Skip to main content

Panpsychism: Categorical Analysis

Who this chapter is for

You will learn how UHM's position (pan-interiority) differs from classical panpsychism and Hoffman's Conscious Realism. The analysis is conducted through the categorical apparatus: five ontological positions are compared as functors from the category Hol\mathbf{Hol} to the category of phenomenal properties.

About notation

In this document:

The Strasbourg problem: where does consciousness come from?​

In 1994 in Strasbourg, at the conference Toward a Science of Consciousness, David Chalmers posed a question that split the science of consciousness into two camps: why do physical processes accompany subjective experience? Why can a "zombie" not exist — a being physically identical to a human but devoid of inner experience?

This question — the hard problem of consciousness — generated three response strategies:

  1. Eliminativism (Dennett): there is no problem, subjective experience is an illusion
  2. Emergentism (IIT, GWT): consciousness emerges from complexity
  3. Panpsychism: consciousness is fundamental — it always existed

Panpsychism is the most radical and most ancient of these strategies. Its idea is maximally simple: if consciousness cannot arise from non-conscious matter (since the mechanism is unclear), then consciousness is a fundamental property, inherent in all matter from the very beginning. An electron possesses mass, charge, spin — and, possibly, some elementary "inner experience".

The appeal of this position lies in bypassing the hard problem. If consciousness is fundamental, there is no need to explain how it "emerges" from physics. The problem, however, shifts: if an electron possesses experience, how does the electron's experience combine with the experience of other electrons into a unified human experience? This is the combination problem — the main difficulty of panpsychism.


Definition of panpsychism​

Panpsychism (from the Greek πᾶν — all, ψυχή — soul) is a metaphysical position:

∀X∈Ob(Phys):Consciousness(X)≠∅\forall X \in \mathrm{Ob}(\mathbf{Phys}) : \mathrm{Consciousness}(X) \neq \varnothing

where Phys\mathbf{Phys} is the category of physical objects.

In ordinary language: every physical object possesses at least a minimal form of consciousness or proto-consciousness. A stone, an atom, a thermostat — all "experience something".

But this simple thesis splits into many incompatible positions: what exactly does "consciousness" mean? Full-fledged experience (as in humans) or something minimal? And how does the minimal become full-fledged? Below we consider the main variants.


Variants of panpsychism​

1. Eliminative panpsychism (Strawson)​

Galen Strawson (b. 1952, son of philosopher P.F. Strawson) in the article Realistic Monism: Why Physicalism Entails Panpsychism (2006) proposed a radical argument: if physicalism is true and consciousness is real, then consciousness must be a property of matter at the fundamental level. Strawson rejects emergentism as "magic" — in his view, a genuinely new quality cannot arise from that which does not possess that quality.

Claim: Everything possesses consciousness in the full sense (L2 in CC terminology).

Category Panelim\mathbf{Pan}_{\mathrm{elim}}:

Ob(Panelim):={X∣C(X)>0}\mathrm{Ob}(\mathbf{Pan}_{\mathrm{elim}}) := \{X \mid C(X) > 0\}

where CC is the consciousness measure.

Critique from UHM — formal refutation:

∃X:ΓX=I/7⇒C(X)=0\exists X : \Gamma_X = I/7 \Rightarrow C(X) = 0

The maximally mixed state Γ=I/7\Gamma=I/7 has P=1/7P=1/7, canonical R=1R=1, Φ=0\Phi=0 and the chosen score C=ΦR=0C=\Phi R=0 [T/D]. It fails Cap2\mathrm{Cap}_2 because its purity/integration cuts fail. These arithmetic facts do not prove maximal physical chaos, absence of every possible self-model, or absence of experience without the phenomenal bridge [I/H].

Conclusion: Eliminative panpsychism contradicts the UHM formalism. Systems with zero consciousness exist. Not everything possesses experience. [T]

Correction: the thesis refuted above is not Strawson's

Galen Strawson ("Realistic monism: why physicalism entails panpsychism", Journal of Consciousness Studies 13(10–11): 3–31, 2006) argues that a physicalist who takes experience to be real must accept micropsychism — "at least some ultimates are intrinsically experience-involving" — and, since he would "bet a lot against there being such radical heterogeneity at the very bottom of things", that all of them are. He explicitly rejects the thesis this section attributes to him: "Panpsychism certainly does not require one to hold the view that things like stones and tables are subjects of experience — I don't believe this for a moment". He ascribes no reasoning or reflection (L2) to ultimates, and he names William James's objection to many subjects composing one larger subject as the problem he has to face. Two consequences [I]: (1) the refutation above refutes "everything is L2", a thesis no contemporary panpsychist defends; against Strawson's actual thesis — every ultimate has an experiential aspect, composites need not be subjects — UHM's universal L0 with a threshold for L2 is a near relative, not a refutation; (2) "eliminative" is this page's label, not Strawson's: he calls his position realistic monism, or real physicalism.

2. Constitutive panpsychism (as analysed by Chalmers and Goff)​

Philip Goff (b. 1977, Durham University) and David Chalmers (b. 1966, NYU) worked out a more refined position — micro-subjects (elementary bearers of proto-experience) combine into macro-consciousness — without either of them holding it as his own view: Chalmers analyses constitutive panpsychism without endorsing it, and Goff, who proposed the phenomenal-bonding solution for it (2016), defends cosmopsychism in Consciousness and Fundamental Reality (2017); see the combination problem, stated precisely. Goff's popular book Galileo's Error: Foundations for a New Science of Consciousness (2019) argues for panpsychism in general. (Earlier editions presented both authors as defenders of this position; that attribution is retracted.)

Claim: Micro-subjects exist, and their combination generates macro-consciousness.

Category Panconst\mathbf{Pan}_{\mathrm{const}}:

Ob(Panconst):={({Xi},⊕)∣Xi — micro-subject,⊕ — combination operation}\mathrm{Ob}(\mathbf{Pan}_{\mathrm{const}}) := \{(\{X_i\}, \oplus) \mid X_i \text{ — micro-subject}, \oplus \text{ — combination operation}\}

Combination problem:

The central difficulty of constitutive panpsychism: there is no definition of the operation ⊕\oplus such that:

C(X1⊕X2)=f(C(X1),C(X2))C(X_1 \oplus X_2) = f(C(X_1), C(X_2))

for any function ff. Why? Because the experience of the whole is not reducible to a function of the experiences of the parts. You see a red apple — but your experience of the "red apple" is not the sum of the experience of neuron-1 and the experience of neuron-2. Between individual proto-experiences and the unified macro-experience there is a gap that nobody has filled.

UHM's approach to the combination problem:

UHM composes holons by the tensor product:

H1⊗2:=(Γ1⊗Γ2,φ12)\mathbb{H}_{1 \otimes 2} := (\Gamma_1 \otimes \Gamma_2, \varphi_{12})
Status: a criterion of when, not an account of how

Even with correlation in place of Φ12\Phi_{12}, UHM offers at most a criterion — whether a correlated composite, once represented on D(C7)\mathcal{D}(\mathbb{C}^7), passes the four-condition window — and does not explain the constitution: how exactly micro-experiences unite into a single experience. This is a reformulation of the problem, not its solution in the philosophical sense; what the machinery does and does not answer is assessed below.

Does a composite score detect correlations?​

Why it fails. For a product the spectrum does factorise: the eigenvalues of A⊗BA \otimes B are the products λiμj\lambda_i \mu_j, the partial traces return Γ1\Gamma_1 and Γ2\Gamma_2 exactly, and the mutual information is zero — step (c) is false, and in step (d) there is no correlation to lose. On a product the data of steps (a) and (b) coincide (the spectrum is the product of the spectra and, for non-degenerate spectra, the eigenvectors are the products of the eigenvectors), so μ\mu loses nothing there. Nor is Φ12>1\Phi_{12} > 1 a mark of correlation. Taking Φ12\Phi_{12} as the integration measure of the joint 49×4949 \times 49 matrix (the page defines no other), purity and diagonal purity are both multiplicative on products, and the identity P=Pdiag(1+Φ)P = P_{\text{diag}}(1 + \Phi) gives

1+Φ(Γ1⊗Γ2)=(1+Φ1)(1+Φ2).1 + \Phi(\Gamma_1 \otimes \Gamma_2) = (1 + \Phi_1)(1 + \Phi_2).

A worked pair: the state Γw=12∣u⟩⟨u∣+12⋅I/7\Gamma_w = \tfrac12 \lvert u\rangle\langle u\rvert + \tfrac12 \cdot I/7 with u=(1,…,1)/7u = (1, \ldots, 1)/\sqrt7 passes the purity, reflection and integration thresholds (P=5/14P = 5/14, R=2/5R = 2/5, Φ=3/2\Phi = 3/2), and two uncoupled copies have Φ(Γw⊗Γw)=21/4\Phi(\Gamma_w \otimes \Gamma_w) = 21/4 at I(1:2)=0I(1{:}2) = 0; two holons at Φ1=Φ2=1\Phi_1 = \Phi_2 = 1 give exactly Φ12=3\Phi_{12} = 3. (Checked numerically on 200 random pairs of 7×77 \times 7 states: the identity holds to a relative error below 10−1510^{-15} and ∣I(1:2)∣≤2⋅10−15\lvert I(1{:}2)\rvert \leq 2 \cdot 10^{-15}; regression test test_phi_of_a_product_factorises in website/scripts/check_core_numbers.py.)

What replaces it.

  • Stands [T]: for a correlated joint state ρ12≠ρ1⊗ρ2\rho_{12} \neq \rho_1 \otimes \rho_2 the marginals do not determine the joint state; the information the partial traces discard is I(1:2)=S(ρ1)+S(ρ2)−S(ρ12)=D(ρ12 ∥ ρ1⊗ρ2)>0I(1{:}2) = S(\rho_1) + S(\rho_2) - S(\rho_{12}) = D(\rho_{12} \,\|\, \rho_1 \otimes \rho_2) > 0 (a standard identity of quantum information theory), and CC-7 says when the stationary state of coupled holons has I>0I > 0: when the coupling has a correlating part at the parts' attractors, under its assumption (ND) — not for every coupling ([T for almost every anchor]; the earlier "[T] guarantees I>0I > 0 for interacting holons" is retracted, 2026-09-25). This holds for every correlated pair, classically correlated ones included, so it does not single out integrated experience.
  • Hypothesis [H]: the phenomenal functor, extended to composite states, is colax-monoidal but not monoidal exactly on correlated joint states. It is not proved: FF is defined on D(C7)\mathcal{D}(\mathbb{C}^7), and applying it to a state on C7N\mathbb{C}^{7^N} first requires the aggregation channel that the theory does not fix (assumption 1 below). Research programme [Pr]: fix that channel, then prove or refute the colax property.

A phenomenal interpretation requires a map on the composite together with a physical and experimental criterion for when two reports describe one subject.

What has been resolved and what remains
  • Retracted [✗]: "Resolved [T]: emergence as irreversibility of the coherence map μ\mu" and "Resolved [T]: emergence threshold Φ12=1\Phi_{12} = 1 (T-129a)". Both rested on the theorem retracted above: a product of two L2 holons already has Φ12≥3\Phi_{12} \geq 3 at I(1:2)=0I(1{:}2) = 0, and T-129a concerns single states on D(C7)\mathcal{D}(\mathbb{C}^7), not composites.
  • Stands [T]: a correlated composite has a joint state that its parts do not fix; the discarded information is I(1:2)>0I(1{:}2) > 0 (a standard identity). When coupled holons are correlated at all is CC-7, [T for almost every anchor].
  • Open [H]: the colax-monoidal restatement on correlated states, pending the aggregation channel.
  • Open [Pr]: the constitutive mechanism — how exactly a correlated composite is experienced as unified experience. This is an analogue of the hard problem: UHM can at most state a condition, not the content of the unified experience.

3. Panprotopsychism (Chalmers)​

Chalmers (2010, The Character of Consciousness) proposed a softened version: everything possesses not "experience" but "proto-mental" properties — something that is not itself consciousness, but under the right combination generates it.

Analogy: H₂O. Neither hydrogen nor oxygen is wet by itself. But their combination is wet. The "proto-wetness" of hydrogen + the "proto-wetness" of oxygen → the wetness of water. Analogously: the "proto-experience" of an electron + the "proto-experience" of another electron + the right combination → human experience.

Category Panproto\mathbf{Pan}_{\mathrm{proto}}:

Ob(Panproto):={X∣Int(X)≠∅,C(X)=0}\mathrm{Ob}(\mathbf{Pan}_{\mathrm{proto}}) := \{X \mid \mathrm{Int}(X) \neq \varnothing, C(X) = 0\}

Correspondence in UHM — level L0:

By the canonical definition [D], L0 is universal — every system with Γ∈D(H)\Gamma \in \mathcal{D}(\mathcal{H}) possesses an inner aspect. What panprotopsychism tracks is its structured (non-degenerate) case:

L0+(Γ):=Γ≠I/7(structured interiority)\mathrm{L0}^{+}(\Gamma) := \Gamma \neq I/7 \quad \text{(structured interiority)} Panproto≅L0+∖L2\mathbf{Pan}_{\mathrm{proto}} \cong \mathrm{L0}^{+} \setminus \mathrm{L2}

L0+\mathrm{L0}^{+} marks the chosen non-reference subclass Γ≠I/7\Gamma\ne I/7 [D/I]. It includes states both passing and failing Cap2\mathrm{Cap}_2; non-reference structure does not imply C=0C=0 or absence of experience. Bare L0 and its interpretation remain separate from that operational gate.

4. Russellian monism (Russell, Strawson)​

Bertrand Russell in The Analysis of Matter (1927) pointed to a fundamental gap in physics: it describes only the structural properties of matter (mass, charge, spin), defined through relations with other objects. But what fills this structure from within? Physics is silent. Russell suggested: the inner nature of matter may be mental.

Russell's category (Russell\mathbf{Russell}):

Ob(Russell):={(Sext,Iint)∣S — structure,I — intrinsic}\mathrm{Ob}(\mathbf{Russell}) := \{(S_{\mathrm{ext}}, I_{\mathrm{int}}) \mid S \text{ — structure}, I \text{ — intrinsic}\}

Correspondence in UHM:

RussellUHMComment
SextS_{\mathrm{ext}} (structural properties)Hamiltonian HH, Lindblad operators {Lk}\{L_k\}Physical laws = structure
IintI_{\mathrm{int}} (inner nature)E-projection ρE\rho_EInteriority = intrinsic

Functor:

FRussell:Russell→Hol,(S,I)↦(Γ,φ)F_{\mathrm{Russell}}: \mathbf{Russell} \to \mathbf{Hol}, \quad (S, I) \mapsto (\Gamma, \varphi)

where Γ\Gamma is constructed from SS and II. Russellian monism is the closest metaphysical position to two-aspect monism of UHM.


UHM's position: Pan-interiority​

Definition​

UHM does not accept any form of panpsychism. Instead it proposes pan-interiority — a position according to which every system possesses interiority (L0 is universal by the canonical definition), every system with Γ≠I/7\Gamma \neq I/7 possesses structured interiority (L0⁺), but not all possess consciousness (L2).

Key distinction

UHM asserts pan-interiority, not panpsychism:

∀Γ≠I/7:L0+(Γ)=true\forall \Gamma \neq I/7 : \mathrm{L0}^{+}(\Gamma) = \mathrm{true}

But not:

∀Γ:L2(Γ)=true\forall \Gamma : \mathrm{L2}(\Gamma) = \mathrm{true}

Theorem (Pan-interiority ≠ Panpsychism)​

L0(Γ)⇏L2(Γ)\mathrm{L0}(\Gamma) \not\Rightarrow \mathrm{L2}(\Gamma)

Proof:

L2 requires R≥Rth=1/3R \geq R_{\text{th}} = 1/3 [T], Φ≥Φth=1\Phi \geq \Phi_{\text{th}} = 1 [T] (T-129) and Ddiff≥2D_{\text{diff}} \geq 2 [D] (T-151) (L2 thresholds).

For the fundamental mode Γ (e.g. an electron):

Φ(Γe)≪1(while R(Γe)=1/(7P)≈1 is a formal artefact at P≈1/7)\Phi(\Gamma_e) \ll 1 \quad (\text{while } R(\Gamma_e) = 1/(7P) \approx 1 \text{ is a formal artefact at } P \approx 1/7)

Consequently, L0(Γe)=true\mathrm{L0}(\Gamma_e) = \mathrm{true}, but L2(Γe)=false\mathrm{L2}(\Gamma_e) = \mathrm{false}: the integration threshold fails. The score R=1/(7P)R=1/(7P) is at least 1/71/7 and close to one near I/7I/7. This evaluates the specified gate, not an electron’s first-person experience. ■\blacksquare

Hierarchy of interiority levels​


Hoffman's Conscious Realism​

Biography and intellectual trajectory​

Donald D. Hoffman (b. 1955) is a professor of cognitive science, philosophy, and logic at the University of California, Irvine (UC Irvine). He began his career with classical psychophysics of visual perception: his early works (1980s–2000s) are devoted to computational modelling of the perception of shape, colour, and objects.

The turning point was the recognition of a paradox: if evolution shapes perception, why should perception be true? Together with Chetan Prakash and Manish Singh, Hoffman formalised this question in the "Fitness Beats Truth" theorem (2009–2015), showing through evolutionary game models that organisms perceiving an "interface" (a compressed adaptive representation) systematically outcompete organisms with "true" perception.

From this result Hoffman arrived at a radical ontological position: space-time is not objective reality but a user interface of conscious agents.

Important classification

Hoffman himself rejects the label "panpsychist". His position is objective idealism (Conscious Realism): conscious agents are the only fundamental reality, and the physical world is a derivative of their interactions. This is closer to Leibniz (monadology) or Berkeley than to Strawson or Goff.

The "Fitness Beats Truth" Theorem (FBT)​

Claim (Hoffman, Singh, Prakash 2015): In evolutionary games (Maynard Smith formalism) on typical fitness landscapes, organisms with the "interface" strategy (compression: many world states → one perceptual category) defeat organisms with "true perception" (isomorphism W→XW \to X).

In plain terms. Imagine two organisms in a forest. The first sees the world "as it is" — distinguishes 1000 shades of green in the foliage. The second compresses: all edible things are green, all poisonous are red. The second makes decisions faster and spends fewer resources on computation. Evolution selects for survival, not truth. Hence: our perception of space-time is an adaptive interface, not a map of reality.

Connection to CC

CC admits a similar interpretation [I]: the agent H\mathbb{H}'s perception of its environment EE is mediated by functor FF, which need not be accurate — functional adequacy is sufficient. But CC does not postulate the illusoriness of space-time: it is emergent, not interfacial.

Full formalism: conscious agent​

Definition (Hoffman, Prakash 2014). A conscious agent (CA) is a sextuple:

C=(X,G,A,W,D,N)C = (X, G, A, W, D, N)
ComponentDescriptionExample
XX (experience)All possible experiences of the agentColours, sounds, emotions
GG (actions)All available actionsMovements, decisions
A:G×W→WA: G \times W \to WAction gg in world ww → new world statePressing a button changes the screen
WW (world)World states (can be another agent!)Surrounding environment
D:X×G→GD: X \times G \to GExperience → decision (choice of action)You see danger → you run
N:W×X→XN: W \times X \to XWorld state → experiencePhotons → "red"

Key idea: WW need not be the "physical world". For two agents C1C_1 and C2C_2, the world of each is the other agent. Physical space-time is the emergent interface of a network of interacting agents.

Composition of conscious agents​

Closure theorem (Hoffman, Prakash 2014): For any two CAs C1C_1 and C2C_2, their interaction forms a new CA: C1⊗C2=C12C_1 \otimes C_2 = C_{12}.

This means that ConsAgents is a monoidal category. Hoffman interprets this as the principle "conscious agents are all there is".

Parallel with CC

The revised CC composition theorem gives conditional persistence for specified stable systems and compatible readouts. It does not derive a universally viable/unique aggregate or phenomenal combination. The product-score identity below holds even for uncoupled states and cannot supply that bridge.

CC's position: pan-interiority vs objective idealism​

Hoffman and CC diverge on a key ontological question:

Hoffman (Conscious Realism)CC (Pan-interiority)
What is fundamental?Only conscious agentsHolons H\mathbb{H} at all levels L0–L4
Is there a non-conscious reality?No — everything reduces to CAYes — L0 (interiority) without consciousness (L2)
Relation of L0 and L2L0 = L2 (everything is conscious)L0 ⊋\supsetneq L2 strictly [T]
Physical worldIllusion (interface)Emergent (real but derivative)
Consciousness thresholdNo threshold (everything is CA)P>2/7∧R≥1/3∧Φ≥1∧D≥2P > 2/7 \wedge R \geq 1/3 \wedge \Phi \geq 1 \wedge D \geq 2 [T]
DynamicsCycle N→D→AN \to D \to A (discrete)Γ˙=LΩ[Γ]\dot\Gamma = \mathcal{L}_\Omega[\Gamma] (continuous)
FalsifiabilityLow (no quantitative predictions)High (22+ predictions)

Functor FHoffmanF_{\text{Hoffman}} (hypothesis) [I]​

Functor construction FHoffman:HolL2→ConsAgentsF_{\text{Hoffman}}: \mathbf{Hol}_{\text{L2}} \to \mathbf{ConsAgents}:

CA componentCorrespondence in CC
XX (experience)Experiential space
GG (actions)Space of CPTP channels {Ψ}\{\Psi\}
NN (perception)Functor FF
DD (decision)Operator φ\varphi
AA (action)Regenerative term R[Γ,E]\mathcal{R}[\Gamma, E]
WW (world)Environment EE in H\mathbb{H}
Status [I]

The functor FHoffmanF_{\text{Hoffman}} is an interpretational hypothesis. For a full proof of equivalence it is necessary to show completeness, faithfulness, and compatibility with composition. This is a research programme.

What Hoffman does better than CC​

  1. Evolutionary epistemology. FBT is a strictly proven theorem providing deep grounds for scepticism towards naïve realism. CC has no analogue of this result.

  2. Accessibility of exposition. The Case Against Reality (2019) is a bestseller; TED talk with 3M+ views. CC is currently accessible only to specialists.

  3. Radicality of the question. Hoffman poses the question "what if space-time is not reality?" with maximum sharpness.

  4. Mathematical elegance. The six-component CA is minimal and convenient for combinatorics.


Precedents and related programmes: combination, boundaries, cosmopsychism, idealism​

This section sets UHM's pan-interiority against the literature the page has not cited so far: the combination problem as the field now states it, the boundary problem, cosmopsychism and analytic idealism. For each it gives the sources, the leading proposals and their standing, and then — marked as interpretation [I] — what UHM's composite-system results actually answer and which assumptions the answer rests on.

Where pan-interiority sits in the field's vocabulary​

The field now works with definitions due to David Chalmers ("The combination problem for panpsychism", in G. Brüntrup & L. Jaskolla (eds.), Panpsychism: Contemporary Perspectives, Oxford University Press, 2016, pp. 179–214). Panpsychism: some fundamental physical entities have conscious experiences. Panprotopsychism: fundamental physical entities have protophenomenal properties — "special properties that are not themselves phenomenal (there is nothing it is like to have them) but that can collectively constitute phenomenal properties". Constitutive views ground macro-experience in the micro-level; emergent views let it arise as something new under contingent laws.

By the interiority hierarchy's own definition, L0 "is not 'consciousness' or 'experience' in the ordinary sense" but "a mathematical property of the object", and the structure at L1 "is not yet perceived"; awareness appears at L2 through reflection and integration. In Chalmers' vocabulary this is a constitutive panprotopsychism with a structural-functional account of awareness [I] — the shape of Sam Coleman's panqualityism (qualities without subjects at the base, awareness constituted functionally), which Chalmers discusses among the combinatorial options. Three consequences follow.

  1. The contrast drawn above — "panpsychism = L2 for all" — is not the contrast with the field: no contemporary panpsychist ascribes reflection to electrons (see the correction in §1). The contrast that stands is UHM's sharp threshold.
  2. The combination problem that reaches UHM is the panprotopsychist one: how non-phenomenal structure constitutes experience. Chalmers argues that this faces a gap of its own, between the non-phenomenal and the phenomenal, on top of the familiar ones.
  3. Earlier editions of another page called the same position "emergentism with an exact threshold" and said that "the combination problem does not arise" (Philosophical foundations of CC). The two readings face different problems: panprotopsychism faces the gap just named; emergentism avoids the combination problem but, on Chalmers' analysis, must treat its threshold as a contingent psychophysical law and faces the problem of mental causation. The corpus asserted both readings; since 2026-09-25 it takes one throughout — the panprotopsychist reading of this section, on which the thresholds say when a system is an L2 subject, not how its structure becomes experience [I] — and the emergentist wording is retracted on that page and in the exercises, as is the claim of a solution in Theories of Consciousness §24, the section overview and the proofs of the interiority hierarchy (§5.1).

The combination problem, stated precisely​

William James posed the problem (The Principles of Psychology, 1890, vol. 1, ch. 6): "Take a sentence of a dozen words, and take twelve men and tell to each one word. Then stand the men in a row or jam them in a bunch, and let each think of his word as intently as he will; nowhere will there be a consciousness of the whole sentence." William Seager named it ("Consciousness, information and panpsychism", Journal of Consciousness Studies 2(3): 272–288, 1995). The Stanford Encyclopedia entry "Panpsychism" (Philip Goff, William Seager & Sean Allen-Hermanson, revised 2022) records the consensus: "It is generally agreed, both by its proponents and by its opponents, that the hardest problem facing panpsychism is what has become known as the 'combination problem.'"

Chalmers (2016) splits the problem into three subproblems plus further aspects, and insists that a solution must address all of them:

SubproblemQuestionSharpest form
Subject combinationHow do micro-subjects combine into a macro-subject?Subject-summing: given any group of subjects and any further subject, the group can apparently exist without the further one
Quality combinationHow do micro-qualities yield macro-qualities?Palette: a handful of fundamental qualities against the vast array of experienced ones
Structure combinationHow does micro-structure yield the structure of experience?Structural mismatch: the structure of experience is unlike the physical structure of the brain
Further aspectsUnity, boundary (Rosenberg 1998), awareness, grainWhy one bounded, homogeneous, aware field rather than a jagged array of micro-experiences

Two authors turned the subject problem into a choice. Sam Coleman ("The real combination problem: panpsychism, micro-subjects, and emergence", Erkenntnis 79(1): 19–44, 2014) argues that "subjects cannot combine": each subject's point of view excludes the others', so fusing micro-subjects could only yield a new, emergent macro-subject — which defeats panpsychism's motive; he recommends giving up micro-subjects and keeping qualities without subjects at the base. Philip Goff (Consciousness and Fundamental Reality, Oxford University Press, 2017) concludes from the same problem that micropsychism fails and defends cosmopsychism with "grounding by subsumption" (below); Daniel Stoljar's review (Notre Dame Philosophical Reviews, 2018.02.09) replies that in the relevant cases "grounding just is grounding by analysis", so subsumption offers no escape. Earlier editions of subsection 2 above presented Goff and Chalmers as defenders of micro-level constitutive panpsychism; neither is (the subsection now says so): Goff's considered view in Consciousness and Fundamental Reality is cosmopsychism, and Chalmers analyses constitutive panpsychism without endorsing it.

Leading proposals and their standing​

ProposalWhoWhat it doesStanding, and whose judgment
Emergent panpsychism: macro-subjects arise under new lawsGregg Rosenberg (A Place for Consciousness, OUP 2004); integrated information theory read as such a lawAvoids combination by making macro-subjects fundamentalPays with mental causation and contingent laws (Chalmers 2016, who also reads IIT this way); the SEP calls Rosenberg's view "a form of layered emergentism"
Phenomenal bondingGoff, "The phenomenal bonding solution to the combination problem" (in Brüntrup & Jaskolla 2016, pp. 283–302)A phenomenal relation among micro-subjects, such as co-consciousness, constitutes a new subject"One of the more promising approaches", yet hard pressed to avoid both a single universal subject and fragmentary subjects (Chalmers 2016)
Quantum holism, combinatorial infusionSeager, "Panpsychism, aggregation and combinatorial infusion", Mind and Matter 8(2): 167–184 (2010); Rodolfo Gambini & Jorge Pullin, Mind and Matter 22(1): 51–94 (2024) and J. Consciousness Studies 32(7): 7–32 (2025)An entangled whole is a new fundamental entity; the supervenience assumptions behind the combination problem fail at the quantum level, where the joint state space grows exponentiallyWorth close examination, but "it does not seem that there is the sort of stable brain-level entanglement that would be needed", and a structural-mismatch worry remains (Chalmers 2016)
Panqualityism with functional awarenessColeman (2012, 2014)Qualities without subjects at the base; awareness constituted functionallyOpen to a quality–awareness gap: qualities without awareness remain conceivable (Chalmers 2016)
Formal closure of compositionHoffman & Prakash, "Objects of consciousness", Frontiers in Psychology 5: 577 (2014)Theorems: the join of two conscious agents is a conscious agentProved by construction inside conscious realism (Hoffman & Prakash 2014); closure of a formalism, not by itself a new subject [I]
CosmopsychismItay Shani (2015); Yujin Nagasawa & Khai Wager (2016); Goff (2017)The cosmos is the one fundamental subject; individual subjects derive from itTrades combination for its reverse, a problem that, in Chalmers' judgment (2016), seems "just as hard as the original combination problem" — below

Overall standing, in Chalmers' judgment (2016): "It is fair to say that no proposed solution has yet gained much support." His conclusion names the avenues most worth exploring — phenomenal bonding or quantum holism for subjects, small palettes for qualities, "principles of informational composition" for structure, and a somewhat deflationary account of awareness — and adds that it is not at all clear whether they can work together.

What UHM's machinery answers, and what it leaves open​

The finite-dimensional composite-system framework supplies joint states in D7kD_{7^k} and CPTP marginal/readout maps. A product state has no mutual information; a separable mixture can still be correlated. Marginals generally do not determine the joint state [T]. These facts distinguish kinds of correlations, but do not identify a subject.

Weak-coupling persistence is conditional on an actual stable uncoupled equilibrium, appropriate hyperbolicity/contraction assumptions, strict margins for the chosen gate and a compatible readout. “Each part is viable” alone does not prove viability of every mean aggregate or an intrinsic composite boundary. The mean of marginals is one symmetric CPTP readout after identifying subsystem frames; the old blanket uniqueness and viability conclusions are withdrawn. See the revised CC composition results.

  • Subject combination. The calibrated Cap2\mathrm{Cap}_2 certificate evaluates a declared seven-dimensional readout and its differentiation variable. Interpreting that operational gate as experience requires the phenomenal bridge [I/H]. T-153's definition does not refute the conceivability objection. T-214's universal prohibition of an internal bridge is withdrawn; Lawvere's evaluator hypotheses do not hold for an arbitrary internal map.
  • Quality and structure combination. A map D7k→D7D_{7^k}\to D_7, its frame and the phenomenal assignment must be specified and calibrated. CPTP readout is nonexpansive in chordal Bures distance; it can discard joint information. A seven-mode ansatz does not prove that the visual field or every substrate is faithfully represented. Exact recovery in T-253 applies to the chosen isometric code, not every substrate state.
  • Unity. If the declared category has a terminal object, its nerve is contractible; the usual positive-degree cohomology of that nerve with constant coefficients vanishes. This is a statement about that classifying space, not all internal/sheaf cohomology, every local coefficient system or all self-models. Actual gluing requires descent and a specified obstruction class. No subject boundary follows from nerve contractibility.

A composite theory therefore needs three additional choices [D/H]: its physical grouping and effective readout; a rule for nested or competing subject candidates; and a phenomenal bridge. A larger tensor score alone does not choose them: it can exceed both component scores even in an uncorrelated product, while the user's marginal is unchanged. The combination problem remains a test of these additional constructions rather than a solved consequence of tensor algebra.

The boundary problem​

The boundary problem asks what fixes the edges of an experiencing subject: why my experience stops at me — not at each of my neurons, and not at me together with my room. Gregg Rosenberg framed it ("The boundary problem for phenomenal individuals", in S. R. Hameroff, A. W. Kaszniak & A. C. Scott (eds.), Toward a Science of Consciousness II, MIT Press, 1998) and developed it in A Place for Consciousness: Probing the Deep Structure of the Natural World (Oxford University Press, 2004). Andrés Gómez-Emilsson and Chris Percy put it in one line: "Once you've proposed a binding mechanism that creates larger, unified, macro 1PPs, what mechanism puts a stop to that process?" (a 1PP is a first-person perspective).

  • Rosenberg's own answer. Higher-level "natural individuals" emerge from lower-level ones, and experience is tied to his "carrier" theory of causation — "a form of layered emergentism, according to which human minds co-exist with the micro-level conscious subjects that give rise to them" (SEP "Panpsychism", 2022). Chalmers (2016) classes it as emergent panpsychism in which high-level individuals exert a small amount of downward causation.
  • Fekete, van Leeuwen and Edelman. Tomer Fekete, Cees van Leeuwen and Shimon Edelman ("System, subsystem, hive: boundary problems in computational theories of consciousness", Frontiers in Psychology 7: 1041, 2016) show that any graded measure of consciousness which counts a system as conscious will also count most of its subsystems, "irrelevantly extended" versions of it and groups of individuals — so it either measures something epiphenomenal or implies "a bizarre proliferation of minds" — unless it rests on intrinsic, "systemic" properties that separate systems whose existence is a matter of fact from systems whose existence is a matter of interpretation.
  • Integrated information theory: exclusion. Experience is "definite"; the conscious complex is the set of units whose integrated information is maximal, and "overlapping substrates that specify less integrated information are excluded" (L. Albantakis et al., "Integrated information theory (IIT) 4.0", PLoS Computational Biology 19(10): e1011465, 2023). Giulio Tononi and Christof Koch draw the consequence for groups: two people talking form an integrated system, but its integrated information is below that of each brain, so "there should indeed be two separate experiences, but no superordinate conscious entity" (preprint arXiv:1405.7089; published as "Consciousness: here, there and everywhere?", Philosophical Transactions of the Royal Society B 370: 20140167, 2015). Chalmers (2016, in a footnote) objects that such a rule makes consciousness extrinsic: intrinsically identical systems may be conscious or not depending on their surroundings.
  • Electromagnetic topology. Gómez-Emilsson & Percy ("Don't forget the boundary problem! How EM field topology can address the overlooked cousin to the binding problem for consciousness", Frontiers in Human Neuroscience 17: 1233119, 2023) propose that topologically closed pockets of the brain's electromagnetic field give observer-independent hard boundaries, and outline a three-stage empirical test. They document the neglect of the problem: a Scopus search found 5 papers with "boundary problem" against 92 with "binding problem".

UHM on the boundary problem [I]:

  1. Nested candidates [H]. Without an exclusion/boundary rule, several groupings can pass a chosen operational gate. This is an unresolved selection problem, not proof that each grouping experiences.
  2. Readout and sector. Revised T-253 permits exact CPTP recovery on a declared isometric code; it does not give a faithful compression of every state of an arbitrary larger substrate to D7D_7. Different admissible sectors/frames can give different Γ\Gamma and canonical Φ\Phi. A calibrated intrinsic sector and dynamics remain part of the empirical bridge.
  3. Terminal object. In the all-dimensional CPTP category the terminal object is the one-dimensional unit with the discard map. It is not the seven-dimensional matrix I/7I/7. Assigning a chosen seven-dimensional reference to a terminal point in another realisation is additional data [D]. Nerve contractibility does not force C=0C=0 for a physical universe or define local subject boundaries.
  4. What can be checked. Once a physical grouping, frame, identifiable state estimator, differentiation variable and temporal protocol are supplied, a Cap2\mathrm{Cap}_2 certificate can be tested. Phenomenality and subject individuation still need their separately stated bridge and evidence.

Cosmopsychism​

Cosmopsychism reverses the direction of explanation: the cosmos as a whole is the one fundamental conscious subject, and individual minds derive from it. It matters for UHM because UHM also has a "one" at the top of its formalism — the terminal object — and because cosmopsychism is the leading alternative for anyone who finds bottom-up combination hopeless.

  • Sources. Itay Shani, "Cosmopsychism: a holistic approach to the metaphysics of experience", Philosophical Papers 44(3): 389–437 (2015): an all-pervading cosmic consciousness is the single ontological ultimate and the ground from which individual conscious creatures emerge. Yujin Nagasawa & Khai Wager, "Panpsychism and priority cosmopsychism", in Brüntrup & Jaskolla (2016), pp. 113–129: the cosmos is the one basic conscious entity, prior to its parts, on the model of Jonathan Schaffer's priority monism ("Monism: the priority of the whole", Philosophical Review 119(1): 31–76, 2010), which argues from quantum entanglement that the cosmos is the one fundamental object. Goff (2017): cosmopsychism with grounding by subsumption.
  • Standing. The SEP (2022) names these authors among those "attracted to cosmopsychism on the grounds that it is better fitted than micropsychism to deal with the combination problem". The price is the reverse problem. Chalmers (2016) called it the decomposition problem — "how does a single subject give rise to multiple dependent subjects?" — and judged that such problems "seem just as hard as the original combination problem"; later, renaming it the "constitution problem", he rated it "very serious" while finding some cosmic avenues "more promising than analogs in the micropsychist case" (Chalmers, "Idealism and the mind-body problem", in W. Seager (ed.), The Routledge Handbook of Panpsychism, Routledge, 2020, pp. 353–373).
  • UHM [I/H]. Terminality is a property of a declared mathematical category, not an identification of the cosmos with I/7I/7 or a proof that the universe lacks experience. The proposal to realise the Universe as a holon requires its own physical state space, effective map, clock, interactions and boundary conditions. In revised T-266, maximisation of a chosen kinematic expression can establish an extremum [C]; assigning the cosmic purity P=3/7P=3/7 is a physical hypothesis [H], not a measured stage or universal cosmic theorem. A purity figure or an assumed trajectory does not prove birth, decay or phenomenal extinction. The universal self-reference ceiling of three is withdrawn (T-136/T-217/T-218); neither nesting nor Kan fibrancy determines a cosmic cognitive depth. UHM can adopt a philosophical stance on cosmopsychism, but this stance is not settled by the finite matrix formalism. The elementary fact that joint correlations are not fixed by marginals supplies no subject criterion by itself.

Analytic idealism​

Analytic idealism, Bernardo Kastrup's view, holds that there is only cosmic consciousness and that individual minds are split-off parts of it. It matters for UHM because it offers the one mechanism for subject boundaries in this section that has a clinical model, and because UHM has a formal analogue of that mechanism.

  • Source. Bernardo Kastrup, "The universe in consciousness", Journal of Consciousness Studies 25(5–6): 125–155 (2018).
  • What it holds. "There is only cosmic consciousness. We, as well as all other living organisms, are but dissociated alters of cosmic consciousness, surrounded by its thoughts. The inanimate world we see around us is the extrinsic appearance of these thoughts. The living organisms we share the world with are the extrinsic appearances of other dissociated alters." The model is dissociative identity disorder: an alter is a region of mental contents cut off from the rest by dissociation, and that cut is the subject's boundary. Kastrup claims that the view escapes the hard problem, the combination problem and the decombination problem.
  • Standing. A minority position. Chalmers (2020, cited above) finds identity cosmopsychism with this kind of fragmentation "a coherent view that is worth taking seriously" but massively revisionary: "It makes us pathological subjects who are entirely unaware of the vast majority of experiences we are having", which "entails a massive failure of introspection". His verdict on idealism as a whole: implausible, yet "not significantly less plausible than its main competitors".
  • UHM [I]. The formal analogue of dissociation is restriction: an individual's self-model acts on its reduced state and cannot reach coherences present only in the joint state (collective unconscious, property 1). The differences are of kind. UHM posits no cosmic subject; its "extrinsic appearance" is the external side of each Γ\Gamma, not the appearance of a cosmic mind's thoughts; and restriction is the normal relation of a part to a whole, not a pathology. Kastrup gives boundaries a mechanism; UHM's partial trace describes privacy but does not say which subsystems are subjects.

Summary​

ProblemStanding in the literature (whose judgment)UHM ingredientWhat it answers [I]What stays open
Subject combinationOpen in the literatureConditional composite dynamics; T-153 operational definitionTests a declared grouping/readoutPhenomenal constitution and boundary selection; T-214 universal no-go withdrawn
Quality combinationOpen (Chalmers 2016)Phenomenal functor on D(C7)\mathcal{D}(\mathbb{C}^7); aggregation of CC-6Nothing yetWhich state carries the composite's content; aggregation only loses distinguishability
Structure combinationOpen; the mismatch argument is "a significant challenge" (Chalmers 2016)Seven collective modes, T-153a [T]An informational routeWhether it yields the structure of experience
Irreducibility of the wholeStandard for correlated states; argued for panpsychism by Seager (2010) and Gambini & Pullin (2024–25)Standard identity; CC-7 [T for almost every anchor] says when coupling produces itThe joint state is not fixed by the partsHolds for any correlated pair; no mark of a subject
BoundaryNeglected (5 papers against 92, Gómez-Emilsson & Percy 2023); IIT answers with exclusionNone; nested subjects allowedA computable test once a sector is chosenThe choice of sector and of aggregation; the proliferation of minds
Decomposition (cosmopsychism, idealism)OpenTerminal nerve result for a declared category; cosmic mapping [H]A conditional formal constructionCosmic phenomenality, sector and boundaries; no identification with I/7I/7 or universal depth cap

Comparative table of panpsychism variants​

VariantAuthorYearClaimCorrespondence in UHMMain problem
Eliminative (this page's label; Strawson: realistic monism)Strawson2006Every ultimate is experiential; stones and tables are not subjects (correction)Universal L0 — a near relative [I]The combination problem, which Strawson concedes
Constitutive micropsychismAnalysed by Chalmers; phenomenal bonding proposed by Goff — neither defends it as his own view (§2)2016Micro-subjects combineH1⊗H2\mathbb{H}_1 \otimes \mathbb{H}_2; a correlated composite (CC-7) — a criterion, not a constitution [I]Combination problem (reformulated, not solved)
PanprotopsychismChalmers2010Proto-mental propertiesL0 — interiorityNo mechanism for L0→L2 transition
RussellianRussell, Chalmers, Goff1927/2010Intrinsic + structureρE\rho_E + (H,{Lk})(H, \{L_k\})No dynamics
Obj. idealismHoffman2014Only CA, physics is interfaceFunctor [I]No thresholds, low falsifiability
CosmopsychismShani; Nagasawa & Wager; Goff2015–2017The cosmos is the one fundamental subjectTerminal object TT — not a subject [I]The decomposition ("constitution") problem
Analytic idealismKastrup2018Only cosmic consciousness; we are its dissociated altersPartial trace as restriction [I]Massively revisionary about introspection (Chalmers 2020)
Pan-interiority (UHM)——All have L0, not all have L2L0⊋L2\mathrm{L0} \supsetneq \mathrm{L2}Interiority is a primitive; does not explain why it exists

Related documents: