Friday, 4 January 2013

Clash of the models? Object-centred and object-driven approaches in online collections

While re-visiting the world of museum collections online for some writing on 'crowdsourcing as participation and engagement with cultural heritage', I came across a description of Bernard Herman's object-centred and object-driven models that could be useful for thinking about mental models designing better online collections sites.

(I often talk about mental models, so here's a widely quoted good definition, attributed to Susan Carey’s 1986 journal article, Cognitive science and science education:
'A mental model represents a person’s thought process for how something works (i.e., a person’s understanding of the surrounding world). Mental models are based on incomplete facts, past experiences, and even intuitive perceptions. They help shape actions and behavior, influence what people pay attention to in complicated situations, and define how people approach and solve problems.'
CATWALKModel House FaceTo illustrate a clash in models, when you read 'model' you might have thought of lots of different mental pictures of a 'model', including model buildings or catwork models, and they'd both be right and yet not quite what I meant:

And now, back to museums...)

To quote from the material culture site I was reading, which references Herman 1992 'The Stolen House', in an object-centred approach the object itself is the focus of study:
"Here, we need to pay attention to the specific physical attributes of the object. The ability to describe the object – to engage, that is, with a list of descriptive criteria – is at the forefront of this approach. A typical checklist of the kinds of questions we might ask about an object include: how, and with what materials, was the object made? what is its shape, size, texture, weight and colour? how might one describe its design, style and/or decorative status? when was it made, and for what purpose?"
In object-driven material culture:
"the focus shifts toward an emphasis on understanding how objects relate to the peoples and cultures that make and use them. In particular, ideas about contextualisation and function become all important. As we have already noted, what objects mean may change through time and space. As products of a particular time and place, objects can tell us a great deal about the societies that gave birth to them. That is, they often help to reflect, or speak to us, of the values and beliefs of those who created them. At the same time, it is also important to remember that objects are not simply ‘passive’ in this way, but that they can also take on a more ‘active’ role, helping to create meaning rather than simply reflect it."
It seems to me that the object-centred approach includes much of the information recorded in museum catalogues, while the object-driven approach is closer to an exhibition.  Online museum collections often re-use content from catalogues and therefore tend to be object-centred by default as catalogues generally don't contain the information necessary to explain how each object relates 'to the peoples and cultures that make and use them' required for an object-driven approach.  If that contextual information is available, the object might be sequestered off in an 'online exhibition' not discoverable from the main collections site.

A complicating factor is the intersection of Herman's approaches with questions about the ways audiences think about objects in museums and other memory institutions (as raised in Rockets, Lockets and Sprockets - towards audience models about collections?).  The object-centred approach seems more easily applicable to individual objects but the object-driven approach possibly works better for classes of objects.  I'm still not sure how different audiences think about the differences between individual objects and classes of objects, so it's even harder to know which approach works best in different contexts, let alone how you would determine which model best suits a visitor when their interaction is online and therefore mostly contextless.  (If you know of research on this, I'd love to hear about it!)

I'd asked on twitter: 'Can mixed models make online collections confusing?'  John Coburn suggested that modes of enquiry online might be different, and that the object-driven attributes might be less important.  This was a useful point, not least because it helped me crystallise one reason I find the de-materialisation of objects online disconcerting - attributes like size, weight, texture, etc, all help me relate to and understand objects.  Or as Janet E Davis said, 'I automatically try to 'translate' into the original medium in my head'.   John answered with another question: 'So do we present objects via resonant ideas/themes/wider narrative, rather than jpg+title being "end points"?', which personally seems like a good goal for online collections, but I'm not the audience.

So my overall question remains: is there a potential mismatch between the object-driven approach that exhibitions have trained museum audiences to expect and the object-centred approach they encounter in museum collections online?  And if so, what should be done about it?


  1. Interesting. I may have to write a post to respond to this after I have thought some more about it.
    Something that I would love to do - and needs to be developed a great deal more than it has so far - is the online equivalent of real world exhibitions.
    Even in the analogue world, there has been a distortion in how we perceive objects depicted in still images that can lead to misunderstanding of them.

  2. Very interesting question. As a material anthropologist, I would argue that in both exhibitions and in online collections, we need more of an object-driven approach. The seminal idea here, at least in my discipline, is that objects have "social lives" and so as objects travel, as they inevitably do in our globalized culture, their contexts, and thus their meanings travel.
    So, for example, a plastic doll made in China might simply be a play toy for 5 year old American toddler, but transported in a load of recycled items to Ghana, it might take on a life as an Akuaba (fertility) doll in local culture. Same object, drastically different meanings, though at first look, we'd simply see a plastic doll only through our own cultural lens.
    The point is that context is key, and too often, online collections (and many exhibitions) ignore the context in which an object was created and collected. Museums (especially history and cultural museums) have a reputation in dealing incorrectly in black and white "truths," or fixing objects in too narrow definitions, and an object driven approach is a step in the right direction in many cases. Of course, we still need to account for the fact that object metadata like measurements and materials can be important. And perhaps the more basic object-centered approach to online collections is called for in some cases. For example, you might not always want to engage in storytelling with your online collections. Perhaps the priority at a given moment is collecting tags and basic identifying information about an object from the public who can interact with your object-centered collections online.
    Nevertheless, digital collections and exhibitions are in many ways better suited for an object-driven approach than physical exhibitions (i.e. interactive maps showing an object's journey; displaying related collections or archival items that help contextualize an object; etc.). Then the problem becomes, which story do you tell? Can you tell all of the stories of an objects' life? Do you privilege one story over another? This approach can muddy the waters, so to speak, but when so much of the focus today is to museums creating good storytelling, it seems to demand a shift toward object-driven online projects.
    Thanks for the food for thought!

  3. Thanks for sharing your thoughts! I'd love to see collections contextualised and related to other objects and information as you describe. I've got an old post (Science Museum linked data) elsewhere about how some of that might worked with linked open data, and some of that is slowly happening at the Science Museum, Cooper-Hewitt, Tate, etc... once the technical issues are solved, the bigger challenge is to find resources (inside or outside the museum) to create and validate those links. Relationships and tags raise similar questions to yours about storytelling and privileging different kinds of knowledge - let's hope that that abundance of options is a problem sooner rather than later!

  4. One method for giving the objects contexts is indexing. With a good vocabulary you are able to link everything to its context an so give more meaning to an individual object, link inside your collection, show the bigger issues/topics you would also take as the basis for exhibitions.

    From the point of view of a museum documentalist, I always see the problem of resources. Objects to be used in exhibitions (on- or offline) will be thoroughly researched. The greater part of museum objects (if we do not talk exclusively about museums with old art collections of highest relevance to the academic field) is inventorized on a much lower level. If you want to put objects online, you have to decide whether you concentrate on objects with high-quality-metadata - or whether you show most of what you have; in the last case there will be many objects with little information (those that will never be shown in what in your text is called an object-driven presentation).

    One big question is: who looks at museum objects online? From my personal experience, our average visitor is more likely to look at online exhibitions. Researchers (whether professional or not), students, and our museum colleagues use online databases. And: people finding objects in our online databases may post them, tweet them, etc - and so again create context.

    I very much like the way, Te Papa works with exhaustive indexing terms plus 'topics' linking objects to pieces of more general and elaborated information. One thing they do really good ist linking (as far as I perceive) each and every object depicted on their websites to the collections search site. And I think the "thematic channels" planned by Europeana are really about the same issue - let's hope they will make user studies... concerning the presenation of museum objects online, there seems to be a bitter lack of evaluation.

  5. You're so right about resourcing! I'd always hoped to be able to publish captions/labels from old exhibitions with collections online, partly to take advantage of previous research and partly because I think the changes in interpretation would be fascinating (if sometimes discomforting for the modern reader), but often they're stored outside the Collections Management System in Word docs or design files (or typed up in folders for older ones).

    And Te Papa ( does a lovely job of giving you more places to go after looking at one page.

    I've been doing some work for Culture24's Let's Get Real project ( which includes a survey asking people about their primary reason for visiting a museum site that day, I'm hoping it'll give us some useful insights into who looks at objects online when we match that to actual behaviour on site.

    I'm sure there's lots of evaluation lurking within museums, but it's remarkably difficult to find it. There are some interesting posts discussing audiences and sharing evaluationon the London Museums Group blog, including reservations about sharing evaluation online.

  6. Hi Mia,
    What springs to my mind as I read you here is why collection item webpages don't at least include pingbacks to referring sites.

    Also! Sembl will be an object-driven presentation, a generator of contextual information about "how objects relate to the peoples and cultures that make and use them". It is now seed-funded (I need to update the website :), and I'll be collecting open GLAM data for the in-game array of images – which will of course be linked back to their collection db pages. Doing my bit to connect the drive to the centre!

    So i disagree with the commenter above that ordinary objects "will never be shown in what in your text is called an object-driven presentation". We just need to create killer apps that entice people to make new meanings and connections.