TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA
75 / 2015, p. 27-101
Katarina Kreceva
Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
Academic literature rarely offers elaborative discussions on the researcher's epistemic choices. The central role of the researcher in qualitative research, however, makes these choices crucial for assuring the data viability and the validity of findings. This article depicts the process of epistemic constitution of the author's master thesis. It offers an insight in the dispositions of epistemic elements that generate a rather unconventional research design which arguably addresses the topic under scrutiny in most adequate manner. The aim is to show how the reflexive interrelations between the topic, methodology and theory influence the researcher's decision making in interdisciplinary studies and more particularly, within a given qualitative research design.
Keywords: epistemology, qualitative research, ethnicity, education policy.
Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologiji naproti
V akademski literaturi redko zasledimo obširne debate o epistemoloških izbirah raziskovalcev, vendar pa so te zaradi osrednje vloge raziskovalcev v kvalitativnem raziskovanju pomembne, saj zagotavljajo relevantnost podatkov in veljavnost raziskovalnih odkritij. Članek obravnava proces epistemološke zasnove avtorjevega magistrskega dela. Daje vpogled v dispozicijo epistemoloških elementov, ki zagotavljajo precej nenavaden raziskovalni pristop, ki izbrano tematiko obdela zelo natančno in na najboljši možni način. Namen članka je pokazati, kako refleksivni odnosi med temo, metodologijo in teorijo vplivajo na raziskovalčevo odločitev za interdisciplinarni študij ter pristop k raziskovanju.
Ključne besede: epistemologija, kvalitativne raziskave, etničnost, izobraževalna politika.
Correspondence address: Katarina Kreceva, Faculty ofEducation, University ofCambridge, 184 Hills Rd, Cambridge CB2 8PQ, United Kingdom, e-mail: kreceva.k@gmail.com.
ISSN 0354-0286 Print/ISSN 1854-5181 Online © Inštitut za narodnostna vprašanja (Ljubljana), http://www.inv.si
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
1. Introduction
This research has been inspired by the constraints faced during my four-year professional experience in endorsing education policy.1 I encountered the po-sitivistic, linear approach to be genuinely applied to all aspects of policy making and research, including development of policy making mechanisms. The linear approach following specific stages of a policy cycle such as: identifying the problem - offering alternative policy recommendations - policy performance monitoring and evaluation appeared as predominant in national and international policy making practices.
Even the most general overview of academic research on policy making, however, disputes this linear models as social management tool detached from various factors such as institutional cultures and individual agency. In contrast to such models, policy research conducted from within the broad spectre of social science disciplines show that the straightforward progression from a problem to a solution is never the case in policy making. Instead, the positivistic approach is often identified to yield paradoxes in policy premises, processes, and outcomes (Apthorpe 1996; Malen & Knapp 1997; Schram 1993).
When discussing education policy in particular, Stephen Ball (1997) in his review of recent education policy discusses a number of different epistemologies which originate from various conceptualizations, research designs and interests. He finds that the wide range of epistemic constructs render the landscape of the social more visible. However, he also concludes that a gap in social inquiry exists towards the underexplored question of what sort of people and 'voices' inhabit the texts of policy analysis...? According to Ball, it is of utter importance not only to explore the social effects of policies but also to "... attempt to capture the complex interplay of identities and interests and coalitions and conflicts within the processes and enactments of policy" (Ball 1997, 271). Evidently, this resonates with an earlier, more general note omnipresent in social structuralism stating that "a category is entrenched by its coherence to other categories... and the process of entrenchment is as much social as it is cognitive ... actions of such entrenchment result as much from the categories as from the understanding and experience the actors have of such categories" (Douglas 1987, 9).
The above stance appears to be particularly relevant in cases of strong affective and ideological dimensions inherent to notions such as ethnicity and inter-ethnic relations. Hence comes the idea of conducting research to the process of ethnic identity construction through education policy in the Republic of Macedonia from a post-structuralist perspective that has been undermined in conventional policy day-to-day activities.
The topic under scrutiny sensibly raises challenges for developing an epi-stemology, first, compatible to the existing technical limitations; and second, adequate to the research focus and the researcher's positioning. Capturing the
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
agency of identities and interests in policy enacting that simultaneously tackles the decision making processes of policy making authorities, requires an in-depth analysis of not only policy documents, but of dynamic institutional cultures and individual agency as well.
Of all the available research methods ethnography and anthropology come to mind as the most adequate due to the following discipline related discussions:
a)	Ability to incorporate abundant data
Given the time constraints contextually embedded in a developing democracy with under-regulated policy making mechanisms, it would be difficult to generate sufficient and reliable data through more structured social research relying on mixed methods or triangulation. However, increasing the range of information and the ability of crosschecking to contribute to the validity of interpretation could be achieved by exploiting complementary data sources. Combining different data sources appears as a common practice in anthropological research.
b)	Ethnography of institutional settings
The contemporary shift of anthropological interests from local cultures to urban, virtual or institutional settings has raised debates on methods and tools to research context-rich multi-layered cultures such as those embedding the policy making processes (Wedel, Shore, Feldman& Lathrop 2005, 39). It has been widely asserted that observation and interlocution as primary tools of ethnography carry advantages over other qualitative research tools for in-depth recording of the "fluid, fractured and complex postmodern nature".
c)	Acknowledgement of situationalism
The two main paradigms in ethnicity studies (primordialism and constructivism) agree that the variety in conditions and processes determine the variety of existing ethnicities. By heavily focusing on contextual specifics, anthropology acknowledges the importance of what is thus established as one of the main predicaments of ethnicity - situationalism.
Parallel to the above widely acknowledged advantages, the anthropologists of public policy also identify some commonly practiced anthropological means such as the "social network theory" and the "extended case method" which bring epistemic inconsistencies to the research field (Wedel et al. 2005, 41). Also, due to the central role of the researcher in field-note taking and data interpretation some methodological concerns have been expressed about the overly influencing speculative ability determining the adequacy of gathered data and the validity of interpretation (Sanjek 1990, Zeldich 1982).
Far from overcoming all the critique on the scientific rigor of anthropological research, the following discussion illuminates the process of composing the most case appropriate epistemology for addressing the topic under scrutiny. The arguments supporting the above listed advantages and concerns are elaborately
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
30
addressed through the constitution of the research design. The specific epistemic elements composed in order to avoid possible inconstancies are being justified through an extensive amplification of the researcher's choices.
2. Research Design
Sampling actors and activities out of the total, infinite flow and designing mechanisms to identify and gather relevant data is one of the crucial factors for ensuring data validity and sound research outcome.
2.1 Demarcation of the research field
Placing the major topic of ethnicity against the institutional framework for education policy making in the Republic ofMacedonia, only one institutional body legitimately embodies these two parameters. The Directorate for Development and Promotion of Education in the Languages of Ethnic Communities (in the following text: the Directorate) is a body within the Ministry of Education and Science of the Republic of Macedonia established according to the state's resolution for respecting the linguistic and cultural identity ofethnic communities. Its main responsibility is advancement and development of education in the languages of ethnic communities. According to the Systematization Act of the Ministry ofEducation and Science ofthe Republic ofMacedonia - the Directorate is chaired by a director and comprised of two departments: (1) the Department for Albanian, Turkish, Serbian, Roma, Vlach and Bosniac languages and (2) the Department for peace and rights of children of all communities.
Figure 1: Illustration of the research field (created by the author K.K.).
the Directorate
1 Director secondary informant
Department for Albanian, Turkish, Serbian, Roma, Vlach and Bosniac languages	B	4 advisers key informants (2 support staff - secondary informants)
		
Department for peace and rights of children of all communities	B	1 Adviser secondary informant
Source: Rulebook for systematization of job positions within the Ministry of Education and Science 2010, 170.
The institutional organization suggests that the most resourceful research field for addressing the research topic may be identified within the frames of the first de artment. This is mainl because:
- It is the only department holding a mandate for enacting policy related to education in the languages and cultures of ethnic communities;
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etnlčnostl skozi Izobraževalno politiko: najprimernejši eplstemologjl naproti
-	According to the Systematization Act the department employs representatives of all ethnic communities2 with consultancy and policy making qualifications and experience;
-	Compared to other organizational units within the Ministry this department is characterized by distinctively low staff turnover;
-	Hierarchically, the Directorate is a meso-level public body under a direct supervision of the Minister of Education and Science. Hence, it is a research-field affected by both micro-level and macro-level policy initiatives;
-	In addition to the features internal to the research field, the researcher has previously cooperated with several representatives of this department on policy related issues. This enables easier access to the research-field, otherwise difficult to obtain due to the existing transparency regulative and long-lasting administrative procedures. It also contributes to gaining trust crucial for this type of qualitative research and equips the researcher with a prior insight in the institutional culture and procedures.
All the above points to an exceptional informational potential situated in this department and its employees. The graphic presentation ofthe delineated research field illustrates that during the field-research there were six employees present in the department. Four were advisers and two were lower level administrative staff. The choice ofkey informants was limited to the four advisers. The other employees from the Directorate were included as secondary informants for purposes of data collection and further clarification and/or crosschecking of ambiguous issues. Although gathered field-notes include data on interaction of key informants with persons external to the research-field as participants in action speech, the research design does not follow the "extended case model" for reasons addressed further in this article.
31
2.2 Ethical considerations
Having delineated the research-filed, several major ethical considerations arise as relevant. Notably, some of these are common to anthropology of public policy, and some stem as case specific. a) Informed consent
The researcher's previous professional engagement has proven beneficial in gaining access to the research field. After submitting an official Letter of Request stating the goal and methods of research, the access was granted by the Minister's Cabinet. Written approval from the Director of the Directorate was also obtained via e-mail3. The researcher's intent to focus on every-day activities was elaborated extensively during the first encounter with the key informants. The request to informed consent with an opportunity to opt-out at any stage of the research process was clearly stated. Given the power relations and the segregation ofduties, it was important at the time to downsize
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
the authority of the prior acquired superior's consent and emphasize the importance of the informants' own will to participate.
All key informants agreed to the terms of the research and the overall experience was that there were no objections regarding information sharing during the field research. However, because the informants' previous experience in policy research was limited to the positivist paradigm and because of the unfolding research design, it is hard to determine to what extent the overall research corresponded to the informants' conception of the research design at the time.
b)	Anonymity and confidentiality (on record / off record information) Considering the public character of the informants and institutions put under scrutiny, journalistic ethics is commonly suggested as the most appropriate ethical approach in applying anthropology of public policy (Wedel et al. 2005). Although everything stated as on record and during action speech was clearly recorded to be potentially quoted or paraphrased, the informants are provided with a status of anonymity through all stages of research and analysis. Due to the exceptionally small number of key informants however, covering the informants' identities from the people familiar with the specific context was a significant challenge. For that purpose, pseudonyms and cover of specific information that might disclose the identity of any informant are used to the greatest possible extent.
Furthermore, while during interlocution certain information was consciously stated as off record, the observations have amounted potentially sensitive data. The off record statements and the potentially sensitive information were not quoted throughout the analysis. Since the time constraints did not allow acquiring informants' approval on the text prior submission, the sensitive character of information was arbitrarily evaluated by the researcher. For that purpose, field notes remain confidential and data protection is applied to all field-work material. Disclosure of documents is limited to those publicly available or obtainable according to the national Law for Access to Public Information in force at the time of the research.
c)	Authority of informants / legitimacy of public institution
Disputed findings in anthropology of professional environments appear resourceful enough for some anthropologists to publish articles on ethics and negotiation over discarded research results (Mosse 2006, Schwegler & Powell 2008). A viable alternative appears possible when the research develops as a cooperative process between the researcher and the informants. Such dialogical mode of ethnography attempts to build a non-hierarchical, non-manipulative research relationships as a solution to the different inter-pretational modes and challenged authorities (Sanjek 1990). However, because the temporal constraints in the specific case hinder such approach, the researcher has to take under special consideration the authority of the informants and the legitimacy of the institution.
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
With this on mind, during data gathering and analysis the researcher tends to be perceptive to the differences between the informant's and her own conceptual stance on ethnicity. When these differ, the issue is not which one is more relevant but how the two can contribute to the understanding of the ethnicity related phenomena. Also, while the researcher has the privilege of approaching ethnicity in a primarily contemplative manner, the informants' approach ethnicity in a predominantly applied manner. Therefore, admittedly, the choices made by the researcher in the given situation may not correspond to choices to be made by the researcher in a different context or from a different position.
2.3 The researcher's positioning
The positioning ofthe researcher is another determinant in qualitative research. In that regards, the history ofa co-worker with the key informants holds an ambiguous status. Sharing a historical and interpretative setting with the informants might be considered as an insider's advantage which ensures willingness to share genuine to the assumption of common understanding. This position is however twofold: On one hand, it equips the researcher with an insight which might increase the relevance of data interpretation without necessarily leading to conceptual harmony. On the other hand, it is a potential trap of overemphasizing shared factors between the researcher and the informants (Corbin Dwyer & Buckle 2009).
Many ofthe statements recorded during field work begin with "Well, you know yourself how it is ...", particularly those remarking the external constraints (higher hierarchy, budget deficiency, inefficient procedures) and those commenting the general state in education (underdeveloped infrastructure, low living standard, etc). The informants tended not to elaborate extensively the assumable shared experience. In order not to rely on personal assumptions, even in cases of such shared experience, incentives for further elaboration were used such as "What do you mean?... did something like that happened recently, since I have been away?, Would you name some specific examples?" However, at times, empathy was used for the purpose of initiating a conversation or a follow-up on sensitive issues, such as in the case of already stated difficulties: "I remember having similar problems. Could you explain how did it work for you? ".
Self-awareness about the researcher's positioning ultimately determines the perspective from which the researcher conceptualizes the informants' experiences, sees connections, causal patterns and influences that are internal to the experience. Practicing such awareness during field-research leads to more dialectical rather than to exclusively insider / outsider role (Burgess 1982, Corbin Dwyer & Buckle 2009, Zelditch 1982). The strengths and challenges faced in exploring this notion of the space between during this research have set an important line of approach.
During the research, the researcher's positioning has been shifting according to three major choices. First, was the choice of getting more immersed into the
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
researcher's role, instead of becoming more immersed into the membership role - avoiding going native. Thus, the observations during field-research were conducted without participative tendency (e.g. I positioned myself centrally in the office at the meeting table; I preferred conversing with informants about current academic experience over previous common professional experience). By undertaking these steps, the researcher avoids leading the informants into shared experiences and obtaining primarily confidential information, at the same time, being cautious not to cause staging of everyday activities.
The second choice, was applying theory guided ethnography in order to increase objectivity during data gathering and interpretation. The process of designing a customized observation framework and analytical tools based on a theoretical overview will be depicted in detail in continuation of this text. The third choice, was limiting the source of information (informants) to the delineated research field avoiding the extended case model. According to the designed data gathering and analytical tools, access to information not available to the informants themselves was identified as a risk to potentially mislead the researcher during data gathering and analysis (Sanjek 1990, Zelditch 1982). Because of that, explorations exceeding the research field were intentionally avoided (with exception to referenced documents) so to limit information only to data on which the informants base their decisions and actions.
2.4 Data gathering
Following the ethnographic method, data gathered during this field research were recorded in a fieldwork diary incorporating: direct notes from the field (action and non-action speech); in situ comments (observations made during field work); additional comments (notes on observations and speech exceeding the research-field); and lists and notes on gathered policy documents.4
In order to increase the adequacy and validity of generated data the research was structured in three phases:
Phase 1 - Initial observation (one week5): The first phase of the research incorporated unstructured observations and interlocution. Data were accumulated loosely - a sort of a screening on the general topic of ethnicity in enacting education policy. The general indexing scheme applied during note taking included: date / name; action / non-action speech; question / answer; unclear / dubious. Phase 2 - Data taxonomy & theoretical overview: grounded on data recorded during the initial observation, initial data taxonomy was developed to reflect the initial findings. The taxonomy was proliferated according to the identified patterns, particularities and exclusions. A literature review on ethnicity studies was conducted in light of these initial findings to pin down the succeeding research. Taking in consideration the theoretical overview, the initial data taxonomy was modified into a customised observation framework and analytical tool.
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
Phase 3 - Theory guided observation (3 weeks): The observations and interlocution during this final phase were semi-structured according to the designed observation framework and indexed in detail.
The three-phased data gathering system posed above builds on the awareness about the dialectics between the analytical induction originating from field-research and the theoretical deduction applied to the field-research. The goal is to create synergy among (l) the local every day interaction charged with context specific values and knowledge; and (2) the wider socio-political issues entrenched within the relevant theoretical discussions.
3. Initial findings - developing an initial data taxonomy
For purposes of epistemic constitution only data gathered during the first phase are indicatively analysed. With all the previously made methodological choices in mind, an analysis consequent to the initial research phase brings to attention the following general assertions:
1)	A peculiar state of affairs exists in regards to on-going activities. Several donor-funded projects in which the Directorate ought to participate have been temporarily withheld and at the time, the Directorate was not involved in other publically funded policy initiatives:
Xl: "No problem, we have time to talk. Nothing is going on right now though. We are waiting for the projects' implementation to start..."
2)	A common understanding among the informants is that this particular state of affairs is owed to institutional policy constraints such as lack of segregation of duties and under-defined responsibilities:
K: "Can you tell me why?" (*the //// project has not started yet)
Xl: "Who knows! We made the project and everything ... we are supposed
to implement it but they are keeping it on stand..."
X2: "And we don't know why. We're supposed to be in charge ..."
3)	Informal power-relations are repeatedly identified as a primary determinant for the current distribution of activities and responsibilities:
Z: [ ".Well, you know. //// is slightly distanced. They are rather closed (for cooperation). We are not working as before. We are waiting to start with the activities, but they are now in charge of coordination."]
4)	A genuine conviction in ethnic group rights and existing structural inequalities is expressed by all of the informants:
X3: "We (*the numerically smaller ethnic communities) come last. Nobody cares! We are most marginalized ..."
5)	The informants' conception of ethnic identity is demonstrated in primarily culturalist terms:
Xl: [".there is not much in those projects for the ethnic communities. It's mostly activities for inter-ethnic communication... Very little tradition ."]
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
When organized in an interpretative frame encompassing causality and relation, the above illustrated assertions indicatively synthesize the following data taxonomy:
Figure 2: Initial data taxonomy (created by the author K. K. duringfield research).
Initial data taxonomy:
-	data on conceptual framework on ethnicity
-	data on institutional structure / networks and policy (including informal power relations)
-	data on attributed responsibilities (informants / others) vs. actual activities
Source: Data generated from initial observations.
In light of the above findings a theoretical overview on ethnicity was conducted with a purpose of identifying the most adequate theoretical framework coherent to the research design and the data emerging from the field.
4. Literature review
Two confronting lines of thought have been building on the dichotomy among primordialism and constructivism addressing the question what constitutes ethnicity?
Speaking in general terms, primodialism asserts that ethnicity is: "... the assumed givens of social existence: immediate contiguity and kin connections mainly, but beyond them the givenness that stems from being born into a particular religious community, speaking a particular language, or even a dialect oflanguage, and following particular social practices."(Geertz, 1963, 110); and constructivism asserts that ethnicity is: "...rational association turning into personal relationship. If rationally regulated action is not widespread, almost every association, even the most rational one, creates an overarching communal consciousness..." (Weber 1978, 389).
Although in opposition, the above approaches have been amounting empirical research based on two common premises: (1) ethnicity is an existing entity: conventions of coding and comparing ethnicity are being developed on basis of enduring substance; and (2) ethnicity is situational: variety within ethnicities is owed to circumstantial fashioning (Brubaker 2004, Cohen 1978, Smith 2009).
In the following discussion I will argue that when these two lines of thought are confronted, they reveal methodological and conceptual commonalities, as well as common conceptual incoherencies which may lead to consolidation of arguments and bring in light an alternative theoretical view on ethnicity.
1. Ethnicity as a discursive phenomenon
A theoretical study on ethnicity in anthropological research conducted by Eriksen (2002) reveals that ethnicity as a term is used for various analytical
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
purposes having "something to do with classification of people and group relationships" (Eriksen 2002, 4). Another theoretical overview (Cohen 1978) comes to the conclusion that the term "ethnicity" is used rather vaguely, with research addressing as ethnic "almost any cultural-social unit... (any term describing particular structures of continuing social relations, or sets of regularized events) " Cohen demonstrates that the same socio-cultural phenomena formerly studied in tribes are later being studied in ethnic groups. The main difference being that instead of accounting for isolated non-western socio-cultural units such as tribes the entity addressed by the newly coined term ethnicity accounts for socio-cultural units defined in relation to other surrounding units, and therefore, arising from a dialogical relationship. The main argument is that the process of relational differentiation self / other introduces an emic approach and inquiry on self-ascription leading to re-examining and reformulation of the previously used anthropological terminology (Cohen 1978, Eriksen 2002). If so, neither Weber, nor Geertz, but Barth's Ethnic Groups and Boundaries (1969) may be considered as a cornerstone for conceptualizing ethnicity in anthropological works.
According to Barth, the term ethnic group in anthropological literature generally designates population which: (1) is largely biologically self-perpetuating; (2) shares fundamental cultural values, realized in overt unity in cultural forms; (3) makes up a field of communication and interaction; and (4) has a membership which identifies, and is identified by others, as constituting a category distinguishable from other categories of the same order (Barth 1998, 10).
By empirically analysing processes related to embodiments of ethnic groups, Barth concludes that their internal constitution results from the process ofarticulation and maintenance ofethnic boundaries. More precisely, the different ways in which ethnic boundaries "are maintained not only by once and for all recruitment but by continual expression and validation" is structuring the interaction which allows persistence of ethnic groups (Barth 1998).
Without getting into a broader discussion about how Bart's ethnic boundaries triggered a repositioning of anthropological interests, it might be asserted that the socio-cultural diacritics on shared identity according to group membership are being incorporated in the conceptual stands of both primordialists and constructivists. Markers such as kinship, religion, language, territorial contiguity, etc. are identified as ethnic group distinguishers in all theoretical approaches. The causality behind the appearance and persistence ofsuch distinguishes vary among the confronting paradigms. The arguments of both paradigms however dwell on applying theoretical and empirical analysis of these codes contextually, on specific ethnic groups - entities.
In pluralist societies, where the ethnic group is situated, the state is an omnipresent mean ofpolitical organization. Due to overlapping terminology
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
and academic interdisciplinarity contemporary ethnicity paradigms struggle over the ambiguous conceptual overlapping of ethnicity, nation, and race. Eriksen (2002) for example, exempts the study of race from contemporary scientific inquiry on the ground of human race being a result of continual interbreeding leading to hereditary physical traits that do not follow clear boundaries. Nevertheless, the author states that "concepts of race can ... be important to the extent that they inform people's actions"(Eriksen, 2002). If however one applies this ontological argument onto ethnicity and ethnic boundaries, it follows that ethnicity may also be seen as only a discursive phenomenon rather than a social / cultural unit with an enduring substance. Ifone asserts (1) that ethnicity dwells on the disposition ofselfvs. others, and (2) that ethnicity is not an entity, but a discursive phenomenon, it follows that ethnicity itself is contained in the discursive constitution of features which ethnically define the boundaries of self / otherness no mater weather such discursive boundaries are caused by social contract or inherent to the community.
2. Ethnicity as an event
Constructivist approaches argue that social structures and particular sorts of social relations (in most cases regulated by the state) ascribe and sustain ethnicity related categorical / group features invoking ethnic affiliation. Primordialist approaches argue that these features are inherent to human communities. Certain ethnic features emerge as important for ethnically related human action depending on historical and contextual circumstances. Hence, both approaches agree that ethnic groups do not exist simply due to social contract or culture respectively, but due to cognitive schemas, cultural idioms, organizational routines and resources, contingent events etc., which relate to the variable of community and constitute what in common sense is addressed as an ethnic group.
In a similar manner Wimmer (2002) speaks of cultural compromise emerging when the actors sharing a communicative space can agree that certain values of classifying the social world make sense. Such a process of negotiating meanings does not depend on convergence of interest, since even from different positions from economic, political and cultural hierarchies, individuals may find - albeit the different reasons - a certain way oflooking at the world meaningful (Wimmer 2002, 8).
This perspective on ethnic groupness differs from the Weberian political relations embedded in rational-choice and defines a construct which is "based on already established and internalized modes of meaning-making, a certain habitus in Bourdieau's words" (Wimmer 2002, 8). Notably, such considerations to a certain extent acknowledge the primordialists' counterarguments of inheritance under the stipulation that inherent ethnic features
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
undergo a continuous process of reframing triggered by events of groupness and vice-versa. Following this line of argument, a review of epistemologies in social scholarship titled The Study of Boundaries in the Social Sciences (Lamont & Molnár, 2002) goes even further to distinguish between two types of boundaries:
a)	symbolic boundaries are conceptual distinctions made by social actors to categorize ... they are tools by which individuals and groups struggle over and come to agree upon definitions of reality...
b)	social boundaries are objectified forms of social differences manifested in unequal access to and unequal distribution of resources (material and non-material) and social opportunities (Lamont & Molnár 2002, 168).
The authors point that while earlier scholarship has predominantly focused on the "social boundaries and monopolization processes in a neo-Weberian fashion", the more recent scholarship focuses on "the articulation between symbolic and social boundaries"(Lamont & Molnár 2002, 169). Hence, ethnicity is no longer analysed exclusively as a situational product - through contextually embedded analysis of boundary characteristics resulting from macro level socio-historical processes. Instead, inquiry is transferred onto the continuum of micro and macro factors influencing activities and events which generate and sustain ethnic boundaries.
Such theoretical framework appears consistent to the initial focus of this research - to capture the interplay of identities and interests that influence the enactment of ethnicity through education policy. Consequent to the theoretical debate, however, the research question addressing the topic must not only focus on the input these identities and interest bring to the constitution ofethnicity, but also on the actions inspired by the reflexive inter-relations between agency and structure.
When accommodated to the theoretical framework elaborated above, the delineated research filed and the identified informants (actors), the research question at this point of the epistemic constitution might be reformulated as:
Generating ethnic identities through education policy making in the Republic ofMacedonia: What strategies do actors choose and why to generate and maintain ethnic boundaries?
5. Designing an Observation and Analytical Tool
Academic literature offers two attempts for development ofuniversally applicable analytical tools relational to the chosen theoretical framework.
One of these is Banton's (2011) endeavour to synthesise all existing research on ethnicity / race / nation into Sixteen Propositions about the Genesis of Social Categories:
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
Figure 3: Sixteen propositions about the genesis of social categories (created by the author K. K.).
Sixteen propositions about the genesis of social categories:
1.	Human individuals have distinctive characteristics
2.	The attribution of significance to such characteristics results in the creation of social categories
3.	Individuals share characteristics with other individuals
4.	The phonotypical characteristics are transformed from one generation to other
5.	Social relations are multidimensional
6.	The familiar demographic environment becomes normative
7.	The common characteristics become the basis for collective action
8.	Categories are interrelated to different degrees
9.	The significance attributed to any particular characteristic is determined by society's relation to its ecological environment
10.	The significance attributed to any particular characteristics is also culturally determined
11.	The close association of categories can be a means of maintaining social inequalities between sections of the population
12.	Roles define rights and obligations in specific relationships
13.	The relevance of social categories depends upon the parties to social relations and their capacity to define the basis of their interaction
14.	Many actions by state institutions serve the ends of particular interest categories
15.	Shared sentiments are given effect in the processes of law-making and law-enforcing that provide foundations for the definition of social roles and reward conformity with social norms
16.	Categories are under pressure, such that, if they are not maintained, they change
Source: Banton 2011, 189-97.
Banton offers no methodological account on which these sixteen prepositions are synthesized. His hypothesis is that due to their overarching nature, the above outlined prepositions could contribute to a development of an objectified, etic, culture-free methodological tool for examining the circumstances in which behaviour is structured by social categories and when it is not. This would reveal how individuals influenced by social categories act in concert with others, creating structures which allow alternatives to social action and / or re-categorization.
Although evident that some of the Banton's propositions include terms such as culture and meaning immanent to primordialism, Banton advocates for constructivist approaches. He argues that social research should "uncover determinants that underlie the consciousness ofthe individuals involved"(Banton 2011, 187). The above set parameters however admittedly omit the cognitive without explicit distancing of conceptions such as culture and meaning from assumptions of social contract and rational choice.
Another, more specific analytical tool may be identified in the Wimmer's (2008) The Multilevel Process Theory. Hereby, the author develops a processual
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etnlčnostl skozi Izobraževalno politiko: najprimernejši eplstemologjl naproti
model for analysing the varying features ofethnic boundaries emerging as a result of the negotiations between actors whose strategies are shaped by the characteristics of the social field. Wimmer argues that the three main characteristics of the field: the institutional order, the distribution of power, and the political networks, determine which actors will adopt which strategy of ethnic boundary making and why. Actions and events resulting from the chosen strategies form "different degrees of political salience and social closure of ethnic boundaries, of exclusion along ethnic lines, of cultural differentiation between groups, and of stability over time"(Wimmer 2008, 970).
Data analysis conducted through this methodological tool potentially consolidates the macro socio-political context with micro ethno-methodology. The model however takes into account only those actions triggered by exogenous factors such as the characteristics of the field overlooking the characteristics inherent to the agency.
Juxtaposing the two presented analytical tools to the overall theoretical discussion on ethnicity and to the research question, it appears that both analytical tools insufficiently address symbolic boundaries. In a case where initial findings point to a predominantly culturalist perception ofethnicity, potentially influential for the informants' actions, it is evident that the above analytical tools would omit important aspects for understanding the emerging field data. Consequently, the most adequate approach would be tailoring an analytical tool that fuses the theoretical framework with the characteristics of the research filed as follows:
1.	Structuring the data according to the Lamont & Molnar's distinction of
symbolic and social boundaries (Lamont & Molnar 2002);
2.	Further proliferating the tool according to the initial data taxonomy.6
Table 1: Observation framework and analytical tool (created by the author K. K).
41
1. Symbolic boundaries		2. Social boundaries		
1. a) Framing/ conceptualizing ethnicity (by informants)	1. b) Interpretations of tasks/responsibilities (by informants)	1/2) Interpretation of tasks/ responsibilities (by others)	2. b) Informal power-relations	2. a) Policy in place
	Operationalizing ethnicity			
Source: Filed research data and Lamont &Molnar 2002.
6. Epistemic Limitations
In order to address the research topic most adequately, the epistemology of this research was primarily composed to record complex, in-depth data, preferring a detailed image of the fluid reality, rather than record hard empirical data leading to generalization and representativeness. As noted under the research design subheading, data is obtained directly in a form of (1) researcher's observations,
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
(2) informants' statements, and (3) policy documents. The abundant data are therefore rather cluttered, gathered by applying a customized semi-theoretical observation framework. Consequently, the logic of the analytical argument is cautious to demonstrate variances in the obtained information, as well as, the relativity of the representational character of the findings. Due to the small sample, disagreements among informants, borderline statements and coding overlaps are put under particular scrutiny. Applying the elaborative observation and analytical tool contributes to systematic and verifiable data gathering and the prior established analytical correlations increase the validity ofdata interpretation (Burgess 1982).
Parallel to the above confines, the specific epistemology also conveys a number of other risks and limitations:
1.	The macro-level historical accounts predominant in ethnicity studies exceed the scope of this epistemology. For purposes of avoiding atemporal, non-contextualized micro field analysis and for introducing the reader to the wider context, analyses should include an illustrative account of socio-historical processes crucial for the constitution of the field's characteristics. To maintain methodological consistency to the greatest possible extent, the context-relevant discussion should be based on policy documents gathered during field research (referenced by the informants).
2.	Compliant to the theoretical framework, the psycho-dynamics recorded under the category of symbolic boundaries should be analytically treated as a form of self-constitution rather than meaningful identity belonging. The dilemma put forward by Banton (2011) about developing a culture free methodological approach should be resolved by taking in consideration the emic to the extent to which it contributes the etic oriented analysis.
3.	A possible dead-end circular conclusion of the type: the articulation between social and symbolic boundaries / the micro and the macro level leads to actors constituting what they apparently designate, should direct the analytical focus towards the specific reasons and outcomes of such findings.
Although the research may be considered small in scale, the scope of generated data and the official authoritative and administrative status of the institution / informants make the deduction of some underlining properties possible. In the specific case, such in-depth data cannot be considered as merely anecdotal because they shed a light on the on-going institutional processes usually placed behind the curtain. Instead of being representative across institutions and different periods however, the validity of the findings should be limited strictly to the delineated research filed at the time of the research. The findings may be at the same time considered as illustrative, symptomatic and therefore contributive to further research and to future analytical purposes.
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
7. Conclusion
The overall discussion demonstrates how the quest for the most adequate epistemology in conducting an interdisciplinary, qualitative research is not a task which can be easily pre-determined. Each epistemic element is dependent on the interrelations of the other elements and unfolds the research into a continuous reframing process.
In the specific case, the researcher identifies and elaborates a number of choices made on the overall disciplinary approach, the research design, the ethical considerations, the theoretical framework and the analytical tools. The choices are evidently made within the range of the subjective (knowledge, skills and positioning of the researcher) and the objective (technical limitations), therefore, varying on a case to case basis. As in much of the anthropological research, this means that the specific research itself is not replicable. However, the hereby displayed insight in the epistemic features provides important information supportive to the quality of the gathered data and the validity of findings.
Having determined the most appropriate epistemic elements consistently, the answer to the research question "What strategies do actors choose and why to generate and maintain ethnic boundaries?" should henceforth easily construe from the interpretative framework. By applying the analytical tool to read the gathered data, the answer should arise from the following analysis: Chapter I: Identifying the social boundaries emerging from the field work gathered data;
Chapter II: Identifying the symbolic boundaries emerging from the field work gathered data;
Chapter III: Pursuing ethnicity in social relations and power distribution by analysing the previously made assertions in regards to social and symbolic boundaries against field work gathered data; Chapter IV: Pursuing ethnicity in actions and events and identifying what strategies do actors choose and practice to generate and maintain ethnic boundaries in education policy making and why? - by analysing the previously made assertions against field work gathered data
To sum up, due to the nature of the interdisciplinary qualitative research, all epistemic elements, (1) are recognized as reflexively interrelated, (2) are being systematically and consistently revised by the researcher throughout the research and analysis; and (3) are made available to the reader for analytical purposes. The ultimate goal is to make the data and the interpretation verifiable and valid by offering on display the researcher's cognitive dynamics, additionally contributing to the strength of the argument.
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
Acknowledgements
/ / This article benefits a great deal from the tutoring of my thesis mentor Trevor T. 44 D. Welland, my previous mentor Caroline Oliver and my evaluators Diane Reay and Caroline Hart.
Notes
1	The author has been working on education policy making and analysis forvarious beneficiaries: an NGO public policy think-tank; The United Nation Development Programme - Country Office in Skopje; and the Sector for European Union within the Ministry of Education and Science of the Republic ofMacedonia.
2	All ethnic communities listed in the country's Constitution except the majority ethnic-Macedonian community.
3	Electronic correspondence is accepted as official communication according to the archival and auditing procedures ofthe Ministry.
4	Transcription and indexing schemes are included as Annex I.
5	The hours spent in the field did not equal the working hours of the civil servants. The fieldwork time frame / schedule was dependent on the dynamics of activities and interlocution of the civil servants and approximately counted for a half working day. Additional amount of time was spent in other departments of the MoES in obtaining relevant policy documents referenced by the informants.
6	A sample ofdata recorded in the observation framework is included as Annex II.
References
Apthorpe, R., 1996. Reading development policy and policy analysis: On framing, naming, numbeing and coding. The European Journal of Development Research 8, 1, 16-35. doi:10.1080/09578819608426651
Ball, S. J., 1997. Policy Sociology and Critical Social Research: A Personal Review of Recent Education Policy and Policy Research. British Educational Research Journal 23, 3, 257-274.
Banton, M., 2011. A Theory of Social Categories. Sociology 45, 2, 187-201.
Barth, F., 1998. Ethnic groups and boundaries: the social organization of culture difference. Waveland Press, Illinois.
Brubaker, R., 2004. Ethnicity without Groups. Harvard University Press, London.
Burgess, R. G., 1982. Some Role Problems in Field Research in Burgess, R. G. Field Research: A Sourcebook and Field Manual. Psychology Press.
Cohen, R. 1978. Ethnicity: Problem and Focus in Anthropology. Annual Review of Anthropology 7, 379-403.
Corbin Dwyer, S. & Buckle, J. L., 2009. The Space Between: On Being an InsiderOutsider in Qualitative Research. International Journal of Qualitative Methods 8, 1, 54-63.
RAZPRAVE IN GRADIVO REVIJA ZA NARODNOSTNA VPRAŠANJA 75 / 2015
K. KRECEVA Raziskovanje etničnosti skozi izobraževalno politiko: najprimernejši epistemologji naproti
Douglas, M., 1987. How Institutions Think. Routledge, London.
Eriksen, T. H., 2002. Ethnicity and Nationalism. Pluto Press, London.
Geertz, C., 2001. Available Light: Anthropological Reflections on Philosophical Topics. Princeton University Press, Princeton.
Lamont, M., & Molnar, V., 2002. The Study of Boundaries in the Social Sciences. Annual Review of Sociology 28, 167-195.
Malen, B., & Knapp, M., 1997. Rethinking the multiple perspectives approach to education policy analysis: implications for policy-practice connections. Journal of Education Policy 12, 5, 419-445.
Mosse, D., 2006. Anti-social anthropology? Objectivity, objection, and the ethnography of public policy and professional communities*. Journal of the Royal Anthropological Institute 12, 4, 935-956.
Sanjek, R., 1990. Fieldnotes: The Makings of Anthropology. Cornell University Press, Ithaca. Schram, S. F., 1993. Postmodern policy analysis: Discourse and identity in welfare policy. Policy Sciences 26, 3, 249-268.
Schwegler, T., & Powell, M. G., 2008. Unruly Experts: Methods and Forms of Collaboration in the Anthropology of Public Policy. Anthropology in Action 15, 2, 1-9.
Smith, A. D., 2009. Ethno-Symbolism and Nationalism: A Cultural Approach. Taylor & Francis, London.
Weber, M., 1978. Economy and Society: An Outline of Interpretive Sociology. (G. Roth & C. Wittich, Eds.) (Fourth ed.). University of California Press, Berkley.
Wedel, J. R., Shore, C., Feldman, G., & Lathrop, S., 2005. Toward an Anthropology of Public Policy. The ANNALS of the American Academy of Political and Social Science 600, 1, 30-51.
Wimmer, A., 2002. Nationalist Exclusion and Ethnic Conflict: Shadows of Modernity. Cambridge University Press, Cambridge.
Wimmer, A., 2008. The Making and Unmaking of Ethnic Boundaries: A Multilevel Process Theory. American Journal of Sociology 113, 4, 970-1022.
Zelditch M., 1982. Some Methodological Problems of Field Studies in Burgess, R. G. Field Research: A Sourcebook and Field Manual. Psychology Press.
75 / 2015 TREATISES AND DOCUMENTS JOURNAL OF ETHNIC STUDIES
K. KRECEVA Researching Ethnicity through Education Policy: Towards the Most Adequate Epistemology
46
Appendices
Appendix I: Transcription and indexing Transcription scheme
At times during observation it was impossible to record speech entirely. All [text written in this type of brackets is speech paraphrased from field notes]. All (*textput in this type of brackets is a researcher's note made in situ, duringfield work). Note: All field notes were made in Macedonian language. The Macedonian to English translation and the paraphrasing are made with an effort to capture the tone and preserve the authenticity of speech to the greatest possible extent.
Indexing scheme
No./No. - Date
K - pseudonym of the researcher X1, X2, X3, X4 -pseudonyms of the key informants Y, Z, A etc. - pseudonyms of the secondary informants i/g - individual speech or action/ group speech or action
a/na - action speech / non-action speech (p/o - question/answer in action speech only) ? - further clarification needed
! - exception to previous observations/notes; and/or contradictory statement
*	- in situ researcher's notes
#	- notes added after observation
///// - data covered for purposes of non-disclosure
Appendix II - Observation framework and analytical tool
I / A: Sample of the data recorded in the observation framework
Source: The Study of Boundaries in Social Sciences (Lamont & Molnar 2002)
Note: This is a random sample selected from the original Classification Matrix with an aim to illustrate the research process. The data is not presented chronologically and all sensitive material is intentionally excluded.
1. Symbolic boundaries			2. Social boundaries		*in situ notes #additional notes
l.a) Framing/ conceptualizing ethnicity (by informants)	l.b) Interpretations of tasks/ responsibilities (by informants)	1/2) Interpretation of tasks/ responsibilities (by others)	2. b) Informal power-relations	2. a) Policy in place	
	Operationalizing ethnicity				
	xl and x2 have		(3/i; g/na) XL	(%; g/na) XL We do not	*?budgeting procedures?
	asked for a vehicle		our budget is now	have a separate budget ac-	
	during the pro-		decreased. We	count. We have been asking	*?Change? Chronicle compari-
	gramming of the		could at least visit	for that since the beginning.	son?
	IPA project stat-		schools before.	Like the Bureau for exam-	
	ing it as of a major		Now we can not	ple. Than we would be able	Influence on representation/
	importance.		do that even once	to plan our activities. There	authority/ communities
			per year.	are more and more budget	
	(3/4g/a)X2: we			cuts with the years. We ask	^Channels of communication
	can see the situ-		K: How does this	for road trips money every	
	ation first hand		influence your	year, and every year we get	Information flowbottom-up and
	and build better		work?	less and less.	top-down
	communication		XL Well every	And they do not allow us	
	with schools		communication	use the Ministry's vehicles	Involvement of stakeholders in
			with schools is	to visit schools in the	policymaking
			now limited to	countryside, but most of	
			phone-calls	our schools are away from	*?comparison with (autonomy)
				Skopje.	
H >