Addendum 113 — R²D and the Black Hole Information Paradox

Recursive Replacement and the Resolution of a Boundary Identity Contradiction

Abstract

Recursively Realized Distinguishability (R²D) evaluates a physical account by asking which boundary defines its states, which realizations are compatible, and which distinctions remain readable. A paradox generated by extending state identity beyond its defining count boundary is resolved by rejecting that extension. Its competing narratives need not be reconciled.

Applied to black holes, this method separates horizon multiplicity, subordinate identity, and preparation–radiation distinguishability. The horizon has a boundary-defined multiplicity whose logarithm receives an area calibration. Its directed enclosing recurrence can be unreadable while its realized count remains defined. Recursive replacement changes state identity; a readable relation can nevertheless persist.

A joint boundary supplies an exact consistency test. Recovery of a specified preparation class from a complete radiation record requires that each admissible radiation record be compatible with only one preparation class. Two finite count constructions show that recovery and nonrecovery are separately compatible with the replacement rule and the same coarse count totals. A further construction shows that neutral cyclic closure does not force a selected preparation–radiation offset to vanish.

The result resolves the boundary identity contradiction and removes compulsory recovery as a consequence of replacement or horizon unreadability. Particular radiation correspondences remain properties of specified enclosing compatibility relations. Neither their preservation nor their failure defines an internal R²D paradox.

1. What counts as a resolution

A paradox is a conflict among claims made under specified conditions. Its resolution can consist of rejecting a premise that makes those claims incompatible. It need not produce a third story in which incompatible ontologies become simultaneously true.

R²D begins with boundary-defined statehood. Its analysis therefore starts before a physical narrative assigns persistent objects, trajectories, storage locations, or causal agents to the readable counts. A representation can express a successful relationship without establishing its accompanying ontology.

“Information” is not introduced here as an additional R²D primitive. When the discussion concerns whether a radiation record distinguishes a preparation, that relation is stated explicitly. When it concerns persistence of a subordinate object, that is a different assertion and is tested against replacement.

The target of the resolution is consequently exact: an account cannot require a boundary-defined subordinate identity to persist unchanged across replacement merely because that identity was previously readable. Such a requirement conflicts with the definition of the replacement. A separate relational recovery condition must be stated through the enclosing compatibility relation.

This follows the boundary-projection diagnostic of Addendum 84 and the claim-specific consistency standard of Addendum 89. [4, 5]

2. The boundary defines the state

At a count boundary \(B_n\), let

\[ \pi_n:\mathfrak M_n\longrightarrow\mathcal M_n \]

classify compatible joint subordinate realizations into readable macrostates. For an admissible state \(i\),

\[ W_{n,i}=\left|\pi_n^{-1}(i)\right|, \qquad S_{n,i}=\ln W_{n,i}. \]

The compatible realization domain and its classification belong to the enclosing boundary. A formal microstate has its meaning relative to that classification.

When the count boundary changes, the subordinate state, source, transition, and path do not automatically retain the same identities in the new domain. Part I expresses the change as

\[ \mu_n\not\equiv\mu_{n+1}. \]

This is recursive replacement. It is stronger than keeping an unchanged microscopic identity hidden beneath a coarser description. [1, Appendix A, Axioms 1–4]

Replacement does not assert that every relation becomes unreadable. Part I explicitly distinguishes replaced identities from coherence of a relation across boundaries. [1, §7.12] Therefore,

\[ \text{replacement of state identity} \]

and

\[ \text{preservation of a readable relation} \]

are logically distinct propositions.

Both must be kept distinct throughout the black-hole argument.

3. What remains readable at the horizon

The R²D horizon is a boundary of recursive readability. Its schematic readability conditions are

\[ \mathcal R(\lambda_H)=1, \qquad \mathcal R(\rho_{H+1\mid H})=0. \]

The directed enclosing recurrence is unreadable from the relevant horizon or exterior read. That does not mean recurrence has ceased or the boundary is unrealized. Part III explicitly distinguishes unreadable recurrence from absence of recurrence. [2, §§9.5–9.11]

The horizon multiplicity is

\[ W_H, \qquad S_H=\ln W_H. \]

Under the area calibration discussed in Part III and Addendum 43,

\[ \ln W_H=\frac{\mathcal A_H}{4\ell_P^2}. \]

Area therefore calibrates the logarithmic multiplicity. It is not a catalogue of individually readable interior identities, and the calibration does not turn those identities into permanent contents awaiting later release. [2, §§9.3–9.5; 3]

The thermal phase-frequency read must also remain distinct from the directed enclosing recurrence. Addendum 49 makes this distinction explicit. Substituting the former for the latter would introduce a bridge that the two quantities do not define by themselves. [6, §36]

These statements identify readable count structure and its limits. None specifies a preparation–radiation recovery rule.

4. The joint boundary

Let \(B_J\) own the classification of readable preparation, horizon-change, and detector records. Work first at a fixed horizon-change class \(h\) and detector setting \(d\). Write a readable joint class as

\[ j=[x,h,r]_d, \]

where \(x\) is a retained preparation record and \(r\) is a complete radiation record at the stated detector resolution.

These labels are features of one enclosing read. They do not restore the components as separately occupied primitive states alongside that read. A retained preparation record is itself a present distinction of the joint classification; it is not the original subordinate state identity carried unchanged through the horizon.

Let

\[ \pi_J:\mathfrak M_{J;h,d}\longrightarrow\mathcal J_{h,d} \]

be the enclosing classification. Its internal multiplicity is

\[ W_{J,j}=\left|\pi_J^{-1}(j)\right|. \]

Define the readable record features

\[ X(j)=x, \qquad E(j)=r. \]

The joint compatible counts are

\[ C_{J,h}^{(d)}(x,r) = \left| \left\{ \mu\in\mathfrak M_{J;h,d}: X(\pi_J(\mu))=x,E(\pi_J(\mu))=r \right\} \right|. \]

The count refers to possible compatible realizations. It is not a probability, a realized frequency, or a trajectory through persistent interior objects.

If \(x_0\) and \(x_1\) are readable preparation distinctions, then

\[ [x_0,h,r]_d\ne[x_1,h,r]_d \]

can hold even though the radiation record is the same. The joint read distinguishes the preparations through its retained record. That fact alone establishes no recovery from radiation.

There is also an essential bookkeeping distinction: omitting \(x\) from a tuple restricts the recorded features. By itself, it does not demonstrate recursive replacement. The following test compares readable features within the same \(B_J\), so it requires no identification of that omission with a physical boundary change.

5. Exact recovery is a compatibility condition

For the specified record classifications, exact recovery means that there is a decoding relation \(D\) on the admissible radiation records such that

\[ \boxed{X(j)=D(E(j))\quad\text{for every }j\in\mathcal J_{h,d}.} \]

This is a statement about distinguishable record labels.

Proposition 1. Exact recovery holds if and only if

\[ \boxed{ E(j)=E(j') \ \Longrightarrow\ X(j)=X(j'). } \]

Proof. If \(D\) exists, equal radiation records give equal decoded preparation classes. Conversely, if every radiation record has only one compatible preparation class, that class defines \(D(r)\). Thus \(X=D\circ E\). \(\square\)

In compatible-count notation, define

\[ \mathcal X(r)=\{x:C_{J,h}^{(d)}(x,r)>0\}. \]

Then exact recovery is equivalent to

\[ \boxed{|\mathcal X(r)|=1\quad\text{for every admissible }r.} \]

Nonrecovery of the chosen distinction means that some admissible record has

\[ |\mathcal X(r)|>1. \]

This criterion concerns exact recovery of the stated preparation labels from the stated records. It does not turn a count table into a quantum evolution law. Distinct statistical tendencies across repeated records are also different from exact recovery from one complete record.

The conditioning must remain explicit. The test above holds within fixed \(h,d\). If \(h\) is unavailable to the radiation reader, recovery must instead be tested using

\[ \widetilde C_J^{(d)}(x,r)=\sum_h C_{J,h}^{(d)}(x,r). \]

A description cannot silently use a horizon label while claiming recovery from radiation alone.

6. Two exact compatibility witnesses

Let the readable preparation labels be \(x=0,1\), the radiation labels be \(r=0,1\), and let \(M\ge2\) be an integer. Consider two alternative enclosing compatibility specifications:

\[ C_{\mathrm{rec}} = M\begin{pmatrix} 2&0\\ 0&2 \end{pmatrix}, \qquad C_{\mathrm{overlap}} = M\begin{pmatrix} 1&1\\ 1&1 \end{pmatrix}. \]

Rows label preparations; columns label radiation records.

In \(C_{\mathrm{rec}}\), each radiation label has exactly one compatible preparation label. The decoder is \(D(r)=r\).

In \(C_{\mathrm{overlap}}\), each radiation label is compatible with both preparations. No such decoder exists.

Both tables have row totals \(2M\), column totals \(2M\), and total multiplicity \(4M\). Thus these coarse counts do not decide whether the preparation distinction is recoverable.

A direct realization of either table assigns one readable joint state to each nonzero entry and gives that state the indicated multiplicity. Every admissible joint state then has more than one compatible subordinate realization. The preparation label can be recovered in the first model while the subordinate realization remains individually unreadable.

Proposition 2. Nontrivial internal multiplicity is compatible with either exact preparation recovery or overlapping preparation possibilities.

The two constructions prove the proposition. Neither requires an unchanged subordinate object to become readable again.

These are alternative compatibility specifications. They are not simultaneous, contradictory claims about one already-fixed table. Their shared coarse totals establish a count-level consistency result; they do not assign either table to an observed evaporation process.

7. Complete cyclic closure does not force pair neutrality

The cyclic construction provides a second test. Let \(G=\mathbb Z_K\) classify ordered closure positions. A proposed \(B_J\)-owned rule can relate readable phase labels by

\[ \varphi_R(r)-\varphi_P(x)=q, \qquad q=u+v\pmod K. \]

For a prespecified allowed pair set \(H\subseteq G^2\), the offset counts are

\[ m(q)= \left|\{(u,v)\in H:u+v=q\}\right|. \]

The neutral-pair choice

\[ H_{\mathrm{neutral}}=\{(u,-u):u\in G\} \]

gives

\[ m(0)=K, \qquad m(q\ne0)=0. \]

For \(K=2\), this supplies the diagonal count pattern. It expresses a particular compatibility rule.

Now specify a complete three-role cyclic relation at \(B_J\):

\[ u+v+w=0\pmod K. \]

The preparation–radiation offset remains \(q=u+v\). Complete neutrality therefore gives

\[ \boxed{q=-w,} \]

rather than requiring \(q=0\).

For \(G=\mathbb Z_2\), the allowed triple set

\[ T_{\mathrm{split}} = \{(0,0,0),(0,1,1)\} \]

is neutral in every complete triple. Nevertheless, its two preparation–radiation offsets are \(0\) and \(1\), giving

\[ m(0)=m(1)=1. \]

For comparison,

\[ T_{\mathrm{neutral}} = \{(0,0,0),(1,1,0)\} \]

also contains two neutral complete triples, while both selected pair offsets vanish. It gives

\[ m(0)=2, \qquad m(1)=0. \]

Proposition 3. Neutrality of a complete cyclic relation does not imply neutrality of a selected pair of its roles.

The two triple sets prove the distinction exactly. The label \(w\) denotes a role in the specified cyclic relation; it has not been identified with a physical interior reservoir. The triples are jointly admitted by \(B_J\). Their component labels do not assemble the enclosing occurrence from below.

Addendum 72 supplies the cyclic composition language. It does not make every chosen pair the entirety of the enclosing closure. [7] Imposing \(u+v=0\) on preparation and radiation therefore requires specifying why those roles exhaust the relevant closure, or why the remaining contribution is zero.

A thermal frequency read or a detector registration cannot substitute for that compatibility specification merely by being associated with a black hole.

8. Where the contradiction would enter

For one fixed \(B_J\), one fixed compatibility relation, and one fixed record resolution, the following statements are incompatible:

1. Every admissible radiation record determines exactly one preparation class.
2. Some admissible radiation record is compatible with two distinct preparation classes.

Proposition 1 establishes the conflict. The replacement rule does not impose both statements. Horizon unreadability does not impose both statements. Complete cyclic neutrality does not impose both statements.

The contradiction is introduced when a description asserts universal recovery independently of its compatibility rule while also allowing overlapping preparation–radiation possibilities under that same rule.

A recovery requirement is admissible as an explicit enclosing constraint. It is not a universal consequence of replacement. Conversely, replacing identity or losing access to a recurrence does not prove that every preparation-sensitive relation is absent.

This separates three assertions that cannot be substituted for one another:

Assertion	R²D status in this argument
The same subordinate occupied-state identity persists unchanged across recursive replacement.	Conflicts with the stated replacement of identity.
A specified readable preparation relation is recoverable from radiation.	Requires the unique-compatibility condition of Proposition 1.
No such relation is recoverable from radiation.	Requires failure of that condition for the specified records.

Neither of the final two assertions becomes universally true merely because the first assertion is rejected.

9. The resolution established by the argument

The boundary identity formulation of the paradox demands that a state retain its identity after the conditions defining that identity have changed. R²D resolves that formulation through its definition of replacement. No reconciliation of rival object narratives is required.

The preparation–radiation formulation is then expressed as a compatibility question. Its two outcomes are separately consistent with the count architecture examined here. An unspecified outcome is an unspecified relation; it is not a contradiction.

This distinction also clarifies the black-hole paragraph in Addendum 84. Its demand for a horizon-to-exterior mapping applies to claims about the actual radiation correspondence. Such a mapping is not a prerequisite for rejecting an inadmissible identity assumption. Conversely, rejection of that identity assumption does not determine the radiation correspondence.

The present resolution is therefore stated positively and at its demonstrated scope:

R²D removes the demand for persistence of subordinate identity across replacement. It represents recovery of a preparation distinction as an enclosing compatibility condition. Replacement, horizon unreadability, and cyclic closure do not impose contradictory recovery requirements.

The argument does not identify all formulations of black-hole quantum evolution with exact decoding of the record labels used here. Its mathematical result applies to the stated distinctions and compatibility relations. Extending a claim to another formulation requires translating that formulation’s conditions explicitly.

Neither the absence of an R²D primitive named “information” nor the absence of a discovered contradiction would establish this result by itself. The content of the argument is the explicit distinction between identity and relation, the recovery criterion, and the compatibility witnesses.

10. What this adds

The prior horizon account established that multiplicity and readable distinguishability can remain when directed enclosing recurrence is unreadable. The present argument adds four exact results:

1. Recovery of specified preparation labels is equivalent to uniqueness of the compatible preparation class for each radiation record.
2. Identical coarse count totals admit both recoverable and nonrecoverable joint descriptions.
3. A recoverable preparation relation can coexist with nontrivial unreadable subordinate multiplicity.
4. Neutral complete cyclic closure can coexist with nonzero preparation–radiation offsets.

These results establish a concrete consistency analysis before any radiation account is promoted into a causal ontology. R²D does not need to reconcile incompatible stories. It requires each asserted state, distinction, and relation to be defined by the boundary that owns its count.

Sources

1. R²D Canon, Part I — A Theory of Dimensionless Recursive Counting. Canon revision 2026-09-25, author-provided part-i.md; §§3.2, 5.1.3, 7.12; Appendix A, Axioms 1–4.
2. R²D Canon, Part III, R2D-Part-III.md, Chapter 9, especially §§9.3–9.11.
3. Addendum 43, Horizon Multiplicity Occurs Before Planck-Area Calibration, especially §§3–11.
4. Addendum 84, Paradox as Boundary Projection, especially §§1, 9–10.
5. Addendum 89, Logical Consistency Before Physical Comparison, especially §§1–2, 6–8.
6. Addendum 49, Horizon-Branch Distinguishability Occurs Before Hawking Recurrence, especially §36.
7. Addendum 72, Closure Phase Occurs Before Complex Amplitude, cyclic composition and its representation.

Sources 3–7 are the author-provided versions in r2d-addenda-current-for-audit.zip.

The propositions and compatibility witnesses in §§5–8 are the mathematical development of this addendum. This addendum does not amend the cited canon or earlier addenda.