✨The Abstraction Fallacy

How are the themes in the abstraction fallacy handled by the concepts and authors in v2-combined?

Alexander Lerchner, The Abstraction Fallacy: Why AI Can Simulate But Not Instantiate Consciousness - PhilPapers

The themes presented in The Abstraction Fallacy—specifically the indispensable role of the “mapmaker,” the distinction between simulation and instantiation, and the ontological status of information—resonate deeply with the systems thinking and complexity science traditions found in the combined collection.

Many authors in the collection have spent decades addressing the same logical traps, often using different terminology such as the “modelling relation,” “epistemic cuts,” and “second-order cybernetics.”

1. The Indispensable Mapmaker: Observer Dependency

A central claim of The Abstraction Fallacy is that computation is not an intrinsic physical process but a “mapmaker-dependent description” requiring an active agent to perform “alphabetization”. This directly aligns with several systems theorists:

• Ross Ashby and the Observer-Defined System: Ashby established that a system is not a physical object but a list of variables chosen by an observer from an infinity of possibilities. Complexity, in his view, is “purely relative to a given observer” because the observer decides which differences “make a difference”.

• Niklas Luhmann and the System/Environment Distinction: Luhmann argues that every system draws its own boundary through its own operations; identity is only possible through this difference. Without an observer making this distinction, the world is merely an “undifferentiated mass”.

• Derek Cabrera and DSRP: Cabrera identifies the “Reification Fallacy,” which mirrors the Abstraction Fallacy by mistaking mental constructs (like complexity or systems) for objective realities of the material world.

2. Simulation vs. Instantiation: The Modeling Relation

The fallacy argues that a syntactic map $(A\to A^{\prime })$ cannot transmute into the physical territory it describes $(P)$. This distinction between the “rules of the mapmaker” and the “intrinsic physics” is a foundational concern for relational biologists and semioticians:

• Robert Rosen’s Modeling Relation: Rosen’s entire framework is built on the “modeling relation”—the encoding of natural phenomena into formal models. He defines “complex systems” (like living organisms) as those possessing “non-simulable models,” meaning no syntactic algorithm can fully exhaust their semantic, physical reality.

• Howard Pattee and the Epistemic Cut: Pattee describes an irreducible divide between “description” (genotype/symbols) and “construction” (phenotype/dynamics). He argues that life emerges from “semantic closure”—a self-referential loop where symbols direct dynamics, but dynamics are required to “read” the symbols. This supports the paper’s claim that a map presupposes a mapmaker.

• Hylton Boothroyd’s Theories vs. Proposals: Boothroyd insists on strictly separating “Theories” (how things behave) from “Proposals” (what we wish to happen), noting that models are “intellectual constructs” rather than physical phenomena.

3. Biological Centrality and Thermodynamic Constraint

The Abstraction Fallacy posits that subjective experience is “physically made of, and fundamentally un-abstractable from, the specific thermodynamic and metabolic dynamics of the experiencing organism”. This “Biological Turn” is echoed throughout the collection:

• Maturana & Varela and Autopoiesis: Living systems are “operationally closed” networks that recursively produce their own components. Maturana argues that an observer cannot biologically distinguish between a perception (externally triggered) and an illusion, as all meaning is enacted internally through the organism’s metabolic structure.

• Terrence Deacon and Constitutive Absence: Deacon defines function and meaning negatively, as “constitutive absence”—what the environment prevents from happening. This aligns with the paper’s view that symbols are not “distilled essences” but depend on the physical constraints of the organism.

• Max Boisot and the I-Space: Boisot treats information as a physical phenomenon subject to thermodynamics. He suggests that organizations (clans or fiefs) use “uncertainty absorption”—managing ambiguity through learning-by-doing rather than premature, rigid codification (alphabetization).

4. Summary of Perspectives

The following table summarizes how different authors in v2-combined address the core components of the “Abstraction Fallacy”:

Theme in Fallacy	Corresponding Concept in Collection	Author(s)
Mapmaker Necessity	Second-Order Cybernetics	Krippendorff , James , Hoverstadt
Simulating $\ne$ Instantiating	Modeling Relation / Non-simulable	Rosen , Pattee , Wilk
Alphabetization Error	The “Blind Spot” / Reification	Frank et al. , Cabrera , Warfield
Semantic Barrier	Operational/Informational Closure	Luhmann , Maturana , Jose
Biological Necessity	Autopoiesis / Life-Maintaining	Maturana , Seth/Block , Juarrero

Dive deeper into how Robert Rosen's mathematical proof of non-computability in living systems specifically supports Lerchner's argument against "computational emergence"?

To understand how Robert Rosen’s mathematical framework supports the refutation of “computational emergence,” we must look at his rigorous definition of why organisms are not “trivial machines” and why their internal causal structure is non-computable.

Robert Rosen’s Non-Computability Proof

Rosen’s primary contribution to this debate is the proof that living systems possess Closure to Efficient Causation, a property that digital simulations can mimic but never instantiate.

1. The Machine vs. The Organism

• The Trivial Machine (Simulation): In a computer, there is a strict separation between software (logical rules) and hardware (the physical processor). The “efficient cause” of the computation—the entity that makes the process happen—is external to the logic itself (the engineer or the power supply).

• The Organism (Instantiation): In a living system, the “software” and “hardware” are the same thing. A cell synthesizes the very catalysts (enzymes) required to maintain its own metabolic functions. Rosen models this as an (M,R)-system (Metabolism-Repair), where the system entails its own efficient causes.

2. The Impredicativity Constraint

Lerchner’s paper argues that symbols are “arbitrary physical tokens assigned by a mapmaker”. Rosen provides the mathematical basis for this “Causality Gap”:

• Linear Entailment: In a computer program, $A$ leads to $A^{\prime }$ because the code says so. The physical switching of the gate is driven by voltage, not by the “meaning” of the symbol. This is Syntactic Simulation.

• Circular Entailment: In a complex system, causal loops are “impredicative.” This means a component $B$ is needed to produce $A$, but $A$ is also needed to produce $B$. This semantic closure cannot be reduced to a sequence of discrete, independent steps (algorithms).

3. Non-Simulability and “Computational Emergence”

Rosen argues that “complexity” is the presence of non-computable models. This directly supports Lerchner’s point that an increase in scale does not lead to a phase transition into sentience:

• The Scaling Fallacy: Functionalists believe that if you add enough syntactic layers, consciousness “emerges.”

• Rosen’s Rebuttal: Because a simulation is a “mapping” that abstracts away the physical degrees of freedom of the substrate to create a “closed” model, it systematically excludes the very “openness” and metabolic dynamics required for life. No matter how large the computer, it remains a simple system (a mechanism) because its behavior is entirely dictated by external state variables and fixed laws.

---

Comparison of Causal Structures

Feature	Computational Simulation (Simple System)	Biological Instantiation (Complex System)
Causality	Linear, driven by “Vehicle Causality” (p)	Circular, closed to efficient causation
Identity	Context-Independent (Hardware/Software split)	Context-Dependent (Structure is the logic)
Model Type	Computable / Simulable	Non-computable / Non-simulable
Outcome	Teleonomy (Mimicry of purpose)	Teleology (Intrinsic purpose/agency)

Would you like to explore how Maturana’s concept of “Structural Coupling” explains how these complex systems interact with their environment without losing their non-computable internal identity?