- Borrowing: Loanwords in the speech community and in the grammar by Shana Poplack
The study of borrowing in particular and of language mixing in general has a long history of research (e.g. Haugen 1950, Haspelmath & Tadmor 2009), and this history is characterized by a pervasive lack of consensus. While there seems to be general agreement that loanword borrowing constitutes the clearly preferred mixing strategy in the languages of the world, there remain many questions that the student of borrowing is confronted with.
Most of these questions are taken up in the book under review and are subjected to empirical scrutiny, based on quantitative and qualitative analysis of a staggering 43,000 tokens of other-language items in the speech of nearly 500 bilingual individuals. The framework of analysis on which the book rests is provided by variation theory, one of whose central concerns is with explaining asymmetry in linguistic form-function relations. Shana Poplack characterizes her approach as sociolinguistic in nature, aiming to reconstruct the actual process of borrowing by bilinguals in the context of their respective speech communities. Based on over thirty years of research on language-mixing strategies, she is also concerned with the status of established loanwords, contrasting them most of all with code-switches and nonce borrowings, that is, items occurring only once in a corpus.
The book is divided into twelve chapters. The introductory chapter provides information relevant for an understanding of the subject matter of the book, including a catalog of definitions of the technical terminology used. Based on a variationist perspective on language contact, Ch. 2 provides a review of this perspective and the way the author applies it to analyze language mixing.
Ch. 3 discusses the data on which the book is based, particularly P’s Ottawa-Hull French Corpus, collected in the 1980s, which provides the main database of analysis. Analyses of the corpus data are then presented in the ensuing chapters, where random-sample techniques are applied in Ch. 4, while Chs. 5–7 use social network methodology. In Ch. 8, a diachronic perspective is added to deal in particular with nonce forms in bilingual production, the main topic of this chapter being how nonce borrowings may turn into loanwords.
Another central topic of the book is discussed in Ch. 9, namely the distinction between borrowing and code-switching and how the two can be separated—a topic that has received quite some attention in previous studies (see Myers-Scotton 2002:153–58 for a discussion and alternative conclusion). In some of those studies, phonetics is argued to play an important role in the integration of borrowed material, and Ch. 10 is devoted to determining the contribution of phonetic features to loanword borrowing as well as to other kinds of language mixing.
While the analysis in the preceding chapters focuses on linguistic variables, diffusion of borrowed words across the speech community in accordance with sociodemographic variables is discussed in Ch. 11, and Ch. 12 concludes the book by presenting a catalog of generalizations on loanword borrowing and how the latter can be separated from other kinds of language mixing.
The text is followed by two appendices, the first of which (Appendix A) contains detailed sociolinguistic and demographic information on the 120 speakers on which the Ottawa-Hull French Corpus is based, while the second (Appendix B) is restricted to a few bibliographical references. There is a general index but no author or language index.
The research on which the book is based has two important characteristics. One is that it relies on quantitative data, drawing on a text basis of sixteen corpora, where all data sets have been stored and annotated. The second characteristic is its crosslinguistic orientation. While the English-French contact situation in Canada accounts for the majority of the analyses presented, the book covers borrowing phenomena in thirteen language pairs, including languages from five different language families. Ch. 7 is especially rich in language data, including recipient languages such as Gulf Arabic, Tunisian Arabic...