Joho the Blog[ia summit] Donna Maurer on Women, Fire and Dangerous Things - Joho the Blog

[ia summit] Donna Maurer on Women, Fire and Dangerous Things

Donna Maurer of Maadmob Interaction Design has a session explaining to IA’s why they ought to take seriously George Lakoff’s Women, Fire and Dangerous Things, a book basically about Eleanor Rosch’s prototype theory. (I’m about to start writing about this in my book.) It’s a difficult book to read, despite its great title.

Categories, she says, have been taken as being existing/natural divisions that can be expressed in clear definitions. But, Lakoff says (says Donna), categories have prototypical examples. We can organize by similarities to the prototypes without having to define the category.

Donna talks about the concept of basic level categories— you generalize up from them and down to the specifics. BLCs are the level you learn first and usually has a short name, e.g., “dog.” It is the “highest level at which a single mental image can reflect the category.” [In English we can’t form an image of furniture but we can of chair.] BLCs depend on the culture. E.g., city dwellers will say “a tree” while a country dweller might say “a maple tree.”

Donna looked for a web site that organized information by laying out BLCs, but she couldn’t find one. E.g., the eBay site’s category list mixes BLCs with more specific categories.

Prototype theory is important to IAs because the classical theory of categories is built into much software. She says this is her most important point: “Recognize that categories occur and you’ll be less stressed about categorization that is not neat.” Messiness is ok. No scheme is going to include everything and be perfect. (Someone from the audience points out that Sotheby’s “Other” category has an “Other” sub-category.)

Other implications for IAs. Donna suggests using less prototypical items to describe edge cases. And, she says, you can derive BLC names by doing research to see what they are for your users. “Basic level items are easily recognized and likely to have good scent.” Card sort, she suggests, with basic level items rather than more granular content elements. “Get people to the basic level of the hierarchy as soon as possible.”

She wonders whether tags are often BLCs, accounting for their popularity. (Livia from the audience says that they probably are basic level if the taggers are typical of the group.)

Q: Does Lakoff deal with the cultural differences?
A: Definitely. That’s what the title is about.

Q: Folksonomies can help us get away from us imposing a taxonomy on a system.

Q: The same category may be basic level for one group and high level for another…

Q: (Christian Crumlish) When you focus on basic level categories for navigation, where should they be on the site, typically?
A: On the home page, if possible.

Q: (me) How stable are BLCs and what size group do they vary over?
A: They’re pretty stable over time, and the size of the group depens on the domain.

[Excellent talk. Exciting to me to see IA’s taking this stuff up.] [Tags: ]

4 Responses to “[ia summit] Donna Maurer on Women, Fire and Dangerous Things”

  1. Yes, it’s great to see talk about cognitive linguistics on the web front. While the material about categorization is surely useful for IAs, cog ling has much to offer UI designers.

    Designers spend their days choosing between possibilities like this:

    * Turn guest access off
    * Guest access is on (turn it off)

    or even this:

    * This is a message in: Brainstorms
    * This is a message in: Brainstorms

    Cog ling provides a solid framework for understanding the differences in these examples and choosing among them. I recommend this introduction by David Lee.

  2. To ell who have written me asking where to buy phendimetrazine online, I have posted a link so you can buy phendimetrazine or phentermine there.

  3. Eleanor Rosch must have some answer to the question how she arrives at prototypes. Even the most messy visual categorization contains answers to yes-no questions, and fuzziness is even more complicated rules: Can you answer “yes” to more than half of the next group of questions … a) Has or could this animal have had 4 legs and b) mostly fur, c) two eyes d) size less than a human etc. (ending up with a complicated pattern and a conclusion saying “This creature is most likely a dog”.)

    We cannot know anything, according to strict Hume-philosophy, but practice and “likeliness” helps a lot making life possible.

    This is a comment to the “Rosch”-reasoning:

    Categories, she says, have been taken as being existing/natural divisions that can be expressed in clear definitions. But, Lakoff says (says Donna), categories have prototypical examples. We can organize by similarities to the prototypes without having to define the category.

  4. Riguarda online video poker games poker software developer ganar dinero portales afiliados casinos texas holdem download…


Web Joho only

Comments (RSS).  RSS icon