Thurnam’s 137 Skulls: The Forgotten People Behind Britain’s Megaliths.

Posted on July 30, 2026 by admin

Long barrows, long skulls; round barrows, round skulls—and a biological division archaeology has known about for more than 150 years

1. The evidence was never lost—it was simply left behind

More than 150 years ago, Dr John Thurnam assembled an extraordinary body of physical evidence from Britain’s prehistoric burial monuments. He did not study a single unusual skeleton and construct a theory around it. He compared large groups of human skulls recovered from two archaeologically distinct types of monument: the earlier long barrows and the later round barrows.

His conclusion became one of the most memorable statements in British archaeology:

“Long barrows, long skulls; round barrows, round skulls.”

Thurnam's 137 Skulls: The Forgotten People Behind Britain's Megaliths. — Thurnam’s 137 Skulls: The Forgotten People Behind Britain’s Megaliths.

The saying survives, but the scale of the evidence behind it has largely disappeared from public discussion.

Thurnam’s two studies contain measurements from 67 long-barrow crania and 70 round-barrow crania—a combined sample of 137 prehistoric skulls. The long-barrow series had a combined average cranial index of approximately 70.5. The round-barrow series averaged approximately 81.

A cranial index is calculated by dividing the maximum breadth of a skull by its maximum length and multiplying by 100. A low index indicates a relatively long, narrow skull. A higher index indicates a shorter, broader skull.

This does not mean that one skull was literally twice the length of another. It describes the proportional relationship between length and breadth. Nevertheless, a difference between population averages of 70.5 and 81 is substantial.

More importantly, the skull measurements did not exist in isolation. They followed changes in monument form, burial practice and associated material culture.

Thurnam classified the long barrows as the earlier monuments. Their primary deposits usually contained whole or disarticulated skeletons, often gathered together at the broad end of the mound. Metal objects were absent from the original burials, and cremation was rare and incomplete.

The round barrows belonged to a later archaeological tradition. Cremation was far more common, bronze objects appeared, and individual burials became increasingly prominent. Thurnam regarded the contrast as evidence of two different populations or chronological communities rather than a mere change in architectural fashion.

Victorian researchers surrounded these measurements with racial labels and social interpretations that cannot simply be carried into modern science. Thurnam’s ideas about chiefs, slaves, sacrifice and cannibalism must be treated as historical hypotheses, not proven explanations.

But rejecting his Victorian terminology does not make his measurements disappear.

A skull measured at a cranial index of 68 remains proportionally long, whether it was measured in 1869 or yesterday. The correct response is to remeasure the surviving specimens, establish their precise archaeological contexts, radiocarbon-date them and obtain ancient DNA.

Instead, modern accounts commonly reduce the question of Britain’s megalithic builders to a simple label: early farmers.

That label may describe an economy, but it does not identify biological ancestry, monument designers, engineers or the social group whose dead were selected for burial inside the monuments.

Farming can be adopted. Domestic animals can be traded. Cultivation can spread between neighbouring populations. Finding agricultural practices in a landscape does not automatically prove that every monument in that landscape was conceived and built by a newly arrived population.

Thurnam’s evidence therefore raises a question that archaeology has never adequately resolved:

Why did the people buried in long barrows possess a consistently elongated cranial form, while the people buried in later round barrows possessed much broader skulls?

That is not a Victorian curiosity. It is a modern research question waiting to be reopened.

2. The 67 long-barrow skulls: what Thurnam actually measured

Thurnam separated long barrows into two principal forms: unchambered long barrows and chambered long barrows.

The unchambered monuments were generally enormous earthen mounds with lateral ditches and primary burials concentrated beneath the broader and higher end. The chambered monuments incorporated stone-built compartments containing multiple human remains.

From ten unchambered long barrows, Thurnam obtained 27 measurable crania. Their indices ranged from 63 to 75, with an overall average of only 69. He described them as remarkably long and narrow.

The lowest value, 63, came from the nine-skull series at Norton Bavant. This was an exceptionally elongated skull even by the standards of the wider long-barrow sample.

Unchambered long-barrow crania

Site	Measured skulls	Index range	Mean index
Winterbourne Stoke	1	75	75
Tilshead East	5	68–74	71.5
Bowl’s Barrow, Heytesbury	4	65–70	67
Fyfield, Giant’s Grave	1	69	69
Tilshead Lodge	2	66–68	67
Figheldean	1	67	67
Netheravon	1	69	69
Tilshead Old Ditch	1	68	68
Stonehenge Long Barrow 165	2	70.5–71	71
Norton Bavant	9	63–73	68.5
Total	27	63–75	69

The table is striking because the pattern is not confined to one cemetery. It appears across multiple Wiltshire long barrows excavated at different times.

The four skulls from Bowl’s Barrow averaged 67. The two from Tilshead Lodge averaged 67. The single measurable examples from Figheldean, Netheravon and Tilshead Old Ditch produced indices of 67, 69 and 68 respectively.

Thurnam’s 27-skull unchambered series included 21 skulls he regarded as male and six as female. Although Victorian sex assessments require modern checking, this indicates that the result was not created by measuring only one sex.

The chambered long barrows produced another 40 measurable crania from seven sites.

Chambered long-barrow crania

Site	Measured skulls	Index range	Mean index
Uley	2	71–74	72.5
Littleton Drew	7	68–74	71
West Kennet	4	67–73	70
Nympsfield	2	74–75	74
Rodmarton	5	71–74	73
Charlton Abbot’s	17	68–77	71
Oldbury	3	68–74	71
Total	40	67–77	71.5

The chambered sample is slightly broader on average than the unchambered series, but it remains predominantly elongated.

West Kennet produced four skulls ranging from 67 to 73, averaging 70. Littleton Drew produced seven skulls averaging 71. Charlton Abbot’s supplied the largest chambered sample, with 17 skulls and a mean of 71.

We must not claim that every one of these 67 people was strictly dolichocephalic. Thurnam’s published tables provide site ranges and averages rather than all 67 individual measurements. At sites such as Charlton Abbot’s, where the range reaches 77, some individuals were evidently broader-headed.

What can be stated securely is that:

the combined long-barrow population averaged approximately 70.5;
every site series had a long-headed, sub-long-headed or predominantly long-headed average;
the strongest unchambered series were extraordinarily elongated;
the pattern occurred across both earthen and chambered long-barrow traditions.

This is not the result of selecting two spectacular museum skulls. It is a regional population pattern involving 67 measured crania from 17 monumental burial sites.

3. The 70 round-barrow skulls: the population changes

The round-barrow evidence is even more dramatic because Thurnam published a categorical distribution for the complete sample.

He assembled 70 skulls from later round-barrow burials. Their cranial indices ranged from 74 to 89, with an average close to 81.

Among those 70 skulls, Thurnam found:

no dolichocephalic examples;
no sub-dolichocephalic examples;
12 intermediate skulls with indices of 74–76;
14 broader skulls with indices of 77–79;
44 brachycephalic skulls with indices of 80–89.

Distribution of the 70 round-barrow skulls

Cranial category	Index range	Number	Proportion
Dolichocephalic	70 or below	0	0%
Sub-dolichocephalic	71–73	0	0%
Orthocephalic	74–76	12	17%
Sub-brachycephalic	77–79	14	20%
Brachycephalic	80–89	44	63%
Total	74–89	70	100%

The significance is not that every round-barrow person had an almost spherical skull. The majority were broad-headed, while a minority occupied an intermediate range.

The critical result is that not one of the 70 entered Thurnam’s long-headed or sub-long-headed classes.

The complete contrast

Burial population	Measured crania	Approximate mean index	Long/sub-long skulls
Long barrows	67	70.5	Predominant, but exact individual total unavailable
Round barrows	70	81	0 of 70
Later secondary burials inside long barrows	12	about 79	Predominantly broader-headed

The third row is particularly important.

Long barrows were sometimes reopened and reused by later people. Thurnam examined 12 skulls from secondary burials inserted into Wiltshire long barrows. Their average index was approximately 79.

Some were accompanied by later pottery, including decorated drinking cups. These were not part of the original long-barrow burial population.

This produces a remarkably controlled comparison.

Inside the same monument:

the primary burials were long-headed;
the later inserted burials were substantially broader-headed.

Therefore, the difference cannot easily be dismissed as the effect of soil, regional geography, monument location or measurement technique. The cranial form changes with the archaeological phase.

The same point is illustrated by individual named round-barrow skulls.

Ten named Wiltshire round-barrow examples

Site	Cranial index
Kennet Hill	74
Morgan’s Hill	75
Stonehenge, Barrow 150	78
Stonehenge, Collection 266	79
Wilsford	80
Winterbourne Stoke	81
Roundway	83
Codford	83
Stonehenge, Barrow 150	84
Stonehenge, Collection 265	84

These ten had an average index of approximately 80.1.

The burial evidence changes as well. Thurnam’s long barrows were dominated by inhumation and collective deposits. Cremation occurred only exceptionally. In the later circular barrows, cremation was far more common and bronze artefacts became part of the archaeological record.

The evidence therefore points to more than architectural development.

We see simultaneous changes in:

skull proportions;
treatment of the dead;
collective versus individual burial;
cremation frequency;
monument design;
associated technology.

Whether this represents migration, population mixing, social selection or long-term biological change remains open to investigation.

What is not scientifically reasonable is to pretend that no physical distinction existed.

4. From Cro-Magnon to the long-barrow population

The elongated cranial form did not originate in Britain’s long barrows.

Long, narrow skulls are present much earlier in the European Upper Palaeolithic record. Specimens traditionally associated with Cro-Magnon, Barma Grande, Grotte des Enfants, Chancelade, Předmostí and other Upper Palaeolithic sites frequently possess substantial front-to-back cranial length.

“Cro-Magnon” is a historical label for early European Homo sapiens, which was once seen as a separate modern species. Nevertheless, the term remains useful when discussing a recognisable Upper Palaeolithic pattern of tall, robust bodies and large, often elongated crania.

The 2025 European cranial study by Pavel Grasgruber provides modern measurements for many of these specimens. It identifies considerable variation rather than a single universal Cro-Magnon type, but it also confirms that several Upper Palaeolithic individuals possessed conspicuously tall or narrow cranial forms.

Barma Grande 5 and Chancelade 1 stand out as tall, narrow cranial outliers. Early western European specimens including Cro-Magnon 1 and Grotte des Enfants 4 show differences from some central and eastern European Upper Palaeolithic skulls, although the study stresses that limited sample size prevents a simple biological classification.

Our own matched body-and-cranium database includes several relevant individuals:

Upper Palaeolithic/Mesolithic specimen	Reconstructed stature	Cranial index
Barma Grande 5	187.0 cm	68.6
Grotte des Enfants 4	185.7 cm	76.8
Předmostí 3	183.0 cm	71.8
Sungir 1	183.2 cm	76.6
Oberkassel 1	175.1 cm	72.7
Romito 4	173.2 cm	74.9
Bichon 1	169.9 cm	75.8
Chancelade 1	165.7 cm	69.6

Note: Villabruna 1 and M50 provide two complementary pieces of evidence. Villabruna 1 directly connects early R1b1 with a non-round-headed cranium, having a cranial index of 75.97. M50 directly connects basal R1b with exceptional stature, having a published estimate of 181.63 centimetres. The complete combination of early R1b, exceptional stature and strongly long-headed cranial morphology has not yet been demonstrated in one securely documented individual, but both components of the working hypothesis are now represented within the early R1b record.

This demonstrates two important points.

First, an elongated skull does not automatically indicate exceptional stature. Chancelade 1 was strongly long-headed but comparatively short, while Grotte des Enfants 4 was very tall despite having a broader cranial index than Barma Grande 5.

Second, the combined pattern of tall stature, robust anatomy and elongated cranial form unquestionably existed among some Upper Palaeolithic European populations.

The central question is whether Britain’s long-barrow population represents biological continuity from these older European hunter-gatherers.

Cranial morphology is substantially inherited and has long been used to investigate biological affinity, population continuity and descent. Individual skulls may be affected by development, nutrition, cultural practices or burial pressure, but these factors cannot reasonably explain away a repeated population-level pattern involving 67 measured crania from 17 long-barrow monuments.

The evidence becomes still stronger because the cranial form changes with the archaeological phase. The primary long-barrow burials were predominantly long-headed, while the later secondary burials inserted into the same monuments were substantially broader-headed. The later round-barrow population was broader-headed again.

This is not a random collection of unusual skulls. It is a structured biological pattern associated with different burial populations and different archaeological periods.

The working continuity hypothesis is therefore:

Britain’s long-barrow population retained a substantial biological inheritance from older European hunter-gatherer populations, including the elongated cranial form visible among parts of the Upper Palaeolithic population traditionally described as Cro-Magnon.

This does not require every long-barrow individual to have been identical to every Upper Palaeolithic specimen. Ancient populations contained variation, just as modern populations do. The argument concerns biological continuity across populations, not the claim that one unchanging physical type survived for tens of thousands of years.

The conventional narrative proposes that Anatolian-derived farming populations migrated progressively across Europe and eventually reached Britain. However, genetic resemblance alone does not establish that migration route, and our analysis of more than 14,000 calibrated radiocarbon records has not identified the archaeological trail that such a mass population movement should have produced.

There is no coherent sequence of increasingly later settlement sites running from Anatolia through southeastern Europe, central Europe, France and finally Britain. Nor is there the expected concentration of early sites along the proposed migration corridor. Instead, the radiocarbon evidence shows extensive activity already present in northwestern Europe during the period in which the supposed farmer migration is claimed to have occurred.

Genetic components described as Anatolian-related may reflect shared ancestry, contact, intermarriage or limited gene flow. They do not automatically prove that a large farming population crossed Europe, replaced indigenous communities or constructed Britain’s monumental landscape.

It is therefore entirely possible that people later classified archaeologically as “farmers” were substantially descended from older European hunter-gatherer populations who adopted cultivation, livestock and new material practices without being biologically replaced.

The important distinction is this:

Agriculture describes what people did. Cranial morphology helps establish who they descended from, while radiocarbon evidence tests whether the claimed migration actually occurred.

The long-barrow measurements therefore support biological continuity with Europe’s older inhabitants, while the radiocarbon record provides no corresponding settlement trail for the conventional Anatolian farmer migration narrative

5. The R1b question: what the archaeological record permits us to predict

R1b is not a blood group. It is a Y-chromosome haplogroup passed primarily through the paternal line.

The traditional popular narrative often associates the major expansion of R1b in western Europe with Bronze Age Steppe-related or Bell Beaker populations after approximately 2500 BC.

That explanation may describe a major later expansion of particular R1b branches, especially those descending from R1b-M269.

It cannot mean that R1b itself was absent from Europe before the Bell Beaker period.

Ancient DNA has now identified R1b among European hunter-gatherers thousands of years before the proposed Bell Beaker migration.

Villabruna 1: R1b in Upper Palaeolithic Italy

The earliest confirmed R1b individual currently known is Villabruna 1, discovered at Riparo Villabruna in northern Italy.

Villabruna 1:

was an adult male;
belonged to the Late Upper Palaeolithic Epigravettian population;
was directly dated to approximately 14,000 years ago;
carried the early paternal lineage R1b1, or R-L754;
had estimated stature results ranging from approximately 167.2 to 169.1 centimetres.

The average stature reported for European Late Upper Palaeolithic males was approximately 165.6 centimetres. Villabruna 1 therefore fell within the taller part of the known male range for his period.

Villabruna 1 places R1b in western Europe around nine thousand years before the conventional Bell Beaker horizon.

His discovery separates two questions that are too often confused.

The first is when R1b originally appeared in Europe.

The second is when particular later branches of R1b expanded and became dominant.

The expansion of some R1b branches during the Bronze Age cannot be presented as the first appearance of the wider R1b lineage in Europe.

M50: a tall Mesolithic R1b hunter-gatherer

The strongest direct body-and-DNA crossover currently in our database is I4655, also recorded as SCCL_50 and M50, from Schela Cladovei on the Romanian bank of the Danube.

This individual:

belonged to basal R-L754/R1b;
dates to approximately 7059–6571 calibrated BC;
had a right tibia measuring 412 millimetres;
received a published stature estimate of 181.63 centimetres.

This is direct evidence that an exceptionally tall Mesolithic European man carried basal R1b roughly four thousand years before the conventional Bell Beaker horizon.

M50 is particularly important to our investigation because his stature fits the physical pattern being examined.

The evidence already shows that some Upper Palaeolithic European populations possessed tall stature, robust skeletal proportions and elongated cranial forms.

M50 now demonstrates that exceptional stature was also present in at least one securely identified early R1b hunter-gatherer.

It does not prove that M50 possessed an elongated skull, because his individual cranium has not yet been securely matched to published cranial measurements.

Nor does it prove that Britain’s long-barrow population was predominantly R1b.

What it proves is that R1b existed among European hunter-gatherers early enough to have formed part of the biological inheritance from which later long-barrow populations could have developed.

Villabruna 1 establishes the presence of R1b in Upper Palaeolithic western Europe.

M50 demonstrates the continued presence of basal R1b among Mesolithic hunter-gatherers and provides a direct association with exceptional male stature.

The responsible hypothesis is therefore:

Early R1b paternal lineages were present among European hunter-gatherers long before the Bell Beaker period, and Britain’s long-headed long-barrow population may have included descendants carrying those lineages.

That is a prediction capable of being tested.

It is not yet a universal conclusion.

The evidence ladder

Statement	Present status
R1b existed in Upper Palaeolithic western Europe	Directly demonstrated by Villabruna 1 approximately 14,000 years ago
Villabruna 1 was relatively tall for his period	Supported by published stature estimates of approximately 167.2–169.1 centimetres
Basal R1b existed among Mesolithic European hunter-gatherers	Directly demonstrated
Some early R1b individuals were exceptionally tall	Directly demonstrated by I4655/M50, estimated at 181.63 centimetres
Long-headed European populations existed before the Neolithic	Directly demonstrated by Upper Palaeolithic and Mesolithic cranial measurements
Britain’s primary long-barrow burials were predominantly long-headed	Strongly demonstrated by Thurnam’s series of 67 crania
Britain’s long-barrow population included males carrying R1b	Plausible and directly testable, but not yet demonstrated across the population
All long-headed prehistoric people carried R1b	Not demonstrated
All early R1b individuals were tall or long-headed	Not demonstrated
R1b first appeared in Europe with Bell Beaker migrants	Contradicted if interpreted as the first European appearance of the wider R1b lineage
Particular later R1b branches expanded during the Bronze Age	Possible, but separate from the earlier presence of basal R1b

The absence of a securely identified round-headed early R1b individual in our present register strengthens the working association between early R1b, greater stature and elongated cranial form.

However, the absence of contradictory evidence is not the same as proof.

The archaeological record currently presents two independently demonstrated patterns.

The first is the presence of R1b paternal lineages among Upper Palaeolithic and Mesolithic European hunter-gatherers.

The second is the persistence of elongated cranial morphology from older European populations into Britain’s primary long-barrow burial population.

The working hypothesis is that these two patterns may intersect.

Britain’s long-headed long-barrow population may have included descendants of older European hunter-gatherers carrying early branches of R1b.

This is not an attempt to project the later Bell Beaker model backwards into an earlier period.

It is a testable prediction based on:

the confirmed presence of R1b in Upper Palaeolithic Italy;
the confirmed presence of basal R1b among Mesolithic Danube hunter-gatherers;
the exceptional stature of M50;
the existence of elongated cranial forms among older European populations;
the predominantly long-headed character of Britain’s primary long-barrow burials.

Aveline’s Hole: the missing British crossover

A third and potentially critical connection comes from Aveline’s Hole in Somerset, Britain’s largest known Early Mesolithic cemetery.

The male genetic sample I3004, also recorded as SB 337B3, came from a human tibia directly dated to approximately 8600–8300 BC.

The current genetic annotation assigns this individual to an R1b-derived paternal lineage. This would place R1b in Britain more than five thousand years before the conventional Bell Beaker horizon.

Aveline’s Hole also produced directly measured long-headed crania.

The reconstructed skull M1.11.301 had a maximum cranial length of 186 millimetres and a maximum breadth of 134 millimetres. Its cranial index was 72.0, placing it securely within the dolichocephalic, or long-headed, category.

The surviving evidence does not presently demonstrate that the R1b tibia and the long-headed skull belonged to the same person. They must therefore not be combined as though they represent a single securely reconstructed individual.

Nevertheless, their occurrence within the same tightly dated Early Mesolithic burial population is highly significant.

Aveline’s Hole provides the missing geographical and chronological bridge between the other two early R1b findings.

Villabruna 1 demonstrates that R1b1 was present in Upper Palaeolithic Italy approximately fourteen thousand years ago and was carried by an individual with a non-round-headed cranium.

M50 at Schela Cladovei demonstrates that basal R1b was carried by an exceptionally tall Mesolithic man whose published stature estimate was 181.63 centimetres.

Aveline’s Hole places an R1b-derived paternal lineage within a British Early Mesolithic cemetery where securely long-headed individuals were also buried.

These are three complementary observations:

Individual or site	Genetic evidence	Physical evidence	Present status
Villabruna 1, Italy	Early R1b1/R-L754	Cranial index 75.97; non-round-headed	Individual genetic and cranial crossover
Aveline’s Hole, Britain	I3004 provisionally assigned to an R1b-derived lineage	At least one cemetery cranium had an index of 72.0	Site-level crossover; genetic call requires confirmation
M50, Schela Cladovei	Basal R-L754/R1b	Published stature estimate of 181.63 cm	Individual genetic and stature crossover

The three findings cannot yet be merged into a claim that one securely identified person possessed R1b, exceptional stature and an extremely long skull.

They do, however, substantially strengthen the prediction that these characteristics intersected within the older European population.

The evidence now shows:

early R1b associated with a non-round-headed individual in Upper Palaeolithic Italy;
a provisional R1b male within a British Mesolithic cemetery containing long-headed crania;
and basal R1b associated with exceptional stature in Mesolithic southeastern Europe.

The Aveline’s Hole result is particularly important because it places the proposed association directly within Britain thousands of years before the construction of the long barrows.

Its current Y-chromosome assignment carries a technical quality warning and must therefore remain provisional. However, it is no longer an isolated result without archaeological context. It occurs within a burial population whose surviving cranial evidence independently includes the long-headed form predicted by the hypothesis.

Revised evidence ladder

Statement	Present status
R1b existed in Upper Palaeolithic western Europe	Directly demonstrated by Villabruna 1
An early R1b1 individual possessed a non-round-headed cranium	Directly demonstrated by Villabruna 1
Basal R1b existed among Mesolithic European populations	Directly demonstrated by M50 and other early samples
Some early R1b individuals were exceptionally tall	Directly demonstrated by M50
Long-headed people were buried at Aveline’s Hole	Directly demonstrated by cranial measurements, including an index of 72.0
An R1b-derived male was buried at Aveline’s Hole	Provisionally supported by I3004, but the contamination-flagged genetic call requires confirmation
The Aveline R1b male was the measured long-headed individual	Not demonstrated; the DNA sample came from a tibia that has not been matched to the measured cranium
Early R1b and long-headed morphology occurred within the same British Mesolithic cemetery	Supported at site-population level
Britain’s long-barrow males carried R1b	Strengthened as a testable prediction, but not yet directly demonstrated
R1b first appeared in Europe with Bell Beaker migrants	Contradicted if interpreted as the first European appearance of R1b

The next scientific step

The surviving long-barrow skulls identified by Thurnam should be relocated in museums and archaeological collections.

Their original labels, monument locations, chamber positions, burial phases and excavation histories should be reconciled with modern catalogues.

Where preservation allows, the research programme should include:

direct radiocarbon dating;
renewed cranial measurement;
three-dimensional scanning;
ancient-DNA sampling;
Y-chromosome identification;
mitochondrial identification;
isotope testing;
comparison with the known prehistoric body-size records.

The results should then be compared by individual.

Cranial form, stature, skeletal proportions, genetic lineage, burial position and archaeological date should no longer be studied as separate categories of evidence.

A confirmed group of primary long-barrow males possessing elongated skulls, exceptional stature and early R1b paternal lineages would provide the missing biological connection between Upper Palaeolithic populations, Mesolithic hunter-gatherers and Britain’s megalithic communities.

Until that work is carried out, the R1b connection remains a strong research hypothesis—not a licence to replace one simplistic migration narrative with another.

What can already be stated with certainty is that R1b did not first appear in Europe with the Bell Beaker phenomenon.

It was present in Upper Palaeolithic Italy approximately fourteen thousand years ago.

It survived among Mesolithic European hunter-gatherers.

And at least one of those securely identified Mesolithic R1b men stood approximately 181.63 centimetres tall.

6. Our new method: estimating a person’s height from the skull

Long-barrow excavations often preserved skulls while postcranial bones were lost, mixed, damaged or separated from their original individuals.

That creates an obvious problem. Traditional stature reconstruction depends primarily on complete long bones such as the femur and tibia. Without those bones, the person’s height is usually left unknown.

Our project has therefore begun testing whether external cranial dimensions can provide a preliminary stature range.

The method is based on adult male prehistoric individuals for whom both cranial measurements and independently reconstructed body heights survive.

The three principal skull measurements are:

M1: maximum cranial length;
M8: maximum cranial breadth;
M17: cranial vault height.

These three dimensions are multiplied to produce an external cranial-size proxy. The cube root then converts that volume-like product back into a linear measurement.

Current three-dimensional model

Estimated stature in centimetres:

89.11 + 0.5537 × ∛(M1 × M8 × M17)

This is not a measurement of actual brain volume. It is an external three-dimensional cranial-size proxy.

The current calibration sample contains 14 securely or probably matched adult males.

Specimen	Known stature	M1	M8	M17
Barma Grande 5	187.0	204	140	153
Grotte des Enfants 4	185.7	198	152	131
Sungir 1	183.2	188	144	130
Předmostí 3	183.0	202	145	133
Oberkassel 1	175.1	194	141	137
Předmostí 9	173.3	196	145	134
Romito 4	173.2	195	146	132
Villabruna 1	172.9	181	137.5	133.5
Rochereil 1	172.6	190	138	138
Le Peyrat 5	171.5	194	144	126
Bichon 1	169.9	190	144	122
Arene Candide 3	169.3	185	144	126
Arene Candide 5	167.1	203	142	146
Chancelade 1	165.7	194	135	149

The present model has an average in-sample error of approximately 4.9 cm. That is useful for broad categories—short, medium, tall or very tall—but it is not accurate enough to claim an exact stature.

A simpler two-dimensional fallback is available when M17 is missing:

Estimated stature = 107.402 + 0.002442 × (M1 × M8)

In leave-one-out testing, this model produced an average error of approximately 5.9 cm:

Validation result	Performance
Within ±5 cm	57.1%
Within ±7 cm	71.4%
Within ±10 cm	78.6%
Mean absolute error	5.9 cm

This level of error is why our database reports a range rather than presenting a cranial estimate as a direct measurement.

Historical records demonstrate the problem.

Historical case	Body-based stature	2D skull estimate	Difference
Coldrum male composite	164.5 cm	172.5 cm	+8.0 cm
Halling Man	166.5 cm from femur	172.3 cm	+5.8 cm
Ipswich Man	180.0 cm	174.9 cm	−5.1 cm

The Coldrum result is also a group composite rather than one individual. None of these historical cases has been used to alter the model.

For most of Thurnam’s 67 long-barrow skulls, only cranial indices, ranges and site averages are currently published in the material we have extracted. A cranial index alone cannot produce a height because it gives a proportion rather than the skull’s absolute dimensions.

We therefore need the original M1, M8 and M17 measurements from Crania Britannica, museum catalogues or surviving skulls.

Once recovered, those dimensions could provide the first systematic stature estimates for many of Britain’s long-headed long-barrow occupants.

Applying the method to skulls without surviving body measurements

The most important purpose of the model is not to recalculate the heights of individuals whose skeletons already provide stature estimates.

Its real value lies in the cranial-only discoveries for which the skull survives, but the associated long bones are missing or cannot be securely identified.

Our present register contains several such cases.

Where maximum cranial length, maximum cranial breadth and cranial vault height survive, the preferred three-dimensional model can be used.

Where only maximum length and breadth are available, the less precise two-dimensional fallback model must be used.

The following results are therefore predictions rather than known statures.

Skull or site	M1 length	M8 breadth	M17 height	Cranial index	Model used	Estimated stature	Approximate height
Aveline’s Hole skull M1.11.301	186 mm	134 mm	Not available	72.0	Two-dimensional	168.3 cm	5 ft 6 in
Carnon calvaria	184 mm	137 mm	Not available	74.46	Two-dimensional	169.0 cm	5 ft 6½ in
Langwith Man	192 mm	135 mm	127 mm	70.31	Three-dimensional	171.5 cm	5 ft 7½ in
Engis skull	198 mm	140 mm	Not available	70.71	Two-dimensional	175.1 cm	5 ft 9 in
Dartford cranium	207 mm	150 mm	Not available	72.46	Two-dimensional	183.2 cm	approximately 6 ft

These estimates reveal a considerable range of possible body sizes among the surviving long-headed skulls.

The relatively small Aveline’s Hole and Carnon crania produce estimates of approximately 168 to 169 centimetres.

Langwith Man produces an estimate of approximately 171.5 centimetres.

The larger Engis skull produces an estimate of approximately 175 centimetres.

The exceptionally large Dartford cranium produces an estimated stature of approximately 183 centimetres, or around six feet.

These figures must not be treated as direct measurements.

The two-dimensional model has a mean leave-one-out error of approximately 5.9 centimetres. A practical working range of roughly six centimetres on either side of each estimate is therefore more responsible than presenting a single exact height.

The estimates could consequently be expressed approximately as:

Aveline’s Hole: 162 to 174 centimetres;
Carnon: 163 to 175 centimetres;
Langwith: 165 to 177 centimetres;
Engis: 169 to 181 centimetres;
Dartford: 177 to 189 centimetres.

These are working archaeological ranges, not formal statistical confidence intervals.

The importance of Aveline’s Hole

The Aveline’s Hole calculation requires particular care.

The measured skull had a cranial index of 72.0 and produces a two-dimensional stature estimate of approximately 168.3 centimetres.

However, this skull has not been securely matched to the male tibia from which the provisional R1b genetic result was obtained.

We cannot therefore claim that the R1b male himself stood 168 centimetres tall.

What can be stated is that the Aveline’s Hole cemetery contained:

a provisionally identified R1b-derived male;
at least one securely long-headed skull;
and a measured cranium whose dimensions suggest an individual of approximately medium stature.

The genetic, cranial and stature evidence remains associated at the cemetery-population level rather than the securely identified individual level.

What the estimates suggest

The cranial-only results do not indicate that all long-headed individuals were exceptionally tall.

Instead, they suggest that long-headed prehistoric populations included individuals ranging from approximately average stature to potentially more than six feet tall.

This is an important distinction.

Cranial shape and cranial size are not the same measurement.

A person may possess a strongly elongated skull without possessing an exceptionally large skull or exceptional body height.

The Dartford cranium combines an elongated form with unusually large absolute dimensions and consequently produces the tallest prediction.

Aveline’s Hole possesses a similarly long-headed form but smaller absolute dimensions and therefore produces a much lower stature estimate.

This is precisely why cranial index alone cannot estimate height.

The index reveals the proportion of the skull.

The complete measurements reveal its physical size.

Present conclusion

The model now provides provisional stature estimates for five cranial discoveries where a secure body-based comparison is unavailable.

The results range from approximately 168 centimetres to approximately 183 centimetres.

They demonstrate that Britain and western Europe’s surviving long-headed crania did not represent one uniform body size.

Some appear to have belonged to individuals of ordinary or medium stature.

Others, particularly the Dartford individual, may have belonged to exceptionally tall people.

Further recovery of the original cranial measurements from long-barrow collections could extend this analysis from five isolated examples to a much larger prehistoric population.

7. Who built Britain’s megaliths—and what must happen next?

Thurnam’s measurements do not identify the name of a Stonehenge architect.

They do something more fundamental: they identify a distinctive biological population occupying Britain’s early monumental burial tradition.

The 67 long-barrow crania were predominantly long-headed. The 70 later round-barrow skulls contained no dolichocephalic or sub-dolichocephalic individuals in Thurnam’s classification. Later burials inserted into long barrows were broader-headed and averaged approximately 79.

The change in human morphology follows the change in monument and burial practice.

This creates a serious problem for the simplified story that “incoming farmers built Britain’s megaliths.”

Modern genetics supports substantial migration into Britain during the Neolithic. It does not automatically prove that every long barrow, causewayed enclosure, stone setting and later phase of Stonehenge was conceived by one biologically uniform immigrant population.

The word farmer identifies subsistence practice.

It does not identify:

paternal lineage;
cranial morphology;
ancestry proportions;
social leadership;
specialist engineering knowledge;
the population selected for monumental burial.

People can adopt agriculture without losing their ancestry. Incoming farmers can mix with indigenous hunter-gatherers. Different groups can cooperate, exchange partners and technologies, or occupy different positions within the same society.

The traditional construct therefore contains a hidden logical jump:

Farming appears in the archaeological record; therefore incoming farmers designed and built the monumental landscape.

That conclusion must be demonstrated, not assumed.

What the evidence currently permits us to say

Finding	Assessment
Long-barrow and round-barrow skull populations were markedly different	Strongly demonstrated
The original long-barrow population was predominantly long-headed	Strongly demonstrated
Later secondary burials inside long barrows were broader-headed	Strongly demonstrated
Long-headed morphology has Upper Palaeolithic European precedents	Demonstrated
Early R1b existed among Mesolithic European hunter-gatherers	Demonstrated
Long-barrow people descended partly from older hunter-gatherers	Strong continuity hypothesis
Long-barrow males probably included early R1b lineages	Testable hypothesis
Long-barrow people alone built every British megalith	Not demonstrated
Incoming farmers alone built every British megalith	Not demonstrated

This is the crucial distinction.

The evidence does not justify declaring that every long-headed person was a monument builder or that one paternal lineage created an entire civilisation.

But it equally does not justify removing the long-headed burial population from the discussion and replacing them with an undefined category called “farmers.”

The surviving human remains offer a direct route to resolving the question.

A serious research project should:

locate all surviving skulls from Thurnam’s 17 long-barrow series;
identify individual museum and excavation numbers;
separate primary deposits from later intrusions;
rescan and remeasure the crania;
recover M1, M8 and M17 for stature modelling;
radiocarbon-date the individuals directly;
sequence their DNA;
compare their ancestry with Mesolithic, Early Neolithic and Bronze Age populations;
test whether R1b occurs among the long-headed males;
compare their reconstructed statures with our 429-person prehistoric body-size database.

This investigation could confirm, modify or reject the Cro-Magnon–Mesolithic–long-barrow continuity hypothesis.

That is how science should work.

The measurements have existed since the nineteenth century. Modern radiocarbon dating, 3D imaging and ancient DNA now allow us to test what Thurnam could only observe.

“Long barrows, long skulls; round barrows, round skulls” was not merely a Victorian slogan.

It was a concise description of a measurable change in Britain’s prehistoric population.

The physical evidence has waited more than 150 years for archaeology to finish the investigation.

PODCAST

Author’s Biography

Robert John Langdon, a polymathic luminary, emerges as a writer, historian, and eminent specialist in LiDAR Landscape Archaeology.

His intellectual voyage has been interwoven with stints as an astute scrutineer in government and grand corporate bastions, a tapestry spanning British Telecommunications, Cable and Wireless, British Gas, and the esteemed University of London.

A decade hence, Robert’s transition into retirement unfurled a chapter of insatiable curiosity. This phase saw him immerse himself in Politics, Archaeology, Philosophy, and the enigmatic realm of Quantum Mechanics. His academic odyssey traversed the venerable corridors of knowledge hubs such as the Museum of London, University College London, Birkbeck College, The City Literature Institute, and Chichester University.

In the symphony of his life, Robert is a custodian of three progeny and a pair of cherished grandchildren. His sanctuary lies ensconced in the embrace of West Wales, where he inhabits an isolated cottage, its windows framing a vista of the boundless sea – a retreat from the scrutinising gaze of Her Majesty’s Revenue and Customs, an amiable clandestinity in the lap of nature.

Exploring Prehistoric Britain: A Journey Through Time

My blog delves into the fascinating mysteries of prehistoric Britain, challenging conventional narratives and offering fresh perspectives grounded in cutting-edge research, particularly LiDAR technology. I invite you to explore some key areas of my research. For example, the Wansdyke, often cited as a defensive structure, is re-examined in light of new evidence. I’ve presented my findings in my blog post Wansdyke: A British Frontier Wall – ‘Debunked’, and a Wansdyke LiDAR Flyover video further visualises my conclusions.

My work also often challenges established archaeological dogma. I argue that many sites, such as Hambledon Hill, commonly identified as Iron Age hillforts, are not what they seem. My posts Lidar Investigation Hambledon Hill – NOT an ‘Iron Age Fort’ and Unmasking the “Iron Age Hillfort” Myth explore these ideas in detail and offer an alternative view. Similarly, sites like Cissbury Ring and White Sheet Camp receive re-evaluations based on LiDAR analysis in my posts “Lidar Investigation Cissbury Ring through time” and “Lidar Investigation White Sheet Camp,“ revealing fascinating insights into their true purpose. I have also examined South Cadbury Castle, often linked to the mythical Camelot56.

My research also extends to ancient water management, including the role of canals and other linear earthworks. I have discussed the true origins of Car Dyke in multiple posts, including Car Dyke – ABC News Podcast and Lidar Investigation Car Dyke – North Section, which suggest a Mesolithic origin 2357. I also explore the misidentification of Roman aqueducts, as seen in my posts on the Great Chesters (Roman) Aqueduct. My research has also been greatly informed by my post-glacial flooding hypothesis, which has helped explain landscape transformations over time. I have discussed this hypothesis in several posts, including AI now supports my Post-Glacial Flooding Hypothesis and Exploring Britain’s Flooded Past: A Personal Journey

Finally, my blog also investigates prehistoric burial practices, as seen in Prehistoric Burial Practices of Britain and explores the mystery of Pillow Mounds, often mistaken for medieval rabbit warrens, but with a potential link to Bronze Age cremation in my posts: Pillow Mounds: A Bronze Age Legacy of Cremation? and The Mystery of Pillow Mounds: Are They Really Medieval Rabbit Warrens?. My research also includes astronomical insights into ancient sites, for example, in Rediscovering the Winter Solstice: The Original Winter Festival. I also review new information about the construction of Stonehenge in The Stonehenge Enigma.

Other Blogs

2024 Prehistoric Britain Blog Review

8000-year-old Cro-Magnon skull

Vikings, Doggerland, and the Last Survivors of Europe’s Lost Waterworld

The Great Ancient DNA Illusion: How Statistical Models Became Archaeological “Facts”

Posted on July 25, 2026 by admin

Introduction – A Revolution Built on Probabilities

Over the past decade, ancient DNA has revolutionised archaeology. For the first time, scientists can recover fragments of genetic material from people who lived thousands of years ago, offering remarkable new insights into ancestry, population relationships and prehistoric life. It is one of the greatest scientific advances ever applied to our understanding of the past, and its contribution cannot be overstated. (The Great Ancient DNA Illusion: How Statistical Models Became Archaeological “Facts”)

Yet alongside this revolution has emerged a growing problem.

Ancient DNA is often presented to the public as if it provides direct and unquestionable evidence of prehistoric events. Newspaper headlines confidently announce the arrival of new populations, sweeping migrations and the replacement of entire peoples, creating the impression that these conclusions are simply “read” from the DNA itself.

The reality is considerably more complex.

Ancient DNA does not arrive as a complete genetic blueprint. Most prehistoric DNA survives only as tiny, degraded fragments that must be reconstructed using sophisticated statistical techniques. Radiocarbon dates are refined using Bayesian probability models. Individuals are assigned to ancestral populations through comparative statistical analysis. Finally, these results are interpreted within existing archaeological frameworks to produce historical narratives.

Each of these stages is scientifically valid and often essential. However, each also introduces assumptions, probabilities and modelling. By the time the public reads that a migration has been “proved”, the conclusion has passed through several layers of statistical interpretation before becoming an archaeological fact.

This distinction matters.

(The Great Ancient DNA Illusion: How Statistical Models Became Archaeological "Facts") — (**The Great Ancient DNA Illusion: How Statistical Models Became Archaeological “Facts”**)

Science advances by continually questioning its own models, especially when new discoveries challenge long-held assumptions. Over the past week, we have examined a series of remarkable discoveries that do exactly that. The identification of Mesolithic R1b lineages in Britain, together with an increasing number of pre-Beaker R1b individuals across Europe, raises an important question that deserves careful consideration.

Have our statistical models begun to shape the stories we tell about prehistory, rather than simply helping us to interpret the evidence?

This article is not an attack on ancient DNA research. On the contrary, it is a defence of good science. Ancient DNA remains one of archaeology’s most powerful tools, but it should never be confused with certainty. As we shall see, there is a crucial difference between genetic evidence and the statistical models used to interpret it—and recognising that difference may fundamentally change how we understand Europe’s prehistoric past.

1. Ancient DNA Is Not a Photograph of the Past

One of the greatest misconceptions surrounding ancient DNA is the belief that it provides a complete genetic snapshot of an individual who lived thousands of years ago. In reality, ancient DNA is nothing like reading a modern human genome. Time, burial conditions and natural chemical processes progressively destroy DNA after death, leaving researchers with only tiny surviving fragments from which they must reconstruct the original genetic sequence.

Unlike modern DNA samples, where virtually the entire genome can be sequenced, ancient specimens are often highly fragmented and contaminated by bacteria, fungi and even modern human DNA introduced during excavation or handling. Laboratories therefore begin by extracting only the authentic ancient fragments before using specialised computer software to align these short pieces against the modern human reference genome. The result is not a complete genome but a partial reconstruction based on the evidence that survives.

The quality of that reconstruction varies enormously from one individual to another. Some exceptional specimens preserve millions of readable DNA positions, while others may contain only a few thousand. These positions are known as Single Nucleotide Polymorphisms (SNPs), the individual genetic markers used to identify ancestry, relationships and inherited traits. The fewer SNPs that survive, the greater the uncertainty in the final interpretation. Two individuals may therefore appear equally represented in a published database, yet one may be based on a near-complete genome while the other relies on only a small fraction of the available genetic information.

To overcome these limitations, researchers compare ancient DNA against large databases of modern and ancient reference populations. Sophisticated statistical algorithms estimate which missing genetic markers are most likely to have been present, a process known as imputation. This is a powerful and entirely legitimate technique, but it remains a statistical prediction rather than a direct observation. As the amount of surviving DNA decreases, the reconstruction becomes increasingly dependent upon probability rather than recovered evidence.

The same principle applies when assigning an individual to an ancestral population or identifying a Y-chromosome haplogroup. These classifications are not usually determined by a single defining mutation but by assessing how closely an incomplete genetic profile matches previously identified populations. Every assignment therefore carries an associated confidence level. In many cases, these confidence values are extremely high, but they are rarely absolute. The public, however, almost never sees these probabilities. Instead, tentative statistical conclusions are frequently presented as definitive historical facts.

None of this diminishes the extraordinary value of ancient DNA research. Without these statistical methods, much of prehistoric genetics would remain inaccessible. However, it is essential to recognise the distinction between recovered evidence and reconstructed evidence. Every ancient genome represents a combination of preserved DNA, statistical modelling and informed scientific interpretation. Before archaeologists begin discussing migrations, population replacements or cultural change, the genetic evidence has already passed through several stages of reconstruction, each introducing a degree of uncertainty.

Understanding this distinction is fundamental to interpreting ancient DNA responsibly. The science is exceptionally powerful, but it is not a direct photograph of the past. It is a carefully reconstructed image whose clarity depends upon the quantity and quality of the surviving evidence, the statistical methods employed, and the assumptions built into those models. Appreciating that uncertainty is the first step towards separating what the DNA actually tells us from the historical narratives later constructed around it.

2. The Bayesian Revolution

One of the least understood aspects of modern archaeology is the role played by Bayesian statistics. While ancient DNA often dominates the headlines, Bayesian modelling has quietly become one of the most influential tools for constructing archaeological chronologies. It is an exceptionally powerful statistical method that has transformed the interpretation of radiocarbon dating, but it is also frequently misunderstood.

Contrary to popular belief, radiocarbon dating does not usually produce a single calendar year. Instead, every radiocarbon result is expressed as a probability distribution covering a range of possible dates. Depending upon the quality of the sample and the calibration curve, this range may extend over several decades or even centuries. In other words, the radiocarbon result itself is not a precise date but a statistical estimate with varying levels of probability.

This is where Bayesian analysis enters the picture.

Bayesian statistics allows archaeologists to combine radiocarbon dates with other sources of information. For example, if it is already known that one archaeological layer lies beneath another, or that a sequence of burials occurred over time, these relationships can be incorporated into a statistical model. The Bayesian algorithm then recalculates the most likely date ranges that satisfy both the radiocarbon evidence and the archaeological sequence. The result is often a considerably narrower chronological window than the original radiocarbon dates alone.

This represents a remarkable scientific advance. Rather than treating every radiocarbon date in isolation, Bayesian modelling uses all available evidence to produce the most probable chronology. Used correctly, it can significantly improve our understanding of archaeological sites and has become an indispensable tool in modern research.

However, Bayesian modelling also introduces an important distinction that is often overlooked.

Every Bayesian model begins with a set of prior assumptions. These assumptions may include the order of archaeological layers, the duration of cultural phases, the relationship between samples, or the accepted chronology of a particular archaeological culture. The statistical model then calculates a new set of probabilities—the posterior probabilities—that best fit both the radiocarbon evidence and those prior assumptions.

This does not mean the conclusions are wrong. In many cases they are entirely justified and scientifically robust. What it does mean is that the final chronology is no longer based solely upon direct radiocarbon measurements. It is a statistical reconstruction that combines measured evidence with informed archaeological assumptions.

This distinction is crucial because Bayesian outputs are often presented to the public as if they were direct observations. A published date range may appear highly precise, yet that precision frequently reflects the strength of the statistical model as much as the underlying radiocarbon evidence itself. Few readers realise that changing the assumptions within the model can alter the resulting chronology.

The same principle now extends far beyond radiocarbon dating. Bayesian methods are increasingly used throughout archaeology, from estimating population movements and cultural transitions to integrating genetic, environmental and archaeological datasets. As these models become more sophisticated, archaeology has gradually shifted away from relying solely on direct observations towards interpreting the past through increasingly complex statistical frameworks.

None of this diminishes the value of Bayesian analysis. It remains one of the most important advances in archaeological science. But it is essential to recognise what it actually produces. Bayesian modelling provides the most probable interpretation of the available evidence—not direct evidence itself. Understanding that distinction is fundamental if we are to separate measured data from the statistical models used to explain it.

3. When Models Begin Reinforcing Models

Modern archaeology has become increasingly dependent upon sophisticated statistical techniques. Individually, these methods are scientifically sound and have transformed our understanding of the past. The problem arises when the output from one statistical model becomes the starting assumption for the next. Over time, a chain of individually reasonable analyses can unintentionally create a feedback loop in which the original hypothesis appears to gather ever-increasing support.

Consider how a typical prehistoric individual is interpreted today.

The process begins with a degraded ancient DNA sample recovered from an archaeological site. Because much of the genome has been lost over thousands of years, statistical reconstruction is used to estimate the missing genetic information. The resulting genome is then compared against previously identified reference populations to determine the individual’s closest genetic affinities.

At the same time, the skeleton is dated using radiocarbon analysis. Rather than relying solely on the measured radiocarbon range, Bayesian modelling is frequently used to combine those dates with archaeological assumptions about site sequences, cultural phases or burial relationships. This produces a more precise—but still statistical—chronology.

Finally, the reconstructed genome and the refined chronology are interpreted within existing archaeological models. If the individual’s ancestry resembles a recognised population associated with a particular migration, the result is often presented as further evidence supporting that migration. Future ancient DNA discoveries are then compared against this growing body of interpreted data, using the same reference populations and the same chronological frameworks.

The process can be summarised as follows:

Ancient DNA
        ↓
Statistical reconstruction
        ↓
Bayesian dating
        ↓
Reference populations
        ↓
Migration hypothesis
        ↓
Future DNA interpreted using the same model
        ↓
Model appears confirmed

None of these individual steps is inherently flawed. Statistical reconstruction is essential when dealing with incomplete genomes. Bayesian analysis is one of archaeology’s most powerful chronological tools. Reference populations are indispensable for interpreting genetic relationships. Each method has been developed for good scientific reasons.

The difficulty arises when the assumptions embedded within earlier stages gradually become accepted as established facts within later stages. If the reference populations themselves were originally defined using a particular migration model, and future samples are classified by comparison with those same populations, the model inevitably begins to reinforce itself. New discoveries are no longer assessed entirely independently—they are interpreted within the framework created by previous statistical analyses.

This is not scientific misconduct, nor does it imply that archaeologists deliberately manipulate evidence. It is a recognised challenge in many scientific disciplines where models are repeatedly refined using earlier model outputs. Unless alternative hypotheses are actively tested, there is always a risk that one interpretation becomes increasingly self-supporting simply because new evidence is examined through the same analytical lens.

The history of science contains many examples of this phenomenon. Established models often appear increasingly secure until new evidence emerges that was never anticipated by the original framework. The recent discovery of Mesolithic R1b individuals illustrates exactly why this matters. These discoveries were unexpected because they lay outside the assumptions of the prevailing migration narrative. Rather than fitting neatly into the existing model, they force archaeologists to reconsider some of the assumptions upon which that model was originally built.

Good science depends on continually testing its own foundations. Statistical models are invaluable tools for interpreting incomplete evidence, but they must never become immune to challenge. Their purpose is to explain the evidence—not to determine in advance what the evidence is expected to show.

4. Britain’s Oldest R1b Changes the Starting Point

For more than two decades, the dominant interpretation of Britain’s prehistoric genetic history has been relatively straightforward. The R1b Y-chromosome lineage, now the most common paternal lineage in Britain and much of Western Europe, was widely regarded as having arrived with populations associated with the Bell Beaker phenomenon around 2500 BC. Within this framework, the appearance of R1b was seen as one of the principal pieces of evidence supporting a substantial migration into Britain during the Early Bronze Age.

That interpretation rested upon the evidence available at the time.

Recent discoveries, however, have fundamentally altered the starting point of the discussion.

The identification of an R1b lineage from Aveline’s Hole in Somerset, dating to the Mesolithic, demonstrates that R1b was already present in Britain thousands of years before the emergence of Bell Beaker culture. Instead of appearing around 2500 BC, R1b is now documented in Britain several millennia earlier, forcing archaeologists to reconsider one of the key assumptions underpinning the traditional migration narrative.

This discovery should not be overstated. A single Mesolithic R1b individual does not demonstrate uninterrupted genetic continuity from the Mesolithic to the present day. Nor does it prove that later migrations did not occur. Human populations have always moved, mixed and evolved, and no serious interpretation should suggest otherwise.

What the discovery does demonstrate is something equally important.

It shows that the previous assumption—that R1b first appeared in Britain with the Bell Beaker phenomenon—was incomplete. The starting conditions upon which many migration models were constructed have now changed. If R1b already existed within Britain before the Beaker period, then its later frequency can no longer be interpreted simply as evidence for its initial arrival. Instead, archaeologists must distinguish between an existing indigenous component and any later additions introduced through migration.

The implications extend well beyond a single archaeological site. Statistical models are only as reliable as the assumptions upon which they are built. If one of those assumptions changes, then every interpretation derived from it deserves to be re-examined. Models that once began with the premise that Britain contained no R1b before 2500 BC must now incorporate evidence showing that this is no longer the case.

This is how science is meant to progress. New discoveries refine existing theories rather than undermine the scientific process itself. The discovery at Aveline’s Hole does not invalidate ancient DNA research or the study of prehistoric migrations. Instead, it demonstrates the importance of continually testing established models against new evidence.

Most importantly, it changes the question archaeologists should now be asking. The debate is no longer whether R1b was present in Britain before the Bell Beaker period—that question has been answered by the evidence. The more important question is how widespread that earlier R1b population was, how it related to later populations, and what proportion of Britain’s paternal ancestry genuinely reflects continuity rather than later admixture.

Changing the starting point does not determine the final answer, but it changes every calculation that follows. That is why the discovery of Britain’s oldest R1b represents far more than a single genetic result—it requires the foundations of the existing migration model to be reconsidered.

One aspect of ancient DNA that deserves further explanation is the treatment of quality-control assessments within published genetic databases.

Ancient human remains vary enormously in the quality of DNA they preserve. Burial environment, groundwater chemistry, microbial activity, repeated handling, excavation history and the age of the specimen all influence how much authentic ancient DNA survives. As a result, some genomes are reconstructed from exceptionally well-preserved material, while others inevitably contain greater uncertainty.

For this reason, databases such as the Allen Ancient DNA Resource (AADR) assign quality assessments to individual samples. These include categories such as Pass, Questionable and Critical, together with a range of technical measurements relating to contamination, sequencing quality and confidence in the reconstructed genome.

These assessments are an essential part of good scientific practice.

However, it is important to distinguish between a quality-control warning and the rejection of a sample.

A quality flag indicates that additional caution is required when interpreting that individual. It does not automatically remove the sample from the published archaeological record, nor does it necessarily invalidate every conclusion derived from that genome. Instead, it provides researchers with the information required to judge the reliability of each result alongside its archaeological context and other independent lines of evidence.

This issue is particularly relevant for some of the earliest British prehistoric remains. Britain’s cave environments have often proved less favourable for long-term DNA preservation than many continental burial contexts, resulting in a number of early British genomes carrying higher quality-control warnings than better-preserved material recovered elsewhere in Europe. Such preservation differences are an expected consequence of taphonomy rather than evidence that British prehistoric individuals should automatically be excluded from analysis.

Accordingly, this investigation has not attempted to conceal or ignore quality assessments. Where such warnings exist, they should form part of the interpretation. Equally, they should not be confused with formal rejection of a published sample. Scientific interpretation requires weighing the genetic evidence together with archaeological context, radiocarbon chronology, preservation quality and the wider geographical distribution of comparable discoveries.

Most importantly, the conclusions presented in this investigation do not depend upon any single individual.

Whether one particular sample is ultimately confirmed, revised or reclassified as analytical techniques continue to improve, the wider pattern remains unchanged. Pre-Bell Beaker R1b lineages are now recorded across multiple regions of Europe by numerous independent excavation teams and laboratories. It is this cumulative archaeological and genetic pattern—not the interpretation of any one specimen—that forms the basis of the discussion presented throughout this article.

As ancient DNA technology continues to advance, individual samples will undoubtedly be refined, reassigned or, in some cases, rejected. That is a normal and healthy part of scientific progress. The purpose of this investigation is therefore not to argue that every published assignment is beyond question, but to demonstrate that the growing body of evidence now warrants a broader re-examination of the demographic models used to explain Europe’s prehistoric genetic history.

5. Then More Early R1b Appeared Across Europe

Had the discovery of Britain’s Mesolithic R1b at Aveline’s Hole remained an isolated case, archaeologists might reasonably have regarded it as an exceptional anomaly requiring further investigation. Science often encounters unusual discoveries that ultimately prove to have little wider significance. However, that is no longer the situation.

Over the past few years, the number of securely identified pre-Beaker R1b individuals has steadily increased across Europe. Instead of a single unexpected discovery, researchers are now faced with multiple individuals recovered from widely separated regions, all dating to periods long before the Bell Beaker expansion traditionally associated with the arrival of R1b in north-western Europe.

The evidence now extends far beyond Britain.

Pre-Beaker R1b lineages have been identified in Britain, France, Germany, Denmark, the Czech Republic and across parts of the Balkans. These discoveries span different archaeological cultures, different environments and thousands of kilometres of geography. While each individual must be interpreted within its own archaeological context, together they demonstrate that early R1b was distributed far more widely than many migration models originally assumed.

This growing body of evidence is important because scientific confidence increases when independent discoveries begin pointing in the same direction. A single sample can always be questioned. Two or three may still be regarded as unusual. However, as discoveries accumulate across multiple countries, different excavation teams and independent laboratories, the likelihood that they all represent isolated anomalies steadily diminishes.

The geographical distribution is equally revealing. Rather than clustering around a single point of origin or a single archaeological culture, these early R1b individuals are scattered across much of Europe. Such a distribution is more consistent with a lineage that was already present across parts of the continent before the emergence of the Bell Beaker phenomenon than with one suddenly appearing everywhere after 2500 BC.

This does not mean that later migrations did not occur, nor does it suggest that Bell Beaker populations played no role in spreading particular R1b subclades. Human populations have always migrated, mixed and expanded. The archaeological and genetic evidence clearly demonstrates repeated episodes of movement throughout prehistory. What these discoveries challenge is the simpler assumption that R1b itself was entirely absent from north-western Europe until the arrival of Bell Beaker communities.

Perhaps the most significant consequence is methodological rather than historical. As each newly discovered pre-Beaker R1b individual is added to the ancient DNA record, the statistical foundations of existing migration models become increasingly difficult to maintain in their original form. The baseline assumptions are changing because the evidence is changing.

In science, patterns matter far more than isolated discoveries. Today, the appearance of early R1b across Britain, France, Germany, Denmark, the Czech Republic and the Balkans can no longer be dismissed as a collection of unrelated anomalies. Together they form an emerging geographical pattern that deserves serious investigation.

The question facing archaeology is therefore no longer whether pre-Beaker R1b existed—it demonstrably did. The challenge now is to determine how widespread these populations were, how they were connected across Europe, and how much they contributed to the genetic landscape that later archaeological models attributed almost entirely to Bronze Age migration.

6. The Statistical Illusion

For years, the Bell Beaker migration hypothesis has been presented as though the ancient DNA record were a complete picture of prehistoric Europe. It is not. Like every archaeological dataset, it represents only the individuals who survived, were excavated and were selected for genetic analysis. The question is therefore not whether the database is useful—it undoubtedly is—but whether it can be treated as a statistical census of prehistoric Europe.

To answer that question, we examined every published prehistoric male dated before 2500 BC contained within the Allen Ancient DNA Resource.

The results are surprisingly straightforward.

Pre-2500 BC males	1,351
Confirmed pre-Beaker R1b	43
Observed R1b frequency	3.18%

Unlike many previous discussions, these figures are not derived from statistical modelling or selected case studies. They are direct counts from the published ancient DNA database.

At first glance, 3.18% appears small. In reality, it has profound implications.

Previous chapters estimated the Mesolithic population of Europe at between 250,000 and 500,000 people. If the observed frequency of 3.18% is applied conservatively to those population estimates, it represents an illustrative minimum of approximately 8,000 to 16,000 R1b individuals living across Europe before 2500 BC.

This is no longer a discussion about a handful of exceptional skeletons. It is a population measured in many thousands.

Communities of this size would have been capable of maintaining regional populations, exchanging technology, establishing long-distance trade networks and contributing genetically to neighbouring populations over many generations. They represent a substantial indigenous component of prehistoric Europe rather than isolated anomalies.

Equally important is what this means for the traditional migration narrative. If thousands of R1b individuals were already distributed across Europe before the Bell Beaker horizon, then the later dominance of R1b no longer requires a single overwhelming migration to explain its presence. Indigenous populations already existed upon which later demographic expansion, cultural diffusion and regional admixture could act.

The argument becomes stronger still when sampling bias is considered.

A second independent archaeological database containing 725 prehistoric skeletons demonstrates that hundreds of excavated individuals are absent from the published genetic record. Ancient DNA is therefore not a census of prehistoric Europe but a selective archaeological sample. Preservation conditions, excavation priorities and research objectives all influence which individuals eventually appear in genetic databases.

The consequence is unavoidable. The 43 confirmed pre-Beaker R1b males should not be interpreted as the total prehistoric R1b population. They represent the minimum number currently visible within a highly selective sample. When even this conservative dataset identifies an observed frequency of 3.18%, the mathematical implication is that prehistoric Europe already contained many thousands of R1b individuals long before the Bell Beaker period.

This changes the debate fundamentally. The question is no longer whether pre-Beaker R1b existed—it demonstrably did. The question is whether a continent already containing thousands of indigenous R1b individuals requires a later population replacement to explain the genetic evidence, or whether existing populations, interacting through long-established exchange networks and gradual demographic expansion, provide a more parsimonious explanation.

7. If Not Replacement, Then What?

By this stage, several important conclusions have emerged from the evidence presented throughout this blog.

The ancient DNA database is not a census of prehistoric Europe. Bayesian chronological modelling depends on prior assumptions. Confirmed pre-Bell Beaker R1b individuals are now distributed across much of Europe thousands of years before the traditionally accepted migration horizon. Statistical analysis demonstrates that these individuals were unlikely to represent isolated anomalies, while independent archaeological evidence shows that the published genetic database contains only a fraction of the excavated prehistoric population.

Taken together, these findings raise an important question.

If the traditional model of wholesale population replacement is no longer the only explanation consistent with the available evidence, what alternatives should now be considered?

The first possibility remains the conventional interpretation: large-scale migration accompanied by substantial population replacement. Human migration is a well-documented feature of history, and there is no reason to reject the possibility that movements of people contributed to the changing genetic landscape of prehistoric Europe. However, once measurable indigenous R1b populations are demonstrated before 2500 BC, migration alone can no longer be assumed to explain the entire pattern.

A second possibility is gradual admixture.

Rather than one population replacing another, incoming groups may have mixed with long-established regional populations over many generations. Such a process would naturally produce increasing frequencies of particular Y-chromosome lineages without requiring the near-complete disappearance of those already living across Europe. Genetic expansion through assimilation is a well-recognised demographic process and is consistent with populations interacting over centuries rather than decades.

A third possibility is regional survival.

Europe has always been geographically diverse. Mountain ranges, coastlines, forests and river systems created natural barriers that encouraged local continuity alongside occasional contact. Some regions may have experienced substantial migration, while others retained much of their earlier population. Such a model would explain why genetic continuity appears stronger in some areas than others and why archaeological traditions often persist despite changing material culture.

A fourth possibility is cultural diffusion.

Ideas frequently travel faster than people. Pottery styles, metallurgy, farming techniques and religious beliefs can spread through trade, exchange and social interaction without requiring large-scale migration. The Bell Beaker phenomenon itself displays many of the characteristics of a cultural network, appearing across an enormous geographical area while exhibiting considerable regional variation. If existing communities adopted new technologies and social practices through exchange, cultural change need not imply wholesale demographic replacement.

Finally, Europe may simply have experienced multiple episodes of migration, interaction and assimilation over thousands of years.

Human history is rarely explained by a single event. Climate change, flooding, expanding trade networks, technological innovation and changing social structures would all have encouraged repeated movements of people across the continent. Under such circumstances, the genetic landscape observed today would be the cumulative result of many demographic processes acting together rather than the consequence of one catastrophic migration.

The evidence presented throughout this investigation does not require the rejection of migration as a historical reality. People have always moved, traded, intermarried and established new communities. What it challenges is the assumption that a single migration event provides the only satisfactory explanation for the genetic and archaeological evidence.

Once indigenous R1b populations, sampling bias, demographic modelling and the statistical evidence presented in the previous chapters are taken into account, prehistoric Europe begins to look considerably more complex than a simple story of invasion and replacement.

Perhaps the greatest lesson from ancient DNA is not that one theory has finally solved European prehistory, but that the past was almost certainly more complicated than any single model can adequately describe.

I agree. In fact, I think the book has naturally built towards a final conclusion.

The structure now looks like this:

Ancient DNA Is Not a Photograph of the Past
The Bayesian Revolution
When Models Reinforce Models
Britain’s Oldest R1b Changes the Starting Point
Pre-Beaker R1b Across Europe
The Statistical Illusion
If Not Replacement, Then What?
Conclusion – Time to Rethink European Prehistory

I wouldn’t make Chapter 8 long. Around 1,200–1,500 words would be enough. It shouldn’t introduce new evidence. It should simply pull together everything the reader has already seen.

Something like this:

8. Time to Rethink European Prehistory

Every scientific theory begins as a hypothesis.

Some hypotheses survive repeated testing and become stronger with each new discovery. Others require modification as new evidence accumulates. The history of science is not the history of certainty, but of continual refinement as better data become available.

The Bell Beaker migration hypothesis transformed prehistoric archaeology by incorporating ancient DNA into the study of Europe’s past. It provided an elegant explanation for the widespread distribution of R1b lineages after 2500 BC and rapidly became the dominant model for understanding the later Neolithic and Early Bronze Age.

Yet this investigation demonstrates that the foundations of that model are no longer as secure as they once appeared.

The first problem is statistical.

Ancient DNA databases do not represent complete prehistoric populations. They represent only those individuals whose remains survived, were excavated and were selected for genetic analysis. Every conclusion derived from those databases must therefore recognise the limitations imposed by preservation, excavation and research priorities.

The second problem concerns chronology.

Bayesian modelling has undoubtedly improved archaeological dating, but every Bayesian model depends upon the assumptions that define it. When previous interpretations become the priors for new analyses, there is always a risk that established ideas reinforce themselves rather than being independently tested.

The third problem is genetic.

Forty-three confirmed pre-Bell Beaker R1b individuals are now known from across Europe. These individuals pre-date the traditionally accepted migration horizon by centuries and, in many cases, millennia. They are geographically widespread and cannot reasonably be dismissed as isolated anomalies.

When placed within estimated Mesolithic population figures, even the conservative observed frequency recorded in the published DNA database corresponds to many thousands of R1b individuals living across prehistoric Europe. Such populations require explanation in their own right.

Finally, archaeology itself presents a more complex picture than a single migration narrative suggests.

Material culture changes at different rates from genetics. Trade networks expand and contract. Technologies spread between communities. Populations mix, divide and reconnect over generations. Human history is rarely explained by one event, one migration or one cultural horizon.

None of this proves that migration did not occur.

Human migration is one of the constants of prehistory. Europe has always been shaped by movement, exchange and interaction.

What the evidence presented in this book demonstrates is something more modest but, perhaps, more important.

The current evidence no longer requires a single replacement model to explain the emergence of R1b across Europe.

Instead, the available data are equally consistent with a far more dynamic prehistoric landscape in which indigenous populations, regional continuity, repeated migrations, long-distance trade and cultural diffusion all contributed to the genetic and archaeological record we observe today.

Perhaps the greatest lesson from this investigation is methodological.

Science advances not by defending established ideas but by continually testing them against new evidence. Ancient DNA has revolutionised archaeology, yet it remains only one line of evidence. Genetics, archaeology, anthropology, geology, palaeoclimatology and statistics must all be considered together if we are to reconstruct Europe’s past as accurately as possible.

The purpose of this blog has not been to replace one certainty with another.

It has been to demonstrate that important questions remain unresolved, that assumptions deserve re-examination, and that the archaeological record is considerably more complex than the simplified narratives often presented to the public.

The prehistoric peoples of Europe were not merely passive recipients of change arriving from elsewhere. They were active participants in a continent that had already been interconnected for thousands of years through trade, migration, adaptation and cultural exchange.

As new discoveries continue to emerge, the story of prehistoric Europe will undoubtedly evolve again.

The evidence presented here suggests that evolution has already begun.

Appendix A – Confirmed Pre-Bell Beaker R1b Individuals Included in the Present Analysis

The following table lists all 43 confirmed pre-Bell Beaker R1b individuals identified in the Allen Ancient DNA Resource (AADR v66.1, 1240K) and included in the statistical analysis presented in this investigation.

Individual ID	Site	Country	Date	Y-DNA
I6912	Brunn-Wolfholz	Austria	5500–4750 BCE	R1b1a1b
I14169	Makotřasy	Czechia	4300–3500 BCE	R1b
I14173	Makotřasy	Czechia	4300–3500 BCE	R1b
I14176	Makotřasy	Czechia	3700–3500 BCE	R1b
I15826	Praha-Jinonice	Czechia	3634–3382 cal BCE	R1b
I15650	Hostivice-Palouky	Czechia	3800–3400 BCE	R1b
I15648	Mužský-Hrad	Czechia	3598–3371 cal BCE	R1b
PNL001	Plotiště nad Labem	Czechia	2919–2875 cal BCE	R1b1a1b1a1a2a5a~
OBR003	Obříství	Czechia	2913–2786 cal BCE	R1b1a1b1a1a2a
VLI015	Vliněves	Czechia	2900–2650 BCE	R1b1a1b
STD002	Stadice	Czechia	2885–2639 cal BCE	R1b1a1b1a1a2a
VLI092	Vliněves	Czechia	2885–2636 cal BCE	R1b1a1b1a1a2a
VLI011	Vliněves	Czechia	2884–2636 cal BCE	R1b1a1b1a1a2b1
KON003	Konobrže	Czechia	2900–2600 BCE	R1b1a1b1a1a
NEO866	Lundby-Falster	Denmark	3633–3380 cal BCE	R1b
BOU38	Aven de la Boucle	France	3626–3369 cal BCE	R1b
I8055	Xanton-Chassenon	France	3081–2901 cal BCE	R1b
I0559	Quedlinburg-9	Germany	3646–3528 cal BCE	R1b
I1590	Blätterhöhle Cave	Germany	3644–3528 cal BCE	R1b
I1594	Blätterhöhle Cave	Germany	3338–3024 cal BCE	R1b
I2762	Barcehalom	Hungary	2916–2881 cal BCE	R1b1a1b1b
I18101	Kunhegyes-Nagyállás-halom	Hungary	2950–2600 BCE	R1b1a1b1b
JK2804	Cannas di Sotto	Italy	3371–3103 cal BCE	R1b1b
I6699	Teleor-3	Romania	5292–5000 cal BCE	R1b1a1b
PIE004	Pietrele Măgura Gorgana	Romania	4701–4544 cal BCE	R1b1b
PIE017	Pietrele Măgura Gorgana	Romania	4708–4537 cal BCE	R1b1b
PIE023	Pietrele Măgura Gorgana	Romania	4603–4447 cal BCE	R1b1b
PIE019	Pietrele Măgura Gorgana	Romania	5000–4000 BCE	R1b1b
PIE064	Pietrele Măgura Gorgana	Romania	4589–4409 cal BCE	R1b1a1b
PIE026	Pietrele Măgura Gorgana	Romania	4546–4370 cal BCE	R1b1b
PIE042	Pietrele Măgura Gorgana	Romania	4539–4370 cal BCE	R1b
PIE013	Pietrele Măgura Gorgana	Romania	4536–4362 cal BCE	R1b
I23123	Urziceni	Romania	4400–3500 BCE	R1b
I12823	Smeeni	Romania	3300–2500 BCE	R1b1a1b
I10499	Rahman	Romania	2896–2677 cal BCE	R1b1a1b1b
I10500	Rast-Măgura-Barburlui	Romania	2893–2674 cal BCE	R1b1a1
ATP3	El Portalón Cave	Spain	3516–3365 cal BCE	R1b1a1b
ART038	Arslantepe	Turkey	3365–3102 cal BCE	R1b1a2a
I3035	Fox Holes Cave	United Kingdom	4000–3500 BCE	R1b1a1b1a1a1c1a2b
I2611	Summerhill	United Kingdom	3092–2905 cal BCE	R1b1a1b1a1a2c1a1f1a1
M96	Schela Cladovei	Romania	7250–6500 BCE	R1b
M95	Schela Cladovei	Romania	7125–6603 cal BCE	R1b
OC	Ostrovul Corbului	Romania	7022–6485 cal BCE	R1b

Data source: Allen Ancient DNA Resource (AADR), Version 66.1 (1240K). Table compiled from the filtered dataset used in the present analysis, including all confirmed pre-Bell Beaker R1b individuals dated before the Bell Beaker horizon.

PODCAST

Author’s Biography

Robert John Langdon, a polymathic luminary, emerges as a writer, historian, and eminent specialist in LiDAR Landscape Archaeology.

Exploring Prehistoric Britain: A Journey Through Time

Other Blogs

2024 Prehistoric Britain Blog Review

8000-year-old Cro-Magnon skull

Vikings, Doggerland, and the Last Survivors of Europe’s Lost Waterworld

The Cheddar Man Hoax

Posted on May 29, 2026 by admin

CSI Cheddar Man – Examining the Evidence

Before discussing skin colour, eye colour, hair colour, race, ancestry, or the famous BBC reconstruction, we must first examine the actual evidence.

Not the headlines.

Not the artwork.

Not the political arguments.

The DNA report itself.

Like any forensic investigation, the first question is simple:

How reliable is the evidence upon which the conclusions are based?

If the foundations are uncertain, then every subsequent interpretation must be treated with caution.

This chapter, therefore, examines the report exactly as a forensic investigator might examine witness testimony in a courtroom: statement by statement, assumption by assumption, conclusion by conclusion.

Exhibit A: Eye Colour

Report Statement

“There is 1 locus (LOC105374875 rs12896399) with low coverage (1x), hence a heterozygote is possible.”

CSI Observation

The report immediately acknowledges uncertainty. One of the key genetic markers used in the prediction was observed only once. In modern genetic analysis, a single read provides far less confidence than multiple independent observations. The authors, therefore, cannot be certain whether they are observing the complete genotype or only part of it.

Report Statement

“Prediction includes a range that includes what the 1x coverage found (ancestral G allele) and the possibility of an A-derived allele being present.”

CSI Observation

The prediction is therefore not based solely upon observed evidence. It is partly based on hypothetical genetic combinations that may exist but have not been directly observed.

Report Statement

Blue eye: 0.564–0.711

Intermediate eye: 0.189–0.143

Brown eye: 0.247–0.145

CSI Observation

The software does not produce a single answer. It produces three competing probabilities. Brown remains possible. Intermediate remains possible. Blue remains possible. The model itself is expressing uncertainty.

Report Statement

“Final prediction: Intermediate (blue/green) eye colour.”

CSI Observation

Notice what has happened. Multiple probabilities have now been converted into a single descriptive conclusion. Yet the underlying model still contains alternative outcomes.

Report Statement

“It is certainly not a brown-eyed or clear blue-eyed individual.”

CSI Observation

This is perhaps the most revealing sentence in the entire eye-colour section. The report explicitly states that Cheddar Man was neither brown-eyed nor clear blue-eyed. Yet many subsequent headlines simply described him as ‘blue-eyed’. The report and the headlines are not saying the same thing.

Exhibit B: Hair Colour

Report Statement

“There is 1 locus PIGU rs2378249 with low coverage (1x), hence a heterozygote is possible.”

CSI Observation

Again, we encounter the same problem. One of the markers used to predict hair colour was only observed once. The authors, therefore, acknowledge that an alternative genotype may exist but remain undetected.

Report Statement

Blond: 0.009

Brown: 0.692–0.741

Red: 0.006–0.012

Black: 0.292–0.237

CSI Observation

The model strongly favours brown hair. Black hair remains a secondary possibility. Red hair and blond hair are assigned very low probabilities.

Report Statement

“Final Prediction: Dark Brown/Black hair colour.”

CSI Observation

This conclusion is reasonable based on the probability table. However, it is important to note that the strongest probability favours brown hair rather than black hair.

Report Statement

“This individual would be perceived as having dark brown hair. However, black cannot be ruled out.”

CSI Observation

The report itself is considerably more cautious than many later reconstructions. Dark brown is presented as the preferred interpretation. Black remains one possible alternative.

CSI Question

If the report favours dark brown hair, why do many reconstructions depict almost jet-black hair?

Exhibit C: Skin Pigmentation

Report Statement

“There are 3 loci missing.”

CSI Observation

This is arguably the single most important sentence in the entire report. Three pigmentation markers contain no data whatsoever. They are absent from the genetic record.

Report Statement

“The profile does contain 2 loci with low coverage (n=1x).”

CSI Observation

In addition to the three missing markers, two additional markers were observed only once. The prediction, therefore, contains both missing information and uncertain information simultaneously.

Report Statement

“When factoring in possible genotype combinations, a prediction range may be generated.”

CSI Observation

The report is not calculating a single answer. It is generating multiple possible outcomes because the complete genetic profile is unknown.

Report Statement

“The range consists of assuming the two loci with low coverage are correct as homozygote for their sequenced allele…”

CSI Observation

The model must now make assumptions. It assumes that the single observed allele is representative of the complete genotype. This may be correct, but it remains an assumption rather than a direct observation.

Report Statement

“…and omitting the 3 missing loci from the prediction model as they have no coverage…”

CSI Observation

The model is now attempting to predict appearance while simultaneously lacking three pieces of relevant information.

Report Statement

“…to including these SNPs with their ancestral… and also their derived allele counterparts.”

CSI Observation

The software, therefore, models multiple genetic scenarios because the true genotype is unknown.

Report Statement

Intermediate: 0.152–0.038

Dark-Black: 0.848–0.962

CSI Observation

These values are not observations. They are statistical outputs generated from a model operating with missing loci, low-coverage loci, and multiple possible genotype combinations.

Report Statement

“If we omit the three missing alleles, our tool produces 0.752 and 0.248 probabilities…”

CSI Observation

This may be the most damaging sentence in the entire report. By changing the assumptions, the model produces dramatically different results. This demonstrates how sensitive the prediction is to missing information.

Report Statement

“The missing loci certainly impact on this prediction.”

CSI Observation

The authors explicitly acknowledge that the missing data affects the outcome.

Report Statement

“It is therefore best to have some allele present to infer the most probable range for Cheddar Man.”

CSI Observation

Notice the wording. The report is not observing the complete genotype. It is inferring a probable range from incomplete information.

Report Statement

“It is unlikely that this individual has the darkest possible pigmentation, but it cannot be ruled out.”

CSI Observation

Again, the report is discussing possibilities rather than certainties.

Report Statement

“Better sequencing coverage would clarify to what degree this individual has a dark complexion.”

CSI Observation

This sentence alone demonstrates that the issue remains unresolved. If better sequencing could alter the prediction, then the current prediction cannot be regarded as definitive.

Exhibit D: The Reliability Problem

At this point, a pattern emerges.

The report repeatedly uses the language of uncertainty:

“possible”

“heterozygote”

“prediction range”

“cannot be ruled out”

“better coverage would clarify”

“missing loci”

“infer”

These are not the words of certainty.

They are the words of probability.

The authors themselves repeatedly acknowledge uncertainty.

Yet by the time the public encountered Cheddar Man, much of this uncertainty had vanished.

CSI Verdict

The report does not present a photograph.

The report does not present certainty.

The report presents a statistical model operating on incomplete ancient DNA.

The eye-colour prediction contains low-coverage markers.

The hair-colour prediction contains low-coverage markers.

The skin-colour prediction contains both low-coverage markers and completely missing loci.

The authors repeatedly acknowledge these limitations.

The real mystery, therefore, is not Cheddar Man.

The real mystery is how a report built upon uncertainty, assumptions, probability ranges and incomplete genetic information became a single face presented to millions of people as historical reality.

How DNA Predictions Actually Work

Before continuing our investigation, we need to understand what the Cheddar Man report is actually doing.

Many people imagine that scientists recover ancient DNA and simply read the result:

Eye colour: Blue

Hair colour: Brown

Skin colour: Dark

Unfortunately, it is nothing like that.

The reality is considerably more complicated.

Understanding this process is essential because many of the report’s conclusions depend not on direct observation but on statistical prediction.

DNA Does Not Contain Labels

DNA does not contain a sentence that says:

“This individual had blue eyes.”

Nor does it contain a sentence that says:

“This individual had dark skin.”

Instead, scientists examine specific locations within the genome known as SNPs (Single Nucleotide Polymorphisms).

Think of these as tiny switches.

Certain combinations of switches are statistically associated with particular traits.

For example:

Some combinations are more common among people with blue eyes.

Some combinations are more common among people with brown eyes.

Some combinations are more common among people with darker pigmentation.

The prediction system, therefore, works backwards.

Instead of observing the colour directly, it attempts to infer the most likely appearance from a collection of genetic markers.

Why Multiple Markers Are Needed

No single marker determines eye colour.

No single marker determines hair colour.

No single marker determines skin colour.

Instead, multiple genes interact.

The software, therefore, examines many locations simultaneously and combines them into a probability model.

This is important.

The software is not identifying a colour.

It is calculating the likelihood of different colours.

This is why the report produces percentages.

The model is effectively asking:

“Based upon the markers available, how often does this genetic pattern occur in people with blue eyes?”

The answer might be:

71%.

That does not mean the individual had blue eyes.

It means the model estimates a 71% probability.

Why The Results Are Not Yes Or No

Most people expect science to provide definitive answers.

The problem is that biology rarely behaves that way.

Imagine a genetic profile that resembles:

70% of blue-eyed people

20% of green-eyed people

10% of brown-eyed people

The software cannot honestly answer:

Blue.

Instead, it produces probabilities.

This is exactly what we see in the Cheddar Man report.

What Is Coverage?

This is where the report becomes particularly important.

When scientists sequence DNA, they do not read a marker once and immediately trust the result.

Instead, they attempt to read it repeatedly.

Each successful read is known as coverage.

For example:

1x coverage = read once

5x coverage = read five times

10x coverage = read ten times

30x coverage = read thirty times

The more times a marker is observed, the greater the confidence that the result is correct.

Why Ancient DNA Is Different

Modern forensic DNA usually comes from living individuals or recent biological samples.

Ancient DNA is very different.

Cheddar Man died approximately 10,000 years ago.

During that time, the DNA has been exposed to:

moisture
bacteria
chemical degradation
temperature change
physical decay

As a result, much of the original DNA has been damaged or lost.

Scientists, therefore, recover fragments rather than complete genetic sequences.

Why 1x Coverage Is A Problem

Imagine tossing a coin.

If you toss it once and obtain heads, can you conclude that the coin always lands heads?

Of course not.

You need multiple observations.

The same principle applies to DNA.

If a marker is observed only once, scientists cannot be completely certain they have captured the full genetic picture.

This is exactly why the report repeatedly states:

“A heterozygote is possible.”

The authors acknowledge that a second allele may exist, but it was simply not observed.

What Is A Heterozygote?

Every person carries two copies of most genetic markers.

One inherited from their mother.

One inherited from their father.

Suppose the possible variants are:

and

The possible combinations are:

If scientists only observe one G because coverage is 1x, they cannot know whether the actual genotype is:

This uncertainty forces the software to model multiple possibilities.

Why Missing Loci Matter

The report repeatedly refers to missing loci.

A missing locus means:

No data.

No observation.

No result.

The software, therefore, has a choice:

Ignore the marker completely.

Or estimate possible outcomes using statistical modelling.

Neither option is ideal.

This is why the report repeatedly discusses prediction ranges.

Why The Predictions Change

One of the most revealing passages in the report shows that changing the assumptions yields dramatically different probabilities of pigmentation.

This happens because the model is attempting to fill gaps in incomplete information.

In simple terms:

Change the assumptions.

Change the outcome.

That does not mean the model is wrong.

It means the model is sensitive to missing information.

The Critical Distinction

This is the single most important point in the entire debate.

The report does not directly observe:

Eye colour.

Hair colour.

Skin colour.

Instead, it observes genetic markers and then uses a statistical model to estimate the most likely appearance.

The final result is therefore not an observation.

It is an inference.

And the reliability of that inference depends entirely upon the quality and completeness of the DNA available.

This distinction will become crucial when we examine how a report filled with probabilities, assumptions, missing loci and low-coverage markers eventually became a single face presented to the public as historical reality.

The Wider Dataset – Do The Other Hunter-Gatherers Support The Headlines?

Having examined the Cheddar Man report in detail, the next question is obvious:

Was Cheddar Man unique?

Or do the other Mesolithic genomes support the popular narrative that Europe’s early hunter-gatherers were a uniform population of dark-skinned, blue-eyed people?

The answer is no.

The wider dataset reveals considerably more variation than most newspaper headlines ever acknowledged.

Exhibit 1: La Braña (Spain)

La Braña is often cited alongside Cheddar Man because the DNA results appear superficially similar.

Report Conclusion

Eyes:
Intermediate (green/hazel)

Hair:
Black/Dark Brown

Skin:
Dark to Dark-Black

CSI Observation

At first glance, La Braña appears to support the Cheddar Man narrative.

However, the same limitations remain present:

phenotype prediction rather than direct observation
probability modelling
incomplete ancient DNA
reconstruction through statistical inference

Most importantly, La Braña does not demonstrate that all Western Hunter-Gatherers looked alike.

It merely shows that another individual produced a similar probability profile.

One additional example is not proof of a continental population.

Exhibit 2: Loschbour (Luxembourg)

Loschbour is where the simplistic narrative begins to break down.

Report Conclusion

Eyes:
Intermediate blue/green

Hair:
Dark Brown/Black

Skin:
Intermediate

CSI Observation

This result differs significantly from both Cheddar Man and La Braña.

The same prediction system now produces a substantially lighter complexion.

If the popular narrative were correct, we would expect all three individuals to cluster tightly together.

Instead, variation immediately appears.

The dataset itself, therefore, demonstrates that pigmentation diversity already existed among Western Hunter-Gatherers.

Exhibit 3: Sven (Carsington Pasture Cave)

Sven is rarely mentioned in media discussions.

Report Conclusion

Dark Hair

Intermediate to Dark Skin

CSI Observation

Again the prediction does not fit a simplistic black-versus-white framework.

Instead we find another combination occupying the middle ground.

The reality is considerably more complex than the headlines suggest.

What The Authors Actually Concluded

Perhaps the most overlooked part of the entire study is the authors’ own summary.

Rather than claiming a single uniform appearance, the report identifies variation between individuals and suggests that pigmentation characteristics may have varied geographically and temporally.

This is a very different conclusion from the one often presented to the public.

The Statistical Problem

The media effectively reduced the dataset to:

Cheddar Man = dark skin

Therefore:

All Mesolithic Europeans = dark skin.

This is not how science works.

Three or four individuals cannot be assumed to represent an entire continent.

Even less can they be assumed to represent thousands of years of population history.

The actual dataset demonstrates variation, not uniformity.

The Reconstruction Problem

The public rarely encountered the probability tables.

They rarely encountered the caveats.

They rarely encountered the missing loci or low-coverage markers.

Instead they encountered reconstructed faces.

Once a face is produced, uncertainty tends to disappear.

The reconstruction becomes remembered as evidence, even though it is actually the final stage of a long chain of interpretation.

CSI Verdict

The wider dataset does not support the simplistic narrative often promoted in newspapers and documentaries.

Instead it demonstrates:

• significant variation between individuals

• differing pigmentation predictions

• differing eye-colour probabilities

• differing skin-colour probabilities

• and the continued presence of uncertainty throughout the dataset.

Far from proving that all Western Hunter-Gatherers shared a single appearance, the study itself suggests a much more diverse and complex picture.

The further we move away from the original DNA and towards the final reconstructions, the more certainty appears.

Yet the underlying evidence remains probabilistic, incomplete and open to multiple interpretations.

That chapter then sets up Chapter 4 perfectly, where you introduce the Cro-Magnon evidence and ask the obvious question:

If blue eyes, lighter pigmentation alleles, and long-headed Cro-Magnon populations already existed in prehistoric Europe, why did the public end up with a reconstruction that many people instinctively associate with a modern African phenotype rather than a Mesolithic European hunter-gatherer?

That’s where the argument starts to become genuinely interesting.

The Discovery

The Skeleton That Rewrote British History

Few archaeological discoveries in Britain have generated as much public attention as Cheddar Man.

Found in 1903 within Gough’s Cave in Cheddar Gorge, Somerset, the skeleton was immediately recognised as something extraordinary. Unlike the fragmented remains often recovered from prehistoric sites, Cheddar Man was remarkably complete and quickly became one of the most important human discoveries ever made in Britain. Later radiocarbon dating placed the individual in the Early Mesolithic period, approximately 10,000 years ago, making him one of the oldest nearly complete Homo sapiens skeletons ever found in Britain.
For over a century, Cheddar Man remained primarily an archaeological curiosity. He was important because of his age, his preservation, and what he could tell us about the hunter-gatherers who recolonised Britain after the retreat of the last Ice Age. Few outside archaeological circles paid much attention to him. (The Cheddar Man Hoax)

That changed dramatically in 2018.

Suddenly, Cheddar Man was no longer simply Britain’s oldest skeleton. He became a global news story.

Newspapers, television documentaries, museums, and websites across the world announced a remarkable discovery:

Britain’s earliest known inhabitant had dark skin and blue eyes.

The headline spread rapidly.

For some, it was a fascinating scientific revelation. For others, it challenged long-held assumptions about Britain’s ancient population. Social media exploded with debate. Politicians, commentators, journalists, and activists all weighed in on what the discovery supposedly meant.

The reconstruction itself was striking. The face presented to the public showed a man with very dark skin, dark curly hair, and vivid blue eyes. It was an image designed to challenge expectations, and it succeeded spectacularly. Within days, the reconstruction had become one of the most recognisable prehistoric faces in the world.

Yet buried beneath the headlines was a much more interesting question.

What exactly had scientists discovered?

At first glance the answer appears straightforward. Researchers extracted ancient DNA from the petrous portion of Cheddar Man’s skull, one of the densest bones in the human body and often the best source of preserved genetic material. Using modern sequencing techniques, they recovered enough DNA to investigate ancestry, physical traits, and population relationships. The resulting analysis was then used to create a facial reconstruction for television and museum audiences.

Case closed.

Or so it seemed.

However, archaeology has a long history of transforming cautious scientific observations into definitive public narratives. The journey from excavation trench to newspaper headline is rarely straightforward. Data must be interpreted. Models must be constructed. Probabilities must be assigned. Artists must make decisions. Journalists must simplify complex science into a few memorable sentences.

At every stage, uncertainty can become reduced until the final public story appears far more certain than the original evidence ever allowed.

This is particularly important in the case of Cheddar Man because the public generally encountered only the final reconstruction. Few people ever read the genetic reports, the supplementary data, or the technical discussions surrounding the limitations of ancient DNA analysis. Most simply saw the finished image and accepted it as a scientific fact.

But science does not operate through finished images.

Science operates through evidence.

This raises a crucial question.

Was the famous reconstruction simply a visual representation of the available evidence?

Or did the reconstruction become more definitive than the evidence itself?

To answer that question, we must set aside modern assumptions, political arguments, and media headlines. Instead, we must approach the case exactly as a forensic investigator would.

We begin not with conclusions, but with evidence.

The skeleton is our first witness.

The DNA is our second.

The prediction models are our third.

And the media narrative is our fourth.

Only when all four witnesses have been questioned can we determine whether the famous image of Cheddar Man represents a scientific certainty—or merely one possible interpretation among several.

The investigation begins.

The Media Story

How a Scientific Study Became a Global Headline

Before examining the DNA, we must first examine the story that most people believe.

After all, very few members of the public have ever read a genetics paper. Even fewer have examined supplementary DNA datasets. What most people know about Cheddar Man comes from newspaper articles, television documentaries, museum websites, and social media posts.

In other words, they know the story rather than the evidence.

That distinction is important because the story that emerged in 2018 was remarkably simple.

Britain’s oldest known inhabitant had dark skin and blue eyes.

It was a powerful headline.

It was memorable.

Most importantly, it challenged modern assumptions about ancestry and appearance.

The Natural History Museum described Cheddar Man as a Mesolithic hunter-gatherer with dark skin and blue eyes, and further suggested that populations across Mesolithic Europe possessed similar characteristics.

The accompanying reconstruction reinforced that message visually.

The public was presented with a man possessing very dark skin, tightly curled dark hair, and striking pale blue eyes. The image rapidly became one of the most widely reproduced prehistoric reconstructions ever created.

For journalists, it was the perfect story.

It contained science.

It contained surprise.

It challenged conventional expectations.

And perhaps most importantly, it could be summarised in a single sentence.

Unfortunately, science rarely fits into a single sentence.

The moment a scientific finding is compressed into a headline, something is inevitably lost. Complex statistical probabilities become certainties. Nuance disappears. Caveats vanish. Alternative interpretations are quietly forgotten.

This process is not unique to archaeology.

It happens throughout science.

However, the Cheddar Man story provides a fascinating opportunity to observe the process in real time.

Consider the wording used throughout media reports.

The public repeatedly encountered phrases such as:

“Cheddar Man had dark skin and blue eyes.”

Not:

“The available genetic markers suggest.”

Not:

“The prediction model indicates.”

Not:

“One possible reconstruction.”

Instead, uncertainty largely disappeared, replaced by definitive statements.

This transformation is subtle but important.

Scientific investigations generally deal in probabilities.

Media reports generally deal in conclusions.

The gap between those two approaches is often where misunderstanding begins.

The Natural History Museum article provides a useful example. While discussing pigmentation, the article states:

“He is just one person, but also indicative of the population of Europe at the time. They had dark skin and most of them had pale coloured eyes, either blue or green, and dark brown hair.”

This is a significant claim.

A single individual has effectively become representative of an entire population spanning thousands of kilometres and many generations.

Yet anyone familiar with human populations knows that variation is the rule rather than the exception.

Even within modern populations, physical appearance varies enormously. It would therefore be surprising if Mesolithic Europe, stretching from Iberia to Scandinavia, displayed no comparable diversity.

This raises our first forensic question.

How much of the reconstruction is supported directly by Cheddar Man’s DNA?

And how much derives from broader assumptions regarding Mesolithic populations?

The distinction matters.

Because the public was not shown a range of possibilities.

They were shown a face.

A single face.

A definitive face.

A face that appeared to settle a debate before most people had even seen the evidence.

The reconstruction itself introduces another layer of interpretation.

Facial reconstruction is not the same as photography.

The artists responsible for the reconstruction openly acknowledge that the process combines science and artistic judgement. Measurements are taken from the skull. Tissue depths are estimated using modern comparative datasets. Missing details must be inferred. Hair texture, skin tone, facial expression, and numerous subtle characteristics require interpretation.

This does not make facial reconstruction unscientific.

Far from it.

But it does mean that every reconstruction contains assumptions.

Those assumptions may be reasonable.

They may even be highly probable.

Yet they remain assumptions nonetheless.

This creates an important chain of events:

DNA is analysed.

Genetic markers are selected.

Prediction models generate probabilities.

Scientists interpret those probabilities.

Artists interpret those interpretations.

Journalists simplify the result.

The public receives a finished image.

By the time the process is complete, the final product may appear far more certain than the original evidence ever was.

This is the central issue facing any forensic investigation of Cheddar Man.

The question is not whether the scientists acted in good faith.

Nor is it whether the artists were skilled.

The question is much simpler.

Did the certainty presented to the public accurately reflect the certainty contained within the underlying genetic evidence?

To answer that question, we must leave the headlines behind and examine the primary evidence itself.

The next witness is the DNA.

The Blue-Eye Mystery

A Trait That Should Not Exist?

If there is one feature that has consistently puzzled geneticists, anthropologists, and evolutionary biologists, it is blue eyes.

Unlike skin colour, which varies continuously across human populations, blue eyes are unusual. They are rare globally and concentrated largely within populations of European ancestry.

For most of human history, the assumption was simple.

Blue eyes evolved relatively recently in northern Europe.

The logic appeared sound. Northern Europe receives less sunlight than equatorial regions. Fair skin and lighter pigmentation were thought to be evolutionary adaptations that improved vitamin D production in low-light environments. Consequently, blue eyes were often viewed as part of the same package of traits.

Then the ancient DNA evidence arrived.

Suddenly, that simple narrative began to unravel.

The genetic analysis of La Braña in Spain suggested that blue eyes were already present amongst European hunter-gatherers thousands of years before the arrival of later farming populations.

Cheddar Man appeared to tell a similar story.

The implication was profound.

Blue eyes may not have originated amongst later northern European populations at all.

They may have been inherited from a much older ancestral population.

Where Do Blue Eyes Come From?

Modern genetic studies suggest that most blue-eyed individuals share a common mutation affecting the regulation of the OCA2 gene.

This gene plays a major role in melanin production, the pigment responsible for skin, hair, and eye colour.

Research discussed in earlier studies suggested that all modern blue-eyed individuals ultimately trace their eye colour to a common ancestral mutation that altered how melanin was expressed within the iris. Rather than producing the high concentrations associated with brown eyes, the mutation reduced melanin production, creating the blue appearance we recognise today.

The important point is that blue eyes are not produced by a separate blue pigment.

There is no blue colouring within the iris.

Instead, blue eyes result from reduced melanin and the way light scatters through the eye.

In other words, blue eyes are a genetic modification of the ancestral brown-eyed condition.

The question, therefore, becomes:

When did this mutation first appear?

The Problem for Traditional Narratives

The discovery of blue-eyed Mesolithic hunter-gatherers created an immediate problem.

If blue eyes already existed amongst populations such as La Braña and Cheddar Man, then the trait must predate many of the later migrations traditionally used to explain European pigmentation patterns.

This does not automatically tell us where the mutation originated.

However, it does demonstrate that the standard picture of blue eyes emerging alongside fair-skinned farming populations is incomplete.

Blue eyes clearly existed earlier.

The question is how much earlier.

And within which population?

Darwin's Children — The Cheddar Man Hoax

The Neanderthal Question

This is where the investigation becomes considerably more controversial.

Langdon research notes that Neanderthals possessed several pigmentation variants not commonly found in modern populations and that some Neanderthal individuals appear to have possessed lighter pigmentation traits, including red hair and fair skin.

Furthermore, it is now widely accepted that interbreeding occurred between Homo sapiens and Neanderthals, leaving a measurable genetic legacy within modern Eurasian populations.

This naturally raises an intriguing question.

Could traits such as blue eyes have originated within Neanderthal populations and later entered the Homo sapiens gene pool through interbreeding?

At present, the answer is unknown.

The evidence currently available does not allow such a conclusion to be demonstrated.

Nor does it allow it to be completely dismissed.

What can be said is that the existence of blue-eyed hunter-gatherers long before many later population movements suggests that the history of eye colour is likely to be far older and more complex than the simplified narratives often presented in popular media.

The important point is not whether the Neanderthal hypothesis is correct.

The important point is that alternative explanations exist and deserve investigation.

Science progresses by examining possibilities, not by prematurely closing them.

A Trait Looking for an Explanation

By now, the original Cheddar Man story is beginning to look very different.

The popular narrative suggests a simple conclusion:

Cheddar Man had blue eyes.

Yet the supplementary report itself actually describes an intermediate blue-green eye colour rather than clear blue eyes.

The wider European evidence reveals that similar eye-colour patterns already existed elsewhere in Mesolithic Europe.

The genetic history of blue eyes appears considerably older than once believed.

And the ultimate origin of the mutation remains a matter of continuing investigation.

Rather than answering questions, blue eyes create new ones.

Where did the mutation originate?

Why did it spread?

Why did it survive?

And what does its presence reveal about the populations that occupied Europe at the end of the Ice Age?

To answer those questions, we must move beyond pigmentation alone and examine the people themselves.

The next witness is not a genetic marker.

The next witness is the Cro-Magnon population.

The Forgotten Evidence – The Skull Itself

Looking Beyond Pigmentation

By this point in our investigation, we have spent considerable time discussing skin colour, eye colour and hair colour.

That is understandable.

After all, these are the characteristics that dominated the headlines.

Yet there is an obvious problem.

Pigmentation is only a tiny fraction of what makes a person recognisable.

If two individuals share the same skin colour but have completely different skull shapes, facial structures, and body proportions, they can appear entirely unrelated.

Conversely, individuals with different pigmentation but similar skeletal anatomy may appear closely connected.

This raises an important question.

Why did the public debate surrounding Cheddar Man become almost entirely focused on colouration while largely ignoring the actual skeleton?

Because, unlike the pigmentation predictions, the skull is not a probability.

The skull is evidence.

How Anthropologists Originally Identified Ancient Populations

Long before the development of ancient DNA analysis, anthropologists classified prehistoric populations using skeletal morphology.

They examined:

skull shape
cranial length
cranial height
forehead profile
eye socket shape
nasal aperture
cheekbones
jaw structure
dentition

These characteristics allowed researchers to compare populations across both time and geography.

The presenter in the Cheddar Man documentary notes that Mesolithic hunter-gatherers were traditionally recognised by precisely these skeletal features, and that many displayed robust facial anatomy, powerful mandibles, and pronounced cranial characteristics.

In other words, before DNA entered the discussion, researchers already possessed a substantial body of evidence regarding the physical appearance of these populations.

That evidence did not disappear simply because genetic testing became available.

The Long-Headed Europeans

One of the most frequently recorded observations concerning Europe’s Upper Palaeolithic and Mesolithic populations was their tendency towards long skulls.

Traditional anthropologists referred to these populations as dolichocephalic.

The famous Cro-Magnon skeletons discovered in France were described as possessing:

long skulls
broad faces
low-set square eye sockets
strong jaws
powerful facial structure

These features became defining characteristics of Europe’s earliest modern populations.

Although many of the racial classifications used by nineteenth-century anthropologists have since been abandoned, the measurements themselves remain part of the archaeological record.

The skulls have not changed.

The measurements have not changed.

Only the terminology has changed.

The Curious Absence of Morphology

This is where the Cheddar Man story becomes rather strange.

The reconstruction was widely discussed because of its skin colour.

Yet comparatively little public attention was paid to the underlying skull morphology.

Most newspaper articles focused on:

black skin
blue eyes
dark hair

Very few discussed:

cranial proportions
facial architecture
jaw structure
orbital shape
skeletal robustness

Yet these are arguably the characteristics that define a face far more than pigmentation alone.

A person’s skull determines:

head shape
facial width
eye placement
nose shape
jaw profile
chin structure

These features dominate the appearance.

Skin colour merely overlays them.

What The Skull Suggests

The surviving skull formed the foundation of every reconstruction produced.

Without it, there would be no face to reconstruct.

Genetics can estimate pigmentation.

The skull determines the individual.

This distinction is often forgotten.

A pigmentation model can suggest probabilities.

A skull provides measurable anatomy.

Consequently, any attempt to reconstruct Cheddar Man should begin not with skin colour but with the physical evidence of the skeleton itself.

This is exactly how forensic reconstructions are normally performed.

The skeletal structure comes first.

The pigmentation is added later.

Yet much of the public discussion surrounding Cheddar Man reversed this process.

The pigmentation became the story.

The skeleton became secondary.

The Bigger Picture

The result is that many people now associate Cheddar Man with a skin colour rather than a population.

Yet the archaeological evidence suggests he was part of a wider Western Hunter-Gatherer community that occupied much of post-glacial Europe.

These populations possessed physical characteristics extending far beyond pigmentation alone.

To reduce them to a single skin-tone category risks losing sight of the broader biological reality.

The real lesson of Cheddar Man may therefore be that the most important evidence was present from the beginning.

Not the DNA.

Not the reconstruction.

Not the headlines.

But the skull itself.

And it is to that skull, combined with the genetic evidence, that we now turn for our final reconstruction.

The Cro-Magnon Connection

The Evidence Everyone Forgot

By this point in our investigation, we have spent considerable time examining pigmentation.

Eye colour.

Hair colour.

Skin colour.

Yet there is an obvious problem.

Pigmentation is only a tiny fraction of what makes a human being look the way they do.

A person’s overall appearance is determined by dozens of other characteristics, including skull shape, facial structure, height, robustness, muscle attachment, cranial capacity and body proportions.

This raises an important question.

Why did the public discussion surrounding Cheddar Man become almost entirely focused on skin colour while largely ignoring the rest of the skeleton?

The answer may be because pigmentation generates headlines.

Skull morphology does not.

Yet if we wish to understand who Cheddar Man really was, the rest of the skeleton deserves equal attention.

The Wider Western Hunter-Gatherer Population

The genetic analysis places Cheddar Man within the Western Hunter-Gatherer population that occupied much of Europe following the end of the last Ice Age.

La Braña, the famous Spanish individual frequently cited alongside Cheddar Man, also belongs to this broader population.

This is important because it means that Cheddar Man was not an isolated individual.

He was part of a much larger population network stretching across post-glacial Europe.

The question, therefore, becomes:

What did these people actually look like?

Unfortunately, this is where the public narrative becomes surprisingly selective.

The newspapers focused on skin colour.

The documentaries focused on eye colour.

The headlines focused on race.

Very little attention was paid to the wider physical characteristics of the people themselves.

The Long-Headed Europeans

Long before the development of DNA analysis, Victorian anthropologists devoted enormous effort to measuring prehistoric skeletons.

Their methods were often imperfect, and many of their conclusions are now outdated. However, one observation repeatedly appeared throughout their studies.

Many early European skeletons possessed long skulls.

These were traditionally described as dolichocephalic populations.

The classic Cro-Magnon skeletons discovered in France displayed long skulls, large cranial capacities, strong facial structures and robust physiques. They differed sufficiently from modern Europeans that early researchers initially regarded them as a distinct variety of humanity.

Today, archaeologists classify these individuals as fully modern humans.

However, the physical characteristics remain.

The skulls did not suddenly become round.

The facial structure did not disappear.

The underlying morphology still exists within the archaeological record.

An Uncomfortable Coincidence

This is where the story becomes interesting.

The two most famous Mesolithic individuals associated with pale eyes and darker pigmentation are:

Cheddar Man
La Braña

Both belong to the Western Hunter-Gatherer population.

Both lived before the arrival of later Neolithic farming populations.

Both are repeatedly used as examples of Europe’s original post-glacial inhabitants.

Yet when modern articles discuss these individuals, almost all attention is directed towards pigmentation, while the wider physical characteristics of these populations receive little mention.

The result is a strange distortion.

Readers are encouraged to focus on a handful of pigmentation markers while largely ignoring the broader biological picture.

It is rather like trying to identify a suspect from their eye colour while ignoring the rest of their face.

More Than Skin Deep

The irony is that the genetic evidence itself repeatedly warns against oversimplification.

The pigmentation models generate probabilities.

The eye-colour predictions contain uncertainty.

The skin-colour predictions contain missing loci.

The hair-colour predictions contain alternative outcomes.

Yet despite all this uncertainty, the public discussion became increasingly focused upon a single characteristic:

Skin colour.

The wider physical appearance of Europe’s hunter-gatherers largely disappeared from view.

This is particularly surprising because prehistoric populations are often identified through combinations of traits rather than isolated characteristics.

Height matters.

Facial structure matters.

Cranial morphology matters.

Body proportions matter.

Population history is rarely written in a single SNP.

The Cro-Magnon Question

This brings us to a question that rarely appears in discussions of Cheddar Man.

If Western Hunter-Gatherers formed part of a wider European population with roots stretching back into the Upper Palaeolithic, how much continuity existed between these Mesolithic groups and the populations traditionally described as Cro-Magnons?

The question is not whether they were identical.

Ten thousand years of separation would make that unlikely.

The question is whether the pigmentation evidence has distracted attention away from broader population continuity.

Because if the discussion focuses solely on skin colour, an enormous amount of archaeological and biological evidence remains unexplored.

The reality may be that Cheddar Man was neither the modern racial symbol presented by some journalists nor the simplistic caricature presented by his critics.

Instead, he may represent part of a much older European population whose appearance was considerably more diverse and complex than modern labels allow.

The Real Lesson

The greatest weakness of the public narrative is not its discussion of pigmentation.

Pigmentation is important.

The weakness is that it discussed little else.

The public was shown a face.

A skin colour.

A pair of eyes.

A headline.

What they were not shown was the uncertainty within genetics, the diversity within Mesolithic populations, or the broader physical characteristics of the people themselves.

In short, they were shown a conclusion before they were shown the evidence.

And that brings us to the final stage of our investigation.

How did a series of probabilities become one of the most famous prehistoric faces ever created?

The next witness is the reconstruction itself.

Reconstructing Cheddar Man – The Final Verdict

The Most Important Piece of Evidence

Throughout this investigation, enormous attention has been devoted to pigmentation.

Dark skin.

Blue eyes.

Brown hair.

Yet the most important piece of evidence has been sitting in front of us from the beginning.

The skull.

Unlike the pigmentation predictions, which are generated from incomplete DNA and statistical modelling, the skull is real evidence. It belonged to a living individual and survives today.

If a forensic scientist were reconstructing a suspect, they would begin with the skull and only then consider pigmentation.

The public discussion surrounding Cheddar Man largely reversed this process.

What The Skull Tells Us

The skull determines:

overall head shape
forehead profile
facial width
eye socket shape
nasal structure
jaw shape
chin shape
cheekbone position

These characteristics influence appearance far more than eye colour alone.

Traditional descriptions of Upper Palaeolithic and Mesolithic European populations frequently describe robust facial anatomy, long skulls and broad facial structures associated with the populations historically referred to as Cro-Magnons.

Whether every characteristic applied directly to Cheddar Man is less important than the broader observation:

The skull itself provides the foundation of the reconstruction.

Not the pigmentation.

What The DNA Actually Suggests

Having established the skeletal framework, we can then add the genetic evidence.

The DNA suggests:

Eyes

Not bright blue.

Instead, the report concludes:

Intermediate (blue/green)

and specifically states:

Not a clear blue-eyed individual.

Hair

Brown hair was the strongest prediction.

Dark brown was the author’s interpretation.

Black remained possible but was not the primary result.

Skin

The highest probabilities fell within darker pigmentation categories.

However, three loci were missing, and two others had low coverage, yielding a range of probabilities rather than certainty.

The CSI Reconstruction

If we reconstruct Cheddar Man using only the evidence presented during this investigation, the result would be:

A young Western Hunter-Gatherer male, approximately 5 feet 5 inches tall, with a long Mesolithic skull, robust facial features, broad eye sockets, a prominent nose, brown to dark-brown hair, pale intermediate blue-green eyes and skin pigmentation likely darker than many modern northern Europeans but impossible to define precisely from the available DNA evidence.

That reconstruction is less dramatic than the famous museum version.

It is also closer to the evidence.

The Verdict

After examining the genetics, the supplementary report, the pigmentation models and the wider Mesolithic evidence, the verdict is surprisingly simple.

The famous reconstruction cannot be proven.

Nor can it be disproven.

The evidence supports the possibility of darker pigmentation.

The evidence supports pale eyes.

The evidence supports brown hair.

What the evidence does not support is the level of certainty with which these features were presented to the public.

The DNA produced probabilities.

The scientists produced interpretations.

The artists produced a face.

The media produced a certainty.

And that certainty became the story.

The real lesson of the Cheddar Man case is not that the reconstruction is necessarily wrong.

It is that one possible interpretation became presented as though it were the only interpretation.

The skull is real.

The DNA is incomplete.

The reconstruction is a hypothesis.

And as every good investigator knows, a hypothesis is not the same thing as a proven fact.

PODCAST

Author’s Biography

Robert John Langdon, a polymathic luminary, emerges as a writer, historian, and eminent specialist in LiDAR Landscape Archaeology.

Exploring Prehistoric Britain: A Journey Through Time

Other Blogs

2024 Prehistoric Britain Blog Review

8000-year-old Cro-Magnon skull

Vikings, Doggerland, and the Last Survivors of Europe’s Lost Waterworld

Tag: Archaeogenetics

Long barrows, long skulls; round barrows, round skulls—and a biological division archaeology has known about for more than 150 years

1. The evidence was never lost—it was simply left behind

2. The 67 long-bar­row skulls: what Thurnam actually measured

Unchambered long-bar­row crania

Chambered long-bar­row crania

3. The 70 round-bar­row skulls: the population changes

Distribution of the 70 round-bar­row skulls

The complete contrast

Ten named Wiltshire round-bar­row examples

4. From Cro-Magnon to the long-bar­row population

5. The R1b question: what the archaeological record permits us to predict

Villabruna 1: R1b in Upper Palaeolithic Italy

M50: a tall Mesolithic R1b hunter-gatherer

The evidence ladder

Aveline’s Hole: the missing British crossover

Revised evidence ladder

The next scientific step

6. Our new method: estimating a person’s height from the skull

Current three-dimensional model

Applying the method to skulls without surviving body measurements

The importance of Aveline’s Hole

What the estimates suggest

Present conclusion

7. Who built Britain’s megaliths—and what must happen next?

What the evidence currently permits us to say

PODCAST

Author’s Biography

Exploring Prehistoric Britain: A Journey Through Time

Further Reading

Other Blogs

Introduction – A Revolution Built on Probabilities

1. Ancient DNA Is Not a Photograph of the Past

2. The Bayesian Revolution

3. When Models Begin Reinforcing Models

4. Britain’s Oldest R1b Changes the Starting Point

5. Then More Early R1b Appeared Across Europe

6. The Statistical Illusion

7. If Not Replacement, Then What?

8. Time to Rethink European Prehistory

Appendix A – Confirmed Pre-Bell Beaker R1b Individuals Included in the Present Analysis

PODCAST

Author’s Biography

Exploring Prehistoric Britain: A Journey Through Time

Further Reading

Other Blogs

CSI Cheddar Man – Examining the Evidence

Exhibit A: Eye Colour

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Exhibit B: Hair Colour

Report Statement

Report Statement

Report Statement

Report Statement

CSI Question

Exhibit C: Skin Pigmentation

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Report Statement

Exhibit D: The Reliability Problem

CSI Verdict

How DNA Predictions Actually Work

DNA Does Not Contain Labels

Why Multiple Markers Are Needed

Why The Results Are Not Yes Or No

What Is Coverage?

Why Ancient DNA Is Different

2. The 67 long-barrow skulls: what Thurnam actually measured

Unchambered long-barrow crania

Chambered long-barrow crania

3. The 70 round-barrow skulls: the population changes

Distribution of the 70 round-barrow skulls

Ten named Wiltshire round-barrow examples

4. From Cro-Magnon to the long-barrow population