- Shopping Bag ( 0 items )
1 The emergence of overseas varieties of English
It is probably true to say that mainly regional forms of English were taken to the colonies which England founded in the core 200-hundred-year period between the early seventeenth and the early nineteenth century. Those who served in the overseas settlements were very largely from the lower ranges of society, irrespective of whether one is talking of deportees in early Australia, indentured servants in the early anglophone Caribbean, emigrants and adventurers of various sorts in many other colonies, the sailors who worked on the ocean-going ships, or the bailiffs and other members of the colonial service industry. The only people from the educated middle classes and higher would have been senior officials in the administration, clerical and educational staff or army officers stationed overseas. Given this situation, any treatment of colonial English is likely to be concerned with varieties which are not similar to, or even near, the current or recent historical standard of British English, even granting that the notion of 'standard' had a less clear profile in previous centuries than it does today.
The present book sees its justification in a number of aims which have been set by the editor and the contributors. The first is the attempt to bring into focus just what input varieties were probably operative in individual colonies. The second is to examine the extent to which dialect mixing and/or language contact have been responsible for the precise structure of overseas varieties in areas with multiple immigration patterns. The third aim is to attempt an evaluation of the different reasons for extraterritorial varieties having the form which they show. Dialect input is only one source for colonial English, as shown in the following list of factors determining its shape:
Dialect input and the survival of features from a mainland source or sources.
Independent developments within the overseas communities, including realignments of features in the dialect input.
Contact phenomena where English speakers co-existed with those of other languages.
An indirect influence through the educational system in those countries in which English arose without significant numbers of native-speaker settlers.
Creolisation in those situations where there was no linguistic continuity and where virtually the only input was a pidgin, based on English, from the preceding generation.
In the study of varieties of English, linguists have sometimes favoured one of the above explanatory factors to the exclusion of others. But even a cursory glance at the forms of English overseas shows that accounting for their structure means taking more than one factor into account and according them relative weight on the basis of considered linguistic arguments. This attitude characterises many of the contributions of the present volume which attempt to afford dialect input arguments greater weight in the discussion of the genesis of overseas varieties and so redress an imbalance which they perceive in the linguistic descriptions of the varieties they are involved in without, however, seeking to adopt an ideological standpoint in favour of dialect retention in extraterritorial forms of English.
1.1 What constitutes dialect input?
In an investigation of the nature of the present one a major concern is determining just what constitutes dialect input to extraterritorial varieties. A simple starting point would be to contrast unusual features in these varieties with those attested in present-day British dialects and simultaneously consider whether there is historical continuity between the area in Britain where a feature or features is/are found and the overseas site at which these seem to reappear. This task is not as easy as might be imagined. Even in an area like Newfoundland (Clarke, this volume) which was for a considerable time isolated from the rest of the anglophone world and which has only two dialect input sources - South-East Irish English and South-West British English - the matter is far from decided as there was interaction between the two mentioned areas in the British Isles (Hickey 2002a). Furthermore, one finds dialect spread across communities overseas. Indeed there may well have been internal migration in an overseas region, such as the Caribbean, which opens up the possibility of features diffusing there, for instance it is conceivable that features of Irish English spread outward from Barbados (and Montserrat) when speakers from here shifted to different parts of the Caribbean, travelling as far as the south-east coast of the United States (Montgomery 2001: 129).
Another approach in determining what features may have existed in English formerly and been transported overseas involves the examination of historical documents. Here one can avail oneself of the many more or less prescriptive commentaries on English from the early modern period - going back to the sixteenth century - and comb through them for mention of features not part of present-day standard British English but attested overseas. Such sources range from occasional comments in travel literature (such as that by Fynes Morison for Ireland in his Itinerary of 1617-26; see Hughes 1903) to whole glossaries of provincialisms (cf. collections of words from the dialect of Forth and Bargy by Vallancey and Poole; see Dolan and Ó Muirithe 1996). Two other significant sources are more or less genuine representations of rural speech in fictional literature and more pointed caricature of accent such as the satirical treatments of Irish English in Restoration drama (Bliss 1979).
Of greater value are dedicated works on English pronunciation and grammar. For instance, for Ireland, Thomas Sheridan's Rhetorical Grammar of the English Language (1781) is an important guide to pronunciation in Dublin in the late eighteenth century. Here he admonishes 'the well-educated natives of Ireland' for various traits of speech which he regards as nonstandard (Sheridan 1781: 146). By these means a glimpse of contemporary usage is gained. Similar prescriptive works exist for speech in Britain for the entire early modern period.
Among the other sources of data being utilised by linguists of late are emigrant letters. These are documents written usually by individuals without too much education and hence without overt conceptions of correct English. They often provide glimpses of what the speech of emigrants was like as the correspondents were in most cases individuals who had already left their country, settled overseas and were writing back to those still at home. Such emigrants' letters have been examined by Michael Montgomery in his investigation of Ulster Scots influence on early American English (Montgomery 1995, 1997a) and similar collections of letters have been used when examining Irish English in the south of Ireland by Markku Filppula (Filppula 1999) and the present author (eighteenth-century letters illustrating southern Irish English of that period). In addition to this there is a study of regional British English as attested in emigrants' letters; see Bermejo-Giner and Montgomery (1997).
In the nineteenth century there are a number of studies which were written as a consequence of the rise of linguistics as an academic discipline. Such works are frequently diachronic in nature and offer insights into conservative speech in Britain which may well have been taken overseas to the colonies. A notable instance of such a source is the comprehensive work on English pronunciation by Ellis (5 vols. 1868-9) and of course the invaluable dialect grammar and dictionary by Wright (1905).
1.2 The relative value of dialect features
When considering possible British/Irish sources for overseas varieties one must not be misled by similarities which are so common that they are of little value. There is a general principle that the more widespread a feature the less it is indicative of a connection between homeland and colony. An apt example of this is diphthong flattening (Wells 1982: 149f.), by which is meant that the movement of the tongue at the end of the rising diphthongs /ai/ and /au/ is much reduced, if not entirely absent, hence one has pronunciations like wife [waːf] and house [hɑːs]. This phenomenon is very common indeed; it is found in South Africa and the southern United States, two regions which are definitely not linked historically, as well as in the north of England, e.g. faan [faːn] 'fine' (Ihalainen 1994: 213). Other instances are final cluster simplification, particularly postsonorant deletion, as in mend /mɛn/ and the alveorisation of /ŋ in unstressed syllables, typically in the progressive form of verbs, e.g. talking /tɔːkn / or the assimilation of sibilants to nasals as in wasn't [wɒzṇt] → [wɒdṇt], found in south-east Ireland and parts of the southern United States (Troike 1986). It is the very general nature of such features which diminishes their diagnostic value when considering historical connections.
The opposite case, so to speak, is represented by camouflaged forms which are dialectally significant but often difficult to recognise as they show a surface similarity to constructions found in more standard varieties of English but are used differently. Two examples can be taken to illustrate this. In African American English come with V-ing, as in She come acting like she was real mad, looks like a normal case of the verb of motion but this is in fact a special use as a kind of auxiliary verb indicating indignation on the part of the speaker (Wolfram and Schilling-Estes 1998: 173; Spears 1982). The second instance, from Irish English, is the use of never to mark the past as in She never called us which does not have the meaning 'She did not call us on any occasion' but refers to a particular occasion which is obvious if a specification of time is added (optionally): She never called us that evening (common in northern England and in Scotland as well).
Dialect features are characterised not only by presence or absence but by relative frequency. The consideration of frequency has led to many insights in recent years, especially on the level of syntax. For instance, in their study of relatives, Tottie and Rey (1997) and Tottie and Harvie (2000) found that the lack of wh-relatives and the frequency of zero relatives points to the dialect background of African American English (Tottie and Rey 1997: 244) and shows a system not unsimilar to that of Middle Scots examined by Romaine (1982).
Furthermore, early attestations of overseas varieties can be illuminating in this context. Howe (1997: 267ff.) maintains that earlier African American English (as incorporated in the ex-slave recordings collected in Bailey et al. 1991) is more conservative than modern African American English and more akin to nonstandard southern white English, setting itself off from creole patterning in this respect.
1.3 Internal ranking in dialects
Dialect features show internal ranking, that is, not all features are of equal significance for the status of the dialect. Some are group-exclusive, i.e. a community of speakers uses a variant which is not found in adjacent communities. Within a community dialect features can of course be unevenly distributed; in general, those associated with sections of the community far removed from the standard are taken as highly indicative of that community in that they contribute significantly to its unique profile.
Furthermore, an implicational relation may hold in a dialect too, that is, the presence of one feature may imply the presence of another much as with implicational universals in phonology such as voiced stops implying voiceless stops in any given language. For instance, if a speaker of Irish English has the habitual aspect as in She does be home of a Saturday, then he/she is certain to have the immediate and resultative perfectives as in She's after eating the cake and She has the book read, respectively.
Features which show implicational relations are usually those which are markers in the sociolinguistic sense, i.e. they tend to disappear from speakers' speech on style shifting upwards. As such they tend to play a role in perceptual dialectology (how speakers themselves see a dialect) and surface in linguistic stereotypes (Hickey 2000a).
1.4 Linguistic constraints on variability
The use of features in a dialect may be subject to constraints on variability. On the one hand there are independent constraints which are traceable to some extralinguistic factor, ultimately of social origin. But there are also linguistic constraints on variability such as the relatively rare occurrence of final cluster reduction before a vowel, e.g. find [faɪn] but find out [faɪnd aʊt] or the nonoccurrence of diphthong lowering before voiceless segments in Canadian English (tight is [təɪt] and not [taɪt] although tide is [taɪd]). The instance of final cluster reduction just quoted provides a good example of this kind of variability (see Wolfram and Schilling-Estes (1998: 249-56) for a detailed discussion of this in American English). In the history of English one can see that sequences of bilabial nasal and voiced stop have not existed for centuries, i.e. comb is pronounced without the final /-b/ (Jespersen 1909: 216). If the nasal is followed by a voiceless stop then it is retained, e.g. damp with final /-p/. Stops following velar nasals have experienced a similar development: voiceless stops are retained, cf. think, but voiced ones are generally lost in word-final position, cf. sing [-ŋ] (the stop can, however, be retained in north-western forms of British English; see Upton and Widdowson 1996: 34f.). Deletion after velar nasals is not always the case word-internally, contrast singer [-ŋ-] with longer [-ŋg-]. However, the cases which are linguistically interesting from the point of view of present-day varieties are those where an alveolar nasal is followed by a voiced stop. Here the stops are realised in standard English, e.g. cold, card, wind, all with final /-d/. In relaxed colloquial styles the final voiced stop can be deleted when followed by a further stop, e.g. cold meat [-lm-], but there are dialects where this deletion holds for careful styles, e.g. Dublin English. In these cases the stop deletion is a dialect feature and not just an aspect of fast speech.
1.5 Dialect survival
There is a characteristic topography which goes with dialect survival overseas. In general, inaccessible, mountainous or isolated coastal regions keep the features which were characteristic of the input varieties. Appalachia and Newfoundland are two classic examples of this kind of situation as the Ozark Mountains region. Indeed, there may well be interconnections between such regions as Christian, Wolfram and Dube (1988: 2) postulate for Appalachia and the Ozarks (see map in Carver 1987: 119; he notes, for instance, the occurrence of poke 'bag, small sack' in the Appalachians and the Ozarks, see pp. 176f.). The Outer Banks of North Carolina provide an example of an isolated coastal region with dialect features not found in mainstream varieties of American English (Wolfram and Schilling-Estes 1997: 5-15; this volume). Features in such areas tend to be retained which are not necessarily characteristic of the country they are part of. For instance, rhoticism - the Southland 'burr' - in the Otago region of the South Island of New Zealand is not typical of the rest of the country. Such locations exist in the contemporary anglophone world and may have existed historically, but have since disappeared, e.g the baronies of Forth and Bargy in the extreme south-east of Ireland (Hickey 1988).
1.6 Dialect diaspora
Movement away from one area to a smaller, more remote one is what one can term 'dialect diaspora'. This situation is found in a few cases in the anglophone world and has been the subject of investigation by a number of linguists (notably Shana Poplack, Sali Tagliamonte and John Singler for diaspora forms of African American English). The linguistic interest of such areas derives from their separation from the core area and hence their lack of participation in later developments in this latter area. A case in point is offered by the Americana settlement in Brazil which consists of African Americans who left the southern United States in the wake of defeat after the American Civil War (Montgomery and Melo 1990: 195). Certains features which are regarded as prototypical of present-day southern United States speech, such as diphthong flattening in the PRICE lexical set,1 are not found here. The conclusion which can be drawn is that this phonetic feature is a recent phenomenon, postdating the movement of African Americans to Brazil. Indeed, researchers like Guy Bailey are of the opinion that diphthong flattening is a fairly recent phenomenon (Bailey and Ross 1992: 528; Montgomery and Melo 1990: 206-8).
There are other African American diasporas, notablly on Samaná peninsula in the Dominican Republic and in Nova Scotia (Poplack 2000: 4-10; Poplack and Tagliamonte 2001: 10-38; 39-68). Data from these locations form a central object of investigation in the chapter by Poplack and Tagliamonte, this volume. The return to West Africa by African Americans in the newly founded state of Liberia in the nineteenth century (it was proclaimed a republic in 1847) has been investigated by John Singler along with the development of African American English after this displacement from the core area in the southern United States (Singler 1991). On the repatriated African Americans in Sierra Leone, see Ehret (1997: 174-6). For an examination of the language of an expatriate community of African Americans in Sierra Leone on the basis of letters, see Montgomery (1999).
Dialect features can also offer information about migration routes within a country. In the movement of African Americans from the south to the north in the United States there were two basic streams, one which involved African Americans from North and South Carolina moving up along the coast to Washington, DC, Philadelphia and New York, and one which involved those who took a midwestern route up into St Louis, Chicago and Detroit. It has been noted (Wolfram and Schilling-Estes 1998: 115) that the latter group are less likely to show the shift of [ð] to [v], as in brother [-v-], smooth [-v], than are their counterparts at eastern seaboard locations.
1.7 Ebb and flow
When viewing dialect survival in the context of the current book, a number of caveats are called for. Perhaps the most important involves what the present author in another context has called ebb and flow (see Hickey 2002b). By this is meant that certain features which appear to be historical continuities in a remnant area may well not represent a straight line of development from the earliest days of the dialect. An instance (from Britain) which shows this clearly is velarised /l/ in syllable codas. This existed in Old English and was responsible for the breaking of vowels as in West Saxon eald 'old' /æəɫd/. Syllable-final velarisation continued into Middle English and led to vocalisation of the lateral, something which is still obvious from the spelling of such words as talk or walk in present-day English which retain the /l/ in writing. But the velarisation of /l/ in popular London English, cf. milk [mɪʊk], would appear to be a recent phenomenon which was not common in the nineteenth century. The explanation for this would seem to lie in the pendulum movement among speakers' preference for a velarised /l/ in their speech. A swing of the pendulum in favour of this realisation can be seen in present-day Southern Irish English where pronunciations like field [fɪːəɫd] are becoming increasingly common (as attested amply in the recordings for A Sound Atlas of Irish English), although traditionally Irish English has been known for an alveolar /l/ in all syllable positions (Hickey 1986a).
There are various motivations for the phenomenon of ebb and flow. It may occur between generations of speakers and, if general across a broad section of the population, it may become established in the speech community. It may also be the result of dissociation between groups in a society where the preference of a feature by one group may lead to its being avoided by another (Hickey 2000b).
1.8 False leads
Another caveat concerns features which seem to have a single identifiable source. The clarity of such cases often masks other sources which might be considered. A case in point is a-prefixing as in They were out a-playing on the strand. Some authors have pointed to parallels in Irish and Scottish Gaelic (Majewicz 1984) in which there is a structural parallel; consider the Irish rendering of the English sentence just given: Bhí siad amuigh ag imirt ar an trá [was they out at playing on the strand]. But this obvious parallel would appear to be coincidental. The structure a-V-ing is well attested in British English during the colonial period, deriving historically from on V-ing with phonetic reduction of the preposition on much as in asleep from an earlier on sleep. This may well be the source for those varieties of American English which show this structure as Montgomery (2000), who is sceptical of the Celtic origin, rightly points out. In addition, such text collections as A Corpus of Irish English (Hickey 2003c) has only a very few attestations in the many historical texts for Irish English which it contains.
1.9 The likelihood of sources
Competing sources for dialect features require that one considers more general aspects of language development in trying to reach a decision about which source is the most likely in a particular situation. An example of this is provided by vowel epenthesis in Irish English and Afrikaans English. The epenthetic vowel in question is a shwa in words with final /-lm/ clusters, i.e. with heavy codas consisting of more than one nonhomorganic sonorant, hence film is typically [fɪləm]. Branford (1994: 486) in his discussion of English in South Africa mentions the presence of the same feature in Irish English and suggests that it might be a source. But the number of Irish settlers in South Africa was only about 1 per cent, so hardly significant in the genesis of varieties of English there. However, Afrikaans shows a similar epenthesis and studies of the geographical distribution of epenthesis (Hickey 1986b) confirm that it is a low-level phonetic phenomenon with a typically areal spread, for instance it is found in Dutch and in the adjacent German dialects of the northern Rhineland. Its occurrence in Afrikaans - as a transported feature of Dutch, of course - would suggest its appearance in South African English is the result of an areal spread from the former language, given the close contact between Afrikaans and English in South Africa.
1.10 'Colonial lag'
Historically, commentators on varieties of English outside Britain tend to highlight their conservative nature. For the dialect of Forth and Bargy, mentioned above, there are remarks from as far back as 1577 by Richard Stanyhurst on the similarity between that variety and Chaucerian English which for Stanyhurst would have been a vague reference to an antique form of English (Miller and Power 1979). Latter-day writers refer to the language of the Elizabethan era or to that of Shakespeare and frequently claim that dialects tend to maintain this still (there are many such references to Irish English, for example, and to Appalachian English; Montgomery 1998, 2001: 107-9). Precisely what such labels mean is frequently not specified, the power of the argument seems to derive from its very vagueness. Nonetheless, a certain antiquity is the point being made and the situation where colonies seem to fall behind developments in the mainland is often labelled 'colonial lag' (Görlach 1987).
But a closer look at allegedly conservative dialects reveals that they are not simply preserved versions of earlier forms of the language on the mainland but have themselves gone through processes of their own. Such processes can be inherited, i.e. overseas varieties continue processes initiated at their historical source (Branford 1994: 477). This is clearly the case with the raising of short front vowels in varieties of English in the southern hemisphere. Furthermore, varieties at new locations obviously undergo independent developments which may be triggered by language/dialect contact or result from internal motivation within the language or triggered by the new society using it. In addition, the specific nature of an overseas variety may rest substantially on dialect mixture, given settlers from different regions. In such cases the attention of linguists has rested on the nature of the mixture and the results it engendered; see the contribution by Gordon and Trudgill in the present volume.
1.11 Distributional patterning
Recent literature on varieties of English has concentrated on elements which were inherited by forms of the language which arose at new locations. Ongoing changes, such as the raising of short vowels just alluded to or the lowering and retraction of diphthongs, also to be found in the anglophone southern hemisphere, are just two examples of features inherited by varieties arising overseas. Another aspect of this complex is whether the realisations and rules are categorical or variable in their application. Furthermore, there may be hierarchies of constraints which are to be found with realisations and these may reflect the situation in the source dialect, indeed such hierachies may be the clearest indication that a certain dialect is the source for another, as Poplack and Tagliamonte have shown conclusively in their work of this subject (see Poplack and Tagliamonte, this volume, and the contributions in Poplack 2000).
1.12 The neglect of distinctions
Finally, one can mention that the neglect of distinctions, present in more standard forms of English, can be characteristic of a particular variety. A clear example of what is intended here is provided by the use of the so-called 'extended present' of Irish English (Filppula 1997). By this is meant the use of a present form of a verb to encompass an action which stretches back into the past. In such cases, for instance in sentences with the temporal adverbial since, e.g. He has been here since we moved to Dublin, English requires the present perfect. However, Irish English only uses the present and so neglects the tense distinction found in standard English, e.g. We're living here for ten years now. A significant source for this usage in Ireland (it is also found in Scotland) may well be Irish where an equivalent to the present perfect of English does not exist.
1.13 Folk dialectology
The last topic to mention in this section is what has come to be known as folk dialectology. By this is meant examining how nonlinguists conceive of dialect distinctions and dialect areas. Preston (1993a: 338-44) gives an analysis of 138 south-eastern Michigan respondents' outlines of the dialect area 'southern' in the United States. The broad view was generally correct but it covered a large area, much greater than what linguists would regard as 'southern'. The respondents differed according to class affiliation and age. Both younger and lower-middle-class respondents regarded 'southern' as covering a larger area than did older and upper-middle-class respondents. Preston (1993a: 344-56) also conducted investigations into what nonlinguists viewed as areas of 'correct' speech and of 'pleasant' speech.
The value of such investigations lies in the information it gives us about speakers' attitudes to varieties other than their own. This in turn can help in accounting for such sociolinguistic movements as accommodation (Trudgill 1986) and dissociation (Hickey 2000b) and ultimately assist linguists in explaining externally motivated language change.
2 The spread of English
The dissemination of English beyond the island of Britain has a history which is over 800 years old, beginning with Ireland in the late twelfth century. The early settlement of Ireland by the English has largely been a matter of internal concern for scholars engaged in Irish English but it is of interest here because it is the earliest example of language mixture involving transported English and the insights gained here are of relevance to the examination of later instances of English at overseas locations. The situation of imported English to Ireland in the late twelfth century (from west Wales) provides information about the planting of English into a multicultural context in which Norman French was the superstrate and Irish the substrate of the host country (Hickey 1997a; this volume: chapter 3).
For the establishment of forms of English outside Europe the early settlement of Ireland is also of considerable significance. In many instances the route for English across to the New World was out of Ireland rather than directly from England. This began with the deportation of politically undesirable Irish to Barbados in the early 1650s, continued with the departure of religious dissenters from Ulster to North America, chiefly Pennsylvania, throughout the eighteenth century and also involved the seasonal migration of Irish to Newfoundland on the eastern coast of Canada up until the first decades of the nineteenth century (Hickey 2002a). In numerical terms the major Irish exodus was in the middle of the nineteenth century with large-scale emigration to the eastern United States. Each of these situations resulted in different kinds of linguistic influence in the host regions and are dealt with in the relevant contributions in this volume. Furthermore, the transit route taken by immigrants is of significance. For instance, many of the Irish travelling to the New World used Canada as a point of entry and passed through the Ottawa Valley on their way to the United States (Charbonneau 1997; O'Gallagher 1984).
|1||Dialects of English and their transportation||33|
|2||Scots and Scottish English||59|
|3||Development and diffusion of Irish English||82|
|4||The emergence of American English : evidence from seventeenth-century records in New England||121|
|5||The language of transported Londoners : third-person-singular present-tense markers in depositions from Virginia and the Bermudas, 1607-1624||158|
|6||Remnant dialects in the coastal United States||172|
|7||Back to the present : verbal-s in the (African American) English diaspora||203|
|8||'Canadian dainty' : the rise and decline of Briticisms in Canada||224|
|9||The legacy of British and Irish English in Newfoundland||242|
|10||The English dialect heritage of the southern United States||262|
|11||Solving Kurath's puzzle : establishing the antecedents of the American midland dialect region||310|
|12||English dialect input to the Caribbean||326|
|13||South African English||363|
|14||English transported to the south Atlantic Ocean : Tristan da Cunha||387|
|15||English on the Falklands||402|
|16||English input to Australia||418|
|17||English input to New Zealand||440|
|18||English input to the English-lexicon pidgins and creoles of the Pacific||456|
|19||Englishes in Asia and Africa : origin and structure||503|
|20||South Asian Englishes||536|
|21||South-East Asian Englishes||559|
|App. 1||Checklist of nonstandard features||586|
|App. 2||Timeline for varieties of English||621|
|App. 3||Maps of anglophone locations||627|
|Glossary of terms||654|