John Swygert

Ivory Tower Publishing

September 12, 2026

Research Note

This paper represents the first direct experimental application of the relational-synthesis framework developed in the preceding TSTOEAO work. Rather than asking relational reasoning to create missing physics, it tests whether independently established thermodynamic constraints can be combined without overcounting to expose a sharper joint boundary—and, just as importantly, to show exactly where the synthesis must stop.

Abstract

The preceding TSTOEAO research sequence progressively narrowed the framework from broad gravitational hypotheses toward increasingly strict tests of relational scientific reasoning.

Cross-domain transport demonstrated that known mathematical structures can survive translation between physically different systems, but did not generate new physical laws. Dynamic relational invariants identified forbidden transitions, but their realization remained dependent upon domain-specific dynamics. Intra-domain constraint generation reproduced correct physical bounds, but those bounds ultimately remained encoded in the assumptions from which they were derived.

The sixth paper therefore shifted the research objective from attempting to create information through abstraction to identifying latent consequences contained across multiple independently established relationships.

The present paper performs the first end-to-end relational-synthesis experiment under that revised standard.

The selected domain is finite-time information erasure in nonequilibrium statistical mechanics.

Three independently developed constraint families are brought into a common relational representation: (1) the Landauer free-energy cost of erasing information; (2) thermodynamic uncertainty relations connecting precision to entropy production; and (3) finite-time thermodynamic speed limits constraining entropy production when a transformation must be completed within a specified duration.

Do these independently established constraints jointly force a stronger bound than any one provides separately?

For a bit erased with final error probability ε, the equilibrium informational contribution gives

ΔF_erase = k_B T [ln 2 – H_b(ε)]

where

H_b(ε) = -ε ln ε – (1-ε) ln(1-ε)

For an isothermal process, entropy production Σ contributes additional irreversible work,

W ≥ ΔF_erase + TΣ

If independently valid precision and finite-time bounds require Σ/k_B ≥ B_precision and Σ/k_B ≥ B_speed, their relational intersection implies

W ≥ k_B T [ln 2 – H_b(ε) + max(B_precision, B_speed)]

The maximum rather than the sum is required because both inequalities constrain the same entropy-production quantity and therefore cannot automatically be counted as independent additive costs. This composite inequality is the locked relational prediction of the present experiment.

The synthesis is mathematically valid under the joint domains of applicability of the component inequalities. It also identifies an important boundary: no universal relationship between bit-error probability and a thermodynamic current’s uncertainty can be inferred without an additional device-specific mapping. The result is therefore a demonstration of constraint synthesis; historical novelty remains a separate question requiring literature verification.

01 Research Continuation

This paper continues the sequence: Relational Compression in Gravitational Domains; From Relational Compression to Relational Form Invariance; From Relational Form Invariance to Algebraic Relational Invariance; From Algebraic Relational Invariance to Predictive Cross-Domain Transport; From Cross-Domain Transport to Intra-Domain Constraint Generation; and From Constraint Generation to Relational Synthesis.

The present paper performs the concrete synthesis experiment required by Paper 6.

02 The New Standard

Paper 6 proposed that TSTOEAO should no longer attempt to generate missing physical information from abstraction alone. Instead, it should ask whether independently established scientific relationships contain joint consequences that are not explicit when those relationships are considered separately.

known relations → common representation → constraint intersection → candidate consequence → independent verification

03 Why Information Erasure Is a Suitable Test

Information erasure is governed simultaneously by several well-developed bodies of theory. It possesses an informational boundary, an energetic cost, irreversible entropy production, finite-time restrictions, stochastic fluctuations, measurable error, and experimentally realizable devices. It therefore provides multiple valid relationships that constrain the same physical process from different directions.

04 The Target Process

Consider a physical memory containing one classical bit. Initially:

p₀ = (1/2, 1/2)

The erasure operation attempts to reset the bit to logical state 0. At completion:

p_τ = (1-ε, ε),     0 ≤ ε ≤ 1/2

Perfect erasure corresponds to ε = 0. No erasure corresponds to ε = 1/2.

05 Binary Entropy

Define the binary entropy in natural logarithmic units:

H_b(ε) = -ε ln ε – (1-ε) ln(1-ε)

H_i = ln 2,     H_f = H_b(ε)

ΔH_removed = ln 2 – H_b(ε)

06 Landauer’s Informational Floor

For an isothermal reset under the appropriate thermodynamic assumptions, the minimal reversible energetic contribution associated with reducing the logical entropy is [1,5]

ΔF_erase = k_B T [ln 2 – H_b(ε)]

For perfect erasure, ε → 0 and H_b(ε) → 0, so ΔF_erase → k_B T ln 2.

07 Imperfect Erasure

Allowing finite error lowers the ideal informational floor. As ε → 1/2, H_b(ε) → ln 2 and therefore ΔF_erase → 0.

greater logical certainty → greater minimal thermodynamic cost

08 The Reversible Limit Is Not the Finite-Time Limit

Landauer’s limit describes a reversible floor. Real operations are generally performed in finite time. Finite-time operation produces additional entropy. Let Σ ≥ 0 denote total entropy production. [1,5]

W ≥ ΔF_erase + TΣ

09 First Relational Coordinate

Define normalized work w = W/(k_B T) and normalized entropy production σ = Σ/k_B. Then

w ≥ ln 2 – H_b(ε) + σ

This becomes the common energetic coordinate into which the remaining constraints will be mapped.

10 Constraint Family One

C_L:  w – σ ≥ ln 2 – H_b(ε)

This represents the information-energy boundary.

11 The Precision Problem

A physical erasure process is stochastic. Repeated nominally identical operations need not produce identical microscopic trajectories. Many devices therefore involve a measurable current J such as particle transfer, charge transfer, molecular transitions, heat flow, or probability current. Its mean is ⟨J⟩ and its variance is Var(J).

12 Relative Current Uncertainty

R_J = Var(J) / ⟨J⟩²

Small R_J corresponds to a relatively precise current; large R_J corresponds to a noisy current.

13 Thermodynamic Uncertainty Relations

For classes of nonequilibrium stochastic systems satisfying the assumptions of the relevant thermodynamic uncertainty relation, precision cannot be made arbitrarily large without entropy production. [2,3]

σ ≥ B_precision

For the familiar steady-state current form this is often represented schematically as

σ ≳ 2/R_J = 2⟨J⟩²/Var(J)

The precise coefficient and form depend upon the specific TUR and its assumptions.

14 Important Scope Restriction

There is no single universal TUR applicable without qualification to every arbitrary finite-time erasure protocol. Therefore this paper does not assume σ ≥ 2/R_J universally. Instead, B_precision denotes the lower bound supplied by an appropriate independently validated uncertainty relation for the selected stochastic model.

15 Constraint Family Two

C_P:  σ ≥ B_precision

high stochastic precision → thermodynamic cost

16 The Speed Problem

Erasure must also occur within a duration τ. A reversible operation generally requires increasingly slow control. Finite-time thermodynamics therefore introduces an additional restriction.

17 Thermodynamic Speed Limits

For suitable classes of stochastic dynamics, the initial and final probability distributions cannot be transformed into one another arbitrarily quickly at arbitrarily small entropy production. [4,6]

σ ≥ B_speed(τ, p₀, p_τ, M)

Here M represents the kinetic or geometric information required by the particular speed-limit theorem.

18 Typical Finite-Time Structure

B_speed ~ D²/(K τ)

Here D is an appropriate statistical or thermodynamic distance and K contains kinetic information such as mobility, activity, or diffusivity. The precise expression is model dependent.

19 Why the Kinetic Factor Cannot Be Removed

The same pair of probability distributions may be connected by physically different devices with different mobilities, barrier heights, transition rates, friction coefficients, diffusion constants, and control protocols. Therefore a universal finite-time energetic cost cannot generally be obtained from p₀, p_τ, and τ alone.

20 Constraint Family Three

C_S:  σ ≥ B_speed

faster transformations → additional irreversible requirements

21 The Three Constraint Surfaces

C_L: w ≥ ln 2 – H_b(ε) + σ

C_P: σ ≥ B_precision

C_S: σ ≥ B_speed

Each describes a different boundary of the same erasure process.

22 The Relational-Synthesis Question

What follows when all three constraints must be true simultaneously?

23 The Admissible Region

Ω* = Ω_L ∩ Ω_P ∩ Ω_S

The process must satisfy every constraint simultaneously.

24 Intersecting the Entropy-Production Bounds

σ ≥ max(B_precision, B_speed)

25 Why the Bounds Are Not Automatically Added

It would be tempting to write σ ≥ B_precision + B_speed. That is not generally justified. Both expressions are lower bounds on the same quantity. Unless their dissipative mechanisms are independently additive, summing them may double-count the same entropy production.

σ ≥ max(B_precision, B_speed)

26 The Locked Relational Prediction

P_TSTOEAO:  W ≥ k_B T [ln 2 – H_b(ε) + max(B_precision, B_speed)]

This is the locked relational prediction.

27 Interpretation

The bound contains two conceptually distinct layers: the informational floor k_B T[ln 2 – H_b(ε)] and the unavoidable irreversible penalty k_B T max(B_precision, B_speed).

28 The Composite Trade-Off

logical certainty + irreversible operational constraint → minimum work

error → informational cost

precision → entropy-production cost

speed → entropy-production cost

29 Competing Operational Bottlenecks

If B_precision > B_speed, the process is precision-limited. If B_speed > B_precision, it is speed-limited.

30 The Relational Seam

B_precision = B_speed

This defines a seam in parameter space where neither precision nor speed dominates the irreversible cost.

31 Why the Seam Is Scientifically Interesting

The crossover may identify an operating regime where further improvement in speed begins to dominate energetic cost more strongly than further improvement in precision, or vice versa.

B_PS: B_precision = B_speed

32 A Device-Design Interpretation

Suppose an engineer specifies target logical error ε, required processing time τ, required current precision R_J, and temperature T. The device cannot simultaneously achieve these targets with arbitrarily small work.

W_min ≥ k_B T [ln 2 – H_b(ε) + max(B_precision, B_speed)]

33 Where TSTOEAO Must Stop

One might attempt to identify ε directly with R_J and thereby produce a universal relation between logical error and current fluctuations. No such identification has yet been justified.

34 Logical Error Is Not Automatically Current Uncertainty

ε describes the probability of a logically incorrect final state. R_J describes fluctuations in a specified physical current. They are different observables.

ε ≠ R_J  in general

35 The Missing Mapping

To obtain a direct four-variable relation among W, τ, ε, and R_J, one must specify a device-dependent relation:

R_J = F(ε; D)   or   ε = G(R_J; D)

Here D contains the device architecture and stochastic dynamics.

36 This Is a Boundary, Not a Failure

The absence of a universal mapping tells us exactly where the relational synthesis stops. The constraints genuinely meet through σ, but logical error and current precision do not automatically meet directly.

shared process ≠ identical observable

37 Relational Hypergraph

The system can be represented as a hypergraph with vertices {W, T, ε, σ, τ, J, Var(J), M}. Hyperedges include R_L(W,T,ε,σ), R_P(σ,J,Var(J)), and R_S(σ,τ,p₀,p_τ,M).

38 The Central Shared Variable

The intersection occurs because both operational constraints touch σ.

B_precision → σ ← B_speed

σ → W

39 Relational Closure

ε → ΔF_erase → W

precision → σ → W

speed → σ → W

The combined implication produces the composite work bound.

40 Independent Conventional Verification

Start with W ≥ ΔF + TΣ and ΔF = k_B T[ln 2 – H_b(ε)]. If the selected system satisfies Σ ≥ k_B B_precision and Σ ≥ k_B B_speed, then σ must exceed both bounds and therefore exceed their maximum.

P_conventional = P_TSTOEAO

41 Classification of the Mathematical Result

The relational synthesis is mathematically valid. Historical novelty is a separate question requiring dedicated literature analysis.

classification: successful relational synthesis; novelty undetermined

42 What Has Actually Been Generated

The experiment did not create new thermodynamic information. It took independently specified boundaries and produced their joint admissible region.

strongest jointly guaranteed irreversible penalty = max(B_precision, B_speed), not their sum

43 Why This Matters

An incautious synthesis could overstate the minimum cost by summing lower bounds on the same entropy-production quantity. Relational accounting prevents that double counting.

44 Information Provenance

The term ln 2 – H_b(ε) comes from logical entropy reduction. B_precision comes from the uncertainty relation. B_speed comes from the finite-time restriction. The max operation comes from simultaneous lower bounds on the same entropy production.

45 The Assumption Graph

A_L → C_L,     A_P → C_P,     A_S → C_S

valid synthesis requires A_L ∩ A_P ∩ A_S ≠ ∅

46 Domain Compatibility Is Mandatory

A TUR derived for a nonequilibrium steady state cannot automatically be combined with an arbitrary finite-time driven erasure protocol. Likewise, a speed limit derived for overdamped diffusion cannot automatically be applied to a discrete electronic memory. The actual device must instantiate all selected theorem assumptions. [2,3]

47 The Proper Experimental Version

A genuine physical test should select one explicit device model, such as a two-state Markov memory, a colloidal particle in a controlled double-well potential, a single-electron memory, or a molecular switch. Every bound must be valid for the same dynamical model class.

48 Why the Present Paper Does Not Fake That Step

No specific device is imposed here. Therefore no universal numerical coefficient for B_speed or B_precision is claimed. The present result is the structural synthesis that must hold whenever compatible instances of the three component constraints are simultaneously valid.

49 Failure Boundaries

No common validity domain means no legitimate synthesis. No justified mapping from current precision to logical error means no universal error-precision equation. Different derivations do not imply independent physical costs.

50 When Addition Would Become Valid

Addition would require a physical decomposition Σ = Σ₁ + Σ₂ with independently justified contributions satisfying Σ₁ ≥ B₁ and Σ₂ ≥ B₂. Only then does Σ ≥ B₁ + B₂ follow.

51 Shared-Variable Constraint Rule

Multiple lower bounds on the same quantity intersect by the strongest bound unless independent additive channels are demonstrated.

This is ordinary mathematical logic, not a new physical law.

52 Additive-Channel Rule

Bounds may be summed only when the physical quantity is independently decomposed into corresponding additive contributions.

53 Error as an Informational Coordinate

The error probability changes the reversible information-removal floor ΔF_erase(ε). Speed and precision instead constrain the irreversible contribution TΣ.

error modifies the floor; speed and precision modify the excess cost

54 A Four-Way Trade Space

W_min = W_min(ε, τ, R_J; T, M, D)

The geometry contains device-dependent parameters and is therefore not universally reducible to four observables.

55 The Pareto Surface

Instead of a single optimum, the system may possess a Pareto boundary in the admissible space of work, error, speed, and precision.

P = ∂Ω_admissible(W, ε, τ, R_J)

56 A TSTOEAO Interpretation

In TSTOEAO terminology, the required reset provides a gradient; finite error tolerance, finite-time operation, and stochastic precision create boundaries; entropy production represents physical cost; and the feasible device region represents Encoded Equilibrium under those constraints.

Gradient → Boundary → Cost → Admissible Equilibrium

57 The Framework Does Not Replace Thermodynamics

Every numerical physical constraint still comes from thermodynamics and stochastic dynamics. TSTOEAO contributes, if anything, through organization of their relations.

58 Did Relational Synthesis Work?

Yes, in the methodological sense. It correctly produced the composite maximum-bound structure and prevented unjustified summation. It has not yet established new physics.

59 The More Interesting Result May Be the Missing Edge

There is no universal edge ε ↔ R_J. That missing relationship prevents reduction of the full problem to a universal four-variable bound and becomes a precise target for domain-specific investigation.

60 Missing-Edge Principle

When relational closure requires a connection not supplied by the premises or established physics, the connection must remain missing until physics supplies it.

61 The Next Device-Level Question

How does final logical error ε depend upon measurable stochastic current statistics for a specified physical memory?

62 A Possible Device-Specific Mapping

ε = G(⟨J⟩, Var(J), τ, {k_ij})

Only after deriving such a mapping can a TUR be converted into an explicit error-dependent bound.

63 Why Such a Result Could Be Stronger

W ≥ k_B T [ln 2 – H_b(ε) + max(B_precision(ε), B_speed(ε,τ))]

This would produce a genuine energy-speed-error trade-off for that model.

64 The Crossover Prediction

B_precision(ε) = B_speed(ε, τ)

Solving this may define a crossover time τ*(ε) or error threshold ε*(τ).

65 Why This Is the Stronger Next Experiment

A fully specified model permits exact equations, numerical coefficients, simulation, experimental comparison, literature comparison, and the possibility of identifying an overlooked operating boundary.

66 The Candidate Latent Consequence

The strongest candidate produced by the present relational map is the possibility that a device possesses a precision-speed crossover seam that partitions its operating space into distinct dominant-cost regimes. This remains conditional: if one bound dominates everywhere, no crossover exists.

67 Possible Computational Implementation

For a specific two-state device, define Θ = (ε, τ, T, k₀₁, k₁₀, …), calculate compatible Landauer, precision, and speed bounds, and construct the joint surface. Measured work can then be compared with the bound through a residual E_W = W_obs – W_bound.

68 A Stronger Experimental Protocol

Choose a specific physical two-state memory; specify its stochastic dynamics; select compatible Landauer, TUR, and finite-time speed-limit relations; determine their common validity region; construct the composite bound; lock crossover or residual predictions; solve or simulate the master equation independently; then compare with published theory and experimental data.

69 Falsification Conditions

The synthesis fails if the component assumptions have no common physical domain, if a valid full dynamical simulation violates the proposed device-specific bound, if a claimed crossover never occurs, if equivalent results already exist when historical novelty is claimed, or if ordinary analysis exposes every candidate relation just as directly with no methodological gain.

70 What Has Been Learned

Constraint intersection can strengthen knowledge of an admissible region. Shared-variable bounds combine differently from independent additive costs. Missing relational edges can be as informative as existing ones. Domain compatibility must be established before synthesis.

71 Current Status of TSTOEAO

relational mapping + constraint synthesis + boundary detection

TSTOEAO has not yet demonstrated a new fundamental physical law.

72 The Next Locked Target

For a specified two-state stochastic memory, do compatible Landauer, uncertainty, and finite-time constraints produce a previously unnoticed crossover boundary or tighter operating bound that survives exact master-equation analysis?

73 Why the Negative Results Still Matter

Each failed stronger claim removed ambiguity. The program now knows that abstraction cannot create absent dynamics, analogy cannot establish mechanism, common mathematical form does not imply common physics, correct prediction may merely rediscover known information, valid relations cannot be combined outside their common validity domain, and different lower bounds cannot automatically be added.

74 The Central Result

W ≥ k_B T [ln 2 – H_b(ε) + max(B_precision, B_speed)]

Its scope is conditional upon simultaneous applicability of the underlying thermodynamic results.

75 The More Important Boundary Result

No universal direct relation between logical error and current uncertainty follows from these three constraints alone.

Additional physical structure is required.

76 Refined Relational Synthesis Principle

When independently valid constraints share physical variables and possess a nonempty common validity domain, their intersection defines a joint admissible region whose boundaries may reveal consequences not explicit in the constraints considered separately.

77 Relational Non-Overcounting Principle

Constraints derived independently must not be treated as additive unless the constrained physical quantity is demonstrably decomposable into independent additive contributions.

78 Missing-Edge Principle

When relational closure requires a connection not supplied by the premises or established physics, the connection must not be inferred merely from participation in the same system.

CONCLUSION

The sixth TSTOEAO paper proposed a shift from information generation to relational synthesis. The present paper performs the first concrete experiment under that standard.

The selected system was finite-time information erasure because several independently developed physical theories constrain the same process. Landauer’s principle constrains the reversible informational cost. Thermodynamic uncertainty relations constrain the entropy production required for precision. Finite-time thermodynamic speed limits constrain the entropy production required for rapid transformation.

The operational constraints require σ ≥ B_precision and σ ≥ B_speed. Therefore σ ≥ max(B_precision, B_speed), and consequently:

W ≥ k_B T [ln 2 – H_b(ε) + max(B_precision, B_speed)]

The synthesis survives independent conventional reasoning, but it also exposes its own boundary. Logical error probability and current uncertainty are not universally interchangeable observables. Without a device-specific mapping, ε ↔ R_J cannot be assumed. Likewise, the speed and precision penalties cannot automatically be added because both may constrain the same entropy production.

The strongest result is therefore not a claim that TSTOEAO has discovered a new thermodynamic law. It is that the relational-synthesis protocol can identify the shared variable through which independent constraints interact, construct the resulting joint admissible region, prevent invalid double counting, identify a candidate crossover boundary, and reveal the precise missing relationship preventing further closure.

The next scientific threshold is sharply defined: choose one explicit stochastic memory, derive its error-current relation, use compatible Landauer, uncertainty, and speed-limit results, lock the predicted crossover, and then solve the full dynamics.

If the crossover is already known, the exercise is a rediscovery. If it is false, the relational candidate fails. If it is correct and previously unnoticed, TSTOEAO will have produced its first serious example of latent scientific consequence discovery.

Not information from nowhere. Not analogy mistaken for mechanism. Not a bound hidden in its own assumptions. But established pieces assembled carefully enough that their intersection tells us something we had not previously seen.

The information was already there.

The scientific question is whether the relationship was.

Now the synthesis itself must be tested.

References

1. Landauer, R. (1961). Irreversibility and Heat Generation in the Computing Process. IBM Journal of Research and Development, 5(3), 183-191. https://doi.org/10.1147/rd.53.0183

2. Barato, A. C., & Seifert, U. (2015). Thermodynamic Uncertainty Relation for Biomolecular Processes. Physical Review Letters, 114, 158101. https://doi.org/10.1103/PhysRevLett.114.158101

3. Gingrich, T. R., Horowitz, J. M., Perunov, N., & England, J. L. (2016). Dissipation Bounds All Steady-State Current Fluctuations. Physical Review Letters, 116, 120601. https://doi.org/10.1103/PhysRevLett.116.120601

4. Shiraishi, N., Funo, K., & Saito, K. (2018). Speed Limit for Classical Stochastic Processes. Physical Review Letters, 121, 070601. https://doi.org/10.1103/PhysRevLett.121.070601

5. Faist, P., Dupuis, F., Oppenheim, J., & Renner, R. (2015). The Minimal Work Cost of Information Processing. Nature Communications, 6, 7669. https://doi.org/10.1038/ncomms8669

6. Aurell, E., Gawędzki, K., Mejía-Monasterio, C., Mohayaee, R., & Muratore-Ginanneschi, P. (2012). Refined Second Law of Thermodynamics for Fast Random Processes. Journal of Statistical Physics, 147(3), 487-505. https://doi.org/10.1007/s10955-012-0478-x