Sources of Power
by Gary Klein
I've been sitting on this book for weeks, waiting for my other projects to quiet down so I could spend the time to review it. It's a really fascinating book about how people make difficult decisions under trying circumstances.
I'm not that familiar with the field of decision making, although it has a lot of apparent overlap with the problem solving studies that I've read (Lave & Wenger, Suchman, Scribner, etc.). In fact, he acknowledges that problem solving and decision making tend to blur in natural settings (p.141).
In this book -- which is well written and accessible for a lay audience despite being based on rigorous qualitative research -- Klein tackles the question: why do people work so well and make such good decisions under difficult decisions? To investigate, he describes several careful naturalistic studies of decision making
- under time pressure,
- in high stakes activities (firefighting, warfare),
- by experienced decision makers,
- with inadequate information,
- when goals are unclear,
- with poorly defined procedures,
- involving "cue learning," the ability to recognize patterns and make distinctions,
- in context,
- under dynamic conditions, and
- in teams
(pp.4-6)
That's a lot of constraints, but together they make for a focused set of studies that cut across several domains and enable Klein to make some strong claims about decision making.
For one thing, he becomes very skeptical about formal decision-making processes (p.29), arguing instead for a renewed valuing of tacit knowledge: "Intuition grows out of experience," he argues (p.33), sounding a lot like researchers who have studied tacit knowledge in education and HCI and workplace studies.
He also argues that "mental stimulation" -- the ability to model consequences of different solutions mentally before trying them out physically -- is a central skill (p.45), although limited by the mind's memory constraints (p.53) and the tendency to rationalize conflicting evidence (p.65). He claims that successful decision makers operating under these constraints tend to come up with a
single strategy for solving the problem; unsuccessful ones often follow a more formal comparative evaluation, which is appropriate for deliberation but not for decisionmaking under these constraints. Similarly, "forcing functions" of the environment lead plan structure: stable environments and plentiful resources lead to complex interconnected plans, while rapidly changing environments with scarce resources lead to modular, limited plans that were more fault-tolerant and segmented (p.145). (Klein doesn't go into the implications for netwar, but they seem obvious.)
Klein identifies eight aspects of expertise: "things that experts can see that are invisible to everyone else." I'll just block quote these:
- Patterns that novices do not notice.
- Anomalies -- events that did not happen and other violations of expectancies.
- The big picture (situation awareness).
- The way things work.
- Opportunities and improvisations.
- Events that either already happened (the past) or are going to happen (the future).
- Differences that are too small for novices to detect.
- Their own limitations. (pp.148-149)
Klein also points out that expert decision makers generate counterfactuals as well: "explanations and predictions that are inconsistent with the data" (p.154).
Not surprisingly, Klein advocates for increasing initiative in organizations (such as military and fire-fighting units) by decreasing central control, pushing decision making out to the people who are most immediately engaged with the decision making context (p.224). He stresses that this means commanders have to learn how to communicate intent while leaving discretion to those who implement that intent (p.225). (This is the distinction between command and control that we're seeing a lot about in netwar and power-to-the-edge approaches.)
Towards the end of the book, in Ch.14, Klein draws some regrettable analogies between team and individual minds and between team development and the development of individual cognition. These analogies are problematic, given what we know about early cognitive development, but present only a minor flaw in the book.
Klein ends in Ch.17 with a discussion of the naturalistic research he conducted and a stirring defense of qualitative research in general.
Overall, I was impressed by the book and would even consider assigning readings from it for a qualitative research class. Klein's clear descriptions, well chosen illustrations, and careful research make for a solid piece of work that sheds light onto decision making and complements the sociocultural literature on problem solving.
