Image as Interface: Consequences for Users of Museum Knowledge

Photographs of objects are ubiquitous in the work and presentation of museums, whether in collection-management infrastructure or in Web-based communication. This article examines the use of images in these settings and traces how they function as interfaces and tools in the production of museum knowledge. Because images are not only the main material presented but also become multilayered objects on which to act in order to access or produce knowledge, they play a key role in the involvement of users with museums. This development is analyzed in the context of the Tropenmuseum (an ethnographic museum in Amsterdam, the Netherlands) based on an ethnographic study of visual practices at and about the museum. Drawing on science and technology studies and new media studies, our theoretically driven analysis demonstrates how images as interfaces provide networked contexts for museum knowledge. The various dimensions of images as interfaces in museums are explored through the questions: How are users engaged by these interfaces? Which skills and strategies are needed for this engagement? What are the consequences of visually mediated interfaces for users of digital knowledge in/about/from museums, archives, and other collections? These developments are discussed in terms of their consequences for how museums view their role. Introduction A strong trend in the design and presentation of digital collections is to use images as interfaces. In the context of museums, archives, and libraries, these digital images traditionally have a documentary function; they 664 library trends/winter 2011 posit objects as referents that one might witness during a visit (ethnographic objects, books, maps, etc.). In these contexts, the way images take on instrumental and authoritative roles remediates tradition and supports new practices (Bolter & Grusin, 2000). Comparing, producing, sharing, annotating, searching, and viewing such images are increasingly important epistemic strategies. While all photographs carry traces of their context of production and use, the photographic images considered in this chapter have a particular embedding at the intersection of digital technologies and electronic networks. In this context images are portable, spontaneously produced, and easily translatable across technological platforms. Furthermore, this intersection also means that images can be related to each other within databases or with many other resources on the Web and that they serve as support for mediated social interactions such as discussion, annotation, or photosharing. The images are therefore not only the main material presented but become themselves forms of engagement and of embedding that shape access to and the production of knowledge. In this article we analyze how these new practices are developing in the context of the Tropenmuseum (an ethnographic museum in Amsterdam). The Tropenmuseum is a useful case for our analysis for at least two reasons. First, the museum makes extensive use of a Web-based collection database of images in a system called The Museum System (TMS). The database not only structures much of the institutional work processes within the museum, but also (re)defines what can count as the collection and the ways in which other users can interact with the collection via digital images. Second, the database is progressively configuring images as interfaces to other kinds of information and to other kinds of activities. At the Tropenmuseum, the main institutional investment in collections management has been in the development of the Web-based image database, TMS. This collection database was put into use in 2000 and carried a number of promises: of modernizing the museum, of improving management, and of enabling the museum to become a better caretaker of its collections. In addition, the museum explicitly aimed at using the database to change user interaction with the collections for employees and for museum and museum website visitors alike. The hope was that the networked database would diminish the number of times museum employees would need access to depots by replacing the practice of handling physical objects by that of consulting a collection database. The database was also introduced with the goal of making the museum collections available to a wider audience (including its original community) via the Web and to help multiply the number of visitors to the website and the museum. While TMS is an extremely important factor in transforming interactions with the Tropenmuseum collections, other projects are also changing the role of users. Following international trends in the museum world, 665 de rijcke & beaulieu/image as interface the museum is currently investing in other new, distributed infrastructures for visual knowing. Several of these initiatives focus on involving new users by using images as interfaces. Our analysis of the consequences of deploying images as interfaces for users of museum knowledge is guided by the following questions: (1) How are users engaged by these interfaces? (2) Which skills are needed and how are they learned? (3) What are the consequences of these visually mediated interfaces for users of digital knowledge? Because they cut across issues of digitization, such as metadata, outreach, personalization, and user engagement (Marty, 2007), images treated as interfaces enable us to explore the links between digital culture and museum practices. These images thus help constitute new forms of knowledge production in relation to practices of user engagement via digital forms, including the new values that may arise from such intersections (Mason, 2007). This article is based on ethnographic fieldwork at the Tropenmuseum in 2009. The fieldwork consisted of systematic participant observation; open-ended interviews with museum employees and visitors; a detailed scrutiny of new Web-based initiatives in relation to the museum collection; and an examination of official policy documents, relevant archival material, and funding applications relating to digitization and information management. Conceptually, we draw on two bodies of work: new media theory and science and technology studies (STS). Together, they allow us to analyze mediation processes and the dynamics of technologies involved in manipulating and circulating images. We use new media studies to analyze the co-existence of different frameworks of mediated interactions with images. This approach enables us to scrutinize these interactions in relation to other spheres of visual culture and to the history of representations (Cartwright, 1995; van Dijck, 2005; de Rijcke, 2008a, b). STS emphasizes the importance of innovation and embedding of new forms of knowledge, including material and institutional aspects (Beaulieu, van Heur & de Rijcke, 2010; Hand, 2008). Furthermore, we take from technology studies the practice of understanding the term ‘users’ as those interacting with a technology—in our case, an interface (cf. Oudshoorn & Pinch, 2003). This enables us to treat a variety of actors more symmetrically without using labels that presuppose that those in the museum are producers and those outside are consumers. It also heightens our awareness of potential changes in existing configurations of users and images as interfaces. Our argument is threefold. First, we demonstrate that images become increasingly active objects, which have many functions besides being viewed. This leads to a revised concept of how to interact with an image. This has consequences for how museums then think about images in situ. Second, we argue that images as interfaces provide a networked context for digital knowledge, creating the conditions that can lead to interactions 666 library trends/winter 2011 that exceed the limits of single images, single collections or institutions, and even of single platforms. This has consequences for how museums view their role. Third, having images as interfaces reinscribes museums and other institutions in contemporary visual culture where media coverage and user-generated content is increasingly relevant. The Image as Interface How can we understand the fact that the images are not only the main material presented in a museum but are themselves increasingly multilayered objects on which to act in order to access knowledge? First of all, the particularities of the visual culture within museums need to be taken account. According to Hooper-Greenhill (2000), museums have always ascribed a large role to the visual because of their focus on the museum collection and the visual display of objects: “The power of display as a method of communication lies in its capacity to produce visual narratives that are apparently harmonious, unified and complete. These holistic and apparently inevitable visual narratives, generally presented with anonymous authority, legitimised specific attitudes and opinions and gave them the status of truth” (p. 151). Hooper-Greenhill argues that “display” practices tend to enforce one-way communication and are difficult to modify because they are built into the structures and practices of institutions. Simultaneously she notes that more recent trends emphasize two-way communication, more openness to the voices and expertise of visitors and users (Hooper-Greenhill, 2000; see also Jörgensen, 2004; Simon, 2009). In addition, other tools and settings that also support such trends (such as information infrastructures, digitization, and new kinds of platforms for Web-based interactions) are now being integrated into museums. The material culture of institutions is therefore changing in response to the use of digital images, which have a particular physicality and, like printed photographs, require an adapted environment for preservation, manipulation, and display (think of servers, scanners, screens, and lighting conditions). Our analysis therefore considers change and continuity in the use of images for knowledge production in and about museum collections. Modes of visual mediation are clearly influenced by material culture and historical trajectories in museums. In the case of the Tropenmuseum, the focus on the visual is deeply ingrained in the organization’s digital archiving practices. The museum divides these practices into three levels: the first level is “basic registration,” which is followed by “registration” and “documentation” (Beumer, 2009, p. 9). The production of “digital images of physical objects” is a crucial element of basic registration. Earlier, analogue ways of documenting the collection used “paper documentation” in the form of various kinds of inventory cards, sometimes accompanied by an explanatory drawing (p. 9). The Tropenmuseum also has a long institutional tradition of photographic documentation and has always had 667 de rijcke & beaulieu/image as interface a large number of analogue photographs of its physical objects. Ever since the museum started working with the Web-based collection database, The Museum System, both the paper documentation and the analogue photographs have been digitized with all of the information combined in digital media. The photographic collection has always been documented on inventory cards. These so-called UDC cards included copies of the historical photographs further annotated with a description of the scenes plus additional data relating to their origin (Beumer, 2009, pp. 32–37). One of the central aims of digital archiving in the museum is to make the collection more manageable, more accessible, and less prone to deterioration. The last is based on the hope that “the objects and [printed] photographs themselves [will] no longer function as an ‘information system’” (p. 38). Digital archiving is therefore motivated by belief in the substitutability of digital images for physical objects as well as by faith in the information management gains to be acquired through digitization. To varying degrees the Tropenmuseum uses conventions of a certain type of photorealism in the twentieth century about the mechanical objectivity of optical photography (Daston & Galison, 2007). With the shift to digital photography in a database setting, the hopes of replacing interaction with objects and printed photos by interactions with digital images became prominent. While there is a change in media, such a shift need not mean a break. Interestingly, Bolter and Grusin (2000) stress the dynamics of remediation between various media without positing breaks or revolutions created by different imaging techniques. This approach is sensitive to the specific ways in which digital imaging relates to optical photography and print-on-paper documentation. This points to the need to analyze the use of images in museums in terms of media history, though it is important not to forget that such remediation itself changes the context and meaning of images. Our analysis of changing visual forms in museums and of images as interfaces embraces the visual culture in museums and digitization as part of the institutional agenda, along with an awareness of mediation that includes relations to users as well as attention to materiality or technology. In addition, we pay particular attention to the interactive functions built into visual material in databases as well as to the networked setting in which these images circulate. In today’s museums, libraries, and archives, existing practices for the production, handling, and dissemination of images of objects are increasingly blending with new, networked technologies for visual knowledge production. Interaction and manipulation in a networked setting are integral to these practices, and they emphasize intervening rather than observing (Hacking, 1983; Lynch, 1991). The screenshots (figs. 1–5) exemplify how the image becomes an interface that invites interaction. The collections can be searched. For each item, there is a photograph and catalogue information—information that resembles 668 library trends/winter 2011 what was previously inscribed on catalogue cards. While the description is static, the image has built-in functionality. Users are invited to interact with it, either with the image itself in visual terms (zoom, crop, move) or with the image as a digital file in a networked setting (print it, e-mail it, preserve it). They can also make it part of their own selection and create their own space in the database. It is also easy to take it out entirely and have it travel to other settings and to other media—including this publication. These possibilities are important for the way in which knowledge can be created. Furthermore, they constitute an understudied form of visual knowing. In their study of representational practices in scientific atlases, Daston and Galison (2007) identify intervention as an emerging mode of representation but only in relation to individual images. As we have shown, however, understanding these practices is not solely a question of looking at individual jpeg files, nor of narrowly tracing a shift from photographic to digital aesthetics. The databasing and networking of these images and the role that such infrastructures play within particular institutions are key elements in this new way of knowing. We now turn to the specific way in which images as interfaces are embedded in practices in museum settings. In order to see both the difficulties and potential of such uses of images, we focus on skills that are needed to engage in these practices. This focus brings to the fore what Figure 1. Screenshot of Tropenmuseum online collection database (http://collectie .tropenmuseum.nl/nindex.asp?lang=en) Figure 2. Results of an advanced search, using the search term “vodou”. Figure 3. One of the items from the results page of the “vodou” search, displaying several possibilities for interaction with the image. Figure 4. The same image as in figure 3, after clicking on the zoom button. Figure 5. The same image, after using the e-mail functionality displayed in figure 3. The image is accompanied by a selection of the annotations and descriptions on the object in the Tropenmuseum database. 671 de rijcke & beaulieu/image as interface people need to learn as well as what people can learn when engaging with databases of images in a multilayered, networked context. While skills are often deployed in combination, we discuss them separately here for analytic purposes. We first analyze the skills needed for seeing and interacting with networked images. This is followed by a detailed scrutiny of the skills needed for producing visual knowledge and for interacting with multiple platforms that support visual material. Finally, we examine how people develop skills that enable them not only to understand images but also to distinguish between various sources of visual information. We specifically consider a range of users inside and beyond the museum who use images as interfaces. This inclusive approach will enable us to consider practices in relation to images as interfaces without designating them a priori as inside or outside the museum or as involved in the production or use of knowledge. This opens up the possibility that these very distinctions are themselves being reconfigured. Skills for Interacting with Images as Interfaces Visual material has always played an important role in archival, library, and museum documentation practices. As is the case with everyday seeing, which is developed and trained by interaction in the world around us (Hacking, 1983), the skills used in interaction with visual material in institutions are also not simply there as givens. They need to be acquired and mastered. Importantly, existing practices and expertise help reshape the new skills needed for visual knowing and for interaction with digital images in a networked setting (cf. Hand, 2008). New interfaces mold and extend existing viewing habits (cf. Alaç, 2008; Daston, 2008). By focusing on changing skills, we are able to show that the transformations we describe are not simply a question of databases providing information effectively through digital media—as the modernization tale of computerization would have it. Rather, we are witnessing changes in how people interact with information, in the evaluation of what constitutes information, and, ultimately, in the production of knowledge. The skills needed to create and interact with these new interfaces help make meaning as a result of distributed actions between users and images. These actions are enabled—and perhaps sometimes also constrained—by the specificities and possibilities of a networked interface. A fundamental characteristic of networked practices of seeing is that the images are aligned on-screen with other digital material (Rubinstein & Sluis, 2008). Therefore, viewing skills alter not only when we move from analogue to digital imaging, but also as a result of this “windowed” and networked viewing (Friedberg, 2006). In addition, the specificities of working with/behind computer screens should also be taken into account (Alaç, 2008) as well as the ways in which different interfaces support different kinds of interaction with the visual material. 672 library trends/winter 2011 Actively Seeing and Interacting with Visual Sources As part of our fieldwork at the Tropenmuseum, we interviewed most of the curators on staff, each with his or her own area of expertise. In discussing the role of images in their daily work routines, we identified the skills they needed to interact with the databased material. In the words of one curator: For the primary task of documenting and validating the collection, I absolutely need TMS. . . . In the past, the photographs of objects in TMS were not always of a very high resolution which hampered the use of the zoom function and this caused problems for some images, for instance when an entire sword is photographed and the photographer needed to step back to capture the object in its entirety. . . . In the past, the focus lay more on quantity instead of quality when it came to photographing the collection. In practice it turns out that you definitely need quality, otherwise you cannot properly examine the objects. The idea is that TMS facilitates scrutiny of the entire object and that it replaces a visit to the depot. In many cases the system indeed suffices, but if a marionette, for example, is only photographed from the front, it does not work for me because I need to see the side as well, to see the ornaments in the crown, because that gives me a clue as to which character I am dealing with. In order to know this object visually, interaction with the zooming possibilities of digital photographs in the database is essential. The curator knows the interface and how to explore the object by changing the resolution on the screen. In some instances these skills in interacting with digital materials prevail over the skills needed to work with the material objects themselves. The interfaces change the setting, tools, and objects with which the curators make knowledge. Yet it is important to consider that these skills are not limited to an individual’s know-how. The episode above points to the ways in which the encounter with the digital image is only part of the network needed for skilled vision to work. Indeed, the potential of digital photography is not enough. The curator’s ability to see properly, to see enough of the object, and to apprehend it in sufficiently detailed views, depends on the particular instantiation of digital technology that was implemented in the institution. A “focus on quantity,” which was the result of institutional priorities, affects the possibilities for looking at and knowing the digital image. Institutional decisions on how to pursue digitization affect how the user is able to see and learn from an image. In many cases, the Tropenmuseum curators worked at other ethnographic museums before coming to Amsterdam. When the curator mentioned above started working at the museum, one of his tasks was to develop a new museum section for his area of expertise. Institutional responsibilities, infrastructures, and particular areas of expertise all shape interactions with digital images: 673 de rijcke & beaulieu/image as interface Curator: “I worked at another ethnographic museum for 14 years and knew my sub-collection by heart which partly had to do with the fact that this was a collection of ‘merely’ 17,000 objects. In the Tropenmuseum, my sub-collection has three times this amount of objects. . . . In the beginning, I had difficulties finding out what exactly was in the collection and what I could use [for the development of the new section for which he is responsible].” Ethnographer: “And that had to do with the amount of objects?” C: “Yes, but it was also related to TMS. . . . Right now, I cannot tell the difference anymore, but back then I felt that there was a difference in search terms. I couldn’t use the terms I was used to in my former workplace, I really needed to make a shift. For the new display I also did not want to use the most famous objects. But I did not have enough time to pull this off so I only partly succeeded.” The curator believes that this partial success not only had to do with the contextual use of keywords but also with the intricacies of the process of changing from analogue archiving to working with a digital image database: C: “A number of objects are not yet photographed or were not photographed at the time. So from time to time I now see things and think: ‘Oh, this would have been something I could also have used.’ But this simply has to do with the fact that we’ve been working through the backlog [inhaalslag] these past four years.” This exchange reveals different ways of interacting with the images in the database. The predominant mode is through keywords attributed to the various objects. Such an approach is almost too banal to mention being so fundamental to the indexing and information retrieval systems that have been central in museums for the past century. This mode of interacting with information, and therefore with objects, is perhaps dominant in digital databases that are efficient at manipulating information in this specific way. Note, however, how a different interaction with the database leads to different knowledge about the collection. “From time to time, I now see things . . .” points to a browsing behavior that leads to discovery, where one first sees something and then knows it. This contrasts with already knowing a relevant category, name, or keyword, and then calling up the image of the object to look at it. This example illustrates two important points about the skills deployed in the use of visual material in databases at the museum. First, effective use of digital information sources require a specific sensibility to the particularities of databases of collections. Their size, the quality of images, and the way in which digitization was implemented are all elements that shape how users can interact with the visual material and must be taken into account: users must learn to see in context (Alaç, 2008; Goodwin, 1995). While much of this contextual knowledge may remain implicit in day-to-day activities, our fieldwork enabled us to make clear that when us674 library trends/winter 2011 ers know about the mediation of images, they are better able to see with them. The second important element illustrated by this interview is the way in which interfaces shape what can be known. Searching on keywords will call up certain images for further consideration, but this strategy relies on a priori knowledge of relevant keywords. An interface that supports visual browsing would enable “seeing” to precede or to stimulate formalized knowledge of labels and categories. Skills for Producing Visual Knowledge and Interacting with Platforms In this section, we turn to the ways in which particular platforms that support visual material have come to be used at, with, and for the Tropenmuseum. The cases discussed here enable us to address the changing skills of individuals and of institutions, the former through visual “user-generated content” (Cox, 2008; Petersen, 2009; Van House, 2002) and the latter through the interaction of the museum‘s collection of images with other platforms. The Tropenmuseum recently became a partner of the Wikimedia Foundation, the organization behind Wikipedia and the Wikimedia Commons. This cooperation developed in the context of a project called Wiki loves art/NL (WLANL). The initiative sought to stimulate amateur photography in museums with the goal of getting more photographs of cultural heritage on Wikipedia pages under a Creative Commons license. In June 2009 a group of forty-six museums in the Netherlands opened their doors to the public for special sessions that allowed participants to make photographs of designated objects from their collections. Participants uploaded their images on Flickr, which thus served as a conduit for the photographic material. A jury, consisting of the organizers and a number of museum employees, decided which photos would subsequently be used on the Wikipedia pages and who would receive an award for “best photo.” In a blog post on WLANL, U.S.-based museum exhibit designer Nina Simon noted that participating museums were especially interested in making their content digitally accessible without breaking any copyright laws, while the Wikimedia foundation was primarily involved to obtain useful data (Simon, 2010). Many photographers were more concerned with “freely making pictures for their own use (or their portfolio)” and “quite a few came to do their own thing and they had ample opportunity to do so” (de Lusenet, 2010), as one of the Dutch participants pointed out in reaction to Simon’s post. Clearly, multiple interests and motivations were served by this event. What is relevant for our purpose is that the circulation of images via these platforms makes possible multiple uses and appropriations without causing them. Flickr serves as a pipeline from amateur photographers to Wikipedia, while institutional actors (from Wikimedia to the Tropenmuseum) 675 de rijcke & beaulieu/image as interface maintained a gatekeeper function. Not only do we see a shift toward the digital in the material structures that support storage and display of photographs, but in this case both personal and institutional visual resources take the shape of networked databases. There are of course differences in the way various databases (TMS versus Flickr) are set up, and in their different possibilities for interaction, but nevertheless we see an alignment of the way visitors and institutions organize their visual knowledge about the museum. Furthermore, the intersection of the multiple agendas of museums and of visitors via Flickr and Wikipedia points to new ways of negotiating what it means for a digital image of a museum object to be or to become public. A photograph in this initiative was treated as a creation to share with other viewers; an opportunity to document the collection; and the production of copyright-free data. The WLAN activity reconfigures the public/private dynamics of visual knowledge in interesting ways: the museum opens its doors for a private session for amateur photographers; amateur photographers make their personal snapshots public; and there are complex shifts in ownership, copyright, and right to publicize, as the images are produced, uploaded, selected, and further circulated. The ways of working of different groups become aligned in this project. The skills of amateur photographers for producing visual knowledge about the collection is linked to the aspirations of the museum and of Wikimedia for greater production of copyright-free images, while the photographer’s work is arguably enhanced through the visibility it gains in the course of this interaction. Different parties use each other to leverage a greater impact of their own skills. A second example of the different kinds of interaction with platforms that support visual material also involves Flickr, but was initiated by visitors to the Tropenmuseum. As part of our fieldwork on the practices involved in the Tropenmuseum networked image database, we interviewed the moderator of the Tropenmuseum group on Flickr. Also taking part in the interview was his girlfriend, an enthusiastic amateur photographer herself. Both the moderator and his girlfriend carry their cameras with them whenever they can. This was also the case when they went on their first visit to the Tropenmuseum about two years ago. The couple continues to go back to the museum on a regular basis. Both are enthusiastic users of photo-sharing opportunities on the Web. After their first visit to the museum, the moderator wanted to upload to Flickr the photos he had taken. He explained that this was not in an attempt to advertise the Tropenmuseum, but simply because of his interest in photography, though he also partly ascribes this to his professional background as a teacher. He likes to inform and educate people. In this case he does so by writing annotations to the photos, for instance on particular exhibitions. At that point, he noticed that no Flickr Tropenmuseum group existed, although 676 library trends/winter 2011 there were other users who uploaded photos of the Tropenmuseum to Flickr. He decided to create a group called “Tropenmuseum,” and deliberately opted for an open structure by not creating strict rules for joining the group. As to the skills needed to produce visual knowledge and interact with platforms like Flickr, the moderator remarked on the policy of the use of tripods not being allowed in the museum. This means that visitors are not likely to use long exposure times without a tripod, because they need to resort to other photographic means to capture the often stark differences between dark backgrounds and the beautifully lit objects on display. This is one way in which the Tropenmuseum itself shapes the kinds of photographs (and we would argue, photographers) who create visual material about the museum. Interestingly, the moderator also referred to another, nonphotographic skill that he developed while working with images as interfaces. By uploading his images to Flickr and by serving as the moderator of the Tropenmuseum group, he was stimulated to expand his network beyond national borders, even though he was not at first fluent in English. Because part of this network is now made up of international contacts, his use of English has greatly improved. After a while, he told us, he even started to think bilingually. In the case of the Tropenmuseum, the interfaces offered by the networked image database, but also by platforms such as Flickr, can be seen as creating sites of new literacies and creativities (Burgess, 2009; Petersen, 2009) that redefine the role of cultural producers (Bruns, 2008; Jenkins, 2006). As we have seen, Flickr provides a space for interested museum visitors to share their photos, to add reviews of the exhibitions they went to, and to provide descriptions of parts of the collection on display that caught their eye. The Flickr group serves as an open, accessible podium for interaction and exchange, while the photographs act as interfaces for an international group of interested users, brought together by one enthusiastic moderator. These and other participatory new media phenomena, although diverse, play an important role in shaping various aspects of knowledge production. In the case of Flickr, the conjunction of different photostreams and, importantly, metadata, annotations, and descriptions may in the long run lead to a renegotiation of epistemic authority in relation to different types of visual material. The co-existence and closeness of various understandings of images on the Web make it all the more important to analyze what enables us to determine how images can come to be trusted and to be useful and how to generate instances of such trusted