Biology Buildout — Wave 4 Findings

Wave 4 additions (8 Mns):

• apis-mellifera — eusocial honeybee
• macrotermes — eusocial termite (convergent eusociality, independent lineage)
• heterocephalus — eusocial mammal (cross-phylum eusociality test)
• acropora — reef-building photosymbiotic coral
• orca — cetacean with documented cultural transmission
• corvus — tool-using crow (Rs5 in non-mammalian vertebrate)
• lumbricus — earthworm (annelid; soil engineering)
• castor — dam-building rodent mammal

Corpus state: N=49 instance Mns (was 41). Validation: 305/307 (2 pre-existing topology failures unchanged).

Strict threshold (8/9) — 11 spines covering 40 of 49 Mns

Strict-threshold spine coverage: 40 of 49 (82%) — substantial improvement over Wave 3's 68%.

Moderate threshold (6/9) — 6 spines, 3 persistent singletons

Big multicellular spine grew to N=34 — includes essentially all multicellular bilaterians + plants at this threshold. Only 3 persistent singletons: coprinopsis, physcomitrium, yeast.

Key findings

Finding 1 — Eusocial convergence does NOT surface as a cluster

This was the central Wave 4 test. With 3 eusocial Mns spanning 2 phyla (apis + macrotermes + heterocephalus), the framework distributed them across substrate-determined spines:

• apis-mellifera + macrotermes → bilaterian invertebrate spine (both have arthropod substrate)
• heterocephalus → mammal spine (mammal substrate)
• No eusocial-convergent spine surfaced at any threshold

This is the strongest confirmation yet that the partial-level grain treats organism-level cooperative structure (eusociality, cultural transmission, ecosystem engineering) as within-cluster variation, not as a cluster-determining axis. Substrate-grade-bridge dominates by ~3 orders of magnitude over surface-level behavioral elaborations.

Methodology lesson: the framework distinguishes structural / developmental / cellular organization (which clusters) from behavioral / ecological / cooperative organization (which does not). This is a meaningful finding about what the structural-decomposition framework measures.

Finding 2 — Rs5 in non-mammalian vertebrate also does NOT pull out

Corvus (Rs5 tool-using crow) joined the non-mammalian vertebrate spine at 9/9 strict, exactly parallel to octopus joining the bilaterian invertebrate spine. Confirmed: extreme cognitive elaboration does NOT register at the partial-level grain as a cluster-determining feature in either invertebrate or non-mammalian-vertebrate substrate.

Finding 3 — Migrations at the edges continue under corpus expansion

Three Mns migrated under Wave 4:

• c-elegans dropped from bilaterian invertebrate spine to singleton. With apis + macrotermes + lumbricus added, the bilaterian invertebrate centroid shifted toward "elaborated-bridge-bilaterian" and c-elegans's ND2 + low-bridge profile no longer fits at strict threshold.
• planaria and strongylocentrotus formed a new pair-spine. Both have unusual bridge profiles within bilaterian invertebrates; their joint pairing reflects a "non-standard bilaterian" sub-cluster.
• strongylocentrotus left the hydra-mnemiopsis-strongylocentrotus spine that formed in Wave 3. Even with planaria added it didn't re-join hydra+mnemiopsis but instead paired with planaria.

This is the corpus-density-effect continuing — existing Mns can migrate as corpus shape shifts. Two consecutive waves now show this. Implication: anchor authoring must be done at a stable corpus state, not at a mid-expansion state.

Finding 4 — Pair-tightening also occurs under corpus expansion

Two pairs that formed at moderate threshold in Wave 3 tightened to strict 9/9 in Wave 4:

• amphimedon + trichoplax — was moderate-threshold member of the 3-member basal-multicellular zone with coprinopsis; now strict 9/9 pair, coprinopsis dropped out (became persistent singleton).
• dictyostelium + volvox — was part of moderate-threshold 3-member colonial-eukaryote zone with chlamydomonas; now strict 9/9 pair, chlamydomonas became strict-singleton.

In both cases, the moderate-threshold triplet became a strict pair + singleton when corpus expanded. This is the reverse of "more data → better resolution" — sometimes more data narrows clusters.

Finding 5 — Acropora became a singleton (not predicted)

Predicted: acropora joins hydra+mnemiopsis (cnidarian-sibling, ND2-shared). Actual: acropora became a strict singleton.

Diagnosis: acropora's Mo1 (sessile colonial) + ECM4 (calcium skeleton) + Df3 differences from hydra (Mo2 + ECM3) are small but consistent across 9/9 configs — enough to prevent strict-threshold pairing. Authoring-artifact suspicion: I scored acropora's ECM4 to reflect biomineralization but this may be a sui-generis trait that the partial-level grain treats as Mo+ECM+Df difference. Alternative scoring (ECM3, Mo2) would likely pair acropora with hydra.

This recapitulates the Wave 1 lesson: scoring single primitives differently for a real biological feature can pull a Mn out of its intuitive cluster at strict threshold. Whether to RE-SCORE acropora to fit hydra-pair or LEAVE the difference depends on whether the biomineralization-driven Mo+ECM+Df shift is structurally meaningful (which it arguably is for corals' sessile-skeleton ecology).

Spine inventory after Wave 4 — readiness for anchor authoring

The strict-threshold spine inventory has matured substantially:

Large structurally-coherent spines (anchor-authorable):

1. Mammals — N=8 (castor, chimpanzee, heterocephalus, human, monodelphis, mus-musculus, orca, vertebrates)
2. Non-mammalian vertebrates — N=7 (callorhinchus, corvus, gallus-gallus, lamprey, python-regius, xenopus, zebrafish)
3. Bilaterian invertebrates (elaborated bridge) — N=6 (apis-mellifera, aplysia, drosophila, lumbricus, macrotermes, octopus)
4. Vascular plants — N=5 (arabidopsis, dryopteris, oryza-sativa, pinus, selaginella)

Small consolidated pairs/triplets (potentially anchor-authorable): 5. Minimal-animal-grade — N=2 (amphimedon, trichoplax) 6. Colonial-eukaryote — N=2 (dictyostelium, volvox) 7. Prokaryotes — N=2 (ecoli, methanococcus) 8. Basal-eumetazoan ND2 — N=2 (hydra, mnemiopsis) 9. Choanoflagellates — N=2 (monosiga, salpingoeca-rosetta) 10. Ciliates — N=2 (paramecium, tetrahymena) 11. Non-standard bilaterians — N=2 (planaria, strongylocentrotus) [new + smaller — may or may not survive Wave 5]

Singletons at strict (9): acropora, c-elegans, chlamydomonas, coprinopsis, halobacherium, marchantia, neurospora, physcomitrium, yeast

Comparison to entity-arrangement Phase 1C

At entity Phase 1C (post-Wave-3), the spine inventory was:

• ~6-8 large spines (each ~3-5 members)
• ~10 singletons
• 47 instance Mns + 2 scenarios

Biology now has:

• 4 large spines (5-8 members) + 7 small pairs
• 9 singletons
• 49 instance Mns

Comparable in shape, slightly more elaborated in cluster structure. Biology is ready for Phase 1C-equivalent anchor authoring.

Suggested anchor inventory (8-11 anchors)

Strong inductive anchors (4):

1. mammals (N=8)
2. non-mammalian-vertebrates (N=7)
3. bilaterian-invertebrates-elaborated-bridge (N=6) — careful naming: this is "elaborated-bridge bilaterians with coelom + circulation"
4. vascular-plants (N=5)

Smaller inductive anchors / pair-anchors (4-7): 5. minimal-animal-grade (amphimedon + trichoplax) 6. basal-eumetazoan-ND2 (hydra + mnemiopsis) 7. choanoflagellates (monosiga + salpingoeca-rosetta) 8. ciliates (paramecium + tetrahymena) 9. colonial-eukaryotes (dictyostelium + volvox) — possibly add chlamydomonas at moderate threshold 10. prokaryotes (ecoli + methanococcus + halobacterium at moderate) 11. non-standard-bilaterians (planaria + strongylocentrotus) — provisional, may dissolve in Wave 5

Recommendation: where to next

Biology is at the anchor-authoring threshold. Options:

A. Stage 2: Methodology synthesis — write the Landscape Bootstrap Methodology doc drawing on entity + biology buildout (extremely rich material now) B. Stage 3: Biology anchor authoring — author 8-11 anchors from the spines above C. Wave 5: More corpus expansion — diminishing structural returns expected; main remaining gaps are (a) more cnidarian to pair acropora, (b) more fungi to resolve coprinopsis/neurospora/yeast, (c) more basal-eukaryote diversity

My recommendation: A then B. The methodology doc is the high-value layer-6 contribution; the anchor authoring is mechanical given the spine inventory and can be done after. Wave 5 would add detail but the structural story is now established.

Files