It’s always a great start to a day when the first two links you click inspire a flurry of fresh thought. I have been getting stuck into some PhD writing this week, and fast losing myself in the doldrums of theory. So waking this morning to a little bit of inspiration was just what I needed.
The first shot of inspiration, that woke me far more than a coffee would, was this super-cool research on What Makes Paris Look Like Paris? (on Openculture via Jasper Visser & Seb Chan). I remember as a child someone telling me that all cities had colours; that some cities were grey, and some brown. Some were blue. These dominant colours reflected the materials that had been used in construction, the fashions that had shaped the way the city was constructed, the natural resources that were available to the builders. And so I often look out at new cities as I approach them, and watch to see what colour they are. This research reminds me of that, for it analyses Google Street View imagery to look for the common visual elements of a city, like its architectural features. It turns out that not only do cities have colours, they also have distinguishing architectural features.
Imagine what kind of new information and understand lies within our collections, if similar techniques were deployed. What kind of features are common to paintings of Paris? Are there common colours used to depict Paris, or are the architectural elements captured in the research above visible? And what else can we learn about our collections through these kinds of techniques? It is easy to think of this (for me) in terms of art collections, but I am sure there is much that can be found in archeological data and other museological data. (Not that most museum data is that great, as Mia Ridge recently discovered when playing in the Cooper Hewitt datasets.)
There is some work being done in this area, of course, but I’m interested in what else we can find in our collections using these kinds of techniques. This morning, I also watched What do they have? Alternate Visualizations of Museum Collections in which Piotr Adamczyk, when he was at the Met, talks about the possibilities for new information that might be possible in collections data. During the question time, he speaks of his interest in using data to look at provenance and figure out the history of the object in order to visualise who had it, when and where. For me, this is exactly the sort of flow of information that I would find so interesting about collections. I am really interested in the power structures and power makers in any sector. When I met curator Helen Molesworth recently, I asked her what I would discover about her influence on permanent collections, were I to look across the course of her career; who and what she had collected consistently or in different institutions over her life as a curator. It was a question that floored her, because it was one no one had ever asked before. But to me, this is the interesting stuff of museums. Who are the individuals that change the shape of our collections, and indirectly then, the shape of our material wants and expectations? Who has shaped the art market by collecting the works of an individual and increasing their value for other collections? Which individuals have really changed the shape of our cultural heritage, its value and its impact? Who has championed the work of previously unknown artists, and turned them into a hot commodity?
So my other early moment of inspiration in starting the day was watching Koven Smith’s MuseumNext talk, which was just gone online. In it, Koven speaks about curators using algorithms to produce collection narratives – interpretive algorithms. Now it seems to me that this idea starts to coincide with the work being done above, whereby collection researchers and communicators working in a museum could have a focus on a whole collection, and how it relates to the rest of the world, rather than only having curators (or researchers) whose focus is on exhibitions and material culture.
When I last wrote about big data and museums, I quoted from Mia Ridge, who mentioned that there are probably lots of other people who can do great things with museum data, much more than museums can and potentially should. And I agree with that. But I also wonder if making sense of our collections at a macro level with these sorts of techniques and possibilities isn’t also something museums should be doing. I don’t know about that, but I do think it’s something to think about.
What do you think? What would you like to see visualised using museum collections? Are there new ways of looking at the work we do that technology is making possible in ways that weren’t previously available? And should this work be done within the museum, or is it just the responsibility of the museum to enable others to do it?
10 thoughts on “On data visualisation + algorithmic curating”
perhaps somewhat related, I have always found Edward Tufte’s work good to think about:
Museums should absolutely be trying stuff like this. But there is still a divide between the technology people who understand the possibilities and the museum science people who know what we should be looking for. We need to bridge that gap to really unlock all of the knowledge that is locked up in our collections just waiting for us to discover it.
So is it a question, then, of hiring different people? Of putting people to work together on such problems? On bringing external people into the museum with expertise on visualisation (for instance) and getting them to work through these things under curatorial advisement? How do you think we bridge that gap?
It’s a tough question to answer and I don’t really believe that there is an “institutional” solution to the problem. I come from a field that is very much about self-education. The only way to really learn anything in computer science is to explore. It’s messy, inefficient and sometimes slow, but aside from some basic theory you can pick up in a few classes it really is the only thing that works (apologies to my CS professors from way back when, but you all know it’s true).
I would imagine that curating is a skill that really only develops through a similar exploration (I can’t imagine an educational program that does more than provide a head start in a field that requires so much familiarity with the subject matter). So I’m assuming that museum professionals are capable of exploring the possibilities and learning without direct guidance all the time. The only thing I can imagine has stopped so many from doing so in the case of digital tools so far is either a lack of time or a lack of interest.
I think perhaps the best thing we can do is show people examples of what can and is being done elsewhere, like the Paris thing. The real problem that most people have with digital tools is that they don’t know what’s possible because, as a society, we’re not used to thinking about letting machines do some of our thinking for us. We don’t always trust the possible outcomes, and we don’t consider the things that our senses and human limitations cause us to miss because, until someone or something makes that manifest for us, we can’t really experience those things. They’re just not in our vocabulary of tangible ideas we think about.
People aren’t interested in exploring the possibilities because they just aren’t aware of how cool those possibilities can be. Once we show them something that grabs their attention, I would hope that they would be motivated to pursue those possibilities. It seems the problem right now is more one of motivation than anything else.
Matt this is really interesting. I agree with you that motivation is a big barrier here, and a lack of awareness of what is possible. I also think the fact that for many people working in museums the objects really are *the* thing probably gets in the way here. Why would I care to investigate things with boring data when I can go and play with “the real thing”? I mean, I am absolutely thrilled about what lies beyond what we’ve already done when it comes to this sort of stuff, but sit me down in front of a database or ask me to write lines of code, and I’m nothing. I am powerless in that space. The problem then is even greater if I am someone who is powerless and also without vision for what could be possible.
I had a conversation this week with a woman who has worked in museums for many years, including being a director at a very small institution, and she was adamant that museum collections online actually had very little to do with museums; that they were encyclopaedias at best, and should have a name that separated them from museums so that people didn’t get confused. That attitude, which might not be uncommon, is a real barrier to this sort of stuff. If dealing with data in any capacity beyond just managing collections internally and maybe telling people what stuff you have in that collection is not considered museum work, how do you move beyond that? Do you move beyond that? My instinct is that we have to, but what if it isn’t for museums to be doing this at all; if it is just for them to be enabling others to make sense of our collections in exciting ways? But then I don’t think museums can claim to be knowledge institutions if there is a refusal to work in the space where knowledge is so commonly transferred and made.
Reblogged this on The Kinetic Museum and commented:
Suse is headed in the right direction here–check it.