We know quite a bit about dog biology. We know that, though a chihuahua and a chow chow are dramatically different in size and shape, dogs share all but 0.15 percent of their genome. We know that a dog's sense of smell is so refined it can sniff the passage of time. We know that dogs were not always dogs.

When humans tamed wolves, dogs became the first domesticated animal. Dogs are, if not our best friends, at least our oldest companions. But we don't know the full history of dogs. Not clearly.

To find answers, scientists have mined modern dog genomes as well as ancient bones. In 2015, researchers traced dogs' peak genetic diversity to Asia, suggesting that the first dogs appeared in the region where Nepal and Mongolia are today. In 2016 another team of scientists argued that humans domesticated dogs from wolves not once but twice. These experts hypothesized that people in Asia transformed wolves to dogs, as did a separate group of people in ancient Europe.

Scientists have proposed "basically every region" in Europe, the Middle East and Asia as the birthplace of the domesticated dog, said Stony Brook University genomicist Krishna Veeramah. In a new study in the journal Nature Communications, Veeramah and his colleagues write that the most plausible explanation was a single instance of domestication. But they did not make any claims as to where dogs split from wolves, he noted.

The wolf-dog split is a tricky one. "Dogs and wolves are promiscuous, and the species boundary between them is a leaky one," said Adam Boyko, an expert on dog genetics at Cornell University in New York and an author of the 2015 study. "To further complicate it, it seems like dogs were domesticated from a wolf population that has likely gone extinct in the wild."

The new work focused more on the when than the where. Dogs separated from wolves between 36,900 to 41,500 years ago, a date that Veeramah and his colleagues reached by calculating the rate of canine mutations over time. They compared DNA from ancient specimens with modern genetic data to create what's known as a molecular clock.

Within the last three years, scientists have been able to tease more information from ancient bones by focusing on a section of skull called the petrous bone. You can feel it just beneath your skin - it's the bone that curls around the ear like a nubby stalactite. The petrous bone is the densest in the body. As an archaeological specimen, its density thwarts bacterial infiltration and degradation.

It used to be that teeth or long bones were prized specimens, Veeramah said. "Now everyone just does everything with the petrous bone," he said.

One of Veeramah's colleagues found a petrous bone in a German archaeological site named Herxheim (a Neolithic pit full of human bones that may have been willing sacrifices or, possibly, butchered by ancient cannibals). The dog bone, radiocarbon dated to 7,000 years old, "had just incredible amounts of DNA," Veeramah said. It represented close to 70 percent of the dog's genome.

The biologists also analyzed a 5,000-year-old dog skull from Cherry Tree Cave in Germany. They combined this data with that from a dog bone found in Newgrange, Ireland, which was included in the 2016 dual-domestication report. They compared these three with snippets of genetic info from 5,600 wolves and modern dogs, plus nearly 100 complete canine genomes. By tracking the mutation rates, the scientists determined dogs probably split from wolves about 40,000 years ago. Dogs divided into two groups - European and Asian groups - about 20,000 years afterward, the authors said.

The DNA revealed that the German animals "looked like modern dogs, in particular like European dogs," Veeramah said. The study was "concordant" with the idea of a single origin for the wolf-dog split, he added.

"Given the high degree of sharing of sweeps," which is to say genetic signatures, "between these ancient samples and modern samples, it seems clear that these dogs descend from a single domestication origin," Boyko said in an email. It does not rule out a separate domestication event, he said, but that event would have "contributed little if any genetic material to these genomes."

University of Oxford evolutionary biologist Greger Larson said that Veeramah's paper "explicitly comes after our 2016 paper," which proposed two separate domestication events. Larson said that the older study presented a hypothesis, not an ironclad fact, based on the deep genetic differences between dogs in Europe and Asia.

The new research failed to slam the door on the 2016 hypothesis, in Larson's view. "They don't provide an alternative scenario" for the "ridiculously deep split" between Eastern and Western dog genetics, he said. Nor does it counter curiosities in the archaeological record. Though it's possible to find ancient dog bones in the far corners of Eurasia, Larson noted, the middle of the continent is suspiciously empty.

"We didn't have a smoking gun," Larson said. "And they don't have a smoking gun."

Both parties were polite in their scientific disagreement. Veeramah and Larson each lauded the other's work, even if they diverged on the conclusions. Larson said that it was "outstanding" to have the new analysis of the German dog bones. Veeramah complimented Larson's ambitious program at Oxford University to collect and share information on modern and ancient dogs, calling it a "wonderful project."

The new research also demonstrated the meandering path of domestication. As wild predators transformed into floppy-eared, tail-wagging pooches, along the way the dogs picked up multiple genes for a starch-rich diet. Both the 7,000-year-old bone and the 5,000-year-old skull lacked duplicate genes seen in modern dogs. These were selected much later in dog evolution, then, than expected.

Veeramah plans to study dogs later in antiquity, during the Roman empire. "Dogs were doing all sorts of things," he said. "They were involved in battle, in war."

We know that dogs were important to us, Boyko said, but we still don't really know why we domesticated dogs before all other animals. "If we replayed the history of man, would dogs still wind up at our side? If we didn't have dogs, how different would modern man look?" he asked.

These questions border on the philosophical, Boyko said, but knowing the "where, when, how and why" of dogs would give us "really useful insights" into domestication. And as for the potency of domestication, he said, you could argue we have done it to ourselves, too.