Well, since you wanted the long explanation, here it is:
Ill start with the process of domestication (highlighting the study talked about in my earlier post). It is generally accepted that dogs are the descendants of Eurasian grey wolves, and there is evidence that they may have originated in east Asia. The dog may be a very young species at only 14 000 years old, although genetic evidence suggests dog may have began to diverge from the wolf as long as 135 000 years ago. However, even though dogs are the recent descendants of wolves and are very similar to wolves in terms of both behavior and morphology, they are in several ways quite different from wolves. Dogs display a large number of characters not seen in wolves, such as piebald coats, floppy ears, and curly tails. Wolves are also seasonal breeders, with both males and females becoming fertile once per year. Dogs are not seasonal breeders, and most female dogs (with some rare exceptions) come into season twice per year, and male dogs are fertile year round. Dogs also have smaller skulls relative to their overall size than wolves and are able to form social bonds for a longer period of time during their early lives than are wolves.
Interestingly, many of the morphological and physiological differences that exist between dogs and wolves may not have been intentionally selected for by humans, and could have been a result of selection for tameness in dogs. Thats where the article I posted earlier comes in, the experiment which involved Siberian foxes (Vulpes vulpes) demonstrates how this could have happened. The experiment was started in the 1940's by the Russian geneticist Dmitri Belyaev, who studied the process of domestication using a population of fur farm foxes. The foxes used in the beginning stages of the experiment were difficult to handle, very afraid of people and generally behaved like wild animals. The experimenters began to selectively breed the foxes for one trait - tameness around people. At the age of one month, an experimenter would offer food to each fox kit while trying to pet and handle it. This was done twice - while the kit was alone and while it was with other fox kits. This routine was repeated monthly until the kit was seven to eight months old and at that point, each kit was assigned to one of three classes based on how tame it was. Class III foxes attempted to flee from experimenters or tried to bite them. Class II foxes were not friendly to the experimenters, but allowed themselves to be touched. Class I foxes were friendly towards the experimenters and would often approach them. After six generations of breeding only tame foxes, a new class, Class IE, ("domesticated elite") had to be added. These foxes were very dog-like and actively sought out human attention and would lick experimenters and wag their tails like dogs. After twenty generations, 35% of the experimental foxes were domesticated elite and today, 70-80% of the foxes are.
Since the foxes in the experiment were being selectively bred for a behavioral trait (tameness), the experimenters hypothesized that physiological changes in the systems governing the fox's hormones and neurotransmitters would also occur, as an animal's behavior is often mediated by these chemicals. Indeed, that is exactly what happened. As the experiment proceeded, a steady drop in the hormone producing activity of the domestic fox's adrenal glands was measured. For example, after several generations of selective breeding, the basal level of corticosteroids in the blood of the domesticated foxes was far lower than that of the control group of non-domesticated foxes. Changes in the neurochemistry of the domesticated foxes were also noted, as they had higher levels of serotonin in their brains compared to the control group of foxes.
After several generations of selecting for tameness, new traits only rarely seen in wild foxes began to become more common in the domesticated population. For example, after ten generations, several of the domesticated foxes had piebald coloured or brown mottled coats. Later in the experiment, it was noted that several of the tame foxes had floppy ears, short tails or curly tails. Even later, changes in the skull morphology of the foxes was noted as well, as skull measurements showed that the cranial height and width of the domesticated foxes tended to be smaller than those of control group foxes. The domestic foxes also had shorter and wider muzzles than the control group ones.
Many of the differences between the domestic foxes and the wild foxes are similar to the differences seen between domestic dogs and wolves. Wolves do not have floppy ears, curly tails, or piebald colored coats, but many dogs do. Skull size is also one of the main ways dogs differ from wolves and selecting foxes only for tameness changed their skull size. The results of this experiment seem to suggest that many of the unique characters seen in dogs and not wolves are a result of the selection of dogs for tameness. But, how does selecting animals for a behavioral trait change their overall morphology like this? It has already been noted that selecting animals for a behavioral trait can change the amount of hormones and neurotransmitters produced by the animals because an animal's behavior is often controlled by such chemicals. The early development of an animal is also, in part, controlled by these chemicals, so a small change in the animal's endocrine and neurochemical systems may result in changes to the early development of the animal.
Several developmental differences in the domestic foxes compared to the wild foxes were noted in the experiment. The domestic fox kits' eyes opened earlier than the control foxes, and they also responded to noise earlier than the wild foxes. The domestic foxes also began to show a fear of unknown stimuli starting later in life than the non-domestic foxes. This means that the domestic foxes had a longer window of socialization than did the wild foxes. In canids, the window of socialization in which the animal can form social bonds begins when its ears and eyes open and it can explore its environment and closes when it begins to fear novel stimuli. Dogs differ from wolves in the length of their window of socialization. In wolves, this window closes at about three weeks, and it dogs it closes at 8 to 12 weeks, depending on the breed.
Several of the morphological changes seen in the foxes seem to be a result of changes to their early development. Floppy ears and curly tails, for instance, are characteristic of young fox kits and these traits are carried over into adulthood in many of the domestic foxes. The changes noted in the morphology of the fox's skulls may also be a result of early developmental changes, but this conclusion cannot be made since only the skulls of adult foxes were studied. Overall, however, many of the changes to the domestic foxes resemble paedeomorphosis, or the retention of juvenile traits in adults.
Many researchers consider dogs to be paedeomorphic wolves, meaning that they have retained characteristics that are typical of juvenile wolves as adults. For example, the floppy ears that characterize most dog breeds may be paedeomorphic trait, as very young wolf pups have floppy ears, which straighten shortly after birth. Even the erect-eared dogs, such as huskies and German shepherds, have ears which straighten up later than do the ears of wolf pups. The curled sickle tail of most domestic dogs is also a neotenous trait. Adult wolves typically have straight tails that are carried at a downward-pointing angle, whereas wolf pups, like many adult domestic dogs, have tails that are carried up above the back.
The bark of domestic dogs is another juvenile trait. Adult wolves can and do bark as an alarm call, but they rarely do. However, wolf pups bark more often than adult wolves, which makes them similar to many domestic dogs. Adult dogs also have skull characteristics that make them rather similar to four-month-old wolf pups (Coppinger, R. and Coppinger, L., "Dogs: A Startling New Understanding of Canine Origin, Behavior and Evolution." 2001). When a dog reaches four months of age, its head's growth rate slows down relative to the growth rate of the rest of its body. A four month old wolf's head will continue to grow at a more rapid rate relative to a dog's until it is about seven or eight months old. The result is that adult wolf-sized dogs have head sizes that are similar to that of a juvenile wolf's.
On to Artificial vs. Natural selection:
Today, it is clear that the evolution of the dog is being shaped by artificial selection because many dog breeders select dogs for certain working behaviors or, more often, a certain desired look. It is often assumed that humans domesticated dogs by taking wolf pups, taming them and selecting the best tempered wolves for breeding. Gradually, tamer, easier to handle "wolves" (dogs) were produced from this process. However, this may not have been the case, because how and why wolves became associated with human beings is not yet known. The archaeological record gives very little indication as to how the dog diverged from the wolf. Debates on this topic often focus on the issue of intentionally, or whether or not humans consciously began to breed wolves to produce dogs. A scenario outlining how the dog may have evolved by a process of natural selection is described in detail by Coppinger and Coppinger, and Budiansky, S. (The Truth about Dogs: An Inquiry into the Ancestry, Social Conventions, Mental Habits and Moral Fiber of Canis familiaris, 2000). They point out that Palaeolithic and early Neolithic people would have been very unlikely to have raised young wolf pups as pets, since raising wolf pups is an extremely difficult and time consuming task. Wolf puppies must be imprinted on humans starting at about 10 days to ensure that they are not afraid of humans as adults. Such a young puppy would need to be bottle-fed by humans, and pre-historic people probably would not have possessed the equipment needed to do it. Domesticating wolves would have necessarily involved a huge number of wolves as well, and there is no evidence that Palaeolithic people ever kept large numbers of wolves. For example, the experiment described above where domesticated silver foxes were produced in relatively few generations from a starting population of what were essentially wild foxes involved thousands of animals. It is unlikely that prehistoric people could have ever kept very many wolves captive in their settlements.
The dog, then, may be a species that was the result of natural selection. A new "niche" was opened for animals about 10 000 years ago as some groups of humans began to settle into permanent settlements instead of living as nomads. Animals could live around these settlements and scavenge for food bits left around by people. Dogs may have "self-domesticated" when they started scavenging waste from the human settlements that began to appear at this time. For any animal to succeed in such a niche, it would have to be comfortable living and eating in close proximity to humans, as an animal that feared them would not be able to survive in human settlements and would return to living away from people. Wild dogs that were less cautious around people than others may have began to live near humans, and natural selection would gradually produce "proto-dogs" from these animals that did not fear humans and could live in close proximity to them. Indeed, today in several parts of the world dogs do live like this. Many villages across the world are home to dogs that live as scavengers and who are not intentionally cared for by humans. The first primitive dogs were likely very similar to these scavenging village dogs.
These early canids that lived on the fringes of human society scavenging scraps may be the ancestor of early dog breeds. These animals, which were already tame around humans, could be taken and used as hunting companions, or livestock guardians and different dog types likely evolved from these proto-dogs, not wolves. Artificial selection by humans also further shaped the evolution of the dog, as dogs with traits desired by humans would have been better cared for than others, and would have been more likely to survive and breed.