Chapter 7: Constitutional Review as Path Homology

Beyond Rules: The Shape of the Legal Space

The standard view of the Constitution is that it is a set of rules — prohibitions (Congress shall make no law…), grants of power (the judicial Power shall extend to…), and structural provisions (separation of powers, federalism). Constitutional review, on this view, is the process of checking whether a statute violates one of these rules.

But rules are local. They tell you what is prohibited at each point — what the government may not do in any specific instance. They do not tell you about the global shape of the space of permissible legal states. They do not tell you whether the rights guaranteed by the Constitution form a connected web or a disconnected collection of isolated islands. They do not tell you whether a new statute, which passes every local rule test, might nonetheless create a global inconsistency — a contradiction that emerges only when you follow a chain of reasoning through the full space of legal states.

The Constitution does more than impose rules. It defines the topology — the global shape — of the space of permissible legal states. Statutes must exist within this topology. A statute that violates the Constitution is one that requires the legal space to have a shape that the Constitution prohibits.

The Constitutional Subcomplex

Definition (Constitutional Subcomplex). The constitutional subcomplex \mathcal{C} is a subcomplex of \mathcal{K} consisting of all vertices and simplices that satisfy the constitutional constraints:

\mathcal{C} = \{ \sigma \in \mathcal{K} \mid \Phi_k(\sigma) = \text{true}, \; k = 1, \ldots, K \}

where each \Phi_k is a Boolean predicate encoding a constitutional provision:

• \Phi_{\text{EP}}: Equal Protection — J_{\text{law}} is invariant under protected-class transformations at the vertices of \sigma.
• \Phi_{\text{DP}}: Due Process — J_{\text{law}} is well-defined on the quotient space (procedural regularity).
• \Phi_{\text{1A}}: First Amendment — the speech dimension is unconstrained by government action (within established exceptions).
• \Phi_{\text{SoP}}: Separation of Powers — the legislative, executive, and judicial components are in their proper domains.

The constitutional subcomplex is the region of the judicial complex where all constitutional constraints are simultaneously satisfied. It is the “constitutionally permissible zone” — the subspace of legal states that the Constitution allows.

A statute that is constitutional lives inside \mathcal{C}: its effects are confined to the constitutionally permissible zone. A statute that is unconstitutional extends outside \mathcal{C}: it creates legal states that violate one or more constitutional constraints.

But this description — inside versus outside — is too crude. The real question is not whether the statute’s effects are “inside” the constitutional subcomplex but whether the statute changes the topology of the constitutional subcomplex. This is the insight that transforms constitutional review from a rule-checking exercise into a topological computation.

Why Path Homology?

The judicial complex \mathcal{K} is directed: citations go from later cases to earlier cases, and authority flows from higher courts to lower courts. Standard simplicial homology — the kind computed by Ripser and other persistent-homology software — operates on undirected complexes. Computing standard homology on a directed citation network requires “forgetting” edge directions, which destroys the temporal and hierarchical constraints that are legally essential.

This creates two failure modes:

1. False negatives. A directed cycle detected by path homology may disappear when directions are forgotten — it may become a boundary in the undirected complex. Standard homology would miss a genuine legal inconsistency.

2. False positives. Standard homology may detect undirected cycles that are not traversable in the directed graph — they require following a citation backward in time. Standard homology would produce a false alarm.

The appropriate homology theory is path homology, developed by Grigor’yan, Lin, Muranov, and Yau. Path homology is defined on directed graphs and digraphs: its cycles are directed paths that return to their starting vertex, and two directed cycles are homologous if they can be deformed into each other through directed homotopies. This is exactly the structure needed for legal reasoning, where the order of citations matters and temporal flow is irreversible.

The Formal Construction

Definition (Path Homology of the Judicial Complex). An elementary p-path in \mathcal{K} is a sequence of vertices e_{i_0 i_1 \cdots i_p} = (c_{i_0}, c_{i_1}, \ldots, c_{i_p}) such that each consecutive pair is connected by a directed edge c_{i_k} \to c_{i_{k+1}}.

The path complex \Omega_p(\mathcal{K}) is the free abelian group generated by all elementary p-paths.

The boundary operator \partial_p: \Omega_p \to \Omega_{p-1} is the alternating sum of face maps (omitting each intermediate vertex in turn), restricted to allowed paths — those whose faces are themselves paths in \mathcal{K}.

The path homology groups are:

\widetilde{H}_n^{\text{path}}(\mathcal{K}; \mathbb{Z}) = \ker \partial_n / \text{im} \, \partial_{n+1}

Let us unpack this definition.

p-paths. A 0-path is a single vertex. A 1-path is a directed edge c_i \to c_j. A 2-path is a pair of consecutive directed edges c_i \to c_j \to c_k. A directed cycle is a 1-path that returns to its starting vertex: c_0 \to c_1 \to \cdots \to c_m \to c_0.

The boundary operator. The boundary of a 2-path c_i \to c_j \to c_k is the alternating sum of its faces:

\partial_2(c_i \to c_j \to c_k) = (c_j \to c_k) - (c_i \to c_k) + (c_i \to c_j)

This is the directed analogue of the boundary of a triangle in simplicial homology. But there is a crucial difference: the face (c_i \to c_k) is only included if the directed edge c_i \to c_k exists in \mathcal{K}. If it does not, the face is absent from the boundary formula, and the 2-path contributes a non-trivial element to the path homology.

Cycles and boundaries. A cycle is a p-path whose boundary is zero: \partial_p \alpha = 0. A boundary is a p-path that is the boundary of a (p+1)-path: \alpha = \partial_{p+1} \beta. The path homology group \widetilde{H}_n^{\text{path}} measures the directed cycles that are not boundaries — the directed cycles that cannot be “filled in” by higher-dimensional directed paths.

What Path Homology Detects

\widetilde{H}_0^{\text{path}}(\mathcal{K}; \mathbb{Z}) counts the number of directed-path-connected components — the number of groups of vertices such that you can reach any vertex in the group from any other by following directed paths. If \widetilde{H}_0^{\text{path}} has rank greater than 1, some regions of the judicial complex are unreachable from others via directed paths (even if they are connected in the undirected graph).

\widetilde{H}_1^{\text{path}}(\mathcal{K}; \mathbb{Z}) detects non-trivial directed cycles — directed loops that cannot be contracted within the directed graph. These are the legal inconsistencies we are most interested in: sequences of legal reasoning that start and end at the same point but with different Hohfeldian labels.

\widetilde{H}_2^{\text{path}}(\mathcal{K}; \mathbb{Z}) detects “directed bubbles” — directed 2-cycles where a collection of directed cycles forms a closed surface in the directed graph. These correspond to self-contained legal regimes whose internal reasoning is sound but whose boundary violates a constitutional constraint.

The Main Result

Theorem (Topological Constitutionality). Let \mathcal{C} be the constitutional subcomplex and let \ell be a proposed statute that modifies the judicial complex \mathcal{K} to \mathcal{K}_\ell. The statute \ell is constitutional if and only if the inclusion \mathcal{C} \hookrightarrow \mathcal{K}_\ell preserves the path homology:

\widetilde{H}_n^{\text{path}}(\mathcal{C}; \mathbb{Z}) \cong \widetilde{H}_n^{\text{path}}(\mathcal{C}_\ell; \mathbb{Z}) \quad \forall \; n \geq 0

where \mathcal{C}_\ell = \mathcal{C} \cap \mathcal{K}_\ell is the constitutional subcomplex in the modified judicial complex. Equivalently: \ell is constitutional if and only if it does not create new non-trivial directed cycles in the constitutional subcomplex.

The Proof

Proof sketch. A non-trivial directed cycle in \mathcal{C}_\ell that does not exist in \mathcal{C} represents a directed sequence of legal reasoning — applications of the statute combined with constitutional provisions, following the temporal and hierarchical order of legal authority — that returns to the starting vertex but with altered Hohfeldian structure. This is a legal contradiction: a sequence of legally valid steps, each following proper authority, whose net effect violates a constitutional guarantee.

More precisely: let \gamma = (c, c_1, \ldots, c_m, c) be a directed 1-cycle in \mathcal{C}_\ell that is not a boundary in \mathcal{C}. Then \gamma represents a legal path that:

1. Starts at a vertex c satisfying all constitutional constraints.
2. Follows directed edges (each respecting temporal and hierarchical order) corresponding to the statute \ell and existing provisions.
3. Returns to c.
4. But the Hohfeldian labels transported around \gamma have changed — non-trivial directed holonomy.

Because the cycle is directed, it cannot be dismissed as an artifact of ignoring legal hierarchy: each step follows proper authority. This is a genuine inconsistency.

The “only if” direction: if the path homology is not preserved, there exists a new non-trivial directed cycle, which represents a constitutional violation (as above). The “if” direction: if the path homology is preserved, no new directed cycles are created, and the statute’s effects are topologically compatible with the constitutional constraints. \square

What the Theorem Means

The theorem gives a precise mathematical meaning to “unconstitutional”:

An unconstitutional statute creates a topological obstruction — a non-trivial element of \widetilde{H}_1^{\text{path}}(\mathcal{C}_\ell) / \widetilde{H}_1^{\text{path}}(\mathcal{C}). The obstruction is a directed cycle that exists in the modified constitutional subcomplex but not in the original.

A constitutional amendment is a topological surgery — it modifies \mathcal{C} itself, potentially creating or destroying path-homology classes. The Thirteenth Amendment destroyed the path-homology classes that permitted slavery. The Nineteenth Amendment destroyed the path-homology classes that permitted gender-based disenfranchisement. Each amendment changed the topology of the constitutional subcomplex.

Judicial review is the process of computing path homology. Given a proposed statute, the court computes \widetilde{H}_n^{\text{path}}(\mathcal{C}_\ell) and checks whether it equals \widetilde{H}_n^{\text{path}}(\mathcal{C}). For a finite directed graph, this is a finite linear-algebra computation — specifically, it requires computing the Smith normal form of the boundary matrices, which is polynomial in the size of the graph.

RUNNING EXAMPLE — THE DIGITAL PRIVACY ACT AND PATH HOMOLOGY

Judge Rivera analyzes the Digital Privacy and Public Safety Act through the lens of path homology.

Step 1: Identify the constitutional subcomplex \mathcal{C}. The relevant constitutional provisions are: - First Amendment: freedom of speech, freedom of the press, freedom of association - Fourth Amendment: protection against unreasonable searches and seizures - Due Process: procedural regularity

The constitutional subcomplex \mathcal{C} is the subgraph of decided cases that satisfy all three provisions. It includes cases establishing the right to anonymous speech (McIntyre v. Ohio, 1995), the warrant requirement for digital data (Carpenter v. United States, 2018), and the right of journalists to protect their sources (Branzburg v. Hayes, 1972, partially).

Step 2: Model the statute’s effect on \mathcal{K}. The Digital Privacy Act adds new edges to the judicial complex: - An edge from “user identity data held by platform” to “law enforcement possession of identity data” — this edge has zero weight (the statute makes the transfer automatic, without a warrant). - An edge from “anonymous online speaker” to “identified speaker” — this edge represents the mandatory identity verification requirement.

These edges modify the constitutional subcomplex: \mathcal{C}_\ell = \mathcal{C} \cap \mathcal{K}_\ell.

Step 3: Compute path homology of \mathcal{C}_\ell. Rivera traces directed cycles in the modified subcomplex:

Cycle 1 (First Amendment): c_{\text{anonymous speech}} \to c_{\text{identified by Act}} \to c_{\text{government knowledge of identity}} \to c_{\text{chilling effect}} \to c_{\text{anonymous speech}}

This cycle starts with the constitutional right to anonymous speech (McIntyre), follows the Act’s identification requirement, arrives at government knowledge of the speaker’s identity, recognizes the chilling effect on speech, and returns to the anonymous speech right. But the Hohfeldian labels have changed: the speaker started with a right to anonymous speech and returned with a no-right — the Act has extinguished the anonymity right by making identification mandatory.

Is this cycle a boundary? For it to be a boundary in \mathcal{C}, there would need to exist a 2-chain whose boundary is this cycle. But the directed edge from “anonymous speaker” to “identified speaker” did not exist in \mathcal{C} before the Act — the Constitution does not permit warrantless compelled identification of speakers. The cycle is a new element of \widetilde{H}_1^{\text{path}}(\mathcal{C}_\ell) that does not exist in \widetilde{H}_1^{\text{path}}(\mathcal{C}).

Cycle 2 (Fourth Amendment): c_{\text{digital data privacy}} \to c_{\text{warrantless disclosure}} \to c_{\text{government possession}} \to c_{\text{Carpenter warrant requirement}} \to c_{\text{digital data privacy}}

This cycle starts with the established Fourth Amendment protection of digital data (Carpenter), follows the Act’s warrantless disclosure provision, arrives at government possession of the data, and returns to the Carpenter holding. The holonomy is non-trivial: Carpenter says a warrant is required; the Act says it is not. Traversing the cycle changes the Hohfeldian label from “right to privacy” to “no-right.”

Step 4: Constitutional determination. \widetilde{H}_1^{\text{path}}(\mathcal{C}_\ell) \ncong \widetilde{H}_1^{\text{path}}(\mathcal{C}). The Act creates two new non-trivial directed cycles in the constitutional subcomplex. The statute is unconstitutional.

Rivera has not merely “applied a test” or “balanced interests.” She has computed a topological invariant and found that it has changed. The unconstitutionality is not a matter of judicial philosophy or political temperament — it is a mathematical fact about the directed graph structure of the legal system.

\widetilde{H}_0^{\text{path}}: Connectivity

The zeroth path homology group \widetilde{H}_0^{\text{path}}(\mathcal{C}; \mathbb{Z}) counts the directed-path-connected components of the constitutional subcomplex. A statute that makes some constitutional states unreachable from others via directed paths (respecting legal hierarchy) creates an \widetilde{H}_0^{\text{path}} violation — even if the undirected graph remains connected.

This is a real legal phenomenon. A statute that creates an administrative process so burdensome that constitutional rights are de facto inaccessible — even though they remain de jure available — may not create a new directed cycle (no \widetilde{H}_1^{\text{path}} violation) but may disconnect the constitutional subcomplex (an \widetilde{H}_0^{\text{path}} violation). The right still “exists” as a vertex, but no directed path from the citizen’s current position can reach it.

\widetilde{H}_2^{\text{path}}: Constitutional Bubbles

The second path homology group \widetilde{H}_2^{\text{path}}(\mathcal{C}; \mathbb{Z}) detects “constitutional bubbles” — directed 2-cycles where a statute creates a closed directed surface of legal states that are internally consistent but globally incompatible with the Constitution.

Consider a statute that creates a self-contained regulatory regime (say, a new administrative agency with its own enforcement procedures, its own standards of review, and its own remedial structure). The internal directed reasoning of this regime may be sound — every directed cycle within the regime is trivial. But the boundary of the regime, where it interfaces with the rest of the legal system, may violate constitutional constraints.

This is captured by \widetilde{H}_2^{\text{path}}: a 2-cycle is a closed directed surface whose boundary is a 1-cycle. If the 1-cycle (the boundary of the regulatory regime) is non-trivial in \mathcal{C}, the regime violates the Constitution at its interface with the rest of the legal system — even though its interior is constitutionally sound.

Completeness Conjecture

Conjecture (Completeness of Path-Homological Review). The full set of path homology groups \{\widetilde{H}_n^{\text{path}}(\mathcal{C}; \mathbb{Z})\}_{n \geq 0} provides a complete invariant for constitutionality: a statute is constitutional if and only if it preserves all path homology groups of the constitutional subcomplex. Since \mathcal{K} is a finite directed graph, all \widetilde{H}_n^{\text{path}} are finitely generated abelian groups and are exactly computable.

If this conjecture is true, constitutional review is reducible to a finite computation: compute the path homology groups of \mathcal{C} and \mathcal{C}_\ell, compare them, and declare the statute constitutional if and only if they agree. The computation is polynomial in the size of the directed graph.

If the conjecture is false — if there exist constitutional violations not captured by path homology — then path homology provides a necessary but not sufficient condition for constitutionality. Even in this case, the path-homological test would detect a large class of constitutional violations and would provide a formal framework for constitutional analysis that is currently lacking.

Strict Scrutiny, Intermediate Scrutiny, Rational Basis

American constitutional law employs three “levels of scrutiny” to evaluate statutes that impinge on constitutional rights:

Strict scrutiny (for racial classifications, fundamental rights): the statute must be narrowly tailored to achieve a compelling governmental interest. Very few statutes survive strict scrutiny.

Intermediate scrutiny (for gender classifications, commercial speech): the statute must be substantially related to an important governmental interest.

Rational basis review (for most economic and social legislation): the statute must be rationally related to a legitimate governmental interest. Most statutes survive rational basis review.

In the topological framework, these levels of scrutiny correspond to different thresholds for how much topological change is permissible:

Strict scrutiny requires that \widetilde{H}_n^{\text{path}}(\mathcal{C}_\ell) \cong \widetilde{H}_n^{\text{path}}(\mathcal{C}) exactly. No new directed cycles are permitted. The statute must be topologically trivial — it must not change the shape of the constitutional subcomplex at all.

Intermediate scrutiny permits small topological changes — new directed cycles that are “small” in a precise sense (they enclose a small area of the complex, or they involve few vertices). The statute may create new directed cycles, but they must be confined to a small neighborhood and must not propagate throughout the constitutional subcomplex.

Rational basis review permits any topological change that does not disconnect the constitutional subcomplex or destroy fundamental directed cycles. The statute may create new cycles and even change the rank of \widetilde{H}_1^{\text{path}}, as long as the zeroth homology group (\widetilde{H}_0^{\text{path}}, the connectivity structure) is preserved and the most important 1-cycles (those representing fundamental rights) remain non-trivial.

This topological reinterpretation of the levels of scrutiny makes them precise. Instead of debating whether an interest is “compelling” versus merely “important” versus merely “legitimate,” we compute the topological impact of the statute and compare it to the threshold for the applicable level of scrutiny.

How Amendments Change the Topology

A constitutional amendment does not merely add a rule. It modifies the constitutional subcomplex \mathcal{C} itself, potentially changing its path homology.

The Thirteenth Amendment (abolition of slavery) destroyed a family of path-homology classes. Before the amendment, the constitutional subcomplex contained directed paths consistent with the Hohfeldian configuration of slavery (one person having a property right in another, the enslaved person having a no-right). After the amendment, these configurations were removed from \mathcal{C}, destroying the directed cycles that enclosed them.

The Fourteenth Amendment (equal protection, due process, privileges and immunities) added new topological constraints. The Equal Protection Clause created a new predicate \Phi_{\text{EP}} that must be satisfied at every vertex of \mathcal{C}, shrinking the constitutional subcomplex by removing vertices that violate equal protection. This shrinkage destroyed some directed cycles (those passing through racially discriminatory vertices) and potentially created others (new tensions between equal protection and other constitutional provisions).

The Nineteenth Amendment (women’s suffrage) destroyed the directed cycles that permitted gender-based disenfranchisement. The path from “eligible voter” through “state election law” back to “eligible voter” previously had different Hohfeldian labels for men and women; the amendment forced the labels to be gender-invariant, removing the non-trivial holonomy.

Each amendment is a topological surgery — a controlled modification of \mathcal{C} that changes its path-homology groups in a specific, intended way. The amendment process (supermajority in both houses of Congress plus ratification by three-fourths of the states) is the political mechanism for authorizing topological surgery.

A Case Study in Constitutional Phase Transition

The overruling of Plessy v. Ferguson (1896) by Brown v. Board of Education (1954) is the paradigmatic example of a constitutional phase transition — a discontinuous change in the topology of the constitutional subcomplex. The topological framework provides a precise description of what happened.

Before Brown. The constitutional subcomplex \mathcal{C}_{\text{pre-Brown}} contained vertices and edges corresponding to the “separate but equal” doctrine. Specifically:

• Vertices: legal states where racial segregation in public facilities was permissible, as long as the segregated facilities were “equal.”
• Directed edges: citation paths from Plessy to subsequent cases applying the separate-but-equal doctrine to schools, transportation, restaurants, and other public accommodations.
• Hohfeldian structure: the state had a liberty to segregate (no duty to integrate), and African Americans had a no-right to integrated facilities (no claim against segregation, as long as facilities were “equal”).

The path homology of \mathcal{C}_{\text{pre-Brown}} included certain 1-cycles: directed loops through the separate-but-equal region that were non-trivial (they enclosed the segregated legal states as a distinct topological feature).

After Brown. The Supreme Court held that “separate educational facilities are inherently unequal.” This was not merely a new legal rule — it was a topological surgery on \mathcal{C}:

• The “separate but equal” vertices were removed from the constitutional subcomplex. Legal states involving public school segregation no longer satisfied \Phi_{\text{EP}} (the Equal Protection predicate).
• The directed edges connecting Plessy to subsequent segregation cases were severed — or more precisely, their weights were set to infinity (the doctrinal path from Plessy to constitutional validity was blocked).
• The Hohfeldian structure was inverted: African American students now had a right to integrated education, and states had a correlative duty to desegregate.

The path homology changed: \widetilde{H}_1^{\text{path}}(\mathcal{C}_{\text{post-Brown}}) \ncong \widetilde{H}_1^{\text{path}}(\mathcal{C}_{\text{pre-Brown}}). The 1-cycles enclosing the segregated region were destroyed (the segregated vertices were removed), and new 1-cycles were potentially created (the tension between the desegregation mandate and states’ rights produced new directed circuits).

The transition cost. Brown was not merely a doctrinal change — it was a phase transition with enormous cost. Hundreds of subsequent cases had relied on Plessy. School systems, bus systems, and public accommodations across the South had been designed around the separate-but-equal framework. The transition cost — the sum of edge-weight modifications required to implement the topological surgery — was measured in decades of litigation, legislative action, and social upheaval.

The framework makes the transition cost computable:

\text{Cost}_{\text{overrule}} = \sum_{i=1}^{n} w_{P_i} \cdot |\Delta w(c_i, \cdot \, ; P)|

where the sum is over all cases c_1, \ldots, c_n that relied on Plessy, w_{P_i} is the precedential weight of each case, and |\Delta w| is the magnitude of the edge-weight change. For Brown, this sum was enormous — arguably the largest transition cost in American constitutional history.

Why the Transition Was Necessary

The topological framework also explains why Brown was necessary. The separate-but-equal doctrine created a non-trivial Wilson loop in the constitutional subcomplex. Starting from the Equal Protection Clause (“no State shall deny to any person within its jurisdiction the equal protection of the laws”), traversing the Plessy doctrine (“separate but equal satisfies equal protection”), and returning to the Equal Protection Clause produced a non-trivial holonomy: the Hohfeldian labels were different at the end of the loop than at the beginning.

Specifically: the Equal Protection Clause assigns all persons a right to equal treatment. Plessy assigned some persons (African Americans) a no-right to integrated facilities. The Wilson loop was W(\gamma) = s (jural negation: right \to no-right). This non-trivial Wilson loop meant that the pre-Brown constitutional subcomplex was internally inconsistent — the Equal Protection Clause and the separate-but-equal doctrine could not coexist without contradicting each other.

Brown resolved the inconsistency by choosing which element to preserve: the Equal Protection Clause (a constitutional provision) over Plessy (a judicial interpretation). The topological surgery removed the inconsistent vertices, collapsed the non-trivial Wilson loop to the identity, and restored gauge invariance to the Equal Protection region of \mathcal{C}.

The Finite Computation

For a finite directed graph G with n vertices and m edges, the path homology groups \widetilde{H}_p^{\text{path}}(G; \mathbb{Z}) are computable by finite linear algebra:

1. Enumerate all elementary p-paths (sequences of p+1 vertices connected by directed edges).
2. Construct the boundary matrices \partial_p.
3. Compute the Smith normal form of each boundary matrix.
4. Read off the homology groups: \widetilde{H}_p^{\text{path}} = \ker \partial_p / \text{im} \, \partial_{p+1}.

The computational complexity is polynomial in the number of paths, which in turn is bounded by n^{p+1} (the number of (p+1)-tuples of vertices). For p = 1 (the most important case for constitutional review), the complexity is at most O(n^3) — the cost of computing the Smith normal form of a matrix with at most n^2 rows and n^3 columns.

For a realistic constitutional subcomplex — even one with millions of vertices — the computation is feasible using standard sparse-matrix algorithms. Path homology implementations for directed graphs exist and have been validated on graphs of comparable size.

Machine-Checkable Constitutionality

The topological constitutionality theorem has a radical implication: constitutionality is machine-checkable. Given a directed graph representation of the constitutional subcomplex and a specification of how a proposed statute modifies the graph, a computer can compute the path homology before and after the modification and determine whether the statute is constitutional.

This does not mean that constitutional review should be fully automated. The construction of the constitutional subcomplex — which vertices satisfy \Phi_{\text{EP}}, \Phi_{\text{DP}}, \Phi_{\text{1A}}, \Phi_{\text{SoP}} — requires legal judgment that cannot be fully automated. The attribute-vector scoring, the regime-boundary identification, and the constitutional-predicate evaluation all involve human decisions.

But given the constitutional subcomplex, the topological test is automatic. This separates the task of constitutional review into a human component (constructing the subcomplex) and a computational component (computing the path homology). The human component requires legal expertise. The computational component requires linear algebra.

RUNNING EXAMPLE — RIVERA’S OPINION

Judge Rivera writes her opinion striking down the Digital Privacy and Public Safety Act. Her analysis follows the topological framework, translated into the language of constitutional law:

“The Act creates two constitutional defects. First, by requiring social media companies to verify and disclose user identities without a warrant, the Act creates a pathway from constitutionally protected anonymous speech (McIntyre v. Ohio Elections Commission, 1995) to government identification of speakers that is inconsistent with the First Amendment’s protection of anonymous political speech. This pathway generates a contradiction: the First Amendment guarantees a right to speak anonymously, but the Act extinguishes that right by making anonymity impossible. Starting from the constitutional guarantee, following the Act’s requirements, and returning to the guarantee produces a different legal configuration than the one we started with — proof that the Act and the Constitution are structurally incompatible.”

“Second, the Act’s warrantless disclosure provision is inconsistent with Carpenter v. United States (2018), which held that the government must obtain a warrant to access certain categories of digital data. The Act permits warrantless access to identity data, creating a contradiction with Carpenter’s holding. This contradiction is not resolved by the Act’s claim that identity data is ‘less sensitive’ than location data — the Fourth Amendment’s warrant requirement does not depend on the government’s assessment of data sensitivity.”

“The state has articulated legitimate interests in public safety and fraud prevention. Under strict scrutiny, however, these interests must be served by the least restrictive means. The Act is not narrowly tailored: less restrictive alternatives exist (targeted warrants for specific users, administrative subpoenas with judicial oversight) that would serve the state’s interests without creating the constitutional contradictions identified above.”

Rivera does not use the words “path homology” or “directed cycle.” But her reasoning — identifying contradictory paths through the legal system, checking whether less disruptive alternatives exist — is precisely the computation that the topological framework formalizes.

Original Meaning vs. Living Constitution

The oldest debate in constitutional theory is between originalism (the Constitution means what its ratifiers understood it to mean) and living constitutionalism (the Constitution’s meaning evolves with changing social understandings). The topological framework is neutral between these positions — but it reframes the debate in precise terms.

Originalism corresponds to a fixed constitutional subcomplex: \mathcal{C} is determined at the time of ratification and does not change (except by formal amendment). The path homology of \mathcal{C} is fixed, and every statute is tested against this fixed topology. New cases may explore new regions of the complex, but the topological constraints do not change.

Living constitutionalism corresponds to a slowly evolving constitutional subcomplex: \mathcal{C} changes as social understandings change, through judicial interpretation rather than formal amendment. The path homology evolves, and statutes are tested against the current topology rather than the original topology.

The topological framework does not resolve this debate. But it makes the disagreement precise: originalists and living constitutionalists disagree about whether \mathcal{C} is fixed or evolving. Given either assumption, the topological test is the same — compute the path homology and check preservation. The debate is about the input to the computation (which \mathcal{C} to use), not about the computation itself.

The Role of Judicial Judgment

The topological framework does not eliminate judicial judgment. It relocates it. Instead of asking “is this statute constitutional?” — a question whose answer depends on how one interprets ambiguous constitutional text — the framework asks:

1. What is the constitutional subcomplex? This requires identifying which vertices satisfy the constitutional predicates \Phi_k. This is a judgment call: does a particular legal state satisfy equal protection? Due process? The First Amendment? These are the traditional questions of constitutional law, and they require traditional legal analysis.

2. How does the statute modify the complex? This requires specifying which new edges or vertices the statute introduces. This is also a judgment call: what legal states does the statute create? What new paths does it open? What existing paths does it close?

3. Does the path homology change? This is not a judgment call. Given the answers to questions 1 and 2, the path homology computation is automatic — it is finite linear algebra on a directed graph.

The framework concentrates judicial judgment in questions 1 and 2 (constructing the subcomplex and modeling the statute’s effects) and automates question 3 (computing the topological test). This is a gain in precision: instead of the entire constitutional analysis being a matter of judgment, only the input construction is judgmental — the inference from inputs to conclusion is mechanical.