Abstract
Drawing from a combination of specific, empirical research projects with different theoretical backgrounds, a workshop discussed one methodological aspect often somewhat overlooked in African Studies: comparison. Participants addressed several questions, along with presenting overviews of how different disciplines within African Studies approach comparison in their research and naming specific challenges within individual research projects. The questions examined included: Why is explicit comparative research so rare in African Studies? Is comparative research more difficult in the African context than in other regions? Does it benefit our research? Should scholars strive to generalise beyond individual cases? Do studies in our field require an explicit comparative design, or will implicit comparison suffice? Cross-discipline communication should help us to move forward in this methodological debate, though in the end the subject matter and specific research question will lead to the appropriate comparative approach, not the other way round.
African Studies in Germany has gained momentum in recent years. Interdisciplinary meetings at which scholars discuss theoretical and empirical puzzles regarding the continent take place on a regular basis. One of the major intellectual and financial contributors has been the DFG's Priority Programme (SPP) 1448, “Adaptation and Creativity in Africa: Technologies and Significations in the Production of Order and Disorder”. 1 The SPP seeks to enrich the research on current institutional transformations in Africa. Biennially, scholars from up to twelve projects examine how actors in Africa deal with different economic, political and social challenges resulting from globalisation processes. In particular, they look at creative adaptations that cause specific forms of institutional (dis)order. Moreover, they are interested in the flow and circulation of technologies, understood as material technologies (for example, medicinal drugs), socio-political technologies (for instance, models of power-sharing or stateness), and technologies of the self (for example, notions of personal responsibility and independence). As a follow-up to the first biennial results conference in Maputo (October 2012), some of the researchers – in a workshop earlier this year – continued to discuss comparison as a methodological aspect often overlooked in the field.
DFG = Deutsche Forschungsgemeinschaft (German Research Foundation). For more information on SPP 1448, “Creativity and Adaption”, and the individual projects, please visit www.spp1448.de.
The workshop addressed how different disciplines within African Studies approach comparison and explained the specific challenges of individual research projects. The combination of anthropologists, geographers, historians, cultural studies scholars and political scientists proved to be an interesting mix. They represented most of the SPP projects. 2
DFG SPP 1448 “Adaptation and Creativity in Africa” Workshop, 4 March 2013, GIGA German Institute of Global and Area Studies, Hamburg. – The report is based on the contributions of the following participants (in alphabetical order): Andrea Behrends, Thomas Bierschenk, Sebastian Elischer, Christoph Haferburg, Lena Heinze, Laurence Marfaing, Frank Mattheis, Andreas Mehler, Katharina Newbery, Norman Schräpel, Michael Stasik, Johannes Vüllers, Julia Willers, Franzisca Zanker.
Do We Need Systematic and Structured Comparison?
A brief overview of the major African Studies periodicals proves that explicit comparative research remains rare. 3 From there, our discussion quickly moved to the question of whether comparison would benefit our research. The main concern raised was that asymmetric comparisons (with a stronger focus on one of the two or more cases) allow only a limited contribution to science. In order to contribute to a “higher level of analysis”, one must be able to distance oneself from specific cases. Focusing on just one case with only an implicit comparative approach means running a high risk of, at worst, distorting the results – meaning, what is said to be specific to the case in question.
The three leading African Studies journals – African Affairs, Journal of Modern African Studies (JMAS) and Africa (IAI) – only published a few contributions in 2012 that could be considered explicitly comparative. Each journal published just one strictly comparative article in 2011. Additionally, the last comparative article published by Africa Spectrum was in 2011. (Analysis was provided by Andreas Mehler during the workshop.)
This position was challenged by those arguing that it can be very difficult to make “complete” symmetrical comparisons when it comes to case studies in Africa. A researcher almost always has more knowledge on one “case”, making comparative research inherently asymmetric. Moreover, to insist on strict systematic comparisons can result in the omission of much of the context of a case from the analysis, which can also limit or distort the findings. In addition, comparative research design does not necessarily generate more interesting findings. In fact, sometimes the results can be somewhat banal. From this point of view, systematic comparisons are not a necessary condition for generating knowledge. Additionally, asymmetrical comparisons can result simply from the general ethnographic research approach to first ask “Why is this specific case interesting?” and then to embed the phenomenon into the global context. Following this approach, the project “Oil and Social Change in Niger and Chad” finds that extracting crude oil not only transforms but completely restructures socio-political arenas in new African oil states. It was noted by participants, however, that the step of embedding case-study findings into the global context is often missing.
Based on the assumption that all findings in the humanities are drawn from at least implicit comparisons, it is important to clarify the tensions inherent in comparison, including tensions between implicit and explicit, qualitative and quantitative, synchronic and diachronic, as well as ex ante and ex post comparative approaches. The latter distinction concerns going into the field with a comparative research design elaborated in advance, including selecting cases based on abstract criteria rather than developing a comparative design based on the results of fieldwork.
Do We Use Comparative Methods throughout the Research Process, or can Comparisons be Drawn Ex Post?
A vibrant debate took place on the issue of how it is possible – if at all – to use a comparative research approach in an ex post manner. This approach has long been used by ethnographers, whereby the object of comparison only becomes clear during the research process and can then be used to shed new light on the cases in question (e.g. Bierschenk and de Sardan 1997; Sørensen 2010: 45). For other disciplines, this approach is difficult to understand, as it would be impossible for them to undertake research without knowing the tertium comparationis in advance, as this forms the basis of their analysis. Nonetheless, despite the difficulties and challenges of conducting research across different field sites (reiterated by team members from all of the projects), some very useful findings have resulted from research using this approach to comparison. “Translations of the ‘Adaptation to Climate Change’ Paradigm in Eastern Africa”, for example, contributes to the understanding of how the global paradigm of adaptation to climate change is translated by actors in Ethiopian, Tanzanian and Rwandan institutions and how these translation processes, in turn, shape environmental governance and society–environment relations in the three countries.
Another project that uses an ex post comparative approach is “Translating Global Health Technologies”. Focusing on access to medicine in Uganda and Rwanda, it examines the development of new institutional arrangements regarding medical care and scientific research in these two countries.
What about Comparison is Specific to African Studies?
With regard to the African context, historians and geographers raised the concern that an “imperial ideology” in comparisons (and rankings) based on an occidental framework constrains scientific work. This is linked to a further issue: Comparisons are often based on socially constructed knowledge and/or units of comparison. Whilst, arguably, comparisons are still possible, one must take their constructed dimension into account. Indeed, this is one of the benefits of a more “grounded” research design. Thinking of comparison as an end product or as a predetermined conceptualisation can affect our understanding of what is actually taking place. Two projects explicitly address the constructed dimension of their units of comparison: “Changing Stateness in Africa: Practices and Imaginations from Cameroon and Ghana” explores how order and disorder are produced by the institution of the state, the expectations of some of its subjects/citizens, and external reference models. Focusing exclusively on the actor level of social interaction, “Entrepreneurial Chinese Migrants and Petty African Entrepreneurs: Local Impacts of Interaction in Urban West Africa” examines interactions between Chinese and African entrepreneurs in Accra (Ghana) and Dakar (Senegal) and the development of social practices and norms in that context.
The pragmatic constraints of comparative research were deliberated throughout the workshop. It is difficult to compare research from different sites when conceptualisations and research topics are interpreted in a different manner in the course of fieldwork, partly also because access varies. It is even more problematic when the research design contains a systematic case selection that a lack of data does not really allow for. It was suggested that the limited amount of explicitly comparative research published might also be due to the fact that scholars tend to feel more comfortable reviewing only specific cases and would have to conduct asymmetric reviews for articles that cover several field sites across different countries.
Can We Even Communicate with Each other (across Disciplines)?
The comparative approaches within the disciplines are partly based on very different premises. Political scientists try to isolate explanatory factors in order to better explain a phenomenon. They argue that generalising or even showing the uniqueness of a case can be done only by using an explicit and systematic comparative design (e.g. Collier 1993; George and Bennett 2005; Lijphart 1971). Geographers have noted the political message that can result from a comparative research design, especially when using “space” as a socially constructed unit of analysis (see e.g. Belina and Miggelbrink 2009; Griggs 2000). This is something the project “‘Festivalisation’ of Urban Governance” keeps in mind while looking at the dynamics of urban governance and the re-ordering of spaces in Johannesburg and Durban following the FIFA World Cup; the researchers are examining the significations of (dis)order and re-negotiations of state–society interlinkages resulting from the event.
Historians, especially those specialising in African Studies, have increasingly paid attention to histoire croisée. Comparing processes, perspectives and understandings, this method explores the entanglement of cases otherwise portrayed as isolated. In addition, reciprocal comparisons offer opportunities to overcome a priori normatisations, over-determined universalisms or unidirectional accounts of interrelations (e.g. Austin 2007; Coquery-Vidrovitch 2006). Cultural studies assemble matters and methods of various disciplines and hence take a varied approach to comparison. They point to the difficulties related to processes of cultural differentiation, inter-culturality and trans-nationality, as well as acknowledging the problem of methodological Eurocentrism. One answer has been to focus on “cultural transfers” and mutual exchange (e.g. Middell 2000; Siegrist 2003). Like histoire croisée, reciprocal comparison was challenged by advocates of other disciplines on the grounds that it is not comparative in the narrow sense.
While intercultural comparison had been the hallmark of anthropology until far into the twentieth century, comparison has recently gone somewhat out of fashion in the discipline. Contemporary anthropologists, for the most part, instead focus on embedding micro-processes into larger phenomena (exceptions being Gingrich and Fox 2002; Moore 2005). Two significant epistomological challenges arise – not only in anthropology, but in any comparison where a qualitative approach is adopted: First, units of analysis are usually interpreted as “constructed”; second, the more detailed a description is, the more difficult it becomes to compare it to something else.
Discussing comparative methods on an interdisciplinary basis is not easy, but our discussion between the two “extremes” represented by political scientists and ethnographers was less polarised than expected. However, differences between our projects were still pronounced, as they juxtaposed, on one side, an approach based on a Popper-inspired tradition of generating knowledge by falsifying theories or hypotheses with, on the other side, a grounded-theory approach of entering the field without preconceived hypotheses, albeit with leading questions and sensitising concepts. The project “The Local Arena of Power Sharing: Patterns of Adaptation or Continued Disorder”, for example, follows a Popper-inspired approach. It examines whether (and, if so, under what conditions) national power-sharing agreements effectively enhance peace on the local level, focusing on local arenas in Burundi, DR Congo, Kenya and Liberia. By contrast, the project “Roadside and Travel Communities: Towards an Understanding of the African LongDistance Road (Sudan, Ghana)” takes a grounded-theory approach. By examining roadside and travel communities and the socio-technological orders created by their interaction in four road/roadside settings in Sudan and Ghana, this project has developed a new approach to comparative research. The researchers use four “windows of enquiry”, each focusing on one “window” in particular but simultaneously working towards gathering generic insights into all “windows”, thereby enabling exponentially thickened comparisons.
It was agreed that it is a mistake to associate particular epistemologies with particular disciplines, and that inductive and deductive reasoning should be used in due proportion. After all, the aim of comparative area studies is to reconcile and/or combine generalisation and specification.
Conclusions
The projects within the SPP all take a different approach to comparative research and engage with comparison to different degrees. Differences between the disciplines remain – the theoretical and methodological plurality being a strength of SPP 1448 – yet, debate remains vital. African Studies may be avowedly multidisciplinary, but it largely lacks fundamental collaboration between the different disciplines (Bryceson 2012: 298–299). Given the fact that no participant claimed that implicit comparisons are entirely absent from his or her work, should we all make our comparisons explicit? Is comparison more meaningful for testing theories or for developing them? How do we evaluate ex ante or ex post comparisons?
At the most basic level, we all need to be aware of and reflect on comparison within our research. Debating reasons for and against systematic and structured comparisons can help us understand the nature of our research questions and the best way to address them. More specifically, those using systematic comparison to identify causality need to deal with possible omissions of in-depth case specificities. Single-case and largely asymmetric comparative studies should specify the type of comparative approach they are employing and highlight why the studied case could be interesting for other scholars, including those who aim to generalise on a higher level. We can learn from the way the “other” disciplines approach comparative research in Africa, but in the end the chosen theory and methodology must stem from our epistemologies, ontologies and research questions. The subject matter and specific research question(s) lead to the relevant comparative approach, not the other way round.