Accessing the Self in Self-Access Learning: Emotions and Feelings in Language Advising

The importance of the affective dimension and the role of beliefs, self-efficacy and learners’ voices in language learning are recognized in the literature (Arnold, 1999; Brewer, 2006; Ogasa, 2010). Although emotions and feelings seem to play an important role in self-directed language learning (Bown & White, 2010; Candas & Eneau, 2010), little is still known about how to support the affective dimension throughout the self-directed learning process (Aoki, 1999). Clearly, the cognitive and the metacognitive, the subjective and affective dimensions of learning need to be addressed, in a self-access centre, in order to support learners on their road to autonomy. Language advising provides the appropriate arena for this. Within the professional and interpersonal relationship between advisors and learners (Ciekanski, 2007), it is easier to reflect on the affective implications of learning and to help learners to cope with them. Ongoing research into emotions and feelings in advising contexts shows that affect and subjectivity occupy a large proportion of learners’ (and advisors’) discourse. This paper makes a case for integrating reflection on the affective and subjective dimensions of learning, both in the research and in the practice of language advising.


Background
Language advising has become an integral part of many self-access centre set-ups and is recognised as a useful way of ensuring the learners' access to their own perceptions, beliefs and learning experiences, and of facilitating them in their self-directed learning processes.Influenced by the humanistic approach of Carl Rogers, the principles of advising these last few decades have largely focussed on the cognitive and metacognitive aspects of the learning process.More recently, however, as Canagarajah (2003) points out, there has been a 'social turn' in the literature, focusing on sociocultural factors and their impact on autonomy (understood as individual construction), and on learning in general.Moreover, the growing interest in the literature on the sociocultural aspects of learning such as learners' agency, and the embodying of their experiences in individual histories, open up previously unexplored areas for research on learners involved in autonomous learning.
The increasing interest in the literature on the affective and subjective dimensions of learning places an obligation on us to pay attention to and to integrate these aspects into language advising.What if a learner mentions negative feelings?What if negative emotions are not explicitly mentioned but come to the surface?What if positive emotions prevent learners from adopting a more objective focus on their own learning?We would have to admit that for most language advisors, with a background based more on pedagogy than on psychology, dealing with feelings and emotions presents a challenge.
In this paper we present the first findings of a research study on the role of emotions and subjectivity in language advising, conducted with adult learners and university students in self-access settings in France and in Germany.This study aims both at gaining a better understanding of the complex relations between emotions and cognition in self-directed language learning processes and at helping advisors to focus, at times, on affective aspects in the learners' discourse and to address them in order to support the autonomization process.
First, we will briefly present different forms of learners' support in selfaccess centre sets-up and illustrate the traditional approach to language advising.Next, we will discuss two research studies on affect in self-directed language learning (Bown & White, 2010;Candas & Eneau, 2010), taking into account both their research approach and their findings.Afterwards, we will illustrate our research approach and some preliminary findings which emerged from the discourse analysis of advising sessions with two different learners.Finally, we will draw our conclusions and make some recommendations for a research agenda.

What Support for Learning in Self-access Facilities?
In the literature on self-access language learning (SALL), a distinct shift in priorities can be recognized from an "emphasis on materials and resources" to "access to the self" (Everhard, 2012;Murray, 2011).Little (1997) identifies the necessity, within a self-access setting, of providing the learner with "access to self" in terms of their capacity to apply "to the task in hand those processes of analysis, planning, reflection and evaluation […] defined as central to the development and exercise of autonomy" (Little, 1997, p. 36).
Besides this move away from the focus on equipping and organizing selfaccess spaces with materials and resources (Gardner & Miller, 1999), a new focus has been emerging on supporting learners in the reorientation required for the change from learning in a teacher-directed mode to learning in a more selfdirected mode.Thus, self-access centres need to be perceived more as "dispositifs d'apprentissage" (self-directed learning facilities), providing, besides materials and resources, opportunities and support for learners engaging in self-directed learning (Linard, 2010). 1his support for learners assumes different forms.Most self-access centres provide study guides, tutors and language advisors, and encourage learners to create learning groups, in order to promote the social dimension of self-directed learning.These supports mostly centre on the cognitive and metacognitive aspects of the learning process: how to recognize one's learning needs, to define objectives, to choose materials and tasks, to monitor one's own learning process and to evaluate learning progress (according to Holec's (1981) definition of learner autonomy).
The traditional approach to language advising focuses on three main areas of supporting learning and fostering learner autonomy, these being: (i) listening to and observing learners' perceptions / beliefs (about learning, about the language, about themselves) ("écouter et observer les représentations"); (ii) providing conceptual and methodological information (e.g., about language learning, about learning) ("apporter des informations conceptuelles et méthodologiques"); and (iii) providing psychological support ("soutenir psychologiquement") (Carette & Castillo, 2004, pp. 78-79).However, according to Gremmo (1995, p. 45), providing psychological support may be difficult for the advisors, if learners themselves are not aware of their problems or if they lack motivation.A deeper insight into affective aspects of the self-directed learning process may help advisors to better recognize and address psychological and motivational issues in an appropriate way.

Affect in Self-directed Language Learning
In spite of the increasing interest in the more individual aspects of language learning, such as motivation, learner biographies, learner voices and identities, in the literature on learner autonomy (e.g.Dörnyei & Ushioda, 2009;Chik, 2007), very few investigations focus on affect in self-directed language learning.
Bown & White (2010) investigate awareness and control of affect in selfdirected learning processes, focusing on self-regulation 2 and processing of emotions in three different learners.Their analysis of a series of one-to-one sessions with instructors (native speakers) showed that these dialogues gave rise to strong emotions and that whereas one learner succeeded in regulating his emotions, reframing frustration and focusing instead on what he had accomplished, the two others had to deal with anxiety and frustration, so that they had to sacrifice, at times, their language learning "in order to cope with negative feelings" (Bown & White, 2010, p. 435).
In their investigation, Bown & White identify three steps in the "intelligent processing" 3 of emotions in self-directed language learning, which are: (i) the perception stage, in which learners become aware of their emotions; (ii) the stage of reflection on emotions, and (iii) the self-regulation stage, in which learners manage their emotions in order to facilitate learning (Bown & White, 2010, p. 434).
A relevant methodological finding of their research is that in order to investigate self-regulation of affect in self-directed learning, a process-oriented approach is appropriate.Nevertheless, in this field, further investigation is needed which also takes into account the effects of advisor-learner interaction on affect in self-directed learning.
In their investigation of learner autonomy and affect, Candas & Eneau (2010) analyze the learning sessions of learners in self-access mode and interview 2 According to Dörnyei's definition of self-regulation "as a process-oriented construction, focusing on self-regulatory mechanisms, involving regulation of the self by the self, to meet particular goals or manage achievement" (Dörnyei 2005, mentioned in Bown & White, 2010, p. 434), in "selfregulation of affect, the psychological self is involved in overcoming self-doubt, managing different forms of anxiety, or generating positive emotions for example, to enhance learning or achievement.Investigating the self-regulation of affect involves focusing on learner-initiated processes and strategies which manage and change affect in a productive way to enhance achievement" (ibid.). 3 For the notion of "intelligent processing", see Goetz et al., 2005.the same learners, focusing, among other things, on affective aspects of learning.
The data they obtained, triangulated with the results of questionnaires which extrapolate learner profiles and learning strategies, give interesting insights into paradoxical attitudes towards affect in the learning process.On the one hand, Candas & Eneau's findings show how, in spite of learners' difficulties in verbalizing and reflecting on their emotions, affect has a significant influence on the learning process, independent of the learner's profile and language competence.On the other hand, whereas the use of affective strategies, such as looking for interesting texts, watching funny videos, avoiding boring and demotivating tasks, or taking into account one's own mood while learning, occupy a significant place in learning management overall, clearly learners accustomed to other-directed learning, as compared with more self-directed learners, allow pleasure and fun to play less of a role in their learning (since they do not consider tasks generating pleasure and fun to be real "learning tasks") ("pour les étudiants habitués aux contrôles externes de leur activité, se faire plaisir ne revêt pas un statut de formation" Candas & Eneau, 2010, p. 150).
Based on these discoveries, Candas & Eneau believe that it would be worth rethinking the role of affective strategies in traditional classifications, such as Oxford's (1990) or Hrimech's (2000), and to consider including them among the direct learning strategies, beside cognitive and metacognitive strategies (Candas & Eneau, 2010, p. 159).
Both from a research point of view and from a practical one, it is worth focusing more closely on the learners' difficulty in expressing their emotions while learning and on the way the pedagogical dialogue in advising sessions could help them to deal with this dimension of their learning process.

Language Advising which Promotes 'Access to Self'
In the manifold panorama of self-access centres all around the world, language advising is defined and practised in many different ways.In some centres it is a mandatory part of a self-directed learning programme, while in others it is an optional service.It may be offered only for a specific language, for example, for English as a foreign language, or as a cross-language service.It can be provided by teachers trained as advisors or by learners' peers.It can be offered face-to-face or online.
Ciekanski defines language advising as: … a professional as well as an interpersonal relationship that concerns learning in its cognitive and subjective, as well as personal dimensions.
… even if advisors share the same professional definition of what an advising relationship is, this definition is constantly renegotiated in relation to the context and to each learner.The notion of collaboration is fundamental to the pedagogical approach to autonomy, and collaborative practices between advisor and learner are encouraged by the very structure of the advising interaction.(Ciekanski, 2007, p. 125) The relationship between advisor and learner is an expert-novice relationship and therefore asymmetric as far as place, knowledge and activity are concerned; however, it is essentially dialogic and interactive.In her analysis of the language advising discourse undertaken on 31 advising sessions between four learners and four expert language advisors4 , Ciekanski identifies in all advising sessions: (i) pedagogical sequences, focused on the learner's analysis of their learning activity, the advisor's feedback, evaluation, decisions for further learning (between 40% to 60% of the advising time); (ii) organizational sequences, concerning making an appointment, or negotiating about resources, learning partners, etc. (from 25% to 40% of the advising time); and (iii) conversational sequences, consisting of more personal conversations about learning, language and about the learner's feeling involved in the learning process (between 2% to 25% of the advising time).
Based on these findings, and on several years of experience as language advisors and trainers of advisors, we started to investigate the nature and role of emotions and feelings in the language advising discourse.

Emotions and Feelings in Language Advising: A Research Approach
Our research aimed at investigating the role of affect (emotions and feelings) in self-regulated learning processes and at gaining a better understanding of the complex relations between emotions and cognition in autonomization processes.In particular, we wanted to investigate how emotions and feelings are addressed by learners and by advisors in language advising sessions, in order to help advisors to recognize expressed and unexpressed emotional aspects in the learners' discourse and to be able to address them in support of the autonomization process.
Since the research is still in progress, we would like here just to illustrate our methodological approach, to offer some details about the corpus we are analyzing and give some insights into initial research findings.
The corpus consists of the transcripts of individual advising sessions recorded in self-access settings in a higher education context (Freie Universität Berlin and Université Nancy 2) and in adult education (CNAM, Paris) for a total of eight sessions, four in German and four in French, with different learners.
These constituted three German students in Berlin, who were learners of Spanish and Italian, and two French learners of English in Paris and Nancy. 5  The advising sessions were selected as being representative of different learning situations and different learner attitudes towards their learning, namely: i) adult learners involved for the first time in a self-directed learning program; ii) students preparing for examinations or wanting to improve their written or oral competence; iii) learners experiencing frustration and anxiety, and iv) learners experiencing success and satisfaction with their learning.
The advising sessions were recorded and transcribed taking into account only the oral code. 6On the basis of the transcripts, we conducted a discourse analysis for each session, investigating turn-taking, speech acts, speakers' attitudes, and interaction.
Afterwards, our analysis focused on the expression of emotions.We took as our starting-point Damasio's distinction between a) emotions as observable, neurophysiological, transitory reactions to a stimulus, and b) feelings (or emotions) as the non-observable, private experience of emotions (Damasio, 2002, p. 15) and coupled this with Plutchik's circumplex of basic emotions (Plutchik, 1980), with some modifications.We first looked at the expression of emotions through direct or indirect verbal reference "ich finde es frustrierend (I find it frustrating)"; "es ist so ein Horror (it's horrific)", speaking of an examination, but also on the basis of paraverbal and suprasegmental signals, such as intonation, speech speed, stuttering, laughter, etc. (see Kehrein's (2002) criteria for tracing emotions in speech).
Finally, we investigated subjectivity in learners' and advisors' discourse, especially with regard to the learning process.This was done according to Kerbrat-Orecchioni's (1980) classification, taking into account only explicit lexical occurrences of subjectivity in the discourse.Kerbrat-Orecchioni distinguishes several types of verbs and adjectives revealing subjectivity.We took into account specifically affective and evaluative terms.Affective verbs and adjectives express the speaker's feelings or attitudes towards a state of affairs (such as "I love it", "I am satisfied", "it's funny").With evaluative terms the speaker expresses a personal judgment on an object or a state of affairs (such as "it's big", "it's small").Kerbrat-Orecchioni distinguishes two kinds of evaluative adjectives: axiological and non-axiological adjectives.Axiological adjectives express a value judgment (such as "it's crucial", "it's nice", "it's excellent").Some examples which emerged from our corpus are displayed in Table 1.
In our investigation, we took into account all the occurrences of subjective verbs and adjectives both in the learner's and in the advisor's discourse, categorizing them according to the typology mentioned above.
A preliminary finding of our research was that the occurrence of subjective and affective discourse depends on individual and contextual dimensions: some learners are more likely to use affective discourse than others.Also, some advising session contexts are more likely to develop the utterance of affective discourse than others.For the purposes of this article, we offer findings from two sessions from the French and two sessions from the German corpus.

French corpus, learner 1
Learner 1 from the French corpus (FL1), a lawyer, 50 years old, has to deal with international clients and wishes he could work in the UK.He learned English for 7 years, more than 30 years ago.In the first session analyzed (EC2 7 ), he has experienced his first self-directed learning session: his main emotions are enthusiasm and satisfaction.In the second session analyzed, the last of the series in his self-directed learning program (EC4), he is very satisfied with his learning conditions and he is still motivated by self-directed language learning.He feels he is progressing in language and in learning and he has developed several learning strategies.In EC2 we found 65 expressions or traces of emotions from FL1 ( 22negative and 43 positive) out of a total number of 361 conversational turns, including the advisor's and the learner's turns: (e.g."il est très sympathique (he is very nice)", "on a discuté de son père (we talked about his father)", "il est corporate law (he is in corporate law)", "j'ai appris pas mal (I've learned quite a lot)", "ben disons que je craignais de pas le comprendre (well, let's say I was afraid I wouln't be able to understand him)".In EC4 8 , we found 41 expressions or traces of emotions (6 negative and 35 positive) from FL1, from a total of 135 conversational turns, including the advisor's and the learner's turns.Traces of emotion decrease from EC2 (beginning) to EC4 (end of the learning session) from 65 to 40.For some examples, see Tables 3 and 4 in the Appendix.
If we compare positive and negative emotions, FL1 shows more positive emotions than negative ones.The number of traces related to negative emotions decreases (almost 75%) from 22 to 6.We may therefore assume that FL1 feels more secure in his learning and in the advising session.Reporting on learning helps him to self-regulate his learning emotions by reflecting on his language and learning competences.
With regard to subjectivity, the results of the discourse analysis of EC2 and EC4, as displayed below (see Tables 6 and 7 in the Appendix), show that the 'subjective' in the discourse was, in each session, more dominant than the 7 EC2 is the second advising meeting. 8EC4 is the fourth advising meeting.
'objective', both for the learner and the advisor.However, the proportion of objective discourse increased in the learner's discourse from EC2 to EC4 (15% to 18%) which is congruent with the advising goals.

German corpus, learner 1
Learner 1 (L1), from the German corpus, GL1, seeks out help from the advising service in order to make a learning plan for preparing for an examination she has already failed twice.L1 expects external, other-directed help from a private teacher, the advisor and/or other persons.She would like to get things done, to be corrected in the right way.Throughout the two sessions she focuses on past negative experiences (in classroom learning at the university, at private schools, with private teachers).She feels frustrated because, having learned the language abroad, she lacks input on academic writing, which is required at the examination.At the same time, she is not capable of describing in detail what her language gaps really are.She keeps reviewing the grammar on her own and finds it frustrating; she does not speak in classroom situations because she feels she is less fluent than her classmates and she is not interested in the topics the teacher proposes.This advising session is defined by L1's negative emotions and attitude.
Fear, frustration, anticipation and negative expectation, anger with regard to particular teachers and classroom situations, the feeling of being unappreciated by teachers are the most frequently recurring emotions.
L1 expresses annoyance and fear with regard to her upcoming examination ("das ist so ein Horror (it's horrific)", "das ist so eine große Hürde (it's such a big hurdle)") and frustration about her previous failure and her competences ("ich kann das so wenig (I can do so little)").She also has a negative perception of herself as a language learner, which frustrates her even more, since she already has a PhD in another subject and she feels her language teachers look down on her.
In the first session (EC1), we found 169 expressions or traces of emotions by GL1 (154 negative and 15 positive) from a total number of 397 conversational turns, including the advisor's and the learner's turns.In the second session (EC2), which took place two weeks after the first session, we found 124 expressions or traces of emotions (97 negative and 27 positive) by GL1 on a total of 381 conversational turns, including the advisor's and the learner's turns.For some examples, see Tables 5 and 6 in the Appendix.
Since the two sessions were held within a very short time period, few differences can be observed in the emotional note: negative emotions still dominate; however, positive emotions, such as interest and satisfaction ("das war für mich auch neu, so heranzugehen, und das fand ich eine ganz gute Hilfe (it was for me a new way of approaching it and I found it quite helpful)") increase and we even found an expression of gratitude ("danke, ja (thanks, yes)").
As in the case of FL1, in these two sessions, the subjective part of the discourse dominates for the learner (see Tables 9 and 10 in the Appendix) without a significant change (58% in EC1 and 60% in EC2), whereas the adviser's discourse is mainly objective (40% subjectivity in EC1 and 39% in EC2).It seems that the advisor tries to compensate for the great amount of emotional involvement and bring more objective topics into the discourse, such as specific questions about learning activities, materials and plans.
The advisor's discourse deserves analysis, too.Our first findings show that the advisors either try to counterbalance a learner's strong emotions, mitigating frustration or asking for clear examples to explain this, or at times empathize with the learner, echoing his/her emotions.

Conclusions
Although specific dialogue between advisor and learner is usually described in the literature as ‛objective' discourse, helping learners to develop criteria for autonomous learning and sufficient detachment to describe and analyze the learning situations (see Abé, Gremmo & Régent, 1981), our findings show that the place of affect and subjectivity, both in learners' and advisors' discourse, is more extensive than objectivity, partly because of the interpersonal dimension of the advising sessions Moreover, emotions and subjectivity are strongly present even in the pedagogical sequences of advising sessions, when learners report on their learning activities, evaluate learning progress or failure or discuss future learning steps.
Avoiding the personal dimension in advising sessions would be nonsensical; however, many advisers do not feel at ease dealing with the psychological aspects of learning.
It becomes clear from investigation of affect and self-regulation in selfdirected learning that the expression of emotions and subjectivity in language advising are areas that should be integrated into the research agenda and into the training of language advisors in order to identify ways of supporting the 'self', both in self-access and in self-directed learning in general.A 75: j'ai fait un petit peu j'ai regardé un petit peu hier j'ai mis des moins des plus en fait euh ce qui m'a euh ce qui m'a euh enfin ce qui m'a comment dire euh ce que je suis bien actuellement c'est Task listening (oui) donc euh A 109 : excusez moi mais il faut que j'enlève mon pull il fait chaud disappointment, frustration A 73 : oui je pense que je pense qu'effectivement il a un débit assez assez lent mais parfois dans des dans des explications là je je suivais plus apprehension, fear, frustration A 167 : et puis là j'ai un manque de vocabulaire et puis après c'est un problème de prononciation fear, expectation A 55 : ben disons que je craignais de pas de pas le comprendre annoyance, disappointment, frustration A 51: ben je trouve que je suis très hésitant (hum hum) et que j'ai pas beaucoup de vocabulaire A119 : j'ai eu du mal à suivre parce qu'en fait il me manque beaucoup de vocabulaire

Analysis
of discourse content in advising sessions offers illuminating insights into learners' experiences.For the researcher, they provide precious firsthand information, which should be triangulated with other data, such as data from interviews with learners and advisors following the advising sessions or learning sessions, learner biographies and learner logs, within a process-oriented research approach.Notes on the contributors Maria Giovanna Tassinari established the Centre for Independent Language Learning (CILL) at the Freie Universität Berlin, Germany and has been running it since 2005.Her research interests are learner autonomy, advising and multilingualism.Her PhD on "Learner autonomy in foreign language learning: components, competencies, strategies" was awarded with the Bremer Forschungspreis des AKS (Arbeitskreis der Sprachenzentren) in 2011.Maud Ciekanski is an associate professor at the Université de Lorraine, France and holds a PhD in language education on advising and self-directed language learning from the CRAPEL (Nancy 2).Her current research domains are language advising and technology-mediated language learning.Since 2008 she has been on the Editorial Board of the French journal ALSIC (Apprentissage des Langues et Systèmes d'Information et de Communication).