Pottery is a vital source of information about trade and industry in Roman
Britain, especially the movement of food.25 Easily broken, pottery’s tendency to absorb traces of foodstuffs meant that it was routinely dumped in huge amounts. It was also used as containers for cremations and grave goods, and for storage, especially of coins in hoards. Most pottery vessels found in complete or semicomplete form are recovered from graves or hoards.
The Roman army established its own potteries to produce kitchen-ware and occasionally more sophisticated material. They bought in from local suppliers, introduced new forms, and helped stimulate production. Early military sites like Longthorpe and Usk have produced sherds that include imported wares and such specialized Roman forms as mortaria and lamps [164, 165].
A mortarium found at Grantham, but produced at the Hartshill-Mancetter industry, a major force in Romano - British ceramics that dominated the market in central and northern Britain.
The two larger, first-century lamps were probably made in Italy or Gaul. The smaller, wheel-thrown lamp was designed to imitate bronze. Found at Billingsgate, it was probably made in London and was discarded after the loss of its handle.
These forms rapidly became widespread, especially in towns where demand was serviced by imports and local industries. A shop stocking imported samian and a clay oil-lamp factory in West Stockwell Street were part of an extensive ceramics industry based in and around Colchester from the mid-first century. Large-scale pottery production tended to be the preserve of more rural areas, such as the massive Oxfordshire industry of the second century and later. A notable exception was the Nene Valley industry, which promoted the growth of the large ‘small town’ at Water Newton.
The vast majority of pottery used in Roman Britain was ‘grey kitchen-ware’, manufactured in astronomical quantities, usually very close to where they were used. Only under exceptional circumstances were such pieces transported over longer distances. At a local level small potteries, such as the ‘Sugar Loaf Court’ potter in late first-century London, serviced no more than a few farms or a neighbourhood. Other kitchenware industries, such as the black-burnished potters of the southwest and Thames Estuary area, were on a much bigger scale. Analysis of the wares and plotting of find-spots can pinpoint an industry’s market reach, making it clear that water transport and army contracts could have had a dramatic effect on how far these products travelled.
Some specialized classes of pottery were more widely distributed. Normally made by potters who made nothing else, mortaria were produced in a number of different forms, all retaining the distinctive gritted bowl for mixing foodstuffs. Early examples were imported, but by the late first century a number of Romano-British industries, like the Brockley Hill potteries, had begun to produce them. In the second century, mortaria were being made in Hartshill-
Mancetter (Warwickshire), Colchester and the Nene Valley. Amphorae, as they had been since the Iron Age, were almost invariably imported for their contents, which ranged from wines to fish sauce and olive oil. Some carry legible traces of writing that record what was inside and occasionally the city of origin.
Fine wares in the western provinces were dominated in the first and second centuries by the Gaulish samian industries. The origin of the term ‘samian’ is uncertain, but possibly comes from the red-slip wares manufactured on Samos, or the Latin word samiare, ‘to polish’. The regional version of red-slip samian wares popular in the Roman Empire was produced in south Gaul in the first century and central Gaul in the second century [ 166]. In the late second century, East Gaulish samian took advantage of the demise of Central Gaul, but was usually inferior in quality. A Romano-British samian industry started up at Colchester c. 160, but the project was soon abandoned, apparently unable to compete with Gaulish samian. In the third century, the quality of East Gaulish samian deteriorated further until the industry collapsed altogether. Samian is found on almost all sites of early Roman date. Forts tend to produce the largest range of forms, but even modest farmsteads are likely to produce fragments of samian cups.
A mid-second-century samian bowl found at Alcester (Alauna), imported from central Gaul. Small towns, just like their larger counterparts, had access to goods imported from across the Roman world.
167. Pottery fragments from London.
A group of pottery pieces from the London wharfs gives an idea of the prolific range of wares found in the provincial capital, and reflected to a varying extent throughout Britain.
Samian’s value to us lies in the use of distinctive decorative styles and potters’ name-stamps. Because samian has been recovered in large amounts from levels of known date, it has helped date associated wares that lack identifying features. Samian imports peaked in the late first century, but continued to enter in substantial quantities until the early third century, when the industry fell into terminal decline and effectively disappeared. Other fine wares were shipped into Britain in far smaller quantities, including the Gallo-Belgic wares Terra Rubra and Terra Nigra, and a variety of beakers decorated with a trailed slip, which survive as the most conspicuous remnants of a massive North Sea trade into eastern and northern Britain. They started to arrive in the late first century as seaborne trade in London reached its climax, and were still arriving nearly two hundred years later.
During the second century, Romano-British pottery industries became increasingly significant. Hand-made black-burnished kitchenware (BBi) manufactured in the Poole Harbour area was transported in vast quantities to the northern frontier from the second through the fourth centuries, supplemented by the wheel-thrown BB2 wares of the Thames Estuary. The Colchester and Nene Valley industries were producing fine wares that resembled Lower Rhineland material. Since Colchester had also been the location of a short-lived attempt to establish a samian industry, clearly derived from prototypes originating in East Gaul, it is quite possible that the Colchester potteries developed in part from the arrival of immigrant potters from the
Rhineland. Colchester’s samian might have failed, but its colour-coated cups and beakers were widely used until the middle of the third century.
The Nene Valley pottery industry had been around since the army passed that way in the mid-first century. To the west, a widely dispersed group of potters in Oxfordshire was producing mortaria and bowls that copied samian forms. By the third century, the Oxfordshire and Nene Valley industries controlled much of the market for fine wares in Britain. Some imports still arrived, but in very reduced quantities. Alice Holt, near Farnham, was the centre of a substantial kitchenware industry that monopolized the market in the late third and fourth centuries in the southeast. Distribution of the material shows that proximity to the Thames and its tributaries played a major role in dictating which sites these wares reached.
One of the abiding enigmas of Roman Britain is, given the extent that pottery production had reached by the end of the fourth century, how quickly pottery ceased to be made. Although some fifth-century pottery production continued, along with limited importation of Mediterranean fine wares to western Britain, neither the demand for pottery nor the skills to make it seem to have survived the disruption of the period. There remains no easy explanation for this.
STONE AND SCULPTURE
The only easily available stone in southeast Britain is flint. Found in irregularly-shaped nodules, it has to be carefully compacted in lime mortar to be of any use for building. The Central Zone coincides with a band of Jurassic limestone that stretches from Somerset right up through Lincolnshire. Easily worked, the limestone was widely used and exported, usually by sea, to the south and east. 23 The late third-century fort at Bradwell-on-Sea, for example, was built partly from Clipsham pink stone, transported from quarries northwest of Water Newton (Cambridgeshire), and must have been taken on a journey by river and sea over 300 km (190 miles).24
North wall of the third-century coastal fort, showing a clear join between two working parties using completely different styles of laying stone.
In the north, abundant local varieties of sandstone made the construction of forts and the manufacture of thousands of carved architectural details, altars and inscriptions possible. Since sources of stone have generally remained useful, evidence of quarrying is fairly rare. However, many quarries have been identified in the Hadrian’s Wall zone, some of which have produced inscriptions left in the stone face by the military masons. It has been estimated that 3.7 million tonnes of stone were needed to complete the Wall.26 Much the most dramatic evidence can be found at Limestone Corner [168], between the forts of Carrawburgh and Chesters [169]. Here, the forward ditch of the Wall was cut from the living rock by chiselling out holes for the insertion of wooden wedges. Soaked in water, these expanded and split the rock. Eventually, the work proved too much and it was abandoned, leaving a number of the blocks in situ, still with their wedge holes.
I68. Limestone Corner (Northumberland).
Close-up of an abandoned block in the forward ditch of Hadrian’s Wall.
169. Chesters (Northumberland).
The Hadrianic cavalry fort at Cilurnum straddled Hadrian’s Wall, not visible here apart from a short stretch between the River Tyne and the fort’s east wall. Most of the internal buildings and defences remain buried. The detached baths lie close to the river bank for drainage reasons, and to reduce the risk of fire. Like all of the Wall’s facilities, the fort and its buildings were built from locally quarried stone.
More exotic stone was sometimes imported from the Continent. The monumental Trajanic inscription found at Caerleon seems to have been carved from marble quarried in Tuscany. It may even have been carved at the source in the year 98, since when it arrived the text was clumsily altered to adapt Trajan’s titles to those of 99-100. Fishbourne Palace was fitted out with veneers made of marble shipped in from the Pyrenees, Haute Garonne, Skyros and Turkey, showing that if the money was available, stone could be obtained from effectively anywhere in the Roman Empire. Given the practicalities, however, such material was usually only imported in very small quantities, though the Carrara marble cladding used for the monumental arch at Richborough is an exception.
Sculptors are known by name on a few pieces, for example, Sulinus the scultor [sic] on an altar at Bath. Another Bath altar names a lapidarius (stonemason) called Priscus.27 Carving stone was not restricted to people who did it for an occupation. This much is obvious from some of the most clumsily executed examples. The Jurassic limestone that crops up so much in the Central Zone is so soft when freshly quarried that it is easier to carve
Than wood. Private individuals with a reasonable amount of dexterity were quite capable of producing their own carved reliefs, like Juventinus, who carved a figure of the god Romulus in a gabled relief and proudly recorded his own name on it.28 In the remote north, the relief from High Rochester depicting Venus [170] is a marvellous combination of classical content and amateurish execution.
170. High Rochester (Northumberland).
Carved stone relief depicting Venus with her handmaidens. Despite the classical subject matter, the exaggerated features and wooden postures are hallmarks of an inexpert sculptor. Height 67 cm. Third century.
TIMBER
Timber was so fundamental in the Roman world that its name, materia, has become our general word, ‘material. Timber was used in enormous quantities in the forts and fortresses of the first century, and also in towns, for housing and for the first phases of some public buildings, like Silchester’s basilica and forum. It was also used for the construction of ships and carts, as well as for tools, writing tablets, and even sculpture. As fuel it kept bathhouses going, as well as the furnaces that smelted ores into metal and pottery kilns.
Prehistoric communities were already adept at exploiting and clearing woodland. If Roman exploitation was on a grander scale, a new programme of woodland management must have been begun quickly. Just how much effort was involved is clear from a mutiny that broke out in the army on the
Rhine in the year 14, caused in part by the work. 29 By 75 to 85, the legion at Caerleon was gathering timber for a fortress calculated to have needed wood from at least 150 ha (371 acres), with additional stocks needed for replacements and maintenance.30 The contemporary fortress at Inchtuthil is estimated to have consumed over 16,000 cubic m (650,000 cubic ft) of wood, though all that survives are postholes, trench slots and iron nails. 31 A phenomenal 5,500 ha (13,591 acres) of managed woodland has been estimated as the minimum to service the later masonry legionary baths at Caerleon.32 These measurements are impressive but meaningless, since their significance depends entirely on whether or not they resulted in the exhaustion of stocks, leading to the consolidation of forts in stone, or if demand was met from managed woodland.
References by ancient authors to carpentry, and the evidence of timber surviving in waterlogged deposits, show that carpenters’ skills varied as much as those of sculptors. In large-scale, heavy-duty applications like wharfside timbers, the joints were crude. Large iron nails were used indiscriminately to hold the framework together. Conversely, evidence from structures like the Southwark warehouse [ 171] and early military buildings at Carlisle show that Roman carpenters were quite capable of producing neat dovetailed joints.33 Trenails were used for dowelled joints when necessary, and remains have been found at, for example, the Blackfriars and County Hall ships, found in London, as well as at Frocester villa. The limited survival of wood means that we have no idea how roofs were constructed, except that they were strong enough to support heavy ceramic tiles.
Waterlogged conditions preserved part of a timber warehouse that had been built into a specially dug pit, accessed by a ramp. This means it was probably designed to store perishable foodstuffs. The floor was made of planks laid on sill-beams. Probably late first to late second century.
TILE AND BRICK
So long as the resources (suitable clay and water) were available, tile and brick could be manufactured on-site or nearby. The legions and fleet manufactured their own tiles, usually stamped with the unit’s titles. The best known are the tiles made by the XX legion at their works depot, located at Holt (Clwyd) close to the fortress at Chester, and those made for the fleet, classis Britannica. Tiles were also produced in London for the procurator, while some of the major towns like Gloucester also had their official tileries. Private and commercial tile production gives us a better idea of the market. The Ashtead villa seems to have been run as a
Commercial tiling concern. A tilery sited nearby produced distinctively decorated flue-tiles, used at other locations, such as Cobham and Walton Heath, in the region. Other tileries have been found at Chelmsford and London, which seem unrealistically distant for freighting such heavy and fragile goods, so the Ashtead tilers may have built kilns closer to where the tiles were needed.
The Kent tiler, Cabriabanus, used a roller-die to impress his name on his tiles, thus identifying his work. It also produced a relief surface that helped the tile adhere to mortar and was invisible during the tile’s functional life span. His work has been found at several places in Kent, including the Darenth and Plaxtol villas, so he must have been a jobbing tiler, taking commissions as and when they arose.
Flue-tiles were often distinctively marked. Usually a tile-comb was used to create diagonal and crossed strips that have no significance beyond creating a roughened surface for mortar. The Ashtead tiler decorated his work with animal scenes [ 172], even though these were invisible once the tiles were installed. Most roof - and other tiles carry no marks at all, except occasionally for numerals now interpreted as batch-marks; one brick from Woodchester carries the numerals ‘XXXXIIH’ and ‘XXXXVI’. The tile-kiln i Great Cansiron Farm, Hartfield (East Sussex) is not specifically associated with a villa house, but its most probable function was serving that market. One of the tegulae found here was inscribed with the numbers ‘CCXV’ and ‘CCXIIII’, probably referring to kiln-loads34 In general, similarities of tile-comb patterns and other design details at individual sites suggest that most villas were serviced by a kiln built nearby and operated by a single team.
Box-flue tile from the kiln on the villa estate. The distinctive animal chase pattern distinguished the products of this factory, which have been found at several other sites in the area, including Beddington, 13.5 km (9 miles) to the east.
London, also lying on a major tidal river opposite the Continent, started out its life as a home to traders and commerce. Not surprisingly, it has produced unparalleled evidence for trade in the Roman period in Britain. A sensational discovery at Southwark in 2002 could not have been more appropriate. This was a dedication to the Spirits of the Emperors and the god Mars Camulos by one Tiberinius Celerianus, who came from a north Gaulish tribe called the Bellovaci [173].35 He tells us that he was a moritix, or moritex, and a Londoner. Moritix seems to be the colloquial Celtic term for a merchant seafarer, where Celtic and Latin both used the same root for the word for ‘sea’ (in Latin, mare and in Welsh, mor).
Inscription from the temple precinct dedicated to Mars Camulos and the Imperial Spirits by the moritix, Tiberinius Celerianus. The inscription is incomplete, but parallels from the rest of the Empire suggest that Celerianus had been honoured as Primus Omnium, ‘first citizen of all Londoners’. (Museum of London).
Lucius Viducius Placidus was a negotiator, or merchant, from Rouen. In 221, he paid for an arch at York that was probably an entrance to a temple precinct, and possibly a shrine. Placidus also left a dedication at the shrine of the goddess Nehalennia at Colijnsplaat, near the mouth of the Scheldt across the North Sea.36 What Placidus was trading in can only be guessed at, although pottery is likely, but he evidently dealt in goods that came from the Rhineland. Marcus Secundinius Silvanus also left a dedication to Nehalennia on the other side of the Scheldt estuary, at Domburg, recording his occupation as a negotiator cretarius Britannicianus, ‘pottery merchant on the Britannia trade.37 Another variant was the negotiator Britannicianus moritex, Gaius Aurelius Verus, who made a dedication to Apollo at Cologne.38 Cologne is known to have been a source of goods that ended up
In Britain. Pottery was certainly among them, along with glass and pipeclay figurines, some stamped by the manufacturer Servandus of Cologne.39
At Vindolanda, Gavo was responsible for supplying some of the food and textiles used at the fort. He is recorded on an account that lists the goods and their prices. Atrectus worked as a brewer ( cervesarius). He was probably a civilian trader who serviced the market for beer in the fort. 40 Such individuals are freak survivals from what must have once been a very large number of men who made a living out of the cross-Channel and North Sea trade, or who serviced demand on a more local scale. The sea merchants took financial risks in shipping goods across waters that were not only susceptible to major storms, but also to piracy. The results of their collective efforts are clear from the large quantities of ceramic material in particular that was shipped into Britain. The Vindolanda tablets exemplify how impossible it is to have any really meaningful idea of trade and commerce in Roman Britain. Gavo’s account includes wool, beans and honey - substances that have virtually no chance of surviving in the archaeological record.
COINAGE AND UNITS OF EXCHANGE
Coinage made the sophisticated levels of trade in the Roman period possible. Rome also introduced standards of lengths, weights and measures, making exchange more reliable and consistent. The Carvoran dry-measure [174], for example, is thought to have been used for measuring corn. Its stated capacity is 17.5 sextarii, about 9.5 litres (16.7 pints). Since it carries an imperial inscription for the years 90 to 91, it may have been used for assessing tribute or for doling out official rations to military units.
174- Carvoran (Northumberland).
Bronze dry-measure from the fort at Carvoran (Magnis), bearing the name (deleted) and titles of Domitian for 90-91. Magnis was not built before Trajan’s reign (98-117), so the measure was probably lost then or later.
The so-called ‘Celtic’ tribal coinage was largely silver and gold, but the weights, denominations and purity of issues varied immensely. Celtic coinage was demonetized after the invasion and generally disappeared, though very occasionally some pieces appear in hoards of Roman date. 41 In the Roman world, currency was based on intrinsic value. Debased silver, for example, was worth less than purer silver, even if the denomination was the
Same. Throughout the period, the purity and weight of silver and gold coinage fluctuated either because silver was debased, or because gold and silver was struck in smaller weights. A range of brass and copper small-change coins was issued for everyday circulation. Soon after the invasion, Roman coinage became the only circulating coinage in Britain. To begin with, soldiers were probably the most prolific users of coin. Barter is likely to have dominated transactions in the early part of the Roman period, and probably remained a significant factor in how the economy worked. Nevertheless, all communities had access to coinage, and all of them used it to some extent. Roman coins found their way into hoards in Scotland, showing that even beyond the frontier tribal peoples had become accustomed to coinage as a means of storing wealth.
The government spent much of its time trying to recover the gold and silver bullion paid out to its employees. The coins were of high value, so the recipients either chose to save or hoard the bullion, or exchange some of it at a moneychanger for brass and copper coins [ 175]. Since taxes had to be paid in bullion, taxpayers had to take base-metal coinage to the moneychanger and exchange it. The moneychangers, naturally, took a cut on the exchange rate. So, in practice, gold and silver always circulated at a premium.
175. Hadrian (117- 38) and Commodus (180-92).
Brass sestertii of Hadrian and Commodus. During the second century, the sestertius was the most important base-metal denomination. Many sestertii remained in circulation until the late third century, when inflation meant that they were worth more as metal. Consequently, the coins were melted down to manufacture goods or to make copies of the debased silver issue.
There was a continual loss inherent in the system. Hoards that were
Never recovered (the ones which survive today) resulted in permanently lost bullion. Some was used to manufacture jewelry or other precious metal products. Other bullion found its way out of the Empire, being used to buy goods or buy off barbarians. Strain was put on available bullion stocks and suppliers when emperors chose to offer their employees, usually soldiers, substantial pay-rises or donatives, particularly once the Empire’s expansion had been terminated and new sources were no longer available. Under these circumstances, the government debased the silver to make it go further. Unfortunately, the average member of the Roman Empire was sophisticated enough to tell the difference, and promptly hoarded any old silver he had. This only put a further strain on the system through a simple rule of coinage: bad money drives out good.
By the end of the third century, apart from the efforts by Carausius to issue good silver in Britain, Roman silver coinage was a bad joke. Bronze coins issued with a silver wash masqueraded as silver, but fooled no one. The result was steady inflation. From the late third century on, there were numerous attempts to reform the coinage with a variety of new coin types. Silver was coined on and off in the fourth century, but it was no longer produced regularly. Instead, gold dominated late Roman coinage, along with abundant bronze issues that individually represented tiny fractions of the gold coins. The Romans had discovered the difficulties in maintaining dual standard coinage, since the bullion value of gold and silver can never be maintained in an exact relationship. Different factors affect the supply, and therefore the price, of each. The same problem would recur in eighteenth-century England, with a similar result until the coinage reform of 1816. Very little silver was struck between 1715 and 1816, and instead Great Britain, then one of the most powerful nations in the world, used gold denominations and copper fractions of those gold coins.42
Throughout the Roman period, these problems were exacerbated by the erratic production and distribution of coinage. Britain suffered from periods when official coin was in short supply, and this was usually made good by bouts of copying or forging generally base-metal coins. This is useful evidence that Britain was dependent on coinage. The most notorious phases were between 43 and 64, when the army seems to have been responsible for issuing copies of official base-metal coins [ 176], the late third century, and episodically in the fourth century, when copying was so prolific and so variable in quality that many people must have been engaged in its production. However, forging to some degree always took place
Throughout the Empire. In late eighteenth-century Britain, shortages of copper halfpennies and farthings were made good not only by the forging of official issues, but also by the manufacture of tokens. These forgeries were also sold as commodities in their own right. Since the copper was worth about half the face value, they could be made and sold at a profit. Unscrupulous employers and merchants bought the counterfeit coins and used them to pay wages and bills. It is both possible and likely that the same activities went on in Roman Britain. The government in the 1770s and 1780s tolerated the forgeries because the economy could not function without them, and doubtless the provincial authorities of Roman Britain felt the same way.
Copy of a copper as of Claudius. Distinguished by poor style and ragged appearance, these copies were manufactured in Britain, probably by the army, after the conquest of 43 to make good the lack of official coinage. The recoinage of Nero (see [31]) made the copies redundant and they fell out of use.
Until the late third century, official coinage used in Britain was struck mainly at Rome, and occasionally in the first century at Lyons. Thereafter, coinage was produced at a variety of provincial mints. These included London between 286 and 325, but most coinage came from Lyons, Arles and Trier. Coinage entered Britain through pay for soldiers and officials, payment for goods and services, and as loans. The London mint, founded by Carausius, produced the first official coinage in Britain, along with the ‘C’ mint (its exact location is unknown). The ‘C’ mint was closed in 296, but London’s remained in use under the Tetrarchy and Constantine. After c. 325, any coinage produced in Britain was unofficial, and circulated alongside the legitimate issues.
177, i?8. Constantine II (317-37) and Crispus (317-26).
Bronze coins of Constantine II and Crispus as heirs designate of Constantine I. Both coins were struck at London, indicated by the letters ‘PLON’ on the reverse, and were found in the Langtoft (East Yorkshire) hoard, discovered in 2000. Diameters about 2 cm.
We know nothing about the process of supply, but studies of hoards, coin types and die-links (coins struck from the same physical dies) have shown that bullion coin moved freely about the Empire. However, base-metal coin tended to circulate only in the provinces in which it was distributed. Two as types of the year 154 are usually only found in Britain, suggesting that the issues were sent to Britain for use rather than to anywhere else. Relatively few official base-metal coins of the early third century seem to have found their way to Britain, in contrast to other areas, such as Italy and North Africa, where they were abundant. Britain continued to make use of second-century coins, which became increasingly worn and which often turn up in late third-century contexts.
Variations in coin-supply and inflation meant that the number of coins lost varied enormously. Since earlier coins had higher intrinsic values and, in the case of base-metal coins, were larger and less easily lost, they usually remained in circulation longer. Base-metal coins from the 270s on were of trivial individual value, and were produced in abundance. Until this was recognized, density of occupation was often assessed according to the number of coins found on a site. Nowadays special mathematical formulae have been devised that help to compensate for this problem. Archaeologists and numismatists alike now know that a site occupied more or less continuously will produce dozens, or even hundreds, of third - and fourth-century coins for every first - and second-century coin.
Causes of discontent before the Boudican revolt were ‘the confiscation of money given by Claudius to leaders of the Britons’, and the recall of loans by others, including Seneca.43 Seneca’s loan was equivalent to 40 million sestertii, enough to pay the annual wages of seven legions. It is unlikely that much of this money was in circulation. Either it will have been hoarded by the Britons, melted down and converted into other items, spent on buying goods from the Empire, or used to pay taxes. Whether or not any of this money was ever recovered is unknown. Cash also trickled into general circulation through casual expenditure by soldiers and other officials, and when the state made payments for goods and services. This increased greatly after 64, when Nero started to produce base-metal coin in abundance. Coins of this date and later are far more common in Britain. Perhaps one of the most evocative deposits is the sacred spring of Sulis-Minerva at Bath. Unlike a hoard, which was the property of a single person, the coins deposited here were thrown in by thousands of different pilgrims over many decades. Ever mindful of the value of coinage, most of the pilgrims took care to make offerings from extremely worn base-metal coins, or even forgeries. Each of these coins had clearly spent many years in active circulation.
The supply of coin, or production of copies to make good shortfalls, continued throughout most of the fourth century. But from the 380s there was a tail-off, and, unlike earlier periods, copying seems to have declined as well. The last official coins to reach Britain were struck around 402. Thereafter, any other coins that arrived came in such small numbers that they must have been brought by individuals, rather than as part of official consignments. The cessation of the coin supply, and the failure to make up the difference with local copies, had dramatic implications for Britain’s trade and economy. Coinage would not be used so much in Britain again until the seventeenth century.
CONCLUSION
This chapter has only touched on a few aspects of the Romano-British economy, although the subject inevitably recurs throughout this book. There are many different factors that provide us with vivid evidence for the existence of a thriving economy, but it is impossible for us to know which of those factors stimulated others, and vice versa. The army’s role is evident in the distribution of goods, such as black-burnished ware, and it probably also played a large part in determining the availability of samian ware. Although samian was abundantly available in towns and smaller settlements, this availability may have been a side effect of military trade in the commodity. If so, then any changes in military buying patterns could have been responsible for the decline in the whole industry by affecting margins.
Coinage was clearly used in considerable quantities, and made the exchange and storage of wealth easier. Its availability probably helped provoke economic activity, but the extensive bouts of copying in the third and fourth centuries show that such availability cannot have been the only factor. Coinage was available in greatest abundance during the late third century and into the fourth, mostly in the form of small bronze coins of indeterminate intrinsic value. But the greatest intensity of identifiable economic activity was in the first and second centuries, the period when the army was at its most active, and when the major towns in particular were experiencing intensive periods of development. Part of the reason for the discrepancy is that the buying power of individual coins at this time was much higher than later on, since the inflationary pressures of debasing bullion silver coinage had not really hit home. It has been suggested though that a conspicuous reduction in coin supplies in the early third century may be linked to the apparent stagnation in urban development. 44 However, the reduction in coin supplies of this period mostly affected base-metal issues, and not those of the silver, which are well represented, for example, in the massive early to mid-third-century hoards. Moreover, this is the period from which we have specific evidence of individuals engaged in the North Sea trade.
These examples illustrate perfectly how difficult it is for us to draw any firm conclusions about how the Romano-British economy functioned. Material from almost every significant Roman site points not only to major levels of imports in the first and second centuries, but also to widespread
Access to those trade routes. This was the time when towns developed from ramshackle, commercial street frontages into more pretentious places. At the same time, mosaics and more elaborate architecture were very rare in the countryside. But in the fourth century, towns were either stagnating or decaying, and trade with the Continent had declined. Public building was effectively over, though some homeowners were still in a position to invest in improved houses. Conversely, in the countryside, especially in the Central Zone, a significant proportion of the villa estates evolved into expensively appointed miniature palaces. This is so incongruous in the context of the rest of Roman Britain that an obvious explanation is difficult to find. But as rural establishments, villas were generally linked to agricultural production, and it is Britain’s grain that we know was of major importance to the Empire at the time. Britain was relatively peaceable and isolated from the major barbarian incursions of the period. This may have completely readjusted the economy, benefiting a particular sector in the community that spent the money on itself.
3°3
At least 90 per cent of the Romano-British population lived in the countryside. Rural Roman Britain also exhibited the greatest diversity. Some of the rural population lived in caves or remote roundhouse farmsteads that had scarcely changed for hundreds of years. By the fourth century, wealthy people lived in country houses that were architecturally pretentious and surrounded by estates [ 179]. But even in the lowlands, less than one in six known rural settlements can be classified as ‘villas. Roman types of rectangular houses were commonplace, but the ancient Iron Age roundhouse continued in use almost everywhere. The landscape ranged from the ancient woodlands of the Weald to the bleak moorlands of the north, and from the mountains of Wales to the fenlands of East Anglia. The centre, east, south and southeast were the most densely populated and productive. The Roman world affected everywhere, making all of the countryside integral to understanding Roman Britain.
Part of the third-century baths suite, in its fourth-century form. The octagonal chamber was the frigidarium, and the long room was the tepidarium. Although the house never became lavish, its longevity, expansion and improvements reflect a pattern found across southern Britain.
The descriptions by Caesar, Strabo and Diodorus Siculus create a general picture of Britain as an intensely rural world with a thriving and productive population, even if some of what they wrote is obviously wrong. Tacitus says that the aftermath of the Boudican revolt produced desperation because the participants had recklessly failed to plant crops.1 This tells us that the tribes involved were accustomed to organizing a cycle of ploughing, cultivation and harvesting. To do this at all meant a systematic approach to life, defined plots of land, and the wealth of empirical knowledge essential to successful agriculture. These traditions stretched far back into the past. The Roman world brought new markets, increased demand from an urban and military population, and improvements in the form of new tools. But the essence of rural Britain probably did not change very much.
In recent years, environmental data has provided evidence for crops and land usage in a wide variety of landscapes and regions. Aerial photography and land surveying have made it possible to identify and plan field systems, and, together with archaeological exploration in advance of road and building work, have shown that rural populations and settlements were denser than previously thought. Villas are no longer interpreted in isolation as architectural phenomena, but as single components of much more complex rural communities.
I8o. Distribution of villas in Britain.
Although villas are known in all three areas, they occur predominantly in the Central Zone. (After Hingley and Miles, and Jones and Mattingly).
CROPS AND LAND USAGE
Primeval Britain was mainly wooded. Forest clearance made it possible for Bronze Age peoples to start carving up the landscape into areas for cultivation, while also developing woodland management, allowing timber to be grown for a variety of structural and manufacturing purposes. In the Roman period, woodlands were used for the structural and fuel needs of forts and towns, and also for industries such as pottery and iron-smelting. The Weald of Kent, for example, was a major iron-smelting area and it remained wooded throughout the period. A piece of woodland in Kent, called Verlucionum, was the subject of a legal case heard in London in the year u8.2 Villas of the region were clustered along the northern part of Kent, the River Medway, and its tributaries.
In the areas opened up to agriculture by forest clearance, a number of changes took place after the Roman conquest. Animal bones are ubiquitous finds at Roman settlements, showing that cattle, sheep, pigs, venison, fowl, and even horses were butchered for meat, with cattle and pigs apparently growing in popularity over the period [ 181]. The need until modern times to drive herds of animals from their homes to places of consumption means that it is impossible to know where they came from. In the eighteenth and nineteenth centuries, it was commonplace for cattle to be driven to London from as far away as Wales. In antiquity, when the roads were if anything better, the same must have happened. Sheep kept for wool were likely to have been clipped with shears, one of the new tools introduced during the Roman period, possibly encouraging the apparent increase in sheep. Field systems were well established in prehistory throughout large parts of Britain, but now they were more likely to be divided up with hedges, or even stone walls, perhaps a reflection of the legal control of land tenure, enshrined in provincial and civic records. Land was reclaimed through drainage, mainly around the Wash.
Fragment of a mosaic panel depicting a hunting scene. Fourth century. (Somerset County Museums Service).
The Iron Age staples of spelt, bread and emmer wheat, six-row barley, oats and beans were augmented with rye and a range of more exotic plants like coriander, plums and cherries. Working the land itself became easier with the improvement of plough technology. Asymmetric ploughshares were more effective at cutting and turning the soil, helped along by coulters (iron blades fixed in front of ploughshares to cut the soil vertically). Crops were now treated in drying-ovens, a characteristically Roman innovation and found at numerous sites, often with food traces, and stored in granaries instead of Iron Age pits. However, traditional methods remained in use, too.
RURAL SETTLEMENT
The countryside varied enormously, from fertile agricultural territory
Within a day or two’s journey from a town, to remote moorland in the military zone. This was reflected in the nature of settlement, of which the villas are the best known. In the Western Zone, which included Wales, the southwest, and much of the north and northwest, rural settlement changed relatively little from the Iron Age. Usually these remoter settlements produce far fewer finds of the material goods so typical of Roman-period settlements further east. But they were not unaffected by Roman contact. Milestones from Cornwall show that provincial administration reached right into the southwest.3 Even small numbers of characteristically ‘Roman’ finds suggest that contact did exist, and that the Roman impact on remote rural communities was more significant than is now obvious from the archaeological record.
In the Western Zone and much of the north, villas were generally absent, and native farmsteads usually contained with some sort of enclosure were widespread. These farmsteads had been characteristic of much of Iron Age Britain, but in most of the developed part of the province they gave way to Roman building types. They ranged from the characteristic Cornish ‘round’, where a handful of dry-stone houses clustered together in a circular or elliptical enclosure, such as at Chysauster [ 182], to the roundhouses of the north in round, or even rectangular, enclosures. In Scotland, beyond the Roman province, the ‘substantial roundhouses’, including the brochs and duns, remained. These monumental structures, with walls that could be 5 m (16 ft 5 in) thick, seem to have been the alpha houses of the region. They perhaps reflect a hierarchy in which towns and villas were simply not part of the way status was expressed. In the south and east, for example, Roman building forms, particularly the villas, became the housing type of choice amongst the elite, while Iron Age forms remained lower status. Some Scottish broch residents still had access to Roman goods, like the occupants of Fairy Knowe (Stirlingshire), where Roman coins and glass were used alongside imported fine wares, like samian, and the thoroughly Roman cooking device, the mortarium.4
There were about nine stone roundhouses at Chysauster, each with its own courtyard for outhouses. Little changed here throughout the Roman period, and the village is an important reminder that for much of rural Britain, villas and polite Roman living were unthinkable luxuries.
Contact probably resulted from a mixture of individual trading relationships and personal choice, rather than general economic trends. In northern Britain, the army was the most influential factor. Here villas are almost non-existent, although there are exceptions, such as the villa at Ingelby Barwick, near Stockton-on-Tees. In the north, the military towns, such as Corbridge and Carlisle, and the vici that clustered around the forts were dependent on the army. At the end of the period, the military towns seem to have decayed before the forts, while earlier there is no case of a vicus surviving the closure of a fort as had happened in the south. The army imported the goods it needed, while any demands it made locally on produce will have had a marked effect on settlements that were probably only just viable. The economic and social ties that maintained towns and villas were never really established in the north.
In Wales, villas were rare except in the southeast, where the land was more agriculturally productive. Some Iron Age farmsteads, such as at Whitton (Glamorgan), were able to develop into modest villas. Elsewhere in Wales, settlement became more archaeologically visible, but the difficult landscape prevented this from becoming intensive. Although roundhouses were often replaced with rectangular buildings, the structures remained simple. The land was simply too marginal to support anything better, with the army’s demands perhaps making this worse. On the other hand, the villas that did exist must have depended on a military market. In south Wales, most of the villas had been abandoned by the middle of the fourth century. The legionary fortress at Caerleon seems to have been given up, since the II legion is only testified at the much smaller fort of Richborough in the fourth century. The removal of an entire legion is bound to have had a dramatic effect on the economy in south Wales, though the nearby civitas capital at Caerwent seems to have survived until the end of the century.
Living on the margin is reflected in the nature of settlement towards the Midlands and Wroxeter. Wroxeter was one of the remoter civitas capitals, and it had to wait until Hadrianic times before it was endowed with a forum and basilica. The area was able to support a significant town, but roundhouses remained common nearby and there were very few villas. In the Peak District, on the border between the ‘civilian’ south and the ‘military’ north, some people were living in caves in the Carsington area. However low-grade the places these people lived in, they still had access to the Roman goods that make them identifiable in the archaeological record.
In the rest of Britain, much of the landscape was dotted with villas, but they were interspersed with more basic Roman-type farmsteads or even roundhouses. Frocester Court (Gloucestershire) remained until the late third century no more than a Romanized Iron Age settlement [183]. Roundhouses gave way to timber rectangular houses there, but not until the late third century were the inhabitants able, or willing, to move into a substantial masonry villa. Nearby, at Standish, a similar settlement apparently never reached the villa stage.
183. Frocester Court (Gloucestershire).
Until the late third century, Frocester, like many other rural sites, remained a simple Romanized farmstead with outbuildings.
Areas without villas include surviving woodland, such as the New Forest and the Weald. On Salisbury Plain, Iron Age traditions seem to have continued almost unabated. Perhaps this was an imperial estate, managed and farmed for the emperor with the economic effect of the removal of the surplus from the local economy. Imperial estates undoubtedly existed, but no evidence to substantiate any in Roman Britain has ever been found, apart from an inscription found reused in the villa at Combe Down (Somerset).5 It has been suggested that the monumental building at Stonea [184] in the Fens, along with its attendant settlement, may have been the headquarters of an imperial estate established under Hadrian, and administered by soldiers to help supply the northern frontier. 6 Where villas developed they generally did so on the better quality land, and had access to main roads and urban markets. The richest concentrations are in the Mendips and Cotswolds, an area also of relatively high urban density. Cirencester was one of Roman Britain’s largest towns, and it lay close to the colony at Gloucester. It also sat on the Fosse Way, which connected it to Bath and Ilchester to the south. The roads joined the major towns to a number of smaller towns, and spurs drew the villas of Somerset [ 185] and the Severn Estuary into the network.
Hypothetical reconstruction of the so-called ‘tower’ that formed the centrepiece of the settlement.
This villa, which grew to a substantial size in the fourth century, benefited from its close location to the Fosse Way, which enabled it to be incorporated into the provincial communications infrastructure. This geometric mosaic was laid in the villa during the fourth century, and is shown being recorded shortly after its discovery in 2002. Ploughing has damaged part of it.
The individual size and level of architectural embellishment at villas is usually related to location and date. For example, those on the edge of more marginal areas like the Fens or Wales tend to be smaller and less well appointed. Villas are much more common features of the third century and afterwards in Britain, but many had structural histories that stretch back at least into the second century, and sometimes the first. Villas often overlie, or are adjacent to, Iron Age farmsteads, but did not always follow on immediately. The Halstock (Dorset) villa was begun in the mid-second century on an Iron Age site that had fallen out of use a century before. 7 At Park Street (Hertfordshire) an early villa soon followed a late Iron Age rectangular house. The greatest period of villa development was in the fourth century, but by the 380s and afterwards villas were in decline. A few remained in use into the fifth century, but invariably in reduced circumstances (see [258]).
Virtually without exception, villas lay on or close to land suitable for agriculture. This does not necessarily mean that any specific villa’s economy depended exclusively on the exploitation of land, but agriculture probably
Always played an important part. Although the sites chosen vary considerably, villas always had access to water from springs or rivers, and were well-drained. The Darenth Valley (Kent) villas form a prime example, and they parallel another series along the Medway a few miles east. Villas in the Darenth Valley include Lullingstone, Darenth and Farningham, all of which lay a few metres from what was then a wider and deeper river than it is today.
The villa at Great Witcombe (Gloucestershire) could not have been more differently located [ 186]. The house lies on a steep northeast-facing slope, astride one of the many springs. The location is so beautiful and striking that this must have been the primary attraction. The springs provided water for living and also for the baths, while the slope provided all the drainage needed. The steep hillside meant that the house had to be built along a contour, and its wings terraced into the slope. Buttresses had to be set up against the walls to prevent subsidence. The water threatened to undermine the foundations, and had to be controlled by being channelled under the house. The buttressing rather implies a detailed appreciation of potential structural problems, but the evidence from other sites collectively suggest that builders took chances, and where problems occurred, remedial measures, assuming there was time, were taken.
I86. Great Witcombe (Gloucestershire).
The villa as it might have appeared in the fourth century. The house, which overlooks the same dramatic views that it did in antiquity, seems to have incorporated a water shrine, but this does not mean that its prime function was as a temple.
WHAT IS A VILLA?
The world ‘villa’ meant, in theory, a ‘farm’, but is better translated as a rural dwelling of some pretension, often, but not always, with agricultural associations. A house outside the walls of a city may be a villa, or simply an extramural townhouse. Likewise, a house within the walls, evidently functioning as a farm, may be really a villa, rather than a townhouse (see [95]). The ‘villa’ proper means not just the main house, but also all its peripheral structures, such as barns. It follows then that a village a couple of miles away, manned by the villa estate workers, is also technically part of the villa, as are any field systems in the vicinity. Some of the more modest rural houses were perhaps regarded by their aspirant (or downright bourgeois) owners as villas, but seen by their wealthier neighbours as trumped-up hovels. Conversely, there is no doubt that elsewhere in the Empire the owners of some extravagant villas maintained the conceit that their rural palaces were modest bucolic retreats. Clearly, defining a ‘villa’ in both our terms and those of the Romans is not an easy task. It is made more difficult because in any one instance we do not know if a villa was owned primarily to make money or to express status. We do not even know whether the residents owned the premises or were tenants of Romano-British, or Continental, landlords. We also do not know in any one case what the borders of a villa estate were, and whether villages or lesser houses nearby belonged to it.
Although simple stone rectangular houses started appearing before 43, the process was limited, localized and lacking in conspicuous aspiration. In parts of Britain during the late first century, rural villas and simple farmsteads proliferated. The appearance of increasingly numerous rectangular, multi-roomed houses in the south and the Midlands by the late first century is one of the most conspicuous changes of the period. Where villas are known in abundance, the key point is that being made of stone they tend to leave conspicuous traces (even when robbed out). They are easier to identify from the air and as the result of a chance find. Prehistoric housing forms are more difficult to detect, especially when a Roman house was later built on top. They remained in use even in areas where the villas
Were at their densest, while simple rectangular houses clustered together in villages, probably accommodating estate workers and their families.
Most Roman houses had rectangular plans with more than one room. Wattle-and-daub, timber, thatch and earthen floors could all be features of a Roman-period rural rectangular building. It would be stretching the point to classify a two - or three-roomed rectangular hut as a villa. Virtually all villas are based on wings, with a series of square or rectangular rooms and corridors, with even the largest villa being simply a bigger version of a smaller one. However, Roman-type houses were given to elaboration. This is a crucial distinction from the roundhouse. Roundhouses came and went, and new ones were similar to the old. No one ever extended a roundhouse, installed mosaics or fitted it with a bathhouse.
World History
