A rather remarkable deficiency

There’s a much ballyhooed article from Science going around that promotes the surprising conclusion that dogs were first domesticated in Europe. Dan Graur points out that there is one little problem with the data:

The take home message of the Thalmann et al. paper is simple: Dogs were not domesticated in the Middle East or China as previously claimed; they were domesticated in Europe. Let me repeat the main result of this paper: Dogs were domesticated in Europe; previous claims on the domestication of dogs in the Middle East or China are wrong and have been refuted.

Interestingly, on page 873, it is written:

“Notably, our ancient panel does not contain specimens from the Middle East or China, two proposed centers of origin (5, 6).”

So, the origin of dogs was moved from the Middle East or China to Europe by the simple expedient of omitting any sample from the Middle East or China.

Well, the paper does have the primary prerequisites for getting published in Science: superficially sexy data sets, involving a familiar large and photogenic animal, an unexpected result, with a high probability of drawing the attention of the mass media. That’s what we mean by “significant research,” right?

Say, isn’t that also the formula for a TED talk?


  1. octopod says

    So there aren’t any dog or wolf specimens >13ka from China or the Middle East. There are in Europe. They used those for their mtDNA tree and SURPRISE! the tree was rooted in Europe.

    Not to be a horrible paleo-snob, but if this is what happens when biologists try to deal with sampling across time, or use data of nonzero age, I think I understand why it’s taken so long for people to start incorporating fossils in their phylogenetic trees…!

  2. alric says

    I don’t see anything nefarious going on. Those were the samples available and the calculated european domestication (35k) is more ancient than the middle eastern archeological dates.

    Researchers that favor the middle eastern hypothesis would still be able to find and analyze DNA samples for comparison with the european ones.

    I don’t have access to the paper but I bet their conclusions are phrased to reflect this.

  3. gillt says

    @6 octopod. Did you take a look at Figure 1 in the paper? I see India, Japan, China, Oman, and others included.

    “These mtDNA assemblies from ancient canids were compared with complete mitochondrial genome sequences from 49 wolves, 77 dogs, including divergent dog breeds such as Basenji, and Dingo; three recently published Chinese indigenous dogs; and four coyotes totaling 148 mitochondrial genomes.”

    Past mitochondrial and Y chromosome analysis that suggested a non-European location for the onset of domestication were more limited in sampling of modern or ancient wolves or prehistoric dogs and had weak statistical support for phylogenetic branching points.

  4. carlie says

    < I think I understand why it’s taken so long for people to start incorporating fossils in their phylogenetic trees…!

    They keep getting pissed off when the actual physical record doesn’t match up with their pretty gene trees coded with all their assumptions.

  5. jakc says

    It seems unlikely that only one group of humans discovered the value of domesticating the wolf. It makes fewer headlines, but isn’t the appropriate claim limited to dogs also being domesticated in Europe? And, even if wolves were domesticated by a single group of humans and then spread only from that group, it defies belief that the domesticated group would not have a significant admixture from undomesticated wolves. As a cladistics analysis that would mean that the last common ancestor of modern dogs is not I’m the group first domesticated, but far prior to that, implying that domestication is not a single event but a continuing event.

  6. beergoggles says

    It’s funny I was just reading an article about how the University of Central Florida is fighting a court order to release the documentation regarding that lying bastard Regnerus’ study. And then I come here and read this post. It’s stuff like this that makes me think there needs to be better transparency in the pre-publication review process.
    Or they can just hit them up for all the related documentation afterwards and watch them squirm I guess..

  7. moarscienceplz says

    C’mon PZ, stop harshing our buzz! Everybody knows all the really cool advancements were made by white, male, preferably blond, preferably blue-eyed Europeans (before they they became even cooler by emigrating to the good ol’ U.S.A. (U.S.A! U.S.A! U.S.A!)

  8. Rob Grigjanis says

    So, the origin of dogs was moved from the Middle East or China to Europe by the simple expedient of omitting any sample from the Middle East or China.

    A dingo ate their samples.

  9. barbara4 says

    I just can’t be all that upset by this paper’s limitations. The authors themselves pointed out the problem of lack of samples from SE Asia — no wolves, no appropriate fossils. Given the intensive human use of the area, the conditions that don’t favor fossilization as well as in some other areas, the tendency to grind up fossils for traditional medicine, and the political limitations, I consider think it plausible that that limitation is real. The authors also discuss why the earlier report of the eastern Asian origin of dogs may be faulty. The authors may be wrong, of course, but this isn’t a “well, that’s stupid!” situation. The authors did a reasonably good job and pointed out its limitations. Now presumably someone will come along with another study, and back and forth it will go until a consensus is reached. We all like to think we’ve produced the last word, but we haven’t.

    “moarscienceplz” — While scientists are strongly prejudiced toward present new!, unexpected! conclusions, we’re not nearly so prejudiced toward having blue-eyed blonds be the source of all. In fact, I would expect that if there’s prejudice toward a conclusion in the next article in this mini-controversy, it will be toward a non-European (new! unexpected!) origin.

    It will be interesting to see what comes next!

  10. alwayscurious says

    Notably, references 5 & 6 are from Nature & Science. The next step should be to demonstrate how changing the assumptions underlying the calculations serve to change (or not change) the location of dog domestication.

    Strike that. Maybe asking genetics “Where?” is asking too much. Say we trace back dog domestication to relatives of an ancient strain of wolf, how do we know that strain of wolf lived in eastern Europe at the time of domestication instead of eastern Asia? They aren’t calling roaming wolf packs for nothing!

    From my skimming, it appears that references 5&6 only looked at genetic evidence in existing dogs & wolves. Then used computer simulations to trace those lineages backwards. This was then compared with the archaeological records to render final verdict. But the latest paper actually claims to have analyzed historical DNA from fossils of known origin. This creates a stronger, direct link between archaeology & genetics. This hardly settles the question, but it does raise the evidence bar up a notch.

  11. David Marjanović says

    It seems unlikely that only one group of humans discovered the value of domesticating the wolf.

    Discovering the value isn’t enough to actually succeed in such a difficult, time-consuming endeavor. AFAIK, all known dogs are more closely related to each other than to any wolf.

  12. jakc says


    Well one thing hunter-gatherers usually have a lot of is time. And let’s not forget that some of the pressure for domestication comes from the wolf. A wolf who can get humans to tolerate it rather than kill it or chase it away, gets a free meal – humans are pretty good hunters. It is less obvious what humans get initially from domestication. I think the question is not the difficulty but the motivation.

    As for the idea that are dogs are more alike than like wolves, I am not clear what the point of that statement. Surely it doesn’t mean that dogs and wolves have been reproductively isolated for a significant time? People are still crossing dogs and wolves, so at the least, for it to be correct, we have to exclude wolf-dog hybrids. The main dog and gray wolf lineages diverged at some point in the past, but there has been continuing admixture since. I don’t see how a purely dog clade can be identified in order to be distinguished from a wolf clade.

  13. chigau (違う) says

    jakc #19
    So you got your information about how good humans-as-hunters are compared to wolves-as-hunters from …
    video games?

  14. jakc says

    Wow. So Chigau, you’re under the impression that hunter gatherers colonized the entire planet, except Antarctica by being bad hunters? That the extinction of megafauna had nothing to do with humans? That say, routinely killing animals that wolves wouldn’t take on isn’t a sign of human hunting skills?

    As for work: the Bushman down in the Kalahari, the closest we have to the old hunter gatherers, spend ten to fifteen hours a week gathering and hunting. That leaves a lot of time for other things. Now farming, that’s labor intensive. Modern society is labor intensive. It’s why we have higher population densities and better toys. You turn Las Vegas into a hunter gatherer society and there is going to be a lot of dead people pretty quickly. But yeah, hunters and gatherers spend less time working than you think