A 3,700-Year-Old Sheep Virus in Medieval Books Fits the Flood Timeline

Parchment, pathogens, and the same dating strategy we have used all along

Holy books have a way of telling more than the scribes intended. A recent Popular Mechanics report, built on a Science Advances study, describes sheeppox virus DNA recovered from medieval parchment and from Bronze Age sheep teeth. The magazine headline leans on a modern Greek outbreak that killed hundreds of thousands of animals. The science underneath is more interesting than the headline. Researchers rubbed centuries-old gospel pages with ordinary conservation erasers, pulled DNA from the crumbs, and reconstructed genomes of a livestock plague that shepherds once blamed on witchcraft.

We have been saying for years that genetic change is measurable, that the rates we can actually count do not require millions of years, and that the Bible already told us when the post-Flood world began. Dr. Nathaniel Jeanson’s pedigree work on human Y-chromosome and mitochondrial DNA has pressed that same point with living people. This sheeppox paper does not set out to confirm Genesis. Several of its own numbers do anyway. Where the authors leave the biblical timeline, they do so by assumptions, not by the DNA sitting on the page.

What they actually found

Sheeppox virus (SPPV) is a large double-stranded DNA virus in the capripoxvirus group, related to goatpox and lumpy skin disease virus, and more distantly to the human smallpox virus. It is not an RNA flu. It does not reinvent itself every season. It attacks sheep with fever, lesions, and damage to lungs and gut. In a flock that has never seen it, losses of 75 to 90 percent are not unusual. Milk, wool, and meat all fall. Skins taken from sick animals can carry the scars, and they can carry the virus.

The team, led from University College Dublin, sampled parchment from manuscripts that include the York Gospels (about A.D. 990 to 1020), the Corpus Glossary, and other British and continental books from roughly the eighth to the fifteenth centuries. The method is almost humble. Conservators already clean pages with PVC erasers. The researchers extracted DNA from that eraser dust, built sequencing libraries, captured capripox sequences with tiled RNA baits, and authenticated the reads by the damage patterns that mark truly ancient DNA. They also powdered pulp from Bronze Age sheep teeth taken from pastoral sites on the Eurasian steppe, as far as the Altai, dated near 1700 B.C.

They report 21 ancient SPPV genomes. Multiple pages of the same book can be positive, which means infected animals were being turned into writing material often enough to leave a trail. Some calfskin and goatskin pages also carried SPPV reads, which may be contamination in the workshop or occasional spillover. The oldest high-quality genome still has the same nine inactivated genes that modern SPPV carries. After that early loss of function, the gene count stays remarkably steady. The authors themselves contrast that stability with human smallpox, which kept shedding genes. That contrast is not a side note. It is one of the most important biological results in the paper, and it runs the opposite direction from molecules-to-man evolution.

That is the data. Books made from animal skins are biological archives. The virus was in medieval Europe. It was already in steppe flocks by the time those Bronze Age teeth formed. None of that requires a war with Scripture.

The numbers that already agree with Scripture

Genesis gives a clear post-Flood clock. Adding the ages in Genesis 5 and 11 on the Masoretic numbers, and tying that line to the known date of Solomon’s temple, places the Flood near 2348 B.C., about 4,370 years before our day. Creation stands near 4004 B.C. Livestock as we meet them in the post-Flood world come off the Ark and spread with the families of Noah. A virus of domestic sheep has to live inside that window, not in a 10,000-year Neolithic dreamtime.

Now look at what the paper’s own clock produced.

Using the dated parchment and the dated teeth as tips on the tree, the authors ran a time-calibrated analysis in BEAST and estimated a mean SPPV substitution rate of 9.98×10−69.98 \times 10^{-6}9.98×10−6 substitutions per site per year. With that rate, the sampled SPPV genomes coalesce at about 4,080 years before the present. The 95 percent range runs roughly 5,137 to 3,636 years. The median sits on the Flood. The range is the Flood window, not the age of dinosaurs.

That is the same kind of result Jeanson has been publishing for human paternal DNA. High-coverage father-son studies hide a fast Y-chromosome rate, on the order of three mutations per generation. Applied to the global tree of living men, that rate brings Y-chromosome Adam to about 4,500 years, the Flood bottleneck, not a 200,000-year evolutionary Adam. Mitochondrial pedigree rates do the same service for the maternal line and land near 6,000 years. Jeanson’s method is not a different chemistry. It is the refusal to throw away the rate you can measure in the present in order to protect a long age you assumed in the past.

The sheeppox team measured a rate on samples whose dates, from the medieval books at least, are historical. They then watched their SPPV tree collapse to about 4,080 years. That is consistency, not coincidence. A post-Flood world full of expanding flocks, seasonal contact on the steppe, and later European herds packed tight enough to feed monasteries with skins, is exactly where a host-adapted poxvirus would leave genomes this close together.

The Bronze Age teeth near 1700 B.C. also fit. That date is only about 650 years after the Flood. Shem’s descendants are already spreading. Pastoral societies on the steppe are exactly what Genesis 10 and 11 lead us to expect once the families move east. We do not need a separate “invention of wool” five millennia before Abraham to explain a sick sheep in Siberia.

The stability finding fits as well. God built kinds with room to adapt, not with a license to become other kinds. We have written about that built-in adaptability after the Flood: regulatory switches, standing variation, and rapid sorting as animals filled emptied landscapes. A poxvirus that lost nine genes early, then stopped remodeling its core, looks like a pathogen that finished its host fit in the first centuries after the animals left the Ark, then rode the flocks with only modest drift. That is not evolution from a primordial soup. That is a created system absorbing a curse and remaining itself.

What “shedding genes” actually means

When the paper says smallpox kept “shedding genes,” it does not mean the virus invented new machinery. It means working genes were broken.

A poxvirus genome is a long DNA instruction set, about 150,000 letters in sheeppox. Many of those genes are not the core copy-and-assemble kit. They are extras that help the virus handle a particular host: dampening immune alarms, surviving in a given tissue, or infecting more than one species. A gene is counted as inactivated when a mutation ruins the instruction. The usual wrecking tools are a frameshift (an insertion or deletion that scrambles the reading frame), a premature stop codon (the sentence of the gene is cut off early), or a broken start so the protein is never made. The letters may still sit in the genome as a stump. The function is gone.

That is loss of information. A working gene is specified programming. A stop codon or a frameshift destroys that specification. The virus has fewer usable tools, not more.

Natural selection can still favor the broken copy. Once a virus is locked onto one host, extra host-range genes can become baggage. Breaking them can save resources or change how the virus talks to that host’s immune system. The flock gets sicker. The specialist thrives. None of that writes a new gene. Selection kept a damaged instruction because the damage happened to fit a narrower niche.

The sheeppox data make the direction unmistakable. Modern SPPV has nine dead genes. Those same nine are already dead in the oldest high-quality ancient genome, from a Bronze Age sheep about 3,700 years ago. After that, the number of working genes stays about the same down to living isolates. Sheeppox did its losing early, then stopped. Human smallpox, the authors note, kept inactivating genes as it narrowed onto people. Two members of the same viral family. Same kind of change. Information going down, not up.

If evolution is supposed to mean a climb in functional information, this is counter-evolution. Adaptation by shedding genes is the virus breaking tools it already had. That can make a specialist. It cannot build the tools. It cannot turn a microbe into a man. It cannot turn a generalist pox into a fundamentally new kind of creature. It is the same pattern we keep documenting after the Flood: created systems, standing variation, and loss-of-function tweaks that sort a kind into a new climate or a new host. The curse bites. The design still holds. The information does not grow.

Our strategy and Jeanson’s strategy

The strategy we use at A Flood of Hope, and the strategy Jeanson has pressed in the Answers Research Journal and in Traced, is not complicated.

The sheeppox paper follows steps 2 and 4 for the SPPV clade itself. It breaks steps 1, 3, and 5 when it leaves that clade and starts dating the whole capripox group.

The authors estimate that major capripox lineages (sheeppox, goatpox, and lumpy skin disease) diverged between about 11,500 and 3,700 years ago. In some model runs the median SPPV split from the others sits near 5,000 years. In the widest runs the root is pushed toward 11,500. They then pair that deep split with conventional dates for sheep domestication and the movement of flocks off the steppe into Europe.

That pairing is the hinge. It is not forced by the parchment. It is forced by three assumptions.

Assumption 1. Archaeological dates older than the Flood are treated as raw calendar truth.The 1700 B.C. teeth can be accepted on a biblical timeline. The 11,500-year root cannot. To get that root the model must treat the conventional Holocene as a real 11,000-plus-year corridor and then run the same clock backward through it. Radiocarbon, pottery seriation, and “Neolithic package” dates are themselves calibrated inside the same long-age system. Using them to date a virus is circular when the question on the table is whether that system is true.

Assumption 2. The substitution rate fitted on 1,000 to 4,000 years of tips remains constant all the way to the genus root.Virus work already shows a time-dependent rate pattern. Short windows look fast. Long windows look slow, because the fastest sites saturate and because purifying selection removes many changes before they become fixed. The paper’s own rate, near 10−510^{-5}10−5 substitutions per site per year, is a short-to-medium window rate. Stretching it to 11,500 years is the move Jeanson refuses to make with human DNA. If you keep the fast empirical rate, the deep capripox differences do not buy you extra millennia. They tell you the lineages were already distinct, or became distinct quickly, inside a short real history.

Assumption 3. Every difference among sheeppox, goatpox, and lumpy skin disease must be explained as mutation after a single split, with no created diversity at the kind level.That is the evolutionary default. Scripture does not require it. God made kinds. Pathogens that track those kinds can carry designed differences from the start, just as dog breeds carry standing variation that is not a record of a million-year wolf-to-poodle march. Once you allow created diversity plus rapid post-Flood sorting, the 11,500-year root is no longer a discovery. It is an artifact of forbidding the Bible’s categories.

Assumption 4. Domestication must be a long Holocene process rather than post-Flood husbandry.Abel kept flocks. Noah knew clean and unclean animals. Job’s wealth is measured in sheep and camels. The Bible does not describe a 5,000 B.C. experiment in which hunter-gatherers slowly invent wool. It describes families who already knew livestock walking off an Ark into a scoured world and filling it. Intensive herding on the steppe after the Flood is not a problem for Genesis. Calling that herding 11,000 years old is.

These assumptions are problematic because they decide the answer before the genomes speak. The parchment DNA does not know the Holocene. The teeth do not know the Neolithic. The clock the authors fitted to their own dated SPPV samples already pointed at 4,080 years. Only the decision to keep going, past the Flood and past Creation, produces the 11,500-year headline.

Why the biblical timeline is the right frame

First, it is the record of the God who was there. “For in six days the Lord made heaven and earth, the sea, and all that in them is, and rested the seventh day” (Exodus 20:11). The Flood followed in the days of Noah, “and all flesh died that moved upon the earth” (Genesis 7:21). The animals that walked off the Ark were the stock of every kind we have now. A sheep virus that coalesces near 4,000 years is living in the world those verses describe.

Second, the method is the honest one. Jeanson’s Y-chromosome work and this parchment work agree on the procedure that matters: count change on samples whose ages you have some reason to trust, then stop when the tree meets the history God wrote. The secular habit is the reverse. Assume the deep past, install it as a calibration, then announce that genetics has independently confirmed the deep past. That is not independence. That is a loop.

Third, the biology is what we keep finding when we look without the loop. Soft tissue in dinosaur bone. Carbon-14 in coal and diamond. Heat still in the crust. Ice that reads as post-Flood, not as a million-year archive. Mitochondrial clocks in people, flies, worms, and fleas that land in thousands of years, not in millions. Now a livestock poxvirus whose sampled family tree meets at the Flood. The pattern is not fragile. It repeats.

Fourth, the remaining differences among capripox viruses are not a threat. They are the sort of within-kind diversification we expect when flocks, herds, and people move hard and fast after a global reset. Goatpox and lumpy skin disease can be sister pathogens in the same created grouping, sorted onto goat and cattle kinds as those animals spread. Nine genes already silenced by the time of the oldest tooth means the host fit was largely finished early, and it was finished by breaking instructions, not by writing new ones. That is adaptation after the Flood, the subject we have taught from this same desk. It is not molecules-to-man evolution. When the change we can actually see is the loss of information, the last thing we should do is call it evidence that information built itself.

Fifth, the medieval books themselves are a quiet mercy. Scribes copied gospels onto the skins of sick sheep and, without knowing it, stored a witness. “The grass withereth, the flower fadeth: but the word of our God shall stand for ever” (Isaiah 40:8). The ink preached Christ. The parchment kept a record of the curse that still touches the flocks. Both belong to the same Lord.

What this does not mean

It does not mean the Greek outbreak is ancient DNA crawling out of a library. That outbreak is a modern pastoral disaster, and vaccines and quarantine are still wise. It does not mean every date in the paper is wrong. The York Gospels are a tenth-century book. The virus on those pages is a tenth-century problem. It does not mean Jeanson measured this virus in a father-son pedigree. He did not. The consistency is in the dating strategy and in the SPPV coalescence the authors themselves computed.

It also does not mean we treat Popular Mechanics as a commentary on Genesis. The magazine told a good story about witches and holy books. The hope is not in the story. The hope is in the God who judged the old world by water, who kept a remnant, and who will keep His word until the last page.

The better reading

Read the study this way and the pieces sit still.

Medieval parchment preserved SPPV because infected sheep were common enough to become books. Bronze Age teeth preserved it because post-Flood shepherds on the steppe were already living with the same plague. The measured rate, applied to those samples, brings the SPPV family together near 4,080 years, which is the Flood on the biblical clock. Gene loss finished early and then stopped, which is what a completed adaptation looks like, and it is adaptation by broken instructions, not by new ones. That loss of information is the opposite of the evolutionary story. The only number that fights Scripture is the 11,500-year capripox root, and that number is not a genome. It is an assumption that the Holocene is long, that the clock never changes, and that created kinds may not carry designed differences.

We will keep using the strategy that works. Measure what can be measured. Refuse the circular calibration. Let Genesis name the bottleneck. The books on the monastery shelf, the teeth in the steppe, and the Y chromosomes in living men are not three different histories. They are one history, and it is short, judged, and still open to mercy.

“But Noah found grace in the eyes of the Lord” (Genesis 6:8). The same grace holds the world now, virus and all, until He makes it new.

References and sources

L’Hôte, L., et al. “Three thousand five hundred years of sheeppox virus evolution inferred from archaeological and codicological genomes.” Science Advances.

Rayne, Elizabeth. “A 3,700-Year-Old Virus Hiding in a Medieval Book Just Killed 500,000 Animals.” Popular Mechanics, September 15, 2026.

University College Dublin / EurekAlert press materials on the parchment SPPV study (Louis L’Hôte; Kevin G. Daly).

Jeanson, Nathaniel T., and Ashley D. Holland. “Evidence for a Human Y Chromosome Molecular Clock: Pedigree-Based Mutation Rates Suggest a 4,500-Year History for Human Paternal Inheritance.” Answers Research Journal 12 (2019): 393–404.

Jeanson, Nathaniel T. Traced: Human DNA’s Big Surprise. Master Books, 2021.

Jeanson, Nathaniel T. Mitochondrial pedigree papers in Answers Research Journal applying measured rates across humans and other kinds.

Wright, Robert L. “Adaptation After the Flood.” AFloodOfHope.com, May 17, 2026.