Research on the relationship between language and cognition in bilinguals has often focused on general effects that are common to bilinguals of all language backgrounds, such as the positive effects of bilingualism in various areas of cognitive development (e.g., Bialystok, Reference Bialystok, Kroll and De Groot2005; Karmiloff-Smith, Reference Karmiloff-Smith1992). However, there are also language-specific effects in the relationship between language and cognition in bilinguals that emerge in the form of cross-linguistic influence and, in many cases, these cross-linguistic effects do not appear to be confined to purely linguistic (e.g., phonological, morphological, syntactic, semantic) phenomena. For example, bilinguals’ choice of words for referring to objects and actions, as well as their choice of syntactic and discursive structures for referring to events and situations, often reflect ways of conveying meaning and intentions that are specific to particular language backgrounds.

The theme of this special issue is conceptual transfer, an exciting area of research in bilingualism and second language acquisition that deals with cross-linguistic differences and cross-linguistic influences in the mental construction and verbal expression of meaning. This is an area of research that has been garnering increased attention in recent years and whose goals and scope have been coming into increasingly clearer view (see Jarvis, Reference Jarvis2007; Jarvis & Pavlenko, Reference Jarvis and Pavlenko2008). It is not always easy to see the big picture in this area of research, however, given that the relevant published work has appeared primarily in isolated studies (e.g., Athanasopoulos, Reference Athanasopoulos2009; Brown & Gullberg, Reference Brown and Gullberg2008; Pavlenko, Reference Pavlenko2003) rather than in contextualized collected volumes like the present special issue. There also exist some misconceptions about the objectives of this area of research, the assumptions it rests on and its relationship to linguistic relativity (e.g., Lucy, Reference Lucy1992; Whorf, Reference Whorf and Carroll1956), thinking for speaking (Slobin, Reference Slobin and Perdue1993, Reference Slobin, Gumperz and Levinson1996) and other related research frameworks (e.g., Levelt, Reference Levelt1989; von Stutterheim & Nüse, Reference von Stutterheim and Nüse2003). With these considerations in mind, my purpose in putting together this special issue has been to compile a carefully chosen selection of new, well-designed, theory-grounded empirical investigations that collectively illustrate the range of goals, the breadth of scope and the emerging trends of this area of research, and which also, of course, add to the body of knowledge of this phenomenon by bringing new findings to light. I have also sought out contributors who are not only interested in cross-linguistic effects arising from potential cognitive (e.g., mental processes) and conceptual differences across speakers of different languages, but who are also cautious in interpreting their data. The studies in this issue are furthermore noteworthy with respect to the diversity of L1s (Dutch, Greek, Japanese, Russian, Spanish, Turkish) and L2s (English, German, Swedish) represented, as well as with respect to the attention given to cross-linguistic effects in both directions, from L1 to L2 and L2 to L1.

Although not all of the studies in this special issue deal explicitly with the notion of conceptual transfer, this nevertheless represents the context of the present forum, and it is therefore necessary for me to explain what this means in more detail. The term conceptual transfer can be understood at three different levels: as an observation, approach and hypothesis. First, at an informal level, it refers to the observation that language learners, L2 users, bilinguals and multilinguals from different language backgrounds sometimes express objects, events, qualities and relationships in conceptually different ways. Anecdotal evidence for this abounds, such as in my own informal observation that bilinguals from different L1 backgrounds seem to differ relatively predictably with respect to whether they refer to their breakfast cereal with singular versus plural pronouns (e.g., This is good! versus These are good!), as if they view it slightly differently. Observations like these are not simply anecdotal, however; a number of them have also arisen in the context of empirical investigations. They include observations about how the same stimuli are often expressed by bilinguals (or language learners or multilinguals) from different language backgrounds as exhibiting different attributes (e.g., Sera, Berge & Del Pintado, Reference Sera, Berge and Del Pintado1994), different spatial relationships (e.g., Ijaz, Reference Ijaz1986; Jarvis & Odlin, Reference Jarvis and Odlin2000; Pavlenko, Reference Pavlenko2003), and differing characteristics related to temporal relationships, motion and event structure (e.g., Hohenstein, Eisenberg &Naigles, Reference Hohenstein, Eisenberg and Naigles2006; von Stutterheim, Reference von Stutterheim, Foster-Cohen and Pekarek Doehler2003). They also include empirical observations about how bilinguals from certain language backgrounds tend to express some types of information that bilinguals from other language backgrounds tend not to, such as information about whether an event was personally witnessed (Dragiev, Reference Dragiev2004) or whether an event is currently in progress (e.g., Carroll & Lambert, Reference Carroll, Lambert, Byrnes, Weger-Guntharp and Sprang2006; von Stutterheim & Lambert, Reference von Stutterheim, Lambert and Hendriks2005). In short, conceptual transfer as an observational phenomenon is essentially the observation that some cases of cross-linguistic influence seem to go beyond mere issues of structural and semantic well-formedness, and instead extend to conceptual meaning – particularly to the types of meaning that require a person's assessment of physical realities, social conditions and other real and imagined contexts.

At a more formal level, conceptual transfer refers to an approach to research. More specifically, it refers to the investigation of cross-linguistic influence through the perspective of the recent empirical and theoretical developments in cognitive linguistics (which also draws from cognitive psychology). Some of the more prominent theories of cognitive linguistics include Cognitive Grammar (e.g., Langacker, Reference Langacker2008), conceptualist semantics (e.g., Talmy, Reference Talmy2000), construction grammar (e.g., Croft, Reference Croft2001), metaphor theory (e.g., Kövecses, Reference Kövecses2005) and mental space theory (e.g., Fauconnier & Turner, Reference Fauconnier and Turner2002). The cognitive linguistics paradigm also includes theories about the nature of mental concepts and their relationships with word meanings and the functions and meanings of grammatical constructions (e.g., Lakoff, Reference Lakoff1987). Approaching cross-linguistic influence through one or more of these theoretical perspectives is the meaning of conceptual transfer that best characterizes the present special issue. Although the specific objectives of the articles in this collection vary broadly, what unites them is their grounding in an overarching cognitive framework that views language use as being governed not just by well-formedness constraints, but also by more basic cognitive systems and abilities, including perception, memory and categorization (Langacker, Reference Langacker2008, p. 8).

Work on conceptual transfer has not yet approached cross-linguistic influence from all of the theoretical perspectives contained within the paradigm of cognitive linguistics. Most studies that have investigated cross-linguistic influence within the framework of cognitive linguistics (and which could therefore be described as studies of conceptual transfer) have so far been grounded in either theories of concepts (e.g., Jarvis, Reference Jarvis1998; Pavlenko, Reference Pavlenko2003) or theories of conceptualist semantics (e.g., Cadierno, Reference Cadierno, Robinson and Ellis2008; Inagaki, Reference Inagaki2001). The latter is particularly true of studies that have pursued Slobin's (Reference Slobin and Perdue1993, Reference Slobin, Gumperz and Levinson1996) notion of thinking-for-speaking, which is the notion that speakers organize and structure the content of their thoughts in language-specific ways just prior to verbalization, even if at a more general level their patterns of thinking and perception might not be language-specific. Although thinking-for-speaking could be investigated within the framework of just about any of the theories of cognitive linguistics, Slobin himself has concentrated on how it relates to Talmy's work on the cognitive semantics of motion events (e.g., Slobin, Reference Slobin, Strömqvist and Verhoeven2004, Reference Slobin, Maeder, Fisher and Herlofsky2005, Reference Slobin, Hickmann and Robert2006), and this is also true of most of the work in this line of research, including the articles in the present issue by Brown and Gullberg, and by Daller, Treffers-Daller and Furman.

The article in the present issue by Flecken is also explicitly tied to Slobin's thinking-for-speaking framework, but instead of relying on Talmy's theory of conceptualist semantics, is grounded in the hypotheses of von Stutterheim and Nüse (Reference von Stutterheim and Nüse2003) concerning the relationship between grammatical aspect and the mental perspectives that speakers (including both monolinguals and bilinguals) take on events during speech-planning processes. The article by Bylund and Jarvis is also motivated by the work of von Stutterheim and colleagues (e.g., von Stutterheim, Reference von Stutterheim, Foster-Cohen and Pekarek Doehler2003; von Stutterheim & Nüse, Reference von Stutterheim and Nüse2003), but adopts Langacker's (2008) theory of Cognitive Grammar as its theoretical foundation. Finally, the two remaining articles in the present collection – i.e., the ones by Pavlenko and Malt, and by Athanasopoulos, Damjanovic, Krajciova and Sasaki – are motivated mainly on the basis of previous empirical findings, but they, too, are situated within the same overall cognitive framework: The article by Pavlenko and Malt is motivated by previous research in cognitive science on object naming (e.g., Kronenfeld, Armstrong & Wilmoth, Reference Kronenfeld, Armstrong, Wilmoth and Dougherty1985; Malt, Sloman & Gennari, Reference Malt, Sloman and Gennari2003), and the article by Athanasopoulos and colleagues uses linguistic relativity (e.g., Lucy, Reference Lucy1992; Roberson, Davies & Davidoff, Reference Roberson, Davies and Davidoff2000) as its point of departure.

Before turning to the third meaning of conceptual transfer, I will attempt to clarify the relationship between conceptual transfer, on the one hand, and thinking-for-speaking, linguistic relativity and related frameworks, on the other. To be sure, there is a great deal of overlap between conceptual transfer and thinking-for-speaking, but not a complete overlap. Thinking-for-speaking was originally proposed and has been investigated largely in relation to contrastive linguistics and child L1 acquisition (see, e.g., Berman & Slobin, Reference Berman and Slobin1994) – i.e., in contexts where bilingualism and second language acquisition were not under examination. Of course, since the publication of Slobin's (1993) influential paper describing its relevance to second language acquisition, thinking-for-speaking has also been successfully applied to the investigation of cross-linguistic influence. The consequence of these facts is that thinking-for-speaking has a broader scope than conceptual transfer in the sense that it does not deal solely with transfer. However, from another vantage point, thinking-for-speaking is narrower than conceptual transfer because thinking-for-speaking is based on the assumption that language specificity occurs only during speech-planning processes, whereas the scope of conceptual transfer is not restricted by such an assumption. The relevant speech-planning processes are referred to by Levelt (Reference Levelt1989) and von Stutterheim and Nüse (Reference von Stutterheim and Nüse2003) as macro-planning and micro-planning, where macro-planning involves decisions about what information to convey about a situation or event and at what level of detail, whereas micro-planning involves structuring and linearly ordering the selected non-linguistic meanings in a way that is compatible with the structures and categories available in the speaker's language (see Slobin, Reference Slobin and Perdue1993, Reference Slobin, Gumperz and Levinson1996). Although Levelt (Reference Levelt1989) holds that language specificity occurs only at the level of micro-planning, thinking-for-speaking allows for language specificity at both levels of speech planning, and this is also the position of von Stutterheim and Nüse (Reference von Stutterheim and Nüse2003). Conceptual transfer, too, allows for language specificity at both levels, but goes even further than this, in two ways. First, it deals not just with micro-planning and macro-planning – which are essentially processes of selecting and structuring information in the mind – but it also deals with possible differences in how information is stored in the mind to begin with (i.e., what information is even available for selection). Second, whereas Levelt (Reference Levelt1989), Slobin (Reference Slobin and Perdue1993, Reference Slobin, Gumperz and Levinson1996) and von Stutterheim and Nüse (Reference von Stutterheim and Nüse2003) all explicitly assume that language specificity involving conceptual meaning takes place only while the speaker is preparing to communicate his or her intentions, the conceptual transfer framework allows for the possibility that language specificity might also be found in non-linguistic behavior (e.g., perception, patterns of attention, judgments of similarity, event recall).

Language specificity in non-linguistic behavior is, of course, the domain of linguistic relativity, which is also part of the broader domain of cognitive linguistics (e.g., Pederson, Reference Pederson, Geeraerts and Cuyckens2007). Some of the earlier work on conceptual transfer made strong connections between conceptual transfer and linguistic relativity (e.g., Jarvis, Reference Jarvis1998; Odlin, Reference Odlin2005; Pavlenko, Reference Pavlenko1997), but more recent work has emphasized that the overlap between the two frameworks is smaller than originally assumed (e.g., Jarvis, Reference Jarvis2007; Jarvis & Pavlenko, Reference Jarvis and Pavlenko2008). For one thing, studies on linguistic relativity are interested primarily in non-linguistic behavior, but studies on conceptual transfer concentrate mainly on linguistic behavior (both receptive and productive) in order to determine how conceptual influences associated with the L1 or any previously acquired language might affect the acquisition and use of another language. In other words, linguistic relativity focuses more on the effects of language on cognition, whereas conceptual transfer focuses more on the effects of cognition on language use – particularly the effects of patterns of cognition acquired through one language on the receptive or productive use of another language. Another important difference is the fact that conceptual transfer includes within its scope the processes of conceptual macro-planning and micro-planning, as mentioned earlier, which are not generally of interest to relativists. One final difference is that the scope of conceptual transfer is not restricted to cases where language-specific cognitive patterns are necessarily traceable to structures in the source language (i.e., the L1 or any previously acquired language) per se; rather, like pragmatic transfer, conceptual transfer also deals with differences resulting more generally from a person's long-term integrated membership in a particular speech (or discourse) community (see, e.g., Pavlenko, Reference Pavlenko2003).

So far, I have described conceptual transfer as an observation and as an approach. The third and final way it can be understood is as a hypothesis. More specifically, it refers to the hypothesis that certain instances of cross-linguistic influence in a person's use of one language originate from the mental concepts and patterns of conceptualization that the person has acquired as a speaker of another language. This is, of course, a rather general hypothesis that requires further operationalization and definition of terms in order for it to be testable. This needs to be done within the context of individual studies, so I will not attempt to do so here. However, I will explain how the terms “concept” and “conceptualization” are to be understood throughout this collection of articles, after first emphasizing that work on conceptual transfer is usually not directed at settling the question of whether speakers of different languages have different concepts or conceptualizations, but rather at deriving testable hypotheses from existing theoretical and empirical work in cognitive linguistics concerning such differences, and testing whether the language use (reception or production, or both) of language learners, bilinguals or multilinguals is consistent with those hypotheses.

The term concept should be understood throughout this special issue as referring to a mental representation of an object, quality, action, event, relationship, situation, sensation or any other perceivable or imaginable phenomenon for which the mind creates a mental category. As described by Barsalou (Reference Barsalou2003), Lakoff (Reference Lakoff1987), Murphy (Reference Murphy2002) and others, concepts are generally made up of multiple images, impressions or image schemas that are acquired through various senses (e.g., visual, aural, olfactory, tactile), and which are organized according to how prototypical, or representative, they are of a particular concept. Concepts also often include a knowledge component that allows a person to judge whether something is a member of a particular category (or concept) on the basis of how it functions, even in the absence of any formal similarity to existing members of the category. Concepts are furthermore amenable to change as a person's life experiences, types and levels of expertise, and social and communicative practices change over time. The fact that concepts do not have fixed category boundaries also means that the way a person categorizes something can differ in accordance with differing task demands. However, despite the dynamic nature of concepts and the differences that can exist across individuals even within the same speech community regarding the particular make-up of their conceptual inventories, individuals’ active participation in a speech community results in a high level of consistency at least in the concepts that have been acquired during the process of language socialization. Elsewhere, Aneta Pavlenko and I (Jarvis & Pavlenko, Reference Jarvis and Pavlenko2008, pp. 112–173) have referred to such concepts as language-mediated, and these include both lexicalized concepts (e.g., Murphy, Reference Murphy2002) – such as the concepts referenced by words such as bird and take – as well as grammaticized concepts – such as those associated with grammatical aspect, tense and number (cf. Slobin, Reference Slobin, Bowerman and Levinson2001). The latter tend to involve relational templates, or schemas, with empty slots that can be filled with lexicalized concepts (e.g., Langacker, Reference Langacker2008).

The term conceptualization in this collection of articles is used in a way that is compatible with its use by Langacker (Reference Langacker2008), Levelt (Reference Levelt1989), Slobin (Reference Slobin, Gentner and Goldin-Meadow2003), von Stutterheim and Nüse (Reference von Stutterheim and Nüse2003) and other scholars working in the cognitive linguistics framework. On one level, it refers to the macro-planning and micro-planning processes referred to earlier, where a speaker prepares a preverbal intention for verbalization. On a more general level, it refers to the process of constructing a temporary, complex and dynamic mental representation of a situation, event or state of affairs. Whereas concepts refer to representations that are stored in the mind (presumably in long-term memory), conceptualization is the process of selecting individual concepts and organizing them together (presumably in working memory) in a way that reflects a particular perspective of a specific situation. Elsewhere (Jarvis, Reference Jarvis2007), I have made a distinction between two types of conceptual transfer – concept transfer and conceptualization transfer – where the former refers to cross-linguistic effects arising from differences in the structure or internal make-up of the concepts stored in the minds of speakers of different languages, and where the latter refers to cross-linguistic effects arising not from different concepts, but from differences concerning which concepts are selected and how they are organized in a particular conceptualization of a situation or event. Evidence for the former might be found in language-specific differences in bilinguals’ performance on tasks involving similarity judgments (see the article in this issue by Athanasopoulos, Damjanovic, Krajciova and Sasaki), typicality judgments (e.g., Malt & Sloman, Reference Malt and Sloman2003), object categorization (e.g., Athanasopoulos & Kasai, Reference Athanasopoulos and Kasai2008) and object naming (see the article by Pavlenko and Malt in this issue). Evidence for the latter might be found in language-specific differences in the elements of meaning that bilinguals use to describe situations and events, and in the ways that they combine those elements of meaning (see the articles in this issue by Bylund and Jarvis; Flecken; Brown and Gullberg; Daller, Treffers-Daller and Furman).

It is, of course, impossible to demonstrate conclusively whether cross-linguistic effects have indeed arisen from the conceptual level and, for this reason, Pavlenko and Malt are careful to say only that their results show cross-linguistic effects in naming patterns, without implying that these effects necessarily arise from conceptual differences. The other articles in this issue do refer explicitly to concepts and conceptualization, but as I said earlier, their purpose is not to settle the question of whether cross-linguistic effects arise from language-specific concepts and language-specific patterns of conceptualization, but rather to examine whether the predicted linguistic consequences of hypothesized differences of these types can in fact be found. The study by Flecken goes one step further in using eye-tracking data to show that the linguistic patterns produced by her participants do in fact correspond with their patterns of attention.

One final note before turning to a brief description of the contents and organization of the present special issue is that, even though the distinction between concept transfer and conceptualization transfer is of theoretical value, it is not always possible in practice to distinguish between the two. This is especially true in the case of patterns of event construal – a term that is defined at length by Langacker (Reference Langacker2008, pp. 55–89), and which is used frequently in the literature (including in each of the last four articles in this issue) in a way that could be described as referring to the linguistic manifestation of how an event has been conceptualized by an individual. In the case of event construals, cross-linguistic effects could theoretically arise either from differences in the concepts (including grammaticized concepts) that are stored in the minds of bilinguals representing different language backgrounds, or from differences simply in which concepts they happen to select (and how they organize them vis-à-vis one another) when describing an event. When it is impossible to distinguish between the two, the more conservative interpretation is the latter one, as demonstrated in the article by Daller, Treffers-Daller and Furman.

The organization of articles in this issue is as follows. The first two studies deal with categorization processes in the domain of lexicalized concepts, or mental categories that correspond with lexical items and which reflect word meanings. The first is the article by Panos Athanasopoulos, Ljubica Damjanovic, Andrea Krajciova and Miho Sasaki, which is based on Athanasopoulos's earlier work (2009) on the perception of color categories by Greek–English bilinguals. The researchers’ follow-up study in this issue investigates the perception of color categories by Japanese–English bilinguals. Japanese has two color terms that correspond with what English speakers call blue, and the purpose of this study is to determine how bilinguals who are proficient in both languages perceive the relevant hues. The results show that the bilinguals who use Japanese more often are more sensitive to the distinction between the two blues than are the bilinguals who use English more often. This finding seems to suggest that the internal make-up – or perhaps the accessibility – of a perceptual category is determined largely by whether a person is immersed in a discourse community where that particular perceptual category is frequently expressed.

The second study is the article by Aneta Pavlenko and Barbara Malt, which is an investigation of how Russian–English bilinguals name drinking vessels such as cups, glasses and mugs. The bilinguals in the study were shown pictures of a number of drinking vessels and asked to name them in their native language, Russian. The results show that the bilinguals’ naming patterns depend largely on how old they were when they moved to the US. The influence of English on their Russian naming patterns is strongest among the early bilinguals, but even the late bilinguals exhibit some influence from English. Together, this study and the one by Athanasopoulos and colleagues suggest that experience with a second language (or perhaps any language beyond the L1) can result in altered patterns of mental categorization in a person's native language. The study by Pavlenko and Malt also shows, however, that bilinguals do not resort simply to just L1 or just L2 categorization patterns, but instead seem to exhibit the effects of the attraction of both languages pulling in separate directions at the same time, a process that can be described in relation to the notions of conceptual restructuring, convergence or shift, depending on how the combined effects of the conceptual systems underlying the two languages play out in the language knowledge and use of any given bilingual (see Jarvis & Pavlenko, Reference Jarvis and Pavlenko2008, pp. 153–173).

The remaining four studies in this issue deal with event construal. The first study on event construal is the article by Emanuel Bylund and myself, which follows up on claims by von Stutterheim (Reference von Stutterheim, Foster-Cohen and Pekarek Doehler2003) concerning the relationship between grammatical aspect and patterns of event construal. The participants were Spanish–Swedish bilinguals who were given two tests in their native language: a grammaticality judgment test and a video-description task. Our main finding is that there is indeed a significant correlation between the bilinguals’ sensitivity to aspectual distinctions (on the grammaticality judgment test) and their reference to event endpoints (in the video-description task). We interpret these results as showing the effects of conceptual entrenchment, where the Spanish speakers who are most affected by their immersion in Swedish discourse show the strongest Swedish-like tendency to rely on a bounded-event construal in both receptive and productive tasks.

The next study on event construal is the article by Monique Flecken, whose purpose is to further examine the relationship between grammatical aspect and event construal, this time with respect to bilinguals’ choice of aspect rather than their reference to endpoints. In order to establish that aspectual choices reflect conceptualization patterns, Flecken also uses eye-tracking evidence to show what the bilinguals attend to as they describe events. The participants in the study include Dutch–German bilinguals, as well as speakers of both Dutch and German who are not bilingual in both languages. Neither language has historically had a grammaticalized progressive aspect, but this is currently becoming grammaticalized in Dutch, suggesting that Dutch speakers are likely to refer to the ongoingness of actions more than German speakers do. The results of the participants’ performance on a video-description task confirm this expectation, but, unexpectedly, the highest level of reference to ongoingness is found in the bilingual group. The main finding of the study, though, is that the Dutch–German bilinguals not only express ongoingness more, but they also fixate more on the action of events than the other two groups do. Here, again, we see an important connection between the language structures people produce and the mental conceptualizations they appear to be constructing. The degree to which this outcome is due to cross-linguistic influence is difficult to determine, but Flecken explains that incompatibilities between Dutch and German may be partly to blame for the bilinguals’ elevated references to ongoingness, which appear to be the result of weakened semantic restrictions concerning where ongoingness is not conventionally expressed in Dutch.

The third study on event construal is the study by Amanda Brown and Marianne Gullberg, who examine Japanese–English speakers’ expression of Path in motion events, where Path is encoded differently into the lexical semantics of the two languages. Participants include Japanese–English speakers, as well as relatively monolingual speakers of each language. The data are oral animated cartoon retells, which the bilingual participants produced in both languages. The results show that the bilinguals produce more Goal expressions than either of the monolingual groups. On the surface, this does not appear to be the result of cross-linguistic influence; however, like Flecken, the authors of this study interpret this finding as possibly reflecting an indirect consequence of cross-linguistic influence, whereby the juxtaposition of the specific languages the bilinguals have learned has redirected their attention and resulted in a type of restructuring that is specific to that particular pairing of languages. The other main results of this study show more direct effects of cross-linguistic influence: Whereas the Japanese monolinguals tend to use more Path verbs and fewer Path adverbs, the English monolinguals show the opposite pattern. The bilinguals’ patterns tend to fall between both monolingual groups, particularly in the use of adverbs, with more Path adverbs than the Japanese monolinguals but fewer than the English monolinguals. This is true of bilinguals’ performance in both languages, but one very important finding is that the bilinguals are somewhat more Japanese-like when they perform the task in Japanese, and somewhat more English-like when they perform the task in English. Overall, the researchers interpret these findings as suggesting that the acquisition of a second language results in a certain degree of restructuring regarding processes of preverbal planning in both the L1 and L2, the outcome of which can be seen as a trend toward convergence in the preverbal-planning processes associated with both languages in the minds of individual bilinguals.

The final study on event construal is the article by Michael Daller, Jeanine Treffers-Daller and Reyhan Furman. Like the study by Brown and Gullberg, this study also deals with the construal of motion events in relation to Path, this time in picture stories told by Turkish–German bilinguals, as well as by monolingual German and Turkish speakers. Similar to the results of Brown and Gullberg, the results of this study show that the bilinguals, in both languages, use various available Path options with frequencies that are between those of monolingual Turkish and monolingual German speakers. One important difference, though, is that in the Brown and Gullberg study, there is no difference between the bilinguals living in the L1 versus the L2 environment, whereas in this study, the bilinguals who had returned to Turkey are more similar to the Turkish monolinguals, whereas the bilinguals still living in Germany are more similar to the German monolinguals. It is important to note, though, that the two groups of bilinguals in this study are more different from each other than the two groups of bilinguals in the Brown and Gullberg study in terms of how long they had lived in either environment and how proficient they are in one language versus the other. Accordingly, Daller, Treffers-Daller and Furman interpret their findings not only in relation to cross-linguistic influence, but also in relation to language dominance. The authors also interpret their results as suggesting that the patterns of conceptualization and expression that bilinguals routinely are exposed to and rely on in the dominant language environment become so cognitively entrenched that these same conceptualization patterns are activated when they plan utterances in the non-dominant language.

These six studies produce important new insights into cross-linguistic influence and illustrate well what some of the main issues are in the investigation of conceptual transfer. They show that speakers of different languages can often be found to differ in patterns of language use that are related to processes of categorization and conceptualization, and they likewise show that such patterns acquired through one language often carry over into the use of another language. These studies have also alluded to some of the ways in which conceptual transfer tends to differ from linguistic transfer. First, even though bidirectional effects have been found in various types of linguistic transfer (see, e.g., Pavlenko & Jarvis, Reference Pavlenko and Jarvis2002), these bidirectional effects are generally complementary rather than convergent. The studies in the present collection show that conceptual transfer, on the other hand, may be characteristically bidirectional and convergent in a way that appears to reflect bilinguals’ attempt to accommodate both languages within a single system. This makes sense in light of the research on aphasia and agnosia, for example, which suggests that even though bilinguals do appear to create separate linguistic systems for their two languages, they have to rely on a single conceptual store (e.g., Paradis, Reference Paradis2004). Another theme of these studies is the importance of entrenchment, a notion espoused by Langacker (Reference Langacker2008), for example, in describing how linguistic “units emerge via the progressive entrenchment of configurations that recur in a sufficient number of events to be established as cognitive routines” (p. 220). Indeed, a major part of what conceptual transfer entails is probably best understood in terms of cognitive routines related to categorization and construal that have become routinized through a person's active participation in a discourse community where certain notions tend to be expressed frequently and saliently in particular ways.

Major questions remain concerning how concepts are stored in long-term memory and how conceptualization takes place in working memory. There are separate lines of research being conducted on conceptualization and memory, and at some point these two perspectives need to be brought together firmly, as both have important implications for cross-linguistic influence. Important bridgework also needs to take place between conceptual transfer and what von Stutterheim and her colleagues refer to as principles of information organization. What are those principles? Also, how, where and in what form are they stored in memory and how are they invoked? Among other things, how do they relate to mental schemas that play a central role in related theoretical frameworks? One of the biggest questions for conceptual transfer, which was brought up by Odlin (Reference Odlin, Doughty and Long2003), is how deep it runs. That is, is it a relatively shallow phenomenon occurring only in processes of micro-planning, as Levelt (Reference Levelt1989) has suggested, or does it run even deeper, perhaps even in areas of general cognition that have been of interest to the relativists? Finally, we need to ask whether conceptual transfer can be avoided, or at least overcome at some point, during the process of acquiring another language. Is it possible, for example, to develop separate conceptual systems for our two languages, or is some sort of conceptual convergence inevitable? Likewise, which of our concepts and conceptual patterns might represent conceptual universals, and are therefore not relevant to cross-linguistic influence. These are some of the questions that need to be asked and need to be answered. And, fortunately, this is one area of research where the journey is as enjoyable and fascinating as I imagine the eventual destination will be.