If you haven’t seen it already, the trailer for Cage’s upcoming movie “Knowing” ends in a great line: “What do you do if you know Armageddon is coming and no one believes you?” The premise of the movie is the discovery of a prophetic piece of paper in a 50 year old time capsule that shows every major disaster…but ‘what happens when the numbers run out’? (The characters assume ‘the end’, it could be just the author ran out of paper to write on, or time in the elementary school class that authored it.)
The similarity to events that occur here several times a month is uncanny. Typically, one of the team will return from a meeting with the customer (user, prime contractor, program office) muttering, and pacing…in general showing a generous amount of frustration. The explanation normally centers around something like: “We can see that designing the automation that way will put the user in the ‘Surprise Mode Switch’ dilemma, and that will inevitably lead to an accident!”, or “They want to use their favorite design from a recent post on ManyEyes, but it has a False Emergence to it that will confuse users and will inevitably lead to…” Right, you’re sensing the theme. The good news: The list of these well proven, predictable system design errors is well known to good Cognitive Systems Engineers. For example, they form the core of Woods, Hollnagel 2006: Joint Cognitive Systems, Patterns in Cognitive Systems Engineering. Right, the other shoe…The bad news is that users, traditional systems designers, psuedo-designers, etc. etc. are *not* aware of these patterns, and refuse to accept them. Its a bit like not believing the earth is round, only much harder to prove to the unbelievers (of course that meme took quite some to to become a majority opinion as well).
It seems there are more than a few forces at work that make that the discipline frustrating:
a. traditional processes (e.g. DoD Acquisition Standards, CMMx initiatives, etc) and traditional training programs (e.g. Systems Engineering, Software Engineering, Design Programs, PMP, etc.) still teach ‘the world is flat’…and represent the dominant position among the decision makers/overseers. i.e. it takes an iconclast to accept Cognitive Systems Engineering into their long standing environment.
b. Each generation of system developer, not being exposed to the patterns of these “design disasters in waiting” is left to rediscover each and every one of them.
c. Users will move heaven and earth to succeed at their work, often despite the flaws designed into their tools. Woods calls this a “User Paradox”…so the system developers might never even see evidence of the flaw. (sometimes the traditionally trained accident investigators don’t even see it post hoc! they just arent trained to know what to look for.)
d. Ever changing technologies (e.g. each generation of programming language and environments) promises such great and wonderful things. Those promises distract the focus from what to do with them, and often become ends in themselves (rather than focusing on the characteristics of the tools being designed for users doing real work).
We used to say “Sometimes Cognitive Systems Engineering is like knowing tomorrow’s winning lottery number, but no one will believe you to buy the ticket.” Maybe now we’ll start saying “Being a Cognitive Systems Engineer is like being Nicolas Cage in ‘Knowing’.”
Stumble it!