For decades scientists have debated how, where and when the wolf became the dog. Now a new study hints that dogs were domesticated just once, challenging a previous claim about how many times humans befriended canines.

In a paper published this week in Nature Communications Krishna Veeramah at Stony Brook University and colleagues argue that dog domestication occurred once, sometime between 20,000 and 40,000 years ago.

Early efforts at nailing down the time and place of domestication varied wildly. One set of analyses published in the late 1990s suggested dogs and wolves diverged some 135,000 years ago in the Middle East. A 2009 paper placed the divergence much more recently, at 16,300 years ago in southern China. Others have located the series of events that led to canine domestication in Europe, rather than the Middle East or Asia.

Whereas methodological concerns plagued those early studies, modern genomic techniques combined with the ability to extract DNA from well-preserved fossils are inching scientists ever closer to the elusive question of Fido's origins.

At first glance, the findings should not seem all that controversial. Veeramah's timing is well within contemporary estimates. But last year a study published in Science made the argument that dog domestication actually occurred twice: once in Europe and once in Asia.

It all comes down to doggy demography.

In his 2016 study Greger Larson, a biologist at the University of Oxford, compared the genomes of modern dogs from across Eurasia with the genome of an ancient dog whose remains were unearthed at an Irish archaeological site, and estimated that Western and Eastern dog populations diverged between 6,400 and 14,000 years ago, during the Neolithic period. That is later than the earliest archaeological evidence for dogs in both places, which means modern dogs share an ancestor that may in fact be younger than the dogs that had already inhabited both Europe and the Far East during the earlier Stone Age period, the Paleolithic.

Larson takes the findings to suggest that some European wolf gave rise to west Eurasian domestic dogs whereas some Asian wolf gave rise to east Asian domestic dogs. At some point, one of those lineages died out and the other expanded. Based on an analysis of mitochondrial DNA, a collection of genetic material that propagates from mothers to their offspring within each cell’s mitochondria, Larson argued it was the European dog population that was replaced as Asian dogs migrated west with their human masters.

Veeramah is quick to point out Larson's analysis hinged largely on the genome of one ancient pooch, extracted from a 5,000-year-old fossilized ear bone preserved at a Neolithic site in Ireland called Newgrange. The current study utilizes that genome as well as two others extracted from fossils uncovered from Neolithic sites in southern Germany—a canine cranium dated to 5,000 years and a skull fragment from an older site, dated to approximately 7,000 years.

By including additional samples of ancient origin as well as ones from a diverse set of modern dogs from around the world, Veeramah's team was able to more finely hone the domestication story. Indeed, the researchers found evidence for genetic continuity in European dogs, rather than replacement by Asian ones, for at least 7,000 years. “We have a 7,000-year-old sample that was nearly indistinguishable from modern domesticated dogs,” says University of Michigan biologist Amanda Pendleton, a co-author on the study. Genetically, “it looks just like the dog chewing your left shoe in the closet.”

But Larson contends the new paper is generally in agreement with his. “It provides [evidence for] exactly what we did, in terms of the split between East and West,” he says, and the two papers place the dog–wolf divergence at around the same time as well. The key trend, for Larson, is taken from the archaeological record. “You get very old dogs on both sides of the Old World but young dogs in the middle,” he says. To know whether his hypothesis of dual domestication events is ultimately correct would require uncovering Paleolithic dog fossils from Europe to see how similar—or different—they are from the younger Neolithic ones. If his hypothesis is right, the oldest dogs in western Europe would look very different from their Neolithic counterparts.

Veeramah says the baseline assumption should be a single domestication process in a single place, although he is wary about saying exactly where that might have occurred. “The burden of proof is higher” to support a scenario in which multiple domesticated dog lineages emerged independently in different parts of the planet, he says.

Part of the trouble with accurately charting the evolutionary history of domestic dogs is that animal carcasses do not fossilize very well in some climates (like the tropics), which skews the archaeological record toward those places that do reliably turn up high-quality fossils. And not all ancient human cultures took equal care to preserve the remains of their dogs, even in places where fossilization is a possibility. Moreover, it is not enough for a bone to fossilize; it has to become preserved in such a way that DNA remains extractable. So although there are Paleolithic fossils as old as 30,000 years, finding one with usable DNA fragments is a separate matter entirely.

For the moment, the precise nature of Fido's family tree remains somewhat shrouded in mystery. As more and more fossils are found and their DNA is sequenced, however, the origins of domestic dogs will inevitably become clearer. And then we'll be able to say with greater certainty just how that dog chewing your left shoe in the closet became your best friend.