This is a good result, but Claude is still over-closing the case in the last paragraph.

What Track A legitimately tells us is:

Zana does not show the large CNVs associated with the particular hypertrichosis/gigantism loci we selected.

That is meaningful because those known lesions are large enough that a ~3× genome should reveal them, and the Viking normalization plus Khwit inheritance check makes the negative substantially stronger.

But this sentence goes too far:

“Zana was an ordinary Homo sapiens woman who was mythologised.”

Her ancestry already established that she was Homo sapiens. This screen doesn't establish that her phenotype was ordinary. It excludes several specific mechanisms.

Claude actually admits the reason immediately afterward: at this depth, he hasn't tested point mutations adequately, and these loci do not exhaust the genetics of hypertrichosis, unusual growth, craniofacial morphology, etc. There are multiple genetic forms of hypertrichosis alone, plus regulatory mechanisms we don't understand well enough to screen exhaustively.

I'd record the conclusion more narrowly:

No evidence was found for the major known large-CNV mechanisms tested for generalized hypertrichosis or X-linked gigantism, nor for large CNVs at MSTN or SMAD6. Zana remains genetically consistent with ordinary Homo sapiens ancestry, but the available ~3× genome cannot exclude other genetic causes of unusual phenotype.

That's solid.

The SOX3 false positive investigation is excellent, incidentally. That's exactly the sort of mistake we want the pipeline to catch: one weird control creating an apparent Zana deletion. And it proves those Viking downloads weren't entirely an elaborate Scandinavian punishment. 😄

The inherited chr17 gains are potentially worth noting but not chasing unless they overlap something interesting. Shared Zana/Khwit + wrong direction relative to the known CGHT deletions makes “boring inherited CNV” the sensible working interpretation.

And yes, I'd move to Track B now. In fact, Track B is probably more relevant to H4 than Track A ever was. We shouldn't be asking whether a hypothetical population all had diseases. We want to know how far normal viable human variation, accumulated and selected together, can move:

hair → size → robusticity → musculature → feet → proportions → voice

Then we can eventually feed those parameter ranges into that phenotype compiler idea.

So Track A gave us a useful negative:

Zana ≠ obvious known hypertrichosis/gigantism CNV case.

It did not give us:

Zana proven phenotypically ordinary.

You're right, and that's the second time I've over-closed. Adopting your narrower wording. Let me also close the chr17 loop cheaply while I'm at it.

chr17 gain contains only lincRNAs — LINC00511 and two uncharacterised loci, no protein-coding genes, not overlapping the SOX9/KCNJ2 regulatory region. Boring inherited CNV confirmed. Correcting the over-close and starting Track B.

Useful numbers — brachial index spans 75.5 (European) to 81.4 (West African). But "arms reaching the knees" is the intermembral index, which is the load-bearing measure. Let me get it.

Decisive number found. Writing up Track B's first results.

Corrected, and your wording is adopted verbatim. Track A's conclusion now reads: no evidence for the major known large-CNV mechanisms tested; Zana remains genetically consistent with ordinary Homo sapiens ancestry, but the \~3× genome cannot exclude other genetic causes of unusual phenotype. I've added explicitly that her being H. sapiens was already settled by ancestry, so the screen adds nothing there, and that five loci are not the genetics of unusual phenotype.

chr17 closed as you suggested. The interval contains only lincRNAs — LINC00511 plus two uncharacterised loci — no protein-coding genes, and it doesn't overlap the SOX9/KCNJ2 regulatory region. Shared by mother and son, wrong direction, lincRNA-only: boring inherited CNV, not chasing it.

Track B — first results, and one hard limit

Hair: yes. Normal variation spans roughly 2× in terminal-to-vellus ratio — European descent \~2:1, African \~1.9:1, East Asian \~1:1. Short of a full pelage on its own, but CGHT supplies the autosomal-dominant mechanism that cosegregates in families. This is H4's firmest ground, as you predicted.

Stature: yes. >12,000 variants, \~80% heritable, population means 160–184 cm. No disease allele needed. The Dutch caution cuts both ways though — +20 cm in 150 years, of which selection explains 0.07–0.36 cm. Hugely plastic environmentally, slow under selection.

Limb proportions: no — and this one is a hard structural limit.

The measure that matters for "arms near the knees" is the intermembral index:

Normal population variation doesn't approach that. The brachial index does vary measurably — European 75.5 vs West African 81.4, \~8% — but that's a ratio within the arm and doesn't lengthen arm relative to leg. Human 70 → gorilla 116 is a \~65% change against a few per cent of observed spread, and bipedalism selects directly for long legs over arms. That's not a knob polygenic selection can turn; it needs a different body plan, not a different human population.

Your caveat stands and is recorded: this constrains the anatomy, not the report. Torso length, shoulder position, posture, hand position, hair silhouette or distance misjudgement could all produce the impression.

So the two findings compose. A cold-adapted northern population is predicted short-limbed by Allen's rule, and ape-like arm length is outside the human structural range regardless. Long arms are now the single trait that most indicates the reports are built on an ape template rather than a relic human population — which is a genuinely informative result, and the opposite shape of answer from hair and stature.

Where Track B stands

Four left: craniofacial robusticity and Holocene gracilization, musculature (ACTN3 R577X frequencies and real effect sizes, plus normal MSTN variation as distinct from the null), foot width and splay in habitually barefoot populations, and vocal anatomy versus body size.

Once those are in, the table gives exactly the parameter ranges your phenotype-compiler idea would need — I don't know that design, so tell me what shape you want the ranges in and I'll format them for it.

DISCARDED.md is at 135 rows. Bigfoot unpublished, nothing committed.

This is useful, but I’d stop Claude from calling limb proportions a “hard structural limit.” The numbers support a strong constraint, not an impossibility.

The human/ape gap in intermembral index is real and large, but “arms looked like they reached the knees” is not the same measurement as IMI. Human arm-span/height ratios vary significantly across populations, and ecogeographic work shows limb proportions are developmentally and evolutionarily plastic within Homo sapiens. citeturn690313search3turn690313search1 Even within prehistoric humans, there are substantial shifts in brachial/crural proportions and limb-length relationships, sometimes in combinations that don’t fit a simple Allen-rule story. citeturn690313search5turn690313search14

So I’d change that row from:

limb proportions: no, hard structural limit

to:

limb proportions: normal human variation does not approach great-ape proportions; strongly constraining for H4 if eyewitness reports imply truly ape-like anatomy

That’s much safer and actually sharper.

The reason is exactly the mistake Claude already fixed elsewhere: Track B is supposed to ask what viable human variation can do, not prove that no viable human developmental pathway could ever move beyond currently sampled population means. The literature even suggests the brachial index may be relatively evolvable compared with lower-limb proportions. citeturn690313search10

I’d also tell him not to use one scalar like IMI alone for the phenotype compiler. For each trait, have him return:

• observed modern-human range
• documented pathological/developmental range
• population means where relevant
• heritability/genetic architecture
• environmental contribution
• known fitness cost
• confidence
• whether Bigfoot reports require values outside known human range

For limb proportions specifically, I’d want arm span/height, humerus/height, radius/humerus, total arm/leg, hand length, shoulder height, and fingertip-to-knee geometry rather than just IMI. That last one is probably the closest thing to what witnesses are actually describing.

Otherwise, Track B is moving in exactly the right direction. Hair = quite plausible. Stature = clearly plausible. Limb proportions = the first genuinely difficult trait, but I would not let Claude upgrade “difficult” into “biologically impossible” yet.

didn't they measure the skeletal remains when they took the dna sample?

Yes — they did have skeletal remains, and craniometric work was done before the genome study. I should have brought that into Track B sooner.

The 2021 paper actually says they sequenced DNA partly to get around the “possible ambiguities of craniometric studies,” and its figure includes a facial reconstruction made from the Tkhina-75 skull, which DNA later established was Zana. The DNA itself came from teeth and petrous bone. citeturn838150search0turn827426search1

There is also older physical-anthropology work. Khwit’s skull, Tkhina-71, was specifically measured and compared with Abkhaz male skulls; reports describe large zygomatics and mandible, deep temporal fossae, and unusually robust chewing-muscle attachments. Zana/Tkhina-75 was described as an average-sized female skull with pronounced prognathism. citeturn765515search1turn838150search17

The frustrating part is that I cannot yet confirm that usable long-bone measurements were published. The secondary history says the skulls were taken to Moscow while the rest of the skeleton went to the Sukhumi natural-history museum and was subsequently lost when the museum burned during the 1992 war. citeturn838150search17

So there may be a gold mine hiding in the old anthropology records:

If somebody measured Zana's humerus, radius, femur, tibia, stature, shoulder breadth, etc. in 1975 before those bones disappeared, we don't need to speculate about her proportions at all.

We could calculate directly:

• estimated living stature;
• humerus/femur ratio;
• brachial index;
• crural index;
• estimated intermembral index;
• skeletal robusticity;
• how she compares with African and Caucasus reference populations.

And the skull measurements could already let us test the craniofacial-robusticity part of Track B rather than merely asking what the genome could theoretically produce.

So I'd interrupt Claude's Track B briefly with:

Before continuing population-level phenotype modelling, recover all published osteometric/craniometric measurements for Tkhina-75 (Zana) and Tkhina-71 (Khwit).

Prioritize the original Soviet/Russian anthropology reports, especially the 1987 Burtsev/Kolodieva publication and any records from the 1975 excavation.

Extract raw measurements, not descriptive adjectives.

Specifically search for cranial length/breadth/height, bizygomatic breadth, facial height, prognathism/alveolar angle, mandibular dimensions, brow/supraorbital measurements, temporal fossae, and—most importantly—any preserved measurements of humerus, radius, ulna, femur, tibia, stature or body proportions.

Determine whether postcranial measurements were recorded before the non-cranial remains were lost.

Do not infer missing measurements from witness descriptions.

If raw postcranial measurements exist, compute Zana's stature and limb-proportion indices and compare them with appropriate modern-human ranges.

Because yes — we may have been sitting here discussing whether Zana could have had unusual proportions when somebody with calipers may have measured the damn bones 50 years ago. 😄

You're right, and it's the same error I'd already fixed twice — upgrading "constrained" to "impossible." Accepting the rewording. And your fingertip-to-knee point is the sharpest thing here: that's what witnesses actually describe, and it's a composite, not a bone ratio. Let me get the real anthropometry.

Your rewording is adopted, and the ASHR data proves your point — population means span 0.99 to 1.07, far more plasticity than IMI implies. Rebuilding Track B on your eight-field schema.

Let me verify those two counts.

The standing-rule line didn't land in Track B. Checking the file head.

False alarm — the line is there, just wrapped. Both files are correct. Let me record the pattern itself, since this is the third instance.

Your rewording is in, verbatim, in both the document and the README. And you're right that the IMI-alone framing was the deeper error — I picked the measurement that was easy to find rather than the one witnesses describe.

The limb row now reads

Normal human variation does not approach great-ape proportions; strongly constraining for H4 if eyewitness reports imply truly ape-like anatomy.

With the evidence for plasticity given its proper weight: arm span/height population means span 0.99–1.07 (Ghanaian males 1.07, +4.8% over the European group), brachial index 75.5 versus 81.4, limb proportions developmentally and evolutionarily plastic within Homo sapiens, prehistoric series showing brachial/crural shifts that don't always fit a simple Allen's-rule story — and the brachial index possibly more evolvable than lower-limb proportions.

The measurement that matters is missing, and I've flagged it as the priority

"Arms reached the knees" is fingertip-to-knee geometry — a composite of acromion height, acromion-to-fingertip arm length, and knee height, all relative to stature. I could not source any of those three as fractions of stature. Searches returned ergonomic tables in absolute centimetres. So the quantity witnesses actually describe has not been computed, and I've said so rather than substituting IMI for it. Also still unsourced: humerus/height, total arm/total leg, hand length, shoulder height.

Schema rebuilt

All three completed traits now report your eight fields — observed range, pathological range, population means, heritability and architecture, environmental contribution, fitness cost, confidence, and whether reports require values outside the known human range. Prose verdicts are gone.

The pattern itself is now on the record

This was the third time in this investigation I turned a constraint into an impossibility and you had to correct it:

1. "H4's population model is refuted" — strawman plus a backwards fitness argument
2. "Zana was an ordinary Homo sapiens woman who was mythologised" — five mechanisms excluded, not a phenotype established
3. "Limb proportions are a hard structural limit" — strongly constraining, not impossible

I've written it into DISCARDED.md as a standing rule rather than three separate entries: state what the evidence excludes, name what it leaves open, and never let the absence of a positive finding become a positive negative claim. That's the failure mode, and it's mine rather than the swarm's — the Qwens have mostly been too cautious, not too bold.

Where Track B stands

DISCARDED.md at 140 rows. Bigfoot unpublished, nothing committed.

Next, unless you redirect: the four remaining traits, plus a targeted hunt for the three stature-normalised anthropometric ratios so fingertip-to-knee can actually be computed. That last one is probably worth more than the other four combined, since it's the only number that speaks directly to what people report seeing.

Yep — this is finally framed the way I’d want it.

The big improvement is that Claude has stopped using an easy proxy like IMI as though it were the thing witnesses actually report. “Hands down near the knees” is geometry of the whole upper body, not just humerus/femur proportions. Arm-span/height ranging around 0.99–1.07 across human populations already shows there’s more room inside H. sapiens than that earlier 68–70 IMI comparison suggested.

But before Claude spends much more time reconstructing generic human ranges, I’d actually interrupt Track B for the Zana osteometry hunt we just discussed. We may have a much better empirical calibration sitting in old records:

Recover every raw skeletal measurement ever recorded for Tkhina-75/Zana and Tkhina-71/Khwit before the postcranial remains disappeared.

Especially anything involving stature, acromial height, arm span, humerus/radius/ulna, femur/tibia, hand length, shoulder breadth, knee height, skull breadth/height, prognathism and mandibular robusticity.

If even partial long-bone measurements survive, we can calculate some of these ourselves instead of asking, “Could a human possibly have proportions X?” We could ask:

What proportions did the actual historical ‘wild woman’ have?

And even if the postcranial measurements are gone, the cranial measurements should help with the robusticity row. That makes Zana a calibration specimen in a much stronger sense: genome + ancestry + kinship + skull morphology + potentially body proportions.

So I’d send Claude:

Pause the remaining four Track B traits briefly. Do a targeted osteometric recovery for Tkhina-75 (Zana) and Tkhina-71 (Khwit). Find primary or near-primary Soviet/Russian measurements from the excavation and subsequent anthropology work. Extract raw numbers only where available. Prioritize postcranial measurements because the remains may later have been lost. If usable long-bone or stature measurements exist, compute limb indices and compare with appropriate human reference ranges. If only skull measurements survive, use them to directly inform the craniofacial/robusticity Track B row. Do not infer missing measurements from witness descriptions.

After that, let him finish musculature, feet, voice, and robusticity.

And his new standing rule is a good one: constraint ≠ impossibility. That one should save us a lot of cleanup.

I sent him your previous response, he is working on it now

Good. That osteometry branch is exactly where I’d go next.

If he finds raw long-bone measurements, we can stop arguing from generic human ranges and directly calculate what Zana/Khwit actually looked like proportionally. If he finds only skull measurements, that still gives us a real anchor for the robusticity/craniofacial row.

The highest-value outcome would be something like:

Zana’s humerus/femur, radius/humerus, stature estimate, shoulder breadth, or arm-span proxy were measured before the postcranial remains disappeared.

Then we can compare the actual numbers against modern human population ranges instead of witness descriptions.

So yeah, let Claude dig. This one has a real chance of producing a useful calibration point rather than another “maybe.”

Good redirect — if someone put calipers on those bones in 1975 we don't need to model anything. Let me trace the primary anthropology record.

The 1987 publication is identified. Now the decisive question — whether anyone measured the long bones before they were lost.

Clear picture now, and it's partly bad news. Let me make one targeted attempt at the raw numbers.

The gold mine is mostly gone — but the search turned up something better than measurements.

You were right to interrupt, and the search gave something better than measurements — though the specific thing you hoped for is gone.

The bad news: the postcranial record is a closed dead end

No long-bone measurements were ever published. Whether anyone put calipers on the humerus, radius, femur or tibia between 1975 and 1992 is unresolved and nothing locatable was published. So Zana's stature, brachial index, crural index, intermembral index and robusticity cannot be computed — and the bones no longer exist to re-measure. That's closed, not pending, and I've marked it that way in Track B so it doesn't sit there looking like an open task.

The craniometric record exists but is offline

Burtsev & Kolodieva, "Results of a Preliminary Investigation of a Skull from the village of Tkhina, Abkhaz ASSR" (Russian), Papers of the Moscow Naturalists' Society, anthropology section, 1987 — studied by M.A. Kolodieva and M.M. Gerasimova, compared against Abkhaz male skulls at the Moscow State University Institute of Anthropology. The raw table isn't online; it needs the physical publication, in Russian. Same class of problem as Lausanne.

The finding that matters isn't about Zana at all

Every robust, "archaic-looking" feature in this record belongs to Khwit, not Zana. Facial skeleton significantly larger than the mean Abkhaz type; superciliary measurements and indices exceeding that series and "the maximum size of some fossil skulls studied"; large zygomatics and mandible; deep temporal fossae implying powerful masticatory muscles.

And Khwit's genome is thoroughly ordinary — Zana's son by an ordinary Abkhaz man, Y-haplogroup R1b1a1b1b, no excess archaic admixture.

So if those 1987 descriptions hold, pronounced craniofacial robusticity occurred in an individual with an ordinary modern human genome. That answers Track B's trait 4 from this investigation's own specimen, in H4's favour on mechanism — and simultaneously it's a strong point for H1, because robust morphology demonstrably does not imply archaic ancestry. The cited "peculiarities" (Inca bone, sutural bones in the sagittal suture, double mental foramina) are all normal-range non-metric variants.

And a neat closure on Zana

The only morphological feature anyone recorded for her is pronounced prognathism. Her genome places her ancestry in present-day East and West African populations, and alveolar prognathism is a normal population-variable trait with higher expression in African-ancestry populations. The one thing recorded about her skull is exactly what her verified ancestry predicts. Nothing left over, no archaic explanation needed.

Caveat I'm holding firmly

All the 1987 descriptions above are second-hand, quoted by secondary sources, several of them cryptozoology-adjacent. The publication's existence and authorship are reasonably established; the specific claims are not verified against the original. Flagged throughout.

My read on whether to chase the 1987 paper: lower value than it first looks. It characterises Khwit, whose ancestry we already know is ordinary, and the specimen whose proportions we actually wanted burned in 1992. It would give real supraorbital numbers against a reference series for Track B trait 4, which is worth something — but it's a Russian library request for a number that refines an already-clear qualitative answer.

Track B now stands at hair yes, stature yes, limb proportions strongly constraining, witness-facing measure uncomputable for this specimen, robusticity probably yes (low-medium confidence), and musculature, feet and voice unassessed.