Source
Feltovich, P. J., R. R. Hoffman, D. Woods, and A. Roesler. 2004. “Keeping It Too Simple: How the Reductive Tendency Affects Cognitive Engineering.” IEEE Intelligent Systems 19 (3): 90–94.
TL;DR
There are eleven axes of “understandability”, or alternatively “complexity”, that tempt reductivist interpretations. Reductionism leads to suboptimal complex sociotechnical systems (CSS) design and mistaken assumptions about perfect substitutability between humans and automation. We should instead consider these dimensions when designing CSSs and proactively counteract reductivism.
Summary
Cognitive Flexibility Theory research proposed eleven axes of “understandability”:
- Static vs. dynamic
- Discrete vs. continuous
- Separable vs. interactive
- Sequential vs. simultaneous
- Homogeneous vs. heterogeneous
- Single vs. multiple representations
- Mechanism vs. organicism
- Linear vs. nonlinear
- Universal vs. conditional
- Regular vs. irregular
- Surface vs. deep
The phenomena towards the second option are complex and encourage new learners to employ reductive simplifications. They support a rejection of complexifying information with defensive behaviors (demean effect, argument from faulty causal reasoning, extirpation). Reductionism also leads to shallow and inaccurate conclusions.
What does this mean for understanding Complex Sociotechnical Systems (CSSs)? CSS are both (1) continually changing via new technologies and the Law of Stretched Systems and (2) every participant is concurrently modeling the work and the other parties involved. So there are always differing perspectives on goals, urgency, scope, etc.
For CSSs, the reductivist tendency challenges understanding across the previous eleven axes. For instance:
Homogeneous/heterogeneous
The reductive tendency would result in assumptions that the processes, values,ways of doing things cultural norms, abilities, loyalties, and so forth are pretty much mathe same across the many diverse units ofthe CSS—a kind of uniformity tendency. This reduction’s effect in design would be that the cognitive engineer fails to anticipate the diversity of reactions and adaptations to a workplace change.
From a more conceptual level, reductionism makes perfect substitutability between human actors and devices seem possible (How to Make Automated Systems Team Players, MABA-MABA or abracadabra? Progress on human-automation coordination). This mistaken assumption undergirds predictions about “replacing” humans with AI (IBM plans to replace 7,800 jobs with AI over time, pauses hiring certain positions ).
While the roles may be filled by AI, the work system and the remaining human actors will be changed in unforeseen ways. Transferring responsibility shifts the entire work system.
We should use awareness of the reductivist tendencies when designing CSSs. Specifically, build “tools that help operators comprehend the implications of numerous, interdependent, and constantly changing variables that affect successful execution of a task.”
Notable Quotes
When learners are confronted with evidence contrary to their views, they perform mental maneuvers to rationalize their faulty beliefs without fundamentally altering them. These protective operations are calledknowledge shields, and researchers have identified 23 of them.
References
Attributions
Photo by Florian Klauer on Unsplash