School of Social and Political Science

Blog content

Dr Carin Runciman and Boitumelo Matlala 2023

Discussions of decolonisation have been reinvigorated in recent years following pressure from student movements in different parts of the world such as the #FeesMustFall and #RhodesMustFall protests in South Africa and the Why is my curriculum white? campaign at University College London. This, in turn, has led some scholars to reckon with the coloniality of knowledge and power that is embedded in University teaching in both the Global North and South.

Research methods are central to this discussion, given their significance in producing knowledge about society. However, it is notable that much of the contemporary discussion of decolonising methodology and research methods tends towards discussing qualitative research methods. This can give the appearance that quantitative methods are either exempt or peripheral to this discussion.

Some of the reasons for this may lie in the history of quantitative methods in the coloniality of knowledge, where oftentimes indigenous, black and minority identities and experiences have been erased by simply not counting them. Given the importance of statistical data to state planning projects, this erasure has been significant and with long-term consequences. In turn, this underscores the potential power of quantitative data to be a powerful tool in the hands of oppressed and marginalised populations, as Zweiner-Collins and her colleagues have argued.

What then does it mean to decolonise survey research? In this blog post, we share our experience of using survey research within a project that aimed to produce a decolonial approach to understanding democracy in South Africa.

Democracy from below

The ‘Democracy from Below’ project is funded by the National Institute for the Humanities and Social Sciences. The aim of the project is to expand the canon of democratic theory by advancing the understanding of the conceptualisation of democracy within indigenous knowledge systems in South Africa and how this intersects with the history and experiences of working class Black African South Africans.

To do this, we adopted a mixed-methods approach, using a multi-lingual online survey and qualitative interviews to explore the meanings of democracy. The aim of the survey was twofold. One, to develop a macro analysis of the concepts used to express democracy and democratic ideals within South Africa’s indigenous languages and, two, to experiment with ways to measure the understanding and practice of democracy that was not framed solely through measures of liberal democracy. The survey was, therefore, an experimentation of the possibilities and limitations of survey research in engaging with indigenous knowledge systems.

Our positionalities, as a non-indigenous and indigenous researcher working together, in some ways reflect the complexities of what we sought to capture in the survey. The survey design attempted to capture a pluriversal approach to understanding democracy. We operated with an understanding that the democracy that we sought to measure neither replicated nor rejected liberal democracy. Therefore, our task in designing the survey was to find ways to measure the extent to which tenets of liberal democracy remain important while also attempting to construct survey questions that allow us to measure the strength and extent of alternative expressions of democracy. This introduced a tension between adapting and using survey questions from major international surveys such as the Afrobarometer or the European Social Survey, for comparability and designing new questions that could express alternative understandings of democracy.

One of our concerns was whether the survey design could decentre traditional survey questions on liberal democracy enough to allow alternative conceptions to be expressed while, at the same time, not knowing if we had constructed survey questions that accurately captured the alternative theories and practices that both researchers had documented in previous qualitative work.

Conceptual and linguistic challenges

We were deeply aware that our choice to conduct an online survey meant that we had significant power in shaping how our participants may understand democracy. Inadvertently, by including standard questions about liberal democracy we risked reinforcing normative definitions of democracy and ideas of how it should work. This could have deflated sections of the survey that sought to elicit alternative ideas about democracy – leading respondents to either skip questions in these sections or compromise the level of honest responding.

To balance these pitfalls, a sizeable portion of the survey sought to unmask other factors that influence people’s conceptions of democracy, such as community or personal beliefs/values. To safeguard the space for expressing alternative ideas of democracy, we also posed open-ended questions that allowed participants to answer freely.

A critical area of debate was even about the term democracy itself within the questions. In Sesotho, Setswana and Sepedi the most commonly used word is an adaptation of the concept in English – demokrasie. While in isiZulu the concept intando yeningi (the will of the majority) has evolved to describe the systems or the processes of liberal democracy. In turn, this reflects that isiZulu, the most commonly spoken language in South Africa, has been subject to greater attempts at standardisation in comparison to other languages. In contrast, inkululeko (freedom in isiXhosa) arguably has a more expansive meaning that is less constrained by liberal democratic ideals in comparison to intando yeningi.

What we want to make clear here is that this is not simply a process of translation, it is a serious reckoning with indigenous knowledge systems and appreciating how these have been shaped by history. As a result, we have to be aware that although we attempted to standardise the questions across all the languages, the reality is that this is not entirely possible and we have to be attentive to these differences in our analysis.

Considerations about language and translation and the potential (dis)connections with indigenous knowledge systems were a continual challenge throughout the survey design. As the survey was co-created between a non-indigenous and indigenous researcher, English had to be the primary language of the survey construction. This, arguably, creates epistemological limitations.

One clear example of this was the use of a five-point Likert scale. In many of South Africa’s indigenous languages, it is difficult to distinguish between ‘strongly agree’ and ‘agree’. The relevance of this distinction was not always clear when considered in indigenous languages. The difficulties of translating the Likert scale also surfaced broader assumptions behind some of the survey mechanics, for instance the viability of a linear assessment of the strength/intensity of attitudes and views.

This is something that requires more reflection for those conducting survey work in South Africa or in other parts of the world where such measures are not mutually intelligible. It highlights the need for survey work undertaken in a decolonial frame to be conscious of the interaction between methods and indigenous knowledge systems.

What is being decolonised?

Reflecting on this work has caused us to question what, if anything, did we decolonise in this project? We cannot claim to have decolonised survey research, but we do think we have given important consideration to some of the steps that this has involved. Perhaps, more importantly, the decolonisation process has been more about us. Despite our different positionalities, a commonality between us has been that our educations have reflected the coloniality of knowledge, and arguably, we have been rewarded for reproducing rather than challenging it. This project opened up a space where we could challenge ourselves and our domain assumptions about survey research within the broader epistemic decolonisation movement and provided a space to experiment with what using a survey in such a project yields.

_

Photo by Emily Morter on Unsplash