LINGUIST List 32.3132

Mon Oct 04 2021

Review: Cognitive Science; Syntax: Lopez (2020)

Editor for this issue: Jeremy Coburn <>

Date: 18-Jun-2021
From: Dennis Ott <>
Subject: Bilingual Grammar
E-mail this message to a friend

Discuss this message

Book announced at

AUTHOR: Luis Lopez
TITLE: Bilingual Grammar
PUBLISHER: Cambridge University Press
YEAR: 2020

REVIEWER: Dennis Ott, University of Ottawa


In the author’s own words, the slim book under review aims to provide “a first approximation to understanding” how “multilingual individuals must navigate a linguistic input of considerable complexity and eventually build up a linguistic competence system in which the various [grammatical] features fit in their proper place” (45). This modestly stated goal summarizes by the same token the central theoretical claim of the monograph: multilingual speakers possess a single, integrated competence system/I-language, as opposed to some configuration of multiple, more or less encapsulated mental grammars. To make this case, López largely focuses on code-switching (C-S) patterns in bilingual speakers, drawing on a growing body of literature; parts of the discussion synthesize and extend research by López and collaborators on C-S conducted in the UIC Bilingualism Research Laboratory (, which he directs.

CHAPTER 1 introduces and motivates the claim that multilingual speakers have a unified system of linguistic knowledge, as opposed to multiple fully or partially independent grammatical systems. López argues against this latter, “common-sense view” of multilingualism on both conceptual and empirical grounds. While the individuation of “languages” is intuitively appealing, we should not impute objective reality to such intuitions (López illustrates the futility of counting languages with a discussion of post-Creole continua, but of course the point stands more generally); and empirically, inter-linguistic grammatical dependencies in code-switching (C-S)—where a lexical item from “one language” enters into a syntactic relation with a lexical item from “the other language,” e.g., when a Basque negation licenses a Spanish NPI (or vice versa)—show that whatever codes are perceived to be involved, they are not siloed systems. The best way to proceed, López argues, is on the basis of the hypothesis that “the I-language of bilinguals is not substantially different from [that of] monolinguals” (2).

CHAPTER 2 surveys the theoretical landscape and serves as a bridge to the presentation of López’s own model. It begins with a discussion of “separationist” models of multilingualism, i.e., theories that postulate two (or more) encapsulated lexicons and/or phonological systems, as exemplified by MacSwan’s work (e.g. 1999). López rejects these models on empirical grounds, citing again the aforementioned cross-language dependencies in C-S, further examples of which are provided in this chapter. He then goes on to sketch an alternative model couched in Distributed Morphology, dubbed Minimalist Distributed Morphology (MDM). On this integrationist model, multilingual speakers have a unified linguistic competence system, which consists of several lists related by the generative procedure: in addition to an invariant inventory of conceptual items, one list contains grammatical features and roots, another vocabulary-insertion rules (VIRs).

CHAPTER 3 develops López’s MDM model in greater detail and provides some requisite background on general syntactic theory. His assumptions include the idea that roots are assigned categories by “little-x” heads and that certain heads constitute phase heads; specifically: C, Voice (distinct from the categorizer v), and K(ase), the head of nominals. Building on prior joint work with González-Vilbazo, López argues that these phase heads have a privileged role in C-S, in that choice of the phase head determines properties of its complement. For instance, choice of a “Spanish” light verb (= Voice) dictates VO order in its complement VP, whereas choice of a “German” Voice head results in OV order. López illustrates how the realizational MDM model can account for C-S within nominals, specifically how a K-head from a concord-less language combined with other nominal material from a concord language effects the suppression of concord in the overall NP. The upshot of the discussion is that multilingual speakers differ from monolinguals essentially only in degree, not qualitatively: their I-language is a unitary system, which incorporates a wider range of functional categories, VIRs, and postsyntactic operations compared to “monolinguals”.

CHAPTER 4 develops in further detail the claim that multilingual competence is rooted in a single, unified lexicon. López again draws on “mixed” dependencies in C-S to argue against separate lexicons and describes how his unified MDM model can account for these dependencies. The flipside of this productivity are constraints on C-S, such as when it is observed that a Basque negation is incompatible with Spanish INFL and vice versa; this is explained by the differential feature structure of the two negators. A conceptual consequence of the one-lexicon architecture is that there is no theoretically relevant distinction between C-S and “borrowing” (as argued by some); these only remain as “descriptive labels for what is the output of the same [generative procedure]” (69). Specific case studies offered in this chapter concern the occurrence of “gender transfer” when new roots are introduced into the lexicon, and competition among functional items and inflectional affixes from different “languages” in (rare) cases where there is no free variation (which López surmises are due to different featural specifications of the items involved and different application conditions of VIRs, respectively).

CHAPTER 5, the longest chapter of the book, presents a detailed case study of gender and concord in Spanish/X C-S, including a Spanish/German amalgam dubbed “Esplugish” and studied in great detail in González-Vilbazo 2005. This choice of topics is particularly appropriate in view of the fact that gender and concord phenomena constitute a rare area of C-S research that has been investigated in quite some detail by different researchers, i.e., where the empirical database is relatively solid (though not free from contradictory findings, as López concedes). This allows López to demonstrate in detail the workings and virtues of his MDM model, specifically the significant role played by the Subset Principle in deriving surprising asymmetries in Esplugish D–N collocation: while Spanish determiners happily combine with German nouns, the inverse case is intricately constrained (discriminating between nominative and non-nominative D-heads, among other factors). Conversely, López argues, approaches that posit multiple separate lexicons—one per each “language”—overpredict the combinatorial options, since this very architecture precludes the possibility of competition.

CHAPTER 6 continues the elaboration of the MDM model, defending the idea that even the phonological competence of multilinguals does not incorporate separate “language”-specific subsystems, but is fully integrated. The chief witnesses for this claim are cases where phonological properties of one language are seemingly superimposed on lexical items from the other language, e.g., when Spanish/Catalan speakers apply a distinctly “Spanish” rule for clitic-cluster simplification to Catalan clitics, or when prosodic-phrasing rules of Japanese are applied to a mixed Japanese/Brazilian Portuguese utterance.

CHAPTER 7 begins an exploration of psycholinguistic issues that continues in the following chapter. The chapter is devoted to lexical matters: from the perspective of the MDM model, what does it mean to learn a word? In the general case, López argues, this feat amounts to no more than the association of a root with a concept and a phonetic form—or multiple such forms, in the case of multilingual speakers. The acquisition of additional exponents is no different from the general process of lexical acquisition and implies nothing more than a single (distributed) lexicon. The remainder of the chapter adduces psycholinguistic evidence for this conclusion and highlights drawbacks of competing models.

CHAPTER 8 continues the psycholinguistic theme, focusing on matters of acquisition and processing. The discussion centers on effects of ‘interference’ and ‘convergence’ (e.g., when US Spanish speakers use gerunds as nominal modifiers, evidently an import from English), which, López argues, bolster further the picture of a single, integrated grammatical system underlying multilingual competence. This jibes with observations about C-S in children, which, just like its manifestation in adult language, bears the hallmarks of a single integrated system; this López takes to support the idea that multilingual acquisition, too, is an integrated process (contrary to separationist views). The chapter closes with a discussion of various claims about the putative processing cost of C-S.

CHAPTER 9 presents a brief discussion of related works that subscribe to a (more) lexicalist view of C-S, by way of which López also acknowledges antecedents of his own non-lexicalist model. López then addresses the nature of “code-blending,” where spoken and signed language are used simultaneously. The discussion in this chapter mainly serves to highlight the fact that many rather fundamental questions in code-switching/blending research remain at present unsettled.

CHAPTER 10 summarizes the central claims and conclusions of the book and contrasts them with prima facie similar ones reached in the context of “translanguaging” research.

Two appendices conclude the book. APPENDIX A discusses previous approaches to constraints on code-switching. APPENDIX B elaborates briefly on the “post-Creole continuum” discussed in chapter 1.


Despite its nominal brevity of about 200 pages, López’s book covers an impressive amount of ground in the pursuit of (at least) three interrelated goals. One is to argue against the traditional, common-sense based separationist view of multilingualism. Another is to highlight the benefits of a realizational model of morphosyntax for C-S research and provide specific case studies to demonstrate its viability. The third, and perhaps most important, goal is to extend an invitation to theoretical linguists to discover the empirical treasure trove that is C-S and embrace it as a useful analytical testbed, as opposed to some negligible deviation from the “ideal speaker–hearer”. Judged by these goals, the book is undoubtedly an impressive success.

López is to be commended for providing the first detailed articulation of the manifold ways in which the traditional, separationist view of multilingualism (in its various incarnations) is misguided. In fact, I believe a stronger point can be made: the separationist view cannot even be coherently stated. On the one hand, if we accept some individuation of languages/codes based on what will necessarily be arbitrary criteria, we are led to recognize that we are all massively multilingual. As once memorably emphasized by Chomsky in an unorthodox context ( we constantly use different codes in communication, e.g., when talking to our friends vs. our grandparents. On the other hand, adopting a cognitive-internalist perspective on language forces us to abandon talk of “languages” as though they were actual objects in the world and instead define technical notions (such as I-language) as descriptors of a scientifically identifiable reality. As forcefully argued in Chomsky 2000 in the context of this and other confusions, common sense does not beget metaphysics; categorization of linguistic competence by means of mono-, bi-, and multi- prefixes remains as a mere artifact, convenient in ordinary language use but with no place in linguistic theory.

Once the focus is shifted to linguistic capacity, it follows, as highlighted in the book’s conclusion, “that code-switching … is an epiphenomenon” (191). As López is careful to point out, this recognition does not make C-S any less interesting to the theorist; on the contrary, its study is no different from investigations in any other grammatical system (compare the conventional designation of a particular pattern of C-S as ‘Norwegian American’), but enriched by experimental variables not proffered by the individual “languages” involved when considered in isolation (see Hoot and Ebert 2021 for a pertinent recent example)—viz., the many potential grammatical conflicts navigated by the speaker in the use of C-S. The various case studies in this book make an impressive case for how the focus on these conflicts and their resolution not only serves the study of C-S, but can shine a new light on general theoretical issues in the study of language, ranging from the micro level (e.g., gender must be represented separately from the nominal root) to the macro level (a realizational, DM-type model emerges as the superior framework for the study of morphosyntax generally).

The epiphenomenal character of C-S entails that it should be investigated using the same tried and tested methods that have advanced the study of “monolingual” competence; López cogently supports the validity of data obtained by introspection (as opposed to a dogmatic limitation to naturalistic/corpus data) by pointing out that ultimately all information about grammars is elicited in this way. Through no fault of his own, the discussion reflects the current sparsity of such introspective evidence: key tokens of data, summarized in the concluding chapter, recur throughout the book; and more than once only a single datum or two is used to make a rather sweeping point. How much the reader is willing to be moved by some of the arguments developed here will thus depend in no small part on his or her receptiveness for qualitative data without secured quantitative footing. For this reason alone, my feeling is that López’s book may have more of an impact on the theoretical-linguistics community than traditionally-minded bilingualism researchers.

The only minor criticism I could level at López’s treatise is its somewhat unbalanced organization. Compared to the central chapters in which López’s principal theoretical hypotheses are developed and applied, the later psycholinguistic chapters have a bit of a “lit review” character, making at times for a slightly tedious read. This impression is amplified by the fact that it is not always clear that the studies discussed even focus on the same object, as expressed by López in a caveat that opens Chapter 8 to caution the reader that “what I call ‘syntax’ and what the psycholinguistics literature calls ‘syntax’ are often quite different things” (146). Particularly striking in this respect is Chapter 9’s discussion of ‘code-blending’, the—sometimes only partially congruent—bimodal use of signed and spoken language: to my mind, it is not obvious that the phenomenon even falls within the purview of a competence model (as opposed to performance; an issue that López addresses in a different context [182]). Relatedly, in the discussion of the cognitive “cost” incurred by C-S, López’s integrationist view is said to predict “that stopping code-switching should entail the cost of inhibiting a system”; but neither the notion of “cost” assumed here nor the relevant linking hypotheses are sufficiently elaborated to permit an evaluation of this claim which, at worst, could be read as a tacit admission of separationist premises. Finally, the motivation for excluding the material contained in the two appendices from the main text is obscure.

My impression is that additional discussion of the MDM architecture’s predictions concerning the constrained productivity of C-S in various areas would have strengthened López’s case more than the somewhat cursory discussion in these later chapters. To be sure, the near-exclusive focus on morphosyntax is well-motivated: it is this area of research where the most data are available, allowing López to show the workings and benefits of his model most perspicuously and in great technical detail. But there is a much broader range of topics and questions where C-S may provide just the right testing ground to enable genuine advances in understanding. To mention just two examples, there has been woefully little research into movement dependencies and ellipsis in C-S (with some notable exceptions, such as Ebert 2014 or González-Vilbazo and Ramos 2018), despite obvious bearing on issues such as the copy theory of movement and ellipsis parallelism. Particularly promising with regard to these and other investigations are scenarios of diglossia involving languages exemplifying opposite ends of a particular parametric spectrum, such as different alignment systems and word order (as in the Basque/Spanish case), where much is to be gleaned from the ways in which speakers resolve inevitable grammatical conflicts arising between the “languages” involved. Perhaps the book would have benefitted from casting a somewhat wider net by devoting another chapter or two to outlining such avenues for future research, even if relevant data are sparse at present.

The above remarks should not distract from the immense merits of this monograph. It is high time that theoretical linguists embrace the empirical riches offered by C-S, and this book may just be the nudge that sets the ball rolling. López’s lucid discussion offers a glimpse into this new world, most of which remains uncharted territory, furnishing a compelling illustration of the ways in which C-S research can emancipate itself from the shackles of traditional orthodoxy, to the great benefit of the study of language as a whole.


Chomsky, N. 2000. New horizons in the study of language and mind. CUP.

Ebert, S. 2014. The morphosyntax of wh-questions: evidence from Spanish–English code-switching. PhD dissertation, UIC.

González-Vilbazo, K. 2005. Die Syntax des Code-Switching. PhD dissertation, University of Cologne.

González-Vilbazo, K. and S.E. Ramos. 2018. Codeswitching. In J. van Craenenbroeck and T. Temmerman (eds), The Oxford handbook of ellipsis. OUP.

Hoot, B. and S. Ebert. 2021. On the position of subjects in Spanish: evidence from code-switching. “Glossa” 6(1), 73. DOI:

MacSwan, J. 1999. A minimalist approach to intrasentential code-switching. Garland.


Dennis Ott is Associate Professor in the Department of Linguistics at the University of Ottawa, Canada. His research focuses primarily on formal syntax; specific interests include A'-movement, dislocation, ellipsis, head movement, locality and connectivity effects, and the interaction of grammar and pragmatics.

Page Updated: 04-Oct-2021