Reviews Links & Letters 3, 1996 159 comes to teaching and expanding gram- ported here. The top-down approach sug- mar, vocabulary and with gested by discourse anaiysis by no means advanced students. Here a discourse rules out the need to adopt a bottom-up approach promises to make useful con- approach with its focus on lexico-gram- tributions to help teachers to become maticai forms when this helps the learner more aware of precisely why such learners to acquire more linguistic competence. remain very non-native-like in their In fact, elsewhere Cook has argued for a linguistic production. Modality (in place for a focus on form through repeti- grarnmar), relexicalisation (in vocabulary) tion and learning by heart, and this is and prorninence (in phonology) are just thought-provoking coming from a spe- three areas that spring to mind after ciaiist in discourse analysis! Indeed, it is reading McCarthy. probably true to say that McCarthy hirn- Before finishing, 1would like to make self would seem to be a proponent of a a criticism of what is generdy a very help- baianced approach to teaching ful and practical book. Whereas the book and learning. We can learn a lot from dis- does contain some hints for classroorn course analysis, but we do not have to practice and references to published rnate- throw the baby out with the bathwater. riais, these are not exemplified in the text. John Bradbury Also, does McCarthy only have know- Departament de Filologia Anglesa ledge of materials that have been publis- Universitat Autbnoma de Barcelona hed by the same publisher? Would it not have been more useful for teachers to see References examples of a variety of activities from different sources and then helped to assess COOK,Guy (1989). Discourse. Oxford: Oxford these in light of the insights discourse University Press. analysis throws up? An earlier book for -(1994). «Repetition and learning by heart: teachers on the same topic by Guy Cook an aspect of intirnate discourse, and its implicxions~.In ELTJoumal Volume 48,2. does this much better. COULTHARD,R.M. (1985). An Introduction To surn up, this book provides the to Discourse Analysis. London: Longrnan. language teacher with an excellent intro- MCCARTHY,M. and O'DELL,F. (1994). duction to what discourse anaiysis is and English Vocabulary in Use. Carnbridge: its relevance to his or her teaching situa- Cambridgt: University Press. tion. The thinking practitioner is encour- REDMAN,S. aid ELLIS,R. (1989). A Way with aged to pick and choose what he or she Words Book l. Carnbridge: Cambridge rnay find useful from the research re- University Press.

Suzanne ROMAINE.Language in Society. An Introduction to Sociolinguistics. Oxford: Oxford University Press, 1994. xv +235 pages.

In the preface, Romaine explains why she work is definitely that while the field of called her book language in society as sociolinguistics is a multi-faceted one, no opposed to hnguage aandsocies, this being subdivision can be made that does not «to emphasize the fact that the study of look at the socio and linguistic as a holis- society must accord a place to language tic concept. within it at the same time as the study of The first chapter is an introduction language must take account of society.~ to this idea. In it the terms language and (p. ix ). The underlying rnessage of this dialect are introduced and lively exarn- 160 Links & Letrers 3, 1996 Reviews ples are given to illustrate how not only horne language, terrns traditionally used linguistic but prirnarily social criteria in surveys and censuses where societal rnust be used to decide what constitutes a bilingualisrn exists, which usuaily take the language and what a dialect. Indeed the nation-state as a reference point and are terrn variety is preferred as it avoids dern- carried out, ironically enough, by govern- ma by definition. Political, cultural and rnent bodies. As is rnade clear frorn the geographical factors are al1 at play and abstract at the beginning of Chapter 2, ultirnately rnight even be the deciding fac- taken frorn Genesis 1 1: 1-1 1, rnultilin- tors over the linguistic. One's language gualisrn is still seen as the curse of Babel, or dialect in the traditional sense are ter- something divisive and therefore detri- rned as one's regional dialect and indica- mental to a unified power. While no alter- te the speaker's geographical origin; social natives are offered to the terrninology of dialect or accent tells us what a person's these questions, particular insight is rnade status is within that society, and register into the way the typical census questions reflects what we are doing and where. The are answered. When asked which is their notions of speech cornrnunity and corn- , subjects may, subcons- rnunicative cornpetence are also introdu- ciously or not , choose languageslvarie- ced in this chapter, with the forrner being ties with which they wish to be identified, defined as «a group of people who do not be it for prestigious, ethnic, religious, poli- necessarily share the sarne language, but tical or other reasons. These same choices share a set of norrns and rules for the use are rnade in everyday life. The author goes of language.))(p. 22) and the latter as «the on to discuss domains and the individual's conventions they share about their spe- use of a particular language according to ech variety~(p. 22). These conventions the given variables; for example interlo- refer both to grarnrnatical knowledge and cutors, place, topic and on a societal leve1 appropriateness of use, which is usually diglossia, where «each language or variety culturally transrnitted. The significance in a rn~ltilin~ualcornrnunity serves a of one's culture is stressed here, as is the specialized function and is used for parti- human trait of classification in order to cular purposes.)) (p. 45). She explains understand the world we live in. We use how, in cases of bilingualisrn without language to help us understand basic con- diglossia, the cornpetition for status as cepts such as love, work, time and rnoney official language and the choices rnade in and thus construct our lives accordingly. everyday life by those who are free to The way we classify depends on our cul- choose, can lead to language shift and ture, on the rnetaphors we live by and on ultimately language death. She also looks our belief networks. briefly at code-switching and the variety In the second chapter, Rornaine of reasons for which a speaker rnay do so. moves on to language choice in a bilin- In the chapter that follows Rornaine gual or rnultilingual context. In sorne takes a more in-depth look at language cases severa1 languages coexist in relative and social class. First, she develops the harrnony, in others the rnarginalisation idea that ((urbanizationtends to prornote of rninority languages and peoples leads linguistic diversity as well as uniforrnity)) to bitter controversy. It is the author's (p. 68) with the standard language unify- airn in this chapter to examine the bilin- ing a diverse population. She then reviews gual speaker's reasons for switching or severa1 well known studies on social choosing one variety over another. She stratification and linguistic variation, begins once again by casting doubt on the such as post-vocalic Ir/ in New York City. narrow definitions of rnother tongue and The class of the speaker, the arnount of Reviews Links & Letters 3. 1996 1G1 attention paid to speech, accornrnodation indeed ridiculous, but what Rornaine is made to the hearer, gender and age, advocating here isn't a burn-your-bra type are al1 factors which have been shown to linguistic revolution, but an intelligent condition individual variation. Then she transition to a language of equals. She touches on social nenvork studies which highlights the fact that this transition can have grown out of the desire by sorne only be made if society changes too, and researchers not to take for granted that that what rnust be sought first is a posi- individuals can be necessarily grouped tive self-irnage arnong wornen. She also into social classes. This kind of study has focuses on single-sex and mixed-sex con- shown that a person's socialising habits versational interaction and provides lively have a direct effect on the way shelhe examples from the field of conversational speaks, and that can understanding of analysis which show us that not only the network structure leads to insight in the words we use but the way we speak is cul- process of language shift in a bilingual turally determined. cornmunity.)) (p. 83). It is argued that the In Chapter 5, the author examines enforced standardization of a language linguistic change and its social rnotiva- promotes the class division referred to and tions. She starts by looking in sorne detail ties in with the earlier idea of our need to at the isoglosses separating the Low and classify in order to understand the world High varieties of German, and shows how we live in. After looking at colonisation in a historical sense linguists have long from the point of view of the European been interested in the developrnent of lan- as linguistic imperialist, the author guage, primarily in geographical terms, concludes that «To establish a linguistic but that as in the case of the German order is to declare a social order» (p. 95) sound shift al1 rnay not be as clear cut as whereby the literate are empowered with the early studies claimed and that other a standard language with which they factors, such as sociolinguistic patterns govern the norrns and therefore create and time, must be considered. She rnakes vast inequalities. constant reference to the ideas of «change

In Chapter 4 the author turns her frorn above and changeU frorn below to attention to some of the research findings refer to the differing points of departure related to language and gender. She begins for the dihsion of linguistic innovations by claiming that the early research done in through the social hierarchy.)) (p. 140 ). this field was rnistaken in its approach, For exarnple, a contrast is made between one which saw wornen as being different the non-rhotic pronunciation of post- from rnen, just as in the Bible wornen are vocalic Ir/ in RP British English," a chan- seen as derivative of men. Rornaine argues ge which originally seems to have come that this received wisdorn is reflected in from below, and the rhotic speech now our narning practices and is ((symbolicof prestigious in the United ~titesarnong an order in which men come first)) the rniddle-class which can be seen as a (p. 103). Her angle is that being fernale change from above. The dihsion of post- or male is not only genetic but learnt too, vocalic Ir1 in the US and Britain is also i.e., we learn through socialization to be given some consideration frorn a histori- what we are. She goes on to look at spe- cal and geographical perspective. A brief cific examples of sexist narning language, look is taken at how relatively linguisti- and offers alternative politically correct callv innovative rnen and women are re- terrns. This language reform has been crit- spectively with regard to linguistic change icized and even ridiculed, and frankly and we see from research findings that it some of the suggested terms we see are depends very rnuch on the speech com- 162 Links & Letters 3, 1996 Reviews munity and networks. Romaine then not been recognized as languages in their devotes the rest of the chapter to change own right, usually because they have not in the address , focusing her atten- been standardized. But the uses of pid- tion in particular on the variable and gin and creole languages can be extrernely somewhat arnbiguous usage of the polite varied, and we see, by virtue of a rather forrn ni in Swedish. The criteria used in nice poem written in Hawaii Creole Sweden when choosing which form to English, how the people who speak them use being rnany; age, social class, gender are are fiercely proud of them. and so on. The penultimate chapter focuses on Chapter 6 is dedicated to pidgin and some of the language related problems Creole languages. As always, the author that can arise in school, an institution begins with a loose definition of the which endorses largely middle-class lan- terrns, being careful not to be prescrip- guage and values. This of course is a prob tive. We learn that the term creole is lem even for working-class children usually given to a pidgin which has be- belonging to the dominant culture, but come nativised, and that a pidgin comes is an even bigger problem for minority about in situations where speakers of dif- children among whom the rate of failure ferent language backgrounds need to is considerably greater. Bilingualism has cornrnunicate. First, sorne consideration generally been regarded as a problem as is given to the distribution of pidgins and we have seen, and this negative attitude creoles, that being rnainly in the Third has determined the likelihood of failure. World. Most are based on European lan- Romaine examines and contends the guages, that is they draw the bulk of their deficit theory, and looks at what happens vocabulary frorn these languages, which when there is a rnismatch in styles and are sornetimes referred to as the super- expectations between child and teacher. strates. The grarnrnatical structure is She also takes a critical stand to the way usually influenced by the substrates. The testing ignores the needs of minority chil- name of a pidgin or a creole thus usually dren. Considerable space is given to the indicates their location and their principal history of court action and subsequent lexifier; however these names are often legislation in the United States aimed at unsatisfactory and not always used by the rectifying these linguistic inequalities. She speakers of these languages thernselves. contrasts imrnersion and submersion pro- There is a convenient division of the lan- grarns and concludes that the choice is guages into two groups: Pacific and very much a context-dependent one. Atlantic. But it is shown that there was Finally, she introduces the idea of semi- some interaction between the two, and lingualism, the new deficit theory. And that rnany terms can be found both in so we turn full circle and come back to Pacific and Atlantic pidgins or creoles. the question of cornpetence. She dismis- The author then discusses various hypo- ses the idea that it is sornething that can theses on the origins of pidgins and be rneasured in terms of full or partially creoles, and their structure. She talks full and clairns that cthere is an inverse about the processes of creolisation and relationship between what can be easily decreolisation and introduces the con- measured and assessed quantitatively and cepts of basilect, mesolect, and acrolect. its irnportance for effective comrnunica- Finally, she looks at the languages in their tive skills.» (p. 21). She states the need social context, and at their ovenvhelming for a new definition of language profi- lack of official status. This seerns to be a ciency and questions the traditional policy consequence of the fact that they have of eradication of the minority language Reviews Link; & Letters 3. 1996 163 in favour of the majority one. Her opin- are looking for an introduction to ion is clearlv that of those who believe in sociolinguistics, as the title suggests. Any maintenance programs: the right to one's potentially problematic terms are clearly own language being a basic human one. and carefully defined and plentiful Throughout the book both current examples are given to illustrate the point and early research findings are quoted and in question. The author clearly states at a subtle but strong emphasis is put on the beginning of each chapter the matter empirical studies, suggesting that the to be dealt with, so in this respect it is author is well aware of the opinion held also a book one can dip in and out of. In by some that sociolinguist