So when did the dog start to become 'the dog' and where did it come from,
in terms of genetic lineage? 
First things
first, where did it come from?. The opinion of most people looking
into the subject (and my own personal opinion) is that the dog evolved
from the wolf. This opinion is almost commonly accepted as fact these
days although there are other opinions. Back in the 1950's the well known
animal behaviourist konrad lorenz (who won the nobel prize in 1973
for medicine and psychology) suggested that some breeds of dog were descended
from wolves, and some from the jackal (lorenz, 1954). This suggestion
fostered several ideas and opinions on the subject although he himself
rejected the idea that some breeds were descended from the jackal
once he became aware of the complicated howling repertoire of the jackal
which is quite distinct from patterns found in the domestic dog (serpell,
1995). studies conducted on the dog concerning behaviour (scott and fuller,
1965), vocolisations (lorenz, 1975), morphology
(wayne 1986) and molecular biology (wayne and obrien, 1987) all suggest
that it is the wolf which is the likely progenetor of the domestic dog
as we know it today.
The second question concerning the evolution of the domestic dog, or the domestication, is when did it begin, i.e. when did wolves begin to become domesticated?
Firstly let us consider some facts. the bones of wolves (displaying no morphological evolution toward a dog like state) have been found in association with early hominids bones as far back as 400,000 years ago in an area of kent, in england (serpell, 1995). This by no means suggests that domestication began as far back as this, but does demonstrate that the social spheres of both animals at least overlapped this long ago.
It is in
terms of noticing morphological differences in bone structure,
showing the beginning of a change into the dog as we know it today, that
we can estimate when the divergence was likely to have begun. The earliest
bones found to date showing morphological differences can be estimated
at around 14,000 years old (nobis, 1979). The mandible (jaw bone)
was excavated from a site in germany and the age would date it around
the late paleolithic era. In addition to this an elderly human skeleton
has been excavated in israel lying in a fetal position with the left arm
resting on the thorax of an approximately 4 month old puppy (davis and
valla, 1978). This site was dated at around 12,000 years old. 
So how and why did the domestication begin in the first place?
My speculation for domestication follows the thoughts of many and although these popular theories may seem both rational and likely to some, there are always opinions out there suggesting otherwise. The jigsaw puzzle can be pieced together but unfortunately no one will be able to say exactly how it happened, unless of course, we travel back in time. so, I shall have to speculate like everyone else.
It
is suggested by budiansky (2000) that dogs are social parasites.
In a way this may not be too far from the truth as early man probably
came into contact with wolves as they began to scavenge around human
settlements. The wolves may even have been hunted for food and clothing
and in the process pups may have been introduced to family settlements.
As is strikingly obvious, less controllable wolves would inevitably be
killed, and tameness favoured.
An interesting experiment (belyaev, 1979) showed over a period of 20 years that foxes selected only for tameness and no other trait, began to show interesting dog like traits such as piebald coats, colouring, drooping ears and a diestrus mating cycle.
When wolves
began to scavenge from human settlements or kills, it would confer
a positive advantage for those tamer wolves to have pups closer
to humans. The distance to travel for food would be less, so less
exerting on the mother and pups (provided of course that there wouldnt
be increasing stress if they were beginning to associate more with humans).
Canids have a reasonably long growth and development period. the reproductive
success of canids is closely correlated with the ability of the
mother to provision the pups and herself (macdonald and moehlman,
1983). Wolves have solved the problem by forming packs, other members
of the pack helping to raise the pups. This
behaviour is also seen in the african hunting dog (lycaon pictus).
coyote males help with the feeding of the pups and jackals form pair bonds
to provision mother and pups. Dogs rely on humans to raise their pups,
and owe ther prolification to their ability to have us help them.
The relationship would have been mutually benificial however, with
dogs providing food, fur and sanitation by the removal of waste. In certain
places in africa people still keep dogs for the function of keeping a
baby's bot nice and clean in a more direct cleaning function. The protection
aspect of domesticating
wolves also would surely have played a major part in their societal
integration.
Avery interesting specualtion (coren, 2000) suggests that we may even owe our ability for speech to the humble dog/wolf. The theory goes as such. Around 100,000 years ago (a time which we know wolves hunted in association, in overlapping regions with man) men began to evolve into homo sapiens at which point facial features began to become more rigid, implying a reduction in olfactory acuity, and the larynx began to develop more of a 90 degree bend facilitating a broader range and control of vocalisations, i.e. primitive speech. The suggestion is that as we began to rely less on our olfactory senses, as wolves were beginning to hunt and track for us, our bodies began to change morphologically, allowing us to develop the power of speech. do we owe the power of speech to the dog - is it possible mans best friend actually helped to mould man? Its certinally an entirely plausible theory, and I for one like to think it may have some truth in it. It is true of most things that when a behavioural reliance begins on something, other aspects of behaviour must change to accomodate it.