Words from Collaborating Around Collections at ANU Canberra 8/11/17

[Another life-delayed missive]

Canberra, the compromise Australian capital set relatively equidistant between Sydney and Melbourne. No beaches but instead a city of public servants, research universities, national institutions. Twenty odd years ago I used to pop down reasonably frequently to play techno, and later drum & bass, at the Art School Ball, and one very crazy party inside the Aquarium. I still worry about the effects of all the bass on the fish.

Today was not one of those days.

Today ANU’s Mitchell Whitelaw and Geoff Hinchcliffe had convened a workshop of people from across Australia’s GLAM sector and academia to discuss better ways for institutions and researchers to collaborate with cultural collections. Australia has always been one of the leaders in this area from Australian Museums On Line (AMOL) in the late 1990s to the National Library’s Trove and, for natural history collections — the Atlas of Living Australia (ALA) — collection sharing and collaboration has been a part of the cultural sector’s digital innovation here for well over 20 years hampered by uneven funding and national network connectivity.

I was in a punchy mood. Maybe it was the airport coffee. Or perhaps it’s a bunch of things that have brewing.

Here’s some of what I said.


Last week one of my friends and former team, Tom Phillipson, sent me an image he had excavated from a Flash project for the Powerhouse Museum website that we worked on together back in 2001. Tom is doing a Masters in Curatorial Studies and he was doing some digging into his past for an assignment on online collections.

In that image you can see a more hirsute and shabbily dressed me, back from the days when I could get away with wearing a band t-shirt to work (Tortoise’s TNT album cover), along with Sarah Kenderdine, Peter Murphy, Ingrid Mason, Brett Stanley, and Tom. Server rooms haven’t changed all that much since then, and this image appeared as one of several Easter Eggs we hid inside the microsite we were making.

The ‘About’ team panorama easter egg hidden inside the Powerhouse Museum’s Virtual Museum project (2001)

That photo is pretty much the only evidence that remains of that project. The Internet Archive dutifully archived the site in 2003 but can’t run the Flash binary (yet), and even if it could, the ColdFusion scripts it calls won’t connect to the long-gone database of content.

The project — Virtual Powerhouse — was the first attempt to put the Powerhouse’s collection online by directly connecting to the museum’s collection management system (kEmu). Tom did the front-end design on the site and, being 2001, it was all folding robotic techno-looking. There was a very short-lived Australian youth media web startup called KGrind around a little earlier that had hinted at what might be possible with fast internet — and full streaming video — we emulated some of those aesthetics.

Virtual Powerhouse was a failure. Not only was it made in Flash and a ‘microsite’, but it also only gave access to ‘collection highlights’ — roughly 300 objects that we had taken good photographs of (in glorious 1024px sizes) — and some panoramas done by Peter Murphy of various behind the scenes locations which had to be stitched painfully by hand. Not many people used it — it didn’t satisfy the needs of researchers, and it didn’t give casual web users much of an insight into the breadth or depth of the collection either.

5 years later, in 2006, Powerhouse launched its OPAC2.0 which set a standard for many other institutions around the world for the next few years. OPAC2.0 followed a brief public experiment with tagging with a small project called Electronic Swatchbook, and together these two projects became international beacons for the early potential of large scale online collections, tagging, a public API and downloadable open access metadata in 2008, even the early use of text-mining tools with museum collection information. OPAC2.0 brought a large portion of the Powerhouse’s collection onto the web — far more than could be experienced in the galleries — and along with a lot of the narrative text that registrars and curators had written.

Last year Powerhouse finally overhauled OPAC2.0 — more than a decade after its launch. That would be the equivalent of a century or more in many other fields.


The work at Cooper Hewitt started from a similar collection data starting point. Aaron Cope built a website in Flamework that harvested data periodically from TMS, the collection management system, and in 2012, re-presented it in a prettier, more useful way. Over time, Aaron with Micah Walter & Sam Brenner added computer vision to allow for browsing by colour — forking code that was originally written by Giv Parvaneh for the Powerhouse’s Electronic Swatchbook — and the collection site and its flexible API because the basis of the Cooper Hewitt’s famous new exhibition galleries and interactive experiences.

At the same time as this work, Cooper Hewitt was busy digitising its entire collection. By the end of 2015, nearly 100% of the collection had been digitised, and more importantly the museum had a technical architecture that enabled it to start thinking of exhibitions from the perspective of ‘everything is digitised, everything is already online’ — and refocus the design of exhibitions with that in mind.

Cooper Hewitt had built a “a museum experience that expects a pervasive, ubiquitous network connection” throughout. Few other museums have reached that state — and fewer exhibition design firms, let alone museum IT departments, are prepared for that. But when it is achieved along with 100% digitisation, a whole new range of possibilities for the museum open up. The online collection was at the heart of this transformation — but despite having this at its core, the actual experience of visiting the museum is about as far from a big search box as you can get. That was the point.


Now at ACMI, predictably, I’ve gone through the same process with a new team, rinse and repeat. Provide open access to the collection metadata on Github, build interfaces to the collection, do some experiements with computer vision and machine learning. You can read about some of this over at ACMI Labs. Working with time-based media, heavily encumbered by multiple layers of rights, there are content differences but also new affordances to lean on, insofar as much film, TV and gaming content that the museum might wish to exhibit is also available through streaming services like Netflix, Steam, Playstation Network, XBox Live. Its a different set of challenges and opportunities.


But I’ve come to a point where I think that as a sector it’s time we rethought some of our assumptions and refocused efforts in our digital activites.

Before this gets misinterpreted, I should reitirate that I’m still a huge proponent of museum collections, digitising them, and putting them online. However, much like the public is often unaware that museums hold collections that aren’t on display, the public also assumes we’d already have all our collections online by now. If we had been doing our jobs — at least for the big museums, they might say, “why wouldn’t we have done this already?”. A fully accessible, digitised collection is table stakes for any museum this century. If you haven’t got there by now, and that’s 99.9% of the museum sector, then the chance to rally public support (and funding) behind such efforts may have passed.

The public at large might also fairly assume that museums have had sensible collecting practices and these sensible collecting practices continue into the present day. “Surely we’ve been collecting only the most significant and important markers of our cultural heritage, right?” They would also assume that museums knew and were certain about who they were connecting for — or, collecting on behalf of. “Surely those in charge of acquisition decisions were deeply connected to the full diversity of their communities?”. Ummm, no.

One needs only to think of indigenous artefacts and the ways in which they have often ended up in museum collections through violent colonialism, or the millenia of trade in antiquities as a by-product of war that now stare back at from the cabinets of the wealthiest museums. Or the exclusionary practices of art collecting.

We also know that our museum collections are full of a fair amount of objects that no longer have the value that the curators who originally acquired them thought they would have. That is part of the texture and nuance that museum insiders love — and some of the best museum experieneces are those where you chance upon a particularly quirky or strange set of objects. Sometimes those objects would be deeply inappropriate to collect nowadays, but their presence in museum collections is a reminder of how our societies have changed.

But in asking museums to deliver specific, mesurable, public value from our collections at a larger scale, we might lose some of this diversity.

Consider Ric Grefe’s, then Executive Director of the AIGA, comment on the Cooper Hewitt’s wallpaper collection from September 2011 — “How long can people defend a wallpaper collection?” — then compare that with the enormous success of the Cooper Hewitt’s Immersion Room which not only brought that digitised collection to life, has featured in thousands of Instagrammed moments, and also popped up at the London Design Festival. It is now possible to now argue that many more people would defend that wallpaper collection, simply because it has been made into something novel, tangible, and accessible.

Deaccessioning, especially in art museums, is a fraught process — and this has unfortunately meant that collecting practices have become even more conservative. “We don’t even have space for the objects we already have, and you want us to collect more?!”. One might also now ask if selective collecting practices have failed us — especially when we look around at the vast swathes of born-digital cultural materials we are steadfastly not collecting. But any alternatives remain steadfastly impossible.

In reducing our collecting practices, and not having our existing collections easily accessible, we may have inadvertently slowly begun to burn down the public trust we have accumulated over decades.

Then we might look at our catalogue metadata. This had been long created for internal use only and it shows. It is neither easy nor trivial to change this. At Powerhouse we were lucky in that cataloguing practices had, at least for some database fields, assumed a non-specialist reader. At Cooper Hewitt most records were and remain, despite 100% digitisation, threadbare and it is only through concerted editorial and curatorial work as objects were earmarked for display or inclusion as in-gallery digital objects, that their fields were rewritten or in most cases, written from scratch. At ACMI, only the former ‘lending colllection’ is well catalogued and even that doesn’t measure up to what the public might fight for the same items (films) that are also catalogued in IMDB or Letterboxd.

I’ve long argued that museums can’t wait for metadata to be perfect — it just needs to be released and made public, preferrably with obvious disclaimers that metadata will always be a ‘work in progress’. Now a decade on from that first bulk release of collection metadata from the Powerhouse, I’d argue this should and is standard practice and no museum should be writing big press releases about having done this.

In exhibition development, we transform metadata and ‘documentation’ into something more broadly useful through interpretative practices. On museum websites these began as microsites modelled on the paradigms of museum publishing, and then with the arrival of mobile, splintered into a range of different offerings from apps to ebooks, multimedia guides, and beyond. Like Virtual Powerhouse, none of these microsites or apps have really made the kind of long lasting impact that their host institutions hoped for — other than, perhaps, the most conservative and traditional of all — the Met’s Timeline of Art History. It is probably the only museum microsite that has been consistently resourced for the long term too.

In the last 15 years there hasn’t been a great deal of change in how we display our collections online. At the end of the day we’ve made some pretty front-ends for unmanageable sprawling databases. Some museums — led by the Rijskmuseum’s 2012 redesign have gone big with their images, others have added fancy ways of ‘browsing’, the best of them have even made them work reasonably nicely on mobile devices. But at their core they’re just expensive databases with some images.

They don’t stack up very well against an in-gallery experience in terms of memorability and impact. They aren’t designed for that — they are entirely utilitarian.


I still feel the discomfort (and joy) of first visiting MONA in Tasmania and discovering that you could get access to the MONA collection — but only after you had visited the museum itself and used its mobile concierge — The O. This confidence that you had to first experience the physicality of the collection before you were given access to the 2D simulacra of it remains a bold and important outlier that privileges quality of access more than open access.

A decade ago I, like many of us, argued that the web would give broader access to these collections and create opportunities to have a basic web-mediated experience with objects for those who couldn’t physically visit the museum. That’s still true, but the reality is that it is largely a shrinking cadre of researchers who repeatedly use these collection databases and those who are doing the deepest research work still use them to plan their physical visit requests to see the actual objects. The researchers who might engage in new types of research with digitised collections are still few and far between despite efforts to broaden training in the digital humanities and digital art history.

These days, Google Arts & Culture does a better job than most museums at providing broader access to more collections for a general ‘curious’ online public — and they’ve usefully connected up access to most of the ‘greatest hits’ anyway. And we also know that the majority of visitors to our websites are looking for ‘visiting information’, not long tail of collections.

So was it worth it? Most definitely.

But now it might be time to refocus our efforts on making better in-gallery exhibition experiences; acquiring more objects that matter to the communities we serve; presenting them in ways that engage and excite those communities bringing them closer to the museum; and documenting them in ways that have continuing value and meaning to those same communities.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.