Libri

F. A. J ohansson: Immigrant Swedish . A Study in Multiple Contact Analysis. Travaux de l’Institut de Linguistique de Lund 9. Glccrup, Lund 1973. 180 pp. F aith A nn J ohansson’s thesis for a doctorate1 investigates the Swedish pronunciation of 180 immigrants from nine different . Her 20 informants from each represent the following languages: American English, Czech, Danish, Finnish, Greek, Hungarian, Polish, Portuguese, Serbo-Croatian. The investigation is preceded by a survey of the basic thoughts of contrastive analysis and error analysis (pp. 10-40). These pages are well written and offer a comprehensive bibliography although the important contributions of the German conferences on ‘Contras­ tive Grammar’ in 1969 and 1970 arc not mentioned. The survey is longer than needed in the context of this thesis but it can be read as an instructive introduction by anybody interested in the matter. To avoid interference from written Swedish, the author has chosen an imitation test. The informants are requested to say words and sentences after a Swedish model. The informants, who are at a rather early state of learning, make the impression (at least on the present reviewer) of not being noticeably influenced by the model in the way they pronounce the words and sentences. It seems to require so much of their concentration just to understand what is said that they have no possibility of catching and imitating the nuances of the model. That is why they just seem to provide their normal pronunciation. Be this as it may, the actual pronunciations seem to include just about every imaginable mistake. The author chooses to make a rather rough kind of auditory analysis. The consonants and the stressed of all speakers are considered (but details of various kinds are not included). Statistics of the results of the error analysis are given for the speakers of all the nine languages mentioned above. These results, as well as the phonetic sketches of the ten languages involved, will be of considerable use to teachers of immigrants in Sweden. (In recent years Sweden has become one of the major countries of immigration.) The first of my two major criticisms of the book would be that much lime could have been saved if fewer than 20 speakers from each language had been investigated. If one has heard a couple of speakers from the same source language attempting to speak the same target language (unless they have considerably different ), one has the feeling that one has heard them all. Surely, one does not need 20 individuals to find all errors of importance. Nor should it be of real importance that statistics on 20 informants arc slightly more reliable than statistics on a smaller number, for instance five. The considerable time savings which could have been made could have allowed the author to study in depth various problems concerning her informants. She could have chosen among the following possibilities, (a) A test at which the informants would have had to speak freely instead of imitating could have been added (as pointed outG armng by E. at the public discussion of the thesis) although it is possible that little new would have appeared in comparison with the imitation test (sec my feeling about this test as expressed above), (b) The notations could have been narrower, (c) The vowels and consonants could have been studied in a more detailed way including allophones, unstressed phonemes, phono- tactic facts, phoneme transitions and the length distinction, (d) To some extent, not only ‘segmcntals’ but also ‘suprasegmcntals’ could have been studied, (e) The exact pronun­ ciations of the immigrants when speaking their own languages could have been studied in

1 The thesis satisfies the new and, in some respects, less demanding regulations for a Swedish doctorate. (The former ‘fil. dr.’, which was preceded by a ‘fil. lie.’, was comparable to the higher doctorate of some countries.) At the public discussion of the thesis the present reviewer was the ‘opponent’ appointed by the University of Lund. 254 Libri order to make more detailed comparisons possible, (f) Different stages in the language acquisition of the immigrants could have been studied, (g) While it is true that the author reports that she has compared some of her notations with those of another notator and found the result ‘satisfactory’ (p. 49), it would certainly have been much better if she had done a bit more in this regard, (h) The acceptability test concerning the absence versus the presence of the final segment of the four long close Swedish vowels could have been not only suggested (p. 142) but also carried out. My second major criticism concerns certain generalizations which, if true, would have constituted and impressive climax to the thesis. The author believes that she has found a general tendency according to which substitutions (the errors of the learners) are closer to the tongue’s rest position than the correct sounds of the target language (p. 151). For instance, for the Swedish close vowels the speakers of the other languages generally substitute more open variants (p. 149). However, this fact docs not prove anything as the Swedish close vowels arc particularly close. One would obviously have obtained the contrary result, i.c. substitutions being generally closer than the Swedish close vowels, if one had chosen languages which had still closer close vowels than Swedish. French would be one such case as the main central part of French /i/ and /u/ is closer than the corresponding part of the corresponding Swedish vowels. (Towards the end the Swedish vowels become, on the contrary, so close that they often become fricative.) Also among the open vowels the author believes that she can establish a general tendency towards a neutral position as the Swedish front /a/ is not pronounced sufficiently palatal by the immigrants and the Swedish back /a/ is not sufficiently velar. However, this fact cannot prove any general tendency either. It is only a reflection of the fact that Swedish has two well differentiated open vowels whereas most of the other languages included in the investigation have just one open . The half close and half open vowels are also adduced to prove the tendency towards a neutral articulatory position. In this case the facts are, however, so complicated and uncertain that the author’s statistics cannot be taken as proving anything. (The nine source languages have in this articulatory area one, two or three degrees of aperture in their vowel systems and the exact height of the vowels is more or less different from language to language. In addition, the author’s assumption (p. 57) according to which Swedish /e/ is one step more open than /e:/ is not correct either for or for the model.) Furthermore, the author attempts to exploit the idea of higher vowels having ‘a larger area for articulation’ than lower vowels. She believes that generally more errors occur in the larger area (p. 153). The attempt is equally unsuccessful. Swedish with /i:, !, y:, Y, ui:, 6/ (as in vin, vinn, hys, hyss, hut, hull) is unusually rich in vowels in the higher front area. No wonder that learners with fewer phonemes in this area are in trouble when trying to imitate these Swedish phonemes. (For technical reasons the symbols [ut:] and [d] arc used instead of the symbols I usually prefer.) The author is not convincing either when she tries to use univcrsals as explanatory factors (p. 159). Those vowels which accordingJ akobson to are most ‘basic’ give rise, on the whole, to fewer errors.J akobson ’s univcrals can certainly explain, to some extent, the phonemic structure of different languages. Once this is said, the actual structure of the con­ fronted languages explains the errors. It is certainly meaningless to relate the universal mentioned and the errors observed in this investigation directly. What is shown by the present study, and all similar ones, is that the speakers’ use of the target language is influenced by their knowledge of the sourcein language language specific ways. When (pp. 147-148) the author reviews the limitations of traditional contrastive ana­ lysis, the generalities which she offers are quite acceptable but not really original. The author attempts to justify the lack of detail in her phonetic study by saying (p. 129) that ‘it has not been our aim to try to explain all of the particular difficulties’ but rather that the ‘emphasis of the investigation is placed on observation of common tendencies’ in order Libri 255 to relate the specific results to ‘more general tendencies and hypotheses’. Unfortunately I cannot find that her general conclusions arc of such importance that they compensate for lack of a more detailed phonetic study. On p.25 an example fromW einreich is used as a basis for a discussion of basic principles for the comparison of the phonetics of two languages. It is said that the comparison is made possible inW einreich’s work ‘by a separation of structure and substance’. It is not al­ together clear to me what the author wishes to imply by this statement and what her own standpoint actually is. The ‘separation’ mentioned is a trivial fact and it does not as such make comparisons possible. Generally it can be said that comparative studies will most often have to establish relations between units or realizations whichsimilar arc (but not identical). This is why terms such as identical, same, equal or identification used by the author, and by most others, arevery not appropriate. I have suggested myself the use of the term ‘corresponding’ (seeLinguislische Einheiten, pp. 70-73). (If one lets ‘identical’, ‘same’, etc. mean exactly what they scent to mean, perfectly legitimate comparative studies would be ruled out). On p. 30 the author correctly secs that generative phonology is ‘inapplicable’ to her study. On pp. 119 and 139 it appears that Portuguese informants have substituted [se] for Swedish [a]. However, the sound involved is most probably not [ae] but the one which in the sketch of Portuguese (pp. 113-116) is described as [e]. Ifso, it would only appear before nasal consonants. On p. 143 the major substitutions for Swedish consonants arc provided in a table. One is reminded of the fact that some consonants, viz. /p, t, d, q/, which it would have been interesting to investigate, have been left out of the study. This omission is conditioned by the author’s belief that they were too difficult to interpret auditorily. The reader regrets that she has not been sufficiently audacious to include these consonants. Her results would perhaps have been somewhat more approximate than in other cases but surely still quite interesting. In the pronunciation of the informants recorded, /$/ in particular is more often seriously wrong than it is more or less right. More particularly the exclusion of the consonants just mentioned is thought to be due to ‘the imperfect quality of the recordings and to extreme difficulties in auditory analysis’ (p. 54). Those of the author’s recordings which I have listened to were, however, of good quality. On pp. 145, 146-147 and 158, when discussing numbers of substitutions and degree of difficulty the author takes a somewhat strange consequence of her considering the Southern Swedish model as the target. For /6/, the [JJ-likc substitutions by speakers from all nine languages are not quite of the same kind as most other errors because the [J]-likc sounds arc more or less acceptable in Standard Swedish and it is unlikely that the teachers of the in­ formants have tried to change them./ r For/, the speakers from most languages are said to ‘substitute’ /r/. However, this latter is Standard Swedish, and usually they would not have tried to acquire the/r / of Southern Standard Swedish. Consequently the informants arc said to substitute an /r/ for an/r/, which they have not intended, nor been taught, to pronounce. To my mind, the two main assets of this book are (a) the discussion of contrastive analysis and error analysis (pp. 10-40), which is of general interest, and (b) the sketches of the phonetics of the ten languages involved, as well as the error analysis of speakers from nine different languages, which arc of interest to teachers of Swedish to immigrants. G. H ammarstro .m, Clayton and Umea