ArtMaps: A Technology for Looking at Tate’s Collection

ABSTRACT This article presents ArtMaps, a crowdsourcing web-based app for desktop and mobile use that allows users to locate, move and annotate artworks in the Tate collection in relation to one or more sets of locations. Here the authors show that ArtMaps extends the “space” of the museum and facilitates a new pluriperspectival way of looking at art.

locality. If users disagree with the position of a work on the map, they can move the pin to a different location and leave a comment (Fig. 1).

ARTMAPS: RESEARCH AND DEVELOPMENT
ArtMaps was conceived in response to the need to visualize Tate's collection on a map and facilitate geotagging-related activities. In 2012, at the time of the redevelopment of Tate's website, Tate had 69,775 artworks in its collection, and over 23,000 of these had been indexed with one or more places, including mythical places. Since a number of works could be indexed with the same subject index term, Tate began by mapping them so that all the works indexed with a given term would be located at once. Before this first phase, Tate had geocoded the bottom terms on the place-names index against a number of available databases to return latitude and longitude information (e.g. GeoNames and Google Places). While this had captured a large number of the relationships between artworks and locations, the location information had often been limited (e.g. "India, " "London"), and many works could not be located in this way. A number of index terms were subsequently mapped by hand, though a large number of works still needed to be geocoded. This prompted Tate to suggest the remaining terms should be crowdsourced. The resulting dataset from this phase was the foundation for the ArtMaps project (2012)(2013)(2014)).
An early experiment drew attention to some challenges, including the facts that an artwork could show a number of landmarks; that there may not be an explicit location; that the location may not be a direct representation of a place; that the perspective from which a work is represented may be more significant than the place represented; and that a work may be associated with a plurality of locations. At the time it was felt that capturing users' reflections about these realizations could provide Tate with insight into how users experience and reflect about the encounter with the collection. This realization shifted the initial ambition and led to the development of ArtMaps as a crowdsourcing and a usergenerated documentation tool. An additional functionality This article presents ArtMaps, a crowdsourcing web-based app for desktop and mobile use that allows users to locate, move and annotate artworks in the Tate collection in relation to one or more sets of locations. Here the authors show that ArtMaps extends the "space" of the museum and facilitates a new pluriperspectival way of looking at art. was then added, making it possible for users to add text freely as well as in response to specifi c tasks in order to justify why they would move the position of a work from one location to another.
A number of scaff olded public engagement events were held at Tate Britain, which included the 2012 and 2014 testing of ArtMaps at Patrick Keiller's Th e Robinson Institute (2012), Looking at the View (2013) and Wandering Ruins (2014) exhibitions. Th e testing involved art and technology enthusiasts (2012 and 2014) and families (2012). Feedback showed that, through the setting of tasks, ArtMaps could operate as a "performance" event. Th us participants felt that they could "perform" "ArtMapping" and become "ArtMappers. " Th ey talked about rehearsing, preparing and sharing in a way that is reminiscent of a set of performance rituals. In using the term "ArtMappers, " they indicated that they identifi ed themselves as being part of a group. Participants felt protected by the event framing and used this to their advantage in their social encounters. Moving between mapping, mapmaking, performing and documenting, they used the performance frame to justify the production of their "social" actions in this context. Th is showed the potential for ArtMaps as a tool for group learning, as participants noted that the platform allowed for a "soft " learning approach. As one participant said, "It's not an instruction, it's an association. " Feedback also showed that ArtMaps could operate as an orientation or navigation tool. Participants enjoyed searching terms, fi nding connections, generating trails and exploring those routes and connections that had been generated by others. More specifi cally, they found it interesting that they could compare a specifi c location to what they could see in the artwork (Fig. 2). Th is prompted refl ections on the similarities and diff erences between places and artworks, the people and objects therein, and the positions adopted by the participants themselves in relation to the one taken by the painter of the work. One participant noted: "I found it interesting because the building is missing. So you are in the right place but you are in the wrong place. " Th is showed that ArtMaps could facilitate a "way of looking": as users were asked to fi nd relations between maps, artworks and sites, participants felt encouraged to refl ect on the role of their own points of view in relation to that of the artist or Tate.
Additionally, users pointed out that ArtMaps can lead to an enhanced knowledge of local geography as well as of the artworks encountered. Th ese users indicated that ArtMaps gave them unexpected insights into familiar neighborhoods. One user said: "It's phenomenal cos I'm genuinely surprised there are even two paintings related to this area. " Th is suggests that tools like ArtMaps can be used to widen participation and generate and share knowledge about heritage (Fig.  3). Another participant noted: I thought it would be much more engrossing for the children to be . . . gazing at the screen all the time, but actually it took you outside of that digital world and to the world around you and we talked about things we hadn't talked about before.
Rather than acting as a substitute for the museum experience, ArtMaps seemed to complement it, in the sense that participants felt that encountering new artworks through ArtMaps prompted them to return to the museum to have a better look at the original works (Fig. 4). One participant stated: "Never was I interested in Turner until today. Now I want to study Turner more." ArtMaps also acted as a  Tate) sharing tool, setting up the museum as a social space. Th us two participants mentioned that in ArtMaps, "Everyone's learning together. " Interestingly, participants made use of a whole variety of social media to further disseminate their ArtMaps, and one participant posted her work to Facebook noting that, through social media, ArtMaps takes art to people that would not necessarily access it. It broadens people's horizons. I have got friends on my Facebook that have probably never been to an art gallery so by doing projects like this you are actually taking art to them.
Feedback at subsequent workshops showed that families worked collaboratively and enjoyed responding to art creatively outside the museum (Fig. 5).
An unscaff olded gallery observation led by Cristina Locatelli indicated that, among visitors to Tate Britain during the period analyzed, the percentage of younger visitors that were attracted by the ArtMaps kiosks was greater than the percentage of over-60 visitors attracted to the kiosks, hinting at the diff erent interests and levels of technological confi dence of these two audiences and suggesting that ArtMaps has the potential to reach younger audiences. Individual feedback indicated that Art-Maps could facilitate the recollection of personal memories of places, potentially fostering a stronger sense of belonging to a place. Th is in turn could prompt the sharing of personal stories, as in the case of this participant who wrote on ArtMaps: Th is is "Postmans [sic] Park"-the site of the original monument. From an idea by the artist GF Watts to celebrate the heroism of ordinary people you can see lots of plaques to commemorate heroic deeds, and a small model of the man himself. Th ere is also a play by "Lone Twin" TC inspired by this [sic] monument. I was a postman at the once ECDO offi ce which used to be nearby. Be your own hero. marka. 2013.
Th e educational potential of the platform was tested at the Pimlico Academy in London and online three times by Laura Carletti between 2013 and 2014. Th e latter was part of the Telling Tales of Engagement Prize. Th e aim of the former was to explore the use of ArtMaps for wayfaring and mapping in a mobile learning context facilitating the generation of diff erent roles [2]. Th e aim of the latter was to target the digital public and to investigate the crowdsourcing potential of the platform. Th e 21 participants who completed all the tasks and the fi nal questionnaire "agreed" or "strongly agreed" that using ArtMaps facilitated new ways of engaging with locations and artworks (85 percent); increased their understanding of a particular artist's The final public engagement event in March 2014 built on findings from the first event and findings from a workshop held by Cristina Locatelli with the Migrant Resource Centre (Fig. 6), reexamining how people can explore places through art by creating mobile journeys for others to experience.
This event used another WordPress-based platform developed at Horizon, Wander Anywhere, and structured the event as a collaborative design for user experiences on location. The starting point was the Ruins Lust exhibition at Tate Britain, which examined the fascination with ruins of the past, present and future as a subject for art. Participants were divided into interest groups around three London-based artworks selected from the show: Keith Coventry's Heygate Estate (1995), Rachel Whiteread's Clapton Park Estate, E5 from the series Demolished (1996) and Joseph Gandy's An imagined view of the Bank of England in ruins (1830). The event confirmed findings from previous workshops and provided further evidence that people's understanding of the locations where they devised their trails, as well as those they explored on others' trails, were subsequently seen in a new light. Participants, who had been prompted to reconsider the overlooked within the urban landscape in light of its (often hidden) history, were then motivated to revisit the work of artists who were the starting point for those participants' ideas. Feedback indicated that participants felt they were involved in a social activity and employed various forms of social media to plan, share, test, prompt engagement and disseminate experiences throughout and beyond the event.
We have seen how the research and development for Art-Maps showed that the tool can facilitate access for those who do not habitually visit museums, extend the gallery experience outside the museum, allow for encounters with items not ordinarily on display, stimulate collaboration and group discussion, facilitate mobile learning, generate dialogue and interaction outside of the interface and, through crowdsourcing, potentially produce valuable and original knowledge for the museum. Here we suggest that ArtMaps also promotes a new pluriperspectival way of looking at art through its relationships with places and, vice versa, facilitates the perception of places through their relationship with art. In other words, ArtMaps encourages the production of relational knowledge through observation (i.e. the viewer is standing in front of a digital work that has been removed from its gallery context and looks at it), immersion (i.e. the viewer is immersed in a distinct but related physical place), and relational mapping (i.e. the viewer is mapping the two distinct fields, an artwork and a place, seeing one in relation to the other and bringing them together onto a cartographic map).

ARTMAPS: A NEW WAY OF LOOKING AT ART
Over the years, a number of museums have been inviting their audiences to contribute ideas, recollections and even personal artifacts to enhance and extend the experience of their collections [3,4] and to promote collaborative experiences [5]. With the introduction of ubiquitous computing, more complex participatory encounters have been produced, leading to novel ways of looking at and experiencing art. Apart from ArtMaps, other projects, such as the Imperial War Museum's Putting Art on the Map [6] and Museum of London's Streetmuseum [7], to name a few, facilitate novel ways of encountering collections, encouraging users to locate or place artworks on a map, and compare and contrast past and present views of them. This spatial form of engagement with art is crucial in terms of how we see ubiquitous computing facilitating a new way of looking at art.
We know from Ross Parry that, ever since the Renaissance, museums have been in some ways "knowledge made spatial, " involved in "spatial architectonics, " i.e. "the systematic construction of knowledge through space" [8]. We also know that because of this, museum visits have tended to be framed in space and time as something separate from everyday life and that, with the advent of computers, museums have been challenged with the task of reconsidering what they are as well as "when and how a museum visit could occur" [9]. Art-Maps, like Putting Art on the Map and Streetmuseum, is made possible by developments in ubiquitous computing and social media that are grounded in the promise of the delivery of a novel kind of space, a "mixed reality" [10] that merges physical and digital spaces. Not only do such platforms therefore facilitate visiting a collection from outside the museum, bringing the museum to the user, they also make possible the juxtaposition of the collection, experienced digitally, with physical environments from the user's everyday life, practiced physically, in space. Through such platforms, museum visits can occur in conjunction with other kinds of mobile experiences, linking physical and digital sites, hence extending the museological experience outside the conventional museum space. The space of the 21st-century museum, herewith, has become juxtaposed with the physical world, which means that the 21st-century museum can be experienced everywhere and at any time. This is not the first time that the discovery of a technology, in our case ubiquitous computing, has changed the way we look at art. We know from Jonathan Crary that during the first few decades of the 19th century a new form of observer emerged [11] who was formed by the introduction of a set of new methods for managing the work of large populations that "brought in new procedures of individualisation" [12] reliant on the accumulation of knowledge about people [13]. For Crary, the invention of a number of optical devices transformed "the status of the observing subject" [14], informing the course of visual culture in the 19th century. He suggests that "these apparatuses are the outcome of a complex remaking of the individual as observer into something calculable and regularizable and of human vision into something measurable and thus exchangeable" [15]. An even earlier shift is described by Svetlana Alpers as a move toward a calculable and regularizable visual culture, which, she believes, occurred as a consequence of the introduction of a new style of painting in the Netherlands in the 17th century. At this time, she notes, painters moved away from an imitative, Albertian model-in which a picture consisted of "a framed surface or pane situated at a certain distance from a viewer who looks through it at a second or substitute world" [16]-to a descriptive form of painting from which it is no longer possible to detect the viewpoint of the painter because there is seemingly no frame [17]. The introduction of a new technology played, once again, a crucial role in this context. Thus, for Alpers, one could, for example, observe "the effect of nature, seen in a camera obscura" in Jan van der Heyden's work and "a circular field of vision" in the case of Jacob van Ruisdael's panoramic landscapes [18]. The scene is more complex in Diego Velásquez's Las Meninas (1656), in which the painter famously represented himself looking out of the painting toward the position of the looker [19]. Still, from a mirror in the painting, we can see that the lookers captured in their act of onlooking are the royal couple, King Philip IV and Queen Mariana, though some have argued that the mirror in fact reflects what is in Velásquez's painting within the painting [20]. For Alpers this picture captures both the Albertian world model and the one introduced by 17th-century Dutch painting, representing, at the same time, "the world prior to us made visible" and "we prior to the world and commanding its presence, " so that we see, as an unreconciled image, "a replication of a world and a substitute world that we view through a window frame" [21].
Two centuries later, we are again witnessing the emergence of a new kind of observer who is consuming and generating information [22] and who is looking at art relationally [23] and, increasingly so, by overlaying physical and digital environments in what could be described as a "mixed reality" [24]. For this 21st-century observer, mobility is crucial, as are mobile interfaces' ability to facilitate "implacement" [25], which, in Jason Farman's interpretation, helps toward locating "our situated nature and our sense of proprioception with others and with objects in a space" [26], offering "context for embodied being-in-the-world" [27]. This new kind of observer experiences both the museum collection and the physical environment from which the observer views the collection by mapping. Through mobile interfaces the observer can now make maps (i.e. generate objects, as Putting Art on the Map, Streetmuseum and ArtMaps show) while mapping (i.e. generating knowledge by looking at the relationships among people, place and art).

MAPPING STRATEGY
As mapmaking and mapping are sometimes seen as synonymous, we need to pause to describe the subtle differences in the way that scholars have understood these terms and practices, which are significant for what occurs while looking at art through ArtMaps. Mapmaking means creating "systems of signs" [28] or "social constructions of reality" [29]; mapping means generating knowledge about the environment while we are "on the move" within it [30]. ArtMaps aims to facilitate the creation of user-generated documentations of what occurs while mapmaking (i.e. creating systems of signs) and mapping (i.e. generating knowledge while on the move through a bird's eye, cartographic view). These may confirm but also destabilize the cartographic, totalizing view of the entirety of Tate's collection visualized through a map, by instead presenting or invoking a plurality of perspectives about the relationships between artworks in Tate's collection and the places that are variously related to them. While it is indeed Tate's collection, or rather the digitized simulacra, that we see on the map, it is other places, rather than Tate itself, that are foregrounded, so that Tate becomes the frame of the ArtMapping experience. ArtMaps not only facilitates this unreconciled pluriperspectival experience, it asks viewers to map their positions within it.
We know from past research that the use of different, often even minimal, forms of role-play constitute an effective strategy for facilitating the design and subsequent experience of mixed-reality environments [31]. We also know that geographical processes of encounters with art, or what have been described as "geography-art intersections, " can come to form shared practices wherein knowledge is advanced in both fields [32]. Mapping and mapmaking also play a major role in this context, and we can learn more about why this may be so by looking again at what occurred in the 17th and 19th centuries. Seventeenth-century Dutch painters, as Alpers notes, developed a fascination with maps that appeared increasingly in paintings, as was the case with Johannes Vermeer's well-known Art of Painting (1666), which, according to Alpers, "illuminates the resemblance between pictures and maps" [33], defining "an impulse to record or describe the land in pictures" [34]. Crary, too, discusses the significance of maps in 19th-century art and cites the example of Jacopo de Barbari's View of Venice (1500), which he describes as a bird's-eye view or ichnography, a "pre-Copernican, synoptic and totalizing apprehension of the city as a unified entity" [35]. He then juxtaposes this view to Canaletto's Piazza San Marco, looking east from the northwest corner (c. 1755), which discloses, in his opinion, "a field occupied by a monadic observer, within a city that is knowable only as the accumulation of multiple and diverse points of view" [36]. Barbari's view of Venice is roughly comparable to the view offered by Google Maps that constitutes the entrance point into Art-Maps. They are both a bird's-eye view. They both promise a synoptic experience of a territory. The works we look at through this process, however, do not conform with the map or the place, Tate, in which they are located as a collection, nor do they necessarily represent a location. To look at an artwork through ArtMaps, the user therefore has to position him-or herself in relation to three points of reference-the Google Maps interface; the place in which the artworks are pinned on ArtMaps (the place that may be represented in the work or with which the work may be associated); and the artwork itself-which, like Canaletto's view of Venice, may not be an accurate reproduction of a location, but rather, as we have seen also in some of the feedback offered by ArtMaps users, the product of an (unsynthetized) composition entailing different points of view. To look at art through places means to embrace a variety of viewpoints from which to observe it-the processes involved here are both displacement and, to use Casey's term, "implacement. " To achieve this pluriperspectival way of looking at art, users may need to adopt different roles and strategies to augment the sense of their presence in both the physical and digital space. While enhancing their being in a set of places, users create environmental relationships between themselves and what is around them [37], because integral to an understanding of what constitutes presence is the relationship between a subject and what is before or in front of him or her. Presence is always contextual to the environment a subject finds himor herself in and, as no environment is ever complete, since environments are forged by living things [38], environments are best described as processes rather than as fixed locations, which means that one is only present insofar as one is part of and in relation to a set of environmental processes. Through ArtMaps, not only do users position themselves in relation to a physical environment, they position themselves in relation to an artwork that is in itself absent from that environment in space and time. According to John Berger, not only does every image embody "a way of seeing" [39], but also when we "see" a landscape, we locate ourselves within it so that if we "saw" the art of the past, "we would situate ourselves in history" [40]. The "presencing processes" that take "place" while people are using ArtMaps thus allows those users not only to re-perform these different ways of seeing an environment or constructing their space, but also to connect views generated of this environment by other individuals, primarily artists, at different points in time. ArtMappers therefore perform the presumed act of viewing taken on by the artist. This in turn allows them to map the memory of that act onto the "present" landscape.

CONCLUSION
According to Crary, each technology produces a different kind of observer: The lack of optical unity in the planar fields of vision generated by the stereoscope produces a fragmented observer [41]. The photography, like the camera obscura, creates an assumed unity of the viewer [42]. Ubiquitous computing produces a pervasive viewer that is capable of persistently overlaying different (physical and digital, geographical and cultural) views to map a terrain through art, and vice versa.
Nicholas Serota draws attention to the fact that since the 1920s, "artists have created environments which establish independent space to be encountered by the viewer, rather than simply modifying the space which is given by the museum" [43]. In this context, Serota mentions the "comparative" strategy introduced by Rudi Fuchs for documenta 7 in Kassel in 1982 and subsequently at Castello di Rivoli in 1984. According to this curatorial view, different works by the same artist are dispersed and exhibited together with works of other artists so that "works by Baselitz, for example, might be seen initially alongside Beuys, then alongside Chamberlain" [44]. ArtMaps extends this curatorial approach outside the museum.
The Internet has always promised "a museum without walls, with no sense of place" [45]: ArtMaps reverses this position and presents a museum without walls that does allow users to orient themselves in the physical world through art and in the digital world through associated digital footprints [46]. As users annotate Tate's collection, relating art to places and places to art, they write the art that is at Tate within the territory they pass through in their everyday lives.