Shana Poplack. 2018. Borrowing. Loanwords in the Speech Community and in the Grammar. New York: Oxford University Press. Pp. 272. $US $99 (hardcover).

Shana Poplack is, arguably, Canada's best known sociolinguist, enjoying a well-merited world-wide reputation due to the remarkable number of publications she and her numerous co-authors have produced on language mixing over the past 40 years. As Pieter Muysken points out in the foreword of the book, “I had been thinking that Shana Poplack, if she had the energy, should assemble her various writings on lexical borrowing into a single book. And, like rubbing Aladdin's lamp, there it was!”

The book is organized as follows. Following Muysken's foreword is a preface in which Poplack acknowledges the collaboration of her many colleagues and students as well as the numerous funding agencies which have allowed her to pursue her research. In Chapter 1, Poplack discusses the rationale of the work. The aim is not to provide new data but instead to synthesize, build on and reinterpret more than 40 years of research on language mixing strategies in the discourse of bilingual speakers, to characterize the phenomena of lexical borrowing in the speech community as well as in the grammar, both synchronically and diachronically. Poplack adds that the focus is not on the product of borrowing but on the process. For her, the process of borrowing is essentially one of integration into the structure of the recipient language. This integration must take into account the inherent variability in both donor languages (L_D) and recipient languages (L_R).

“Chapter 2 reviews the variationist perspective on language and outlines its specific applications to the study of language mixing” (p. 8). It discusses such issues as the primacy of the speech community, the optimal subjects for a variationist study of language mixing, namely the adult bilingual whose linguistic repertoire is stable, the type of data that best reflects the systematic form of language – the vernacular, the Principle of Accountability, the Comparative Sociolinguistic Method, and the Principle of Diagnosticity.

Chapter 3 briefly revisits the Canadian Ottawa-Hull region mega-corpus of French (2.5 million words, 120 carefully selected informants from five different communities, providing nearly 20,000 tokens of some 2,000 different English-origin types), the results of which are discussed in Chapter 4. The chapter also briefly describes two diachronic corpora, speech drawn from the Récits du français québécois d'autrefois, which are audio recordings of spontaneous language material produced by 44 elderly Québécois born between 1846 and 1895, and the Corpus du français en contexte : milieux scolaire et social, which consists of the oral recordings of 166 bilingual francophone teenagers from the Quebec community of Mont Bleu (Ottawa-Hull region), collected between 2005 and 2007.

As mentionned, Chapter 4 reviews the results obtained in the Ottawa-Hull region of Canada, originally described in Poplack et al. (1988). This research, whose results are quite well-known, has undoubtedly set the standards for all ensuing large-scale investigations of language-mixing and, more specifically, borrowing – and the standards are very high indeed!

Chapter 5 discusses the problem of variability in borrowing and how to deal with it. Based on the hypotheses developed in Poplack and Meecham (1998), which state that if the constraints on variability of L_D-origin forms are parallel to those constraining their L_R counterparts, the L_D-origin forms can only be borrowings and not code-switches.

Chapter 6 investigates the use of syntactic criteria to disambiguate ambiguous elements – specifically bare forms – in the analysis of intra-clausal mixing of French with Wolof and Fongbe, basically revisiting the results discussed earlier in Poplack and Meecham (1995) as well as English-Igbo data from Eze (1997, 1998). Again, the results show that the very great majority of borrowed bare forms “display quantitative parallels to their relevant L_R counterparts, specifically at conflict sites, parallels that are far too detailed to be due to chance” (p. 96).

Chapter 7 reviews a series of studies based on eight typologically distinct language pairs. Results again confirm that the overwhelming majority of language-mixing types are lone L_D-origin items and that they systematically behave like L_R items, that is, they are grammatically integrated to L_R. They also show that true code-switching is exceedingly rare and represents only a minor strategy in most bilingual interactions.

Chapter 8 represents an innovation in that, for the first time, it traces “the diachronic trajectory of nonce forms in bilingual production over a real-time period of 61 years and nearly a century and a half in apparent time” (p. 11). Poplack compares data taken from three diachronically related spoken-language corpora of Quebec French: Récits du français québécois d'autrefois (Poplack and St. Amand 2007); a sub-sample of the Ottawa-Hull French corpus (48 speakers born between 1893 and 1965) and Le français en contexte : milieux scolaire et social (2005–2007) (Poplack and Bourdages 2005). Results show that the overwhelming majority of L_D-origin material consist of lone items and that almost all of these have been attested in the L_R lexicon. A very small proportion are unattested nonce L_D items, which disappear after their first mention. The few nonce loans that do advance chronologically are integrated into L_R morphology and syntax, this integration occurring abruptly, at the first mention of the item.

In Chapter 9, Poplack confronts the borrowing process with multiword code-switches produced by the same French-English bilinguals.

Chapter 10 addresses “the question of whether speakers marshal phonetic integration as a strategy to distinguish language-mixing types” (p. 12). Poplack convincingly shows that “individuals do not systematically integrate their L_D words phonetically into L_R”, this being the case whether the items are nonce borrowings or well-attested loanwords. “This confirms that phonetic and morphosyntactic integration are independent. Only the latter is a reliable metric for distinguishing language-mixing types” (p. 12).

Again, based on the Ottawa-Hull corpus, Chapter 11 assesses “the effects of age, gender, social class membership, level of education, individual bilingual proficiency, minority versus majority status, and neighborhood of residence on the adoption and distribution of borrowed material” (p. 13). Although various social and individual factors variably play a role, Poplack finds that community norms are the most important and that they crucially outweigh individual abilities. Chapter 12 concludes with some reflections on the implications of this work for the study of language mixing (p. 13).

This book is a very impressive piece of work. It neatly summarizes the results of many years of research on language-mixing strategies accomplished by Poplack and her colleagues, and convincingly argues in favour of the theoretical distinction between borrowing and code-switching, demonstrating the primacy of the former and relegating the latter to a minor language-mixing strategy. Throughout the book, the methodology employed to gather and analyze the various data is always well presented; the following argumentation is always clear, straightforward and easy to follow. Specialized technical linguistic terms are consistently defined and, in fact, kept to a minimum. Numerous examples, taken from the various corpora, are generously provided, allowing the reader to better understand the stakes at hand. Quantitative results are clearly presented in a variety of text, statistical or numerical tables, and line or bar graphs. References to relevant previously published material is amply provided throughout the text. Each chapter ends with a discussion reviewing the essential points made in the preceding pages.

As the French say: “Après les fleurs, le pot” (After the flowers, the pot). In her analysis of the Ottawa-Hull corpus, Poplack looks at the data according to part of speech and finds that, in accord with previous studies, the overwhelming majority of borrowed vocabulary falls in the category of nouns (64%), followed by verbs (14%), interjections and frozen expressions (12%), adjectives (8%) and conjunctions (1.5%) (p. 48). Unfortunately, Poplack fails to define or to give examples of what she considers to be a “frozen expression”. Flikeid (1989: 217), in her study of Nova Scotian Acadian French, points out that a great number of “frozen phrases” such as come to think of it, enough is enough, I suppose so, never mind, etc., occur regularly and repeatedly in the corpus. These can occur freely at various points of the sentence; others must occupy a specific slot, as in (1):

Some English sequences can even be part of an otherwise French expression, as in (2):

It is not totally evident whether Poplack would consider the above as “frozen expressions” or as bona fide code-switches, since they maintain L_D structure. Since these sequences are often recurrent across speakers, Flikeid proposes that they be viewed as a special type of borrowing. A clarification on the part of Poplack on this issue would be most welcome.

On another note, one cannot help wonder why, in her study, Poplack fails to mention a particularly problematic but relatively well-known set of data of Louisiana French (LF). This data speaks to the borrowing of English participial and infinitival verb forms into LF, where they maintain neither their original L_D structure nor integrate to L_R structure. The following examples are taken from a study of LF by Dubois and Sankoff (1997):

Particle verbs can also occur bare, as in the following example, taken from Rottet (n.d.):

Dubois and Sankoff (1997) state that 75% of verbs produced in their corpus are “bare forms”, as in the above examples. Rottet (2018) confirms that many of these forms have been attested in dictionaries as early as 1936. In light of this, one cannot consider them as nonce loans, nor can they rightfully be considered as code-switches according to Poplack's own definition, since they do not maintain L_D morphological structure. It is, of course, theoretically possible that LF-origin verbs variably delete infinitival or participial suffixes. If that were the case, then “bare forms” such as in the above examples must be considered borrowings, since they match L_R variable structures. Indeed, Picone (1994: 273–74) admits that, in both Louisiana French (LF) and in Louisiana English (LE), there is some evidence of reduction or deletion of inflections for past participles, as in:

Unfortunately, I know of no quantitative study on either LF or LE verb form variation which could bolster arguments for bare form integration either into LF or into LE structure. Indeed, if both L_R and L_D verb structures were to prove to be variable, bare forms such as above might therefore correspond to either L_R variation or to L_D verb structure variation. The dilemma is thus far from being solved. Nevertheless, the fact that some of these bare verb forms are clearly historically attested show that they have the property of ‘listedness’ and should therefore, pace Poplack, be considered as borrowings, even if they do not respect L_R morphological structure.

In spite of the above shortcomings, I consider this book to be well-researched, well-analyzed and well-discussed. I believe it to be a landmark in the history of research on linguistic borrowing, and it should be part of the personal library of anyone interested in (variationist) sociolinguistics in general and, more specifically, in language-mixing strategies.