In 1914, quarry workers at Oberkassel near Bonn uncovered a late Paleolithic double human burial. Among the bones was a dog, and for most of a century it stayed a footnote. Then a reanalysis led by Luc Janssens, summarized by University College London in 2018, put the grave at roughly 14,223 years old, found remains of two dogs rather than one, and read the younger animal's teeth as a record of repeated illness, most likely canine distemper, severe enough that the team concluded it could not have survived without sustained human nursing. Somebody kept a dying puppy alive through a Pleistocene winter, and the animal ended up in a grave with two people. Every genome sequenced since has made that story older and considerably less tidy.

One origin, and a wolf population nobody has found

The headline finding of modern dog genomics is a negative one: no living wolf population is the ancestor. In 2020 a team led by Anders Bergström sequenced 27 ancient dog genomes and reported in Science that all the dogs sampled share a single common ancestry distinct from present-day wolves, with limited wolf gene flow since domestication. The source population appears to be gone. A 2022 follow-up in Nature, built on 72 ancient wolf genomes spanning the last 100,000 years, narrowed the geography without closing the case: dogs are overall closer to ancient wolves of eastern Eurasia, which points east for the domestication event, while dogs of the Near East and Africa draw up to half their ancestry from a second population related to modern southwest Eurasian wolves. None of the 72 wolves was a direct match for either ancestry. We know roughly where to look and have not yet found the animal.

Five kinds of dog before the first farm

The other surprise in the 2020 data is chronological. By 11,000 years ago, before agriculture had spread and while much of the world was still Mesolithic, at least five major dog ancestry lineages already existed, running from the Levant to Lake Baikal to the Americas and out toward the forebears of the New Guinea singing dog. Dog diversity is not a Victorian invention layered onto uniform stock; it is Paleolithic in depth, and the breeds we argue about today are a thin recent film on top of it.

The gene that made a four-pound dog possible

Then there is the part that belongs to a chihuahua library specifically. Body size in dogs is governed to an unusual degree by a single locus, and in 2022 Jocelyn Plassais and colleagues reported in Current Biology that the decisive variant sits not in the IGF1 growth-factor gene itself but on the opposite strand, in a long non-coding RNA now labeled IGF1-AS. Two alleles, one predisposing to small size and one to large. The small-size version is the ancestral one, and a Pleistocene Siberian wolf that lived roughly 53,000 years ago carried a copy of it.

Read in that order, the finding rearranges the usual assumption. Humans did not invent the small dog by grinding wolves down over centuries; the raw material was already segregating in wolf populations long before anyone thought to select for it, and breeders located the variant and concentrated it. The chihuahua sits at the far end of that concentration, per our facts pillar, which is a curiosity, not a license: the variant explains why small dogs are possible, not why ever-smaller dogs are advisable. Our size guide and health library cover what happens when miniaturization is pushed past what the skeleton and the blood sugar can carry.

The American chapter, and where the chihuahua fits

Dogs reached the Americas with people. The 2018 study led by Máire Ní Leathlobhair in Science found that pre-contact American dogs formed a single lineage most likely originating in Siberia, tracing to a common ancestor around 14,600 years ago. The unsentimental part is what came next: that lineage has essentially vanished from modern American dogs of European descent, persisting only in modest fractions in Arctic breeds, and its closest detectable living relative is a transmissible canine cancer lineage traced to one dog thousands of years back. Population history rarely flatters anyone.

Which puts the chihuahua in an interesting position. The American Kennel Club describes the breed as among the oldest of the Americas, with a lineage reaching back to pre-Columbian kingdoms, and the maternal evidence supports it: a 2013 mitochondrial study led by Barbara van Asch in Proceedings of the Royal Society B found that the most common haplotype in their chihuahua sample also appeared in a pre-Columbian Mexican dog. That is a direct maternal thread from ancient Mexico to the dog on your sofa. It is also one thread. Mitochondrial DNA tracks mothers only, while the nuclear picture shows heavy replacement across the Americas. The defensible summary: a documented pre-Columbian maternal inheritance inside a genome that is substantially post-contact, which beats both the romantic version and the debunking version, and our origins guide works through the Techichi question in detail.

Frequently asked questions

Are dogs really descended from wolves?

Yes, and that part is not in dispute; the open question is which wolves. The ancient-genome work indicates dogs descend from a wolf population that no longer exists, and the 2022 survey of 72 ancient wolf genomes found no direct match among them, so the specific progenitor group is still missing from the record.

How long have dogs lived alongside people?

The Bonn-Oberkassel burial sets a firm floor of roughly 14,200 years: a dog that lived with humans, fell ill, was cared for, and was buried among them. Genetic divergence estimates run earlier and are actively argued over in the literature; the archaeology is simply what is securely dated.

Is the chihuahua an ancient breed?

The lineage has ancient roots; the breed as a registered standard does not. Maternal DNA ties chihuahuas to pre-Columbian Mexican dogs, but a breed with a written standard is a nineteenth and twentieth century construction, and no breed has existed unchanged for millennia. Ancient ancestry and ancient breed are two different claims.

Does the ancient size gene mean very tiny dogs are natural?

It means the variant is old, not that extreme miniaturization is harmless. The IGF1-AS allele permitting small body size was circulating in wolves long before domestication, but stacking every factor that produces a two-pound dog is recent work, and the fragile bones, dental crowding, and low blood sugar risk that follow are veterinary problems rather than heritage. Any tiny puppy showing weakness, tremors, or collapse needs a veterinarian promptly.

Fourteen thousand years ago someone in what is now Germany nursed a sick puppy through weeks it had no business surviving, and that dog went into a grave with two people. Everything after, the five Paleolithic lineages, the missing wolves, the body-size variant older than the last glacial maximum, the maternal thread running from pre-Columbian Mexico to a modern living room, is commentary on that grave. The dog asleep against your leg is the current draft of a fourteen-millennium document, and she is, genetically and taxonomically, every bit the dog buried at Oberkassel; our taxonomy piece is there should anyone at a barbecue suggest otherwise.