1. Introduction
The proposal that the language we speak influences how we think and understand reality, more commonly referred to as the principle of linguistic relativity (LR) or the Whorfian hypothesis, was for many decades seen as a question not worthy of linguistic enquiry. It has often been equated with the idea that speaking a particular language determines the way you think (i.e. linguistic determinism) and has for this reason been regarded as scientifically disproven and fundamentally flawed. In recent years, however, LR has become a vogue topic again. The role of language in the conceptualization of domains such as space, motion, time, objects and substances, color and emotions is currently being explored under the premise of LR (Davidoff, Davies & Roberson Reference Davidoff, Davies and Roberson1999; Levinson Reference Levinson, Gentner and Goldin-Meadow2003; Cardini Reference Cardini2010). The theoretical orientations within which LR is investigated range from studies that argue that differences in linguistic structure lead to differently organized cognitive structures, to formulations postulating that different linguistic categories simply make certain aspects of reality more salient than others (without altering underlying cognitive structures or processes).
The majority of studies that have been conducted within the paradigm of neo-Whorfianism focus on crosslinguistic comparisons of monolinguals: bilingual speakers have been studied only to a limited degree within this paradigm. This is a glaring omission, especially considering Grosjean's (Reference Grosjean1982) assertion that more than 50% of the world's population is bilingual, and Aronin & Singleton's (Reference Aronin and Singleton2008) view of multilingualism as the ‘new linguistic dispensation’ in a changing and increasingly globalized world. Historically, the relationship between language and cognition from the bilingual perspective had been largely ignored (for exceptions in earlier work, see Brown & Lenneberg 1954; Ervin Reference Ervin1955), but this is an area which is now rapidly making progress. Evidence of this can be found in special issues of journals such as Bilingualism: Language and Cognition (2011), Vigo International Journal of Applied Linguistics (2011),Footnote 1 and conference panels organized around the topic (American Association of Applied Linguistics 2007; International Symposium on Bilingualism 2007, 2009). The four books included in this review are evidence of the recent growth of the field and contribute to the study of language contact in the individual, and its effects on the linguistic system, the conceptual system and various cognitive functions. All the volumes acknowledge the role that the Whorfian hypothesis has played in the debate on language and cognition, but afford different prominence to it. The inclusion of the Whorfian hypothesis in a discussion on language contact in the individual is important, as it opens up the possibility that such contact might have implications not only for the linguistic system, but also for various cognitive processes. What also makes LR interesting in studies on bilingualism and cognition is that the crosslinguistic differences between the languages bilinguals know and use could be responsible for maintenance of or changes in their conceptual systems. These possibilities are explored in this review, which will evaluate how the four books conceptualize the relationship between language and cognition in bilinguals, and how they approach this topic in relation to the Whorfian hypothesis.
The books will be discussed in order of date of publication, so that readers can see how the field has progressed, from the publication of Jarvis & Pavlenko (2008) to the two newest volumes in the review (Cook & Bassetti 2011; Pavlenko 2011).
2. Content overview
The earliest publication, Jarvis & Pavlenko (2008), offers a general introduction to crosslinguistic influence (CLI)Footnote 2 at both the linguistic and conceptual level. In seven chapters, the book covers the history of the study of CLI, the methodology used to investigate it, the phenomena of linguistic transfer, conceptual transfer (where differences between languages lead to L1-induced transfer on the conceptual as well as the structural level) and conceptual change (where the acquisition of an L2 causes conceptual changes in the L1) and the factors that interact with CLI, finally drawing conclusions and making recommendations for further research.
The volume edited by Han & Cadierno (2010) clearly delimits the type of bilingualism in focus by placing the LR hypothesis in the context of adult second language acquisition (SLA). The eight studies in this volume use the ‘thinking-for-speaking’ paradigm as the specific formulation of LR, covering topics such as the effects of verbal as well as non-verbal (gesture) ‘thinking-for-speaking’ on the reading of space and motion events (Cadierno; Coventry, Valdés & Guijarro-Fuentes; Hasko; Stam), and on linguistic features such as glossing and the use of articles and grammatical morphemes (Ekiert; Han; Stringer). In Odlin's chapter the focus is on conceptual transfer; it includes a discussion of what might count as evidence of transfer at the conceptual level.
Pavlenko (2011) offers a snapshot of current research investigating CLI on verbalized as well as non-linguistic thought. In the eight chapters, the subjects include bilinguals who acquired their L2 at school, immigrants acquiring an L2 in their adoptive countries, and simultaneous bilinguals. The book does not give prominence to the debates surrounding the Whorfian hypothesis but, instead, uses them to frame the discussion on thinking and speaking in two languages. Pavlenko provides a historical overview of language and thought in Chapter 1, and a synthesis of the findings of recent studies on ‘thinking and speaking in two languages’ in Chapter 8. One chapter, by Athanasopoulos, actively engages with the more traditional tenets of the Whorfian hypothesis in investigating strictly non-verbal cognition, while the rest examine the ‘thinking-for-speaking’ paradigm in different instantiations ranging from ‘thinking, speaking and gesturing’ (Bylund; Gullberg; Schmiedtova, von Stutterheim & Carroll) to ‘thinking and naming’ (Malt & Ameel; Pavlenko).
Cook & Basseti (2011) frame their volume around the relation between language and cognition in bilinguals. The bilinguals investigated include those acquiring an L2 in formal contexts, L2 speakers living in an L2 context, and simultaneous bilinguals. The first part of the volume (Part A) provides a background to the LR principle in monolingual contexts, viewed from different traditions of scholarly enquiry. This includes the anthropological tradition (Lucy) and that of cognitive linguistics (Evans). Part B addresses bilingualism and cognition. As cognition is a key concept in this volume, there is an active engagement with the Whorfian hypothesis and its consequences for bilinguals in a number of cognitive processes, such as color similarity judgment, memory of spatiality and the effect of emotions on recall and autobiographical memory (see the chapters by Athanasopoulos; Coventry et al.; Knickerbocker & Altarriba). A further feature of this volume is that it offers, in Part C, a discussion of the practical implications of research on bilingual cognition for education, translation, advertising and intercultural communication. For example, Cook, in his chapter on language education, proposes that in the light of research indicating that the acquisition of an L2 might change bilinguals’ conceptualization patterns, teaching should not focus on ‘forcing the concepts of the L2 on the learner’ but, instead, should allow for the development of a blend of the L1 and L2.
This brief overview has already given an idea of the complexity of the field, as well as the exciting research opportunities it offers. The next section will show how each of the volumes contributes to the investigation of language contact in the individual, with a particular focus on verbal and non-verbal cognition.
3. Discussion and evaluation
Each of the four books addresses the question of bilingual cognition in its own way, a fact that indicates the plethora of methodologies available for the investigation of the relationship between language and cognition in bilinguals. The individual studies include different language pairs – Danish and Spanish/German (Cadierno in Han & Cadierno), Spanish/Swedish (Bylund in Pavlenko), Chinese/English (Chen & Su in Cook & Bassetti), Russian/English (Jarvis & Pavlenko) – a range that helps to further our understanding of how language differences might influence bilingual cognition. The dominance of English is still prevalent, and is shown by Cook & Bassetti (p.169) to be both an area of concern and a research focus for future enquiry.
The varied framings of the interaction between language and cognition in bilinguals found in the four books mean that their findings are informed by different theoretical orientations, and contribute in different ways to researching CLI in language and cognition. One of the major contributions of Jarvis & Pavlenko is to show how ‘language transfer’ and the Whorfian hypothesis intersect in the area of conceptual transfer and change, via a thorough and insightful epistemological-historical account of language transfer phenomena and a clear statement of their position on the LR hypothesis and how it is approached in their book. They acknowledge the theoretical roots of their notion of ‘conceptual transfer’ (primarily Whorf, but also Piaget and Vygotsky), but do not subscribe to exactly the same views. Their view is similar to that of Whorf, in that they argue that some concepts are not universal but are formed through the language socialization process. These ‘language-mediated concepts’ are thus affected by CLI. However, Jarvis & Pavlenko differ from Whorf in that their investigation of conceptual transfer starts with language and, via cognition, ends with language (p. 116), whereas the more traditional approach to LR starts with language and ends with cognition. Subsequently, their chapters on conceptual transfer and change are clearly formulated and insightful, especially with regard to their typology of conceptual change (p. 156), which has been further developed by Pavlenko (2011: 246). Chapter 2, on methodology, is particularly useful, as it gives guidelines for rigorous empirical investigations and attempts to provide a methodology for isolating, identifying and measuring the effect of one language on another.
Another significant feature of this volume is Chapter 6, on factors that influence transfer, which shows a nuanced understanding of the effects of one language on another, and points to the factors that interact with CLI. These are divided into five broad categories: linguistic and psycholinguistic factors; cognitive, attentional and developmental factors; factors related to the learning environment; factors related to cumulative language experience and knowledge; and factors related to language use (pp. 174–207). However, perhaps partly reflecting the state of the field at that time (2008), the book suffers from a lack of studies that focus on multilingual contexts (a point that I will return to later in the review) or childhood bilingualism. As Jarvis & Pavlenko provide a synthesis of existing research, the studies they refer to in their volume mostly focus on adult SLA and situations of immigration or migration, where formerly monolingual speakers have to learn an L2 in order to integrate into mainstream society. This shortcoming is, however, pointed out by the authors themselves (pp. 234–235). This book provides the most comprehensive account of CLI, as it includes CLI not only on language but also on cognition, unlike earlier works on the phenomenon (e.g. Gass & Selinker Reference Gass and Selinker1983; Odlin Reference Odlin1989). It should be noted, however, that certain areas of research have received much more attention since the publication of the book. One such area is the effects of language on non-verbal cognition (see, for example, Thierry et al. Reference Thierry, Athanasopoulos, Wiggett, Dering and Kuipers2009; Boroditsky, Fuhrman & McCormick Reference Boroditsky, Fuhrman and McCormick2010; Trueswell & Papafragou Reference Trueswell and Papafragou2010), which we can expect to be extended to bilingual speakers (see Athanasopoulos Reference Athanasopoulos2009). Jarvis & Pavlenko (2008), however, can be credited for their role in this growth. The book is recommended as an invaluable source for researchers, teachers and students in the fields of both bilingual cognition and linguistic transfer research, as it provides an excellent historical overview, guidelines on the use of research methodology and a thorough synthesis of CLI research.
Han & Cadierno (2010) make a commendable attempt to investigate bilingual cognition in the context of adult SLA. Given the clear delimitation of the type of bilingualism investigated, this collection of papers could have made for a clearly focused volume which would have been of interest to SLA researchers, language teachers and practitioners. Unfortunately, this potential is not realized. This is the only volume in this review that explicitly refers to ‘linguistic relativity’ in its title, so one might expect LR to be actively and accurately investigated. However, the title is somewhat misleading because, of the four volumes, this book engages the least with LR. The book seems to equate LR with ‘thinking-for-speaking’, while in fact thinking-for-speaking is a particular way of investigating aspects of LR. The editors seem to subscribe to the view of ‘weak’ vs. ‘strong’ versions of linguistic relativity (p. xvi), characterizing Whorf's view as linguistic determinism. In fact, Whorf never proposed that language determines thought, but rather that:
The linguistic relativity principle. . .means in informal terms that users of markedly different grammars are pointed by their grammars toward different types of observation and different evaluations of externally similar acts of observation and hence are not equivalent as observers but must arrive at somewhat different views of the world. (Whorf Reference Whorf and Carroll1956b: 221; my emphasis)
The strong vs. weak conceptualization of LR is considered to be a specific reading rather than a reflection of the work of Whorf (see Pavlenko 2011). Many of Whorf's (Reference Whorf and Carroll1956a) assertions were not deterministic: he states, for example, that ‘. . .there are connections but not correlations or diagnostic correspondences between cultural norms and linguistic patterns’ (1956a: 158) (see also Athanasopoulos in Cook & Bassetti, Pavlenko 2011 and Cook 2011).
This inconsistency in the interpretation of LR is a feature not only of the Preface but also of several of the chapters that follow. Hasko (p. 35), for example, refers to ‘the weaker Neo-Whorfian view’, thus buying into the idea that Whorf's original thesis was a stronger view (i.e. linguistic determinism). Stam (p. 61) alludes to Lucy as a proponent of the strong version of LR, whereas, in fact, Lucy (Reference Lucy1992: 3) rejects the strong vs. weak dichotomy and argues that the goal of investigation into linguistic relativity should not be to disprove the linguistic determinism position, or to prove that language influences thought, but instead to investigate precisely what the influences are of language on thought, how important these influences are, and how and when they operate. Han (p. 161) and Odlin (p. 192), however, are more careful in their account of LR and thinking-for-speaking. Following Slobin's (Reference Slobin, Gumperz and Levinson1996) original thesis, Han calls thinking-for-speaking a modified version of LR, while Odlin, in his chapter on conceptual transfer, discusses the different orientations that can be adopted in the investigation of the hypothesis.
The volume does make a contribution in terms of methodology by reporting on longitudinal studies. In particular, Stam's longitudinal case study based on multimodal visual data is an important contribution. This study found that the L2 affected the path of motion in gestures used when speaking the L1 and shows how conceptualization develops over time, offering an alternative to the more common cross-sectional research design. Han's choice of participant in her own longitudinal study is interesting in that the subject chosen is not the typical university student, but an older Chinese/English professional. Han interprets the results of the study (the participant's use of articles and plurals in L2, which is more typical of Chinese than English), to mean that even after prolonged exposure (12 years) to English in an English-speaking environment, ‘the subject's mind largely remained L1-relativized’ (p. 181). This interpretation, however, can be questioned. L1 influences on the L2 linguistic level do not necessarily lead to effects at the conceptual level, and no evidence other than morpheme analysis is offered to show that these effects take place at such a level. Generally, the volume does not deliver the hoped-for exploration of LR and, in fact, some of the studies could have been more fruitfully examined within the paradigm of linguistic transfer. Han's study is not the only one that fails to deliver clear evidence that CLI operates on more than the linguistic level. Stringer, in his discussion of glosses, argues that differences in the mental lexicon will lead to differences in thinking-for-speaking, without clearly distinguishing between concepts and words. This is a reminder of Casasanto's (Reference Casasanto2008: 67) view that often the only evidence offered in earlier investigations of linguistic relativity that people who speak differently think differently, is that they speak differently.
In contrast, both Cook & Bassetti (2011) and Pavlenko (2011) discuss the complexities surrounding the Whorfian hypothesis in a systematic and nuanced way. Pavlenko offers a historical account of language and thought, but emphasizes that, following the tradition of Lucy (Reference Lucy1992), the goal of the volume is not to provide evidence for or against linguistic relativity, but to contribute to the scholarship around language and cognition by expanding its focus to bilingual speakers. Cook & Bassetti, on the other hand, actively engage with the hypothesis. Both volumes constitute significant contributions to the field, exploring a range of language pairs (including Swedish/Spanish, Polish/English, Italian/German and Dutch/French) in a range of different domains (event construal, gesturing about motion, object naming, color perception, emotion, etc.). Both also report on studies that indicate that language contact affects the individual's conceptualization, verbal and non-verbal cognition in a number of different conceptual domains and cognitive functions. Bylund, in Pavlenko's volume, argues that grammatical aspect may make salient certain aspects of an event when speaking, while Malt & Ameel show that simultaneous bilinguals form object-naming patterns that do not entirely conform to those of monolingual speakers of the two languages they use. Besemeres, in the Cook & Bassetti collection, points out how bilingual narratives can aid the understanding of the interplay between language and emotions in bilinguals (see also the chapter in the same volume by Knickerbocker & Altarriba, on emotions). These two volumes, however, acknowledge that this effect of language on cognition or conceptualization is tied up with other factors, including age of onset of acquisition, proficiency and use of both languages, and any other languages known (Cook & Bassetti 2011: 177–178), and try to offer suggestions (both in terms of methodology and theory), on how these factors might be isolated. Athanasopoulos, in his chapter on bilingual cognition and color in Cook & Bassetti, invites further investigation of this phenomenon and believes that advances made in methodology such as ‘neurophysiological techniques that may reveal very early perception of color in the visual processing stream’ (p. 258) can be utilized more accurately to disentangle the effect of language from other variables. Pavlenko lists possible predictors of conceptual restructuring in bilingual cognition, such as age of L2 acquisition, length of exposure and frequency of language use (p. 248), which should be considered, isolated or controlled for in order to determine more clearly the effect of language differences on bilingual cognition (cf. Chapter 6 in Jarvis & Pavlenko 2008).
Both volumes make a significant contribution to the field, and handle the debate around linguistic relativity in a sound, scholarly manner. In particular, the opening and closing chapters by Pavlenko provide an interesting juxtaposition of the historical routes taken by the field, and its current state. Cook & Bassetti's volume also provides insight into possible practical concerns around bilingual cognition, which makes their volume particularly useful for language practitioners as well as scholars and students.
However, a criticism of both volumes is the lack of focus on childhood bilingualism. Only the chapter by Siegal et al. in Cook & Bassetti addresses the issue of bilingual cognition in children with a specific focus on theory of mind. This is probably a reflection of a shortcoming in the field itself. Studies on linguistic transfer in bilingual children (Müller Reference Müller1998; Nicoladis Reference Nicoladis2003) and crosslinguistic studies on the cognitive development of children (Gopnik, Choi & Baumberger Reference Gopnik, Choi and Baumberger1996; Gopnik Reference Gopnik, Bowerman and Levinson2001) have opened up possibilities for exploring bilingual cognition in children. More research is needed on the effect of language on cognition for simultaneous and early bilinguals. Very little is known about the interaction of languages and cognitive development in bilingual children from a Whorfian perspective. More is known about the effects of bilingualism on general cognitive development (see Bialystok Reference Bialystok and Bialystok1991, Reference Bialystok2001) than about the influence that one language exerts on another and the effects of bilingualism on language-specific cognitive behavior.
A second omission even in the insightful and excellent accounts of Cook & Bassetti and Pavlenko is the question of bilinguals in multilingual contexts. Here, in particular, the absence of attention to Africa and Asian countries outside East Asia (such as India) is disappointing. Not only are there no studies reporting on the African context; no African language is included in any of the language pairs under investigation. Such an extension of the book's subject-matter would have been desirable for a number of reasons. Firstly, it would have added to the ecological validity of studies on language contact in the bilingual. Secondly, these multilingual contexts provide possibilities for reflection on our current theoretical agenda. For example, Banda (Reference Banda2009: 5), following Myers-Scotton (Reference Myers-Scotton1993), argues that in Western contexts monolingual communities live side by side, while in African contexts multilingual communities live side by side. He further argues that the acquisition of L2s in African contexts entails learning two or three languages outside formal environments, and the former colonial languages in school. In terms of language use, Banda (Reference Banda2009) argues that one or two local languages, together with the former colonial language, are usually used in these contexts as lingua franca. Terminology that is thus currently used in a commonsense way in studies on bilingualism and subsequently bilingualism and cognition, such as ‘length of stay’, ‘proficiency and use of both languages’ (Cook & Bassetti 2011: 177; Pavlenko 2011: 248–241) cannot be transplanted to these contexts without further definition. These concepts will need to be problematized; for example, length of stay in a context where multilingual communities live side by side will have to be replaced not only by length of exposure but also by intensity of exposure. Proficiency in two distinct languages will be difficult to measure in communities where two languages are used in hybrid and fragmented ways for a variety of different functions. Which baseline norm will be used to establish levels of proficiency? Will it be the monolingual native speaker (a rarity in the majority of African countries), or will we see the adoption of Cook's multicompetence thesis (the idea that bilinguals have a language competence different from that of monolingual speakers of either of the languages they know)? A discussion of the extent to which our knowledge about bilingualism and cognition is constrained by the specific contexts where research on this topic is currently carried out would certainly have enriched these volumes.
Another problem common to all the volumes is that they suffer a bias due to the sampling of one particular population in bilingual research, namely, the student population. In Han & Cadierno, for example, four out of the six empirical papers have students as participants. In Pavlenko, out of the six papers that report on ongoing empirical work, five have either students or people with academic qualifications as participants, while in the remaining paper this information is not given. Students might be a convenient group for researchers and, like any other group of participants worthy of investigation, but by no means are they the only group. Currently, what we know about bilingual cognition is largely confined to this group.
4. Conclusion
Language-specific cognition in bilinguals is a rich and exciting research field, but also a challenging one, due to the complex nature of bilingualism and the need to draw from different disciplines in the study of language and cognition. All the volumes in the review reflect the vibrancy of the field. Significant contributions have been made in methodology, the number of different language pairs under investigation, the domains investigated, and the different theoretical orientations to this phenomenon. The studies in these volumes seem to indicate that bilinguals do perceive and negotiate their way around the world differently from monolinguals, and that not only bilingualism but, more especially, the differences between the languages bilinguals know can act as a variable for these differences.
A number of problematic issues are also reflected in the four volumes, one being what counts as evidence for any conceptual effects of crosslinguistic influence. LR should not be used simply because it is a vogue topic. The investigation of LR should reflect its richness and complexity and be accompanied by the kind of research it deserves. The Han & Cadierno volume, in particular, fails to give an adequately rich and complex account of LR. A systematic and thorough investigation of LR can also give valuable information about CLI in general, such as the kinds of evidence to use in support for CLI on the linguistic or conceptual level and the best methods and theoretical paradigms to use in such an exploration. Cook & Bassetti do succeed in reflecting this complexity.
These four books also demonstrate that our view of bilingualism itself needs to be expanded. There seems to be an overreliance on student populations as participants in studies on bilingual cognition. The pool of participants should be extended to include children, people living in multilingual contexts, bilinguals who are not literate in an L2 or even in any of the languages they speak, and people who learn hybrid urbanized varieties, together with the languages spoken at home as L1, while acquiring an L2 for school. Research into bilingual cognition acknowledges diversity, but diversity within bilingualism is not reflected in the most up-to-date volumes in the field. In light of Evans & Levinson's assertion (Reference Evans and Levinson2009) that the only true universal property of language might be the capacity of the human mind to deal with diversity in language, the field of bilingual cognition can only truly progress if this notion of diversity is further extended to not only investigate different languages and different conceptual domains, but also different contexts and different types of bilingual language users. Diversity is not only a theoretical concern: language practitioners and language teachers work in increasingly diverse contexts with bilinguals who are not always literate, university-educated individuals. The four volumes all contribute in different degrees to the field of CLI in language and cognition, though our knowledge of this phenomenon is currently restricted to a limited range of contexts.
Marcelyn Oostendorp holds a Ph.D. in linguistics from the Department of General Linguistics, Stellenbosch University, South Africa. Her research work focuses on the effects of the L2 as language of teaching and learning on conceptual fluency in higher education.
