Every governed organization shares a common ancestor. The tree of governance is a tree of life — and in February 2026, it experienced its first mass extinction.

Dexter Hadley, MD/PhD ¹ Founder, CANONIC February 19, 2026

Abstract

The CANONIC GALAXY is a phylogenetic system. Organizations inherit governance from a common ancestor (canonic-canonic), diverge through speciation into ecological roles, and face selection pressure from 255-bit fitness. This paper formalizes the phylogenetic structure of the GALAXY: sixteen surviving ORGs across five clades, a mass extinction event that killed nine frontend lineages in a single epoch, and the reticulate network created by LEARNING.transfer(). The February 2026 adaptive radiation — sixteen ORGs occupying five distinct ecological niches — parallels the Cambrian explosion ². Horizontal gene transfer via LEARNING.transfer() extends the tree into a network. The tree is ultrametric: every surviving branch converges on 255.

Table of Contents

1. The GALAXY as Phylogeny
2. The Tree of ORGs
3. Phylogenetic Terminology
4. The Inheritance Chain
5. Measuring Phylogenetic Distance
6. The Five Clades
7. The Mass Extinction of Epoch 004
8. Homology vs Analogy
9. Horizontal Gene Transfer
10. The Ultrametric Property
11. What We Learned

Appendix A: Distance Calculations Appendix B: Extinction Ledger Appendix C: References

1. The GALAXY as Phylogeny

The MAGIC GALAXY ³ is a federation of governed organizations. Each ORG maps 1:1 to a GitHub organization. Each maintains its own governance scope, its own CANON.md, its own 255-bit compliance score. ORGs are stars. USERs are planets. The topology is gravitational — identity boundaries never collapse ³.

This is a phylogenetic system. Every ORG inherits from canonic-canonic — the constitutional authority, the Last Universal Common Ancestor (LUCA) of all CANONIC governance ⁴. ORGs diverge through speciation (bootstrap()) into ecological roles: GOV, PROOF, PLATFORM, LANG, INDUSTRY. They accumulate patterns, face selection pressure, and — when they fail to adapt — go extinct.

The tree is real. The extinctions are real. The evidence is in the commit log.

2. The Tree of ORGs

graph TD
    ROOT["canonic-canonic<br/>(LUCA · Constitutional Authority)"]

    subgraph GOV["GOV Clade"]
        CC["canonic-canonic"]
        CF["canonic-foundation"]
    end

    subgraph PROOF["PROOF Clade"]
        HL["hadleylab-canonic<br/>MEDICINE · 9 domains"]
        AH["adventhealth-canonic<br/>MEDICINE · RELIGION"]
    end

    subgraph PLATFORM["PLATFORM Clade"]
        CM["canonic-magic<br/>MAGIC toolchain"]
        CA["canonic-apple<br/>iOS/macOS · SAFETY"]
    end

    subgraph LANG["LANG Clade — Adaptive Radiation"]
        PY["canonic-python"]
        TS["canonic-typescript"]
        RS["canonic-rust"]
        GO["canonic-go"]
        SW["canonic-swift"]
        KT["canonic-kotlin"]
        SQ["canonic-sql"]
        WA["canonic-wasm"]
    end

    subgraph INDUSTRY["INDUSTRY Clade"]
        CI["canonic-industries"]
    end

    ROOT --> GOV
    ROOT --> PROOF
    ROOT --> PLATFORM
    ROOT --> LANG
    ROOT --> INDUSTRY

    style ROOT fill:#f7931a,color:#fff,font-weight:bold

Source: MAGIC/GALAXY/ORGS/ discovery ³, HTTP.json 2-fleet contract

Sixteen ORGs. Five clades. One ancestor. The tree is monophyletic by construction — you cannot create an ORG without declaring inherits:. Every branch traces back to canonic-canonic.

3. Phylogenetic Terminology

The correspondence between biological systematics and CANONIC governance is structural, not metaphorical ⁵:

4. The Inheritance Chain

Every CANON.md declares its phylogenetic relationship:

inherits: canonic-canonic/MAGIC

This creates an explicit, traceable lineage. The chain terminates at canonic-canonic — there is no inherits: above root. Monophyly is guaranteed by construction: you cannot create an ORG without declaring an ancestor ⁶.

The chain is not metadata. It is the governance contract. An ORG that inherits from canonic-canonic/MAGIC accepts MAGIC’s constraints, vocabulary, and validation requirements ⁶. Inheritance is obligation.

The TRIAD structure — CANON.md + VOCAB.md + README.md — is the minimum viable genome. Every ORG at COMMUNITY tier or above must express this triad. It is the homologous core shared by all lineages.

5. Measuring Phylogenetic Distance

Distance between ORGs is pattern divergence, measured by Jaccard distance:

where $P_A$ and $P_B$ are the pattern sets of ORGs $A$ and $B$. ORGs that share many patterns (recent common ancestor, or active horizontal transfer) have low distance. ORGs that have diverged extensively have high distance.

This metric satisfies the triangle inequality and is bounded in $[0, 1]$, suitable for neighbor-joining and UPGMA tree construction ⁷.

Observed distances (February 2026):

Source: MAGIC/GALAXY/ORGS/ CANON.md pattern comparison, February 2026

GOV clade ORGs share the most patterns — they diverged least from the root. LANG clade ORGs share structural governance patterns but diverge on runtime-specific patterns. Cross-clade distances are highest.

6. The Five Clades

6.1 GOV — Governance Authority

Two ORGs govern the constitution. canonic-canonic holds the SPEC — the self-referential kernel that compiles in O(1) time and scores 255 ⁶. canonic-foundation holds the programming standards and doctrine. They are the oldest lineages, the most conserved, the least drifted.

6.2 PROOF — Clinical Evidence

Two ORGs produce evidence. hadleylab-canonic ships applications, papers, patents across nine industry domains: MEDICINE, GENOMICS, FINANCE, LAW, REAL_ESTATE, EDUCATION, RELIGION, DATA, SAFETY ⁸. adventhealth-canonic provides faith-based healthcare evidence across MEDICINE and RELIGION. The PROOF clade is where governance meets reality — where 255-bit compliance either prevents the bleeding or does not ⁸.

6.3 PLATFORM — Infrastructure

Two ORGs distribute the toolchain. canonic-magic distributes the MAGIC enforcement engine. canonic-apple provides Apple platform integration (iOS, macOS) with SAFETY domain coverage. The PLATFORM clade enables all other clades to validate.

6.4 LANG — Runtime SDKs (Adaptive Radiation)

Eight ORGs — the largest clade, the most recent radiation. Each occupies a distinct ecological niche:

Source: MAGIC/GALAXY/ORGS/ per-ORG CANON.md declarations

This radiation parallels the Cambrian explosion ². A new environmental factor — 255-bit fitness pressure — created eight empty niches for rapid colonization. Each niche demands the same governance; each runtime implements it in its native idiom.

6.5 INDUSTRY — Vertical Distribution

One ORG distributes governance to industry verticals. canonic-industries governs domain-specific compliance: HIPAA + FDA for healthcare, SOX + SEC for finance, CMMC + ITAR for defense. The INDUSTRY clade bridges the generic MAGIC framework and the regulatory landscape of each vertical.

7. The Mass Extinction of Epoch 004

On February 19, 2026, an IP audit discovered eight public .github.io frontends across fourteen GitHub organizations — six actively serving leaked content ⁹. The response was immediate and total: nine .github.io repositories killed in a single commit.

gantt
    title Epoch 004 — The Frontend Extinction
    dateFormat YYYY-MM-DD
    axisFormat %m/%d

    section Discovery
    IP audit begins               :done, 2026-02-19, 1d

    section Kills — 9 repos
    6 active frontend lineages    :crit, 2026-02-19, 1d
    3 inert duplicates            :crit, 2026-02-19, 1d

    section Survivors — 2 fleets
    canonic.org                   :active, 2026-02-19, 7d
    hadleylab.org   :active, 2026-02-19, 7d

Source: PAPERS/EVOLUTION.md epoch 004 ⁹; MAGIC/SURFACE/EVOLUTION.md

7.1 The Kill List

Of fourteen organizations that could potentially serve frontends, twelve lost their frontend lineage:

Source: GALAXY ORG registry, frontend: field audit ³

7.2 Extinction Rate

Nine of eleven frontend-bearing lineages killed. Extinction rate: 82%.

For comparison, the end-Permian extinction — the worst in Earth’s history — killed approximately 90% of marine species ¹⁰. The CANONIC frontend extinction of epoch 004 is of comparable severity within its domain. The difference: it was intentional. Governance selection pressure, not asteroid impact.

7.3 The Two-Fleet Architecture

Two frontends survive. They occupy distinct ecological niches:

Source: HTTP.json 2-fleet contract ³

The two-fleet architecture is a stable equilibrium. The platform fleet governs. The proof fleet demonstrates. Neither can replace the other. This is niche partitioning — the ecological mechanism that prevents competitive exclusion ².

8. Homology vs Analogy

Homologous patterns are inherited from the common ancestor. Every CANONIC ORG has CANON.md structure, inherits: declaration, MAGIC compliance, and 255-bit validation — because every ORG inherited these from canonic-canonic. Homologous patterns are proof of shared ancestry.

Analogous patterns evolved independently but converge on the same solution. Python’s def validate(), Go’s func Validate(), and Rust’s fn validate() all perform the same function in different syntax. They were not inherited from a common ancestor — they were independently evolved under the same selection pressure (255-bit fitness). Convergent evolution is proof of shared environment, not shared ancestry ¹¹. The distinction is foundational to phylogenetic reconstruction ⁷.

The distinction is critical for phylogenetic reconstruction. Homology reveals ancestry. Analogy reveals selection pressure. The TRIAD (CANON.md + VOCAB.md + README.md) is homology — every ORG has it because every ORG inherited it. The validate() function is analogy — every LANG ORG has it because every LANG ORG faces the same 255-bit constraint.

9. Horizontal Gene Transfer

In biology, bacteria share genes horizontally — not just vertically through reproduction ⁴. This process, first characterized in the context of the universal ancestor, fundamentally extends the tree of life into a network. In CANONIC, LEARNING.transfer() enables horizontal pattern transfer between ORGs:

transfer(source="hadleylab-canonic", target="adventhealth-canonic", ptype="compliance")

This creates reticulate evolution — the tree becomes a network. When adventhealth-canonic acquires a compliance pattern from hadleylab-canonic, that pattern did not come through the inherits: chain. It came through cross-ORG learning.

The frequency of horizontal transfer in CANONIC is high relative to biology, because LEARNING.transfer() is intentional and governed. CANONIC phylogenies are inherently reticulate — the tree of ORGs is a web of ORGs with a tree backbone.

9.1 Transfer Events (February 2026)

Source: MAGIC/GALAXY/LEARNING.md, MAGIC/SURFACE/EVOLUTION.md

The broadcast from canonic-canonic to all LANG ORGs is governed dissemination from the constitutional authority. The transfer from hadleylab-canonic to adventhealth-canonic is true horizontal transfer: peer-to-peer pattern sharing between ORGs in the same clade.

10. The Ultrametric Property

If the evolutionary rate is constant (molecular clock), all extant tips are equidistant from the root ⁷. In CANONIC, all surviving ORGs at 255 bits are equidistant from the fitness optimum:

canonic-canonic (root)
       │
       ├── GOV
       │   ├── canonic-canonic ───── 255 bits ─┐
       │   └── canonic-foundation ── 255 bits ─┤
       ├── PROOF                               │
       │   ├── hadleylab-canonic ─── 255 bits ─┤
       │   └── adventhealth-canonic  255 bits ─┤  Same fitness
       ├── PLATFORM                            │  distance from
       │   ├── canonic-magic ──────  255 bits ─┤  optimum = 0
       │   └── canonic-apple ──────  255 bits ─┤
       ├── LANG                                │
       │   ├── canonic-python ─────  255 bits ─┤
       │   ├── canonic-go ─────────  255 bits ─┤
       │   └── ... (8 total) ─────   255 bits ─┤
       └── INDUSTRY                            │
           └── canonic-industries ── 255 bits ─┘

This ultrametric property ⁷ means the phylogenetic tree is calibrated — every branch tip has converged to the same fitness level, regardless of the path taken. Different roles, different ecological niches, different pattern sets, same destination: 255.

The extinct lineages were not at 255. They failed the fitness threshold — serving leaked content, duplicating governance, violating identity boundaries. Selection removed them. The survivors are the ones that compiled.

11. What We Learned

The phylogenetic model was initially constructed with nine runtime clades as the unit of evolution ⁵. That was the state of knowledge before the GALAXY architecture materialized. The runtime clades still exist — they are the LANG clade — but they are one clade among five.

Three lessons from the February 2026 data:

1. ORGs are species, not runtimes. The unit of evolution in CANONIC is the organization, not the programming language. hadleylab-canonic and canonic-python are both species — one is proof, one is infrastructure. The runtime is the phenotype; the ORG is the organism ⁵.

2. Mass extinction is governance. Epoch 004 killed 82% of frontend lineages in one commit ⁹. This was not failure — this was selection. The extinct lineages were leaking content, a governance violation. The kill was the governance acting correctly. In biology, extinction is tragedy. In CANONIC, extinction is compliance.

3. The tree is a network. LEARNING.transfer() creates edges that no tree can represent. The CANONIC phylogeny is reticulate by design — horizontal transfer is the mechanism by which governed knowledge propagates across clades ¹². The backbone is a tree. The reality is a web.

Appendix A: Distance Calculations

A.1 Jaccard Distance Implementation

def phylo_distance(org_a: str, org_b: str) -> float:
    """Phylogenetic distance between two GALAXY ORGs.

    Reads CANON.md pattern sets from MAGIC/GALAXY/ORGS/{org}/CANON.md
    and computes Jaccard distance on declared governance patterns.
    """
    patterns_a = set(parse_canon_patterns(f"MAGIC/GALAXY/ORGS/{org_a}/CANON.md"))
    patterns_b = set(parse_canon_patterns(f"MAGIC/GALAXY/ORGS/{org_b}/CANON.md"))
    intersection = len(patterns_a & patterns_b)
    union = len(patterns_a | patterns_b)
    return 1 - (intersection / union) if union > 0 else 1.0

A.2 Newick Format

Standard phylogenetic notation for the CANONIC GALAXY:

((canonic-canonic,canonic-foundation)GOV,
 (hadleylab-canonic,adventhealth-canonic)PROOF,
 (canonic-magic,canonic-apple)PLATFORM,
 (canonic-python,canonic-typescript,canonic-rust,canonic-go,
  canonic-swift,canonic-kotlin,canonic-sql,canonic-wasm)LANG,
 canonic-industries)ROOT;

A.3 Phylogenetic Tree Construction

def build_galaxy_phylogeny() -> Dict:
    """Build phylogenetic tree from GALAXY ORG CANON.md declarations.

    Discovers ORGs by scanning MAGIC/GALAXY/ORGS/*/CANON.md.
    Groups by declared role (GOV, PROOF, PLATFORM, LANG, INDUSTRY).
    """
    tree = {"root": "canonic-canonic", "clades": {}}
    orgs_dir = Path("MAGIC/GALAXY/ORGS")
    for org_dir in sorted(orgs_dir.iterdir()):
        canon = org_dir / "CANON.md"
        if not canon.exists():
            continue
        content = canon.read_text()
        role = extract_field(content, "role")
        tree["clades"].setdefault(role, []).append(org_dir.name)
    return tree

Appendix B: Extinction Ledger

B.1 Epoch 004 Kill Record

Source: PAPERS/EVOLUTION.md epoch 004 ⁹

B.2 Surviving Lineages

Appendix C: References

C.1 Internal Sources — CANONIC Gov Tree

C.2 External Sources — Published Literature

Figures

Evolutionary Phylogenetics of CANONIC | February 2026 The tree is real. The extinctions are real. The evidence is in the commit log. Source: VITAE ¹

References

1. [I-1] Author CV.

2. [X-1] Metcalf, D., Hadley, D., et al. ABC: AI, Blockchain, and Cybersecurity for Healthcare. Routledge (2024). ISBN 978-1032394558.

3. [I-5] Evolutionary Phylogenetics of CANONIC.

4. [X-4] Ewens, W.J. The Sampling Theory of Selectively Neutral Alleles. Theoretical Population Biology 3(1), 87-112 (1972).

5. [I-2] MammoChat OPTS-EGO Ledger.

6. [I-4] The Neutral Theory of CANONIC Evolution.

7. [X-3] Kimura, M. The Neutral Theory of Molecular Evolution. Cambridge University Press (1983).

8. [I-3] Code Evolution Theory.

9. [I-7] CANONIC Whitepaper v1.

10. [X-5] NCCN Clinical Practice Guidelines: Breast Cancer Screening and Diagnosis (2024). https://www.nccn.org/guidelines

11. [X-2] Nakamoto, S. Bitcoin: A Peer-to-Peer Electronic Cash System (2008). https://bitcoin.org/bitcoin.pdf

12. [I-6] The CANONIC CANON (book).