Boundaries Define Scientific Objects
The count domain comes before the object
R2D begins with the principle:
Counting requires distinguishability.
But distinguishability alone is not enough.
A scientific count also requires a boundary that determines what counts as a distinction.
R2D therefore makes the boundary primitive to the scientific object.
A boundary defines the count domain.
The boundary does not merely surround an already-defined object.
It participates in defining what the object is.
A Boundary Is Not Necessarily a Physical Surface
In ordinary language, a boundary often means a wall, membrane, surface, or spatial edge.
That is not the primitive R2D definition.
A boundary is the relation that determines:
- which differences are distinguishable;
- which subordinate realizations are mutually compatible;
- which compatible realizations belong to the same readable state;
- and which completed subordinate processes count as realizations of an enclosing occurrence.
A membrane can define a boundary.
So can an ensemble definition.
So can an experimental measurement.
So can the distinction between one molecular state and another.
The primitive requirement is not a physical surface.
It is a count domain.
The Boundary Defines the Object
Consider a single coin.
At one boundary, the readable states are
\[ H \qquad \text{and} \qquad T. \]
The coin is the object.
Now consider ten coins together.
At the ensemble boundary, the readable states might instead be
\[ (10H,0T),\; (9H,1T),\; \dots,\; (0H,10T). \]
The ten-coin ensemble is now the object.
The individual heads and tails have not disappeared physically.
But they no longer possess the same state meaning at the ensemble boundary.
The boundary has changed.
Therefore the count domain has changed.
Therefore the scientific object has changed.
The Same Physical System Can Support Different Objects
This distinction becomes especially important in science.
Consider water.
At one boundary, the scientific object may be a water molecule.
Its readable states belong to the molecular state domain.
At another boundary, the scientific object may be a glass of water.
Its readable states belong to an enclosing thermodynamic or mechanical state domain.
The water is not a poorly resolved water molecule.
And the molecule is not highly resolved water.
They are different scientific objects because they are defined at different count boundaries.
R2D therefore rejects the assumption that increasing or decreasing resolution always reveals more or less detail about one ontologically fixed object.
Instead:
\[ \boxed{ \text{change boundary} \rightarrow \text{change count domain} \rightarrow \text{change scientific object} } \]
There Is No Privileged Physical Scale
Statistical mechanics itself provides an important clue.
The same combinatorial mathematics can describe an ensemble of:
- spins;
- electrons;
- molecules;
- proteins;
- molecular motors;
- cells;
- coins;
- or other distinguishable systems.
Nothing in the combinatorics declares one of these objects to be the uniquely fundamental scale.
A protein can be the object.
A molecular state within that protein can be the object.
An ensemble of proteins can be the object.
A muscle fiber can be the object.
Which one is being counted depends on the boundary.
Science is boundary dependent. It does not privilege one object or one scale.
Boundary-Relative Statehood
This leads directly to the most important definitions in R2D.
A state does not first exist independently of the boundary and then receive a label such as “microstate” or “macrostate.”
Those terms describe a relationship to a boundary.
At boundary \(B_n\):
- a macrostate is a distinguishable state readable at \(B_n\);
- a formal microstate is one compatible joint realization of the immediately subordinate peer boundaries.
The distinction is relational.
A state can therefore have macrostate meaning at one boundary and microstate meaning relative to another.
Immediately Subordinate Peer Boundaries
Suppose the boundary being considered is
\[ B_n. \]
Let its immediately subordinate peer boundaries be
\[ B_{n-1,\alpha}. \]
Each subordinate peer possesses its own state domain.
A realization of the enclosing boundary involves some joint realization of those peers.
Not every formally imaginable combination must be compatible.
The enclosing boundary \(B_n\) determines which joint peer realizations belong to its admissible count domain.
Schematically:
\[ \text{peer realizations} \rightarrow \text{compatibility at } B_n \rightarrow \text{state domain at } B_n. \]
This is why the enclosing boundary is not simply a passive container.
It defines the relation under which subordinate realizations count together.
Compatibility Comes Before Multiplicity
Once the boundary has defined which subordinate realizations are compatible, those realizations can be classified into readable states.
If many compatible subordinate realizations are classified as the same readable state, that state has multiplicity.
Thus the logical order is:
\[ \boxed{ \text{boundary} \rightarrow \text{compatible realizations} \rightarrow \text{classification} \rightarrow \text{multiplicity} } \]
Multiplicity therefore does not create the boundary.
The boundary defines the count domain in which multiplicity can be counted.
Resolution Changes the Count Boundary
This has an important consequence for observation.
Suppose a subordinate transition is not individually readable at \(B_n\).
One might imagine that better experimental resolution simply reveals that hidden transition while leaving the scientific description unchanged.
R2D says something more precise.
To read the subordinate transition directly, the observer must move to the subordinate boundary where that transition is distinguishable.
But once the boundary changes, the state classification changes too.
The transition that had microstate meaning relative to \(B_n\) now has macrostate meaning at \(B_{n-1}\).
Therefore:
\[ \boxed{ \text{increased resolution} \rightarrow \text{change of boundary} \rightarrow \text{change of state meaning} } \]
Observation has not merely revealed more detail about the same formal state.
It has changed the count domain from which the state is defined.
The Causal Question Changes Too
Return to the ten-coin example.
At the ten-coin boundary, one may ask:
Why does the ensemble occupy one macrostate rather than another?
To resolve an individual coin transition, the observer changes to the single-coin boundary.
The question is now:
Why does this coin realize heads rather than tails?
These are not the same causal question at different resolutions.
They belong to different count domains.
This matters because bottom-up causation is often inferred by resolving smaller and smaller mechanisms.
R2D warns that increasing resolution changes the boundary from which causality is being defined.
A subordinate mechanism can be necessary for the realization of an enclosing state without therefore being the source of the organization that distinguishes the enclosing state.
A Boundary Is Not Coarse-Graining
The difference from ordinary coarse-graining is fundamental.
A coarse-graining picture often suggests:
\[ \text{same underlying state} \rightarrow \text{less information}. \]
R2D instead proposes:
\[ \text{new boundary} \rightarrow \text{new classification} \rightarrow \text{new state domain}. \]
The subordinate realization can remain physically consequential.
But its internal path, source identity, and state classification need not remain readable as the same objects from the enclosing boundary.
Changing the boundary is therefore not merely forgetting information.
It changes the ontology of the count.
Boundaries Make Recursive Replacement Possible
Once statehood is boundary dependent, recursive replacement follows naturally.
A subordinate state is readable at its own boundary.
When an enclosing boundary becomes the count domain, the subordinate distinctions no longer remain the states being counted.
Their compatible realizations instead participate in defining the multiplicity and state structure of the enclosing object.
Thus:
\[ \text{subordinate state domain} \rightarrow \text{boundary change} \rightarrow \text{enclosing state domain}. \]
The subordinate object has not been erased.
Its count meaning has changed.
R2D calls this operation recursive replacement.
Scientific Objecthood Is Boundary Indexed
The broader implication is straightforward.
A particle can be real.
A molecule can be real.
A protein can be real.
A cell can be real.
A muscle can be real.
A planet can be real.
A spacetime metric can be real.
R2D does not require any of them to be illusions.
It requires only that their scientific identity be indexed to the boundary that makes their states distinguishable.
A scientific object can therefore be:
- completely real;
- measurable;
- predictive;
- and causally useful
without being primitive at every scale.
This is why R2D is not merely a theory about particular objects.
It is a proposed theory of scientific objecthood.
The Boundary Comes First
The primitive sequence can now be expanded:
\[ \boxed{ \text{boundary} \rightarrow \text{distinction} \rightarrow \text{compatible realization} \rightarrow \text{state classification} \rightarrow \text{count} } \]
The scientific object appears only after the count domain has been defined.
This is the inversion at the foundation of R2D:
The boundary does not merely tell us where the object is.
The boundary helps define what the object is.