Date: Tue, 07 Dec 2004 17:07:55 +0100 From: Philip Shaw Subject: New-Dialect Formation: The Inevitability of Colonial Englishes
AUTHOR: Trudgill, Peter TITLE: New-Dialect Formation SUBTITLE: The Inevitability of Colonial Englishes PUBLISHER: Edinburgh University Press YEAR: 2004
Philip Shaw, University of Stockholm, Sweden
This monograph argues that under certain circumstances the formation of a new variety of a language (a 'new dialect') is deterministic, that is, it depends only on the dialect characteristics of the initiators and not on extraneous factors like status and identity. Support for this thesis is largely from a corpus now known as ONZE -- Origins of New Zealand English (hereafter NZE) -- which derives from the spoken reminiscences of some 325 first- generation Anglo New Zealanders (born in New Zealand between 1850 and 1900) recorded between 1946 and 1948. Trudgill argues that the corpus represents an almost unique record of the moment before a new variety is born. Some support for Trudgill's thesis also comes from other varieties of English, particularly other Southern Hemisphere ones with nineteenth- century roots. The findings of the ONZE project are described thoroughly in another publication of which Trudgill is an author -- Gordon et al 2004 -- and this book aims to use those findings to make a general point.
The first chapter establishes that most colonial varieties of European languages (and of Hindi) derive their distinctive features in part from dialect mixture among the founding settlers. The similarities among the Southern Hemisphere varieties are due to similarities in the dialect mix among the original settlers, who came from (Southern) England, Scotland and Ireland in that order of frequency. To these observations Trudgill adds the evidence that new dialects are formed by quite young children who are relatively impervious to considerations of status or prestige. Hence, Trudgill can argue that under tabula-rasa conditions (i.e. where there is no local variety of the incomers' language) new-dialect formation is deterministic, affected only by the initial linguistic mix.
The second chapter aims to establish relevant features of nineteenth- century British English, to see how it differed from early NZE, which seems to have become focused around 1900. The first generation of 'colonial' children has no peer-group variety to adopt, and thus follows its parents' variety more closely than elsewhere, so that sound changes in progress are held up for a generation. As a result the ONZE sample includes features which would have died out in the speech of the sample's contemporaries in Britain and Ireland. Thus the ONZE sample can be used, along with written sources on nineteenth-century English, to define the English of the first settlers, and the chapter proceeds to reconstruct the accents spoken by those settlers. The argument can be exemplified by considering h-dropping, the 'English' English feature by which word-initial /h/ in stressed content words is lost, so that 'ill' and 'hill' are homophonous. In the Southern Hemisphere the phenomenon is nowadays absent or rare. In the ONZE recordings about 25% of speakers exhibit h-dropping, which suggests that it was less common in nineteenth- century English varieties than it is today. This is corroborated by dialect studies over the last hundred years, suggesting that only a minority of the first settlers exhibited h-dropping, which meant, deterministically, that the feature did not survive into modern New Zealand speech.
The remaining five chapters develop the proposed model of new-dialect formation, referring to Trudgill 1986. This involves six key processes, which Trudgill illustrates in chapter 3 from New World Spanish and French and Fiji Hindi: mixing of people with different dialects; levelling, in which demographically minor variants are lost; unmarking -- linguistically unmarked forms survive in preference to marked ones; interdialect development -- new forms arise from the interaction of dialects; reallocation, where each of two surviving forms becomes specialised to a different allophonic or social niche; and focussing, in which the new variety becomes stable, with its own norms.
These processes operate over three stages. Chapter 3 describes Stage I of the process, rudimentary levelling in the speech of the first settlers. The speech of the ONZE informants does not include instances of many localised phenomena (such as the merger of /v/ and /w/) which must have been in the speech of their parents -- the first settlers. Trudgill infers that these features were levelled out early because of their extreme minority status, perhaps especially if they had a low status, since among these adults status would presumably be an operative factor. At this stage interdialect development also occurred, the ONZE tapes including for example "a high level" of hypercorrect initial /h/ and even /hw/ in words with no historical /h(w)/ like 'apple' and 'witch'.
Chapter 4 describes the speech of Stage II, represented by the ONZE corpus speakers themselves, who do not sound like each other or like modern New Zealanders. Because there was "no common peer-group dialect for them to accommodate to", parental varieties had an unusual influence on first- generation children's speech, but even so they were not followed precisely. The ONZE speakers, in fact, combine features from various input varieties in very unusual ways. They also vary very much in their own speech, with vowel realisations varying apparently at random between values typical of different British varieties. The third way in which they vary is that individuals born and brought up in the same place speak in very different ways. It seems that the accommodation which caused rudimentary levelling at Stage I did not operate at Stage II. However there was 'apparent levelling' in which some fairly widespread features of nineteenth-century British English (such as Scottish centralised KIT, and Northern English retention of the same vowel in FOOT and STRUT) failed to make it into the ONZE varieties. Trudgill argues that this is not because the features had been diluted by accommodation among adults, as in Stage I, but because a feature had to have a given threshold frequency in the input data (of the whole community, not just the parents) to be adopted by children.
Chapter 5 describes Stage III in which koinéisation (levelling, unmarking, reallocation) and focusing deterministically formed NZE from something like the variable and varied speech of the ONZE interviewees. The second generation of New Zealand-born English speakers were faced with fewer variants and simply selected those that were most common, which, coincidentally, were often, but not always, typical of the South East of England. Where non-southeastern forms were in the majority in the ONZE- type varieties, they were chosen for NZE. Examples are h-retention and retention of front realisations of START/PALM. Sometimes the majority for the Southeastern- type form that survived was very small: for example, modern NZE consistently has a rounded vowel (like RP) in LOT words even though only 53% of ONZE informants used such a vowel. A small majority this size did not always guarantee focussing. In ONZE 57% of speakers used a short vowel in CLOTH words but in modern NZE both long and short vowels can be heard in this set, according to Trudgill. Moreover, other factors than majority status also operated. About two-thirds of ONZE informants had the shwa/I distinction in unstressed syllables('Lenin' = 'Lennon'). Nevertheless NZE does not, showing that "unmarkedness may sway the balance in favour of [large] minority variants" (120).
The Southern Hemisphere Englishes are similar to one another because a similar mix of actual dialect features was available in the feature supermarket (Trudgill's equivalent of Mufwene's feature pool), but also because of drift -- potential for change inherent in the starting system. In Chapter 6 Trudgill gives a number of examples of parallel developments taking place after the varieties had split, such as HAPPY tensing (tense [i] rather than lax [I] in unstressed open final syllables) and Glide Weakening (centering of the endpoint of rising diphthongs).
Chapter 7 makes explicit the point that the previous chapters have made implicitly: social factors can be dispensed with. Like Labov and Croft, Trudgill argues that adaptation and the speech of the individuals in contact are all that is needed to explain language change. The drift phenomena cannot be the result of the prestige of British varieties because some have not affected Britain, and the developments have generally gone further in the Southern Hemisphere than in Britain. Proponents of an explanation based on prestige would also have to explain why other prestige innovations in Britain have not been adopted in NZE. Similarly stigma cannot have been strong in the relatively egalitarian conditions of New Zealand, even less so before compulsory schooling in 1877, so it cannot be invoked as the cause of the failure of h-dropping to survive. Nor should we consider identity or ideology as possible causative factors in this situation. Why should New Zealanders talking to other New Zealanders need to signal their identity? Trudgill accounts for the unity of NZE by the mobility of its speakers.
The book's argument is convincing and Trudgill's style is very clear and unpretentious. Nevertheless the reader can have a feeling that Trudgill's agenda is not completely overt. The reasons may be his strategy of not foregrounding the context of his argument and some unclarity about the strength of his claim.
The first means that he appears to presuppose readers who need the short- vowel phonology of varieties of English explained in detail, but are so familiar with current developments away from purpose-oriented explanation in sociolinguistics that they can piece together where Trudgill's ideas fit in from passing references to Labov and others. This reader at least was frustrated not to have a broader view of such trends presented here, particularly as it would have helped to differentiate this book from Gordon et al 2004.
The strength of the term 'deterministic' is not clear. It is not formally defined, and sometimes it seems to mean only that social factors -- speaker intentions of some sort -- do not apply, while at other times it seems to mean that the outcome is wholly determined by the initial condition. The case for the second meaning is not made, as there seems to be some chance involved as well as in the cases of unmarking and variability in CLOTH words, contrasted with deterministic elimination of variability in cases like LOT.
Trudgill uses a methodological criterion that one should not invoke speaker intentions, status, identity, etc. unless one has to. But he is not necessarily consistent in eliminating unnecessary entities. He gives both rudimentary levelling at Stage I and the threshold effect at Stage II as explanations of the absence of features in the ONZE data, even though the evidence for both entities is the same, and seems to justify this by saying the processes are plausible (as of course they are), without showing that they are necessary.
Once one admits a process just because it is plausible one is laid open to arguments that social factors are plausible too. Trudgill does not explain why drift failed to carry h-dropping forward, as it did Diphthong Shift, or why the unmarking involved in h-loss did not overcome statistical tendencies, as it did for the unstressed /I/- shwa merger. In fact the frequency of hypercorrect /h/ insertion among the ONZE speakers and the coincidence of loss of h-dropping with the rise of schooling in the last third of the nineteenth century would provide evidence for an argument that h-dropping was stigmatised. This argument might seem as plausible as the deterministic one for this feature, and if one is to reject it on methodological grounds the rigour has to be consistent
Nonetheless the case that prestige and identity can be largely dispensed with in this context is made. On that basis the book is very successful. It is written with Trudgill's habitual lucidity, handles masses of data clearly, reveals very wide reading, and, as I hope my summary shows, makes a generally very impressive case. Certainly one finishes it convinced that the story presented here of the development of NZE is more convincing than any of its rivals. Trudgill's abilities applied to such wonderful data as the ONZE material have resulted in a fascinating book which enlarges our understanding of the processes which created colonial Englishes.
Gordon, Elizabeth, Lyle Campbell, Jennifer B Hay, Margaret Maclagan, Andrea Sudbury, and Peter Trudgill. 2004. New Zealand English: Its Origins and Evolution Cambridge: Cambridge University Press
Trudgill, Peter. 1986. Dialects in Contact. Oxford: Blackwell.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Philip Shaw has degrees in English and linguistics from the universities of Oxford, Reading, and Newcastle upon Tyne. He has taught at universities in Thailand (Chiang Mai, Silpakorn), Germany (Bonn), England (Newcastle), and Denmark (Aarhus School of Business), and is currently working in the Department of English at Stockholm University. He is co-author (with Gunnel Melchers) of 'World Englishes: an Introduction' (Arnold 2003) and has published widely in journals. He is interested in the structure and uses of English worldwide, particularly in business and academic settings and across cultures.