British Archaeology banner

Cover of British Archaeology

Issue 67

October 2002

Contents

news

Hopeful dead clutching their tickets to heaven

Long survival of York’s Roman fortress defences

Drinking den below streets of Edinburgh

All the emotions on display in Southwark Roman cemetery

Treasure Act brings in the gold and silver once again

In Brief

features

Roads from Rome
Hugh Davies discusses Roman roads as a transport system

Shipwreck to slavery
Mike Parker Pearson on the story of an 18th century sailor

Great sites
Rosamund Cleal on the Neolithic site on Windmill Hill

letters

The origins of industry, Tolkien’s inspiration and museums

issues

George Lambrick on the power of public support

Peter Ellis

Regular column

books

Viking Weapons and Warfare by J Kim Siddorn

Prehistoric Cooking by Jacqui Wood

European Landscapes of Rock Art edited by George Nash & Christopher Chippindale

The Prehistoric Archaeology of Ireland by John Waddell

Vikings and the Danelaw edited by James Graham-Campbell, Richard Hall, Judith Jesch & David Parsons

Image and Power in the Archaeology of Early Medieval Britain edited by Helena Hamerow & Arthur MacGregor

favourite finds

Rob Ixer on a lump of lead ore that made a nice paperweight

 

ISSN 1357-4442

Editor Simon Denison

features

Roads from Rome

Roman roads can reveal much about Roman life when viewed as part of an integrated transport system, writes modern-transport expert Hugh Davies

Many people are pleased and proud to know that they live near a Roman road. There is considerable enjoyment to be had from walking along preserved lengths of such roads, while names such as Watling Street and Fosse Way have become integral parts of our heritage.

Archaeologists are certainly familiar with the physical remains. Roads or streets are often encountered during excavation. Yet such familiarity can, I believe, lead to lack of interest, even boredom. After the extent and nature of the structure have been dutifully recorded, excavators move on to more interesting features, such as buildings or burials, which seem to offer greater scope for analysis and deduction.

This means that we may not be getting the full benefit from the extensive physical legacy of the Roman road system, which I believe can make a great contribution to our general understanding of the Roman period.

I became involved in researching Roman roads about six years ago. Until my retirement I had worked for over 25 years at the Transport Research Laboratory in Crowthorne, Berkshire, studying many aspects of modern transport and traffic, including the building of new motorways. During this time I had occasionally walked along a Roman road called the Devil's Highway, south of Bracknell, near my home. After retirement I had more time to explore my neighbourhood, and a chance remark by a friend - namely, that the road had once linked London and Silchester - fired my interest in Roman roads and led to a PhD at Reading University and ultimately to a book.

My lifetime's work in modern transport planning led me to view Roman roads not as isolated monuments but as roads linking towns and settlements, forming an integrated transport network. During my study,I found that there were many widely-held misunderstandings about Roman roads, as well as gaps in our knowledge of what Roman roads were built for, and how they were used.

Transport pioneers

There can be little doubt that the road network built by the Romans throughout their empire was a major achievement. Britain, one of the provinces furthest from Rome, was provided with a road system which in total length is comparable with our modern trunk road and motorway network. The quality of construction was no less impressive, and was not bettered until the days of Telford and McAdam in the 19th century.

There are many similarities between the problems faced by the modern road designer and his Roman counterpart. As direct a route as possible needs to be planned linking the end points and any intermediate destinations, while avoiding steep or marshy ground and finding suitable crossing points for rivers. There are of course some differences. A conquering force would have had few worries about the opposition of landowners. Also, the Romans were less troubled than we are by steep gradients, as average traffic speeds were not expected to be high, say no more than 3 mph for wagons. However, in most respects, little has changed in the designer's task.

Roman road-building has an excellent reputation, and on the whole it is justified. Examination of the remains of many roads over the last century or so has shown that in Britain, Roman roads were largely well-metalled, with good drainage, and aligned so that they took a fairly direct line between origin and destination.

Yet there are many misconceptions about how the roads were designed. The reputation of the Romans for order has led some to believe that the road network demonstrates a grand geometric strategy, with roads imposed on the landscape in a series of regular, rectangular patterns. However, at many towns, such as Silchester, Verulamium, Leicester and Ilchester, roads which arrive at an oblique angle change direction at the town gate or boundary to match the street grid. This indicates that the grids predate the roads, and that roads were built to serve settlements, at least the larger ones, rather than being arbitrarily imposed across the landscape. At Lincoln, by contrast, the principal street does pass straight through, with the town lying almost directly in line with the approach alignment from each side, suggesting that the road and the town were part of the same overall design.

It is also often assumed that the road cross-section conforms to a regular design standard, with a metalled width of 20ft (just under 3m). According to this pattern, the metalled road is built on top of a raised bank, the agger, with drainage ditches running along each side. High class roads have parallel road lanes, unmetalled, alongside the central agger, with boundary ditches on the outside to define the road area. The metalling is also assumed to be of standard design, with large stones at the base, over which are laid successive layers of smaller stones with a running surface of gravel or paving.

Almost none of this regularity is evident in practice. Road widths vary greatly, between about 12ft and 30ft, with some even narrower and wider examples. Ditches are often absent, or only dug on one side, while metalling varies in width, depth and design. Parallel lanes and outer ditches are only sometimes present. The material forming the metalling can be gravel, flint, cobbles or limestone blocks. This variability reflects a practical awareness of local materials and topography, and the volume and type of traffic that individual roads were required to carry.

Roman roads also changed over time. The network observed today through survey and archaeology is the cumulative result of four centuries of development. At the start of the period, the conquering Roman army would not have waited for well-engineered roads to be built before moving into enemy territory - it would have used any suitable local tracks. As the front line advanced, the road builders would have followed, initially with a rapidly-built - probably lightly-metalled - road to guarantee the army's supply line. More substantial roads would only be built, over the top of the earlier road, once a region was secured.

Possible evidence of one such early military road has been found both south and north of the Thames, presumably built to link the legionary base at Colchester with the invasion beach-head at Richborough in Kent. A lightly-metalled road was found south of the river at Spital Street, Dartford, interpreted as an early line for Watling Street; while north of the river, at Old Ford, London, archaeologists found a single-carriageway unmetalled road. Traffic damage to this road was soon dealt with by the introduction of light metalling. Some mid-1st century pottery was found. It could be argued that the experience gained in building this section of road taught the engineers that an unmetalled road was unsuitable for conditions in Britain. Later, a series of wider, grander roads was built over the top, culminating in a three-lane carriageway designed for heavy volumes of traffic.

Romans in Scotland

Similar evidence for an early military road was found in northern Scotland, probably dating to Agricola's campaign in the early 80s AD. The most northerly Roman road known in Britain linked Stirling up the Gask Ridge to Kirriemuir, linking a series of forts. At Stirling, a substantial gravel road rested, in places, on a foundation of large stones. But the road diminished in quality as it travelled north, until finally at Kirriemuir it was composed of a layer of clay with a gravel surface, overlying turf. Had the fortress at Inchtuthil been completed and held for a sustained period, a more permanent road would no doubt have been built.

All roads require repair; but while some Roman roads may never have had more than one road surface, many had around 10 surfaces. What were once thought to be complex, multi-layered structures, can often now be seen as a series of roads, each comprising perhaps a foundation and a surface, each laid directly on its predecessor.

As new roads were laid directly over the old, they tended to become wider over time. However, some roads, particularly streets in towns, became narrower. In general, town streets were twice as thick as country roads, but 3ft narrower, reflecting both wear of traffic and the pressure of roadside development. But few rules can be applied - Britain's Roman roads are endlessly variable.

Individual examples of road repairs have been found. At Stamford, on Ermine Street, a pot-hole was found in 1956 which had been carefully repaired using rammed limestone reinforced with small flagstones. In Cirencester, a roadside building was excavated with a portico that had been progressively raised to keep pace with successive layers of metalling placed on the street. However, the ill-effects of rising street levels could sometimes be alleviated by a parallel rise in the level of the adjacent ground surface. This occurred when a new building was built directly over the top of the remains of its predecessor. At Milk Street in London, both street and adjacent buildings were seen to rise independently, with an intervening drain being raised to keep pace with them.

There were regional differences too. In the south and east, gravel was the predominant material used for roads, sometimes supplemented by cobbles or flints, while bridges, culverts and drains tended to be of timber. In the west and north, roads were on average wider. Larger stones were more often used, usually in the foundation, with some evidence for paving slabs near and within fortresses, and stone was more likely to be used for culverts and bridges. It appears that a heavier form of construction was preferred where the military remained active.

The earliest roads would have bypassed the future site of Londinium, which did not exist at the time of the invasion. The fact that the Roman road network we are familiar with today appears to be centred on London is a result of the rapid rise and dominance of the town. As Roman towns began to expand, roads developed to meet the increased demand for travel between them, and the dominant size of London meant it became the focus of the road system, a role it has retained to this day.

Beyond London

It may, however, be possible to discern a diminishing in the importance of London away from the south and east. Watling Street, for example, forms the main street of most of the settlements it passes through (although its course through Verulamium is far from direct), but by the time it reaches Wroxeter it has lost this role and is connected to the town by a minor street. Wroxeter's main street was formed by the road running north-south along the Welsh Marches, linking the fortresses of Caerleon and Chester.

Today, it is hard to imagine traffic without thinking of congestion. But congestion is not a new problem. Watling Street at Rochester was generally only 9ft wide, sufficient for one-way traffic only. Yet seven layers of metalling suggest heavy traffic and thus, probably, congestion. Many Roman towns in Britain had street plans approximating to the classical rectangular grid pattern, and the cross-road alignment of junctions would have helped traffic to flow smoothly. Yet a number of awkwardly aligned street junctions have been excavated.

At Milk Street in London, for example, a staggered cross-roads was found. A wagon crossing the junction would have had to make two right-angled turns, each of which would have needed to use the full width of the road (wagons need twice their width in order to turn through a right-angle). The wagon would thus have had to wait until any other wagon, or pedestrian, had moved out of its way. Congestion was inevitable.

The steepness of the gradients on some Roman roads suggests that there was no demand for rapid passenger travel using wheeled vehicles. Anyone wishing to travel fast in the Roman period would have needed to use horseback. But while average speeds may not have been high, the roads provided consistent and reliable travel opportunities. The quality of metalling and drainage shows that they were intended to be usable throughout the year, not just in dry weather.

Far from being the result of some grand plan or strategy, with straight lengths of road imposed arbitrarily on the landscape of Britain, we should see the Roman road network as the result of centuries of adaptation, as demands for travel changed over time.

While it may not be easy to discern the precise circumstances which led a particular road to be built in a particular style at a particular place, the realisation that these decisions were not arbitrary gives us the opportunity to make more use of Roman roads to assist in developing our understanding of this period in British history.

Hugh Davies's book, 'Roads in Roman Britain', was published this summer by Tempus at £16.99 (0-7524-2503-X)

CBA web:

British Archaeology

Jan/Feb 2005
Mar/Apr 2005
May/Jun 2005
Jul/Aug 2005
Sep/Oct 2005
Nov/Dec 2005
Jan/Feb 2006
Mar/Apr 2006
May/Jun 2006
Jul/Aug 2006
Sep/Oct 2006
Nov/Dec 2006
Jan/Feb 2007
Mar/Apr 2007
May/Jun 2007
Jul/Aug 2007
Sep/Oct 2007
Nov/Dec 2007
Jan/Feb 2008
Mar/Apr 2008
May/Jun 2008
Jul/Aug 2008
Sep/Oct 2008
Nov/Dec 2008
Jan/Feb 2009
Mar/Apr 2009
May/Jun 2009
Jul/Aug 2009
Sep/Oct 2009
Nov/Dec 2009
Jan/Feb 2010
Mar/Apr 2010
May/Jun 2010
Jul/Aug 2010
Sep/Oct 2010
Nov/Dec 2010
Jan/Feb 2011
Mar/Apr 2011
May/Jun 2011
Jul/Aug 2011
Sep/Oct 2011
Nov/Dec 2011
Jan/Feb 2012
Mar/Apr 2012

CBA Briefing

Fieldwork
Conferences
Noticeboard
Courses & lectures
CBA Network
Grants & awards

CBA homepage