Beyond the PDF: Expanding Art History Digitally with British Art Studies

British Art Studies, a born-digital, open-access journal, co-published by the Paul Mellon Centre for Studies in British Art (PMC) and the Yale Center for British Art (YCBA), launched in November 2015. This article documents how and why British Art Studies was established and the ethos that has motivated the editorial team to always think “beyond the PDF” in creating a new digital publication for the field of art history. Drawing on case studies of articles and features published in the first seven issues of British Art Studies, the purpose of this article is to reflect on the creation and development of the journal alongside the broader critical and methodological questions that have guided and shaped our approach to publishing. Co-written by the current editorial and production team, this article explores how establishing and maintaining British Art Studies has enabled us and our authors to join the growing conversation about the intersection and cross-fertilization of art-historical research with digital tools and technologies.


Introduction
The peer-reviewed article in a specialist journal is a cornerstone of academia and governed by well-tested and respected methods. Conventions that determine the format and tone, apparatus such as referencing, and the process of peer review have become enshrined and play an explicit role in how scholars are evaluated and promoted within academic institutions. The journal article is often seen as a testing ground for an argument before it is applied to a longer format, typically the monograph, an equally well-established cornerstone of academia. These protocols, methods, and standards, which were developed with and for the printed format, have become almost sacrosanct.
The digital age is disrupting and challenging these tried-and-tested methods of scholarship and publication. As scholars increasingly take advantage of digital tools and resources for conducting and disseminating research, new digital forms of publishing are pushing the boundaries of traditionally conceptualized articles, journals, and books. 1 It was at this moment of creative disruption, generated by the digital humanities, that British Art Studies was conceived and launched. The potential of the digital platform to rethink how we develop, create, and publish art-historical research has motivated us from the very start to think beyond translating the existing conventions of print publishing into the electronic format. Two important questions have guided, and continue to guide, our strategy: first and foremost, what can we do digitally that we cannot do in print? And second, what digital tools do we need to support the research processes and outcomes of scholars? These may appear to be obvious questions, but they help us avoid the risk of forcing scholarly research into pre-existing digital formats and standardized templates. In other words, our goal is to find digital tools that can support research, rather than the other way around. These questions provoke not only a myriad of answers, but also open up unexpected possibilities and challenges, depending on the nature and scope of individual research projects. By asking ourselves and our authors these questions at the outset of each project, we have sought to move away from the assumption that publishing digitally is merely a method of dissemination.

Establishing British Art Studies
British Art Studies was born digital. Its look and aesthetic, as well as its usability, have been of central importance as we worked with our developers and designers, the London-based company Keepthinking, to develop a journal that does not simply replicate the formats and aesthetics of a print journal but uses the digital platform to create an environment in which text, sound, and still and moving images can be brought together to facilitate and present art-historical research in new ways. This does not simply have an effect on the "end product"how an article looks or is presented; it also impacts the processes by which the research is carried out, as well as the internal work of the editorial team. It requires us to ask: What tools do we need to employ? What other models exist? Who else do we need to work with?
At this creative point of disruption, the working practices of publishers and authors alike are being transformed. As editors, we have increasingly moved away from standard journal publication models whereby authors submit polished texts that are selected through a process of peer review and editing prior to publication. While we still publish some material in this way, we are increasingly also calling proactively for project proposals that start with a set of ideas for how to utilize the British Art Studies platform in innovative ways. This necessarily means that the editorial team becomes involved at a much earlier stage, working with authors not only as reviewers and publishers, but also as collaborators and creative producers. By workshopping projects with authors and peer reviewing not just finished articles but ideas and concepts, British Art Studies puts the emphasis on the process of getting thereand on exposing this as part of the research process. This is an approach shared by other open-access publications, particularly those with a commitment to publishing multimedia scholarship, as the editors of Vectors presciently wrote in 2011: The affordances of digital media for processfor the understanding born of doingare tremendous. The power of process resides less in the end products themselves than in the very act of making. We will need more finely grained accounts of the processes involved in the production of multimedia scholarship in order to evaluate properly the labor required in such research. 2 At British Art Studies we are experimenting with ways to express and expose these "finely grained accounts" of the processes of scholarship, which traditionally have been suppressed and hidden from view in the end-product of the published article; the case studies presented below illustrate this. The results open a new potential of openness and transparency, where the visibility of approach and method purposefully (and, we hope, productively) reveals the flaws, failures, and problems of research, as well as its successes and certainties. Through this approach we embrace speculation, the possibilities of interpretation, and the need to revise and revisit arguments after publication. With its tradition of negotiating diverse media and forms, art history as a discipline has much to gain and to offer in developing such a methodology. As has long been recognized, the transactions between image and word, and between object and viewer, upon which the field is built are already open to many permutations. Undoubtedly, digital publishing can play an important role in expressing and visualizing these relationships.

British Art Studies and "Digital Art History"
We do not see British Art Studies as a "digital art history" journal, but rather as a digital publishing platform for art history and art historians, artists, curators, conservators and others whose work intersects with visual arts. We see this outlook as a critical part of the mission and one that is also critical for art history, a field that has been slow to participate in digital publishing: we want to encourage not only those who are already versed in digital tools, but also those who are new to the digital realm, by providing the tools, resources, and support with which authors can navigate new terrain. We want to bring digital publishing into the fold of art history, rather than hiving off the journal into an even more specialist subdomain, although we recognize, utilize, and champion the field-changing work being done under the rubrics of digital humanities and digital art history. Thus, our ambition is that this journal is of interest and relevance beyond the confines of "digital art history" or even "British art." Inspired by the work of art historians such as Pamela Fletcher and Anne Helmreich, who have used digital tools to ask new questions about their topics of researchin their case the art market in nineteenth-century Londonwe seek to demonstrate the purpose and possibilities of this kind of work to the field of art history beyond a community that would recognize themselves as digital humanities experts. As Pamela Fletcher has observed: "Digital art history is notor does not have to bea separate track within the discipline, available only to the technologically inclined or gifted." 3 In this vein, we do not see British Art Studies as progressing on a separate track, but as part of ongoing conversations and experiments in what it means to practice and publish art history digitally. Each issue of the journal, every article and collaboration, tests different possibilities and has the potential to make digital methods more accessible and inclusive.
The expression "digital art history" is increasingly a topic of debate. Recently, art historian Claire Bishop, in her article "Against Digital Art History," polemically argues that "digital art history has a fraught relationship to history and interpretation." She asks: Does the data set exist in history before being sequenced digitally or is it only actualized once it has been laid out via the digital archive? Are the assembled historical 'facts' found or produced? What's the relation between what's empirically observable and what's true?
Her issue is that technology "is presumed to provide objective access to reality in a way that subjective interpretation cannot." 4 Bishop is wary of what she perceives as a hierarchy inbuilt within digital art history whereby "empirical data" trumps "subjective analyses." "Theoretical problems," Bishop writes, "are steamrollered flat by the weight of data." This stark assessment provides a useful jolt; a thorn in the side that counters celebratory assessments of the seemingly limitless possibilities of digital methods for our field. It usefully highlights the attendant problems of not critically assessing digital tools and technologies, but it perhaps overstates the separation between the work being done by computer technologists to produce and analyze data, and art and cultural historians who supposedly provide the interpretation of it. Johanna Drucker, in her contribution to the 2013 Visual Resources issue dedicated to digital art history, convincingly argued for a distinction between digitized art history (a practice built on the use of online resources) and the use of analytic techniques enabled by computational technology. 5 In our experience, these challenges, problems, and questions are being raised and debated through the critical conversations of those working with digital resources and technologies within the field of art history. To take an example from issue 4 of British Art Studies: in their article, "The Temporal Dimensions of the London Art Auction, 1780-1835," Matthew Lincoln and Abram Fox used the Getty Provenance Index Sales Catalog Database, a database of painting auction records that date between 1780 and 1835, in order to tease out the temporal structures in the early flourishing of art auctions in London, and test their relationship to the schedule of the Royal Academy's exhibitions and the larger "Season" of London society. 6 At the forefront of their article is a section on "Data and Methodology" that clearly lays out the gaps in their data and the attendant issues of historical interpretation. "Like any archive," Fox and Lincoln write, "this database is not a perfect representation of all historical auctions as they actually occurred; we face that intractable historiographic problem in tandem with all historians." 7 We are finding that these problems are not necessarily being "steamrollered" by those working with datasets and computational technologies, but are rather being opened up, shared, and presented to the reader for critical assessment. The potential of working with "data" (in the myriad forms that it takes) is not just pertinent to digital art history or the digital humanities; it is the topic of crucial conversations more broadly in the digital age. In the rest of this article, we explore how British Art Studies is a platform for such conversations, and highlight some of the key issues, challenges, and possibilities that we have encountered in establishing it as such.

Open-Access Publishing
British Art Studies operates in the context of a small group of open-access digital journals for art history, and we have built on the pioneering work of journals such as Nineteenth-Century Art Worldwide, Tate Papers, Journal of Historians of Netherlandish Art, and Journal 18. It also comes out of an institutional publishing context: the Paul Mellon Centre and the Yale Center for British Art are well-established publishers of printed books and exhibition catalogues, in collaboration with Yale University Press. British Art Studies was therefore launched in the context of a well-respected foundation of scholarly publishing by two research centers and an academic press. However, the journal has allowed us to do something quite different to the print list: to develop and disseminate research in flexible formats and at no cost to the reader.
With the launch of British Art Studies, we effectively replaced a series of printed anthologies that we had published since 1995. Twenty-four volumes of the series Studies in British Art had been produced with Yale University Press, addressing a variety of fields and subjects, ranging from the Tudor period to the twentieth century. 8 The individual volumes often arose from scholarly initiatives developed by the two centers, and provided a means of capturing new research on focused topics by scholars at different stages of their careers. Studies in British Art captured some of the major developments and emphases in the field as they were shaped over some two decades. While the printed format was both appropriate and necessary to the series, the increasing availability of digital tools provided us with an opportunity to rethink our publishing modelsboth from the author's and the user's end: by reinvesting the funds used to support Studies in British Art, we were able to develop an open publishing model that was free to all readers with Internet access and that offered a more nimble and responsive approach to scholarship. Whereas the print volumes were necessarily shaped as coherent collections on focused topics, which encouraged authors to write essays of fairly uniform lengths and formats, the digital journal allows us to be more open and spontaneous and to accommodate different methods and modes of writing. Each issue can assemble a variety of research topics, thereby allowing us to react responsively to developments in scholarship as they arise.
Issue 3 of British Art Studies, published in summer 2016, was our first "special" issue, and represented a transitional essay collection that had initially been conceived as a printed anthology for Studies in British Art. The conception of the volume, titled British Sculpture Abroad: 1945 to Now, began at a workshop at the Yale Center for British Art, bringing a group of scholars together to consider the histories of postwar British sculpture outside of Britain. The structure of the book developed as a simple chronology: five sections, each dealing with roughly one decade, consisting of a contextual essay followed by a number of case studies. Transitioning this format to the digital journal was fairly straightforward, and the basic parameters remained the same. But if there was an apparent conceptual alignment between print and digital, the ability to publish online opened the issue to a wider range of content and different kinds of scholarship. The issue included picture essays, exhibition histories, film clips, and short statements, as well as more standard essays. In a print publication, it can be difficult to respond to changing content; word counts, numbers of illustrations, and deadlines have to be predetermined and maintained, and there can be a push and pull between the editorial process and print production, all of which also determine the cover price of the end product. When publication is brought in house, as it was with British Art Studies, and linked directly to the editorial process, the shifting of content and scope can be dealt with more fluidly. British Sculpture Abroad involved 30 contributors in 10 countries. Having the flexibility to accommodate longer and shorter essays, a variety of approaches, and the ability to substitute, add, or omit content where needed, meant that the volume could stay on track and proceed, rather than being hampered, slowed down, or derailed by changes. In other words, rather than having to shoehorn the content into the publication format, the digital platform was adapted to suit the nature of the content.
British Art Studies extends and applies an open access policy supported institutionally by the Paul Mellon Centre and the Yale Center for British Art. In the realm of art history, restricted access to images is one of the fundamental barriers to research; our institutions have supported policies aimed at bringing those barriers down. In 2011, Yale University announced an open access policy granting license-free and royalty-free access to images of hundreds of thousands of works of art in the Yale collections. When the Yale Center for British Art released its collections online, images of works in the public domain were made available on a free and open access basis: users can download high-resolution files suitable for publishing directly from the collections website, without having to ask for files or permission. The images can be used for any purpose.
Following a similar ethos, British Art Studies authors publish their work in the journal under the Creative Commons CC-BY-NC 4.0 license, which allows authors or any other users to re-publish and adapt the materials for non-commercial purposes, so long as proper attribution is made. During a time when universities in the UK, US, and elsewhere are increasingly concerned with making publicly funded research freely available, this has the practical benefit of allowing any scholars who publish with us to deposit their articles in institutional or subject-based open access repositories. 9 At the same time, it supports a way of thinking more commonly found in the creative industries, associated with practices such as remixing and sharing. It also facilitates the transmission of research to broader audiences.
Open access is an ethos which, as we have come to discover, does not simply result in the "end product." In seeking models for best practice, we looked to arbiters such as the Directory for Open Access Journals (DOAJ), a community-curated online directory that indexes and provides access to high quality, open access, peer-reviewed journals. The rigorous DOAJ application and review process prompted us to consider topics such as preservation, identification, distribution, and licensing, and in doing so to reflect on the transparency of our own working practices. These issues are not, however, specific only to serial publications such as British Art Studies. Programs such as the Getty Foundation's Online Scholarly Catalogue Initiative have produced indispensable documentation about the working processes, successes, challenges, and failures of publishing in this form. 10 The "openness" and generosity of these publishers has led to invigorating international discussions that are reflected in the titles they produce. 11

Preservation and Sustainability
The guarantee of permanence and stability of the digital platform is one concern that potential contributors frequently voice. A service agreement between British Art Studies and ITHAKA Portico is in place to provide perpetual access to PDF versions of the journal should we as the publisher ever cease to support the title. However, while Portico will accept a variety of file formats, it currently has no means by which to render enriched multimedia content to a user. Thus, the long-term accessibility of the texts published by British Art Studies is secured, but a means by which to support continued access to interactive or multimedia content is as yet unresolved. In hoping to exploit the digital platform to its best advantage, we have always intended the primary reading format of British Art Studies to be HTML, and only released secondary versions in PDF and XML in response to specific feedback from our readers and distributors. However, there is some discord between this ambition and the requirements of preservation service providers, who favor the static PDF as both an access and archival format. While permanent long-term archiving of PDF files is widely supported, the technology and expertise required to replicate the digital ecosystem produced by the mingling of HTML, Javascript, and CSS is not yet fully established. As such, digital publications that privilege the augmented reading experience enabled by HTML are currently excluded from the benefits and large subscriber bases offered by aggregator platforms such as JSTOR.
The need to balance the enthusiasm and expectations of our collaborators against that of maintaining a solid and sustainable infrastructure is also a point of quandary. While software such as Adobe's Flash Player or the Unity game development platform offer the possibility to create rich interactive environments, the fact that these tools are not based on open web standards and require support from third-party plug-ins has made us hesitant to make use of such apparatus. As the editors of Vector have stated, "if authors or teams do choose a proprietary tool or platform, they should be able to provide a strong rationale for the choice they have made and articulate its advantages and shortcomings." 12 As such, we have chosen to direct our development budget towards well-documented protocols such as those produced by the World Wide Web Consortium (W3C), which promote consistency in the design of code, and provide a stable matrix upon which we can confidently build.

Images and the Digital Environment
Since the launch of British Art Studies, the evolution of the International Image Interoperability Framework (IIIF) has offered the single most significant addition to the art historian's tool set. 13 IIIF is a set of shared specifications for interoperable functionality in digital image repositories defined by the W3C Web Annotation Data Model. 14 IIIF and other forms of interoperable data and image delivery standards represent a move toward transcending the silos created by digital asset management systems and institutional websites. In this way, British Art Studies has echoed larger institutional efforts to open up image collections and connect them in innovative ways through Linked Open data to support research and sharing. The Yale Center for British Art is a leader in the adoption of interoperable standards and is a founding member of the IIIF consortium, as well as a member of the American Art Collaborative 15 and, along with the Paul Mellon Centre, of PHAROS. 16 The IIIF specification makes it easy to parse and share digital image data, migrate across previously unconnected technology systems, and provide enhanced image access for scholars and researchers. A series of open-source applications, including the Universal Viewer and Mirador, 17 have been created to present these resources, and offer a level of functionality that would be otherwise unachievable through bespoke development. The framework provided by IIIF allows British Art Studies to view, zoom, and reassemble image resources that can be incorporated within the body of published texts, making both image-based and textual primary documents, such as exhibition catalogues, available for consultation independently of the interpretative arguments of our authors. While our use of the IIIF framework has so far been limited to a fairly straightforward presentation of digital facsimiles, the potential to enable rich scholarly annotation of these resources is something that British Art Studies is keen to explore with future contributors.
By virtue of being an integrated part of the Yale Center for British Art and the Paul Mellon Centre's institutional infrastructures, British Art Studies is able to carry out all picture research for its authors and pay any fees associated with obtaining high-resolution media files and their copyright clearances. But the journal also actively uses existing exemptions to copyright to facilitate scholarship and public critical discussion. These exemptions are known in the US as "fair use" and in the UK as "fair dealing." The dominant practice in scholarly print publishing has been (and still is), however, to avoid using such exemptions lest copyright owners pursue litigation and test the existing case law. Pushing against this trend, digital publishing makes it easier for arts journals to experiment with fair use: for print publishers, there is always the threat of having to pulp an entire print run of books, whereas online publishers can revise a published layout in moments.

Collaboration and Conversation
Digital publishing is necessarily collaborativeas, of course, is print publishing, which requires the expertise and labor of editors, production editors, copy editors, and printers to produce the final publication. However, we have found that working in digital publishing has brought the conversation about the varied skills and expertise of our in-house team as well as external technical experts to the forefront of our collaborations with authors. The collaborative process means for each project the author and the members of our team need to work out not only what can be done and what needs to be done, but also who can and should do it. What will the authors and the journal team learn in this process? What can we all take away and share beyond the publication of the "article"? Working through these questions is inherent in almost every British Art Studies project, although we have not devised a standardized way to document or publish this information alongside the final output. One such approach is offered by the open-access journal Nineteenth-Century Art Worldwide, which in 2016 published "Digital Humanities and Art History," a series of articles sponsored by a grant from the Andrew Mellon Foundation. Emily Pugh, Elizabeth Buhe, and Petra ten-Doesschate, the team that instigated this series and oversaw the development and production of the articles, wrote a separate article, Nineteenth-Century Art Worldwide's "Digital Humanities and Art History: Reflections on Our First Articles," the purpose of which was to "reflect on the process as a whole, recounting our efforts and sharing what we and those who have worked on these articles learned in producing them." 18 This spirit of openness acknowledges the failuresor, to put it more productively, the experiments that did not work quite so well or turned out differently than expected. It uses the publishing process as part of the research process.
British Art Studies has encouraged a critical reflection on the methodology and processes involved in creating a digital research project. An example is the exercise undertaken by Catherine Roach in an article proposal that challenged readers to become more critically aware of the illusionary artifice that a digital reconstruction presents. In the process of reconstructing the interior of an exhibition staged at the British Institution in 1823, Roach observed that "reconstructions should be crafted with an awareness of their limitations and presented in a way that makes those limitations transparent to the viewer" and noted that the process nevertheless allowed an unprecedented opportunity to revive the ephemeral nature of a temporary installation. 19 Describing the practice of reconstructing the "visual narratives" of the exhibition, Roach reflected that the development of the video presented here was a collective endeavour … The designer, George Voicke, created the video from notes, source images, and mock-ups that I provided; this process was overseen by Tom Scutt, Digital Manager at the Paul Mellon Centre, who also worked on production of the final video and provided crucial editorial suggestions. The result is a visualization of the British Institution galleries in 1823 that is meant to be evocative instead of precise. 20 In negotiating the tension between working with partial and indefinite historical material and the requirements of software developers who needed a definitive answer to each question about the space's specifications, this particular article allowed both the author and our editorial team to reflect on the heightened process of exchange required to successfully undertake this form of research. We have found the results most fruitful when insights that emerge from employing digital tools inform the scholarly argument rather than simply using digital illustrations or visualizations to support pre-existing conclusions.
Collaboration is embedded in the very format of our journal by virtue of the possibilities offered by the digital platform to work with others, sharing knowledge and expertise, and also co-authoring, co-researching and co-publishing multi-vocal articles. Several of the "features" that have become regular to the journalthe Conversation Piece, the Cover Collaboration, and the One Object and Look First piecesare formats conceived by the editorial team to support multi-author projects and to prompt potential authors to think about how certain research questions and ideas are better served by including a range of perspectives and voices. The ideal of the lone scholar still predominates in the arts and humanities, where professional rewards and promotions are often evaluated through models which place the most prestige on the single author scholarly monograph. As colleagues working together across research centers (which also house art, library and archive collections), we are keenly aware of the diversity of working methods in our field. Yet, scholarly publishing in the humanities has been slow to adapt to representing these practices in research outputs. Co-authored articles showcasing pioneering findings are common in the sciences, but no similar model exists for collaborative arts and cultural research. Academic events, such as conferences and symposia, often raise questions and issues for our field which are debated in conversation, but our methods of publishing these events as proceedings often eradicates the cross-fertilization of ideas at these gatherings, atomized into individual contributions that demonstrate what was presented by an individual but not how their ideas were shaped by the responses to their presentation.
The Conversation Piece model that has frequently been featured in British Art Studies is one way in which we have used the digital platform to encourage multivocal perspectives on our field. Beginning in the first issue, with the purposely provocative title "There's No Such Thing As British Art," Richard Johns, an art historian at the University of York, proposed and "convened" this discussion, developing it first at a conference and then writing a provocation that he sent to conference participants as well as others who had not attended the original event. 21 In addition to the different perspectives of the contributors thereby assembled, the Disqus commenting tool built into the journal platform was added to the bottom of the feature so that readers could contribute their ideas as well. However, we found that the commenting function was little used (yielding only two responses). This attempt to draw readers into direct and public debate with journal content has perhaps been one of the least successful aspects of the platform. Whereas comment functions offered by the online versions of newspapers are notoriously populous and opinionated, it would appear that there is not an appetite among our readers to intervene in this way in the context of an academic journal, even when online. We continue to offer the Disqus feature and to think about how we might encourage readers to use itor conversely, whether discussion in the wake of releasing a feature might be facilitated and made public by the journal in other (perhaps more conventional) ways, such as through letters to the editors.
The Cover Collaboration feature also offers a primarily visual spacea virtual exhibition space of sortswhich we have designed to counteract the ironic situation that art history digital publications are often not all that visual or aesthetically pleasing. Often giving preference to text and only using small, thumbnail images, digital scholarly platforms have sometimes made timid use of their potential to create absorbing image environments with which to provide vivid visual support to the research questions being posed. This is a vestige of the PDF model and the limitations of printing technologies which, until relatively recently, were expensive and meant that most art history publications were illustrated only in black and white, or with a limited number of illustrations. It also seems to reflect a persistent nervousness towards image permissions and copyright in the digital sphere. We were determined that British Art Studies would make a visual impact and have a strong visual identity -this aspect of digital craft and creativity was paramount in the initial discussions with our developers and designers. We are keenly aware that digital publishing can present visual narratives and art-historical arguments in ways that are not only compelling and satisfying, but also have important implications for how research questions and arguments are communicated.
Working with artists, curators, and technologists, the Cover Collaboration feature of the journal presents full-screen images (still and moving) and accompanying interpretive text. From a selection of ten or so images, different designs are loaded at random on refreshing or returning to the cover, with the hope of piquing initial interest (while recognizing that readers might access the journal from any number of points) and inviting the reader into the journal. The Cover Collaboration for issue 2, for example, used moving images to communicate the unique details and stylistic traits that define the work of architect Louis I. Kahn, as seen in his design for the Yale Center for British Art. A series of GIFs made by David Lewis, an architectural historian then undertaking a postdoctoral fellowship at the Yale Center for British Art, capture the gentle modulations of light, texture, and atmosphere both inside and outside of the building, foregoing static representations such as stills and floor plans for a more dynamic means of illustrating architectural history. 22 In some instances, the customary format of text with accompanying illustrations has been insufficient to represent the complex and multifaceted nature of an object, archive, or discourse. This is particularly the case for articles which have brought together technical analysis, art history, and images generated through the process of conserving an object, such as Rebecca Hellen and Elaine Kilmurray's contribution, which connected the research carried out in the conservation studio with research from the archive, presenting new information about Sargent's working methods. 23 Or Roxane Sperber and Jens Stenger's discussion of "Canaletto's Colour," which detailed the changes to the artist's grounds, painting technique, and palette when working in England, using a slider image comparison tool to allow readers to compare the supporting illustrations, thus making the argument more visual and visible. 24 An early experiment in issue 1, "Deakin: Double Exposures," featured a discussion between an art historian and an archivist in film format across five different short features. 25 The combination of sound, illustration, and moving image footage allowed us to create a narrative that better reflected the serpentine relationships and society of the photographer John Deakin, and to present the material in a format that can travel beyond the confines of the British Art Studies domain. Those films have subsequently been screened in workshops and gallery events, providing opportunities to re-engage with the content as a point of departure for future research.
We have found that rich and digitally ambitious scholarship rarely arrives as a fully formed and polished proposal. Recognizing that the working processes and tools will be unfamiliar to many scholars, pedagogic frameworks such as panel discussions and skillsharing workshops have proved successful. One such initiative is "Objects in Motion," a workshop that approximated the characteristics of a studio or lab, which has been generously funded by the Terra Foundation for American Art. Our aim with this project is to nurture articles that are collaboratively authored, bringing together scholars from different locations, disciplines, and stages in their careers, so as to generate dialogue and innovative approaches. The Terra funding will be used to facilitate the research phase for developing the "Objects in Motion" series, and has covered the costs of an initial residential three-day workshop in May 2018, with the hope that some participants will wish to continue this exchange over a period of two years in order to develop their research for publication. Pamela M. Fletcher of Bowdoin College joined us to give a keynote and facilitate a skill-building workshop, after which each participant had the opportunity to present on the nature and anticipated scope of their projects and publication aims.

Peer Review and Citation Issues
There has been much recent effort to counteract the once widespread assumption that digital publications are not as rigorous as those produced in print. The field of art history is witnessing progressive changes in these assumptions, particularly with the support of organizations such as the College Art Association (CAA), which in 2016 published guidelines for the evaluation of digital scholarship. 26 British Art Studies is a peer-reviewed journal, but we have adopted a peer review model that is flexible as needed and appropriate to the particular content. The traditional models of double-blind peer review were developed and honed for printed materials and reflect a different publishing environment. Some of the questions and challenges that we have encountered so far include: At what point in the development of a work do we initiate peer review? As already described, not all "articles" arrive as finished submissions. Do we wait until we have completed a project (often as a product of substantial labor and financial investment)? Whom do we ask to review? Increasingly, we need a wider pool of reviewers to help us to assess material which has been produced collaboratively, across disciplines and using a range of technologies. How do we review multimedia content? Is the traditional model of "blind" or anonymous review the most appropriate, especially for collaborative projects?
While striving to preserve rigorous scholarly standards and the input of the opinion of external voices in developing the content of the journal, we have also come to realize that we need to be part of a process of developing new methods and standards of peer review for digitally produced and published art history. For larger and complex projects, we have begun to peer review at the proposal stage, asking experts for their opinion on the scholarly merits of the proposal, as well as its contribution to digital innovation. This involves reaching out to a wider field of experts to assess the various facets of a project and assist the editorial team when deciding whether to commit time and resources to a submission. How every submission has been reviewed for British Art Studies is clearly stated in the information that accompanies each article. We are aware that for digital publications to be trusted by authors and respected by readers with the same amount of weighting as printed materials, the processes of review needs to be transparent and standardized across digital journals. This will be one of the main tasks of a newly formed consortium of digitally published, open-access art history journals that will meet annually at the College Art Association.
Through the British Art Studies platform we have also addressed the issue of citations pertaining to digital publications. Responding to a dissatisfaction surrounding only being able to provide a link to an entire article (in the absence of page numbers), a more precise method of citation was developed whereby every paragraph of text, image and object in an article is associated with a permanent DOIa Digital Object Identifierwhich allows even greater precision and accuracy when locating a quote or source. Assigning a unique alphanumeric string to each element within an article allows a granular level of referencing and provides a persistent link to content no matter where it is located online. This means, for instance, that were British Art Studies to change its domain name, the DOIs could be redirected to the material's new location. Digital publishing has required us to come up with this creative solution for citation, using the potential of digital tools to solve the challenges of citing online articles.

Conclusion
It is still early days for British Art Studies. As we write this, we are preparing to publish the ninth issue. With each issue of the journal, the platform is adapting to the needs of our authors and readers. We are learning what works and what has been less successful. Somewhat ironically, our decision to move "beyond the PDF" was revised by a request from our international Advisory Board to introduce a PDF function to allow for British Art Studies content to be read offline. Our original argument for not including PDF downloads was that the journal was designed to be experienced on screen, especially given the amount of multimedia content. Persuasive arguments about the unevenness of access to fast internet, the usefulness of PDFs for personal research collections, and the desirability of PDFs for teaching packs and repositories, led us to capitulate and introduce a download function. These downloadable documents clearly state that they are just reference copies and do not include all functionalities, but the provision of PDFs still cedes a certain amount of control over reader experience from the author and editorial team.
Where readership is concerned, it has been gratifying to observe engagement to an unprecedented degree of accuracy. While quantitative measures of a printed book's success are calculated in unit sales, number of reviews, and subsequent citations, the very granular statistical information offered by Google Analytics has enabled us to monitor engagement in a much more focused manner, and to think about how we might commission content in response. For instance, while we know that in excess of 60,000 unique users have accessed the journal since its launch, it is also possible to rank articles in order of popularity. 27 Such data shows us that 6 of our 10 most read articles have been Conversation Piece or One Object features, and that the material published in film format captures reader attention to a proportionately longer duration than long-form text. It is also possible to filter and segment the reports by categories such as geography, device, web browser language, and age range (as perceived by Google). This information could, in the future, help us to set and monitor goals with the hope of engaging particular target audiences.
The digital can be disruptive, often in incredibly productive ways. British Art Studies is teaching us about different ways of doing things; new ways of working, presenting, recording, and preserving. It helps us to reflect on the kinds of research we support, produce, and publish as research centers in the twenty-first century.

Disclosure Statement
No potential conflict of interest was reported by the authors. TOM SCUTT is Digital Manager at the Paul Mellon Centre for Studies in British Art in London, and Digital Editor of British Art Studies. Tom identifies and initiates projects that will support the Paul Mellon Centre's strategic objective to embrace the use of technology in the study of cultural heritage materials. In doing so he leads the organization's digital audience engagement strategy, manages procurement of digital services, and coordinates the work of external designers, developers and partner agencies. Additionally, he oversees the digitization of the organization's research collections, including its Photographic Archive.