Different ways of seeing: Exploring audience reactions to images of probation supervision

The Photovoice methodology

While participatory visual methods have existed for many years in various guises and under different names, Photovoice first emerged as a distinct research method in the field of health promotion and community development research. Wang and Burris (1997: 369), who developed the method, described how Photovoice ‘uses the immediacy of the visual image to furnish evidence and to promote an effective, participatory means of sharing expertise and knowledge’. The aim is to enable people to visually document their lived experiences, enhance their critical awareness through discussion of the images and communicate with policymakers (Wang and Burris, 1997). Photovoice is grounded in several theoretical frameworks, including feminist thinking, Freire’s education for critical consciousness, participatory documentary photography and transformative pedagogy (see Delgado, 2015; Latz, 2017; Mitchell et al., 2017 for detailed discussions). Accordingly, Photovoice advocates are committed to the principles of empowerment, collaboration, inclusivity and social action. Photographs are the main source of evidence in a Photovoice project and allow co-researchers to construct their own understandings of their lives. The method has become increasingly popular over the past two decades and has been used to study topics like mental illness (Mizock et al., 2014), urban life (Delgado, 2015), illicit behaviour (O’Hara and Higgins, 2019) and motherhood (Murray and Nash, 2017).

Photovoice is designed to promote communication between everyone involved in a research project in order to produce meaningful insights into co-researchers' lived experiences. Unlike conventional research involving photography, co-researchers rather than researchers choose and collect images relating to the issues of inquiry. The resulting images are viewed as powerful forms of knowledge that facilitate communication and portray individual and community needs in a creative and visceral way (Baker and Wang, 2006). The method involves providing co-researchers with cameras to photograph their experiences of the phenomenon in question, then discussing and analysing those pictures either in focus groups or individually. The knowledge acquired through this process prioritises the perspectives of co-researchers; that is, those most intimately connected to the research topic. The producers of the photographs become the expert in creating and interpreting the images and are also empowered to decide which photographs become part of the research process and which are excluded (Fitzgibbon and Healy, 2019: 20).

Photovoice was developed with the understanding that photographs constitute valuable sources of information in their own right (Rose, 2012). Photographs can provide insights that are either unique or add to spoken and written expressions. The rich information elicited from photographs, as well as focus group discussions about them, provides in-depth understandings of co-researchers’ realities. Ideally, the process leads to co-researchers being understood and integrated into the research not in a passive way, but as co-creators of knowledge. Photovoice also allows for a conscientious creation of narratives through various interpretations of the ‘truth’ elicited by the particular framing of the images and the accompanying discussions between co-researchers and researchers and between co-researchers (Burles and Thomas, 2014). This process may limit the number of ‘truths’ that emerge from the data analysis but produces an array of definitions and understandings that create a composite reality of the phenomena being researched (Fitzgibbon and Stengel, 2018).

Social transformation is at the heart of Photovoice research and it is widely accepted that visual methods are capable of reaching stakeholders in ways that verbal methods cannot (Murray and Nash, 2017; Wang and Burris, 1997). Proponents argue that marginalised groups are often silenced by oppressive actors, structures and systems; thus, failure to facilitate meaningful engagement between co-researchers and audiences can constitute another form of silencing (Mitchell et al., 2017). While evidence of concrete change has been documented (see e.g. Delgado, 2015; Flanagan et al., 2016), most scholars caution against romanticising or overstating the impact of visual research on policy and practice (Mitchell et al., 2017). This is because relatively little is known about how audiences make sense of visual data, or whether their attitudes and behaviour are influenced by the viewing process. The shortage of audience studies also means that a critical step in the data analysis process is being neglected since audiences do not passively absorb content but actively interpret it in ways that are influenced by the image itself, their pre-existing knowledge and the social setting of the exhibition (Latz, 2017;Mitchell, 2011; Rose, 2012).

Latz and Mulvihill’s (2017) research, which analysed audience responses to a Photovoice exhibition on the experiences of community college students, is a notable exception. They proposed that the viewing process is shaped by an interplay between personal (e.g. previous experiences, beliefs and motivations); sociocultural (e.g. whether people attend an exhibition alone or within a group, prevailing ideas about the topic in question) and physical (e.g. sights, sounds and smells) contexts. These contexts are mediated by identities linked to audience members’ reasons for attendance (e.g. ‘professionals’ might attend because the topic is relevant to their work). Because the audience and co-researchers had similar backgrounds, Latz and Mulvihill (2017) found that audience members tended to view the photographs through the lens of their own student experiences, which elicited empathetic responses (personal). Audience preconceptions about community colleges were also positively transformed by the viewing process (sociocultural). Lastly, the physicality of the images, combined with the intimacy of the subject-matter, provoked powerful responses in audience members (physical).

Research design

Having reviewed the emerging literature on audiencing, this section describes the research design. The sample was recruited through a probation-funded, community-based rehabilitation project that works with adults under probation supervision in Dublin, Ireland. The eight men who decided to participate were invited to a day of photography and discussion in May 2016. After a briefing session where the study and their rights as co-researchers were explained, the men were provided with a digital camera and information on how to use it. They were given an hour to take 10 pictures that captured their experiences of supervision, producing 84 images in total. The photographs were printed and returned to the co-researchers who were asked to choose the most meaningful image and provide this picture with a title or caption. The men chose 31 images for further discussion and captioning. Next, co-researchers took part in a focus group where the meanings of the pictures were explored and discussed. These discussions were recorded and transcribed.

An artist then carefully selected and curated 36 images for the exhibition, using her own skills and knowledge to group the photographs in a meaningful and visually appealing way. She was not told which of the images had been identified as significant by co-researchers. The exhibition was held in November 2016 at the probation-funded programme where the fieldwork was conducted. An invited audience of over 30 stakeholders attended, including policymakers, probation officers, rehabilitation workers, academics and criminology students. To encourage them to reflect on and discuss the photographs, the audience was given print-outs of the images and asked to record their interpretations. Twenty-four audience members provided 656 responses to the images, averaging 18.2 responses per photograph (the average number of responses per photograph is provided as an indicative measure of audience engagement). Ethical considerations were to the fore at all times, and the study received ethical approval from the researchers’ universities and the Irish Probation Service.

Lessons learned

Curating the exhibition

Mitchell et al. (2017: 77) proposed the term ‘circulating the vernacular’ to describe how photographs created by ordinary people (the vernacular) must be viewed repeatedly to have impact. To paraphrase Mannay (2010), messages lie dormant until observation permits images to speak. With this in mind, reflections on our experience of curating a Photovoice exhibition are presented next.

Public (or semi-public) showcases of co-researchers’ photographs are a core feature of the Photovoice process and are believed to produce several beneficial outcomes. Robinson (2013) contends that such exhibitions have the potential to be creative and empowering for co-researchers. The photographs also serve as a medium for transmitting critical information about social issues, which might otherwise remain unheard, to those in power. Sharing research data in this way enables researchers to ‘give back’ to co-researchers, stakeholders and communities (Delgado, 2015). For visitors, exhibitions offer immediate and first-hand insights into the social world of marginalised groups which can raise fundamental questions about privilege, justice and opportunities (Moxley et al., 2012).

By providing a space for visual representation and re-representation, exhibitions can reach audiences and create impact beyond the traditional academic community and provide a compelling platform to communicate research outcomes (O’Neill, 2004). At its best, art’s infrastructure (whether a space for a temporary gallery at the probation project as in our case or an established public location like a museum, gallery or studio) and means of communication create a forum for community discussion and intense personal expression, encouraging audiences and co-researchers to exchange knowledge and develop reflective capacities. This is particularly important when dealing with unconventional and enigmatic problems like offender supervision whose solutions require active engagement from multiple stakeholders.

The preparation and curation of images for exhibition thus requires careful consideration because the ways that images are presented can influence how audiences ‘see’ and respond to them. To achieve maximum impact, images must be exhibited in a suitable environment and displayed in a format that is sensitive to the cultural setting, exhibition space and audience (Mitchell, 2011). Ultimately, our choice of a local setting proved highly effective, not only because its convenience ensured high attendance but because the audience’s familiarity with the neighbourhood promoted intense engagement with the images and elicited a sense of empathy with co-researchers’ experiences.

Photographs must be grouped and ordered in ways that are impactful, meaningful and aesthetically appealing (Delgado, 2015). The audience should also be guided purposively through the exhibition so that, individually and collectively, the images tell a coherent story. We were fortunate to have an artist as part of the research team who provided invaluable advice about the technical and aesthetic aspects of curating an exhibition. She selected 36 photographs for exhibition, which were displayed in groups of six (Figure 1). Carolyne Kardia explained her selection as follows,

The pictures were arranged to metaphorically ‘speak’ to one another, to enable not only the ostensible subject matter to be evident but also the subtle almost unconscious meanings to emerge which are often unintentional. They could be called ‘mistakes’ but somehow convey the complexity of the co-researcher's experience (e.g. the visual structure or composition of a building or wall in a picture may not only convey the recognition and meaning of the place but by its placement in the image also physically shut out the viewer) so the actual tactility of the image is both understood as a narrative but also on a kinaesthetic, or experiential, level too. By emphasising this aspect of the images both for the co-researcher and the viewer it allows the process of taking the photograph and subsequently reviewing the image to become a ‘transformative’ experience, allowing the image to re-present itself, depicting previously unrecognised feelings and experiences. The groups of pictures were selected to facilitate and explore these interconnections between subject matter and similar experiences of each of the co-researchers. To use the example of the building, a photo of a church may have been taken not only for its place as a retreat and sanctuary, but also as an institution with potentially repressive aspects. Each photograph can be interpreted on many different levels and the aim of the curation of the exhibition was to subvert over-simplification of their meanings, to allow their complexity to be exposed and to give the images as well as the co-researchers a voice which adequately reflects this complexity.

OPEN IN VIEWER

Figure 1.

Image from exhibition.

While this meant that co-researchers were not involved in the selection process, at least one image from each co-researcher was selected to ensure that all voices were represented. Ideally, a collaborative selection process is preferable but was not possible on this occasion due to time and resource constraints. After much debate, we also decided not to include the captions provided by co-researchers in the exhibition to allow the images to speak for themselves. This allowed viewers to interact with the photographs in an unmediated way and enabled us to study how audience members viewed and made sense of the images (on this point more generally, see Mitchell et al., 2017).

Additionally, compositional issues such as the size, shape and quality of the photographs can affect audience responses (Rose, 2012). On the artist’s advice, images were printed at the highest resolution on A3-size foamex-boards, 3 mm deep. The images were bled to the edge of the board to maximize their impact and give them greater presence. Foam-board mounts facilitate audience engagement because there are no barriers between the viewer and image (as is the case with picture-frames). From a practical viewpoint, foam-boards can be easily and quickly affixed to walls using Velcro buttons and removed with minimal damage. They are also light and durable and can be transported to other venues.

As O’Hara and Higgins (2019) observe, co-researchers often express anxiety over whether the images they produce are ‘good enough.’ Similar concerns may arise among researchers, with some choosing not to exhibit images that are blurred, dark or overexposed. We decided to include several such images in our exhibition because the focus group discussions suggested that so-called ‘mistakes’ can produce powerful and illuminating responses. Jack intended to capture a straightforward self-portrait in the bathroom mirror that showcased his transformation from drug addiction to health (Figure 2). On seeing the blurred photograph, he immediately concluded ‘that’s gone wrong’ and set it aside. However, the researchers and other co-researchers were struck by the artistic merits of the image, and the ensuing discussion prompted Jack to explore its personal significance. He explained, ‘[I wanted] a picture of meself, that looked so well like. From when I went in, there was probably only ten stone left of me like so basically just keeping the appearance of meself, looking after meself’. Jack ultimately decided that the photograph was not a ‘mistake’, concluding ‘to be honest, I didn’t think the pictures would have brought that much out of me. Made me think that much’. This example highlights how the amorphous character of photographic ‘mistakes’ can stimulate profound reflection, and unlock previously unspoken or unconscious feelings.

OPEN IN VIEWER

Figure 2.

‘Your health is your wealth’ by Jack.

The audience was also struck by this image and 19 members provided observations on its meaning. The vast majority identified it as ‘autobiographical’ [AM21], leading them to speculate on the significance of identity, past, present and future. As one observed, ‘The man in the mirror. Who am I? Blurring of me, my identity, who I am, what I do’ [AM18]. Significantly, many detected notes of uncertainty and confusion in the image, illustrated by the following statements: ‘All a blur, not seen for me’ [AM2] and ‘looking at life through Perspex’ [AM7]. Although the audience was not aware of his past, Jack had only recently been released from prison and may well have been experiencing doubts and trepidation about his future. This shows how visual artefacts – including so-called mistakes – can generate interesting, if unintended, synchronicities between representation and reality.

Advocates frequently emphasise the transformative potential of visual methods, arguing that they empower co-researchers by enhancing self-awareness and providing the human and social capital needed to engage in social action and advocacy work (Delgado, 2015; Wang and Burris, 1997). To achieve this, the images, and their underlying messages, must be effectively communicated to audiences who can effect change, such as policymakers. This section explores the extent to which intended messages reached the target audience; in this case, probation policymakers, probation officers, rehabilitation workers, academics and criminology students.

The images generated almost universally positive responses from the audience which largely tallied with the photographers’ intended messages. The probation project was perceived primarily as a source of help and support, a space that fostered creativity and self-expression and provided a refuge from the outside world. For instance, a photograph of the sign at the entrance to the probation project was perceived to signify ‘[a] safe place to go to’ [AM4] and ‘this probation project which helped me to make positive changes in my life’ [AM17]. Images of the artwork produced by project participants received particularly effusive responses, illustrated by responses to Jack’s photograph of the art-room which generated a total of 16 comments (Figure 3): ‘Creativity. Wonderful. My mess, my space, my expression’ [AM18] and ‘colour, chaos and creative, a happy space’ [AM14]. Just two audience members offered ambiguous views of probation supervision. For instance, one commented ‘probation work – payback for past offences’ [AM16] in response to an image of a rubbish bag lying alongside a gardening tool, a statement that has connotations of redemption and punishment.

OPEN IN VIEWER

Figure 3.

‘It's never too late to learn’ by Jack.

Audience interpretations were broadly consistent with the messages that the photographers intended to convey. In the focus groups, co-researchers expressed overwhelmingly positive attitudes towards supervision and explained that many of their images represented the help and support they received through the project. In particular, they cited the emotional and practical assistance provided by staff and peers as well as the opportunities for growth and education offered by the project. As Jack, who intended his photograph to represent the transformative potential of art, explained ‘it’s not basically about art. It’s just…about change, like, learning and change, like to change your life’. These findings suggest that visual methods can be used effectively by marginalised groups to communicate their experiences to stakeholders. Like the co-researchers who captured the images, the audience perceived probation supervision as strengths-based, welfare-oriented and rehabilitative in nature.

Enhancing insight and expanding knowledge

Visual methods encourage co-researchers to question preconceived ideas, reflect on personal experiences and engage in intense emotional dialogues, even around challenging, sensitive or abstract subjects. As Rose (2012: 305) explained, Photovoice inspires co-researchers to ‘talk about different things, in different ways’. Although Rose (2012) was discussing the impact of Photovoice on image-makers, it was clear from our study that the technique also prompted audiences to engage with the lifeworld of the photographers. As an audience member stated in relation to one group of photographs, ‘really sombre – gives a great insight into their viewpoint on the community and their place as outsiders within it’ [AM21]. Certain photographs elicited particularly strong responses; for instance, audience responses to Figure 5 such as ‘Brilliant. Crystal maze’ [AM11] and ‘I love this but I don’t know why’ [AM8].

The image-makers and audience could both be described as insiders since they occupied the same penal field, albeit in different positions. While insiders bring rich understandings to the interpretation of data, there is a risk that familiarity with the field may eclipse novel or unexpected findings by rendering the everyday invisible. Mannay (2010) posits that visual methods can address this issue by making the familiar strange and encouraging researchers and co-researchers to reflect on taken-for-granted assumptions (see also O’Hara and Higgins, 2019). Our findings suggest that the same may be true for audiences, generating new insights into the supervision experience.

Ultimately, the audience recognised that supervision experiences do not occur in a vacuum but must be considered in light of probationers’ relationships with society, including the wider criminal justice system (see also Farrall, 2016). Probationers’ journeys to, and through, supervision usually involve contact with many criminal justice agents, including police, lawyers, judges and prison officers as well as probation officers. These encounters can shape supervision experiences for better or worse and thus merit further exploration. Photographs of the local courthouse were widely perceived as a symbol of failure, condemnation and degradation (e.g. ‘Where we end up! Labelled’ [AM1]), but lawyers were seen as important sources of support for probationers. Audience members remarked ‘someone to talk for me?’ [AM2] and ‘free legal aid which was on my side’ [AM17] in response to Thomas’ image of a solicitor’s office which received a total of 19 responses (Figure 4).

OPEN IN VIEWER

Figure 4.

‘Help and Freedom’ by Thomas.

Similarly, co-researchers’ understandings of supervision transcended legal boundaries to encompass wider criminal justice contexts. Like the audience, co-researchers described court appearances and imprisonment in largely negative terms, framing these experiences as ‘wasted time’. Conversely, lawyers were seen as advocates and providers of help and support. Although Thomas was not able to attend the focus group discussion, Daniel who captured a similar image commented:

That’s me solicitor’s office and me solicitor has got me away with [laughs] too much. He’s absolutely brilliant. He usually gets me sent to the hospital or out of charges so he’s good. That’s why I took a picture of him. And he’s done a lot of things for me to change my life around, you know what I mean. He actually got me into treatment before when I was on drugs.

Shedding light on stakeholder concerns

While the exhibition helped co-researchers to communicate with stakeholders, the audience did not passively absorb the content of images but actively interpreted them through the lens of their sociocultural and experiential frames of reference, in ways that sometimes altered the intended meaning. Rose (2012: 30) defines ‘audiencing’ as the ‘process by which an image has its meaning renegotiated, or even rejected, by particular audiences watching in specific circumstances’. Meanings are not absolute but are always subjectively constructed through an interaction between observer and object in particular socio-cultural settings at specific times and places (Åkerlind, 2012). A range of factors can influence perception, including the image itself, the social setting in which the image is viewed and the observer’s life experiences, social identities and cultural assumptions (Latz and Mulvihill, 2017). As Spencer (2011: 19) notes, ‘we “read” the images in front of our eyes through the pictures we have in our heads’. Thus, audiencing studies can be used to shed light on stakeholders’ core concerns, values and beliefs.

Analysis showed that audiences actively attributed meaning to images in ways that were shaped by their socio-cultural background. For instance, many identified themes in the photographs that called to mind the idea of redemption through suffering. The redemption script is a potent narrative within Western culture which embodies the idea that negative life experiences can ultimately be transformed into positive outcomes (McAdams, 2006). This narrative has become deeply ensconced in rehabilitation theory and practice in recent years, thanks to the work of scholars like Maruna (2001). The significance of this script is best illustrated by responses to Adam’s evocative image of a broken pane of glass lying among weeds which elicited 19 observations from the audience (Figure 5); for instance, ‘shattered lives but green shoots and hope around the corner’ [AM14], ‘a shattered life can also be beautiful’ [AM3] and ‘hope through difficulty’ [AM9]. Interestingly, this image was one of the audience’s (and researchers’) favourites but was not singled out for comment by the photographer.

OPEN IN VIEWER

Figure 5.

‘Untitled’ by Adam.

Despite the overall convergence between probationer and audience interpretations, some divergence was found with respect to individual photographs that may be explained by cultural assumptions. For instance, Conor captured a close up of a goldfish bowl that is located in the waiting area of the probation project building (Figure 6). For him, the fishbowl signified the time and space he could now devote to relaxation and positive thinking due to his participation in the project and move away from crime. He vividly recalled his first visit to the project, explaining ‘basically when you come here for the first time, you’re asked to sit downstairs while someone comes down to you. You can sit there, sit in the chairs and look at the fish, it’s relaxing’.

OPEN IN VIEWER

Figure 6.

‘Relaxing; Time looking at the fish’ by Conor.

The audience also responded strongly to the image, providing 19 responses in all. Contrary to Conor’s intention, the majority identified negative themes in the photograph. For one, the image elicited thoughts of confinement and surveillance: ‘Colours, vivid. But contained, restricted. No freedom. Only the fishbowl space. The fishbowl of life where I can be watched and monitored’ [AM18]. Likewise, the comments – ‘Glass cage!’ [AM10] and ‘little fish, small pond. Needs more space’ [AM14] – evoke feelings of restriction and confinement. Others spoke of alienation and separation from the surrounding world, evident in the comment, ‘like a goldfish – on the inside looking out’ [AM3]. Finally, a sense of futility pervaded some observations, e.g. ‘going round in circles’ [AM16, AM23]. One audience member [AM21] even described the photograph as among the ‘saddest’ in the series, seeing in it a representation of the probationer’s circumstances. Just four identified positive emotional themes, reflected in the comment ‘I would love to be like a goldfish in beautiful surroundings’ [AM19].

Though rare, these examples prove interesting in terms of what they reveal about the processes of perception and knowledge production, highlighting how images are interpreted through the lens of shared cultural understandings – a process that sometimes enlightens but at other times misleads audiences as to the intended message. The phrase ‘living in a goldfish bowl’ has a particular cultural resonance that conjures thoughts about a lack of privacy and constant surveillance. However, the image had a different personal meaning for the photographer, reminding him of a quiet and tranquil space. Conversely, photographers’ messages were more readily communicated when the audience and photographer attributed the same meaning to familiar cultural tropes, as discussed above in relation to the theme of ‘redemption’. These findings show that audience members are active participants in the sense-making process and are thus worthy of study in their own right. For instance, audience studies could be used to provide insight into stakeholders’ core concerns, and to produce multi-layered understandings of research topics.

Managing inter-subjectivities

The last issue to be discussed concerns the challenges of incorporating audience studies into Photovoice research. As noted above, several groups, each with a unique perspective, were involved in the process, namely co-researchers, researchers, an artist and audience members. The knowledge produced through this process was fluid, dynamic and open-ended; the product of multiple layers of interpretation, created over time as the meaning attached to the images was reflected and refracted through various viewpoints. Each group engaged with the project for different reasons and their disparate motivations may have shaped their interpretations: the co-researchers wanted to share their experiences of supervision, the researchers to contribute to knowledge about supervision, the artist to set up an attractive and engaging exhibition and the audience to learn more about probationers’ perceptions of supervision. As shown above, the images generated a variety of responses that were sometimes at odds with each other and at other times showed remarkable consistency. Yet, the various perspectives always reverberated through one another; for instance, it is likely that the audiences’ interpretations were influenced at least to some extent by the particular groupings and order created by the artist. Likewise, it was clear that observers influenced one another’s responses, creating an ongoing conversation about the images as they were viewed (for example, one provided a humorous caption ‘tyred!!’ [AM6] for a photograph of a car with a punctured tyre while another offered a similar but less entertaining ‘tired’ [AM7]).

Spencer (2011) contends that the challenges involved in navigating multiple meanings, while complex, should not deter researchers from using visual methods. In our experience, the use of visual methods raises important questions about how to analyse and present data in ways that reflect different ways of seeing. Ultimately, we decided to foreground the photographers’ ways of seeing and added our interpretations only to contextualise the findings. In addition, a separate analysis was conducted on audience responses to explore how they interpreted the images (and by extension, how they perceived supervision) and whether their interpretations reflected the messages that co-researchers intended to convey. To demonstrate the analytic procedure, the process is now described in relation to one photograph, namely Luke’s photograph of the front page of a tabloid newspaper which reported on the murder of a member of a well-known criminal family who was killed as part of an ongoing feud (Figure 7).

OPEN IN VIEWER

Figure 7.

‘Live the life you love’ by Luke.

As discussed above, photographs were coded and analysed using Mizock et al.’s (2014) visual matrix. In this case, the concrete dimension was coded as a literal description of the image: ‘Front page of tabloid newspaper with story of gangland murder’. Next, the symbolic dimension was derived from the caption provided by the co-researcher and a thematic analysis of his commentary about the image. Luke, who captioned the photograph ‘live the life you love’, described gangland crime as a backdrop to everyday life within his community. When asked to explore his feelings about it, he said ‘[I] don’t feel anything, to be honest. It’s bad but it’s what happens. […] They chose that life so…’ When another co-researcher commented that many of the murder victims were ‘innocent’ and killed only because of familial links to criminal figures, Luke responded ‘that’s what they say’. His apparent unwillingness to countenance any sympathy for the murder victim clearly conveys his rejection of criminality and the criminal lifestyle. Afterwards, the researchers contributed the abstract dimension; for example, they identified shades of the desistance literature in Luke’s rejection of criminality (see e.g. Maruna, 2001).

The written observations provided by the audience were then coded and overarching themes identified. This image provoked a strong and varied reaction from the audience, generating 21 responses in total. The majority thought the photograph represented a potential turning point in the photographer’s life, believing the news story prompted him to reflect on his life choices. This is illustrated in the following comments: ‘The end? Inevitable? Not all bullet proof?’ [AM1] and ‘I want to get away from this life’ [AM19]. Interestingly, just three observations reflected Luke’s intended message (e.g. the statement ‘another day, another story’ [AM2] suggests a resignation to the frequency of these crimes).

Although photographs have not always been treated as data in Photovoice research, their value is increasingly recognised (Mizock et al., 2014). This image and responses to it illustrate the variety of interpretations that emerge when the same image is examined from different angles, and from different insider perspectives. This shows that photographs, and their interpretations, are always socially constructed; thus, the viewing process may sometimes lead to incongruous opinions about the meaning of a particular image (Creighton et al., 2018). While this complicates the analytic process, researchers should not be deterred from using visual methods because each way of seeing enriches the analytic process and provides different insights into the research topic. As O’Hara and Higgins (2019) observe, a combination of visual and textual data allows for triangulation between sources thereby increasing the accuracy of interpretations.

Discussion

This article explored audience responses to an exhibition of photographs taken by people under probation supervision. In doing so, it aimed to contribute to the emerging literature on audiencing, an aspect of the Photovoice methodology that is currently under-researched. The findings suggested that visual methods can be used effectively by marginalised groups to communicate their experiences to stakeholders and help to forge empathic connections between photographer and audience. Our experience indicated that effective practitioner–client relationships strengthen the sympathetic resonance between image-maker and observer even further. To illustrate, a probation worker correctly guessed which co-researchers produced which photographs when we showed her the entire set of photographs. Nonetheless, positive outcomes are not inevitable and depend on the willingness of the target audience to engage with the exhibition and accept constructive criticism (Johnston, 2016). While our target audience engaged enthusiastically, other audiences could be less receptive to image-makers’ messages. In such cases, Alexandra’s (2015, cited in Mitchell et al., 2017) concept of ‘political listening’ may offer a solution. Recognising that consensus is not always possible, she argued that Photovoice exhibitions should be positioned as safe spaces for the exchange of views.

On a less positive note, none of the image-makers attended the exhibition either because they had moved on from the project or preferred to view the exhibition by themselves. It is likely that the five-month delay between the photography session and exhibition also played a part in their decision not to attend. Additionally, their reticence may reflect the feelings of discomfort that can be elicited by the exhibition process. As Mitchell et al. (2017) discovered, co-researchers often experience anxiety about communicating their experiences to powerful others. To ensure that co-researchers’ voices are represented in the knowledge exchange process, we recommend that the gap between the photography session and exhibition is kept to a minimum and that adequate preparatory time is spent building engagement and confidence. When participants contribute to exhibitions, the effect can be positive for all parties.

On some readings, the decision to ‘show’ rather than ‘tell’ might be interpreted as a silencing of our co-researchers, compounded by the use of an artist to select the exhibition images. While relatively rare, some divergence between audience and co-researcher understandings was observed, resulting in misinterpretation of the intended messages. Captions are generally used in Photovoice exhibitions to foreground co-researchers’ interpretations and limit the space for misinterpretation on the part of the audience (Latz and Mulvihill, 2017). On reflection, we accept that our decision not to use captions risked exacerbating the already marginalised position of our co-researchers. At the same time, audience and co-researcher interpretations were broadly consistent, suggesting viewers received the intended messages, which somewhat mitigates this risk. Additionally, we subsequently circulated an article to key stakeholders (including senior policymakers) which explored co-researchers’ interpretations of the images. Admittedly, this was an indirect way to share co-researchers’ experiences but stakeholder feedback suggested a genuine engagement with the material. Nevertheless, our experiences highlight the difficulties of facilitating true participatory research, particularly when time and resources are limited (though it must be remembered that Photovoice can accommodate different participation levels, from consultation to control).

Given the opportunity for unmediated engagement with the images, the audience responded powerfully – and positively – to the images and the people who created them. Their responses suggested a sympathetic understanding of the co-researchers, portraying them as people seeking peace and redemption, hopeful but constrained by circumstances that were sometimes beyond their control. Likewise, probation supervision was framed in positive terms as a source of help and support that provided probationers with an opportunity to develop latent talents. The exhibition also encouraged the audience to reflect deeply on co-researchers’ lived experiences, helping to make the everyday visible and extending their vision beyond the legal borders of supervision. The audience did not passively observe the images but actively interpreted them in ways that were shaped by their sociocultural and experiential frames of reference, including their occupational background and insider status (see also Latz and Mulvihill, 2017). As Spencer (2011: 16) notes, ‘visual representation is always political’ because audiences tend to interweave ideological and artistic interpretation. While we did not set out to assess change in attitudes towards probationers or probation supervision, studies suggest that Photovoice exhibitions can alter practitioner attitudes in positive ways (see e.g. Flanagan et al., 2016).

The images appeared to evoke metaphorical forms of reasoning among audience members. Like Spencer (2011: 16), we discovered that ‘poetic use of imagery creates feelings and texture; the imagery speaks directly to the individual’s inner self evoking memories, reflections and feelings’. This is a significant finding because metaphors are not just linguistic embellishments but represent an essential cognitive structure that helps people to order and make sense of their experiences (Johnson, 1987). In fact, even simple metaphors can affect the way people think about crime and punishment (Thibodeau and Boroditsky, 2011). The positive metaphors suffusing the photographs in our study thus offer an antidote to the mostly negative metaphors that characterise political and public discourse on crime and punishment.

In conclusion, it is widely agreed that visual researchers should analyse the internal narrative of an image as well as its content, that is, the explanation offered by the image-maker. However, our findings show that a full appreciation of visual data also requires attention to its external narrative, that is, the meanings attributed to imagery by audiences.