The R²D Second Law
Irreversibility is nonreciprocal realization around closure
A completed local closure can return exactly to its starting state.
Its native multiplicity can also close exactly:
\[ \Delta S_{n,\alpha,\mathcal C}=0. \]
But this does not require the two opposed directions around the closure to be realized with equal occurrence-count efficiency.
R2D identifies this distinction as the primitive origin of irreversibility.
Structural closure can be exact while occurrence realization is nonreciprocal.
The R2D Second Law then makes an additional cross-boundary statement:
Local occurrence-count nonreciprocity is read at the enclosing boundary as an increment in nondirectional background.
Formally,
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C} = \delta_{\mathcal C}\Phi_{n+1} \geq0. } \]
This is the Second Law of Recursive Counting.
Closure Does Not Require Reversibility
Consider a completed local closure
\[ \mathcal C_{n,\alpha}: i \rightarrow k \rightarrow i. \]
The state has returned.
Therefore its native multiplicity closes:
\[ \boxed{ \Delta S_{n,\alpha,\mathcal C}=0. } \]
But there is another question:
Were the two opposed legs realized equally often?
The answer need not be yes.
This is why R2D separates closure from reversibility.
Directional Occurrence-Count Efficiency
Let
\[ \nu_{n,\alpha,H} \]
be the realized occurrence count of one oriented leg of the closure.
Let
\[ \nu_{n,\alpha,L} \]
be the occurrence count of the opposed return leg.
Define the directional occurrence-count efficiency
\[ \boxed{ \eta_{n,\alpha,\mathcal C} = \frac{ \nu_{n,\alpha,L} }{ \nu_{n,\alpha,H} }. } \]
Choose the orientation such that
\[ 0< \eta_{n,\alpha,\mathcal C} \leq1. \]
If the two opposed directions are realized equally,
\[ \nu_{n,\alpha,L} = \nu_{n,\alpha,H}, \]
then
\[ \boxed{ \eta_{n,\alpha,\mathcal C}=1. } \]
The closure is occurrence-count reciprocal.
Nonreciprocal Realization
If
\[ \nu_{n,\alpha,L} < \nu_{n,\alpha,H}, \]
then
\[ \boxed{ 0< \eta_{n,\alpha,\mathcal C} <1. } \]
The structural cycle still closes.
But its two opposed directions are not realized with equal count efficiency.
This is the primitive R2D definition of irreversible realization.
Thus:
\[ \boxed{ \text{irreversibility} \neq \text{failure of state closure}. } \]
Instead,
\[ \boxed{ \text{irreversibility} = \text{nonreciprocal occurrence realization around closure}. } \]
Occurrence-Count Loss
R2D represents this directional inefficiency logarithmically.
Define
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C} = \ln \left( \frac{ \nu_{n,\alpha,H} }{ \nu_{n,\alpha,L} } \right). } \]
Because
\[ \eta_{n,\alpha,\mathcal C} = \frac{ \nu_{n,\alpha,L} }{ \nu_{n,\alpha,H} }, \]
we can also write
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C} = -\ln \eta_{n,\alpha,\mathcal C}. } \]
For reciprocal realization,
\[ \eta=1, \]
so
\[ \boxed{ L_\tau=0. } \]
For nonreciprocal realization,
\[ 0<\eta<1, \]
so
\[ \boxed{ L_\tau>0. } \]
Therefore,
\[ \boxed{ L_\tau\geq0. } \]
“Loss” Does Not Mean Destruction
The word loss requires care.
R2D does not propose that occurrence count is annihilated.
Nor does it propose that multiplicity is destroyed.
Nor does it propose that direction disappears.
The quantity
\[ L_\tau \]
measures a logarithmic deficit in reciprocal directional realization.
It says that one opposed leg is realized less efficiently than the other.
Thus:
Occurrence-count loss is nonreciprocity, not count destruction.
Exact Multiplicity Closure Can Coexist With Irreversibility
This gives one of the cleanest relations in Part I:
\[ \boxed{ \Delta S_{n,\alpha,\mathcal C}=0 } \]
while simultaneously
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C}>0. } \]
The first statement says:
native possible multiplicity closes exactly.
The second says:
realized occurrence count is directionally nonreciprocal.
These are different count relations.
So an irreversible closure does not require a failure of the underlying multiplicity structure to close.
Possible Count and Realized Count Behave Differently
This distinction returns us to the separation between multiplicity and occurrence.
Multiplicity is possible count:
\[ W. \]
Occurrence is realized count:
\[ \nu. \]
The closure condition
\[ \Delta S_{\mathcal C}=0 \]
belongs to possible count.
The irreversible condition
\[ L_\tau>0 \]
belongs to realized occurrence count.
Therefore:
\[ \boxed{ \text{possible-count closure} \neq \text{occurrence-count reciprocity}. } \]
Background Is a Different Kind of Count
R2D now introduces background.
Let a local peer boundary () have a positive background count
\[ b_{n,\alpha}>0. \]
Define its logarithmic coordinate
\[ \boxed{ \Phi_{n,\alpha} = \ln b_{n,\alpha}. } \]
Background is not another local directional state.
Within peer boundary \(\alpha\), it is common to the local macrostates.
Thus, for a local transition
\[ i\rightarrow k, \]
\[ \boxed{ \Delta\Phi_{n,\alpha,i\rightarrow k}=0. } \]
Background therefore does not independently favor one local macrostate over another.
Background Is Nondirectional Locally
This gives background its defining local property.
At the boundary where it is background,
\[ \boxed{ \Delta\Phi=0. } \]
It does not supply a local directional tilt.
So background is fundamentally different from
\[ \Delta P, \]
\[ \Delta R, \]
or
\[ \Delta A. \]
Those quantities distinguish directions within the local state domain.
Background does not.
Background is nondirectional relative to the boundary at which it is background.
But Background Is Boundary Relative
Now consider two peer boundaries,
\[ \alpha \qquad\text{and}\qquad \beta. \]
Their local background coordinates might be
\[ \Phi_{n,\alpha} \]
and
\[ \Phi_{n,\beta}. \]
Each is nondirectional within its own peer boundary.
But the enclosing boundary can compare them.
Define
\[ \boxed{ \Delta\Phi_{n+1|n,\alpha:\beta} = \Phi_{n,\alpha} - \Phi_{n,\beta}. } \]
Equivalently,
\[ \boxed{ \Delta\Phi_{n+1|n,\alpha:\beta} = \ln \left( \frac{ b_{n,\alpha} }{ b_{n,\beta} } \right). } \]
If this quantity is nonzero, then the enclosing boundary can distinguish a relation between the two subordinate backgrounds.
So:
What is nondirectional background at one boundary can become part of a distinguishable relation at an enclosing boundary.
Background is therefore boundary relative, just like state identity itself.
The Enclosing Background Increment
Now consider one completed representative local closure.
Let
\[ b_{n+1}^{-} \]
denote the enclosing background count before the associated realization,
and
\[ b_{n+1}^{+} \]
the enclosing background count afterward.
Define the logarithmic background increment
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} = \ln \left( \frac{ b_{n+1}^{+} }{ b_{n+1}^{-} } \right). } \]
This is an enclosing-boundary count relation.
It is not the same object as the local occurrence-count loss.
The Second Law relates the two.
The Second-Law Postulate
R2D postulates that the local logarithmic occurrence-count loss and the enclosing logarithmic background increment are two boundary-specific readings of one recursively organized realization.
Thus,
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} = L_{\tau,n,\alpha,\mathcal C}. } \]
Substituting the occurrence counts,
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} = \ln \left( \frac{ \nu_{n,\alpha,H} }{ \nu_{n,\alpha,L} } \right). } \]
Therefore,
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} \geq0. } \]
Reciprocal Realization Produces No Background Increment
If
\[ \eta=1, \]
then
\[ L_\tau=0. \]
The Second Law gives
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1}=0. } \]
Thus:
\[ \boxed{ \eta=1 \iff \delta_{\mathcal C}\Phi_{n+1}=0. } \]
Reciprocal local realization corresponds to no Second-Law background increment.
Nonreciprocal Realization Produces a Positive Background Increment
If
\[ 0<\eta<1, \]
then
\[ L_\tau>0. \]
Therefore,
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1}>0. } \]
So:
\[ \boxed{ 0<\eta<1 \iff \delta_{\mathcal C}\Phi_{n+1}>0. } \]
The primitive irreversible relation is therefore nonnegative.
This Is Not Bottom-Up Transfer
The equation
\[ \delta_{\mathcal C}\Phi_{n+1} = L_{\tau,n,\alpha,\mathcal C} \]
must not be read as
\[ \text{lost local count} \longrightarrow \text{background substance}. \]
No destroyed count is physically transported upward.
The local quantity and enclosing quantity are defined at different boundaries.
They are two boundary-specific readings.
Thus:
\[ \boxed{ \text{local nonreciprocity} \longleftrightarrow \text{enclosing background increment}. } \]
The equality is between logarithmic count coordinates.
It is not arithmetic equality between raw counts.
The Causal Direction Does Not Reverse
The Second Law also does not reverse the causal inversion developed earlier.
R2D does not say that local occurrence-count loss causes the enclosing background.
The enclosing-to-local causal architecture remains
\[ \boxed{ \chi_{n+1}^{(r)} \Longrightarrow \mathcal C_{n,\alpha}. } \]
The Second Law says how one realized relation is read differently at the two boundaries.
As with recursive carry, cross-boundary correspondence must not be confused with upward causation.
Productive Output Does Not Define Irreversibility
The enclosing constraint can be decomposed as
\[ C_{n+1} = C_{n+1}^{\mathrm{out}} + C_{n+1}^{\Phi}. \]
The productive component is associated with retained distinguishable output:
\[ C_{n+1}^{\mathrm{out}} \lambda_{n+1}. \]
The background-associated component is associated with realization read as background:
\[ C_{n+1}^{\Phi} \lambda_{n+1}. \]
But productive output does not determine whether a closure is irreversible.
The irreversibility criterion remains
\[ \boxed{ L_\tau>0. } \]
For example,
\[ C_{n+1}^{\mathrm{out}}=0 \]
does not imply
\[ L_\tau=0. \]
A system can retain no productive distinguishable output and still realize its opposed closure legs nonreciprocally.
Zero Productive Output Is Not Reversibility
This distinction is important physically.
A process can produce no useful output and still be irreversible.
Likewise, productive output can be present without itself defining the degree of irreversibility.
Thus:
\[ \boxed{ \text{productive output} \neq \text{irreversibility}. } \]
Irreversibility is determined by the reciprocity of occurrence realization around closure.
The Background-Associated Constraint Is Not the Background Increment
Two quantities must remain distinct:
\[ C_{n+1}^{\Phi}\lambda_{n+1} \]
and
\[ \delta_{\mathcal C}\Phi_{n+1}. \]
The first belongs to the decomposition of the total constrained enclosing realization.
The second belongs to the Second-Law correspondence with local occurrence nonreciprocity.
Part I does not postulate
\[ C_{n+1}^{\Phi}\lambda_{n+1} = \delta_{\mathcal C}\Phi_{n+1}. \]
A later physical mapping may relate their physical reads in a particular system.
That is a separate empirical question.
The R2D Law and Second Law Do Different Jobs
The R2D law is
\[ \boxed{ \Delta A_{n,\alpha,\mathcal C} = C_{n+1} \lambda_{n+1} \rho_{n+1|n,\alpha}. } \]
It relates:
local recurrence-weighted asymmetry
to
enclosing recurrence-weighted constrained distinguishability.
The Second Law is
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C} = \delta_{\mathcal C}\Phi_{n+1}. } \]
It relates:
local occurrence-count nonreciprocity
to
enclosing nondirectional background increment.
Thus:
\[ \boxed{ \text{R²D law} \neq \text{R²D Second Law}. } \]
They are complementary postulates.
Asymmetry Is Not Irreversibility
Another distinction follows.
Closure asymmetry is
\[ \Delta A_{n,\alpha,\mathcal C} = \Delta P_{n,\alpha,\mathcal C} - \Delta R_{n,\alpha,\mathcal C}. \]
Occurrence nonreciprocity is measured by
\[ L_{\tau,n,\alpha,\mathcal C}. \]
These are not the same relation.
A closure can be asymmetrically organized while its opposed occurrence counts remain reciprocal.
Thus:
\[ \boxed{ \Delta A_{\mathcal C}\neq0 \not\Rightarrow L_\tau>0. } \]
R2D therefore separates:
directionality
from
irreversibility.
Direction tells us which realization is favored.
Irreversibility tells us whether opposed realization occurs with equal directional efficiency.
Closure, Direction, and Irreversibility Are Three Different Questions
For one completed cycle, ask:
Did the state return?
For a closure, yes.
Is the enclosing relation directional?
Possibly:
\[ \Delta A_{\mathcal C}\neq0. \]
Are the opposed realization counts reciprocal?
Possibly not:
\[ L_\tau>0. \]
These questions are independent enough that they require separate mathematical quantities.
This is one of the reasons R2D does not identify irreversibility simply with “the system moved downhill.”
Nothing Is Destroyed
The R2D Second Law can therefore be stated without invoking destruction.
Native multiplicity closes:
\[ \Delta S_{\mathcal C}=0. \]
The local state closes.
Occurrence continues.
Direction can remain.
What fails to be reciprocal is the efficiency with which opposed directions realize occurrence count.
The logarithmic difference is
\[ L_\tau. \]
At the enclosing boundary, that same recursive relation is read as
\[ \delta_{\mathcal C}\Phi. \]
Thus:
Nothing must disappear for irreversibility to exist.
The primitive irreversible relation is a change in how count is realized and subsequently classified across boundaries.
Why This Resembles the Thermodynamic Second Law
The structure is deliberately suggestive.
Thermodynamic irreversibility is associated with nonnegative entropy production.
R2D has now produced a primitive nonnegative count relation:
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} = L_{\tau,n,\alpha,\mathcal C} \geq0. } \]
But Part I has not yet identified
\[ \Phi \]
with thermodynamic entropy.
It has not yet introduced temperature or heat.
The thermodynamic Second Law is therefore not assumed here.
Instead, Part II asks whether thermodynamic entropy production is a physical read of this more primitive recursive-count relation.
The Empirical Burden
The Second Law of Recursive Counting is a postulate.
The definitions of
\[ \eta \]
and
\[ L_\tau \]
follow from occurrence counts.
The statement
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} = L_{\tau,n,\alpha,\mathcal C} } \]
is an additional R2D claim.
A physical realization must therefore establish independently:
- the relevant closure;
- its opposed occurrence counts;
- the resulting \(L_\tau\);
- the appropriate enclosing boundary;
- and the corresponding background increment.
Only then can the proposed correspondence be tested.
The later thermodynamic program carries that empirical burden.
The Primitive Second-Law Architecture
The complete structure can be written:
\[ \boxed{ \begin{array}{c} \text{local closure} \\[4pt] \Downarrow \\[4pt] \Delta S_{\mathcal C}=0 \\[8pt] \text{opposed occurrence realization} \\[4pt] \Downarrow \\[4pt] \eta_{\mathcal C} = \dfrac{\nu_L}{\nu_H} \\[8pt] \Downarrow \\[4pt] L_\tau = -\ln\eta_{\mathcal C} \geq0 \\[8pt] \Longleftrightarrow \\[8pt] \delta_{\mathcal C}\Phi_{n+1} \geq0 \\[4pt] \text{enclosing background increment} \end{array} } \]
The native structure closes.
The realized directional efficiencies need not.
The enclosing boundary reads the difference as background.
Core Equations
Native multiplicity closure
\[ \boxed{ \Delta S_{n,\alpha,\mathcal C}=0 } \]
Directional occurrence-count efficiency
\[ \boxed{ \eta_{n,\alpha,\mathcal C} = \frac{ \nu_{n,\alpha,L} }{ \nu_{n,\alpha,H} } } \]
Reciprocal realization
\[ \boxed{ \eta_{n,\alpha,\mathcal C}=1 \iff L_{\tau,n,\alpha,\mathcal C}=0 } \]
Nonreciprocal realization
\[ \boxed{ 0< \eta_{n,\alpha,\mathcal C} <1 \iff L_{\tau,n,\alpha,\mathcal C}>0 } \]
Occurrence-count loss
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C} = -\ln \eta_{n,\alpha,\mathcal C} = \ln \left( \frac{ \nu_{n,\alpha,H} }{ \nu_{n,\alpha,L} } \right) } \]
Background increment
\[ \boxed{ \delta_{\mathcal C}\Phi_{n+1} = \ln \left( \frac{ b_{n+1}^{+} }{ b_{n+1}^{-} } \right) } \]
The Second Law of Recursive Counting
\[ \boxed{ L_{\tau,n,\alpha,\mathcal C} = \delta_{\mathcal C}\Phi_{n+1} \geq0 } \]
The Key Distinction
The R2D Second Law does not say that entropy-like substance is created because something has been destroyed.
It says:
A local structure can close exactly while its opposed realization counts are nonreciprocal.
That logarithmic nonreciprocity is read from the enclosing boundary as nondirectional background.
So the primitive second-law architecture is
\[ \boxed{ \text{exact structural closure} + \text{nonreciprocal realization} \longleftrightarrow \text{enclosing background increment}. } \]
This separates structural closure, directionality, irreversibility, and background before thermodynamic entropy has been introduced.
Next
Why R²D Is Not Coarse-Graining →
Why the Mathematics Comes Before Units →