Collaborative audio-visual rhetoric: A self-reflexive review of collaboration in anthropological film projects

Through a self-reflexive review of collaboration in two anthropological film projects conducted during the period from 2013 to 2018, we address the dilemmas that are at stake when producing and distributing audio-visual images of vulnerable subjects (for example, refugees), and the consumption of so-called “subject-generated” imagery. The poetical and political strategies of audio-visual forms of anthropological (re)presentation as well as the existing shortcomings related to those topics, such as the ‘refugee crisis’, which is caught in the current context of our media-saturated society is explored. We suggest the notion of “shared anthropology” as a framework for exploring the critical potential of collaborative audio-visual rhetoric. From this perspective, we discuss the films Unless the water is safer than the land, in which performers re-enact the narratives of refugees; and Passager, a collaborative film project between an audio-visual anthropologist and a young Afghan refugee who left his country in search of safer living conditions. We explore how these films can be conceptualised as distinctive forms of “collaborative audio-visual rhetoric”.


Introduction
In Participatory critical rhetoric: Theoretical and methodological foundations for studying rhetoric in situ, Middleton et al. (2015) convincingly argue that rhetorical scholars are increasingly turning to fieldwork and related ethnographic, performative and qualitative research methods to gain access to and analyse a broad range of everyday rhetorical practices. They argue that participatory critical rhetoric -as a distinct framework for studying 'situated' rhetoric -is a theoretically and methodologically sound approach within critical rhetorical studies. They also emphasise that participatory critical rhetoric can potentially enhance our knowledge of the inevitable presence of rhetoric in everyday life, by adding to our 'archive of rhetorical practices and texts' and by focusing on how rhetorical critics should engage in dialogue, collaboration and interaction with rhetors and (implied) audiences at the moment of rhetorical invention itself. As such, participatory critical rhetoric opens possibilities for engaging with marginalised and vulnerable voices that often are not being heard (Middleton et al. 2015). Middleton et al. (2015) refer to participatory critical rhetoric, as a framework that 'privileges a participatory epistemology that opens the critic to the intersecting corporeal, affective, embodied, emplaced and often fleeting dimensions of rhetoric that unfold in the everyday spaces of rhetoric' (Middleton et al. 2015:160; also in Degerickx et al. 2020). We argue that these are the dimensions that are crucial in addressing the dilemmas that are at stake when producing and distributing audiovisual rhetoric of vulnerable subjects, and more specifically in the consumption of so-called "subject-generated" imagery. In this article, we explore the notion of "shared anthropology"' as a framework for understanding and assessing the critical potential of collaborative (rather than participatory) audio-visual rhetoric.
If we accept the importance of the senses and lived-experiences in the process of our meaning-making, then the study of (audio-visual) rhetoric should also be more inclusive in the way that it integrates the contingency and diverse forms of interpretation and sense-making that are contextualised through specific narratives. In that vein, we argue, it is necessary to integrate more collaborative work of multiple authorship with those who are the "subjects" of (audio-visual) anthropological research. Such critical work embodies the diversity of aesthetics and ways of meaning-making, not only from a perspective of "a lived experience" of research participants, but in relation to the lived experiences that they create by and through their personal narratives.
The aim of this self-reflexive review is therefore to address the complexities of producing and (re)presenting images and narratives of refugees from the perspective of (audiovisual) participatory rhetoric. We discuss the films Unless the water is safer than the page 03 of 17 Number 34, 2020 ISSN 2617-3255 land, in which performers re-enact the narratives of refugees, and (more elaborately) Passager, a collaborative film project between a visual anthropologist and a young Afghan refugee who left his country in search of safer living conditions. We explore how these films can be conceptualised as what we would coin 'collaborative audiovisual rhetoric'.

The rhetorical turn in anthropology
Anthropology has already for a long time been confronted with both epistemological and methodological self-reflection and self-criticism. Starting from the recognition that it is impossible to accurately and completely (re)present the lived experiences of their research "subjects", ethnographers increasingly aim at comparatively relating their own cultural frame to that of research "participants" (or collaborators) and emphasise the interactive relationship in the research process (Pinxten 1997;Jackson 2013).
The current understanding of ethnographic research as an "interactive encounter", and anthropology as a comparative discipline is of crucial importance as 'the informant and the ethnographer are producing a common construct together' (Pinxten 1997:31; also in . This creates both a very context-specific, as well as a fragile exchange between the researcher and the research participants. From this perspective, shared anthropology aims at creating and representing anthropological knowledge through not only collaborative ethnographic "research", but also through collaborative ethnographic "writing", either textually or through other media such as film in the case of audio-visual anthropology. As such, the notion of shared anthropology leads to an increasingly critical approach towards the production, circulation and consumption of ethnographic narratives and an exploration of alternative forms of ethnographic (audio-visual) representations.
Anthropologists have increasingly been influenced by the rhetorical turn in the social sciences, starting from the realisation that writing, crafting and developing persuasive ethnographic narratives are fundamental aspects of their discipline (Rutten, Van Dienderen & Soetaert 2013b). This self-reflexivity can of course be related to the "rhetoric of science" in general, involving the (albeit reluctant) recognition that all scientific writing and reporting of research is rhetorical up to some point (Gross 1994;McCloskey 1990 In their influential project on rhetoric and culture, Ivo Strecker and Stephen Tyler (2009) argue that the discipline of anthropology still needs to fully come to terms with the 'crisis of representation' that was elaboratively explored in the seminal volume Writing culture: The poetics and politics of ethnography by James Clifford and George Marcus (1986). Furthermore, they argue that the discipline of anthropology needs to recognise and even embrace the pervasiveness of persuasiveness (Herrick 2005), in other words, the inevitability of rhetoric in 'presenting' fieldwork (as a form of rhetorical invention) and 'persuading' an (implied) audience (Rutten, Van Dienderen & Soetaert 2013b).
We want to clarify that we obviously do not refer to the more negative understanding of rhetoric as "mere" persuasion (implying that ethnographic accounts are always "mere" rhetoric), but we emphasise the inevitable and pervasive "rhetoricity" of how meaning about culture and lived experience is constructed and negotiated.
As such, we need to be self-critical when producing and presenting audio-visual images and narratives of research subjects and acknowledge that rhetoric is the means with which we "describe" culture, and at the same time, also the means with which we "interpret" cultural practices and lived experiences (Strecker & Tyler 2009;Rutten et al. 2013b). In this vein, understanding the "crisis" of representation needs to be reframed as understanding the inevitable "rhetoric" of representation, acknowledging that the description and delegation of cultural knowledge and lived experience is always rhetorical and as such is not only a "reflection" of a particular reality, but at the same time also a "selection" and a "deflection" (Burke 1966).

Audio-visual and sensory modes of knowing
In both the fields of rhetoric and anthropology, textual language has largely been the dominant mode of study and communication. The visual, in general, has for a long time often been regarded as more emotional and unclear, and as such, less suitable for academic and scholarly research. There has always been a tense relationship between anthropology and ethnography on the one hand and media, such as photography and film, on the other. Within anthropology, visual representations were regarded as suspect and potentially questioning the credibility of the scholarly work as '[n]othing was more threatening to assertions of ethnographic authority than an anthropologist with a camera' (Grimshaw & Ravetz 2005:5).
In a critical review of 20 years of Visual Anthropology, Jay Ruby (2005) gives an overview of the discussions in the field of visual anthropology that still are relevant today. He distinguishes between three approaches to the field: 'visual anthropology as ethnographic film, a cultural study of pictorial media and an inclusive anthropology of page 05 of 17 Number 34, 2020 ISSN 2617-3255 visual communication ' (2005:159). Ruby (2005) argues that an anthropology of visual communication problematises how (traditional) ethnographic films 'communicate', implying that '[a]nthropologists desiring to communicate their anthropological ideas via film have to confront the problem of how film communicates for any purpose before they can be assured that they can accomplish their desired goal' (Ruby 2005:165).
He approaches the visible and pictorial worlds as social processes, 'in which objects and acts are produced with the intention of communicating something to someone' (Ruby 2005:165). Furthermore, he constructs his view of visual anthropology based on 'the belief that culture is manifested through visible symbols' (Ruby 2005:165).
From this perspective, the importance of examining all manifestations of "the visual" as forms of culturally based communication and interaction is emphasised (Ruby 2005:165).
However, moving away from a more "visual-centred" approach to anthropology, we concur with those scholars who argue that visual rhetoric and visual anthropology should be acknowledging other sensory modes of knowing; advocating that "the visual" cannot be demarcated from sensory and bodily faculties. David MacDougall (2006:268) even considers the 'quest for a "pictorial representation of anthropological knowledge"' a rather 'conservative' approach that starts from a positivistic point of view and he emphasises the importance of also taking the non-visible and the larger domain of the senses into account. As such, we align our work with the "sensorial turn" in anthropology and ethnographic research (Ingold 2000;Pink 2009;MacDougall 1997MacDougall , 2006, which refers to an increasing problematisation of ignoring "experience" and the senses as fundamental aspects of anthropological knowledge and exploring possible ways to produce and represent such experiential and sensory ethnographic findings. Arnd Schneider and Chris Wright (2006)  However, there seems to be a lack of research in which specifically subject-generated knowledge in such collaborations is thoroughly discussed, and more specifically, a critical exploration of the notion of collaboration itself, the aesthetic strategies and page 06 of 17 Number 34, 2020 ISSN 2617-3255 the ethical dilemmas related to this. As such, we also need to expand the rhetorical perspective in anthropology to debates regarding the incorporation of sensory and experiential modes of knowing in ethnographic audio-visual narratives and think through the implications for both the politics, as well as the poetics of anthropology and the aesthetics of ethnographic film.
The "refugee crisis" as a visual rhetorical trope It is an understatement to claim that "the refugee" and "the refugee crisis" have become influential tropes in our contemporary visual economy, with the Mediterranean Sea being an 'iconic and lucrative topos', which implies that 'on a daily basis, European citizens consume night shots of wet people helplessly stretching out their hands to their white saviors' (Bellinck & Van Dienderen 2019:61).
This visual economy, according to Chouliaraki and Stolic (2017), has been critically addressed but at the same time also perpetuated by (independent) media, artists, activists and visual anthropologists, whose work often leads to 'impressive but of interest' and are presented as 'stereotyped distortions', without offering room for critical reflection on how they are represented (Bellinck & Van Dienderin 2019:61). The main issue raised is that viewers miss crucial information about "how" and "by whom" the film is produced. As such, Bellinck and Van Dienderin (2019:61) argue that the film falls prey to an image-positivistic logic (which the authors refer to as taxidermic) that 'represents clichés, stereotypes, and established values as part of a cultural hegemony' that maintains an uneven power balance, rather than questioning it.
Bellinck and Van Dienderin (2019:69) explicitly refer to this as a blind spot and based on their practice-led research they argue for a speculative documentary that explicitly engages with the 'multiple and mutable' realities that visual anthropologists engage with and create, and how these are inevitably marked by gaps. We fully concur that the aim should not be to ignore these gaps, but to "come to terms" with them. Hereafter, we discuss how we tried to come to terms with such gaps from the perspective of collaborative audio-visual rhetoric. We start from a self-reflexive approach, meaning we make explicit the narrative strategies that were chosen and, as such, reveal the interrelationship between the process and the political conditions of the fieldwork and the product of the anthropological visual representation. We aim to clarify how different conditions led each project to adjust the degree of collaboration to its research subjects and that "collaboration" should not be an aim as such, but is highly dependent on the scope, focus and specific context in which the film was produced. We explore what it implies to "share the anthropology" rather than "merely" sharing a camera in a visual ethnographic study.

Unless the water is safer than the land (2015)
In 2013, Arjang Omrani was asked to make an ethnographic film about stories of fleeing, in collaboration with minor-aged unaccompanied refugees in Cologne, Germany.
The producer's idea was to create a workshop, teaching the refugees how to use a camera and subsequently persuading them to make films about their own stories.
However, this raised a number of critical questions, because making a film, whether documentary or fiction, is not only about learning how to handle a camera, but also, more importantly, about having a vision on the medium itself and understanding the possibilities and potentialities that exist within the scope of applying a particular technology. Furthermore, there are important differences between the various roles in such a collaborative film project where one changes from subject to author and co-author. Indeed, holding a camera does not necessarily or straightforwardly imply that one is no longer the subject of an audio-visual rhetorical account. Clearly, the danger was that this project could easily end up being a naive representation incapable page 08 of 17 Number 34, 2020 ISSN 2617-3255 of mediating the experiential aspects of such profound and tragic stories. As in many visual projects made about or with participation or collaboration of vulnerable subjects, the project would at best generate emphatic feedback from an (already interested) audience, with the danger of intensifying existing stereotypes about refugees; or at worst it would merely offer a platform for voyeurism.
The project was made a few months before the first arrivals of Syrian refugees at the European borders, which had brought the topic to the top of the news headlines. At that stage, however, people were less informed and aware of the dire conditions and the difficulties and suffering many refugees had to endure. Owing to the pervasive media attention that followed the first "wave" of refugees, 1 people are now more Having these concerns in mind and based on the selection of the personal narratives of the participants, a series of public performances was filmed that were acted by performers, and not the refugees themselves. As this film was to be screened to a German audience, there was collaboration with German members of a youth theatre community to perform refugee roles. As such, the refugees themselves did not reproduce the stereotypical (bodily) images that people usually see through mainstream media. Furthermore, since the film had the intention of being reflexively critical (especially for the European audience) instead of being merely narrative, the concern was not to cause any obstacles for the asylum cases of the refugee participants as they were still being processed by the authorities. In communicating about lived experiences, we can question whether we would need to see or get to know the "real" people in the film or any other form of presentation in order to consider their story as documentary page 09 of 17 Number 34, 2020 ISSN 2617-3255 realistic or truthful. Why would it be important to know which country a person was born in or to which group of people they belong to when a story can be anyone's story? By applying sensorial and haptic strategies of storytelling, the aim was thus to prevent the film narrative from becoming a medium for objectifying someone's life, and furthermore shifting the argument into different dimensions with which the audience might need to self-identify themselves with, or rather align it with the bigger context of their own social and global perspectives.
The actors represent the stories of any human being in desperate need of help and care. By way of performance, one can argue, the particularity of the refugee stories has become universally recognisable. They are communicating the stories that could be representative for any human being who is in need of help and care. However, this time they look like the audience's own children, sisters, brothers or classmatestelling stories not usually heard or seen in relation to European bodies -stuck under a container's floor or asking for help and food from people in the street as an immigrant. The aim of this form of storytelling was to invite the audience to sensory commitment and embodied experiences that 'are not engaged with a naive sense of experience as "direct" … but rather mediated and qualified by our engagements with our bodies and things' (Sobchack 2010:4 Europe, while drowning, striving, struggling and lamenting. These sympathetic and empathetic responses during the earlier stages of the crisis, later on, faded away under the shadow of suspicion or hostility, especially after the November 13th [2015] terror attacks in Paris' (Chouliaraki & Stolic 2017:2). Chouliaraki and Stolic (2017:2-3) argue that 'this combination of empathy with suspicion is an established pattern in the representation of human mobility ... [that] identifies the refugee as a fundamentally ambivalent media figure'.
The extensive media coverage made the so-called "refugee crisis" a part of daily discourse throughout Europe. This media discourse -through its use of news headlines or iconic photographs and videos -needs to be approached as a form of knowledge that produces and reproduces collective audio-visual rhetoric and metanarratives about refugees and "their" as well as "our" experiences. From this perspective, Jackson (2013)  How can one develop a subjective reading of this lived reality that offers a counterbalance both to the fact-claiming media and the already established "truth" about this situation, without re-producing the existing stereotypes? Arjang Omrani explored these questions in the film Passager.
Asef is from Afghanistan. Omrani met him in June 2016 in the camp at Athens' Piraeus port. After meeting a few times, they lost track until they met again in September 2016. They discussed the idea of making a film about the refugee crisis and whether Asef would be interested to take part in that project. Asef accepted the offer and they exchanged numbers to remain in contact. Omrani tried to contact Asef a few times, but his phone was switched off. After a few days he called back on Viber and was informed that Asef was detained on Kos Island after he had been caught while trying to get on a plane with fake documents in order to fly to Germany. This was the starting page 11 of 17 Number 34, 2020 ISSN 2617-3255 point of the collaboration. Omrani tried to instruct Asef on phone to film where he was and to explain his conditions. Asef sent the video back via Viber, which is the opening shot of the film. Eventually, he was released and returned to Athens, where he regularly met with Omrani to discuss how the project will progress.
During the collaborative film-making process, Asef was constantly trying to find a way to leave Greece towards the west of Europe. This proved to be an unsuccessful mission during the entire six months from when he had reached the Greek shore by boat until the time the project started in September 2016. Despite these difficulties, he was determined to find his way out and therefore he was always on hold to be called by smugglers to tell him where to meet in order to cross the border. This was a major obstacle for following him during the process of the project, which made Omrani wonder whether he should record the film entirely by himself. Certainly, this was a major risk, but as Asef showed interest in learning, they decided to start teaching sessions on cinema, montage and storytelling, in such a way that Asef could start filming and expressing his diaries anytime he felt like. They tried to create a simplified version of what Omrani had been working on with his students over the past few years in his audio-visual anthropology seminars, but at the same time recognised the importance of being focused and practical, based on the project's needs. These needs included aspects such as the technology that would be used, forms of narrative, style of montage and also creating a clear intention of what the film could be about and how the audience should be addressed. What was it that Asef wanted to share with any potential viewer?
This latter question was mainly produced through an inter-subjective exchange.
Therefore, there was a continuous conversation about different aspects related to fleeing in general, as well as Asef's personal story. The aim was also to make Asef reflect and draw links between the mediatised refugee "crisis" in general, and his own personal situation, as well as to be very aware of how these inevitably influence one another. He was asked to reflect about these issues through his diaries and personal experiences. Very soon it was concluded that the film should solely represent Asef's portrait in an auto-ethnographic form as 'a vehicle and strategy for challenging imposed forms of identity' (Russell 1999:276). Asef hoped that the film would somehow be helpful in the path he was taking, whether in providing opportunities for further studies, or jobs, or something else. Therefore, Omrani decided not to be physically present in the film. He felt that any sort of presence (as a protagonist, a mentor, a researcher or filmmaker) in the film would distract from Asef's story. The aim was to avoid that the film's narrative loses its pace by integrating a self-reflexive style to address the collaborative methodology. This had already been aptly phrased by Van Lancker if this is not revealed in the film itself, is nonetheless the basis of the film's performance'.
Furthermore, this collaborative method, as Van Dienderen (2008:176) suggests, 'implies that social or cultural-scientific research is based on the conviction that reality is shaped by the concept of continual negotiation'. Certainly, the intention of Omrani has not been to neglect his presence in the collaboration, but to prevent this from distracting from the main target of the project and to dim the poetical and empathic aspects of the narrative.
Asef's mobile phone camera was used to record the film. There was a concern about the quality of the images and the sound, which did not turn out the be an issue. In addition, this choice had significant advantages since there was no need to overload Asef with technical information of handling a camera. He used the technology that he was used to and had already been using for a while; therefore, he had better control and he could easier associate with what he was about to record, compared to handling a camera that he had no experience with. Moreover, there was literally no time nor an opportunity to teach him and let him explore and get accustomed to a "professional" camera. Additionally, carrying a camera meant an extra burden that needed to be kept safe, that occupied space and attracted unnecessary attention. Finally, using the mobile phone's camera represents the position of the medium that is globally applied by people, and in this case the refugees, in registering and sharing key moments in their lives, which was very much affiliated to the form and style (or the poetics) of the work.
The diary form of the narrative implied that the montage had to be chronological, and Asef's role as storyteller was therefore significant. Omrani was not present during any filming, which gave Asef more freedom, but also the responsibility to choose 'the fragments of [his] reality [he] deem[s] significant to document' (Elder 1995:94

Concluding reflections
Any form of collaborative work in the context of anthropology should have a tendency towards shared anthropology. Within the context of ethnographic film, this implies a form of collaborative audio-visual rhetoric. This not only "empowers" vulnerable subjects (as in "giving voice to") but emphasises the importance of shared meaningmaking through sensory experiences. The kind of knowledge to be produced by or with research participants should aim to enhance their knowledge in a broader sense about the project that is being conducted as well as about the anthropologists involved, and how they think and perceive themselves, their world and obviously their ways to understand and judge their interlocutors.
Therefore the self-reflexivity of the researchers, first of all, should be towards and for the people they are working with, as 'it modifies and regulates the hierarchy positions among those involved in the project and leads us to a more democratic, relationship and knowledge production' (Omrani 2013:12). Research participants or collaborators should know how they are being perceived by researchers, and to a greater extent by those who are going to "consume" the work. We strongly believe that every shared anthropological project should have this concern at the core of its agenda. Jean Rouch (in Rouch & Feld 2003), as one of the first (audio-visual) anthropologists to develop collaborative work and a pioneer of shared anthropology, asked for feedback from his local informants before finalising his montage. We wonder what would have happened if the informants had the chance to watch the work in Europe and among a European audience. How would this affect their intention and their feedback on the work? We suggest that this important aspect should be considered in any collaborative work, whether the collaborators are holding the camera, or co-authoring a work that is filmed by another party. Together with Asef, Omrani reviewed and discussed the major (problematic) assumptions and potential questions, they anticipated, that the audience might have: how should those questions be addressed? How could the audience be persuaded and through reflecting on his life story how could Asef express the aspects and elements that are either left untold or have been told differently?
These kinds of projects could help to generate another form of the famous 'parallax effect' coined by Fay Ginsburg (1995); the importance of subject-generated projects that offer a counterbalance to the media discourse and other individual representations page 14 of 17  (1998:138), to 'create a better position to address conflicting views of reality', but explicitly by challenging the assumption that the observers and observed are strictly separated.
Although we cannot consider Passager as a purely subject-generated film as Omrani had been the initiator of the project and the editor of the film, Asef still found the space and means to mark his presence and create significant moments in the film to express himself as the author. Maybe he was not completely aware of or optimistic about the result of the project judging by his voice throughout the film. Yet, we assume that it had been a worthy practice, at least to gain self-confidence and be ready when another opportunity arises.
With the comparative analysis of these two films, we would like to highlight how different conditions led each project to adjust the degree of collaboration with its subjects. More specifically, in the first project, the final work turned out to be less authored collaboratively and consequently more abstract in form, while in the second project, the very specific condition (and obviously the willingness and motivation of the collaborator) forced the project to take a collaborative path. While a collaborative approach that integrates the subject's role as a co-author of the work and not "merely" a research subject, is the ideal, it is, however, vital to consider that not every project has the capacity to be authored collaboratively, and the degree of involvement vary with each project.
At the same time, it is also impossible to set "criteria" for the degree of collaboration.
As prospects for "equality" in expertise are limited, one must be sceptical of how such characteristic can be quantified and measured. We therefore propose that various types of collaboration with subjects can take place and be assessed based on each party's degree of involvement in the authorship and development of the work, and thus its audio-visual rhetoric. However, a corresponding straightforward and selfreflexive procedure that documents the production process remains vital.