Before humans milked cows, herded goats or raised hogs, before they invented agriculture, or written language, before they had permanent homes, and most certainly before they had cats, they had dogs.

Or dogs had them, depending on how you view the human-canine arrangement. But scientists are still debating exactly when and where the ancient bond originated. And a large new study being run out of the University of Oxford here, with collaborators around the world, may soon provide some answers.

Scientists have come up with a broad picture of the origins of dogs. First off, researchers agree that they evolved from ancient wolves. Scientists once thought that some visionary hunter-gatherer nabbed a wolf puppy from its den one day and started raising tamer and tamer wolves, taking the first steps on the long road to leashes and flea collars. This is oversimplified, of course, but the essence of the idea is that people actively bred wolves to become dogs just the way they now breed dogs to be tiny or large, or to herd sheep.

The prevailing scientific opinion now, however, is that this origin story does not pass muster. Wolves are hard to tame, even as puppies, and many researchers find it much more plausible that dogs, in effect, invented themselves.

Imagine that some ancient wolves were slightly less timid around nomadic hunters and scavenged regularly from their kills and camps, and gradually evolved to become tamer and tamer, producing lots of offspring because of the relatively easy pickings. At some point, they became the tail-wagging beggar now celebrated as man’s best friend.

Some researchers question whether dogs experience feelings like love and loyalty, or whether their winning ways are just a matter of instincts that evolved because being a hanger-on is an easier way to make a living than running down elk. Raymond Coppinger, a professor emeritus of biology at Hampshire College, noted in his landmark 2001 book, “Dogs,” that “best friend” is not an “ecological definition.” And he suggested that “the domestic house dog may have evolved into a parasite.”

Researchers also point out that of the estimated one billion dogs in the world, only a quarter of them are pets. The vast majority of dogs run free in villages, scavenge food at dumps, cadge the odd handout and cause tens of thousands of human deaths each year from rabies. They are sometimes friendly, but not really friends.

Modern dogs are different from modern wolves in numerous ways. They eat comfortably in the presence of people, whereas wolves do not. Their skulls are wider and snouts shorter. They do not live in pack structures when they are on their own, and so some scientists scoff at dog-training approaches that require the human to act as pack leader.

Wolves mate for the long haul and wolf dads help with the young, while dogs are completely promiscuous and the males pay no attention to their offspring. Still, dogs and wolves interbreed easily and some scientists are not convinced that the two are even different species, a skepticism that reflects broader debates in science about how to define a species, and how much the category is a fact of nature as opposed to an arbitrary line drawn by humans.

Tracing the Origins

If current divisions between species are murky, the past lies in deep darkness. Scientists generally agree that there is good evidence that dogs were domesticated around 15,000 years ago. By 14,000 years ago, people were burying dogs, sometimes along with humans. But some biologists argue, based on DNA evidence and the shape of ancient skulls, that dog domestication occurred well over 30,000 years ago.

And as to where the process occurred, researchers studying dog and wolf DNA — most of it modern but some from ancient sources — have argued in recent years that dogs originated in East Asia, Mongolia, Siberia, Europe and Africa.

One reason for the conflicting theories, according to Greger Larson, a biologist in the archaeology department at the University of Oxford, is that dog genetics are a mess. In an interview at his office here in November, he noted that most dog breeds were invented in the 19th century during a period of dog obsession that he called “the giant whirlwind blender of the European crazy Victorian dog-breeding frenzy.”

That blender, as well as random breeding by dogs themselves, and interbreeding with wolves at different times over at least the last 15,000 years, created a “tomato soup” of dog genetics, for which the ingredients are very hard to identify, Dr. Larson said.

The way to find the recipe, Dr. Larson is convinced, is to create a large database of ancient DNA to add to the soup of modern canine genetics. And with a colleague, Keith Dobney at the University of Aberdeen, he has persuaded the Who’s Who of dog researchers to joina broad project, with about $2.5 million in funding from the Natural Environment Research Council in England and the European Research Council, to analyze ancient bones and their DNA.

Robert Wayne, an evolutionary biologist at U.C.L.A. who studies the origin of dogs and is part of the research, said, “There’s hardly a person working in canine genetics that’s not working on that project.”

That is something of a triumph, given the many competing theories in this field. “Almost every group has a different origination hypothesis,” he said.

But Dr. Larson has sold them all on the simple notion that the more data they have, the more cooperative the effort is, the better the answers are going to be. His personality has been crucial to promoting the team effort, said Dr. Wayne, who described Dr. Larson as “very outgoing, gregarious.” Also, Dr. Wayne added, “He has managed not to alienate anyone.”

Scientists at museums and universities who are part of the project are opening up their collections. So to gather data, Dr. Larson and his team at Oxford have traveled the world, collecting tiny samples of bone and measurements of teeth, jaws and occasionally nearly complete skulls from old and recent dogs, wolves and canids that could fall into either category. The collection phase is almost done, said Dr. Larson, who expects to end up with DNA from about 1,500 samples, and photographs and detailed measurements of several thousand.

Scientific papers will start to emerge in 2016 from the work, some originating in Oxford, and some from other institutions, all the work of many collaborators.

Dr. Larson is gambling that the project will be able to determine whether the domestication process occurred closer to 15,000 or 30,000 years ago, and in what region it took place. That’s not quite the date, GPS location and name of the ancient hunter that some dog lovers might hope for.

But it would be a major achievement in the world of canine science, and a landmark in the analysis of ancient DNA to show evolution, migrations and descent, much as studies of ancient hominid DNA have shown how ancient humans populated the globe and interbred with Neanderthals.

And why care about the domestication of dogs, beyond the obsessive interest so many people have in their pets? The emergence of dogs may have been a watershed.

“Maybe dog domestication on some level kicks off this whole change in the way that humans are involved and responding to and interacting with their environment,” he added. “I don’t think that’s outlandish.”

Shepherding the Research

Dr. Larson is no stranger to widely varying points of view. He is an American, but recently became a British citizen as well. His parents are American and he visited the United States often as a child, but he was born in Bahrain and grew up in Turkey and Japan, places where his parents were teaching in schools on American military bases.

He graduated from Claremont McKenna College in California and received his Ph.D. at Oxford. In between college and graduate studies, he spent a year searching for the bed of an ancient river in Turkmenistan, and another couple of years setting up an environmental consulting office in Azerbaijan. He had an interest in science as an undergraduate, and some background from a college major in environment, economics and politics, but no set career plans. Instead, his career grew out of intense curiosity, a knack for making friends and a willingness to jump at an opportunity, like the time he managed to tag along on an archaeological dig.

He was staying in Ashgabat, Turkmenistan, and a local man who had helped him rent an old Soviet truck to explore the desert told him some Westerners were arriving to go on a dig, so he wangled his way onto one of the trucks.

“I think everybody there thought I was with somebody else,” Dr. Larson said.

By the time the group stopped to rest and someone asked him who he was, it was too late to question whether he really belonged. “I was a complete stowaway,” he said.

But he could move dirt and speak Russian, and he had some recently acquired expertise — in college drinking games — that he said was in great demand at night. By luck, he said, the researchers on the dig turned out to be “the great and the good of British neolithic archaeology.” One of them was Chris Gosden, the chairman of European Archaeology at Oxford, who later invited him to do a one-year master’s degree in archaeology at Oxford. That eventually led to a doctoral program after he spent some time in graduate school in the United States.

The current project began when he became fed up with the lack of ancient DNA evidence in papers about the origin of dogs. He called Dr. Dobney, of the University of Aberdeen in 2011, and said, “We’re doing dogs.”

After receiving the grant from the council in England, he and Dr. Dobney organized a conference in Aberdeen, Scotland, to gather as many people involved in researching dog origins as they could. His pitch to the group was that despite their different points of view, everyone was interested in the best possible evidence, no matter where it led.

“If we have to eat crow, we eat crow,” he said. “It’s science.”

A 32,000-Year-Old Skull

Mietje Germonpré, a paleontologist at the Royal Belgian Institute of Natural Sciences, is one of the many scientists participating in the dog project. She was one of a number of authors on a 2013 paper in Science that identified a skull about 32,000 years old from a Belgian cave in Goyet as an early dog. Dr. Wayne at U.C.L.A. was the senior author on the paper and Olaf Thalmann from the University of Turku in Finland was the first author.

It is typical of Dr. Larson’s dog project that although he disagreed with the findings of the paper, arguing that the evidence just wasn’t there to call the Goyet skull a dog, all of the authors of the paper are working on the larger project with him.

In November in Brussels, holding the priceless fossil, Dr. Germonpré pointed out the wide skull, crowded teeth and short snout of the ancient skull — all indicators to her that it was not a wolf.

“To me, it’s a dog,” she said. Studies of mitochondrial DNA, passed down from females only, also indicated the skull was not a wolf, according to the 2013 paper.

Dr. Germonpré said she thinks dogs were domesticated some time before this animal died, and she leans toward the idea that humans intentionally bred them from wolves.

She holds up another piece of evidence, a reconstruction of a 30,000-year-old canid skull found near Predmostí, in the Czech Republic, with a bone in its mouth. She reported in 2014 that this was a dog. And she says the bone is part of evidence the animal was buried with care. “We think it was deliberately put there,” she said.

But she recognizes these claims are controversial and is willing, like the rest of the world of canine science, to risk damage to the fossils themselves to get more information on not just the mitochondrial DNA but also the nuclear DNA.

To minimize that risk, she talked with Ardern Hulme-Beaman, a postdoctoral researcher with the Oxford team, about where to cut into it. He was nearing the end of months of traveling to Russia, Turkey, the United States and all over Europe to take samples of canid jaws and skulls.

He and Allowen Evin, now with the National Center for Scientific Research in Montpellier, France, also took many photographs of each jaw and skull to do geometric morphometrics. Software processes detailed photographs from every angle into 3-D recreations that provide much more information on the shape of a bone than length and width measurements.

Dr. Germonpré and Dr. Hulme-Beaman agreed on a spot in the interior of the skull to cut. In the laboratory, he used a small electric drill with a cutting blade to remove a chunk the size of a bit of chopped walnut. An acrid, burning smell indicated that organic material was intact within the bone — a good sign for the potential retrieval of DNA.

Back in Oxford, researchers will attempt to use the most current techniques to get as much DNA as possible out of the sample. There is no stretch of code that says “wolf” or “dog,” any more than there is a single skull feature that defines a category. What geneticists try to establish is how different the DNA of one animal is from another. Adding ancient DNA gives many more points of reference over a long time span.

Dr. Larson hopes that he and his collaborators will be able to identify a section of DNA in some ancient wolves that was passed on to more doglike descendants and eventually to modern dogs. And he hopes they will be able to identify changes in the skulls or jaws of those wolves that show shifts to more doglike shapes, helping to narrow the origins of domestication.

The usual assumption about domestic animals is that the process of taming and breeding them happened once. But that’s not necessarily so. Dr. Larson and Dr. Dobney showed that pigs were domesticated twice, once in Anatolia and once in China. The same could be true of dogs.

Only the Beginning

Although the gathering of old bones is almost done, Dr. Larson is still negotiating with Chinese researchers for samples from that part of the world, which he says are necessary. But he hopes they will come.

If all goes well, said Dr. Larson, the project will publish a flagship paper from all of the participants describing their general findings. And over the next couple of years, researchers, all using the common data, will continue to publish separate findings.

Other large collaborative efforts are brewing, as well. Dr. Wayne, at U.C.L.A., said that a group in China was forming with the goal of sequencing 10,000 dog genomes. He and Dr. Larson are part of that group.

Last fall, Dr. Larson was becoming more excited with each new bit of data, but not yet ready to tip his hand about what conclusions the data may warrant, or how significant they will be.

But he is growing increasingly confident that they will find what they want, and come close to settling the thorny question of when and where the tearing power of a wolf jaw first gave way to the persuasive force of a nudge from a dog’s cold nose.

“I’m starting to drink my own Kool-Aid,” he said.