I'm in particular curious about the period spanning from the post-Roman era until shortly after the Norman conquest.

Just a casual 500 years then!

To start with, we have to be careful using terms like "serf" when it comes to looking at the Early Medieval English. Within the context of the period, a "serf" or servus is quite literally a slave. Unfortunately we don't really have anything approaching census data for pre-Conquest England, so we mostly have to rely on Domesday Book of 1086 and try to extrapolate backwards. At the time of Domesday, roughly ten percent of all 'households' listed in England were slaves, and the figure is likely to have been even higher in the 9th and 10th centuries. Slaves were property, but ownership was not restricted to the nobility. Alfred's Doomboc legal codex implies that any free man could own slaves. Conditions of slavery were covered by an increasingly sophisticated legal system. Slaves, for example, were allowed to own property and save money, were given designated holidays every year, and were prohibited from working on Sundays. A popular conception is that slaves were Welsh, but this wasn't necessarily always the case. Penal slavery was a punishment listed for a number of crimes throughout the period, most typically being caught in the act of theft. In the 7th Century legal code of Hlothhere of Kent, a thief caught red handed could be sold as a slave overseas. Interestingly in contemporary Wessex, it was illegal to sell an English slave out of the the country.

The most common peasantry were tenant farmers, who occupied a number of social strata but can mostly be divided into two broad classes: the villein or villager, just over some thirty percent of the population at Domesday, was the largest and more prosperous class, who on average held a virgate of land (a quarter of a hide or ploughland, very, very roughly some 30 acres); and the smallholder, cottager or borderer who comprised just under thirty percent of households at Domesday, and worked somewhere between 5 and 15 acres depending on opportunity and status. All of these classes of peasantry are free men, but are commonly and frustratingly confused with the class of Freemen, around some fourteen percent of households at Domesday but likely previously much higher, who were wealthier peasants who owned their land outright.

One of the most important rights of the peasantry that was curtailed by the 1066 Conquest was the ability to forage and hunt in woodlands. Excavations of butchery remains at the 10th Century burh at Stafford suggest that game was a major source of meat to the contemporary population. Similarly, the 10th Century Ælfric's Colloquy also suggests that hunters of boar, deer, hare and birds were commonly-encountered members of the community.

Another important right was military service. Prior to the 870s in Wessex, military service was the province of the thegnly classes and Freemen peasants. Following Alfred's fyrd reforms in the face of the Danish threat, however, this right and obligation was extended to all free men. On land, the best and most capable within a community might be chosen for service in the fyrd, while in coastal communities, individuals would serve as Lithmen in maritime service. Military service was highly prestigious and could bring lucrative rewards and status. The Lithmen who crewed English warships in the tenth and eleventh centuries were high-status individuals in the settlements of the Cinque Ports region; fyrd warbands could capture considerable loot and captives from successfully defeating a Viking army or raiding party, and following the Battle of Brunanburh in *c.*926, the entire town of Malmesbury was awarded Freeman status by Æthelstan in recognition of the valiant effort of that town's warband among the Wiltshire fyrd. Freeman status could be quite lucrative: in conferring outright ownership of land, it absolved the landowner from any rents or service obligations save direct taxation liable to the king.

Freeman numbers post-1066 average at about 14% of the population, and the figure is higher in former Danelaw areas than in Wessex. There's considerable debate as to why this is the case; one school of thought holds that this higher proportion is a result of the Danes themselves, either establishing themselves as 'freemen' through conquest, or through a wave of Scandinavian colonists. A more prevailing argument is that Freeman levels had previously been similarly high across England, but that the West Saxon peasantry increasingly lost this status as the state greatly tightened its control in response to the Danish threat in the 870s and 880s. Either way, the numbers of Freemen peasants declines quite steeply in the wake of the Conquest.