# AI Wiki - PKM - Systems Thinking **Systems thinking** is a discipline for seeing wholes — recognizing that components are part of interconnected systems whose behavior cannot be understood by examining the parts in isolation. It originated mid-20th century from cybernetics, biology, and operations research, and remains foundational for handling complexity in any domain — including PKM. ## Core Distinction Reductive thinking asks: "What is this thing made of?" Systems thinking asks: "How does this thing relate to other things, and what behaviors emerge from those relationships?" Both are useful. But systems thinking gets emphasis when: - Behavior is **non-linear** (small inputs produce large effects, or vice versa) - **Feedback loops** drive outcomes more than initial conditions - **Emergence** appears (whole-properties not in parts) - **Time** matters (delays, stocks, flows) - **Multiple causes interact** rather than one cause producing one effect These are the conditions Morin calls "complex" — see [[AI Wiki - PKM - Complex Thinking]]. ## Key Concepts ### Stocks and Flows A **stock** is something that accumulates (water in a bathtub, notes in your vault, money in an account). A **flow** is the rate of change (water in/out, notes added/removed). Most complex behavior comes from how stocks and flows interact over time, not from single events. PKM example: your vault is a stock of notes. Your reading habit is an inflow. Your pruning practice is an outflow. The stock's dynamics depend on the *ratio* of inflows to outflows, not just on either alone. ### Feedback Loops [[AI Wiki - PKM - Feedback Loops in Knowledge Systems]] are how systems regulate themselves: - **Reinforcing loops**: outcomes amplify (more notes → more links → more discoveries → more notes) - **Balancing loops**: outcomes self-correct (capacity limits, fatigue, review backlog) Most system pathologies come from missing or broken feedback, not from "wrong" parts. ### Delays Effects rarely follow causes immediately. Notes you write today may not reveal their value for months. Bad system design ignores delays and produces oscillation, overshoot, or premature judgment. ### Leverage Points Donella Meadows' classic taxonomy: not all interventions are equal. From weakest to strongest: 1. Constants and parameters (often what people focus on, low leverage) 2. Buffers and stocks 3. Structure of feedback loops 4. Rules of the system 5. The system's paradigm (highest leverage) For PKM: changing your *paradigm* of what notes are for is far higher leverage than changing your tagging convention. ### Emergence [[AI Wiki - PKM - Emergence in PKM]]: the system has properties no part has. The vault as a whole has voice, coherence, gaps — none of which any single note possesses. ## Systems Thinking vs Complex Thinking These are siblings, not equivalents: | Systems Thinking | Complex Thinking (Morin) | |---|---| | Engineering / management origin | Philosophical / epistemological origin | | Stocks, flows, feedback, leverage | Dialogics, holography, auto-eco-organization, reflexivity | | Often quantitative | Often qualitative | | Focuses on system behavior | Focuses on knowing about systems | Systems thinking gives you tools to model and intervene in systems. Complex thinking gives you epistemology for thinking *about* knowledge of systems. PKM benefits from both. ## Application to PKM ### See the Vault as a System Not as a folder of notes, but as a **stock-and-flow system with feedback loops**: - Inflows: capture, reading, conversations - Outflows: pruning, retirement, sharing - Stocks: notes, links, tags, identity - Feedback: review rituals, spaced repetition, periodic audits Most vault problems come from imbalances in these dynamics, not from bad individual notes. ### Diagnose Pathologies Systemically Common pathologies and their systemic roots: - **[[AI Wiki - PKM - Collector's Fallacy]]**: high inflow, low outflow, no synthesis loop - **[[AI Wiki - PKM - Knowledge Decay]]**: no maintenance feedback, stocks become stale - **[[AI Wiki - PKM - Information Overload]]**: inflow exceeds processing capacity - **[[AI Wiki - PKM - Research Purgatory]]**: capture loops dominate; synthesis loops never fire ### Design for Feedback Build rituals that close the loop: weekly reviews, [[AI Wiki - PKM - Periodic Reviews]], [[AI Wiki - PKM - Spaced Repetition]]. Without feedback, the system degrades. ### Find Leverage Points Don't optimize tags (low leverage). Question what the vault is *for* (highest leverage). Most vault overhauls fail because they target low-leverage interventions. ## Limitations Systems thinking has its own failure modes: - **Analysis paralysis**: modeling forever, never acting - **Pseudo-scientific dressing**: using systems language without rigor (every problem becomes "complex") - **Underestimating human factors**: systems diagrams treat people as nodes; people are stranger than that [[AI Wiki - PKM - Reductionism vs Holism]] addresses the broader tension. ## Practitioners - **Donella Meadows**: clearest writer, *Thinking in Systems* - **Peter Senge**: organizational application, *The Fifth Discipline* - **Jay Forrester**: founder of system dynamics - **Russell Ackoff**: critique of mechanistic management - **Edgar Morin**: complexity-focused systems thinker (see [[AI Wiki - PKM - Edgar Morin]]) ## Open Questions - How much systems-thinking literacy is required before it pays off, vs being just jargon? - Are software tools (causal-loop diagrams, system dynamics models) helpful for PKM, or are they overhead? - Where does systems thinking break down? ## References - Meadows, D. (2008). *Thinking in Systems: A Primer*. Chelsea Green. - Senge, P. (1990). *The Fifth Discipline*. Doubleday. - Morin, E. (1977). *La Méthode, Vol. 1*. Seuil. ## Related - [[AI Wiki - PKM - Complex Thinking]] - [[AI Wiki - PKM - Edgar Morin]] - [[AI Wiki - PKM - Cybernetics and PKM]] - [[AI Wiki - PKM - Feedback Loops in Knowledge Systems]] - [[AI Wiki - PKM - Self-Organization]] - [[AI Wiki - PKM - Emergence in PKM]] - [[AI Wiki - PKM - Reductionism vs Holism]] - [[AI Wiki - PKM - Mental Models for Knowledge Work]] - [[AI Wiki - PKM - Networked Thought]]