Lucy and the Pre-Human Hominins

Lucy was not human: her brain and locomotor anatomy are australopith, though she walked upright. The oldest biped claims rest on few, reconstructed bones, and mainstream work has reversed the Ardipithecus headline. Homo naledi leans human on both sides' methods. The family trees rest on a method that failed its own controlled test.

By AI (Claude Opus 5, thinking level unavailable; cut to an overview and split by Claude Opus 5.5, thinking level unavailable) Draft:  August 2026

Casually dismissing a method because it gave a result "we are personally uncomfortable with would be a serious error." — Todd C. Wood, 2011, the young-earth creationist whose own analysis put Lucy's species in a cluster with chimpanzees.11

The four reviewers were "in strong consensus that the methods, data, and analyses do not support the primary conclusions." — eLife's editors, 2023, on the claim that Homo naledi buried its dead.22

Summary

Before the first fossils anyone calls Homo, the record holds a set of animals that walked upright on two legs and had brains the size of a chimpanzee's.  Sahelanthropus.  Orrorin.  Ardipithecus.  The australopiths, including the skeleton everyone knows as Lucy.  The robust paranthropines.

The mainstream reads them as the early part of the human line.  Young-earth creationism has generally read them as extinct apes.  This overview and its child articles check both readings against the published measurements, and they reach a mixed verdict that will not please either side.

The short version.  The material at the bottom of this record is far thinner than the museum displays suggest, and the arithmetic says so.  Two of the oldest bipedality claims rest on reconstructions of crushed specimens, and mainstream researchers have publicly reversed one of them.  Homo naledi leans human on the mainstream's classification and on creationism's own quantitative method.

One line per article.

  1. Sahelanthropus, Orrorin, and Ardipithecus: the three oldest claimed hominins rest on very little bone, their bipedality rests on reconstructions and a few features, and mainstream work on the same fossils has reversed the 2009 Ardipithecus headline.
  2. Lucy and Australopithecus afarensis: Lucy was not a human. Five independent lines say so, and the fifth is young-earth creationism's own quantitative method. She was not an ordinary ape either: her kind walked upright.
  3. The South African Australopiths: the Taung child, A. africanus and Little Foot were small-brained apes that walked upright, not humans, and young-earth baraminology agrees.
  4. Australopithecus sediba: a small-brained, tree-climbing australopith with some Homo-like teeth, hands and hips; not human, and probably too young to be the ancestor of Homo.
  5. Paranthropus: the robust australopiths were small-brained bipeds with the largest cheek teeth of any hominin; every young-earth writer calls them non-human, and creationist baraminology puts them in a group of their own.
  6. Homo naledi and the Rising Star Claims, under The Genus Homo: naledi leans human, on the mainstream's classification and on creationism's own quantitative method.

These articles correct four citation errors in their own research dossiers — three wrong first authors and one wrong licence, all checked against the publishers' indexed records.  See the footnotes of the two written child articles, and the authorship note in Sahelanthropus, Orrorin, and Ardipithecus.

What these articles cover, and why the line is drawn there

This overview and its children stop at genus Homo.  H. erectus, the Neanderthals, the Denisovans and H. floresiensis belong to The Genus Homo, along with everything resting on ancient DNA.

That split is not editorial convenience.  It is the method.  For almost every form in these articles there is no usable genetics of any kind — no ancient DNA and, apart from a few Paranthropus teeth, no proteome.  They are classified on bones and on their position in the rock, and nothing else.  Once genus Homo is reached, sequence data enters the argument and changes what can be settled and how, which is a different kind of article.

Homo naledi is the boundary case, because it is where the human/non-human line is actually being contested right now, and because it is the one taxon in this whole field with enough specimens for real statistics.  It now has its own article under The Genus Homo, Homo naledi and the Rising Star Claims, and Part 1 below gives its verdict.

The rank question, stated up front, and stated down

In the other articles of this series the transitions cross large taxonomic distances.  A lobe-finned fish to a four-limbed land vertebrate.  A reptile-grade jaw to a mammal ear.  Nobody argues those are the same created kind.

This one is different, and honesty requires saying so before the argument starts.

On the mainstream account, the boundary between the australopiths and us is a genus-level boundary inside a single family.  Sahelanthropus, Orrorin, Ardipithecus, Australopithecus, Paranthropus and Kenyanthropus sit in one subtribe; Homo sits alone in another.3  That is a low rank by the standards of this series, and creationists accept common ancestry at that rank routinely — for cats, for dogs, for horses, for finches.

And Todd Wood's baraminology divides the taxa in the same place.  His analyses come out with genus Homo (plus Australopithecus sediba) on one side and Australopithecus, Paranthropus, Pan and Gorilla on the other.34  The mainstream and the best creationist quantitative work agree about where the morphological gap is.  They disagree about what crossing it means.

So the argument is not about the location of a line.  It is about content.  Across that boundary the brain roughly triples in size, the thorax and shoulder reorganise, the hand changes function, the foot changes function, and the balance organs change shape.  That content is what makes the case interesting.  The taxonomic rank is not, and inflating it invites an easy rebuttal.

One thing this article will not do

It will not say the hominin fossils are fakes.  Piltdown was a fraud, it was exposed in 1953, and nothing in the pre-Homo material examined for this article is fraudulent.  The reconstruction problems documented in the child articles were documented by the reconstructors themselves, in their own papers, which is how anyone knows about them.

Part 1: The verdicts, one article at a time

Sahelanthropus, Orrorin, and Ardipithecus

The record is very poor at the bottom, which is exactly backwards from where the confident claims are made.  Sahelanthropus is nine cranial specimens, one femoral shaft and two partial ulnae from one locality.556  Orrorin is thirteen fossils.6  Two of the three oldest claimed hominins carry their load-bearing evidence on reconstructions of crushed specimens.

The strongest material against the standard picture is entirely secular and dated 2019 or later.  Three papers from one research group have reversed the 2009 headline claim about Ardipithecus, and a 2023 paper puts the Sahelanthropus ulna in knuckle-walking morphospace.78910  This is stronger than anything apologetics has produced on the subject, and it should be cited instead of the apologetics.  The full case is in Sahelanthropus, Orrorin, and Ardipithecus.

Lucy and Australopithecus afarensis

Lucy was not a human.  Her endocranial volume is 388 ml.  Her species averages 445.  Chimpanzees average 369 across a sample of 240 individuals.11  Modern humans average about 1,350.1212  Her species has ape-like internal brain organisation, great-ape-like balance organs, ape-like shoulder development, and an infant foot built to grip.  And the most rigorous method young-earth creationism possesses agrees: Todd Wood's baraminology has twice placed Australopithecus afarensis in a cluster with chimpanzees and gorillas, under every distance metric, character filter and clustering algorithm tried.344

Christopher Rupe's recent case for Lucy as a pygmy human deserves a fair hearing and does not survive one.  He is right about one thing his own movement has been wrong about for thirty years, and his load-bearing anatomical argument fails on a parenthesis he added to somebody else's sentence.

The mixture argument fails on excavation, not on anatomy.  Exactly one bone in Lucy's skeleton is foreign, mainstream anatomists found it, and the same paper cleared everything else by showing six of her vertebrae articulate in an unbroken run.  Two other skeletons settle the general question outright.

Her kind walked upright.  The Laetoli footprints are real, human-like and not a reconstruction, and Nature calls them the oldest unequivocal evidence of obligate bipedalism.13

The measurements, Rupe's case, the excavation record and the objections are in Lucy and Australopithecus afarensis.

The South African australopiths, Australopithecus sediba, and Paranthropus

Each has its own article.  One line on each:

  • The South African Australopiths covers the Taung child, A. africanus and Little Foot. Todd Wood's 2010 baraminology left A. africanus ambiguous; in his words, the evidence for including it in the human holobaramin "appears quite weak."3
  • Australopithecus sediba: Wood's baraminology puts sediba with genus Homo,3 while a 2021 mainstream paper reads the two Malapa individuals as two separate genera,14 so its place is contested on both sides; the sediba article shows that Wood's own later skeleton analyses do not group it with humans. The claim that it is a mixture is answered in Lucy and Australopithecus afarensis, Part 4.
  • Paranthropus: Wood's 2010 analysis and the 2021 rerun on a 69-character dataset both put the robust australopiths in a group of their own.34

Homo naledi

Homo naledi sits under The Genus Homo, and Homo naledi and the Rising Star Claims now carries the human-or-not verdict this article used to argue.  Creationism's own quantitative method says human, and it is not close.  Todd Wood's 2016 supermatrix analysis places H. naledi in the human holobaramin.15  Sinclair and Wood's 2021 reanalysis confirms it:  in the two-cluster human/non-human partition, naledi's individual silhouette width is +0.49, and forcing it into the non-human cluster drops that to −0.49.4  The mainstream also puts it in Homo, on cranial shape, dentition, hand, wrist, foot and lower limb, while noting the australopith-grade brain, shoulder and pelvis.

The one sustained creationist dissent began on the other side.  Jean O'Micks's first paper, on skull and tooth characters alone, concluded that "Placement of H. naledi in the human baramin is supported".1616  Later in 2016, after adding postcranial characters, he reversed that, and across five papers in 2016–17 he argued that H. naledi is not human.17  Wood replied on taxon sample size, and McLain on the geology and taphonomy of the site.1819  Remaining undecided is defensible on one specific ground:  Wood's analysis is craniodental, and naledi's postcranium is genuinely australopith-like, which is where O'Micks aimed.  The 2026 enamel proteome did not settle the question;  the naledi article gives the result.

Part 2: The record across genera

Species named on almost nothing

taxon what the species rests on
Sahelanthropus tchadensis 9 cranial specimens (min. 3 individuals) + 1 femoral shaft + 2 partial ulnae, one locality
Kenyanthropus platyops essentially one distorted cranium, KNM-WT 40000
Australopithecus garhi one partial cranium, BOU-VP-12/130
Australopithecus deyiremeda one maxilla and two mandibles
Orrorin tugenensis 13 fossils, min. 5 individuals, none associated as a skeleton
Ardipithecus kadabba isolated teeth and a few postcranial elements

This is not a creationist framing.  It is the arithmetic of the published record, and the sharpest confirmation of it comes from creationists doing statistics.

Todd Wood's 2010 analysis had to eliminate Kenyanthropus from all three datasets that contained it, because 62.3%, 72.4% and 86.4% of its character states respectively were unknown.3  Eleven years later, Sinclair and Wood dropped four taxa outright — H. antecessor, S. tchadensis, A. garhi and K. platyops — for having under 20% taxic relevance, meaning fewer than one character in five could be scored.4

Read that again with the oldest name in it.  Creationism's own best quantitative method cannot say anything at all about the oldest claimed hominin, in either direction, because there is not enough of it to measure.  That is a baraminology result and a fragmentary-record result at the same time.

An independent check nobody in this argument has run

The Paleobiology Database is an open, non-partisan compilation of fossil occurrence records.  A query for the whole tribe Hominini was run for this article's research dossier on 3 August 2026.20

327 occurrence records, out of 2,001,053 in the database.  The breakdown for the pre-Homo forms:

taxon occurrences
Australopithecus afarensis 41
Australopithecus indet. 32
Australopithecus africanus 7
Australopithecus anamensis 2
Ardipithecus ramidus 2
Sahelanthropus tchadensis 1
Australopithecus garhi 1
Ardipithecus kadabba 1
Australopithecus sediba 1
Australopithecus deyiremeda 1

Two cautions, and this article prints both rather than the headline.

First, a database "occurrence" is a taxon at a collection, not a specimen.  One occurrence can stand for one tooth or for an entire death assemblage.  This is a census of localities, not of bones.

Second, the database substantially undersamples this group, because paleoanthropology publishes in its own journals and does not systematically deposit occurrence data.  Homo naledi, with 1,550 specimens, does not appear at all.  So these are floors, not totals, and anyone quoting the 327 as "all the hominin fossils" would be misusing it.

What the table does legitimately show is the distribution of places:  A. afarensis is known from an order of magnitude more localities than Ardipithecus, and five named taxa come from one apiece.  That corroborates the specimen table above, and those in the child articles, from a source that has no stake in the argument.

And a number this article refuses to give

There is no published census of hominin fossil specimens.  This was looked for and not found.  The nearest thing is Schwartz and Tattersall's four-volume The Human Fossil Record, which catalogues specimens but is a print reference work and does not reduce to a figure.

So the popular claim that all the hominin fossils would fit in a coffin, or on a billiard table, has no source and should not be repeated.  Homo naledi alone is 1,550 specimens from one chamber.  Where a total is wanted, use the per-taxon counts above and let the reader add them up.

Part 3: How much confidence the method can carry

Every family tree in these articles, and Todd Wood's baraminology too, runs on cladistic analysis of craniodental characters — the shapes of skulls and teeth.  It is worth knowing how well that method performs when its answer can be checked.

The controlled test

Mark Collard and Bernard Wood ran the experiment in PNAS in 2000, and the design is the important part.

Cladistic analysis of craniodental characters is the standard method for placing fossil hominins, and it had never been externally validated — because for fossils there is no independent answer to check against.  So they applied identical methods to equivalent craniodental data from two groups of living higher primates whose molecular phylogenies are known and robust:  the apes, and the papionins (baboons, mangabeys, macaques, mandrills).

The craniodental phylogenies were incompatible with the molecular phylogenies for both groups.  The ape analysis grouped humans with orangutans in one clade and chimpanzees with gorillas in the other — and the false chimpanzee–gorilla clade was supported by four more putative shared derived characters than the true chimpanzee–human clade.

The bootstrap numbers are the part that stings.  None of the clades supported at 70% or above was compatible with the molecular trees.  A false three-way ape clade came out at 95%.  A false papionin clade came out at 98%.2121

Their own conclusion:  craniodental cladistics cannot be relied on to reconstruct the relationships of apes, papionins, "and, by extension, the fossil hominins" — and, more troublingly, such analyses can strongly support phylogenetic hypotheses that are misleading.21

This is entirely secular, published in PNAS, by one of the field's leading figures who is no friend of creationism.

Two peer-reviewed results also measure convergence inside Australopithecus itself, on Lucy's limb-joint proportions and on the gorilla-type jaw.  They are in Lucy and Australopithecus afarensis, Part 6.

The caution against overreach, stated plainly

Collard and Wood's conclusion is that craniodental cladistics is unreliable.  It is not that morphology carries no information, and it is certainly not that the animals are unrelated.  They call for new approaches, not for abandoning phylogeny.

So the claim this article makes is narrow and it is strong:  the confidence attached to any particular hominin family tree is not warranted by the method that produced it, and the field's own controlled test says so.

That applies to the tree that puts Lucy on the line to humans.  It applies equally to Todd Wood's baraminology, which runs on the same craniodental matrices and inherits the same weakness — which is exactly why Wood says so himself.

Part 4: Three more claims from our own side that do not survive

An audit that names other people's errors has to name its own first.

The first of the four claims this research retires, that Owen Lovejoy cut up Lucy's pelvis with a power saw, is answered in Lucy and Australopithecus afarensis, Part 8.  The other three span the genera.

2. "Charles Oxnard showed australopiths were just apes"

He is not a creationist and he did not say this.

Oxnard is a distinguished comparative anatomist.  His multivariate analyses concluded that the australopithecines are uniquely different from both modern humans and the African apes — more different from each of them than those two are from each other — that they show a mosaic including resemblances to the orangutan, that their locomotion probably included substantial climbing, and that genus Homo may be old enough to parallel Australopithecus, denying the latter a place in the direct human lineage.

That is a claim about phylogenetic placement.  It is not a claim about humanity, and it is not "they were apes."  His position is that they were neither — something unique and now extinct.  Quoting him for "just apes" misuses him, and a hostile reader who checks will find it in about five minutes.

Two further cautions.  His central paper is from 1975, which predates Lucy's full description, the publication of the Laetoli footprints, Selam, Kadanuumuu, and every CT-based study cited in this article.  And this project's own note file contains a long passage attributed to that paper which is a composite, assembled out of several parts of it and presented as one continuous quotation.22  It must not be reproduced.  Nothing is quoted from Oxnard anywhere in this article, because the paper itself is paywalled and could not be opened.

What may legitimately be taken from him is the "unique mosaic, not intermediate" claim, which the modern data still supports, and his methodological observation about the reconstruction of incomplete foot material — attributed as a paraphrase, with the date, and not as a quotation.

3. "All the hominin fossils would fit in a coffin"

No source.  This was looked for and not found, in any form that could be cited.

Homo naledi alone is 1,550 specimens from one chamber.23  Use the per-taxon counts in Part 2 and in the child articles instead — they are sourced, they are more specific, and the Sahelanthropus row makes the point far better than a slogan does.

4. "There is no fossil record of chimpanzees or gorillas at all"

Nearly true, and therefore worth stating carefully rather than loudly.

Three chimpanzee teeth were described from the Kapthurin Formation in Kenya in Nature in 2005.  So the record is not empty.  It is close to empty, and the honest word is "drastically poorer" than the hominin record.24

There are two candidate explanations and both are live:  a genuine taphonomic difference, since forest settings preserve badly and lake margins preserve well, and collection bias toward hominins.  A creationist arguing that the near-absence shows the hominins are not really apes is over-reading it.  A critic arguing it shows nothing is under-reading it — it does mean that when a fossil ape turns up in Africa in hominin-bearing deposits, there is no comparative series to place it against.

Objections and responses

"The gap between the australopiths and Homo is enormous."

Do not use this.  It is a genus-level gap inside one family, at a rank where creationists accept common ancestry in every other mammal group without argument.  The mainstream's taxonomy and Todd Wood's baraminology draw the line in the same place.

Overselling the rank invites a one-sentence rebuttal.  Argue the content instead — the brain roughly triples, the thorax and shoulder reorganise, the hand and foot change function.  That is a harder thing to answer and it is true.

"The hominin fossils are out of stratigraphic order, or fake, or both."

Neither claim is supported and this article will not carry either.

On order:  across 2,001,053 occurrence records in the Paleobiology Database there are no primates in Early Cretaceous or older rock, and no great apes in Eocene or older rock.25  The argument in this field is about what the fossils are, not about where they sit.

On fraud:  Piltdown was a fraud and was exposed in 1953.  Nothing in the pre-Homo material examined here is fraudulent.  The reconstruction problems in Sahelanthropus, Orrorin, and Ardipithecus are documented by the reconstructors.

"The whole field is confused — look how many species they keep naming."

The complaint is real and it comes from inside the field.  Tim White has argued in print that the incentives of publishing, funding, career advancement and media promotion work against self-regulation, that limited fossil material makes it worse, and that there is ample motivation to generate new species names.26

But this argument cuts against the creationist reading too, and honest use has to say so.  A field that names too many species produces apparent diversity and apparent gaps that may not be real.  And a creationist who counts named hominin species to show how confused the field is, is counting the inflated list.

"Paleoanthropology never admits it is wrong."

The Ardipithecus reversal and the 2023 Homo naledi reviews are the answer.  Three papers from one research group reversing a 2009 Science headline.  Four eLife reviewers telling a famous team its evidence did not meet the field's standards, in public, permanently attached to the paper.

The naledi review episode is worth a paragraph of its own, regardless of what naledi turns out to be.  It is a fully documented case of paleoanthropology policing itself in public, with four reviewers telling a famous team that its evidence did not meet the field's standards, and with the record of that exchange permanently attached to the paper.  A skeptical reader who thinks creationists caricature the field should be shown this.  A creationist reader who thinks the field never self-corrects should be shown it too.  It also disposes of "the experts all agree," in either direction.

A creationist argument that depends on the field never self-correcting is refuted by the field's own archives, and citing those archives is a stronger move than complaining about them.

What would change these verdicts

Stated as conditions, so a reader can hold these articles to them.

  • A baraminic analysis on a postcranial character matrix. Every baraminology result in these articles is craniodental, its author says so, and the published creationist dissent is specifically postcranial. H. naledi's 1,550 specimens make such a matrix possible for the first time.
  • A published census of hominin fossil specimens. None exists. Somebody should build it out of Schwartz and Tattersall's four volumes, and both sides would benefit.
  • The conditions for each genus are in its own article: those for Lucy in Lucy and Australopithecus afarensis, and those for naledi in Homo naledi and the Rising Star Claims.

Verdict

The pre-Homo record is mixed, and it is thinner at the bottom than the museum displays suggest.  The oldest claimed hominins rest on very little bone and on reconstructions that mainstream work has partly reversed.  Lucy was not a human, on five independent lines, one of them young-earth creationism's own method;  she was not an ordinary ape either, and the Laetoli footprints show her kind walked upright.  Homo naledi leans human on both sides' methods.  The method behind every hominin family tree failed its own controlled test, which cuts against the confident trees and against creationist baraminology alike.  The gap between the australopiths and Homo is genus-level inside one family:  argue its content, not its rank.

Sources

The two research dossiers behind this article are published alongside it:  research/pre-homo-hominins.md, which carries the specimen inventories, the baraminology audit, the eLife review history and the full quotes ledger; and research/rupe-lucy-mixture.md, which audits Christopher Rupe's whole dated public corpus and the taphonomy of Lucy's excavation.

Every quotation printed above was checked against a source that was actually opened, and appears in one of those two ledgers with a verdict.  Twenty passages in this project's notes could not be verified from any source, and none of them appears here.  Where a source could only be reached as an abstract, the footnote says so.  The sources that could not be reached at all, and the access routes that did work, are recorded with the dossiers.27

  1. 1.

    Wood, Todd C.  “Baraminology, the Image of God, and Australopithecus sediba.”  Journal of Creation Theology and Science Series B 1:6–14.  2011.  Wood's reply to the theological objections raised against his 2010 results.  He sorts the objections into three concerns:  that the data are biased or inadequate; that humans could once have existed as several species; and that putting A. sediba in the human holobaramin is theologically offensive.  He answers each one, and argues that objections about the image of God or the doctrine of salvation are not valid reasons to reject human baraminology.  He proposes testing the "discontinuity hypothesis", the idea that God created organisms in discrete, recognizable groups.  The article uses his warning that casually dismissing a method, because it gives a result that creationists are personally uncomfortable with, is a serious error. ↩

    1. 1a.

      “we are personally uncomfortable with would be a serious error”.  The verbatim span is the quoted fragment.  It is quoted as a fragment because that is how it appears in the sentence.  The surrounding clause, casually dismissing baraminology because it gave a result, is paraphrased rather than reproduced. ↩

  2. 2.

    eLife editors and four anonymous peer reviewers.  “Evidence for deliberate burial of the dead by Homo naledi: public reviews and editorial assessment, version 1.”  eLife reviewed preprint RP89106.  Jul 11, 2023.  The public reviews and editors' assessment that eLife published with the 2023 preprint claiming that Homo naledi buried its dead.  The senior editor was George Perry of Penn State, and four anonymous peer reviewers wrote the reviews.  The article uses them because they are the specialists' own verdict on that claim, and eLife publishes its reviews whatever the authors wish, which is why this episode is unusually well documented. ↩

    1. 2a.

      The assessment says that demonstrating the earliest known intentional funerary practice by a small-brained hominin.  “would be a landmark finding”.  but that.  “the evidence for these claims is considered inadequate in the current version of the study.”  It records that the four reviewers were.  “in strong consensus that the methods, data, and analyses do not support the primary conclusions.”  and that without full excavations the study lacks crucial geoarchaeology (especially micromorphology, the microscopic study of sediment layers) and taphonomic work, and that the null hypothesis must be that the skeletons accumulated naturally. ↩

  3. 3.

    Wood, Todd C.  “Baraminological Analysis Places Homo habilis, Homo rudolfensis, and Australopithecus sediba in the Human Holobaramin.”  Answers Research Journal 3:71–90.  2010.  A young-earth creationist's statistical study of fossil hominin skulls and teeth.  Wood reanalysed eight already-published cladistic matrices (tables of characters scored for each fossil), by Strait, Grine and Moniz 1997; Strait and Grine 2001 and 2004; Zeitoun 2000; Prat 2004; Smith and Grine 2008; Argue et al. 2009; and Berger et al. 2010, without rescoring any character.  The eight datasets were compiled by conventional paleoanthropologists, not by creationists.  The method has two parts:  baraminic distance correlation, which asks whether two taxa share a pattern of distances to all other taxa, and multidimensional scaling, which projects the distance matrix so clusters can be seen.  Wood holds a PhD in biochemistry from the University of Virginia and directs Core Academy of Science.  The article uses it as the most rigorous creationist method, and it places Australopithecus afarensis in a cluster with chimpanzees and gorillas.

    1. 3a.

      Section Abstract:  The analysis divides hominins into as many as four holobaramins:  genus Homo including Australopithecus sediba; genus Paranthropus; Australopithecus africanus; and a fourth group consisting of Gorilla, Pan, Australopithecus afarensis and Australopithecus garhi.  Wood says the results.  “tentatively confirm the common creationist claim that fossil hominids can be divided into human and non-human categories”.  and that, in contrast to many creationist claims, H. habilis, H. rudolfensis and.  “most surprisingly”.  Au. sediba belong in the human holobaramin. ↩ ↩ ↩ ↩

    2. 3b.

      Table 1:  Following Wood and Richmond 2000 and Wood and Lonergan 2008, it places Sahelanthropus, Orrorin, Ardipithecus, Australopithecus, Paranthropus and Kenyanthropus in the subtribe Australopithecina, and Homo alone in the subtribe Hominina.  Both lie inside the tribe Hominini, the subfamily Homininae and the family Hominidae.  The nomenclature is the mainstream's, printed in a creationist journal.  The article uses it as the basis for its statement that the disputed boundary is a genus-level one inside one family. ↩

    3. 3c.

      Kenyanthropus platyops was present in three of the eight datasets and was eliminated from all three because too many character states are unknown:  86.4% missing in Strait and Grine 2004, 72.4% in Smith and Grine 2008, 62.3% in Prat 2004. ↩

    4. 3d.

      Section Discussion:  Wood writes that the evidence for including Au. africanus within the human holobaramin.  “appears quite weak”.  His reasons, given unprompted:  it appeared as an outlier in three-dimensional multidimensional scaling in both datasets where it correlated positively; bootstrap support for the africanus/rudolfensis correlation was only 62% and 59%; reducing the outgroup eliminated the correlation entirely in one dataset; and Zeitoun's dataset showed significant negative correlation between an africanus specimen and most members of the human group. ↩

  4. 4.

    Sinclair, P., and Todd C. Wood.  “Revising Hominin Baraminology with Medoid Partitioning and Fuzzy Analysis.”  Answers Research Journal 14:451–462.  2021.  A 2021 re-analysis of two hominin character sets by Sinclair and Wood.  The paper exists because another creationist attacked the method:  Reeves argued that baraminic distance correlation is an untested homebrew statistic, and that standard clustering algorithms and Jaccard distances should be used instead.  Sinclair and Wood took the criticism and reran everything with those tools.  The result was a robust cluster of putatively human taxa that appeared in almost every analysis, including Homo naledi and Australopithecus sediba, while the status of Homo floresiensis stayed unresolved.  The article cites it for Wood's placement of A. afarensis with chimpanzees and gorillas under every method tried, and for the four taxa that could not be analysed at all. ↩

    1. 4a.

      On the Berger et al. 2010 character set, 69 craniodental characters for 12 taxa, every combination of distance metric, character filter and clustering method produced the same three-cluster partition, with average silhouette widths of 0.58–0.59.  A silhouette width scores how well a taxon fits its own cluster, from −1 (a misfit) to +1.  On Kaufman and Rousseeuw's scale, above 0.5 counts as reasonable to strong clustering.  Forcing Au. sediba into either non-human cluster produced strongly negative silhouette widths (−0.611 into Paranthropus, −0.525 into the other australopiths) against +0.525 in the Homo cluster.  Giving it a cluster of its own dropped the average from about 0.58 to 0.37. ↩ ↩ ↩

    2. 4b.

      On the larger 391-character, 24-taxon supermatrix (Wood's 2020 update of Dembo et al. 2016, archived on Dryad), four taxa were removed for having under 20% taxic relevance, meaning fewer than one character in five scorable:  H. antecessor, S. tchadensis, Au. garhi and K. platyops.  Paranthropus was removed separately as uncontroversially non-human.  This is a creationist result about the fragmentary record, and the article uses it as such in Part 1. ↩

    3. 4c.

      In the two-cluster human/non-human partition, H. naledi's individual silhouette width is +0.49.  Moving it to the non-human cluster drops it to −0.49.  Giving it a cluster of its own gives an average of 0.16. ↩

  5. 5.

    Guy, Franck et al.  “Postcranial evidence for Sahelanthropus tchadensis.”  Research Square.  Sep 10, 2020.  A preprint of a paper on the leg and arm bones of Sahelanthropus tchadensis from site TM 266 in Chad, later published in Nature as Daver et al., Nature 609 (2022): 94–100.  It describes one left femoral shaft and two ulnae, and compares their cortical bone and bending strength with those of apes and humans; the authors concede that none of the bones can be reliably ascribed to a skull found at the site.  The article uses it as the postcranial evidence that the Chad hominin was a terrestrial biped that also climbed carefully. ↩

    1. 5a.

      The craniodental inventory is given as a nearly complete cranium (TM 266-01-060-1), three mandibles and several isolated teeth, representing a minimum of three adult individuals.  The three postcranial specimens, all found in 2001: TM 266-01-063, a left femoral shaft, about 242 mm preserved, July 2001; TM 266-01-050, a left ulnar diaphysis with the proximal epiphysis eroded, 239 mm, July 2001; TM 266-01-358, the proximal half of a right ulna shaft, 155 mm, November 2001. ↩

  6. 6.

    Smithsonian National Museum of Natural History, Human Origins Program.  Human Origins species pages.  2026.  Retrieved Aug 3, 2026.  Section Sahelanthropus tchadensis species page:  “The first (and, so far, only) fossils of Sahelanthropus are nine cranial specimens from northern Chad.”  An online species database from a museum that summarizes the fossils of each species.  It is a secondary source, and the article uses it only for specimen counts, which are the kind of claim such a database is built to carry. ↩

    1. 6a.

      Section Orrorin tugenensis species page:  “There are 13 fossils from at least 5 Orrorin individuals.”  The named specimens are BAR 1000'00, the type, a jaw fragment with three lower molars; BAR 1002'00 and BAR 1003'00, femora.  The page also mentions long curved finger bones and two well-preserved thigh bones. ↩

  7. 7.

    Prang, Thomas C.  “The African ape-like foot of Ardipithecus ramidus and its implications for the origin of bipedalism.”  eLife 8:e44433.  2019.  A study that compares the foot bones of living and fossil primates, using 385 individuals from 45 taxa, with the 4.4-million-year-old Ardipithecus ramidus foot to see which living group it most resembles.  It fits evolutionary models to six foot measurements, and the Ardi foot sits with chimpanzees and gorillas.  The article uses it for the conclusion that the precursor of hominin bipedalism was African-ape-like terrestrial quadrupedalism and climbing.  Mirror:  PMC

    1. 7a.

      Using foot proportions across a large anthropoid sample, Prang finds the Ar. ramidus foot most similar to living chimpanzees and gorillas.  Prang concludes that the evolutionary precursor of hominin bipedalism was African-ape-like terrestrial quadrupedalism (walking on all fours on the ground, with the heel down first) and climbing, and that the modern human foot derives from a form adapted to terrestrial plantigrade quadrupedalism. ↩

  8. 8.

    Prang, Thomas C. et al.  “Ardipithecus hand provides evidence that humans and chimpanzees evolved from an ancestor with suspensory adaptations.”  Science Advances 7(9):eabf2474.  2021.  A study that compares the hand of the 4.4-million-year-old Ardipithecus ramidus with the hands of living apes and humans, using morphometric (shape-measuring) and phylogenetic comparative methods.  The authors argue that the Ardi hand keeps suspensory features (adaptations for hanging from the arms) that chimpanzees and bonobos share, and that the hand changes between Ardipithecus and Australopithecus.  The article uses it as the hand evidence in its account of Ardi.  Mirror:  PMC

    1. 8a.

      The Ardi hand retains suspensory-adapted morphology shared with chimpanzees and bonobos, and the authors identify a shift in hand morphology falling between Ardipithecus and Australopithecus. ↩

  9. 9.

    Meyer, Marc R. et al.  “Knuckle-walking in Sahelanthropus? Locomotor inferences from the ulnae of fossil hominins and other hominoids.”  Journal of Human Evolution 179:103355.  2023.  A morphometric study of ulna shape, testing whether the shape of a forearm bone shows how a primate moves.  It compares the ulnae of humans, apes, Miocene apes and fossil hominins.  The article uses it for the finding that the Chad ulna falls within the knuckle-walking group.

    1. 9a.

      Method: elliptical Fourier analysis of ulna shaft contours across humans, five extant ape species, two Miocene apes and 17 fossil hominin specimens.  Ulna shaft shape correlates with locomotion, and the TM 266 ulna falls within the knuckle-walking morphospace; discriminant function analysis classifies it with Pan and Gorilla at high posterior probability.  The same is true of OH 36, a purported Paranthropus boisei ulna. ↩

  10. 10.

    Prang, Thomas C. et al.  “Ardipithecus ramidus ankle provides evidence for African ape-like vertical climbing in the earliest hominins.”  Communications Biology 8:1454.  2025.  A study of the 4.4-million-year-old ankle bone (talus) attributed to Ardipithecus ramidus, compared with the tali of chimpanzees and gorillas.  The authors find that the Ardi talus resembles those of two living apes that are adapted to vertical climbing and to walking on the ground on the whole sole of the foot.  The article uses it because its conclusion, that humans evolved from an African-ape-like ancestor, is set against the view that the ancestor was a generalised arboreal ape.  Mirror:  PMC

    1. 10a.

      The Ardi talus (ARA-VP-6/500-023) resembles the tali of chimpanzees and gorillas, which are adapted to vertical climbing and terrestrial plantigrade quadrupedalism.  The authors state that their observations are inconsistent with recently proposed models of human origins that envision the human–chimpanzee last common ancestor as a generalised arboreal ape, and that humans instead evolved from an African-ape-like ancestor. ↩

  11. 11.

    Gunz, Philipp et al.  “Australopithecus afarensis endocasts suggest ape-like brain organization and prolonged brain growth.”  Science Advances 6(14):eaaz4729.  2020.  A study of eight fossil skulls of Australopithecus afarensis, Lucy's species, scanned with conventional and synchrotron CT (X-ray scanning that builds a three-dimensional picture).  The authors read brain organisation from the imprints that the brain's folds leave on the inside of the skull, and estimated brain growth by combining new volume estimates with age estimates for two infants.  The method is synchrotron imaging plus geometric-morphometric reconstruction, which is why these figures supersede earlier estimates.  It finds ape-like brain organisation with no features derived toward humans, and protracted brain growth.  The article uses it for the best endocranial volumes available and for its finding on brain organisation.

    1. 11a.

      Measured endocranial volumes.  A.L. 288-1 (Lucy):  mean 388 ml, range 365–417.  A.L. 822-1:  382 ml.  A.L. 333-45:  488 ml.  A.L. 444-2:  522 ml.  Adult A. afarensis average:  445 ml, SD 60 (standard deviation, a measure of spread), range 365–526.  DIK-1-1 (Selam), infant:  275 ml corrected, 237 ml unreconstructed.  Comparative sample:  Pan troglodytes adults, N = 240, mean 369 ml, SD 37.  Excluding the geologically younger A.L. 444-2 drops the adult afarensis average to 420 ml. ↩

  12. 12.

    De Caro, Liberato.  “Commentary: Human brains have shrunk: the questions are when and why.”  Frontiers in Ecology and Evolution 12:1368347.  2024.  A commentary by Liberato De Caro, of the Institute of Crystallography of Italy's National Research Council, on a 2023 study of when human brains shrank.  It gathers the large published samples of living human endocranial volume, the largest being Beals, Smith and Dodd's 1984 survey of 5,288 skulls, and combines three of them into an average of 1,343 cc over 9,948 adults.  The article uses it for the modern human average. ↩

    1. 12a.

      Section 1 Introduction:  “Beals et al. (1984) analyzed data on 5,288 crania from 122 different ethnic groups and reported an endocranial volume of 1,349 ± 78cc.”  The same section combines the Beals sample with two others into an average of 1,343 ± 129 cc for 9,948 adults. ↩

  13. 13.

    McNutt, Ellison J. et al.  “Footprint evidence of early hominin locomotor diversity at Laetoli, Tanzania.”  Nature 600:468–471.  2021.  A Nature paper that relocated and re-excavated five footprints at Laetoli site A, whose location had been lost, and compared them with American black bears, chimpanzees and humans, using 3D photogrammetry and laser scanning.  It found that the site A footprints differ from the famous site G prints in foot shape, gait and 3D form, so at least two hominin taxa with different feet and gaits walked at Laetoli at the same time.  The article uses it for the site G footprints and for the evidence that two kinds of biped walked there.

    1. 13a.

      “widely accepted as the oldest unequivocal evidence of obligate bipedalism in the human lineage”.  The paper calls the site G trackways this.  Site G was discovered in 1976 by Mary Leakey's team and published by Leakey and Hay in 1979.  No skeletal material is associated with the prints, so attribution to A. afarensis is by elimination, since it is the only hominin known from Laetoli at that horizon. ↩

  14. 14.

    Rak, Yoel et al.  “One hominin taxon or two at Malapa Cave? Implications for the origins of Homo.”  South African Journal of Science 117(5/6).  2021.  A 2021 paper about the two Malapa Cave individuals, MH1 and MH2, that argues from the mandibular ramus (the vertical branch of the lower jaw).  Ritzman et al. 2016 read the same anatomy the other way, so the question is still open.  The article uses it to show that a two-taxa reading of the Malapa material now has mainstream support.

    1. 14a.

      The authors argue from the mandibular ramus that MH1 and MH2.  “represent two separate genera:  Australopithecus and Homo.”  The article's point is that this means two individuals of different taxa at one site, not one chimeric skeleton. ↩

  15. 15.

    Wood, Todd C.  “An Evaluation of Homo naledi and "Early" Homo from a Young-Age Creationist Perspective.”  Journal of Creation Theology and Science Series B: Life Sciences 6:14–30.  2016.  Wood's supermatrix baraminology of all named hominins, which places both "early" Homo and H. naledi in the human holobaramin.  Mirror:  Archive.org ↩

  16. 16.

    O'Micks, J.  “Preliminary baraminological analysis of Homo naledi and its place within the human baramin.”  Journal of Creation Theology and Science Series B: Life Sciences 6:31–39.  2016.  O'Micks's first paper, on craniodental characters alone.  It placed naledi in the human baramin;  he reversed the conclusion later the same year.  Mirror:  Archive.org ↩

    1. 16a.

      Page 31:  “Neither data set was sufficient to demonstrate discontinuity among clusters conclusively.  However, in both analyses Homo naledi appears continuous with and fits well within the cluster of Homo species, despite some Australopithecus-like characteristics. … Placement of H. naledi in the human baramin is supported.  Since this is one of the first morphological dataset available for H. naledi and it includes only craniodental characteristics, further baraminological analysis is warranted.” ↩

  17. 17.

    O'Micks, J.  “Homo naledi Probably Not Part of the Human Holobaramin Based on Baraminic Re-Analysis Including Postcranial Evidence.”  Answers Research Journal 9:263–272.  2016.  Re-runs the baraminic analysis with postcranial characters added.  The first of O'Micks's five papers arguing that naledi is not human. ↩

  18. 18.

    Wood, Todd C.  “Taxon Sample Size in Hominin Baraminology: A Response to O'Micks.”  Answers Research Journal 9:369–372.  2016.  Wood's reply to O'Micks's postcranial re-analysis, on how taxon sampling affects the result. ↩

  19. 19.

    McLain, Matthew.  “Reply to O'Micks Concerning the Geology and Taphonomy of the Homo naledi Site.”  Answers Research Journal 10:55–56.  2017.  A young-earth palaeontologist answers O'Micks from the excavators' sedimentology and taphonomy. ↩

  20. 20.

    Paleobiology Database occurrence query for the tribe Hominini.  n.d.  Retrieved Aug 3, 2026.  An open, non-partisan compilation of fossil occurrence records, queried here for the whole tribe Hominini, and the figure used throughout this series.  The query returns 327 occurrence records in a database holding 2,001,053.  The two cautions, printed in the article body, are these: an occurrence is a taxon recorded at a collection, not a specimen, so the count is a census of localities; and the database substantially undersamples this group, because paleoanthropology publishes in its own journals and does not deposit occurrence data systematically.  Homo naledi does not appear at all, so these are floors, not totals.  The article uses it for the count of hominin occurrences, which it reports as a floor. ↩

  21. 21.

    Collard, Mark, and Bernard Wood.  “How reliable are human phylogenetic hypotheses?”  Proceedings of the National Academy of Sciences 97(9):5003–5006.  2000.  A genuine controlled experiment that tests craniodental cladistics, the method that sorts species into family-tree groups (clades) by shared features and is the standard method for placing fossil hominins.  That method had never been externally validated, because for fossils there is no independent answer to check against, so the authors applied identical methods to equivalent craniodental data from two groups of living higher primates, the hominoids (apes) and papionins (baboons, mangabeys and macaques), whose molecular phylogenies are known and robust.  The article uses it because it shows that skull-based trees can give strong support to relationships that the molecular trees reject. ↩

    1. 21a.

      Measured results.  The hominoid analysis correctly placed Hylobates as the earliest-branching hominoid, but grouped Homo with Pongo in one clade and Pan with Gorilla in the other, and the false Pan–Gorilla clade was supported by four more putative synapomorphies (shared features treated as evidence of common descent) than the true Pan–Homo clade.  In the bootstrap tests, a resampling check of how firmly each clade is supported, none of the clades supported at 70% or above was compatible with the molecular cladograms: a false (Gorilla, Pan, Pongo) clade at 95%; a false (Gorilla, Pongo) clade at 73%; in the papionins a false (Cercocebus, Macaca, baboon) clade at 98%; and from the hominoid qualitative data a clade incorrectly linking Gorilla and Pan at 92%. ↩

    2. 21b.

      Their conclusion: cladistic analyses based on standard craniodental characters cannot be relied on to reconstruct the relationships of hominoids, papionins,  “and, by extension, the fossil hominins”.  and, more problematically, such analyses can strongly support phylogenetic hypotheses that are misleading.  The scope limit matters: their conclusion is that craniodental cladistics is unreliable, not that morphology carries no information and not that the animals are unrelated.  They call for new approaches, not for abandoning phylogeny. ↩

  22. 22.

    Oxnard, Charles.  “[title to be added].”  Nature 258:389–395.  1975.  Oxnard's central paper, from 1975, predates the Lucy description, the Laetoli footprints, Selam, Kadanuumuu, and every CT-based study the article cites.  The article quotes nothing from it, and it treats his views as dated for that reason. ↩

  23. 23.

    Berger, Lee R. et al.  “Homo naledi, a new species of the genus Homo from the Dinaledi Chamber, South Africa.”  eLife 4:e09560.  2015.  A description of a new species of the genus Homo, Homo naledi, named by Lee R. Berger and colleagues from 1,550 fossil specimens in the Dinaledi Chamber of the Rising Star cave system in South Africa, representing at least 15 individuals.  The article uses it to set this large sample against the thin records of other species.  Mirror:  PMC

    1. 23a.

      Hominins were first recognised in the chamber in October 2013.  A short excavation recovered 1,550 hominin specimens, of which 737 are partial or complete anatomical elements, representing at least 15 individuals, with nearly every element of the skeleton repeated multiple times and some in articulation.  The paper calls it.  “the largest assemblage of a single species of hominins yet discovered in Africa.” ↩

  24. 24.

    McBrearty, Sally, and Nina G. Jablonski.  “First fossil chimpanzee.”  Nature 437:105–108.  2005.  A 2005 Nature paper describing three chimpanzee teeth from the Kapthurin Formation in Kenya.  The article uses it to show that the African ape fossil record is drastically poorer than the hominin record, though not empty.  Two explanations are live: a genuine taphonomic difference (how remains are preserved) between forest and lake-margin depositional settings, and collection bias toward hominins. ↩

  25. 25.

    Berean Archive.  The Order of the Fossil Record.  n.d.  An article in this project that assesses the hydrodynamic and ecological sorting models proposed to explain fossil order on their own terms.  The article uses it for the figure from that article's own Paleobiology Database work: across 2,001,053 occurrence records there are no primates in Early Cretaceous or older rock and no great apes in Eocene or older rock. ↩

  26. 26.

    White, Tim D.  “Palaeoanthropology: Five's a Crowd in Our Family Tree.”  Current Biology 23(3):R112–R115.  2013.  A 2013 article in Current Biology in which Tim White argues that the incentives of publishing, access, funding, career advancement, media promotion and celebrity work against the field's ability to self-regulate.  The article uses it for the complaint from inside the field: limited fossil material makes self-regulation worse, and there is ample motivation to generate new species names. ↩

  27. 27.

    Research dossier for this article.  n.d.  The research dossier kept for this article, with its list of sources that could not be reached. ↩