Outsiders and learners: Negotiating meaning in comparative European social work research practice

This paper draws on two experiences of undertaking comparative research in England, Sweden and Italy, and offers a discussion of the kinds of dilemmas raised in relation to language and meaning surfaced by these. Its primary focus is to examine the reflexive construction of meaning, both in relation generally to the workings of an international research team and, more specifically, how such construction of meaning impacts on the process of qualitative interviewing across linguistic and cultural boundaries. The article argues that a culture of reflexive research practice is helpful for approaching national differences, to facilitate understanding in internationally mixed research teams and to support and empower participants in research, in second languages and differing cultures. Overall, the paper attempts to advance some tentative ideas that other international researchers may be able to consider and deploy in relation to their own research ambitions.


Introduction
European social work research received further recognition and endorsement with the establishment of the European Social Work Research Association, in Bolzano 2014. Given this focus, it is an opportune moment to direct attention to the strengths and struggles of undertaking comparative European social work research. It is to this renewed debate that this paper aims to contribute.
It has been noted with regularity that although much has been said about why comparative research is desirable, very little links theoretical considerations to methods and techniques (Ha¨ma¨la¨inen, 2014). This paper's focus is directed towards concepts and practices. Its primary aim is to consider how issues of language, and therefore of meaning, can be accommodated when working comparatively across linguistic divisions -an issue that all comparative European research faces to a minor or greater extent.
The paper will focus on issues of language and meaning in comparative research. Firstly, it offers a little background: the execution of two small European research projects which led to the authors' focus on comparative issues. Then it considers previous research in the field, particularly the body of work concerned with reflexive and team approaches in comparative contexts. Next, the paper turns its attention specifically and substantially to issues of language in international research endeavours, particularly the language issues in conducting interviews, and the language issues in multinational research teams. This is followed by a consideration of issues of reflexivity in international research groupings generally, and the authors' particularly. Some additional consideration of language, identity and culture is then developed, before concluding with a brief review of how knowledge from the team can inform the research process itself.
The paper will give some attention to the prevalent literature of comparative research, considering such issues as context, and using the problematic of 'what actually are the elements of comparison?' Much has been said on 'national' as a context in itself, but in many ways this does not tell us much (Baines and Cunningham, 2013). 'Context' for comparison in social work can be more to do with e.g. different welfare regimes, law, types of organisations and activities. Really the national is less the point than those factors and what meanings they prescribe and delimit for those who work with (in) them. The core of this article, though, is an examination of the reflexive construction of meaning, in relation to both an international research team, and the manner of its inflection in the process of qualitative interviewing across language and cultural boundaries. It argues that a culture of reflexive research practice is fundamental to approaching national differences, to best facilitate understanding in internationally mixed research teams and to support and empower participants in research, in second languages and second cultures.

Background
The background to this paper is the authors' involvement over some years in a range of international projects (Frost and Campanini, 2005). The two particular comparative research projects that generated the observations this paper discusses, were undertaken in Sweden, Italy and England. Both were financed (with seed money of 4000E for each) by The European Association of Schools of Social Work. The first project examined social work students' beliefs about their readiness to practice, in their final year of social work education, and after one year in practice. It was a qualitative, small-scale, pilot study with five students from each country interviewed twice, alone or in groups. The findings suggested considerable similarities as well as clear differences between the students' experiences and the way social work education prepared the students in different contexts (Frost et al., 2013a).
The second project considered the issue of resilience in child protection social work, from the perspective of how people stay, rather than why do people leave. A total of 57 workers in England, Italy and Sweden were interviewed using qualitative methodology within a hermeneutic paradigm, and employing semistructured in-depth methods (Frost et al., 2013b).
A single non-national ('outsider') researcher undertook the interviews in the first project, and a non-national interviewer was used, alongside a local (national) researcher, in the second (see below for discussion). Just for the reader's interest: the findings, after thematic analysis, suggest that 'surviving' in child protection work was enhanced by organisational factors, relationships, interest, commitment, supervision and education (both initial and ongoing) (Frost et al., 2013b).

Previous research
Previous research on comparative methods in social work research typically includes four statements: (a) it is important; (b) it is difficult; (c) it is not so common and (d) if it is done it needs careful consideration of how it affects meaning and understanding, especially when it comes to developing local social work practice (see references below). Comparative research is often mentioned as a 'traditional' mode of research, widely used in many fields of inquiry and perhaps especially in social sciences (Azarian, 2011), a claim that is regularly substantiated with reference to classic theorists such as Durkheim (1982Durkheim ( [1895), who claimed all sociological analysis as necessarily comparative. A further regularly asserted claim is that even if social work in individual countries is 'deeply connected with and shaped by country-specific social, economic, political and cultural factors' (Ha¨ma¨la¨inen, 2014: 193), comparative research is needed to clarify the theoretical and institutional diversity of social work through the identification of similarities and differences between the systems and traditions found across nation states. (Ha¨ma¨la¨inen, 2014: 192) However, a great deal of the literature of comparative research points to a plethora of problems involved. The highlighted difficulties include, for example, actual 'comparing', e.g. how countries really can be compared, how language often makes comparisons difficult and how the context creates challenges in producing any meaningful analyses of the data (Paso, 2014).
The acknowledgement in much of the literature that there seems to be a lack of (often combined with the need for) comparative studies is only rarely the prequel to rectifying this deficit with thorough reviews of the existing literature, which are strikingly absent (Shardlow and Cooper, 2000;Shardlow and Wallis, 2003).
When social work comparative research is found, it tends to focus on standardised and rather policy-driven approaches, which can also be something of a limitation. There is a tendency to mainly build on only four areas: (i) theoretical studies; (ii) particular sectors across nations, such as issues of social policy, housing policy or child protection policy; (iii) evaluating policy effectiveness across states or (iv) a concentration on country comparisons with a focus on certain welfare provisions (Harris, 2007).
Additionally, the existing literature of comparative research in social work is full of warnings or discussions in relation to methodological limitations. The advice frequently proffered is to rely on multiple methods in cross-national research in order to avoid major methodological pitfalls (Hantrais, 2014;Quilgars et al., 2009;Smeltser, 2003). This is helpful up to a point, but multiple/mixed methods have never ceased to be highly controversial epistemologically and often impractical in reality (see e.g. Bryman, 2012).
Usefully for the work here, is the seam of existing literature on the importance of reflection and reflexivity in the research process: nationally and internationally; individually and in teams (Findlay and Gough, 2003;Hammersley and Atkinson, 2007;Siltanen et al., 2008). As a research team, it became quickly evident that to be able to build an effectively functioning international group where difficulties such as meaning and language could be surfaced was a crucial element for producing worthwhile findings. As Hammersley and Atkinson note: the concept of reflexivity acknowledges that the orientations of researchers will be shaped by their socio-historical locations, including the interests that these locations confer on them. (Hammersley and Atkinson, in Baines and Cunningham, 2013: 75) However, a specific literature that highlighted a subjective and reflexive stance, allowing the connecting up of culture, language and identity with interpretation and creation of meaning in the processes of qualitative research, proved mostly illusive. Mangen's work in the 1990s offered a good analysis of some of the struggles with cross-national qualitative research in general that is mindful of complexity and particularly the difficulties with working across languages without being unethical or overambitious: the strengths of qualitative approaches lie in attempts to reconcile complexity, detail and context. Critical for protagonists is the integration of 'reflexivity', by which is meant the ability of researchers to take stock of their actions and their role in the research process. This is a particularly urgent task in research crossing cultural boundaries. (Mangen, 1999: 110) As well as the obvious problems of e.g. working in a single language as exclusionary and hierarchical, he also presents, dovetailing with the authors' work, concerns that the language issues are intensified where there are 'emotional' issues under consideration. Leaving aside the psychosocial challenge that all issues are 'emotional' (Frost, 2008b), certainly the research projects underpinning this paper surfaced personal histories and experiences of 'survival', preparedness, resilience etc. This locates the enquiry in this category of (potentially) 'emotional' even by quite rationalist criteria.
However really it is more recent studies that look at the role of reflexive approaches as a countermeasure to the challenges of international comparative research, not only in terms of understanding their disparate contexts (Baines and Cunningham, 2013) but also in relation to building reflexive multinational research teams (Siltanen et al., 2008).
Siltanen et al. are particularly helpful in challenging a static essentialist notion of reflexivity, as for example, a single person reviewing existing ideas to better understand them, and deploying instead a more constructionist and relational approach: a process is identified. Through sharing observations and understandings of the research interview and indeed of the broader processes at work, a research subject, research data and a relational research team emerge, bought into existence and as a contingent and fluid entity (2008). Mechanisms for facilitating this emergent construction of knowledge and identity are examined: e.g. interviewing in twos and analysing data in the group. The whole premise of 'working separately together' is an appropriate maxim for the comparative research team. Even though Siltanen et al.'s discussion did not encompass international diversity per se, the focus on diversity in the team is apposite for such work: 'In research teams, elements of this ''otherness'' are present within the team and here too, researchers capacities for reflexivity can be significant resources for recognising, understanding and interpreting diversity of experience' (Siltanen et al., 2008: 48).
As well as highlighting the need for a reflexive stance in relation to understanding the 'other' of the research process -the interviewee -that the team consists of 'others' is important to build into the reflexive process, Siltanen et al. argue. How much more so then, when the 'others' of this research, both the subjects and the team, are from different cultural and linguistic arenas.
The international aspect of reflexivity is given a particular emphasis by Baines and Cunningham (2013). For these researchers, the importance of reflexivity is to be able to both enter into the world of, and keep an identificatory distance from, the subject of the research: a challenging balance. Some of the techniques mentioned -the importance of interviewing with one 'outsider' and one 'local' present; the importance of analysing findings in the team -were recognised as promoting both a counterbalance to and a synthesis of the need to identify with and maintain objectivity in relation to the world of the researched (Baines and Cunningham, 2013).
It is such reflexive and comparative approaches, particularly, that this work has attempted to build on. The paper has attempted to learn from all of the above, and remains mindful of the strengths and limitations that have been articulated. Its particular intention is to focus at the praxis level, particularly on reflexive approaches to reducing the struggles of qualitative comparative work, and additionally to adhere to the fundamental working principle for this project: that the international team should become the model for the work overall. The paper will now offer some consideration of how specific challenges in the authors' comparative research -language, culture and identity -were addressed.

Taking language into account
Perhaps the most often claimed difficulty with cross-country/cross-cultural comparisons is related to language. This is of course particularly true -unavoidable in reality -in qualitative research, where all forms of cultural knowledge are fundamental for interpretation (Quilgars et al., 2009). This is a multifaceted and ever-expanding discussion. Post-structuralism's fundamental tenet of linguistic construction/determinism has in many ways made all discussion of language per se redundant: identity, culture, reflexivity are all discussions of language. And in relation to research interviews, the discursive construction and reconstruction of the interview subject are a fundamental given (Yuval-Davis and Kaptani, 2009). Here, though two different aspects of language are addressed: the language deployed when interviewing people with a different background than the interviewer and the language constituting the multilanguage team.
Thinking firstly of the language of the interview and comparing this team's experiences to the studies referred to above, the essential problematic that emerges is 'obstacles to understanding' (Mangen, 1999). One core dilemma is whether to interview people only in their own language and use interpreters in order for the researchers to understand them: for some an important ethical consideration (Temple, 2006;Temple et al., 2006). The fundamental point is that respondents draw on a range of narratives through which they make sense of their experience, and express these within the interview context, which then in itself constructs meanings from them. Is this process damaged where the interviewees have less than optimum access to their whole range of expression and nuance? With interpreters, the struggle with meaning is shifted to a different context: that of the translation, literally, and in the sense of how the interviewer reads off meaning from the interview at the time. Both are unreliable tools, for co-constructing 'findings', with many possible pitfalls (Liamputtong, 2010). Realistically too, small-scale comparative research rarely has sufficient funding to allow access to professional translators (the authors' being no exception). Available aids to translation tend to be a mishmash of ad hoc contacts such as fellow students and friends, whose understanding of the questions, relationship to the answers and fluency differentially represents the interviewees' narratives. In the case of this study, the unequivocal position of using English as the common language for all interviews was adopted. This meant that Swedish and Italian students would discuss their experiences in a second language (and likewise, two of the researchers). This was not completely unrealistic: it is a prerequisite for both Italian and Swedish university students to have a working knowledge of English, in order to be accepted for social work education, and some of the core course literature on the social work programmes has to be read in English. However, there is no doubt that students in the first project and child protection social workers in the second had different levels of confidence in their English skills. The stipulation also probably excluded some possible subjects. To use Mangen's (1999) phrase, some 'tactical sampling' was imposed by deciding to use only English.
For the research team, an awareness of language issues and their possible impact became a fundamental part of the work. In the interviewing and in data analysis work, the ramifications of choosing to work in a single language were foregrounded: particularly whether the quality suffered (Zulauf, 1999). In reality, most interviews consisted not of an interviewee speaking, or struggling to speak, in English, augmented by a lively and creative group process of negotiating meaning through trial and error. For example, in Italy, interviews were undertaken in groups. This allowed participants to help each other when they ran out of words in English and introduced discussion and dissent as to what was 'really' being said by any one of them. Helpfully, negotiated meanings emerged ('No -she means . . .', '. . . no she doesn't, it's more like this . . .' '. . . no the Italian word she used is nearer to this one in English . . .').
However, the research team were also conscious in one example of limitations. In one of the initial Swedish interviews, the student expressed unease with her language limitations diminishing her ideas. For pragmatic reasons, a focus group was impossible here, so the support with and refining of meaning that happened very naturally in the Italian group were not unavailable, to the detriment of the material and the frustration of the student. The remainder of the Swedish students had very good levels of English and were satisfied with their capacity to use it.
The research team allowed the edges of their role as dispassionate interviewers to blur a little by also intervening with language struggles in the interviews, where appropriate and where they could. All the members of the project group had a little knowledge of Italian and/or Swedish (Bowler, 1997, in Mangen, 1999, and of relevant concepts in these languages that could help the students express themselves. This was found to be an encouraging and useful resource in expected and less obvious ways. Certainly it was useful that the student could be prompted past a specific block in expression but also, and less predictably, the 'model' of the uncertain researcher struggling with language 'gave permission' to the interviewee to continue with their similar struggle. The issue of language, second languages and meaning in comparative qualitative research emerged in this work as both pivotal and full of contradictions. Built into the structure of the research itself is the belief that doing qualitative interviews with students and new social workers -project 1 -will be able to catch their narratives of experience and understand their meanings of being a newly examined/qualified social worker. A core strength of qualitative approaches is to 'reconcile complexity, detail and context' (Mangen, 1999: 110). Using a second language, can severely curtail the nuances and the detail. For two thirds of the sample, their capacity to construct a detailed version of their complex experience and understandings was circumscribed.
After project 1, the team developed their ideas in relation to approaching the dilemma of language. The second project experimented with some different approaches. In project 1, the 'local' researcher had been excluded, because in some cases, they were also the tutor/lecturer for the student sample being interviewed. The team were concerned the interviews would be compromised by the power dynamics of having one's lecturer present, especially as questions of how students perceived the effectiveness of their social work education were addressed. However in the second project, this power relationship did not apply. These were qualified child protection workers, and though some may have been ex-students of a researcher, few were. Taking into account the above and drawing on the experience of other researchers (e.g. Baines and Cunningham (2013 -see above), the interviews were conducted with the 'local' and international researcher present. This further facilitated negotiations over language, meaning and understanding in the interviews and data analysis. The productive negotiation of meaning in a focus group, described above, might also include the researcher offering possible interpretations: 'did you mean this?' '. . . Or more like that . . .?' Having local and international interviewers in the second project also made the role of the researchers more viable. Four out of five, interviewers did not have English as a first language (similarly, in every interview, one researcher was, culturally, the novice: a point returned to below). This in itself needed careful orchestration in the research process. In an obvious way, intellectual skills and academic 'performance' may be compromised when not using a first language (Temple, 2006). More specifically, interviewing represents particular challenges as well as the advantages touched on above. The interviewers with less confidence in their English skills cited the semi-structured topic guide as invariably helpful -a linguistic as well as comparative framework to organise material. Additionally, the inevitable process of co-producing meaning in the interview between researcher and researched (Kvale, 2007), alluded to above, was acknowledged within the team as transcending simple language differences. What the interviewer brings to the co-production of meaning in any interview is always individual and language is only one dimension of this. A common structure for the interview secured some stability, comparability and sense of coherence to this fluid situation.

Language and the research team
Moving on to consider language, meaning and the analysis of collected data represent a shift now in the paper to thinking more intensively about the international research team itself, in qualitative comparative research. It might be worth noting that this was a team with a great deal of history of working together on e.g. EU Erasmus projects and of writing together (1) (2).
In relation to making sense of the projects' data, even though the interviews were recorded and then transcribed into written text by the researcher coming from the country in which the interviews had taken place, all of the research team analysed all the material.
The local researcher could offer specific knowledge e.g. with unknown concepts, or when at times the respondent had lapsed into their local language to express themselves. For instance, the Swedish researcher used the word 'school' in order to talk about their social work education, but the English students interviewed did not understand that as meaning 'university' and the next section of the transcript initially seemed to be disconnected and somewhat irrelevant. The English researcher was able to fathom this.
Words and concepts could be clarified 'locally', but the issue of meaning making in the data analysis was as negotiated and dynamic as the group interview process, outlined above, had been, in which tentative and partial ideas were constantly reformulated through group discussion (Quilgers et al., 2006).
The process of data analysis in the whole team also produced diversity and negotiation not just of language but of the core themes emerging in the data. What, singly, seemed noteworthy could be commonplace, and what went unnoticed could be remarkable, given other (national) perspectives.
The level of challenge to and co-construction of interpretations of meaning was facilitated by the team context. However, it also promoted engagement with a further significant dimension of researching in international groups: the issue of power. Initially this seemed to be a concern specifically about language: the differential authority attained in the research group in relation to relative skills in handling English. Groups, though, are complex and multidimensional, and the interpersonal dynamics themselves are vulnerable to power relations, and 'rational' issues segue into the relational, impacting on how a research team will be able to function (Bion, 1968).
Many authors have discussed the 'colonial' aspects of using English as a first language for research (Ortiz, 2004). Although not entirely avoidable, in this familiar research team this potentially sensitive issue was confronted, in an attempt to avoid inadvertently instigating unequal power relations or indeed introducing an analytical hierarchy in the data work. The working group's years of involvement had included situations in which reliance on each other's expertise, in skills including but not confined to, language had featured. Complimentary academic expertise: in social policy for one participant, evaluation and therapy in another, theory and relational dynamics in the third, meant the roles of expert and apprentice were spread evenly. Just in terms of research, the Swedish participant here had by far the most experience and knowledge.
However, language 'expertise' had tended to accrue to the English member. She had in the past proofread and/or advised on written work from the other two. In order for this imbalance not to impact on the research process, this became part of the group's reflexive approach, and was explicitly discussed throughout the project. It also induced a level of self-policing to avoid 'playing the expert'/'playing the novice', in terms of language and meaning.
Issues of 'authority' were kept under review, in relation to language. The ongoing discussion facilitating this also surfaced a range of other observations and understandings of the workings of an international team engaged in international qualitative research. The next section offers some tentative thoughts on these.

Working internationally in groups
Small-scale qualitative research in itself, and additionally undertaking this in an international team, demands a high level of collective reflexivity, facilitated, in this team, by shared past experience.
Mangen captures our position succinctly: To summarise, the strengths of qualitative approaches lie in attempts to reconcile complexity, detail and context. Critical for protagonists is the integration of 'reflexivity', by which is meant the ability of researchers to take stock of their actions and their role in the research process. This is a particularly urgent task in research crossing cultural boundaries. (1999: 110) Continuing the discussion in the last section above, the, perhaps most obvious, starting point of any reflection within an international team is in relation to language in the broadest sense (including discourse, meaning etc.) and power dynamics. In this research project, English is the core language, and the only common language, and in itself has different rational and affective resonances for each member. The English speaker on one hand barely has to think about language: as stated above, can be effortlessly knowing and easily fall into being 'in charge' of meaning and indeed action: writing proposals, interview guides, papers etc. On the other hand, having ordinarily English incompetence in all other languages is shaming in this otherwise multilingual group. The Italian can also speak English and French, and manage in Spanish and German, with a little Swedish, and can, if only briefly, address a conference almost anywhere in Europe. For the Swede having excellent English means academic and geographical freedom. It is the language in which professional worlds are open: status publications, research bidding, world conferences, academic visits to the USA, Africa, Australia, Hong Kong: the 'glittering prizes' of the academic world require English. For the Italian, English also offers access to some of this, but perhaps more ambivalently. Like much of Southern Europe, the ease of acceptance of English as the lingua franca is less nationally subscribed to than Scandinavia; academic success abroad and nationally less seamless. The learning of English is not so embedded or taken for granted: more of a struggle. Southern Europe is often excluded from research projects for precisely these linguistic reasons.

Language and identity
The above considerations of language permeate issues of identity: national, professional and individual. This would be expected, indeed unavoidable, in some versions of identity theory. This is too vast a subject for such a paper as this, but it may be worth registering how, within certain perspectives, language and identity are indeed mutually constitutive. For example, post-structural approaches to identity would have language and discourse as determining (Hall, 1992). A social work professional's identity is positioned by discourses of caring, welfare policy, gender, economics, philosophy and populist/media stories of 'do-gooders' and 'idealistic lefties', etc. These are all the ways in which social work is defined and spoken of, reflecting hierarchies of power, in a particular culture at a particular time. Even in the kinds of modernist identity theory, social work has traditionally utilised far more, language still has a pivotal role in (volunteristic) identity construction. Symbolic interactionism from Goffman to the present requires the ongoing construction of an identity (narrative; reflexive; presentational) in which language is definitional. For example, how can you present yourself -in a Goffmanesque version of identity as self-presentation (Goffman, 1959) -if you can only express some limited range of yourself? How can you be witty, erudite, deep and/or sparkling, without a sophisticated vocabulary? Everything gets oversimplified. Within the research team, Goffman's work was familiar, and helpful for structuring reflections on identity, language and the team. A consciousness and carefulness in these matters were attempted. The team aimed for openness towards language-power dynamics, to raise them and give them serious consideration. The easy binary of 'intellectual colonialism' versus 'international engagement' has been under constant and dynamic challenge and even the affective elements -e.g. how a limited choice of words can make you feel stupid; how insecure only speaking one language renders you -shared. The openness of the team in surfacing and examining the meanings of language for themselves also of course rehearsed the language-power dynamics in the interviews, allowing for better preparation in terms of skills and sensitivity to the interviewee's struggles around language. Language invariably represents difference in a range of ways: in international teams and their research, but in this project, attention was also given to identifying commonalities and/or equivalencies (e.g. of expertise) with the aim of everyone developing dynamic research identities and becoming credible international researchers.

Commonalities
As research teams vary widely, equally will commonalities and the relative value with which they are perceived. To offer some illustrative examples of the work under discussion here: firstly, that ethical and theoretical assumptions shared were important. For example, in relation to establishing both approaches and inspiration, the team shared an interest in contemporary research writing around cross-cultural working, and an agreement that this work proved as useful as the more standard texts on cross-national research. The former tends to foreground issues such as meaning, reflexivity, ethical engagement, participants 'voice' etc., which seemed compatible with language/meaning issues and the social work context of the project (e.g. Liamputtong, 2010).
Secondly, the researchers were privileged to have a high degree of disciplinary commonality (social work primarily, with secondary interests in social policy and sociology) shared experience (trained in and practitioners of social work prior to undertaking doctorates and academic careers) and similar current work role (prof/ assoc profs, teaching, researching and writing currently). This is far from necessarily the case in many international teams, where often there are interdisciplinary diversities as well as international and status ones, a problem identified in e.g. Lewis and Brannan's work (2012). Overall, there was a very high level of shared understanding to counterbalance language differences. Also (see below) the team had experience of working in each other's systems and hence some understanding of core differences in education and service delivery. Perhaps equally importantly each member had the confidence of expertise in their own areas, and respect for the expertise of the others.

Developing cultural knowledge
As with the terms 'identity' and 'language', at this point in (intellectual) time, volumes could be written on the subject of 'culture' and its meanings, and here there is only the space to briefly draw on some of these ideas. As has been argued previously (Frost, 2008a) in relation to 'national' but also all other demarcators of identities, all aspects of culture, including cultural competence and intercultural being, are: processes not essential traits; fluid not static; constructed and reconstructed though systems of meaning and language; contested not settled (Hall, 1992). The extent that the notions of 'national' and 'cultural' both overlap and disaggregate is equally complex and cotested (Sarup, 1996).
The research team here over more than a decade have become 'internationalised', and particularly 'encultured' in the three countries represented.
In a variety of ways, and sometimes through the conscious intervention of the other team members, some cultural knowledge and cultural competence have developed. For example, independently, study visits, language learning and student projects have been undertaken. The team collectively has generated opportunities for each other to develop, by extended visits, exchanges, conversations, hospitality etc.
Via processes of mutual constitution and identification interests in each other's countries: e.g. the politics, books and films and music (very bondingly), the food, the news have been pursued. Acting as conduits for each other's cultural knowledge, by offering opportunities as various as advice and help with leisure and holiday choices to invites to teach have extended this further.
In terms of commonalities and culture, language and identities, then, the reflexive team approach overall has allowed for a greater understanding and a greater trust (in the sense discussed by Malina and Easterby-Smith, 1999) as having a key role in cross-cultural teams), more capacity for flexibility within power relations, taking different positions at different times and respecting each other's expertise. This certainly seemed to improve outcomes. Not just did a reasonable standard of nuanced and robust findings from the collected data ensue (clearly the primary aim of any research) but also here the sensitivities accessed within the team in relation to e.g. securities and insecurities around language, confidence and authority and so on, were not just respected but then productively taken into the arena of interviewing itself.

Taking reflection back into research interviews
This final section develops the point that learning from within the research team impacts on the research itself. What difference did the awareness gained from high levels of reflexivity, discussion and experience of working cross-nationally in a small team make?
Through examination of individual and team concerns, a consciousness was developed of the ways in which social processes of defining 'insiders and outsiders' inflect status in academia and within groups. These terms: 'outsider'; 'foreigner', as with other core concepts in this paper (e.g. identity; culture) have a wealth of poststructural and postcolonial literature underpinning their contemporary usage. This paper only has the space to comment that these are fluid, contingent and socially constructed terms of much contestation (Sarup, 1996). Their relevance is evident in research, where this construction may be exacerbated by 'foreignness' as a reality, as well as a metaphor. Here the team wanted to avoid the problem identified by Zulauf in her international research of being seen by the institutions she wished to research in as the 'foreign' researcher, unknown and with no particular institutional reputation, and therefore struggling to gain access (1999).
In these two projects, the role of the 'local' researcher in their own university or regional social service department was to render the 'foreign' as known. Emotional and physical access to the students was provided in ways which to some extent ameliorated a sense of outsider 'intrusion'. The local researcher personally asked their students to be interviewed, set up the space, met them, introduced them, provided biscuits and coffee and said things like: 'This is Helen and we have been working together for more than 10 years; she works with a similar group of students to you, and has also been a social worker like you', the subtext of which is something like: 'she's not a ''foreigner'', but one of us. I can vouch for that'. Secondly, using the combined techniques of cross-national interviewing, explained above, and informed reflexivity, the analysis and discussion of the research data were more fluid and open, more challenging and dialogical than if the research had only been conducted locally in one country, and each team member had only bought national material to the table. That each researcher in some context was both an 'outsider and learner', as the title of this paper highlights, offered more room for discursive meaning to emerge rather than be unilaterally fixed on.
Previous knowledge of students from each culture as well as the interview material meant a tentative discussion mostly avoiding attempts to oversimplify or overgeneralise. There was an informal cultural interpretation function already established within the research team, which helped with establishing a sense of meaning. The team were able to offer some confirmation, express surprise or comment on the relative aspects of what seemed to be emerging from the interview material. Examples from the research are things like Italian students' self-motivation, Swedish students' professional pride, the English culture of complaint.
However being foreigners, and therefore subject to different academic and practice systems and expectations, also had a range of practical difficulties. For example in the studies under discussion, one such compromise relates to the access to interviewees. In the first study, the team had agreed on a broader sample: larger numbers from each of three universities involved. However because of international differences in patterns of teaching and learning, when one of the researchers became available in a second country, term was already over, and in another, on the very last course of the whole program. That there were no other reasons to attend college when the researchers could visit, meant that students had to attend particularly for these research interviews, which severely limited the number of participants who could be recruited. As discussed above, that interviews where in English also impacted on the sample of interviewees were available to the project. One of the solutions mentioned in the literature (Zulauf, 1999) is that the researcher's institution and/or reputation may help to facilitate access. So it did in this case. The fact the researchers were known, researchers/teachers brought some trust to the students who choose to come: the lack of status of the 'foreign' interviewer referred to by Zulauf in the previous section above was overwritten by the reputation of the local researcher. The numbers were still small, but not negligible.

Conclusion
Comparative research in itself is notoriously complex and 'imperfect'. The reality is often hemmed in with all kinds of compromises and accommodations. There is a far greater gap between the ideal and the reality of research (Hantrais and Mangen, 1996;Zulauf, 1999). The aim of this paper has been to critically analyse the struggles, compromises and learning of the reflexive international team in undertaking qualitative comparative research. It has presented some thoughts on the learning undertaken by the team in relation to small-scale research projects on preparedness and survival in social work practice, in Italy, Sweden and England. It focused particularly on issues of language (and languages) and the construction of meaning in research: both in the focus group and in the research team. The dynamics and development within the team itself have also been given some consideration, in terms of identity and culture. Paired working practices have been considered, as has 'foreigness' in international research projects and its potential mitigation by a more inclusive team approach. Certainly, the paper illustrates throughout the comment above: that compromises and accommodations are the meat of such international research. However, and despite the evident limitations discussed, the potential learning from such work more than compensates.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The two particular comparative research projects that generated the observations this paper discusses, were undertaken in Sweden, Italy and England. Both were financed (with seed money of 4000E for each) by The European Association of Schools of Social Work page 3.