In the Preface to this volume, Joel Walters explains why we need “yet another model of bilingual processing” (p. viii): The field of bilingualism is divided between researchers focused on the social situating and functions of bilingualism and those concerned with mental structures of bilingualism. Yet bilingual speakers function linguistically at the intersection of these two realms; Walters is concerned to construct a map of cognitive processes at that intersection. In other words, Walters takes on an enormous task: to construct a psycholinguistic model of bilingual processing that includes the wealth of social and pragmatic information that plays a role in bilingual speech and interaction.
Chap. 1 asserts the difficulty of describing bilingualism, comparing it to a “mesmerizing piano-violin duet” (1). While utterances are linear, the knowledge underlying them is “densely layered and largely non-sequential” (1), which must be accounted for in an integrative model. The Sociopragmatic-Psycholinguistic (SPPL) model thus undertakes to account for specific bilingual phenomena such as codeswitching, interference, and translation, as well as for functions affected by bilingualism, such as computation and cursing. Following a brief overview of the various approaches he wishes to integrate, Walters describes terminological differences between the fields he is attempting to bridge, and argues for integrative work nonetheless.
The second chapter, the review of literature, begins with an acknowledgment that the chapter will be selective: There is too much work on bilingualism to review in one chapter. Walters selects ten researchers or teams and classifies their work into four categories: (i) those whose work is linguistic and structural, at least implicitly touching on both sociopragmatic and psycholinguistic areas, (ii) those whose work is ethnographic, (iii) those whose work is laboratory-based, and finally, (iv) models of monolingual processing adaptable to bilingualism.
The work of Carol Myers-Scotton is described as having the “widest scope” (34) in that it examines motivations and constraints of codeswitching and the cognitive organization and processing that underlie it. Walters describes the stages of development of Myers-Scotton's model, and what he sees as its strengths and weaknesses. Curiously, he makes no mention of the Markedness Model (Myers-Scotton 1998), which explicitly connects social motivations to speakers' linguistic choices and thus is highly sociopragmatic; instead, he concentrates on more psycholinguistic aspects of her work. In the same section, the work of Michael Clyne is assessed as broad in scope and functional in orientation, giving some attention to processing issues, while that of Peter Auer, “a card-carrying functionalist” (48), is not oriented to processing concerns.
It is a mark of the difficulty of his integrative task that Walters undertakes to “dig deeply” (58) in the work of sociolinguistic researchers for their notions on psycholinguistic processing. In the cases of both Rampton and Zentella, processing seems not to have been their focus; Walters's attempt probably indicates the size of the disciplinary divide he is trying to bridge. Similarly, the psycholinguistic models discussed in the rest of the chapter are examined for any attention they might give to sociopragmatic concerns, as well as for the processing issues (control, activation, inhibiting) that the models address. Walters concludes the chapter by observing that researchers from all four groups have shown willingness to alter their models to address concerns from outside their primary focus areas, but a new integrative model is still needed.
Chap. 3 outlines Walters's own model. Two large modules are accessible at all stages; they contain information about language choice and affect. Five components are arranged sequentially, containing “social identity information,” “context-genre information,” intentional information, a formulator, and an articulator. The assumptions behind this model are that bilingualism needs its own specific processing model, and that identity information and sociopragmatics need to be integrated into speech production models. This information should be organized into autonomous but interacting modules; the model should permit both sequential and parallel processing. Finally, the system needs redundancy as a check at all levels.
Throughout the rest of the chapter, Walters gives data that support each of the components of the model, such as the individual case history of “Yulia,” an Israeli of Russian origin, showing how identity issues affect her language use. Walters's discussion of the evidence for each of his components is broad, in that he both cites studies that focus on the phenomena covered by a particular component (such as genres or the roles of topics) and also describes the kinds of research methods he sees as most useful in exploring these phenomena. Appropriately, he sees the usefulness of multiple approaches in data collecting and analysis, a position similar to that of the pragmatic researchers Kasper & Rose 2002.
Chap. 4 focuses on processing mechanisms in the SPPL model. Walters briefly defines the four mechanisms he sees as central: imitation, variation, integration, and control. Having justified the “boxes” of his model diagram in chap. 3, Walters devotes chap. 4 to the “arrows.” For each of the four mechanisms, Walters “walks through” the arrow linking one of two modules (Language Choice or Affect) to one of five components (Identity, Context/Genre, Intentions, Formulator, or Articulator). For example, the Language Choice module is linked by imitation to the Intentions component. An illustration of this link is the case of the Chinese linguist who wrote to Walters in English to request reprints of publications, including in his request a very Chinese-L1-like lengthy apology.
In the second part of chap. 4, Walters expands on the four processes themselves, citing other theories, frameworks, and examples in which functioning of these processes is central. Thus, in discussing imitation, Walters cites Speech Accommodation Theory/Communication Accommodation Theory (Giles & Coupland 1991) as a (sociopragmatic?) framework in which imitation plays a central role. Tannen 1989 offers a more “linguistic” approach, but imitation is nevertheless central. Walters sums up: “Acts of linguistic choice, motivated internally by the social needs and externally by pressures of social comparison, draw heavily on processes of repetition and imitation” (173). Both imitation and its partner mechanism, variation, are essential to the bilingual phenomena Walters is concerned with – codeswitching, interference and translation. Two other processes, however, are required to create the relationship between imitation and variation, according to Walters, these being integration and control. These “executive mechanisms” are also discussed from several different theoretical standpoints, and a version of Perceptual Control Theory for bilingualism is sketched.
Chap. 5 turns the focus to the bilingual phenomena Walters hopes to account for with his model: codeswitching, interference, translation, and bilingual computation and cursing. Walters distinguishes between intentional and performance codeswitching, the former a matter of skillful choice motivated by identity or contextual considerations, the latter likely to be accompanied by hesitation phenomena or other evidence of dysfluency. Both types can be accounted for with the SPPL model. Walters's discussion of interference is somewhat less conclusive: He calls for an integrated investigation of interference combined with fluency, since we need to understand these notions together; we can infer that the SPPL can provide a model to guide this research. Similarly, the model can help distinguish between interpretation, which is “more ostensibly social and interactive,” and translation, “more individual and psycholinguistic” (210). The discussion of the sociopragmatics of interpreting seems somewhat limited; Walters seems to assume “conference interpreting” as the norm, whereas articles in Mason 1999 show the variety of sociopragmatic challenges in dialogue interpreting that arise if one also considers medical, police, and court interpreting. Nevertheless, Walters's discussion of the psycholinguistic/cognitive aspects of interpreting shows that the SPPL model may well foster further research in this area. Similarly, the SPPL model raises research questions for exploring bilingual computation and cursing. The chapter ends with a discussion of methodology appropriate for the research programs proposed.
Chap. 6 connects the SPPL model with real-world issues of language acquisition, loss, and disorders, showing how specific questions provoked by the model could lead to increased attention to questions of identity and affect in language teaching, the nature of fluency, language attrition vs. gaps in acquisition, and so on. Understanding these phenomena would benefit researchers and speakers alike.
In each chapter Walters focuses on a different part of the model, defining components, understanding processes, elucidating bilingual phenomena themselves; thus, each chapter functions as a different lens to turn onto the object, the SPPL model itself. It is helpful that each chapter contains a chapter summary; the reader can lose focus in the breadth of discussion as Walters moves from model architecture, to evidence for that particular architecture, to research results and methodology, for each box or arrow of the model. While this thoroughness is appropriate for his goals, the periodic return to the overall view of the model that the chapter summaries provide makes these sections particularly welcome. Although the volume is impressive in the scope of its inquiry, I suspect that it will be of greater interest to psycholinguists than to sociolinguists; nevertheless, the work serves as useful reminder of the wisdom of casting frequent glances across disciplinary borders.
