The argument arrives in a familiar shape. A creationist, often confident and sometimes well-read, asserts that the fossil record contains no genuine transitional forms between apes and humans, that the boundary between human and non-human is sharp and self-evident, and that genetics, taxonomy, and structural homology all confirm a discontinuity that evolution cannot bridge. The argument sounds, at first hearing, like it might be engaging with real science. In reality, it is a performance of certainty erected on a foundation that collapses the moment you ask a single, precise question: which fossils, exactly, fall on the human side of your line, and which fall on the ape side?
That question exposes a crisis at the heart of creationist palaeontology. The leading creationist authorities, from Answers in Genesis to the Discovery Institute to Creation Ministries International, have examined the same fossil specimens that evolutionary biologists have examined. They have access to the same published morphological data, the same endocranial volume measurements, the same limb-proportion analyses, the same stratigraphic context. And they reach contradictory conclusions. Some place Homo habilis among the humans. Others place it firmly among the apes. Some treat Homo erectus as fully human. Others regard it as a separate, non-human kind. Homo naledi, discovered in the Rising Star Cave system in South Africa in 2013 and 2015, has been sorted onto both sides of the creationist line by different creationist writers, sometimes within the same organisation across different publication dates.
This is not a minor dispute about classificatory convention. It is a structural refutation of the creationist claim. If a clear, objective boundary between human and non-human existed in the biological and fossil record, creationist authorities would agree on where it falls. The fact that they do not, that they sort the same physical specimens into different categories depending on which writer you consult, demonstrates conclusively that no such boundary can be reliably identified. The disagreement among creationists is itself the evidence. You cannot claim there is an obvious, unbridgeable discontinuity in the fossil record and simultaneously be unable to place the relevant specimens on the correct side of it.
This essay makes three connected arguments. First, that the australopithecine-to-Homo fossil record is gradational in the most straightforward empirical sense, with each successive species showing a mosaic of ancestral and derived characters that resists any clean categorical division. Second, that the internal contradictions of creationist classification prove, by the creationists’ own standards, that no sharp boundary exists. Third, that the molecular evidence, specifically human chromosome 2, the shared endogenous retroviruses embedded in the genomes of humans and other great apes, and the nested hierarchy of genetic similarity, corroborates the fossil record so thoroughly that the creationist position requires the simultaneous rejection of palaeontology, genetics, and molecular biology. These are not three separate pieces of evidence. They are three independent lines of enquiry pointing to the same conclusion.
1. What Creationists Actually Claim
Before examining the evidence, it is worth being precise about what the creationist argument actually asserts, because the claim has a sophisticated and a naive version, and they require somewhat different treatment.
The naive version is the one you encounter most often in popular creationist material: the fossil record contains no transitional forms between apes and humans, the intermediates predicted by evolutionary theory are simply absent, and palaeontologists have been searching for them for 150 years without success. This version is straightforwardly false, and its falsity can be demonstrated by pointing to a museum. The sophisticated version is more interesting. It concedes that there are fossil hominins with intermediate-looking features but insists that these specimens can be cleanly divided into two groups: fully ape (the australopithecines and related forms) and fully human (Homo species), with no genuine intermediates that would require common descent. The sophisticated version further argues that the anatomical similarities between humans and other great apes, and the genetic similarities, are explicable by common design rather than common ancestry.
Both versions share the same foundational assumption: that there is a clear, discoverable boundary between human and non-human biological forms, and that a dispassionate examination of the evidence will locate it. The argument then proceeds to claim that evolutionary palaeontology has failed to find any specimen that genuinely straddles this boundary, and that the specimens typically cited as transitional, habilis, ergaster, naledi, and others, are, on closer examination, either fully ape or fully human. The creationist’s confidence here is absolute. The boundary, they insist, is self-evident, the intermediates nonexistent, and the discontinuity robust. This is the claim we are going to test, not against the creationist’s worst arguments, but against their best ones. And the instrument we are going to use, at least initially, is not evolutionary biology at all. It is the creationist literature itself.
2. The Gradational Record from Australopithecines to Homo
The hominin fossil record, contrary to what creationist popular literature asserts, is one of the best-documented evolutionary sequences in palaeontology. It is not perfect, because no fossil record is perfect, and the preservation of terrestrial primates is always going to be incomplete given taphonomic realities. But the sequence from the early australopithecines through to anatomically modern humans is documented by thousands of specimens from dozens of sites across Africa, Europe, and Asia, spanning roughly seven million years of geological time.
The genus Australopithecus appears in the fossil record around four million years ago. The best-known species, Australopithecus afarensis, represented most famously by the partial skeleton known as Lucy (AL 288-1, discovered in Ethiopia in 1974), combines a small brain of roughly 400 to 500 cubic centimetres, comparable to a modern chimpanzee, with a postcranial skeleton that shows unambiguous evidence of bipedal locomotion. The pelvis is broad and bowl-shaped, far more similar to a modern human pelvis than to a chimpanzee pelvis. The knee joint has a valgus angle consistent with striding bipedalism. The foramen magnum, the hole at the base of the skull through which the spinal cord passes, is positioned further forward than in any quadrupedal ape, indicating an upright head posture. These features are not ambiguous in the slightest: afarensis walked bipedally through the East African savanna and woodland environment while retaining a brain size in the ape range and upper limb proportions that suggest some continued arboreal activity.
Hominins with significantly larger brains appear in the fossil record from roughly 2.5 to 2.8 million years ago, and the taxonomic boundary between the latest australopithecines and the earliest members of the genus Homo is precisely the region where the creationist argument about discontinuity is supposed to be most compelling. In reality, it is the region where the gradational nature of the record is most obvious. Australopithecus sediba, discovered at Malapa in South Africa in 2008, dates to approximately 1.98 million years ago and combines a small australopithecine-sized brain with a hand capable of a precision grip and legs adapted for an energy-efficient striding gait. Palaeoanthropologist Lee Berger and colleagues described it as exhibiting a unique mosaic of primitive and derived characters that illuminate the transition from Australopithecus to Homo. Whether sediba itself sits on the direct ancestral line to modern humans remains contested among specialists; what is not contested is that it occupies precisely the morphological territory between the genera that the creationist insists is empty.
Homo habilis, first described by Louis and Mary Leakey, Philip Tobias, and John Napier in 1964 from specimens found at Olduvai Gorge in Tanzania, has a brain volume ranging from roughly 510 to 680 cubic centimetres, substantially larger than the australopithecines but well below the modern human average of approximately 1,350 cubic centimetres. The hand bones associated with some habilis specimens show features consistent with a power grip and a precision grip combined, suggesting manipulative capacity beyond that of modern apes. Habilis is associated with the earliest Oldowan stone tool industry, simple but deliberately knapped flakes and choppers that represent a technological threshold no other non-human primate has crossed in the archaeological record. The species also retains long arms relative to its legs, a feature more consistent with the australopithecines than with later members of Homo. This combination, a larger brain, early tool use, and retained primitive postcranial proportions, is precisely what you would expect of a transitional form in the evolutionary sequence. It is also precisely what makes habilis so troublesome for creationist taxonomy, because the species does not fit cleanly into either of the two categories the creationist framework requires. It is too brainy and too tool-using to be simply an ape, and too small-brained and too primitively proportioned to be simply a human.
Homo erectus, which appears in the fossil record from roughly 1.9 million years ago and persists, in some populations, until perhaps 100,000 years ago, shows a further step along the same gradient. Brain volumes in erectus populations range from approximately 600 to over 1,100 cubic centimetres, with later specimens approaching the lower end of the modern human range. The postcranial skeleton of erectus is, in its proportions, essentially modern. The skull retains heavy brow ridges, a pronounced sagittal keel, a receding forehead, and a nuchal torus that are absent in anatomically modern humans, but the architecture below the neck is strikingly similar to our own. Homo erectus populations made the Acheulean hand axe, a bifacially worked tool of considerable sophistication, controlled fire, and spread out of Africa into Eurasia, reaching China and Southeast Asia by at least 1.6 million years ago.
Then there is Homo naledi, described by Lee Berger and colleagues in 2015 from nearly 1,550 fossils representing at least 15 individuals recovered from the Dinaledi Chamber in the Rising Star Cave system. Naledi presents a combination of features that forced a reassessment of simple linear models of hominin evolution. The skull is small, with a brain volume of roughly 465 to 560 cubic centimetres, comparable to the australopithecines or early habilis. The hands and feet are strikingly modern in overall proportions, better suited to manipulation and obligate bipedalism than those of any australopithecine. The shoulders and chest are more primitive, resembling earlier hominins in their overall configuration. Dates obtained from the Dinaledi Chamber specimens in 2017 placed them at approximately 236,000 to 335,000 years ago, meaning that a small-brained hominin with this mosaic of features was alive in Africa at the same time as archaic Homo sapiens. The deliberate deposition of bodies in the cave, now further supported by evidence from the Lesedi Chamber, suggests mortuary behaviour previously thought to be exclusively associated with modern cognition.
The picture that emerges from these specimens, and this is a very selective survey of a record that includes dozens of additional species and thousands of additional specimens, is not one of a clear boundary with humans on one side and non-humans on the other. It is a picture of gradual change in multiple anatomical systems, occurring at different rates in different lineages, producing a mosaic of characters that resists any simple binary classification. This is what evolutionary theory predicts. It is emphatically not what the creationist narrative predicts.
Stephen Jay Gould addressed the creationist misuse of the fossil record directly and with characteristic precision: “Since we proposed punctuated equilibria to explain trends, it is infuriating to be quoted again and again by creationists, whether through design or stupidity, I do not know, as admitting that the fossil record includes no transitional forms. Transitional forms are generally lacking at the species level, but they are abundant between larger groups.” The hominin fossil record, spanning the transition from small-brained, partially arboreal australopithecines through to large-brained, fully bipedal, tool-using anatomically modern humans, is precisely the kind of large-group transition where the fossil record is, by Gould’s own reckoning and by any objective palaeontological assessment, richly populated with exactly the intermediates the creationist claims are absent.
3. The Creationist Classification Disaster
Now we come to the argument that is, in some ways, more decisive than any amount of fossil evidence, because it turns the creationist’s own standards against them. If the boundary between human and non-human is real, objective, and scientifically identifiable, then creationist authorities examining the same evidence should reach consistent conclusions about which specimens fall on which side. They do not, and the divergence is not marginal.
Consider the fate of Homo habilis in the creationist literature. Answers in Genesis, the organisation founded by Ken Ham and one of the most influential creationist bodies in the English-speaking world, has at various times and through various authors placed habilis firmly in the ape category, arguing that the species is merely a variant of Australopithecus and that its larger brain size and tool associations do not elevate it to human status. This position is represented in articles published on their website by creationist anatomist David Menton, who argues that the skeletal proportions of habilis are sufficiently ape-like to exclude it from the human family. However, other creationist researchers, including some associated with the Creation Research Society, have placed habilis on the human side of the line, arguing that the evidence of tool use and the expanded brain volume warrant its inclusion within the human kind. The same physical specimens, the same measurements, the same stratigraphic associations: two directly contradictory categorical conclusions reached by researchers operating within the same theoretical framework.
The same disagreement applies to Homo erectus. Most mainstream creationist organisations treat erectus as fully human, often arguing that the specimens assigned to this taxon represent post-Flood populations of modern humans displaying normal intraspecific variation or the results of genetic bottlenecks after the dispersal from Babel. Ken Ham’s organisation has explicitly argued this position in multiple publications, treating Peking Man and Java Man as ordinary humans and suggesting that the skull features which distinguish erectus from modern humans fall within the range of normal human variation. This is an empirically extraordinary claim, given that the mean endocranial volume of early Homo erectus sits roughly 40 percent below the modern human average, but the claim is made with complete confidence. Yet other creationist researchers, approaching the same specimens, have been sufficiently troubled by the morphological differences to exclude erectus from the human category, treating it as a separate created kind that is now extinct and that was never human in the morally or theologically relevant sense.
Homo naledi has produced the most acute version of this classification crisis. When the species was first described in 2015, the small brain size led many creationist commentators to assign it to the ape category, with some dismissing it as simply an australopithecine variant or a pathological population of Homo habilis. When the dating results of 2017 revealed that the species lived within the past 335,000 years, well within the timeframe that most young-earth creationists assign to the post-Flood world and therefore to the period of human existence on Earth, the classification problem became acute. A species that existed during the time of humans, in Africa, that appears to have engaged in deliberate mortuary behaviour, and whose hands and feet show striking modern features, cannot be comfortably assigned to the pre-human ape category. But its tiny brain makes it equally uncomfortable in the fully-human category. Different creationist writers have responded differently, with some reclassifying it as human, some maintaining the ape classification, and some simply avoiding systematic engagement with the specimen altogether, which is itself a revealing methodological choice when the evidence is supposed to be self-evident.
The significance of this pattern cannot be overstated. The creationist argument rests on the claim that there is a genuine, objective, biologically real discontinuity between humans and non-humans. If that discontinuity were real and identifiable, then researchers examining the same specimens would converge on consistent classifications, subject only to the normal uncertainties of working with incomplete fossil material. Evolutionary palaeontologists do, in fact, reach considerable consensus about the broad outlines of hominin phylogeny, disagreeing principally at the level of species relationships and the precise topology of the family tree rather than about which general grade of organisation a specimen represents. Creationist researchers, by contrast, reach directly contradictory conclusions about the most fundamental question their framework requires them to answer: is this specimen human or not? The boundary they insist is obvious cannot be located by the people most committed to finding it.
One might respond that this merely reflects the inherent difficulty of the evidence, that even evolutionary biologists disagree about taxonomy, and that disagreement among creationists does not, by itself, refute the claim that a boundary exists. This objection has surface plausibility but fails on examination. Evolutionary biologists disagree about the precise species-level taxonomy and phylogenetic relationships within the hominin clade, just as they disagree about the species-level taxonomy of any diverse biological group. What they do not disagree about is the overall trajectory of change, the direction of the morphological gradient, or the existence of intermediate forms between australopithecines and modern humans. The disagreements concern the topology of the branches, not the existence of the tree. Creationist disagreements, by contrast, concern the most fundamental classification their entire argument requires: is this specimen human? That is not a dispute about fine-grained phylogenetic detail. It is a dispute about the very boundary the argument is supposed to demonstrate. When your own experts cannot agree on whether the key evidence falls on the human or the non-human side of your supposedly clear boundary, the honest conclusion is that the boundary is not there.
4. Human Chromosome 2 and the Genomic Signature of Common Ancestry
The fossil record is one line of evidence. The genome is another, entirely independent line, and it points to the same conclusion with a precision that palaeontology alone cannot achieve.
Modern humans have 46 chromosomes, arranged in 23 pairs. Our closest living relatives, the chimpanzees, gorillas, and orang-utans, have 48 chromosomes, arranged in 24 pairs. This difference initially seems like it might support the creationist claim of a sharp discontinuity: humans have a different number of chromosomes from other great apes, which might suggest we belong to a separate created kind. The problem for the creationist position is that the reason for this difference is written into the structure of our second-largest chromosome with extraordinary and unambiguous clarity.
Human chromosome 2 is the result of an end-to-end fusion of two ancestral chromosomes that, in their unfused form, still exist separately in the other great apes, corresponding to chimpanzee chromosomes 2A and 2B. The evidence for this fusion is multiple and independent. The first line of evidence is the presence of telomeric sequences, the repetitive DNA sequences (TTAGGG repeats) that normally cap the ends of chromosomes, in the interior of human chromosome 2, at precisely the location where the two ancestral chromosomes would have joined if they fused end-to-end. These internal telomeric sequences are a direct molecular signature of the fusion event and are not found in the interiors of non-fusion chromosomes in any comparable length or organisation. The second line of evidence is the presence of two centromeres, the regions of a chromosome where spindle fibres attach during cell division. Normal chromosomes have one centromere. Human chromosome 2 has one fully functional centromere and one vestigial centromere, partially inactivated by accumulated mutations, at precisely the position where the centromere of one of the ancestral chromosomes would be expected if the fusion model is correct. The third line of evidence is that the gene content and order along human chromosome 2 matches the gene content and order along chimpanzee chromosomes 2A and 2B almost perfectly, when the chimpanzee chromosomes are placed end-to-end in the orientation predicted by the fusion model.
This is not an inference from circumstantial evidence. It is a direct, structural, molecular demonstration that two ancestral chromosomes were fused in the hominin lineage after it diverged from the lineage leading to chimpanzees. The fusion event reduced the chromosome count from 48 to 46, and the molecular evidence of the fusion, the internal telomeric sequences and the vestigial centromere, is preserved in the genome of every living human being on the planet. This evidence was predicted by evolutionary theory before the human genome was sequenced in full. The prediction was that, if humans share common ancestry with the other great apes, we should be able to find a human chromosome that is the product of a fusion of two ancestral ape chromosomes, and that the fusion chromosome should contain internal telomeric sequences and a vestigial centromere at predictable positions. The prediction was confirmed in precise detail when the sequence was obtained. No alternative explanation has been offered by creationist biology that accounts for this pattern without invoking common descent or the theologically uncomfortable alternative of deliberately deceptive design by a creator who planted false evidence of ancestry throughout every human genome.
The creationist response, where one is offered at all, typically takes one of three forms. The first is to dispute the interpretation of the internal telomeric sequences, arguing that they could have been present for reasons other than a fusion event. This argument fails because internal telomeric sequences of this length and organisation are not found in comparable positions in the interiors of non-fusion chromosomes, and because the sequences are located precisely where the fusion model predicts they should be, not in some arbitrary location. The second response is to argue that the vestigial centromere is not actually vestigial, that it serves some undiscovered function. This argument does not engage with the evidence for fusion but merely asserts that the structure might have some other explanation, which is a promissory note rather than a refutation. The third response, which is what most popular creationist material does, is to dismiss the whole line of evidence without engagement, directing readers elsewhere and hoping they do not look too closely. Human chromosome 2 is, by itself, sufficient to establish common ancestry with the other great apes beyond any reasonable scientific doubt. But the chromosome fusion evidence is far from the only molecular argument available.
5. Endogenous Retroviruses and the Nested Hierarchy
The human genome contains approximately 8 percent retroviral sequences: the remnants of ancient viral infections that became integrated into the germ-line DNA of our ancestors and were subsequently inherited by all their descendants. These endogenous retroviruses (ERVs) are, for the most part, inactivated and harmless, though some are now known to have been co-opted for physiological functions including the formation of the placenta. Their interest for the question of human origins is not primarily functional but genealogical, and what they reveal about genealogy is decisive.
When a retrovirus integrates into the genome of a cell, the precise chromosomal location of integration is essentially random across billions of possible insertion sites in the mammalian genome. The probability that two independent retroviral integrations in two different organisms would produce insertions at precisely the same chromosomal location is therefore vanishingly small, well beyond the threshold of coincidence even in a single instance and effectively impossible when multiplied across hundreds of instances. When, however, two organisms share a common ancestor, and a retroviral integration event occurred in that ancestor’s germ-line, all of the ancestor’s descendants will carry the same ERV at the same chromosomal location, because they inherited it from the same source through ordinary biological inheritance.
Humans and chimpanzees share a very large number of ERVs at identical chromosomal locations. Humans and gorillas share a somewhat smaller number. Humans and orang-utans share fewer still. The pattern of shared ERVs produces a nested hierarchy that exactly mirrors the phylogenetic tree inferred from morphology, from chromosome structure, and from the comparison of functional gene sequences. This is precisely what evolutionary common descent predicts and precisely what independent creation of separate kinds does not predict. If humans and chimpanzees were separately created, there is no reason whatsoever why they should share broken, inactivated viral sequences at identical chromosomal locations. The probability of this occurring by independent chance is not merely small but effectively zero when considered across the full set of shared ERVs. The only parsimonious explanation is common ancestry.
The nested hierarchy of genetic similarity more broadly, extending beyond ERVs to the full set of shared gene sequences, pseudogenes, and non-coding elements, constitutes further powerful evidence. The degree of DNA sequence similarity between humans and chimpanzees, approximately 98.7 percent across comparable genomic regions (the figure is somewhat lower when insertions and deletions are included, settling around 95 percent for whole-genome comparisons, but the precise percentage does not alter the argument), is not in itself the most significant observation. What is most significant is that the pattern of similarities and differences, when mapped across multiple species, produces a perfectly nested hierarchy that exactly matches the phylogenetic tree predicted by evolutionary theory. Humans are most similar to chimpanzees, somewhat less similar to gorillas, less similar still to orang-utans, less similar again to gibbons, and so on outward through the primate order and beyond. This is the pattern produced by branching common descent with modification. It is not the pattern predicted by independent creation, where there is no principled reason why the degree of genetic similarity should track perfectly with the morphological and palaeontological evidence of relationship.
The creationist response to shared ERVs has evolved over the past two decades, and the evolution of that response is itself revealing. Early responses often denied that the shared sequences were retroviral in origin at all. As the genomic evidence accumulated and made this denial untenable, the response shifted to arguing that the shared ERVs serve functional purposes, implying that a common designer placed them there for a reason. This argument has become progressively less persuasive as genomic research has confirmed that the vast majority of shared ERVs between humans and chimpanzees are inactivated and non-functional, carrying the mutational signatures of long-term degradation rather than functional maintenance. Some ERVs have been co-opted for placental function, but these represent a small fraction of the total, and their co-option is itself consistent with the evolutionary expectation that natural selection would exploit available genetic material for new functions over millions of generations. The common-design response, even granting its most favourable interpretation, cannot account for the nested hierarchy of ERV sharing that precisely tracks the phylogenetic tree. A designer who placed specific ERVs in specific chromosomal locations in humans and chimpanzees but not in gorillas, and then placed a different subset of ERVs in humans, chimpanzees, and gorillas but not in orang-utans, would be producing a pattern that is indistinguishable from the pattern produced by branching common descent. At that point, the design hypothesis has made itself unfalsifiable and, as a consequence, scientifically vacuous.
For a broader discussion of how the molecular and morphological evidence converges on our shared ancestry with other primates, see our earlier essay on why “we did not evolve from monkeys” is a more complicated claim than it sounds.
6. The Taxonomy and Homology Arguments
The creationist argument sometimes takes a more philosophically sophisticated form, appealing to the distinction between homology and analogy in comparative anatomy and arguing that shared structures between humans and other apes are consistent with common design as readily as with common descent. This argument, associated particularly with Intelligent Design proponents at the Discovery Institute such as Stephen Meyer and Douglas Axe, deserves careful treatment because it is not obviously wrong on its own philosophical terms. The claim is that structural similarity alone does not prove genealogical relationship; architecturally similar structures could reflect a designer working with a common template, just as different aircraft share aerodynamic features because of the constraints of flight rather than because of genealogical descent from a common ancestor aircraft.
There are several problems with this argument, and they are not minor ones. The first is that it addresses morphological homology in isolation from the molecular evidence, which it cannot explain. The aeroplanes-share-features analogy breaks down entirely when applied to shared endogenous retroviruses and the chromosome 2 fusion, because there is no functional reason for a designer to place identical broken viral sequences at identical chromosomal locations in multiple species. The analogy between aeroplanes sharing wing structures, which is a functionally constrained design choice reflecting aerodynamic necessity, and humans sharing inactive retroviral insertions with chimpanzees at identical genomic positions, which reflects no engineering necessity whatsoever, is not a good one. The wing placement is required by physics; the ERV placement is required by nothing except genealogical inheritance.
The second problem is that the common-design argument predicts the wrong distribution of homologies. If structures are similar because a designer chose to use a similar template, then there is no reason why the degree of similarity should correlate with the degree of ecological or morphological relationship. A designer working freely could produce radically different degrees of similarity in any combination of species. But what we actually observe is the nested hierarchy: the degree of morphological, and much more precisely genomic, similarity tracks perfectly with the phylogenetic tree inferred from multiple independent lines of evidence. Humans are most similar to chimpanzees, next most similar to gorillas, and so on outward. This is not the prediction of common design working with a shared template. It is the signature of common descent with modification, and the distinction between those two explanations is not subtle.
The third problem is that the common-design argument cannot account for the distribution of vestigial structures and poor design elements in the human body. The recurrent laryngeal nerve descends from the brain through the chest to loop around the aorta before returning up to the larynx, taking a path several times longer than necessary and one that makes sense only as a relic of the ancestral fish anatomy in which the nerve ran a direct course to the gills before the vertebrate neck elongated and the heart descended into the thorax over evolutionary time. In a giraffe, this same unnecessarily looping path extends over several metres. These are not the signatures of optimal engineering; they are the signatures of evolutionary constraint, of a developmental process that cannot rewire itself wholesale but can only modify what is already present. Common design does not predict these structural absurdities. Common descent, with its requirement that structures be modified from existing forms rather than redesigned from scratch, predicts them precisely.
Richard Dawkins captures the logic of this point effectively in the context of the broader argument from comparative anatomy: “There is no doubt that whales and dugongs come with their dry-land history written all over them. If they had been deliberately created for the sea, they’d be very different, and a lot more like fish than they are. Animals that have their history written all over them are among the most graphic pieces of evidence we have that living things were not created for their present ways of life but evolved from very different ancestors.” The same principle applies with full force to human anatomy. Our bodies bear the marks of our evolutionary history in the vestigial structures we carry, in the sub-optimal design elements that make sense only as modifications of ancestral forms, and in the molecular signatures that link us to the other great apes beyond any reasonable doubt. These are not the hallmarks of special creation. They are the hallmarks of descent with modification operating under the constraint of what history has provided.
7. The “Kinds” Framework and Why It Cannot Define the Boundary
Creationist biology, particularly in its young-earth variant, typically organises biological diversity using the concept of “kinds” (baramin in the technical creationist literature, derived from the Hebrew min), which are supposed to represent the originally created units of biological diversity. The baramin concept has been elaborated by a small number of creationist researchers into a formal discipline called baraminology, which attempts to use statistical and analytical methods to identify the boundaries of created kinds. It is worth examining this framework in some detail, because it represents the creationist’s best attempt to provide a scientifically rigorous definition of the boundary their argument requires.
The baraminological literature on hominins is informative precisely because it is so inconsistent. Different baraminological analyses, using different methods and datasets, produce different conclusions about the number and membership of hominin baramin. Some analyses place all hominins, including the australopithecines, within a single baramin, which would mean that all fossil hominins from Australopithecus afarensis through to modern humans constitute a single created kind and that the variation observed in the fossil record represents variation within that kind. Others identify separate baramin for the australopithecines and for the genus Homo, placing the boundary between the genera. Still others identify Homo habilis as a possible bridge between baramin, which is baraminological code for acknowledging that the specimen does not fit cleanly into any single category the framework can accommodate.
The irony of the baraminological project, when applied to hominins, is that the more rigorously creationists try to make it work, the more it resembles standard cladistic analysis and the more it converges on similar conclusions: the hominin fossil record shows gradational change across morphological space, and the boundaries between species or grades are matters of degree rather than of kind. This is the conclusion that evolutionary palaeontology reaches by different methods. When an alternative analytical framework, one specifically designed to demonstrate the reality of created kinds and the absence of evolutionary continuity, converges on the same observation of gradational morphological change, the observation has been robustly corroborated by independent approaches rather than undermined by them.
There is a deeper philosophical problem with the kinds framework as well. The creationist claims that kinds are objectively real and biologically discoverable, that there is a fact of the matter about which organisms belong to the same kind. But the persistent disagreement among baraminologists about the membership of hominin baramin suggests that, if such a fact exists, the methodology is not capable of reliably discovering it. And if the methodology cannot reliably identify the boundary even within its own theoretical framework, then the claim that the boundary is real and self-evident collapses into empty assertion dressed in technical language.
The broader philosophical problem with the concept of kinds, as we explore in greater detail in our essay on the myth of kinds as a scientific concept, is that biological variation is continuous and hierarchically nested in exactly the way that common descent predicts and the kinds framework struggles to accommodate. The creationist needs kinds to be discrete, bounded, and mutually exclusive categories. Biology presents organisms that are continuous in their variation, graded in their morphological differences, and nested in hierarchical similarity patterns. The two pictures are not compatible, and the fossil record of human evolution is one of the clearest demonstrations of that incompatibility in the entire history of palaeontology.
8. What the Creationist Must Reject to Maintain the Position
It is worth pausing to take stock of what the creationist position requires, in terms of the rejection of established evidence, in order to maintain itself. This is not a rhetorical exercise. It is a genuine accounting of the epistemic cost of the position, and that cost is very high indeed.
The creationist must reject the palaeontological evidence: the thousands of hominin fossil specimens from dozens of sites across multiple continents, the consistent pattern of gradational morphological change from small-brained partially arboreal australopithecines through to large-brained fully bipedal anatomically modern humans, and the repeated demonstration by independent research teams working with different specimens and different analytical methods that this pattern is real and robust. This rejection requires either that the specimens have been systematically misidentified across the entire international community of palaeoanthropologists, that the radiometric dating methods are all systematically erroneous in ways that have somehow escaped detection by physicists and geologists for a century, or that the pattern is an illusion produced by confirmation bias operating simultaneously across thousands of researchers in dozens of countries with competing professional incentives. None of these alternatives has been seriously sustained in the creationist literature because none of them is capable of being sustained.
The creationist must reject the molecular evidence: the chromosome 2 fusion with its internal telomeric sequences and vestigial centromere, the shared endogenous retroviruses at identical chromosomal positions across the great apes, and the nested hierarchy of genomic similarity that precisely tracks the phylogenetic tree predicted by evolutionary theory. This rejection requires an explanation for why an independent designer would place the molecular signatures of common descent throughout the human genome without there being any common descent, a position which is not a scientific hypothesis but a claim that the evidence has been deliberately arranged to deceive, which raises theological problems of its own that creationist writers rarely address directly.
The creationist must also, as we have seen at length, explain why their own authorities cannot agree on the fundamental classification question that their argument requires to be answerable. If the boundary between human and non-human is real, objective, and scientifically identifiable, then why do the scientists most committed to identifying it reach contradictory conclusions about the most important specimens? This is not a challenge from outside the creationist framework. It is a challenge that the creationist framework generates internally, and it has no satisfactory answer within that framework.
The question is not whether evolutionary theory is without its own open problems and genuine uncertainties, because it is not, and no honest scientist would pretend otherwise. The question is whether the creationist alternative, the sharp boundary between human and non-human kinds, is supported by the evidence. And the answer that the evidence returns, including the evidence produced by creationist researchers themselves, is unambiguous: the boundary cannot be drawn, because there is no boundary to draw.
9. Steelmanning the Creationist: The Best Version of the Argument
Fair dealing requires engaging with the most sophisticated version of the creationist position, not merely the most popular one. The most sophisticated version, advanced by some Intelligent Design theorists rather than young-earth creationists, concedes most of the palaeontological record and much of the molecular evidence, and instead argues at the level of information theory and developmental complexity. The argument, associated principally with Stephen Meyer’s work in Darwin’s Doubt and Signature in the Cell, is that the appearance of genuinely novel body plans and cognitive capacities in the fossil record requires an infusion of specified complexity that undirected evolutionary processes cannot generate. The hominin fossil record, on this reading, shows not the gradual accumulation of small changes but the eventual emergence of distinctively human cognitive and cultural capacities, including consciousness, language, and symbolic thought, that represent a qualitative discontinuity not captured by the morphological measurements on which evolutionary palaeontologists focus their analyses.
This argument has several problems, and they are not minor ones. The first is that the claim of the sudden appearance of human cognitive capacities does not survive scrutiny of the archaeological record. The emergence of symbolic behaviour, language-related anatomy, and complex tool use in the hominin lineage is itself demonstrably gradational. The production of Oldowan tools by Homo habilis represents a threshold of manual and cognitive sophistication not found in any other primate lineage. The Acheulean hand axe tradition of Homo erectus represents a further advance, requiring the ability to hold a target shape in mind and work toward it through a sequence of planned steps across multiple knapping events. Evidence of pigment use and potentially symbolic marks appears in the African Middle Stone Age at sites dated to between 70,000 and 100,000 years ago. Fully modern behavioural complexity, including figurative art and long-distance trade networks, appears in the archaeological record of anatomically modern humans in Africa and Eurasia from roughly 40,000 to 50,000 years ago. This is a gradual and documented sequence of increasing behavioural complexity, not the sudden discontinuity that the Intelligent Design argument requires.
The second problem with the information-theoretic argument is that it has not been quantitatively demonstrated to the satisfaction of the relevant technical community. Meyer asserts that the information content of complex biological systems cannot be generated by mutation and natural selection, but the assertion rests on probability calculations that assume the space of functional sequences is vanishingly small relative to the total sequence space. This assumption has been challenged by the experimental protein evolution literature, which has shown that functional proteins are far more densely distributed in sequence space than Meyer’s calculations require, and that evolutionary searches through sequence space are far more efficient than random sampling. The argument depends on a premise that is actively contested within the relevant technical literature, and it is not appropriate to treat a contested premise as an established conclusion.
The third problem is that the information-theoretic argument, even if it were successful on its own terms, would not establish that the relevant complexity was provided by the God of any particular religion, by a designer of any specific character, or by an intervention at any specific point in human evolutionary history. It would establish only that some aspect of the development of complex biological systems is not yet fully explained by current evolutionary theory, which is a different and much weaker claim than its proponents typically present. The gap between “current evolutionary theory does not fully explain the origin of symbolic cognition” and “therefore a divine creator produced the human species in its current form, bearing a special moral status and an immortal soul” is not a logical step. It is a leap across a chasm of assumption, and no amount of sophisticated information-theoretic language makes the leap shorter or the landing less precarious.
10. The Deeper Stakes: What the Boundary Is Really About
The creationist insistence on a sharp boundary between humans and non-humans is not, at its root, a scientific claim. It is a theological one. The boundary is required not because the evidence demands it but because the doctrine requires it. If humans were created separately and distinctly, in the image of God, then human moral significance, the immortal soul, the special relationship with the divine, and the authority of the moral law derived from that relationship all have a secure metaphysical foundation. If humans are instead continuous with other animals, the product of the same unguided process of mutation and selection that produced chimpanzees and gorillas and every other organism on this planet, then that metaphysical foundation is, at best, complicated and, at worst, in need of reconstruction from entirely different materials.
This is the genuine anxiety driving the search for the boundary. The creationist is not examining the evidence and then reaching a conclusion about human origins through dispassionate enquiry. The conclusion about human origins has been reached on theological grounds, often long before the person encountered the scientific evidence, and the subsequent engagement with palaeontology and genetics is a defensive operation to protect a prior commitment. When the evidence does not cooperate, the evidence is rejected. When the evidence is ambiguous, the ambiguity is amplified and presented as a challenge to evolution rather than an honest acknowledgement of genuine scientific complexity. When creationist authorities disagree with one another about the classification of key specimens, the disagreement is not treated as evidence that the classification framework is inadequate but is quietly attributed to the inherent difficulty of working with fragmentary fossil material, a difficulty that evolutionary palaeontologists face equally and resolve with considerably more consistency.
None of this requires that the people making these arguments are personally dishonest. Many creationist researchers are sincere, intelligent, and genuinely believe that they are following the evidence wherever it leads. The problem is not personal dishonesty but a methodology that allows the conclusion to precede the investigation, which is not science regardless of how many technical terms are deployed in its service and how many peer-reviewed-looking journals are created to publish its results. An argument that starts from the required conclusion and works backward to find supporting evidence is not strengthened by rigour in the subsequent analysis. It is condemned at the level of methodology before the analysis even begins, and no amount of impressive-sounding morphometric data can rescue a project that is, at its foundation, apologetics dressed in scientific clothing.
The scientific consensus on human evolution is not a conspiracy, not a reflection of scientists’ personal philosophical commitments contaminating their professional conclusions, and not a case of an establishment suppressing inconvenient evidence. It is the result of 150 years of investigation by thousands of researchers from dozens of countries and scientific traditions, examining millions of specimens and millions of base pairs of genomic sequence, all pointing in the same direction. The consilience of independent lines of evidence, including fossil morphology, stratigraphy, comparative anatomy, chromosome structure, endogenous retroviruses, and the nested hierarchy of genetic similarity, is not the kind of consilience that emerges from shared error or coordinated deception. It is the kind that emerges when multiple independent methods of investigation encounter the same underlying reality and find, to a high degree of consistency, that reality behaving in precisely the way that evolutionary theory predicts it should.
11. Moral Seriousness About What We Are
There is a final point worth making, and it is not a scientific one but a philosophical one that deserves to be stated clearly rather than left as a subtext. The creationist often argues that evolutionary continuity with other animals implies a diminishment of human dignity, that if we are one branch of the primate family tree rather than a separate creation, we have no more moral significance than any other branch. This argument, which is designed to make evolutionary biology seem threatening to human values, rests on a confusion between genealogy and moral status.
The moral significance of a being does not derive from its ancestry. It derives from its capacities: for suffering, for reason, for relationship, for self-awareness, for the construction of meaning. Evolutionary biology tells us that these capacities developed gradually over millions of years of hominin evolution, from simple cognitive precursors in our ape ancestors through the increasingly sophisticated forms documented in the archaeological and fossil record. That they developed gradually, through an unguided process, does not make them less real, less morally significant, or less worthy of protection and celebration. A capacity for suffering is no less real, and no less morally relevant, because it evolved rather than was installed by divine fiat. A capacity for reason is no less extraordinary because it has antecedents in other primate cognition.
What evolutionary biology does tell us is that the sharp categorical boundary between the morally significant human and the morally negligible animal is harder to draw than theological tradition assumes, and that this difficulty is not a problem to be solved by better creationist taxonomy but an invitation to think more carefully about the actual grounds of moral consideration. If what matters morally is the capacity for suffering and the capacity for something like self-directed life, then those capacities are distributed across the animal kingdom in varying degrees, and our moral responsibilities extend further than a theology of special creation would suggest. This is not a diminishment of human dignity. It is an extension of moral seriousness to the other beings who share our ancestry, our genome, and, in varying degrees, our capacity for experience.
The boundary that creationists cannot draw, the sharp line between human and non-human in the fossil record and in the genome, is not a gap in the evidence for evolution. It is a gap in the creationist argument, and it is a gap that the creationist’s own literature, by its internal contradictions, has widened rather than closed. The evidence, examined without a predetermined conclusion, tells a coherent and deeply compelling story: a story of gradual change, of mosaics of ancestral and derived features, of populations slowly diverging over millions of generations, of a lineage that eventually produced beings capable of asking, with extraordinary sophistication, where they came from. The answer that biology provides is, when you sit with it honestly, more remarkable than the alternative. We are not separately created beings parachuted into a world that belongs to us by divine decree. We are the improbable survivors of an ancient and indifferent process, and we are, so far as the evidence currently suggests, the only part of that process that has ever looked back at itself and tried to understand what it is.
The line cannot be drawn because there is no line. There is only the record of life, changing through time, branching and diverging, producing extraordinary complexity out of simpler forms across depths of geological time that the human mind struggles to encompass in their full magnitude. The australopithecines, the habilines, the erectines, the Neanderthals, and the anatomically modern humans are not separate creations arranged neatly on either side of a theological boundary. They are chapters in a single continuous story, and we are, for the moment, the most recent sentence written in it.
References
Berger, L. R. et al. “Homo naledi, a new species of the genus Homo from the Dinaledi Chamber, South Africa.” eLife, 2015.
Berger, L. R. et al. “New Homo naledi fossils, dating, and the depositional context of the Rising Star Cave.” eLife, 2017.
Coyne, J. A. Why Evolution Is True. Oxford University Press, 2009.
Dawkins, R. The Blind Watchmaker. Longman Scientific and Technical, 1986.
Gould, S. J. “Evolution as Fact and Theory.” Hen’s Teeth and Horse’s Toes. W. W. Norton, 1983.
Ijdo, J. W. et al. “Origin of human chromosome 2: an ancestral telomere-telomere fusion.” Proceedings of the National Academy of Sciences, 88(20), 9051-9055, 1991.
Johanson, D. and Edey, M. Lucy: The Beginnings of Humankind. Simon and Schuster, 1981.
Meyer, S. C. Darwin’s Doubt. HarperOne, 2013.
Singer, P. Animal Liberation. New York Review of Books, 1975.
Venema, D. and McKnight, S. Adam and the Genome. Brazos Press, 2017.
Wood, B. and Collard, M. “The human genus.” Science, 284(5411), 65-71, 1999.