Abstract
This article considers research into barriers to learning (including HIV/AIDS) in a small, rural town in KwaZulu-Natal, South Africa. A variety of qualitative participatory research methods were used, including photovoice, a method in which research participants take photographs and then decode these together with the researchers. Rich, thick data was obtained using photovoice, and the researchers found this method particularly useful for dealing with the ‘unspoken’ and working with marginalised people. This method, particularly because of the emotive nature of photographs, is also a potentially powerful political tool in exposing and exploring deepening levels of poverty and crisis experienced by the marginalised in post-apartheid South Africa. Photovoice as a method, however, raises issues of ethics and researcher-researched power dynamics; in particular, whether it is ethically acceptable to use photographs from consenting participants in light of the imbalance of power between the subject and the researcher, particularly in the context of HIV/AIDS. This article explores ethical and power issues in using the photovoice method.
Keywords
In 2004, staff members in the Faculty of Education and School of Psychology at the University of KwaZulu-Natal in South Africa embarked on a two-year research project entitled Mapping HIV/AIDS as a Barrier to Learning and Participation in Basic Education. Using a case study approach (the research focussed on a small, rural town and the surrounding area), the project intended to examine the extent to which HIV and AIDS are barriers to learning, and how they interact with other exclusionary factors or other barriers. The project hoped to extend our knowledge of HIV and AIDS in education at a local level and inform school and community policies and strategies in the community.
The project involved adult and child learners, parents/caregivers, teachers, school management staff, and members of the school governing bodies at two pre-primary schools, six primary schools, three secondary schools, one so-called special school, and three adult basic education classes in Richmond, a small rural town about 35 km south of Pietermaritzburg, the capital of the province of KwaZulu-Natal, South Africa. Government officials and the broader community were also involved in the study, as were local NGOs, health facilities, political representatives, church leaders, traditional leaders, and other local leaders within the broader community. Members of an HIV and AIDS support group were also involved.
The small town focussed on in the research serves the farming and forestry communities, and it is surrounded by semi-formal and informal settlements, which house about 70,000 people, and outlying farms and rural settlements. There are more women (53%) living in the area than men, roughly in keeping with the ratio elsewhere. Also in keeping with general population trends in the province (and the country) is the fact that the population is relatively young—11% were under the age of 5 and 46% were under the age of 20 at the time of the research (Richmond Municipality, 2002). The area has been hard hit by the HIV and AIDS pandemic and is situated in a province with the highest HIV-positivity rate—nearly 40% in 2005 (Department of Health, 2007). In addition, when we undertook the research, 38% of the economically active population in the area was unemployed. Not surprisingly, household income was low, with 77% of households earning less than R1500 per month (Richmond Municipality, 2002).
The area came to prominence during the 1980s and 1990s as a flashpoint of political violence, first between the United Democratic Front (the African National Congress (ANC)-supporting, anti-apartheid movement within the country at the time) and Inkatha (an apartheid-backed, ethnic-based bantustan party), and later between the African National Congress and the United Democratic Movement (a splinter group from the ANC) (Aitchison, 2003; John & Rule, 2006; Taylor, 2002). John and Rule, two colleagues who were also involved in the research project, argue that the violence has had far reaching effects, including displacement, fractured families, increased levels of substance abuse, and a high level of sex work—all of which obviously impact on levels of HIV and AIDS. They show how the private discourse of deterioration, fragility, and despair uncovered by the research is overlaid by a public discourse of recovery, stability, and hope (John & Rule, 2006). Thus, as far as the media and government officials are concerned, the area was experiencing a period of peace and reconstruction. Our research findings tell a different story.
Research Approach and Methods
The research project was intended as an in-depth case study of the kinds of barriers to learning faced by children and adults within a particular community, and the ways in which these barriers intersect with and/or reinforce each other. It was suspected that HIV and AIDS would be a significant barrier, and that issues of stigmatisation and discrimination connected to HIV and AIDS meant that this would be a particularly difficult barrier to research. The in-depth case study approach was chosen because it is particularly appropriate under these circumstances, and the research was intended to inform school and community policies and strategies.
In part, the choice of this approach was in response to an article written in 2002 by Rob McBride, who argued for more in-depth research in education for development because it encapsulates much that emancipatory researchers are interested in, such as ‘ordinary’ people, power, and the relationship between the researcher and the researched:
Like much qualitative research, [in-depth research] is interpretive in that it deals with the perceptions of ordinary people in their daily lives, including both routine and problematic activities. Much in-depth research is small scale, dealing with a single case. It recognises the reality of power in the research situation, between the researcher and the researched (whom it endeavours to put on as level a footing as possible). (p. 144)
The researchers involved in Mapping HIV/AIDS as a Barrier to Learning and Participation in Basic Education argued that using an in-depth case study approach and participatory research methods would complement the more numerous quantitative, macro-level studies on barriers to education and would overcome some of the difficulties associated with researching HIV and AIDS—a ‘hidden’ or ‘taboo’ subject because of stigmatisation and discrimination.
The researchers saw the use of participatory methods as essential. Participatory methods attempt to overcome the power imbalances between the researcher and the participants in the research. Rather than situating researchers as professionals acting upon rural, often unschooled, community members-as-objects, the use of creative, multiple participatory tools gives participants a voice. For example, the interaction between group members around specific tasks, such as completing a timeline, generates shared information and enables new perspectives to be gained. According to Babbie and Mouton (1998), the key components in this approach are “participation, engagement, involvement and collaboration—with participants involved in the research as equal partners. The participants are co-researchers whose insider ‘local knowledge’ is valued for sense-making” (p. 64).
Regarding specific participatory research techniques, Seidel and Coleman (1999) make fruitful use of narratives and storylines in ‘envoicing’ HIV-positive men and women in rural KwaZulu-Natal. Huber and Gould (2003a, 2003b) use children's drawings, including maps and timelines, to investigate school non-attendance of orphans in Tanzania. Many of these methods (e.g., timelines, drawings, in-depth interviews, among others) were used in the research, precisely because of their potential to uncover the ordinarily hidden or unspoken.
The Photovoice Method
The photovoice method was developed by Caroline Wang, with Mary Ann Burris (Wang & Burris, 1994, 1997), and further developed with various colleagues (Wang, Morrel-Samuels, Hutchinson, Bell, & Pestronk, 2004; Wang & Redwood-Jones, 2001). Photovoice as a research method rests on the assumption that people themselves can best identify and represent their own realities, and it was developed as a participatory research method primarily for undertaking participatory needs assessment (Wang & Burris, 1997). The photovoice method involves giving research participants (individuals or groups) cameras for a period of time with which they are encouraged to photograph anything they think is of significance to the study. The photographs are then developed and, in a workshop setting, the participants are asked to talk about the photographs they have taken. In this way, participants are able to record and reflect on what they see as the most important information, and they then enter into a dialogue among themselves and with the researchers. This dialogue enhances the researchers' understanding, as well as the participants', and can be used to reach policymakers (Wang & Burris, 1997).
A number of claims have been made about the photovoice method. The initial developer of the method, Caroline Wang, describes photovoice as a participatory action research strategy which empowers communities to influence policymakers (Wang, 1999), whilst a key argument is that photovoice “shift[s] power in the research process from the researcher to the researched” (Nelson & Christensen, 2009, p. 36) and can lead to tangible outcomes to the benefit of the individual participants and their communities:
The structure of Photovoice lays the foundation for participants to function as partners and eventually leaders as the participatory research process progresses …. This process resulted in shared power between the “traditional academic researcher” and the community, fostered trust, and promoted a sense of ownership in the community, which in turn builds capacity for social change. (Newman, 2010, p. 57)
Wang and Burris (1997) developed the concept using theoretical literature on critical consciousness (in particular, the work of Brazilian educationist Paulo Freire), feminist theory, and documentary photography. They take this further through the method of having the research participants take the photographs themselves. The method also arises from what they see as the limitations of photovoice, which include the difficulty of interpreting photographs and possible subjectiveness of the photographs (e.g., people might self-censor or misrepresent for their own interests). Essentially, the method involves a number of steps:
Training.
There can be one or more training sessions. The first should include a discussion of the cameras, ethics, and power (including the need to ask for permission); the different ways photographs might be seen; and the fact that copies of the photographs should be given back to community members. Training should also be given in using the technology (e.g., how to look after the camera and how to use it), but Wang and Burris (1997) recommend that technical advice should be minimized “to avoid stifling people's creativity” (p. 378). A discussion of how the photographs will be used should also be held.
Taking the photographs.
Participants take photographs that they believe best depict the issue at hand, for example, community needs.
Participatory analysis.
This is a three-stage, participatory process involving selecting, contextualizing, and codifying. The participants choose which photographs they consider most significant. They then contextualize these by telling stories about the pictures. Wang and Burris (1997) use the acronym VOICE to describe this stage—“voicing our individual and collective experience” (p. 381). The stories can be spoken to others or written down. In the codifying stage, participants use three dimensions to consider the images: issues (that are pragmatic, immediate, and tangible and can be targeted for action), themes/patterns across the images, or theories that can be developed. Community needs can then be collectively prioritized and action determined.
Since Wang and Burris' first discussion of the method in 1994, photovoice has become an increasingly popular research tool; Catalini and Minkler (2010) report that “the practice of photovoice is growing rapidly and the photovoice literature is proliferating” (p. 447). Two recent articles have reviewed studies using photovoice. Hergenrather, Rhodes, Cowan, Bardhoshi, and Pula (2009) reviewed a total of 31 studies that used photovoice (cut-off date, April 2008), whilst Catalini and Minkler (2010) reviewed 37 articles (cut-off date, January 2008).
These reviews show that photovoice has been used with participants of all ages, in different parts of the world, and to consider a variety of issues, though most issues relate to health and social justice (Catalini & Minkler, 2010; Hergenrather et al., 2009). Although most of the studies use a fairly similar process, Catalini & Minkler's review found that the majority alter Wang's original method to suit their specific project. Thus, not all include training or group discussion of the photographs (although most include both). Hergenrather et al.'s review suggests that some of the further developments of the method by Wang, in particular the group discussion of the photographs, have been increasingly adopted. For example, they found that about a third of the studies they considered used the SHOWED method of root-causes questioning—What do you
Both reviews found that most studies reported at least one outcome, usually involving the identification of community issues, the identification of key advocates or policy makers who could be approached, or individual empowerment. Catalini and Minkler (2010), however, found little attempt in the studies they reviewed to evaluate the long-term impact of the method on individuals or communities.
In recent years, several manuals on the method have been developed by organisations that have used the method and wish to encourage others to do so (Blackman & Fairey, 2007; John Humphrey Centre for Peace and Human Rights, 2010; Palibroda, Krieg, Murdock, & Havelock, 2009; Photovoice Hamilton, 2007). These are primarily guides to using the method as developed over time by Wang and her colleagues, and these guides make no significant changes.
Photovoice in Our Research Project
The Method
The photovoice method as used in our study consisted of four basic steps:
Explanation of the research project and its purpose, distribution of the cameras (we used disposable cameras with a built-in flash), and briefing of the participants in their use (including the ethical issues involved). The participants were asked to take photographs depicting their work as community volunteers, members of an HIV support group, and community health workers.
Taking of photographs by the participants over a period of time.
Collection of cameras and development of film.
Workshops with participants to talk about the photographs (e.g., why they took them and what they thought the photographs showed).
It should be noted that very few people in the small town we focussed on have access to a camera, and photographs of oneself or one's loved ones are few and far between. This provided both advantages and disadvantages. Most people had never handled a camera before and required basic training (e.g., click the button to take a photograph; don't move the camera when taking a photograph; take pictures with the light behind the photographer; be careful not to put fingers/thumbs in front of the lens, etc.). On the other hand, the novelty and excitement of being trained and taking photographs was very apparent, as was the willingness of people to be photographed.
The participants were told that they could take a certain number of ‘personal’ photographs, which they would be allowed to keep at the end. This was, firstly, because we felt uncomfortable about providing a technology which was so keenly welcomed, only to remove it with no tangible benefits to the photographers themselves, and, secondly, to increase the chances that the cameras would indeed be returned to us intact!
We used the photovoice method with three separate groups of research participants during the research: (a) drop-in centre volunteers, (b) support group members, and (c) community health workers.
The drop-in centre volunteers
The local Day Care and Support Centre, locally known as the drop-in centre, was officially opened in September 2003 as a joint project of the government Departments of Health and Social Welfare. In addition to the director, about 18 volunteers staff the drop-in centre, cooking meals for the children who ‘drop-in,’ identifying further vulnerable families, providing home-based care, and delivering food parcels (in accordance with state policy). Volunteers also help community members through the process of applying for official documentation, such as birth certificates or identity documents and social grants. The Centre serves the communities in relatively formal, urban areas of the town.
A focus group held in September 2004, which consisted of 14 volunteers from three different areas, was also used to brief the volunteers about the photovoice method. Three disposable cameras were distributed, one to each of the areas. The volunteers were shown how to use the cameras and were instructed to share among themselves the 27 photographs on each camera. The volunteers were instructed to ensure that they asked for permission before they photographed anyone. They were given one month to take their pictures. The cameras were subsequently collected by the researchers and the photographs developed.
A workshop was held in December 2004 for the volunteers to look at the photographs they had taken and to write about and present each photo. Follow-up questions were asked by the researchers present. This workshop was recorded on audio cassette and video, and subsequently transcribed.
The support group members
The Support Group involved in the project is a support group started in 2000 for people infected and affected by HIV and AIDS. Members of the group are actively involved in educating people in their own communities about HIV and AIDS and offering support and home-based care to others infected and affected by the syndrome. The public clinic provides medical supplies needed by the members of the group for this work. They also run a vegetable garden at the local clinic, which provides them with fresh vegetables for themselves and the families they help. They also seek ways to increase their own income and the income of the group, for example, through beadwork.
The Support Group had been identified from the start as an important group of research participants. When an initial meeting was held with key members regarding the research process and their possible participation, the use of the photovoice method was discussed and group members were keen to try it. A workshop on handling the cameras was subsequently held with members of the group. The group members were also given their brief, which was to take photographs that depicted something of their work in the community. The need to ask for permission before taking a photograph of a person was again emphasised.
Members were given the cameras for three weeks. These were collected and the film developed. A workshop was then held to decode the photographs. Instead of asking participants to decode all of the photographs, as had been done with the drop-in centre volunteers, the researchers asked them to pick five photographs they would like to talk about. This was done purely because of time constraints.
The community health workers
In terms of the government's primary health care policy, community health workers (CHWs) are community members who are trained to provide a very basic level of health care (e.g., ensuring TB medication is taken, weighing babies, providing basic home-based care) and education to members of their communities. They are generally paid a small stipend for this work, and they are supported by the local clinic. Community health workers are supposed to be between 25 and 55 years of age, and may be men or women. They are supposed to have completed ten years of schooling.
We chose to work with CHWs in a rural settlement, near the small town where the research was being conducted, in which land is held communally under the control of a traditional leader. As with the drop-in centre volunteers, a group of the CHWs from the rural settlement attended a focus group at which they were briefed about using the camera (only four CHWs were involved, so only one camera was given). After a few weeks, the camera was retrieved and the film developed. A workshop was then held at which the CHWs wrote about the photographs. During the break, their texts were translated to the researcher (in this case, myself), and participants then presented their photographs and were asked questions about the pictures and their comments on them.
The Data
The data consist of the photographs taken by the participants; their own written notes about each one; and the transcriptions of their presentation of the photographs and the question-and-answer session following their presentation. In fact, we made little use of the photographs per se. Catalini and Minkler (2010) reported that, in the studies using photovoice that they had surveyed, the photographs were rarely used as a source of data, and the discussions about them generally provided the main source of data.
Analysing the Data
In their review of studies using photovoice, Catalini and Minkler (2010) reported that one of the weaknesses in the literature concerned analysis, with most studies failing to provide sufficient information about how they moved from the photographs to the findings, although “the photovoice process is often valued for its ability to uncover rich descriptive information” (p.441).
We used conventional tools of analysis on the texts produced—the written explanations of the photographs by the participants and the transcribed presentations and discussions. The texts were loaded and thematically coded using NVivo. The codes used were derived from an intensive process by the broader researcher group. The data set as a whole provided a range of information, including some basic quantitative demographic data regarding household membership, household income, and school attendance. For example, of the 20 households portrayed in the volunteer photographs and explored in the discussion, 13 were female-headed households; three were headed by children under the age of 18; five were headed by women over the age of 60; and half of the households had six or more members. Of the 20 households, three received no income at all, nine received intermittent income from occasional, temporary work, and six received government grants (although 16 were eligible). Furthermore, 18 of the 20 households could be seen as disjointed or fractured, with one or both parents dead or missing.
The photographs and the participants' analysis provided significant contextual data. But to what extent did the method provide data to answer the actual research questions? The data revealed that the most significant barriers to school attendance were abandonment and disjointed families, death in the family and/or being orphaned, and poverty, unemployment, or underemployment. Political violence, illness, and the lack of certification (such as birth certificates, death certificates, and identity documents) necessary for accessing government grants were also cited as problems.
Many of the stories told by the volunteers about the households in which they work suggest high levels of child mobility—children being moved from one home to another depending on current circumstances and who is prepared to take them in. The stories told by the volunteers also suggest a significant level of child abandonment, usually with another member of the family (primarily the child's grandparent). Abandonment has in many cases compounded poverty, which has had the direct result that many children are simply receiving too little to eat. This affects both their ability to attend school and their ability to perform at school. This problem was something the volunteers emphasised.
Many of the households portrayed in the photographs had been affected by the political violence that swept the area in the 1990s. Firstly, many children now live without one or both parents, either because they were killed during the violence or are now in jail as a result of their involvement in the violence. In some cases, children live with adults/caregivers who suffer the psychological scars of the violence. Secondly, it is not difficult to imagine the psychological trauma experienced by children during the period of the violence, particularly those who personally experienced it.
The photographs thus revealed the ways in which and the extent to which the context—of disjointed families, high unemployment levels and low income, and the legacy of violence—directly impacts on the ability of children to attend school. Furthermore, the photovoice method also allowed an exploration of the complex ways in which the barriers coalesce. For example, the inability to attend school because of poverty is compounded by abandonment by the child's mother. This is then compounded by the death of the granny. The child is subsequently cared for by a family of four children, themselves orphans (and presumably caring for the child at least in part in order to access food parcels from the Drop-in Centre). In another example, political violence has led to displaced children, but also resulted in deserted houses into which other displaced families have been able to move.
Ultimately, the method provided useful data, including quantitative, for answering our research questions, literally making visible the ways in which this community experiences barriers to learning (including HIV and AIDS) and their interrelationship. Our analysis has been reported on more fully elsewhere (Jacobs & Harley, 2008). My concern here is rather with the deeper ethical issues related to the photovoice method.
Power and Ethics
Kellehear (1993) argues that “all research has social implications” (p. 12) and Schratz and Walker (1995) argue that research is social action, not disconnected from it. This research project was perhaps more conscious of this fact than is often the case, although the extent to which the research was seen as part of an overtly political project varied considerably amongst members of the research team.
I, myself, do the research I do in order to help bring about change in the world to benefit the oppressed. Thus, my interest in this particular research project was primarily political; in particular, I wished to highlight the deepening levels of poverty and crisis in ordinary communities which have been brought about by the South African government's neoliberal economic policy or, in a paraphrase of John and Rule's (2006) argument, to show the private discourse that is overlaid (silenced?) by the public discourse.
The political transition in South Africa is widely held up as a glowing example of a successfully negotiated settlement, as a result of which a ‘rainbow nation’ now basks in a variety of political, social, and economic freedoms denied them under apartheid. What the political settlement actually did was to entrench the power of capital (Bond, 2000; Greenberg, 2004). True, that power is now shared among different races (not all of the wealthy are white—indeed there is a growing, and increasingly conspicuous, black middle class), but, nevertheless, the poor under apartheid remain poor, if not poorer, under the new regime, with the gap between rich and poor continuing to grow (Ntshingila, 2006; People's Budget Coalition, 2005; South African Institute of Race Relations, 2010, 2011).
Within two years of the first democratic election in 1994, the Growth, Employment and Redistribution (GEAR) strategy was adopted as the new government's macroeconomic policy. With the help of World Bank economic advisors, the policy continued a process of economic reform, which included trade liberalisation and privatisation, started decades before by the apartheid government. Since then, what has amounted to a structural adjustment programme has been implemented. However, ironically, this has not been imposed by the International Monetary Fund, but rather voluntarily applied by the ANC-led government (Bond, 2000; Pithouse, 2006).
The result has been catastrophic for the poor. The privatisation of basic services, such as water and electricity, and the insistence on cost-recovery has made the Constitution's guarantee of the right to clean water and a clean environment a joke. As many people who were connected to water and electricity infrastructure in the great roll-out of the second half of the 1990s have now been disconnected. In many cases, this has been done not by the authorities, but through the prepaid system widely imposed on poor communities, which means that those unable to pay ‘disconnect’ themselves (Bond, 2002; Coalition against Water Privatisation (SA), The Anti-Privatisation Forum (SA), & Public Citizen (SA), 2004; McDonald, 2002; Muller, 2004). Trade liberalisation has seen the decrease of tariffs to well below those required in terms of the General Agreement on Tariffs and Trade (GATT) (Nicholson, 2001), which has resulted in the retrenchment of hundreds of thousands of workers and the collapse of sectors such as clothing and textiles (Save Jobs Coalition, 2005). Job shedding, accompanied by jobless growth (for the economy is indeed growing), means an unemployment rate of around 40% (Congress of South African Trade Unions, 2008). There are now over 2 million so-called ‘discouraged’ work seekers—those who want a job but have not looked for work in the last four weeks because they have given up hope of finding one (South African Institute of Race Relations, 2011). And, not surprisingly, rural areas are worst hit.
The results of these economic policies are an increasing level of poverty among the lowest income groups. In a 2004 study, the Human Sciences Research Council (HSRC) showed that households living in poverty had sunk deeper into poverty (People's Budget Coalition, 2005). In 2005, Global Insight Southern Africa published figures showing that the number of desperately poor people had risen from 1.9 million to 4.49 million between 1994 and 2002 (Ntshingila, 2006). Whilst 2002 marked a definite high point in desperate poverty, and since then there has been a decline in the number of people living on less than a dollar a day (the world standard measure of absolute poverty), by 2009 over 42% of South Africans were living in relative poverty. This means that some 20 million people earn less than the calculated poverty income (South African Institute of Race Relations, 2010, 2011).
Within this context, HIV and AIDS inevitably impacts in devastating ways, and this, too, has been well documented. Indeed, as Hein Marais argued in a 2005 publication on the impact of AIDS in South Africa (which argued that it is the poor who are most affected), the discourse of the impact of AIDS often replaces/overlays a deeper contextual analysis—the suggestion is that without AIDS, communities would ‘cope.’ AIDS somehow pushes them over the edge. Whilst by no means questioning the AIDS impact (indeed, he provides powerful evidence of what he calls the ‘mangling’ effects of AIDS), Marais suggests that it has been over-privileged. He is critical of the way in which AIDS is separated out, “hoisting it beyond the other factors that generate wretchedness” (p. 47), and he is concerned that this deflects attention away from the main causes of, for example, food insecurity. He is also scathing in his analysis of much of the literature on the AIDS impact, which he believes is overly simplistic, particularly the notion of households. He suggests that more nuanced research at the ‘household’ level needs to be done—something which, as we have seen, the photovoice method in our research has begun to do.
Marais (2005) is particularly critical of the fetish of ‘coping,’ of ‘resiliency,’ which so much AIDS impact writing displays, and the ways in which this plays into policy and interventions that ignore the bigger picture:
To be sure, some studies have indicated that a partial recovery in consumption levels can eventually occur, suggesting that the households have overcome the shock and are again ‘coping’. But to describe as ‘coping’ the activities of households sunk in impoverishment is to unmoor the discussion from ethics. By any humane definition of the word, such households are not ‘coping’; a ‘successful coping strategy’ becomes an oxymoron. Regaining a precarious and chronically insecure form of household ‘viability’ cannot reasonably be declared a success … . coping strategies actually are not about success—they're about failure. They can enable one to survive, but not to transcend the circumstances that trapped one in the path of mishaps in the first place. Implicit in the discourse of ‘coping’ is an acceptance, an endorsement even, of the way things are, a patronizing gloss on a reality of privation and marginality. (p. 55)
And yet, despite this well-documented (inevitable) trend of making things worse for the poor, the state continues to argue that South Africans are better off, are ‘coping’ fine, although there is growing official concern about unemployment rates, and some stop-gap, temporary measures (such as the food parcels the drop-in centre volunteers, amongst others, hand out to destitute families) have been introduced. Social welfare grants, accessed by millions, continue to be the only way many poor people survive.
As I have said, the trend towards deepening and widening poverty is well documented. How, then, could this research hope to make any kind of meaningful contribution in this regard? I would argue that the research, for me, had several strengths which I believed could make it a useful part of a political project:
the fact that the research was taking place in an area greatly affected by the political and economic context I wanted to write about;
the fact that it aimed to look at a number of issues critically linked to the political economy of the area (and the country), and, in particular, the ways in which these were linked (e.g., unemployment/underemployment and AIDS);
the use of an in-depth, case study approach, which allowed for a very close analysis of how macro forces play out, are negotiated, and are constituted at the local (and even household) level;
the use of participatory methods, which addressed my unease about researcher-researched power relations (although it could in no way be described as a participatory research project) and which were more likely to result in so-called ‘thick’ data;
the inclusion of community volunteers as research participants (as opposed to only teachers, learners, school management committees, among others.)—in other words, an attempt to get beyond elites and power-holders; and
the use of the photovoice method.
Photovoice appealed to me not only because this method spoke to my own feelings and concerns about researcher-researched power dynamics, but also because of the power of the image and its potential for political purposes. As Schratz and Walker (1995) have commented: “Pictures, it seems, have a power that words often lack” (p. 76).
Rob Walker (1993), who has written extensively on the subject, speaking about photographs in general, rather than photovoice in particular, asserts that photographs allow an exploration of affective factors which are often not captured using more conventional language-bound methods:
One of the reasons why I am intrigued by the use of photographs in educational research is that their use touches on the limitations of language, especially language used for descriptive purposes. In using photographs the potential exists, however elusive the achievement, to find ways of thinking about social life that escape the traps set by language. (p. 72)
Writing with Schratz, Walker furthers this argument:
In social research pictures have the capacity to short circuit the insulation between action and interpretation, between practice and theory, perhaps because they provide a somewhat less sharply sensitive instrument than words and certainly because we treat them less defensively. Our use of language, because it is so close to who we are, is surrounded by layers of defence, by false signals, pre-emptive attack, counteractive responses, imitations, parodies, blinds and double-blinds so that most of the time we confuse even (perhaps, especially) ourselves. In this maze of complexity pictures can have a use. (Schratz & Walker, 1995, pp. 76–77)
More than simply overcoming the barriers of language, photographs take the viewer to a particular place at a particular time in ways words cannot: “A photograph … creates an immediate if vicarious sense of being there that is stronger than most readers will get from reading an ethnographic description or selected interview transcripts” (Schratz & Walker, 1995, p. 74).
One of the criticisms of using photographs in research is not simply that they can be falsified (as they certainly can), but that they are highly subjective—as Paul Byers (1966) famously said, “cameras don't take pictures.” In a sense, as I have discussed above, this is exactly the point of photovoice—to shift the subjectivity of the researcher to the researched, to allow the research participant to determine what is useful and important to represent in a photograph. In any case, photographs are not necessarily any more or less subjective than written or oral data:
Like our field notes and other forms of empirical data, photographs may not provide us with unbiased, objective documentation of the social and material world, but they can show characteristic attributes of people, objects, and events that often elude even the most skilled wordsmiths. (Prosser & Schwartz, 1998, p. 116)
Photovoice is thus an extraordinarily powerful method for a number of reasons, as reported on above. It provides a view from the research participants' perspective of their daily lived experience (or their view of their daily lived experience) as it relates to the research topic, and thus purely as a research method it is of considerable value.
But from the perspective of a socially-concerned researcher, it is the shift in power photovoice allows, from the researcher to the researched, which is perhaps its greatest (and most famous) strength. This shift happens at two levels. On one level, participants become the researchers, with something to show and to say:
I: Aunty Greta, did you enjoy taking the photos?
P: It was a lovely method. With the camera in your hands, you feel like you want to show others what is going on in the communities. You see people inside houses; big families staying in one room; wet floors because of the rain. You see only children staying in the houses: no mother, father, all died, and you see no food. If you go there … you will cry when you see how children are staying: no food, no parents, one child looking after the other. One example I know of [is] a house in Inhlazuka. Oh I cried because of the situation. The parents are dead, the children are staying alone. I wish I had the camera then.
At another level, the participants also see the photographs as giving value to what was unvalued before; thus, they, as the photographers, have been given the power to give things value. The ‘empowering’ nature of photovoice is something that many researchers who have used the method have reported (Duffy, 2001; Koltz, Odegaard, Provost, Smith, & Kleist, 2010; Nelson & Christensen, 2009; Virgi & Mitchell, 2011).
Thus photovoice, by allowing the researched to create an image (in and of itself a powerful representation of lived experience), is obviously a potentially powerful political tool. And yet, having used the method, I am left with many unanswered questions about the nature of researcher power and the nature of research ethics.
Some of my concerns relate to any research project, anywhere, particularly the way in which the ethics of research is being subsumed into a bureaucratic process which depoliticises it. This is a point Kellehear (1993) makes:
When the findings have been analysed and are part of a final report, thesis or article, dissemination and implications of dissemination are important topics of ethical discussion. I am not referring here to submissions which may be taken to institutional committees—sometimes referred to as ‘ethics committees’. Where these exist, researchers may or even must avail themselves of their scrutiny and advice. And even though I think that the proliferation of such formal devices has its merits, the disadvantage is that this may abrogate the researcher from the responsibility of seeing ethics as part of the ongoing process of research … .
As ethics committees focus on research subjects (animals or humans), these committees often give the false impression that research ethics apply mainly to that particular interface … . The stated priorities of ethics committees can give the false impression that ethics is about ‘what we do to others’ rather than the wider moral and social responsibilities of simply being a researcher. (Kellehear, 1993, pp. 13–14)
Webb (2006) also raises concerns about ethics committees, warning that there is a danger that once the ‘obstacle’ of the ethics committee has been cleared, the researcher regards ethical issues as having been dealt with.
But this issue of how ethics in research is so often narrowed to informed consent coalesces with the specific ethical issues related to the photovoice method. The issue of ethics is something which has been emphasised by Caroline Wang herself. In an article written specifically about the ethics of the method, Wang and Redwood-Jones (2001) give very clear guidelines on what they consider to be minimum best practices regarding photovoice ethics. These include:
Getting consent from both the participant and the subject(s) of the photographs. They advocate a series of consent forms for participants and those they photograph, prior to taking the photographs, at the time of taking the photographs, and after to release prints for publication. They also advocate written material for the participants on the project, method, and who to contact.
Ensuring that ethical issues are discussed with participants at the time they are briefed and trained in taking photographs.
Providing letters about the method and the project where photographs will be taken, for example, in a school or workplace.
Providing prints to the participants to give to those they photographed.
However, in their reviews of studies using the method, neither Catalini & Minkler (2010) nor Hergenrather, et al. (2009) specifically reflected on the ethics of what the studies were doing or indeed how the studies themselves dealt with the issue. It is not clear whether these studies specifically included getting consent and providing written explanation about the method and the project or whether participants were given copies of the photographs they took. Catalini and Minkler (2010) reported that most of the projects they looked at included some kind of basic training, but they did not reflect on whether this specifically included training in issues of ethics (or simply in the technical skill of using a camera). In some cases they found little or no training at all.
In considering some of the studies using photovoice that have been published after the period covered by these two reviews, I am struck by the general lack of discussion of the ethics involved in the process. I specifically looked at studies including participants who would normally be considered ‘high risk’ in that they are generally considered as particularly vulnerable—children, people living with HIV and AIDS, people living with disabilities, refugees, abused women, and addicts. Some of these studies do not report at all on the process of informed consent or of discussing ethical issues with participants (Nelson & Christensen, 2009; Virgi & Mitchell, 2011), which does not mean this was not done, of course, but does suggest that the researchers did not experience any kind of ethical dilemma during the course of the research process. A few other studies did discuss ethics (Duffy, 2011; Green & Kloos, 2009; Newman, 2010; Rosen, Goodkind, & Smith, 2011), but confined themselves to discussions about the handling of the informed consent process and/or reporting that training had included a discussion of the ethics of taking photographs. The manuals on using photovoice that have appeared in recent years (John Humphrey Centre for Peace and Human Rights, 2010; Palibroda et al., 2009; Photovoice Hamilton, 2007) all include specific sections on ethics, but these are derived almost entirely from Wang and Redwood-Young's 2001 article specifically addressing the issue of ethics.
Thus, on the whole, since Wang and Redwood-Young published their article those who have used the method tend to argue that as long as they follow these guidelines, all will be well (see Rosen et al., 2011), which is a bit like the view that as long as one has complied with the requirements of an ethics committee, ethical issues have been dealt with. However, in recent years some articles have raised concerns about ethical issues related to the method, which have been summarised by Prins (2010). Prins agreed that some studies give few details about how they actually used the method, and she argued that they “present a somewhat romanticized view of participatory photography's transformative results” (p. 427), tending to underplay the risks and problems of the method. She was critical of the way the technology is presented as acultural; rather, photovoice shapes and is shaped by the particular sociocultural setting within which it is used. She raised in particular the following concerns:
the possibility that photographs taken are not in fact taken by the people who claim to have taken them;
the role the method plays in raising hostility, controversy, and in-group tensions;
the potential for the method to increase the visibility of individuals, thus subjecting them to the scrutiny/surveillance of others; and
the potential for violating local social norms.
Drawing on her own research using the method, whilst acknowledging the often-reported strengths of the method, Prins (2010) reported that the participants experienced great resistance and suspicion, as well as ridicule, from the community in which they were photographing. She emphasised the critical importance of a careful discussion about power and ethics, within the particular cultural context, at the start of the process.
In our project, we followed Wang and Redwood-Jones' (2001) guidelines. We gave the participants and the people who were photographed a letter (in isiZulu) explaining the project and the method and their right to anonymity and to withdraw at any time. The participants were asked to read this letter to people who were unable to read. Participants and the people who were photographed were asked to sign a consent form saying that they understood the contents of the letter and that they agreed/did not agree to participate in the project. The participants were instructed to always ask for permission before they photographed anyone. We also provided a set of prints for the participants to give to the people they had photographed. In addition, as is required by our institution, we submitted (and received) an application for ethical clearance by the University's ethics committee.
Thus, we fulfilled the basic requirements of ethical practice according to the method and our institution's research regulations. But some concerns remain—concerns which to some extent might extend to any research project, but which are primarily related to the nature of images per se. Many of the photographs taken in the project were of people, and, as we have seen, of people who are living extremely vulnerable lives—people in poverty, in crisis; children who have been abandoned, have suffered unimaginable loss; and children who have been traumatised by violence. It is their faces we look at when we look at the photographs. Writing about the marginalised and quoting from interviews can be powerful, but, as discussed above, not as powerful as an image. This is the very strength of the photovoice method, but also its most disturbing feature. Schratz and Walker (1995) have pointed out that it is the specificity of photographs—that they are of these people, at this time, in this place, and in this circumstance—that causes unease, and they said that “perhaps this explains why social research has been reluctant to use pictures, for this very identification of people poses a threat to conventional notions of objectivity and raises ethical issues that research finds difficult to address” (p. 90).
There are also issues, for me, connected to the technology of taking the photographs, and not just the photographs themselves. Cameras create a distance, however momentary this may be, between the viewer and what is being viewed. The technology objectifies. And photographs last much longer than the moment they capture. We know this—which is why even our photovoice pictures have a posed quality about them. Can they ever really be a true reflection of reality? And once the photographs exist do they not take on a new reality of their own? Once in the public space, can that reality not be manipulated, be used (and misused)? The abuse of photographs of ‘the poor’ by NGOs, for example, was powerfully made by members of a shack settlement in Pietermaritzburg, where I live:
We get very angry that the NGOs raise money on these issues that affect us. They come here time and again and take lots of photos of our shacks and our children, they write their reports to their donors—but they do not consult and work with the people; we never see them again. This is a kind of exploitation. (Butler, 2007, p. 46)
Temple and McVittie (2005) have argued that the fate of visual products resulting from research projects has been given very little attention; attention has instead focused on the production of such material. They say that questions such as who owns this material, how it can be used, and who is responsible for the eventual fate of such material need to be discussed.
One of the intentions of the photovoice method, as we have seen, is to attempt to overcome the power imbalance between the researcher and the researched. As discussed above, the method is very successful in doing this—but only to a certain extent. By giving certain categories of community members (the volunteers or support group members or CHWs) potentially powerful technology to ‘image’ the reality of life in the community, are we not simply replacing one level of power (ours) with another (theirs)? Are we not indeed creating a new level of power, which by extension then even further displaces from the locus of power those who are most powerless (further ‘disempowering’ them)? And if these ‘researched,’ these ‘research participants,’ then take a photograph of another (in this project, even less powerful) person, is this not simply replacing the power relationship of researcher-researched? How much more of an ‘Object’ can one become than being the subject of a photograph? As Prins (2010) said, “photography is a practice of power” (p. 435). And is getting consent really enough? To what extent can the subjects of the photographs taken in the project really be said to be in a position to exercise full choice in the matter, for example, when they are being photographed by the people who bring their food parcel each month, who provide them with support and help in caring for the sick and dying? And when the letterhead on the piece of paper they are given (and may not be able to read) is that of a university?
I leave these as questions—I don't know the answers.
Conclusion
But perhaps, after all, the ethical problems this experience has raised for me are not really the result of the method, the technology, or the pictures themselves. Perhaps they are problems I have created because I was involved in the wrong political project. Because I was targeting the wrong audience for the wrong reasons.
Perhaps photovoice should instead be used to allow others—community members, the poor, women, whomever—to take photographs, and then show them to themselves. Perhaps the photographs should then be used in a truly democratic space as codes, and be decoded, in the Freireian sense, to uncover the multiple layers of oppression the photographs truly represent. Perhaps this would allow true conscientisation, the building of understanding and of resistance. Perhaps that's the political project.
Footnotes
Acknowledgements:
I would like to acknowledge the work of Suhana Jacobs, a Masters student involved in this project. Suhana not only undertook much of the fieldwork with me, but our conversations in the car to and from the research area helped us work through our thinking on the technique. I would also like to thank Mark Butler for reading through the second-but-last draft of this article and providing really helpful comments and critique.