who was motivated to come to Ireland in 1171, as Duffy (1997: 69) puts it, “more to bring the pioneers there back into line than to conquer the Irish”. The expedition of Henry II to Ireland in 1171 demonstrated a combination of military strength and relationship-building. A contemporary report describes an entourage of 400 ships, and Lydon ([1972] 2003: 42) estimates a force of 500 knights and 4,000 other soldiers. Though there were some military encounters along the way to Dublin, the relatively bloodless nature of the campaign helped Henry in a symbolic effort to “make an impact on the Irish”, as Lydon (2003: 45) notes, by shows of courtly magnificence which included “entertaining the Irish to a lavish feast at Christmas in the course of which they were introduced to many strange and exotic dishes”. Gilbert (1865: 27) records that Henry II “appears”

12   

   Geography, demography, and cultural factors

not to have known English, but to have used “French interpreters” to communicate with the English speakers below him. We see evidence of an awareness of language matters, however, since interpreters were used when dealing with Irish speakers during Henry’s visit, in order to ensure, according to Lydon (2003: 44–5fn.), that Irish people declaring fealty to Henry would fully understand “the resulting new relationship” with the English king. The outcome of these military and other activities was that a great number of the Irish elite, with some notable exceptions in Ulster, had sworn loyalty to Henry as lord of Ireland; Church leaders also expressed their loyalty to Henry following a council held in 1172. The colonisation which followed had a profound effect on the social geography of Ireland. To illustrate the local effects of such settlements, we may cite Otway-Ruthven’s (1965) case studies of a number of early 14th century manors. Information is available from 1304, for example, about the population of the manor of Cloncurry, Co. Kildare. Distinguishing among the various classes of farmers and tenants, Otway-Ruthven (1965: 80) concludes that this manor included 191 people of English descent (made up of 112 burgesses and 79 other tenants) and 111 of native Irish descent, “almost all betaghs [low-ranking tenants] and cottagers”. Similar results are found in other areas, though in some cases, such as Moycarkey in Co. Tipperary, the balance was much more heavily tilted towards the English: here Otway-Ruthven (1965: 81) estimates the presence of 39 English tenants as opposed to just 9 of Irish descent. Summarising this period, Otway-Ruthven (1980: 109) argued that “the Norman settlement of Ireland was no mere military occupation supported by the settlement of English and French burgesses in a few towns, but a part of that great movement of peasant colonization which dominates so much of the economic history of Europe from the eleventh to the fourteenth century”. Details of such settlements lie outside the scope of this review, but see, among others, Glassock (1987) for a general review, and Simms (1986), Mitchell and Ryan (1997: 303–315), and Stout and Stout (1997: 53–60) for geographical perspectives and maps. It is not surprising that many of the initial settlers came from south Wales and the southwest and southwest midlands of England. Otway-Ruthven (1965: 78) notes that Strongbow “seems to have enlisted men from all the lordships along the coast of south Wales”, but documents additional settler names from Devon, Cornwall, Exeter, and as far as Lancashire. Duffy (1997: 59) describes the settlement as one which “included people whose backgrounds lay scattered throughout Britain, northern France, and the Low Countries”. The mixed background of what we can describe, following Mufwene’s (2001) development of the term, as the “founder population” for Irish Middle English naturally paves the way for dialect contact and mixing, decreasing the chances that the fledgling Irish English would reflect any specific dialect of British English. Despite this diversity



First contacts   

   13

in what we might see as linguistic or ethnic terms, however, it should be borne in mind that, in political terms, this settlement was English: as Lydon (2003: [7]) bluntly explains, “the people who settled in Ireland after the invasion were almost entirely English, and it is well to recognize that fact and not disguise them as Norman, Anglo-Norman or anything else”.

1.2.2 L inguistic relations in medieval Ireland Two diglossic societies were brought together in 12th-century Ireland. In the native society, Christianity had already established the use of Latin as a H language, and Irish had developed a H language variety for religion, law, learning, and literature by the time of the late Middle Irish and Classical or Early Modern Irish period (the latter dating roughly from 1200). Linguistic practice among the new colonists, however, is a subject of some speculation. In England, the Norman conquest in 1066 had established French as a H domain language alongside Latin. This role for French did not, however, imply that English was in danger of being supplanted by French as the language of the general population. It is well documented that the Norman conquest was, as Kibbee (1991: 9) describes it, “not a national migration, but rather a military conquest”: Kibbee’s review of population figures suggests that the Norman element following the conquest was “roughly 1.3%” of the population. Studies such as those by Short (1980) show that even among the elite in English society, the use of French as a first language was declining by the 12th century: by the time the future King Henry II had married Eleanor of Aquitaine in 1152, according to Kibbee (1991: 14), “the descendants of the first Norman French invasion were becoming totally assimilated”. We can thus think of the colonists who came to Ireland in terms of a social pyramid in which the top members may have spoken French or French and English as their native languages, but reserved Latin and French for H language domains, especially those in which writing was required: discussion and texts pertaining to French in medieval Ireland are provided by Bliss and Long (1987), Shields (1975– 76), and Picard (2003b). Successively lower members of society were more likely to speak English. It is possible that Welsh or Flemish could have been spoken as L languages among those coming to Ireland on the early expeditions, but we have no direct evidence to this effect. The new settlers were too few in number and lacking in resources to provide the basis for a self-sufficient economy. Two societies, native Irish (or Gaelic) and English (often now referred to – if problematically, as we have noted above – as Anglo-Norman, and referred to in later periods as Old English or Anglo-Irish) entered into a relationship in which each had reason to maintain its own iden-

14   

   Geography, demography, and cultural factors

tity and community boundaries. Between these societies, however, social, economic, and political ties were also necessary. We get one tantalising glimpse of contact between these two societies in the Norman French poetic work depicting the coming of the English settlement, long known as the “Song of Dermot and the Earl” (Orpen 1892), but now re-edited by Mulally (2002) as the “Deeds of the Normans in Ireland”. The early part of the text includes an account of its anonymous Norman author speaking directly with Diarmait Mac Murchada’s personal interpreter, Maurice Ua Riacáin, whose name is given in French as Morice Regan. As Mulally (1988: 328) points out, this poem is one in which “the English language does not impinge at all on our poet, though ‘the English’ are its heroes”. The picture of a Norman French speaking chronicler speaking to the interpreter for the Irish king Diarmait Mac Murchada suggests the kind of elite bilingualism that may have been characteristic of this period: French, Latin, and Irish commanded prestige in elite domains, with Irish and English at L domain level for the general Gaelic and Anglo-Norman populations respectively. These two societies did not, however, exist in a state of balanced parallel diglossia. Relationships were unstable, due to continuing conflicts over political and economic supremacy, and unequal, in so far as English in time became associated with colonial power while Irish became associated with a social definition of nationhood and tradition that was to take different forms over the succeeding eight centuries. We can chart the rise of English in Anglo-Irish society following the earliest attestations in the records of the Dublin Guild Merchant Roll (Connolly and Martin 1992), which date predominantly from the first half of the 13th century. The guild roll is written in Latin and is largely a list of names and payments, so there is little opportunity for the development of any continuous English text. At times, though, the scribe uses English of, the, or the name of a trade, so that Connolly and Martin’s (1992) edition yields references such as Torkaill of Kardif (p. 7), Gregory the Kene (p. 15), and Reginaldus the Letherkervere (p. 52). Longer texts are found in the Great Parchment Book of the Corporation of Waterford from 1365–7 onwards (Gilbert 1885a). By the late 15th century, English had become the main language in the Statute book of Galway (Gilbert 1885b), and had started to supplant French in the statutes of the Irish parliament (Morrissy 1939). This period can be seen as one in which, as in England, the English language was supplanting French as a H language. In Ireland, however, the rise of prestige for English was accompanied by a decline in the Anglo-Irish society which supported it. By the second half of the 13th century, English interests in Ireland had suffered from poor internal management, military conflicts with Gaelic society, and the neglect of Anglo-Irish economic interests in order to satisfy English military and political ends in Britain and France. Hand (1972) details ways in which the Irish version of English law



First contacts   

   15

had started to diverge from English practice since the late 13th century, showing fragmentation between Irish and English legal systems as evidence that (p. 403) “the medieval conquest was not a process of displacement so much as of superimposition” of the English system. An early 14th-century complaint from the reign of Edward III provides insight into the way in which these separate societies operated at the time: “marcher law” has left behind little documented evidence, but refers to clear accommodations between the two systems. The text of the complaint, given by Mac Niocaill (1964: II, 336fn), is in French; it is followed by the translation given by Hand (1972: 413). A de primes, pous le conquest ensa deuz maners dez gens ad este e est en Irlaunde, c’est a ssavoir Engleys e Yrois, entre queux trois maners des loys ad est[e] usee dount chescoun est contrariaunt a autre c’est a ssavoir commune loy, ley yrois e le de marche; e par la ou diversete de ley est, semble a nous qe les gens ne pount estre d’une ley ne de une commune. Since the conquest there have been two kinds of people in Ireland and there still are – the English and the Irish – and amongst them three kinds of law had been used, each of which conflicts with the others – common law, Irish law and marcher law; and it seems to us that where there is diversity of law the people cannot be of one law or one community.

Concern at the possible loss of the English colony can be seen, for example, in a representation made in 1341 to Edward III, in which it is declared that one third of the original settlement territory “is now come into the hand of your Irish enemies and your English lieges are so impoverished that they can hardly live” (Watt 1987: 367). The plague epidemic which swept Europe between 1348 and 1350 came to Ireland in 1348 with devastating effect and persisted intermittently for years afterwards. The effect of the plague hit the Anglo-Irish community and religious enclosures hardest, due to their greater population densities, and seriously damaged the economy. A council in Kilkenny thus made representations to Edward in 1360, in which it is argued that the colony had been neglected in various ways and was “en poynt d’estre perdu” [‘on the point of being lost’] (Richardson and Sayles 1947: 19): their fears may have been realistic. This combination of a new acceptance for English in H domains, together with the real possibility that people of English descent whose interests now lay primarily in Ireland might assimilate to the Gaelic majority in political or military allegiance as well as in language and other social practices, gave rise to a discourse over language in which English and Irish were seen in conflict as part of a wider struggle of national loyalties. A parliamentary tradition of complaint about the assimilation of English colonists to Gaelic society starts with a statute from 1297, written in Latin. This statute does not mention language specifically, but it does reflect the outlook of what Lydon (1987: 273) calls “a government unable

16   

   Geography, demography, and cultural factors

to cope with an increasing burden of defence and peace-keeping”, declaring, as translated by Berry (1907: 211), that Englishmen also as degenerate in modern times, attire themselves in Irish garments and having their heads shaven, grow and extend the hairs from the back of the head and call them Culan, conforming themselves to the Irish as well in garb as in countenance, whereby it frequently happens that some Englishmen reported as Irishmen are slain.

Note here that use of the Irish word culan is a linguistic reflection of the cultural assimilation which the statute aims to suppress. Variations on the word degenerate are common in this tradition of complaint, and we should interpret them in the light of Duffy’s observation (1997: 142) that “gens is the Latin for a people, a nation; to become ‘degenerate’ is to lose a sense of belonging to that nation”. Various such complaints, in Latin or French, continue through the 14th century: see, for example, Berry (1907: 211, 412, 417–8), Gilbert (1885a: 292), and Crowley (2005: 4–5). Statutes enacted at a parliament in Kilkenny in 1366, written in French, point to two elements of concern for the colony. In addition to measures directed against assimilation in various cultural practices, a statute directs that “every Englishman use the English language, and be named by an English name” (Berry 1907: 435). Here the opposition is between English and Irish, but a further statute (Berry 1907: 437) is aimed at enhancing cohesion within AngloIrish society, directing “that no difference of allegiance henceforth be made between the English born in Ireland, and the English born in England, by calling them English hobbe, or Irish dog, but that all be called by one name, the English lieges of our lord the King”. A similar statute from Waterford in 1384–85, written in English (Gilbert 1885a: 292), directs that a fine be payable “if ony man dwelling within the lyberte of the same citie shal curse, diffame, or dispice ony citsayn of the saide citie in calling him Yrishman”. Legislation in the 15th century continues the tradition of the 14th in trying to maintain community boundaries, as in the Dublin statute from 1457–58 which prohibited the lodging of Irish men or “men with bardys above the mouth” (Gilbert 1889: 280–81) or the 1466 regulation which required the use of Latin (which we may take to be an intended defence against Irish) in written communication with Irish merchants and those who “understandith not Englys” (Gilbert 1889: 323). An ordinance from Waterford adopted in 1492–3 (Gilbert 1885a: 323) shows both an exhortation for those associated with the English colony to use English in court proceedings and a recognition that members of native Irish society (those “of the countre”) would naturally speak Irish, declaring that



First contacts   

   17

no manere man […] of the citie or suburbes duellers [dwellers], shall enpleade nor defende in Yrish tong ayenste ony man in the court, but that all they that ony maters shall have in courte […] shall have a man that can spek English to declare his matier, excepte one party be of the countre; then every such dueller shalbe att liberte to speke Yrish.

Some indication of the difficulties in trying to enforce the legislative approach to efforts to maintain English may be seen from the acts of the Parliament of 1495, which reaffirmed the Statutes of Kilkenny but specifically excluded those statutes which pertained to “the language of Irish” (Statutes 1786: 77). Broadly speaking, we can see 15th century Ireland as including three major elements: native Irish society, older English colonists who had now become more truly Anglo-Irish and whose loyalties to England could be tempered or set aside by local alliances with Irish society, and a small group who formed a loyal colonial presence. The attempt to fix social divisions geographically can be seen in the notion of the “English pale”, a term which was used to denote the territory in which colonial control was felt to be relatively secure. The origins of the Irish historical sense of pale, derived ultimately from Latin pālus ‘a stake’ (OED), are unclear. Lydon (2003: 200) points out a similar meaning in earlier English usage, as in the authorisation given in 1378 to the town of New Sarum (in the area of modern Salisbury) to construct “un grant Pale”, i.e., a “trench […] and wooden fence” around it. Murphy and Potterton (2010: 265) conclude that the immediate sense in Irish usage derives from the “Pale of Calais”, which was designed to fortify the residual English stronghold at Calais. Grummitt (2008: 5) clarifies that the term “Pale” had been used since 1436 to denote the land at Calais held by the English king, though it was not in common usage until the 1490s. The connection between Calais and Ireland is direct in the person of Sir Edward Poynings, who had held high office in the administration at Calais in the early 1490s and took up the position of lord deputy in Ireland in 1494. Thus when one of the first acts of the parliament convened by Poynings in Drogheda in 1494–5 called, as quoted by Ellis (2007: 447), for “diches to be made aboute the Inglishe pale”, we find not only a linguistic development in a colonial context, but the beginning of a culturally indexical notion which encapsulates in a single phrase the delimitation of an area where, in theory, English law and culture prevailed over native Irish society. Though the ditch referred to by the act of parliament relied on private undertakings and was never fully constructed, many fortifications (including towers and castles as well as banks and ditches) played a role in delimiting core areas of the Pale in counties Dublin, Meath, Louth, and Kildare. It is a mistake to think of the Pale as simply an area surrounding Dublin, since, as Murphy and Potterton (2010: 264) point out, some of the market towns within the Pale owed their very

18   

   Geography, demography, and cultural factors

strength to their distance from Dublin and from each other. Maps given by Murphy and Potterton (2010: 266–275) connect the political design of the Pale with historical and archaeological evidence: further maps are given by Mitchell and Ryan (1997: 319) and Edwards and Hourican (2005: 80–81). Rather than thinking of the delimitation of the Pale in absolute terms, though, it may be best to see it in the terms given by Lyons (2003: 853), who describes it as “a shifting medieval frontier whose borders coincided with geographical, cultural, administrative, political, and military boundaries between English and Irish regions in Leinster”.

1.3 Irish English in transition The reduction and geographical retraction of the Anglo-Irish colony, as well as the trend towards Gaelicisation which the medieval parliaments addressed, have led to a view in which Irish English is seen as a dead language at some point before 1600. Bliss (1977a: 26), for example, has stated that “by the end of the fifteenth century little English was heard except in the Pale […] and in the towns”, and that “by the middle of the sixteenth century spoken English was in a state of almost total eclipse” (Bliss [1976] 1978: 546). Bliss (1977a: 26) also uses the early 17th century evidence of Fynes Moryson, which is discussed here below, to conclude that by Moryson’s time “through most of the country Mediaeval HibernoEnglish was effectively extinct”. If Irish English had died, say, by the time of Henry VIII, we might be encouraged to ignore the earlier period of Irish English and start with, roughly speaking, the Elizabethan settlement referred to by the Edgeworths. Yet once we allow for the effects of diglossia in limiting the number of written texts in English which were compiled within the medieval colony, and when we consider the important economic and political role of the towns in colonial society, we come to a very different picture of the maintenance and modernisation of Irish English. With regard to the role of towns in early Irish English, we should bear in mind that the account given by Thomas (1992) demonstrates solid archaeological or documentary evidence for no less than 56 such walled towns. Allowing for towns where the evidence is not definitive, the total possible number is over 100. These towns form a clear network of settlements, especially in the east and south, with other walled towns as far west as Galway and as far north as Coleraine. The map given by Whelan (1997: 183) demonstrates what he describes as “a curtain of walled towns” which, by 1350, “ringed the Pale and the south-eastern river valleys”. If we understand these towns as focal points for networks of English speakers with common political and economic interests as well as a common language, we can suggest that they had a stronger linguistic significance than their



Irish English in transition   

   19

strict population numbers would imply. We may, then, find a clearer guide to the development of Irish English in the view of Canny (1980: 169), who, citing “the remarkable resilience of the anglicized community in Ireland”, argued against Bliss’s portrayal and found it “probable that Hiberno-English culture was experiencing a revival rather than a decline in the decades approaching 1534”. Admittedly, the picture of the “state of Ireland” which was given to Henry VIII in 1515 was not encouraging (see State Papers 1834, II, iii: 8). The only areas counted as “subjett unto the Kinges lawes” were parts of the counties of Louth, Meath, Dublin, Kildare, and Wexford, and even in these counties, it was stated that “all the comyn peopplle […] that obeyeth the Kinges lawes, for the more parte ben of Iryshe byrthe, of Iryshe habyte, and of Iryshe langage”. Outside these loyal areas, it is said that the English people are “of Iryshe habyt, of Iryshe langage, and of Iryshe condytions, except the cyties and wallyed tounes”. Here we note the reference to cities and walled towns as reserve areas for English. Despite the pessimistic nature of this report, we should not take it as indicative of the total assimilation of the Anglo-Irish community; the report also opined that “thEnglyshe folke” of the counties which lie outside the King’s control “wolde be right gladde to obey the Kinges lawes” if the king could provide them with sufficient protection. Changes subsequently instigated by Henry VIII sought to eliminate the tripartite division – linked by unstable and changeable allegiances and practices – of Gaelic society, Anglo-Irish society, and English rule. Henry’s initial efforts were not particularly successful, and from 1534 onwards, we see a strengthening of the legislative tradition of fighting Gaelicisation across a wide range of activities, including language. A directive in 1536 from Henry VIII to the town of Galway (Hamilton 1860: 17), for example, instructs the citizens to sell goods only in market towns, “to shave their lips, to let their hair grow over their ears, and wear caps”, not to give protection to the king’s enemies, and “to learn English”. The values associated with English and Irish are made explicit in “An Act for the English order, Habite, and Language” passed by the Irish parliament in 1537 (Statutes 1786: 119–125; note also Crowley 2000: 21–23). Based on the assumption that “there is againe nothing which doth more conteyne and keep many of his subjects of this his said land, in a certaine savage and wilde kind and maner of living, then the diversitie that is betwixt them in tongue, language, order, and habite”, the parliament directed that “the said English tongue, habite and order, may be from henceforth continually (and without ceasing or returning at any time to Irish habite or language) used by all men that will knowledge themselves according to their duties of allegeance” (Statutes 1786: 120). This association of loyalty and the use of English, and the desire to build a stable community around these values, is further seen in the direction (pp. 121–122)

20   

   Geography, demography, and cultural factors

that every person or persons, the King’s true subjects […] to uttermost of their power, cunning, and knowledge, shall use and speake commonly the English tongue and language, and that every such person and persons, having childe or children, shall endeavour themselfe to cause and procure his said childe and children to use and speak the English tongue and language, and […] shall bring up and keep his said childe and children in such places, where they shall or may have occasion to learn the English tongue, language, order and condition.

Measures of this kind resemble those of earlier times, both in their symbolic and political assumptions. They were also as ineffective as their predecessors. Crucially, this legislation still accepts the notion of separate political societies within Ireland. Yet problems for English authority stemmed from both the continual possibility of the rebellion of native Irish society, and from the uncertain loyalties of Old English society itself. Henry II had early expressed the need to rein in the independence of the first English colonists in Ireland; similar considerations motivated the future king John to visit as lord of Ireland in 1185. By the time of Henry VIII, the self-interest of the Anglo-Irish colony was still a problem: Brady (1989: 27), for example, argues that “having once established their military and political dominance in the island, the descendants of the Anglo-Norman conquerors had exploited their superiority not to defend or extend English law, but simply to increase their own power and wealth”. Thus it is that the Parliament of 1541 takes on particular significance in charting the course of the historical and political developments which frame the Irish linguistic environment. Though preparations for this parliament are well documented in State Papers (1834) and elsewhere, there is still uncertainty as to the names of all the participants (see State Papers 1834 III (iii): 306–7fn and Bradshaw 1979: 239fn). The major theme which emerges from this material is the care which was taken to ensure the participation of nobility from both the Anglo-Irish and Gaelic Irish populations. This broad base of attendance signalled what Bradshaw (1979: 240) describes as “a revolutionary constitutional innovation”, since it showed “the inauguration of a single polity of subjects composed of Englishry and Irishry alike”. This new conception of Irish society was given further impetus by the policy known to historians as “surrender and regrant”, under which the Gaelic nobility and a smaller number of lords from among the Old English were encouraged to surrender their hereditary claims of land and control in order to have them re-granted with new charters by king Henry. Ostensibly, this policy might bring about little practical change at the beginning, but by breaking the link between land ownership and earlier historical claims, the policy aimed to bring about what Maginn (2007: 955) describes as “the eventual incorporation of Ireland’s English and Gaelic populations into an expanding English state”.



Irish English in transition   

   21

A fundamental legal element in the objective of incorporating the native Irish nobility into a single political order is found in the act which declared Henry VIII as king of Ireland. Before this time, the English kings had followed the principle established by Henry II, that the king of England would hold the title of Lord of Ireland. A detailed description of the events leading to the naming of Henry VIII as King of Ireland has been given by Anthony St Leger, the Lord Deputy of Ireland at the time (State Papers 1834 III (iii): 304). According to this account, a “solempne preposition” in praise of Henry was read out first by Sir Thomas Cusake, which was answered in kind by the King’s Chancellor. The Earl of Ormond then “briefly and prudentlie declared, in the Irysshe tong”, the effect of these declarations to the assembled lords, “greatly to their contentation”. Following this ritual overture, the bill naming Henry and his heirs as king of Ireland was read; once “being redd, and declared to them, in Irisshe, all the hoole Howse most willinglye and joyouslye condissended and agreid to the same”. The bill was further approved by the house of commons, which included “dyvers knightes, and meny gentilmen of faier possessions”. Later commentators (e.g. Cahill 1938; Ó Cuív 1951; Bliss 1978; Crowley 2005; Hickey 2007) have accepted a view, as expressed by Ó Cuív (1951: 13), that the use of Irish on this occasion “was […] sheer necessity, for it seems that the Earl of Ormond alone of the Anglo-Irish Lords knew English”. I have argued elsewhere (Kallen 1994a, 1997c), though, that the historical evidence does not support this view. There is, for example, no other evidence of the use of Irish in the Parliament of 1541: see Statutes (1786), State Papers (1834), Hamilton (1860), and Brewer and Bullen (1867). Even the incomplete evidence of correspondence with the nobility in the State Papers gives an indication of a role for English at this social level. We may note William de Burgo’s letter of 1541 to St Leger in English, swearing his allegiance and lamenting his earlier degeneration (State Papers 1834: 290–291), and a letter of the same year regarding McGillapatrick of Upper Ossory (State Papers 1834: 289), about whom it is stated that “his sonne hathe bene, this yere and more, in your Inglisshe paale, and is well brought up, and speketh good Inglisshe”. Further discussion along these lines is also given by Palmer (2001: 140–1), who states that for the aristocracy under the new regime, “some command of English was useful, or even imperative, for those confronting an expansionist English administration”. Lydon (2003: 223) also takes the parliament of 1541 as significant on the basis that in it “Gaelic lords for the first time sat with their Anglo-Irish peers”, pointing out that “these Gaelic lords were themselves a sign of the new times, bearing English titles and pointing the way to a degree of Anglicization which was eventually to bring to an end the Gaelic order”. In short, we would be better to see the use of Irish in the 1541 Parliament not as a utilitarian concession to the loss of English, but as a symbolic, political recognition of the

22   

   Geography, demography, and cultural factors

importance of Irish in the attempt to incorporate Gaelic society (including those who wavered between the two societies) into a common political order under English rule. Since English would be the dominant H language in this new order, knowledge  – though not necessarily wholesale adoption  – of English would become important for the Gaelic elite at an early time. Many subsequent changes in political structure and economic life lie beyond the scope of this review, yet we must note the establishment of new English settlements which changed patterns of communication, ultimately giving rise to the conditions for large-scale language shift. Queen Mary initiated a scheme in 1556 to settle English colonists in Counties Laois and Offaly, which were renamed as Queen’s County and King’s County. This scheme involved 160 families, and was largely unsuccessful. Elizabeth I, however, continued in this vein, with the confiscation of lands in Munster, and the apportionment of territory to English “undertakers” who were to clear the land of the native Irish population and settle English people in their place. Smyth (2006: 99) graphically demonstrates the spread of new population centres on the matrix of medieval settlement: for discussion and maps see also Lennon and Gillespie (2000: 58–9), Edwards and Hourican (2005: 156–8), and Smyth (2006: 85). Jefferies (2003: 877) estimates that the Munster plantation had brought in 14,000 settlers by 1611, while the settler population was to rise to an estimated 22,000 by 1641. Putting the Munster plantations in the more general context which includes the Ulster plantations, discussed with special relevance for our purposes by Corrigan (2010: 114–121), Smyth (2006: 100) estimates that at the start of the 17th century, immigrants of English or Scottish descent constituted 5% of the total population in Ireland, yet by 1641 the figure would be closer to 20 percent, rising to 27–30% by 1700. Even though these figures are heavily weighted by the Ulster plantations, the building up in the rest of the country of new communities and new patterns of land use and economy, in which English was used as the everyday language, could only have had an accelerating effect on language shift. Anecdotal reports of language use in the 16th and early 17th centuries century give evidence of interaction between the two languages. We may note, for example, Dominicke Linche’s efforts in 1569 to establish a school in Galway where English would be taught (Kelly 1897), sets of bilingual letters to and from the Irish leader Cormac O’Conor in 1567 (see Hamilton 1860: 328), and a letter of 1583 from Sir Morogh ne Doe O’Flaherty addressed to the Mayor of Galway and to Barnaby Goche. This letter, which notes the lack of food or drink in Connacht at the time, was written in Irish, and includes what Hamilton (1867: 441) calls “an English interpretation”. Some leaders chose between the two languages for pragmatic reasons: James Fitzmaurice, who wrote letters in 1579 to various nobles urging a rebellion against the queen (Hamilton 1867: 172–3), wrote to some in



Irish English in transition   

   23

English, and others in Irish. Evidence of Gaelicisation, language contact, and, at the same time, the distance between English (as a H language) and Irish is seen in the report of James Stanihurst, a Dubliner who had been educated at Oxford and Lincoln’s Inn in London. Stanihurst’s (1577) description of Ireland declared (p. 3v) that English was spoken in the Pale and in Wexford, and in “all the ciuties and townes”, though elsewhere the “native language is Irish”. He continued the view that recent generations had lost the English language to Irish, and in so doing had compromised the values of the English conquest. In Stanihurst’s analysis, “the Irish language was freee dennized in ye English pale” because the Anglo-Irish “became not all togither so wary in Keeping, as their auncesters were valiant in conquering” Ireland. Stanihurst paid particular attention to the continued use of “the dregs of the olde auncient Chaucer English” (p. 3) in Fingal in north Co. Dublin and in the baronies of Forth and Bargy in Co. Wexford (see also Chapter 4 of this volume). Stanihurst’s account gives an early indication of language contact in the archaic dialect, since he claimed (1577: 2v) that speakers of this variety “have so acquainted themselves with the Irishe, as they have made a mingle mangle, or gallamaulfrey of both the languages”. Note that it is this older English, and not any contemporary variety, about which Stanihurst makes his observation on language mixing. Thus while this dialect (or L variety in our terms) may have shown the effects of bilingual contact, it was Stanihurst’s claim (p. 27v) that, elsewhere, “the citizens, townesmen, and the inhabitants of the english pale […] differ little or nothying from the auncient customes and dispositons of their progenitors”. Christopher St Lawrence, Lord of Howth and compiler of the text known to us as the “Book of Howth” (Brewer and Bullen 1871), gives us insight into perceptions of the Anglo-Irish community at this time. St. Lawrence, who had also studied at Lincoln’s Inn and was, as McGowan-Doyle (2011: 20) says, “operating at the centre of English administration”, went with some of his allies in 1562 to present letters to queen Elizabeth. His note of the event (Brewer and Bullen 1871: 201) is that “the Queen asked the Lord of Houthe whether he could speak the English tongue. Belike, such was the report of the country made to the Queen”. St. Lawrence’s resentful tone reflects both his assumption of English as the natural language for him, and for others of his class, to use, and a view that frequent reports of Gaelicisation and cultural treachery had distorted the Queen’s understanding of Anglo-Irish society. Written in the early 17th century, the report of Fynes Moryson, who lived in Ireland as chief secretary to Lord Mountjoy, the Lord Deputy of Ireland, between 1600 and 1603, shows little sympathy towards native Irish or Anglo-Irish customs (other than an appreciation of Irish whiskey). It does, however, give an insight into the bilingualism of the time, and the use of language in establishing social

24   

   Geography, demography, and cultural factors

boundaries. Moryson reported that the native Irish population “disdayned to learne or speake the English tounge”. Though this loyalty to Irish may have been expected, what Moryson found more surprising is that “the English Irish and the very Cittizens (excepting those of Dublin where the lord Deputy resides) though they could speake English as well as wee, yet Commonly speake Irish among themselves” (Kew 1998: 50; cf. Hughes 1903: 213). Moryson also noted “the Cittizens of Watterford and Corcke having wyves that could speake English as well as wee” (Kew 1998: 50), and further observed “wemen of the meere Irish but also of the old English Irish, who could speake English aswell as ourselves, yet durst not speake it with us if their husbands or their Fathers were present”. The suggestion that women of both societies, Anglo-Irish and Gaelic (where the term “mere Irish” uses mere in the historical sense of ‘pure’), could speak English “well” is itself important; so is the view that Anglo-Irish men “commonly” spoke Irish amongst themselves. It is possible that by this time, women had adopted greater adherence to linguistic prestige norms which favoured English of a kind closer to the type spoken by Moryson. We may also see in the use of Irish for intra-communal conversation, and in the efforts of husbands and fathers to discourage women within their households from speaking English to an outsider like Moryson, the use of language to maintain social boundaries. As a final illustration of H language multilingualism during this era, we may cite O’Flaherty (1846: 281–2), who wrote in 1684 and recorded, in relation to an indenture from 1615, that Richard Lynche in Galway was at that time in the practice of writing legal instruments for the “natives of Iar-Connaught”. Sometimes the documents were in Irish, “often” in Latin, and “generally” in English. In the latter case, the documents also include certificates which state that the contents had been explained to the relevant parties in Irish. Subsequent Irish history also forms a chain of events which have linguistic consequences, but which cannot be considered in detail here. Religion becomes the focal point of conflict, with Catholic resistance to plantation culminating in rebellion in 1641. This uprising has been the subject of much analysis by geographers and historians alike: note, for example, Smyth (2006: 103–65). Depositions given by witnesses and victims of the violence inflicted in this uprising have themselves become a source of linguistic information, since a collection of 19,000 manuscript pages dating from this period has recently become available in transcription and digital images. Very few attempts to analyse this material linguistically have been made thus far, but the 1641 Depositions website opens many possibilities for further research. The demography of this period shows both a substantial decrease in overall population levels, which are estimated to have dropped from between 1.8 and 2.1 million people to no more than 1.3 million (see Smyth 2006: 159–61 for details),



Irish English in transition   

   25

and a redistribution of land and population in which land was confiscated from Catholic ownership, much of the Catholic population was transplanted to poorer land in Connacht and Co. Clare, and significant land grants in the confiscated areas were made to “adventurers” and soldiers who had served with forces loyal to Oliver Cromwell. Determining precise population figures for the Cromwellian plantation is problematical; Edwards and Hourican (2005: 161) point to a figure of over 1,000 adventurers and a plan to grant land to 35,000 soldiers, though it appears that “fewer than a quarter” of these soldiers ever actually settled in Ireland. Smyth (2006: 161) estimates that over 10,000 Catholic landowners forfeited their land at this time, and a total of perhaps 45,000 people underwent transplantation to Connacht between 1653 and 1655. These displacements, which have been mapped by Edwards and Hourican (2005: 162) and Smyth (2006: 169), gave further power to the spread of English, not only by providing new populations for whom the L domain of everyday communication was dominated by English, but by establishing English-dominated H domains in new relationships of law, administration, and land ownership. We can gain indirect insight into language distribution following the Cromwellian settlement by analysing the “Census of 1659” (Pender 1939, 2002). As Pender (1939) makes clear, the method and purpose of this work, incomplete and not published at the time it was compiled, is something of a mystery. Smyth (1988: 56) describes the census as actually constituting “an abstract of poll-tax returns for 1660”, and suggests that the returns significantly underestimate the true population of the time (see also Smyth 2002: xxix). What is significant for linguistic history, though, is that the population in this census is returned in what look like linguistic or ethnolinguistic categories: “English”, “Irish”, “Scots”, and, in the barony of Bargy, Co. Wexford, “Old English” (cf. earlier comments and discussion in Chapter 4). Thus while it would be inaccurate to take this head count as a linguistic census, it is also true that if we allow for bilingualism, especially within the Anglo-Irish population (who may have been counted as “Irish” on the basis of their long standing in the country but should not be assumed to be monoglot Irish speakers), and do not try to associate language exclusively with ethnic descent, we can still gain a general view of language distribution at this time. Here we may cite, for example, Aalen’s (1997: 23) commentary on the mapping of 17th century surnames, which points to a high concentration of “Old English” names in the southeast, with lesser concentrations elsewhere in eastern parts of Munster, and with the “old Pale area […] still recognisable to the north of Dublin”. Whelan (1997: 185) further interprets the census data to demonstrate an English “Dundalk-Ennis-Cork-Waterford quadrangle” which “bears the strongest imprint of the manorial village structure”. For maps, see further Smyth (1988: 75, 78; 2002: xlv, liv, lvi, lix; 2003: plate 2h, plate 2i) and a map of Co. Dublin in Kallen (1985).

26   

   Geography, demography, and cultural factors

With these caveats in mind, it is useful for us to examine some patterns in the 1659 “census” figures. The general predominance of the Irish majority is reflected in broad figures. In the 27 counties for which records exist, it is only in Ulster that large communities returned as English or Scots are to be found. Even within Ulster, an outlying county such as Donegal shows only 28 per cent in the English and Scots category. Co. Dublin, with an “English” population of 45 per cent, is the only non-Ulster county with an English population greater than 20 per cent. Counties such as Galway and Mayo, for which no data exist, would be expected to show even higher percentages of Irish returns. Yet if we turn our attention to the urban areas, the pattern shifts towards one in which the central core of the city or town contains a substantial number –  sometimes a majority – returned as English, while the surrounding liberties and suburbs are much more solidly Irish, as is the surrounding countryside. These urban focal points for English, linked by a commonality of political and economic interests, form not only a network for the further addition of value for English in H language domains, but a path along which we could expect various kinds of bilingualism in which English speakers interacted with their bilingual and Irish-speaking neighbours. To illustrate the patterns which emerge from the “census”, Table 1 shows a list of towns or cities with populations returned at over 600 in counties which are now part of the Republic. Some areas in the census are not included, due to problems in the way results were recorded. For each city or town, Table 1 gives the total population (Total) and the percentage of that total returned as English within the central core (Urban%), and in the suburban area or adjacent townlands (Suburb%). The (Area%) column then gives the percentage which the English population represents out of the total population of the combined city and suburbs. The (English %) column gives the percentage of the English in the entire city or town as a portion of the English population in the county as a whole, while the (County%) column shows the percentage which the English population represents in the population of the county as a whole. In some cases, denoted by NA, statistics are presented in such a way that it is impossible to make the relevant calculations. These figures may include soldiers and their wives, most, though not all, of whom are classed as English. Figures which include Northern Ireland are given in Kallen (1994a: 158). The disproportionate centring of the English population in the central urban areas is brought into clear relief in these figures. In Kinsale, Co. Cork, for example, 43 per cent of the urban population is returned as English, while even in the suburbs, 28 per cent are returned as English. Yet this population, which represents 13 per cent of the English population in the county as a whole, constitutes only two per cent of the total population of the county. An “English” majority in the central area of Limerick is counterbalanced by an English population of



Irish English in transition   

   27

only 6 per cent in the suburban areas; this population, at 41 per cent, represents almost half of the English population returned for Co. Limerick as a whole, but only three per cent of the total population in the county. Though Dublin represents something of an exception, since its urban English population accounts for 30 per cent of the Co. Dublin total, it nevertheless follows the pattern of an urban core with the largest proportion of its English population, surrounded by suburban areas that are strongly “Irish” in the returns: some such districts, such as Raheny in north Dublin and Crumlin in south Dublin, show “Irish” returns of 85% or more. Details on the history of Irish in Dublin, including maps based on the 1659 “Census”, are given by Mac Mathúna (1991). Table 1: Population returned as English, “1659 census” (Pender 1939)

Town

Total

Urban %

Suburb %

Area %

English %

County %

Carlow Dublin Kilkenny Dundalk Drogheda Wexford New Ross Cork Kinsale Youghal Limerick Waterford Athlone Sligo

1,517 21,827 1,722 2,536 1,605 902 618 4,826 2,197 1,111 3,105 1,607 948 1,398

48 74 39 29 NA 44 46 62 43 48 53 57 40 27

13 25 13 07 NA 18 0 28 28 06 06 14 14 09

26 45 25 13 60 38 39 33 38 42 26 39 56 15

52 66 30 18 53 21 15 25 13 07 41 47 NA 39

07 30 02 03 10 02 02 04 02 01 03 05 NA 03

Even considering the under-estimations and methodological weaknesses and uncertainties associated with this census, what matters most for using this material to chart the spread of English in Ireland, is that it supports the anecdotal and other evidence of the cities and towns as part of a network of English-language communication and interests, which provided many points of local contact with networks of Irish speakers. Unfortunately, the political and economic concerns which motivated the census did not call for any account of bilingualism, so we can only use this material as a complement to other sources and suggest that by the middle of the 17th century, there were ample opportunities for contacts of various kinds between English and Irish. The remaining years of this transitional period, particularly since the military victory of British forces led by William III in 1691, saw the enactment of laws

28   

   Geography, demography, and cultural factors

which became known collectively as the “penal laws”. This body of legislation aimed to restrict Catholic political participation and land ownership, as well as the activities of the Catholic church, including education. The destabilisation of Gaelic society which resulted from these developments allowed further inroads from English, reaching all levels of society. The status of Irish as a H language could be maintained in some areas, since those areas where Catholic landowners or middlemen were able to maintain their position in the 18th century included what Dickson (2005: 262) describes as “modest patrons in the countryside and town prepared to support poets and scribes, employing them as tutors and clerks”. In this vein, Ó Murchú (1970: 26) cites examples such as that of Sir James Cotter, who died in 1705. Cotter was “said to have spoken five or six languages fluently” but maintained Irish as the language of home; the important Irish text Párliament na mBan [‘The Parliament of Women’] (Ó Cuív 1970) was dedicated by its author to Cotter’s son in 1697. Evidence from within the Gaelic tradition, even in the early 17th century, shows the negative reaction to the spread of English. Conell Ma Geoghagan of Westmeath, who translated the “Annals of Clonmacnoise” from Irish into English in 1627, complained in English about the demise of support for the Irish literary tradition, observing (Murphy 1896: 7–8) that there were many septs in ye kidome that lived by it [the manuscript] & whose profession was to Chronicle and keep in memory the state of the K.dome [kingdom] […] & now because they canot enjoy that respect & gaine by their said profession […] they set naught by the sd [said] knoledg [knowledge], neglect their Bookes, and choose rather to put their children to learn eng: [English] than their own native Language.

Other manifestations of resistance to the spread of English can be seen in works such as the Irish poem “Faisean Chláir Éibhir” [‘These fashions on the plain of Éibhear’] by Brian Mac Gíolla Phádraig, who lived from ca. 1580 to 1652. Here the poet, cited with original and translation from Ó Tuama and Kinsella (1981: 89–91), laments a changing social world and comments on language specifically. The reference to smáilBhéarla ‘grimy English’ draws a connection between the decline of Gaelic high culture and the spread of English in the L domain. Is cor do leag mé cleas an phlás-tsaoilse: mogh in gach teach ag fear an smáilBhéarla ’s gan scot ag neach le fear den dáimh éigse ach ‘hob amach ’s beir leat do shárGhaelgsa’ A trick of this false world has laid me low: Servants in every home with grimy English but no regard for one of the poet class save ‘Out! and take your precious Gaelic with you!’.



Irish English in transition   

   29

More indirect satire crafted in works such as the 17th century Pairlement Chloinne Tomáis [‘Parliament of Clan Thomas’] (Williams 1981), the later Stair Eamuinn Uí Chléire [‘The history of Eamuinn Uí Chléire’] by Seán Ó  Neachtain (newly edited and translated by Mahon 2000), and the poem “Suirí Mhuiris Uí Ghormáin” [‘The courtship of Muiris Uí Ghormáin’] by the 18th century Louth poet Peadar Ó Doirnín (De Rís 1969: 28–9). Pairlement Chloinne Tomáis is of special interest here, not only because of its textual complexity, but because of what its attitude to language reveals about linguistic relations in the 17th century. The first part of this manuscript was evidently composed in the early 17th century, with a second part composed after the Cromwellian conquest. Existing texts come to us from manuscripts dating to the early 18th century: Williams (1981) provides commentary, text, and translation. As Williams (1981: xxv-xxvi, xlii-xlvi) shows, the author of the first part of this work demonstrates an intimate familiarity with the local effects of the Munster plantation in Kerry, while the author of the second shows equal familiarity and involvement with the effects and aftermath of land distribution following Cromwell. Both parts of the manuscript provide an insight into the changing social and economic relationships of the time, and both give an insight into the role of English. Williams (1981) argues that sources of inspiration for the author of the first part of the text include the relatively obscure English satirical pamphlet known as Martin Mark-All, Beadle of Bridewell, as well as Alexander Barclay’s English translation of Sebastian Brant’s Narrenschiff or “Ship of Fools”. This section includes a satirical dialogue between an English-speaking tobacco merchant (referred to as Roibín an Tobaca) and one Tomás an Trumpa in which Tomás, to the delight of his Irish-speaking comrades, speaks a kind of broken English that is only humorous to those whose English is good enough to appreciate the poor quality of Tomás’s English: “What the bigg great órdlach for the what so penny for is the la yourselfe for me?” says Tomás in asking how much tobacco he can buy for a penny (see Williams 1981: 40, 97 for texts). In terms of diglossia, the rhetorical stance of this text, which we can presume was shared with the intended readership, is to assume Irish as a H language and to mock the L domain English used by low-level commercial functionaries such as Roibín the tobacco seller. This mockery extends to those whose desires to associate with incoming norms attracted them to English, no matter how poorly they spoke the language. The literary inspiration for this part of the manuscript which is pointed out by Williams (1981) leads us to an insight into H level bilingualism of the time, since the disparaging view of English (in the L domain) is conveyed within a narrative that draws inspiration from literary works in English (in the H domain). The second part of the manuscript contains no such dramatic scenes, but reveals a greater number of English loanwords, inserted into literary Irish with varying effects. As Mac Mathúna (2007a, b; 2012) makes clear, the use of

30   

   Geography, demography, and cultural factors

English insertions and loanwords for a variety of purposes, while not typical of Irish writing in the 17th and 18th centuries, was not uncommon. The 18th century saw the continuing development of bilingualism at various levels. Ó Murchú (1970: 27), for example, notes the linguistic loyalties shown by the 18th century Irish-language poet Donnchadh Caoch Ó Mathúna, who “found it necessary to declare his allegiance to Irish values when he was derisively described as an Englishman because he was overheard speaking English at the market in Cork”. At a different social level, we get a glimpse of local linguistic relations from an account of some illegal activities in Galway in 1739 (MacLysaght 1944: 107–108). This account focuses on a bilingual trial in which it was alleged that Mathias Kincannon had appeared with a book in his hand and a stole around his neck in order to carry out a fraudulent marriage ceremony between Luke Lyons and Jenny Finn. Two witnesses confirmed this accusation, but a third, Margarett McDonnough, “not understanding English did not know what was said” at this occasion. It was, however, further noted that McDonnough “knows said Kincannin well and saw him often say Mass at Mr. Anthony Lynch’s house”. In this case, we can see Kincannon as a bilingual speaker at the centre of a network which includes both speakers of English and speakers of Irish. We may also cite the report of John Bush (1769), writing more in the manner of Moryson, on his trip through Ireland. It was Bush’s view that (p. 34) “English is the universal language of the country among people of any fortune, and very few of the lowest class are met with that cannot speak it”. Elaborating, Bush (pp. 34–35) maintained that “you’ll meet with thousands of the lowest rank, who speak both English and Irish with equal ease; and […] perhaps, they really speak better English than the same class in England”. Bush attributed this superiority to the learning of English in school within this stratum of society. Bilingualism in this period also saw the strong development of a “macaronic” song tradition, in which verses or lines of English and Irish were mixed, sometimes with contradictory or humorous effect, and sometimes satirising those who were using English: “Suirí Mhuiris Uí Ghormáin”, cited above, is one such satirical example, while Ó Muirithe (1980) provides a collection of such texts.

1.4 L anguage shift and linguistic realignment The Irish Parliament, which had been established in the late 13th century, was abolished by “An Act for the Union of Great Britain and Ireland” which was passed by the respective Parliaments in Britain and Ireland in 1800, and took effect in 1801. This legislation established the United Kingdom of Great Britain and Ireland. In the next century, following years of agitation and a period of guer-



Language shift and linguistic realignment   

   31

rilla warfare that started in 1919 and is commonly referred to as the War of Independence or the Anglo-Irish War, a peace treaty was signed in 1921 and ratified in 1922 which established the Irish Free State in the 26 counties that now constitute the Republic. Two major linguistic themes emerge from this period. One is the popular shift to English, which, given the history we have discussed thus far, can be understood as the culmination of the loss of Irish as a H language in its traditional domains and as a large-scale switch to English as the L domain language of first-language communication. The second is the contradictory rise of Irish as a H language, not in the domains of traditional Gaelic society, but as part of a nationalist movement, similar to 19th century national language movements elsewhere in Europe, that associates Irish with national identity and has made support for the language a defining part of the linguistic agenda in the Republic. Both elements are essential to an understanding of the social and cultural standing of Irish English in the Republic today. We get a glimpse of language shift in progress from the parochial survey compiled by William Shaw Mason (1814–19) from reports of Protestant clergymen in response to a detailed parish questionnaire. Mason’s question as to “the genius and dispositions of the poorer classes; their language, manners and customs, &c” provoked a wide range of responses. Seventy-seven reports can be analysed for evidence of linguistic practice, 59 of which are from territory in the presentday Republic (including Donegal and Cavan). Incomplete though the reports are (with none, for example, from Galway or Mayo, and only one from Sligo or Wicklow), they do give an overall picture of transitional bilingualism. The selections below illustrate a number of common themes in this material, notably (1) the role of markets and towns as central places for English; (2) inter-generational shift, with younger speakers more inclined to speak English; (3) bilingualism in which Irish is the language of in-group communication and English used for communication with outsiders; and (4) the association of English with higher prestige groups. (1) Kilgarriff, Co. Cork. “The language of the common Roman Catholic peasantry is Irish; Protestants of the lower order speak both English and Irish: in town [Cloghnakilty] English is frequently spoken by both”. (II: 311). (2) Lismore, Co. Waterford. “Many of the common people now speak English, and the next generation will probably speak it; for the children are universally sent to school”. “Of the genius of the common people little can be ascertained, for they are very shy in converse with their superiors, and confidentially communicate only to each other”. (I: 554). (3) Middleton, Co. Cork. “The English language is generally understood and spoken; the lower class commonly speak Irish”. (III: 271).

32   

   Geography, demography, and cultural factors

(4) Tracton Abbey, Co. Cork. “The language generally spoken is a corrupted dialect of the Irish, differing in many respects from that spoken in Connaught. […] The English language is, however, gaining ground, in consequence of its being taught in the schools, and of the frequent intercourse of the inhabitants with the towns of Kinsale and Cork”. (III: 471). (5) Rathdrummin, Co. Louth. “The Irish language is in general use, but it is rare to find any person who does not speak English well; which language is still increasing amongst them, as may be collected from the children being always able to explain and interpret, where the parents do not speak English”. (I: 624). Table  2 summarises 59 reports which can be categorised to reflect degrees of language change or maintenance; in anticipation of the later territorial division, the Ulster data include only the Ulster counties now in the Republic. Haphazard though the coverage may be, it gives evidence of both the spread of English in the east and the retention of Irish, particularly in parts of Munster. Table 2: Summary of language reports in Mason’s (1814–19) parochial survey

Province

Irish dominant

Bilingualism: Irish dominant

Bilingualism: English ­dominant

Bilingualism: other

English ­dominant

Leinster Munster Connacht Ulster

0 3 1 0

7 6 2 1

10 2 2 3

5 7 0 0

8 1 0 1

Education came particularly to the fore as a vehicle for language shift during the 19th century. The penal laws had restricted the availability of education for Catholic children, and a system of unofficial “pay schools” developed during the early 18th century. A combination of the surreptitious nature of this enterprise, the lack of funding, and the unstable tenure of teachers and classes alike meant that these schools often assembled in outdoor areas protected by hedgerows. Even when the schools were sited in what McManus (2002: 70) describes as “a bewildering assortment of buildings”, they continued to be referred to popularly as hedge schools. Because the hedge schools had no central organisation or management, we must be cautious in generalising about the hedge school curriculum. Certain trends in language use are nevertheless clear. Classical education, especially in Latin, but to a lesser degree in Greek and Hebrew, was highly valued and provided a basis for boys who would attend Irish colleges on the Continent for edu-



Language shift and linguistic realignment   

   33

cation to the priesthood. Smith’s (1774: 418) account from Co. Kerry, that “I have in my survey met with some good Latin scholars who did not understand the English tongue” and that “Greek is also taught in some of the mountainous parts, generally by persons who pick it up as mendicant scholars”, may or may not be typical, but it illustrates the perceived link, discussed also by McManus (2002: 125–130) and more generally by Dowling ([1935] 1968), between the hedge school and classical H languages. The societal tensions between Irish and English were naturally reflected in the hedge schools. While anecdotal evidence highlights the importance attached to classical languages, and the documentation by McManus (2002) of prominent Irish-language poets who were hedge schoolmasters helps to support her view (p. 132–133) that “the hedge school masters held Irish culture in the highest esteem”, a number of other factors served to increase the role of the hedge schools in the spread of English to the younger generation. These factors included the growing economic advantages of English in Irish trade and employment; its importance in Britain and North America (to which emigration was increasing); a negative image which had developed as a backlash against the link between Irish-language education and Protestant proselytising (see Crowley 2005: 36–49 for details); and the much greater availability of books in English. Thus the trend, as documented in many of Mason’s (1814–19) reports, was for the school to be seen as a centre, among other things, where English would be learned and used. Considering this mix of Irish, English, and Latin, Bliss (1977b: 18) suggests that the hedge schools may have influenced some aspects of the development of Irish English. Hedge school teachers who were not native speakers of English may have passed on some pronunciations and lexical uses which were based on spelling and the application of rules from Latin. Evidence from McManus (2002: 91, 114) suggests that mastery of “sesquipedalian words such as ‘antitrinitarians’ and ‘coessentially’ ” was not only a goal of the education of children, but a reflection of the value schoolmasters may have put on such words in order to demonstrate their high standard of education. As the penal law restrictions were lifted, education supported by Catholic and Protestant institutions began to expand. A parliamentary commission investigated the state of Irish primary education in 1824, and found over 10,000 schools in operation; Akenson (1970: 57) estimates that the “great majority” of Catholic children, between 300,000 and 400,000, were attending hedge schools. McManus (2002: 57), citing work from Joseph Lee, gives a suggestion that this provision would have seen approximately one third of the population between the ages of 5 and 15 in school. Since this provision did not meet the needs of the population systematically, a state-supported national school system for education at primary level was established in 1831. Language was hardly considered an issue

34   

   Geography, demography, and cultural factors

in this system. Akenson (1989: 537) cites the “virtual non-existence” of rules on the use of Irish at official level during this time, though one educational inspector in the 1850s did make a case, which was not followed up, that using Irish would be beneficial in Irish-speaking districts. Akenson’s view that “it is a mistake to conclude that the [education] system was a British machine for destroying the Irish language” suggests – and the demographic evidence supports this view – that by this time, no specific instruments were needed to create an education system in which literacy and English were tied together. This direction should not surprise us. The same popular opinion which had associated English with economic success in the hedge schools continued unabated: Ó Súilleabháin’s (1940) study of oral tradition concerning the use of bataí scoír (‘tallysticks’) to punish children for speaking Irish in school shows a practice which spans the era of the hedge school into the National School era. Moreover, even if a decision had been made to provide Irish-language education on a large scale, the new educational system would have had difficulties in finding an adequate supply of materials and a sufficient supply of suitably trained Irish teachers. The trend away from Irish was given further impetus by deaths and migration which followed the famine that started in 1845 after loss of the potato crop due to potato blight. Though a number of food shortages had hit Ireland in the first part of the 19th century, the Great Famine, in which it is now estimated that 1 million people died, had a profound impact on every aspect of Irish society. The impact of the Famine on language reflects the especially strong demographic effect of the Famine in those areas where Irish retained vitality as a community language. Whelan’s (1997: 85–8) account of the development of potato farming in Ireland maps the vulnerability of the west and south to potato blight (see also Kelly 2000: 92–93). These areas coincide to a large extent with the areas where Irish was used intensively as a community language (see, for example, maps by Fitzgerald 1984), and from which emigration during the Famine period was at its highest levels (see Miller 1985 and a review in Kallen 1994b). Thus the Famine, having hit many Irish-speaking districts much harder than the rest of the country, furthered the depopulation of Irish-speaking areas and drew a connection between knowledge of English and the hope of material wellbeing. The remainder of the 19th century saw the continued spread of English as both the dominant H language and, ultimately, the dominant L language of Ireland. Census data, which begin to be reliable, within limits, from 1861, have been analysed retrospectively by Fitzgerald (1984) to extrapolate the distribution of language going back to the generation born in 1771–81: compare also discussions in Ó Cuív (1951), Ó Murchú (1985), and Kallen (1988, 1994a). The essential theme in these figures is the continual rise in the percentage of English speakers in successive generations from the late 18th century onwards. The broad picture



Language shift and linguistic realignment   

   35

is only complicated by the different rates at which specific districts did or did not take part in language shift. This change took place after the Edgeworths pronounced on the use of English as the inheritance of those whose ancestors had brought the language to Ireland, and set the stage for the comments we have quoted above from Hume. The second theme of this period, the movement to restore the prestige of Irish and to revitalise the conditions of its use, has formed an essential part of language relations in the Republic. The developing pressures in favour of English had received various responses from within Irish literary tradition. Foras Feasa ar Éirinn “A Basis of Knowledge about Ireland”, written ca. 1634 by Seathrún Céitinn (who was known in English as Geoffrey Keating) provided a focal point against many of the anti-Irish arguments in circulation during the upheavals of the 17th century: see Crowley (2005: 52–55) for a concise discussion as well as Cunningham (2000) and Morley (2012) for detail. Antiquarian interest in Irish grammar and literature developed in the late 18th century with works such as those of Vallancey (1773). Charlotte Brooke’s Reliques of Irish Poetry (Ní Mhunghaile 2009) was published in 1789, and was pioneering in the presentation of Irish literature in English translation. This work had a major influence in renewing interest in Irish at the level of H domain literature. The foundation of literary and historical groups such as the Gaelic Society in 1807, the Iberno-Celtic Society in 1818, and the Ossianic Society in 1853 strengthened the appreciation of Irish, working against the prevailing trend to language shift and somewhat tangentially with nationalist political movements of the time, which were conducted largely through English. The joining together of loyalty to Irish with the demands of modern political nationalism is exemplified in the works of Thomas Davis, leader of what became known as the Young Ireland movement. Davis was a co-founder of the journal The Nation, and his essays developed a view of the Irish language and nationhood which had parallels in other 19th century European views on language, people, and nation. Davis ([1843] 1945: 69) was an early critic of the National Schools, arguing that “they are not national, they do not use the Irish language, nor teach anything peculiarly Irish”. In his two-part essay “Our National Language”, Davis (1945: 71–3) articulated the nationalist link with Irish: a people without a language of its own is only half a nation. A nation should guard its language more than its territories – ’tis a surer barrier, and more important frontier, than fortress or river. […] Nothing can make us believe that it is natural or honourable for the Irish to speak the speech of the alien, the invader, the Sassenagh tyrant, and to abandon the language of our kings and heroes.

36   

   Geography, demography, and cultural factors

Though Davis died in 1845, notions which he had articulated continued to develop. The foundation of Conradh na Gaeilge (the Gaelic League) in 1893 gave new impetus to the move to promote Irish, since it had a more popular appeal than the earlier antiquarian societies, and was able to exert pressure, for example, in making Irish a compulsory matriculation subject in the National University of Ireland, which had been given a legislative framework in 1908. The position developed by Douglas Hyde, a co-founder of Conradh na Gaeilge and a Protestant, can best be understood in the light of more general principles within Fishman’s (1972: 45) theory of language and nation. In Fishman’s view, language is often used by emerging groups to demonstrate that “the heirs of past greatness deserve to be great again” and that “the purported continuity of the language” can be “the authenticating device for finding, claiming, and utilizing one’s inheritance”. One of Hyde’s most frequently-cited essays illustrates these principles in action. Hyde ([1892] 1991: 527) starts by declaring that “if we take a bird’s eye view of our island today, and compare it to what it used to be, we must be struck by the extraordinary fact that the nation that was once, as every one admits, one of the most classically learned and cultured nations in Europe, is now one of the least so”. This past was necessarily Irish speaking, for in Hyde’s view (1991: 530), apart from the north-east of Ulster, “the bulk of the Irish race really lived in the closest contact with the traditions of the past and the national life of nearly eighteen hundred years, until the beginning of this century”. The dream of restoring a link between nation and language – the claiming of the cultural inheritance, in Fishman’s terms – is seen in Hyde’s conclusion (1991: 533) that if Home Rule be carried, […] the Irish language […] shall be placed on a par with – or even above – Greek, Latin, and modern languages in all examinations held under the Irish government. […] We shall insist that in those baronies where the children speak Irish, Irish shall be taught, and that Irish-speaking schoolmasters, petty sessions clerks, and even magistrates be appointed in Irish-speaking districts. If all this were done, it should not be very difficult […] [to] bring about a tone of thought which would make it disgraceful for an educated Irishman – especially of the old Celtic race […] to be ignorant of his own language […] at least as disgraceful as for an educated Jew to be quite ignorant of Hebrew.

Though the popular use of English, as well as the development of English as a vehicle for literary expression, continued to grow, the language question remained a point of controversy where national issues were concerned. Conradh na Gaeilge had been founded as a non-political, non-sectarian body, which, as Garvin (1986: 73) argues, offered Protestants the opportunity “of claiming an Irish identity that was unchallengeable without having to change their religion”. Later developments, however, saw a greater convergence between



Language shift and linguistic realignment   

   37

Catholic nationalism and the Irish language movement. Periodicals such as The Leader, founded in 1900, provided a forum for the exploration of such issues. Here we may cite the views of “P. F.” (1908: 332) who argued not only that “to adopt the English speech is to adopt the English thoughts and feelings in regard to these […] objects, namely God and Erinn”, but that “accordingly, although Irish speakers of English may still retain some dim, rudimentary relics of the old Gaelic tradition of Faith and Patriotism, they also have that in the very fibres of their mental and moral nature, which is essentially destructive of those remains”. Though not all correspondents for The Leader agreed with this position, views such as those of “P. F.” articulated a linkage of Irish, national independence, and Catholicism that rang true for many. Ultimately, Conradh na Gaeilge adopted a position in 1915 which advocated Irish political independence. This change of position caused a split within the movement which led to the resignation of Douglas Hyde. The equation of political nationalism and adherence to Irish continued to develop, so that when the Irish Free State was established following the implementation of the Anglo-Irish Treaty in 1922, the revitalisation of Irish became a founding principle for the Free State government. Hyde himself went on in 1938 to become the first President of Ireland, following the adoption of the 1937 constitution. The pre-eminent role given to Irish in public debates on language and nation did not mean that expressive, as opposed to purely practical, values for Irish English were neglected in the years leading to political independence. Though texts and discussion in Bartley (1942, 1954), Bliss (1979), and Earls (1988), among others, show the many ways in which early uses of Irish English in literature had often been mocking and had portrayed Irish characters with a minimal command of English or developed linguistic stereotypes that reinforced a more widely negative view of Irish people, literary uses of Irish English took a more positive turn during the 19th century. Edgeworth’s ([1800] 1964) relatively naturalistic use of Irish English in Castle Rackrent is a landmark in this development, and other 19th century writers sought a variety of ways to present the language, especially, of rural speakers from areas which were only then shifting to English. Texts of this kind, exemplified by Banim and Banim (1825) and Griffin ([1829] 1919), are found below in Chapter 7. Work by William Carleton, a native of Co. Tyrone who spent most of his life in Dublin, had special impact during this period: for linguistic discussion see, for example, McCafferty (2005b). The trend to use Irish English for literary purposes was developed later by writers such as John Millington Synge, Lady Gregory, W. B. Yeats, George Fitzmaurice, and others associated with the “Irish literary revival” of the late 19th and early 20th centuries. The use of Dublin vernacular and other types of

38   

   Geography, demography, and cultural factors

Irish English for literary purposes was developed, too, by writers such as Oliver St. John Gogarty, Sean O’Casey, and James Joyce.

1.5 E  nglish and the Republic The task of nation building following the Treaty of 1921–22 was complicated by the discrepancy between the ideological attachment to Irish which had become part of the nationalist movement in general and the strength which English had obtained by this time in both H and L domains of language usage. We can appreciate the problem by reference to Bauman’s (2001: 90) characterisation of what he terms “the nation-building stage of modernity”, in which “nation-building meant the pursuit of the ‘one state, one nation’ principle, and so ultimately the denial of ethnic diversification among the subjects. From the perspective of the culturally unified and homogeneous ‘state nation’ the differences in language or custom found on the territory in the state’s jurisdiction were but not-yet-fully extinct relics of the past”. The linguistic contradiction for Irish nation building in the formation of the Free State was that while a plausible claim could be made within the new state to recognise Irish as the national language, the relegation of English as a “relic of the past” which a strict “one state, one nation” policy would have advocated was untenable. Even Davis ([1843] 1945: 73–75) reassured his readers that they would not be asked to forsake the English language, but simply to agree that Irish should “be cherished, taught, and esteemed”. Though a detailed treatment of the position of English and Irish in contemporary Irish society would lie outside the scope of this introduction, we will consider in the rest of this chapter both the traditional associations between language and nationhood, and what I see as a postmodernist perspective (discussed also in Kallen 2002b) arising from the changing positions of English and Irish in an era of increased globalisation.

1.5.1 English, Irish, and the nation The widespread use of Irish English as an L language, together with the dominance of English in H language domains, was sufficiently great in 1921–22 that English, or a specifically Irish version of English, could plausibly have been designated as the official language of the new state. Yet at the level of official policy, the Free State constitution, and the 1937 Constitution which provides the basis for the Republic of Ireland, gave a special position to Irish as the “national language” in the establishment of official languages. Article 8 (1) of the 1937 Consti-



English and the Republic   

   39

tution (Bunreacht 1937) states that “The Irish language as the national language is the first official language”, while Article 25 (6) states that “in case of conflict between the texts of a law enrolled […] in both the official languages, the text in the national language shall prevail”. The formulation which elevates Irish to the position of “first official language” on the basis of its position as “the national language” is an explicit statement of the nationalist ideal which links Irish with the nation and sees the state as the political realisation of this concept of nationhood: it makes no claim that Irish is the most widely-used language. A practical concession to English is given in Article 8 (2) of the 1937 Constitution, which recognises English as “a second official language”, or in Irish mar theanga oifigiúil eile ‘as another official language’. Constitutional or subsequent legal references to English do not refer to any specifically Irish form of English, but simply to some generalised notion of the language. The importance attached to Irish has given rise to policies which have, to some degree, established a secure domain for the language within the Republic. The 2011 Census (This is Ireland 2012: 98) returns 41.4 per cent of the population over the age of three, totalling 1,774,437 speakers, as able to speak Irish; this figure represents a dramatic reversal of the situation which could be extrapolated from the language decline at the start of the 20th century. According to the State Examination Commission website, Irish was taken as a subject in the 2012 Leaving Certificate examination (the final qualification at secondary-level education) by 42,965 students, representing 82% of the total numbers sitting the exam. Broadcasting in Irish reaches a national audience in radio and television, and the public signage of the linguistic landscape (documented in Kallen 2009 and Moriarty 2012) maintains niche positions for Irish alongside the more dominant English. When the state acts in the area of language and nation building, whether through policy, education, or other domains, its concern is largely with the status and promotion of Irish. The marginal position of English in the ideology which is expressed in the constitution can not, however, be understood only in political terms. The long tradition of negative attitudes towards Irish versions of English which we can see in the satirical expressions of early writers never completely lost its grip on some sections of opinion, which can at times see the Irish version of English as inferior relative to some (usually undefined, but implicitly British) norm. This attitude, which Croghan (1986) has referred to as “brogue-speak”, is reflected in a wide range of works. Burke (1896: 187), though dispassionately discussing aspects of the development of Irish English, explained distinctive features of the lexicon in Irish English as “partly owing to our imperfect grasp of English”. Clery (1921: 552) went further in taking a hostile tone towards Irish English, arguing that “we certainly have not learned how to speak English, for we have not acquired its sounds […]. Like the Chinese with their ‘pidgin’ English, we have merely learnt

40   

   Geography, demography, and cultural factors

how to make ourselves understood by a system of mis-pronounced English words, incapable of literary development”. More recently, Moore (2011: 42) refers to “ongoing moral panic” in public perceptions of new phonological realisations in the Dublin area (on which, see references to “Dortspeak” in Chapter 5). Even positive views of distinctively Irish usage in English may ultimately return to the assumption of Irish as the true language of the nation. The 1965 government White Paper on the Irish language (Athbheochan 1965: 6), for example, recognises a role for English in Irish culture, but in so doing stresses the centrality of the Irish language, remarking that “English, of course, has also contributed to our national heritage but the English we speak still bears the imprint of the attitudes of mind and modes of expression which prevailed when Irish was the language of general use”. This reference to English assumes the existence of strong transfer elements from Irish – a point which will be discussed in Chapter 5 –  and implies that it is through such transfers that Irish English can retain a connection to cultural authenticity. Such a view is developed further by Henry (1974, 1977), who does not criticise Irish English for failing to meet some external definition of standard English, but instead argues that contact-influenced English, which he terms “Anglo-Irish” and describes (Henry 1977: 36) as “language forming on the same base as corresponding Irish structures, with native intonation and pronunciation and a foraging for English materials”, constitutes “a new language which could develop a literature in its own right” (1977: 23). Praising the authors of the Irish literary revival for their attempts to develop this new Anglo-Irish language, Henry (1974: 23) attacks the “education machine” in Ireland for having rejected Anglo-Irish and causing it to be “forced into the straight-jacket of a mode, style and grammar at once related to it, distinct from it and alien to it”. The adoption of this alien “Standard English”, according to Henry (1974: 25), is incompatible with authentic Irish expression, since the “standard” either imports directly from England or excludes anything truly native: Ireland is fast becoming a mere resonance chamber for Standard English; produced in Southern England, received in Dublin and sent in a kind of miming ritual by a voice from Dublin. […] We do not innovate: if we do, the innovation is rejected from the Standard language, that is English, or referred to the dubious peripheral area of Irish drollery. In other words, the undoubted and very considerable and varied creative power of the Irish people in respect of language finds virtually no outlet into written Standard English.

In ideological battles over language as historically defined within the Republic, then, Irish English is open to criticism from different angles: as a tongue which is alien to the concept of nation itself, as a language which has been imperfectly learned for utilitarian purposes under the pressure of colonisation, or as a contact vernacular whose most expressive elements reflect an Irish substratum



English and the Republic   

   41

that is rendered invisible by a standard version of English which lacks cultural authenticity. Henry (1974: 33) incorporates elements of all these criticisms in his declaration that “an Irishman speaking English is not in the same category as an Englishman. The language has only been rented out to him”. The different status of Irish and Irish English within the language policy of the Republic is tied up with the question of language standardisation. The evolution of this policy can be accounted for to a large extent by studies in the theory of standard languages. Garvin (1959: 29) pointed out that standard languages demonstrate “unifying, separatist, and prestige functions”, in which “unifying” functions can unite different dialects “into a single standard language community”, while the “separatist” function can “set off a speech community as separate from its neighbors”. Crucial attitudes towards the standard language include “language loyalty”, “pride”, and “awareness of the norm”. Haugen ([1966] 1972) elaborated on these points, in noting that standardisation requires action both at the level of language (for which norms are codified and functions elaborated) and of society (which selects norms and gives them general acceptance). Irish at the foundation of the state had fulfilled these social and political conditions for recognition at the standard level. Irish had inspired loyalty and pride in significant sections of the population; there was a willingness to undertake a dialectal unifying process which would result in a standard version of the language that could be used in schools, administration, newspapers, and elsewhere; and there was no doubt that the use of Irish would set Ireland apart from its neighbours. Though the partition between the Irish Free State and Northern Ireland left open the question of how extensively Irish would apply as a national language, partition had the ultimate effect of giving Irish a role in distinguishing the territory of the Republic from the nearest neighbouring political jurisdiction. Irish English, on the other hand, had no such position. It inspired no loyalty or pride from a national perspective since it did not constitute – to use a phrase from Fishman’s (1972: 44) theory discussed above – a “link to the glorious past”. There was no movement to codify a standard Irish English that was different from standard English anywhere else; and it would be difficult to construct any political or economic value in formalising a distinctively Irish form of English. Reflections on the possibility of giving official standing to a distinctively Irish form of English in the Republic are, of course, possible. Henry’s (1974, 1977) argument in favour of an Anglo-Irish developed and elaborated from L domain speech to a literary standard celebrated the work of authors in the past, though it did not affect contemporary policy. Croghan (1990: 17) discusses the possibility of elevating Irish English as a national standard, but concludes that “the proposal that Hiberno-English is the real national language would be generally greeted with incomprehension: some, including those who would not give any

42   

   Geography, demography, and cultural factors

support to the national claim for Irish, might suspect a lack of patriotism”. More recently, Dolan (2002: 154) has called for governments in the Republic and in Northern Ireland to offer “support for the main language of the country, HibernoEnglish”, on the basis that “the richness of the linguistic varieties in the island should be publicly celebrated”. For reasons we have seen thus far, though, politics and history have not worked to bring about such a role for any distinctively Irish form of English.

1.5.2 New directions on language and the nation The discourse of nation building with respect to language offers only a partial picture of developments since the foundation of the State. We have noted above the increase in the number of self-reported Irish speakers as evidence of success in the revitalisation of Irish  – both as regards actual usage and as a reflection of positive attitudes towards the language. Looked at another way, though, we should also attach importance to the decline in the number of Irish speakers in Gaeltacht areas. The 2011 Census (This Is Ireland 2012: 41) returns 66,238 Irish speakers living in Gaeltacht areas, representing less than two percent of the total population in the Republic. Ó Riagáin (2007: 229) estimates that no more than five percent of the population speak Irish as their “first or main language”. Even among pre-school children in Irish-speaking homes, truly monolingual Irish speakers are virtually non-existent: Brennan’s (2004: 12–13) study of Irish-speaking pre-school children in the Aran Islands Gaeltacht, for example, shows code switching and the use of English to be common in the children’s linguistic environment. Public language use, whether in publishing, broadcasting, or everyday speech, is strongly dominated by English. These developments suggest a new role for Irish – not as an L domain community language (as at the start of the 19th century), but as a second language, promoted and often learned in the H domains of school and other public usage, and used in varying degrees in L domains. This new direction could be seen as part of a post-modernist development in the Irish linguistic order, since it breaks the fundamental link between language and nation as understood in arguments such as those of Thomas Davis. Instead, it exemplifies the trend identified by Lyotard ([1979] 1984: 14) in which “the old poles of attraction represented by nation-states, parties, professions, institutions, and historical traditions are losing their attraction”. We can see indications of new developments as early as Greene’s ([1978] 1982) argument that Irish identity within the European Economic Community would be “greatly strengthened by the movement of Irish from a ritual role to that of a working language at home and in the Community”. Though it was not until 2007 that Irish was adopted as an



English and the Republic   

   43

official language of the European Union, this move did represent a step towards the internationalisation that Greene anticipated. A new definition of linguistic relations also followed the Good Friday Agreement or Belfast Agreement (Agreement 1998), which committed the governments of the Republic and the United Kingdom to a number of undertakings in regard to language in the wider context of what is generally referred to as the “peace process”. Cross-border bodies were established to support Irish (Foras na Gaeilge, based in Dublin) and Ulster Scots (the Ulster Scots Agency or Boord o UlstèrScotch, based in Belfast). The inclusion of the entire island within the remit of Foras na Gaeilge represents a new configuration of the links between language and nation, since it uses an agreement between the governments of two states to create a body whose constituency includes large sections of the population in Northern Ireland which have neither historical attachment to nor practical experience of Irish. This move, in principle at least, creates a new role for Irish, going beyond its perceived status as an ethnolinguistic marker in order to take a place as part of the broader cultural heritage open to society at large. A further re-analysis of Irish linguistic relations stems from the increase in immigration to Ireland during the late 20th century. Gallagher (2007) estimates that more than 200 languages are now spoken in Ireland, while the 2011 Census (This is Ireland 2012: 35, 91–92) shows 514,086 residents (roughly 11% of the population) speaking a language other than English or Irish at home: for those born outside the Republic, Polish, Lithuanian, French, and Russian were the most commonly spoken. Contemporary multilingualism is reflected in the expanded range of languages which may be examined at Leaving Certificate level (including Arabic, Japanese, Lithuanian, Latvian, Polish, Romanian, and Russian), and in the increasing visibility, illustrated in Kallen (2010), of immigrant languages in the linguistic landscape. The implications of this immigration for Irish are sketched by Ó Conchubhair (2008: 236–7), who suggests that “the old arguments conceived during the 1880–1920 revival period are neither applicable nor practical” for Irish, since so many new immigrants lack a historical tie to the language. Instead, he argues, “the new multicultural, multilingual Ireland offers an opportunity to reaffirm the centrality of Irish […] to any reconfigured definition of Irish identity”. McCubbin (2010) sees in immigrant discourse over language and nation a challenge to what he terms the “ethnically essentialist terms” which characterise most policy and public debate thus far. What links the perspectives of Greene (1982), the 1998 Agreement, and Ó Conchubhair (2008) is both the assumption that Irish is still crucial in a concept of Irish nationhood, and a recognition that the model in which support for Irish relies on a perceived natural connection to Irish – the kind of attachment which Shils (1957) refers to as primordial – does not fit current conditions. When we con-

44   

   Geography, demography, and cultural factors

sider both the internationalisation of Irish and the demographic switch by which Irish has become a second language in most communities of the Republic, we can see that while policy towards Irish remains based in history, it does not represent a return to the position of the past. The redefinition of the role of Irish outside the categories of 19th century linguistic nationalism naturally affects the position of Irish English. Just as the Constitution in the Republic makes no provision for a specifically Irish kind of English, the 1998 Agreement, which is largely concerned with national identities, makes no mention of Irish English. It is assumed in the text of the document that all communities speak English (though it is recognised that some also speak other languages) and that English is not in need of protection. Neither is Irish English mentioned in the subsequent St Andrews agreement of 2006–07, which took further steps to implement support for Irish and Ulster Scots (Agreement 2006). In this sense, Irish English remains in the ideological vacuum which was created with the Constitution of the Free State. We would be wrong, however, to take the lack of overt mention of the position and status of Irish English in language policy documents as a sign that it is marginal in Irish society. On the contrary, we should consider Bauman’s view (2001: 11–12) that once a community “starts to praise its unique valour, wax lyrical about its pristine beauty and stick on nearby fences wordy manifestoes calling its members to appreciate its wonders and telling all the others to admire them or shut up – one can be sure that the community is no more”. Understood from this view of language and culture, we can suggest that for most speakers, the position of Irish English is so secure as to require no further debate. The language seized on in the Irish literary revival and celebrated by Henry (1974) as “Anglo-Irish” did not take hold as a “third language” and has not been cultivated at L or H levels. What we find instead today is a broad family of Irish English varieties. Some varieties are strongly local, and in some areas local dialect features may be especially inclined to reflect the effects of bilingualism and language shift from Irish. Other varieties are more generally widespread. All relate in one way or another to an Irish version of international standard English which is similar to, but different from, standard English elsewhere. These different varieties of English have become internationally known through recent presentations of Irish culture in literature, film, the works of a wide range of singers, and the internet, and have made general features of Irish English increasingly available to a world audience: see Walshe (2009), for example, for a detailed discussion of Irish English in film. In ways that are not unlike Greene’s aspirations for Irish in an international context, this international exposure of Irish English has also helped to take Irish English out of the domain of local tradition and place some aspects of it into the international mainstream.



English and the Republic   

   45

Irish English in the Republic of Ireland is thus in a contradictory position. It has official status as the second national language, but it has no codified standard. Dialectal features of Irish English are widespread and socially accepted in a range of language domains, and some literary movements have celebrated these distinctive features. In policy and ideological debates, Irish English takes second place to Irish, since Irish claims symbolic value and a link to perceptions of cultural authenticity. In practical terms, however, Irish English without doubt has the dominant position, as measured both by the number of speakers and by the role of English across a wide range of H and L domains. It is to the complexity of this variety of English that we now turn.

2 Phonetics and phonology – She’d bate the hen that crowed on the turrace of Babbel (p. 199) – For spuds we’ll keep the hat he wore And roll in clover on his clay By wather parted from the say (p. 372) – So she says: Tay for thee? (p. 145) – A nigg for a nogg and a thrate for a throte (p. 374) – Lord help you, Maria, full of grease, the load is with me! (p. 214) – And why do we say that, you may query me? Quary? (p. 442) – […] Finglas fields and Santry fields and the feels of Raheny and their fails and Baldoygle (p. 142) – He wented to go (somewhere) while he was weeting (p. 223) – […] for ’tis I that have the peer of arrams that carry a wallop (p. 445) – Please by acquiester to meek my acquointance! (p. 145)  – James Joyce, Finnegans Wake (Joyce [1939] 1975)

Few works of modern literature have been analysed in as much word-for-word detail as James Joyce’s Finnegans Wake. Joyce’s use of Irish English phonology has received less attention than other aspects of his wordplay and multifaceted referencing, but it is no less significant. The ten examples cited above all hinge on a single salient feature of Irish English: the historical retention of [eː] in words, most of which used [ɛː] in Middle English, that have switched to [iː] in the pronunciations that were considered standard by Joyce’s time. The title of the work is a play on the same alternation between [eː] and [iː], since, as Wall (1986: 17–18) points out, the pronunciation indicated by could refer to a wake held after the death of an individual, or could reference the word weak, using the older [weːk] pronunciation. The first four examples above show relatively transparent respellings of beat, sea, tea, and treat to indicate the use of [eː]. The next requires the reader to interpret the spelling as [ɡreːs], to denote the word grace, thus making a connection to Hail Mary, full of grace, the Lord is with thee in the Catholic Rosary. The following example plays on the contrast between [iː] in queer and [eː] in the word quare, which should be viewed as an independent lexical item and is discussed in Chapter  4. Where we have next quoted Joyce’s list of north Dublin suburbs, we can see a subtle gradation in spellings, from and the spelling of feels which share the same meaning (with being potentially associated with both fields and feels) and the [iː] vowel, blending into fails, which introduces a new meaning and plays on the knowledge that fields is not pronounced dialectally with [eː]. The final three examples carry this theme further, using spellings for words that take [eː] in the standard language. Whether Joyce is simply making a playful reference to the interchange between [iː] and [eː] in Irish English dialect or referring to hypercorrection by speakers who are avoiding the dialectal [eː] pronunciations is open to speculation: Edgeworth



Typological overview   

   47

and Edgeworth ([1802] 1803: 148) had long ago referred to such hypercorrection in declaiming, “From all persons, of whatever sex, rank, or pretensions, who call tables teebles, and chairs cheers, good Lord deliver us!”.

2.1 Typological overview Our first approach to Irish English phonology will be to place it within the phonology of English as a world language. Some commentators have argued that Irish English phonology shows no great discontinuities with that of other Englishes: Lass (1990: 148), for example, describes Irish English “not as a ‘contact-English’ in any important sense (regardless of the fact that it began as a second-language variety), but as a perfectly normal first language, internally evolved variety, with only marginal contact effects”. Irish English certainly bears comparison with other Englishes when described in terms of the lexical sets articulated by Wells (1982). More recent overviews such as those of Hickey (2004b), Upton (2004b), and Schneider (2004) also show a great many similarities between Irish English and other varieties, while at the same time allowing that some features may be distinctive of English in Ireland or of places where Irish English has been brought. Taking account of cross-varietal comparisons, we start with a brief examination in Section 2.1 of Irish English in a typological perspective, looking not just at what makes Irish English distinctive, but at how Irish English compares with other related varieties. In Section 2.2, we will look at some more distinctive features in detail. Throughout this discussion, transcription is in accord with the Handbook of the International Phonetic Association (IPA 1999); further background may be found in works such as Ladefoged (2006) and Collins and Mees (2008), which both include recordings. We start with seven of the most commonly-cited features that can be used to classify types of English: (1) rhoticity, the quality of retaining /r/ in the syllable coda of words like car and park, contrasting with so-called non-rhotic varieties which delete /r/ in this position, (2) labiovelar fricative retention, which keeps a phonemic contrast between /w/ as in Wales and /ʍ/ in whales, (3) h-dropping (referring to the non-realisation of /h/ in word initial position in words such as Henry and hello), (4) yod-dropping, which denotes the non-use of /j/ in the initial sequences of words such as music, suit, Houston, coupon, tuna, and student, (5) use of a “dark” or velarised /l/, realised as [ɫ], especially in syllable-coda position as in milk or real, and (6) the generally predictable correlation of vowel length with quality distinctions, so that the /i/ in bead [bi:d] is predictably longer than the /ɪ/ in [bɪd] bid, as opposed to the Scottish Vowel Length Rule described by Aitken (1981), in which length is conditioned more directly by the phonological and morphological environment of the vowel.

48   

   Phonetics and phonology

Taking these features in turn, we can start by saying that Irish English as a whole is firmly rhotic, realising a variety of /r/ sounds (including [ɹ], [ɻ], and [ʀ]) in syllable codas. There are some exceptions – from both speakers who may be influenced by prestige norms such as RP and, in a different pattern, working class speakers in Dublin – but these types have not been studied systematically. With regard to working class Dublin speakers, it is possible that a simple rhotic/ non-rhotic distinction is misleading, and that there may instead be a cline of r-colouring to vowels, ranging from strong /r/ articulation through intermediate stages and ultimately to /r/ deletion. Regardless of these exceptions, there is no evidence that a general change towards non-rhotic pronunciations is in progress. Many Irish English speakers retain the contrast between /w/ and /ʍ/, though there are also pairs of words in which the contrast is lost, especially for younger speakers. A full sociolinguistic study on this point could provide details which we lack at this time, but we can be even more certain with regard to the h-dropping pattern of British English dialects, which is not a feature of Irish English. Moreover, in the Republic, the name of the letter is almost universally pronounced as /heɪtʃ/. Rahilly (2006) describes this pronunciation as being “stigmatised” in Northern Ireland, where /heɪtʃ/ is a Catholic shibboleth in opposition to Protestant /eɪtʃ/, but in the Republic, despite the use of /eɪtʃ/ by some speakers and in the H domain of broadcast news, no such stigma operates. Relative to many varieties, Irish English shows relatively little yod dropping. Words with non-coronal syllable onsets such as music, few, and cute routinely show [j] after the initial consonant, although there may be lexical exceptions: Houston may be pronounced [ˈhustən], and, in a slightly different vein, the deletion of /j/ in you is frequently cited as a feature of traditional dialect in Cork and Kerry. Henry (1958: 113), for example, cites [dųˑsiᵄ] d’you see and [haυu] have you from Glencar, Co. Kerry, [ˈletuˑ] let you from Ross Carbery, Co. Cork, and [ɪϕuˑ] if you from Ballymakeery, Co. Cork. Following coronal stops, in words such as Tuesday, student, astute, institution, due, and duke, yod dropping does not occur: a palatoalveolar consonant, giving [ˈtʃuzdeɪ], [ˈstʃudn̩t], [dʒuk], etc. is much more frequent than realisation with a glide, as in [ˈstjɪudn̩t] or [djɪuk], while outright yod dropping is rare. The coronal nasal /n/ presents a more variable picture: knew, new, newspaper, neurosis, and numerous are generally palatalised; minute is sometimes palatalised as [ˈmɪnjət]. Nevertheless we note, for example, that Leahy (1915: chapter 3, p. 4) cited [nuː] new and [nuːz] news for some Cork city speakers, Henry (1958: 113) cites [ˈnuːmərəs] numerous from Glencar and [nuː] knew from Kilcommon, Co. Tipperary, manufactured may be realised as [ˈmanəfaktʃərd], and nude usually shows yod dropping. The evidence from older and traditional dialect sources is significant, since it indicates an early basis for variable yod dropping. Fricatives and liquids typically show yod dropping: suit, assume, lute,

Consonants   

   49

illuminate, slew, and lewd, give [sut], [ˈəsum], [lut], and so on. Pronunciations such as [ˈfʌɹnɪt.jɚ] furniture (Leitrim; JK) show a split between the consonant and the palatal element: the [t] is not palatalised at all, but the glide remains in tact at the start of the next syllable. Henry’s (1958: 113) attestation of [ˈυalu] value in Glencar pushes yod dropping one step further, since this word is more usually realised with palatalisation, as in [ˈvalju] or [ˈvaʎɪʊ]. The distinction between so-called clear and dark /l/ is not usually made in Irish English, where /l/ is typically realised as clear (not velarised) in all syllable positions, yielding [lɪp] lip, [fʊl] full, and [mɪlk] milk. Words such as calm, palm, and almond in Wells’s PALM set generally do not include /l/ at all, yielding [kaːm], [paːm], and so on. Phonemic contrast between, say, palm and Pam or calm and the river Cam may be maintained through the use of a longer, slightly retracted, vowel in the PALM words. Use of velarised [ɫ] among younger speakers was, however, reported by Wells (1982: 431), citing observations by Fidelis MacÉinrí. Moylan (2009: 36), who follows work by Dónall Ó Baoill in viewing the historic lack of [ɫ] as in inheritance from 17th century English settlement, reports that the “velarizing tendency is now much stronger and seems to be spreading fast”. My experience would confirm this view, though little quantitative evidence has been presented on the precise nature of the spread of [ɫ] in social or linguistic terms. Vowel length in Irish English, apart from those areas affected by the influx of speakers of Scots and Scottish English, is typologically southern, in that length and quality usually go together. Thus while foot and food in Scottish-type English (including parts of Northern Ireland, as discussed by Corrigan 2010) can have the same short vowel, yielding [fu̶t] and [fu̶d], the southern type builds on a quality distinction and predictably has [fʊt] foot with a short vowel and [fuːd] food with a longer vowel.

2.2 Consonants 2.2.1 The labial group Most speakers of contemporary Irish English produce sounds of the /p, b, f, v, w, ʍ, m/ group in ways that are not significantly different from general patterns found elsewhere. Traditional dialect, however, presents a different picture. Henry (1957: 59), whose study of traditional dialect in Roscommon is based on fieldwork done in the 1940s, states quite simply that “in good dialect usage the variants of the f- and v-phonemes are bi-labial fricatives (ϕ, ϕ̡, β, β̡ …) as in Ir[ish]”. The use of [ϕ] extends to /ʍ/ in this data as well. Examples cited by Henry (1957: 43, 60, 69) thus include free [ϕriː]; film [ˈϕiləm]; whistle [ϕwiʃ l̡]; why [ϕo̤i]; fen,

50   

   Phonetics and phonology

when [ϕɪn]; voice [βɑːis]; moving [ˈmuːβən]; and bravest [ˈbreːβɪʃt]. Though Henry’s (1958) wider study of traditional dialect does not discuss this feature in detail, it is attested in examples such as fly [ϕl̪υɪᵄ] and calving [ˈk̹ɑːβᵞɪŋ] in Ross Carbery (pp. 113, 121). The use of [ϕ] in words like [ϕɑr] for, [ˈkɑϕən] coughing, and [loːϕ] loaf thus overlaps with [ϕo̤i] why and [ϕʷeːər] where, and can create homophones such as [ϕɪn] fen, when, [ϕiːl] feel, wheel, [ϕo̤it] fight, white, and [ϕo̤il] file, while. Henry (1957: 60) also notes the use of [β] or [βʲ] rather than [v] in [βɑːis] voice and [ˈbreːβɪʃt] bravest. This use of [β] overlaps with the use of [β] in /w/ words, as in [tβelβ] twelve and [dβelː] dwell, which Henry (1957: 36) states to “have been occasionally heard from good dialect speakers”. Though Henry (1957), noted a decline in these pronunciations, with younger speakers tending towards [f], [fʲ], and [hw] as realisations of /f/ and /ʍ/ respectively, and [v] and [vʲ] realising /v/, Ó Baoill (1990: 160) notes the continued use of [ϕ] rather than [hw] in the English of Irish-speaking districts. Like Henry, Ó Baoill draws a connection between the use of bilabial fricatives in contemporary Irish-speaking districts with a historical analysis that attributes the feature to transfer from Irish (compare also Barry 1982: 119): further historical discussion of the phonology of bilabial consonants in Irish English is provided in Chapter 5.

2.2.2 T  he coronal group The single most distinctive area of Irish English consonant phonology today concerns a set of phonemes defined by articulation with the coronal area, including the tip and blade, of the tongue: /ϴ, ð, t, d, s, z/. The purely interdental fricatives [ϴ] and [ð] rarely occur in Irish English outside of Ulster English. The phonetic realisations of /ϴ/ and /ð/ in syllable initial position usually come in three different types when the fricatives are not used: an alveolar stop ([t], [d]), a dental stop ([t̪] and [d̪]), and an affricate which typically combines a dental stop and an added fricative ([t̪ϴ] and [d̪ð]). The dental realisations are often transcribed as [T] and [D] in Irish grammatical tradition. This set of realisations is also permissible in syllable final position, though the stopping of /ϴ/ and /ð/ can also interact with phonological variation that applies to /t/ and /d/, such that glottalisation, dentalisation, and lenition (discussed below) may also apply to realisations of /ϴ/ or /ð/. As a result, the phonemic contrast between /t/ and /ϴ/, as well as that between /d/ and /ð/, in pairs of words such as tin and thin, debt and death, bat and bath, den and then, header and heather, paths and pads, or breathe and breed may be neutralised. The lexical distribution and phonological patterns by which these oppositions are neutralised or maintained are complex, and we will discuss some relevant points below.

Consonants   

   51

Realisation patterns for /ϴ/ and /ð/ depend on a mixture of geographical, social, and phonological factors. Henry’s (1957: 55–57) transcriptions in Roscommon consistently show dental realisations (here transcribed as [t̪]), as in [bəˈt̪iŋk] bethink, [ϕeːt̪] faith, and [ˈbrəd̪ər] brother, yet his more wide-ranging (1958) survey also shows alveolar realisations, particularly from southern and midland areas. Here we may note (Henry 1958: 123) such pronunciations as [nɑˑrt] north, and [səiəd] scythe from Trumera, Co. Laois; [tɑːt] thought and [dɛˑr] there from Emper, Co. Westmeath; [souət̞] south and [wiˈdout̞] without from Athea, Co. Limerick; and [troˑt] throat from Melleray, Co. Waterford. The latter word is often also realised as [troːt̪], and Ó hÚrdail (1997: 148) gives evidence of related exchanges of dental and alveolar consonants, as in [t̪ϴiːt] teeth and [t̪ϴrut] truth. Further evidence of variation comes from two transcriptions presented by Barry (1982: 129–130). In one transcription, a Cork speaker (recorded by Anthony Lunny) uses [d̪] in all 19 occurrences of /ð/, but realises the three occurrences of /ϴ/ as [wɪt̪] with, [t̪ɾu] through, and [tɾi] three. This pattern contrasts with that of a speaker from north County Dublin, recorded by Máire Ní Rónain. Here the pattern is more evenly distributed, with eight realisations of [d̪ð], six of [ð], and five of [d̪], as exemplified in [d̪ðɛər] there, [ðe] they, and [d̪ǫːz] those. The one example of /ϴ/ occurs before /r/, as a dental stop in [t̪ʰɾi] three. Detailed examinations in such transcriptions suggest that any geographical trends are no more than tendencies. Evidence from Ó hÚrdail (1997) and Hickey (2004a, b) shows that alveolars and dentals can be found in the same region, as determined by a variety of interacting social and linguistic features. One complicating factor in the realisation of coronal sounds is the frequent dentalisation of /t/ and /d/, particularly before /ər/ and /r/. This feature is frequently noted in Ulster English (see Harris 1985a, McCafferty 2007, and Corrigan 2010), but is found throughout Ireland. Dentalisation of /t/ and /d/ yields forms such as [t̪rɒχ] trough and [kast̪ərˈʀeːtəd] castrated from Glenamaddy, Co. Galway (Henry 1958: 122) alongside [ˈlad̪ər] ladder and [ˈmɑd̪əʀn̩] modern (Henry 1957: 57). The widespread pronunciation [ˈat̪ɚ] after in Dublin (JK) reflects both consonant deletion and dentalisation. Joyce’s wather ‘water’ and thrate ‘treat’ examples cited above use spellings to denote dentalisation of /t/ by the same rule. This dentalisation of /t/ and /d/ can yield results which are close to or identical to the stopping of /ϴ/ and /ð/, thus potentially neutralising oppositions between three and tree or udder and other. Maintenance of the phonemic distinction can also be found in various ways. Ó hÚrdail (1997: 140) observes Limerick speakers using [t̪] in true in contrast to [t] in through, but cites (p. 142) “an elderly speaker in Tuam”, Co. Galway, using “alveolar plosives in all cases” of /t/ and /ϴ/, except before /r/ and /ər/, where dental alternatives were used. Though these are isolated examples, they suggest two different approaches to the relevant phonemic patterns:

52   

   Phonetics and phonology

the first maps [t̪] before /r/ as an allophone of /t/, contrasting it with [t] as an allophone of /ϴ/, while the second appears to merge /t/ and /ϴ/, choosing [t̪] as the allophone before /r/ and [t] elsewhere. Additionally, Nally (1971: 36) points out that morpheme boundaries may block dentalisation, so that, for example, plaster uses dental [t̪] before [ər], but brighter does not. Wells (1982: 431) also notes that the dental feature used by some speakers in their stopped realisations of /ϴ/ and /ð/ can be lost before alveolars such as /s/ and /l/, thus yielding [feːt̪] faith with a dental, but [ˈfeːtləs] faithless with an alveolar, and allowing faiths and fates to merge in the pronunciation [feːts]. Information of this kind is only suggestive: a full picture of the social and geographical distribution of these variations can only be ascertained by more systematic fieldwork and analysis of data. Irish English is further distinguished by a unique system for the treatment of alveolar stops, especially in the realisation of /t/ in (a) single-syllable words where the preceding vowel receives stress (as in cat), as well as longer words in which the stressed vowel immediately precedes /t/ (as in about), (b) intervocalic position following a stressed syllable (as in letter), and (c) multisyllabic words, as in ticket, where /t/ is in the coda of a weak syllable which follows the stressed syllable. In Table 3 below, we will symbolise these groups by the key words BAT, BUTTER, and BUCKET. On the assumption that the phonetic realisations in question all involve a weaker closure of the vocal tract than that used in producing the alveolar stop [t], the sounds which are used are often understood by linguists to represent a process of lenition, or weakening. Though I argue in Kallen (2005a) that the term “lenition” is not well defined in phonological theory and should be used with caution in describing this kind of data, we will sometimes use it here as a simple way of uniting a disparate set of allophones under a common heading. Many speakers in Ireland, as in other parts of the world (see the reviews in Schneider et al. 2004), use realisations of /t/ such as the voiced tap in [ˈlɛɾɚ] letter and [aʊɾ əv] out of and the glottal stop in [ɡɛʔn̩] getting and [lɪʔl̩] little. Though these realisations do not feature in Henry’s (1957, 1958) depictions of traditional rural dialect, Leahy (1915) refers to the voicing of /t/ in words like Protestant and catechism, suggesting an older history for this realisation than might be expected. These allophones are now common in Irish English, but by no means universal. Unpublished studies reported in Kallen (2005a) associate the use of [ɾ] with male speakers more than female speakers: one such study of young middle class speakers (Byrne 1996) showed males using [ɾ] in 14% of possible realisations in their reading style, dropping to 5% in a more formal wordlist style. Usage of [ɾ] by female speakers in both contexts in this study was negligible. The use of [ʔ] may be sensitive both to speaker gender and to the effects of style shifting. Pardy’s (1987) unpublished study of inner-city Dublin adolescents, also reported in Kallen (2005a), shows, for example, that while female speakers used [ʔ] in 63%

Consonants   

   53

of relevant cases of /t/ in their casual speech style, this percentage fell to 30% in reading passages and 10% in wordlists. Males, on the other hand, showed only 45% use of [ʔ] in casual speech, but retained higher levels in the more formal styles, showing 38% and 16% in reading passages and wordlists respectively. While the “Westland Row” study of inner-city Dublin teenagers which Seán Devitt and I carried out in 1993 (reported in Kallen 2005a) did not offer much opportunity for evidence of style shift, it too showed a strong tendency for female speakers to favour [ʔ] relative to males in casual speech (showing 32% and 17% respectively). These differences were levelled at 25% for both groups in more formal classroom discourse (see Kallen 2005a: 67 for detail). What is most distinctive in Irish English is the realisation of /t/ as a fricative, shown, for example, in [ˈmat̞ɚ] matter and [hæt̞] hat. This sound segment is sometimes referred to by linguists as a “lenited” or “slit fricative” /t/: Pandeli, Eska, Ball, and Rahilly (1997: 68) describe it as an “apical alveolar fricative formed with a broad central channel”. Pandeli et al. (1997: 69) also note no less than 15 different symbols which have been suggested as transcriptions for this sound. The symbol used here is [t̞], which follows Ó Baoill (1990, 1997) and is recommended by the IPA (1999: 16). This notation combines the phonemic place features of /t/ with the IPA diacritic for lowering – indicating that the tongue has lowered from a potential point of contact with the alveolar ridge. A similar allophone of /d/ is found in [d̞], though this sound occurs less frequently than its voiceless counterpart. The voiceless sound was noted by Hume ([1877–78] 1878: 12) as a feature of Dublin speech, and by Leahy (1915: Section B, pt. 2,p. 3) as common among speakers in Cork city. Barry (1981b: 68, 88) viewed the use of [t̞] as diagnostic of non-Ulster speech, and while Henry (1958: 124) notes it in Co. Cavan [ˈɡʲɛt̞əm] get him, his survey shows [t̞] and [d̞] occurring no further north than Cavan, and not being found in the southwest of the country: see Kallen (1997a: xviii, 2000: 55) for maps. More recent fieldwork by Corcoran (1998), however – based on interviews with speakers from nine points in the Republic, ranging from Letterkenny, Co. Donegal and Belturbet, Co. Cavan in Ulster through to Co. Offaly and Dublin, and including Cork city and county in Munster – shows [t̞] as a feature among younger speakers throughout the Republic. Corrigan (2010: 44) notes its occurrence in South Ulster English, which may represent a recent development. Typical examples from traditional rural dialects include Henry’s (1958: 124–125) [ˈʀabət̞] rabbit (Loughrea, Co. Galway), as well as [fut̞] foot and [brɛd̞] bread from Emper, Co. Westmeath. Modern examples abound and are discussed by Wells (1982: 429–30), Kallen (2005a), Hickey (2007: 322–325), and others. The [t̞] segment is relatively rare in world Englishes, though it, or sounds closely related to it, has been noted in parts of the midlands and north of England, Scotland, Newfoundland, and Australia. This pattern of distribution for [t̞] may itself tell a story, since it is often

54   

   Phonetics and phonology

suggested that the influence of Irish immigrants could be a source for the use of the sound outside of Ireland. Discussion on this point is given, especially, by Horvath (1985), Clarke (1986, 1997b, 2010: 54, 146), Docherty, Foulkes, Milroy, Milroy, and Walshaw (1997), Chirrey (1999), Tollfree (2001), Kallen (2005a), and Jones and Llamas (2008). While the use of [t̞] as an allophone of /t/ is, as Wells (1982: 429) says, “one of the most conspicuous features of Irish English”, the use of [h] in similar positions is in some ways more remarkable, since it violates a general constraint in English syllable structure rules (see Giegerich 1992: 162–3). Henry (1958: 124–5) records various examples of word-internal [h] as in [ˈprẹhɪz] praties ‘potatoes’ (Frosses, Co. Donegal), [lɛhər͡ʒ] letters and [lihl] little (Kilclare, Co. Leitrim), [o̤ɪ ˈɡɒhət] I got it (Loughrea), [ˈlahəd ˈbɒhəmz] latted bottoms (Trumera), and [lɛˈhim] let him (Feakle, Co. Clare). The pattern of [h] use in Dublin, while phonetically comparable, diverges significantly from traditional rural dialects. In the Westland Row study, for example, [h] does not occur word-internally in words such as [ˈlɛt̪ər] letter, [ˈbɛt̪ər] better, [ˈleʔər] later, or little (realised three times as [lɪʔl̩]). The use of [h] in word-final position by speakers in this study, however, is prominent: out of 108 relevant occurrences of /t/ following a stressed vowel in a simple syllable coda (i.e. a coda with no other consonants), 35% were realised with [h]. Examples include [aˈbaʊh] about; [mih] meet and [mɛh] met, [ah] at, [ɪh] it, [ˈaʊhəyə] out of you, and [ˈʃtrihan] street and. Some words show variable realisation: of 10 uses of the word right in this sample, seven show final [h] and three final [ʔ]. To illustrate this range of variable realisations of /t/, Table 3 displays results taken from four sources: Henry (1958), the unpublished studies of Byrne (1996) and Corcoran (1998) mentioned previously, and the Westland Row study described in Kallen (2005a). This table concentrates on /t/ realisations in BAT, BUTTER, and BUCKET words in which /t/ is the only consonant following an immediately preceding vowel: we will refer to this environment as the “simple coda”. There is a theoretical controversy as to whether single intervocalic consonants following stress (as in letter) should be classed as ambisyllabic (belonging to the syllables of both the preceding and following vowel) or whether phonological processes operate to capture such consonants into the coda of the first syllable. As Harris (2010) points out, there is also a question as to whether the syllable or the metrical foot (starting with a stressed syllable and continuing until the next stressed syllable) is the correct unit of analysis in determining sites for lenition. Such theoretical questions lie outside the scope of this treatment. Table 3 simply presents the raw numbers of occurrences for realisations for /t/ in simple syllable coda position, categorised according to the BAT, BUTTER, and BUCKET environments. Other phonological environments may block the pattern we are examining here. We will not, for example, find leni-

Consonants   

   55

tion before stress (as in attack or attend), and complex codas which include following consonants (in words such as bets or width) either block lenition completely or severely limit it. Words like best, rent, or checked will not show a full range of realisations, though we can cite some use of [t̞] following /r/, perhaps especially among younger speakers, in words like heart, hurt, and parting. Dentalisation as discussed above often means that BUTTER words (such as water and letter) show dental [t̪] rather than [t̞], [h], or [ɾ], as might be expected. Connected speech may also create new environments for lenited realisations: word-final [ɾ], for example, does not occur with words in isolation, but may arise when /t/ is followed by a vowel in connected speech. NA denotes that no relevant examples are gives in the data set. Table 3: Summary of /t/ realisations in post-stress simple coda position

Study Henry (1958) t t̞ h other Byrne (1996) t t̞ ʔ ɾ Ø other Corcoran (1998) t t̞ h ʔ ɾ Ø other Kallen (2005) t t̞ h ʔ ɾ Ø other

BAT

BUTTER

BUCKET

4 19 13 2

2 5 12 4

2 3 NA NA

42 54 2 6 7 4

26 65 NA 19 NA NA

51 55 4 2 4 17

137 268 7 31 10 14 12

30 156 7 12 84 0 30

42 87 1 6 0 12 2

22 1 38 32 7 1 1

2 NA NA 2 NA NA 2

NA NA NA NA NA NA NA

56   

   Phonetics and phonology

The studies summarised in Table 3 were conducted with different methodologies. Henry (1958) relies on traditional dialect interviews and observations, and the reporting of results focuses on points of interest, rather than accounting for all realisations /t/. Byrne (1996) and Corcoran (1998) rely on results from tools such as word lists and reading passages with a finite number of sample words, while the Westland Row study relies entirely on spontaneous conversation or classroom discourse. This diversity means that no strict comparison is possible from putting these results together. Taken as a whole, however, they do illustrate, from complementary points of view, wide variation in the realisation of /t/ across specific phonological environments. The [ɾ] tap which features significantly in the more recent data is absent from Henry’s (1958) traditional dialect material. While Byrne (1996) and Corcoran (1998) both show speakers favouring [t̞] most clearly in the BUTTER group, this relative frequency is more strongly pronounced in Byrne’s results; speakers in the Westland Row study, however, hardly ever use [t̞], favouring within the BAT group [h] and [ʔ] instead. More detailed quantitative work is needed to ascertain how this variation relates to factors such as geographical and social variation, speech style, connected speech phenomena, and the possibility of change in progress. Other, less common, realisations of /t/ include [ϴ], [ts], and [ɹ], the latter of which arises from what Wells (1982: 370) terms the “T-to-R rule”. Examples of [ϴ] as an allophone of /t/, rather than /ϴ/, are reported by reported by Henry (1958: 124–127). Most of these are found in the environment of /r/, and include [ˈwɒϴər] water (Glangevlin, Co. Cavan), [sϴroːk] stroke and [ˈkoulϴər] coulter (Emper), and [dɑːϴər] daughter and [sϴreːmz] streams from Glenpipe, Co. Kilkenny. The [ts] realisation is less common, but it is significant that Henry (1958: 124–5) cited “a transitory following s” as a feature of traditional dialect, as shown in [suts] soot (Glenamaddy) and [ha>t̞s] hat (Trumera). More recently, Barry (1982: 126) noted this kind of affrication in final and initial position (as in tax, bat, did, and day), describing it as a “feature of Dublin speech [which] has spread widely among the young”, while Hickey (2007: 324) ascribes syllable-initial [ts] to “younger female speakers in Dublin”. The relationship between older dialect use and any newer developments calls for systematic investigation. The use of [ɹ] as a realisation of /t/ does not appear to be a feature of traditional dialect, but it can be heard in Dublin: [wəɹ ɪz ɪh pɛh] what is it, pet? (Kallen 2005a: 60), [ɡɛɹ ɒf wɪl jə] get off, will you! (Hickey 2007: 323), and [pʊrənaɪspæk] put an ice pack (Corrigan, Edge, and Lonergan 2012: 20) are attested examples. Though the feature is not especially common, it plays a role in language awareness in catch phrases such as excirrah and delirrah ‘excited and delighted’, which can be used by speakers who might not otherwise use the t-to-r rule. Generally associated with the north and midlands of England, this [ɹ] realisation has also been noted in Tyneside (Wells 1982:

Consonants   

   57

374, Docherty et  al. 1997), Liverpool (Knowles 1973), and Derby (Docherty and Foulkes 1999). Lenition as described here may also affect other consonants. The description of Liverpool English by Knowles (1973), for example, includes realisations of /t/ such as [t̞] (or something close to it), [ɾ], and [ɹ] via the T-to-R rule, which are also found in Ireland, but Knowles (1973: 324–326) also notes the lenition of /k/ and /ɡ/. Evidence of similar realisations in Ireland could reinforce a view that there has been regional influence between these areas. Henry (1958: 124–127) gives evidence of just such a phenomenon, notably the realisation of /k/ as either a palatal fricative, seen in [ˈkʲiçən] kicking (Glangevlin, Co. Cavan), [ə çeˑç] a cake and [baç] back (Loughrea), and [ˈçriçər] cricket (Athea), or as the uvular fricative of [ˈbuχət] bucket (Bawnboy, Co. Cavan), [ˈχruχəd] crooked (Trumera), [ə ˈχlɑχən ˈhɛn] a clocking hen (Killakee, Co. Dublin), and [ˈkruχəd] crooked (Ballymakeery). Henry (1958) also notes the use of a “fricative” form of [g]; using the IPA notation [ɣ] for this sound, we find, for example, [ɣeˑət̞] gate (Trumera), [ə ɣudˑ ˈɛɡˑ] a good egg (Baltinglass, Co. Wicklow), and [ba̘ɣ] bog (Feakle). In a similar vein, Hickey (2007: 325) cites the pronunciations [weːəx] wake and [ˈtʃɒxlət̞] chocolate from Co. Limerick. Though these uses of fricatives are not restricted to the simple coda environments which trigger so much variation in the realisation of /t/, they suggest a pattern of lenition which was formerly more widespread in Irish English than is generally the case today.

2.2.3 Palatals and velars The discussion of yod dropping can be related to wider issues in the use of palatal and velar consonants. Burke (1896: 698) used spellings such as opportunity and fortune to denote the use of velar consonants in Meath, Kildare, and Carlow; he held these pronunciations distinct from those indicated by and which he described as otherwise “universal”. These observations are supported by Henry (1958: 127–9), who associates this use of velars with the north midlands, but admits that it may be found elsewhere. Forms he cites include [ˈkriʃkən] Christian and [ˈkwɛˑʃkən] question (Emper), and [ˈnæːkərl] natural (Bawnboy). Nally (1971: 37) also cites /k̹uːn/ tune and /ɡ̹uː/ dew from Westmeath. Henry (1958: 129) notes a reverse process, whereby velars in certain environments can become alveolars, as in [tlutʃ] clutch (Emper) and [tlamp] clamp (Bragan, Co. Monaghan). Though Wright (1905: 68) stated that this use of initial [tl] had not been adopted in any Scottish or Irish dialects, he did note it as a feature of northern, midland, southern, and southwestern dialects of England, and further claimed that “it occurs as an individualism among educated people in all parts of England”.

58   

   Phonetics and phonology

While the advancement of velars before high front vowels in English (as in key and kit vs. car and cool) is common in English generally, palatalisation before low vowels such as /æ/ and /a/ is less common. Corrigan (2010: 46) notes [kʲ] and [ɡʲ] before low front vowels in Mid Ulster and Southern Ulster English, though her contemporary data show it as being absent among Belfast females and male speakers of Ulster Scots. In traditional dialect, Henry (1958: 115) describes it as a “Northern and Eastern feature fading to the West and South”. Though the feature is today associated only with areas near or close to the Ulster dialect zone – it is, for example, typified in the pronunciation [ˈkʲævən] Cavan  – Henry (1958: 115) demonstrates its sporadic existence across a wider span, including, for example, [k̹a̘tl̹] cattle (Nobber, Co. Meath), [kʲaˑr] car and [kʲaˑf] calf (Emper), [kʲapš] caps (Trumera), and [kʲaˑrd̥] card (Kilmore Quay, Co. Wexford). Both Henry (1958) and Ó Baoill (1997) relate this kind of palatalisation to a wider pattern, in which the consonants of English reflect phonetically what is a phonemic division in Irish between palatalised and velarised consonants, seen in minimal pairs such as [bʲoː] beo ‘alive’ versus [bᵞoː] bó ‘a cow’. Evidence of this kind suggests that a much more geographically and phonologically widespread system of palatalisation and velarisation may have operated at one time than can be found today. Similarly, palatalisation of /s/ and /z/ to [ʃ] and [ʒ] in stop, stick, sleep, or rottenest, cost, fist, etc. is frequently ascribed to the influence of Irish, on the assumption that English consonant phonemes – notably /t/ and /d/ but in some cases /p/ and /k/ as well – were associated by Irish speakers with their corresponding Irish palatal phonemes. The palatalised pronunciations of English would thus arise from the operation of a rule of Irish phonology which requires all consonants in a syllable onset or coda to share the same value for palatal or velar. English /t/ would thus dictate the use of [ʃ] in adjacent position, and would call for the use of [ʒ] with English /d/. Thus Leahy (1915: Section B, chap. 8, pp. 6–9) describes the use of [ʃ] in combinations of /sl/, /sn/, /sm/, /sp/ and /sk/, illustrated by [ʃlip] slip, [ʃnɛːl] snail, [ʃpeɪːk] spike, and so on. Henry’s (1957: 52, 61–2, 230–3) observations in Roscommon give us particular insight, because they illustrate the lexically irregular nature of the process. Palatalisation is usual before /l/, /n/, and /t/, as in [ʃtik] stick, [ʃtei] sty, [ʃlutː] slut, [ʃnoː] snow, [nɪʃːt] nest, [ˈleʃnən] listening, [ˈdaʒlər] dazzler, [ˈɡuʒl̩] guzzle, and [ˈrɑʒn̩] rosin. Henry (1957: 52fn) says that “the combination st is not used by good dialect speakers”, though Henry’s data also show palatalisation being blocked when /st/ is followed by /r/ as in [st̪rap] strap. The picture with /s/ before /p/ and /k/ is less predictable. Though Henry (1957: 52) implies that palatalisation before /p/ is possible, he gives no such examples, citing instead words such as [speːd] spade and [spuːn] spoon. Words with /sk/ include palatalised [ʃkeːm] scheme ‘play truant’, [ˈϕʷɪʃkər] whisker, and [ˈʃkilət] skillet, in contrast with non-palatal [ska̘ːld] scald, [skoːp] scope, and [skutː] scut.

Vowels   

   59

2.3 Vowels The lexical sets systematised by Wells (1982) provide a way to compare and contrast different varieties of English, linking present-day variation to the history of the language. A catalogue of lexical sets and their expected vowels, however, runs the risk of oversimplifying the situation, because there can be considerable variation within each set, measured both by the phonetic realisation of particular phonemes and by the membership of words in particular sets. Nevertheless, this section starts with a general overview of the most widely-used realisations for each of Wells’s lexical sets. Sets as defined by Wells are given in capital letters, followed by words which illustrate the set. The most common vowel realisations are given next, though it must be emphasised that it would be impossible to account for all the existing phonetic variations in this overview. (See also Hickey 2004b for a different, but related, approach to Wells’s lexical sets in Irish English.) Most of the comments which follow are based on Irish English outside the Ulster dialect area: Ulster features such as the fronting of /u/ and /ʊ/ to /u̶/ and use of the Scottish Vowel Length Rule are discussed in Corrigan (2010). High vowels FLEECE creep, deep, seen, grieve, she, Easter, machine, police NEAR cheer, here, ear, fierce, weird, nearly, serious, steering KIT ship, tip, give, mix, king, busy, women, system, rhythm GOOSE loop, too, moon, you, lute, duty, coupon, grew, beauty CURE boor, poor, allure, ensure, gourd, tourist, rural, Euro FOOT put, butcher, cuckoo, wool, good, wolf, woman, should

iː iːr ɪ uː uːr ʊ

Mid vowels (may sometimes be lowered) FACE name, patron, station, spectator, beige, weigh, eight, steak eː ~ eɪ SQUARE mare, bare, hair, pear, where, scared, Mary, dairy, fairy eːr DRESS press, kept, method, bread, realm, friend, merry, Terry ɛ~ ɛ̞ GOAT boat, bone, home, wrote, no, poultry, bold, though, sew oː ~ oʊ FORCE more, hoarse, door, four, court, divorce, moral, glory oːr ~ ᴐr NORTH or, for, origin, York, horse, morning, quart, warm ᴐr ~ ɒr ~ ɑr THOUGHT taught, daughter, ought, Paul, autumn, tall, chalk, salt ᴐː ~ ɒ ~ ɑ CLOTH off, cough, moth, boss, Ross, coffee, sorry, origin, orange ᴐː ~ ɒ ~ ɑ STRUT cup, bus, shrub, money, wonder, love, double, blood ʌ ~ ᴐ̈ ~ ʊ NURSE hurt, shirt, Byrne, assert, person, earn, heard, work ʌr ~ ɛr ~ ᴐ̈r

60   

   Phonetics and phonology

Low vowels (START may be raised) TRAP apple, cab, glad, tax, happen, marry, lapse, panda, Pam BATH after, path, dance, grant, aunt, calf, half, ask, castle, fast LOT stop, pot, job, god, box, solve, Tom, swan, waffle, watch PALM calm, balm, palm, psalm, father, lager, almond, soprano START far, bar, sharp, part, garb, heart, farm, safari, Sahara Diphthongs PRICE right, write, nice, mind, sign, island, buyer, sky CHOICE void, joist, poison, noise, join, coin, oil, boy, toy MOUTH doubt, loud, south, house, gown, round, how Unstressed vowels happY cosy, seriously, calorie, taxi, nasty, coffee, money, abbey lettER paper, figure, murder, sugar, teacher, anchor, Esther, acre commA sofa, arena, opera, Gina, drama, Brenda, mania

æ~a æ ~ aː ɑ~ɒ aː ~ æː aː ~ ɑː ~ ɛː aɪ ~ ɑɪ ~ əɪ ᴐɪ ~ ɒɪ ~ ɑɪ aʊ ~ ɑʊ ~ɛʊ i ~ iː ɚ ~ ɐr ə~ɐ

2.3.1 V  owels and lexical incidence The use of lexical sets in phonology always involves a tension between the phonetic realisations which are found in particular sets and the assignment of words to the sets. Words of the type book, hook, cook, nook, took, crook, and brook, as well as hooker (a type of boat associated with Galway) and cooker (which can refer either to a kitchen appliance or a cooking apple), may all be pronounced with [uː] by some Irish English speakers, though there are many speakers who use [ʊ] for all of them. We could either say that these words add phonological variation within the FOOT set (which is otherwise realised with /ʊ/), or we could select these words and put them into a lexical subset, which we could label as the BOOK subset. This subset is also identifiable in dialects of the North of England. With the subset approach, we could then map the lexicon and phonology more clearly, and say that some or all of the BOOK words may be assigned to the GOOSE set (using /u/) for certain speakers. Since Irish English has a phonemic contrast between /u/ and /ʊ/ and since speakers appear to use /u/ or /ʊ/ in the BOOK set for each word independently, we will understand the differences between speakers as a matter of lexical assignment, and say that the words of the BOOK subset are all variable in their assignment to the GOOSE or the FOOT set. The lexical element in vowel assignment is sometimes very clear: a speaker may, for example, use [ʊ] for cooker to denote a cooking appliance, but [uː] in cooker to denote a cooking apple. (The word snooker, on the other hand, is universally assigned to the GOOSE set in Irish English.) Assignment

Vowels   

   61

of BOOK words to the GOOSE set is more common in traditional dialect and older urban varieties, and the direction of change seen in younger speakers is towards assignment to the FOOT set. Even though most GOOSE assignments for this lexical subset may be recessive, the category is still productive: I have heard a native Galway speaker assign such recent international imports as cookie and the film title Tootsie to the GOOSE class with [uː], in essence treating them as part of the BOOK subset. Similar questions are raised with assignments to the DRESS and TRAP sets. The DRESS set is realised in a largely uniform way with [ɛ] or a more lowered realisation with [ɛ̞]: the precise social and phonological distribution for the lowered variant, which may come close to [æ] and appears to be more common among younger speakers, awaits further study. Within the DRESS set as a whole, though, we may also note two other groups: one in which the vowel may be raised to [ɪ], especially preceding nasals but also before some other consonants (discussed in Section 2.3.3 below), and a second containing words such as any, many, anyone, anyway, and anything where [a] is used instead of [ɛ]. Since this latter group is relatively small and exerts little influence on the system of phonemic oppositions (pairs such as any and Annie being limited), it may be simpler to assign these words, where applicable, to the TRAP set rather than to identify a subset in which to contain them. The FOOT and STRUT sets also represent potential contrasts that can be realised in different ways. If we take /ʊ/ as the basic realisation in the FOOT set, and a lower vowel such as [ʌ], [ᴐ̈], or [ə] as the target for STRUT words, there is a very good chance that any speaker of Irish English will display a phonological contrast between two lexical sets. The choice of words which illustrates this contrast may, however, vary considerably. Pairs such as put and putt could both be pronounced with [ʊ] or with [ʌ], or the two words could be distinguished by a contrasting choice between the two vowels. A speaker who does not distinguish between put and putt may nevertheless distinguish, for example, between could and cud, or use different vowels in some combination of words such as cuddle, budget, puddle, butcher, would, should, woman, or wool. Harris (1996: 15–17) details the development of the split between PUT words, which have tended to retain high vowel realisations, and CUT words, which have seen vowel lowering since, roughly, the mid 17th century. In reviewing data from Irish English outside of Ulster, Harris (1996: 22–23) takes the view that systems here show a mix of features, lacking “a full-blown PUT–CUT split” but displaying partial “lowering of historically short u”. In the absence of large-scale systematic study of the pattern in this area, we can illustrate it with a few examples. Bertz (1987: 39) views the opposition between look and pull with [ʊ] and luck and dull with [ə] as highly variable in Dublin: in

62   

   Phonetics and phonology

the informal speech of speakers with primary school education, the potential contrast is “always neutralized” (implicitly, in favour of [ʊ]), while speakers with second-level education show “free variation between neutralization and opposition”, and the formal speech of those with higher education “never neutralizes”. A more complicated picture comes from Henry’s (1957) treatment of material from Roscommon. Here STRUT words are generally realised by [ᴐ̈], especially following or preceding labial consonants, as in bubble, cup, cupboard, humble, rub, summer, stubble, up, couple, flood, plobber ‘large fat head or face’, and something. The few non-labial words with [ᴐ̈] include colour, hundred, and sun. Unrounded central [ə] is indicated in brother, other, and us. Lexical assignment, however, has put book, bosom, cook, cookers, rook, and souple in the GOOSE set with [uː], while nothing, once, and one use the [ɑ] of LOT, and onions shows the raised vowel [ɪ͂] commonly used in KIT words. Harris (1996: 22) shows yet another approach, based on unpublished work by Anthony Lunny on the English of west Cork. As seen below, words in this area can be grouped into [uː], [ʊ], and [ᴐ̈] realisations, with inconsistency in the phonological conditioning of each. [uː] book move noon

cook coop tooth

[ʊ] drum up some son

grub uncle but

[ᴐ̈] look took full cud up would

could wood stood flood sugar

Harris’s (1996: 24) suggestion that this mixture of features in the PUT–CUT split in Irish English reflects an early stage of historical developments that have proceeded further in other English varieties is, I think, compelling. Lexical assignment of the low vowel sets is complicated by the lack of minimal pairs in this area. Lexical contrasts such as Sam and psalm or Pam and palm are relatively rare, which makes it difficult to know how contrasts are constructed on the basis of vowel backness or length. Phonological diversity in some familiar words can make it difficult to assign a word to a single set: father can readily be found with any of [æ], [a], [ɑ], or [ɒ] (all of which are usually long); Henry (1957: 26) also cites father with [ɛː]. The LOT and START lexical sets include words which use the back vowels [ɑ] or [ɒ], but START, especially, also includes [a] realisations as in [ˈɡardn̩] garden, [hart], heart, etc. Words of the TRAP, BATH, and PALM group are usually realised with [æ] or [a]: pronunciations such as [ɡrɑːs] grass and [dɑːns] dance in the BATH set are extremely rare in Irish English. In all these sets, vowel length may also be a factor, with vowels of the PALM group often

Vowels   

   63

being longer than those of the other groups. For many speakers, minimal pair contrasts are thus maintained between [pam] Pam and [paːm] palm or [sam] Sam and [saːm] psalm. Here we may also note the traditionally widespread pronunciation of the first letter of the alphabet as [a] (hence in the TRAP set), which has now been replaced for most speakers with the /e/ of the FACE set.

2.3.2 V  ariation in the front vowels: MEAT and CLEAR Words such as meat, speak, and leave (usually spelled with ) had the vowel /ɛː/ in Middle English. These words contrasted with Middle English /eː/ words such as meet, keep, and sweet (where spellings are common). In what Wells (1982: 193) refers to as the FACE merger, some words with /ɛː/ and /eː/ became merged into one category to be found in the FACE set (generally using /eː/). Words such as wait, day, grey, whey, weight, and eight are also included in this group. By a further process which Wells (1982: 194–195) terms the FLEECE merger, other words with Middle English /ɛː/ (such as meat, speak, and leave) merged with /eː/ words (including keep, green, and meet), but further raised to /iː/ and form the main basis of the Present Day English FLEECE set. The Irish English retention of /eː/ in words that have otherwise joined the FLEECE set with /iː/ is one of the most salient features of traditional dialect: it is this feature which our selections from Finnegans Wake address. Though this feature is now recessive, it is worth exploring in some detail. In order to do so, we will identify four further lexical subsets. It will be useful to distinguish between a MEAT subset, potentially realised with /eː/ and frequently (though not exclusively) derived historically from Middle English /ɛː/, and a KEEP subset, which includes other sources of the /iː/ pronunciations in FLEECE words, including Middle English /eː/. We will also make a parallel distinction between a CLEAR subset that includes Middle English /ɛː/ in words such as fear, ear, and weary, and is distinguished from the DEER subset that forms the rest of Wells’s NEAR set. Sheridan’s suggestions (1780: 59) to “attain a just Pronunciation of English” note that Irish speakers “pronounce the words tea, sea, please as if they were spelt tay, say, plays; instead of tee, see, pleese”. Sheridan noted, however, that in “combinations of ee and ie, the Irish never mistake; such as in meet, seem, field, believe”. Patterson (1860) listed over 100 words from Belfast in which [eː] was commonly used instead of the [iː] which he felt to be correct. This list shows little evidence of localisation to Belfast, and includes many words which would plausibly have fit on a similar list from other parts of the country at the time: Patterson’s [eː] words include eat, bean, steal, cream, dean, eagle, bleach, measles, fever, either, seize, intrigue, and scheme. Leahy’s (1915: section B, chap.  V, pp. 5–9)

64   

   Phonetics and phonology

analysis went a step further, in characterising the use of older [eː] in terms of what we would call social variation. Leahy (1915: section A, chap. VII, p. 4) identified three classes of speaker in Cork city: Class III, the most prosperous and elite group (descended from later Protestant settlers and Catholics who had become prosperous in the 18th and 19th centuries), Class II, the “badly educated or illiterate descendants” of Old English Catholics and late 17th-century native Irish migrants to the city, and Class I, descendants of suburban dwellers of the 18th century and more recent migrants from the countryside. Class I speakers would be expected to show the strongest effects of transfer from Irish. Citing sea, deal, heat, wheat, and clean as examples, Leahy shows Class I speakers using the older [ɛː] in all such words (thus [sɛː] sea, etc.), while Class II speakers use a “frontal” and “close” [eː]. Class III speakers in Leahy’s account show the later-developing [iː]. Noting the use by all speakers of [iː] in words such as sweet, field, and we, Leahy further noted diversity in related words of our CLEAR and DEER subsets. Class III, which had adopted the FLEECE merger in using [iː] for the MEAT subset, is seen to use [ɛː] in both [klɛːr] clear and [pɛːr] peer, thus retaining and extending the older vowel realisation in words before /r/. Class II in this account similarly merges the two vowels, but uses [eː] for both. Class I reflected the diversity of change in progress, showing both the [ɛː] used by Class III and the [iː] later associated with the NEAR set. No group in this account uses vowel differences to contrast clear and peer. Later work by Henry (1957, 1958) gives evidence of both /ɛː/ and /eː/ realisations within the MEAT and CLEAR subsets. The /ɛː/ realisations cited by Henry (1958: 110) all come from Northern Ireland (as in [bɛˑət] beat from Tartaraghan, Co. Armagh), but since Henry’s published samples are not exhaustive, we should not rule out such possibilities in the Republic. Henry’s (1957) Roscommon study includes at least 36 words which use [eː] in the MEAT or CLEAR subsets; [ɛi] is used in one cited pronunciation of me. MEAT words of this kind include clean, dream, east, heat, increased, pleased, sea, sheaf, speak; conceit, decent, eye, key, seized, sty, and wheat. The smaller CLEAR subset with [eː] includes beard, ear, fear, and queer. Henry (1957) also shows variation within these sets: dream and fear, for example, are realised with [iː] as well as [eː], while words such as team and weave, which could potentially be realised with [eː], are found exclusively with the [iː] of FLEECE words. Further examples from Henry (1958: 110) include [hweˑət] wheat, [meːəl] meal, and [kreˑəm] cream (Kilmore Quay), and [beˑəʀd] beard (Glencar). Since the word listings in Henry (1957, 1958) are not exhaustive, it is not possible to quantify the extent of variation which may exist within the original data, but this material gives ample attestation of [eː] pronunciations at mid-century. We should also note that some words which belong to the FLEECE set for historical reasons have migrated to the MEAT subset with [eː] for some

Vowels   

   65

speakers. Leahy (1915: section B, chap. IV, p. 9) noted that while all classes used [iː] in sweet, field, and we, the word feet showed [iː] only with Class III: Class II used [eː] and Class I [ɛː]. Ó Grianna (1918: 3), an Irish-speaking writer from the Donegal Gaeltacht, also argued firmly that his mother pronounced sleep as “slape”. There is no doubt that words of the MEAT and CLEAR subsets are dwindling in favour of /iː/ and /iːr/ in the contemporary language. The widespread and salient use of the older forms, however, has left Irish English with a dialectal legacy which many speakers are aware of and may use for various purposes. Some such realisations have taken on an independent lexical existence: bate (