Giddens, Sen and IsiXhosa-Speaking Women Traders: Theoretical Grafting to Enhance Analysis

ABSTRACT

Grafting selected elements into a primary theory from another theoretical approach can sometimes enhance the theoretical framework to facilitate more refined data analysis. Because of the possible risks of theoretical fragmentation that may result from incorporating elements of different theories into a single framework, we opted for grafting selected elements of one theory into our primary theory, to carefully retain the logical coherence of the main theory. In this paper, we draw on a study of isiXhosa-speaking women traders in Cape Town’s Langa Township (hereafter referred as Langa) to argue that grafting elements of Sen’s capability approach (CA) into Giddens’s structuration theory (ST) enhanced our theoretical framework. This improved the quality of our analysis and enabled us to generate more nuanced findings. Our qualitative study with 25 African women traders investigated how these women exercised agency in responding to structural constraints and opportunities that affected their trading businesses. We used participant observation and in-depth interviews to collect data. Sen’s CA (grounded in social justice) expanded on Giddens’s notion of structures as constraining and enabling. For this aspect of our theoretical framework, CA informed by Sen’s research on women’s empowerment projects amplified the intersection of gender, class and race constraints that affected the women traders. Furthermore, CA’s emphasis on quality of life and well-being outcomes was useful in helping us assess the transformative capacity of the women traders’ agency. This dimension complemented Giddens’s ST and therefore our theoretical framework was enhanced by grafting in these elements from Sen’s CA.

KEYWORDS:

• structuration theory
• structural constraints and opportunities
• capability approach
• grafting
• theoretical framework
• women’s agency

1. Introduction

When developing a theoretical framework to analyse data about a complex social phenomenon, it is sometimes necessary to selectively graft elements of one theory into another theory. The grafting process can enhance the theoretical framework to undertake a more nuanced analysis. On its own, without the grafted elements, the dominant theory would not be able to achieve this analytical purpose. Massey et al. (1993) noticed in their research that when grand or macro theories are used to analyse (and explain) specific situations they can benefit from this kind of selective grafting to meet the nuanced analytical demands dictated by a given context under investigation. For example, Massey et al. (1993) argue that while grand theories such as world systems theory and dual labour market theory help to explain international migration, they focus on broader residual effects of globalisation and market penetration across national boundaries, whilst neglecting micro-level decision-making processes at the household level. The complex multifaceted contemporary migratory processes cannot always be explained by a single theory and require the selective inclusion of other theoretical elements to analyse the micro-level nuances. This is how grafting selective elements from another theory into the dominant theory to enhance the theoretical framework can lead to more nuanced analysis and produce more sophisticated findings. In our community-engaged research, which aimed to inform development intervention practices, we have likewise found theoretical grafting necessary.

“Plugging in” elements from another theory to strengthen an established theory requires grafting. The notion of grafting in theory originates in horticulture. Grafting in horticulture entails joining a portion of one plant (e.g. a bud) onto a stem, root, or branch of another plant so that a union is formed and they grow together. We are aware of the possible risks of using more than one theory in a single theoretical framework for analytical purposes in applied research. One such risk is theoretical fragmentation. The internal logical coherence of the different parts of a theoretical framework, a feature of a good theory (Higgins 2004), can be lost when mixing and matching elements from other theories. If a theoretical framework constitutes elements from more than one theory and the different parts are not coherently connected then the resultant theoretical fragmentation is likely to lead to flawed analysis, which defeats the purpose of adapting the primary theory. For example, theoretical fragmentation can restrict the nature of the investigation of a phenomenon to the extent that it ignores the intensity of relationships with other relevant concepts such as analytical variables or categories that are regarded as antecedents or consequences, which help explain the phenomenon (De Aguiar Rodrigues and De Carvalho-Freitas 2016).

The problem of theoretical fragmentation is one of the reasons why we prefer the grafting model. With theoretical grafting, the “stem” and “root system” of the primary theory are retained and only a few selected elements of one theory are carefully grafted into the primary theory, taking care not to disrupt the logical coherence of the main theory.

The aim of the paper is to demonstrate and reflect on examples of theoretical grafting that were included in an adapted theoretical framework in order to analyse human behaviour in a specific context. The study¹ on which this paper is based investigated varying degrees of agency among 25 isiXhosa-speaking women traders in Cape Town’s Langa township. We argue in this paper that the adapted theoretical framework produced nuanced findings partly because we grafted very selectively some theoretical elements from Amartya Sen’s capability Aapproach (CA) into Anthony Giddens’s structuration theory (ST), thereby enhancing our analytical capacity. Giddens’s ST guided the study as the primary theory, but selected elements from Sen’s CA complemented specific aspects of our main theory. Giddens is more of a theorist than Sen, in that he provides theoretical explanations of how society functions through, for example, social reproduction; thus, his contribution is at a broad ontological level. In contrast, Sen’s CA is not a full-fledged theory; rather, it is often described as a “broad normative framework for the evaluation and assessment of individual well-being and social arrangements, the design of policies and proposals, and proposals about social change in society” (Robeyns 2005, 94; see also Sen 2009).

ST is pertinent to the study because it delineates the conceptual and contextual framework of structure (structural constraints and structural opportunities) and of agency as recursive practices that people engage in to change their life circumstances and thereby shape social systems (Giddens 1984). With ST as our primary theory, we had a well-developed theoretical lens for understanding how the lives of women traders in our study are shaped by societal structures and how they in turn influence structures (even in small ways) through systems of social reproduction. For example, societal structures (such as race, class, ethnicity and gender) and institutional rules and resources (such as global and local institutions of trade, commerce and economic systems) shape the women’s routinised and daily trading activities whilst they in turn reproduce and adapt trading practices and related structures. We selected aspects of CA for our study firstly because it is rooted in principles of social justice, and secondly because it was designed to evaluate poverty alleviation, human development and women’s empowerment projects. This paper is not a critique of Giddens’s ST per se, but rather a discussion of how CA complemented ST and why this was necessary for the purposes of our analytical work. This is not by any means the first time elements of another theory have complemented ST. For example, in Liwane and Rossouw’s (2019) study, elements borrowed from humanist paradigms complemented ST to analyse and explain the influence of social structures on educators’ agency in public education in South Africa. Another example is Kachale, Erasmus, and Sonnenberg’s (2019) study, which drew on ST and sustainable development frameworks to understand the relationship between fuelwood energy consumption and environmental sustainability.

At the centre of the discussion in this paper are the following three examples of theoretical grafting: (1) Sen’s CA complemented Giddens’s notion of structures as constraining and enabling by incorporating principles of social justice that help distinguish enabling and constraining environments; (2) we applied CA to assess transformation, which is one of ST’s dimensions of agency, by integrating Sen’s notions of “quality of life” and “well-being outcomes” as benchmarks to gauge improvements emanating from the women’s trading businesses; (3) Sen’s CA provided a gender lens for evaluation research aimed at women’s empowerment, which we used to complement Giddens’s delineation of structural constraints.

We begin by providing some background on the study and our rationale for choosing ST and CA. This is followed by the main discussion which demonstrates how we drew on Sen and Giddens to develop our theoretical framework for the study and how in turn our adapted framework informed the research questions, analysis and findings.

2. Some background information about the study

We chose Langa as a research site because it is an example of a South African high-density township where poverty is endemic. Langa is located 12.4 km south east of the Cape Town city centre and was established in 1927 as one of the three areas where black Africans could reside according to the Native Urban Areas Act 21 of 1923 that restricted their movement to prescribed areas in the colonial city. Langa has a rich cultural and historical heritage, which includes resistance to the apartheid government (City of Cape Town 2017). Table 1 shows the (most recent) 2011 South African Census records of Langa’s population as 52 401, with 16 958 people per square kilometre, and thus classified as a high-density township. According to Statistics South Africa (2011), characteristic poverty indicators in Langa include: limited access to piped water and toilets, protracted housing problems and low levels of education. Table 1 shows that only 33.1 per cent of the residents achieved a Matric (high school) level of education, and 57.5 per cent live in formal dwellings. Of the 17 402 households, nearly half are female-headed. In our study, 13 out of 25 women were heads of their households.

Table 1. Selected statistics about Langa pertinent to the study’s demographic and poverty indicators.

Key characteristics	Statistics	Key characteristics	Statistics
Total population	52 401	Number of households	17 402
Working age (15–64)	71.7%	Female-headed households	41.7%
Population density	16 958 persons per square kilometre	Formal dwellings	57.5%
Matric aged 20+	33.1%	Piped water inside dwelling	49.6%
		Flush toilet connected to sewerage system	72.4%

Source: Statistics South Africa (2011).

Our study highlighted several other dimensions of poverty, for example violence, ill health, and crime and gender inequality, that were prevalent in the lives of women who participated in this study (see Mpofu-Mketwa 2020).

We adopted a case study methodology, with in-depth interviews and participant observation to investigate how isiXhosa-speaking women traders in Langa exercised agency in responding to structural constraints and opportunities that affected their businesses. A non-probability purposive sample of isiXhosa-speaking women traders (n = 25) participated in the study. A sample size of 25 participants out of a possible 50 licensed women traders (based on informant information) was included in the study. The sample was appropriate for qualitative fieldwork as it provided rich in-depth information, which allowed for case clustering and the identification of patterns in the data (see Mpofu-Mketwa 2020). Our unit of analysis was individual isiXhosa women traders in Langa, hence we purposively chose women with similar backgrounds including class, language, cultural heritage, business type and geographical location. With these demographic similarities among the participants, we could focus on similarities and differences in ways in which they responded to structural constraints and opportunities that they as individuals encountered in their daily lives. The in-depth interviews were originally conducted mainly in isiXhosa; the English quotations included in this paper are translations. Miles, Huberman, and Saldana’s (2014) thematic coding approach guided the qualitative analysis of data, which consisted of field notes and transcripts. We used NVivo text analysis software to analyse the textual data.

Table 2 provides selected socio-demographic information about the study’s participants. Notably, of the 25 participants only seven completed high school, a little over half (14/25) were 41–55 years old, and the vast majority (22/25) sold food. The women traders operated their independent survivalist businesses mostly at the Langa taxi rank or surrounding areas, with occasional employed or voluntary assistance from community members or friends. Our qualitative data revealed further that a few braai/barbecue and butchery businesses employed two to three assistants as the work is labour intensive. At the time of the study, their businesses had been in operation for about three months to 10 years. In Langa, traders usually belong to one of the traders’ associations, which made it easier to gain access to the women who participated in the study (Mpofu-Mketwa 2020).

Table 2. Socio-demographic details.

Age
26–30	5
31–40	2
41–55	14
56+	4
Total	25
Highest level of education attained
Primary school	2
Incomplete secondary school	16
Completed secondary school	6
Post-secondary school training	1
Total	25
Marital status
Married	12
Widowed	3
Divorced	2
Never married	6
Separated	1
Living with a boyfriend	1
Total	25
Membership to a traders' association
Active member	8
Inactive member	4
Non-member	13
Total	25
Type of business
Catering/cooked food	8
Braai/barbecue	5
Open-market meat butcher	3
Fruits and vegetables	3
Spaza shop/small grocery shop	2
Flea market/clothing	1
Hairdressing	1
Shebeen/home-based liquor	1
Combination clothing and meat	1
Total	25

3. Rationale for using ST and the need to graft elements from CA

As the primary theory, Giddens’s ST informed the four key analytical concepts at the centre of the study, namely structure, structural constraints, structural opportunities and agency. The rationale for choosing ST stems from the nature of the observations and questions that the we were asking at the outset of the study. We both have long associations with and an in-depth knowledge of Langa.² We noticed that among the many women street traders some thrived while others struggled, amidst similar broad structural constraints and opportunities. ST was appropriate because it explains individual and collective agency within the constraints and opportunities or resources provided by the social order (Fuchs 2003; Giddens 2009; Hubbard 2000; Lieblich, Zilber, and Tuval-Mashiach 2010; Ritzer 2008). Thus, through the concept of social reproduction the theory explains “how day-to-day activities of social actors draw upon and reproduce structural features of wider social systems” (Giddens 1984, 24, 25). The structural properties of social systems are both the medium and the outcome of the practices recursively organised by human agents across time and space (Giddens 1984). Whilst ST is certainly useful in explaining how social structures influence human actors and how they in turn exercise agency in shaping structures, for the purposes of our study, we needed to adapt ST to accommodate the analytical demands dictated by the specifics of Langa’s socio-economic and trading environments. Grafting selected aspects of Sen’s CA into Giddens’s ST was further facilitated by Giddens engaging capabilities for transformation in ways that sufficiently aligned with Sen’s notion of capabilities. As already mentioned, this theoretical complementarity made grafting feasible and enhanced our theoretical framework in order to meet the analytical needs of our community-engaged research. For example, Giddens’s rendering of structures as allocative and authoritative resources evokes transformative capacity. According to Giddens (1984, 33): “Allocative resources refer to capabilities—or, more accurately, to forms of transformative capacity-generating command over objects, goods or material phenomena. Authoritative resources refer to types of transformative capacity generating command over persons or actors.”

Giddens’s assertion of societal structures as resourceful and inherently capable of transformation is plausible and useful in explaining the broader social system, at an abstract level. Aspects of ST needed to be enhanced and elements of Sen’s CA seemed a suitable match for grafting to work so that the adapted theoretical framework could effectively analyse the time and place-based contexts of Langa. At a theoretical level and logically there appeared to be a match, but until we actually worked with our adapted theoretical framework and performed the analysis, we were not certain that it would succeed.

We now turn to the specific aspects of CA that were grafted into ST.

4. Selected aspects from CA grafted into ST

Aspects of CA provided the following examples, which we chose to graft into ST to strengthen the theoretical framework: (i) the notion of instrumental freedoms, (ii) quality of life and well-being outcomes as measures of transformation and (iii) gender dimensions of agency and constraints.

Instrumental freedoms are at the core of Sen’s notion of development and CA (Robeyns 2006). Sen (1999) uses the concept of instrumental freedoms to underscore the essence of development as creating enabling societal institutions that are crucial for expanding people’s capabilities or opportunities to live lives that are free from all forms of deprivations. According to Sen (1999), the enabling social institutions include: (i) political freedoms—e.g. civil rights broadly and opportunities to determine who governs and how; (ii) economic facilities—e.g. opportunities to utilise natural resources, to produce or to exchange goods and services; (iii) social freedoms—e.g. opportunities for education, healthcare, social welfare; (iv) transparency guarantees—e.g. opportunities for transparency and to build trust in society; and (v) protective security—e.g. opportunities to access social safety nets. Sen’s CA thus advocates for developing an enabling environment that is rooted in social justice principles, including equity.

CA furthermore foregrounds well-being outcomes and quality of life improvements as measures of transformation. Through the CA, Sen and others have developed an evaluative framework for quality-of-life assessments (Robeyns 2005; Sen 1999). These measures helped us to evaluate transformation among women traders’ micro trading businesses, as we outline below in relation to our research findings.

It would be remiss for a study of African women traders in the Global South to exclude the gender dimensions of agency and constraints. CA offers, moreover, an intervention-orientated approach that has evolved in practice and over time in the context of human development and women’s empowerment initiatives (Alkire 2008; Alkire and Ibrahim 2007; Kabeer 1999; Nussbaum 2000; Nussbaum and Sen 1993; Robeyns 2003; Sen 1999). As such, it provides a basis for evaluating the progress of development initiatives by assessing: (1) the extent to which women are free to exercise their agency, (2) how enabling their structural environments are, and (3) the specific aspects of women’s well-being that improved or deteriorated (Nussbaum 2000; Sen 1999). Importantly, given the focus of our study, CA does not assume that development initiatives address poverty equally in households and society. It is attentive to vulnerable groups such as women, children and people with disabilities.

CA contributed significantly to our study by providing a multidimensional understanding of poverty and development, which is inclusive and revolves around freedom. In the following section we discuss examples of how Sen’s CA actually complements Giddens’s ST in the context of our theoretical framework and driven by our analytical needs.

5. How Sen and Giddens work together for the purpose of our study

We illustrate how grafting CA into ST enhanced our theoretical framework, which in turn informed our data collection, facilitated nuanced analysis and ultimately generated our findings. The three examples show how: (i) Sen’s concept of instrumental freedoms further develops Giddens’s notion of structures as both constraining and enabling; (ii) CA, as an evaluative framework, is tailored to assess well-being outcomes and quality of life improvements at the micro level, which are underdeveloped in ST; and (iii) Sen’s gender inclusive approach to development compensates for its under-emphasis in the way ST explains social reproduction in capitalist societies (see Kahlert 2012, 59).

We now discuss each of these three examples.

5.1 Integrating instrumental freedoms

Giddens (2010, 86) refers to structures in society as “rules and resources,” which can enable or constrain individuals. The conception of structures as rules and resources has sometimes been referred to as a “cultural toolkit” that human actors tap into to transform their lives (Giddens 2010; Lieblich, Zilber, and Tuval-Mashiach 2010). In any social order or structure, where there is enablement there are opportunities. This refers to the ability of social, political, economic and legal systems (including institutions) to provide resources that create opportunities for human actors to effect change in their lives (Giddens 2010, 86). In our analysis, we found that this conceptualisation of opportunities is useful in an ontological sense, but it is insufficient for assessing development in a more practical, evaluative and context-specific way.

Within ST, structures can indeed be enabling and are possible reservoirs of opportunities. However, because the theory is not explicitly rooted in principles of social justice and is not a normative framework for human development, we needed a direct and integrated approach which CA offers for “real opportunities to achieve valuable states of freedom.” In this respect, CA complements ST by framing development as an expansion of opportunities (instrumental freedoms) or the removal of multiple forms of deprivations (“unfreedoms”) for a more socially just society to evolve. Sen’s concept of instrumental freedoms makes a significant contribution to our study because of its theoretical and analytical capacity to integrate the expansion of substantive freedoms and opportunities across a wide spectrum in the area of public policy. Table 3 illustrates Sen’s five instrumental freedoms with examples derived from our data.

Table 3. Sen’s instrumental freedoms with examples from our study.

Sen’s instrumental freedoms
Political freedom	Transparencyguarantees	Economic facilities	Social arrangements	Protective security
Opportunities forpolitical participation, free speech,public scrutiny and elections	Opportunities to hold government accountable and to protect its citizens from abuses of power	Opportunities to acquire the economic means for survival and to avoid a life of poverty	Opportunities for basic education, healthcare, building cohesive community relations	Opportunities for social security and welfare programmes that prevent destitution
e.g. women’s rightsto vote andparticipate inimproving community life	e.g. anti-corruption strategies against nepotism in securing trading stalls	e.g. availability of merchandise supplies and market for buyers	e.g. statutory and non-governmental support in promoting well-being and skills development	e.g. social security provision through the Department of Social Development and social capital from community networks

Instrumental freedoms are based on the premise of espousing democratic participation in different institutions including “markets and market-related organisations, governments and local authorities, political parties and other civic institutions, educational arrangements, the media and other means of communications” (Sen 1999, 9–11). Giddens refers to these institutions as structural properties of social systems (Giddens 1984). These conditions are necessary for a conducive trading environment.

By assessing how political freedoms promote, for example, economic security, social opportunities for education, and healthcare for human development we were better prepared to analyse the women’s trading environment. Given the socio-historical context of South Africa that engendered poverty and inequality in township communities, the concept of opportunities derived from CA is relevant to evaluate the progress of development policies in a post-apartheid democratic South Africa. For example, in 1996, the South African government adopted the Growth Employment and Redistribution (GEAR) macro-economic policies aimed at increasing labour market flexibility and government investment, privatising state assets to reduce government spending, and creating significant job opportunities (Cohen 2010; McGrath and Akoojee 2007; Ranyane 2015). It was hoped that the economic growth would trickle down to the poor. However, the literature documents the failure of the policies to offer equitable distribution of economic gains or the “trickle-down effect” (McGrath and Akoojee 2007). By contrast, there were massive job losses, increasing unemployment that pushed unemployed workers, mostly black women, into informal trading with very limited government economic support for their survivalist businesses (Cohen 2010; Ranyane 2015). Most of the women traders in our study reported job retrenchments and outsourcing of labour as factors leading them to operate trading businesses.

It is therefore even more important for multiple stakeholders to engineer government and society’s accountability (Robeyns 2006). Sen’s notions of “transparency guarantees,” “economic facilities” and “protective security,” for example, demonstrate the pertinence of an accountable government that protects its poorest citizens by providing social security for survival in response to poverty. Giddens’s references to “rules and resources” and a “cultural toolkit” are undoubtedly useful concepts for analysing the enabling as well as constraining nature of society. Adding instrumental freedoms to the analysis amplified society’s obligation to provide multidimensional opportunities for poverty alleviation.

By defining opportunities as different resources and forms of enablement that enhance the participants’ life chances in social, economic and political spheres, we were able to ask qualitative interview questions about the various structural opportunities available to women traders such as religious organisations, family support systems, local government educational institutions and non-governmental organisations’ support for skills development. Examples of these interview questions included:

1. Are there any places or organisations that are available in your area to teach or provide information about running businesses?

2. How do you compare your opportunities as a woman in operating a trading business with men traders?

These questions illustrate how we were able to substantially expand on Giddens’s abstract notion of resources provided by social systems. For example, Giddens delineates resources into allocative resources and authoritative resources in a way that resonates with Sen’s conception of opportunities articulated in instrumental freedoms, particularly political freedom and transparency guarantees. Authoritative resources pertain to “non-material resources involved in the generation of power, deriving from the capability of harnessing the activities of human beings.” These resources “result from the dominion of some actors over others” (Giddens 1984, 373). Thus, power is central to shaping social transformation. Sen’s notion of political freedom echoes this by delineating political freedom as opportunities to participate in socio-economic activities, free speech, public scrutiny, and elections (Sen 1999). By creating opportunities to hold governments accountable and to protect their citizens from abuses of power, transparency guarantees complement authoritative resources in seeking transformation through freedom from political hegemonies. In our study, the women traders’ narratives reflected these linkages, for example through their perceptions of their civil rights and empowerment in personal and public spheres and challenging unequal distribution of power and income in the households and trading stalls.

Giddens’s conception of allocative resources refers to “material resources in the generation of power, including the natural environment and physical artefacts” these resources derive from “human dominion over nature” (Giddens 1984, 373). The transformative capacity of humans to exercise command over goods and services (Giddens 2010) is similar to three of Sen’s freedoms, namely economic facilities, social opportunities and protective security. Examples include: economic facilities which provide citizens with opportunities to participate in the labour market; social opportunities such as basic education, healthcare, and cohesive community relations; and protective security opportunities through social security programmes. These examples all constitute allocative resources, as indicated in Table 3.

There are clear points of convergence between Giddens’s allocative resources and Sen’s instrumental freedoms. Grafting CA’s instrumental freedoms into our theoretical framework expanded ST’s concept of enablement and resources, thereby helping us to solicit findings about structural opportunities in more context-specific ways than we would have done with ST alone.

Below we discuss how our enhanced theoretical framework facilitated nuanced analysis, which in turn shaped the findings. The focus here is not on the details of the key findings per se; rather, we illustrate how specific findings emerged with help from the enhanced theoretical framework.

5.1.1 Economic facilities

Following the demise of apartheid, market mechanisms of demand and supply for goods and services and the labour market presented an enabling economic environment for the women traders. Notwithstanding the impact of post-apartheid macro-economic policies on the labour market stated above (e.g. the African National Congress (ANC) government’s post 1996 macro-economic policy, GEAR), findings from our study indicated that post-apartheid democratic elections paved the way for a new dispensation and ushered in a more enabling economic environment for black South Africans than under apartheid, when it was common for traders to experience police harassment for example. In our study, some women traders reported the availability of flexible market opportunities, access to some financial services, and constant merchandise supply from nearby farmers as well as the neighbouring wholesalers in the Epping industrial area adjacent to Langa (Mpofu-Mketwa 2020). The labour market also equipped some of the women traders with business skills that they acquired from previous work experience in the retail and restaurants sectors, which they applied to their own trading businesses (Mpofu-Mketwa 2020). By delineating the integrated nature of political and economic facilities offered by Sen’s concept of instrumental freedoms, we were thus able to undertake a more nuanced analysis of the Langa trading environment.

5.1.2 Social opportunities

Social opportunities identifiable in the women’s trading and social environments included formal municipal services such as social housing, public health services, refuse removal and the designation of the Langa taxi rank area as a market hub. Less formal networks such as traders’ associations and community and family social support networks provided opportunities which included trading skills, market access and social capital. The CA moreover advocates for normative values, such as the values of “ubuntu” that guide social relations and foster well-being outcomes, in ways that an ST lens would not have adequately detected. Ubuntu is “a philosophy of being that locates identity and meaning making within communalism rather than individualism. An ubuntu value system espouses the connectedness amongst individuals who view each other as deeply interdependent” (Makalela 2018, 827). A strong sense of belonging to a collective influences the individual through reciprocal and interdependent relationships. For example, in our study community interdependence and interpersonal interactions based on “ububele” (or kindness), “inhlonipo” (or respect) and the African belief that “umntu ngumntu ngabanye abantu” (a person is a person through other people) are key to achieving the type of social bargaining necessary for running trading businesses in African contexts. These values were also intrinsic in creating opportunities for the women traders to exercise their social identity and cultural practices. Thus, the concept of instrumental freedoms enabled the detection of specific and pertinent facets of ubuntu relevant in the township cultural context and provided the theoretical tools for a more nuanced analysis of the women traders’ opportunities.

5.1.3 Protective security

Protective security as freedom encompasses institutional arrangements aimed at preventing vulnerability and falling into abject poverty. In South Africa, the Department of Social Development (DSD) provides a social protection system aimed to alleviate and reduce poverty, vulnerability, social exclusion and inequality (Department of Social Development 2021). At one time or another, all the women traders in our study said they had accessed and received state welfare grants (Mpofu-Mketwa 2020). Examples included monthly old age pensions, child support grants and disability grants. Notwithstanding the women’s perceptions that state grants were inadequate, they reported that the grants supplemented their household incomes during hard times. While the South African socio-historical structural context engendered inequality, poverty and unemployment, CA provided us with a pragmatic analytical tool for assessing the extent to which state social protective provisions for poverty alleviation were supportive.

5.1.4 Transparency guarantees

Transparency guarantees refer to the freedom to deal with one another governed by guarantees of disclosure and openness (Sen 1999). Transparency guarantees were manifested in the various ways that community members appraised government institutions that served them. The women traders demonstrated their activism against poor state service delivery by participating in housing protests as forms of collective agency (Mpofu-Mketwa 2020). Some participants also reported exposing nepotism and corrupt practices by some leaders of the local traders’ association in the way trading stalls in Langa were allocated (Mpofu-Mketwa 2020). Community residents were bound by norms of fair practice, with the women traders alluding to “fighting” corrupt practices as a way of establishing and maintaining fair structures or practices in their trading system. Incorporating in our theoretical framework transparency as an instrumental freedom tied to social justice highlighted these aspects in the women’s trading environments, thereby complementing ST.

5.1.5 Political freedom

Although some women traders freely expressed their allegiance to political parties of their choice, they were not actively involved in political activities beyond occasionally attending rallies and wearing T-shirts depicting political party logos (Mpofu-Mketwa 2020). Opportunities relating to political freedom therefore presented mainly in the form of the women’s perceptions of their empowerment, including the right to participate in socio-economic activities outside the home and soliciting statutory services to protect themselves against gender-based violence (Mpofu-Mketwa 2020). Some women traders actively participated in community structures such as residents’ associations, street committees and traders’ associations, which organised around local issues that affected their members (Mpofu-Mketwa 2020). Belonging to traders’ associations, for example, helped to garner bargaining power with the municipality and protection of their trading rights (Mpofu-Mketwa 2020).

Overall, the women traders in our study linked their freedom to a broader socio-political freedom to participate in political and socio-economic activities in South African society. These findings demonstrate that although progress in social institutions is slow in achieving gender equality and combating gender-based violence, gender consciousness was well developed among the women in our study. Thus, opportunities for gender equality and knowledge about rights to statutory protection from gender-based violence demonstrate the value of integrating CA’s notion of political freedom in our theoretical framework by highlighting the extent of women’s empowerment.

The above discussion illustrates some of the specific ways in which CA helped us investigate how the women traders actively addressed their community’s constraints and resources, thereby uncovering the pertinent nuances of economic facilities, political opportunities, social arrangements, protective security and transparency guarantees. We would probably have missed these nuances had we relied solely on ST’s abstraction of opportunities as rules and resources.

5.2 CA as an evaluative framework for quality-of-life assessments

While ST helped us to understand how the social world is interwoven within societal structures and human agency, the CA helped to illuminate (and assess) the human development practice of enhancing the social world. As an evaluative framework, the CA gauges poverty, inequality and quality of life (Robeyns 2005). Quality of life assessments focus on how development projects impact on the poor in relation to achieving well-being outcomes such as getting nourishment, and gaining confidence and autonomy in decision-making (Sen 1999). These aspects were crucial in our study to assess the extent of the post-apartheid improvement in quality of life associated with the women’s agency in running their trading businesses. ST plays the ontological role in the theoretical framework by explaining the mechanics of social reproduction, and ultimately explaining how society works. As such, it is not concerned with the pragmatic aspects, such as quality of life measurements at the micro level that were crucial in our study to ascertain how agency contributes to poverty alleviation. To assess the quality of life and well-being outcomes, we asked the following interview questions, for example:

1. What goals have you set for yourself which you expect to achieve from the operation of your business?

2. How does your family feel about you operating your business?

In response to question 1 for example, one woman stated:

You need to start to be independent because in the homes that we live, marriages end. You get divorced or a boyfriend leaves you, if you have a boyfriend … Do not depend on a boyfriend or a husband if you are married. (in Mpofu-Mketwa 2020, 127)

This response reflected the goals of self-reliance, independence and a sense of financial security that motivated the participants to run trading businesses. The evaluative component was crucial to the study and informed our data collection and analysis.

5.2.1 Multi-dimensional approach to poverty—quality of life

Evaluation of improvements in quality of life and well-being outcomes calls for understanding poverty in its multifaceted nature, beyond income deprivation. While Giddens does well to highlight class-based constraints, constraints associated with power, allocative and authoritative resources to explain causation of poverty, in ways that are relevant for our study, the CA strengthened the analysis by illuminating other multi-dimensional aspects of poverty that we evaluated to measure transformation. These included unequal distribution of power and income in households, hunger, lack of clothing, shelter, illness, insecurity, and lack of access to clean water and sanitation (Sen 1999). Writing on the application of the CA in development practice with women, Robeyns (2006) demonstrates how CA delineates less salient capabilities or opportunities such as physical health, mental well-being, social relations, education, domestic work, shelter and environment leisure activities, and time autonomy. Using education as an example of a development strategy, Giddens portrays the benefits of education in terms of investment in human capital and means to earning income. Sen, however, goes further to advocate education as a tool for expanding people’s capabilities beyond the labour market to include empowerment, awareness of rights, and psychological well-being, especially for women in the Global South (Robeyns 2006). For some women traders in our study, quality of life improvements that derived from running trading businesses included the following: knowledge acquisition, gaining autonomy in decision making, peaceable familial relationships, gaining a sense of aesthetic fulfilment from business ownership, developing a conduit for practising cultural values and other various well-being outcomes (Mpofu-Mketwa 2020).

Another advantage of working with the CA is that it does not specify or prioritise certain capabilities over others; rather, it allows the poor and marginalised to determine for themselves what they perceive as valuable capabilities. Thus, Sen has subjected this endeavour of developing a rank of capabilities to inclusive democratic processes and social choice procedures (Robeyns 2005, 2006). For the women traders, their narratives illuminated how trading activities improved aspects of their lives and which improvements they perceived as more valuable (Mpofu-Mketwa 2020). They constructed their own meanings about their well-being, quality of life and capabilities that mattered to them. Capabilities that were deemed valuable were formulated empirically, thereby making the application of the CA geographically and methodologically context specific.

Drawing on Giddens’s notion of transformative capacity of agency, it was important to evaluate improvements in quality of life and achievements in well-being outcomes from trading businesses. Transformation encompasses making a difference to pre-existing circumstances through actions. Giddens (1984), for example, asserts that people act to change pre-existing structures, systems and institutional environments by using resources and rules provided by the structures to either maintain or slightly change the structures, thus highlighting their transformative capacity. Despite the survivalist nature of all the trading businesses in our study, evaluations of the women’s’ well-being outcomes achieved from the businesses showed improvements in quality of life (Mpofu-Mketwa 2020). For example, the income generated ranged from a net monthly income of less than R2000 to R5000 per month. Ten out of 25 women traders earned between R4000 and R5000, which is significant in the context of poverty and multiple deprivations (Mpofu-Mketwa 2020). Tangible life improvements for the profile of participants who generated the most income included savings, purchasing affordable houses or cars, home improvements and children’s education (Mpofu-Mketwa 2020). Other improvements included performing costly traditional rituals or obtaining business and life skills training such as driving, thus enhancing opportunities for class mobility and enjoying better familial relationships (Mpofu-Mketwa 2020).

The CA thus enhanced the ST as a theoretical framework in that evaluations of transformation shed light on other well-being outcomes beyond income such as improving difficult family relationships, practising cultural beliefs and sending remittances to extended family members in the rural Eastern Cape (linked to social identity and family obligation), thereby accessing empowerment opportunities and making autonomous decisions. It was necessary to evaluate these multi-faceted well-being outcomes because in South Africa’s deprived communities, poverty is more than just income deprivation. Poverty incorporates social ills such as crime, food insecurity, domestic violence, and the eradication of social values associated with “ubuntu” such as neglecting to care for extended family and elderly parents (Mosoetsa 2013; Schneiderman et al. 2020).

Analysing the data through the lens of CA revealed how their trading businesses transformed the women traders’ quality of life and addressed multi-dimensional needs beyond just income and consumption of commodities and other related economic variables such as Gross National Product (GNP) and economic growth (Mpofu-Mketwa 2020).

5.3 Gender sensitivity in researching women’s capabilities

In ST Giddens foregrounds class stratification in a capitalist society to explain social reproduction. Giddens (1984) used the example of how rebellious working-class schoolboys situated in a poor British neighbourhood in the 1970s engaged in a classroom disruptive and counter-school behaviour. Eventually, the lads became school dropouts and were employed as low-level manual labourers, and reproduced their working-class culture (Giddens 1984).

A class analysis of social reproduction in ST makes sense in explaining Willis’ example. It does not work as well for gendered stratification because if we draw our understanding of gender-based violence and patriarchal exclusionary practices solely based on the duality of structure, this would imply that victims of gender-based violence are active agents of the violence they suffer at the hands of their abusers. In South Africa, for example, gender-based violence operates cumulatively at familial, cultural and societal levels. Cultural practices that condone violence against women and legislative systems that exclude women from participating in socio-economic activities in the global south are well documented (Moffet 2008). In our study, we drew on Giddens’s understanding of structures as an “institutional feature of society” to frame how markers of societal organisation such as class, gender, race, culture, ethnicity, power and patriarchy influenced isiXhosa-speaking women traders’ practices, activities and boundaries that rewarded or sanctioned these practices. Giddens’s conceptualisation of structure undoubtedly helped us to understand the relationship between social action and broader societal organisation across time and space. For example, Giddens’s explication of structures as rules and resources includes gender dimensions. Regarding culturally appropriate rules that regulate respect and support in social relations, including family and marital relations, supporting extended family members indicates agreement between Sen and Giddens insofar as broadening the sense of structures as enabling and expanding opportunities beyond income.

However, for our study, which assessed human development, women’s empowerment and the opportunities for enablement, it was important to include theoretical elements that could sufficiently inform and pinpoint specific contextual factors in the women’s trajectories to improving their quality of life. We once again turned to Sen’s CA to complement ST in order to deepen the understanding of patriarchal hegemonies for the purpose of practical women’s empowerment interventions.

Informed by our theoretically blended conceptualisation of structural constraints, we included in our qualitative interviews gender-sensitive questions such as:

1. What are your challenges as a woman in your family and community?

2. As women, we sometimes feel we need to be freer to make decisions about our lives; what challenges from your family do you face when making decisions about your business or other aspects of your life?

The above questions sought to solicit responses about how the women traders experience gender-sensitive constraints and constraints associated with power that came from institutions such as the family or the municipality. For example, the response below demonstrates some of the gender-based challenges:

The problem is the money I get from here; I cannot save it as I wish … He [husband] does not give me money for the children. So, I need to take the money and look after the children; milk, nappies, clothes and everything. (in Mpofu-Mketwa 2020, 169)

Notably, our findings support the results of studies (see Robyeyns 2006) that used CA in poverty alleviation projects with women by revealing inequalities in the distribution of income in households. As an inclusive and gender-sensitive approach to human development, CA strengthened ST by incorporating gender-specific constraints such as the lack of access to financial capital, sanitation needs, shelter, vocational skills and health (Kabeer 2016). In addition, CA advocates for collective agency for women in the form of membership in community associations, to garner more bargaining and organisational power to enable them to challenge cultural norms and practices that disadvantage women (Kabeer 2016).

These gendered dimensions of poverty highlight the intersectional nature of oppression based on race, class and gender which incrementally reinforce poverty. Thus, the women traders in our study experienced constraints that emanated from them being black, from low socio-economic status and from being women. For example, some of the women traders in our study reported leaving school early, before acquiring formal aspirational qualifications for alternative career paths, because of the cultural practice of “ukuthwala” (a traditional custom of arranged marriages) (Mpofu-Mketwa 2020). Some of these women also reported experiencing verbal, emotional and financial abuse from their husbands/partners (Mpofu-Mketwa 2020).

With help from CA, we uncovered how the specific and pertinent cumulative effect of poverty, familial circumstances, limited schooling opportunities, cultural practices, patriarchy and structural violence constrained the women traders. Limited opportunities for schooling (i.e. less than three years of high school education) were associated with less innovative entrepreneurial skills and practices, which impacted negatively on their earning capacity. Class-based constraints derived from ST that were evident in the study included living in poor-quality housing such as informal settlements, and poor cash flows exacerbated by the reluctance of financial institutions to support working-class women, with no collateral, in their business ventures. Constraints associated with power entailed competition amongst traders and with more established competitors, aggressive customers, bureaucratic municipal regulations and controls, and unequal distribution of resources within some households. We also drew on ST to help us understand how the women exercised agency in responding to these constraints. Alongside these constraints, most women traders often experienced what Giddens terms material constraints related to physical well-being and the emotional labour of caring for sick family members. Sen’s concepts of “instrumental unfreedoms” and “ill beings” as well as quality of life measurements and multidimensional poverty indicators armed us theoretically and methodologically to undertake more nuanced analysis of the context-specific constraints that women traders experienced. Gendered constraints included combining trading activities with household labour and care responsibilities, unequal distribution of income, and the lack of autonomy the women had in their households regarding their time and their ability to pursue other socio-economic activities.

Giddens (1984) discusses power as sanctions of various kinds, which vary in intensity from the direct application of force or violence to the mild expression of disapproval. Social actors use power to control resources or influence the circumstances of others in society (Giddens 1984). Giddens explains the causes of domination by those holding power, for example through discourses of surveillance, capitalism, hegemony, reification and socialisation (Giddens 1984). These explanations of causes of domination are very useful to understand power dynamics at macro level. To gain a micro-level understanding of the manifestation of unequal power relations across societal institutions, communities and down to the level of households, Sen offered us a more context-specific micro analysis that was useful for our analytical purposes (Robeyns 2006; Sen 1999, 2000). While ST helped us understand the nature and incidence of power, the CA helped us assess people’s capabilities in micro contexts such as households.

Power struggles manifested in our study at the familial level, for example when some of the women’s spouses control their income in the household context and at a community level when the local municipal law enforcement members apply sanitation and regulation mechanisms to confiscate or destroy their merchandise (Mpofu-Mketwa 2020). While some women rejected power constraints emanating from traditional cultural practices of “ukuthwala” for example, by leaving forced marriages, others accepted the practice as honouring their cultural values (Mpofu-Mketwa 2020). In analysing the women’s responses to constraints associated with power, some women were persuaded to accept the cultural norms of “ukuthwala” as part of the natural order. This constitutes what Giddens refers to as reification or “forms of discourse which treat properties of social systems as objectively given in the same way as natural phenomena,” thereby denying social actors the power to change their circumstances (Giddens 1984, 180).

Sen thus complemented Giddens in enhancing our understanding of the concept of constraint by identifying constraints germane to women in local contexts, such as the distribution of power in households, which aligns with Giddens’s notion of constraints associated with power and authoritative resources. Sen’s advocacy for a pluralist approach to development, by addressing a multitude of deprivations such as social inequalities and social injustices (see Robeyns 2006; Sen 2000), is pertinent to the South African context of the study. However, the CA is limited in illuminating the ontological aspects of social reality, which is a feature of ST. Consequently, CA is likely to neglect how social injustices, social inequalities and unequal power relations historically came into existence, which ST would explain, for example, through discourses of surveillance, authoritative resources and hegemony.

6. Conclusion

In our assessment, had we relied exclusively on Giddens’s ST for our theoretical framework this study would have generated some interesting findings. Nonetheless, these findings would have been far less nuanced than the more specific and pertinent final results generated. These nuanced findings emerged largely because we developed a theoretical framework that grafted aspects of the CA into ST and made it “fit for purpose.” In this paper, we drew on our study of isiXhosa-speaking women traders to illustrate both the appropriateness and pragmatism of theoretical grafting. The grafting of carefully chosen theoretical elements of Sen’s capability approach into Giddens’s structuration theory enhanced the theoretical framework of this study, which improved the quality of our analysis and enabled us to generate more nuanced findings about the social phenomenon under investigation. However, we end with a warning: Theoretical grafting certainly can enhance analysis, but it should be used cautiously to avoid theoretical fragmentation.

Acknowledgements

The authors acknowledge and thank the women traders from Langa who participated in the study. We express our sincere gratitude to the SARS editorial board and the anonymous reviewers who reviewed our first and subsequent submissions; their comments were invaluable in strengthening the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Margaret McNamara Memorial Fund; UCT Humanities Completion Grant; Eric Abraham Refugee Scholarship. Next Generation Social Sciences in Africa (Research grant and Dissertation Completion Grant) and Mellon Mays Undergraduate Fund (MMUF) Predoctoral and Travel & Research Grants as funders.

Notes

1 This paper draws on the first author’s doctoral research. The second author was the thesis supervisor.

2 The first author has lived in Langa and undertook research in the community. The second author has family in Langa, is actively involved in a civil society organisation in the community and has undertaken research there.