# RCHM × Hodge Conjecture — Research Round 1
## Witness Defect, Deformation Handoff, and Rigid Obstruction

**Status:** Internal research note  
**Version:** Round 1 / v0.1  
**Date:** 2026-09-17  
**Methodological base:** Relational Constraint–Handoff Methodology (RCHM) v0.1

---

## 0. Research Goal

This round applies RCHM to the Hodge conjecture.

The purpose is not to restate the conjecture, but to decompose its unresolved content into structurally distinct defects:

1. lack of an algebraic witness;
2. failure of an existing witness to deform with the Hodge class;
3. rigid Hodge states with no deformation path at all;
4. higher-order formal obstructions hidden beyond tangent-space comparison.

The central RCHM question is:

$$
\boxed{
\text{Does every legally admissible Hodge state possess an algebraic witness?}
}
$$

---

## 1. Static Reformulation

Let $X$ be a smooth projective complex algebraic variety and fix codimension $p$.

Define the rational Hodge-class domain

$$
\mathcal H^p(X)
=
H^{2p}(X,\mathbb Q)
\cap
H^{p,p}(X).
$$

Define the algebraically witnessed domain

$$
\mathcal A^p(X)
=
\operatorname{Im}
\left(
cl^p:
CH^p(X)_{\mathbb Q}
\to
H^{2p}(X,\mathbb Q)
\right).
$$

One always has

$$
\mathcal A^p(X)
\subseteq
\mathcal H^p(X).
$$

The Hodge conjecture asks whether

$$
\boxed{
\mathcal A^p(X)=\mathcal H^p(X).
}
$$

Under RCHM, this is naturally expressed as a witness-completeness problem.

Define the static witness defect

$$
\boxed{
\mathfrak D^p(X)
=
\mathcal H^p(X)/\mathcal A^p(X).
}
$$

Then the conjecture is equivalent to

$$
\mathfrak D^p(X)=0.
$$

This quotient is only the starting point. It does not yet explain why a legal Hodge state should possess a source witness.

---

## 2. Family-Level Handoff Structure

Let

$$
\pi:\mathcal X\to S
$$

be a smooth projective family.

Let

$$
\alpha_s\in H^{2p}(X_s,\mathbb Q)
$$

be a locally parallel transported rational cohomology class.

Define the Hodge locus

$$
H_\alpha
=
\left\{
s\in S:
\alpha_s\in H^{p,p}(X_s)
\right\}.
$$

This is the locus in which the transported class remains legally admissible as a Hodge class.

Now introduce a parameter space $M_\alpha$ for algebraic-cycle witnesses whose classes equal $\alpha$ on the corresponding fiber:

$$
q:M_\alpha\to S.
$$

For $m\in M_\alpha$ over $s=q(m)$, let the corresponding algebraic cycle be $Z_m$ and require

$$
[Z_m]=\alpha_s.
$$

Then necessarily

$$
q(M_\alpha)\subseteq H_\alpha.
$$

The RCHM handoff question becomes

$$
\boxed{
q(M_\alpha)
\stackrel{?}{=}
H_\alpha.
}
$$

Interpretation:

> Wherever the class remains Hodge, can an algebraic witness follow it?

This converts the static Hodge conjecture into a deformation-handoff problem.

---

## 3. First-Order Handoff Defect

Assume

$$
m\in M_\alpha,
\qquad
s=q(m).
$$

The Hodge-admissible infinitesimal directions are

$$
T_sH_\alpha.
$$

The infinitesimal directions along which the existing algebraic witness can deform are

$$
dq_m(T_mM_\alpha).
$$

Define the first-order handoff defect

$$
\boxed{
\Delta_1(\alpha,m)
=
T_sH_\alpha
\big/
dq_m(T_mM_\alpha).
}
$$

If

$$
\Delta_1(\alpha,m)=0,
$$

then every first-order Hodge-admissible deformation direction is covered by the witness component through $m$.

If

$$
\Delta_1(\alpha,m)\neq 0,
$$

then there exist first-order directions along which the cohomology class remains Hodge but the chosen witness does not deform accordingly.

Thus

$$
\Delta_1
$$

measures a genuine relational handoff defect.

---

## 4. Why Tangent-Space Equality Is Not Enough

A key correction appears immediately.

Even if

$$
\Delta_1(\alpha,m)=0,
$$

one cannot conclude that the witness extends to all higher orders.

A Hodge locus may be nonreduced, or its tangent space may overestimate the actual geometric freedom.

Therefore tangent-level agreement does not imply formal witness completeness.

The correct hierarchy is

$$
\Delta_1,\Delta_2,\Delta_3,\ldots
$$

where $\Delta_k$ measures failure of witness lifting to order $k$.

Equivalently, introduce deformation functors

$$
\operatorname{Def}_{cycle},
\qquad
\operatorname{Def}_{Hodge}.
$$

For every local Artin algebra $A$, there is a natural comparison map

$$
\operatorname{Def}_{cycle}(A)
\to
\operatorname{Def}_{Hodge}(A).
$$

The stronger condition is formal surjectivity:

$$
\boxed{
\forall A,
\quad
\operatorname{Def}_{cycle}(A)
\twoheadrightarrow
\operatorname{Def}_{Hodge}(A).
}
$$

This is the appropriate RCHM notion of

$$
\textbf{formal witness completeness}.
$$

---

## 5. Dominant Witness Handoff Criterion

Let $H$ be an irreducible component of $H_\alpha$.

Let $M$ be an irreducible proper component of the relevant cycle parameter space such that

$$
q(M)\subseteq H.
$$

Suppose there exists

$$
m\in M
$$

such that

$$
dq_m:T_mM\to T_{q(m)}H
$$

is surjective.

Then locally,

$$
\dim q(M)=\dim H.
$$

Since $q$ is proper, $q(M)$ is closed.

Because

$$
q(M)\subseteq H,
$$

and both are irreducible with the same dimension, one obtains at the reduced set-theoretic level

$$
\boxed{
q(M)=H.
}
$$

This gives a practical criterion:

$$
\boxed{
\text{one witness component}
+
\text{full infinitesimal handoff}
+
\text{properness}
\Rightarrow
\text{component-level coverage}.
}
$$

This does not by itself solve higher-order scheme-theoretic issues, but it provides a concrete bridge from local deformation control to global component coverage.

---

## 6. Three Distinct Obstruction Classes

RCHM separates the unresolved Hodge problem into at least three classes.

### 6.1 Anchor Defect

A legal Hodge state exists, but no algebraic witness is known.

Formally,

$$
\alpha\in\mathcal H^p(X),
$$

but

$$
\alpha\notin\mathcal A^p(X)
$$

is not ruled out.

This is the most direct form of witness absence.

---

### 6.2 Handoff Defect

An algebraic witness exists at one point, but it cannot be shown to follow all legal Hodge deformations.

This occurs when

$$
q(M_\alpha)\subsetneq H_\alpha
$$

or infinitesimally when

$$
\Delta_1(\alpha,m)\neq 0.
$$

Higher-order versions occur when first-order lifting succeeds but formal lifting fails.

---

### 6.3 Rigid Defect

Suppose

$$
\dim H_\alpha=0.
$$

Then the legal Hodge state has no nontrivial deformation direction.

In RCHM language,

$$
\operatorname{Esc}(H_\alpha)=\varnothing.
$$

However,

$$
\boxed{
\operatorname{Esc}(H_\alpha)=\varnothing
\not\Rightarrow
\alpha\text{ is algebraic}.
}
$$

This is a crucial methodological warning:

$$
\boxed{
\text{elimination of escape freedom}
\neq
\text{production of a witness}.
}
$$

Rigidity proves lack of deformation freedom, not algebraicity.

---

## 7. First Global Decomposition

The unresolved structure can therefore be schematically written as

$$
\boxed{
\text{Hodge obstruction}
=
\text{Anchor defect}
+
\text{Handoff defect}
+
\text{Rigid defect}
+
\text{higher-order formal defect}.
}
$$

The terms are not intended as additive numerical invariants yet.

They are structural classes of failure.

---

## 8. Calibration Against the Codimension-One Case

For

$$
p=1,
$$

the Lefschetz $(1,1)$ theorem identifies integral Hodge classes of degree $2$ with first Chern classes of line bundles.

Thus

$$
c_1:
\operatorname{Pic}(X)
\to
H^2(X,\mathbb Z)
$$

provides a direct witness interface.

After rationalization,

$$
\mathcal A^1(X)=\mathcal H^1(X).
$$

Hence

$$
\mathfrak D^1(X)=0.
$$

RCHM interpretation:

$$
\boxed{
\text{legal Hodge state}
\Longleftrightarrow
\text{line-bundle witness}
}
$$

is mediated by a canonical interface

$$
c_1.
$$

This suggests a central higher-codimension question:

$$
\boxed{
\text{What structure can play the role of a higher-codimensional replacement for }c_1?
}
$$

This does not require a literal generalization of $c_1$.

The search target is any sufficiently canonical witness interface that converts legal cohomological states into algebraic sources.

---

## 9. Deformation Handoff vs. Bridge-Object Handoff

Two broad research strategies emerge.

### 9.1 Deformation Handoff

Start with one algebraic witness and propagate it through the Hodge locus.

Goal:

$$
q(M_\alpha)=H_\alpha.
$$

Main tools potentially include:

- variation of Hodge structure;
- infinitesimal variation of Hodge structure;
- Hilbert and Chow schemes;
- normal-bundle deformation theory;
- semiregularity;
- obstruction theory;
- period calculations;
- formal deformation theory.

---

### 9.2 Bridge-Object Handoff

When a direct witness is difficult to construct, introduce an intermediate geometric object $Y$:

$$
X
\longleftrightarrow
Y
\longrightarrow
\text{known algebraic class}.
$$

Then transport or pull the algebraic class back to the original target.

Abstractly:

$$
\alpha_X
\leftarrow
\beta_Y
\leftarrow
Z_Y.
$$

This produces a relational bridge:

$$
\text{Hodge state on }X
\to
\text{geometric structure on }Y
\to
\text{algebraic witness}.
$$

Under RCHM, bridge objects are not auxiliary decoration; they are new interface domains.

---

## 10. Why Hypersurfaces Are a Good Next Test Domain

Even-dimensional smooth hypersurfaces provide a useful controlled laboratory because both sides can often be made explicit.

The Hodge side can be studied through:

$$
\text{periods},
\quad
\text{Jacobian rings},
\quad
\text{IVHS},
\quad
\text{Hodge loci}.
$$

The witness side can be studied through:

$$
\text{linear cycles},
\quad
\text{complete intersections},
\quad
\text{Hilbert schemes},
\quad
\text{normal bundles}.
$$

Thus the following quantities may be computed or estimated:

$$
T_sH_\alpha,
$$

$$
dq_m(T_mM_\alpha),
$$

$$
\Delta_1(\alpha,m),
$$

and eventually

$$
\Delta_2,\Delta_3,\ldots
$$

This makes hypersurfaces a suitable first domain for actual RCHM defect calculations.

---

## 11. Next Research Target

The next round should focus on a concrete family of even-dimensional hypersurfaces.

The working objective is:

$$
\boxed{
\text{compute or structurally classify }
\Delta_1
\text{ and the first nonzero higher-order defect}.
}
$$

The key diagnostic question is whether an apparently excess Hodge deformation direction corresponds to:

1. a genuinely missing algebraic witness;
2. a different witness component not yet included;
3. a nonreduced or fake infinitesimal direction;
4. a higher-order obstruction;
5. a signal that a bridge-object construction is required.

This gives a concrete research loop:

$$
\text{Hodge tangent}
\to
\text{cycle tangent}
\to
\Delta_1
\to
\text{higher jets}
\to
\text{witness classification}.
$$

---

## 12. Round 1 Conclusion

RCHM changes the Hodge conjecture from a single global equality

$$
\mathcal A^p(X)=\mathcal H^p(X)
$$

into a hierarchy of handoff questions.

The main output of Round 1 is the decomposition

$$
\boxed{
\text{legal state}
\to
\text{anchor witness}
\to
\text{deformation handoff}
\to
\text{formal lifting}
\to
\text{global component coverage}.
}
$$

The central methodological correction is:

$$
\boxed{
\text{rigidity is not algebraicity}.
}
$$

Eliminating all escape directions may reduce the problem to a rigid state, but a rigid legal state may still lack a known source witness.

Therefore the next useful quantity is not merely the size of the Hodge locus, but the mismatch between legal deformation freedom and witness deformation freedom:

$$
\boxed{
\Delta_1(\alpha,m)
=
T_sH_\alpha/
dq_m(T_mM_\alpha).
}
$$

The next round should attempt the first explicit RCHM defect calculation in a hypersurface family and determine whether the residual directions are genuine, formal, or witness-dependent.