General

Abstract

language of our own: the genesis of Michif Oxford: Oxford University Press, 1996). Some contributors found that the cases they looked at did not appear to be cases of grammaticalization. Van Rossem finds no evidence that the Negerhollands affix ' -so' was ever a lexical item. According to Smith there is more evidence to suggest that a grammatical marker in Saramaccan (focus marking we) was transferred from Fon (one of the source languages) rather than being the result of grammaticalization of the English word ' well'. Syea finds that a clitic (la) in Mauritian Creole represents a case of degrammaticalization as opposed to grammaticalization, since it evolved from an affix. The majority of contributors found evidence of grammaticalization, but we also see differences. Huber, Poplack and Tagliamonte, Baker, and Muhlhauser show evidence of grammaticalization in pidgins as opposed to borrowings from one of the source languages. Grant and Mufwene show that nature of the grammaticalization can differ even amongst closely related contact languages. Grant looks at Grand Ronde Chinook Jargon, which is reported as having many more cases of grammaticalization than other varieties of Chinook Jargon. Mufwene looks at a number of Englishbased Creoles and shows that they do not always select the same morphemes for the same grammatical functions and that grammaticalization proceeds at different rates. Kouwenberg considers the conventionalization of one of a number of variants in Berbice Dutch Creole as grammaticalization. Bruyn tries to bring order to what could be described as chaos by proposing three types of grammaticalization. Namely, ordinary (gradual and language internal), instantaneous (much faster than the ordinary) and apparent (transfer after grammaticalization in another language) (p. 42). These categorizations are workable if we are sure that we are dealing with essentially completed processes, but Kihm, and Poplack and Tagliamonte show data which indicate that some of the processes may not be complete. Kihm describes a problematic reflexive in Kriyol as a case of half-hearted grammaticalization. Maybe it is not half-hearted, just incomplete. Poplack and Tagliamonte, after an interesting statistical analysis of past time marking in Nigerian Pidgin English, conclude that although there is evidence of grammaticalization over the past couple of centuries, the forms considered have not yet fully attained the status of grammatical markers. The findings reported in the articles are almost as varied as the types of contact languages. This makes the issue more intriguing and underlines the need for analyses of data from other contact languages. We may eventually find that grammaticalization in contact languages is not as different from that in noncontact languages as we now suspect, but only time and further research will tell. One blot on the otherwise high quality of the book is the level of personal critical comments that occur in two articles. There is even one case of an unjustified accusation of faulty transcription that was left unedited because it was thought that it made a valid general point (n. 6, p. 160). I think the general point could have been made in a general way. Healthy debate should always be encouraged but we should debate issues not personalities. This, however, does not detract from the high quality of this well-compiled selection of articles which should be recommended to anyone even indirectly interested in contact language issues. The book is most accessible to linguists but may also be of interest to other academics and students.