Skip to main content

Museums and the Web

An annual conference exploring the social, cultural, design, technological, economic, and organizational issues of culture, science and heritage on-line.

Rhizomatic Art Stories - Balancing between Innovation and Usability

Annette Rosenvold Hvidt and Merete Sanderhoff, SMK National Gallery of Denmark, Denmark

http://www.smk.dk/kunsthistorier

Abstract

In 2008 SMK, the National Gallery of Denmark, initiated a 5-year digital strategy. A core project of this strategy is Art Stories: an innovative online resource that offers stories and background material about artworks, artists and how they connect. The ambition is to make art history come alive on the Web, utilizing the possibilities offered by digital media. Art Stories was developed over a three-year period in a highly iterative process. Our vision has been to create an 'art history on the terms of the Web' which implies structuring the content in a complex, 'rhizomatic' way. We have repeatedly invited users to discuss and test the concept, hoping to ensure an engaging and intuitively navigable online art experience. This paper goes through the process of developing the site in dialogue with users – but also sometimes against their immediate demands and expectations.

Keywords: Art history, persuasive design, usability, iterative process, lessons learned

1. A new approach to presenting art history online

Usually, art historical material presented online looks like a book that has been transferred to the Web: lots of text with small illustrations of the artworks mentioned. Our own museum website used to look like this:

Fig 1: Screenshot from a presentation of the Danish 18th century artist Nicolai Abildgaard in the old websiteFig 1: Screenshot from a presentation of the Danish 18th century artist Nicolai Abildgaard in the old website

In 2008, SMK was given a unique opportunity to change this and create a dynamic, interactive online art experience. This is the story of how we went from the look above to the one below – aided by users, and guided by a strong vision of art history in terms of the Web.

Fig 2: Screenshot from the presentation of Abildgaard in Art StoriesFig 2: Screenshot from the presentation of Abildgaard in Art Stories

2. What is Art Stories?

In the spring of 2008, SMK was fortunate to make contact with a fund called Nordea-fonden which encouraged us to present them with a concept of an entire digital programme for SMK, including a new website, Web TV productions, improved access to online collections, and a new database system to support it all. The concept SMK Digital was met with enthusiasm by Nordea-fonden, which granted 22 million DKK, equaling 5 million USD, over a period of 5 years. A donation on this scale for a digital programme has never before occurred in Denmark.

Art Stories is a core project in SMK Digital. It is conceived as a research-based online resource founded on the principle of serendipity (Chan, 2007). Layers of information are hidden in the artworks, and content can be entered from many different angles. Art Stories is intentionally in the plural; its stories are told by many – often opposing – voices within the art world. Lesser known artists and artworks are shown side by side with more familiar material in order to challenge the traditional notion of a canonized body of 'highlights'. Selecting which artists and artworks to feature out of a collection of 9,000 paintings and sculptures and 240,000 works on paper, spanning from the 14th century to the present, has been a sensitive task. We have endeavored to draw in underexposed content such as fragile, light-sensitive drawings that can rarely be on physical display, and artworks by unheeded artists usually stowed away in the dark cellars of SMK, side by side with artworks that are considered core national heritage.

Fig 3: In Art Stories we have tried to feature unheeded artworks; for instance, those by the female 19th century painter Elisabeth Jerichau Baumann, who has never been acknowledged alongside her contemporaries within Danish art history.  Fig 3: In Art Stories we have tried to feature unheeded artworks; for instance, those by the female 19th century painter Elisabeth Jerichau Baumann, who has never been acknowledged alongside her contemporaries within Danish art history. The idea is to present art history as a 'rhizome' (Deleuze & Guattari, 1988) – a non-hierarchical network of connections crossing traditional art historical categories, thus echoing the structure of the Web. We have deliberately eschewed a linear structure, hoping users will get lost in the content (in a positive 'serendipitic' sense).

The site is built around three categories: Artworks, Stories, and Artists. These elements are related via Relations. When users navigate from one element to another, they create their own trails through art history and can share their trails with other users via social media. Instead of perceiving SMK's collections as a 'walled garden', the resource links from SMK to related artworks and content in museums and institutions all over the world.

Fig 4: A user-generated track through Art Stories that can be saved, printed, and sharedFig 4: A user-generated trail through Art Stories that can be saved, printed, and shared

The central dilemma we have tried to solve is: How do we develop an innovative digital art experience based on a complex, rhizomatic structural principle, and at the same time keep it intuitively navigable and user-friendly?

Did we achieve this? That's up to the users to judge. Art Stories version 1.0 launched on January 20th 2011, and as we are writing this paper, fresh feedback from users is rolling in. The paper presents results from the five pre-launch tests and surveys, with a focus on the process of developing the concept and design in dialogue with users – but also sometimes against their immediate wishes and expectations. At the time of MW2011, we will present results from post-launch tests and surveys to uncover how Art Stories is actually being received and used.

Our vision

Our vision of Art Stories (Hvidt and Sanderhoff, 2009) can be summed up as follows:

  • Research based – a website featuring reliable art historical content drawing on research made by SMK's art historians and conservators.

Fig 5: Stories about how museum conservators work and what they discover are a central feature in Art Stories. We use a simple curtain tool that you drag with the mouse in order to explore hidden layers beneath the visible surface of paintings  Fig 5: Stories about how museum conservators work and what they discover are a central feature in Art Stories. We use a simple curtain tool that you drag with the mouse in order to explore hidden layers beneath the visible surface of paintings

  • Durable – a resource that users can return to again and again for more online art experiences and knowledge. We wanted a distinct yet classic design to support usability and durability.
  • Persuasive – lots of different interaction possibilities and functionalities to draw users in. In the process, we were inspired by principles from Persuasive Design (Fogg, 2003) which we have loosely based the design on. The idea is to activate and 'persuade' users to browse around, discover, open hidden layers, and feel inspired to dig deeper and get lost in a positive sense among the many layers offered.

    Fig 6: Selected=Fig 6: Selected artworks can be blown up in full screen zoom, allowing users to come closer than is possible in the gallery and study details such as the artist's brushwork

  • Visual – images work as entrances or starting points to ever deeper layers of information. Users should always be able to have a rewarding experience just from browsing images

Fig 7: A range of late 19th-early 20th century portraits of artists' wives from SMK and other collections are shown side by side  Fig 7: A range of late 19th-early 20th century portraits of artists' wives from SMK and other collections are shown side by side

  • User-centred – the site features multiple ways into the content, and users create their own 'trails' through art history by engaging with it. Ultimately, we'd like users to become aware that there is no such thing as a singular 'art history', but on the contrary, many parallel 'stories about art' that coexist and throw light on each other.
  • Connected – SMK artworks and content connect with related material all over the Web. Art Stories is based on associative relations, and the Web enables us to make these associations real – even between artworks that are separated by time, space, bricks and miles.

Fig 8: All content from outside SMK is linked to their sources so they are always just one click away. Here, Bruegel's famous painting 'The Triumph of Death' from Prado in Madrid alongside paintings with similar subjects from Danish museums.  Fig 8: All content from outside SMK is linked to their sources so they are always just one click away. Here, Bruegel's famous painting 'The Triumph of Death' from Prado in Madrid alongside paintings with similar subjects from Danish museums.

  • Multivocal – Art Stories cultivates a multivocal approach that is currently being tested everywhere in SMK's educational material. A core principle is to stress that scholars have different opinions about the same artists and interpret the same artworks differently. Senders are named and linked up with their online employee profiles, making it transparent who's talking.

To put it briefly: Art Stories is designed to stimulate users' passion for art!

Targeting adult users who want more

User surveys and studies conducted at SMK in recent years – both at the physical venue and online – have shown that our adult users demand more content about the artworks, presented in appealing ways. For several years now, SMK has been offering tailor-made art experiences for children and young people, online as well as onsite (Nielsen, Nygaard and Wilde, 2008), but nothing specifically catering for adult users. Art Stories is designed with two target user groups in mind:

  • Online users outside Copenhagen – We wanted to offer our online users engaging online art experiences that would work independently of a visit to the physical museum. Being the National Gallery of Denmark, SMK has the mission to be a museum for everyone in the country. Art Stories offered a chance to reach new users who don't stop by Copenhagen on a regular basis but are interested in art and art history. First and foremost, Art Stories is conceived as a destination in itself.
  • Regular onsite visitors – We wanted to offer our onsite users more engaging art historical content to support their visits to SMK onsite. We know from surveys that this user group mainly consults smk.dk to check out visiting information, but when asked they demand more background about the artworks on display. Art Stories broadens the scope of the onsite visit by delving deep into works from SMK's collections and relating them to art in other collections.

We assume that whereas the second group needs to be guided towards the site, the first group is more likely to come across Art Stories via, for instance, Facebook groups and other social media that they feel at home in. These assumptions will be tested post-launch.

  • Educators – A third target group that we also expected to reach were high school, college and adult education teachers who might utilize Art Stories content in their courses.

Process from concept sketches to launch

The development of Art Stories went through the four major stages sketched below.

1. Conceptual development: 1 year (August 2008 – August 2009)

Conceptual sketches, user scenarios, abstract design and IA. We collaborated with a Web bureau that specializes in user-centred design.

2. Request for proposals: 1 month (August 2009)

We asked four different Web bureaus to propose how to design Art Stories. We chose to work with a bureau called Oncotype. While the other bureaus advised us to drop the vast amounts of text on the grounds that it's a no-go on the Web, Oncotype understood that this was a crucial project precondition and showed serious engagement in tackling it.

3. Developing a prototype: 9 months (September 2009 – June 2010)

In collaboration with Oncotype, we developed an interactive design in multiple layers. Oncotype worked in the CMS Ruby on Rails enabling them to make swift sketches in a clickable system so we could immediately see the dynamic consequences of alterations and adjust iteratively. Prototype version 01 was user-tested in December 2009. Based on user feedback, we adjusted the site and had a prototype version 02 ready for testing in June 2010.

4. Implementing Art Stories in smk.dk: 7 months (June 2010 – January 2011)

SMK launched a new website (http://www.smk.dk) in March 2010, developed in the Open Source CMS Typo3 by MOC Systems. Art Stories was planned to launch half a year later, and a space was reserved under 'Explore the Art' in the main menu. The prototype was transferred from Ruby on Rails to Typo3 in close collaboration between developers from Oncotype, MOC Systems, and members of the SMK Digital team. This turned out to require more time than expected, pushing the launch of Art Stories to January 2011.

Overview of user tests and surveys

During the development of Art Stories we conducted two online surveys, one focus group test, and two usability tests using the think-aloud protocol. The results were used to guide us regarding the basic issue we were tackling. How far could we move in innovative directions with the structure and navigation of the site, and still keep our users interested and motivated?

1. First lessons learned

In 2008-09, SMK held a special exhibition of the Danish 20th century surrealist artist Wilhelm Freddie, one of the artists who would later be featured in Art Stories. In order to start getting some user feedback on the concept and content, we created a subsite (http://freddie.smk.dk/) for this exhibition in collaboration with the Web bureau Spild Af Tid (Danish for 'waste of time'). This turned out NOT to be a waste of time! The conceptual ideas we had of a platform with many different entrances, a dynamic/organic look, and lots of text and images at the same time, turned out to be a bigger challenge for the Web designers than any of us had imagined.

A survey about the Freddie subsite was distributed to SMK Digital's online user panel, consisting of smk.dk users who have agreed to act as test users on SMK Digital projects. Their replies showed that the site didn't provide a satisfying experience. The playful collage-like design that was being tried out removed focus from the artworks and furthermore had the unfortunate side effect that many users didn't notice that they could watch videos and play games. These were new features that we wanted many users to discover and try out. The Freddie subsite mainly showed us what not to do. The two basic lessons learned in this process were:

  • We needed more advice on how to create the basic structure and navigation of the site we had in mind.
  • We must leave the idea of a funky organic-looking design behind since it completely overshadowed the artworks.

2. Feedback on content

In 2009-10 SMK held a special exhibition of the Danish 18th century history painter Nicolai Abildgaard who would also later be featured in Art Stories. In order to test some of the formats that we were working on for the site, we conducted an online survey, testing an artist text bio and a short biographical video. As these new formats were implemented in the framework of SMK's old website, the user feedback was only valid as to content and content presentation, not the design. Knowing that on the face of it SMK Digital's online panel consists of potentially motivated users, the results of the survey showed us that we still had a long way to go, but we were headed in the right direction.

The most important lessons learned from the second survey were:

  • Text presentation – We needed to work intensively on making the massive amounts of text in Art Stories much more appealing to access and navigate. Users commented that the text was boring to read and rigid to navigate. Seeing that one of our primary goals was to motivate users to delve deeper into the content, we needed to come up with a design that could hold all this text and still radiate an air of easy accessibility and catching enthusiasm.
  • Videos – We were on the right trail with short biographical videos about the artists. Users loved the video about Abildgaard and felt motivated to learn more about the artist after viewing it. One user put it this way: "It was liberating to see such an easy and cool take on the artist." Some users commented on the pace of the video and suggested we slow down a bit and linger more on each artwork. This suggestion was implemented in the following productions.

Fig 9: Still from the video about Danish artist Nicolai AbildgaardFig 9: Still from the video about Danish artist Nicolai Abildgaard

3. Moving towards a design

One of the first things Oncotype did when offered the job of designing Art Stories was to kill one of our absolute darlings. We had envisioned an interface built as a rhizomatic network of connected art historical dots you could navigate between. However, with all the content we had planned to upload, Oncotype predicted that the rhizome would in no time grow enormous and impossible to navigate. Instead they suggested that we work with layered content.

Fig 10: Abstract design by Advice DigitalFig 10: Abstract design for Art Stories before we dropped the idea of a rhizomatic interface

In October 2009, we had been working intensively for two months on developing the concept and structure for Art Stories, and conducted a focus group test on the paper sketches, navigational principles, wording and overall concept. The most important lessons learned in the focus group test were:

  • Concept – The group instantly felt like clicking on sketched buttons such as 'Zoom', 'X-ray' and 'Hotspots' to see what might happen, and they generally liked how the concept invited them to explore the content instead of telling them what to see and think.
  • Structure – The concept of individual user trails that would automatically form when moving through the material appealed to the group.
  • Design – The presentation of text was not optimal. Text and images in layers was a new experience to the group, and they found it difficult to navigate. They liked, however, that we were working towards integrating text and images to provide a different experience than 'reading a book with illustrations'. One user commented: "It's so cool: The story comes out of the image!"

Fig 11: Presentation of text with open note, hotspot link: We have worked intensively with making text appealing to read and navigate. Text is dynamically interconnected with the images described, and the text features hidden layers, lots of links, and ilFig 11: Hotspots feature special zones of interest in the artworks

4. Getting there

By December 2009, we had arrived at our first version of a clickable prototype in Ruby on Rails which was tested in individual one-hour sessions. The most important lessons learned in the first round of individual user tests were:

  • Design – The clean, minimalistic design that Oncotype had developed appealed to the users. It allowed the artworks and special functionalities to stand out, and radiated credibility and quality. The biggest design challenge of all, however, was still to make space for huge images and huge text amounts at the same time. Users were divided on the topic of layered content. Some felt confident navigating between layers of image or text; others were annoyed that text would cover parts of the images and vice versa.
  • Navigation – The experimental navigation continued to cause confusion. Users didn't intuitively grasp that elements would open in a vertical downwards sequence they could scroll through.
  • User-generated content – Some test users spontaneously demanded options to contribute their own content; for instance to be able to upload their own images, comment on stories, or even create their own stories. When testing the site post-launch we will ask users what kinds of options they'd like for contributing content. The effort to implement these will be coordinated with the general museum website where MySMK – a space for users to interact with each other and the museum content – is coming up in 2011.

5. Final challenges before implementation

We listened to our users and followed their advice on many issues. But we didn't want to give up on our core ideas, such as making space both for large images and vast amounts of text, and interconnecting all content : these demanded an innovative navigational structure. Instead, we kept trying to enhance the usability of these features. In order to test the results of our iterations and get fresh ideas from users, we conducted a final user test in June 2010 before implementing Art Stories into our brand new museum website. The most important lessons learned in the second round of individual user tests were:

  • Navigation – Users needed more guidance in order to navigate the site. The front page with one featured trail didn't explain the overall concept to users, and they got confused by the abundance of menus featured on the front page. Furthermore, they couldn't intuitively work out the horizontal image scroll and the vertical text scroll. Up till now, Oncotype had been reluctant to add visible markers in the design, but guided by the user feedback, we decided to implement arrows, scroll bars, and taglines.
  • Design – Half the users still got really annoyed that they couldn't read about an image and see it in its entirety at the same time. Oncotype ended up solving this by implementing a zoom tool that could both enlarge and reduce the image. This enabled users to freely control the experience of looking and reading at the same time – and it meant that we finally reached a satisfying solution to our biggest challenge!

Fig 12: The enlarge-reduce function VideosFig 12: Presentation of text with open note, hotspot link: We have worked intensively with making text appealing to read and navigate. Text is dynamically interconnected with the images described, and the text features hidden layers, lots of links, and illustrations

Users continued to be very pleased with the introductory artist videos, commenting that they offered a quick overview of an artist's life and work. They appreciated that we had used different speakers which gave an immediate impression of multiple voices telling the stories.

Fig 13 & 14. Video still + Stories image stream featuring art and music examples: In Art Stories associations often spread to other kinds of artistic practices. We have worked together with two music labels to set appropriate music to art history so usersFig 13 & 14. The enlarge-reduce function allows users to freely control the experience of looking and reading at the same time

3. First user comments post launch

On January 20, 2011, Art Stories was launched, and the initial reception of the site exceeded all expectations. The Danish media covered the launch extensively, giving the site a great kick-start. Both national newspapers, online art magazines, and the local TV-station featured reviews and background interviews about the site. In Denmark, it's unusual for an online initiative about art to create a stir in the media, so it really came as a surprise when journalists started calling to ask questions about Art Stories.

Users have blogged and twitted about it, mostly leaving excited comments. One twitter put it this way: "SMK, I like. Those trails across the content are cool. I think I'm gonna use several hours on this site." We've received critical user comments too, especially about navigation and the limited possibilities for sharing and contributing content. A user on Twitter said: "Hey information architects/UX'ers, I don't want to understand how you think in order to search out content. You must adapt to me."We're in contact with the users who've commented on the site and will invite them to participate as test users in February.

Also within the Danish museum world there has been great interest. On launch day, the Danish Ministry of Culture featured Art Stories on the front page of its website, and positive comments have been rolling in from colleagues at other museums ever since launch. A curator from an art museum outside of Copenhagen wrote an e-mail saying: "Inspiring, appetizing, delicious, and amazingly informative and serious at the same time. It's really funny and really good. It's the best I've seen that this medium can do."

Expressions such as these of course reassure us that the three years of development and iterations we've put into Art Stories haven't been in vain. But we know that we have a lot of work ahead of us making Art Stories still more functional and versatile in meeting different user demands.

Future challenges

Three major challenges stand out post-launch:

1. How do we keep our users loyal and keep them coming back for more?

We're working on developing strategic tools to help users stay loyal via feeds and newsletters. Upcoming user tests will offer us a chance to learn more about how users would like to be informed and reminded about the resource, and what might persuade them to keep returning for more.

2. How do we make space for user-contributed content without selling out on the core quality about Art Stories that it's a research-based resource?

We've distributed a survey to SMK Digital's online panel, and will conduct individual user tests in February, asking how users would like to interact with the content available in Art Stories. We don't know yet how to integrate user-generated content alongside the research-based material that is central to it; neither do we know how to make it attractive for users to spend their valuable time contributing content to this platform. Another crucial challenge is to motivate the researchers and conservators at SMK, maybe even from colleague institutions, to contribute their knowledge and material to Art Stories so the content keeps evolving.

3. We need an English version.

Denmark is a small language area, and very few people outside the country speak or read Danish. If we want to take real advantage of the fact that we've put an art historical resource out on the Web, it's vital that we offer an English version, as an absolute minimum. The challenge is that we don't have the funds to get all the content translated, but we still believe it's crucial. In order to get started on this huge task, we're planning to have a representative slice of Art Stories translated into English so we can present the site properly at Museums and the Web in April.

4. Conclusion

Art Stories is a type of online resource about art history that we haven't seen before in Denmark. We've been extremely fortunate to have the necessary time and resources to work highy iteratively and develop a new approach to presenting art history on the Web. We're grateful to the users for their dedicated and invaluable feedback. We're also pleased that we stood our ground on certain issues and kept trying to make our original vision turn into something real. Launching Art Stories has shown us that it's circumstances like these that make the whole difference. They have enabled us to create an online resource about art that has the potential to become a destination in itself.

Acknowledgements:

The authors are deeply grateful to: Nordea-fonden, Liza Burmeister Kaaring, Anne Skovbo, Oncotype (http://www.oncotype.dk/), MOC Systems, Michael Junge, Malene Rørdam, Helene Brøndholt Nielsen, Sebastian Campion, Axel Kellermann, Mathilde Schytz Marvit, Fabulab, Martin Strange Hansen, Henric Wallmark, Angela Spinazze, Carrie Davenport Anderson, Christopher Pott, Jakob Fibiger Andreasen, Iben Reither, Ragnar Lundø, Sofie Regitze Kattrup, Sarah Grøn, Birthe Nobel Øby, Naxos Records Denmark, Dacapo Records, Advice Digital (http://advicedigital.dk/), E-types (http://www.e-types.com/46408/), and Kollision (http://www.kollision.dk/).

References

Chan, S. (2007). "Tagging and Searching – Serendipity and Museum Collection Databases". In J. Trant and D. Bearman (Eds.) Museums and the Web: Selected papers from an international conference. Toronto: Archives and Museum Informatics. 87-100. Also available at http://www.archimuse.com/mw2007/papers/chan/chan.html

Deleuze, G. and F. Guattari (1987). A Thousand Plateaus. Capitalism and Schizophrenia. Minneapolis: University of Minnesota Press.

Fogg, B.J. (2003). Persuasive Technology: Using Computers to Change What We Think and Do. Massachusetts: Morgan Kaufmann Publishers.

Hvidt, A. R. and M. Sanderhoff (2008). "Digital Art Stories". In A.S. Warberg Løssing (Ed.) Digital Museum Education – From a User Perspective. Copenhagen: The Heritage Agency of Denmark. Available in Danish at http://www.kulturarv.dk/publikationer/publikation/artikel/digital-museumsformidling/ (consulted 01/31/2011).

Nielsen, A.T.S., T. Nygaard and E. Wilde (2008). "Teens Connect to Art and Each Other at Young People's Laboratories for Art". In J. Trant and D. Bearman (Eds.) Museums and the Web: Selected papers from an international conference. Toronto: Archives and Museum Informatics. 115-122. Also available at http://www.archimuse.com/mw2008/papers/nielsen/nielsen.html

Cite as:

Hvidt, A.R., and M. Sanderhoff, Rhizomatic Art Stories – Balancing between Innovation and Usability. 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/rhizomatic_art_stories