Monthly Archives: September 2009

Lucy Suchman on framing technology

This is a sort-of rough edit of my live-blogging notes for Lucy Suchman’s talk today at the IT University of Copenhagen. The talk was entitled “Human-machine reconfigurations – expanding frames and accountable cuts”

Lucy Suchman is an antropologist by training, and has worked at the legendary Xerox PARC research facility for many years. Suchman is here to talk about framing, and how we think about the way we frame technology. She presents a photo of a computer screen and an engineer’s hand pointing to something on the screen. This is an example of framing the human-machine interaction. Here, it’s just the hand – the body part interacting with the machine. This frame cuts out a lot of context in order to focus on the specific interactions.

What Suchman wants to draw our attention to, is the way that we make these frames in how we relate and think of technology and our interaction with technology, particularly in relation to research on technology: Where are we to make the cut between human and machine in a given research frame? How do we make those frames? How might we expand those frames? How can we take responsibility for the cuts we make?

She points to the cover illustration of her latest book “Human-Machine Reconfigurations” (2007). The illustration shows a “Device for washing hands” where the framing of the device blurs the boundaries between human and machine – which parts belong to whom?

Suchman says that this illustration is a good way to illustrate the word “reconfiguration”, which she finds to be a vital part of using technology. She cites Donna Haraway’s notion of technologies as ‘materialized figurations’ (from her book, “Modest Witness”). That is: Technologies take part of our activities and practices and materialize them. Configuring a tool to fit with a certain activity or practice.

Designing, then, is most of all a question of reconfiguring the relationship between human and machine, between practice and the materialized figuration of that practice. For instance, between the practice of drilling a hole and the specific drill matching the practice of drilling and containing certain assumptions as to how drilling works.

In all of this, the question Suchman focuses on is “how are persons and things configured and reconfigured in relation to one another? And how might they be figured together differently?”

She uses the example that when roboticists are designing human-like machines, they are expressing their notions of what it means to be human – of human practices – in their design. Suchman wants to show a series of examples of such reconfigurations that she’s worked with while at Xerox.

She shows an age-old magazine ad from back when people still believed in “the paperless office”:

“Why do this…” (picture of paper napkin with the proverbial good idea scribbled on it)
“… When you can do this?” (picture of two persons sitting with a laptop between them at a lunch table)

The ad suggests that people would always prefer the laptop since it offers much more technological power. But rather than assuming the complete displacement of paper technology by digital technology, Suchman and her research associates focused on how to compare the particular affordances of these two media, focusing on the interoperabilities and incompatibilities between the two media. This proved to be a much more challenging and fruitful approach, partly because the relationship between paper and digital media was the central focus of the work at Xerox – it is the “Document company”, after all.

They learned that it is vital to focus on the social arrangements within which design takes place. If you want to change the way things are designed, you have to change the context, offering designers the opportunity to engage in meaningful relations with the potential users.

Suchman presents another example where the Xerox researchers were examining customer complaints in relation to a new xerox copier. The machine proved notoriously difficult to use, and they tried to map the issues people had with these photocopiers by hanging out by the photocopier, talking to users trying to make duplex copies.

But the problems regarding the machine proved too tricky to study “in the wild”, so they ‘captured’ the machine and brought it back into the lab at PARC to test it. So they got their colleagues to try out the new copier, filming their efforts on video. They filmed a memorable sequence of two famous computer scientists failing to get the machine to do duplex copies: “They theorized, and tried their best. Spending an hour and half making prints, filling the room with paper but unable to make a single two-sided print.”

This immense difficulty of using the device stood in stark contrast to Xerox’s own advertising, which remarked “all you have to do is push the green button.” Thus, “the marketing campaign tried to obscure that any learning was required to use the more advanced functions of the machine.”

In the end, Suchman did a careful mapping of user rationale against actual use against the design rationale of the copier to discover how differently the technology had been framed by the designers compared to the users.

Then, Suchman shows a short bit of video from a study of the work flows at ground operations centre at an airport (the place where they handle communication and coordination of aircraft once they’re on the ground).

They did a careful examination of which sources of information the ground controllers consult in order to gather the information necessary to coordinate planes: Video screens, flight tables, radio contact, talking to one another in the control room, and so on.

What they found was that this sort of utilizing different information sources is unremarkable everyday stuff to the controllers. This led to a new understanding of what an information system is:

– multiple, partial information sources
– assembled into a working system through the skilled practices of their use

Their conclusion was that it is only by having the professional knowledge to use a number of partial information sources in conjunction that these information sources became useful. In this way, the information system is a configuration of both information sources and the skilled practices of the controllers. As Suchman’s colleague, the eminent interaction analyst Charles Goodwin, noted in his later work on the study (“Professional Vision” (1994)):

practices … used by members of a profession [that] shape events in the domains subject to their professional scrutiny. The shaping process creates the objects of knowledge that become the insignia of profession’s craft.

A third story from Xerox. Suchman did participant observation at a big law firm in Palo Alto in order to explore how the lawyers used and stored paper records. She set up a video camera pointing at a lawyer’s file cabinet, and asked him to “please record what you do when you use your file cabinet.”

She shows a short video clip with the lawyer going through his file cabinet to find a specific kind of Non-Disclosure-Agreement for one of his colleagues.

Suchman found that the lawyer acted as a librarian, helping the other lawyer find and give context to the specific document that he wouldn’t have had, had he found it on his own in the company online document repository. They also studied the way that lawyers and temporary filing workers worked to code comparable documents.

She was surprised to find that the lawyers considered their coding to be better because it required ‘subjective’ interpretation and professional judgement. They considered the temporary workers’ coding to be poor because of they sought to be ‘objective’ and thus without the necessary interpretation.

In studying how both groups coded the documents, they found that all jobs contain elements of routine work and knowledge work, and it is impossible to simply separate the routine work from the knowledge work. Instead, it is a much more delicate process to find out how to best apply automation in relation to these elements.

As she ends her lecture, Suchman quotes Bruno Latour’s famous passage from Pandora’s Hope on the “gun in the hand”:

You are different with a gun in your hand; the gun is different with you holding it. You are another subject because you hold the gun; the gun is another object because it has entered into a relationship with you.

(Latour, Pandora??s Hope, page 179)

In short: We get different entities when we put technology and people together. She ends by quoting the feminist physicist Karen Barad (in her book “Meeting the universe halfway” (2007)):

Agency is not an attribute but the ongoing reconfigurings of the world.

– which obviously matches the points that she’s been making very well.

Q: What are the big challenges in your work today?
A: Still very interested in Artificial Intelligence, including the reconfiguration of military technology such as pilot-less aircraft and their interfaces. The military and entertainment complexes are growing together in this space.

Also: Remotely controlled robots for use in surveillance, sentry duty and so on. which poses the interesting AI-question: How do you determine whether a given person is a friendly or a un-friendly?
We won’t see the full-blown autonomous robot soldier anytime soon, but the remote controlled robots will certainly be possible and big part of the near-future.

Also: Wants to write about her time at XEROX Parc and what she’s learned about what innovation is based on her experiences there: What constitutes an innovation? She posits that it is all about the framing…

Bootstrapping complexity

So, last week I posted my remix of Kevin Kelly’s book “Out of Control”. And soon after putting the remix online, I sent a note with a link to Kevin Kelly to make him aware of the remix, hoping that he would approve.

He did approve. Much more than I expected. And it didn’t take him long to reply:

I LOVE the remix! I wish you had been my editor. There is only one thing missing from this fantastic remix – a better title. I was never happen with the book’s title and now that it is more focused, the need is even greater. What would you call it?

Whoa! Initially, I hadn’t considered changing the title as I wanted to make it as clear as possible where the material came from. Good titles are notoriously difficult to find, and I’m sure that Kevin has thought quite a bit about this one.

Considering the remix as a new whole work, I found that it was the notion of bootstrapping and self-organization that had kept me reading the book initially: the recurring patterns of self-sustaining systems, which I knew were to be summed up at the end of the book. What appealed to me was the fact that the book not only describes self-organisation but also invites further experimentation.

So I picked my title with that in mind: “Bootstrapping Complexity” plays on the fact that the book not only describes how complexity comes about but also how complex a venture self-organization really is. In this way, the title meant to signal a positive empowerment to explore self-organization – both by reading the book and by experimenting on the basis of the book.

I’ve updated the remix with the new title. The new PDF version is here.

Out of control – remixed

This summer, I read Kevin Kelly‘s book Out of Control. It is a fascinating book full of fascinating ideas reaching across the board from artificial intelligence, evolution, biology, ecology, robotics and more to explore complexity, cybernetics and self-organising systems in an accessible and engaging way.

But as I read the book, I also found it suffering from a number of frustrating flaws: Not only is it way too long-winded, it is also almost completely void of meta-text to help the reader understand what Kelly is trying to do with his book (having read the book, I’m still wondering).

Indeed, reading the book I got the feeling that Kelly was trying to combine several different books into one: There is a fascinating study of self-sustaining systems. But there is also a sort of business-book take on network economy. And an extended meditation on evolution and postdarwinism.

I’m sure that to Kelly, all of these things are tightly interconnected. But he doesn’t explain these interrelations very well to the reader. His central argument is that as technology becomes ever more complex, it becomes more akin to biological systems (eco-systems, vivisystems, interdependent and co-evolving organisms). But because the individual chapters are set up as essays on their own, there is often little to tie these wildly different ideas together.

I would have preferred a much shorter book, more narrowly focused on the idea of self-organising systems. The whole text of the original book is easily available online at Kelly’s own website, so I thought: Why not remix the online text to make such a book?

So I did.

I’ve put my remix up here. The PDF version is available here. Comments are most welcome.