Canonical Functionalism: Defining Functional Structure without Observer-Relative Semantic Maps
Pith reviewed 2026-05-22 02:04 UTC · model grok-4.3
The pith
Consciousness-relevant functional organization is the minimal state-transition structure from states with identical future behaviors under all continuations.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The central claim is that a system's consciousness-relevant functional organization should be identified with its canonical functional structure: the minimal state-transition structure obtained by identifying internal states that have identical future behavior under all possible continuations. A state is functionally defined by its complete counterfactual role, that is, how the system would evolve and respond from that state under possible future interactions. The framework does not identify which systems are conscious or claim that functional organization suffices for consciousness; it identifies the canonical object over which such theories should be formulated.
What carries the argument
Canonical functional structure, the minimal state-transition structure formed by equivalence classes of states sharing identical future behaviors under all continuations; it supplies the observer-independent basis for functional descriptions.
If this is right
- Objections involving lookup tables or simulations do not refute functionalism unless they show that the canonical structure is absent or altered.
- Functionalist theories must now specify which additional structural features, if any, are required beyond preservation of the canonical structure.
- The debate shifts from whether functional organization is sufficient to whether the relevant canonical structure is preserved in a given case.
Where Pith is reading between the lines
- This definition could allow direct comparison of artificial and biological systems by applying the same state-equivalence procedure to their internal dynamics.
- Measures of complexity or integration could be applied to the resulting canonical structure to generate new predictions about which systems satisfy functionalist criteria.
- The approach suggests testing whether systems engineered to match canonical structure but differ in surface semantics produce matching behavioral or report-based outcomes.
Load-bearing premise
A system's internal states possess well-defined, observer-independent future behaviors under all possible continuations that permit partitioning the state space into equivalence classes without external semantic interpretation.
What would settle it
A demonstration that two physical systems with identical canonical functional structures nevertheless require different semantic interpretations to match functionalist predictions about consciousness, or that no such equivalence partitioning can be constructed for any actual dynamical system.
read the original abstract
Computational functionalism about consciousness is often criticized for relying on observer-relative interpretations of physical systems. This paper proposes a mathematical refinement of functionalism that avoids this problem. The central idea is that consciousness-relevant functional organization should be identified not with arbitrary input-output mappings, semantic labels, or externally imposed computational descriptions, but with a system's canonical functional structure: the minimal state-transition structure obtained by identifying internal states that have identical future behavior under all possible continuations. On this view, a state is functionally defined by its complete counterfactual role: how the system would evolve and respond from that state under possible future interactions. We call this position canonical functionalism. The framework does not claim to identify which systems are conscious, nor to show that functional organization is sufficient for consciousness. Rather, it identifies the canonical object over which functionalist theories of consciousness should be formulated: the task is to specify consciousness-relevant invariants, measures, or structural conditions over canonical functional structures, rather than over arbitrary semantic interpretations or superficial behavioral profiles. This reframes familiar objections about lookup tables, simulations, unfolding, and observer-relative computation: such cases do not by themselves refute functionalism, but force the functionalist to specify whether the relevant canonical structure is preserved, and if not, which additional structural features are missing.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript proposes 'canonical functionalism' as a mathematical refinement of computational functionalism about consciousness. It defines the consciousness-relevant functional organization of a system as its 'canonical functional structure': the minimal state-transition structure obtained by quotienting internal states into equivalence classes based on identical future behavior under all possible continuations. The paper does not claim to solve the hard problem or identify which systems are conscious; instead it reframes standard objections (lookup tables, simulations, unfolding) as questions about whether a given physical system preserves this canonical structure rather than an arbitrary semantic interpretation.
Significance. If the definition can be made fully intrinsic and shown to yield unique, extractable structures from real dynamical systems, the proposal would supply a clearer, less observer-relative target for functionalist theories in philosophy of mind and computational neuroscience. The absence of worked examples, uniqueness theorems, or demonstrations on concrete state spaces currently limits its immediate utility for guiding empirical or modeling work.
major comments (2)
- [Abstract] Abstract and central definition: the construction presupposes a fixed set of 'possible continuations' and an interaction interface to define future behavior, yet the manuscript does not derive this interface from the system's intrinsic physics or dynamics. Different choices of observables or input channels can induce different partitions of the same underlying state space, reintroducing a form of modeler-dependence at the level of the equivalence relation itself.
- [Definition of canonical functional structure] The claim that canonical structure is 'minimal' and 'observer-independent' is load-bearing for the reframing of objections, but no formal statement is given showing that the equivalence relation is uniquely determined by the transition function alone without additional stipulations on the continuation set.
minor comments (1)
- [Abstract] The abstract refers to 'unfolding' and 'lookup tables' without citing the specific arguments (e.g., Putnam, Maudlin, or Block) being addressed, making it harder for readers to assess the reframing.
Simulated Author's Rebuttal
We thank the referee for the careful reading of our manuscript and the constructive comments provided. We address each of the major comments in turn and indicate the revisions that will be incorporated in the next version of the paper.
read point-by-point responses
-
Referee: [Abstract] Abstract and central definition: the construction presupposes a fixed set of 'possible continuations' and an interaction interface to define future behavior, yet the manuscript does not derive this interface from the system's intrinsic physics or dynamics. Different choices of observables or input channels can induce different partitions of the same underlying state space, reintroducing a form of modeler-dependence at the level of the equivalence relation itself.
Authors: We agree with the referee that the manuscript presupposes an interaction interface without deriving it from the intrinsic physics of the system. This could indeed lead to different partitions depending on the choice of observables or input channels. To resolve this, we will revise the central definition to include an explicit construction of the continuation set based on the complete causal structure of the system as determined by its physical dynamics. Additionally, we will include a lemma demonstrating that the equivalence relation remains invariant under physically equivalent choices of interface. revision: yes
-
Referee: [Definition of canonical functional structure] The claim that canonical structure is 'minimal' and 'observer-independent' is load-bearing for the reframing of objections, but no formal statement is given showing that the equivalence relation is uniquely determined by the transition function alone without additional stipulations on the continuation set.
Authors: The referee is right to note that no formal statement is provided to show that the equivalence relation is uniquely determined by the transition function. The current presentation leaves the continuation set somewhat implicit. In the revision, we will provide a precise mathematical definition of the canonical functional structure as the quotient by the largest bisimulation relation with respect to all possible input sequences, and prove its uniqueness as the minimal structure preserving the transition dynamics. This will be added as a new theorem in the paper. revision: yes
Circularity Check
No circularity: canonical structure defined directly via state equivalence
full rationale
The paper introduces canonical functional structure as a direct mathematical definition: the minimal state-transition structure from equivalence classes of states with identical future behavior under all possible continuations. This construction is presented as a primitive refinement of functionalism without any derivation chain that reduces to fitted parameters, self-citations, or prior results by the same authors. No equations, ansatzes, or load-bearing citations appear in the central claim, and the framework explicitly disclaims identifying conscious systems or proving sufficiency. The derivation is therefore self-contained as a proposed redefinition rather than a closed loop.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption Physical or computational systems possess internal states whose future evolution is well-defined under every possible continuation.
invented entities (1)
-
canonical functional structure
no independent evidence
Lean theorems connected to this paper
-
IndisputableMonolith/Foundation/AbsoluteFloorClosure.leanabsolute_floor_iff_bare_distinguishability unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
two states are equivalent if no possible future input history can distinguish them... Can(S) = R_S / ~_S
-
IndisputableMonolith/Foundation/ArithmeticFromLogic.leanembed_injective unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
canonical realization theorem... isomorphic to a canonical machine determined uniquely by the behavior B
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
Reference graph
Works this paper leans on
-
[1]
Ned Block. Troubles with functionalism. In C. Wade Savage, editor,Perception and Cognition, volume 9 ofMinnesota Studies in the Philosophy of Science, pages 261–325. University of Minnesota Press, Minneapolis, 1978
work page 1978
-
[2]
Can only meat machines be conscious?Trends in Cognitive Sciences, 30(4):298–308,
Ned Block. Can only meat machines be conscious?Trends in Cognitive Sciences, 30(4):298–308,
-
[3]
doi: 10.1016/j.tics.2025.08.009
-
[4]
Patrick Butlin, Robert Long, Eric Elmoznino, Yoshua Bengio, Jonathan Birch, Axel Constant, George Deane, Stephen M. Fleming, Chris Frith, Xu Ji, Ryota Kanai, Colin Klein, Grace Lindsay, Matthias Michel, Liad Mudrik, Megan A. K. Peters, Eric Schwitzgebel, Jonathan Simon, and Rufin VanRullen. Consciousness in artificial intelligence: Insights from the scien...
work page 2023
-
[5]
Chalmers.The Conscious Mind: In Search of a Fundamental Theory
David J. Chalmers.The Conscious Mind: In Search of a Fundamental Theory. Oxford University Press, New York, 1996
work page 1996
- [6]
-
[7]
What is consciousness, and could machines have it?Science, 358(6362):486–492, 2017
Stanislas Dehaene, Hakwan Lau, and Sid Kouider. What is consciousness, and could machines have it?Science, 358(6362):486–492, 2017. doi: 10.1126/science.aan8871
-
[8]
Adrien Doerig, Aaron Schurger, Kathryn Hess, and Michael H. Herzog. The unfolding argument: Why IIT and other causal structure theories cannot explain consciousness.Consciousness and Cognition, 72:49–59, 2019. doi: 10.1016/j.concog.2019.04.002
-
[9]
The symbol grounding problem.Physica D: Nonlinear Phenomena, 42(1–3): 335–346, 1990
Stevan Harnad. The symbol grounding problem.Physica D: Nonlinear Phenomena, 42(1–3): 335–346, 1990. doi: 10.1016/0167-2789(90)90087-6
-
[10]
Toward a universal theory of consciousness.Neuroscience of Consciousness, 2024(1):niae022, 2024
Ryota Kanai and Ippei Fujisawa. Toward a universal theory of consciousness.Neuroscience of Consciousness, 2024(1):niae022, 2024. doi: 10.1093/nc/niae022
-
[11]
The abstraction fallacy: Why AI can simulate but not instantiate consciousness
Alexander Lerchner. The abstraction fallacy: Why AI can simulate but not instantiate consciousness. Manuscript, PhilPapers/PhilArchive, 2026. URL https://philpapers.org/ rec/LERTAF. Manuscript
work page 2026
-
[12]
David K. Lewis. Psychophysical and theoretical identifications.Australasian Journal of Philosophy, 50(3):249–258, 1972. doi: 10.1080/00048407212341301
-
[13]
Taking AI welfare seriously, 2024
Robert Long, Jeff Sebo, Patrick Butlin, Kathleen Finlinson, Kyle Fish, Jacqueline Harding, Jacob Pfau, Toni Sims, Jonathan Birch, and David Chalmers. Taking AI welfare seriously, 2024
work page 2024
-
[14]
Oxford University Press, Oxford, 2015
Gualtiero Piccinini.Physical Computation: A Mechanistic Account. Oxford University Press, Oxford, 2015
work page 2015
-
[15]
Hilary Putnam. Minds and machines. In Sidney Hook, editor,Dimensions of Mind: A Symposium, pages 138–164. New York University Press, New York, 1960
work page 1960
-
[16]
MIT Press, Cambridge, MA, 1988
Hilary Putnam.Representation and Reality. MIT Press, Cambridge, MA, 1988
work page 1988
-
[17]
John R. Searle. Minds, brains, and programs.Behavioral and Brain Sciences, 3(3):417–424,
-
[18]
doi: 10.1017/S0140525X00005756
-
[19]
Anil K. Seth. Conscious artificial intelligence and biological naturalism.Behavioral and Brain Sciences, pages 1–42, 2025. doi: 10.1017/S0140525X25000032. Accepted manuscript
-
[20]
Cosma Rohilla Shalizi and James P. Crutchfield. Computational mechanics: Pattern and prediction, structure and simplicity.Journal of Statistical Physics, 104(3–4):817–879, 2001. doi: 10.1023/A:1010388907793. 23
-
[21]
Mark Sprevak. Computation, individuation, and the received view on representation.Studies in History and Philosophy of Science Part A, 41(3):260–270, 2010. doi: 10.1016/j.shpsa.2010. 07.008
-
[22]
Triviality arguments about computational implementation
Mark Sprevak. Triviality arguments about computational implementation. In Mark Sprevak and Matteo Colombo, editors,The Routledge Handbook of the Computational Mind, pages 175–191. Routledge, Abingdon and New York, 2018. doi: 10.4324/9781315643670-14. 24
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.