🧬 biology

Topology as Logic: Structural Role Geometry Across Formal, Software, Biological, and Prebiotic Systems

This pre-registered study demonstrates that dependency topology correlates with functional load-bearing organization across seven diverse substrates—from formal mathematics and software to biological and prebiotic systems—revealing a measurable "structural role geometry" where betweenness-based persistence outperforms degree-based metrics in identifying operational logic.

Original authors: Vladi Ivanov

Published 2026-06-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Vladi Ivanov

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine you are looking at a massive, tangled ball of yarn. To the naked eye, it's just a mess of strings. But what if you could look at that ball and instantly know which specific threads are holding the whole thing together, and which ones are just decorative loops that could be cut without the ball falling apart?

That is essentially what this paper is trying to do, but instead of yarn, it's looking at systems—like computer chips, mathematical proofs, old banking software, and even the nervous systems of animals.

Here is the simple breakdown of what the author, Vladi Ivanov, discovered:

1. The Big Question: Can Shape Tell Us How Things Work?

Usually, we think of "logic" as rules, words, or code. We think of "shape" (topology) as just how things are connected. They seem like totally different languages.

The author asked: If we ignore the words and rules, and just look at the shape of the connections, can we find the "skeleton" of the system? In other words, can the map of connections tell us which parts are the "bosses" (the critical load-bearing parts) and which parts are just "helpers"?

2. The Two Ways to Measure Importance

The paper introduces a crucial distinction between two ways of measuring a node (a point in the network):

The "Popular Kid" (Degree): This counts how many direct friends a node has. If a node is connected to 50 other things, it has a high "degree."
- The Trap: In many systems, the "popular kids" aren't actually the ones doing the heavy lifting. They might just be connected to a lot of things that don't matter much.
The "Bridge Builder" (Betweenness): This measures how often a node sits on the shortest path between two other nodes. It's the bridge that traffic must cross to get from A to B.
- The Discovery: The paper argues that Betweenness is the real measure of "load-bearing logic." It finds the critical bridges that hold the system together.

3. The "Carry Chain" Analogy (The Digital Circuit Proof)

The strongest proof comes from a digital circuit (a tiny computer chip).

The Setup: Imagine a 4-bit calculator. It adds numbers.
The "Popular Kid" (Degree): The input wires (where you type numbers) have the most connections. They are the "popular kids."
The "Bridge Builder" (Betweenness): The "carry" wires (which pass the "1" from one digit to the next) have fewer connections, but they are the only path for the calculation to finish correctly.
The Result: When the author looked at the shape, the "Degree" metric got it wrong (it thought the inputs were the most important). But the "Betweenness" metric got it right (it identified the carry wires as the critical logic).
The Metaphor: If you cut the input wires, the machine stops receiving data. But if you cut the "carry" wires, the machine stops thinking. The shape of the network revealed the "thinking" part, even without reading the code.

4. Testing Across Seven Different Worlds

The author didn't just look at one thing. They tested this "Shape = Logic" idea on seven very different things to see if it held up:

Digital Circuits: Found the critical "carry" paths.
Math Proofs (Lean & Coq): Found that the core mathematical libraries (like Algebra) act as the "bridges" holding the whole proof system together.
Old Banking Code (COBOL): Found that the specific programs that handle money transfers are the structural "hubs," even in messy, old code.
Animal Brains (C. elegans to Fruit Flies): Found that certain nerve cells act as "command centers" in both worms and flies, despite 600 million years of evolution. The "shape" of their connections stayed the same.
Prebiotic Chemistry: Found that in a model of early life chemistry, one specific molecule acts as the "hub" that makes the whole reaction work.

5. What This Means (and What It Doesn't)

What it DOES mean:
In all these systems, the "shape" of the connections (specifically the "Bridge Builder" metric) perfectly matches what human experts know is the most important part of the system. If you remove those "bridge" nodes, the system's logic collapses. The geometry is the operational logic.

What it DOES NOT mean:

It doesn't mean you can read a math proof just by looking at a graph. You still need to understand the symbols.
It doesn't mean the shape creates the meaning. The meaning comes from the rules; the shape just reveals where the rules are most critical.
It doesn't mean this works for everything yet. Some tests were small or needed more data (like the Coq math proof test), but the direction is clear.

The Bottom Line

The paper argues that structure reveals function. Just as you can tell a bridge is important by looking at where the weight is distributed, you can tell the "logic" of a computer, a brain, or a math library by looking at the "bridges" in their connection maps. The author calls this "Topology as Logic."

They pre-registered all their tests (like writing down their guesses before looking at the data) to prove they didn't just get lucky. The results suggest that across hardware, biology, and math, the "load-bearing" parts of a system always look the same geometrically.

Technical Summary: Topology as Logic

Problem Statement

The paper addresses a fundamental disconnect between two frameworks for understanding information-processing systems: classical logic, defined by symbolic manipulation and inference rules, and topology, defined by geometric positions and connections. While these appear categorically different, all information-processing systems (from proof assistants to neural networks) possess a dependency graph. The central research question is whether the geometry of this dependency graph—specifically its hub structure across multiple relational layers—can recover the "operational logic" of the system.

The authors define "operational logic" not as the reconstruction of symbolic semantics or theorem content, but as the identification of load-bearing structural organization: the specific nodes and pathways whose removal or perturbation would cause the system's logical organization to collapse.

Methodology

The study employs the IRDME (Inter-layer Relationship Discovery and Mapping Engine) framework and the Functional Proximity Law (FPL) to analyze seven independent substrates.

Experimental Protocol

Pre-registration: All seven experiments were pre-registered with explicit quantitative hypotheses before analysis. Hypotheses were hashed and archived in a public repository to ensure transparency and prevent p-hacking.
Multilayer Representation: Systems are modeled as flat arrays of relations typed by "layer" (e.g., physical wiring vs. simulation state, declared imports vs. enacted usage).
Metrics:
- Degree Hub Persistence: The correlation of node degrees ( $k$ ) across layers.
- Betweenness Hub Persistence: The correlation of betweenness centrality ( $b$ ) across layers.
- Rank Divergence: The analysis of how node rankings shift between layers to identify structural boundaries.
Validation: Statistical significance was determined using permutation-based tests (e.g., degree-preserving edge rewiring) to ensure findings were not artifacts of degree sequences.

Substrates Analyzed

The evidence base spans seven pre-registered experiments across diverse domains:

Digital Circuits: A 4-bit ripple carry adder (H_LOGIC_DIGITAL_v1) and the ISCAS85 c432 benchmark (H_LOGIC_F3_ISCAS85_v1).
Formal Mathematics: Lean 4 mathlib4 (H_LOGIC_EXTRACTION) and Coq Corelib (H_LOGIC_COQ_STDLIB_v1).
Legacy Software: COBOL banking code (F2).
Neural Connectomics: Cross-species analysis from C. elegans to Drosophila (F12c/F13).
Prebiotic Chemistry: A published autocatalytic network from CatReNet (X9 v2).

Key Results

1. The Load-Bearing Logic Signal (Betweenness vs. Degree)

The most significant methodological finding is the divergence between connection-density logic (degree) and load-bearing logic (betweenness).

Digital Circuit Evidence: In the 4-bit adder, degree-based hub persistence between physical wiring and state-correlation layers was weak ( $r \approx 0$ ). Input nodes had high fan-in but low logical weight. Conversely, betweenness-based persistence was strong ( $r = 0.771, p = 0.004$ ). The "carry chain" nodes, which are critical for global arithmetic propagation, dominated betweenness in both layers despite having moderate degree.
Null Validation: Degree-preserving edge rewiring confirmed that degree correlation is an artifact of degree sequences, whereas betweenness persistence ( $r = 0.762$ ) remained significantly above the null distribution ( $p < 0.001$ ), proving it captures specific edge structural information.
XOR Divergence: XOR gates ranked high in physical wiring but low in state-correlation, demonstrating that topology can identify "logic gates without logical weight" in global computation.

2. Cross-Substrate Replication

Formal Mathematics:
- Lean 4: Confirmed strong betweenness-based role geometry ( $r = 0.777, p = 0.004$ ), identifying Algebra and Analysis as dual hubs.
- Coq: Provided a directional replication ( $r = 0.49$ ) but was statistically underpowered ( $n=17, p=0.29$ ).
Digital Circuits (Scale Replication): The ISCAS85 c432 benchmark confirmed the pre-registered degree hypothesis ( $r = 0.426, p = 0.002$ ). The betweenness finding ( $r = 0.34$ ) was exploratory but consistent with the 4-bit results.
Other Substrates:
- Neural: Showed conserved hub grammar across 600 million years of evolution, with a Spearman-Pearson discrepancy suggesting rank-order conservation (betweenness-like) rather than absolute connection count.
- COBOL & Prebiotic: Identified conserved hub structures in banking code and autocatalytic networks, though sample sizes were small ( $n=14$ and $n=13$ respectively).

Limitations

The paper explicitly acknowledges several constraints:

Sample Size: The Coq ( $n=17$ ), COBOL ( $n=14$ ), and prebiotic ( $n=13$ ) datasets are small, limiting permutation test power. Findings are presented as "pattern-establishing" rather than definitive.
Data Nature: The COBOL and prebiotic datasets are representative structural models, not raw scraped data from single codebases or wet-lab measurements.
Protocol Gaps: In the ISCAS85 c432 experiment, the pre-registered hypothesis tested degree, not betweenness; the betweenness result is exploratory. Similarly, the 4-bit betweenness result was post-hoc.
Independence: The two formal mathematics experiments (Lean and Coq) are independent corpora, but the two Lean experiments (F9 and H_LOGIC_EXTRACTION) use the same corpus.

Significance and Claims

The paper makes a modest, research-program-level claim: that dependency topology encodes operational logic as geometry. It does not claim that symbolic semantics can be reconstructed from graphs or that theorem content is inferable from hub structure alone.

Core Contributions:

Operational Logic Geometry: In dependency-structured systems, hub geometry (specifically betweenness persistence and rank divergence) reliably identifies the "load-bearing" nodes that domain experts recognize as the system's operational logic (e.g., the carry chain in circuits, command interneurons in biology, core modules in math libraries).
Metric Distinction: The paper establishes that betweenness centrality is the correct metric for detecting load-bearing logic, whereas degree often measures connection density that can be misleading.
Cross-Domain Validity: The convergence of topology-identified hubs with expert-designated operational logic across hardware, formal mathematics, legacy software, neuroscience, and prebiotic chemistry suggests a substrate-independent structural principle.

The authors conclude that "topology as operational logic geometry" is an empirically justified research program warranting further investigation, particularly with larger formal corpora and broader ISCAS85 benchmarks, but it is not yet a universal theorem.