The Birch and Swinnerton-Dyer Conjecture

Field

Number theory / arithmetic geometry

Clay problem statement

For an elliptic curve \(E/\mathbb Q\), the Birch and Swinnerton-Dyer conjecture predicts a deep relation between two public invariants:

  • the algebraic rank of the Mordell-Weil group \(E(\mathbb Q\));
  • the order of vanishing of the Hasse-Weil \(L\)-function \(L(E,s)\) at s=1.

In particular, the rank equality asserts

$$
\mathrm{ord}_{s=1},L(E,s)=\mathrm{rank},E(\mathbb Q).
$$

The full BSD conjecture also includes a leading-coefficient formula involving arithmetic constants such as the regulator, Tamagawa factors, torsion, and the Tate-Shafarevich group. The present CG-source document addresses the rank equality.

Why the problem is difficult

BSD links two very different mathematical worlds.

On the arithmetic side, the Mordell-Weil group \(E(\mathbb Q\)) is a discrete global object. Its rank counts independent rational directions on the elliptic curve.

On the analytic side, \(L(E,s)\) is a complex analytic object whose behavior at the special point \(s=1\) is determined through analytic continuation and global spectral structure.

Major difficulties include:

  • rational points are discrete arithmetic objects;
  • \(L\)-functions are analytic objects defined through global continuation;
  • the critical point \(s=1\) is structurally subtle;
  • algebraic rank and analytic vanishing order live in different public languages;
  • proving equality requires a bridge between arithmetic endpoint data and analytic endpoint data.

Much progress is known in special cases, but the full conjecture remains open.

Coherence Geometry approach

The current CG formulation treats the BSD rank equality as a source-closure problem.

Instead of treating the Mordell-Weil rank and the analytic order of vanishing as unrelated public invariants that must be forced to agree after the fact, the CG-source approach treats them as endpoint readouts of a common coherent source.

For a fixed elliptic curve \(E/\mathbb Q\), the paper constructs a coherent BSD source

$$S_{\rm BSD}(E),$$

with degree layers governed by local coherence, global obstruction vanishing, and primitive terminal readout. The fixed curve determines a finite CG substrate

$$\Sigma_E.$$

Within this substrate, finite primitive terminal capacity implies that the BSD source has a maximal nonzero terminal degree

$$r.$$

The arithmetic and analytic public quantities are then read as endpoint projections of this same terminal degree.

Author’s Intuition

The Birch and Swinnerton-Dyer conjecture has long carried the flavor of two different descriptions of the same hidden structure.

On the arithmetic side, one counts independent rational directions on an elliptic curve. On the analytic side, one studies how the \(L\)-function vanishes at the critical point \(s=1\).

The CG-source formulation asks whether these are not two unrelated quantities, but two endpoint readouts of one coherent source.

In this view, the common object is not first a regulator determinant or an analytic leading coefficient. It is a finite coherent source with terminal degree \(r\). The Mordell-Weil exterior filtration reads \(r\) arithmetically. The analytic vanishing filtration reads r analytically.

The conjectural equality then becomes a synchronization statement:

$$\text{arithmetic endpoint rank}
=
\text{analytic endpoint rank}
=
\text{terminal source degree}.$$

Source-closure formulation

The paper develops three linked components.

Endpoint recovery

The arithmetic endpoint is identified with the Mordell-Weil exterior filtration. Its maximal nonzero exterior degree is

$$\mathrm{rank} E(\mathbb Q).$$

The analytic endpoint is identified with terminal persistence of the analytically continued germ of \(L(E,s)\) in the vanishing filtration

$$\mathcal I_1^k$$

of analytic germs vanishing at s=1. Its maximal persistence depth is

$$\mathrm{ord}_{s=1} L(E,s).$$

Source closure

For a fixed elliptic curve \(E/\mathbb Q\), the CG construction assigns a finite source substrate \(\Sigma_E\). The degree layers of the BSD source are governed by local coherence and global obstruction vanishing.

The constrained-coherence capacity mechanism inherited from the GCT–SCRC–CS-SCT framework gives a finite upper bound on independent primitive terminal coherence directions. By the Exterior-Terminal Support Law, admissible exterior source layers vanish above this capacity.

Thus the BSD source has a maximal nonzero terminal degree

$$r.$$

Rank synchronization

Rank synchronization states that any source-valid CG endpoint rank projection reads this same terminal degree. A source-valid projection counts only source-supported rank data and is nondegenerate on the primitive terminal rank layer it is constructed to read.

The arithmetic endpoint projection reads

$$r=\mathrm{rank} E(\mathbb Q).$$

The analytic endpoint projection reads

$$r=\mathrm{ord}_{s=1} L(E,s).$$

Therefore the two public rank quantities agree:

$$\mathrm{ord}_{s=1} L(E,s)
=
\mathrm{rank} E(\mathbb Q).$$

Relation to the superseded operator formulation

An earlier BSD manuscript, CGI-RSR-000004, developed an operator/determinant formulation involving arithmetic and analytic quadratic structures. That version has been superseded by the current CG-source formulation, CGI-RSR-000032.

The earlier manuscript is not the current technical reference for the BSD page. Its role is historical: it explored a projection-language bridge through quadratic operators, kernels, determinants, and endpoint spectral structure.

The current document shifts the formulation to the native CG-source level. The rank equality is expressed through source closure, terminal degree, exterior support, endpoint recovery, and rank synchronization.

Thus the current public route is:

$$\text{coherent BSD source}
 \rightarrow
 \text{finite terminal degree} 
\rightarrow \text{arithmetic endpoint readout}\quad \text{and}\quad
 \text{analytic endpoint readout.}$$

This replaces the earlier operator-coincidence framing as the active formulation.

Boundary of the result

The current public document addresses the BSD rank equality, not the full leading-coefficient formula.

The full BSD conjecture includes additional arithmetic constants and regulator/leading-coefficient data. Those require further endpoint structure beyond the rank equality treated in CGI-RSR-000032.

The present claim is therefore focused:

$$\mathrm{ord}_{s=1} L(E,s)
=
\mathrm{rank} E(\mathbb Q).$$

The document should be evaluated according to that stated scope.

Status

CG-source solution path under review.

Current public technical record available.

The Zenodo-hosted PDF is the authoritative technical record. This page is descriptive only.

Technical review

Readers are encouraged to evaluate the work according to its stated claims and mathematical arguments.

Particular points of interest include:

  • construction of the coherent BSD source \(S_{\rm BSD}(E)\);
  • definition and finiteness of the fixed curve substrate \(\Sigma_E\);
  • local coherence and global obstruction-vanishing conditions;
  • use of constrained-coherence capacity from the GCT–SCRC–CS-SCT framework;
  • application of the Exterior-Terminal Support Law;
  • existence of the maximal nonzero terminal degree r;
  • endpoint recovery for the Mordell-Weil exterior filtration;
  • endpoint recovery for the analytic vanishing filtration at s=1;
  • source-validity of the arithmetic and analytic endpoint rank projections;
  • nondegeneracy on the primitive terminal rank layer;
  • the rank synchronization step;
  • the identification of both public rank quantities with the same terminal source degree.

Corrections, counterexamples, simplifications, literature pointers, and independent verification are welcome.

Available document

Back to Clay Millennium Problems Home