Widely impugned by an uninformed public – and even by many of those who practice it – with such derogatory terms as “Spanglish” or “Chinglish,” or “Pocho” in Spanish, codeswitching (CS) has emerged from marginal obscurity to become a major topic of interest among linguists of a wide variety of persuasions in the past 30 years. Weinreich (1953) famously denied that a switch between languages within a sentence was possible; the MLA bibliography now lists 900 titles on the subject, half of which have appeared since 1995. The present volume – the third edited by the indefatigable Rodolfo Jacobson, a pioneer in the field since the 1970s – reflects both this growth and the increasing breadth of interest that has occurred along with increasing attention to bilingualism generally in its many aspects and implications.

Designed as a sequel to its predecessor, Codeswitching worldwide (1998), this volume consists of papers from the Fourteenth World Congress of Sociology, held in 1998 in Montreal, Canada, together with some (unspecified) solicited articles. Five of the same authors are present, some partially updating their previous contributions, but overall this collection complements the previous one. The main exception is Carol Myers-Scotton's lead-off paper, “The matrix language frame model: Development and responses,” in which she summarizes changes in her original (1993) model and responds to criticisms of it. Jacobson, in his lengthy Introduction, which summarizes and discusses each chapter, states that “the placement of her chapter at the beginning of this anthology is then intended to set the framework for what codeswitching means in the eyes of sociolinguistic scholars today” (p. 2).

This statement is misleading in two respects, particularly in the context of the present collection. First, though all but one of the chapters refer to Myers-Scotton's matrix frame model, some critically, only one expressly applies the model. Second, as sociolinguists seeking for an understanding of CS will be disappointed to discover, the article itself focuses entirely on technical issues of grammatical constraints, at points presupposing a knowledge of Chomsky's (1995) recent theories, and concluding on an almost Chomskyan note: “The discussion has emphasized that the matrix language is best understood, not as an actual language, but rather as a theoretical construct referring to the abstract morphosyntactic frame that structures bilingual utterances” (54).

Jacobson's chapter, which follows, argues for language “alternation” as a “third codeswitching mechanism” which is found where there is quantitative and structural parity between two languages within codeswitched communication. Jacobson's critique of Myers-Scotton seems to involve a misunderstanding of her use of Chomskyan terminology for clauses (CP, or Complementizer Phrase) and a confusion of the relation of the social determinants of language status to internal grammatical features of CS. The term “alternation” is better reserved as a neutral term referring to any kind of successive use of two or more languages or language varieties, including registers.

It should be noted that while syntacticians and psycholinguists have focused their attention strictly on CS within sentences – intrasentential CS – sociolinguists and anthropological linguists have taken a broader perspective, including in their purview any kind of language alternation. A word about terminology is merited here: Most U.S. linguists use the terms “intersentential CS” for switching between sentences and “intrasentential CS” for switching within sentences, but many non-U.S. linguists restrict the use of “codeswitching” to the former, and refer to the latter as “code-mixing”. This latter usage is unfortunate, however, as it reinforces the public image of CS as “word salad”, a random, grammarless mixture of two languages reflecting a speaker's popularly supposed incompetence in both.

Section 2, “Linguistic aspects: From morphosyntax to semantics,” contains four articles. Jeanine Treffers-Daller discusses the use of past participles in the Romance and Germanic varieties spoken in Brussels and Strasbourg, arguing that structural, rather than sociolinguistic, factors explain the differences in the respective varieties in these two areas. Shoji Azuma analyzes constraints on Japanese-English codeswitching within Chomsky's Minimalist Program framework, concluding that a formal feature analysis can account for most restrictions, functional categories ([+F]) being resistant to switching, but word order can also be a factor in some cases.

Ol'ga S. Parfenova examines the extraordinarily interesting situation of Bulgarian communities in Ukraine, considering the impact of Russian on the Bulgarian spoken there. Statistical analyses of 17 speakers' interview data show that older speakers used the fewest Russisms, and younger urban speakers the most. Of these Russisms, 40% were nouns, 25% adverbs, and 13% verbs. A few examples illustrate various functional categories. The author finds the social-contextual effect of schooling in Russian to have a major influence on relexification, while structural factors serve as constraints. She speculates that Ukrainian policy to foster the teaching of Bulgarian while emphasizing Ukrainization will have variable effects in the future. Parfenova discusses the existence of German and Romanian settlements in Ukraine, and of Bulgarian settlements in Romanian-speaking Moldova, which would offer a fascinating laboratory for examining the relative effects of different language vs. different context on borrowing, codeswitching, and language attrition. It is to be hoped that someone will pursue this opportunity.

In his significant contribution, Ad Backus, rightly noting that “the literature on codeswitching tends to deal more with the morphosyntactic integration of elements from another language than with the motivation for using exactly these elements and not others” (124), seeks to address this shortcoming by proposing a Specificity Hypothesis that “Embedded language elements in codeswitching have a higher degree of [inherent] semantic specificity,” a claim that he tests and illustrates with Turkish-Dutch CS among members of the immigrant Turkish community in Holland. An interesting corollary, deserving further testing, is that “high specificity and low specificity can be equated with higher-level vocabulary and basic-level vocabulary, respectively” (129). Backus recognizes that there is often a cline between these two, and acknowledges that social connotations may also influence language choice.

Section 3, “Codeswitching as oral and/or written strategy,” consists of two chapters. Erica McClure, noting the dearth of attention to differences between oral and written CS, provides valuable documentation of such differences in the Assyrian immigrant community in Chicago. Her analysis is informatively framed in a historical and cultural context, including technological and dialectal impediments to the use of Assyrian in e-mail.

Cecilia Montes-Alcalá gives an extensive classified sampling of switches from her own personal journal over a period of 17 months while she was a college student in California. The examples reflect a shift from primarily intersentential (or clausal) CS to more freely intrasentential CS as her fluency in English progressed. The data provide a rare real-time case study of change in the performance of an individual, but regrettably the author gives little information on herself or the contexts reflected in her isolated journal excerpts.

Sociolinguists will take most interest in the two chapters in section 4, “Emerging ethnicities.” Robert Herbert, in the only contribution explicitly based on Myers-Scotton's model, analyzes how different types of CS may be used in Johannesburg to negotiate identity. The paper is richly illustrated with valuable contextualized examples. Missing only is crucial biographical information on the interlocutors, which would help in judging the accuracy of the investigator's analytical inferences and speculations.

Similarly rich in examples, but with minimal contextualization, is Miriam Ben-Rafael's paper on French-Hebrew switching – which she has labeled “Franbreu” – by French immigrants in Israel. Her categorization of examples by single word or phrase and by type of function provides only occasional glimpses into the purposeful or unconscious reasons for the switches.

Finally, in Section 5, “Communication codes in education,” Diana-Lee Simon illustrates the value of CS in foreign language teaching in two settings: French in Thailand and English in France. More contextualization, using Thai rather than translations, would have been helpful.

It is useful to have a collection such as this eclectic sampler of the state and breadth of research on codeswitching worldwide as of 1998, reflected in the wide variety of languages, places, contexts, and analytic frameworks included, though prospective buyers will be deterred by the exorbitant price. Some chapters could have benefited from the vetting accorded by refereed journals, and several suffer from a frustrating dearth of contextual information. The papers by Backus, McClure, and Herbert provide the most original and significant contributions, and Parfenova's survey invites more in-depth and comparative analysis. Readers looking for rich qualitative data or emergent generalizations will be disappointed, for the most part. Nevertheless, the papers together provide a valuable corpus of examples of CS in a diversity of language pairs, and the 304 items compiled in the References section (343–364), some as recent as 1999, will be a useful resource. Rodolfo Jacobson, succeeding in his goal of giving a partial snapshot of worldwide research on codeswitching approaching the turn of the century, has produced a collection that also reflects the current diversity of perspectives and the regrettable lack of a cumulative consensus in our understanding of this intriguing and challenging phenomenon.