Presenting research results Figure B.1 Jero Tapakan (foreground, left] and Linda Connor (foreground, right] watch a film on cremation together with other residents of Jero's hamlet. Bali, 1979. Photographer: Patsy Asch 6.1 Audiences Perhaps the most important issue to consider when presenting research results - of any kind, not just the results of visual research -is the audience. For much of the time academics normally only have to consider two audiences in their writing and research presentations -their students and their peers. While the nature and context of the presentation may or should cause changes in how the material is presented (writing for a learned journal versus an informal 'work in progress' seminar, for example), the characteristics of the two audiences are generally well-known and presentations to them follow relatively clear and well-established conventions. In other words, while some 140 VISUAL METHODS IN SOCIAL RESEARCH academics are notoriously poor communicators, in principle they should be able to master the conventions to make clear and effective presentations of their research results to their chosen audiences. Some academics and other social researchers similarly need to learn to master the conventions to present results to other audiences - public policy makers, for example - while part of a student's training should help them to communicate effectively.1 The presentation of visual materials introduces some additional complexities. First, there is the 'problem of images' generally, and their poor appreciation as valid research material in some parts of some academic disciplines (MacDougall 1997: 276). Secondly, the multivocality of visual images means they can address different audiences in quite different ways, creating a 'problem of audiences'. In Chapter 4.3.11 mentioned the reader-response or reception theory literature that developed first in literary studies and was later exported to media studies and cultural studies. When these approaches finally reached anthropology the primary focus of attention was ethnographic film and its reception (for example, Crawford and Hafsteinsson 1996). Much of this literature highlights the fact that audiences, particularly student audiences, do not transparently and naturally read ethnographic films, but bring to them previously formed social and cultural understandings. David MacDougall was probably the first to bring anthropologists' attention to the fact that the 'meaning' of an ethnographic film was not inherent within either the film itself or in the intentions of its author(s), but was a negotiable property that lies within a conceptual triangle formed by the (film) subject, the filmmaker and the audience (MacDougall 1978: 422). Following on from MacDougall's and others' insights and allying it to a body of reader-response theory, Wilton Martinez, a doctoral student at the University of Southern California, constructed a meticulously-detailed research project that evaluated undergraduate student response to a select number of ethnographic films (Martinez 1990, 1992). The films were shown to the students as part of an introductory anthropology course and Martinez observed the students watching, issued questionnaires, analysed weekly film reports and essay assignments, and even collected narratives of their dreams (Martinez 1990: 38). His findings were disturbing for those involved with ethnographic film. He concludes that certain films could generate an 'aberrant' response, that far from coming to understand or even sympathize with the film's generally exotic subjects (exotic to the students, that is) the students tended to use the subjects' actions and appearance to confirm inaccurate stereotypes they held of 'primitive' and 'tribal' peoples. Some films, specifically intended to be used in student education, were also found to PRESENTING RESEARCH RESULTS 141 be 'dry' and 'boring' (one example is Asch and Chagnon's The Ax Fight [1975] discussed in Section 6.5.1 below). Even where their attention was engaged by a film, typically because it contained a strong central character and clear narrative drive, such as Nairn's 1974 television film The Kawelka: Ongka's Big Moka, their enjoyment could be patronizing and confirm stereotypes of the simple-minded 'natives' (Martinez 1990: 43-4). Taking a further cue from Umberto Eco and notions of 'open' and 'closed' texts, Martinez also concludes that the more didactic a film is, the more clearly it attempts to establish a clear authorial voice through heavy narration, diagrams and so forth, the more likely a film is to be the subject of an aberrant reading. In contrast, semiotically 'open' films, ones that allow or indeed enforce on the viewer a greater interpretative effort, met with more elaborated and reflexive responses from the students (Martinez 1992: 135-6). In such 'open' films, the space in MacDougall's triangle between subject, author and reader is self-consciously presented as an interpretative space. Martinez's solution to the problem of aberrant readings is thus twofold. First, certain kinds of films are more likely to promote more critical and self-aware responses from viewers;2 secondly, those presenting films to students should help students to develop visual literacy skills, to approach ethnographic films not simply as transparently-represented ethnographic knowledge, but as a particular genre of filmmaking that has a history, a changing set of representational conventions, and a changing set of differentially-placed authors (Martinez 1990: 46; 1992: 152-6). In other words and in my terms, to alert viewers to the external narratives surrounding the films, rather than assuming an unproblematic and automatic transmission of the internal narrative or content. While Martinez is one of the few to have examined the pedagogic value of ethnographic film empirically (indeed, as far as I know he is the only one to have done so), he is not alone in seeking criteria by which the communicative power of visual materials can be assessed. To conclude this section I want briefly to refer to two such assessment models which, while very different in their underlying aims and objectives, have a superficial similarity. Karl Heider's 1976 book on ethnographic film, while now quite elderly, was for a long time influential simply because no other comparative text existed. Like many subsequent authors, Heider deals at length with the problem of defining 'ethnographic film' (1976: 3 ff.). He settles finally on a multi-faceted property of 'ethnographic-ness' - the degree to which a film successfully conveys an ethnographic understanding of the people and activities represented. This he subsequently formalizes into what he calls an 'attribute dimension grid' - a graphic rendering by which fourteen properties or facets can be charted 142 VISUAL METHODS IN SOCIAL RESEARCH (1976: 97-117). These range from technical and production aspects of the finished film itself (basic technical competence, goodness of fit between narration and image), to its relation with external written sources such as study guides and related ethnography. A film which scores highly on all or many of the fourteen analogue scales is overall a 'more' ethnographic film than one which gains low scores for many of the attributes (see Figure 6.2). Heider is well aware of the dangers of over-specifying in this way and indeed the whole model is open to criticism. Nonetheless, his approach has the virtue of defining clearly (if subjectively) what it is he thinks is important about a visual ethnographic presentation, and then using that model to assess specific examples. In this he rather surprisingly shares ground with a very different approach towards evaluating Ethnographic Basis Relation to Print Whole Acts (hp) (Et Whole Bodies Explanation of Distortions Technical Competence IT) Appropriateness of Sound Narration Fit Ethnographic Presence Contextualization Whole People @ ===JeT) i Time Distortions in Filmmaking { 'Continuity Inadvertent Distortion of Behavior wP*gM -^LDwnlo*1 f irmUitM * fiADDON • the online catalogue of archival ethnographic film The! I ADDON du*b thrt* pop-up buttons HADDON dilibiii Other tee-tsyJ search Theie 4» some me] Cototisehtil ttgion: Modrmcaunliy. SubrtgiOn HADI .✓Amazonia Hrclic Circle nuttrelia Caribbean Central N Trie a Central America Central Asia East Rfrita Cast Asia Eastern Europe Melanesia Micronesia Middle East North Africa North America Northern Europe Polynesia Sahara Desert South America )ON Area Search itst begin ycui seurhby ctioojing s pojraphkil npon from one of the Splay itoords.'button. Thi» is the most eftVcliv* v/ny to sttrtt the ; your stilt h further. :ling 1 he button at the bottom of the page. J r I • ci.-i '• . j—| [ Display lecords ] -7-j | Duplny records ] i j [ Display lerank J South fisia Other Searches Southeast Hsia Southern Africa men Africa Western Europe h ^ vj Cj3 -&\ Figure 6.4 Screen shot from HADDON Catalogue Web site, showing selection of geographical area catalogues, paper and computerized, though the larger a catalogue is -such as a major university library catalogue - the easier it is for users to fall into the representational trap of thinking that it is a perfect and complete index of that which it catalogues. As I said above, the HADDON Catalogue is a meta-catalogue, drawing together and standardizing subsections of many other catalogues to address a narrow and specific target audience. Its one great weakness is that it does not contain digitized copies of the films themselves. While this is technically possible, it is not practically possible for the simple reason that very few if any film archives have the resources even to contemplate digitizing their collections (and on the budget available to me for the catalogue project, nor could I). Single films and lengths of film footage are, however, increasingly being incorporated into academic and commercial multimedia packages (see Section 6.5.1 below). 160 VISUAL METHODS IN SOCIAL RESEARCH Whether digitized copies of their still photographs, movie film and videotape exist or not, the social researcher should also consider the possibility that their own material may one day end up in an archive, for use by future researchers. I said above that my own database of digitized fieldwork photographs was for my own private use and was, therefore, organized and indexed for my own convenience. With an intended audience of one issues of presentation are easy to resolve. The principal problem with archiving visual material is to anticipate the needs of unknown future users, a vast potential audience of unknown interests. The only answer is to provide as much information as possible. A complete account of the external narrative surrounding the images should be provided - when and where they were taken, by whom, with what intention, how they have subsequently been used - together with a grounded reading of the internal narrative, not only covering a description of the image's superficial content but also the basis upon which that reading is founded. Many of my field photographs, for example, depict aspects of Indian Jain ritual and ceremonial life, but my reading of the content rests in part on my understanding of Jain social organization. Figure 6.5 (a repeat of Figure 2.14, printed at the end of Chapter 2), depicts a woman in a sari leaning over to serve something to a smiling man, seated in a row of Figure 6.5 A fieldwork photograph. Jamnagar, India, 1983 PRESENTING RESEARCH RESULTS 161 men. It depicts a particular named woman, serving a particular food item to a particular named man, at a particular named feast held at the conclusion of a named period of fasting, on a particular date, in a particular place. It depicts a Jain laywoman honouring a Jain layman in recognition of an austerity he has performed. It depicts a reversal of status with a woman (not a man) gifting food to a man, rather than preparing it as a domestic duty. It depicts ties between migrant Jains in the UK (the woman is British) and India. At each level, the reading of the image becomes increasingly linked to my own sociological analysis -more interpretative and less reliant purely on the internal narrative, as it were. If intended for archival deposit, I should therefore make the basis of my interpretation clear in my description of the image.15 6.5 Multimedia projects In an oft-cited 1945 article, Vannevar Bush, then President of the Carnegie Institution, envisaged a hypothetical machine - the 'memex', 'a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility' (Bush 1995: Section 6). Once a user had stored (on microfilm) all the relevant material, and indexed it, he would use a series of mechanical levers and dry photographic processes to create 'trails' through the information, forging links between relevant items. Bush envisaged the 'memex' primarily as a research tool, though he also describes how a researcher could demonstrate a 'trail' to another party -the example he gives is of proving the technological superiority of the short Turkish bow over the English long bow used in the Crusades (1995: Section 7). Today, despite the claims that Bush's device intellectually prefigured the World Wide Web, users of the Web still have no easy way to link items of 'found' information in the way that Bush envisaged, short of constructing a crude list of 'bookmarks' or 'favorites' in their Web browser. Users can, however, construct their own Web sites (or standalone multimedia corpuses written on removable media such as CD-ROMs) which do permit a greater finesse of linkages. Within their own Web site or other corpus, links between sound, text and image files can be constructed at will, although again such links are largely presented as givens to the user. Consequently, although many multimedia applications within the social sciences (Web-based and free-standing) often have a strong orientation towards promoting and facilitating research by users - allowing them access to unedited text, raw video footage, and so on - 162 VISUAL METHODS IN SOCIAL RESEARCH they can equally well be seen as presentational applications, presenting materials and interpretation in a form that has already been pre-structured, implicitly or explicitly. Nevertheless, one of the advantages of computer-based media is the access it provides to images, and the ability to promote research parity between sound, text and image. There is a price to be paid for this: in the digital world of bits-not-atoms, all bits are equal, but at the cost of the loss of the unique materiality of the atoms. The material difficulty (and economic cost) of publishing large numbers of high quality photographs in a written monograph, and the material impossibility of incorporating moving images or sound on the page are transcended, but in doing so an aspect of their unique materiality is lost.16 Nevertheless, if creator and user of a product are both aware that they are dealing with copies of images, texts and other original source material and aware that part of the external narrative of these objects has been stripped away, then computer-based multimedia carries a high potential for the presentation of research results to particular kinds of audiences.17 In a recent paper, Michael Fischer and David Zeitlyn note that there are several broad models for computer-based multimedia (Fischer and Zeitlyn n.d.). One is strongly time-based and sequential, similar in many ways to a film. For the audience, viewing it is a largely passive experience. A sequence of images, text, graphics and other items follow in predetermined order, although the sequences may be organized into chunks which can be viewed repeatedly or in a variant order. Typical examples would be a marketing presentation to a group of clients using a set of PowerPoint slides or a software 'guide' which uses animation to demonstrate the features of a piece of software. A second model is closer in organization to a book. A series of objects (such as text files) are linked in an overarching sequence, but the user has the opportunity to explore non-sequential links ('hyperlinks'), either to material outside the main body of the text - the equivalent of footnotes - or to other occurrences of similar material - akin to sequentially reading all the page references to a given term in a book's index, without reading the intervening material. Fischer and Zeitlyn point out that this model gives a 'more powerful' book, one which can incorporate moving images and sound. Visual anthropologists have increasingly been attracted to the possibilities this model opens up for adding visual and textual 'footnotes' to ethnographic films. In either case, the model is one of a principal text with associated subtexts. A third model, and one that perhaps best exploits the medium's potential, is what Fischer and Zeitlyn term the 'layered' model. In this PRESENTING RESEARCH RESULTS 163 model, no object or set of objects is necessarily primary, nor does the object or object-set necessarily have any definitive organizational structure or predetermined sequencing. Instead, a set of objects are linked to one another in (metaphorical) layers - a group of photographs, a set of fieldnotes - and also linked to objects in other layers: a photograph of an individual in one layer is linked to a genealogical diagram in which the individual features in another layer, and also to half a dozen fieldnotes that concern that individual in yet another layer. The linkages within the layers may be tight and sequential (for example, all the frames in a length of digitized film footage may be viewed randomly but are probably best appreciated in order) or relatively loosely linked (for example photographs of all the inhabitants of a village could be clustered into household or kin groups, but there is probably little to be gained by viewing the photographs in any predetermined order). This last, the most 'open' model, is potentially the most powerful and useful for the social researcher, but equally the most fraught with potential difficulties - most of which relate to audience and audience use of the final product. In the first 'marketing presentation' model the audience is known or categorizable and their lines of enquiry relatively predictable: how will the launch of a new product affect next quarter's market share? what are the key features of this new version of the operating system? The familiar model of a lecture or talk by an expert aids the audience in making sense of similar material presented in a new medium. The second 'book' model also allows the audience to use familiar concepts, such as footnotes, appendixes and indexes, to navigate an initially unfamiliar product. In contrast, the third layered model makes new demands on the user by presenting a network of information rather than a narrative flow, which places a far greater onus on the user to steer their own course. We may not think serially, as Fischer and Zeitlyn point out, but 'we like to think that we think serially'. Dudley and Petch, who collaborated with Fischer and Zeitlyn on a museum-based multimedia project, recognize in reflecting on their experiences that this type of multimedia is an 'infant forum' for the presentation of material, one whose manipulation is still relatively unfamiliar (Dudley and Petch n.d.). If the strictures advanced by critics such as Ruby for the production of ethnographic film (for example, Ruby 2000: Chapter 10) have any validity, then they apply equally to the production of computer-based multimedia. Not only should the source materials have been created or selected by a social researcher within the context of a carefully-conceived research project, but they should be assembled into a multimedia product 164 VISUAL METHODS IN SOCIAL RESEARCH that is as clearly focussed. A difficult line has to be trodden in this respect with regard to freedom and authority. Multimedia, especially of the 'layered' type, offers a great deal of freedom to the user to explore their own lines of enquiry - indeed, this is the medium's strength. On the other hand, if the user has little or no sense of the research frame within which the project lies, nor any sense of the authority of the product's creator to compile the product, then it is difficult for the user to gauge the validity of the choices that she makes in navigating through the materials. If all authority is abrogated or concealed in the name of the freedom of consumer choice, then the user has no incentive to even bother exploring the product. Simply being told that a multimedia CD-ROM will help the user 'find out about' or 'explore' the issue of homelessness, or indigenous rights, or tribal art is akin to being told to go into a library and start reading.18 Following Ruby's suggestions (see Section 6.3.1), one clear way forward is the path of reflexivity If a social researcher has a specific argument to make, one that she considers objective, universal and which will brook no gainsaying, then computer-based multimedia is probably not the best medium. If a social researcher has an argument to put forward, but is aware that there are variant interpretations of what is presented as supporting evidence (almost always the case with visual materials), then she can usefully employ multimedia to state her own case but also to outline the alternative interpretations and provide access to the raw materials to allow the user to test them all. In fact, this is almost always the case as soon as images are involved as an integral part of an argument, rather than merely as redundant illustrations.19 6.5.1 Interacting with the Yanomamo A clear example is provided by Peter Biella and colleagues in a stimulating multimedia version of an explicitly visual product, in this case an elderly but renowned ethnographic film extensively used in North American anthropology teaching (Biella, Chagnon and Seaman 1997). The CD-ROM Yanomamo Interactive has at its centre Tim Asch and Napolean Chagnon's 1975 film, The Ax Fight. The film itself is unusual in a number of ways, not least because as originally constructed it is a self-conscious multimedia piece in its own right. At its core lies around 11 minutes of unedited film footage depicting a fight between two groups of Yanomamo Indians - an Amazonian group living on the Venezuelan-Brazilian border. This footage opens the film, after the opening titles and a brief snatch of narration, and is then followed by four further PRESENTING RESEARCH RESULTS 165 sequences. First there is a short audio sequence over a blank screen in which the anthropologist and filmmakers discuss what they have just witnessed; then some of the original footage is replayed, but using frame stills and slow motion, as the narration describes the causes and progress of the fight; then there is a series of rostrum camera shots of genealogical diagrams, showing how each of the participants in the fight are related to each other; finally a section entitled 'a final edited version' replays the fight sequence again, without commentary but edited to provide a smooth clean sequence. The aim of the film is to build up a number of layers, commencing with raw 'reality' as the film crew experienced it (they had no forewarning of the fight and filmed with no preparation from a distant vantage point), which is then overlain with layers of personal and sociological detail and analysis. Viewers of this linear, time-based medium have to carry a great deal in their heads as the film progresses, in order to use it as a basis for reading the 'final edited version'. As Biella notes in the CD-ROM version's introductory essay: Despite its effort at precision, to many viewers [the middle] section of the film is unclear. The confusion comes in part from the poor visual quality of the genealogical diagrams and a series of clumsy cuts and pans. More than this, the argument itself is difficult . . . before viewers can even begin to understand the argument they must first associate different faces in the film with genealogical icons, and comprehend the genealogical relationships between the faces. The latter task by itself is difficult to master. (Biella 1997) In short, the potential of what the film could achieve is hindered by the original medium; the multimedia CD-ROM version by contrast, opens up a non-linear space within which the detail can be absorbed at the user's own pace, and the arguments which rest upon that detail can be fully explored. The CD-ROM contains a complete digital version of the film, which can be played right through but also accessed at a number of cue points. Yanotnatnd Interactive is not, however, based upon the book model of multimedia described above - the film as a core text with a number of visual and textual footnotes. While the film is central, other materials - textual and visual - are included as additional layers, linked both to the film and to each other. These layers can be read in their own right, partly in order to test a neo-Darwinian hypothesis advanced by Chagnon and Bugos (1979, also contained on the CD-ROM) concerning the role of kin in disputes. Biella presents some variant forms of this hypothesis in his introductory essay and suggests ways of using the 166 VISUAL METHODS IN SOCIAL RESEARCH filmic, photographic and textual material on the CD-ROM to weigh one up against another. He also suggests ways in which photographs of individuals and their biographical and genealogical data can be used to create a number of internal narratives that interrogate the original film, possibly to the extent of undermining the film's own narrative. For example, a narrative concerning the activities of women in Yanomamo society and their role in disputes can be 'recovered' from the film, despite the fact that the film's narration makes little mention of women in an active capacity, and the camera generally depicts them as reacting to rather than initiating action. Technically, the construction of these narratives are made possible by embedding a variety of links in the textual and visual materials. While there are a very large number of such links, they inevitably serve to constrain and channel the user's navigation of the product. Within the textual materials - the full narration of the film, additional commentaries, associated essays - the links are embedded in a specific and self-evident way: clicking on an individual's name in a text in one window may cause a photograph of the individual to be displayed in another window, or a genealogical diagram to be redrawn with that individual at its centre. Within the visual materials, the links are less evident. Of the many persons and objects within the film and the associated still photographs, Biella and colleagues have chosen to tag individuals - men, women, children - rather than, say, items of material culture. A click on an individual's name in the narration window cues the film to the appropriate point and identifies the individual on screen with a red cross (see Figure 6.6). A click on the word 'hammock' or 'ax' does not produce the same result. This is not a criticism of the multimedia product, merely a reiteration of the point made above (Section 6.4.1) that while text is easily and automatically transformed into hypertext such that all text strings can be matched to all equivalent text strings, computers have great difficulty in dealing with 'hypervisuality' matching one image to another, or automatically linking text string to visual representation. There is a great deal more that could be said about the value (and limitations) of computer-based multimedia for the presentation of visual research materials. Briefly, Yanomamo Interactive succeeds because the specific visual and textual materials are selected and interlinked with a clear aim and purpose. The authors have also considered their audience: they are explicit in stating what materials were used, and what thought lay behind the project - reflexivity - and in acknowledging that those encountering this 'infant forum' (Dudley and Petch n.d.) need clear and straightforward guidance to aid in their reading. PRESENTING RESEARCH RESULTS 167 Blow-by-Bkjw bv Gar, Si-aiiiiU! ttmmar EHBSlEanBlESa I 198280 Kebowa continues ro strike Mohesiwa, undeterred I by the fact that Tarawa thiieatens him with a raised machete. I Yoinakuwa' watches Torawa A woman, possibly Yaukuima, I the sister of Kebdwa and Yoinakuwa, runs across the plaza I with a hemo club over her shoulder. ^......-----—^- Agp: 40 Sex: F IjlKDjJl.'! 2700 Children: 3 Spouses: 2 Figure 6.6 Screen shot from Yanomamo Interactive (1997]; clicking on an individual's name identifies them in the film footage. Reproduced with permission of Harcourt Brace & Company. 6.6 Copyright I get very angry when I see some of the products that are advertised . . . they're stealing the Princess's image, they're stealing her dignity. (Trustee of the Diana, Princess of Wales, Memorial Fund quoted in the Electronic Telegraph, Issue 1089, 19 May 1998) For an image to be stolen, it must belong to someone in the first place. Shortly after Diana, Princess of Wales, was killed in a car crash in August 1997 images of her began to appear on thousands of memorial objects for sale - mugs, plates, plaques - as well as in innumerable print publications commemorating her life. These photographic images presumably 'belonged' to the photographers who had created them or to the agencies for which they worked. They did not 'belong' to Diana or her heirs, unless they of course had taken them or legally acquired them. But later in 1997 the Diana, Princess of Wales, Memorial Fund sought to register the image of Diana as a trademark, filing around 50 images as samples of this 'trademark' with the British Trade Marks Registry (the trademark 168 VISUAL METHODS IN SOCIAL RESEARCH Figure 6.7 Frame still from Paul Henley's film Faces in the Crowd (1994], showing Diana, Princess of Wales, on a visit to Solihull, West Midlands. Photographer: Paul Henley application was later transferred to the executors of Diana's will). While single images of individuals have successfully been registered as trademarks - for example, a distinctive image of the racing driver Damon Hill looking through his visor - the large number of Diana images submitted appeared to be a claim for all and any image of Diana to be covered, preventing almost any use of her image from being used without payment of royalties to her estate (Electronic Telegraph, Issue 1329, 14 January 1999). On the back of this, the Fund embarked on legal action against the American manufacturers of a commemorative 'Diana' doll, claiming that the company was 'exploiting her identity' (Electronic Telegraph, Issue 1089, 19 May 1998). The case served to worry many people in the media that merely creating and reproducing an image may not guarantee ownership of that image. It seems relevant largely to representations of celebrities, those deemed likely to have a financial interest in the use of their representations and, with some exceptions, few academics routinely conduct research with such images. More down-to-earth issues of copyright still remain important however. Many images that the social researcher will PRESENTING RESEARCH RESULTS 169 be using in the course of research will have been created by the researcher herself, and consequently she will normally own the copyright and have relative freedom in her use of the images (although see the earlier discussion in Chapter 5.6 on image ethics). However, any images used that come from other sources may well be subject to copyright and permission will have to be sought for their use in any publication, including 'publication' on a Web site. Permission generally consists of two aspects: seeking permission from the copyright holder to publish the image or to screen a film, and payment of any reproduction or screening fee. In addition, the copyright holder may request (and be entitled to if it is a condition of the permission) a copy of the publication together with an acknowledgement in the publication, or film programme. Finally, the user normally has to pay whatever reproduction or handling charges are involved in making the image available. Normally, for educational uses reproduction fees are waived or a purely nominal charge is made. Researchers requesting reproduction rights should make it explicit in their letter of request that the publication (or Web site) is academic not commercial (assuming this is in fact the case). Reproduction costs and handling charges are, however, fixed and inevitable costs incurred by the individual or organization that holds the copyright and are not normally waived. The normal procedure is to identify the image or footage that is required and then to write to the copyright holder explaining what the image is required for and either requesting a copy of the image or permission to reproduce it from another source (for example, a book plate). A line should be added requesting that reproduction fees, if any, might be waived as the image will be used in a purely academic or educational context. Larger museums and archives normally have their own form to complete, covering much or all of the above. There are, needless to say, many complications that can arise. Most commonly, the copyright holder may not be the individual or institution holding the image. In cases where book plate reproduction is sought, the first course of action is normally to contact the publisher if no other details are given. Alternatively, the caption for the image (or possibly the acknowledgements page of the book) will indicate the source of the image, such as a museum or art gallery. In the case of some film archives the archive acts as a holding institution for other people's material and does not have copyright in the films (for example, the UK's National Film and Television Archive does not own copyright in the majority of the films it holds).20 In cases such as these, the reproduction is normally obtained from the holding institution - and appropriate charges paid; the 170 VISUAL METHODS IN SOCIAL RESEARCH permission to use the image - and again, any rights or reproduction fees - are negotiated with the copyright holder. It is sometimes the case that the copyright holder cannot be located - especially if she or he were an amateur image maker who deposited the material long ago and has since died or moved on. When images such as these are used in academic publications, the publishers may cover themselves with a standard form of words in the book's introductory matter along the lines of 'Every effort has been made to trace all the copyright holders [of the images in this book], but if any have been overlooked, or if any additional information can be given, the publishers will be pleased to make the necessary amendments at the first opportunity' (Evans and Hall 1999: xviii). Protecting one's own copyright in images is quite a different matter. For example, it is not wise to publish high resolution images on the Web without any form of digital or manual watermarking (see Chapter 3.5.1),21 although in practice most network operators (the IT specialist who oversees a network of computers) would probably raise objections to this on account of the storage costs for such large files - several megabytes for photographic quality reproduction. Generally, the low resolution of 72 d.p.i. (dots per inch) which is all that is required for onscreen viewing is too poor quality for ink-on-paper publication. Equally unwise is to loan a photographic print to a friend or colleague for publication without ensuring that appropriate credit will be given (in addition, publishers often have part of a book's production budget set aside for reproduction fees - the front cover image, for example - and it may be worth asking if a reproduction fee is available). Even if an image is not originally published on the Web it may easily find its way there through unauthorized scanning, after which it is extremely difficult both to monitor and to prevent unauthorized reproduction. The best an image maker can do if she sees one of her images used on a Web site without permission is to contact the site's owner and either ask for the image to be removed or for appropriate credit to be made (a request for a fee is unlikely to be met). If that brings no satisfactory result then the researcher should contact the Web host, the company that provides storage for the site's files and access to them. Many Web hosting companies, especially the larger ones with publicly quoted stock, are very sensitive to allegations of Web misuse. Of course, many amateur photographers and film- or videomakers are quite happy to see their work more widely distributed, or are simply not bothered about issues of copyright especially given that there is rarely any financial advantage to be gained. Some anthropologists, however - including myself - are concerned about the ethical implications of allowing images to flow freely through the networks, especially when their content becomes PRESENTING RESEARCH RESULTS 171 divorced from their context. There is a potential for misuse, for example with images of tribal peoples when unwanted or misleading emphasis could subsequently be placed on their nudity or apparent savagery (see also Chapter 5.6 on ethics). Notes 1 The extent to which clear communication is effected or to which presenters of research consciously consider their audiences at all is, of course, extremely variable. My argument here is not so much that social researchers should learn to present their results more effectively, simply that whether they are conscious of it or not, there is an audience. 2 Martinez cites several 'open text' films as examples; the qualities they share include reflexivity - an acknowledgement within the film that a filmic representation is being constructed (see Section 6.3) - and narrativity (see Section 6.2.1). 3 The Finnish photographer Jorma Puranen has actively sought to liberate the long-silenced voices of archival images. By rephotographing images from the 1880s of Sami indigenous people, printing them on plexiglass, placing the panels in the northern Finnish landscape and then photographing the resulting installations, Puranen has loosened the constraining narrative of 'races and types' photography of the late nineteenth century, as well as the narrative of 'archival interest', and has effected an 'imaginary homecoming' for these long-dead Sami (see Edwards 1997: 72-3; Puranen 1999). 4 For another account of selecting images for academic publication, see Harper 1987a: 7-9; see also Wright 1999: 97-105. 5 Harper also stresses that the assignment of images to any of the four categories depends on the overall research frame, not properties inherent in the images themselves. The other two categories, which I do not discuss here and which do not necessarily rely on viewing the images in sequence, are the scientific mode - images treated primarily as evidence or data; and the reflexive mode - images that mediate the relationship between social researcher and subject, as in photo-elicitation (see Chapter 4.4). 6 For Grady, rendering sociology more appealing to students is not a trivial issue. Following Becker (1986) he argues that the reproduction of the discipline is in crisis, producing students - and hence future teachers - who have little appreciation of what he calls the craft of sociology (cf. Epstein 1967) and who are 'insecure, deeply unsure of their work and their voice, and more concerned with justifying their projects than with carrying them out' (Grady 1991: 37, n. 5). Producing visual narratives instead of written essays from secondary sources, Grady argues, would allow students to develop these craft skills and consequently a confident analytical voice. 7 In my experience, some of the least engaging ethnographic films I have seen are those that were edited from research footage by their makers who had not 172 VISUAL METHODS IN SOCIAL RESEARCH at the time intended to make 'a movie' but were unwisely encouraged to do so afterwards. See also Barbash and Taylor (1997: 287) on planning for the audience in advance of shooting. 8 These debates are also found with regard to the internal narrative or content of the film itself. Proponents of the direct cinema and cinéma-vérité movements of the 1960s and 1970s generally eschewed voice-over narration and interviews, claiming that the flow of live action alone should be strong enough to sustain the film and communicate what the filmmaker wished to say; see also the debate surrounding Gardner's Forest of Bliss in Chapter 2.2.1. 9 It is one thing for a filmmaker to decide to write a study guide, another to decide what to put in it. Some I have seen are little more than a background ethnographic essay of a largely factual nature - details of economy, habitat, marriage practices and so forth. This approach would seem to indicate that the film is intended as little more than an illustration of these facts. Instead I would suggest that a study guide should include a full transcript of the film's dialogue, camera angles and editing cuts, together with an essay outlining the circumstances of the film shoot. In this way a viewer can appreciate the film as film - as a constructed representation, as well as the film as ethnography -as a representation of particular people. 10 The catalogues for the Royal Anthropological Institute's ethnographic film library (Woodburn 1982; Willson 1990) contain a list of recommended reading for each film. 11 In Britain for example, the Computers and History of Art Group (CHArt) encourages research and discussion in this area and organizes annual conferences. See also Frauenfelder (1997) for an interesting overlap of agendas. 12 The HADDON Catalogue is available through the Web at . 13 At least for the early footage, my criterion was that material had to be shot on location, in the field, and hence the catalogue includes A.C. Haddon's Torres Strait Island footage (1898). Earlier material of some ethnographic relevance, though shot away from the subjects' normal location, would include Thomas Edison's Sioux 'Ghost Dance' footage (1894), Felix Louis Regnault's Wolof potter footage (1895), and the Lumiěre brothers' twelve shorts of Asante women dancing (1897); see Jordan 1992 for further details. 14 The current draft of Archival Moving Image Materials: a Cataloguing Manual (Balkánsky et al. 1999) puts the position well. While devoting an extraordinary amount of detailed discussion to how cataloguers should differentiate between types of production credit, or how they should deal with films released with different titles in different countries, for the 'Summary' field devoted to a description of the film or television show's content it merely says 'Give a summary of the content of a work. The object of a summary is give the viewer a good idea of what to expect when he or she views the work, thus avoiding unnecessary handling of the film or video.' The brief examples given all focus on content, except for one that mentions the film includes PRESENTING RESEARCH RESULTS 173 'close-up magnified photography'. While the descriptions of a film's content are at best inadequate at many institutions, and hence in the HADDON Catalogue, descriptions of a film's form and structure are almost entirely absent (see also Usai 1994: 48). 15 It would be hypocritical of me to pretend that this is anything other than a wish list, a statement of good intention. While my photographic collection is relatively well-organized, like many other visual anthropologists, I spend my academic life surrounded by cans of disorganized film footage and cassettes of unindexed videotape. Ideally, when budgeting time and resources for a research project an allocation should be made for archiving the materials at the conclusion of the project. Perhaps funding agencies should even insist on this, as the UK's Economic and Social Research Council currently does for quantitative and qualitative (but essentially textual) datasets. 16 In an article on computer-based music, Georgina Born includes some illustrations that are actually photographs of printed pages of an in-house computer software manual (Born 1997: Figures 73-7.7, pages 147,149). Howard Morphy and I, the editors of the volume in which the article appeared, had some difficulty in persuading one or two readers of the manuscript that these images should not be replaced simply by a retyping of the text of the pages (their manifest content). Born's argument in part concerned the fact that these manuals were fluid and often incomplete documents needing handwritten annotations by those who were learning the programs from them and her photographs of the pages showed such annotations. Moreover, the poor quality of the manuals - dog-eared and blurred from much copying - was another aspect of their materiality and a manifestation of the gap that she perceived between the supposedly pure and universal logic of the programming code and the highly socially-embedded circumstances of its communication. Mechanically reproducing the unique documents through a photograph was the closest we could come to communicating some of that unique materiality, where the latent content of the pages was foregrounded by Born's sociological analysis. 17 Strictly speaking, 'multimedia' is a misnomer, the computer being the single medium through which a number of sound, text and image files are displayed. The term acts as a useful reminder however of the media-specific objects - photographs, audiotapes, 16mm films - that have been brought together on the computer. The term is so omnipresent and associated so closely with a new physical object, the CD-ROM or DVD, that I shall continue to use it here. 18 I have written about these issues elsewhere, somewhat overstating the arguments against multimedia, particularly for educational use, to stimulate debate. See Banks 1994 and Biella 1994. 19 This variant interpretation reflexivity model is well-demonstrated in 'Ancestors in Africa', a largely text-based multimedia product by David Zeitlyn (Fischer and Zeitlyn 1999). Zeitlyn's project, together with several others, forms part of a package under the title 'Experience-Rich Anthropology', distributed 174 VISUAL METHODS IN SOCIAL RESEARCH widely in Britain in 1999-2000 on CD-ROM. Many of the other projects include far more visual materials than the one I discuss here, but this nonetheless provides a good model. 20 By contrast, stock-shot libraries (film clips, normally out-takes from completed productions) and picture or photo libraries, have normally cleared the copyright on everything they hold, enabling the user to make a 'one-stop shop' and pay a single amalgamated fee. However, such libraries are normally commercial and oriented to the needs of commercial customers, making them expensive for the student or academic researcher. 21 Terence Wright raises the interesting point that a photograph taken from one source and subsequently manipulated digitally by another photographer or artist could be considered to be a new image, unconstrained by the original owner's copyright (Wright 1999: 164-5). This situation would be compounded when the 'original' was taken by a digital camera, there being no original source apart from an easily manipulable string of digital code to prove where the altered copy had come from. In fact, software similar to that used for digital watermarking is available that can tag electronic documents of all kinds and then indicate subsequent alterations.