Archive for August, 2011

Cultures of Participation Symposium

August 30, 2011 Leave a comment

Gerhard Fischer held a symposium as part of his evaluation for his NSF CDI grant–a great group of people attended, gave feedback, and spoke (Barbara Ragoff from UCSC, Shirley Brice Heath from Stanford, and locals Alex Repenning and Michael Eisenberg). Members of the project participated too: Jason Zietz and Hal Eden gave demos. A great, integrative effort toward enhancing cultures of participation, with an exemplar problem of sustainability and smart use of electricity). Pulling together a diverse group can give feedback you wouldn’t typically expect–it was enlightening to see the tensions between the builders, the number crunchers, and the ethnographers. It seems difficult to alter significantly the course of the grant at this stage, but there’s always opportunities for more grants.

(And thanks WordPress for trashing the longer version of my writeup…I’m now hunting for another blogging tool.)

Categories: Uncategorized

Paul Otlet: The pre-Bush Vannevar Bush

August 24, 2011 1 comment

Folks in CS and HCI love to point to Vannevar Bush as the father of all things Internet and Web related. But even fathers have fathers who set the stage for not-ready-for-prime-time ideas. Paul Otlet can be viewed in that role: a visionary from the early 1900s with a functional search engine–and a plan for an electronic interface to visualize the interconnections in knowledge.

In the early decades of the 20th century, Otlet established a “Mundaneum” (yes, an unfortunate name), a knowledge repository consisting of millions of facts drawn from the literature of the time. Want to Google something in the early 20th century? You could “Otlet” it–mail in your request with a small fee, and get back one or five or twenty or fifty “hits” in response, each on a 3×5 card. He envisioned indexing all knowledge across different media, with access through a phone/screen interface supporting the comparison of many interrelated sources.

To me, a key element of Otlet’s vision resides in knowledge capture not at the level of a document, but at the level of the individual information chunks within the document. It is at this level that Otlet created interconnections between knowledge, such that no one piece of knowledge was an end point but rather a hub to other information. He classified information chunks into four categories: facts, interpretation of facts (opinions?), statistics, and sources. The chunks were linked and ranked in various ways–by topic, by public opinion, by Universal Decimal Classification–into a linked web-like “rĂ©seau” that aspired to be informational and social.

There’s a number of good Otlet and Otlet-related books and papers out there. A couple that appealed to me: NY Times editor Alex Wright’s book “Glut” has a large section overviewing Otlet’s work (in the context of other information management efforts) that draws from several of his NYT pieces, and Boyd Rayward has written a number of books and articles on Otlet (much of it pre-internet, though check out his “Visions of Xanadu” article). And there’s material on Wikipedia on Otlet and the Mundaneum that is a great starting point.

Alas, Otlet was all the rage in his home country of Belgium for a while, but died in poverty in 1944 after watching his collection be shifted into increasingly smaller quarters then be carted away by the Nazis. He was all the rage for several decades though (much longer than, say, Google has been around!). The Mundaneum collapsed in part due to the explosion of information that was to be cataloged. And before you smugly say that such a thing will never happen in this digital age to the Googles of the world, consider whether it’s truly possible to index and cross-reference and meaningfully collect user opinion on every word and phrase and pixel and individual utterance in every document and video and security camera and satellite image and telescopic image that is collected. And if that’s not possible, where do we draw the line? And if there’s a line, there will surely be a next technology–the next Google–to surpass what the current one can do.

What Toulmin claims

August 19, 2011 2 comments

The notion of a claim as a key element in design rationale entered the field of interaction design and human-computer interaction through Carroll and Kellogg’s 1989 “Artifact as theory nexus” CHI paper. In that paper, they describe a claim as the psychological effects (positive and negative) that are caused by an artifact. They outline how claims can be used to guide design through the development and understanding of goals, plans, actions, and evaluation.

But certainly the roots of design rationale predate this artifact paper. Notably, Stephen Toulmin’s classic The uses of argument–the book that several researchers (e.g., Alistair Sutcliffe and Janet Burge) point to as the seminal point in argumentation and design rationale–introduces the notion of a claim as a key part of argumentation. Similar to Carroll/Kellogg, Toulmin spends large stretches of his book noting that claims are interesting because they are falsifiable–they may be true, but perhaps not, and must be re-evaluated as context changes. For example, on p.220 Toulmin makes the summative statement that “we may accept over-hastily the suggestion that a claim to knowledge that proves mistaken must have been an improper claim” as the benefit comes from the creation and analysis inherent to claims. He criticizes our (very human) need for “a God’s-eye view” that is universally true and instead encourages us to embrace the uncertainty as method. So does this apply to claims in the Carroll/Kellogg sense? Short answer: yes; longer answer: maybe (I’ll have to think about it).

According to the preface to the 2003 edition, the Toulmin book originally was meant to “criticize the assumption…that any significant argument can be put in formal terms”. Instead, he ended up with a work that “expound[ed] a theory of rhetoric or argumentation” in a way he calls “informal” but to me seems more analytical. Other books operationalize Toulmin’s concepts in more digestible form–my favorite is The craft of research by Booth, Williams, and Colomb. It’s somewhat ironic that it seems people now view the Toulmin method (and the methods based on it) as too formal and rigid—but it’s quite influential to the notions behind Carroll’s scenario-based design and the IBIS-gIBIS-Phi-QOC-Compendium methods and tools that I discuss in another post.

Alas, Toulmin is now dead. I mention that because he hasn’t been dead for long (since 2009)–though many people assume that the person who wrote the 1958 classic “The uses of argument” has been dead for many decades. In the preface of the 2003 edition of his book (the one I used in this post), he notes as much through a great story…one of many, it seems. Just as others built on and operationalized his work, he points to Aristotle, Descartes, and others as the masterminds behind the concepts he espouses. So add them to the reading list!

Lessons from mobile interfaces class

August 16, 2011 Leave a comment

I just turned in grades for my 5-week mobile interfaces class–CSCI 4830/7000–a cross-listed undergrad/grad course offered during the “Summer B” term at UC-Boulder. It counted as a full 3-credit class, but it was crammed into 3 hours a day, 3 days a week–whew! Some thoughts I want to capture:

– Big thanks to Google for supplying the hardware we used in the class through their University Consortium program. Google is following the path to success that Microsoft followed in the 1980s and 1990s–lots of free stuff for universities to aid in teaching, with the payoff of knowledge and experience with their products and software when students graduate. More broadly, they are doing better than Microsoft of old in that they are providing lots of forums and development environments for free. Ironically, Microsoft has turned into Apple (and Apple has remained Apple-like) when it comes to their mobile platform, with a more closed platform and few/no inexpensive hardware options…and it doesn’t seem to be turning out well for them.
– Not sure if any course is a great fit for such a compressed time period, but mobile interfaces (with a programming emphasis) is better than most: lots of small projects that are individually useful but (if the prof does a good job) collectively meaningful. That’s a great side effect of the current state of things in mobile platforms: small standalone apps are in demand, so you can do something meaningful in a short period of time.
– Projects are nice, but big group projects in a summer class don’t make much sense. Doing it in the future in a month or less? Ditch the group project, or make it an extended/expanded homework.
– I like encouraging/requiring the students to publish to the Marketplace, though again 5 weeks is a tight turnaround. But my current thought is to keep it. It was great to look back and see the success that DeMarcus had with his tip calculator app–1000+ downloads, most still active–and we’ll see if anyone from the summer group can replicate that success.
– I liked having an assignment for which students presented on their favorite platforms. It was good to learn about Blackberry and iPad–but the Magic Cap presentation provided an interesting historical perspective on the development of mobile interfaces and allowed us to compare past and future. I’d be tempted to do more platform presentations (perhaps requiring every group to do a historically important platform and a modern one).
– Amateurs don’t write reusable claims about the apps they develop…and why should they? It’s hard to capture the key design lessons in a way that others will be able to learn from it. It’s a classic CSCW problem: there’s not much benefit for the person writing the claim, even if (big if) there is benefit for people reading them down the road. We need a claims goddess to polish them up and make them presentable!
– It was good to have the grad/undergrad mix, though there needs to be more mixing. The grad students should do some programming, maybe with App Inventor or Droid Draw, to understand better the challenges. The undergrads should do more writing, to encourage them to reflect on their programming.
– Avoid professional paper format for written submissions–the papers won’t be published in anything resembling the form in which they are turned in. And professional paper formats are tough to read, IMO.

Next up: a Maymester offering? Or a cross-listed semester-long offering at VT?