Interactions - September/October 2008

A Concern for Users

Austin Henderson and Jed Harris have noted that many computer systems are constrained by a mechanistic worldview [ 6]. They cite automation projects avoiding errors by drastically reducing options available to users (narrowing language or variety), but in the process crippling communication and organizational flexibility. Henderson and Harris contrasted coherent systems to responsive systems. Coherent systems require consistency and predictability; responsive systems support messiness and improvisation. “In a given system, as responsiveness increases, coherence tends to decrease and vice versa—a classic trade-off. Scaling makes this trade-off sharper. As systems get larger, they have to work harder to maintain their coherence, and this increasingly makes them unresponsive. Conversely, large systems that allow great local responsiveness (such as the World Wide Web) have difficulty maintaining coherence [ 6].”

Henderson pointed out that consistency is an ideology, that other choices are possible: “the core ideology of computer system design is totally permeated with the assumption that computers are rule-following machines, and more generally, that all human activities can and should be described in terms of a consistent set of rules [ 7].”

He argued that “feedback loops…actually make organizations work, and the constant negotiation that these loops entail…computing systems tend to break those loops…so people have to bear the brunt of patching them up, and usually have to fight the computer system to do it.” Henderson and Harris proposed a new approach, which they described as “pliant computing.”

computing” is a deep concern for people who use computers. Henderson sees the relationship between designer and audience changing. As Rheinfrank pointed out, the designer is moving from detached expert to collaborator. And the relationship between designer and constituent is moving from expert-patient to what Horst Rittel called “a symmetry of ignorance (or expertise) [ 8]” in which the views of all constituents are equally valid in defining project goals.

[ 6] Henderson, Austin, and Jed Harris. “A Better Mythology for Computing,” Presentation at CHI 99, Pittsburgh, Pa., May 1999.

Follow Design FOR users Provide input

Participate Design WITH users Combine expertise

Lead

Design BY users Build on:

Scripting Languages

Open Systems Construction sets

Provide feedback

Combine values

Adapted from Austin Henderson [ 7]

[ 7] Henderson, Austin. “Design for Evolution.” Presentation at HITS Conference, II T/ID Chicago, October 2003.

Liz Sanders presented a similar argument with slightly different eras, explicitly introducing the idea of moving beyond human-centered or user-centered design [ 9].

Past

Design Expert

Paradigm driven

Audience Customer User Role

Activity

Current

Human centered

Emerging

Facilitated

Participant

[ 8] Rittel, Horst. “On the Planning Crisis: Systems Analysis of ‘First and Second Generations.’” Bedrifts Økonomen 8 (1972): 390–396.

Consume Shop

Experience Use

Buy Own

Interact

Communicate

Co-create

Adapt/modify/ extend

Design

Make

Adapted from Liz Sanders

Coherent
Rigid
Fragile
Regular
Thin descriptions
Designed by designers
in advance of use by users
enforcing a single view

Responsive

Pliant

Robust

Particular

Thick scenes

Created by participants during use enabling multipile views

Adapted from Austin Henderson

Codevelopment is also a fundamental tenet of open-source software. Eric Raymond wrote, “Treating your users as co-developers is your least-hassle route to rapid code improvement and debugging.” He added, “Even at a higher level of design, it can be very valuable to have lots of codevelopers random walking through the design space near your product.” Raymond famously contrasted “cathedrals carefully crafted by individual wizards or small bands of mages working in splendid isolation” to “a great babbling