A Qualitative Methodology for Minority Language Media Production Research

Identity in Minority Language Media Production

In minority ethnic and linguistic media production, journalists, directors, and other participants are at once professionals and advocates (Husband, 2005). In their production settings their professionalism and identity go hand in hand to define their media output. How these characteristics are to be understood in the context of minority media production becomes relevant to understand the type and impact of media. To try to answer these kinds of questions about the relationship between identity and media production in a minority media production setting, the research methodology to be used has to take into account the contextual, interactional, and implicit ways in which these elements are construed by the production groups. The research methodology to apply, then, should try to be both descriptive and interpretive, and allow for participant reflexivity to model the interwoven structure of identity and media production.

The tools presented here were developed specifically for a research project on identity negotiation in minority language media, but their applicability surpasses that original objective. To that end, the following pages will present the construction of a set of methodological tools and explore its advantages and disadvantages. Following this theoretical debate, there is an account on the assessment of the tools after they were applied. The tools were used for a research project concerned with the negotiation of identity in minority language media settings, and it included the exploration of two groups of indigenous minority (Nasa and Wayuu) and one group of ethnic minority (Raizal Afro-Colombians in the Caribbean) radio producers in Colombia, and four groups of audiovisual television and multimedia minority language media producers in Wales. The main purpose of that research project was to understand how different identity allegiances—to language, territory, traditions, specific art forms or cultural expressions—were prioritized or abandoned for the sake of others in the media output.

When Browne (1996) and Cormack (2007a) pointed at the lack of empirical information on the positive effects of media for language maintenance, they also explained the difficulty of finding a way to get this information. Both Cormack (2007b) and, more recently, O'Connell (2013) have presented a set of possible questions, mainly concerned with the production groups themselves, but some of them geared for the audiences. This research project wanted to work alongside the former, admitting that the latter should also be addressed. The project was concerned with how different identity markers are chosen by media producers and prioritised in minority language media output. The question at hand was, thus, two-fold. What do they understand their identity to be, and how does their media production reflect, or evidence, this identity. To reach that answer regarding identity allegiances, it was necessary to present a research methodology that allowed the interviewees to reflect on both their own perception of what it means to be one of them and how that image is applied when producing/broadcasting content. Thus, the methodological tools needed to be focused on a qualitative and interpretive perspective of identity, rather than a more simplistic numerical factoring usually seen in Census polls.

Following that line of thinking, the first part of the article addresses some basics of qualitative research, and then moves on to a description of the tools selected. Afterwards, the advantages and disadvantages spotted following the applications of the tools are presented, which leads to a final discussion on their further applicability.

Qualitative Approach to the Research of Identity Allegiances

Qualitative research is developed from a phenomenological interpretation that draws from discourse as a way of producing knowledge not limited to the collection of data, but based on exchange and dialogue as bases of participatory construction of knowledge (Servaes, 2006). This poststructuralist view advocates that knowledge is constructed as a social process and that its claims of truth are based on representation, and as such the researcher can only access a version of the truth from a determined vantage point that can be presented for others to consider (Gray, 2003, pp. 20–21), rather than claiming to have found the ultimate truth on a subject. In other words, interpretive research allows for the creation of a narrative, since qualitative research is, in the end, a matter of telling a good story (Holloway & Biley, 2011).

Servaes (2006) defines qualitative research as a “human approach to the understanding of a human world” (pp. 940–941). As such, it aims at answering the questions of how people behave and operate by analysing experiences of individuals or groups, interactions and communications in the making, and traces of experiences and interactions (Flick, 2007, p. ix). This proves especially useful in studies that aim “to examine instances of negotiations of identities that are not necessarily limited to code-switching and to explain what identity options are available to speech event participants, what shapes these options, and which identities are being challenged and why” (Pavlenko & Blackledge, 2003, p. 10).

Furthermore, qualitative approaches may include ideas of participatory research. This interest requires a willingness to question the assumptions made upon the relevance of the research topic, and it presents a constant challenge in the evaluation and assessment of the results of the research process. This is borne in mind acknowledging that “our sometimes tortuous pathways through university administrivia and bureaucracy should be underpinned by the central recurring research question: whose story is it?” (Meadows, 2010, p. 311). To overcome this inherent bias, the methodology adopted needed to provide an opportunity for participants to debate, and not only respond to, the assumptions made by the researcher about their interpretation of their actions. Adding such a contention within the methodology satisfies both ethical and academic responsibilities of the researcher.

Selecting the Methodological Tools

There are a variety of tools at the disposal of the researcher when engaging in qualitative analysis. Silverman (2004) has presented and classified four major methods of qualitative research (p. 11). Two of them, namely observation and interviews, are of especial interest here. They are explained briefly, in order to set the ground to elaborate the points of the methodology presented.

Observation

This method has been used by different scientists throughout history as the first approach to gathering information about the world (Babbie, 2000, p. 258). Observation may prove useful because it allows for the naturally occurring events to take their action, while the researcher concentrates on honing his or her senses to capture as much information he or she can through them. Although it is difficult to ensure that these observations are not tainted by the cultural and academic structure of the observer (Rosaldo, 2000)—the observer bias, it is still a reliable method to provide direct information.

The advantage of observation, especially when researching human beings, comes from the fact that our senses work in more or less the same way as those researched. Common disadvantages of observation are a bias due to the standpoint and the lack of availability of the appropriate set of interpretive tools ahead of the observation, including linguistic skills and overall cultural knowledge of the practices (Babbie, 2000, pp. 264–265). This latter argument is of special relevance here. Although all the interviewees were competent bilinguals—speaking both their minority language and the surrounding majority language, it was through the latter that all conversations took place. This hindrance is not directly surmountable, but their fluency and competence in the majority language enabled them to communicate freely. However, it is important for any further reflection to bear this linguistic distance in mind.

Interviews

Interviews allow for access to the social worlds in which the interviewees are situated (Miller & Glassner, 2004). This takes the interview process as a collection of different voices and contexts for the production of personal narratives, including sometimes oppositional and contrasting views (Tanggaard, 2009). As such,

When interviews allow for exchange, rather than extraction, a dialogue is established, enabling a collective construction of knowledge (Holstein & Gubrium, 2004).

The main advantage of interviews is that they enable interaction between those engaged; they give the respondent the possibility to adjust or modify questions as they are being asked, refuse to respond to them or avoid them altogether, and even counter-question the interviewer. Although the degree of freedom in an interview depends on the specific interest and willingness of the interviewer, it becomes an open method that enables the interviewee to expand on a variety of issues, reach ample detail, and even expand on the aspects that he or she may consider more relevant to debate. It is also a way to reach the personal perspective on issues, especially in the case of one-to-one interviews.

This methodological tool also has a series of disadvantages. Interviewing processes set the two members of the conversation in different positions of power. Because of this power difference, relevant issues concerning language register, formality, convention, and linguistic skills (including language) may have an effect upon the result, leading up to the infliction of symbolic violence (see Bourdieu, 1999). This issue was relevant, insofar as the researcher had to avoid overt or subtle expressions of disgust or discomfort regarding deviation of standard formal speech in the majority language. Common informal and syntactically confusing forms were used in the Spanish of all interviewees in Colombia. This was quite different from the Welsh case, where there were no evident mistakes, apart from a few unnecessary fillers—such as “you know”—in one of the interviews from the Ceidiog production team in Pontypridd and a few borrowings from Welsh in one of the interviews from the Stwnsh Sadwrn group in Cardiff.

Tool Appropriateness

Interviews allow for a self-reporting which enables access to aspects of identity negotiation to be elucidated. However, “the existential presence of a complex hybrid identity structure is not necessarily matched by a personal reflexive awareness of this complexity by the individual” (Husband, 2005, p. 466), so direct questions about identity allegiances (e.g., What is your main identity?) can be considered pointless.

A useful way to overcome this problem is to follow an initial interview with participant observation. Original reports can be corroborated or ascertained, and other issues, such as over-or under-reporting, can be contrasted. The in-situ observation provided information which was noted while accompanying the producers in their everyday practice. An extra challenge was the fact that languages used around the researcher were not always intelligible to him. Not belonging to any of the minorities to be interviewed, the position of a researcher who comes from outside is far more evident, and breaking through to those being researched may prove difficult. The literature on the topic mentions plenty of cases of the difficulties of researchers as outsiders or distanced from the research subjects (see, for instance, Castillejo, 2005; Gray, 2003, pp. 33–34; Hallowell, Lawton, & Gregory, 2005, pp. 42–109; Kvale & Brinkmann, 2009, pp. 74–76), and how our own position may have an effect upon how we analyse the information received. In fact, even in the case of the practice itself, although previous media experience may provide a point of entry to a media production group, it may also add challenges to the ethnographic practices of observation (Patterson & Zoellner, 2010).

These obstacles may be overcome by asking questions to the producers and, in cases where questions cannot be asked, noting as much of the situation as possible. In order to overcome the outsider's bias, a space for debate of the observations of the researcher during participation in their activities could be introduced. This bias comes from the specific interest the researcher has on the subject. “Participant's ideas of why we, as researchers, research certain subjects is therefore of key importance when considering who we interview, and why they might want to be interviewed” (Riach, 2009, p. 369), and conversely have an effect on the information provided by them.

A second set of interviews can then be used to assess this circumstance. Through the response to the observations, possible contradictions, and arising questions made by the researcher, the interviewee has the opportunity to reflect on and evaluate the observations, contest them if necessary, and understand the interpretation process led by the researcher. “The negotiation of meanings and the particular constellation of relationship between interviewer and interviewee are of paramount importance for the meanings produced in a qualitative research interview” (Tanggaard, 2009, p. 1509), because they can help overcome the interpretive bias of the researcher and allow for cross pollination of ideas.

This collective production of knowledge takes place because “any interview situation – no matter how formalized, restricted, or standardized – relies upon interaction between participants who are constantly engaged in interpretive practice” (Holstein & Gubrium, 2004, p. 155), and “the analytical purposes for which research interviews are designed may not be shared by participants” (Bangerter, 2000, p. 459), thus constant debate and interpretation are required.

Furthermore, this way of engaging with research fosters a commitment to a subject-oriented perspective which “leads researchers to collaborate with, rather than investigate the practices of, the study participants” (Pavlenko & Blackledge, 2003, p. 25), and implies a responsibility to share and debate findings made by the researcher with those who take part in the research practice.

Although it seems impossible to overcome all biases, dialogue and construction of rapport with the interviewees allow for research subjects to contest the position of the researcher. Reflexivity becomes a key issue. Those difficult moments during interviews when the researcher's role is brought to the fore as unable to comprehend, as unable to participate, or as resembling the opposite of what is discussed can be defined as “sticky moments” of the interview. Those sticky moments actually “allow us to consider the ways in which research subjects acknowledge and consider their own positionality or biographically created knowledge in relation to their dialogue and practice” (Riach, 2009, p. 366).

The Three-Step Approach: Taking the Theory into Practice

Interviews were selected as the main research tool for research on identity allegiances in minority media production, but they were complemented by an observation of one day in the production environment. The proposed research methodology followed a construction of three steps to the interview process. The three steps are outlined below, and they were undertaken after an initial informal talk between the researcher and those researched. The initial talk was meant to build rapport and give a general overview of the research process, by explaining the purpose of the project, the mechanics of the interviews, the provision of the informed consent forms, and the general guidelines. The three steps were:

Interview to the Double

Although “Interview to the Double” might sound odd, I have used the exact wording presented by Nicolini (2009). On the first step, the objective is to understand how the research subject identifies the practice of his or her production team. The process used to compile this information comes from the work presented by Nicolini (2009), which stems out of previous work by Italian occupational psychologists of the 1970s.

The interviewee is asked to present his or her work in a narrative, describing to the interviewer all the actions he or she undertakes in one single day whilst at work. The way they are prompted to do so is through imagining that there is a double that would come and take his or her place in their job whilst pretending to be him or her for the day. This type of interview is called an Interview to the Double (ITTD) and “offers a glimpse into the modes of justification and rhetorical resources members have at their disposition to make themselves accountable for different practical purposes” (Nicolini, 2009, p. 205); it also provides the interviewer with a general idea of how the interviewee defines his or her job and how the relationships of accountability and responsibility are seen by him or her. The ITTD grants a window into the process that would be contrasted later, during the participative observation. This first step is recorded and transcribed after the whole process, serving as a point of departure which will help organise the other methodological steps.

The overall potential of the ITTD is increased when combined with a participant observation approach, because it may be laden with idealized moral accounts. Nicolini (2009) describes three of the idealized aspects as follows: (a) “the interview elicits the minor common denominator among local practitioners, the aspects of the object of work which are less controversial and less likely to be contested”; (b) “the interview articulates some of the main practical concerns which govern the activity and give the practice a projected directionality and sense”; and (c) “the interview taps into the local lexicon of accountability that novices have to learn in order to produce activities which are observable-reportable, that is accountable” (p. 205).

Thus, ITTD not only provides an idea of how work is undertaken, and how it is expected to be carried out, but also gives a general idea of the aspects of the job which are considered unproblematic.

The specific information given by participants will clearly vary. It is precisely the variations between their specific descriptions of chronological or step-by-step actions where they may provide an account that goes beyond their actual specifics and includes motivations, correlations, reaffirmation of their positioning, amongst many other possibilities.

Since the variety of responses and the information they may include cannot be measured beforehand, there is no way in which this method can be piloted and its results assessed prior to the actual application. Also, research subjects do not come to the interview from an information vacuum about the research project. They have all received the ethical guidelines or a similar form of informed consent, which includes information about the intended objectives of the project (i.e., identity negotiations and language output), and they have received an introductory talk by the researcher. Furthermore, since the linguistic repertoires of those interviewed are always available to them, and their linguistic interests (especially when recorded) include giving visibility to their languages, their own language may make appearances in their rendition of the interview, creating questions for the researcher about the transcription process and how to interpret this information.

Moreover, the interest in the ITTD is that it evidences the aspects of their daily tasks which research subjects consider unproblematic, and those to which they assign specific relevance or importance. The highlighted items are contrasted and contextualised with the observation of production practices.

Participant Observation

The second step includes the participation of the researcher in one day of the production practice itself. It enables the researcher to overcome the bias of self-reporting, thus providing an external evaluation of the actual process undertaken in the media (Cottle, 1998). It is this contrast between the self-reporting ITTD and the observation which proves extremely useful to determine their identity allegiances, because the participants can show evidence of the information they reported as well as have to explain the circumstances in which the report differs from actual activity.

Concomitantly, by participating in the actual production process the researcher has the opportunity to present to the subjects the aspects of their activities which he deems especially interesting, allowing the participants to reflect upon them while continuing their work, rather than afterwards. This enables them to engage in reflexivity “through reintroducing the research subjects into the research process, thus allowing reflexive moments to be woven into the analysis” (Riach, 2009, p. 366). It also allows for an interplay of seeing as well as listening (as recommended by Silverman, 2004, pp. 61–63), which provides an option to assess evident contradictions between reporting and actual actions. Reflexivity also creates the space for debate about how their reporting represents their engagement.

Linguistic and cultural barriers provide certain obstacles for the research observation, making it sometimes impossible to understand the exchange. Clearly, linguistic impairment has a bearing upon the analysis of the information, but circumventions may be set in place to overcome this problem, as already mentioned above.

Semi-Structured Interviews

The third and final step was a semi-structured interview performed on a one-to-one basis or in a collective interview with all the available members of the production team. Although a group interview might allow for more interaction and provide a better picture of the internal relationships of the production team, it was not always possible to arrange for a production team to sit down and discuss their processes and allegiances, because it could have been disruptive to their practices.

Through this interview, the interviewees were prompted to discuss those moments of decisionmaking, about the relation between their descriptions of their jobs following the ITTD and the actual on-site actions undertaken. The interviewees were made aware of the moments in which the researcher had seen the action and negotiation of identity occurring and becoming crystallised, or reified, in the production process. Therefore, they were not caught unawares in the interview. This gesture takes a step “beyond the trend of grafting on tokenistic, self-indulgent or tick-the-box reflexive accounts and instead integrate[s] reflexivity into both our theoretical commitment and our analytical framework” (Riach, 2009, p. 367). The justification for the decisions made, as well as the description of how they were made, provided information regarding how identities are negotiated and how this negotiation eventually becomes reified in media products that are developed and broadcast.

In practice, since asking directly about identity is not useful, the interview covered specific points, rather than having set questions previously defined. Points covered included:

The observation provided the material for the questions, in combination with the issues presented in the ITTD. The interviews were not totally free, since they had a specific interest of information and they were not aimed at analysing rhetorical strategies, nor commonly occurring talk, but rather trying to uncover specific aspects of identity allegiance and negotiation.

Cross Referencing and Analysing Responses

The data obtained from the three-step approach was classified and organised through a selection of various instances. In the first interview, the attention was given to the structuring of the actions and processes (order in which things are done); the presence/absence of a linguistic divide (descriptions, if any, of when a specific language is used and why); and highlighted elements (stressed by means of repetition, reiterative statements, inflections of speech—such as increased tonality, fluctuation, etc.). This provided a basic map of what is done, how relevant it may be, and how much of a routine, or unproblematic, it is. Notice, however, that the speech could be analysed for tonality, expressiveness, or any other rhetorical trope; this is not always required and was not the case in this research.

Participant observation looked at how the information originally presented in the ITTD is actually engaged with, and the main product is a notation of incongruence or emphasis of those reported actions. It also describes those issues not normally covered by the usual routine practice. These issues were brought up on the spot, and referenced for the discussion at a later moment in the in-depth interview.

Finally, the in-depth interviews were used to highlight responses to the matrix of description of action and observation, following on the reflexivity brought into the process. The responses were interpreted and categorised into identification markers defined by the researcher. This categorisation process is heavily dependent on interpretation, and thus may be affected by the mentioned biases. However, instead of fighting against the interpretive nature of the research, the analysis was made available to the interviewees, as will be mentioned below. The categorisation into the already mentioned ascriptions (i.e., indigenous, ethnic, linguistic, professional, etc.) or any constituent elements of the same (i.e., territory, cultural difference, self-identification, etc.) allowed for a mapping of the production group's identification allegiances.

Revisiting the Methodological Tools

The tools mentioned above were used to undertake research about identity allegiances in minority language media production. Colombia and Wales were selected as the countries where the research project was undertaken. Three production teams were chosen in Colombia and four were selected in Wales. In Colombia, the three groups included two radio station production teams, one from the Wayuu and one from the Nasa indigenous communities, and one team of radio producers from the Creole speaking archipelago of San Andrés. Each of the groups consisted of at least two members. In the Welsh case, four audiovisual and multimedia production teams were selected, two of them located in Cardiff, one in Pontypridd, and one in Caernarfon. The groups usually consisted of more than four members, although seldom did all members participate in the whole process.

As mentioned above, I first had to undertake a reflection upon my own identity markers. Being a white, male, middle-class, urban, non-indigenous, academic, majority language speaker (Spanish in the Colombian cases and English in Wales), with almost negligible knowledge in most of the minority languages, had an obvious impact upon the interview settings. The impact was more evident in the reverence of the indigenous Colombian groups than the other settings. Clearly, the Raizal and the Welsh-speakers were more used to the academic definitions of research. However, there was good rapport with the Nasa and Wayuu, because I spent over a week with them, and knew their community from previous research projects. There was no doubt that my position was not invisible. As will be remarked below, Welsh producers often mentioned, and I quote from my interviews, “this is something you will find interesting …,” thus evidencing they had a specific interest in highlighting some aspects that, from their own perspective, suited my research needs. Lack of knowledge in the minority languages themselves hindered the possibility to undertake content or discourse analysis and check if their expected identity markers were present in their work. It further hindered me, in that I was not able to blend in more easily. However, as mentioned above, the interviewees' competence in the majority language was more than sufficient to communicate clearly.

Interview to the Double: Alternative Views

Although the intended purpose of the ITTD was merely to get a grasp of what the common tasks of the given production group were like, and how relatively unproblematic those processes were, the responses varied quite clearly from an hour-to-hour description of actions, to an elaborate speech about their perspective on all sorts of issues.

Whilst the Wayuu and the Nasa production groups, and three of the Welsh groups, answered with a detailed list of activities they did on the specific day, the interviewees from the Welsh production group Green Bay described a day from previous work they had done—highlighting their idea to provide me with what they considered “interesting” information, and those from the Raizal group in San Andrés opted to begin with an assertion of their linguistic identity. This can be evidenced by the following example of a reply to the ITTD, by one of the Raizal Creole-speaking radio producers, rendered here in Standard English for clarity:

This highlights two facts: first, the instruction (i.e., What would someone do, if they were to come and be you for a day?) is only seen as a directive, which quite easily becomes overshadowed by other interests related to what the person wants to say about the topic at hand; second, there is a given expectation of what the researcher may want to hear, dictating the information that the interviewee decides to bring up and how he or she decides to do so.

When prompted to describe their average day, to someone who would come to take their place pretending to be them for a day, respondents were faced with an introspection of what it is they do, and what is it about their job that they deem important to be conveyed to others. However, this sudden description does not arise from a vacuum. All participants were informed about the interest of the research, as stated in the informed consent forms, and they were clearly aware that language was an important issue of debate. Besides that, the researcher had clearly remarked an interest for the language of each community interviewed. Moreover, the presence of recording devices (i.e., a digital audio recorder and/or a video recorder) made people more aware of the fact that this information had the possibility to reach wider audiences.

The responses then varied considerably. Although one would think that it might have proven useful to stop the recording and ask the question again, trying to ensure a response closer to the intended description of unproblematic activities, I thought it was much better to let them express their idea of their everyday job within their terms, even if they went off topic. The result was far richer, because it included information which would have been otherwise difficult to understand. The passionate comments of some of the participants, like the instance highlighted above, and their evident self-awareness of their act of being interviewed, provided much material to understand how they position themselves towards the information they are providing. From clear advocacy towards the language, or against the problems faced by their linguistic or professional community, to the idea of providing “useful” information for the research, their sudden self-awareness during the ITTD positioned them towards the research as a social process.

Although one could argue that research subjects seldom carry that self-reflective identity formation when undertaking their everyday activities (Husband, 2005), they do know that the research process is taking place and that some information gathered from them would be considered of importance. To illustrate this point, look at the following passage from an interview with a member of Ceidiog, one of the Welsh-language production groups:

This type of sticky moment (far more present in the Welsh cases, and highlighted in bold in the quotation above) evidenced an interest on the part of participants to provide accurate and useful information to the research process. Rather than letting the researcher decide upon the information they provide, they set themselves as a first point of entry that determines the validity, importance, and usefulness of their very own comments. Because of this, they also tried to provide cases, quite evident in some of the ITTDs, beyond the remit of their average or standard day. Whilst off the point in the description of their normal, unproblematic daily tasks, they went on to describe problematic situations they had faced, including some self-reflection upon their current endeavours.

Another aspect worth noting about the ITTD was the space it gave for some of them to describe and define their own position regarding the state of affairs of their own media. In the ITTDs of the Colombian Creole production team, the fact that the radio station was closing down became an important part of the description of their day, since it eventually led to the problematic situation of losing their space for debate. This emotional response included more information about the position that the team members have towards their role and clearly sets their mood towards the issue:

In conclusion, the lack of straightforwardness in the responses was, by far, more of an asset than a flaw of the ITTD research tool. Although in cases where the only interest is to define routines not easily grasped any other way (e.g., in the impossibility of conducting observation for some dangerous or unreachable jobs), these kinds of results may be of little use; in the present research, their divergence from the topic brings up more information and allows for more self-reflection, thus making the research process not only a tool of inquiry but also a tool for promoting debate and discussion. This was, in fact, something that was remarked by some of the participants after the interview; the usefulness of granting them with a space to reflect about their job. In fact, one of the participants even made a positive comment in the ITTD itself about this self-reflection instilled by the researcher:

Participant Observation: Being There Not Being Invisible

There is much debate in the literature concerning participant observation, as was already mentioned above. Even more so, on how observation can be difficult to undertake when it is not possible to conceal your difference. Although it is granted that taking the observer role is an evident act, in some procedures, especially when the subjects of observation have been already made aware of it, it does allow for another space of debate and cross-reference. Since the observation here served more as a contrasting element than a description in its own right, lengthier observations were not considered useful, but instead observations of one single day of work were chosen, as previously described. This allowed for corroboration or contradiction of the subject-presented “normal day.” As such, it was the best way to develop the questions for the lengthier interview that would follow.

Arranging for the observation was considerably different in the two countries and with the various teams interviewed. The Wayuu people interviewed live in the north of the country, and the radio station visited is only reachable after an 8-hour drive from the nearest town through unpaved desert paths. The Nasa are easier to reach via public buses after a 2-hour journey on a conventional paved road through the Andes. The Creole Raizal live in San Andrés, reachable by commercial planes to the island. The Welsh production companies were set in Cardiff (2), Pontypridd (1), and Caernarfon (1), all served by trains and acceptable paved roads, but they do require long travelling hours. Whereas in the Colombian cases the trips implied staying in the area where the production team lives and having the visits immediately after the first interviews, in the Welsh cases time had to be put in between the first interview and the observation, in such a way that it would not interfere other scheduling compromises. In that sense, whereas in the former cases the observation was taken as unproblematic, and the teams interviewed had no problem letting the researcher follow their actions the next day, in the Welsh cases, the observer was clearly seen as someone who would have an effect upon their natural working processes, and as such, the observation instances were clearly scheduled for moments when they would not disrupt major undertakings. This predisposition by the Welsh production teams had, again, a dual interest: that of having the chance to allow for the observer to experience something “significantly useful” and also “least intrusive,” according to their perception. Since the participation of the production teams was voluntary, as was their information, and the level of trust was relevant for the responses, there was no interest in trying to go against these decisions and perceptions. In fact, allowing for them to provide these spaces was also useful to comprehend the way they understand research and the role of the observer. They were keen on trying to let the researcher experience something particularly interesting, according to them, while ensuring it was not an instance where the researcher would be in their way.

The Long Interview: All Together Now … Or Not

Precisely because the interest of the research project was to allow the production teams as much freedom as possible, the methodology was designed to accommodate for it. All of the teams underwent the second, lengthier interview. This interview was very open-ended, although it had a guiding principle. The interviews were never the same, and even though certain topics were similar in every case, they were addressed differently, following the entry points given in the ITTD and during observation. The interviews were carried out either one-on-one or as group interviews. In one case, the decision to make it a group interview certainly led one of the participants to become less engaged in the process, and that perspective was all but lost. However, this structured debate proved quite useful in the remaining cases. In one Colombian case and two of the Welsh ones, the group interview allowed for a collaborative debate and even contrasting views, which led to even further reflection by the collective engaged in the research process. Common statements, including “I don't know if the others will agree, but I think that …,” evidence how perceptions about the collective work may be shared, but may also be considered to be unacceptable by some members of the group. One very interesting instance where this happened was when one of the interviewees evidenced a fear in saying something about the identity allegiance of the company, with which the other members might disagree. Here is the contribution of one of the participants in the group interview of the Welsh production company Cwmni Da:

I'm working on something at the moment, and we're going to be encouraging one of the individuals to use English, mainly because the area that the individual is from is predominantly English, though people can understand and can speak Welsh. It wouldn't be natural for somebody from – I hope nobody, sort of, shoots me because of this –for somebody in Holyhead to speak Welsh as their first language. There are lots of people in Holyhead that do speak it, but I think that the majority can understand it, so we would encourage this individual to throw in a bit of English every now and then, because it would make it far more natural. (Long Interview, Cwmni Da, March 14, 2011)

Beyond the anecdotal, the group interview seemed to flow much better because it really set each production team as a group. However, the possibility to undertake this collective debate or a single-person interview depended on the requirements of the production teams. Not all of the teams' members were available at once, and sometimes it was impossible to meet with all of them together within their very busy schedules. What became quite clear is that group interviews worked much better, and should become the staple way of undertaking similar research endeavours.

Final Remarks on Methodology

There is no single error-free way of gathering information. In the specific case of trying to understand identity allegiances, the tools used here seemed to work well enough to gather data on the subject. As presented by Tracy (2010), qualitative research is all about evidencing the positioning of the researcher, and ensuring there is openness and clarity in the processes undertaken. Self-awareness and reflexivity have been an integral part of the process, and highlighting the positive as well as the negative outcomes of the methodological tools used is not only good research practice, but evidence of an ethical commitment to ensure the validity of the information.

Although there is very little space left here to debate the useful elements from the procedure, some of them have become evident in the few examples presented. Each of the production groups presented their own perspective of how they should use their language, or any other cultural identity marker, in their work. Through the ITTD – observation – reflectivity triangle, the groups were able to recognize their own positioning towards their media production practices. One final example that shows evidence of how the production practices of the media production teams have a bearing upon linguistic usage appeared in the final interview at Green Bay in Cardiff, after a question posed about the interchangeable use of English and Welsh by the staff in a Welsh-speaking production company:

The methodological set of tools presented here as a three-step approach were considerably successful at pinpointing the way in which identity, and language as part of it, affects media production practices. Although language as an identity marker was the guiding light for the research process that used this methodology, its applicability in other ethnic, indigenous, or immigrant minority settings seems evident. Following Husband's (2005) and Cormack's (2004) perspectives, learning about minority media production tells us a great deal about media production in general. The tools presented here were designed, precisely, with those ideas in mind.