The domestication of the dog is a complex and unresolved topic. But at this point I am convinced that this is one domestication event which well predates agriculture. To some extent this is common sense. There are tentative archaeological finds of domestic dogs in the New World almost immediately after widespread human habitation of the Western hemisphere, >10,000 years ago. More concretely domestic dog DNA has been retrieved from ~9,250 year old coprolites in Texas. The distinctiveness of the New World dogs is well attested genetically. Eskimo dogs for example are nested in a well diverged clade with "ancient dogs" (e.g., Basenji), indicating their early separation from the main Eurasian stock. Additionally, from talking to a dog geneticist I am to understand that the Eskimo dogs themselves are likely new arrivals, and superseded older dog lineages in the far north. These results suggest to me that by the late Pleistocene many of the hunter-gatherer populations of Eurasia existed in a symbiotic relationship with domestic dogs. I suspect that the dog lineages in fact predate the late Pleistocene, and the earliest tendencies toward coexistence with humans may date to before the Last Glacial Maximum. And yet an interesting sidelight to this is that in Australasia it seems that the native domestic dogs, the singing dog and the dingo, postdate agriculture, and derive from Southeast Asian dogs (arriving ~5,000 years ago). Clearly the story of the dogs of the world has several chapters. There are many debates about the exact location of the "first dogs," whether in the Middle East or East Asia. These arguments often seem to place the date of origin of modern dog lineages in the very early Holocene, around the time of agriculture. This contradicts hints from archaeology, and is difficult to square with the likely arrival of dogs with the First Americans. And yet in hindsight the pre-agricultural constraint of the geographic expanse of the dog shouldn't be too surprising (Sub-Saharan African domestic and feral dogs seem to derive from Middle Eastern dogs who arrived through the Nile Valley during the Bronze Age). Wolves, from whom dogs derive, are creatures of the Palearctic ecozone. A later expansion into tropical Asia and Africa is less surprising, as perhaps the lifestyles of human populations in those regions were not congenial to canine hunting companions before agriculture. The phylogeny of dogs can tell us a lot about human migration. But the presence of absence of dogs can also tell us about the constraints of human societies, and how they reshaped their ecology.