Intercultural Communicative Component (ICC) in the English Second Language (EL2) Curricula: Trends and Challenges of Transformation in South Africa (SA) and the Global World

International news articles published in various languages are usually tailored by the editors for their target readers in the corresponding language. On the one hand, the transformations made to the textual structure of news are intended to offer better accessibility to the information and make it more relevant to the readers in the target culture. On the other hand, the domestication of news provides compatibility with social attitudes, cultural background and shared knowledge of language users. Many such adjustments are made to the structures related to national and cultural identities and their construction. This paper will consider rhetorical and linguistic devices of discursive identity construction in the media coverage of events related to the EU migration crisis in four European languages (English, French, German and Spanish), as well as analyze their potential role in framing public opinion towards the migrant crisis.


Introduction
The demand for efficient use of English as a lingua franca and dominant international language at various levels has given rise to increased interest in intercultural communication (henceforth IC) and communicative EL2 teaching, as well as social transformation. IC entails "interactions between speakers who have different first languages, communicate in a common language, and, usually represent different cultures" (Kecskes 2015:175). This implies that language educators have a double challenge to prepare learners with skills for sharing their linguistic knowledge and intercultural resources in multilingual and intercultural contexts, while interacting with others of diverse cultural, ethnic and racial backgrounds. The purpose of this article is to encourage language educators to promote the plurilingualism perspectives, as well as to facilitate learners' acquisition of plurilingual competence. In addition, it examines how language educators can use the metalinguistic inquiry-based curriculum in the development and assessment of students' critical cultural awareness (henceforth CCA), as well as their construction into competent inter/plurilingual speakers, with the ability to relate to others from other cultures.
Currently, the Common European Framework of Reference for Languages (henceforth CEFRL) advocates a new plurilingualism stance, where a small minority of people such as the elite speak more than one functional language, and describes the plurilingual competence levels that foster IC (Council of Europe 2001:Xii). The model was established in 1999 to deal with the increasing linguistic and cultural diversity (Piccardo 2017). Despite massive global diversity, a truly plurilingual world is still unattainable.
Whereas CCA was originally the key component in Byram's (1997) model of ICC, it is scarcely discussed in the literature pertaining to the framework for intercultural teaching (Nugent and Catalano 2015). While certain scholars analyse the dimension of CCA, they only focus on how to direct students in managing stereotypes and ignore how to integrate the CCA aspect in the EL2 curriculum (Byram 1997, Kramsch 2004, Houghton 2013cited by Nugent and Catalano (2015.
The current article aims to analyse the feasibility of integrating the CCA dimension into the university EL2 curricula and encourage educators to be innovative in facilitating learners' acquisition of Plurilingual competence. Next, the article examines how language educators can use the broad metalinguistic inquiry-based paradigm as a tool to incorporate CCA into the current EL2 within the SA context, as well as to assess learners' cultural purposes and their construction into plurilingual speakers. Finally, it discusses recommendations and implications, for the integration of a plurilingualism stance into EL2 education context in SA.
First, this article outlines a brief overview of the pertinent literature, associated with the culture construct and how it has been approached in language teaching in relation to other traditional concepts and the new plurilingual perspectives in the globalised world. Next, it highlights the need for educators to understand the inter-link between language and culture, as well as how to address the pedagogical challenges and classroom affordances, associated with integrating the plurilingualism stance. It concludes with suggestions and implications for EL2 learners' creative use of multiple languages and social, as well as the psychological intercultural norms within the South African Ubuntu-based context.
The following questions will guide the analysis: Question 1. How can language educators use the social cultural aspects of CCA and pedagogical perspectives of the EL2 university curricula as tools, to facilitate students' acquisition of the ICC/plurilingual competence? Question 1b.What are the challenges of implementation of the ICC component? Question 1c. How can language educators bridge the gap between theory and practices? Question 2a. How can language educators, use the tenets of the metalinguistic inquirybased curriculum to develop learners' CCA and assess their development into inter/plurilingual speakers? Question 2b.What psychological and cultural elements can be used in the curriculum to facilitate students' identity construction?

Conceptual Background: Culture: global and local trends
Global trends in EL2 learning underscore the teaching of culture as a fifth skill, and scholars in intercultural language teaching (henceforth ICLT) emphasise the significant benefit of culture incorporation in the EL2 curriculum (Byram 1997). Educators need to understand the recent approaches to teaching the culture construct, and how to use them in the development of learners ' ICC. Liddicoat (2001) cited by Newton (et al. 2015) delineates two cultural conceptions, namely the static and dynamic views. Whereas these views have contributed to a deeper understanding of culture in L2 teaching, the static view has particularly treated it as peripheral.
Specifically, culture is defined in relation to nation states and fixed 'expanses of space' represented on a map (Streeck 2002:301 cited by Newton et al. 2015 as symbolising: selfcontained factual knowledge, folk dances, festivals, fairs and food ( Phillips 2001:48 cited by Newton et al 2015. Using multilingualism as a lens, the static view focusses on the dynamic process of language acquisition that keeps languages separate. Unlike plurilingualism, which focusses on the interrelationship and interconnectedness of languages, but not exclusively at the individual level, the static view marginalises and trivialises culture by focussing on cultural generalisations and stereotypes (Council of Europe 2001 cited by Newton et al. 2015).
In the 1970's, the goal of EL2 was to develop learners' CC and that notion underpinned communicative language teaching approaches (Hymes 1972 cited by Uso-Juan and Martinez-Flor 2008). CC comprises grammatical, sociolinguistic, discourse and strategic competences (Canale and Swain 1980). According to Byram (1997), the goal of EL2/FL learning is no more the acquisition of CC in English, but the acquisition of ICC, that promotes the notion of plurilingualism, although in more subtle ways. He maintains that culture is inter-linked dynamically to language and IC, so its themes of plurality and hybridity regarding ICC should be integrated in EL2 teaching. Therefore, to address the question of how language educators can facilitate the acquisition of ICC, they need to understand that the notion of culture is important in IC.
This stance to language teaching embodies the dynamic view of culture, which is central to Byram's (1997) model of ICC. Educators need to understand that culture requires tutorials for effective development, as it is concerned with understanding differences in sociocultural norms, values, attitudes, behaviours, self-concept, emotions, beliefs, products, traditions and creating a frame of mutually understood meanings across boundaries.
One of the limitations of Byram's 1997 model is that it describes the sociocultural component of language competence and fails to link it to other competences (such as sociolinguistic, linguistic and discourse competence). In addition, ICC is concerned with intercultural aspects, which are often politically, religiously, and socially influenced. This may sometimes hinder the process of communicative language teaching (CLT); however, language educators need to understand that Byram (1997) delineated four dimensions of sociocultural competence involving five components: attitudes, knowledge, skills (of two types) and awareness. He also transformed them into teaching objectives, such as:  Attitudes of values/beliefs; curiosity and openness; relativising others;  Knowledge of self, and others in communication; of other cultures, processes of interaction-individual and societal skills for interpreting and relating;  Skills for discovering and interacting;  Awareness: critical cultural awareness (Byram1997 cited by Newton and Shearn 2015: 23).
The EL2 educators may use these objectives to tailor eclectically, classroom tasks and activities; ideal for intercultural/plurlingual language teaching (ICLT), learning, as well as assessment hence the model is very influential in ICLT. Above all, for Kramsch (1993( , cited by Newton et al. 2015, the themes of cultural hybridity and plurality, of the significance of shaping and reshaping culture and of the contested nature of cultural identity underpin the proposals for intercultural approaches to language learning. The main challenge in teaching culture from the dynamic perspective is that the speaker's knowledge of culture is implicit, since the cultural beliefs, values, norms, myths and practices are hidden. Moreover, while one of the teaching goals has been to promote multilingualism in SA and ensure that all the 11 official languages are perceived as assets, as opposed to being problems, in practice the state bureaucracy and political leadership seem to be "trapped in a language-is-a problem paradigm" (Alexander 2003:16).

Background: Challenges/ trends and practices
In South Africa, since the first democratic elections in 1994, which coincided with the rapid global transformation, some higher institutions are, specifically faced with the problem of escalating student numbers from disadvantaged backgrounds. Furthermore, previous research indicates that even undergraduate students, who are linguistically competent in English, still lack socio-pragmatic knowledge, as the teaching EL2 syllabi often focus on isolated and decontextualized activities. Some educators even at university level, train learners often to entails a web of interlinked socio-political and historical factors that shape the students' identities, and facilitate the development of learners' cognitive awareness and EL2 proficiency. Therefore, students' cognitive awareness and EL2 proficiency need integration, as they transform each other. Understanding the role of social power is also significant in the social world (Kumaravadivelu: 2006 cited by Maftoon and Shakouri (2013).Furthermore, as facilitators, in addition to developing the acquisition of linguistic competence, lecturers have to focus on pragmatic competence, which is also significant to the EL2 acquisition process.
In addition, Gardener (2002) cited by Stefansson (2013) claims that the cultural context may have a significant effect on the individual's ultimate success in language learning. This is motivated by various elements and attitudes such as values, meaningfulness, awareness, of the learners' learning potential, expectations, ideals and differences in personality. The assumption is that in the cultural, educational context, attitude plays a significant role in determining the individual level of motivation (Gardener 2000 cited by Stefansson (2013). Language educators need to understand that language learning occurs when a new reflective active process takes place hence new information can be linked to prior knowledge. Educators need to consider learners not as failures, but as competent active plurilingual speakers constructing their own proficiency; instrumentally motivated by integrative and autonomous attitude of using metalinguistic and metacognitive skills (Piccardo 2013).
As the acquisition of plurilingual competence is not automatic, language educators need to be innovative to make the plurilingual trajectory less chaotic. For effective development of learners' plurilingual competence, educators need to realise that the plurilingual world is characterised by fragmentariness, irregularities and fuzziness, therefore adopting a new complexity-thinking stance such as plurilingualism (Gibson 1979/1986, cited by Aronin and Singleton (2012 is necessary to transform these plurilingual obstacles into intercultural lessons (Piccardo 2017; Galante 2017). Significantly, educators need to understand the social functional process of the plurilingual perspective. The following key principles for ICLT from classrooms level are recommended:  Using the inter-link between language, culture and communication as a basis, the language educator adopts the dynamic view of culture right from the beginning of the language acquisition process.  The goal is to develop learners' intercultural/plurilingual competence (and linguistic diversity), rather than focussing only on native speaker competence. The former is more holistic and entails shifting learners from passive to active mediators, by integrating intercultural awareness in EL2 teaching, to facilitate learners' development of language proficiency and intercultural skills necessary for identity formulation and global communication. The two competences are paradoxically inter-related, but the latter is inadequate for the development of learners' ICC and less popular, as it entails cloning learners to be passive imitators of native speakers by focussing on grammatical patterns and the acquisition of native-like language proficiency with little reference to culture (Byram et al 2002).  With a view to promote functional multilingualism, learners' bilingual and multilingual practices are supported to motivate students and creatively improve the SLA context.  Language policy in SA often reflects a compartmentalised view of language, due to the failure to acknowledge multilingual /plurilingual goals practices in higher institutions. Language educators may grab the opportunity to empower themselves with more knowledge to facilitate learners development into active critical thinkers.  The knowledge, skills and the ability to learn should be transversal and transferable across languages to create synergies, as well as aim to reach a higher common goal.
constitutional principles, South African indigenous languages and cultural values are often marginalised; bilingual education is sometimes considered harmful, a source of reduced language proficiency; and is associated with linguistic insecurity, psychological problems, as well as exclusion from a language specific community (Baker 1988 cited by Piccardo (2013). Considering that the teachers' self-confidence and awareness of the new trends in the field may determine successful language (Baranova & Skorupa (2011:123) it is maintained that educators need to be willing to learn what and how to teach to confront some of these challenges.
Regarding the second question of how language educators can use the plurilingualism component to improve the development and assessment of students as plurilingual speakers, lecturers can draw inspiration from psycho-cognitive studies. As Brown (1987:33) states, "…L2 learning somehow entails the acquisition of a second identity and because of the extremely social language context, L2 learning is associated with affective features" of language learning, such as attitude and motivation. It suffices to note, however that there are different types of L2 learners and not all of them acquire a second identity. This view is supported in Byram's (1997) ICC framework. It is suggested that positive attitudes towards cultural diversity are more likely to improve intellectual understanding, if students are allowed to explore and evaluate local and global issues. Given the diverse cultural context, the assumption is that English as a dominant language and other languages play a vital role in the harmonisation process of global multilingualism and facilitation of cultural diversity.
The main challenge is that the social economic process entails moving from solid to modernity and is characterised by institutionalised pluralism, variety, contingency and ambivalence (Bauman 1992:187 cited by Piccardo (2013. In understanding adult SLA, the tension lies between the two traditional theories of SLA (the Naturalistic approach and communicative language teaching approach (CLTA), as well as the newer cognitive approach. While the three theories are important for understanding the SLA process, the newer approach is less popular. According to Altenaichinger (2003) cited by Steffasson (2013), the traditional teaching approaches, particularly the CLTA focusses on language competence (knowledge and ability to use the target language) and communicative proficiency, acquired through motivation and awareness.
The cognitive approach focusses on helping the learner to construct a system of knowledge unconsciously, that can be used automatically. The controversies between the two traditional theories and the newer one centre around how language is acquired and learned through social context and studied in interactive encounters (Wagner 1998: 287 cited by Steffasson (2013.
Educators can make the SLA a more interactive and collaborative process, by increasing exposure and motivating the students to assess their CCA by:  Shifting their paradigm from the traditional additive to integrative and collaborative global perspectives,  Being eclectic by acknowledging the learners' wealth of metalinguistic ethnographic processes and cultural knowledge that students manage to draw on through unbalanced linguistic and various strategic repertoires.  Shifting their practices and understanding the linguistic, political, economic and pedagogical challenges,  Being innovative to design tasks that motivate learners' active participation in debates on radical and controversial issues; to deepen engagement and produce positive effects in language acquisition (Morgan 1993:69 cited by Newton and Shearn 2015). Liddicoat et al (2003:20) cited by Newton et al, 2015:43) suggests that noticing input (exposure) is the starting point to appreciate a wide range of authentic texts and sources including oral performatives, visual and written texts/sources, agency exercised through opportunities for interaction, such as drama presentations in role-plays, scenarios and representations. Furthermore, language educators need to know that the pathway for developing intercultural competence starts with a four key learning and teaching process including (Noticing, reflection, noticing and reflection): 1. Awareness raising: Students are exposed to new possibilities, encouraged to notice differences between the input/ affordances and own practices, and to reflect on their culture-general knowledge and noticing what is unfamiliar, as well as making comparisons between the observed communication and their own competences. 2. Experimentation: Students make use of techniques such as code mixing and translanguating. This involves using short metalinguistic and meta-cultural tasks linked to linguistic structures and vocabulary, often based on the students' needs; it helps to increase the students' self-esteem, agency and self-efficacy. 3. Production: Students enter a third place where they experience a paradoxical confrontation that is unique and personal; integrate the information they have acquired in language use through role-plays and communication. The educators focus on teaching the whole person as opposed to merely teaching the language. 4. Feedback: The students actively evaluate their own work; build on their prior knowledge and cultural perspectives; negotiate meaning as they discuss how they felt about speaking and acting in a particular way.
The feedback from the educator allows the learners to undergo plurilanguaging. This is a cyclic process involving exploring and identity construction, experiencing an agentic process of selecting and self-organisation. It also deals with the imbalances of communication, through awareness of chaos, adaptability, as well as negotiation through tolerating ambiguity, interpreting metaphors and nuances creatively (Piccardo 2017). By promoting plurilanguaging, language educators may encourage students to use language as a living tool that is dynamically structured and is emergent to help them shape their cultural world to decolonise their minds. Significantly, Broadbent (2017) regards decolonisation, as a controversial process involving intellectual liberation or achievement of intellectual freedom from the toxic mental oppression and colonial subjugation by transforming students' mind-sets. Similarly, this article considers effective curricula decolonisation as an intricate intercultural and intellectual plan language educators have to realise through interdisciplinary innovative and collaborative effort of all university stakeholders. Educators may design specific tasks creatively to facilitate learners' critical awareness (e.g. of the Ubuntu values), by dismissing myths, and deconstructing stereotypes.
In short, educators may have to cope with the increase in designing tasks involving the use of plurilingual activities such as trying to learn other words through translanguation, and the use of idioms across languages and cultures. They may develop students' awareness that languages are similar. This third place shifts the learners from passive learners to active plurilinguals, and increases their motivation to learn about their colleagues' languages and cultural backgrounds.
Lecturers need to understand that a major goal of ICLT is to facilitate and shift the learner's position from being rooted only in experiences from their existing cultures and identities to an intercultural position between cultures, where the student can negotiate differences and interact across cultures comfortably. The assumption is that due to colonialism and the past apartheid regime, the people's identities, cultures, languages and religions may have been devalued for generations and the 'self' may have suffered spiritually, as a result of colonisation of the mind. There is hence a need for its decolonisation, in order to achieve freedom from mental oppression.
Moreover, educators may have to know the following goals for assessing learners as plurilingual competent speakers:  By demonstrating learning through metacognition/self-monitoring of learning and thinking,  By performing the language they use, speak and write through cultural awareness, language awareness validating identity, agency and inclusiveness (Galante: 2017; Liddicoat 2002 b cited by Newton and Shearn (2015).
Educators may need to ensure that learners follow the steps outlined in this pathway that reflect the process of exploring culture based on the notion of mutuality. This corresponds with the developmental pathways students may take to become proficient plurilingual speakers of EL2, as they reflect on exploration and collaboration which parallel the cultural values, competences and principles, based on the philosophy of Ubuntu in multilingual SA.
The new social cultural identity gained from exposure to others' cultural views and practices facilitates learners' openness to diversity and empathy. They may move away from ethnocentricism and positively contribute to the development of identity construction in relation to the other. The language educator achieves the ultimate goal of raising the learners' sensitivity to cultural differences for a deeper understanding of boundaries and producing intercultural speakers (Newton and Shearn (2015). Language educators thus can use the lecture hall as a sociocultural context to learn along with the students, encourage the latter to use their cultural heritage by code switching, translanguating and plurilanguating to influence creativity, attitude change and motivation, as well as contribute to the development of students' ICC.

Recommendations and implications
Despite some obstacles pertaining to vested economic interests, and possible lack of political will, promoting the multilingualism policy for and respect for pluringualism in SA and African Union is recommended. Language educators may need to examine the feasibility of SA's policy of promoting multilingualism and individual plurilingualism, based on the strategy, "making multilingualism work" (DoE 2002: 18). This may be a catalyst for:  Advancing democratic citizenship employment, mutual respect, tolerance of learners' linguistic and cultural diversity.  Educators to facilitate understanding of the relationship between language and identity.  Giving students a wider perspective of the world, they live in.  Increasing self-dignity.  Obtaining students' autonomy and giving them a wider perspective of the world, they live in.  Developing learners' awareness of the basic values of the South African rainbow nation such as multilingualism, and Ubuntuist values (such as respect, consideration, solidarity and co-operation).  Developing the ability to deconstruct stereotypes, reducing racism and xenophobia.
To understand what second culture teaching entails, language educators may particularly, need to first analyse their own cultures by displaying attitudes of openness and curiosity, as presented in the ICC framework. That way, they may know how to facilitate awareness of the complex nature of adaptation, which according to Brown (1987), includes assimilation, the stages of adaptation, such as culture shock, social distance, as well as the deep-seated affective factors like motivation and attitude.
As Liddicoat et al. (2003) cited by Newton (et al 2015) suggest, language teaching is culture teaching, therefore culture teaching should be central to language learning. The implication from a socio-cognitive perspective is that educators need to know that culture motivates and enables learners to take control of their learning. Educators have to act more as entrepreneurs and facilitators than as authorities, to transform the communication challenges into a font of foundations by tailoring authentic activities focussing on: cooperation, motivation and humanism. To improve the SLA context, educators including tutors should be willing to attend life-long in-service training programmes/ workshops to understand new policies and theories of SLA before promoting student autonomy.
One of the limitations with the intercultural approach is that, to be a proficient intercultural speaker entails acquisition of numerous interrelated skills. To date, however, no curriculum can cover exhaustively, all these aspects such as self /individual learners' identity, needs, knowledge and circumstances. This article hence intends to make language educators realise that in the 21 st century, where the learner uses EL2 with others of different cultural backgrounds, linguistic competence is inadequate to improve academic success; a more sophisticated ICC and different competence than native speaker competence is required.
Researchers in adult language acquisition similarly, contend that the process of language acquisition does not follow a universal sequence, but varies across cultures. Therefore, educators should try to understand the complex interplay of SLA theories and government policy, before implementing the proposed models (Maftoon and Shakouri 2013).
In Europe, available evidence significantly indicates that Esperanto is an "egalitarian" and a more suitable tool for raising learners' metalinguistic awareness, as well as positive attitude to language learning (Tellier & Roehr-Brackin 2017:9).Despite certain limitations, language educators need to understand that the teaching of Esperanto may produce improved cognitive benefits, and some studies have begun to address the issue. Specifically, Bishop's (1997), EKPAROLI project report (1994)(1995)(1996)(1997) outlines some Esperanto benefits. More research is needed to investigate whether the language educators may use the Esperanto model for promoting the identity of the Organisation of the African Union, or even for promoting a functional multilingual democratic, ecological language policy for an Esperanto union for SA's rainbow nation to ensure equal and democratic citizen participation (Christiansen 2009).
According to macro-acquisition research on Black South African English (BSAE), current evidence indicates that the new EL2 variety is becoming more institutionalised and widely used among the black elite in SA. Specifically, the radio corpus on BSAE provides more evidence on the new sociolinguistic reality (Kachru 2006:455 cited by Makalela (2013. However, more research is needed on the propagation of EL2 as a lingua franca through macro-acquisition to formulate artificial languages such as BSAE. Such varieties may be used in multilingual SA, to function as a catalyist for the development of both metalinguistic awareness and language competence in the institutions of higher learning (Piccardo 2017); to reduce the challenges associated with the global EL2 as a lingua franca.
Significantly, instead of seeing language diversity as an obstacle to interaction with indigenous groups, language educators are encouraged to be more perceptive, tolerant and innovative. Lecturers could help students by investing in EL2 and plurilingualism to offer the best opportunity for people who do not have EL1 to learn it well for use in tertiary institutions and mainstream economy (Heugh 2002). This may reduce repeated dropout rates and improve:  The students' standard performance in other subjects,  SA economy and global transformation,  The development of SA English varieties such as BSAE,  The learning of other languages including South African official languages, which are still marginalised (Makalela 2013;Kasanga 2006).
The new stance may offer a newer cognitive dimension on communication by focusing on how speakers, who have different mother tongues and usually different cultures, interact in a common language. While such a stance may have several disadvantages, it may provide a metalinguistic inquiry-based paradigm for assisting students from marginalised backgrounds in meeting the challenges of the current local and global immigrations, as well as the need for education transformation at undergraduate level to enhance critical cultural awareness.
To elicit awareness of plurilingual realisation, a biographical method is recommended, as it is widely used in Sociology, Anthropology and Education (Ferreira-Meyer& Horne 2017). Similarly, the Autobiography tool and language portfolio can help educators integrate other methods of language teaching such as (questionnaires, video-role plays/scenarios/ interviews) and facilitate setting authentic tasks for contrastive analysis of conversation principles/discourse; linguistic universals/cultural scripts/specifics, exploration, cooperation, negotiation and analysis (Piccardo 2013).
Specifically, for assessment, the lecturer could adopt a plurilingual/ intercultural tool recommended by the Council of Europe based on the ABC model of cultural understanding and communication (Finkbeiner 2006and Schmidt (1998cited by Newton and Shaern (2015. The comparative autobiography tool comprises key intercultural experiences intended to help learners analyse and record their intercultural experiences using several prompt questions and sentence starters which have the potential to assess learners' attitudinal capacity of empathy and behavioural aspects of communicative awareness such as address forms, etiquette, politeness forms and conversation routines (Byram 2006bcited by Newton et al 2015. One limitation of this plurilingual tool is that assessment is sometimes difficult, but it could be through group work or individual assignments, conference papers, long essays, dissertations, or theses. This is, therefore an effective teaching strategy, whereby small groups work on local and international projects, set professional networks to construct their professional, cultural and national identities. Learners of diverse linguistic levels of ability may use different learning activities; based on specific topics/themes (e.g., cultural scripts, address terms, etiquette, social norms/values, taboos).
Finally, more research is needed to validate the numerous methods for pragmatic assessment and how to use CALL programmes, as well as modern technology. This may include using participatory digital culture to shift educators' focus from grammar and academic literacy, to community involvement of social-cultural skills acquired through negotiation, and networking. The implication is that for an equitable and effective transformation of the faculty of the humanities EL2 curricula, educators need to create an appropriate social and psychological context, in which learners are made critically aware of their individual identities, beliefs and global perspectives.

Conclusion
In sum, this article is of a conceptual nature therefore, no firmer conclusions can be made. The main argument, however is that effective implementation of the plurilingualism stance, may only take place, if language educators are critically aware of the complex language policies and practices of plurilingualism. Through collaboration with all stakeholders, educators may need to embrace the diverse challenges identified, and creatively use language and cultural norms as a vehicle to integrate it with emergent technology. By using the failures related to multilingualism affordances, educators may shift their additive perspectives and perhaps transform the learning obstacles into building blocks for holistic integrated EL2 curricula. The integration of a plurilingual stance within the ICLT processes may thus require patience, sacrifice, and dedication to empower the students with CCA skills and attitudes, as well as nurture the plurilingualism vision with the Ubuntu spirit of humanism and entrepreneurism for a better global world. The article may hence have implications for promotion of SA's rich cultural heritage, the multilingualism policy for equal access to education, academic and economic success, cultural diversity, as well as political stability in SA and the contemporary global world.

Introduction
The migrant crisis has been in the spotlight of the European press throughout the last three years. International news have highlighted the events in various languages as governments of many European countries faced challenges and looked for ways to address them. Because news media play an important role in the formation and reflection of public opinion, this research will be an attempt to analyze how issues arising out of the migrant crisis tend to be disseminated into different European languages within one selected news media. This study will focus on the Internet version of Euronews. It will consider identity-related information appearing in English, German, Spanish and French versions of news articles related to the European migration crisis, as well as the linguistic devices of discursive identity construction referring to national, cultural, religious and ethnic identities.
The paper is based on the assumption that the version of the news in every language is adjusted for the perception of the target audience in terms of the journalistic tradition and cultural and historical background of the reader, as well as its relevance to them. Such adjustments are permitted by journalistic ethics as long as the information provided is true and does not contradict journalistic standards. Therefore, journalists and translators are usually given the freedom by the editors to decide on such adjustments themselves and to tailor the text to the best of their writing skills and translation knowledge (Bielsa and Bassnet 2009:15). Bielsa and Bassnet (2009) also mention alterations of text and information as strategies of news production, referring to omission, explication and synthesis as "text manipulations" in the sense of handling the text. Valdeón (2014) defines omissions, additions and permutations as three principle strategies of international news production. Bielsa and Bassnet (2009) specify in this process the following textual interventions required from the translator: change of title and lead, elimination of unnecessary information, addition of important background information, change in the order of paragraphs, and summarizing infor mation. Schäffner (1997) refers to the transfer of communicative and discursive action from its initial context into another in media discourse as "recontextualization" and defines the following strategies: selection of information (omissions, additions); restructuring and rearrangement of information; transformation of genre (e.g. interview into report on interview); and incorporation of (selected) information in a new text, including syntactic (and stylistic) adaptation, determined by purpose, practices and policies, and ideologies. It was noticed in the examined research material that different language versions of news tended to omit, generalize or specify identity-related information, which might have been potentially significant for public opinion and attitude towards the migration crisis and the corresponding state policies.
Euronews' homepage looks similar across languages and visually makes the impression that the same content exists in all versions. However, such versions are not meant to be traditional translations of an original article because their content varies. In the process of global news production, the writers may work with various sources in different languages at the same time. This information is spread quickly within a multilingual editorial team and processed into news texts in the target languages.
The objective of this paper is to trace the tendencies for adjustment of information related to the migrant crisis in Europe in the case of Euronews, and to describe their potential relation to the construction and maintenance of European and national identities, as well as the formation of public attitudes towards migrants. The findings suggested in the conclusion rely mainly on the approaches of critical discourse analysis and translation studies and are generally the following: 1. The information adjustment identified contains omissions, additions and alteration of information (for example, generalization). In some cases, omissions and generalizations are made due to the low relevancy of detailed information for the reader; in others, they may relate to patterns of information that might be potentially sensitive in the perception of the target audience. 2. Omissions of potentially sensitive information indicate the relation between information adjustment and discursive identity construction (in particular, identity perpetuation [Wodak et al. 2005]). 3. Although the style of the corpus is impartial and all unverified facts are referred to as such through corresponding words, the analyzed articles contain elements known in critical discourse analysis as "persuasive content features," which may potentially be perceived as tools that facilitate the formation of opinions and attitudes, including those towards migrants.
I expect to encourage further research into this issue in order to consider it from different interdisciplinary perspectives and thus to contribute to the existing discussions on sustainable journalism. By the mixed method of random and purposive sampling, ten Euronews articles covering the issue of the New Year's 2016 assaults in the German city of Cologne were selected. Most of the articles were published in January 2016, and the rest appeared between January 2016 and January 2017. All the analyzed articles are brief columns of similar length, mostly under 300 words. The corpus will be analyzed in four language versions -English (total of 1,607 words), German (total of 2,793 words), French (total of 2,137 words) and Spanish (total of 1,976 words)and a comparative analysis of lexical components and structure of information carried out. The methodology used for this analysis combines the approaches of translation studies and critical discourse analysis. Therefore, we will compare the patterns referring to the same information in different language versions and calculate occurrences of identity-specific information related to countries of origin and social status. Taking into account that the production and translation of multilingual news imply omissions, additions and alteration on lexical, structural and contextual levels, this deductive study will trace the potential role of such discrepancies in the construction of identities and public opinion in different language versions of the corpus. This paper highlights a part of our PhD research on identity-related information in multilingual media (descriptive case-study of Euronews) in which the corpus includes random articles on various topics, including migration, published every first week of every month during the period of over one year. Due to systematic observation of the larger corpus, it was detected that the coverage of the New Year's 2016 assaults in Cologne in the various language versions showed a different focus in terms of identity construction, which in our view could be potentially significant. To verify these assumptions made on the basis of five articles from the corpus of the PhD research, we randomly selected five more articles that appeared first as a result of an automatic search on the English version of the Euronews website.
The examples for this paper are put in chronological order. The criteria for selecting the excerpts in this comparative analysis are the following: the compared excerpts either relate to the same piece of information and are structurally positioned in the same part of the article (headline, beginning, middle, conclusion), or may clearly be identified as a translation from the source language into the target language. The criteria for comparative analysis itself are focused on identity-related information and its structure.
We assume that all the texts selected for the corpus are "trans-edited" from source texts into target texts. One target text may be re-written from multiple sources and languages, whereby some of them are identified (direct speech, author's name specified or reference to any other source) and some not. Comparative analysis of multiple language versions helps us to identify the source text elements, on the one hand, and to detect the applied strategies even if the source text is unknown or unclear, on the other hand.

2.
Adjusting media texts to the target audience: Perspective of discourse and translation studies Almgren (2017:202) writes that "since media coverage affects the public's understanding of immigration, it matters how words and images are used and how information is left out of media narratives." Mass media, including the press, radio and television, form a pa rt of global communication as the encoders, with the decoders being the addressees or target audience. According to Charaudeau (2002), relevancy is one of the principles that define a sociocommunicative situation, which implies that in order to understand each other, both the addressee and the speaker have to share the same view of the world and of themselves. In this understanding, the news text producer and the target audience must share the same identity to some extent as well. Yet this process becomes more complicated in the case of multilingual news. This explains the fact that changes made to texts in global news production are mostly related to identity. Schäffner (1997) mentions that global information is domesticated in the target cultures as "shared knowledge of an event" and always implies background knowledge specific to each culture. The tasks of selection, translation and adaptation entitle a newsmaker to perform a certain cultural mediation, "consciously creating a national cultural identity in relation to other national identities" (Clausen 2003:105). In a comparative study of English and Spanish news versions of BBC-World, Valdeón (2005:115) refers to Katan, mentioning that "translations are commissioned according to whether they fit into the target's culture (distorted, generalized or deleted) perception of the source culture," which would strengthen the ethnocentric perception as translation would originate in the target culture.
As mentioned by Snell-Hornby (2000:12), language is "one of the most potent means of expression of cultural identity, along with non-verbal conventions, norms and rules of conduct to which numbers of a group are encouraged to conform." Discursive strategies of identity construction are expressed through language. Wodak et al. (2005) define the following macro-strategies as having identity-constructing functions: construction, perpetuation (justification), transformation and dismantling (deconstruction).
Wodak et al.'s micro-strategies overlap with the stakes (French enjeu) of discursive strategies described by Charaudeau (2002:301-318): legitimation, credibility and captation, in which the stake of legitimization is used to reinforce the position of the speaking subject, the stake of credibility is used to confirm the truth of the content by manifesting the sincerity, and the stake of captation means ensuring that the addressee perceives or shares the ideas of the speaker. van Dijk's (1995) discursive strategies "employed to manipulate the prevailing models of ethnic events" also share similarities with Wodak et al.'s micro-strategies of discursive identity construction and Charaudeau's stakes of discursive strategies. In particular, strategies such as polarization between Us and Them focus on the problems caused by Them negative socioeconomic consequences of immigration, preferential quotes from selected sources and positive self-representation. (van Dijk 1995:27-45).
For the examples taken from the selected corpus, we will consider how identity -related information is represented in different European languages in terms of discursive identity construction and political rhetoric.

Critics of the "open door policy" and first weeks after Cologne assaults
By the end of 2015, the news about the asylum seekers in Europe was marked by the emphasis on the existing crisis and reports of criticism towards Angela Merkel's "open door policy." German articles tended to be careful in writing about the fall of the chancellor's popularity, either avoiding mentioning it or writing about shifts in popularity using words with positive meanings. The example below contains extracts from an article about the increasing number of refugees arriving in Germany, given in the corresponding language versions: (1) ENG: Angela Merkel has made a point of having her photo taken with refugees and early on in the crisis made it clear they would be welcome in Germany, getting ahead of the rest of the EU on the question. Now some Germans fear she may have got ahead of herself. The English version presents the information with neutrality and distantiation ("some Germans fear") using metaphorical phraseology for euphemization ("may have got ahead of herself"). The French and Spanish versions directly refer to the criticism of Merkel's policy towards refugees: the French version mentions that the German chancellor's popularity was at its lowest point in the last four years. The German version uses positive representation with the expression "sehen das Ansehen Deutschlands im Ausland aufgewertet" ("consider Germany's positive image abroad as overestimated"), and mentions that Merkel is only "auf Platz vier der Beliebtheitsskala" (at fourth place on the scale of favorites). This may be viewed as adjustment with the function of identity perpetuation (Wodak et al. 2005).

GER: Beobachter sehen das Ansehen Deutschlands im Ausland durch Merkels
The Cologne assaults on New Year's Eve 2015 speeded the process of changes to German legislation on migrants and refugees, which followed about a month after January 1. The first news about the attack started to appear on the Euronews homepage some days after the event with the following headlines: The German version above emphasizes Merkel's commitment to decisive action. The major focus is on sexual assaults in all versions of this article. The English version lists this fact at the end of the sentence, but it is followed by first-hand speeches of the witnesses who were harassed; therefore, it does not change the focus of the news. The details of the description of the supposed attackers and their origin vary only slightly in the four language versions. The English version refer to "a crowd of about 1000 young men, many of them drunk" 4 and mention that "police say the assailants were of north African or Arabic appearance, a fact borne out by witnesses". The same allegation about the appearance is mentioned in the other three versions of the article. The English and French versions of the news state that most of the attackers were under the influence of alcohol (Fr. "Plusieurs centaines de personnes, fortement alcoolisées"). This fact does not appear in any of the articles we analyze, except once again in the English and French versions of an article published the same day. 5 Yet all the versions of the news text make reference to the appearance and probable ethnicity or national origin of the men. Their description by the police as having "North African or Arabic appearance" suggests certain potential prejudice. No references to their identity were made, for example, in terms of the language they used or any other means of communication. No documents were mentioned which could have proved their national origin or legal and social status. All of the descriptions were reduced to "North African or Arabic appearance" with unclear definitions of what such "appearance" might be. The French version avoided mentioning the description "Arabic" referring to ethnic origin, but mentioned "North African" in terms of geographic origin. The conjunction "or" in "North African or Arabic" in the German and English versions is even more confusing as it is unclear if they include, supplement or exclude each other. The issue of the individuals' origin and identity being determined by their appearance was emphasized over the fact that they were under the influence of alcohol. No names of the witnesses or police representatives who desc ribed the appearance of the aggressors were given. On the one hand, witnesses' testimonies are intended to increase the credibility of the text (Charaudeau 2014:286-297), but on the other hand, testimonies without names and personal references create the danger of fallacy through argumentum verecundian (misplaced reference to an unqualified authority) (Reisigl & Wodak 2001).
The same day, an article in all language versions mentions that there is no evidence that all the men were refugees: (3) ENG: "There are no indications that there are people involved here who have received accommodation here in Cologne as refugees," said Henriette Reker. GER:"Es gibt keinen Hinweis, dass es sich hier um Menschen handelt, die hier in Köln Unterkunft als Flüchtlinge bezogen haben." FR: "Il n'existe pas d'indication, a dit Henriette Reker, que des gens impliqués dans ces agressions aient reçu un hébergement comme réfugiés ici à Cologne." SPA: La alcaldesa de Colonia, Henriette Reker, afirma que "nada indica que las personas que perpetraron las agresiones estén en los centros de acogida". 6 In discourse analysis, this type of expression ("There are no indications that there are people… who have received accommodation… as refugees") may be regarded as a tool of suggestion, compared to "yes-but" figure (Wodak et al. 2005). Such wording suggests that although there is no evidence that these people were refugees, such assumptions do exist. Yet the Cologne mayor's stance seems to be selective for her city only, as the statement refers to the refugees registered in Cologne only and not throughout the country. This detail was lost in the Spanish translation of the quotation.
Some days after the incident, the words "attacks" and "migrant" appear together in the headline of the English version ("Majority under investigation for cologne sex attacks of 'migrant origin'") 7 . The German and French headlines refer to "foreigners" (Ausländer in German and étrangers in French). The Spanish headline does not contain any such reference. The quotation marks in the English version have the function of distantiation.
The first part of another article contains information about people from North Africa and Arab countries in the German version ("Nordafrikaner als auch Menschen aus dem arabischen Raum"). 8 All versions emphasize the occurrence of sexual assaults. The French version mentions that some of the suspects were identified without any further comments. The Spanish version refers to the suspects as "persons of foreign origin" ("personas de origen extranjero"). Some news also mentions the fact that the assaults caused demonstrations of anti-Islam activists 9 (while the fact that the men were drunk, as well as the assaults themselves, are inconsistent with Islam as religion). In the concluding paragraph of another article, the English version mentions the migrant policy and the French version notes that most of the suspects were asylum seekers ("Parmi eux figurent une majorité de demandeurs d'asile"). 10 There are also some small discrepancies in the stated number of complaints, probably caused by time constraints and different data from source texts.
A quote from a participant of an anti-immigrant demonstration serves as a tool of polarization and as a perceived threat (polarization of then vs. now). "Ich hab nicht mehr das Gefühl, dass das mein Deutschland ist" 11 is translated from the German into the other languages without evident changes ("To me, this doesn't feel like my Germany"). The French writers translated "my Germany" as "Germany that I know" ("J'ai l'impression que ce n'est plus l'Allemagne que je connais"). Quotes from witnesses and those who took part in the violent incident belong to the strategy of legitimation, providing credibility for the text, in which the media is only an observer. Some quotes of the demonstrators suggest direct connection between refugees and the assaults (the word "RAPEUGEE" as blend of words rape and refugee in "Rapeugees not welcome" quoted in the English version 12 ).
The decision of Germany's Chancellor Angela Merkel to change the legislation is not attributed to the Cologne events in the same phrase. However, it is included in the same article in which the assaults are mentioned. The French version mentions that changes to the law were adopted as a response to the criticism of Merkel (which is generally avoided in German versions and in other articles of the corpus as well). The collocation "foreign criminals" generalized in English and "kriminell gewordene Asylbewerber" in German is expressed in the French version as "demandeurs d'asile condamnés par la justice." The Spanish version does not contain such reference to the asylum seekers in the phrase.
(4) ENG: German Chancellor Angela Merkel has hardened her stance. She is promising tighter controls, expulsion for foreign criminals and a reduction in migrant numbers over the longer term in Germany. GER: Die deutsche Bundeskanzlerin Angela Merkel hatte zuletzt schärfere Gesetze verlangt, unter anderem zur Bestrafung kriminell gewordener Asylbewerber. FR: …Face aux critiques, la chancelière allemande s'est prononcée pour un net un durcissement des règles d'expulsion des demandeurs d'asile condamnés par la justice. SPA: Angela Merkel ha defendido reformas para retirar el derecho de asilo y facilitar la expulsión de quienes sean condenados incluso a penas inferiores a dos años. 13 According to the data published by Bloomberg on January 16, 2015, "Germany's welcome for asylum seekers has ebbed after the Cologne attacks, following which scores of women filed criminal complaints for sexual assault". These events also changed the attitude of Germans towards asylum seekers: "Following the Cologne assaults, 60 percent of poll respondents said that Germany can no longer shoulder the flood of refugees, up from 46 percent in December. Thirty-seven percent said the opposite, compared with 51 percent last month." 14 Bloomberg also mentions that, as of January 2016, Merkel's popularity was at its lowest point since the Greek crisis in October 2011.
In the following example, the English and French versions focus the criticism on Merkel's policy, which accelerated changes to the legislation: "Critics are blaming Chancellor Angela Merkel for letting in too many migrants" and "le scandale fragilise la chancelière Angela Merkel." The German version makes a reference to the police, who allegedly had not provided the data on the origin of the suspects and does not include any criticism towards the Chancellor. Unlike the English, French and Spanish versions, the German version lists the nationalities of the suspects: "neun Algerier, acht Marokkaner, fünf Iraner, vier Syrer, ein Iraker, ein Serbe, ein Amerikaner und drei Deutsche" ("nine Algerians, eight Moroccans, five Iranians, four Syrians, one Iraqi, one Serb, one American and three Germans")the list is organized in the descending order of the number of suspects; however, the "three Germans" are put in the end of the list. Notably in the list "one Iraqi, one Serb, one American," the designation of Middle East origin (potential migrant) comes first: (5) ENG: Some of the suspects have been identified by officials as asylum seekers. Critics are blaming Chancellor Angela Merkel for letting in too many migrants. She said changes may be needed.… For many people, the beefed-up police presence around Cologne is not enough to feel safe following the mass assaults. Some 120 women were mugged, threatened and assaulted by roughly a thousand young men. The tendency to avoid mentioning criticism towards Merkel's policy is also noticed in the example below. The English, French and Spanish versions claim that over 60% of German respondents are not satisfied with the job Merkel is doing, whereas the German version uses a generalization referring to 60% of pessimists, without mentioning the chancellor. The French version provided some more of the survey data, indicating that 56% of the respondents estimate that her policy is "bad" and that 48% were afraid of refugees:

GER: Die Polizei soll laut Medienberichten zunächst verschwiegen haben, dass es sich bei etlichen an Silvester kontrollierten
(6) ENG: Public support for German Chancellor Angela Merkel has dropped in the wake of the Cologne sex attacks that were blamed on migrants. A majority of Germans surveyed now think Merkel is doing a poor job dealing with the refugee crisis. More than a million migrants entered Germany last year, and Merkel has resisted pressures from her fellow conservatives to say: "enough". The above examples show that although different language versions may use either specific or generalized descriptions to refer to the alleged origin of the attackers, it is still the central issue in most of the articles. This may be used as an argument for potential polarization between "us" and "them" (van Dijk 1995). It was also observed that if information related to inside culture (the news in German) contains potentially sensitive patterns, it is more likely to be avoided or generalized than in the languages of outside cultures, as it was the case in the mentioned articles.

Changes to the migrant legislation and flashbacks of Cologne attacks
On February 3, 2016, the German government agreed on a set of stricter asylum measures known as "Asylum Package II." On March 2, 2016, these changes were passed by a great majority, but with the harsh criticism of the German socialists who called this package "antidemocratic" (mentioned on the official website 17 ).
Half a year after the attacks, nine people were accused of robbery, although hundreds of women reported sexual assaults. No one was convicted. Throughout all the news coverage, there was a consistent emphasis on the sexual nature of the crimes and the violation of women's rights. Yet the fact that these nine men were accused of robbery and not of sexual harassment is not evident from the English, German and French articles (this fact was mentioned in the Ukrainian version and is implied in the Spanish text): (7) ENG: The large number of accusations of sexual assault and robbery during the end-ofyear festivities caused uproar in Germany, prompting the government to fast-track a reform of its sexual assault legislation. Once done, its hoped perpetrators will be easier to convict. GER: Nach der Silvesternacht waren in Köln fast 1200 Anzeigen eingegangen, davon fast 500 wegen mutmaßlicher Sexualstraftaten. Bislang wurden in Zusammenhang mit den Kölner Vorfällen neun Männer verurteilt. FR: Jusqu'à présent, neuf hommes ont été condamnés pour les incidents de la nuit du Réveillon, pour vols. En février, le chef de la police locale avait reconnu que la plupart des auteurs pourraient ne jamais être retrouvés. SPA: A día de hoy, no hay ninguna condena efectiva. El pasado mes de febrero, el encargado de la investigación ya reconoció que la mayor parte de los agresores jamás podrían ser llevados antes la justicia. 18 Changes to the legislation on sexual assaults were adopted by the German Parliament in July 2016 and became law in November 2017. These changes state that there will be stricter application of the law if there is a clear "no" from the victim. At the same time, the earlier changes to the migrant law made it easier to expel a person from Germany in the case that any assault is reported.
The English version of same article again refers to "North African or Arabic appearance." The German version writes that two of the suspects were not registered as asylum seekers. The Spanish version refers to the "Maghreb origin" of the suspects and the French to the "Algerian and Moroccan origin": (8) ENG: Prior to Friday (May 6), nine convictions had been made for theft, but this is only the second of two cases in which a sexual assault suspect has been identified. Because public attention on migrant issues was coupled with the Cologne assaults being attributed to asylum seekers, a set of important questions were pushed aside. In particular, how are women in Germany protected from sexual assaults if these crimes are committed by local people? Are any changes in legislation necessary to protect women and to bring guilty men to justice to protect future victims? Who was responsible for the organization of the attacks? Who managed to bring such a large number of men, many of them drunk, together in one place and have them act together? Did anyone instruct them to do this?
In connection with the Cologne assaults, the German government adopted stricter laws for migrants and refugees. Yet the people who committed the crimes might have been residing (legally or illegally) in the EU without refugee status for years. Throughout the news coverage, the focus of identities was kept on the national and cultural level through the polarization of cultural values, organized construction of the image of a threat from the refugees and the legitimation of the government's decision. There was little focus on the protection of women's rights or issues of gender-based violence.
Little information appeared in the news about the results of the investigation of the incident in the months that followed. Yet a year later, on January 2, 2017, a tweet from the German police caused controversy and criticism as they used the word Nafri to refer to a group of people they found suspicious: (9) ENG: At issue was the use of the word "Nafris"according to police, an internal term to describe young men from North Africa "who have been attracting attention for years for their readiness to use violence and/or to commit criminal offences. GER: Mit "Nafris" bezeichnet die Polizei "Nordafrikaner" oder Männer mit nordafrikanischem Aussehen. FR: Les autorités ont expliqué que tous étaient des "Nafris" -l'équivalent allemand de "Nord-Af", ce qui a provoqué un tollé sur les réseaux sociaux notamment. SPA: Pero tal vez lo más polémico fuera un tuit de la policía en el que se utiliza el término "nafris" (por norteafricanos) para referirse a las personas controladas, lo que ha provocado una avalancha de críticas, al considerar que hubo un trato discriminatorio por razones de origen étnico. 20 The French version provides the French equivalent for German Nafri, which is Nord-Af, to refer to the ethnic identity. The German version also mentions that the police explained these were men with "North African appearance", without any objective features of such "appearance"which is a potential source of bias.
The English version of the same article, dated January 2, 2017, concludes that, although these assaults caused increased criticism towards Angela Merkel's refugee policy, few of the attackers were actually asylum seekers from Syria or Iraq, but came from North African countries (this fact was not emphasized in the French and Spanish versions of the article). Also notable is that in the earlier articles of 2016, the mention of sexual assaults always took precedence over the mention of robbery in the collocations or even stood alone. In the German version of this article, the order is already switched (zahlreichen Diebstählen und sexuellen Übergriffen: "numerous robberies and sexual assaults").
Of the more than 1,000 people who sexually harassed women in Cologne, none were found or brought to justice. 21 It was established that most of these people were neither asylum seekers nor refugees, but people of "North African descent." Little was said about their legal status or actual residence. The changes to the migrant legislation did not follow immediately from changes in Merkel's popularity, but a year later, public response did slightly change. As of the end of January 2017, the data indicate that "Nearly 60 percent of Germans, according to recent polls, now feel their country can cope with the refugees. That may be in part because refugee numbers have fallen from 890,000 in 2015 to 280,000 last year." 22 The Voice of America (VOA) also quoted Merkel stating that "the initial shock of the refugee impact is starting to wear off." Yet the recent rise in her popularity did not coincide with the changes on refugee policy, but instead with the Brexit process. According to data from the VOA, "analysts say probably the best explanation for her growing popular support are signs… of increased backing for the European Union amid fears of instability." Most texts tend towards synonymic usage of the words "migrants", "refugees", and "asylum seekers". The German and English texts sometimes use such vague descriptions as people of "North African" or "Arabic" origin. The French and Spanish versions, unlike the English and German, seem to avoid mentioning the names of countries of descent (Morocco, Algeria, Tunisia); however, they use the description "Magrebi." Most of the Spanish versions of the analyzed articles are similar to the French versions in content, structure and language. It is likely that most of the Spanish language versions are trans-edited from the French ones. Yet the Spanish versions were softer in their description of the origins of the men and seemed to avoid using nationalities or other identity descriptions whenever possible (e.g., referring to them as origen extranjero>: "foreign origin") or completely omitted the reference to them being asylum seekers. Such exclusion may be explained by the different historical backgrounds of the migration policies in the respective countries. In total, references to North African origin were made eleven times in the German language versions of the articles, nine times in the English, six times in the French and five times in the Spanish ones. References to the countries Morocco, Algeria or Tunisia were made up to two times in English and German, up to three times in French and zero times in Spanish. Unlike other language versions, the Spanish versions contained three references to the region of Maghrebwhich, in our opinion, may be viewed as a contextual synonym for North African. All language versions contained up to three references to Arabic origin.
English language articles contain fourteen references to refugees, ten to migrants and four to asylum seekers (total: 28). For the German versions, the rate is eleven references to refugees (Flüchtlinge), ten to asylum seekers (nine Asylbewerber, one Asylsuchende) and one to migrants (Migranten) (total: 22). The Spanish versions contain sixteen references to refugees (refugiados), five references to migrants (inmigrante) and four to asylum seekers (solicitantes de asilo) (total: 25). The occurrences for the French versions are nine references to refugees (réfugiés), seven to migrants (migrants) and ten to asylum seekers (demandeurs d'asile) (total: 26). In relation to the total word counts the English language versions have the highest rate of references to the regions of origin. All four languages have a relatively similar rate of references to the attackers' status (refugee, migrant or asylum seeker): 1.7% for English, 0.8% for German and 1.2% for both Spanish and French.

Conclusion
A journalist working in international news production carries out the function of cultural mediator. Writing about the events in one's own country or on the international scale implies not only that he or she should provide, as best as possible, truthful and unbiased information, but it must also be taken into account that the news makes an impact on public opinion and has the potential to influence national policy. The decision of the journalists and translators to alter the information so that it differs from the source text is only justified if the facts are not distorted. Such changes are often motivated by differences in the cultural and political backgrounds of the target reader. Yet in covering issues in which national and ethnic identities have to be mentioned, maintaining impartiality is more difficult. This implies a potential dilemma for a news writer: Is it possible to mention the facts without creating bias towards origin, ethnicity, religion or other identities?
Although differences have been detected in terms of the frequency of references to identities, not all of them are as essential as initially expected. The language of the analyzed corpus is in all versions correct and unbiased, all unverified facts are referred to as such through qualifying words ("suspect," "allege," "suppose"), which shows the high degree of professionalism and awareness of the news writers. At the same time, the analyzed articles contain elements known in critical discourse analysis as "persuasive content features", which include the direct description of ongoing events, evidence from eyewitnesses and signals of precision (numbers and places) (van Dijk 1988). On the one hand, news articles follow the principle of being neutral and merely describing events; on the other hand, mentioning a sequence of facts may prompt a reader to commit the logical mistake of post hoc, ergo propter hoc, confusing the chronological order of events for the casual relationship between them (e.g., anti-Islam demonstrations after the Cologne assaults) (Reisigl and Wodak 2001). The fact that some patterns of information happen to be systematically omitted or generalized, as was the case for the fall of Angela Merkel's popularity in the German versions of the articles, suggests that the process of news writing includes conscious or unconscious decisions about patterns that might be potentially sensitive in the perception of the target audience. We believe that the detected tendencies confirm that the way the events are covered may construct public attitudes, whereby the structure of information and collocation of words will be more important than individual words and their frequencies. We do not assume that editorials have any specific intentions in terms of attitude construction, yet the wellestablished traditions of news writing including format and style may contribute to such perceptions.
The principle of letting the reader decide what information means leads to a series of questions: what if media format, as well as changes to the news content and structure made due to cultural and relevance constraints, limit the options for readers' conclusions? What if, by discursive means and rhetorical strategies, the opinion of the reader and reaction of the target audience may be partially predicted before the news appears in print? Because of the constant domestication of multilingual news, different cultures and languages seem to be captured in their own dimension, measuring all world events from their own perspective. As Nash (2005:31) claims, "the words are used not only to describe the reality, but also to construct it." Translating and writing news within specific media formats therefore poses additional challenges for translators and multilingual journalists who have to face the ethical question of whether there is a moment at which the means would no longer justify the ends. Current approaches also propose replacing the "ideal of objectivity" with an ideal of subjectivity: Maeseele and Raeijmaekers (2017:127) write that "this implies that journalists, and news media organizations more generally, make their positions explicit and abandon their claim to truth in order to generate a paradigm shift" and therefore open the floor to discussion.
In conclusion, we would like to refer to research carried out by Nickerson and Winnifred (2008:796-817), who study attitudes towards asylum seekers in Australia and their relation to layers of identities. Through questionnaires and in-depth interviews, Nickerson and Winnifred (2008:807) discovered that individuals who "strongly identified with their nationalities, and those, who perceived national norms as negative exhibited less favorable responses to asylum seekers," while individuals who strongly identified themselves as humans were more positive in their attitudes. Based on these findings, the authors concluded that "activation of the human identity can suppress negativity towards asylum seekers associated with national identification" (Nickerson and Winnifred 2008:811). Supposing that such a conclusion would also be valid for national and cultural identities in Europe, maintaining a balance of references to national and human identities in news writing might be a remedy for preventing the potential construction of prejudice towards migrants and asylum seekers with target readers, and contribute to sustainable journalism worldwide.

About the author
Lesia Ponomarenko is a PhD student at Autonomous University of Barcelona (Catalonia, Spain), Department of Translation and Communication Studies; previously studied translation at National Taras Shevchenko University of Kyiv in 2003-2008 (MA in Philology) and completed joint Master program in German and European studies of National University Kyiv-Mohyla-Academy, Ukraine, and Friedrich Schiller University of Jena,Germany (Master's degree in political science).

Introduction
Current efforts to further integrate the European Union have been met with a revival in regional identities at both the supra-and sub-state level (Paasi 2009). The emergence of such identities has also been compounded by globalization (Hall 1997). The sensitivity to regional identity and the intensification of intercultural encounters at regional level poses challenges for the application of the concept of intercultural competence as a "national" approach. Regional identity is especially fraught in South East European countries, where political borders do not reflect ethnic and cultural lines and been the source of ongoing conflict. In the Albanian context, this sensitivity has an added significance insofar as the Albanian national culture surpasses the current political borders to include Albania, Kosovo, Western Macedonia, and Southern Montenegro. Since 1991, the liberty to travel, relocate, study, and conduct business across political/physical borders has made intercultural encounters and interactions increasingly common. Meanwhile, the wave of liberalization in higher education has intensified the intercultural interaction within Albanian regions in every aspect of lifeespecially university settings. Interaction with other cultures due to the EU integration reforms and the liberalization of higher education system have turned Albanian universities into exciting spaces of intercultural interaction. This provides an ideal environment to explore intercultural communication competence constructs at both the macro and micro levels and investigate their relationship. Intercultural communication competence (ICC) refers to the ability to communicate across different cultural and linguistic contexts. Despite its seemingly simple definition, however, there is seldom consensus on what "competence" and "appropriate" intercultural exchange actually entails. Previous research has focused on characteristics of competence when communicating in intercultural settings and sought to create instruments for measuring ICC. Indeed, the various methodological perspectives on and approaches to intercultural communication have resulted in diverse definitions. Scholars have suggested that a comprehensive model of communication competence requires an investigation of the constructs that lead to competence (Arasaratnam & Doerfel 2005;Arasaratnam 2009) and that variations in the ICC constructs result in different understandings of intercultural competence (Rathje 2007). In addition, the conceptualization of the intercultural communication competence "are highly diverse in their disciplines, terminologies, and scholarly and practical objectives" (Spitzberg & Changnon 2009: 5).
The need for further research on the constructs of intercultural competence across various geographical areas is also well recognized by intercultural researchers by numerous disciplines (Rathje 2007;Arasaratnam & Doerfel 2005;Arasaratnam 2009Arasaratnam , 2014. Research efforts to explore intercultural communication competence and cross-cultural comparative research have been conducted in different cultural contexts. Rathje (2007), for instance, has analyzed the debate on intercultural competence in the German speaking world and proposed a definition based on foundation, scope, application, and goal. Similarly focused on the German cultural context, Fritz, Möllenberg, and Chen (2002) replicated Chen and Starosta's (2000) ICC model. Arasaratnam and Doerfel (2005) have provided a model of competence generated from a multicultural perspective that contributes to a culture-general instrument. A new instrument was introduced by Arasaratnam (2009) to measure intercultural communication competence in diverse cultural contexts. Bennett's Development Model of Intercultural Sensitivity has been revalidated and expanded by Hammer (2011), as well as adapted to examine the cultural and religious differences among Finnish students (Holm, Nokelainen & Tirri 2010) and the extent of intercultural sensitivity of domestic and international students at British universities (Barron & Dasli 2014). Meanwhile, Croucher (2013) has explored the relationship of religious identity and individualist/collectivist variables within the communication traits among French students.
Despite the range of research, however, none have simultaneously explored intercultural communication competence at both the micro and macro level using a large sample. Addressing this gap, this study explores the constructs of communication competence in the Albanian context at the macro (cultural) and micro (regional) scale, examining the demographic variations, as well as the relationship between macro and micro constructs. In addition, this study provides insights on how current instruments behave in a new cultural context. Using exploratory factor analysis, this study generates constructs that lead to perceived intercultural communication competence in a new geographical and cultural context. By investigating communication interactions within and beyond culture, this study contributes to our understanding of intercultural communication competence.

State of research and key terms
Various disciplines have conducted theoretical and practical research on intercultural communication. Originating with Eduard T. Hall's anthropological research on The Silent Language (1959), this research was intensified with Hofstede's notion of cultural dimensions (1984,2001) and Schwartz's argument regarding cultural value orientations (2006). Intercultural communication refers to communication between people from different cultures (Gudykunst 2002). Culture is "a rich complex of meanings, beliefs, practices, symbols, norms, and values prevalent among people in a society" (Schwartz 2006:138).
A variety of approaches have been developed and employed in intercultural interaction research. Gudykunst's (1995) anxiety and uncertainty management theory explains the "causes" of intercultural adjustments, while Kim's (2006; ideological approach reveals four intercultural relations positions with implications for cultural identity. Jackson's (2002) cultural contracts theory explains how power dynamics and cultural boundaries impact the way in which individuals negotiate their identities when interacting with people from different cultures. Meanwhile, the intercultural rhetoric approach developed by Starosta (1984) and Gonzalez (2000; examines the rhetorical implications when messages cross different cultures. The ethnography of communication approach developed by Philipsen (1992) observes the portion of a culture that is devoted to communicative practices. Moreover, the dialectical approach explains intercultural competence from opportunities and constraints in the workplace (Martin & Nakayama 2015).
From the social scientific perspective, the study of intercultural interactions and communication is primarily interest in competence. There is a general consensus that intercultural competence has cognitive, affective, and behavioral dimensions (Spitzberg 1997). Briefly, ICC research can be categorized into three major perspectives.
First, the behavioral perspective of competence. Originally introduced by Ruben (1989), this perspective emphasizes the importance of behavior when conceptualizing competence. Kealey (1989) continued Ruben's behavioral approach and identified seven skills for intercultural competent behavior, while Martin (1993) emphasized the need for investigation of the competence's behavioral enactment dimension. In line with Ruben's view of competence as an individual attitude, Kim (2001) conceptualized intercultural competence as an internal capacity within an individual.
Developed by Wiseman, Hammer, and Nishida (1989), the second perspective uses an individual attitudinal approach to differentiate between cognitive, affective, and conative dimensions designed to measure the ability to develop a positive attitude toward a foreign culture. Along this line, Bennett's (1993b) Developmental Model of Intercultural Sensitivity provides a constructivist view and outlines a "continuum" of cultural awareness, understanding, and adjustment. This model views the experience of cultural difference as a process of the ethno-centric (denial, defense, and minimization) and ethno-relative (acceptance, adaptation, and integration) stages (Hammer 2011).
The third perspective concerns the Model of the Intercultural Communication Competence (ICC), which conceptualizes the ability to appropriately and effectively negotiate cultural identities to achieve a given communication goal (Chen & Starosta 2002). Integrating the cognitive, affective, and behavioral dimensions of competence, this model conceptualizes ICC along two scales: intercultural sensitivity and intercultural communication competence. Intercultural communication sensitivity (ICS) includes the cognitive and affective dimensions of ICC. Defined as "the emotional desire to acknowledge, appreciate, and accept cultural differences," ICS comprises six components: self-esteem, self-monitoring, empathy, openmindedness, nonjudgmental, and social relaxation (Chen & Starosta 1997). Intercultural communication effectiveness (ICE) is the behavioral dimension of ICC and refers to "the verbal and nonverbal ability to attain communication goals." ICE comprises four components: message skills, appropriate self-disclosure, behavioral flexibility, and interaction management (Chen & Starosta 2000;Portalla & Chen 2010).
Research pertaining to regional issues has been conducted from a variety of perspectives. With the resurgence of regional identity, there has been an increased interest in studies of regions and regional identity from scholars in numerous disciplines. This resurgence is particularly overt in the European Union, where the sub-state regions play a significant role on regional identification (Paasi 2009) as a result of policies of cohesion and regional development processes. Indeed, research has shown that regional attachments are positively correlated with the EU support of subnational development (Chacha 2012). Identity scholars have explored the relationship between regional, national, and supranational identities. Using French public opinion data, a study on regional and European identity demonstrated that cultural regional identity showed lower support for European identity, while regional political identity had higher levels of support for the European (supranational) identity (Brigevich 2016). The role of regional identity has also been recognized a as a barrier to regional development in a study conducted in Norway (Semian & Chromy 2014). Similarly, a large survey of university students in Chile suggests that high levels of regional identity are a threat to national unity and that regional and national identities are inclusive and complementary (Asun & Zuniga 2013).
In addition, there has been a research boom to explore the Albanian linguistic development and history in the post-communist era, after 1990. Several of these studies have addressed the relationship between dialect-standard language and practice (Kolgjini 2017;Topalli 1998) and the role of standardized language in national identity (Qosja 2008;Shkurtaj 2014). Extant research suggests an increase sensitivity toward regional and local identities at the political and institutional level. However, researchers have yet to explore how regional identity impacts communication interactions and how this might translate to intercultural competence. This study addresses this gap by investigating the constructs that lead to perceived competence at the cultural and regional level, as well as their relationships and demographic variation. As such, this study is centered on the following research questions: 1. What are the underlying macro constructs that explain the Albanian students' intercultural sensitivity when interacting with people from other cultures? 2. What are the underlying micro constructs that explain the Albanian students' interregional effectiveness when interacting with people from different regions? 3. What differences occur as a result of age and gender? 4. What is the relationship between cultural and regional constructs?
To answer these questions, we selected third perspective's two dimensions of intercultural competence: intercultural sensitivity (ICS) and the effectiveness (ICE). For the intercultural aspect of this study, we adapted the Intercultural Sensitivity Scale developed and validated by Chen and Starosta (2000). This scale is intended to measure the cognitive and affective dimensions of competence. For the interregional part of this study, we adapted the Intercultural Effectiveness Scale (IES) developed and validated by Portalla and Chen (2010), which includes the behavioral dimension of competence.

Subjects
This study disseminated a self-administered survey at five universities. The final sample comprised 460 students from three universities in Tirana, Albania; one in Pristina, Kosovo; and one in Tetovo, Macedonia. The students that attend these universities come from all Albanian regions. Respondents ranged from 18 to 27 years in age, and 62% were women. SPSS statistics software was used to conduct statistical exploratory analyses of survey results.

Instrumentation
The self-administered survey used in this study consisted of two parts: Part I comprised 24 items adapted from the Intercultural Sensitivity Scale, while Part II contained 20 items adapted from Intercultural Effectiveness Scale. Only three demographic questions were included: age, birthplace, and gender. Every item was translated, back translated, and adapted to best address the contextual and cultural differences between the scale's original development and the Albanian context. Corresponding to inter-regional interaction effectiveness, Part II required more adaptation and editing to better fit the intended use. The adaptation to regional items included the use of "dialect" instead of "language," "region" instead of "country," and "regional" instead of "culture." The questionnaire was two pages in length (one page printed on both sides). Space was left before each question for students to record their answers using a number from one to five. On the top of each side of the paper, corresponding with each part of the questionnaire, an explanation table explaining the five-Likert options was included: 1, completely agree; 2, agree; 3, undecided; 4, do not agree; and 5, completely disagree.

Procedure
The questionnaire was administered to students during their regular lecture time. The purpose of the survey was explained to them. Given the nature of the survey, there were no concerns of confidentiality or anonymity. The survey was administered in five universities during the first week of June 2014, with one university visited each day. Classes were randomly chosen at the location, depending on the number of classes in session at the time. Instructions for completing the questionnaire explicitly stated that "other cultures" are defined as any culture other than Albanian, with "Greek, any Slavic, Italian, American, Chinese, or others" provided as examples. In Part II of the survey, the instructions clearly defined regions and dialects in terms of their traditional development rather than current political borders. Various Albanian regions were provided as examples, as they are named in Albanian, including "Diber, Shkoder, Peje, Prishtine, Tetove, Vlore, and Shkup," geographical regions with the respective English names as follows, Diber, Shkoder, Peja, Pristine, Tetovo, Vlora, and Skopje. These explanations were provided so that students clearly understood that our study investigates how they communicate with cultures other than Albanian, as well as the effectiveness of communication within Albanian culture based on regions defined by dialect and not administrative and political borders.

Macro: Intercultural items
Prior to running exploratory factor analysis to determine the underlying factors behind the variables, we determined if our dataset was suitable for the analysis in two stages. First, we checked that patterned relationships exist by referring to the correlation matrix, with each variable having at least one strong relationship (r > +/-.3). We also checked whether our dataset had a multicollinearity problem. The determinant score for our dataset was .030, which indicates the absence of multicollinearity according to the general consensus that the determinant score must be above .00001 (Yong & Pearce 2013). Second, we checked for sampling adequacy. The Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy was .758, which meets the cut-off value of above .5 (Field 2000). Moreover, the anti-image correlation matrix (the superscripted "a") ranges between .857 and .644 in value-well above the cut-off limit of .5. Both the KMO value and the anti-correlation matrix indicate that our datasets are suitable for factor analysis. There were 15 (5%) non-redundant residuals with absolute values above .05, which is significantly lower than the cut-off limit of less than 50%.
The data were analyzed using the Principal Axis Factoring extraction method and orthogonal Varimax with Kaiser Normalization rotation method. Factor analysis works by determining the correlation patterns between observed measures and their underlying common factors (Tucker & MacCallum 1997;Costello & Osborne 2005). Based on this study's research questions, we selected principal axis factoring because the factors are successively extracted up to significant accounted variance. We chose Varimax rotation to minimize the number of variables with high loading in each factor. The intercultural items were considered uncorrelated for the purpose of exploring latent constructs in a new cultural context.
The final number of factors was determined based on eigenvalues, interpretability, and the reliability testing. For the eigenvalues, we used Kaiser's criterion of above 1 (Kaiser, 1960). The Principal Axis Factoring and orthogonal Varimax rotation with an eigenvalue cutoff of 1.0 showed two factors that met this criterion (with five more factors which did not meet it) with cumulative variance of 50.641%. The interpretability of the variables in each loaded factor was decisive after rotation and suppressing low loadings. A rotated factor matrix with factor loadings above .4 was used initially, generating five factors where only three factors had three or more variables each. This was based on the content interpretability of the items, the use of cross loadings in some items, and how the items behaved during reruns. The final decision regarding the "meaningful factors" was made by performing reliability testing on the factors generated through the exploratory factor analysis. The reliability testing helped finalize the factors, with the standardized items having Cronbach's alpha of .748 and .636. The reliability testing added to the interpretability of dimensions within factors with the interitem correlation. For example, we added item 1 and 23 to factor two, they were congruent in terms of content and helped obtain a better Cronbach's alpha value (from .610 to .636).
As a result, the 24 items of the Intercultural Sensitivity Scale were organized into two factors: labeled "Discouraging Attitudes and Behaviors", Factor I has nine items; and Factor II, labeled "Encouraging Attitudes and Behaviors", comprises nine items. Table 1 summarizes the results for inter-cultural items. Note: Items 1, 6, 11, 13, 14, and 17 did not load. Items 13 and 19 loaded above .3 and were extracted as six factors alone with no other item, and item 13 was extracted as one factor with single item.

Micro: Interregional items
The micro analysis followed the process as the macro analysis described earlier.
Here are the coefficients for inter-regional items. Each item had more than one value of r >. +/_3. The determinant score was .042. The KMO value was .792, while the anti-image correlation matrix had values from .862 to .714. There were 14 (7%) of non-redundant residuals with absolute values greater than .05. The Principal Axis Factoring and orthogonal Varimax rotation with an eigenvalue cut-off of 1.0 showed four factors with cumulative variance of 47.772%. The scree plot confirmed three factors as well, a fact that was considered in finalizing the "meaningful factors" for inter-regional items. After rotation using a significant factor criterion of .32, the factor loadings showed four factors as well. The final decision on factors was made on the basis of scree plot, the interpretability and item content, and the reliability tests for a better Cronbach's alpha. Three factors were selected: the first two, as they loaded in Varimax rotation, and then what would appear to be three and four together, as they could not stand separately because each only two items that loaded above .4 in value. Reliability testing also demonstrated that these items work better together. Table 2 summarizes the results for interregional items.

Demographic analysis
We used average factor rating to examine how students' age and gender affected their responses. The average factor rating was determined by summing the students' responses to all the items for a factor and dividing it by the number of items and responses. The average factor ranged between 1 and 5-that is, the available ratings on the five-point Likert scale. We divided age into three age groups: a) students younger than 20 years of age, b) students aged between the 20 and 23, and c) students older than 23. This age categorization was based on the assumption that 20-year-old students are midway through their undergraduate studies, that the second age category (20-23 years old) were the majority and likely at the peak of their college years, while the third category would cover older students. Descriptive analyses were conducted to compare gender and age for each factor on both inter-cultural items (see Table 3) and inter-regional items (see Table 4).

Macro: Intercultural items
There were significant differences based on age and gender in Factor I, Discouraging Attitudes and Behaviors toward other cultures (see Table 3 below). While the first two age categories-"younger than 20" and "20-23"-were not significant, the third age category-students "older than 23"-ranked significantly higher. This difference was significant on items 7 ("I don't like to be with people from different cultures"), 9 ("I get upset easily when interacting with people from different cultures"), 12 ("I often get discouraged when I am with people from different cultures"), and 15 ("I often feel useless when interacting with people from different cultures").
Differences based on gender were less significant. Overall, male students ranked slightly higher than female students. The difference was significant on items, notably item 20 ("I think my culture is better than other cultures"), followed by items 2 ("People from other cultures are narrow-minded") and 9 ("I get upset easily when interacting with people from other cultures"). However, there is a significant difference in ratings for students who did not report their gender: they ranked high compared to those who recorded their gender. The highest differences for the non-reported group were on items 15 ("I often feel useless when interacting with people from different cultures"), 22 ("I avoid those situations where I have to deal with people from cultures different than mine"), 18 ("I would not accept the opinions of people from other cultures"), 4 ("I find it very hard to talk to people from other cultures"), and 7 ("I don't like to be with people from different cultures").
While there were slight differences based on age and gender in Factor II, Encouraging Attitudes and Behaviors toward other cultures, these were less significant than those in Factor I. The differences between age categories were not significant, with the exception of three items: 24 ("I have a feeling of enjoyment toward the differences between myself and people from different cultures"), 10 ("I feel confident when interacting with people from cultures different than mine"), and 5 ("I always know what to say when interacting with people from other cultures"). Table 3 provides a summary of the results for intercultural items.
The differences based on gender were similarly insignificant (see Table 3 below). Within this factor, only item 24 ("I have a feeling of enjoyment toward differences between myself and people from different cultures") ranked significantly lower among male respondents, while item 8 ("I respect the values of people from different cultures") had lower ranking among those who did not report their gender.

Micro: Interregional items
Significant differences based on age and gender were found in the three factors with dialect and region items. Factor I, Dialect and Regional Identification as Obstacle to Interaction, differed according to age (see Table 4 below). The average factor ratings were higher in the age category "older than 23," and the lowest rating in the "20-23" category. While the difference between these two categories were consistent on all the items within Factor 1, some items stood out. Item 6 ("I have problems with grammar when interacting with people with different dialects") had the highest difference between age categories; "older than 23" was the highest (3.42) and "younger than 20" was the lowest (2.69). Item 2 ("I am afraid to freely express myself when interacting with people from different regions") had significantly higher ratings in the "older than 23" age group as well, while the "20-23" group ranked the lowest. Item 10 ("I have problems distinguishing between informative and persuasive massages during interaction with people from different regions with a different dialect than mine") had a significantly different ranking, with the older age category ranking the highest (3.15) and youngest category the lowest (2.64). The older age category had consistently higher ratings for each item with the exception of item 14 ("I often act like a very different person when interacting with people from different regions"), which ranked the lowest. While there were no differences based on gender in Factor I, the average factor ranking was higher for those who did not report their gender. The following items showed significant difference in Factor 1. Item 2 ("I am afraid to express myself when interacting with people from different regions"), in which the "unreported" gender group ranked highest and males ranked lowest. In both item 4 ("I behave differently when interacting with people from different regions") and item 16 ("I always feel a sense of distance with people from regions with different dialect"), the "unreported" gender group were ranked higher while female students had the lowest ranking. For item 14 ("I often act like a very different person when interacting with people from different regions"), the "unreported" gender group ranked high compared to both male and female groups, the latter of which showed no difference. In the gender-based comparison for Factor I, only item 12 ("I often miss parts of what is going on when interacting with people from regions with different dialects") had the highest rating compared to all other items, and there were no differences between the unreported (3.36), female (3.33), and male (3.22) gender groups.
There were no significant differences based on age or gender in Factor II, Respect and Openness to Identification and Interaction (see Table 4 below). There were slight differences in item 18 ("For me, being yourself is the best way of interacting with people from different regions") of Factor II. In terms of age, "older than 23" ranked higher (4.76) than the other two categories. In terms of gender, the female (4.61) group ranked high in comparison to the unreported (4.21) and male (4.46) groups.
There were no differences based on age or gender in Factor III, Dialect and Regional Identification as Enhancement of Interaction (see Table 4 below). There was only one slightly significant difference in terms of gender: in item 1 ("I find it easy to talk with people from different Albanian regions"), the unreported gender group ranked higher than the other two. There were greater differences between age categories in several items for Factor III. In item 19 ("I find it easy to identify with people from different regions") and item 1 ("I find it easy to talk to people from Albanian regions), the "older than 23" age group ranked significantly higher than the other two groups. Item 11 ("I always know how to start a conversation when interacting with people from different Albanian regions") had the highest mean for the "older than 23" category. In contrast, in Item 7 ("I am able to answer questions accurately when interacting with people that use different dialects")which varied significantly among age groups-the "older than 23" group ranked significantly lower than the other two groups. Table 4 provides a summary of the results for interregional items on age and gender. Factor I. Order of age categories from largest to smallest proportion: older than 23, younger than 20, 20-23. Gender categories from largest to smallest proportion: unreported, male students, female students. Factor II. Order of age categories from largest to smallest proportion: o lder than 23, younger than 20, 20-23. Gender categories from largest to smallest proportion: female students, male students, unreported. Factor III. Order of age categories from largest to smallest proportion: o lder than 23, younger than 20, 20-23. Gender categories from largest to smallest proportion: unreported, female students, male students.

Macro/Micro relationship
The correlation analysis showed a strong positive value of r = .504 at the p = .01 level. Figure 1 shows the correlations for the two scales and their respective factors. See the Appendix for Tables 5 and 6, which show the correlations for scales and factors respectively.

Discussion and conclusion
Using a large sample of university students in Albania, this study explored the constructs of intercultural communication competence at the macro (cultural) and micro (regional) scale, investigating the macro/micro relationship and the variations based on demographic characteristics. In doing so, this study posed four research questions: 1) What are the underlying macro constructs that explain the Albanian students' intercultural sensitivity when interacting with people from other cultures? 2) What are the underlying micro constructs that explain the Albanian students' interregional effectiveness when interacting with people from different regions? 3) What differences occur as a result of age and gender? 4) What is the relationship between cultural and regional constructs? The exploratory factor analysis indicated the following. First, at the macro level, the analysis determined two constructs that explain the interaction with people from other cultures: namely Factor I, "Discouraging Attitudes and Behaviors"; and Factor II, "Encouraging Attitudes and Behaviors." Factor I comprised the dimensions of negative attitudes and beliefs, negative behavior, and unsuccessful efforts for intercultural interaction. Factor II comprised three dimensions of confidence and control, respect, and encouraging efforts and behavior. Students ranked low in the Factor I and high in Factor II. As such, the two-factor structure generated by this study did not confirm the original Chen and Starosta (2000) five factor intercultural sensitivity scale structure. Second, at the micro level, the analysis determined three main constructs that explain interactions with people from different regions within the Albanian national culture: Factor I, "Dialect and Regional Identification as Obstacle to Interaction," comprised dialect and region dimensions; Factor II, "Respect and Openness to Identification and Interaction," which had no internal dimensions; and Factor III, "Dialect and Regional Identification as Enhancement of Interaction," which comprised with dialect and region dimensions. Students ranked low in the obstacle (I) factor and high in the respect (II) and enhancement (III) factors. The original intercultural effectiveness scale of six-factors held reasonably well in its application to a new cultural context, with two or more factors from the original scale loaded under one factor in the structure used in this study, and only two items not loading with the original group. As such, application to a new context-Albania-reinforced the inter-dependability between factors, which generated fewer factors in our study.
Third, the demographic characteristics were significant on both the macro and micro scales. Students "older than 23" and those who did not report their gender ranked higher on Factor I of both the macro/intercultural ("Discouraging Attitudes and Behavior") and micro/interregional ("Dialect and Region as Obstacle") level, and low in the factors that encourage interactions on both the cultural and regional scales.
Fourth, there is a strong positive correlation between macro (cultural) and micro (regional) constructs, as shown in Figure 1. Specifically, the "discouraging attitudes and behavior towards other cultures" macro-construct is positively correlated with the "dialect and regional identification as an obstacle to interactions" micro-construct. The macrofactor of "encouraging attitudes and behaviour" is positively correlated with the microfactors of "enhancement" and "respect for dialect and regional identification." This study contributes to the literature on intercultural communication by exploring constructs of intercultural communication competence in a new cultural context. The lower number of factors generated at both the macro and micro scales is better at accommodating the cultural differences and regional variations. As such, a culture neutral measure could be developed with fewer factors that accommodate a larger cultural context, leaving room for culture specific items within these dimensions. Based in the context of Albanian university students, the results of this study suggest that intercultural interaction occurs based on attitude, behavior, and effort, translating into confident/difficult and successful/unsuccessful intercultural interaction. The factors generated by this study reinforce this idea and contribute to a better understanding of how cultural variations produce different levels of competence.
The strong positive correlation between the macro and micro measures indicate that there is no significant variation between cultural and regional perceived competence. Rathje argues for an ICC definition based on a cohesive concept of culture, where "intercultural" is actually produced "culture" by establishing normality among various collectives (2006:263). This is demonstrated by the results of this study: the positive micro-macro relationship suggests that achieving ICC is a "generic competence," as being more competent at the regional level translates to better communication capabilities at the cultural (national and international) level. The results of this study suggest that a comprehensive definition of intercultural communication competence and culture general measurement could be conceptualized and developed as a macro-micro continuum, as this successfully accommodates cultural and regional variations.
Given the exploratory nature of this analysis, we were satisfied with the relatively low Cronbach's alpha values in some of the items, which could be considered as a limitation. However, this was not a confirmatory analysis. The constructs generated offer a better understanding of how intercultural communication competence works in different cultural context. While the variations in the results of this study could be an indication of macro-micro exposure, the impact of age and gender need to be examined further with greater attention to cultural, political, and historical context. Another important direction for future research is to investigate how dialect influences regional identification and communication, as the results of this study indicate higher variation in dialect than region.

About the authors
Nina Gjoci, (PhD, Bowling Green State University) is a lecturer in the Department of Humanities and Communication, European University of Tirana, Albania. Her work focuses on intercultural and international development and public memory.

Introduction
Social media has become a battleground for conflicts over political and social concerns. In the current political climate, issues surrounding race and racial identity have come to the forefront. Since race has tied into many of the sociopolitical debates on social networks in recent years, it is important to research and understand race on social networks. The purpose of this study is to better understand how race is perceived, negotiated, and conveyed on social media. Understanding how racial identity is projected on social media is important due to mass consumption of social networking sites and the assumed visual nature of race, which affects both offline and online relationships. Social media has shown remarkable growth as a major form of communication around the world-in both professional and personal contexts. For example, according to Boyd (2007), social networking sites are used as a form of mainstream socialization that can equate to offline public spaces. In other words, online social networks provide space for the kinds of interactions that used to take place in physical public spaces such as coffee shops. Boyd explains that since social networks provide a place for individuals to meet and communicate in the same way they might in physical locations, they too are a public space.
This public space provides a forum for individuals to explore racial identity. For example, Chan (2017) interviewed college students to better understand how interactions on social networks about race shaped their racial identities. Chan reported that racial information from social networks influences identities of those who use the networks by connecting them to other group members, encouraging pride in their racial identity, and by partaking or not partaking in direct or indirect discussion about race. Through the lens of the Theory of Symbolic Interactionism, Chan's findings indicate that racial identities are influenced by communication on social media.
In another study that assessed the utility of social media for exploration of racial identity, Florini (2013) explained how Twitter provides a space for people of color to connect with one another and share experiences. For example, Florini explained the phenomena of "black Twitter," in which Twitter users rely on linguistic performances with vocabulary, grammar, and textual cues to perform racial identity.
Finally, Brock (2009) claims social media to be a meeting place for people of color to discuss black identity. Brock explains that social media helps society understand race differently because the medium lacks all the physical signifiers of face-to-face communication. Due to individuals' reliance on visual perception in defining race in face-toface interactions, people of color have started to rely on social media as a space to explore racial identity.
On the one hand, the assumed visual nature of race has led to social networking sites being used as a platform for exploration of racial identity for people of color. On the other hand, assumptions individuals make when race is perceived visually may lead to unhealthy discourse on social media, stemmed in problematic definitions of race. Race is often treated as if it is a visual aspect of identity that is "set" at birth, when in reality racial identities are socially constructed and subject to change by members of society (Asante, Sekimoto & Brown, 2016;Condry 2015;diAngelo 2012;Graves Jr. 2010;Maragh 2017). Asante, Sekimoto, and Brown (2016) explain that race, in particular, blackness, is more than outside appearance; in fact, racial identity is constructed through symbols, language, culture, and group experiences. Therefore, racial identity is not about skin color, yet that is how most individuals perceive and understand race. This is problematic because how individuals understand race directly affects their conversations about race and the way in which they project racial identity in the online world. In other words, individuals bring their perceptions of race to online mediums, and the ensuing discourse can affect both online and offline relationships.
Recently, social networks have facilitated heated conversations relevant to race. NFL players kneeling for the national anthem, racialized anger towards police officers and the #BlackLivesMatter slogan, the Standing Rock Sioux Tribe's fight against an oil pipeline, and the 2016 presidential election are examples of current events that led to racialized conversations on social networking sites. Facebook pages, Tweets, and Instagram hashtags bring the words and ideas of individuals to the attention of the public in an unprecedented manner. This provides an opportunity to marginalized individuals who do not often have a voice in society; social media provides a "microphone" that allows them to be heard by large amounts of people all over the world. However, there is a downside. As made apparent with the #BlackLivesMatter movement, marginalized individuals may face animosity, discomfort, and misunderstanding when it comes to issues of race, racial expression, and identity expression on social networks.
Discussions that take place on social networks affect the lived experiences and shape the identities of the members of these social networks. After a long history of racial conflicts, the United States has arrived at a time in which racial conflicts have heightened and need to be addressed. Due to the advances in computer mediated communication and the problematic nature in which individuals define race, which affects offline and online discourse and relationships, there is a need to research how individuals explore their racial identities on social media. Researchers must find ways to better understand how race is perceived, negotiated, and conveyed on social media. Therefore, this study aims to better understand how individuals project racial identity on Facebook, Twitter, and Instagram. Due to the paucity of research exploring the projection of racial identity on social networks, this study was framed as a preliminary investigation into the topic with the following research question examined quantitatively as described in Methods.
Research question: how do Caucasian and non-Caucasian college students display racial identity on social networks?

Literature review
In order to set the framework for this study, an examination of the literature pertaining to social media platforms, online identity construction, and online versus face-to-face communication about race follows.

Social media platforms
Because of their widespread use, it is possible to examine racial identity on three primary social media sites: Facebook, Twitter, and Instagram. According to Duggan (2015), 72% of adults who use the internet use Facebook, 28% use Instagram, and 23% use Twitter. The sites were selected to allow participants to express how various social media sites affected their experiences differently. Unlike Facebook, Twitter focuses on short blogs. In other words, Twitter does not provide the full profile interface that Facebook does. On Facebook, users have the ability to change their profile picture, edit and project large amounts of personal information, control albums of photos shared by individuals and their friends, as well as blog. In addition, the global reach of Facebook offers the opportunity for facilitation of intercultural interactions (Wu & Marek 2018). What Twitter lacks in terms of depth (i.e. no opportunity for individual profiles), it gains as a frequent blogging site whose character limits force users to get to the point. For these reasons both Facebook and Twitter have been chosen for the study. The last social network examined is Instagram. Instagram was included as a platform to examine in this study because it focuses on photos and videos significantly more than Facebook and Twitter. All three social media platforms examined in this study (Facebook, Twitter, and Instagram) are non-anonymous platforms (Zhao, Grasmuck & Martin, 2008). In other words, they are designed to let users validate the online identities of one another in the offline world. Considering their similar nature as non-anonymous platforms and their differences in foci, the three different social media networks will lead to a better understanding of how race is experienced in online communities.

Identity construction on social media
Social networks provide the opportunity for identity construction. According to Boyd and Ellison (2007) social networking sites, including Facebook and Twitter, have been of interest to researchers because users intentionally construct online representations of themselves and engage in impression management. Furthermore, identities are fluid and frequently change with new experiences and environments (Foldy 2012). Foldy described identity construction as a process where identities develop and adapt to an individual's characteristics, actions, and context. Therefore, social networks provide a platform for exploration of identity projection.
Research shows that individuals project identity on social networks in various ways. Some studies have found that photographs were used to display aspects of identity (e.g., Uimonen 2013; Zhao, Grasmuck & Martin, 2008). In addition to photos, Zhao et al. (2008) found that interests, hobbies, favorite movies, artistic tastes, and narratives in the "about me" section of a social media profile are used to construct identities on social media. Boyd's (2007) findings suggest that users primarily rely on their profile, friends list, and comments/blogs to construct their identity on social network sites. Boyd claimed that people use these aspects as a means of identity management. This is because social network users are aware of the connections they have with others in the offline world. Boyd explained that the link between offline and online identities is so close that social network users are likely to present themselves so as to be viewed in a positive light by their peers.
However, there is pressure to adhere to expected norms when performing racial identity on social networks. For example, Maragh (2017) explored black racial authenticity on Twitter and found that discourse pertaining to "acting white" and "acting black" influenced linguistic performances; in other words, individuals felt pressure to act and behave according to the norms of their racial group. Fryer (2006) claimed that people of color who "act white" are sometimes attributed negative labels because they are perceived to be engaging in performance that is characteristic of whites. These studies reveal that there is pressure on people of color to resist "acting white" and instead act in a manner that is "authentic" to their racial group. This is reminiscent of the concept of racial authenticity. The term racial authenticity was first described by Johnson (2003, 3) as historic and political contexts that lead to groups using … "authenticating discourse [that] enables marginalized people to counter oppressive representation of themselves." Johnson explained the difficulty in using terminology such as blackness to give essence to specific identities because the concept is a product of history, politics, and social norms. Therefore, people of color can use social networks, like Twitter, to express identity in a manner that is defiant of the oppressive dominant racial culture and norms. Maragh (2017) pointed out that phenomena like "Black Twitter" can legitimize performances of racial identities that counter the narrative of the dominant culture.
Not only is identity performed, it can also be hidden. In addition to racial identity performances, Maragh (2017) claimed that people on Twitter engage in censorship of certain aspects of identity and highlight other parts of identity. She explained that because there are "rules" for maintaining racial authenticity, Twitter users only perform specific identities. These findings fit with the claims made by Toma and Carlson (2015) that Facebook users are selective in what they choose to post in order to project a specific image. Toma and Carlson claimed that Facebook offers a means of presenting identity selectively because of the ability to choose what is posted. In other words, Facebook users make conscious choices about the way they present themselves on Facebook. Toma and Carlson concluded that Facebook users were likely to represent selected parts of their identity as accurately as possible because their audience might easily detect enhancement of them. Participants believed others might perceive their Facebook profiles as more outgoing, adventurous, relaxed, and calm than their actual selves; this was because their posts and photos were a compilation of the fun and exciting parts of their lives and filtered out monotonous tasks. Facebook users are cognizant of how they project themselves to others, and they make calculations as to how they should selectively present their identity.

Social media, the Internet and race
Finally, it is important to examine differences between face-to-face communication and social network communication about race. Cisneros and Nakayama (2015) argued that social media has changed the way society communicates about race and racial identities. For example, they examine an ambiguous social media blog, which revealed racist discourse about the election of the first Indian American woman to be awarded the title of "Miss America." A dissection of racist remarks about the Miss America title provides the opportunity to show present day society to be just as racially-derogative and prejudiced as in past decades. Social media, especially on anonymous platforms, provides a space for racial discourse to take place in a manner that does not follow the same social norms as face-to-face conversations. For example, Brown (2009) examined hate speech on several of the internet's most visited white supremacist chat rooms. Brown argued that through discourse about separate racial identities, the socially constructed ideology of race becomes more deeply rooted in the perception of people as a reality. Brown's work suggested that the internet can provide an environment for racist discourse that supports the formation and progression of uneducated ideology.
In contrast to the previously cited studies, some research reported that social media platforms provided a space for a more positive exploration of racial identities. For example, Brock (2009) claims online spaces are used to create conversations about what it means to be Black. Online spaces work as a third place, or meeting ground, for people who otherwise would not easily facilitate their conversations. According to Brock, racial identities can be perceived differently online than in face-to-face dialogue, because race is not as easily visually assumed through social media. Florini (2013) found that social media users purposely pursue other methods of racial identity construction when there is a lack of visual representation of race. One method explained by Florini is the use of "signifying," or speaking in a particular means to give the audience cues about their racial identity. Florini explained that Twitter users must project racial identities and make them visible in order for others to recognize them. Social networks like Twitter provide a space for racial identities to be shaped and projected. This may explain Correa and Jeong's (2011) findings, which revealed that minority populations have a greater involvement on social media than their white counterparts. Correa and Jeong also found that self-expression was an important part of the online experience. The findings suggest that social media provides a platform for marginalized populations to express their views.
People can also use social networks to promote positive racial group identities. Chan (2017) claimed that people of color use social networks to project positive images of their racial groups. He explained social media can be used to display pride and empowerment of racial identity, as well as hurt and marginalization. Additionally, Chan reported that people who post about race, or engage with race related content on social media, typically do so to promote positive representation of their racial group.
The literature reviewed reveals that it is natural for individuals to construct identity through social media. However, it is not clear whether or not that identity projection often involves racial identity specifically. There have been very few studies that have explored the projection of racial identity on Facebook, Twitter, and Instagram. This study helps to fill a gap in the literature.

Participants
Data was collected from 347 participants, recruited from a medium sized university in the Midwest in the spring of 2017. Although a convenience sample, college students make ideal participants because a majority belong to the younger generation, and research shows that members of younger generations use social networking sites at a higher rate (Lenhart, Horrigan & Farrows, 2004). Other studies have used college students for social network research as well (Chan 2017;Toma & Carlson 2015;Uimonen 2013). Due to the lack of diversity on the campus in which participants were recruited, students were recruited from both college classrooms and from registered student organizations for students of color (i.e. Black Student Union and Chicano Latin-American Student Association).

Demographics
The survey was distributed to 347 participants. Most of the participants were 18-20 years old (81.8%); 14.7% of the participants were 21-23, and 3.5% participants were 24-29. Gender was fairly equally distributed with 54% of the participants female. The majority of participants were Caucasian (62%), while the rest identified as African (5.8%), African American (6.9%), Asian (4%), Asian American (2%,) Latino (7.5%,) biracial or multiracial (10.4%), or other (1.2% ). Due to the skewed sample sizes, participants were split into Caucasian (n = 214; 61.7%) and non-Caucasian (n = 133; 38.3%) groups. Since white individuals hold privilege in U.S. American society, this category looked at differences between the two groups. Any participant who self-reported as biracial or multiracial identity was placed into the non-Caucasian group, even if one of the racial identities was Caucasian. This choice was made because a multi-racial individual has the potential to experience reality differently than someone who self identifies as Caucasian due to the societal privilege often attributed to Caucasians in U.S. American society. Based on the literature reviewed regarding racial hierarchy and privilege and marginalization of non-Caucasian groups in U.S. American society, this grouping seemed the best fit in answering the research question.

Scale development
Scales used in past research were reviewed to examine the prospect of adapting them for this study. However, existing scales did not examine the projection of racial identity in the context of social media. Instead, they explored perceptions of nationality and bias [i.e. Vandiver, Cross, Worrell & Fhagen-Smith's (2002) Cross Racial Identity scale], self-esteem and feelings of belongingness of adolescents (i.e., Casey-Cannon, Coleman, Knudtson & Velazquez's (2011) Multigroup Ethnic Identity Measure, Collective Self-Esteem Scale-Race, and Multidimensional Inventory of Black Identity), and the development of a non-racist identity (i.e., Carter's (1996) White Racial Identity Attitudes scale). These scales are useful tools for understanding aspects of racial identity. However, they are not useful for exploring racial identity projection in the context of social media, and social media is a vastly different framework than face-to-face communication. For example, individuals would not project their racial identity with the use of pictures in a face-to-face context, yet that would be a very common way to project racial identity through social media. Therefore, newly developed scale items were necessary to measure effectively in this context. However, when crafting questions for each of the scales, previous findings from qualitative studies pertaining to the projection of racial identity in online environments were considered.
Findings from previous qualitative studies confirmed that it was important to ask participants questions about displaying racial identity through photos and text. For example, Chan (2017) found that people of color use social networks to project positive images of their racial groups by sharing media content (i.e. photos, videos, hashtags), posting about accomplishments within the context of racial identity, and commenting on race related content in an effort to positively promote their racial group (i.e. text). Florini (2013) found that Twitter can be used to express racial identity though text. Other studies found that participants used photos to project aspects of identity on social networks (Toma & Carlson 2015;Uimonen 2013;Zhao, Grasmuck & Martin, 2008). More specifically, Uimonen (2013) found that participants used photos to display racial, cultural, and religious identities. These studies suggest that questions related to photo and text depiction of racial identity are important. Therefore, scale items were developed with this previous research in mind.
Relying on themes found in previous research on the depiction of racial identity, two primary scales were developed: the Depiction of Racial Identity Scale, and the Exposure Scale. When developing questions for each scale past research was consulted, as described above.

Depiction of Racial Identity Scale
The Depiction of Racial Identity Scale (DRIS) was developed to measure the extent to which participants in this study used photos and communication/text to convey racial identity. The DRIS is a 6-item Likert scale in which responses ranged from 1 (never) to 5 (always). Although the Cronbach's alpha was high for this 6-item scale (α = 0.881), due to the conceptually distinct nature of the concepts measured (photos and communication/text), the scale was separated into two subscales: the Photos subscale and the Text subscale. Additionally, this division would likely lead to a more meaningful interpretation of the results. Both scales had high alphas: α = 0.910, α = 0.833. The Photos subscale had three items, including "Do you use the photos you post to [Facebook]

Exposure Scale
The Exposure Scale (ExS) was developed to measure the extent to which participants interacted with race on social networks. The ExS is a 9-item Likert scale in which responses ranged from 1 (strongly disagree) to 5 (strongly agree). Although the Cronbach's alpha was high for this 9-item scale (α = 0.889), due to the conceptually distinct nature of the concepts measured (hiding racial identity and interacting/directly communicating with race), the scale was separated into two subscales, the Hidden Identity subscale and the Direct Communication subscale, to allow a more exact analysis of the two ideas.

Data collection
Using systematic sampling, students from 12 sections of the basic communication course were offered extra credit to participate in the study. Additionally, to increase the diversity of the sample, the opportunity to participate was offered to nine student associations for students of color. Students from two of those registered organizations (i.e. the Black Student Union and Chicano Latin-American Student Association) chose to participate. A $10 Chipotle gift card was awarded to a random participant in each student group. The survey software Qualtrics was used to administer the survey; a link was provided to allow a participant to anonymously take and submit the survey online.

Data analysis
Data was analyzed using independent samples t tests with racial group categorization (Caucasian or non-Caucasian) as the independent variable and the four scale scores as dependent variables. Independent samples t tests were conducted to test for significant differences. Additionally, effect sizes were reported to allow a better understanding of the t test results. An effect size of r = .2 indicated a small effect, while r = .5 indicated a medium effect and r = .8 indicated a large effect. To test the reliability of the scales Cronbach's α was used. An α > .7 indicated high response reliability for the scale.

Results
The research question sought to discover ways in which participants signify, indicate, or display their racial identity on social networks. Using four scales, this question was explored on the four dimensions of photos, text, hidden identity, and direct communication. Results for the subscales are described below and included in Table 1.

Photos subscale
There was a statistically significant difference between Caucasian and non-Caucasian participants for photos, t (338) = -6.503, p < .001. The means for Caucasian participants (M = 1.303; SD = 0.714) and non-Caucasian individuals (M = 1.935; SD = 1.078) suggest that Caucasians were less likely to post photos on social networks to display their racial identity. However, results indicated a small effect size of r = 0.3266.

Text subscale
There was a statistically significant difference between Caucasian and non-Caucasian participants for Text, t (343) = -9.646, p < .001. The means for Caucasian participants (M = 1.160; SD = 0.4417) and non-Caucasian individuals (M = 1.838; SD = 0.8594) suggest that Caucasian participants speak with others about their racial identity on social networks less than non-Caucasian participants. Results indicates that means are approaching a medium effect size (r = 0.4448).

Hidden identity subscale
The subscale Hidden Identity examined whether or not participants reported hiding their racial identity on social networks. Results of the t test indicate that there are no significant differences between Caucasian and non-Caucasian groups, t (333) = -.995, p = .320.

Direct communication subscale
There was a statistically significant difference between Caucasian and non-Caucasian participants for Direct Communication, t (338) = -4.947, p < .001. The means for Caucasian participants (M = 1.303; SD = 0.714) and non-Caucasian individuals (M = 1.935; SD = 1.078) suggest that Caucasian participants interact with racial content on social networks less than non-Caucasian participants. However, the effect size is small (r = .2588), indicating that the difference in racial interactions on social networks is very subtle, and would be difficult to detect just by looking at the social networks.

Discussion
The purpose of this study was to understand how college students display racial identity on social networks. Findings suggest that although participants do not intentionally hide their racial identity on social networks, they do not intentionally display racial identity on social networks either. However, non-Caucasian participants in this study were significantly more likely than Caucasian participants to display racial identity with the use of photos, text, and communication. Implications of these findings are discussed.
Previous research reveals that identity is conveyed through blogs and textual displays on social networks (Boyd 2007;Florini 2013). However, results of this study indicate that racial identity is not conveyed intentionally through photos, text, or communication on social networks. Perhaps it is instead true that college students feel that racial identity is ascribed to them by others. Collier and Thomas (1988) claimed that individuals have both avowed and ascribed identities. Individuals wish for others to accept their avowed identities because that is the identity such individuals believe to be a true reflection of themselves. However, at times, others ascribe identities to an individual that may reflect assumptions about that person's identity due to his or her gender, race, social class, etc. When identities are ascribed to individuals, such individuals may either attempt to resist such identities or instead, succumb to them. Chan (2017) has addressed this same phenomena through the lens of Symbolic Interactionism, explaining that individuals learn about the self through interactions with others. This implies that when an identity is ascribed to an individual due to his or her race by people on social networks, that individual is placed into the proverbial "box" of racial identity; therefore, subconsciously and consciously the individual may build his/her racial identity around what others say about him/her.
Results of this study showed that although they did not intentionally project their racial identity, non-Caucasian participants thought about their race significantly more than Caucasian participants. In fact, they thought about their race about half the time when posting, commenting, and interacting with race on social networks, whereas Caucasian participants reported thinking about their race almost never. This difference may indicate that non-Caucasians think about race more while on social media as a result of the identity ascribed to them due to race. These findings fit with past research, which reveals that non-Caucasians on social networks experienced apprehension when posting on these social networks (Chan, 2017). Chan believed that participants experienced apprehension because they feared being perceived as a representative for their whole racial group. Further research could help to explore these notions at a deeper level.
It is also interesting that Caucasian participants strongly disagreed with the notion that they project their racial identities on social media (through photos, text, or communication). It could be deduced that Caucasian participants, being a part of the dominant race in America, do not see their racial identity as an important one to project.
It may also be true that non-Caucasian racial identities and the subsequent experiences of marginalized individuals, are not valued in part because they contradict norms set by a Caucasian dominated culture. Therefore, Caucasian Americans are taught implicit norms surrounding the colorblind narrative-taught to devalue the expression of differences regarding race. Furthermore, it would be advantageous to Caucasians to disregard the importance of their racial identity for the continuation of the colorblind narrative, where the U.S. American public claims that they "don't see race" and therefore are not biased in their interactions with others of different races. For non-Caucasians this is society's way of devaluing the expression of racial identity. The result is a culture that does not give recognition or value to the racial identities of non-Caucasians. For example, the #BlackLivesMatter movement was met with #AllLivesMatter, not only by those who disagreed with the issue, but also by those who did not understand why the hashtag was not inclusive to all racial identities. U.S. American society failed to realize the reason for the expression of racial identity and the desire to draw attention to the experiences of people of color with law enforcement. In this way, race is not only a social construct, but also a political tool that can be used to marginalize opposing worldviews. The failure of Caucasians to view racial identity projection as important creates the norm that racial identity projection is abnormal, and the result is a climate that fails to distinguish and value non-Caucasian racial identities.
It is also possible that Caucasian participants feared that communicating about their racial identity might convey the impression that they align themselves with the values and beliefs of white supremacist groups. The tyranny of the Nazis' regime has given Anglo Saxon Whites a reason to fear the celebration of racial identity. Groups like the KKK use racial identity projection as a way to exclude non-Caucasians and as a catalyst of hatred toward others. Future research should be conducted to better understand why Caucasian participants do not project racial identity.
Regardless of the reason Caucasian participants do not display their own racial identity, it is clear that the expression of salient racial identities of non-Caucasians may be crucial in making progress towards breaking the colorblind narrative. If society can begin to value racial identity and understand the differences in how reality is experienced, progress may be made towards greater inclusivity.

Conclusion
This study examined the use of photos and textual communication, the concealment of racial identity, and interactions with race related content to assess how participants projected racial identity on Facebook, Instagram, and Twitter. Results of this study suggest that racial identity is not intentionally projected by the participants. However, non-Caucasian participants use photos, text, and interactions to convey racial identity significantly more than Caucasian participants. It may be the U.S. American sociopolitical context that leads non-Caucasian participants to be more aware of their race when commenting, posting, and interacting on social networks. In addition to these findings, results of this study revealed that participants do not attempt to hide or filter out their racial identities on social networks. Future research should try to understand why this is the case. Caution must be taken in generalizing the results. However, the findings help to better understand how college aged young adults convey racial identity on social networks. The study has helped to expand the understanding of identity projection and management on social media specifically relating to racial identity.

Researcher reflexivity and limitations
Survey methodology is a difficult tool to use when conducting research about racial identity. Racial identity is a very personal topic, and there is no uniform experience for it. The instruments of the study were created for the purpose of this research and require further reliability testing and validation. It is important that care be taken when generalizing the results of the study. The sample size of 347 participants should not be used as a representative sample of the population. In addition, there was an imbalance of Caucasian (62%) and non-Caucasian (38%) participants.
Identity is unique to individuals, and there are likely many factors beyond racial identity that impact the way respondents may answer. Furthermore, various intersections might prove to have an effect on responses in future research. For example, there may be differences in the way racial identity is projected between those who have the ability to attend college, and those who do not. In addition, compiling non-Caucasian participants into one group can be problematic since they do not necessarily have the same racial experiences. However, non-Caucasian groups might be united in the fact that they do not possess the societal privilege that the Caucasian racial group holds.
Participants were surveyed on their experiences with racial identity projection on three popular social networks (Facebook, Instagram, Twitter). This research is limited to the social networks of the time period. A shift in popular social networks could have an effect on racial identity projection.

Future research
Future research could take a number of different directions. For example, future research could explore the way people feel their racial identities are ascribed to them on social networks and seek to shed light on the limitations to displaying avowed racial identities. It would also be interesting to explore why Caucasians feel that identity projection is unimportant. Alternative forms of research (i.e., focus groups and interviews) may be most useful in such explorations. In addition, different forms of social media (i.e., LinkedIn) might offer new insights on racial identity projection.
It is important to continue to explore racial identity projection on social networks because it is an interesting frontier, with many avenues for future research. Furthermore, findings of continued studies may help researchers and practitioners to understand how others experience the world. Such information may aid educators and other experts in helping individuals to successfully navigate racial identity and discussions about race on social networks. Researchers must continue to find ways to better understand how race is perceived, negotiated, and conveyed on social media. The unwanted animosity created by volatile discussions of social and political debates pertaining to race may someday be remedied with improvements in education pertaining to the social construction of racial identity and the way it is projected, specifically on social networks.

Introduction
Over the last years, two reviews have tried to systematize research on international students: Smith and Khawaja (2011) have provided a review on acculturation models trying to determine the suitability of these models for research on international students. Zhang and Goodson (2011) have analyzed predictors of international students' psychosocial adjustment to life in the U.S.
With this article I propose a different kind of review on international students' research: Instead of listing research topics, theoretic categories and research results, I would like to draw attention to the approaches used to gain scientific insight on international students. With this change in analytic stance, I want to bring up the often implicit assumptions of the approaches that guide the research, and broaden the horizon to more research foci, hoping to encourage a discussion about these different perspectives and their scientific benefit.
For this purpose I will not try to give an exhaustive survey of all research to be found on this topic, but rather concentrate on a few representative studies pointing out the characteristics of their approach. Sometimes I will just choose one study for illustration purposes, sometimes a group of studies depending on how I can best demonstrate my findings in a concise form. I choose two dimensions to structure my analysis. Usually international students are seen as different and this difference somehow leads to a process of change. The first question is therefore the question of difference. What is considered to be the relevant category of difference that guides the research? The second question is what kind of change processes are being investigated?
The reported studies are very different in so far as theoretical and methodical concepts are concerned. I do not claim that they are alike. My focus is the similarity inmore or lessimplicit assumptions, which I will elaborate upon further below.

Perspectives on international students
Three main approaches can be distinguished: the adaptation approach, the biographical approach, and the motivation approach.

The adaptation approach
This approach is dominant in research. Its underlying foundation is the assumption that international students aredue to migrationconfronted with a new environment to which they have to adapt. This assumption is usually accompanied by concepts of "culture" often implicitly conceived as "national culture" (or "regional/ethnic culture"). The migration is therefore viewed as a change from one set of cultural norms to another. Within this approach five different conceptions can be differentiated.

Adaptation as a conflict of norms
The assumption of normative conflicts is very often to be found in studies of the years 1960-1980. The existence of cultural norms is usually considered to be a matter of fact. The term "culture" itself is not defined, but national norms are assumed to guide the behavior of international students. For this reason, these studies investigate the normative conflicts of students coming from one specific cultural region, like Arab students or Chinese students. Difference is located in the cultural origin of the students, be it a specific country, region or status like "developing countries". Arriving in Germany meansaccording to this approachthat the students are confronted with adaptation problems due to new cultural norms, and the degree to which students adapt to the new norms is explored. Adaptation as a change process is not investigated by a comparison of the sample prior and post migration, but as deviation from the a priori defined cultural norms of the home country.
I would like to illustrate this perspective with the study from Tjioe (1972). Her study is particularly interesting in her findings and suitable to show the advantages and shortcomings of this approach. She first reports cross-culture studies to define "East Asian norms" (1972:24ff.). Then she investigates which German norms female Asian students adapt to, and which they reject and demonstrates where the Asian students see differences between their home countries and German culture. It becomes apparent in the statements she quotes, however, that the Asian women do not judge, for example, the behavior of German women in the same way, instead they position themselves in favor of or against "German" behavior.
The focus on different cultural norms (understood as national/regional cultures) has the merit to bring up the differences that international students experience and which the host society often ignores. Tjioe (1972) shows how much the interaction between the German and Asian students is influenced by judgments on each other. It is important to be aware of these perceptions of differences and the difficulties they might entail. However, by concentrating on national/regional cultures the researchers often neglect the differences among the international students themselves, although these differences might be relevant in order to understand the adaptation process in the host society. This approach does not consider the normative conflicts within the families and the societies of origin. The individual experiences leading to certain judgments and positioning towards norms cannot be understood. This brings us to another important point: If the analysis aims at new findings about adaptation as a change process in international students, then there needs to be data on the perceptions and normative positioning of the interviewees before and after migration.

Adaptation as a personality trait
In contrast to the previous approach this one is focused on psychological factors. However, culture -related findings are often included in the concept. In general, the adaptation of international students is investigated through their psychological constitution while studying abroad. This means that a positive psychological constitution is considered to be a sign of a good adaptation. For this approach, it is of particular interest to find factors that predict positive adaptation. The term "culture" is often referred to in these studies without further definition, but as many studies focus on students from a specific country or region, it becomes clear, that the term "culture" is understood in national or regional terms. The study of Searle and Ward (1990) is a major work, which introduces the groundbreaking distinction between psychological and sociocultural adjustment. The first comprises psychological dimensions of well-being and satisfaction and the latter contains the sociocultural capacity to "fit in " (1990:450). On the basis of three theoretical frameworks -"clinical perspectives", which consider personality, life events, and social support; "social learning models"; and "social cognition approaches"the authors investigate "cross-cultural transitions" (1990:449) through a study on students from Malaysia and Singapore who lived between six months and six years in New Zealand. In their discussion (458ff.) the authors refer to the key constructs "cultural distance" and "cultural fit", which they suggest needs further exploration. The construct "cultural distance" is often referred to in the literature. It is important to understand this approach, and therefore I want to give a summary of the studies concerned.
The Cultural Distance Index (CDI) according to Babiker, Cox and Miller (1980) wants to describe "local custom and environment": that is, the specific social surrounding of the students, not national culture (1980:110). The ten items of the questionnaire ask for climate, clothing, language, educational background, food, family structure, etc. (114-115, Appendix A). Despite this aim, the CDI lists scores of national countries. It shows that, for example, Finland, Thailand and New Zealand score relatively close to Scotland, whereas Portugal, Kenya and Denmark lie in the middle (112, Table 4). The CDI only proves that African and Asian students score higher on "anxiety" and go to see the doctor more often than the others. In contrast to that, 'cultural distance' shows no relation to academic achievement, but it seems that "material comfort" is related to academic success (113). What insights does this questionnaire provide? The authors themselves critically reflect on their concepts and findings and point out the shortcomings, such as the fact that for many African countries there was just one respondent. Nevertheless, this study is often quoted to underline the validity of 'cultural distance' as a concept.
The study of Furnham and Bochner (1982) has a different focus. They don't want to pathologize the problems of international sojourners (1982:163ff.) and therefore choose the approach of social learning, arguing that the adequate behavior in a different culture has to be learned (164). They investigated difficulties in specific social situations and friendship networks on the grounds of a priori defined regions of origin (176-177). Respondents were participants in language courses who had been in England between one to five months (176). Results show, that the contact with the host country population appears to be quite difficult. The specific situation of participants in language courses, who just stay a short time in the country and for this reason, usually have little or no intensive contact with the host society, is not discussed. It remains an open question if the difficulties are due to a lack of "social skills" (165) or if the problem is rather the communication in a foreign language and the short length of stay. The approaches of the two studies are quite different. Nevertheless both are often referred to with the aim to underline the relevance of "cultural distance".
The concept of "cultural distance" is, as in the first approach "conflict of norms", an attempt to define collective practice and to compare one set of collective practices to another. Two main problems, which can also be seen in the above mentioned studies, arise: (1) What is part of the collective practice? …Material/physical aspects such as clothes, food, climate, housing? …Or immaterial aspects such as norms and habits like education background, family structure, social skills or language practices? (2) Which collectives do we compare? …Families? …Peer groups? …Local social structures? …Nations or regions?
Up to this point, the approaches 'conflict of norms' and 'personality traits' seem similar. The difference lies in the psychological view that the second approach introduces. Searle and Ward do not only refer to the hypothesis of "cultural distance" but moreover recommend considering the hypothesis of "cultural fit" (Searle & Ward 1990:458): that is, exploring the role that personality traits play. The hypothesis of cultural fit suggests that individual personality traits that resemble host country norms facilitate the adaptation process. This means that individual personality does vary within cultures and an individual positioning towards social norms has to be considered in the adaptation process.
The study of Searle and Ward is a good example of the "adaptation as personality trait" approach, as it discusses its advantages and its problems. This approach opens the perspective to the psychological factors which influence the reaction of the individual to a new environment. The quantitative studies allow for representative samples and complex data analyses, which include psychological and social information. However, the combination of individual psychological factors and social norms brings up questions as well: If "cultural distance" is a relevant concept, it needs a careful definition. Does it refer to collective norms on a national level or does it describe the immediate social surrounding of an individual? And in which way does it determine the individual's behavior? As the (individual) cultural fit is discussed as well, how does it relate to the concept of "cultural distance"?
Ward and Searle demonstrate that psychological and sociocultural adjustment are distinct forms of adjustment, though they are interrelated. But how are they interrelated? Triandis and Suh point (2002:136) out that studies yield different results depending on whether they define 'cultures' or individuals as relevant categories of analysis. Therefore it seems necessary to specify the relation between individuality and collective imprint.

Adaptation through friendships
Several studies are dedicated to investigating the friendships of international students, arguing that adaptation is facilitated through friendships. Change is therefore explored in the form of building up social contacts, and the category that is analyzed in these studies is the nationality of the friendship groups. For this reason, I don't want to summarize only one study as an illustration of this approach, but rather present the different views on friendship groups as characteristic of this approach.
Some studies focus on the difference category "host students" -"home students" (Budke 2003;Hendrickson et al. 2011;Hendrickson 2018;Glass & Westmont 2014), while others stress the importance of international friendship groups (Sobré-Denton 2011; Montgomery 2010). Whether friends from the host country or international friendship groups are more important is rather controversial. Home students would introduce the international students to the host culture and encourage cultural learning, whereas if the international friendship groups play a special role as social support group, they buffer the effects of being a foreigner. Some studies investigate whether the region of origin plays a role in participating in recreational activity and building up friendship networks (Glass et al. 2014;Rienties & Nolan 2014). As the lack of host country friendships tends to be interpreted as a lack of interest in the host culture, yet another group of researchers focuses on the attitudes of the home students (Williams & Johnson 2011;Dunne 2013). They underline that it is often the home students who lack interest in building up friendships with students from abroad. Interestingly, there are studies of international students themselves, who integrate questions concerning friendships in their research. These authors only consider friendships with co-nationals, though reflecting the ambivalence of these relationships. Friendships with students from the host country are given no special importance (Yu 2005;Zhou 2010).
This approach brings up the key factor of social integration: Without relevant social contacts there can be no exchange between the international students and the host society. Friendships as emotional and social factors are an important part in the adaptation process. However, friendship is a vague category not easy to define. The need for social contact and the quality of contact needed for well-being vary individually. A focus solely on nationality does not capture the different types and qualities of the relationships that constitute their emotional meaning. For the development of emotionally satisfying relationships it is important how the international and host country students establish community, that means according to which criteria and needs do they connect to others and build up friendships? The answers to this question could provide new insights on the role of friendships in adaptation processes.

Adaptation as orientation task of higher education institutions
A different stance is taken in the studies of Hosseinizadeh (2005) and Klabunde (2014). Their focus does not lie on the adaptation of international students, but on the challenges for the higher education institutions to provide adequate counseling and support structures. Within the frame of competition for the best international students, it is not the international students anymore whose adaptation processes are investigated but instead the ability of the institutions to adapt to the needs of the international clientele. This is done in both cases by a comparison between German, U.S. and Canadian universities, respectively, and the facilities provided. Teaching and administration staff as well as students are interviewed and programmatic university documents are analyzed.
This approach has the merit to include the higher education institutions as well as their political environment in the research and to point out their role in facilitating the adaptation process. This perspective gives information on institutional structures and practical problems. It investigates the perspective of those people who counsel and advise international students. It would be of particular interest though, not only to compare programmatic proclamations but to evaluate their outcomes.

Adaptation as a plurality-orientation of higher education institutions
Under this headline I refer to various different studies, which all have the same approach of viewing the higher education institutions as the responsible actor to adapt to the plurality within its student and teaching staff body. Sun (2010) investigates the intercultural differences perceived by Chinese students in Germany. Schuman (2012) collected "critical incidents" about intercultural misunderstandings. These studies are concerned with cultural norms like the first approach, but investigate the academic setting, not general social norms. Otten (2006) works out a typology of interculturality orientations of teaching staff, wanting to bring about institutional change. On the basis of the difficulties of Moroccan students in Germany, Kiefer (2014) asks for a difference-sensitive university, where dominance structures are reflected. The Diversity Report (Berthold & Leichsenring 2012) of the CHE Consult explores sociologic factors that can prevent academic dropout.
All of these studies are concerned with the plurality found in international universities and the concepts of the higher education institutions to deal with it. However, and this is the important point concerning this approach, plurality is analyzed on different levels. Whereas Sun (2010) and  are concerned with perceived differences of social practice, Otten (2006) investigates the discursive constructions of difference. Kiefer (2014) and the Diversity Report (Berthold & Leichsenring 2012) observe sociologic factors that constitute difference. All of the factors merit attention as they influence social reality. Particularly the correlations between perceptions, discourses and sociologic factors would provide valuable insights.

The biographical approach
Günther (2009), Yu (2005) and Aits (2008) collected data with qualitative interviews that allow comprehending the biographical context of student experiences and behavior. Change processes can be understood through biographical developments. Rather than the difference of the students being investigated, it is the way in which the individuals themselves deal with perceived or ascribed differences.
The study of Günther (2009) is particularly rich in findings, as she describes personal developments of Guinean students in Germany. She is interested in negotiation processes between social groups and the individual, as well as the margins that the adolescent individuals have to realize lifestyles diverging from their social group (2009:94-95). She identifies three different development patterns (214ff.), depending on the family relations back home. Students whose parents were empathic conflict partners manage to design their own life in Germany and deal positively with experiences of difference or even discrimination. Students, who migrated to fulfill the parents' ambitions and had no margin to try out their own lifestyle, lose their orientation in Germany, and tend to withdraw into their own ethnic community. Students of the third type live in family relationships full of conflict and are therefore still busy with these relationships while studying abroad. Günther concludes that the migration leads to an extension of the psychosocial possibilities of the students (Günther 2009:243), but that the coping with the adolescent migration situation is primarily determined by the socio-emotional resources anchored in their family experiences (245-246) as well as in the host society. She points out that the Guinean students suffer from discrimination in Germany, which limits their development potential. Therefore she sees the need to revise approaches which focus on the cultural adaptation of international students, as it is not cultural practice that determines the margin of adaptation (246).
Günther's findings do not only contradict approaches targeted towards (national) cultural traits, butand this is true for the biographical approach in generalthey allow scientific insights of a different quality: Personal traits, behavior and motivation can be analyzed and explained within the biographical context and individual positions to collective norms become visible and comprehensible. The biographical approach makes it possible to analyze the social context in which individual life styles and action strategies evolve. The subjective meaning that the interviewees give to their social world in a certain context becomes apparent and the factors that influence the individuals' development can be identified.

The motivation approach
The motivation approach examines expected difference as an impulse to study abroad and to change life perspectives in some way or other. Some studies mention motivation as one factor among others, without specifying motivation as a concept (Berthold & Leichsenring 2012:27;Heublein et al. 2004:107). The German Social Survey by the National Association for Student Affairs regularly asks the international students for their reasons to study in Germany. Among the most favored reasons are better career options, the reputation of the German degree and good study conditions (Apolinarski & Poskowsky 2013:45-46). But, as we have just seen in the study of Günther (2009), the reasons for studying abroad are often anchored in biographical context. Alfred Schuetz' fundamental distinction between "in-order-to motives" like future oriented goals and "(genuine) because-motives" like biographical reasons (Schuetz 1953:16f.) adds the biographical dimension as motivational category. I conducted a study which takes up this differentiation and develop, on the grounds of the documentary method (Bohnsack et al. 2010), a typology of study abroad motivation that points out the relation between experiences made within family and society and the goals linked to the study abroad project (Loy 2018a). Other studies work with different concepts: Gong (2003) examines goal orientations based on Dweck (1986, Dweck & Leggett 1988, whereas Boneva andFrieze (et al. 1998) andFrieze andBoneva (et al. 2004) focus on the three motives "achievement", "power" and "affiliation", based on McClellands' (1975McClellands' ( , 1976McClellands' ( , 1984McClellands' ( , 1985 motivational theory. Chirkov (et al. 2007Chirkov (et al. , 2008 investigate the role of self-determined motivation (Ryan & Deci 2000;Deci & Ryan 2000) and values (Tartakovsky & Schwartz 2001) as an important aspect in the adaptation process of international students. While the migration goals appear to be similar in most of these studies, the social genesis of the goals and therefore their social meaning in biographical context seems to be the relevant category that needs more scientific attention (a more detailed analysis is offered in Loy 2018b).
The motivation approach gives research on international students a new direction: The scientific attention is drawn to what drives the students abroad. Hence, it investigates the preliminary conditions of a later adaptation process. The challenge remains to find out if, and how, motivation is related to adaptation processes and which theoretic categories can best be used to describe the motivational structures.

Summary and discussion
We have seen that the focus on normative conflicts reveals the confusions international students experience in their social interactions abroad. The restraint on culture conceived in geographic boundaries, however, ignores the fact that students do position themselves individually towards the norms of their home country as well as towards those of the host country. Furthermore it leaves out other difference categories such as social and educational background, gender, age, etc. The same is true for the personality approach. It has the merit to explore the role of individual personality traits in the adaptation process, but it often perceives personality determined by cultural norms similar to the first approach, without specifying the relation between individual and culture. Yet, it would be of particular interest to describe more precisely the social embeddedness of the individual: In which way is the individual determined by collective norms? What are the collectivities that influence the individual? Which effects have different collective norms on social interaction? The friendship approach takes a closer look on social interaction investigating social relations as one specific factor in the adaptation process. However, the focus on nationality groups limits the analysis to the specific role each (national) friendship group plays in the adaptation process. This might be an important factor, but leaves out the more central question of social relations: How is social relatedness constructed? What different ways are there to build up community which provides social and emotional support? All three of these approaches assume that it is the international student who should adapt to the host society. The orientation approach and the plurality approach make the important shift towards considering the role of the higher education institutions. The 'adaptation as orientation task of the higher education institutions' approach offers a descriptive comparison of academic counseling and support structures for international students. The effectiveness of the programmatic statements, though, is only explored in form of the individual experiences and convictions of the expert interviewees. The 'adaptation as plurality-orientation of the higher education institutions' points more generally to the way the higher education institutions deal with the differences to be found within its student and teaching staff body. Here it is important to be aware of the analytic levels where differences can be found: the level of perception, the level of discursive constructions or the level of sociologic factors that constitute difference. A completely new perspective is adopted in the biographical approach. Here, the change processes in the biography of the students are explored. Influences of the social environment the students experienced in their home country but also in the host country become apparent and can be described, this way offering more precise information on collective norms than the category of 'culture' conceived in geographic boundary does. The motivation approach identifies the goals pursued with studying abroad, butif this focus is included in the analysisit can also reveal the biographical reasons that give the impulse and drive for migration.
It seems that the biographical approach in particular provides new insights on international students' development, as it combines past and present experiences of the students and includes most aspects of the other approaches. Individual personality traits can, in this way, be analyzed together with their social embeddedness, which allows identifying the relevant collectives that form the student's behavior, but also the processes of individualization and emancipation from normative constraints. Sociologic factors of difference can be scientifically captured and motivation factors within the social environment investigated, thus making it possible to relate different motivations to adaptation processes in the host country. Social relations in the home and host country can be explored concerning their emotional quality and areas of conflict, helping to explain interaction between host and home students. This interaction could be included in research and evaluated concerning the same criteria. This approach makes it possible to conceive of student adaptation in the host country in a more general, biographical change process, for which study abroad plays an important role.
About the author Dr. Christine Loy has worked in the international office of the Otto-von-Guericke-University Magdeburg, setting up a tutorial program for international medical students and teaching intercultural communication, before she pursued her PhD degree on the motivation of international students to study abroad. During this time she worked as scientific employee at the Helmut-Schmidt-University (Hamburg). For the purpose of her doctorate she carried out qualitative interviews in Lebanon and Jordan. At the moment, she is teaching a refugee class to prepare the students for vocational education in Germany.

Introduction
In this article the interaction between intercultural competence and L2 acquisition is investigated in the context of study abroad (SA) experience. If the study of second language acquisition (SLA) is to be understood as a scientific analysis of how foreign languages are acquired in different contexts (ranging from formal in-classroom instruction to informal outof-classroom settings), then SA may be perceived as a context for L2 acquisition which combines elements of both formal instruction and immersion. The challenge for SLA in investigating SA is to understand the relationship between how learners process L2 and the sociolinguistic and sociocultural factors that influence this processing. Research results accrued so far (Baker-Smemoe et al 2014, Marijuan & Sanz 2018) suggest that intercultural competence is the strongest predictor of L2 gains during SA, but as will be evidenced in the present article, more variables play a part.
Studying abroad (SA) is a context in which learners enrol on an exchange programme in order to live and study in a host country (Pérez-Vidal 2014a: 20). SA has not been so far defined in terms of the length of the exchange but most studies encompass a length of stay between 3 and 12 months. A common belief is that SA provides unparalleled opportunity of immersion in a target culture and a target language. It is generally accepted that learners return from their SA programmes with improved linguistic abilities, a greater intercultural sensitivity, and a stronger motivation to learn languages. Indeed, some areas of their L2 communicative competence improve and many learners report increases in their motivation to learn languages and cultures, however, as we will discuss further in this paper, the picture of an average sojourner is not as clear as might be expected and further SLA research is needed to help us better understand the relationship between SA and L2 acquisition. SA has thus far been investigated by scholars interested in foreign language education with a view to building a theory of SLA in immersion contexts. One of the first attempts to understand how SA can be framed within SLA research enterprise was Freed's (1995) volume devoted to linguistic and sociolinguistic aspects of SA in European and North American settings. Its contribution was in presenting how the benefits of SA differ from that of formal instruction contexts. Collentine's (2009) research summary published nearly two decades later identified three main themes running through analyses at the junction of SA and SLA: a) cognitive processes connected with L2 acquisition in SA contexts, b) sociolinguistic processes connected with input and interaction while abroad, and c) sociocultural aspects of language learning which witnessed a shift from language-centred to learner-centred perspectives. Consequently, two important volumes on the relationship between SA and SLA by Kinginger (2014) and Pérez-Vidal (2014b) testify to the diversity of SA and open new avenues of inquiry. Among others, they present how non-linguistic individual differences such as foreign language anxiety (FLA) or willingness to communicate (WTC), may in fact influence success in SA-based L2 acquisition.
The aim of the present article is, on the one hand, to demonstrate SA as an SLA research context and, on the other, to discuss intercultural competence as an important research variable to be taken into account in designing relevant studies. This will be done by outlining research areas, tools, and results reported by scholars interested in the influence of SA on L2 acquisition. To begin with, the article looks at terminological considerations and reviews current literature of the field.

Terminological considerations
Even a cursory literature review demonstrates many discrepancies in the terminology used by scholars working in the field of SA and SLA. In this section, the terminology used in the present paper will be unified and clarified.

Study abroad (SA)
SA is known in the relevant literature also as "stay" or "residence" abroad, "in-country study", "overseas language immersion", "academic migration" (Coleman 1997), "student mobility", or even an "L2 sojourn" (Jackson 2016). It refers, very broadly, to various types of study at foreign institutions. The primary division is into whole-programme and withinprogramme mobility (Coleman 2013: 21), the former denoting studying abroad in a full academic cycle for a degree, the latter denoting educational mobility for obtaining credits.
The understanding of SA depends on the context of the research and professional interest of researchers. Although Kinginger (2009: 29) refers to it as a sub-field of applied linguistics, SA attracts not only the attention of linguists, but also scholars concerned with educational policies, economics, psychology, social identity, etc. (Dervin 2011). Most definitions of SA, however, emphasise the educational context of the phenomenon, pointing to it being a component of a university programme (Coleman 1997), undertaken for educational purposes (Kinginger 2009) in a hybrid communicative-learning context (Collentine and Freed 2004). This lack of unanimity across scholars leads to difficulties in comparing relevant research and makes drawing generalisations difficult. Nonetheless, in the present paper, SA in conceptualised as the experience of crossing borders in order to live and learn in a foreign country.

Formal instruction (FI)
SA is very often contrasted with formal instruction (FI). In fact, some researchers (e.g., Pérez-Vidal 2014a) present the two in sharp contrast as ends of a continuum. SA is seen as a naturalistic context allowing for nearly total immersion in the target language and the target culture that is characterised by massive opportunities for sociolinguistically varied L2 input and interaction. FI, on the other hand, is understood as conventional L2 classroom context, i.e. a place in which learners' attention is drawn to language form and meaning, but which gives nearly no opportunities for out-of-classroom L2 interaction. Yet, as research findings unfold, a less black-and-white picture of SA/FI contrasts begins to emerge. For one thing, learners make a lot of use of their previous FI in SA contexts. Second, they do not always benefit from the linguistically rich context of SA accessible to them, since some of them do not look for L2 interaction opportunities available to them, preferring to stick to their native-speaking sojourners (Devlin 2014). Moreover, SA in accordance with its name entails FI especially in countries where the language of the instruction is the language of the target culture (i.e., when sojourners participate in classes conducted in the language of the target country, unlike the situation when e.g. Polish students of English philology travel to a non-English speaking country but take their classes in English).

Intercultural competence
Intercultural competence is both a very broad term and one for which many definitions have been put forward. According to Bennett and Allen (2003: 237), intercultural competence means the ability to 'transcend ethnocentrism, appreciate other cultures and generate appropriate behaviour in one or more different cultures. ' Wilczyńska (2005: 22) suggests that intercultural competence should consist of three elements, namely: (1) the general knowledge of culture as such together with its mechanisms (2) intercultural sensitivity which allows for appropriate interpretation of cultural behaviours, and (3) cultural mediating, i.e. the skill of overcoming communication boundaries.
In short, then, intercultural competence denotes a range of skills that allow successful interaction with a foreign culture. The success of such interaction can be measured by one's ability to establish bonds with people from another culture.

Second language (L2)
L2 will be understood here as the language which is not a person's native language and which is used outside their country. L2 will therefore be used henceforth to denote a foreign language, in particular the language of SA participants' target country, so the language which is acquired/learned during the stay abroad. Unless otherwise stated, the term L2 encompasses here also L3, L4, etc.: i.e., non-native languages acquired by a person who already possesses communicative competence in one or more foreign languages (see Cook 1995). This is, as very often happens in European contexts, because the language of the target country is not the language of the instruction, for example, when students from Romania study English philology in Poland. The situation creates both difficulties and opportunities. For one thing, it becomes difficult to communicate in and live surrounded by Polish, but at the same time it creates opportunities to learn more than one L2.

SA and intercultural competence
Intercultural competence as such, has been repeatedly reported to significantly improve the outcomes of L2 acquisition as intercultural contacts positively impact on L2 learning motivation (Byram 1997, Corbett 2003, Dörnyei & Csizér 2005. Research into the links between intercultural competence and SA has shown that cultural differences, or rather the inability to overcome them, may severely impede the success of one's stay abroad as a result of losing opportunities for interaction with native speakers (e.g., Allen & Herron 2003, Block 2007. A lack of intercultural competence will result therefore in lesser motivation to learn L2 (e.g., Twombly 1995, Wilkinson 1998, Isabelli-García 2006) and more difficulties integrating into the target culture (Baker-Smemoe et al 2014). On the other hand, openness to new cultures and willingness to communicate with them has been proven to be of great value to L2 acquisition leading to significant L2 gains during SA (e.g., , Isabelli-García 2006, Martinsen 2010. We now have hard data from a wide range of studies that when SA participants experience inferiority or anxiety towards the target culture it impedes their interactions with native speakers. For instance, in an early study Twombly (1995) highlighted how making friends in a foreign culture was difficult for the sojourners, and that SA students did not really know how to do it. In the new cultural context of Spain, SA students from America lacked common interests with their classmates, observed a different social structure, and different attitudes toward leisure time, and as they found more dissonance than expected, they reverted to their country-mates and kept in closed groups, sometimes even deciding to leave the country. Wilkinson (1998) also reported how the experience of American students studying in France had not been a short cut to linguistic proficiency, but turned out to be a source of a lot of frustration. Suddenly, English was not perceived as a mode for communication, and the sojourners felt reduced to (1) a minority group and (2) membership in a French-for-foreigners class. As a result, instead of making the most of language learning opportunities, the SA participants in both studies felt troubled and confused, had difficulty integrating into the target culture and thus fewer L2 interaction opportunities.
Likewise, Allen and Herron (2003), who investigated the development of oral and listening skills of American students travelling to France, as well as their integrative motivation and language anxiety, reported that possessing higher levels of intercultural competence helps the participants to make fuller use of the SA experience. The participants faced two main sources of anxiety while abroad: linguistic insecurity and cultural differences. This anxiety increased while interacting with native speakers leading to fewer L2 gains. However, as time passed, the participants demonstrated a significant improvement in their linguistic skills and a decrease in language anxiety. Their integrative motivation remained unchanged. This suggests, that more emphasis should be put during the pre-departure phase on non-linguistic factors, such as intercultural competence, that would help reduce foreign language anxiety. Of note, intercultural sensitivity and cultural adaptation seem to develop more significantly during the second half of the one-year stay, i.e. the longer the stay the more chances for increased intercultural competence (Engle & Engle 2004, Medina-Lopez-Portillo 2004, although even a short-length stay allows almost all students to develop their intercultural sensitivity and to have a better understanding and acceptance of cultural differences (Jackson 2009).
What seems constant in these studies is the fact that although SA participants have the options to establish relationships with and learn language from the local people, they often prefer to spend time with their compatriots. Block (2007: 871) observed that many SA students finish their stay with the realisation that they could never be taken seriously as speakers of the target language. At the same time, those students who do manage to establish successful relationships with the locals and become members of these communities of practice are more likely to move beyond their ethnocentrism toward intercultural sensitivity and ethno-relativism. Isabelli- García (2006) rightly points out that the sheer fact of being surrounded by the target language does not guarantee linguistic development, but that informal relationships contracted by individual learner play a vital role.
Research has repeatedly highlighted the fact that successful immersion in the target culture is of great value to L2 acquisition, particularly with regard to the development of oral proficiency (Collentine 2004, Isabelli-García 2006, Martinsen 2010. For example, Baker-Smemoe (et al 2014), having conducted a large scale study of more than 100 native English speakers who took part in SA programmes in China, Egypt, France, Mexico, Russia, and Spain found that the strongest predictors of L2 gains were cultural sensitivity and social network variables. Those students, who scored high on predeparture Intercultural Development Inventory (Hammer, Bennett & Wiseman 2003): that is, who displayed higher intercultural sensitivity, were the ones who reported greater L2 gains. These findings corroborate with earlier studies (e.g., , Martinsen 2010 which also uncovered the relationship between intercultural competence and SA L2 acquisition success. Even short-term programmes can benefit linguistic development if students are able to deal effectively with the target culture (or other cultures in general). Martinsen (2010) showed how only pre-departure levels of intercultural competence predicted students' L2 gains and Martinsen (et al 2010) confirmed that there is a positive relationship between speaking time with the locals and linguistic gains during shortterm immersion programmes.
In consequence of these observations, many American and European universities offer pre-departure orientation sessions to prepare learners for their intercultural experiences. These sessions for students accepted into SA programmes take different forms in different institutions. They may include:  using (online) digital resources such as videos, podcasts, infographics or presentations which students later discuss with their mentors or complete questionnaires about them  individual or group sessions preparing students for the experience of culture shock, informing them on the strategies for maximising the potential of their stay abroad, or simply discussing with them how to maintain good health  country-specific orientation providing succinct information about their respective destination and the academic system  reading orientation handbooks that provide information on living and travelling abroad safely, experiencing culture shock.
Participants are sometimes asked to prepare "home ethnography" projects to hone their ethnographic skills that might help them get through the initial stages of denial to acceptance of the foreign culture (e.g., Jackson 2006). Pre-departure intercultural training may also include ethnographic projects, experiential learning, exploiting cultural texts, and comparative approaches (Róg 2014). The effects of pre-orientation sessions are usually measured with Intercultural Development Inventory (Hammer, Bennett & Wiseman 2003), showing how participants' inflated opinions of their own intercultural competence and readiness to enter a foreign culture diminish as the sessions progress (Jackson 2009(Jackson , 2018Vande Berg, Paige & Lou 2012).

Other variables influencing L2 gains during SA
The following overview addresses variables that are related to L2 gains during SA. These include language-related variables such as L2 threshold level, the amount of L2 input, and interaction opportunities and 6 non-linguistic variables such as age, aptitude, motivation, willingness to communicate, and personality.

SA and linguistic variables
A substantial body of SLA research has been gathered on the effects of linguistic variables on L2 acquisition during SA (Brecht, Davidson & Ginsberg 1995, Lafford & Collentine 2006, Kinginger 2013, Jackson 2016, Marijuan & Sanz 2018. Researchers have primarily focused on the effects of SA on the development of oral proficiency showing how SA participants' L2 speaking skills developed significantly after the sojourn (e.g., Allen & Herron 2003, Collentine 2004, Martinsen 2010, Mora & Valls-Ferrer 2012, which reflects the fact that oral proficiency is seen as the most desirable outcome of L2 acquisition and, as DeKeyser (2007: 208) observes, the main goal of study abroad. SA has been found to push the development of fluency and discursive abilities (Collentine 2009). For example, SA participants' speech rate and mean length of turn seem to improve significantly (Towell, Hawkins & Bazergui 1996), and so do their listening (Kinginger 2009) and reading abilities (Dewey 2004). Sociolinguistic aspects of language use and learners' pragmatic abilities have also been shown to improve during SA (Perez-Vidal 2016). Among the most-researched linguistic variables are the relationships between initial L2 proficiency and its influence on potential L2 gains during SA, as well as the problem of measuring input, output, and interaction. These two areas are discussed below.

SA and L2 threshold level
Numerous studies have explored the interaction between SA participants' initial L2 proficiency and linguistic gains during their stay abroad. It seems that pre-departure L2 proficiency plays a significant part in whether SA students achieve L2 gains or not. The threshold level for grammatical competence of SA participants was first suggested by Brecht, Davidson, and Ginsberg (1995) who found that grammar and reading scores were best predictors of the acquisition of Russian as a foreign language during SA. Intermediate learners may therefore need a well-developed lexical and grammatical base but also a better working memory capacity so that more input can become intake (Lafford & Collentine 2006: 117). Otherwise, they tend to focus more on the meaning and ignore the form which results in neglecting many L2 grammatical markers. As regards more advanced learners taking part in SA programmes, they may have more cognitive resources and better lexical and grammatical competence to focus on and acquire the L2. Segalowitz and Freed (2004) investigated oral production skills and cognitive abilities of forty university students, 22 of whom received FI in Spanish as a second language, and 18 studied in Spain for a semester. The study looked at changes to their speed and automaticity of lexical access and speed and efficiency of attention control. The SA participants displayed gains in oral fluency, however, it could not have been ruled out that out-of-classroom context was the sole factor responsible for this. The SA participants received formal instruction while abroad and the improvement was hypothesised to be attributed to students' learning abilities and the higher than the at-home group's number of instruction hours. Segalowitz and Freed (2004) concluded that a certain threshold level of L2 lexical access and word recognition is needed for SA participants to benefit the most from the experience. Earlier, Payne and Whitney (2002) hypothesised that the capacity to retain longer segments of input and hold longer strings of language for output may also be key to L2 acquisition. Smemoe-Baker (et al 2014) having reviewed a number of studies hypothesises that, while a certain proficiency threshold is needed, it should be low enough for short-term stays abroad to be measurable. Learners beginning their SA programmes with advanced levels of oral proficiency (measured by OPI, discussed below) appear to benefit less than those starting at intermediate levels. Likewise, Muñoz (2010) found that less proficient siblings improved significantly more than their older and more proficient siblings during year's stay abroad with a family.

SA and input, output, and interaction
Despite the fact that some learners (particularly those at intermediate levels at the programme onset) manage to achieve significant linguistic gains while abroad, some research highlights how SA does not always lead to greater gains than FI (DeKeyser 2007, Collentine 2009, Sanz 2014. The picture remains however unclear as it is extremely difficult to capture the amount, the nature, and the variation in input received by SA participants. As Pérez-Vidal (2017) observes, it is almost impossible to measure the quality and quantity of language interactions that SA students engage in while abroad with instruments other than self-reports which are more often than not unreliable.
One attempt to record L2 contact was  Language Contact Profile (LCP) questionnaire which evolved over the years and became a widely used instrument in various studies and in various contexts (also in academic classrooms and intensive domestic immersion). The pre-test includes questions about language learning experience prior to the departure, while the post-test version requires of learners to assess, among others, the amount of time devoted to learning a foreign language, the number of hours devoted to filling in target language forms, having conversations, etc. The LCP is a fine attempt to record opportunities for input and interaction, yet, as Pérez-Vidal (2017: 649) observes, it "does not lend itself to providing an exact quantification of time on task with target language".
A different research tool is one for measuring ethnic identity (Laroche, Pons & Richard 2009). In one of the parts of the questionnaire the authors ask learners to assess the degree to which they used their L1 and L2. These measures are operationalized as the percentage of usage of both languages for media (e.g., radio, TV, and newspapers) and for interactions with other people. Using the constant sum scale procedure, the authors ask the participants to distribute 100 points between L1 and L2 for each activity.
A yet different approach is the adoption of contact diaries (e.g., Coleman 2015). The input and interactions are tracked by asking learners to meticulously document the frequency of exchanges with the target language speakers, later these reports are visualised through plot diagrams.
These tools, however, still rely on self-reports. Even if they tell us something about the quantity of input and interactions, they say very little about their quality. Put simply, the greatest challenge for researchers is monitoring all the possible exchanges while abroad. The alternative, as Dervin (2014: 25) suggests, might be to connect the participants to microphones that would record their exchanges, but the idea is more than controversial and it is doubtful whether anybody would agree to take part in such research.
The situation is further complicated by the fact, that in most cases, SA programmes such Erasmus+ in Europe or Community Development, Language and Culture in America offer their participants courses in countries, where the academic programme is conducted in a language (usually English) different to the official language of the country. For instance, when Polish students travel to a Spanish-speaking country they take their course in English and communicate with other students in English, yet the linguistic landscape (language displayed on public and commercial signs) and the language used outside their educational institution that surrounds them is Spanish. This creates additional research challenge as it is extremely difficult to disentangle the amount of exposure in a given language that each student receives.
Yet the input and interaction opportunities in both the academic and the official language are much greater than those for at-home students. To simplify the matter, while the at-home students may take their whole course in the target language (similarly to foreign language philology students), they are not pressed to communicate with other students in the target language outside their classrooms. SA students, on the other hand, unless they speak the official language of the country, must still resort to the target language if they want to interact with their hosts. At the same time, they find themselves immersed in the naturalistic context of the official language and the culture of their target country.
Yet, surprisingly, recent research shows that many SA students do not avail themselves of such interaction opportunities (Devlin 2014, Pellegrino 2005, Jackson 2008, Mitchel et al 2015. It appears to run counter to the whole idea of studying abroad, but in fact they avoid interacting with their hosts and resort to sticking to their country mates or other international students as it easier for them to get together with them than with the locals. Because of this, SA programmes sometimes come under criticism or at least scepticism from more sociallyoriented researchers (e.g., Block 2003) who point out to cases where the benefits of SA are lost on students and the whole experiences becomes nothing more than just a vacation.

SA and non-linguistic variables
SA and SLA research has shown that the non-linguistic variables that SA participants bring with themselves to the experience may greatly influence L2 acquisition. Understanding them may help make sense of the variation in proficiency outcomes. Research results accrued thus far suggests that intercultural competence is, apart from social networks variables, the strongest predictor of L2 gains yet other variables are at stake.

Age
Although the relationship between L2 acquisition and age has a long research tradition, the impact of age on L2 acquisition during SA has been investigated by only a handful of studies. Most research so far has focused on older learners, which naturally reflects the trend in the available SA programmes in that there are more adult than young sojourners. The known research compared differences between children and adults (Llanes & Muñoz 2013) and between older and younger children (Muñoz 2010). The L2 gains have been assessed by measuring oral and written fluency, lexical and syntactic complexity, and accuracy. These studies have shown that younger learners are at an advantage when it comes to naturalistic contexts. SA proved more beneficial for children ages 10-11 than for adults aged 19-31 (Llanes & Muñoz 2013). Also, younger children were at an advantage over older children displaying greater L2 gains (Muñoz 2010). Llanes (2012), who compared an at-home group of young learners to their SA peers during their 2-month stay, reported that SA benefited the 11year-old children in terms of fluency, accuracy, and complexity.

Aptitude
The construct of aptitude includes a number of components such as phonetic coding ability, grammatical sensitivity, and memory capacity. The scarce studies regarding the role of aptitude in the context of SA seem to confirm our general knowledge of the role of aptitude in L2 acquisition in that the speed of lexical access and memory control are strongly related to L2 gains. For instance, as regards working memory, Sunderman and Kroll (2009) asked SA participants to complete a translation recognition task and a picture-naming task in order to investigate lexical comprehension and production for learners with and without SA experience. It turns out that a certain level of working memory is needed to reach L2 accuracy in the context of SA (Sunderman & Kroll 2009). Phonological memory appears to be a predictor of oral gains while abroad (O'Brien, Segalowitz, Freed & Collentine 2007) while speed of lexical access and its efficiency (or word recognition and automaticity) have been reported to be related to oral proficiency and fluency . Taguchi (2008) also observed that the speed of lexical access and contact hours were correlated with comprehension.

Motivation
Research on motivation and attitudes during SA revealed generally that learners' willingness to stay and/or live in a foreign country increased after their stay abroad, whereas linguistic self-confidence and the effort to learn languages was higher in FI learners who stayed at home (Pérez-Vidal 2017), although individual case studies of former SA participants' growing motivation to continue learning L2 have also been reported (Róg 2017a(Róg , 2017b. SA participants also display a greater motivation to travel, to meet new cultures, and expect better career prospects (Allen 2013, Juan-Garau & Trenchs-Parera 2014. Also, L2 anxiety has been reported to decrease significantly after SA experience (Allen & Herron 2003;Michiko & Takeuchi 2015).

Willingness to communicate
Willingness to communicate (WTC) is understood as a learner's drive to seek opportunities to interact in the L2. WTC is dynamic and context-dependent. In SA contexts, WTC has been initially shown to be positively influenced by the experience of studying in a foreign country (Dewaele & Wei, 2013) but most studies relate it to the sense of equality/ inferiority with the target culture (Du 2015, Furuta 2015, Róg 2017a. For example, Mystkowska-Wiertelak and Pietrzykowska (2011) report on a survey conducted among Polish students of English philology concerning WTC with other cultures. Apparently, despite Polish students' positive disposition towards other cultures and their aspirations to travel and study abroad, that they do not necessarily wish to become involved in intercultural relations (2011: 130). Polish students proved to be rather unwilling to initiate conversations in a foreign language (e.g., while standing in a queue) as they felt inferior to other cultures, however, the authors (2011: 130-131) point to the fact that this state of affairs could also be true for their mother tongue.

Personality
Many SLA studies emphasised that aspects of a learner's personality (e.g., openness to experience, extroversion/ introversion, or risk-taking) may influence how they acquire L2. In SA contexts, unsurprisingly, having an open personality and willingness to engage in new experiences were found to be prerequisite for L2 acquisition. It has been demonstrated that openness and extroversion were strong predictors of heightened L2 use and that openness was particularly predictive of linguistic proficiency (Ożańska-Ponikwia & Dewaele 2012). Also, while some learners display lesser extroversion and lesser openness, certain SA programme designs may in fact push them to more L2 use and thus greater L2 gains (Dewey 2012, Dewey et al 2014. Some studies found no relationship between personality and L2 success (Baker-Smemoe et al 2014) concluding that perhaps learners with different personality traits employ different learning strategies.

Data collection tools
Traditionally, SA research tools encompass questionnaires and interviews as primary data collection sources. More currently, research landscape changes as scholars make good use of digital technology, such as recordings, (audio and video files), using interview transcripts (coded with e.g. NVivo software), e-surveys (e.g., Google Form or Qualtrics), narratives (SA blogs, online diaries, essays), illustrations provided by SA participants (digital drawings, maps, mind-maps, sociograms) (Jackson, Chan & Tongle 2018). SA students may be contacted by researchers using computer-mediated communication (e.g., Facebook, Skype, or Whatsapp) (Back 2013).
As regards intercultural competence, its level is often measured with the use of either The Development Model of Intercultural Sensitivity (Bennet 1986) or its evolved version the Intercultural Development Inventory (Hammer, Bennett & Wiseman 2003). L2 proficiency is usually measured with the use of oral proficiency interviews. These three tools are described below.

The Developmental Model of Intercultural Sensitivity
The Developmental Model of Intercultural Sensitivity has been developed by Bennett since 1986. Its aim is to provide a framework for explaining how people embrace intercultural differences. A participant's current state of intercultural sensitivity is placed on a continuum of increasing intercultural competence. With increasing understanding and acceptance of cultural differences individuals move from ethnocentrism (understanding of one's culture as being central to reality) towards ethnorelativism (understanding of one's culture as one of the elements organising reality). In this way, they can reflect on foreign values, beliefs and identities, and reshape one's own and integrate new perspectives into their identities in order to become intercultural mediators.

Denial Defence
Minimisation Acceptance Adaptation Integration The progression on this continuum may not be linear and participants may revert back to previous stages. This may follow unsatisfactory intercultural experiences or a prolonged time without intercultural contacts. The DMIS has been found useful by researchers working in the field of SA (e.g., Anderson et al 2006, Jackson 2008.

Intercultural Development Inventory
Numerous researchers (e.g., Engle & Engle 2004, Medina-Lopez-Portillo 2004, Berg 2009, Jackson 2009, Baker-Smemoe et al 2014 also used Intercultural Development Inventory (IDI) developed by Hammer, Bennet and Wiseman (2003). IDI assesses the development of intercultural competence during a 20-minute test comprised of 50 questionnaire items. Each item is placed on a 5-point Likert-type scale. IDI identifies five stages of the development of intercultural competence based on the DMIS: denial, polarisation, minimisation, acceptance, and adaptation. The score, therefore, reflects the current stage of a participant's intercultural sensitivity. IDI is suitable for self-adaptation.

Oral Proficiency Interviews
Oral Proficiency Interiews (OPIs) are used to track changes in SA participants' spoken L2 accuracy and fluency. Although they are very useful for teaching practitioners, as they provide holistic scores, they are not very useful for researchers. Marijuan and Sanz (2018: 188) explain that for one thing, OPIs include multiple language components, and secondly, they may not be sensitive enough to track progress made by more advanced L2 learners. Therefore, OPIs are used by researchers mainly to measure changes in learners' developing fluency, understood as the ability to combine words and sentences into smooth speech (Baker-Smemoe et al 2014).
Most studies (e.g., Allen & Herron 2003, Baker-Smemoe et al. 2014) use mainly American Council on the Teaching of Foreign Languages' (ACTFL) Oral Proficiency Interview, which is a 20-minute face-to-face interview between a certified ACTFL tester and an examinee. It is a globalised and standardised functional speaking test during which the tester evaluates the examinee on a scale from Novice (1) to Superior (10). Although these scores do not reflect linear gains in proficiency, they are easily translated into numerical values, which allows for statistical analysis.

Conclusions
As transpires from the studies reported above, no one L2 acquisition context is superior to others. Learners in formal instruction (FI), study abroad (SA), or immersion contexts display various linguistic outcomes which seem to be related to individual differences between them rather than the learning situation. It is those individual differences that should therefore be studied in relation to different L2 acquisition contexts.
Some consistent research findings with regards to the role of intercultural competence in SA as a learning context may, however, be identified: (1) There is a significant positive relationship between the level of intercultural competence and L2 acquisition, namely the higher levels of intercultural competence, the higher L2 gains.
(2) Individual outcomes of SA are highly variable. A simple exposition to the target culture and the target language guarantees neither increased intercultural competence nor improved L2 acquisition. Success seems to depend on individual differences, such as a learner's attitudes towards the target culture, their willingness to communicate, and their proficiency level on arrival (the threshold level discussed earlier).
(3) Pre-sojourn intercultural training helps increase SA participants' intercultural competence and may consequently add to success in L2 acquisition. (4) SA experience pushes mainly the development of learners' oral fluency and discursive abilities. As participants engage mostly in oral interactions in different sociolinguistic and sociocultural settings their speech rates and awareness of social discourse increase. (5) Measuring the amount of L2 input and output during SA remains a challenge for researchers. Among the still imperfect data collection tools are Language Contact Profile , Language Use in Assessing Ethnic Identity (Laroche, Pons, & Richard 2009), and contact diaries (e.g., Coleman 2015). (6) Oral fluency is measured usually through oral proficiency interviews such as ACTFL's OPI. (7) Changes to SA participants' intercultural competence level before, during, and after study abroad are usually measured through the Developmental Model of Intercultural Sensitivity (Bennett 1986) and Intercultural Development Inventory (Hammer, Bennett & Wiseman 2003). (8) New research tools that help scholars achieve a better insight into the development of intercultural competence and L2 during SA include audio files video files, interview transcripts, SA blogs, diaries, essays, illustrations, written reports, and e-surveys.
The idealistic notions that it is enough to travel abroad and immerse oneself in the language to acquire it have been challenged by current research. Sending learners abroad without enough intercultural guidance is, more often than not, setting them on a collision course with the new culture. A need for a stronger link between SA and SLA research remains an issue. If, as evidenced above, increased levels of intercultural competence denote more openness towards the target culture and the target language, then efforts should be made to enhance intercultural competence levels before learners embark on their journeys. Pre-sojourn orientation sessions which are aimed at preparing students for SA seem to be best opportunities for intercultural training. Optimising pre-sojourn preparation may allow SA participants to make the most of their journeys, both in terms of intercultural contacts and L2 acquisition. Classroom intervention should be aimed, on the one hand, at preparing learners for culture shock, providing country-specific information, and teaching them culture-learning strategies and, on the other hand, at developing their L2 skills in other to achieve the required threshold level that would maximise interaction opportunities during SA.
About the authors Tomasz Róg, PhD, is a senior lecturer of in the Department of Philology, Stanisław Staszic University of Applied Sciences in Piła, Poland. He received an MA in English Studies from Nicolai Copernicus University in Toruń. In 2013 he obtained a PhD from Adam Mickiewicz University in Poznań. He specialises in the field of linguistics, with particular reference to English linguistics. His research interests focus on intercultural communication and linguistic creativity. Tomasz Róg has authored papers and monographs and regularly participates in conferences organised in Poland and abroad. His professional experience encompasses teaching subjects pertaining to linguistics, language teaching methodology, BA seminars and EFL at different levels.
Zofia Moros-Pałys, PhD, received her MA in German Studies from Dalhousie University and Adam Mickiewicz University and a PhD from Adam Mickiewicz University in Poznań. She worked as an assistant professor at the German Department of the Literature and Culture of German Speaking Countries at Adam Mickiewicz University in Poznań (Poland) and as an assistant professor teaching German Philology at Mieszko I School of Education and Administration in Poznań. Since 2016 she has been continuing her scientific research and teaching at the Department of Philology at Stanisław Staszic University of Applied Sciences in Piła (Poland). Her latest research interests focus on the analysis of German literature (literature at the turn of the 18 th and 19 th centuries and contemporary literature and film, especially about collected memory and Eastern Germany stories), contemporary German-Swiss literature and culture as also intercultural competence in learning foreign languages.

Introduction
Today, Sweden is characterised by a higher degree of cultural, ethnic, religious, social and linguistic diversity than ever before in history. Despite there being significant variations in the degree of such heterogeneity, Swedish preschools are increasingly diverse and are obligated to work by the stipulations set in the Education Act (2010:800) and the National Curriculum for the Preschool (SNAE, 2011(SNAE, , 2018. This means ensuring compliance with universal human rights and the fundamental values of Swedish society, including democracy and equality, regardless of gender, culture, ethnicity, religion or any other background. The Education Act stipulates that municipal preschools should be non-confessional, whereas private preschools should be non-confessional in their teaching but may include confessional components beyond this. These rights and values should guide the climate of a preschool, as well as its activities and methods. At the same time, preschools should exhibit openness to, tolerance for and appreciation of cultural, ethnic, religious and social differences, and respect for the unique backgrounds of children. Worth noticing is a change in the curriculum from SNAE 2011, which stated that "the preschool can help to ensure that… children with a foreign background receive support in developing a multicultural sense of identity" (p. 5), to the formulation in SNAE 2018 (p.6), which states that "preschool children should not be unilaterally influenced in favour of one or the other view. Therefore, teaching should be objective, versatile and non-confessional". 2 Considering these stipulations, it seems clear that the curriculum contains ambiguities and contradictory objectiveswhich, in the last instance, must be addressed by the preschool teachers.
Recently, there have been discussions about which food (particu larly meat) preschools should serve. There are calls for vegetarian, vegan and sustainable foodand a heated debate about the presence of halal 3 meat in the preschool. Food, diet and eating are associated with culture and religion, and are surrounded by rules of etiquette, beliefs and rituals. As such, they serve as symbolic distinctions between groups of people, at the same time as they exhibit control of individuals within a group and thereby ensure group cohesiveness. In our case, and to cite Rippin (2012: 28-29): "Halal is also employed in a sense which is opposite to that of haram in its meaning 'sanctified.'" This being said, food, diet and eating are laden with cultural value, religious symbolism, strong feelings and ideological capitalwhich manifest themselves in the preschools and among our youngest.
Should preschools give consideration to dietary or eating requests motivated by culture, ethnicity or religionmore specifically, should they serve halal meat? Here, halal meat refers to meat from slaughter and preparation according to Islamic practices (whereas in Islam, halal has a broader meaning and refers to actions or behaviours that are permissible). By contrast, "haram" refers to impermissible or unlawful actionsand in light of food, this means that pork, blood and wild birds are haram.
Since the curriculum provides limited guidance on this matter and since this is a decision for each preschool to make, it is managed differently. Research by Stier and Sandström (2018) shows that food and dietary requests are frequent topics of debate among preschool teachers and a matter they must deal with in everyday workin relation to children, parents and colleagues with diverging views on this matter.
Against this background, this text explores how a group of 41 preschool teachers at 33 municipal preschools make sense of and approach the cultural, ethnic and religious diversity in the preschool, with focus on questions related to the serving or not serving of halal meatin light of the curriculum stipulations. In the following, these preschool teachers' reasoning when it comes to the dilemmas that arise from the curriculum (i.e., universal and particularistic claims) is focal, but so are the strategies they claim to use as a means to address the challenges pertaining to food and eating in relation to the curriculum.
It should be noted that the meaning of halal (and haram) is not the focus here, and nor is whether it is right or wrong to serve halal meat in preschools. Rather, the focus is these preschool teachers' way of addressing this issueregardless of their understanding or attitudes towards such food. In addition, there are differences in the theories on culture and religion, and whether behaviours are motivated by religion or culture. However, the discussion of such differences falls beyond this text, and religion is treated as a component of culture: i.e. religion is a system of beliefs, world views, symbols and prescribed behaviours within a larger cultural system.
The empirical material comprises focus group interviews with 41 preschool teachers in two highly ethnically and culturally diverse urban areas of metropolitan Stockholm. These preschool teachers make up a highly diverse group of peoplethe majority were not born in Sweden, are bi-or multilingual, and are engaged in intercultural communication with the children, parents and colleagues. Also, their Swedish language competence varies greatly.
We hope to contribute by expanding the understanding of the dilemmas that Swedish preschool teachersand their colleagues in other countriesface as they try to manage dilemmas that result from the curriculum and that are associated with increasing cultural, ethnic and religious diversity in the preschool. We also hope to shed light upon and inspire alternative approaches to intercultural work in the preschool.

Previous research
There is much research on ethnic, cultural, linguistic, religious and social diversity in primary and secondary schools (Banks 1994, Souto-Manning 2007, Garcia 2018. However, such research that focuses on the preschool is more limited (Lunneblad 2006, Stier et al. 2012, Stier & Sandström 2018.
Also, much research is concerned with what preschools are obligated to do or should do (Freeman 1998, Geens & Vandenbroeck 2013, Han & Shelley Thomas 2010, Husband 2012, Keengwe 2010, Kemple 2017. Similarly, other research deals with the epistemologies of curriculum content or discourses on interculturality, intercultural education, differences or othering (Wardle 1998, Vandenbroeck 2007. There are, however, exceptions. Oliveira-Formosinho and Barros Araújo (2011) single out the characteristics of a pedagogical approach for early age groups, i.e. to promote respect for diversity and emphasise the importance of adult-child interactions and family involvement. Stier et al. (2012) and Stier and Sandström (2018) found that preschool teachers have different approaches to multicultural preschool groups. Such differences emanate from different understandings of cultures, modes of reflection and modes of interaction. Joshi, Eberly and Konzal (2005) found that although teachers/preschool teachers define culture as being a set of deeply embedded beliefs and values, in their practice they mainly focus on overt aspects like food and celebrations. Similarly, San Poon, Abdullah and Abdullah (2013) found that preschool teachers claim that their teaching exhibits awareness of multiculturalism, although an in-depth analysis revealed that they were more concerned with outward cultural manifestations. Joshi et al. (2005) studied teachers' and preschool teachers' practices when working with families from cultures different from their own. The most effective strategies for parental involvement were found to be written communication, and parent-teacher conferences/meetings.
Research indicates that much parent-preschool teacher communication and preschool teacher-children communication revolves around the children's food and mealtimes. Harding, Wade and Harrison (2013) and Stier et al. (2012) identify mealtimes as important learning opportunities where socialisation and language development take place. Yet, there are no studies on religiously motivated dietary requests in the preschool. Twiner, Cook and Gillen (2009) conclude that issues of religious identity are overlooked in debates on school lunches, particularly the question of serving halal meat or not. Stier and Sandström (2018) describe preschool situations where cultural and religious differences lead to dilemmas and friction between children, preschool teachers and parents. Here, the curriculum provides little guidance, and the preschool teachers must develop their own strategies to handle children's and parents' expectations and demands, including strategies to avoid difficult situations concerning gender roles, celebration of traditions, music, song, and food and dietary requests. For this reason, this text explores how a group of preschool teachers make sense of and approach the question of serving or not serving halal meatin light of the curriculum stipulations.

Methodology
In this study, data were collected by way of qualitative focus group interviews. This approach enables access to the attitudes, values and reflections of the participants (Krueger & Casey 2000;Kvale & Brinkmann 2009, Wibeck 2015. By allowing the preschool teachers to reflect upon and reason on a set of themes, there was proximity to the everyday life of the preschool.
The focus group interviews were held in Swedish, with one of the authors as the interview facilitator. Every preschool teacher participated in two interviews with about five months between the first and second. In total, 14 focus group interviews were conducted with 41 preschool teachers from 33 preschools. To ensure consistency between the interviews, a manual was used. For the first interview session, the manual contained themes on the understanding of culture, mode of reflection and described mode of interaction as regards multiculturalism. These themes were broken down into sub-themes and discussed among the preschool teachers. The second interview round focused on challenges and inherent possibilities in the preschool's everyday life.
Before the focus group sessions started, the interview facilitator informed the preschool teachers about the study's aim, approach and methodology, and on the purpose and set-up of the interviews. They were also informed about their right to terminate their participation at any time, without giving a reason for such a decision, and the fact that published interview extracts would be confidential. Confidentiality was ensured by using pseudonyms, altering the names of places and events when this was necessary, omitting redundant and irrelevant information, and storing data in a safe location inaccessible to unauthorised individuals. The preschool teachers were invited to raise additional questions or to ask for clarifications. If the participants decided to participate, they completed a consent form. The interview facilitator made back-ups of the consent forms and stored them safely to ensure participant confidentiality.
The focus group interviews began with the facilitator introducing the themes. To come up with multiple ideas and angles on these themes from as many preschool teachers as possible, the facilitators encouraged spontaneous, respectful group interaction and introduced additional questions or asked for clarifications. The sessions were recorded, transcribed verbatim, and translated into English by the authors. After each focus group interview, the preschool teachers anonymously completed a written evaluation, where they assessed how much they had shared their views. They could also provide additional comments.
Drawing upon the interview manual themes, the data (i.e., utterances, themes and stories) were subject to an initial qualitative categorisation. Based on this categorisation, the data were analysed to identify commonalities and differences in how the preschool teachers manage dietary requests from parents in light of contradictory or conflicting curriculum goals. Results were then discussed, and conclusions drawn.
To ensure a high level of authenticity, extracts from the interviews are presented verbatim: i.e., in the actual words of the interviewees. Yet, they are not presented in their original form, since they have been translated into English. For reader clarity and issues of space, some editing has been necessary.

Theoretical approach
To analyse the focus group interviews, we have used a combination of concepts pertaining to the cultural construction and religious symbolism of food and eating and a framework for analysing approaches to multicultural classrooms used in previous studies (Stier et al. 2012, Stier 2016, Stier & Sandström 2018. This framework assumes that the preschool teachers participating in the focus group interviews adhere to different, yet overlapping, approaches to manage the challenges of ethnocultural diversity as they work to mee t the stipulations of the preschool curriculum. These approaches are labelled "instrumental multicultural", "conscious multicultural", "intercultural" and "transcultural". In turn, these approaches have three components: the underlying conception of culture, reflection mode and interaction mode. The underlying conception of culture refers to tacit views on the meaning of cultureviews salient in thought and concrete ways in daily preschool practice. Culture includes, but is not limited to, language, world views, values, norms, traditions, habits, rituals, symbolsand food and dietary requests (Douglas 1966).
Conceptions of and traditions related to food and dietary requests and restrictions are symbolically loaded components of any culture or religion. Neither Jews nor Muslims eat pork, Hindus do not eat beef, while Swedes do eat crayfish. Festivities centre around food, and meals are surrounded by rituals and norms, and what is eaten or not is prescribed by taboos and values. Thus, food is not merely about matteri.e. things we eat for nourishment, survival or health reasons (Barthes 1997;Lévi-Strauss 1969;Douglas 1966Douglas , 1984. Drawing upon Douglas (1966), food and eating are symbols that enable classifications which, in turn, ensure social order within a group and maintains boundaries with other groupsbetween "us" and "them". To put it differently, food is "a code expressing cultural identity and, at the same time, otherness" (Stano 2017: 87). Foods that fall outside these classifications (in our case haram) are seen as dirty, uneatable or taboos, and are potential threats to the social order. Therefore, for reasons of ritual cleanliness, desirable food (e.g., halal) is prescribed, whereas violations to dietary rules are objects of condemnation or sanction. According to Rippin (2012), the word halal can be viewed as "ceasing ritual avoidance behavior" (2012: 28-29). In turn, such questions become ideological and political. It is against this background that the focal place of food, eating and dietary preferences in the preschool should be seen. Meal times and discussions of food are also learning opportunities, and not merely occasions for nourishment (Stier et al. 2012;Harding, Wade & Harrison 2013). In the results section, focus is on how the interviewed preschool teachers manage food, diet and eating based on the demands and wishes of parents, and in relation to the curriculum and law.
Reflection mode pertains to self-knowledge, openness and views on the limits of tolerance and adaptiveness based on cultural, ethnic, religious or social background. Finally, interaction mode is about how the understanding of culture and ways of reflection manifest themselves in the preschool teacher's interaction with children, parents and colleagues. Taken together, these three components impact how the preschool teachers approach their professional practice, that is, how they understand and act with regard to the perceived ethnocultural diversity in the preschool group they work with.
The instrumental multicultural approach is anchored in a shallow understanding of cultural, ethnic and religious heterogeneity (Stier et al. 2012). Cultures are viewed as different, static and internally homogenous entities. The preschool teachers are unwitting bearers of conceptions and stereotypes. Working in a multicultural preschool becomes a question of coping with differences, often with a distant attitude. An example is how they (unknowingly) refer to and exoticise other cultures or focus on cultural totems: e.g., they encourage the children to see, touch or taste things from other cultures without relating these activities to a larger context. In addition, there is a reluctance or inability to set boundaries for what is consistent with the curriculum, the law or norms of the dominant society.
By contrast, the conscious multicultural approach means that preschool teachers are more aware of their role as the bearers of conceptions and stereotypes, and yet are quite unreceptive to alternative views in interaction with children, parents or colleagues (Stier et al. 2012). They are also reluctantly distant when they cope with differences and are anxious when it comes to their boundary-setting when parents pose demands on the preschool. Cultures are viewed as partly dynamic and internally heterogenous entities.
An intercultural approach is based on a view of cultures as highly dynamic, internally varying and (re)constructed in social interaction. This is paralleled by a genuine openness to explore and, to some extent, to work with and make use of alternative cultural repertoires, including language and world views (Stier et al. 2012). This is done by making use of the children's inherent curiosity and capacities, and by stimulating their willingness to transgress cultural and ethnic boundaries and explore other cultures. It means taking the children's queries seriously, coping with cultural variation, challenging one's personal cultural conceptions, exploring intercultural interaction and being unambiguous when it comes to boundary-setting. Preschool teachers who adopt a transcultural approach encourage children to transgress cultural boundaries, beyond the intercultural approach (Stier et al. 2012). The transcultural approach encourages or even urges children to transgress such boundaries. The underlying assumption is that cultures are more similar than dissimilar and that they are constantly constructed and reconstructed in social interaction. Using meta-reflection, the ambition is to lay bare and make use of human similarities and to encourage intercultural exchange and cross-fertilisation.

Overall approach
Using this framework, we can analyse how the interviewed preschool teachers understand culturally, ethnically and religiously prescribed dietary requests, how they reflect upon this matter and how this shows in their interaction with children, parents and colleagues.

Results
In the focus group interviews, questions, dilemmas and considerations pertaining to food and diet requests were a common part of everyday life at the preschool. Moreover, these are highly entangled with cultural values, with conceptions of 'proper' and 'non -proper food', as well as with the ideological underpinnings and legal requirements of Swedish preschools. In concrete terms, these questions, dilemmas and considerations were mainly about balancing what was considered to be best for the children with the wishes of the parents. In particular, the interviews concerned problems with halal meat.

Avoidance strategies
In terms of the interviewed preschool teachers, the instrumental multicultural approach is common. They describe other cultures primarily as different from the Swedish majority culture and view them as static and homogenous. Hence, dietary requests posed by parents are unconsciously viewed as wishes from internally homogenous 'others' and one is often unaware both of one's own cultural food preferences and of the reasons and dilemmas connected to the parents' wishes. Questions about what food to serve or not are dealt with when they occur, rather than in a systematic and prescriptive fashion (compared to questions of gender, which are dealt with more consistently). Not only are questions regarding food dealt with in an ad hoc fashion, decisions related to such requests are made reluctantly and ambiguously, with little consideration for the curriculum or wider implications. On the other hand, there are examples of what has been referred to as a conscious multicultural approachi.e. the fact that the preschool teachers seem aware of the strong religious symbolism of food preferences, and of the interlinkages (justifiable or not) with ideas related to freedom of religion. There is also recognition of the dynamic nature of values and of the diversity of opinions among parents.
Regardless of this, to avoid problematic situations, they are reluctant to set clear boundaries vis-à-vis the parents and therefore come up with strategies to rationalise their actions and to avoid problematic situations. These avoidance strategies become instrumental, i.e. they become concerned with avoiding problems rather than elaborately reflecting on and accommodating their practice implications. In quote 1, the strategy in a problematic situation is to circumvent it, with the pretext (rationalisation) that from a sustainability perspective, vegetables are better than meat: (1) We cannot continue to eat as much meat as we do, it is not sustainable, so we have discussed… when we serve food… who shall have this and who shall have that and then you don't know who shall have halal.
Thus, being highly aware of the religious symbolism of eatable and non -eatable food, and thereby of potential conflicts, reasons are sought in a less-heated sustainability discourse rather than in the curriculum or in a discourse of religious freedom and privilege. The sustainability discourse connotes scientific objectivitya-matter-of-factnesssomething that is underpinned by the curriculum stipulations that the preschool should facilitate sustainability.
Another strategy employed to avoid the problem is to serve halal meat to all children. One reason for this is economical. Since Swedish municipalities have a financing system whereby preschools are paid based on the number of children, children who move to another preschool (where halal meat is served) represent a financial loss. As such, accommodating dietary requests may serve to make the families stay at the preschool: (2) It generates children if the preschool serves halal meat… and many have gone to private preschools… I think what is most interesting is that suddenly halal becomes the meat we serve all children… if you are Christian or if you are not religious at all, then you probably don't want the blessing of this meat.
Therefore, although the preschool teachers can see the disadvantages of serving halal meat to everyone, they are reluctant, unwilling or, perhaps, unable to truly address the problem. Instead, the matter-of-factness of the economic realities is used to justify the unwillingness to address the question of halal meat. Economics, rather than the curriculum or values, define the strategy, the result being that the requirements of the parents are met. At the same time, the preschool teachers are aware of the issues of serving halal meat and may even complain about their own willingness to set boundaries when it comes to the preschool adapting to parental wishes. In a semi-reflexive fashion, and without touching on the question of cultural values and conceptions of food, they claim that serving halal meat is inconsistent with the guiding documents: (3) If you read the guiding documents for the preschool, you will find that they state that the school and the preschool shall be non-confessional… but halal meat is a religious element that we allow in preschool.
The preschool teacher is aware of the inconsistency between parents' dietary requests and the Education Act's stipulation that the preschool be non-confessional. As a consequence, the question of preferences for "Swedish" food is not addresseda question which is loaded with religious symbolism and which is the object of heated discussions in public discourse. Instead, the strategy is one of passiveness and unwillingness to "get involved". Inconsistent with an intercultural approach, they do not set any boundaries, nor involve themselves in discussions with the parents on these matters. Instead, they give way to external pressure, money or the vagueness of the curriculum or religiously motivated claims that are inconsistent with the curriculum. At the same time, they let the latter take precedence over the legal principle of a non-confessional preschool. In their interaction with parents, another avoidance strategy among the preschool teachers is to use noa words instead of the word "halal". 4 In quote 4, the term "halal meat" is replaced by the term "the red bowl": (4) We negotiate with parents about, for example, this thing with halal; then they are given their halal meat and the others are given regular meat… we have criticised that. We don't say halal anymore, we say… "the red bowl".
Once again, this means that the core question and symbolic power of food preferences are not addressed; rather, the issue is diverted by using rhetorical tactics. In quote 5, the preschool teachers are both highly aware of the sensitivity of these issues and engaged in metareflection: (5) About this thing with halal, we have actually made a decision not to call it halal meat, because the word haram is a rather strong negative word and we don't, if I understand it correctly, have a word for it in Swedish.
Thus, they are not in the habit of talking about halal openly; instead some of them refer to halal as "AF" (alternative food): (6) Everything that is not halal becomes haram and haram is "not okay" and then in the end it becomes wrong, we think, that these children are given approved food and the others in some way food that is not approved… therefore we have made a decision to call it alternative food, so we say AFso we have sort of switched.
Thus, one reason for not using the word "halal" but the safe phrase "alternative food" is because halal is associated with its opposite: "haram". Another reason is the connotative objectiveness of "alternative food", which serves to "desacralize" the food. This line of argument indicates a certain meta-reflexivity (consistent with the transcultural approach) and a relative insight into the meaning of halal and haram. Yet it exhibits little awareness of how one's own vocabulary and expressions divert attention away from the core issue: i.e., the extent to which dietary preferences should be met in the preschoolmeaning that they fall short of the intercultural or transcultural approaches. Such a renaming strategy is not without problem. Quote 7 exemplifies how enmeshed and complicated the reasoning becomes when the preschool teachers name the different "sorts of meat" for the children. Here the noa word for halal is "general food" or "variation meat": (7) Yes, we call it halal, but now we have received a letter stating that we must stop saying halal; we must call it "general food"… "variation meat", because there is this thing called haram. Haram is the opposite of halal. And it became very… the children sat looking at those who didn't eat halal and then it was haram and it was ugly.… you shouldn't hide that there is pork. After all, these children are growing up in a multicultural society; you just have to accept that situation.
The last sentence in the quotation suggests that the preschool teacher has reflected on these questionsand has come to the conclusion that the preschool teachers ought to deal with this problem more directly and not avoid it, indicating more of a reflection mode consistent with an intercultural approach. At the same time, this insight does not materialise in preschool practice (i.e., interaction mode), especially not if it means drawing a line in the sand vis-à-vis the parents.

The preschool teachers' efforts to change views
As we have seen, in relation to the parents, the curriculum and the preschool as an organisation, much effort is made by the preschool teachers to deal with the challenges that come with dietary requests. But such dilemmas must also be managed in relation to the children. The curriculum stipulates that children should be heard and encouraged to participate in the everyday life of preschool: (8) In all of this, what are the rights of the child? If I have decided, as a parent, that my child shall be vegetarian. S/he will sit there and see the other children eat and really want to taste it… so what happened with the rights of the child?
Several preschool teachers return to the fact that children are influenced by their parents. In the case of food, the preschool teachers do not think or act according to an intercultural approach. Rather, they refrain from challenging the parents and remain at a conscious multicultural level, whereby they assume that parents want th eir children to follow certain dietary practices, and in this respect the children become bearers of food-related values and religious symbolism.
At the same time, the children are curious and more open to eating different food. At an early age, the symbolic boundaries are less internalised than they are in adults. Therefore, some of the preschool teachers try to make parents loosen their demands that their children must "eat according to the parents' culture". The rationale is that the children should be encouraged to explore and transgress symbolic boundaries: i.e., an intercultural or a transcultural approach. Albeit that it discusses vegan and vegetarian food, this quote illustrates these approaches: (9) We inform the parents that "this happens in the eating situation". Because sometimes it is not about pork in the first place but vegetarian foodif a parent has chosen that. When we say, over and over "this happens in the eating situation and your child doesn't feel good about it"… many times the parent will change his/her mind and say "OK, at preschool it is OK". That is to say: "we do what we do at home, but he or she can have what others are having at preschool".
To convince the parents, the preschool teachers refer to the children's well-being and comfort. The quote also shows how the preschool teachers, in their interaction with parents, deal differently with the concrete meal situation and demands from parents when it concerns vegetarian or vegan food as opposed to halal meat. The reason for this is that the former has less cultural and religious connotations.
Moving on the preschool teachers use food and the meal situation as an intercultural learning opportunity: (10)Meal situations are very good forums for discussion. There is lots about language: "How is that said? How can you say that in English?" There are different dishes, and we talk about how we eat and what they eat in that country. Yesterday one girl saidwe had meatloafand she said, "I want some more chicken!" -I said, "it's not chicken, it's from a cow" -"Cow?" she said and looked very puzzled. I said: "yes, people can eat cow", and we started to rattle off all different animals that people eat in different countries… and had a very exciting discussion.
Together with the children, the preschool teachers are involved in interaction and reflection that broaden both the children's and adults' views on and understanding of different cultures. In this way, the symbolic boundaries of food become visible and, possibly, an object of scrutiny. For the children, this is consistent with the goals of the curriculum. Frequently, the preschool teachers have intercultural discussions with the children about culture, religion and symbolic boundaries. In these discussions, the children often refer to, or unknowingly reveal, their parents' attitudes: (11)The words "halal" and "haram" are very popular among the children, and they talk about so many things that are forbidden. They ask me: "do you eat pork?" or "are you a Christian?" When I say that "yes, I am", they go "aah!" or "what?!" They have so much prejudice from their parents…. You cannot eat with your left hand, and me being left-handed! One child said, "eeeh, you are eating with your left handyou are Sheitan's daughter!" I said, "what? whose daughter?" -"Sheitan's", s/he said. "That's the devil's daughter. My mum says you cannot eat with your left hand because then you are Sheitan's daughter, son or child." "No, you can eat with both hands; it's no problem," I said.
These discussions reveal that the preschool teachers' mode of reflection extends beyond the instrumental multicultural and conscious multicultural approaches. Instead, by drawing the children's attention to the "cleans and dirties" and "acceptables and unacceptables": i.e., the symbolic boundaries of society, they adhere to an intercultural approach. As well, the preschool teachers make efforts to influence the children so that they become more intercultural: (12) I say "I can eat halal meat, I can eat my food, I can eat every sort of food, I have more to choose from, because I can eat everything"… they need a counterbalance since of course at home it is "this is how we do things, this is how it is".
Thus far, we have presented a dozen examples of meal situations where challenges result from divergent views on dietary requests. Values and symbolism permeate these views, and typically such values and symbolism are at odds with one another. Therefore, these dilemmas must be managed by the preschool teachers themselves, something that tends to be done with an unreflective or semi-reflective mode, and an ad hoc-like mode of interaction. The examples given illustrate how the preschool teachers adhere to different understandings of culture and religion and have different modes of reflection and modes of interaction vis-a-vis the children and their parents, as well as to the stipulations of the law, curriculum and the organisational realities of the preschool. A consequence of this is that the preschool teachers handle these issues differently. Yet, a common denominator is that there are two overall approaches: avoidance and attempts to evoke change.

Discussion
Swedish preschools shall ensure compliance with universal human rights and the fundamental values of Swedish society (lpfö 1998/2010). These rights and values should guide the preschools' climate, activities and methods, at the same time as the preschools should exhibit openness to, tolerance of and appreciation for cultural, ethnic, religious and social differences, and respect for the children's unique backgrounds. In the eyes of the preschool teachers, this means that the curriculum contains potential ambiguities and contradictory objectives. It also remains unclear to the preschool teachers how these objectives are to be accomplished. This fact, given the high level of diversity in Swedish preschools, and current debates on migration to Sweden have made food, diet and meal situations sensitive matters. Previous research (Stier & Sandström 2018) shows that preschool teachers must handle the challenges that arise with food, diet and meal situations. Often, they adhere to an instrumental multicultural or conscious multicultural approach in such situations.
Just as in the case with preschool diversity in general (Stier et al. 2012, Stier & Sandström 2018, when the preschool teachers handle dilemmas pertaining to food and dietary requests, they draw upon a conception of culture, a mode of reflection and a mode of interaction. Based on these, four different, yet overlapping, approaches to preschool diversity have been proposed: instrumental multicultural, conscious multicultural, intercultural and transcultural. In this study, the preschool teachers mainly adhere to an instrumental multicultural approachand less so to a conscious multicultural approach (illustrated in quotes 1-7). Cultures, ethnicities or religions are viewed as different, static and internally homogenous entities, and the preschool teachers are relatively unaware of themselves as cultural beings. This includes a limited understanding of matters having to do with ethnic and cultural heterogeneity and how such heterogeneity includes ideas about food and meals and related prescribed behaviours and dilemmas.
Cultural values and religious symbolism related to food are components of any culture or society. Meal times are surrounded by rituals and norms, and what is eaten or not is prescribed by taboos and values. Thusand as we see in the interviewsfood, eating and meals are loaded with symbolic meaning. Food defines boundaries between what anthropologists refer to as clean and dirty and what sociologists and others conceptualise as "us" and "them". Yet, the preschool teachers are relatively unaware of the social dynamics of such boundaries, and simply see and handle them matter-of-factly. This is expressed in how they talk about the religious and cultural aspects of halal meat, while they give little or no attention to the cultural, ethnic or religious underpinnings of food and meals of Swedish society. This affects their outlook on things and their interaction with children, parents or colleagues (Stier et al. 2012;Stier & Sandström 2018). More specifically, it impacts their view on how far the preschool should go to accommodate food preferences of and dietary requests made by parents.
For the preschool teachers, a multicultural preschool becomes a question of coping with differences, with a distant or exoticising mode of interaction and with focus on the culture, ethnicity or religion of "the other". In interaction with parents, there is an unwillingness or reluctance to set clear and unambiguous boundaries when it comes to food preferences and religiously motivated dietary requests, consistent with the curriculum, the Education Act or the overall norms of dominant society. Instead, the preschool teachers adhere to different avoidance strategies: i.e., not challenging the parents on their wishes and demands.
Such strategies include pointing to but not enforcing curriculum stipulations, relabeling, redefining or circumventing the problem or controversy, remaining passive, using 'cooling strategies', using noa words or serving halal meat to all children. In quote 8, the preschool teacher's reflection mode is consistent with an intercultural approach, whereas his/her interaction lacks the clear boundary-setting of such an approach. Additionally, quote 9 is a rare example where the preschool teachers question and argue against the views of the parents, and try to make them change their stance. In this sense, they adhere to more of an intercultural approach. At the same time, this quote revolves around vegan and vegetarian food, not culturally, ethnically or religiously motivated dietary requests.
There are additional examples of the intercultural approach (e.g. quotes 9 -12). Here, the preschool teachers' interaction with the children is more focused on laying bare and discussing cultural, ethnic and religious differencesand relates such discussion to food, eating and the symbolic boundaries and values surrounding food. In these instances, the preschool teachers try to stimulate the children's curiosity and willingness to explore these things. Quotes 9-12 exemplify an intercultural interaction mode. In addition, there are fragments from the transcultural approach (e.g., in quote 9). In this sense, for children food is a tangible and focal part of everyday life to reflect on and discuss. Yet, the preschool teachers need an in-depth understanding of the fact that food is more than matter. It is a symbol that defines cultural, ethnic and religious boundaries and that regulates the relationships between individuals.
Moreover, the children are used as intermediariesto inform the parents about values, methods and food-related ideas. In one respect, this is also an avoidance strategyi.e., the children are used to influence or change the parents' dietary requests. An outcome of this is that the parents better understandand, in turn, accept -the preschool's way of working. Another is that the children become "squeezed" between the expectations and demands of their parents and the preschool.
The avoidance strategies (particularly in their communication with the parents) that the preschool teachers describe are typical for a multicultural-instrumental approach. Hence, their understanding of culture as such and the cultural underpinnings and symbolic power of food is relatively shallow and their mode of reflection largely unconscious. By renaming the food or by serving halal meat to everyone, their mode of interaction is largely void of boundary setting. At the same time, in their interaction with the children, the preschool teachers exhibit more of a multicultural-conscious or intercultural approach. With the children, cultural values and symbolic boundaries are explored in a more relaxed manner. Given the heated discussions on halal meat in Swedish preschools, this is unsurprising.

Conclusions
The results show that the preschool teachersas well as the preschools as organisationshave problems handling dietary requests from parents. Their approaches vary with regard to how they view culture, how they reflect upon these issues and how they interact with children, parents and colleagues. Many of them come across as relatively instrumental and unconscious in their approaches, whereas others are more intercultural in their ways and, for instance, see the value and potential of diversity. Common to all of them is that they use avoidance strategies and use the children to influence the parents' views on food and eating.
In stark contrast to the strong emphasis on gender equality in Sweden, which also permeates the preschools and materialises in their work with gender pedagogy, much less ideological consensus and legal clarity surround questions of culture, ethnicity and religion. Neither the curriculum nor the law is sufficiently specific in these matters.
Why is that? The adherence to multiculturalism may have led to vaguely defined boundaries to what a society should accept in terms of culturally, ethnically and religiously motivated claims on society, whereas boundaries and tolerance for variations with regards to gender and the treatment of boys and girls in preschools are less open to compromise. The preschool teachers have been trained in gender matters and are confident in handling them in relation to parents, whereas there is no professional guide pertaining to cultural, ethnic and religious diversity or dietary requests based on these. Instead, the preschool teachers must develop their own approaches to these questions as they arise in everyday practice. As such, the preschool teachers would benefit from support that can help them to work towards the objectives of the curriculum.
What has been analysed in this study are focus group interview data: i.e., accounts of a limited number of preschool teachers, not what they do in their everyday practice. Yet, the study is valuable and trustworthy since it provides new knowledge on how preschool teachers reason around these matters. In focus group interviews, the presence of others may decrease the participants' willingness to share their thoughts, at the same time as the design allows for more breadth and width when light is being shed on a particular topic. Therefore, the benefits of such this study exceed its limitations.

About the authors
Jonas Stier is a professor of social work at Mälardalen University. He has a great deal of experience in the research field of intercultural studies within a variety of contexts such as preschools, schools and higher education, and has written several books in this subject area. Further, he is strongly committed to public outreach and is a popular lecturer outside academia.
Margareta Sandström is a senior professor of special education at Mälardalen University. She has a great deal of experience researching in the fields of education, special education and intercultural studies within a variety of contexts such as preschools, schools and higher education, and has written several books on this subject area. Further, she is strongly committed to teacher education.