Pavlenko's keynote paper calls for a rethinking of models of the mental lexicon in the light of recent research into emotion and bilingualism. The author makes a convincing case for the inclusion of affective aspects in the study of the mental lexicon. Indeed, the knowledge of the degree of emotionality of a word and of its affective valence is just as important as the knowledge of that word's grammatical class, or its gender. From a pragmatic point of view, one could argue that an L2 user's inaccurate or incomplete understanding of the emotionality and valence of an emotion word, or an emotion-laden word, in the L2 might lead to unwanted illocutionary effects, which might be far more embarrassing than phonological, morphological or syntactical errors.
Pavlenko's view of emotion concepts as prototypical scripts which are themselves embedded within larger systems of beliefs provides researchers with a useful frame of reference for further research in crosslinguistic variation and individual variation. By systematically identifying the loci of difference between emotion concepts in different languages, an accurate picture of emotion concepts can be drawn. This is turn allows a careful comparison of potential overlap in concepts in different languages. Pavlenko's argument that models of bilingual representation and processing need to acknowledge conceptual nonequivalence is very convincing. Her typology of conceptual processes in the bilingual lexicon presents researchers with a perfect framework to start researching conceptual change both at the level of groups of multilinguals with similar language combinations but also at the level of individual multilinguals.
Although Pavlenko mentions L2 learners, she focuses more on L2 users, i.e. those who would traditionally be referred to as bilinguals who typically live in a country where the bilinguals' L2 is the region's majority language.
Pavlenko's paper could also offer a new impetus to research in Second Language Acquisition (SLA) research. The most obvious question is whether classroom instruction could lead to the acquisition of new concepts in the L2. Pavlenko's own work (Reference Pavlenko2003) suggests that this is not necessarily easy. Her highly fluent Russian learners of English at the University of St Petersburg had acquired semantic but not conceptual representations of “privacy” and “personal space”, a concept that does not exist in Russian. Although the learners knew the word in English, they did not identify a situation of an “invasion of privacy” shown to them on film. In contrast, a group of young Russian immigrants enrolled in US universities had acquired the concept of “privacy”, and they were better able to recognize it in a film fragment and flag it up in the retelling. Monolingual Anglo-Americans recognized the concept of privacy straight away and it was more salient in their film retellings. The St Petersburg students had probably never specifically focused on the concepts in English that had no equivalent in Russian. Would such an explicit focus on conceptual nonequivalence in the L2 classroom lead to a faster and richer acquisition of conceptual representations? Empirical classroom research would be needed to answer this question.
A number of questions come to mind for SLA research on the development of concepts. The first issue to be addressed would have to be a clear description of the dependent variable: i.e. how can researchers identify and measure conceptual representation? Non-emotional object concepts are probably more easy to gauge: i.e. containers (Malt & Sloman, Reference Malt and Sloman2003), common household objects (Ameel, Storms, Malt and Sloman, Reference Ameel, Storms, Malt and Sloman2005). These studies have relatively simple designs that allow relatively straightforward quantification and statistical analyses: an object can belong to a limited number of categories, and if certain subgroups of participants systematically assign a object to a particular category, conclusions can be made as to the conceptual representations underlying participants' choices.
Emotion concepts consisting of prototypical scripts are typically more diffuse, more fuzzy, and therefore harder to pin down. It is also relatively unlikely that “pure” examples can be recorded in authentic interactions. People do not like to be too obvious, and may have secret communicative intentions. If a speaker's secret intention is to make someone fall in love with him/her, s/he will try to create a favorable impression with the interlocutor through a variety of verbal and non-verbal means. In other words, there is no immediately identifiable emotion script. The interlocutor might recognize bits of different “friendship” scripts including an idiosyncratic blend of apologies, jokes, compliments, gazes, touches, but remain unaware of the feelings of the speaker and of the true nature of the larger script. Indeed, the speaker may well never refer to the emotion concept of love, and the word itself would not be uttered. This is obviously the territory of novelists, and Kazuo Ishiguro's novel The remains of the day comes to mind, with the soliloquies of Mr Stevens, the butler, to Miss Kenton, the former housekeeper. The reader guesses that Mr Stevens is in love with Miss Kenton but he cannot admit it to himself and thus remains sufficiently vague in order not to betray his feelings.
Linguists have come up with ingenious designs to investigate conceptualization of emotions. Stepanova Sachs and Coley (Reference Stepanova Sachs, Coley and Pavlenko2006) presented a series of situations to their monolingual and bilingual participants, asking them to decide which emotions were described. This systematic approach allows the identification of prototypical scripts and weed out individual differences.
A final obstacle in the description of emotion scripts are intra-individual differences. A person may have certain usual scripts to express certain emotions but may suddenly deviate. For example, employees dealing with customer complaints follow well-established scripts to deal with rude customers, yet I can easily imagine that with especially abusive customers they may suddenly snap and starting shouting back. Non-emotion concepts are less likely to elicit that much variation: the categorization of everyday objects is a relative uncontroversial operation for native speakers.
The second methodological issue has to do with the difficulty of “capturing” an entire emotion script. In other words, where does a script start and when does it end? Can an entire script ever be collected? A script can vary in length from a minimal unit to a whole stretch of discourse. This might be uncontroversial in some cases but in Mr Stevens' case, one could argue that the emotion script leading up to the declaration of love lasts decades, and that in the end no declaration was ever made. Some emotion scripts can also be entirely non-verbal.
The third thorny issue facing SLA researchers in the study of emotion concepts is more of an epistemological nature: how to resist the automatic interpretation of an absence of something in the speech in L2 users as a deficit? The “deficit” hypothesis is obviously often true. The absence of references to “privacy” can be seen as an indication of a lack of conceptual representation among Pavlenko's (Reference Pavlenko2003) Russian learners of English. The relative lack of emotion words in the conversations in French L2 of less proficient learners (i.e. those with lower morpholexical accuracy rates) in Dewaele and Pavlenko (Reference Dewaele and Pavlenko2002) was interpreted as a conscious or unconscious avoidance of emotion talk by less proficient learners because of perceived or real incompleteness of the emotion script and a wish to avoid potential embarrassment to themselves and their interlocutor. However, there is no clear procedure to establish whether an absence (or a low frequency) of some form is the result of avoidance rather than incompleteness. I argued in Dewaele (Reference Dewaele, Hua, Seedhouse, Wei and Cook2007a) that L2 users' deviations from the norm are not necessarily examples of pragmatic failure. L2 users may intentionally avoid certain scripts, just as L1 users do. A L2 user may, for some reason, prefer to follow the L1 emotion script in the L2. It would be wrong however, to conclude that this individual lacked the complete emotion script in the L2. Corpus-based studies are crucial in this respect, as they allow to establish that the absence of a concept is significant (i.e. Pavlenko, Reference Pavlenko2003 and Pavlenko & Driagina, Reference Pavlenko and Driagina2007). A systematic absence indicates a lack of an important interpretive resource. There is a critical difference however between incompleteness and avoidance. Avoidance, as I said, is a choice made by individuals who have the necessary resources at their disposal. Incompleteness, on the other hand, indicates the lack of resources and requires an instructional intervention. This is why we test groups and look at group patterns. Clearly, not all English monolinguals referred to ‘privacy’ and neither did all of the bilinguals – what we see then is individual variation and avoidance. But the fact that none of the L2 learners did so indicates that this is not necessarily a conscious choice (avoidance) but a lack.
I will now formulate a few thoughts about future SLA research in conceptual change. Firstly, longitudinal research on conceptual change in the L2 classroom would be extremely interesting as it could provide useful information on the rate of progress. We know very little about the time-course of conceptual change: is it a trickle growing into a river? Is it a series of small bursts, i.e. growth followed by apparent slow-down or even regression? Is it like the bursting of a dam? The first scenario would probably be the one foreign language teachers would prefer: their teaching producing some constant, measurable effect, and the students seeming to acquire new concepts in the L2 in a linear fashion. The second scenario would reflect typical patterns found in research on sociolinguistic and sociopragmatic competence in the L2, namely U-shaped patterns (Dewaele, Reference Dewaele and Ayoun2007b). The third scenario would be the most worrying one for the language teacher, and the one hardest to interpret for the researcher. Metaphorically one could say that teacher and researcher observe the flow, sitting downstream from the dam, unable to see whether pressure is building up behind the dam. In other words, the pessimist could argue that the teaching of new concepts may be wasted on a learner because there is no apparent change. The optimist may argue that pressure is building up behind the dam, and that at some point there will be radical change.
It is also possible that none of the previous scenarios is “true” in the sense that different people may develop conceptual change in different ways and at different rates.
The second issue for SLA research is that of individual differences both in the concepts being taught and acquired, and in the learners acquiring them.
It is very likely that emotion concepts, and the prototypical scripts they represent, may provoke very different reactions among language learners, and that as a consequence, learners may decide, either consciously or unconsciously, to adopt the script in their repertoire, to adapt it or to reject it outright. The rejection may be linked to a perceived incompatibility with what the L2 learners or L2 users consider appropriate behaviour, based on their L1 experience. The L2 emotion concept may be recognized, but not being appropriated by the L2 user. Besemeres (Reference Besemeres2004) offers an excellent illustration of this in her analysis of the British author and Italian L2 user, Tim Parks, describing the Italian concept of “fare festa a qualcuno” which means “to make a party for someone”, and “combines the ideas of welcoming them and smothering them with physical affection” Besemeres, Reference Besemeres2004, p. 45, citing Parks Reference Parks1996). Parks presents the concept in his humorous rendering of the meetings between his children and their nonni. Besemeres points out that he clearly retains some of that British disapproval in the face of “fare festa”, suspecting his inlaws (and even his wife) of insincerity in their displays of feeling (Besemeres, Reference Besemeres2004, p. 45). This specific Italian emotion concept of “fare festa” offers a good illustration of the difficulties that arise from a foreign language teaching point of view. Italian children will have learned the script through socialization, namely by having experienced it. They would probably have stopped enjoying it as teenagers, and still be too young themselves to use the script with small children. Teenagers learning Italian as an L2 might be able to learn to recognize the concept, but might be unwilling to add this particular script to their L2 repertoire given their typical lack of interest in younger children. Adolescents would undoubtedly be more interested in emotion scripts involving the establishment of contacts of an amorous nature. In other words, many emotion scripts are generation-specific, which raises the issue of appropriateness of inclusion in the foreign language curriculum and of teachability. Tim Parks obviously learnt to recognize the concept of “fare festa” through secondary socialisation after initial instructed SLA.
Summing up, Pavlenko's contribution will undoubtedly offer a fresh impetus to SLA research. I am confident that the methodological and epistemological problems can be overcome through a combination of ingenuity, flair and some inevitable compromises.
You have
Access
Pavlenko's keynote paper calls for a rethinking of models of the mental lexicon in the light of recent research into emotion and bilingualism. The author makes a convincing case for the inclusion of affective aspects in the study of the mental lexicon. Indeed, the knowledge of the degree of emotionality of a word and of its affective valence is just as important as the knowledge of that word's grammatical class, or its gender. From a pragmatic point of view, one could argue that an L2 user's inaccurate or incomplete understanding of the emotionality and valence of an emotion word, or an emotion-laden word, in the L2 might lead to unwanted illocutionary effects, which might be far more embarrassing than phonological, morphological or syntactical errors.
Pavlenko's view of emotion concepts as prototypical scripts which are themselves embedded within larger systems of beliefs provides researchers with a useful frame of reference for further research in crosslinguistic variation and individual variation. By systematically identifying the loci of difference between emotion concepts in different languages, an accurate picture of emotion concepts can be drawn. This is turn allows a careful comparison of potential overlap in concepts in different languages. Pavlenko's argument that models of bilingual representation and processing need to acknowledge conceptual nonequivalence is very convincing. Her typology of conceptual processes in the bilingual lexicon presents researchers with a perfect framework to start researching conceptual change both at the level of groups of multilinguals with similar language combinations but also at the level of individual multilinguals.
Although Pavlenko mentions L2 learners, she focuses more on L2 users, i.e. those who would traditionally be referred to as bilinguals who typically live in a country where the bilinguals' L2 is the region's majority language.
Pavlenko's paper could also offer a new impetus to research in Second Language Acquisition (SLA) research. The most obvious question is whether classroom instruction could lead to the acquisition of new concepts in the L2. Pavlenko's own work (Reference Pavlenko2003) suggests that this is not necessarily easy. Her highly fluent Russian learners of English at the University of St Petersburg had acquired semantic but not conceptual representations of “privacy” and “personal space”, a concept that does not exist in Russian. Although the learners knew the word in English, they did not identify a situation of an “invasion of privacy” shown to them on film. In contrast, a group of young Russian immigrants enrolled in US universities had acquired the concept of “privacy”, and they were better able to recognize it in a film fragment and flag it up in the retelling. Monolingual Anglo-Americans recognized the concept of privacy straight away and it was more salient in their film retellings. The St Petersburg students had probably never specifically focused on the concepts in English that had no equivalent in Russian. Would such an explicit focus on conceptual nonequivalence in the L2 classroom lead to a faster and richer acquisition of conceptual representations? Empirical classroom research would be needed to answer this question.
A number of questions come to mind for SLA research on the development of concepts. The first issue to be addressed would have to be a clear description of the dependent variable: i.e. how can researchers identify and measure conceptual representation? Non-emotional object concepts are probably more easy to gauge: i.e. containers (Malt & Sloman, Reference Malt and Sloman2003), common household objects (Ameel, Storms, Malt and Sloman, Reference Ameel, Storms, Malt and Sloman2005). These studies have relatively simple designs that allow relatively straightforward quantification and statistical analyses: an object can belong to a limited number of categories, and if certain subgroups of participants systematically assign a object to a particular category, conclusions can be made as to the conceptual representations underlying participants' choices.
Emotion concepts consisting of prototypical scripts are typically more diffuse, more fuzzy, and therefore harder to pin down. It is also relatively unlikely that “pure” examples can be recorded in authentic interactions. People do not like to be too obvious, and may have secret communicative intentions. If a speaker's secret intention is to make someone fall in love with him/her, s/he will try to create a favorable impression with the interlocutor through a variety of verbal and non-verbal means. In other words, there is no immediately identifiable emotion script. The interlocutor might recognize bits of different “friendship” scripts including an idiosyncratic blend of apologies, jokes, compliments, gazes, touches, but remain unaware of the feelings of the speaker and of the true nature of the larger script. Indeed, the speaker may well never refer to the emotion concept of love, and the word itself would not be uttered. This is obviously the territory of novelists, and Kazuo Ishiguro's novel The remains of the day comes to mind, with the soliloquies of Mr Stevens, the butler, to Miss Kenton, the former housekeeper. The reader guesses that Mr Stevens is in love with Miss Kenton but he cannot admit it to himself and thus remains sufficiently vague in order not to betray his feelings.
Linguists have come up with ingenious designs to investigate conceptualization of emotions. Stepanova Sachs and Coley (Reference Stepanova Sachs, Coley and Pavlenko2006) presented a series of situations to their monolingual and bilingual participants, asking them to decide which emotions were described. This systematic approach allows the identification of prototypical scripts and weed out individual differences.
A final obstacle in the description of emotion scripts are intra-individual differences. A person may have certain usual scripts to express certain emotions but may suddenly deviate. For example, employees dealing with customer complaints follow well-established scripts to deal with rude customers, yet I can easily imagine that with especially abusive customers they may suddenly snap and starting shouting back. Non-emotion concepts are less likely to elicit that much variation: the categorization of everyday objects is a relative uncontroversial operation for native speakers.
The second methodological issue has to do with the difficulty of “capturing” an entire emotion script. In other words, where does a script start and when does it end? Can an entire script ever be collected? A script can vary in length from a minimal unit to a whole stretch of discourse. This might be uncontroversial in some cases but in Mr Stevens' case, one could argue that the emotion script leading up to the declaration of love lasts decades, and that in the end no declaration was ever made. Some emotion scripts can also be entirely non-verbal.
The third thorny issue facing SLA researchers in the study of emotion concepts is more of an epistemological nature: how to resist the automatic interpretation of an absence of something in the speech in L2 users as a deficit? The “deficit” hypothesis is obviously often true. The absence of references to “privacy” can be seen as an indication of a lack of conceptual representation among Pavlenko's (Reference Pavlenko2003) Russian learners of English. The relative lack of emotion words in the conversations in French L2 of less proficient learners (i.e. those with lower morpholexical accuracy rates) in Dewaele and Pavlenko (Reference Dewaele and Pavlenko2002) was interpreted as a conscious or unconscious avoidance of emotion talk by less proficient learners because of perceived or real incompleteness of the emotion script and a wish to avoid potential embarrassment to themselves and their interlocutor. However, there is no clear procedure to establish whether an absence (or a low frequency) of some form is the result of avoidance rather than incompleteness. I argued in Dewaele (Reference Dewaele, Hua, Seedhouse, Wei and Cook2007a) that L2 users' deviations from the norm are not necessarily examples of pragmatic failure. L2 users may intentionally avoid certain scripts, just as L1 users do. A L2 user may, for some reason, prefer to follow the L1 emotion script in the L2. It would be wrong however, to conclude that this individual lacked the complete emotion script in the L2. Corpus-based studies are crucial in this respect, as they allow to establish that the absence of a concept is significant (i.e. Pavlenko, Reference Pavlenko2003 and Pavlenko & Driagina, Reference Pavlenko and Driagina2007). A systematic absence indicates a lack of an important interpretive resource. There is a critical difference however between incompleteness and avoidance. Avoidance, as I said, is a choice made by individuals who have the necessary resources at their disposal. Incompleteness, on the other hand, indicates the lack of resources and requires an instructional intervention. This is why we test groups and look at group patterns. Clearly, not all English monolinguals referred to ‘privacy’ and neither did all of the bilinguals – what we see then is individual variation and avoidance. But the fact that none of the L2 learners did so indicates that this is not necessarily a conscious choice (avoidance) but a lack.
I will now formulate a few thoughts about future SLA research in conceptual change. Firstly, longitudinal research on conceptual change in the L2 classroom would be extremely interesting as it could provide useful information on the rate of progress. We know very little about the time-course of conceptual change: is it a trickle growing into a river? Is it a series of small bursts, i.e. growth followed by apparent slow-down or even regression? Is it like the bursting of a dam? The first scenario would probably be the one foreign language teachers would prefer: their teaching producing some constant, measurable effect, and the students seeming to acquire new concepts in the L2 in a linear fashion. The second scenario would reflect typical patterns found in research on sociolinguistic and sociopragmatic competence in the L2, namely U-shaped patterns (Dewaele, Reference Dewaele and Ayoun2007b). The third scenario would be the most worrying one for the language teacher, and the one hardest to interpret for the researcher. Metaphorically one could say that teacher and researcher observe the flow, sitting downstream from the dam, unable to see whether pressure is building up behind the dam. In other words, the pessimist could argue that the teaching of new concepts may be wasted on a learner because there is no apparent change. The optimist may argue that pressure is building up behind the dam, and that at some point there will be radical change.
It is also possible that none of the previous scenarios is “true” in the sense that different people may develop conceptual change in different ways and at different rates.
The second issue for SLA research is that of individual differences both in the concepts being taught and acquired, and in the learners acquiring them.
It is very likely that emotion concepts, and the prototypical scripts they represent, may provoke very different reactions among language learners, and that as a consequence, learners may decide, either consciously or unconsciously, to adopt the script in their repertoire, to adapt it or to reject it outright. The rejection may be linked to a perceived incompatibility with what the L2 learners or L2 users consider appropriate behaviour, based on their L1 experience. The L2 emotion concept may be recognized, but not being appropriated by the L2 user. Besemeres (Reference Besemeres2004) offers an excellent illustration of this in her analysis of the British author and Italian L2 user, Tim Parks, describing the Italian concept of “fare festa a qualcuno” which means “to make a party for someone”, and “combines the ideas of welcoming them and smothering them with physical affection” Besemeres, Reference Besemeres2004, p. 45, citing Parks Reference Parks1996). Parks presents the concept in his humorous rendering of the meetings between his children and their nonni. Besemeres points out that he clearly retains some of that British disapproval in the face of “fare festa”, suspecting his inlaws (and even his wife) of insincerity in their displays of feeling (Besemeres, Reference Besemeres2004, p. 45). This specific Italian emotion concept of “fare festa” offers a good illustration of the difficulties that arise from a foreign language teaching point of view. Italian children will have learned the script through socialization, namely by having experienced it. They would probably have stopped enjoying it as teenagers, and still be too young themselves to use the script with small children. Teenagers learning Italian as an L2 might be able to learn to recognize the concept, but might be unwilling to add this particular script to their L2 repertoire given their typical lack of interest in younger children. Adolescents would undoubtedly be more interested in emotion scripts involving the establishment of contacts of an amorous nature. In other words, many emotion scripts are generation-specific, which raises the issue of appropriateness of inclusion in the foreign language curriculum and of teachability. Tim Parks obviously learnt to recognize the concept of “fare festa” through secondary socialisation after initial instructed SLA.
Summing up, Pavlenko's contribution will undoubtedly offer a fresh impetus to SLA research. I am confident that the methodological and epistemological problems can be overcome through a combination of ingenuity, flair and some inevitable compromises.