Which act bears the appearance?

The physical act and its qualitative mode

The decisive question is what bears the appearance already warranted in a current act. This chapter develops the positive principle of concrete performance and mode placement, then makes its alternatives precise enough to test the inference.

Grounding, causation, realization and identity license different inferences. This module examines the alternatives that motivated the current-performance account and the positive principle that supports it. Its finite models isolate implications; their qualitative and epistemic interpretations are stated premises, not experiences generated by a calculation. The models concern the attribution of an actual act and the limits of extending that attribution. The finite checks described here are source-reported checks preserved from the publication.

Use $P$ for an independently nominated physical performance, A for an experience-inclusive act, and $Q$ for its actual qualitative mode. $R$ is a physically operative reason/control component, $V$ a bodily expression, $K$ a current acquaintance, and $K^*$ a composite knowledge performance. $B$ denotes a genuinely exclusive qualitative instantiator outside the proposed mode-bearing $P$. It is distinct from the programme’s unsplit-ground notation $U$. A numerically distinct property instance of $P$ is not such an exclusive bearer.

8.1 Current grounding and evidential parity

An actual relational presentation can be evidence even when its apparent objects are illusory. External inaccuracy is different from absence of appearance. Equally, a response about a relation need not itself contain the sensory counterparts. The appropriate parity principle is evidential: relational subject matter alone is no reason to disqualify otherwise credible current experiential evidence. It does not assign relational testimony the same error rate as unary-quality testimony. Retention, conceptual load and reconstruction can make relevant differences. Joint accessibility and phenomenal unity must also be distinguished (Bayne and Chalmers, 2003).

Nonfloating epistemic grounding

If $K$ is genuine direct current acquaintance with this $Q$, then $Q$ participates in its epistemic basis. A matching response generated independently of that basis is not thereby direct acquaintance with it. Under ordinary causal or synchronic constitutive ordering, an appearance wholly after $K$ cannot supply its present basis. The motivation is contact with this actual occurrence rather than an accidentally or inferentially correct proxy. It names no favoured neural stage and permits a newly constructed $Q$ during current comparison. Direct phenomenal concepts and their content-making basis are a substantive constitutional possibility (Chalmers, 2003; Balog, 2012); unchanged physical vehicles do not automatically preserve that basis.

This principle is compatible with noncausal justification. An abstract theorem can justify a belief without being a material part of it. Current qualitative acquaintance instead concerns a concrete instantiated basis, but concreteness does not by itself make that basis physical. A particular judgment also need not be knowledge merely because $Q$ occurs: its classification may be mistaken.

8.2 Anti-luck constraints and their physical antecedent

Safety is an independently motivated necessary condition on knowledge, relative to a belief’s basis and relevant alternatives; basis individuation is not settled by the principle (Pritchard, 2012). Let $x_0$ be an actual episode and $N(x_0)$ a defended local domain. For an assignment law $L$, let $M_L(x)$ indicate the nominated appearance, and let $B_L(x_0,x)$ mean that an intentional target belief is formed on the epistemically same basis. This relation need not be transitive. Let $Y(x)$ be a physical affirmative vehicle.

Proposition 8.1: Scoped safety constraint

Assume: (i) at $x_0$ the subject knows the current relational appearance obtains; (ii) within $N(x_0)$ every same-target, same-basis belief is true; and (iii) an independently justified physical relation $B^*$ satisfies $B^*\Rightarrow B_L$ for every retained assignment. Then

$$\{x\in N(x_0):Y(x)=1,\ B^*(x_0,x)\}\subseteq\{x:M_L(x)=1\}.$$

Proof. The lower relation supplies the relevant same-basis belief. If its target appearance were absent, that belief would be false within the safety domain, contradicting assumption (ii).

The proof is short because the substantive work lies in $B^*$. Preserving a circuit, checking disposition or physical string does not establish preservation of an experience-constituted basis. An ancillary variable with no physical effect on the reporting process could still enter a proposed appearance law; excluding its relevance needs a source/basis argument. A transported true occurrence also does not create fresh knowledge at the destination, so the proposition cannot be iterated as a truth-to-acquaintance induction. The explicit three-vertex failure and two repair schemas are in Section 2.12.

Sound positive reports need not locate a bearer

Let $C=\{(a,b,a\oplus b):a,b\in\{0,1\}\}$, and let distinguishable buffers $B_1,B_2$ each range over $C\cup\{\bot\}$. Define the common physical report

$$Y(B_1,B_2)=\begin{cases}c,&B_1=B_2=c\in C,\\\bot,&\text{otherwise.}\end{cases}$$

One assignment associates a relational third with active $B_1$; another with active $B_2$. On every positive report both have the reported complete content. Disabling either buffer removes the report, so both physically matter. Yet at $(c,\bot)$ the occurrence assignments differ, and at unequal active values the content/bearer assignments differ. Both abstain publicly. Abstention is not an independently correct negative appearance judgment. This five-by-five model shows why positive soundness, diagonal anchors and difference-making do not provide occurrence completeness or unique carrier selection. Its actual-appearance labels are declarations used to test the inference.

8.3 A finite current scoped-carrier countermodel

Two candidate physical carriers can agree at every accepted current anchor without being physically interchangeable. Fix registers $u,v\in\mathcal P_2=\{0,1\}^2$, a bodily tag $b\in\{0,1\}$, and comparison mode $c\in\{0,1\}$. The registers are denoted $u,v$ here; the checker retains its historical register identifiers. The symbol $V$ denotes bodily expression and $B$ denotes the exclusive qualitative bearer. A state is $s=(u,v,b,c)$. There are thirteen native actions: idle $\iota$, two preparations each of $b$ and $c$, and four preparations each of $u$ and $v$. Their synchronous transition $T_d$ is

$$u'=\begin{cases}p,&d=C^u_p,\\u,&\text{otherwise,}\end{cases}\qquad v'=\begin{cases}p,&d=V_p,\\u,&\text{otherwise.}\end{cases}$$

The tag or mode is replaced only by its corresponding preparation. Crucially, ordinary copying into $v'$ uses the old $u$, including during a preparation of $u$. With $e(p)=\mathbf1_{\{p_1=p_2\}}$, the physical output is $y(s)=(v,e(v))$ when $c=1$ and $\bot$ otherwise. Both law extensions use this one transition/output process. Their hypothetical appearance assignments are

$$\mathcal L_u(s)=\begin{cases}(u,e(u)),&c=1,\\\bot,&c=0,\end{cases}\qquad \mathcal L_v(s)=\begin{cases}(v,e(v)),&c=1,\\\bot,&c=0.\end{cases}$$

These stipulated charts test an inference; they are not measurements that the registers actually present qualities. Both allow first presentation only during the nominated active current comparison.

Proposition 8.2: Scoped agreement and off-anchor disagreement

The model has 64 states and 832 state/action entries. On the eight nominated anchors $u=v$, $c=1$, both appearance laws and the output agree. Across the remaining states, the laws disagree on counterparts in 24 states and on equality in 16 states.

Proof. There are $4\cdot4\cdot2\cdot2=64$ states and $64\cdot13=832$ entries. Four diagonal pairs and two tags give eight active anchors. Of the sixteen ordered $(u,v)$ combinations, twelve have $u\ne v$; with $c=1$ and two tags these give 24 counterpart disagreements. Each equality class has two pair values. Ordered cross-class choices number $2\cdot2\cdot2=8$, giving 16 equality disagreements after the tag factor. At $c=0$, both assignments are $\bot$.

Complementing both bits of both anchored registers, or changing the irrelevant tag, preserves agreement: the checker performs $8\cdot2\cdot2=32$ such checks, with repeated states rather than thirty-two new observations. Both registers have physical routes to reports. Preparing $v=p$ changes the immediate output to $(p,e(p))$; preparing $u=p$ and then idling changes the later output to that value. Each yields four distinct outputs. This establishes efficacy of physical values, not efficacy of phenomenal properties or genuine acquaintance under every action. At each fixed b, the anchors still contain four counterpart values, so b alone cannot determine their complete content under a fixed intrinsic map. This excludes an isolated tag, not a whole body containing the varying fields.

Let $h(u,v,b,c)=(v,u,b,c)$ and interchange the labels $C^u_p,V_p$ while leaving other actions fixed. For every action, $hT_d(s)=T_{h(d)}h(s)$ holds exactly when $u=v$: for idle/tag/mode actions the two sides copy different old sources off the diagonal; for a register clamp one retained coordinate still differs. Therefore

$$\#\{(s,d):hT_d(s)\ne T_{h(d)}h(s)\}=13(64-16)=624.\tag{34}$$

The proposed exchange is not a native interventional isomorphism. Carrier ambiguity here concerns two appearance extensions of common physics; it is not earned by physical symmetry. Off-anchor states have no independently accepted acquaintance metadata, so their differing labels do not establish differing experiences in validated interventions.

The full candidate packages are $\{u,c\}$ and $\{v,c\}$, because changing c changes either complete assignment even at fixed register content. Within an explicitly fixed active-c scope, saying “carrier u” or “carrier v” merely treats that mode as background.

Proposition 8.3: Robust container and shared parts

For an independently warranted nonempty family $\mathcal B$ of complete candidate bearer-part sets, a package $\Sigma$ contains a complete bearer under every candidate exactly when $\bigcup\mathcal B\subseteq\Sigma$. Its smallest robust container is the union. The intersection lists parts common to all candidates but need not be sufficient.

Proof. The containment claim is $B\subseteq\Sigma$ for every $B\in\mathcal B$, equivalent to inclusion of their union. Membership in the intersection says only that a part belongs to every candidate; it does not include each candidate’s remaining parts.

Here the robust union is $\{u,v,c\}$ and the intersection is $\{c\}$. Mode alone cannot determine the four anchored counterpart contents. The union result makes an inclusive current realization compatible with internal-carrier ambiguity, provided occurrence, candidate completeness and physical realization have independent warrant. It neither makes every union component constitutive nor identifies the container with the experience-inclusive act. An open-ended exclusive-bearer extension is outside this finite family. The whole-act/CMP argument below, rather than set containment or these label stipulations, supplies the paper’s separately defended constitutive attribution. The portable check_scoped_carrier.py enumerates the stated physical tables and candidate supports without treating them as consciousness data.

8.4 Causal exclusion does not fix the first stage

A causal argument has four separate commitments: actual qualitative efficacy toward a nominated physical effect; physical causal completeness for that effect; rejection of two ontologically independent sufficient causes systematically overdetermining it; and an interpretation of the remaining dependence as physical grounding, constitution or identity. The first three motivate pressure against an independently extra cause, but do not themselves prove the fourth or reduce qualitative property types. Causal completeness can support physical constitution while leaving reductive and nonreductive views open (Montero and Papineau, 2016).

Even strict event-token identity does not select an earlier stage. Independently fix input pair $(a,b)$, $c=(a,b,a\oplus b)$, and the physical chain

$$I\longrightarrow M\longrightarrow J\longrightarrow R\longrightarrow V.$$

At increasing clock times it computes, retains, retrieves and copies the tuple. Each of the first four stages has six controls: keep its incoming value, disable it, or prepare one of the four tuples. The memory therefore needs five values, not a sensory movie. In the early identity interpretation, active $I$ is an early relational presentation and active $J$ is a later presentation. In the delayed identity interpretation, $I$, $M$ are nonphenomenal for this target and active $J$ is its first presentation. Both identify their actual presentations with existing physically efficacious events, use the same transition functions and admit the same current acquaintance at $J$ when $R$ normally copies it. Serial antecedents are not independent overdetermining apparatuses.

There are $4\cdot6^4=5184$ controlled episodes, with no output disagreement. First-appearance assignments differ whenever $I$ is active, giving $4\cdot5\cdot6^3=4320$ cases. For each input, $J$ is active in $4\cdot36+29=173$ first-three-control combinations: four explicit active retrieval preparations, plus normal retrieval from an active memory; the latter has $4\cdot6+5=29$ cases. Normal copying by $R$ then gives $4\cdot173=692$ shared current-knowledge interpretations. These counts verify the stipulated finite rules; they neither certify knowledge nor show that the tuple has qualitative counterparts. The important result is that granting current acquaintance, closure, efficacy, no independent overdetermination and even token identity still leaves earlier occurrence open. A delayed third at $J$ meets the main target.

8.5 A reader model and a typed realization failure

A reported change need not be a quantified latent-appearance change without a reader restriction. For example, report occupancies .65 and .46 can coexist with latent occupancy .65 in both conditions: the first reader is accurate, while the second misclassifies a fraction 19/65 of positive cases and no negative cases. Its output is $.65(1-19/65)=.46$. Alternatively, accurate readers can track changed latent occupancies. This is an exact illustrative model, not a fit to participants. Adaptation and neural encoding may share the same physical relation variable under either assignment; their correlation does not supply an independently calibrated indicator error. Proposition 2.4 in Section 2.5 cannot intersect different feasible assignment maps.

The analogous realization inference is invalid:

$$\begin{gathered} P\text{ realizes vehicle }V,\quad Q\text{ constitutes the content/basis of }K\\ \not\Rightarrow P\text{ realizes }Q. \end{gathered}$$

For a counterinterpretation, set $V=P$, let $\operatorname{content}(K)=f(V,Q)$, and let independently grounded $Q$ contribute the phenomenal dimension. Vehicle realization and content constitution both hold; a material bearer for $Q$ has not been supplied. A valid stronger premise would say that $P$ realizes the whole direct acquaintance token through a defended literal-constituent/mode-preserving relation. Giving its output vehicle the name “quotation” does not establish that premise.

Literal current embedding is nonetheless a substantive alternative. A complex can physically include current sensory submodules, a relation operation and a control frame rather than merely referring to missing targets. Live inclusion and a detached snapshot differ under post-loading native interventions. The proposal that such embedding constitutes conscious display is stronger than representational reference (Coleman, 2015). Its independently specified awareness role is still required; an unconscious controller can contain and use actual material counterparts. The constitution of phenomenal concepts can explain acquaintance while remaining neutral on physical versus nonphysical qualities (Balog, 2012). A new current counterpart token suffices: demanding the original source token would unnecessarily rule out imagery and reconstruction.

8.6 Possessed reasons, operative expression, and a mixed act

Possessing a reason while acting differs from acting because of it (Davidson, 1963). A person can currently apprehend a colour pair while an independent sensor makes their hand emit the same comparison. Truth and reliable agreement need not show which reason operated. The appropriate expression principle requires a suitable causal or constitutive participation relation when an utterance is explained as based on this particular current acquaintance. It does not require every external truth-maker or epistemic condition independently to move the body.

For an exact common-cause construction, fix physical input $x\in\{0,1\}$ and instructed motor code $c\in\{0,1\}$:

$$R=x,\qquad V=R\oplus c,\qquad\operatorname{decode}(V,c)=x.$$

Separately grant actual qualities $q_x$, genuine current acquaintance with proposition $\psi_x$, and a fixed law linking them to x. All four lawful rows have truthful decoded reports; physical input changes preserve safety and sensitivity, and code remapping preserves semantic content. If $P$ is their common source, these agreements do not establish a $Q/K\to V$ edge. A separate source test can exclude a stale record controller: replace $R=x$ with $R=b$, where b is independently variable, and vary b while independently preserving current $q_x$. The cached controller can be wrong while the current controller remains correct. The lawful common-cause escort passes the current-source test too.

The strongest rival improves on idle parallel knowledge. Grant a genuine contentful reason $R$, actual qualitative grounding and appropriate cognitive binding, and define

$$K^*=(Q,R,\operatorname{Grounds}(Q,R)),\qquad V=f_c(R).\tag{35}$$

$R$ is both a physical component of $P$ and an operative member of this one grounded act. $Q$ is a literal current qualitative constituent, but the exclusive-bearer interpretation assigns it to $B$ and denies that it is a mode of $P$. Whole-through-part attribution can make $K^*$ operative through $R$ without adding a second sufficient physical cause or making every constituent independently efficacious. This is not arbitrary grafting when the actual grounding and binding are granted; those are substantive premises shared with the competing interpretation. It contains one agent and one comparison.

An independent $Q$ override at fixed $R$ may remove factivity or make the report false, but it cuts the strict law linking them. It is not automatically an available native intervention or a nearby same-basis world. Removing $Q$ can leave a belief’s effect intact while changing knowledge to mere belief. Conversely, requiring $Q$ itself to form $R$ adds a causal-basing premise; even that relation can be mediated by a distinct bearer. Thus source-based expression improves act nomination but does not by itself identify the complete physical bearer.

Practical source use

Deliberately selecting present apprehension rather than a remembered answer, knowing competent means and executing that undertaking normally can provide prima facie practical knowledge of the operative reason. This is stronger than merely feeling effective and is defeasible by bypass or confabulation. Ordinary task instructions support such an attribution less directly than an independently grounded practical claim. More importantly, the mixed act in (35) can genuinely use the selected reason through $R$. Practical source use identifies the actual activity; the claim that its whole qualitative basis is a mode of the physical execution still needs the constitutive rule developed below.

8.7 Categorical assembly and physical inherence

Russellian approaches motivate using qualitative or protophenomenal nature as the intrinsic ground of physical structure (Alter and Nagasawa, 2012). This promises explanatory integration beyond an arbitrary correlation, but requires separate steps. Local bases $b_i$, their natures and couplings $L$ may jointly ground component powers and, through composition, a whole native kernel. Grounding composition gives a plural ground; it does not give one presentation. The mental property whose causal role needs explanation must be distinguished from microphysical grounds (Baysan, 2025).

Fix four native units, all local grounds, couplings and intervention kernels. One extension admits a new relational field for each of the six pairs; another admits those fields plus a whole four-member field. Both can be permutation invariant and share one actual pair anchor. They differ in assembly while agreeing on physical power. If pair fields create fresh counterpart tokens, type symmetry cannot transport the same-token anchor. If they reuse the same tokens, every pair can share some field without one field containing all four, as Section 2.2 showed.

There is a positive conditional repair. Let complete assembly-relevant data have automorphism group $G$, and let each fixed actual token $m_i$ have exactly one maximal owner $o(m_i)$. Assume pointwise covariance of same-owner relations and an anchored pair. Every pair in its G-orbit has equal owner values. If the resulting orbit graph connects all tokens, equality along paths gives one actual owner containing all of them. The proof uses the actual owner function, not existential pairwise transitivity. Complete categorical/material symmetry, fixed-token identification and unique ownership each need independent warrant; abstract native-kernel symmetry alone supplies none of them.

Even a directly proposed macro $Q$ grounding the whole power need not automatically inhere in $P$. The relations

$$\operatorname{GroundFor}(Q,D(P))\qquad\text{and}\qquad\operatorname{Mode}(Q,P,\tau)$$

are different. A distinct $B$ could primitively ground that power while denying qualitative inherence in $P$. A definition making categorical bases intrinsic properties of their power bearers would exclude this version, but the inherence premise would then be explicit. Grounding transitivity alone does not establish it.

8.8 Harmony and qualitative powers

Actual qualitative character fixes success conditions for comparison: judging which counterpart appears larger differs from producing whatever bit a controller emits. Unpleasantness supplies an intelligible reason for relief. This fit calls for an explanation. Its normative direction must still be distinguished from actual basing and from a physical causal power. Harmony arguments must also distinguish one global qualitative law from independently redrawn trial labels (Saad, 2022).

Let $G$ be a global appearance-law parameter and $Z$ a physical controller/genotype. Assume a product prior $\mu(G)\nu(Z)$ and physical-history likelihood $L(h\mid Z)$ independent of $G$. Then

$$\Pr(G\mid h)=\mu(G)$$

for positive-probability histories, because marginalizing $Z$ supplies the same factor for every $G$. This does not prevent physical selection increasing harmony within an already aligned world. For two genotypes with fitnesses four and one and initial shares 1/3, 2/3, the first share after t generations is $4^t/(4^t+2)$ under either law. Initial law/controller correlation or genuinely phenomenal evidence can change the law posterior; product independence is not a consequence of epiphenomenality alone.

For four independently ordered quality levels and one fixed permutation of four source levels, a single ordered-pair constraint leaves twelve of twenty-four bijections; repeating that noiseless pair leaves twelve, two disjoint ordered pairs leave six, and a spanning adjacent chain leaves one. Equality/difference preserves all twenty-four. The number 1/24 is a declared uniform-prior calculation, not an objective probability of fortuitous consciousness. A purpose-sensitive controller $a=b\operatorname{sign}(j-i)$, with instruction $b=\pm1$, can act appropriately under one fixed aligned quality law on every pair and both purposes while its physical arguments remain source and instruction. Flexible normative success alone does not identify $Q$ as its power.

A more direct powers proposal

Felt aversiveness makes relief-directed motivation intelligible from its character, rather than from a learned warning code alone. Experienced trying can nevertheless fail, so an intrinsic-power argument should distinguish a motive toward an attempt from successful movement ($M$ørch, 2019). Competing goals and independently specified enabling conditions matter. The pain-asymbolia objection must distinguish changed affective character from failed implementation; a shared pain label does not establish identical total phenomenology, and defining the relevant quality by withdrawal would beg the question. Seeking painful stimulation can express an opposing motive; unchanged aversiveness with intact enabling conditions, no opposing motive and no avoidance attempt would instead threaten strict necessitation. These are distinct philosophical tests, not clinical findings established here. A felt urge to settle a comparison can be an operative motivational facet without containing the complete sensory pair. Efficacy of that motive does not transfer to every qualitative constituent of the scene; a genuine power of $Q$ toward mental recognition also need not be a physical-report power. Fallible attribution of agency is a further reason to separate felt effort from its successful physical efficacy.

The stronger affective target makes the counterparts themselves aversive. Synchronized thermal-pair reports support jointly felt pain with connected/disconnected spatial content, including radiation at a neutral source location (Quevedo and Coghill, 2009). Those observations concern spatial appearance, not measured comparative aversive pull or relief priority. Felt badness, felt pressure, reason-giving force and implemented priority should not be defined by one another (Bain, 2019). For a proposed comparative-power test, relief alternatives require independently matched expected efficacy, delay, cost and side effects, understood targets and control capacity; the strongest felt pressure is not defined as whatever wins the motor competition.

An exact physical construction exposes the remaining inference. Set

$$w=(a,dab,b),\qquad a,b\in\{0,1,2\},\quad d\in\{0,1\},$$

and let a fixed selector choose the greatest active weight with a fixed tie rule. The central channel has no source there and has mixed difference one over $a,b\in\{1,2\}$ when $d=1$, so it is not a sum of separate source functions. At $a=b=2$, changing $d$ changes $(2,0,2)$ to $(2,4,2)$ and makes the central target dominant. Source-only fixed motives need current generation to account for that target. Yet even granting genuine physically efficacious qualitative modes to all active channels, an actual jointly presented field can be assigned either to the whole physical exercise or to exclusive $B$, with the same component powers and selector. A generated source-free motive does not prove whole-field inherence. No such gap-target priority was measured in the thermal study.

If the actual relational $Q$, in virtue of its qualitative nature, exercises the very causal power producing a nominated physical comparison effect, closure/exclusion provides additional placement pressure. Mental-only effects, grounds of another power and efficacy of a whole through one proper physical part do not establish that antecedent. The complete-act route avoids demanding a separate effect from every qualitative constituent.

8.9 Whole-act realization and literal mode preservation

Suppose $K^*$ genuinely guides $V$ through $R$, $Q$ is its literal current constituent or mode, and the whole $K^*$ has complete physical constitution. If the realization preserves those literal parts/modes rather than output-equivalent surrogates, a current physical act bears the actual qualitative basis. No separate $Q\to V$ edge is required. The difficult premise is complete whole-act constitution; physical sufficiency for $V$ realizes less.

The unrestricted rule “every whole with a physical effect is wholly physical” fails. For any effective physical $R$, a fusion $R$ + $B$ can inherit its effect while retaining a distinct constituent. The strongest mixed knowledge act makes $B$ genuinely epistemic, so dismissing it as arbitrary grafting begs the question. Restricting to minimal causal support would instead remove real constituents irrelevant to one tested output, losing the route’s purpose.

Immanent organization improves the nomination. Part roles belong to what the organized whole is doing rather than to an observer’s inventory: an input is this activity’s source, a state its reason, and a movement its uptake because of their participation in the maintained exercise. Hylomorphic accounts explicitly treat organization as immanent activity and perceptual engagement as coordinated exercise (Jaworski, 2020). The role description adds participation in the maintained activity to the specification of transitions. An immanently unified hybrid can also have genuine part roles without an external unifier, so a regress against separate unifiers does not establish exhaustive physical constitution.

Three relation types must remain fixed:

$$\begin{aligned} A=P &: \quad\text{token-performance identity},\\ \operatorname{Constitutes}(M,A) &: \quad\text{constitution, possibly }M\ne A,\\ \operatorname{Mode}(Q,A,\tau) &: \quad\text{literal qualitative mode in a phase.} \end{aligned}$$

From identity, $\operatorname{Mode}(Q,A,\tau)$ entails $\operatorname{Mode}(Q,P,\tau)$ by substitution. Constitution by numerically distinct $M$ does not transfer every predicate to $M$. Complete physical constitution may make the constituted A a physical bearer, but naming $M$ itself requires performance reidentification or an independently warranted mode-preserving realization relation. Neither identifies instantaneous $Q$ with an entire response window.

8.10 The corrected concrete-performance/mode principle

Process identity and multiple-description event individuation are established philosophical options (Place, 1956; Davidson, 1969). Mere systematic correlation is weaker than identity. Likewise, the stronger economy argument against irreducible psychophysical additions cannot simply be inherited when fundamental qualitative modes are permitted (Smart, 1959). The present principle therefore concerns actual instance placement, not a numerical event-saving proof.

Conservative mode placement

Given an independently established literal mode of a particular actual activity, presumptively assign it to that activity’s positively tracked concrete execution, at a fixed activity/phase grain and with its actual executing contributors included. Revise the attribution when independent evidence or a defended ontology identifies another instantiator, an omitted constitutive contributor, a merely supplied ground or a different activity. Descriptive vocabulary and irreducible property type alone do not establish a numerically different bearer.

The rule has three admission conditions. First, the mode is an actual way the activity is undergone or constituted, rather than only an external truth making a result correct. Second, the nominee is tracked through actor, source use, operations, uptake and participation history rather than output agreement. Third, the nomination respects known execution/phase counterexamples. These are grounds for a defeasible identity conclusion. “Complete experience-inclusive execution” must not be inserted into the physical premise.

Its positive warrant is continuity of a tracked instance across descriptions. A concrete property is a way an independently identified activity is; introducing a new property vocabulary need not introduce another performance. Positive execution paths distinguish this locus from its indicator, supplier or copy. At the same time, literal modes must not be transferred to every containing system. Consider the following corrections:

1. A client’s input/output history does not execute remote multiplication. If multiplication is delegated, the remote processor belongs to the calculation; if it merely supplies an operand, it need not be part of a subsequent local calculation.
2. A local signature can be the signing execution while protocol authorization additionally requires an actual remote verifier/ledger operation. Abstract validity rules are not automatically concrete parts; an actual completion operation is different.
3. Biological control does not close the boundary of a prosthetic grasp or instrumental performance. Include the actual jaws/contact or instrument when the selected act requires them.
4. Simultaneous saying and writing the same word can share actor, source, purpose and time while one vocal execution is only a member of the spoken-and-written whole. A continuously tracked single inscription instead supports bodily/intentional reidentification through actual stroke/mark participation.
5. A promise now is not its fulfillment tomorrow. Shared actor and intentional content do not erase the change of temporal grain.
6. A thermometer or copied inscription can track the same value/content without sharing the measured state’s instantiator or original execution. Common ownership and accuracy do not supply token identity.

These cases prevent selecting a desired bearer solely by actor, role or time. They supply constraints on the general rule rather than analogies proving qualitative reduction. In the main application, physically specified native sensory/comparison participation and independently ordinary joint-appearance/activity grounding converge on one current exercise. CMP supports reidentifying that experience-inclusive act with its organized physical engagement at the stated grain. That is the substantive, contestable extension; it is neither observed in a neural measurement nor a restatement that the desired stage appears.

A fair exclusive-bearer alternative

The rival can retain one agent, one well-bound $K^*$, actual $Q$, physical $R$ and one simple generic content law. It denies that $Q$ is a mode of the nominated $P$. It need not add another complete comparator, physical force or separately fitted law for every trial. If the agreed observation laws coincide, their likelihood ratio is one; an independently defended instance principle can still supply a rational preference, but another fit of shared data cannot supply experimental ontology selection.

A copy or view paired bijectively with each performance token may merely redescribe that token. An ontically distinct qualitative trope u with $\operatorname{Instantiates}(P,u)$ likewise permits $P$ as bearer. The genuine rival adds $\operatorname{Bearer}(B,Q)$ and $\neg\operatorname{Mode}(Q,P,\tau)$, or an equivalent nonlifting constitution denial. That substantive placement needs its own individuation/grounding rationale. The immanent account also owes occurrence/content and extension laws; calling its primitive mode free would prejudice the comparison. Raw event or symbol counts depend on the ontology/coding language and do not settle these debts.

The rule can have modest actual warrant without eliminating every coherent rival. Its preference combines independent actual mode, positive role-sensitive execution nomination and a general placement discipline with correction cases. Absence of evidence for $B$ alone would be insufficient. An independently supported sensory executor outside the nomination, a bypassed reason, a distinct phase or a defended instantiator ontology can defeat or refine it.

8.11 Ownership, phase, and the broader programme

Identifying the ordinary experiencer with an animal identified through biological criteria can support literal ownership of current appearance (Olson, 2003). It does not establish which current organized event constitutes that appearance. A sole thinking-subsystem account can avoid duplicate subjects too. Moreover, strict current subject–animal identity has diachronic implications when the animal persists. The existing A3 perspective-continuation account treats a genuine qualification gap as ending a numerical experiential subject episode while biological/person continuity may remain (Rodgers, 2026e). These are different identity types. A subject-role or episodic-viewpoint claim is weaker than strict animal identity and must be named accordingly. Current CMP adopts neither a universal animalist persistence rule nor an equivalence with A3.

World involvement is also open at the correct grain. An external object need not itself perform the native comparison to be a constituent of the perceptual activity. Actual input, coupling and context should be explicit; the conclusion is not internal-only supervenience. Conversely, material source locations and represented locations are not the bearer. Direct cortical and neutral-site pain appearances constrain source-location shortcuts without fixing a unique current field.

The scaffold insight has a legitimate narrower role. Retained results can change future relevance, selection and interpretation; directed reconciliation of an actionable represented discrepancy differs from the mere existence of a physical difference. Thought produces a constrained representation rather than the source itself, and a successful local reconciliation can introduce new differences. This recursive development does not imply that every discrepancy controller has subjective will, that two communicating agents share one field, or that operational information is already qualitative (Rodgers, 2026c,a). It situates the present act within development while preserving the independently required occurrence/constitution distinction.

8.12 How the current connection differs from the conditional routes

The main inference separates occurrence projection from constitutive placement. Within the narrow physically nominated human regime, ordinary sensory experience, source-directed comparative practice and material continuity support projecting another current presentation. Prior task participation and sensory matching do not themselves provide dedicated measurements of jointness or basing. Present $Q$, seeing, current acquaintance and answer correctness are not eligibility predicates. The projected target is tethered to source-directed inspection/comparison, not just $Q$ somewhere in a task window. Biological/operational continuity and current participation are partly ordinary inferences, not complete neural measurements.

CMP then supports the actual activity’s reidentification and literal mode placement. It does not create the occurrence predicted by the first rule. A current-Q interpretation may subsequently support the prediction, but cannot retroactively select the eligible class. An eligible mistaken answer can coexist with joint appearance; a credible field-free eligible case counts against projection.

The three strongest objections therefore have different targets. A serial-only sequence with only a relation proposition challenges jointness. A passive current field accompanying an independently automated answer challenges the activity tether and nomination of the answer-producing operation; split or narrow the activities instead of widening a containing time-window. An exclusive-B hybrid can grant jointness and tether while challenging CMP. A delayed account that first presents both counterparts and their relation at current uptake $J$ satisfies the target and need not be defeated. Keeping these alternatives separate yields a positive current-act connection without asserting first sensory onset, universal machine admission, exact anatomy or reduction of qualitative types.