Levels of Abstraction

Navigating Abstraction (a general response)

in v2 what are the levels and characteristics of the approaches from the highest conceptual level to the lowest implementation details

In the v2-combined collection, the approaches to complexity and systems thinking are structured across four primary levels, descending from high-level paradigm shifts to the technical details of physical implementation.

1. The Paradigmatic / Epistemological Level (Highest Conceptual Level)

At this level, the focus is on how human beings perceive reality and the ethical foundations of inquiry.

• Systems as Mental Constructs: A foundational principle is that “systems” do not exist objectively in the real world; they are epistemological tools—ways of framing and making meaning of a messy reality to design better interventions.

• Ethical Priority: C. West Churchman argues that the highest conceptual level is the “securing of improvement in the human condition,” asserting that any science of management is fundamentally a branch of ethics.

• Biological vs. Mechanistic Mindsets: This involves a paradigm shift away from “clockwork” or deterministic recipes toward a biological or ecological mindset.

• Triple-Loop Learning (Loop 3): This conceptual level addresses the “Power/Ethics” layer, asking why a particular design or decision should be adopted and whether “might is defended by right”.

2. The Methodological / Heuristic Level

This level provides the frameworks used to diagnose the nature of a problem and select appropriate strategies.

• Domain Diagnosis (Cynefin): Determining if a situation is Clear, Complicated, Complex, or Chaotic determines the response model (e.g., “Probe-Sense-Respond” for complex domains).

• Separating “What” from “How”: Methodologies like Soft Systems Methodology (SSM) prioritize logical necessities (the “whats”) over entrenched, messy practices (the “hows”).

• Boundary Critique: Practitioners use Critical Systems Heuristics (CSH) to determine what is included (valued) or excluded (marginalized), identifying who the system serves versus who it harms.

3. The Operational / Modeling Level

This level translates methodologies into structural maps of causality and relationship.

• “System-as-Cause”: The focus shifts to the internal “plumbing” of a system—specifically the stocks (accumulations) and flows (activities) that generate behavior.

• Constraint Management: Alicia Juarrero defines the operational level as managing context-independent constraints (boundaries/gradients) and context-dependent constraints (catalysts/feedback loops) rather than applying force.

• CAD → SIM Process: Alan Kay advises first designing for meaning (CAD) and simulating the semantics (SIM) to observe the phenomena the model exudes before worrying about efficiency.

• Triple-Loop Learning (Loop 2): Addresses “Effectiveness,” asking which options and visions should be decided upon.

4. The Implementation / Technical Level (Lowest Details)

This level deals with the “physics” of the system and the actual mechanics of execution.

• Late Binding: Technical commitments to hardware, data types, or representations are delayed until runtime, allowing systems to grow and adapt while they are running.

• Configurable Switches: At the most granular level, information is physically instantiated into reality via dynamically inert configurable switches or logic gates that are set by formal choice.

• Safe-to-Fail Probes: Implementation is executed through small, parallel experiments designed to reveal the system’s “disposition” without causing catastrophic failure.

• The Italian Flag Tool: Data dependability is managed by explicitly tagging evidence for success (Green), evidence for failure (Red), and recognized ignorance (White).

• Triple-Loop Learning (Loop 1): Focuses on “Design/Efficiency,” asking how processes can be designed and implementations arranged for maximum optimization.

Navigating Abstraction - Specific Advice

Can you develop this answer to include Tim Allens distinction of essence narrative model and mechanism.

see the discussion on Wealth

Integrating Tim Allen’s (T.F.H. Allen) hierarchical framework into the critique of “What Gary Says” provides a rigorous way to determine where Graham Berrisford and Gary Stevenson are actually in conflict. Allen’s hierarchy—Essence, Narrative, Model, and Mechanism—reveals that the two authors often argue at different levels of abstraction, leading to a “category error” in their debate.

Here is an expansion of the critique using Allen’s four levels:

1. The Level of Essence (The Irreducible “System of Interest”)

The Essence is the underlying, irreducible truth or purpose of the system being observed.

• Gary Stevenson’s Essence: The economy is a Stock-Flow system where a massive “blood clot” of wealth (stock) is starving the circulatory system (flow). The essence is the structural instability of the “system-as-a-whole”.

• Graham Berrisford’s Essence: The economy is a Managed System defined by policy, demographics, and external variables like AI. To Graham, the “essence” is not just inequality, but a multi-factor “laundry list” of constraints.

• The Conflict: According to Allen, if you don’t agree on the essence, the rest of the levels will never align. Berrisford critiques Gary for having too narrow an essence, while Gary argues Berrisford’s essence is a distraction from the “physics” of wealth.

2. The Level of Narrative (The Meaning-Making Story)

The Narrative is the story used to make the essence meaningful and to identify the “Hero, Victim, and Villain”.

• Gary Stevenson’s Narrative: A “Neo-Feudal” tale where the “powerful elite” (Villains) extract wealth from “ordinary people” (Victims). This narrative provides the emotional and political “why”.

• Graham Berrisford’s Narrative: An “OECD Reality” tale where well-meaning governments face “unintended consequences” and “capital flight.”

• The Critique: Allen suggests that narratives are often used to hide complexity. Berrisford correctly identifies that Gary’s narrative is “emotionally resonant but logically flawed” because it relies on a zero-sum story that may not account for the emergent properties of a globalized economy.

3. The Level of Model (The Simplified Logic)

The Model is the formal, logical representation of how the parts relate to produce behaviour.

• Gary Stevenson’s Model: A Heterogeneous Agent Model where wealth distribution is the independent variable that dictates interest rates and inflation.

• Graham Berrisford’s Model: A Standard Econometric Model (ONS/OBR) where demographics and productivity are the primary drivers.

• The Systems Critique: Allen and Alan Kay would argue that Berrisford’s reliance on “official statistics” is a “predict and provide” model that is too linear. Conversely, Gary’s model is critiqued for its Category Error: he uses a “household” model to describe a “sovereign currency” system, which Allen would view as a failure of scale-relative ontology.

4. The Level of Mechanism (The Technical Implementation)

The Mechanism is the “physics” or the lowest level of “how” the system is actually changed.

• Gary Stevenson’s Mechanism: An Asset Tax.

• Graham Berrisford’s Critique: He focuses almost entirely on the failure of the mechanism—valuation difficulties, liquidity issues, and administrative costs.

• The Conclusion: From Tim Allen’s perspective, Berrisford’s critique is technically robust at the Mechanism level, but he uses these mechanical flaws to claim Gary’s Essence is wrong. Systems thinkers would call this a “Logical Fallacy”: a difficult “how” (mechanism) does not invalidate a necessary “what” (essence). Even if an asset tax is hard to implement, the structural instability Gary identifies remains an “ontological fact” of the system’s architecture.

Summary Table: Level-by-Level Critique

Level	Gary Stevenson’s Position	Graham Berrisford’s Critique	Systems Verdict (v2-combined)
Essence	Structural Stagnation (Inequality)	Multifactoral Stagnation (AI, Aging)	Berrisford is too “reductionist”; Gary captures the “physics”
Narrative	Neo-Feudalism / Elite Villains	Complex Policy / Unintended Effects	Gary uses a “Strict Father” vs “Nurturant Parent” frame
Model	Household Analogy	ONS/OBR Data Forecasts	Both suffer from “Scale Errors” and “Predictive Fallacies”
Mechanism	Asset Tax	Valuation & Liquidity Failures	True: Berrisford wins on “Complication”; Gary wins on “Complexity”