From: jimruttshow8596
Dave Snowden, founder and chief scientific officer of Cognitive Edge, has pioneered a science-based approach to understanding and decision-making in complex issues, drawing on anthropology, neuroscience, and complex adaptive systems theory [00:00:34]. His work emphasizes the critical role of narratives in making sense of the world and informing action, particularly in uncertain environments.
Narratives vs. Traditional Knowledge Management
Snowden reflects on his past work in knowledge management at IBM in the 1990s, where he found that many attempts focused on “codification,” meaning stuffing information into databases [00:02:05]. This approach was often naive and over-codified processes that were not easily reducible [00:03:08]. He critiques this “naive Newtonianism,” a belief that with enough data, the future can be predicted, much like a 13-year-old’s understanding of physics [00:03:22].
Instead, Snowden and his colleagues argued that decision support was far more complex, requiring consideration of cognitive neuroscience and the actual ways people make decisions [00:02:17]. This realization led to work on narrative and complexity theory, which became central to programs like those at DARPA, focusing on weak signal detection and decision-making in complex policy environments [00:02:27].
Micro-Narratives and Self-Interpretation
Snowden defines narratives as the “watercooler stories,” the tales told at the school gate or in a checkout queue, which truly determine people’s attitudes rather than responses to questionnaires or surveys [00:58:43]. These day-to-day “micro-narratives” are critical because they carry ambiguity [00:59:28].
A key aspect of Snowden’s approach is giving individuals the power to self-interpret their own narratives, rather than relying on algorithms or experts to do so [00:59:04]. This process allows for scaling to very high volumes of data quickly [00:59:13]. The self-interpretation of narratives generates valuable “human metadata” that algorithms alone cannot capture [01:12:00]. This human insight is crucial because humans are good at processing chaos and making collective decisions [00:36:39].
Applications of Narrative-Based Approaches
Weak Signal Detection and Anticipatory Triggers
One significant application is in weak signal detection, such as for counterterrorism [00:29:39]. Narratives can help create “anticipatory triggers” – signals that prompt a heightened state of alert when an event (like a terrorist outrage) is more likely, even if direct prediction isn’t possible [00:29:54]. This technology is being adapted to identify plausibility of abuse in elderly care homes [00:30:00].
Mapping Organizational Culture
When considering mergers or acquisitions, narratives can be used to map the cultures of the two organizations, revealing commonalities or overlaps [00:37:04]. This helps understand underlying cultural attitudes and beliefs [00:29:31].
Cybersecurity and Employee Attitudes
Narratives are used to measure employee attitudes toward cybersecurity, rather than just compliance [00:46:04]. By presenting an infographic of a cyber-security breach, employees interpret it in real-time, providing their situational assessment and a micro-story about the future [00:46:11]. This reveals the attitudes and allows for identification of dispositional states and attitudinal sites [00:46:52].
Community Understanding and Agency
Snowden envisions creating a global network where 16-year-olds act as journalists in their own communities, producing weekly or monthly journals [01:00:18]. This would generate a human interpretation of people’s day-to-day lives, allowing for horizontal integration of ideas and informing policy [01:00:47]. This approach aims to give people agency in their own conditions [01:00:51].
Therapeutic Device
Narratives are also being used as a therapeutic device to help individuals understand and potentially escape narcissistic control in partnerships [01:07:05].
Political Polling and De Novo Parties
The principles of narrative collection and interpretation could be used as a replacement for traditional political polling to find shared values in society or inform the creation of new political parties [01:03:34]. Field workers could gather stories, and people could self-interpret them, removing reliance on expert mediation [01:04:18].
Enhancing Doctrine and Best Practices
By embedding HTML links to real stories within best practice documents, narratives can provide richer context and help people interpret data more effectively [00:59:53]. Narrative-based search can also offer better access to documents [01:00:03].
Distinction from Traditional Methods
Complicated vs. Complex
Snowden emphasizes a fundamental distinction:
- Complicated systems: The sum of their parts; problems can be solved by breaking things down [01:13:11]. These systems tend to be engineered and predictable (e.g., a car engine) [01:13:56].
- Complex systems: The properties of the whole result from interactions between parts, linkages, and constraints, which are not fully known [01:04:06]. In human systems, something beneficial one day can become something else [01:14:49]. Understanding these systems requires probing and parallel experimentation because causes and effects are not self-evident [01:04:09].
Traditional methods often assume a linear approach to causality [00:03:46], assuming inputs define outputs or that futures can be forecast [00:03:55]. However, complexity theory (different from deterministic chaos) highlights that in human systems, one can understand the present and map coherent pathways, but cannot define a specific outcome [00:04:09].
Limitations of Simulation and AI
While tools like agent-based modeling and simulation can be useful for understanding aspects of a system or gaining insight, they often struggle with human systems due to multiple identities and patterns not always based on rules [00:27:25]. Snowden warns against confusing simulation with prediction, or correlation with causation, a mistake some AI practitioners are making [00:27:40]. Murray Gell-Mann famously stated that “the only valid model of a human system is a system itself” [00:28:20].
SenseMaker Software
SenseMaker is a platform designed to gather and analyze micro-narratives [01:01:15]. Unlike traditional employee satisfaction surveys that ask leading questions, SenseMaker asks non-hypothesis questions, such as “What story would you tell your best friend if they were offered a job in your workplace?” [01:02:04].
Individuals then self-interpret their own stories by positioning them on a series of “triads” (triangular signifiers) [01:02:15]. For example, a triad might balance “altruistic,” “assertive,” and “analytical” qualities of a manager’s behavior [01:02:20]. This process:
- Triggers a shift to “novelty receptive processing,” forcing deeper thought [01:02:31].
- Adds multiple metadata points to the original narrative (e.g., six triangles add 18 metadata points) [01:02:45].
- Allows the original narrative to be carried with statistical data, explaining what patterns mean [01:02:56].
- Generates results instantly, enabling real-time feedback and analysis of attitudes or performance (e.g., a 360-degree feedback system where direct reports record and index interactions) [01:05:27].
This process allows for the creation of “narrative landscapes” or “fitness landscapes” that show underlying cultural attitudes, dominant views, and critically, outlier views that might offer different perspectives [00:26:08]. These outliers, who think differently, are often valuable to hunt down for new insights [00:26:00].
Coherence, Dissent, and Diversity
Snowden advocates for “coherent heterogeneity,” meaning maintaining differences that can come together in various ways, rather than homogeneity [00:48:14]. He argues for actively encouraging dissent and diversity within organizations [00:47:30].
SenseMaker’s attitude mapping can measure cognitive and behavioral diversity within an organization, identifying outlier groups worthy of attention, preventing middle management from drowning out crucial perspectives [00:48:47]. When conflicting hypotheses for action arise, rather than trying to resolve them, Snowden suggests constructing “safe-to-fail micro-experiments” around each coherent hypothesis and running them in parallel [00:11:36]. The solution then starts to emerge [00:11:44].
To manage useful diversity versus “crankery,” Snowden’s methods use statistical analysis on clusters of narratives to assess “coherence” [00:50:02]. If there’s tight clustering, coherence is high (possibly overly so, reducing diversity and making systems non-resilient); if there’s little clustering, coherence is low [00:51:50]. The goal is to identify ideas coherent enough to explore, even if different, and distinguish them from nonsense [00:50:02]. The appropriate amount of diversity is situationally dependent: stable ecosystems need less, while destabilizing systems need more for exploration [00:52:25].
Ethical and Aesthetic Considerations
Snowden believes that engineers and those designing tools for human systems should be trained in ethics and aesthetics [00:40:02]. Aesthetics, in this context, relates to abstraction, which is fundamental to human language and allows for “rapid exaptive thinking” – the ability to quickly repurpose things [00:40:12]. Stories of survival and decisionmaking | Parable form stories, as abstractions, provide better moral guidance than values or principles because they define what not to do (negative constraint) [00:40:37]. Understanding or appreciating beauty can make one a more effective decision maker [00:40:47].
Conclusion
Narratives offer a powerful lens for understanding and decision making in complex human systems. By moving beyond traditional, linear approaches and embracing the inherent ambiguity and self-interpretation of human stories, Snowden’s methods, exemplified by SenseMaker, provide real-time insights into attitudes, identify crucial weak signals, and foster resilient, diverse organizations. This approach, rooted in natural science principles, aims to create more humane societies by increasing human agency and informing policy through granular, bottom-up understanding.