Reflecting the interdisciplinary nature of second language acquisition (SLA) and bilingualism, SLA/bilingualism researchers have increasingly borrowed experimental techniques from psychology, linguistics, education, cognitive neuroscience, and other relevant fields. This, along with recent methodological advances in experimental research, has enabled researchers to broaden their research spectrum and diversify areas of interest in an effort to deepen an understanding of language acquisition phenomena. Recent studies on second language (L2) development where event-related potentials (ERPs) – scalp-recorded electrical patterns of the brain activity measured with electroencephalography – were employed to observe neural dynamics are some examples of this spectral expansion (e.g. Tokowicz & MacWhinney Reference Tokowicz and MacWhinney2005; Chen, Shu, Liu, Zhao & Li Reference Chen, Shu, Liu, Zhao and Li2007). The current review is another instance of this intellectual trend in the field, and hence designed to draw more attention to that line of research.

Searching for appropriate books to review on the issue was rather a daunting task because the cognitive neuroscience of SLA and bilingualism (especially in relation to SLA) is currently not so hugely popular an area in the mainstream research, and so relevant books are rare. One critical reason for the selection of the books under review here is that they complement each other, with each volume representing an important route to an understanding of the entire language–brain puzzle; that is, Marianne Gullberg & Peter Indefrey's edited volume is experiment-focused, Michel Paradis's monograph is theory-focused, and the volume by John H. Schumann, Sheila E. Crowell, Nancy E. Jones, Namhee Lee, Sara Ann Schuchert, Lee Alexandra Wood is information-oriented. Gullberg & Indefrey's book is a selection of papers presented at a conference held at the Max Planck Institute for Psycholinguistics, The Netherlands, focusing mostly on reporting experimental findings. Paradis describes his neurolinguistic theory about how bilinguals process two languages. Lastly, Schumann et al., for the most part, focuses on the delivery of information on the neurobiological aspects of learning and SLA. All in all, the books selected for the present review will no doubt offer great resources for those interested in this domain of language acquisition research.

Paradis’ monograph is, as he mentions, the outcome of his past research on neurolinguistic aspects of bilingualism. His informative and thought-provoking monograph provides an overview of neurolinguistic research resulting in a theory dealing with varied neurolinguistic issues that should be taken into consideration in our theory-building efforts. The book is written for researchers and graduate students particularly interested in neurolinguistic dynamics of bilingualism. Although various technical terms are used in the book, it does not seem to be too challenging even for novice readers to understand his main arguments. Interestingly, much space is assigned predominantly to his discussion of neurofunctional systems for language, but not too much space is devoted to linguistic theories; for instance, he briefly mentions Universal Grammar, but does not discuss in detail how specific principles and parameters can be explained by his theory involving neurofunctional systems and subsystems.

According to Paradis's theory, a complex set of independent neurofunctional modules operates in the course of verbal communication and linguistic representations are achieved by means of a systematic cooperation among the modules. The nonlinguistic cognitive system (‘conceptualizer’) responsible for formulating messages to be verbalized and decoding incoming messages to be understood is closely connected to the verbal communication system that contains such neurofunctional mechanisms as the implicit linguistic competence system (i.e. procedural memory and knowledge) subserved by the perisylvian area in the left hemisphere, the explicit metalinguistic knowledge system (declarative memory and knowledge) subserved by the hippocampal area and mesial temporal lobes in the left hemisphere, the pragmatic system whose neuronal activation is evidenced in areas of the right hemisphere, and the motivational/affective system subserved by amygdala. Also incorporated into his neurolinguistic theory of bilingualism are the Three-Store Hypothesis, the Subsystems Hypothesis, the Activation Threshold Hypothesis, and the Direct Access Hypothesis. The Three-Store Hypothesis postulates that a nonlinguistic cognitive system that is not language-specific and includes conceptual–experiential features is separately connected to both languages in bilingual speakers. According to the Direct Access Hypothesis, each word, when it is first heard, is directly accessed and perceived as a word with its meaning in comprehension regardless of its language affiliation. The author notes that ‘language may be identified after the word has been understood, not the other way around’ (p. 204). The Subsystems Hypothesis assumes that the nonlinguistic conceptual system proposed in the Three-Store Hypothesis determines which language is appropriate and activates appropriate features and elements of that language subsystem, which in turn affect the final selection of words and prosody. Under the Activation Threshold Hypothesis it is suggested that the activation of an item occurs only when the amount of neural impulses is sufficient enough to reach its neural substrate. The activation threshold refers to the amount of neural impulses needed to activate the item. The selection of a given language is determined by the raising of the activation threshold of the unselected language. This raising makes it easier for the conceptualizer to access the appropriate language.

The chapters in Schumann et al.'s edited book were initially written as MA theses and Ph.D. qualifying papers. The volume illustrates the neurobiology of aptitude, motivation, procedural knowledge, declarative knowledge, memory consolidation, and attention and how each neural mechanism is related to SLA. The book illustrates SLA and discusses relevant issues in a domain-general neurobiological perspective, providing the intended readership (researchers and graduate students) with an opportunity to gain access to neurobiological research on domain-general learning and its potential relevance to SLA research. Despite their informative display of research findings on the neurobiology of learning and relevant SLA research, their extensive use of technical terms may be quite intimidating to novice readers who have just become interested in this area of inquiry. Briefly summarized in the following are the contents of individual chapters.

In his chapter on the neurobiology of aptitude (chapter 1), Schumann discusses individual differences in brain structures from a Darwinian evolutionary perspective, that is, the environmental selection on neural systems leads to individual variation in brains, and hence aptitude that may, in turn, affect L2 learning. As regards the neurobiology of motivation, Schumann & Wood (chapter 2) point out that the amygdala and the orbitofrontal cortex are closely involved in the stimulus appraisal system at a neural level, facilitating dopamine activity that stimulates brain areas related to motivation and goal-directed behavior. In her chapter on the neurobiology of procedural memory (chapter 3), Lee argues that the proceduralization of linguistic information implicates the activation of the basal ganglia system. Based on the potential learning mechanisms of the basal ganglia system, she also discusses some important aspects of the current L2 research agenda, such as learning fixed expressions (chunking), automatization of syntax and phonology, formation of rules of the target language, fossilization and defossilization, and the issue of learning versus acquisition. Crowell (chapter 4) provides a neurobiological account for declarative memory and learning that involve mechanisms in the hippocampal region. She also illustrates how neural mechanisms in the hippocampo-entorhinal circuits may interact to construct a new neural network during L2 learning. Jones (chapter 5) discusses memory consolidation, focusing on declarative memory of lexical learning in the first language (L1) and language learning in SLA. Memory consolidation refers to how newly learned knowledge is transformed from a new memory into a stable memory. She argues that the Multiple Trace Theory, according to which multiple traces are made in the hippocampus due to repeated retrievals over time, and thus strengthening memories, provides a plausible mechanism for language learning. She delineates the implications of the theory in relation to individual differences in proficiency, adult L2 aphasics, and bilinguals. In her discussion of attention (chapter 6), Schuchert describes attention as a series of moments that involves interactive processes among various neurological functions. She hypothesizes five elements of attention that should be in alignment and attempts to explain the issues of defossilization and noticing. The book ends with Schumann's summary of the chapters and a conclusion.

Gullberg & Indefrey's edited volume consists of seven main chapters, each of which is followed by a brief critical commentary. This inclusion of a short response to each chapter is an excellent feature of the volume because it helps novice readers interact with each article critically, enabling them to read it from a different angle. The volume presents a good number of research findings and reviews with regard to the cognitive neuroscience of SLA and bilingualism. Unfortunately, linguistic theories are not discussed in the main chapters of the volume. What follows is a brief summary of the chapters included in the volume.

David Birdsong provides an overview of theoretical issues and relevant findings with regard to age effects in SLA. Specifically, based on results of previous neuroimaging and ERP studies, he illustrates the correlational link between age-related neurological changes (e.g., volumetric declines in the frontal lobe and prefrontal cortex and dopamine declines) and declines in cognitive functions that likely lead to age effects in L2 learning, production, and processing. H. B. M. Uylings reviews the structural and functional development of prenatal and postnatal human cortex in terms of its relation to particular phases and areas of cognitive development, illustrating diverse critical or sensitive periods that exist for neuroanatomical functions and changes over time. David W. Green, Jenny Crinion & Cathy J. Price discuss convergence, degeneracy, and control in bilingual speakers. Convergence refers to the concept which indicates that the neural representation of an L2 in L2 speakers becomes similar to the representation of that language in native speakers as proficiency improves. Current research (e.g. neuroimaging studies), they claim, offers evidence of convergence, as is also mentioned in Birdsong's chapter. However, they note that individual differences might also occur due to functional degeneracy, which refers to ‘the ability of structurally different elements to perform a similar function or achieve the same outcome’ (p. 112). They also mention that different levels of control are involved in numerous processes that bilingual speakers undergo during task performance. Similarly, A. Rodriguez-Fornells, R. De Diego Balaguer & T. F. Münte illustrate the issue of executive control involved in bilingual language processing. It is the prefrontal cortex, they propose, that activates a top-down control mechanism and interacts with a local bottom-up inhibitory control mechanism in bilingual language processing. Lee Osterhout, Judith McLaughlin, Ilona Pitkänen, Cheryl Frenck-Mestre & Nicola Molinaro report ERP studies in which not much L2 instruction is needed before L2 learners’ brains respond to word-level stimuli and (morpho)syntactically ill-formed sentences. Evidenced, however, is an ERP pattern that differs from that observed in native speakers, which indicates that L2 learners first rote-memorized word sequences and groups of morphemes in chunks, then later decomposed them and found the relevant morphosyntactic rules. Jutta L. Mueller's ERP experiments using Mini-Nihongo, a miniature version of Japanese, found that native and nonnative speakers differed in terms of the occurrence, and/or the distributions, of the early negativity or N400 components elicited in their response to the violations although their performance was almost indistinguishable on behavioral measures. Mueller claims that his findings can be interpreted as supporting strong versions of the critical period hypothesis (CPH), and that nonnative learners utilize a complementary mechanism (e.g., prosodic rather than syntactic processes). Indefrey reports a meta-analysis of 30 hemodynamic studies designed to investigate whether there are any noticeable differences in hemodynamic activation patterns during L1 and L2 processing in bilingual speakers. The meta-analysis indicates that, overall, the regions of the brain activated in L2 processing are similar to those activated in L1 processing (e.g., frontal and temporal regions), that is, there are no L2-specific brain regions activated during L2 processing, although some L2-specific sites seem to exist at the single-subject level.

All three books contribute to an enriching of our knowledge of SLA and bilingualism by putting crucial information together from three different perspectives: theoretical (Paradis), descriptive (Schumann et al.), and empirical (Gullberg & Indefrey). I shall now discuss how the books under review here complement each other and how some specific issues (age effects, the interface question, and cognitive control) are dealt with in the books, followed by a word of caution from Paradis on neuroimaging studies and concluding remarks.

The three books are complementary. Schumann et al.'s volume discusses neurobiological aspects of domain-general learning mechanisms and attempts to explain L2 learning and relevant linguistic phenomena in relation to neurobiological mechanisms of aptitude, motivation, procedural memory, declarative memory, and memory consolidation. On the other hand, rather than dealing with neurobiological mechanisms using all the technical details as seen in Schumann et al.'s volume, Paradis focuses on developing his neurolinguistic theory of bilingualism based on previous research findings and theoretical hypotheses. His theory construction does not particularly appear to require readers' familiarity with technical details of neural systems and related mechanisms but he assumes some knowledge of neurological concepts and notions, although not as much as Schumann et al.'s volume. Gullberg & Indefrey's edited volume is a collection of neurological research on SLA and bilingualism offering the most up-to-date summary of the state of affairs in the field. Compared to the other two books, this volume is more research-oriented and – based on findings from experimental neuroimaging and ERP studies – contains more focused discussions of how neural mechanisms are associated with such factors as age, proficiency, and exposure that are likely to cause interindividual variability in L2 processing and attainment. In other words, whereas the other two books, for the most part, describe a neurological theory, neural mechanisms, hypothetical constraints, and/or the applicability of the mechanisms/theory, Gullberg & Indefrey's volume shows what information cognitive neuroscience can, in reality, provide for our understanding of neural systems in the human brain in relation to L2 processing, production, and learning.

The chapters in Gullberg & Indefrey's book deal directly and indirectly with age effects in greater detail than the others, with several chapters presenting neurological evidence for age-related effects on L2 processing observed in neuroimaging and ERP studies. Thus, for example, Birdsong reviews the literature on how declines in the brain volume and dopaminergic system with increasing age affect neural activation in brain regions related to L2 processing, Mueller presents two Mini-Nihongo studies that seem to lend support to the CPH, and Indefrey discusses a meta-analysis showing the importance of age in acquisition. Unlike their general suggestions regarding the role of the CPH, Paradis is much more specific about age effects in his claims. He argues that age determines the extent to which learners acquire implicit linguistic competence only, and that the CPH does not delimit the scope and scale of metalinguistic knowledge or vocabulary, both of which are developed through controlled language processing, stored in, and retrieved from, declarative memory. He also suggests that the epicenter of the difference between L1 and L2 learning lies in the gradual inability to acquire language incidentally making use of the procedural memory system. He further notes that even if L2 learners’ performance appears identical to that of native speakers of that language, it is rather likely that their performance is a consequence of speeded-up control processing, not of the automatic processing native speakers normally engage in. Unfortunately, he does not elaborate on the issue using neural activation data; this may be due to his unwillingness to use evidence from neuroimaging studies that he considers questionable because of methodologically problematic issues, as suggested in the chapter on neuroimaging studies of the bilingual brain (chapter 6). Nevertheless, his argument that L2 learners (especially late learners) likely rely on metalinguistic knowledge and pragmatics to compensate for their insufficient implicit linguistic competence is evidenced in Mueller's finding that adult L2 learners strategically rely on prosody in L2 processing despite their high proficiency, and further confirmed in Indefrey's meta-analysis of hymodynamic studies. Somewhat oddly, none of the chapters included in Schumann et al.'s book specifically discusses the CPH. Although the volume is designed to illustrate L2 learning via the neurobiology of a domain-general learning mechanism, a section devoted to a focused discussion on the issue of age would have been a valuable addition to their overall explanatory attempt to illustrate SLA in a neurobiological perspective.

With regard to the implicit/explicit language processing in bilingual speakers, Paradis suggests that the acquisition of a new rule occurs through the controlled processing of the rule followed by the automatic processing, not of the same rule, but of an underlying implicit procedure (underlying linguistic competence). He argues adamantly that explicit metalinguistic knowledge does not become implicit linguistic competence as they are different in nature and supported by distinct neuroanatomical substrates, pointing out that ‘implicit competence and explicit knowledge are structurally unrelated and explicit knowledge is relevant to implicit competence development only indirectly, by focusing attention on items to be practiced and monitoring and correcting the output of implicit linguistic competence’ (p. 52). He takes the noninterface position, so to speak, which assumes two different types of knowledge and learning mechanisms (see also Krashen Reference Krashen1981; Paradis Reference Paradis1994). On the contrary, the interface position – regardless of its degree of strength – (DeKeyser Reference DeKeyser1997, Reference DeKeyser, Doughty and Williams1998; R. Ellis Reference Ellis1994, Reference Ellis2005; N. C. Ellis Reference Ellis2005) is supported in Schumann et al.'s volume; that is, explicit knowledge can become implicit. In her chapter on the neurobiology of procedural memory, Lee contends that neuroanatomy supports the interface between leaning and acquisition as the hippocampal region in charge of declarative memory that affects learning is interconnected with the basal ganglia responsible for procedural memory that affects acquisition, sharing the same cortical areas (see Figure 3.4, p. 68, for the schematic representation of the interrelationship between declarative memory and procedural memory). Crowell also suggests, in support of the interface position, that what seems to be a conversion from declarative to procedural knowledge is, in fact, a consequence of a strengthening of connections in the basal ganglia system through practice and sometimes an accompanying weakening of connections in the hippocampal system. Both Lee and Crowell advocate the interface position based on neurobiology and neuroanatomy, but without empirical evidence (e.g. evidence from neuroimaging research in experimental settings) that clearly indicates the connection between the procedural and declarative memory systems. As supportive evidence for his noninterface position, Paradis describes lesion studies on bilingual aphasia and amnesia that show the selective vulnerability of adult L2 speakers to lesions causing aphasia which influences implicit linguistic competence and amnesia which affects explicit knowledge. Lee, Crowell and Paradis mention neuroanatomy as their evidential stronghold, but discuss it in a completely different dimension. Although the interface question is complicated due to its neurocognitive nature, more empirical evidence from various kinds of brain research (e.g. lesion studies, hymodynamic neuroimaging, and ERP studies) would definitely pave the way to a better understanding of the issue from a neurological perspective. Regrettably, Gullberg & Indefrey's edited volume does not directly discuss the issue of interface.

Similarly to the still-controversial debate on the implicit versus explicit knowledge and/or memory (e.g., interface versus noninterface), different theoretical views on the issue of cognitive control are proposed in the volumes. Whereas Schuchert (in Schumann et al.'s volume) suggests that a domain-general attention-control system operates in SLA, involving the activation of such brain regions as the dorso-lateral prefrontal cortex and the anterior cingulate cortex (Rodriguez-Fornells et al. in Gullberg & Indefrey's volume also discuss this cognitive control issue, but focus more on a language-neutral perspective in bilinguals than on domain-generality), Paradis clearly denies the existence of a domain-general central executive presumed to control neurofunctional subsystems or domain-specific control systems, noting that ‘neurofunctional modularity (i.e., the organization of the cerebral substratum of cognitive functions into dedicated neural systems) is more pervasive than has generally been assumed and the hypothetical “central executive” may in reality be an abstraction referring to the summation of operations of local, domain-specific control systems’ (p. 146). He further suggests that neurofunctional modules can cooperate to implement complex cognitive tasks with relevant domain-specific control systems, which does not necessitate a higher-level supervisory system such as a central executive. His viewpoint deviates from the general acceptance of the prevalent notion that there is a unitary domain-general executive system that controls cognitive subsystems/functions. It does not appear excessively nonsensical, however, that a locally activated cognitive control system (language-related control system), for instance, supervises the distribution of neural resources to two languages in a bilingual. A domain-specific control system that deals with neural distributions, in this case, is specific to the language module; that is, it does not involve activating any domain-general central executive. Nevertheless, his argument against the need for a central executive must be empirically examined in diverse aspects of human cognition because research in cognitive psychology has shown some tasks requiring cognitive control involve activating a domain-general central executive. For instance, a domain-general central executive is considered as an important component of Baddeley's (Reference Baddeley1986, Reference Baddeley2007) multicomponent working memory (WM) system. Moreover, some researchers argue that individual differences in WM that lead to performance differences in complex cognitive tasks derive mainly from variation in domain-general executive attention processes (Engle, Kane & Tuholski Reference Engle, Kane, Tuholski, Miyake and Shah1999; Engle, Tuholski, Laughlin & Conway Reference Engle, Tuholski, Laughlin and Conway1999; Kane, Bleckley, Conway & Engle Reference Kane, Bleckley, Conway and Engle2001; Engle Reference Engle2002; Kane, Conway, Hambrick & Engle Reference Kane, Conway, Hambrick and Engle2007); it should be noted, however, that their theoretical explanation, the executive attention theory, is not without controversy (see Oberauer, Süβ, Wilhelm & Wittmann Reference Oberauer, Süβ, Wilhelm and Wittmann2003; Colom, Shih, Flores-Mendoza & Quiroga Reference Colom, Shih, Flores-Mendoza and Quiroga2006; Oberauer, Süβ, Wilhelm & Sander Reference Oberauer, Süβ, Wilhelm and Sander2007 for challenges to the executive attention theory). It is still inconclusive and has yet to be clearly examined whether a domain-general central executive exists/functions as a master supervisory system in bilinguals.

Neuroimaging and ERP studies are, to a varying degree, cited and discussed in all three books under review, but Paradis appears reluctant to make use of neuroimaging studies as the basis of his theory construction. He points out that ‘neuroimaging studies of language in general and of bilingual language functions in particular suffer from a lack of linguistic sophistication’ (p. 183). He argues, in his review of neuroimaging studies, that the tasks are often nonlinguistic and not specifically designed to involve implicit linguistic competence, that the stimuli used are often not language-relevant or language-relevant materials are utilized in nonlinguistic contexts, and that researchers do not distinguish between implicit competence and explicit knowledge in interpreting and generalizing their results. Also problematic, according to Paradis, is the fact that no replication studies have been conducted that employ the same task and technique administered to similar populations. He further argues that ‘until we have the means to decide which experimental task, used with which experimental procedure, according to which protocol, will reveal the relevant areas, any result is moot since it can be contradicted by another result and there is no way to determine which is right (i.e., valid)’ (p. 184). Results must be interpreted with caution and only in consideration of the constraints of the neuroimaging technique employed. The validity and reliability of research findings, and hence the burden of further proof are still in the hands of researchers.

As I mentioned earlier, there are only a few published books that delineate SLA and bilingualism in a neurolinguistic perspective or in terms of research findings in cognitive neurosciences. The volumes selected for this review are three of the very few publications, if not the only ones, that can be found at this juncture, which reflects the current inchoate stage of the cognitive neuroscience of SLA and bilingualism. Around the time these volumes emerged in print, the field of SLA and bilingualism moved forward by making apparent efforts to investigate SLA and bilingualism in neurobiological and/or neuropsychological perspectives (e.g. Tokowicz & MacWhinney Reference Tokowicz and MacWhinney2005; Chen et al. Reference Chen, Shu, Liu, Zhao and Li2007) and this area of inquiry has been discussed and recommended even in interaction research (e.g. Mackey Reference Mackey2006); for instance, if in fact noticing feedback provided during interaction is of critical importance in L2 learning, as Schmidt (Reference Schmidt1990, Reference Schmidt and Schmidt1995, Reference Schmidt and Robinson2001) argues, investigating cognitive processes of noticing may enrich our understanding of the extent to which conversational interaction contributes to SLA. The real contribution of the volumes to the field lies in the possibility that researchers on SLA and bilingualism may become more aware of, and more interested in, this rarely-tapped repository of knowledge on the language–brain relationship. Nevertheless, it is undeniable that the cognitive neuroscience of SLA and bilingualism is still in its infancy and that far more research is no doubt needed to deepen our knowledge on SLA and bilingualism in this respect in the years to come.

In sum, Schumann et al.'s edited volume illustrates fundamental neural mechanisms through microscopic analyses of the neurobiology of domain-general learning mechanisms and how they can explain relevant linguistic phenomena. Furthermore, it shows how these mechanisms are related to L2 development, providing a valuable opportunity for L2 researchers who have little or no background in neurobiology to broaden their knowledge base. However, a somewhat extensive use of technical terms in some chapters may be quite intimidating and frustrating to novice readers. Also, editing should have been more carefully conducted; the entire first paragraph under ‘A language learning scenario’ on page 102 mistakenly reappears on page 103. Paradis’ monograph also provides important insights into bilingualism in a neurolinguistic perspective, focusing mainly on neurofunctional aspects of his theory, but not much on linguistic theories. The methodological issues in the neuroimaging research paradigm that he points out should be considered in future research. Gullberg & Indefrey's edited volume also contributes to our understanding of the cognitive neuroscience of SLA and bilingualism, presenting an array of research findings and reviews of relevant literature. One great feature of their volume is the inclusion of a brief critical response to each chapter, which offers an opportunity for readers to consider different angles from which they can interpret the contents of each article. As in Paradis’ work, it is unfortunate that no particular linguistic theory was at the center of all the discussions in the volume. However, despite their shortcomings, all three books reviewed here have contributed to the introduction of a new chapter to SLA and bilingualism research and generated intellectual excitement for those interested in the cognitive neuroscience of SLA and bilingualism as each offers a unique perspective and a significant review of established and still intensely-debated fundamental issues. As Paradis comments on neuroimaging studies, the current state of affairs in the cognitive neuroscience of SLA and bilingualism may still be at the ‘poking’ stage, but taking into account the theoretical and empirical issues presented in these volumes and incorporating them into future neuroimaging and ERP studies, mutatis mutandis, will surely lead to theoretically more valid and methodologically more reliable outcomes. Lastly, regarding future publications on the cognitive neuroscience of SLA and bilingualism, it is of crucial importance that linguistic theories or language learning theories be discussed in greater detail because it is not brain activation but the relationship between brain activation and language learning that we are interested in. Another issue not covered in the volumes, but probably one that should be included in future publications, is how research on the cognitive neuroscience of SLA and bilingualism contributes to facilitating L2 development, that is, how our knowledge of all these neurobiological and neuropsychological phenomena promotes actual learning. Nonetheless, considering that the cognitive neuroscience of SLA and bilingualism has just begun its own voyage, these volumes are just the beginning of a newly-launched research area, and hence there is much room for theoretical development and much more to discover.