Studies in Second Language Learning and Teaching Department of English Studies, Faculty of Pedagogy and Fine Arts, Adam Mickiewicz University, Kalisz SSLLT 7 (1). 2017. 61-88 doi: 10.14746/ssllt.2017.7.1.4 http://www.ssllt.amu.edu.pl Positive and negative emotions underlie motivation for L2 learning1 Peter D. MacIntyre Cape Breton University, Sydney, Nova Scotia, Canada [email protected] Laszlo Vincze University of Helsinki, Finland [email protected] Abstract The role of basic emotions in SLA has been underestimated in both research and pedagogy. The present article examines 10 positive emotions (joy, grati- tude, serenity, interest, hope, pride, amusement, inspiration, awe, andlove) and 9 negative emotions (anger, contempt, disgust, embarrassment, guilt, hate, sadness, feeling scared, and being stressed). The emotions are corre- lated with core variables chosen from three well-known models of L2 motiva- tion: Gardner’s integrative motive, Clément’s social-contextual model, and Dörnyei’s L2 self system. Respondents came from Italian secondary schools, and most participants were from monolingual Italian speaking homes. They described their motivation and emotion with respect to learning German in a region of Italy (South Tyrol) that features high levels of contact between Ital- ians and Germans. Results show that positive emotions are consistently and strongly correlated with motivation-related variables. Correlations involving neg- ative emotions are weaker and less consistently implicated in motivation. The pos- itivity ratio, that is, the relative prevalence of positive over negative emotion, showed strong correlations with all of the motivation constructs. Regression 1This research was supported in part by a grant to the first author from the Social Science and Humanities Research Council of Canada and grants to the second author from the Ella and Georg Ehrnrooth Foundation, Helsinki and the Finnish Cultural Foundation, Helsinki. 61 Peter D. MacIntyre, Laszlo Vincze analysis supports the conclusion that a variety of emotions, not just one or two key ones, are implicated in L2 motivation processes in this high-contact context. Keywords: integrative motive; L2 self system; social-contextual model; broaden and build theory; positivity ratio 1. Introduction Basic emotions can play a significant role in second language acquisition (SLA) and communication processes, an impact that has been underestimated in both the research and pedagogical literature (Dewaele, 2012; MacIntyre, 2002). In SLA, the literature on individual differences and learner factors has featured a considerable emphasis on cognitive and ability factors such as working memory, strategies, intelligence, aptitude factors, and others, but in SLA they have been discussed as relatively cold cognition, as if emotion played no part in these pro- cesses (Pintrich, Marx, & Boyle, 1993). In contrast, a hot cognition approach ex- amines learning “. . . in which multiple levels of cognition are sparked by moti- vation and emotions in lively sociocultural contexts” (Oxford, 2016, p. 25). The relative lack of emphasis on emotion and the roles it plays in both learning and communication experiences has been something of a historical accident, a hold- over from a time when a serious treatment of emotions was considered irrele- vant by behaviourists, and later when emotions were ignored by cognitively-fo- cussed scholars (Fredrickson, 2013a). However, there is rapidly emerging inter- est in emotion in SLA, partially inspired by applying and expanding on the devel- opments being made in positive psychology (Gabryś-Barker & Gałajda, 2016; MacIntyre, Gregersen, & Mercer, 2016). In SLA, the most notable area in which emotion has been implicated, al- beit indirectly, has been in the literature on motivation. Three of the major L2 motivation research traditions, Gardner’s (1985, 2010) integrative motive, Clement’s (1986) social-contextual model, and to a lesser extent Dörnyei’s (2005) L2 self-system, feature emotions that influence motivational processes. However, even within the SLA motivation literature, the emphasis has not been on basic emotions. Recent research has been addressing the imbalance and working to fit emotions into the theoretical framework of SLA. The present arti- cle will examine the correlations of basic positive and negative emotions, as identified by differential emotions theory (Izard, 2007), with established learner characteristics emphasizing motivation-related variables drawn primarily from the well-established models of Gardner, Clément and Dörnyei. 62 Positive and negative emotions underlie motivation for L2 learning 1.1. Emotion is a primary motive There are good reasons to be concerned with emotion as a core process that im- pacts almost everything we do. In the bookDescartes’ Error, neuroscientist Anto- nio Damasio (1994) built the case that human beings are not thinking machines that feel; rather, we are feeling machines that think. The primacy of emotion emerges in the earliest moments of life as an infant reacts emotionally to every significant stimulus, a process of reacting to events that continues throughout the entire lifespan. All later experiences, including language development, are built on an emotional foundation and are fully integrated with it. At their core, emo- tional reactions are adaptive and drive efforts at coping with life’s changing situa- tions. Active emotions, especially negative ones such as fear or anger, simply take over conscious experience—they cannot easily be ignored (Reeve, 2015). Lan- guages are difficult to learn and the learning process can arouse intense emotions. As teachers, we might ask students to “stop and think” when emotions run high, but have you ever asked a student to “stop and feel something”? 1.2. Defining emotion In psychology, when considered at all, emotion has proven remarkably difficult to define. The definition we will use in the present article was developed by Reeve (2015): Emotions are “. . . short-lived, feeling-purposive-expressive-bodily re- sponses that help us adapt to the opportunities and challenges we face during important life events” (p. 340). Under this definition, emotions have both physical and psychological dimensions, they are reactions to the outside world, they ex- press what is going on inside the body to the outside world, and they exist for a reason—each emotion has a purpose. Reeve’s definition is representative of the field because it allows for emotions to be more than the sum of their parts, re- flecting the complexity of emotional experience. We hasten to note that the ad- aptations provoked by a given emotional experience may be relatively beneficial on one timescale but harmful when considered over a different period of time.2 Although we have chosen Reeve’s (2015) definition to guide our work, the difficulty with defining emotions must be emphasized. A leading scholar in the field, Caroll Izard (2010) surveyed leading emotion researchers but failed to con- verge on a single definition that captures the subject they study. He did, how- ever, generate the following description of emotion. 2 For example, opting for short term comfort can lead to long term problems, as when a panicky student leaves an important examination before finishing the test. It is the interac- tion of emotion with other processes that determines the quality of the adaptation for any particular purpose, considered over a particular timescale. 63 Peter D. MacIntyre, Laszlo Vincze Emotion consists of neural circuits (that are at least partially dedicated), response systems, and a feeling state/process that motivates and organizes cognition and ac- tion. Emotion also provides information to the person experiencing it, and may in- clude antecedent cognitive appraisals and ongoing cognition including an interpreta- tion of its feeling state, expressions or social-communicative signals, and may moti- vate approach or avoidant behavior, exercise control/regulation of responses, and be social or relational in nature. (p. 367) Izard’s (2010) study revealed that theorists working with conceptualizing emo- tion show moderate to high agreement on the structures and functions of emo- tion, and they all agree that “there are rapid, automatic, and unconscious con- nections among emotion, cognition, and action” (p. 366). Historically, biologically-oriented theories have proposed a small number (2-8) of basic, universal emotions. Cognitively-oriented theories have taken a different track, emphasizing the wide diversity of felt emotion, distinctions be- tween similar emotions (e.g., love for your spouse versus love for your parents) and the many words for emotions in a language such as English (Reeve, 2015). At the most basic level, Solomon (1980) recognizes only two types of emotions, positive (pleasant) and negative (aversive), each of which triggers the other. The opponent emotional processes can be considered both complex and inherently conflicted or ambivalent (see MacIntyre, 2007). Izard (2007) offered a broader theoretical approach, differential emotions theory (DET), which features a com- bination of basic emotions and higher-order thought processes. According to Izard (2007), there are only six basic emotions: interest, joy/happiness, sadness, anger, disgust, and fear, which are generated by patterns of physical responding which are more-or-less automatic and hard wired, and may or may not be reg- istered consciously. The exact number of basic emotions has been widely de- bated and different theorists use different types of evidence (e.g., patterns of neural firing, facial expressions, neuro-chemical reactions inside the body) to support their viewpoints. Although the number of basic emotions has been a long-standing concern for emotion theorists (Reeve, 2105), it is the roles that emotions play in situa- tions where languages are learned and used that is likely to be of more concern for the SLA field. Izard (2007) noted that as people mature, the basic emotions are more and more rarely felt because the complexity of the appraisal of the social situation, experiential memory, and self-related cognition are continu- ously modifying basic emotions. Considering the ways in which adults experi- ence day-to-day emotions, Izard goes on to describe ways in which emotion schemas combine low-level physiological (basic emotional) responding with on- going appraisals of the situation and other cognition in a dynamic process, cre- ating the emotional milieu familiar to most adults. 64 Positive and negative emotions underlie motivation for L2 learning In cognitive psychology, a schema is a mental structure to organize infor- mation and interpret events. With respect to emotions, a schema combines the various internal physiological signals (e.g., a fast heart beat) with an interpreta- tion of the social context (e.g., giving my first speech in the L2), urge to act (e.g., I want to quit this speech) and other specific cognitions (e.g., the audience looks confused) to produce an emotional interpretation (e.g., an episode of language anxiety arousal). Emotion schemas allow for the specificity and differentiation of specific emotional reactions, such as a difference in anxiety between com- municating in the L1 versus L2 (see Dewaele, 2012; MacIntyre & Gardner, 1989), as well as allowing combinations of emotions into larger complexes. The specific ways in which schemas are formed over time helps to explain individual differ- ences in emotional reactions, providing a powerful source of motivation. “After the period of early development, emotion schemas (not basic emotion per se) constitute by far the most prominent source of human motivation” (Izard, 2007, p. 265). Even though basic emotions have been found to be ubiquitous across cultures (Ekman, 1972), the repertoire of emotion schemas can vary from per- son to person based on experiences and the ways in which a person learns to differentiate one emotion from another. In this way, emotions provide a basis for both common/shared experiences as well as unique/individual configura- tions of emotions/motivations. Perhaps the single most powerful way to separate emotion schemas is to categorize them as positive or negative felt emotions (Solomon, 1980). As a note of caution, the terms positive and negative reflect the vernacular usage of the terms, that is, whether the emotion is pleasant or unpleasant, welcome or unwel- come. However, it is worth noting that all emotions are adaptive and have the po- tential to contribute to growth and well-being. Even emotions that are unpleasant or unwelcome can lead to positive outcomes, as when anger generates a con- certed effort to overcome obstacles, or an emotion like disgust leads to rejection of foods that are unhealthy or poisonous (Lazarus, 2003). Yet a person will experi- ence positive and negative emotions quite differently because they generate qual- itatively different types of feelings, and because they serve different functions. 1.3. Positive and negative emotions Although emotion has been studied for a long time, the emergence of positive psychology as a recognized field has facilitated interest in positive emotion. Seligman and Csikszentmihalyi (2000) initially proposed three pillars on which positive psychology is founded: positive character traits, enabling institutions, and positive emotion. One of the most significant contributions of positive psychol- ogy to date has been to highlight the differences between positive and negative 65 Peter D. MacIntyre, Laszlo Vincze emotions. Fredrickson’s (2001, 2008, 2013a) broaden-and-build theory pro- poses that negative emotions tend to be focussed and associated with specific thought-action trajectories (anger à destroy obstacle) and positive emotions tend to lead to expansive thinking that broadens a person’s awareness. For ex- ample, people in a positive emotional state will notice more items in their visual field, engage more social connections, and will tend to have urges to act in a greater variety of ways, relative to those with negative emotions. The increased attention and elaborated information processing associated with positive emo- tion has the additional benefit of building personal and social resources for the future (Gregersen, MacIntyre, & Meza, 2016). There is some empirical evidence that, over time, positive emotional experiences produce greater resiliency, re- sourcefulness, social connections, and optimal functioning through broad- minded coping efforts (Cohn, Fredrickson, Brown, Mikels, & Conway, 2009; Fredrickson, 2013a). This leads to an interactive, upward spiral connecting pos- itive emotionality to positive outcomes, a process that has the potential to con- tribute directly to second language development and communication. For a re- cent review of theory development and evidence supporting the broaden-and- build theory, see Fredrickson (2013a). Unfortunately, the role of positive emotions in SLA has not received much research attention (Dewaele & MacIntyre, 2014; MacIntyre & Gregersen, 2012). However, there has been some discussion of positive and/or negative emotion in larger, multi-dimensional models of learner factors, in particular the motiva- tion models by Gardner (1985, 2010), Clément (1980, 1986), and Dörnyei (2005). Space does not permit a full review of each author’s approach, and read- ers are encouraged to consult the original works. · Gardner’s (1985, 2010) integrative motive: Although attitudes are de- scribed as the drivers of motivation in Gardner’s model, there is a place for both positive emotions (including desire to learn the target language and interest in foreign languages) and negative emotions (target lan- guage classroom anxiety and language use anxiety). Integrative motiva- tion is defined by a desire to meet and communicate with members of target language community, and there is a variety of emotion schemas that come to be associated with intergroup interaction. Gardner’s model also features the concept of instrumental orientation toward learning for pragmatic reasons, such as getting a job using the target language. · Clément’s (1980, 1986) socio-contextual model: This model has key fea- tures in common with the integrative motive but is focussed on describing acculturation, using a core emotion-related process labelled “fear of as- similation.” In the case of a minority group learning the language of a ma- jority group, the desire to move toward the new language (integrativeness) 66 Positive and negative emotions underlie motivation for L2 learning is in a state of tension with the fear of losing one’s heritage language and culture. Clément’s model also features a secondary motivational pro- cess, self-confidence, which facilitates language acquisition via the inter- active effects of low anxiety and high perceptions of communicative com- petence. The combination of low anxiety and perceived competence fea- tures prominently in the model of willingness to communicate that was developed later by MacIntyre, Clément, Dörnyei and Noels (1998). · Dörnyei’s (2005) L2 self-system: Dörnyei’s model is centred around three interrelated components: an ideal self that describes what a language learner wants to become in the future, an ought-to self that captures obligations placed on the learner by other people, and a role for prior language experience. Although Dörnyei emphasizes discrepancies in cognition about the present and the future, there is a role to be played by emotional reactions that emerge from perceived discrepancies and the prior experience of positive and negative emotions associated with language learning contexts. The above language motivation models implicate emotion schemas even if the theoretical attention is directed most explicitly toward attitudes, evaluations, patterns of intergroup contact, willingness to communicate, and self-related im- agery. There is a gap in theorizing of motivation in SLA when it comes to the role of emotion and its connection to key processes, including motivation. Among studies in SLA that focus attention directly on emotion, language anxiety has most frequently been studied (see Dewaele, 2012; Horwitz, 2010; MacIntyre, in press). Language anxiety has long been conceptualized as a drain on motivation for language learning and a source of disruption in the learning process (Gardner, 1985; Gregersen & MacIntyre, 2014; Horwitz, Horwitz, & Cope, 1986). Consistent with the idea of an emotion schema described above, research has defined language anxiety as a situation-specific anxiety that devel- ops out of negative experiences with language that lead to the anticipation of further difficulties (MacIntyre & Gardner, 1989). The consequences of anxiety arousal include difficulties in processing linguistic material (MacIntyre & Gard- ner, 1994), lower academic success (Aida, 1994; Horwitz et al., 1986), and dis- ruption of social-communicative processes that lead to language development (Dewaele, 2002, 2007, 2010), including lower willingness to communicate in the target language (MacIntyre,Baker, Clément, & Donovan,2003). Whereas language anxiety and its associated cognitive and emotional pro- cesses exert a generally negative impact on learning and communicating, there is theoretical work that has begun to describe the role of specific positive emo- tions in language acquisition and communication (MacIntyre & Mercer, 2014). Arnold and Brown (1999) argue for bringing research into better balance, “much 67 Peter D. MacIntyre, Laszlo Vincze more attention is given to the question of negative emotions . . . [one] should not lose sight of the importance of developing the positive” (p. 2). In doing so, Dewaele and MacIntyre (2014, 2016) examined language anxiety and foreign language enjoyment as the “two faces of Janus” or “the left and right feet of the learner,” asking whether anxiety and enjoyment are simply opposite ends of one continuum or two different types of experience. Dewaele and MacIntyre (2014) report data from a survey of over 1700 learners that showed anxiety and enjoy- ment correlated modestly (r = -.36) but that the distribution of scores was very different between the two emotions. Further, anxiety and enjoyment showed different patterns of relationships to a number of demographic factors. The au- thors conclude that anxiety and enjoyment are best seen as two interrelated dimensions, each of which with its own trajectory of development over time. “Conceptualized as two separate dimensions, the question becomes one of de- scribing a constructive balance between enjoyment and anxiety, rather than im- plicitly taking them as opposite ends of the same dimension” (p. 262). Indeed, given that learners inevitably experience both communicative difficulties and successes over time, the ratio of positive to negative emotions might be espe- cially important in FL contexts. Fredrickson has labelled thisthe positivity ratio.3 One of the advantages of calculating a ratio of positive to negative emo- tions is that it allows researchers to control for individual differences in base rates of affect intensity; some people report experiencing more intense emo- tions more often than others, who are relatively emotionally quiet and stable (Larsen & Diener, 1987). By examining how positive and negative emotions cor- relate with motivation using both raw scores and relative positivity, we gain a more complete perspective on the connections between emotion and motiva- tion. One of the drawbacks of calculating a positivity ratio, however, is losing the nuances of the contributions of specific positive or negative emotions that pre- dict specific motivational variables. For this reason, we will include both raw correlations between emotions and motivation-related variables as well as a set of stepwise multiple regression equations predicting the motivational variables based on the scores for the specific emotions. 3Positivity Ratio is also the title of Fredrickson’s book for a general audience in which she argues for an artificially precise 2.9:1 ratio of positive to negative emotions for optimal func- tioning, based on work published by Fredrickson and Losada (2005). In a recent publication (Fredrickson, 2014) she has abandoned the specificity of this numerical ratio but retained the more general argument that greater positivity ratios promote healthy functioning and that the idea is worth studying in future research. 68 Positive and negative emotions underlie motivation for L2 learning 1.4. The present study The present study was designed to investigate the relationships between a set of positive and negative emotions and motivational factors in a second language context. The emotions are defined and measured by the modified Differential Emotions Scale (Frederickson, 2013a). The scale measures 10 representative positive emotions (joy, gratitude, serenity, interest, hope, pride, amusement, in- spiration, awe, and love) and 10 representative negative emotions (anger, shame, contempt, disgust, embarrassment, guilt, hate, sadness,feeling scared, and beingstressed). To help respondents focus on the specific emotion schema being assessed, each emotion is defined not by a single word, but rather a clus- ter of three closely related words (e.g., joy is defined by the triad joyful, glad, andhappy). Given the close connection between motivation and emotion, it is important to know how these emotions link to key motivational constructs that have been established in the SLA literature. Specifically we sample from work related to the socio-educational model (Gardner, 2010), the L2 self-system (Dö- rnyei, 2005), and the socio-contextual model (Clément, 1980, 1986) in an effort to connect basic emotions to well-established concepts of language learner mo- tivation factors. For the most part, the gap in theorizing emotion in SLA means that these specific emotions have not been studied directly, so specific predic- tions involving correlations of each emotion with motivation will not be offered. Rather, we are taking a first step in empirically testing correlations among emo- tions and well-established SLA motivation-related variables. The specific motivation concepts included in the present study include in- tegrative orientation (Gardner, 2010) that reflects reasons for language learning to develop relationships with target language speakers. Based on Taguchi, Magid, and Papi (2009), we will also measure both promotional and preventa- tive instrumentality reflecting reasons for learning a language to gain something of value (e.g., a good job) or avoid losing something valuable. From Dörnyei’s L2 (2005) self system, we sampled (a) the L2 ideal self (Dörnyei & Chan, 2013), which describes a vision for future language learning success, (b) the L2 ought- to self, which describes obligations to learn, such as to satisfy parental demands, and (c) L2 learning efforts reflecting the time and energy spent on learning. In- spired by Clément (1986), we sampled concepts related to L2 self-confidence, which reflects the perception of skills in the L2 and low levels of language anxi- ety. Given that prior research has used measures of anxiety and perceived com- petence, both separately and combined together, we will employ measures of both concepts (language anxiety and perceived competence), plus the aggre- gated variable (self-confidence). Clément’s (1986) model places a great deal of emphasis on contact between the language groups that we measured with the 69 Peter D. MacIntyre, Laszlo Vincze perceived quantity and quality of contact with target language speakers. Finally, we used a measure developed by Tropp, Erkut, Coll, Alarcon, and Garcia (1999) to examine psychological acculturation, or identification with the native and tar- get language groups. Research into L2 motivation is becoming more concerned with the learning context. Ushioda (2009) emphasizes the connections between theorizing about motivation and the specific contexts in which the research is conducted. The local context for language use is also likely to be relevant to the patterns of emotion experienced by learners over time. This study was conducted in a location where language groups interact with each other on a regular basis: South-Tyrol, Italy. Given the role that intergroup contact and communication play in the various mo- tivational processes being sampled, we expect that emotions will be correlated with motivation in a context featuring frequent contact between language groups. 1.4.1. The region Although German speakers constitute merely 0.5% of the population of Italy, in the province of South-Tyrol they make up two thirds of the local population with about 300,000 speakers. South-Tyrol is an autonomous province, where both Italian and German are official languages (e.g., Bonell & Winkler, 2006; Oberrauch, 2006). Ac- cording to the autonomy statute, filling any position in public administration requires skills in both official languages. Consequently, applicants must pass the so-called bi- lingualism exam, which has four levels, depending on the qualifications of the appli- cant, and the type of the position applied for. The two language groups have their own school systems. While in Italian language schools the language of instruction is Italian, German is taught as a mandatory second language from the second grade of the elementary school; in a similar way, in German language schools, the language of instruction is German, but Italian is taught as a mandatory second language (Meraner, 2004). However, empirical studies demonstrate that Italian speakers usually have con- siderably poorer skills in German than German speakers have in Italian (e.g., Paladino, Poddesu, Rauzi, Vaes, Cadinu, & Forer, 2009). In addition, relations between the lan- guage groups can be characterized by a certain level of social distance due to the his- torical conflicts (e.g., Eichinger, 2002; Schweigkofler, 2000), most importantly, the Ital- ianization policy led by Mussolini after South-Tyrol became a part of Italy in 1919, which banned the use of German at all levels of society (e.g., schools) and aimed to make the region monolingually Italian (see e.g., Steininger, 2003). In this region, issues of integration between language groups are features of everyday life. The ethnolinguistic vitality of both languages is fairly high (e.g., Vincze & Harwood, 2014), that is, they possess a relatively high status, strong demographic capital and broad institutional support (Giles, Bourhis, & Taylor, 70