Based on the sources, Max Boisot’s ideas—specifically his focus on codification, abstraction, and socio-computational methods—stand in distinct conflict with several other authors who view abstraction as a hindrance, complexity as non-computable, or randomness as the definition of complexity itself.
Here are the specific ways Boisot’s ideas contradict or conflict with the other sources:
1. Abstraction: Tool vs. Trap (Boisot vs. Wilk & Pirsig)
Boisot argues that managing complexity requires moving up the ladder of abstraction to filter information, whereas others argue one must climb down.
• Boisot: Advocates for “Complexity Reduction” (a cognitive strategy) which relies on “codification and abstraction” to filter out noise (“crude complexity”) and focus on regularities[1]. He views the creation of abstract representations as essential to reduce the processing load on the system[1].
• The Conflict:
James Wilk: Directly contradicts this by arguing that “mid-level abstractions” are “conceptual smokescreens” that confer no insight[2],[3]. Wilk advises practitioners to “climb down the ladder of abstraction” to reach “video descriptions” (concrete, uninterpreted facts)[4].
Robert Pirsig: Similarly values the specific and concrete (the “bolt”) over the abstract, advising a “scaling down” of scope to concrete details when stuck[5].
2. Computability: Hybrid Processing vs. Non-Computability (Boisot vs. Relational Biologists)
Boisot believes technology and computation are key to managing complexity, while the biological theorists argue complexity is fundamentally beyond computation.
• Boisot: Proposes “Hybrid Socio-Computational Approaches,” combining human agents with “high-speed parallel computing”[6]. He suggests using technology to scan for weak signals and reduce “trillions of possible patterns”[6].
• The Conflict:
Relational Biologists (Rosen, Pattee, Noble): Define a system as complex only if it possesses “noncomputable or non-formalisable models”[7]. They argue that complex systems cannot be fully captured by syntactic algorithms or simulated by Turing machines[7]. Therefore, Boisot’s reliance on computational pattern recognition would be seen by them as dealing with “complicated” systems, not truly complex ones.
3. The Definition: Regularity vs. Randomness (Boisot vs. Abel)
Boisot’s distinction between types of complexity conflicts with the hard mathematical definitions used by information theorists.
• Boisot: Distinguishes between “crude complexity” (random noise) and “effective complexity” (regularities underpinning a structure)[8]. He implies that “effective complexity” is the target of management.
• The Conflict:
David L. Abel: Explicitly defines “maximum complexity” as “randomness” and the lack of order or pattern[9],[10]. For Abel, if a sequence has “regularities” (which Boisot seeks), it moves toward “Ordered Sequence Complexity,” which is low-information[10]. Abel argues that complexity is “antithetical to order”[10], whereas Boisot seeks the order within the complexity.
4. Logic: Abduction vs. Deduction (Boisot vs. Dettmer & Triz)
Boisot advocates for a fluid, inferential logic, contrasting with the rigid deductive logic proposed by the engineering sources.
• Boisot: Advises using “Scalable Abduction”—an inferential engine that tracks “butterfly events” to anticipate how they might amplify[6]. This is a logic of hypothesis and “best guess” rather than proof.
• The Conflict:
William Dettmer (Theory of Constraints): Relies on “Destructive Deduction” and “Creative Induction” to build rigorous Logic Trees (Current Reality Trees)[11],[12]. Dettmer seeks to trace “root causes” through strict causality[12], whereas Boisot aims to track emergent patterns. Triz: Relies on “Mathematical Objectification” and “tensor calculus” to determine solutions with “exact mathematical precision”[13].
5. Reduction vs. Holism (Boisot vs. Cilliers & Ackoff)
While Boisot advocates “Complexity Reduction” as a valid strategy, holistic thinkers warn that this destroys the system’s essence.
• Boisot: Promotes “Complexity Reduction” to economize on resources[1]. While he also allows for “Complexity Absorption,” he validates the reductionist strategy of simplifying incoming stimuli[1].
• The Conflict:
Paul Cilliers: Argues that if you take a complex system apart (or simplify it to a model), you “destroy the very interactions that define it”[14]. He warns that “framing” inevitably leaves things out, and because of non-linearity, we cannot predict the impact of those exclusions[15].
Russ Ackoff: Argues that the traditional impulse to “cut problems down to size” (reduce/simplify) is “simple-minded” because it ignores the interactions that are critical to performance[16],[17].
6. Intervention: Anticipation vs. Retrospection (Boisot vs. Snowden)
Boisot believes in the possibility of anticipating extreme outcomes, while Snowden argues that understanding usually comes too late for prediction.
• Boisot: Suggests that scalable abduction allows for “anticipation,” enabling agents to track tiny initiating events before they become extreme macroscopic outcomes[6].
• The Conflict:
Dave Snowden: Argues that in complex domains, cause and effect are only visible in retrospect[18],[19]. He contends that we cannot forecast outcomes or identifying specific butterfly effects in advance; we can only “probe-sense-respond” to what is currently happening[19].
Summary of Conflicts
| Topic | Max Boisot’s Idea | Conflicting Source |
|---|---|---|
| Abstraction | Use abstraction/codification to filter and manage complexity[1]. | Wilk/Pirsig: Abstractions are “smokescreens”; climb down to concrete facts[4],[20]. |
| Computing | Socio-computational: Use parallel computing to process patterns[6]. | Relational Biologists: Complexity is non-computable and cannot be simulated[7]. |
| Definition | Effective Complexity (regularity) is distinct from noise[8]. | Abel: Complexity is randomness/noise; order is antithetical to complexity[10]. |
| Logic | Abduction: Inferential tracking of weak signals[6]. | Dettmer/Triz: Rigorous Deduction and mathematical precision[13],[11]. |
| Forecasting | Anticipation: Track “butterfly events” to see amplification[6]. | Snowden: Causal links are only retrospective; prediction is impossible[18]. |
References
[1] Max Boisot.md [2] James Wilk.md [3] James Wilk.md [4] James Wilk.md [5] Robert Pirsig.md [6] Max Boisot.md [7] Relational Biologists - Robert Rosen Howard Pattee Dennis Noble.md [8] Max Boisot.md [9] David L. Abel.md [10] David L. Abel.md [11] Theory of Constraints - William Dettmer.md [12] Theory of Constraints - William Dettmer.md [13] Triz.md [14] Paul Cilliers.md [15] Paul Cilliers.md [16] Russ Ackoff.md [17] Russ Ackoff.md [18] Dave Snowden.md [19] Dave Snowden.md [20] Robert Pirsig.md