Art & Artists

James Davis, Tate, London

http://www.tate.org.uk/collection

Abstract

Suppose three people need something. One is bored and needs inspiration, another is focused and needs concentration, a third is curious and needs information. How could their needs be met? Is it possible to make a single thing that could help them all? Would it be a tool, a system, an application, a service, an environment? This paper describes attempts to imagine this problem, to invent a solution, and to try to build it. The context is that of Tate, a family of British art galleries, and the means by which to deliver artwork information, concentrated study, artistic inspiration, and much that is in between, online.

Keywords: art, Tate, gallery, collections, databases, interfaces

Summer 2008: A simple request

The trouble began with a simple request to redesign the page templates for the collection section of Tate's website (http://www.tate.org.uk/collection). There were seven principal templates and around 30 subsidiary variations that together defined the database-driven pages of this section, which features images and information about 70,000 works of art.

The goal was initially a graphical refresh and functional update, as even though these pages had been originally designed in 1998-2000, they were still working effectively several years later. This longevity turned out to be part of the trouble; such a comprehensive and innovative solution had been delivered at the turn of the millennium that little design and development had happened since. The system worked, was regularly cited as a shining example, and was left exactly as it was.

In reality, the system was lacking. It did not meet contemporary standards for accessibility, reusability, extensibility or technology use as described by Chan (2007), Peacock (2007) and others. The interfaces were clunky and unintuitive, the user experience was patchy, and images were small.

In addition, a large project was underway to catalogue the 30,000+ works on paper by JMW Turner, and following a decision by Tate Trustees, this material became destined for the Web. But to add this swathe of new scholarly material on to the already struggling interface would have done it no justice at all.

In summary, the UK's national collection of British art from 1500, and of international and contemporary art from 1900, was being poorly served on the Web.

A simple redesign of the page templates was not enough. Something more had to be done.

Fig 1: Screenshot of Tate Collection page largely designed in 2000, featuring Ophelia by JE Millais, 1851-2

Looking At Pictures All Day

A research phase was initiated. An intensive survey of museum collection websites from around the world was conducted, and after initially disappointing conclusions, the research field was widened to include websites that had collections of anything, and eventually was widened again, to include any website that might have something to teach us. Inspiring examples from any field were noted, screen captured and saved for later examination.

Fundamentally, museum and art gallery collections online rely heavily on search, with little scope for wandering and chance. Compare to retail collections online, in which un-looked-for objects are presented to you at every opportunity. Even libraries, real physical ones, are exploring innovative ways to present books inspired by digital tagging (Simon, 2009).

It was during this early research that the first obvious but key observation dawned on us: the budgetary and cultural restrictions imposed on museum websites held them back from competing in terms of leveraging the power of relational databases in a way their commercial rivals did. The "real" world had better examples of almost everything.

In order to justify the hours spent clicking fascinating places on the Web, some formal academic research was undertaken. Through study of cultural theorists such as Mieke Bal, Judith Butler, Julian Stallabrass, Norman Bryson, John Berger and others, some ideas began to emerge about the ontological principles at play in the delivery of artworks online. Art history and cultural theory proved very useful in exploring the agencies relevant in this new digital context.

And so it was here that the second key observation came to light. We realised that most Web solutions for displaying art online had been sleepwalking straight into the same paradigm long established in print publishing – that of reproduction. So our artworks online were apparently only a reference to real artworks, they were not artworks in their own right. Yet we suspected that this innate inferiority associated with reproductions need not be a given, that if we thought about it differently we might be able to change it.

This suspicion and new approach was attitudinally opposed to the gallery's own thinking just a decade prior, where the introduction to a printed interpretation guide (Baldwin, 1997) stated, "All reproductions are in black and white because it is hoped that they will not be used as a substitute for visiting the Gallery."

The next area of enquiry concerned visitors. Tate had regularly undertaken in-gallery fieldwork, and working with strategic research consultancies had developed a set of visitor categories. Each had a personality and a likelihood or not of attending a given exhibition, visiting the website and so on. Detailed Web statistics were also analysed to provide a little more insight into visitor paths in and out. The difference between Web and gallery visitors proved particularly interesting when attempting to compare like-for-like.

Fig 2: Comparison of physical visits (left) to digital ones (right) in 2009.

Physical 'visits' to see the painting Ophelia are optimistically guessed as follows: 75% of visitors to Tate Britain may end up in the correct half of the building, and 75% of those may end up in the relevant room and glance upon the painting, which is assumed to be on display 100% of the time. Hindsight suggests this to be a grossly high estimation, but the key point of note is the apparent disparity between this figure and the 42000 who 'visited' Ophelia online that year.

Thinking carefully about audiences led to the third and fourth observations: that visitors differ greatly in their motivations and needs, thus suggesting that a one-size-fits-all approach might in effect be a compromise for everybody; and that while we knew a great deal about our visitors, we knew almost nothing about that far bigger audience: non-visitors.

Finally – and with greater insight forming – we resumed the original assessment of museum collections online, and a more technical observation was made. Databases can force content into a set of rules that do not necessarily benefit the integrity and diversity of the content itself. We will discuss the implications of this later.

Despite the simplicity of insight, it is worth here recalling this complete set of observations, in order to understand the accompanying epiphanies and the design decisions that naturally followed.

• Museums can learn from other fields and commercial sectors
• Online artworks are supposedly not "the real thing"
• Visitors differ greatly from each other
• We know little about our non-visitors
• Database storage might hamper representation of the diversity of content

These observations were pondered, more websites were visited and some things were written down. A summer break helped turn all this into the following realisations.

1. Art Can Be Enjoyed Online

We felt that the 'reproduction as impoverished facsimile' model would always prove itself right if it were believed, and if the commonly held notion of physical artworks having some kind of aura were never challenged. We decided to disbelieve the former and challenge the latter, taking our cues from Berger (1972), Bryson (2001) and Bal (2001). If we believed that artworks could actually be experienced online, then this was surely the first step in making it possible.

We wondered why a never-heard complaint about the photographs on flickr (http://www.flickr.com) was that 'They are okay but you really need to see them in person'. We pondered why an amateur artist's website like DeviantArt (http://www.deviantart.com) could inhabit the top 100 websites on the Web with almost none of its artworks ever having been seen in person except by the artist. Were they aura-less?

Bryson (2001) refers to Benveniste's (1971) linguistic definition of 'deixis' as the essential relation between the indicator (extended in our case to be a website depicting art) and the present instance of discourse (you looking at it now) to describe how the very thing that happens in this relation is not arbitrary but of critical importance.

We concluded there was significant, fruitful space between a reproduction and the real thing, and strove to imagine it and build it.

2.Visitor Diversity Is A Challenge That Can Be Met

A cursory glance at Web metrics, or at complaint e-mails or blog comments, tells us that visitors are hard to categorise. Formal and often costly research confirms this, but does shed useful light on what those categories might be.

But still, their diverse motivations are often opaque, visitor routes in and out are quirky, their responses surprising. Audiences change their minds and modes, just as they wander through a gallery space being unpredictable, or say one thing in front of a colleague and another in front of a partner. The harder we work to know them through research and analysis, the more difficult they are to pin down.

Then, every assumption we do make about their needs on a given, singular Web page becomes a tussle between our perception and their reality. We guess this, they want that. We give that, they select the other. We can only really know that visitors differ from each other, that they are numerous, and we only sometimes meet their needs. We also know that there are hundreds of millions of non-visitors online, people who don't come to the website and have never even heard of us. How do we measure them in order to capture their attention?

We realised that by widening the scope about visitor understanding outside of traditional gallery audience segmentations, we might begin to break in to those hundreds of millions of non-visitors. This seemed so much more possible online than in a physical space.

3. Channels Can Be Powerful and Flexible

We began to envision a flexible Web offering in which visitors could alter our website to meet their needs. If we accepted the challenge of multiple visitor types, and of a new ontological space for art to inhabit online, then we could imagine a suite of tools that offered a combination of linear paths and non-linear explorations of our content.

We developed the idea of perspectives; that is, multiple ways to experience the same content. A small number of flexible perspectives could offer a wide range of ways of experiencing art online.

Three starting points emerged, perspectives from which a visitor could begin to tailor a journey: a scholar, an explorer and a dreamer. More on this later.

As large, complex commercial websites elsewhere began to shift from chunky behemoths to powerful Web applications through technologies such as AJAX, we realised we could take the final step in escaping the worst aspects of the database-template-webpage model, and come up with something more powerful and flexible.

In some ways databases are quite possibly the opposite of art, in that they force rich, diverse content to fit a precise mould or be damned. But art will often prefer to be damned over anything else. Yet we must use databases as a medium through which to experience collections of art online, and this significant challenge must be faced head on.

Stallabrass (2002) alludes to this issue with the user in mind: "the uniformity of look and interface creates a uniformity and passivity of behaviour among net users".

Thankfully, Tate was fortunate in having a Database Architect who understood and appreciated art, and this factor was to prove critical in helping the project negotiate later challenges.

So, a flexible, rich and powerful system, as described by De Caro et al (2010), but based on different ways of looking. Content prioritised by the visitor, and a shift in authorship control away from the institution, shared more with the artist and the visitor.

The very role of the museum was clearly beginning to change, from keeper to sharer.

Autumn 2008: modules & perspectives

All this thinking was worked up into a concept and business case document entitled 'Modular Perspectives'. The perspectives were the ways of looking, and the modules were what you saw.

At the time Netvibes (http://www.netvibes.com) was setting best practice in the field of highly tailorable websites, soon to be matched by iGoogle (http://www.google.com/ig) and the BBC homepage (http://www.bbc.co.uk).

Fig 3: Screenshot of Netvibes experimentally tailored for Tate multimedia and text content, 2008

Modules

These sites featured highly customisable panels or modules, each acting as a conduit for streams of related information. These panels were often movable, deletable, editable and resizeable, allowing visitors to pick and choose based on their interests.

We realised we could partition our collection information into self-explanatory modules, such as images, texts, information and so on. This seemed to offer several advantages, not just flexibility for the end user. It was inherently extensible, so a new feature could be added to the pages in the form of a new module. We could rationalise the dataflow behind the scenes, so one module might display a closely related set of data from one source, a neighbouring module using an alternative stream from somewhere else.

Here follows a list of modules (and options) that were originally suggested for an artwork page:

Table 1: a list of module and options

Other suggested pages with module and option sets were for artists, search and home. These and other pages would later become the more dynamically generated 'points of interest', illustrated later.

Perspectives

We realised immediately that we couldn't simply offer Web visitors a powerful toolset without revealing to them how it could be used, or catering to those who didn't wish to use it. The perspectives idea was leveraged further, and a rationale developed for three very different ways of looking at the collection. A perspective came to be thought of as a particular arrangement, selection and mode of a group of modules, the end goal being that similar content in each perspective could be shown in philosophically and graphically different ways.

Fig 4: Scholar Explorer Dreamer illustration. Original figure sketches by George Jones 1786-1869

Fig 5: Perspectives matrix. Of note is that each perspective is defined by a Lacanian narrator (Tate, visitor, artist) and is driven by a primary actor (study, play, art)

Winter 2008: How do we do this?

It was time to get something built. The team of one quickly became a team of four internal staff, and meetings began to digest the proposal and its technical implications.

The question of methodology was raised at this point, and at the time of writing (early 2011) was still a regular topic of discussion. More on this later.

The highly successful first version of the collection website had been built using a classic waterfall methodology. To simplify, requirements were laid out, features proposed and selected, stakeholders consulted and a development schedule planned and adhered to using a Gantt chart as the guide. We saw no immediate reason to question this proven approach, given both its prior success and that at least one team member had been directly involved back then.

Soon the original concept document had become a comprehensive Gantt chart, and the launch date of autumn 2009 looked and felt realistic.

Everyone seemed confident and work began in earnest.

A brief interlude: Into other parts of the office

It is worth noting here that the collection section of the website was built using a different set of systems and databases than the rest of the website, which was in parallel, and growing rapidly. Other sections had over the years been built using a content management system (Rhythmyx), or by hand in Dreamweaver, or externally often in Flash, and occasionally in Drupal, Moodle, Wordpress and so on.

Significant work was being undertaken to rationalise all of this, and to conduct a thorough redesign of all the webpages outside of the collection.

The same conclusion reached by the collection team – that a simple template redesign was not enough – was being reached by the production team looking at the rest of the website. In short, an opportunity was spotted to bring everything together and do a proper job.

Spring 2009: Multiple collection projects

It became apparent that each of the modules under development in the new collection website was a significant project in its own right – not just in terms of building the channel for delivery and expressing it, but in getting quality data into these channels from an editorial perspective.

It had been noted that our plans to leverage information into richer graphical interfaces would likely throw into sharp relief any data-quality issues that had gone previously unnoticed in the old interface. And so it came to be.

For example, all works in the Tate collection are subject indexed using the extensive in-house taxonomy designed for this purpose. Indexing was a mildly complex procedure undertaken by a skilled person who would analyse the visual, conceptual and other aspects of an artwork and tag it with nodes on a complex tree of around 12,000 terms.

Fig 6: Basic implementation of subject index tree

A major branch of the subject index concerned geography, and so artworks were tagged with locations where possible. The Getty Thesaurus of Geographic Names was used here, although the resulting association was text-based only, and did not allow us to place points on a map.

Fig 7: Salisbury Cathedral from the Meadows, John Constable, 1829

A map module on the new collection website seemed natural given that we had over the last few years already indexed around 22,000 artworks with geographic locations. But this attractive notion proved complex to execute.

First, it was realised that much of the data was of poor quality; second, that when linked up to a mapping API like Geonames (http://www.geonames.org) the mapped locations produced were not as useful as hoped. The problem was that we couldn't automate confirmation of whether a location had been mapped to the correct place or not. Due to the large, fuzzy dataset we decided to crowdsource this process, and built the Art Map application (http://www.tate.org.uk/artmap) to do so.

Fig 8: Screenshot of the ArtMap in phase 1

But because we were crowdsourcing the confirmation of whether locations were right, and not (yet) crowdsourcing the location of artworks on a map (the next step), this first stage proved too abstract for deep public engagement, and takeup was slow. Much later, this confirmation stage would end up being completed by hand.

The main collection development process continued, albeit patchily, because the team members were attempting to complete this project in addition to their daily responsibilities. There was not yet a formal budget nor executive support for other technical activities to be reduced in order to make time, so progress was slow. The Gantt chart on the wall began to drift into fiction.

But suddenly, and after some extensive effort by the Website Technical Architect, a working prototype emerged; it made use of the BBC's Glow javascript library (http://http://www.bbc.co.uk/glow) in which limited modules appeared, in working order, and could be moved around the screen.

While initially spurring the team on, the effect of using and playing with the prototype made us realise we had probably underestimated many of the interaction design challenges that our highly flexible system would present.

Things got difficult. With the Gantt chart slipping off the wall and a new understanding of the challenge we faced, it was time to reassess our methodology.

Autumn 2009: Agile development

As the original launch date came and went, we took a long hard look at how we were working, and decided, prompted by a suggestion from our Website Technical Architect, to switch to an agile development process. It seemed to suit both the unknown aspects of the project's complexity and our ongoing idea revisions. We rebooted the project and began from scratch with writing user stories on cards and beginning the sorting and prioritising activity which others (Ellis et al., 2008) have discussed.

At the same time we formalised the project at executive and director level in the institution, applying for significant staff time and resulting budget to be allocated and succeeding in both. We were back on track, with a new launch date of Spring 2010.

The project became a formal collaboration between two internal departments. Essentially the Tate Online Department led on front-end design and development, while the Information Systems department built the back-end architectures and data flows, so staff with a wide range of experience began to work together on a shared goal.

But our lack of experience in agile methodologies quickly showed. We would complete a phase and then have to research what the next phase ought to be and how best to do it. With no in-house expertise, it became clear that we needed external expertise, likely a formal project manager.

Our institutional drive for quality perhaps clashed with our aims for rapidity, and the project velocity was under question even if our capacity was not. We were making a great product, but would miss the next deadline.

Spring 2010: The low-walled garden

The next year began with a clarified vision, in no small part due to the launch and positive reception of Tate's Online Strategy 2010-12 (Stack 2010). This made clear that our relationship with visitors was shifting dramatically, and their relationships with us online would form a critical new flank of Tate's general activity.

Even delayed, the collection Web application was coming along very well, impressing all those who saw and used it. It was simply lacking integration and many of the planned features, each of which had turned into a project in its own right.

The next issue worth discussing is the development of the Turner works on paper and Camden Town Group catalogues, the latter sponsored by the J. Paul Getty Trust as part of the Online Scholarly Cataloguing Initiative.

A parsing tool had been developed which could automate the transformation of Microsoft Word documents into extensively marked-up XML. In combination with a bespoke asset manager and hierarchy builder for navigation, this would allow cataloguers to continue their practice of writing in a word processor and then simply handing over documents for Web delivery. With this funding, some magic, and a talented external editor, the process proved a success.

Fig 9: Chart detailing architecture and process of research content from cataloguer to website

The scholarly, research-based perspective was designed to present this information in a different, more focused format than elsewhere. But herein lay a difficult problem: how to helpfully present more than one set of information about a single artwork that had multiple contexts. We did not want a research silo, nor a mish-mash, and the resulting solution can best be conceptually described as a low-walled garden. You can see the perimeters of the garden (through navigation) and explore the entirety by following these paths and borders. Yet you can pop into a neighbouring garden with just a small step. This concept defined our methods of linking and switching between research projects in the scholarly perspective and the bulk of collection information in the main explorer perspective.

Summer 2010: Fragile development

With a new project manager appointed, our development methodology changed significantly, becoming a concern in some areas. Our team size was also increased to eight at this point.

Reams of extra documentation began to be produced, including:

• Use case scenarios
• User stories
• Functional requirements
• Success metrics
• Performance indicators
• Product roadmaps
• Stakeholder assessment forms
• Sprint plans
• Project plans
• Release plans
• Programme streams
• Concept frameworks
• Screenflow charts
• Layer models
• Wireframes
• Mockups
• Detailed specifications
• Prioritisation charts

And so on.

While some were useful, you will note that many of these documents sound very much like others, and while from day to day each piece of paperwork seemed perfectly justified and useful in its own right, together and over time they gave the distinct impression of being counter-productive and leaning towards the cost side of the cost-benefit analysis. As work on actual development slowed, there was a nagging sense that we were losing the agility that we had previously benefited so much from. We inhabited documentation land for much of each day.

Fig 10: Abstract layer model

As an example of documentation this diagram was produced to meet an opaque requirement. It did not prove to be of great use.

The second area of concern involved the daily scrum meetings, a popular practice in agile development. These were short gatherings in which the team members would stand in a circle and say what they had done and what they were going to do next. For some weeks these meetings had perhaps a negative effect on the group as a whole, and after checking articles on the subject, such as Stevens (2009), it became clear that our scrums exhibited many of the common problems that such gatherings face. A key example was their initial use more as a reporting mechanism than as an updating tool.

In more general terms we suffered from the tendency of agile projects reverting to waterfall techniques – a team struggling to adapt to new structured management practices with which we were largely unfamiliar.

Art & Artists

One useful and very impactful event happened around this time: a round of lab-based user testing. A small group of users were given a working prototype of the website and were video recorded as they interacted with it. Results were noted and two key conclusions were drawn (amongst others).

1. Users did not intuitively understand why they would want to move and adjust the panels of information around the screen. When this was explained, they did warm to the idea, but it was decided to remove this functionality and focus on flexibility where it really mattered. Soon afterwards, the BBC Glow library was dropped in favour of the more widely supported jQuery (http://www.jquery.com), a positive move in terms of user experience.
2. More significant, however, was that users did not understand what the 'Tate Collection' was, specifically what it contained or did not. Some did not understand that art and artists were included. In the users' defence, it turned out that the Tate Collection already contained many things that were not in the Tate Collection, such as the internationally-owned works by JMW Turner from the Turner Worldwide project. After much discussion and minor curatorial anguish, it was decided to rename this entire section 'Art & Artists', with 'Tate collection' as a subset (and now no capital C). This single linguistic change brought about significant information architecture rearrangements that ripped across the entire website, and while not easy to implement – requiring a major rethink of most other areas – it proved to have a very positive effect.

Autumn 2010: Out of the quagmire

Back in development, the group began to adjust to the agile practices, and the implementation of them began to be tailored to their way of working. It looked as if we might soon meet in the middle.

A couple of pieces of external news validated some technical commitments made earlier, and provided a much needed fillip. First, a Facebook engineer (Jiang, 2010) described the development of Facebook's new 'BigPipe' architecture, which contained some structural similarities to the module production framework our internal technical team had already developed.

Fig 11: Abstracted example of the first concept framework in Art & Artists. Web pages have become Points of Interest determined by a number of external factors

Secondly, a Twitter engineer (Busch, 2010) reported on his company's switch to the Lucene search engine library. This validated the same decision our Systems Developer had made some time prior and from which we were already reaping the benefits.

Of relevance regarding search was the ongoing work around curatorial texts and their style, and the changing practice of registrars and data entry. Tate's Research Department had been developing a guide to consolidate the different stylistic requirements around the gallery.

One of the areas where the search facility faltered was the inconsistency of texts and field data. We had made an early decision to perform a kind of advanced search on behalf of the users, even if they experienced a simple search box. This meant that we polled many fields and indices in order to derive relevance and usefulness.

But some of the information in these databases had originally been written in the 1950s or 60s, later being transcribed and entered. An incredible range of styles, terminologies and punctuation could be found, and there was no imminent opportunity for rationalisation. One by one we have had to identify such quirks and cater to them, a process almost complete by the end of 2010. We hope the introduction of the institution-wide style guide will make this work unnecessary in future.

A recommendation feature called 'Other works you may also like' began to show its strengths as our data improved, and was driven to a large degree by the lead Web Designer's great knowledge of art history as well as technical expertise. A scoring algorithm was tweaked to chart sometimes uncanny journeys through the collection. For the first time since digitization, each artwork webpage became as much a start point as an end point, useful both to promote general serendipity, and to aid conversion of our biggest audience: people parachuting in to these pages from a Google search or shared link.

Winter 2010: Breakthrough and inspiration

Even with the serious methodological challenges faced, the project finally began to flower in a truly successful manner. Dedication from individual Web Designers, Developers and Architects was the main reason, and this effort began to pay off, catching up with the original ambitious scheme.

The slideshow feature became our attempt to close the ontological gap between the artworks in the gallery and their representations online. Each design and interface decision was determined by the artwork and its importance, and in using an early prototype, we realised simple things like how easy it became to let go of the mouse and sit back, how ten seconds between slides would not be enough for some, simply how glorious landscape artwork looked when it took over your screen. Ongoing effort was put into the delivery of as many high resolution images as possible, with months of work required to appease external and internal stakeholders.

Fig 12: Screenshot of functional artwork page close to completion, lacking final visual design, January 2011

As a whole, the product was enjoyable and empowering to use. It could be operated in different ways and in different modes. You could sit back and watch the slideshow, lean forward and adapt the informationally rich features on your page, or perch on the edge of your seat gripped by an academic treatise.

A mood board was erected so that ideas, tips and tricks could be shamelessly borrowed from the world's leading websites and where possible improved or tweaked for our needs. Much work was being done in providing relationships between artworks and artists and other materials across the website, and a shareable breakthrough was made in the neighbouring website relaunch and Content Management System project by removing the hierarchies these taxonomies had originally depended on.

Spring 2011: Now and tomorrow

We write this before 'now' is technically here. The website referred to throughout has not been launched at the time of writing, and while it is scheduled for launch just a few weeks from now, lessons learned from missing earlier milestones must be heeded. It is simply hoped the reader will be able to enjoy the website as soon as possible after reading this.

This paper was written in part ethnographically, in that from 2008, observations were made and data collected with a view to the production of just such a paper. It was largely put together by an editorial team member who only delves into technical enquiries and was not involved in every area of activity. As such it is observational, clearly subjective and will have missed several key events and insights.

When attempting to conclude in a manner that is useful to others, it is worth recalling the hesitation suggested by Dourish (2006) who might be paraphrased thus: an ethnographically-informed paper concluded with a specific attempt to aid others must be treated more carefully than one in which advice can only be inferred incidentally.

Therefore we prefer not to dispense tips, advice or warnings; in fact we recommend you interpret our observations however you wish. However, at least one suggestion may hold shareable validity: there are a lot of non-visitors out there.

When the Google Art Project (http://www.googleartproject.com) was in development – a project Tate partnered on and helped scope and define – we realised how well it would supplement our goals by bringing in so many of these non-visitors.

We will continue to explore innovative ways of sharing our artworks as widely as possible, and to this end, Art & Artists is simply one of many ways to achieve this goal.

Acknowledgements

This project and paper would not have been possible without the dedication and creativity of the following individuals: Applications Developer – Ioannis Iordanidis; Database Architect – Diane Hall; Developer – John Hawthorne; Developer – Suneet Kamath; Freelance Editor – Steve Hare; Freelance Systems Developer – John Sydney-Willet; Head of Tate Online – John Stack; Lead Tester – Andrew Hainault; Online Collection Editor – James Davis; Project Management – Kevin Gordon, Nigel Runnels-Moss; Systems Developer – Kelvin Chappell; Web Designer – Alex Pilcher; Web Designer – Liam Palmer; Web Producer – Tijana Tasich; Website Technical Architect – Rob Adamson.

References

Bal, M. (2001). "On grouping: the Caravaggio corner". In M. Bal Looking in: the art of viewing. Amsterdam: G+B Arts International.

Baldwin, B. (1997). Looking into the Tate. London: Tate Gallery Publishing.

Benveniste, E. (1971). Problems in General Linguistics. Florida: University of Miami Press.

Berger, J. (1972). Ways of Seeing. London: British Broadcasting Corporation and Penguin Books.

Bryson, N. (2001). "Introduction: art and intersubjectivity". In M. Bal Looking in: the art of viewing. Amsterdam: G+B Arts International.

Busch, M. (2010). Twitter's new search architecture. Published Wednesday 06 October 2010. Consulted January 16, 2010. http://engineering.twitter.com/2010/10/twitters-new-search-architecture.html

Butler, J. (1990). Gender Trouble: Feminism and the subversion of identity. New York: Routledge.

Chan, S. "Tagging and Searching – Serendipity and museum collection databases". In J. Trant and D. Bearman (eds.). Museums and the Web 2007: Proceedings. Toronto: Archives & Museum Informatics, published March 1, 2007. Consulted January 12, 2011. http://www.archimuse.com/mw2007/papers/chan/chan.html

Compton, M. (1979). Looking at pictures in the Tate Gallery. London: Tate Gallery Publications Department.

Cosic, V. (2002). "Julian Stallabrass". In S. Cook, B. Graham & S. Martin. Curating new Media. Gateshead: BALTIC.

De Caro, S. et al. "In Search of Novel Ways to Design Large Cultural Web Sites". In J. Trant and D. Bearman (eds). Museums and the Web 2010: Proceedings. Toronto: Archives & Museum Informatics. Published March 31, 2010. Consulted January 16, 2011. http://www.archimuse.com/mw2010/papers/decaro/decaro.html

Dourish, P. (2006). "Implications for design". In Grinter, R. et al. Proceedings of the SIGCHI conference on Human Factors in computing systems. New York: Association for Computing Machinery.

Ellis, D., et al. "Agile Methods for Project Management". In J. Trant and D. Bearman (eds.). Museums and the Web 2008: Proceedings. Toronto: Archives & Museum Informatics. Published March 31, 2008. Consulted January 18, 2011. http://www.archimuse.com/mw2008/papers/jenkins/jenkins.html

J. Paul Getty Trust. (2009) "Online Scholarly cataloguing Initiative" Consulted January 16, 201 http://www.getty.edu/foundation/funding/access/current/online_cataloging.html

Jiang, C. (2010). BigPipe: Pipelining web pages for high performance. Published Friday 04 June 2010. Consulted January 11, 2011. http://www.facebook.com/note.php?note_id=389414033919

Peacock, D. and J. Brownbill. "Audiences, Visitors, Users: Reconceptualising Users Of Museum On-line Content and Services". In J. Trant and D. Bearman (eds.). Museums and the Web 2007: Proceedings. Toronto: Archives & Museum Informatics, published March 1, 2007 Consulted January 12, 2011. http://www.archimuse.com/mw2007/papers/peacock/peacock.html

Stallabrass, J. (2002). "Julian Stallabrass". In S. Cook, B. Graham & S. Martin Curating new Media. Gateshead: BALTIC

Simon, N. (2008). "How to design from virtual metaphor to real experience, and an example". Published Tuesday September 16, 2008. Consulted January 16, 2011. http://museumtwo.blogspot.com/2008/09/how-to-design-from-virtual-metaphor-to.html

Stack, J. (2010). "Tate Online Strategy 2010-12". In Mundy, J. (Ed). Tate Papers Issue 13. London: Tate Online. http://www.tate.org.uk/research/tateresearch/tatepapers/10spring/stack.shtm

Tate. (2005). "Minutes of the meeting of the board of trustees of the Tate Gallery held on Wednesday 16 March 2005 at 10.30am in the board room at Tate Britain". Published March 2005. Consulted January 16, 2011. http://www.tate.org.uk/about/governancefunding/boardoftrustees/tatetrustees_minutes_2005-03.pdf

Tate. (2011). "J.M.W. Turner: Drawings, Watercolours and Sketchbooks." Published January 2011. Consulted January 16, 2011. http://www.tate.org.uk/research/tateresearch/majorprojects/turner.htm

Torp, M. (2009). "Reality Re-checked: Docufiction, re-enactment, and the recontextualisation of the object in recent art". In S. Bjerkhof (Ed.) Reality Check. Copenhagen: Statens Museum for Kunst, 6-38.

Cite as:

Davis, J., Art & Artists. In J. Trant and D. Bearman (eds). Museums and the Web 2011: Proceedings. Toronto: Archives & Museum Informatics. Published March 31, 2011. Consulted http://conference.archimuse.com/mw2011/papers/art_artists

Copyright © 2011 – Archives & Museum Informatics.