Along the river to Stretford beyond Tothele; from there straight along the public road which was called by the ancients 'stret' and is now 'fos'; to the other Stretford under Bubbethorne, to the water which is called Ingelbourne
The Foss Way in the Latin bounds of Brokenborough (Wiltshire), 956408
By far the greater part of modern roads go back at least to the Saxon age, and many thousands of miles of them, the ridgeways, have had a continuous existence going back into times long before history began.
G.B. Grundy, 193318
Roads, bridleways, and footpaths tell the traveller how to get to his destination by a route which avoids some natural obstacles; they provide a surface for wheels or feet, and bridges, fords, etc. to overcome other natural obstacles; and they have boundaries within which to confine the traveller and his animals and thus to prevent conflict between travellers and residents.
In modern Britain these three functions are usually combined, but this need not be so. In the less mountainous parts of modern Greece the climate does not require ordinary roads to have special surfaces, the customs of land tenure do not require them to be fenced, and it is thus very easy to make new roads. The stranger is confronted with a plethora of minor roads not marked on the map; it is embarrassingly difficult to distinguish roads in use from disused or unfinished roads, the public Highway from private tracks, and even between roads and the rest of the landscape. The much more stable road system of Britain is partly the consequence of our climate and multitude of rivers, which require paths and roads to be structures and not mere routes; it also reflects English and Welsh, rather than Scottish or continental, attitudes to rights of way. The idea that ownership of land includes rights to keep the public off it and to be rude to well-behaved trespassers is partly due to the general increase in landowners' rights in the last 200 years; but traces of this idea can be found in documents from earlier periods and in the landscape itself.
A well-developed road system is often supposed to be comparatively modern, a product of what we like to think of as an industrial society. Previously, we are told, there had been a 'pre-industrial society' of self-sufficient peasants; they usually stayed at home, and when they did go in for long-distance transport they used waterways. I have seen a learned article on medieval roads based on the fourteenth-century Gough Map, our earliest road-map, as if the main roads which it shows were the total road-system in existence. This thesis is based partly on general notions of progress and partly on complaints of the state of the roads: if roads were as bad in the sixteenth century as people then said theywere, they must have been worse in the fourteenth century and non-existent in the twelfth.
Archaeological and documentary evidence of roads and transport tell a very different story. Complaints are poor evidence: as our own times show, the public is never content with roads, however good, as long as there is a prospect of getting someone else's money spent on bettering them. Transport has had its ups and downs, but its primitive age passed away long before the earliest records. The pre-industrial society, if it ever existed, must be set far back into prehistory. Even in the Mesolithic period, tools of chert were transported well over 100 miles from the source of this rock in the Isle of Portland (Taylor 1979, p.7). From the Neolithic onwards Britain has had a fully-developed network of major and minor communications. The known Roman roads are but a small part of a system that penetrated to every part of England; they are no more representative of all roads in the Roman period than motorways are representative of all twentieth-century roads. By late Anglo-Saxon times it was possible to transport almost anything by land that could be transported before the eighteenth century; the problem of organizing the upkeep of main highways used for long-distance as well as local transport had partly been solved. Inland waterways were rudimentary at this time and for many centuries after.
Highways include not only main and minor roads joining settlements, but also lanes and ways leading to fields and woods. Their history is not merely of pedlars and pack-horses - important though the latter have been at certain times and in certain places.409 The design of roads, bridges, and town streets stems from the invention of the iron-tyred wheel in the Iron Age. No pack-horse can move heavy timber or large building stones. At least since the Anglo-Saxon period, carts have been part of ordinary farm equipment. Where narrow 'pack-horse' bridges survive, these are usually meant to carry pedestrians and animals dryshod while carts go through the adjoining ford. Anyone who has been to Genoa, the lesser Greek islands or the Near East will know the difference between our landscapes and townscapes, shaped by two thousand years of wheels, and those designed for mules and camels.
I shall use the word 'Highway' to include roads, bridleways, and footpaths. The verges as well as the metalling form part of the Highway and are included in its width, unless otherwise stated. 'Lane' will mean a road (less often a path) confined between hedges, fences, or ditches; 'track' will mean a road not so delimited.
Prehistoric tracks and lanes The actual fabric of prehistoric highways is beautifully preserved in the wooden 'trackways', of Neolithic to Iron Age date, across the peat of the Somerset Levels (Chapter 17). These were a specialized and sometimes very elaborate kind of Highway The technology died out before the historic period, and they have left no mark on the visible landscape.
The banks or cropmarks of disused 'Celtic' or other prehistoric field systems often reveal lanes as well as fields (Fig. 12.1 View). Most settlements and farmsteads were reached by at least one lane. Many of them had other lanes ending among fields; but, in prehistory as later, not every field adjoined a lane, and many could be reached only across other fields.
Where prehistoric field systems are still in use it is likely that the lanes that go with them are prehistoric too. This almost certainly true of many of the lanes of Cornwall. Indeed it is very possible, though difficult to prove, that much of our minor road system, at least in Ancient Countryside, is really prehistoric. Many lanes in Hampshire, Essex, or north-west Dorset are certainly of Anglo-Saxon antiquity, but as yet we have little evidence of how much older they may be. Students of disused field systems should look for instances (C.C. Taylor gives a Dorset example) where a lane between former prehistoric fields on downland is continued below the edge of the downland by an existing lane between fields still in use.
Prehistoric long-distance highways The Neolithic was already an age of industry. Stone tools were factory-made articles originating from particular sources of hard stone. The biggest industry was based on the flint mined at Grime's Graves in Norfolk (Chapter 16). Yet Norfolk is full of stone tools from the Lake District, Cornwall, and Wales. The facts of geology, with which there is no arguing, prove not only that Neolithic people had regular long-distance transport but that they had already developed the light-hearted 'coals to Newcastle' attitude to it with which we are familiar today. Transport increased with the invention of metals; and for every enterprise sending out stone or bronze articles which have survived there were probably ten others sending out perishable goods.
Long-distance roads such as the Icknield Way, the various Ridgeways, and the Pilgrims' Way in Stirrey are usually regarded as prehistoric main highways. This is based partly on tradition, but there is an argument which runs as follows. Neolithic, Bronze and Iron Age people lived on high ground where their settlements, barrows, and field systems are still obvious on what is now downland or moorland. These habitable areas were islands in a sea of uncultivable and impassable wildwood on heavy lowland soils. Main routes followed the populated ridges and crossed the inhospitable lowlands, if at all, by the shortest possible way.
This argument was convincing up to thirty years ago but has been weakened by new discovery. We now know that settlements were not specially concentrated on the high ground; the evidence is better preserved there because of less ploughing in more recent times. On low ground the evidence is at least as abundant but is more difficult to find. The 'Jurassic Way' from Stamford to Banbury was claimed in 1940 as prehistoric on the grounds that most of the sites then known lay close to it; vastly more sites have been found since then, and the Jurassic Way is now seen to avoid the more populous parts of Northamptonshire (Taylor 1979). Moreover, the argument that highways with settlements along their length were necessarily important is unsound. Roman, Anglo-Saxon, and medieval minor settlements often avoided contemporary main roads (Fig. 12.2 View), for very good reason. A trunk road is a Highway for armies. At least two armies out of three will be either hostile or unfed; main roads are dangerous neighbours.
The traditional 'prehistoric tracks' are probably genuine, at least in their general direction. Indeed their exact courses, if not their boundaries, are sometimes proved by Anglo-Saxon records and parish boundaries to be well over a thousand years old. Many are ridgeways, minimizing the need for river crossings although, knowing little of what prehistoric people could accomplish in making bridges and fords, we cannot say how advantageous this was. Sometimes, as with the Icknield Way in Cambridgeshire, they avoid heavy soils on which paving would be needed. Often they provide alternative routes. The traveller along the Chilterns, for instance, could choose between the 'Upper Icknield Way', following the chalk escarpment above the spring-line, and the 'Lower Icknield Way' over the claylands at the foot of the scarp. The upper would have been an all-weather road but had steep hills and many diversions round valleys; the lower saved time and effort but its many fords and sloughs would have made it a summer and autumn route (see also Figs. 2.3, 12.3 View)
It can no longer be asserted that the recognized prehistoric tracks are the whole, or even the most important, of those that existed. There were presumably others, at least as important, on low ground, which can now rarely be distinguished from other roads. And the distinction between prehistoric and Roman roads turns out to be less definite than we should expect.
The most obvious legacy of the Romans to modern England is the roads whose uncompromising red lines stare out of almost every page of the motoring atlas. According to tradition, these roads were laid out hy military engineers to link towns and fortresses across the intervening miles of trackless and houseless wildwood. On this principle the system is simple and we understand it already; it is fun to discover unknown Roman roads or to fill in gaps in partially preserved ones, but there is nothing fundamentally new to be found out.
The first flaw in this interpretation is that the majority of Romano-British villages, villas, etc. are not on known Roman roads. A roadless settlement was no more reasonable then than it is now. The roads that we recognize as Roman are merely the trunk roads among a network of lesser roads which, over much of England, can scarcely have been less dense than it is today. The trunk roads were not a pioneering enterprise: they were added to an existing Iron Age network, and they involved expropriating the lands of people who happened to be in the way. The system was maintained, and presumably adapted, during four centuries of great changes in population and commerce. The problems dealt with by the Roman Highway authorities were not, perhaps, so very different from those of their twentieth-century successors.
How Roman roads were organized The Latin language is rich in words for different kinds of road. In other provinces of the Empire, the Romans had much the same classification of major and minor roads as we have in Britain today. At about the time the Romans were conquering Britain, Siculus Flaccus wrote in his textbook of surveying:
There are state roads (uiae publicae), which are maintained at state expense and bear the names of their originators. These are in the charge of superintendents and are repaired by contractors. For the upkeep of some of them a fixed sum is charged to the [local] landowners from time to time.
There are also local roads (oicina/es) which branch off the state roads into the country and often run through to other state roads. These are maintained in a different way, by the counties ~agi), that is by the head-men of the counties, who normally charge the work of keeping them up to the landowners. Sometimes, we understand, they assign to each Landowner certain sections across his own land, which are kept up at his expense. At the ends of the sections they put up notices to show which landlord has to keep up which section over whose land.
There is a [private] right of way to all [pieces of] land. Sometimes, if there is no local road, the way goes over someone else's land . . . this does not provide a right of way for the general public, but [only] for those who have to get to their [own] fields on these roads. There are also shared roads which branch off local roads; these sometimes form a right of way between the two landowners at their further ends, who share the cost equally at their joint expense.410
Other Roman authors confirm that there was a great variety of Highway authorities responsible for the upkeep of particular roads and classes of roads. Major repairs or making a new main road were the work of emperors, consuls, or benefactors, whose deeds are inscribed on milestones. On the whole this was a civilian matter, but the army took some part in making new roads, especially in
the more distant provinces.411
Britain became one such a distant province in AD 43, and its first twenty years were bloody and precarious. Boadicea's revolt in 61 horrified even the Roman world; two of her three massacres would have been averted had the legions got there in time. We should therefore expect to find more evidence of the army and of military roads than in most other provinces of the Empire.
The Roman era in Britain was as long and as eventful as the interval between the Reformation and the motor-car. After the first century the military front moved away north and west, leaving Lowland England as an ordinary civilian region. It had inherited an Iron Age road system and a Roman military road system, neither of them intended for the purposes of the second century. Alterations to road systems during the Roman period should be looked for. The fate of 'General Wade's' roads in Scotland (p.272) is a parallel for what happens to military roads when civilians do not use them. Many substantially-built Roman roads have disappeared so completely that they can only be found by excavation; did they already fall into disuse in Roman times?
Structure and boundaries A Roman road typically had two side-ditches, between which was a raised embankment (Latin agger) bearing a surface metalled with gravel, stone, or ironworks slag.411 The ditches might be 80 feet apart, the agger 40 feet wide and 3 feet high, and the metalling 20 feet wide. Not all main roads had ditches and agger; and the dimensions varied widely, even between different parts of the same road, for reasons which are little known.
Roads had bridges and paved fords. In difficult terrain there were causeways and cuttings; elsewhere in the Empire there were massive timber substructures on soft ground and tunnels through mountains.
Kinds of Roman main road What we notice about Roman roads is their straightness on the map. They do not have quite the same mindless geometry as in the American Mid-West, where a nineteenth-century road will go through a swamp, or bridge a bend in a river twice, rather than mar its straight course by going round. But the Roman surveyors evidently did not look around for the easiest or quickest or most easily defended way between two points. The route chosen was instead the most direct way, and was influenced only by the more formidable natural obstacles. Gradients, fens, wet hilltops, perils of ambush, and any protests of people whose land lay in the way were usually ignored.
The Romans were masters of surveying (Taylor 1979). Whoever set out the Foss Way evidently knew in which direction Lincoln lay from Exeter, to within a fraction of a degree, and also knew that the Somerset Levels and multiple river-crossings in Nottinghamshire were insuperable obstacles to going there direct. Between the two deviations round these obstacles, the middle 150 miles of the Foss Way never depart by more than 6½ miles from the direct line. The departures are not accidental; they are due to the method of setting out the road on the ground in detail after its general alignment had been determined. This was evidently done by sighting from hilltop to hilltop; the road therefore changes direction slightly every few miles, nearly always on high ground (Fig.12.2 View).
The surveyor of Peddar's Way, starting at the north-west corner of Norfolk, knew to within ½ degree which way to go to reach a certain ford of the Little Ouse, 42 miles away. The road twice deviates very slightly, each time on a hilltop, but returns to the original line on reaching the ford (Fig. 12.3 View).
Roman roads vary in the character of their minor realignments. On Peddar's Way there is one absolutely straight stretch of 19 miles between hilltops; further south, what is probably the same road went for 22 miles ruler-straight into Essex. In contrast, Stane Street, the road west from Coichester, looks amateurish with its very frequent changes of direction, rarely more than two miles straight at a stretch and often much less. These are extreme examples of two kinds of Roman road. Peddar's Way and the Foss Way have rare, sometimes rather abrupt, realignments; Stane Street, and the road from London to Colchester and on towards Norwich, have frequent slight changes of direction.
The reason for the two types of Roman main road has not yet been found (if indeed it has been sought) but I shall risk some guesses. The Foss Way was probably made in the heat of war, to join the military bases of Exeter and Lincoln and to support the front line which lay for a time not far to the west. It was surveyed in a hurry through potentially hostile territory; its planners, having all the manpower of the Roman army to do the hard work, had no need to search for ways round minor obstacles; they could reduce the aligning of the road to a mere problem in navigation. Peddar's Way also appears to be a military road. It has led nowhere in historic times - much ink has flowed in vain attempts to invent a destination for it - but it makes sense as one of a grid of roads built to hold down the territory of the Iceni, Boadicea's tribe. Similar roads were made in all directions from London, even into friendly Sussex, to ensure that there should never be another Boadicea.
Roads of the Stane Street type generally have an evident civilian function. Stane Street itself linked Colchester with lesser towns at Braintree and Braughing. Another road of the same kind went north to the capital of the Iceni at Caistor-by-Norwich and a third went south-west to the London area. Such roads are met by the straighter type of road in such a way as often to suggest that the straighter roads are later. Sometimes they conform to field patterns and sometimes not: Stane Street itself cuts across the earlier hedgerow pattern around Braintree as does the Roman road in Fig.8.4, but in The Saints, Suffolk (Fig. 8.2 ) a field system which we have conjectured to be of the Bronze Age has a 'Roman road' conforming to it.
The matter is illuminated by P.C. Dewhurst's study of what is now the green lane called by Cambridge people The Roman Road, alias Wool Street.412 It begins 4 miles south-east of Cambridge and runs for some 11 miles in the dir(cflon of Colchester (Fig. 12.4 View). In 1959 a pipe-trench was dug along the whole 11 miles and revealed its construction. The part nearest Cambridge looks like a Roman road of the Peddar's Way type and follows the canons of Roman road-building: it had two boundary ditches 42 feet apart, the chalk from which was used to form a raised agger which was metalled with gravel. Where this road crossed the pre-existing Icknield Way the agger was lessened in height to form a level crossing. Wool Street next intersects another Roman road of the Peddar's Way type, leading from the military town at Great Chesterford, in such a way as to establish that Wool Street came first. After this crossing, Wool Street changes its character. Its agger peters out and its straight course disappears. It is still recognizable as a 'Roman road', but becomes a wandering one of the Stane Street type, and in section it is no longer raised but a holloway. The difference is clear even on the 1-inch map but until the excavation nobody noticed it.
Wool Street evidently began as a prehistoric road from Coichester to the small Iron Age town of Cambridge. The rest of it has disappeared but for a length, also a holloway, in and near Chalkney Wood, Essex (p.276).413 Early in the Roman period it was modernized, straightened and given an agger. (A certain cart, jolting over the roadworks, spilt a load of Nottinghamshire coal, which is still there buried under the Roman gravel.) The work began at the Cambridge end and was abandoned 9 miles out, leaving the rest of Wool Street in the Iron Age state in which it still remains. The Great Chesterford road was added later in the Roman era.
Even among main roads, the distinction between Roman and prehistoric is thus not very clear. The original planning of Wool Street has to be set back at least to the Iron Age. The same is almost certainly true of Stane Street and may be of the less straight 'Roman roads' in general. In the century before the Roman conquest, Britons could have learnt about Roman road surveying and have tried their hands at copying it. The very straight Roman roads appear, on the whole, to be the work of the Roman army, although Wool Street proves that some even of these are adaptations of earlier roads. Further research may discover among Roman main roads all the changes and complexities that we should expect of a 400-year-old network.
Minor roads Roads with aggers and other obvious Roman features were the class-A and class-B roads of the Romano-British period. The innumerable minor roads were much more diverse than the main roads. The Roman or late Iron Age grids of fields in south Essex (Plate XIII) were served by grids of lanes; in recent decades these lanes have come up in the world, and juggernaut lorries ceaselessly negotiate their right-angle bends on the way to Tilbury Docks. In the unplanned Roman landscape of the Fens the roads - now revealed only as cropmarks - show the same kind of irregularity as prehistoric minor road systems (p.383). Presumably some pre-existing main roads, such as the Icknield Way in Norfolk (Fig. 12.3 View), which were duplicated by Roman roads, survived unmodernized as minor roads. We can rarely be sure of distinguishing minor Roman from prehistoric roads; who can say how many existing country lanes date from either period?
It is sometimes supposed that the Roman road system went out of use and was later partly restored. This is unrealistic. Roads are highly artificial and survive only through continuous use. A gravel road neglected for five years gets overgrown with bushes; after ten years it becomes a thicket more impenetrable than if it had never been a road. Blackthorn has a nasty habit of growing on disused highways and powerfully induces passers-by to go round some other way.
The survival of so many Roman roads is a strong argument for continuity of communication and settlement from the Roman to the Anglo-Saxon period. Roman roads that survive in bits and pieces are more eloquent than those that are still through routes. The men of London and Colchester might have kept open a way between those towns, even if all the intervening country had become a wilderness. But what of the road between Braughing and Great Chesterford? Both places ceased to be towns; the Roman road between them was not kept up as a through route, and Rockell's Wood has for centuries blocked its course (Fig. 12.5 View). But there were still people in these Essex-Hertfordshire backwoods for whom bits of it were of use as local roads. Every few years, through the darkest of the Dark Ages, there has been somebody from Duddenhoe End and Brent Pelham to take a billhook to the blackthorn on two short stretches of Roman road, which stand out by their straightness amid the maze of lanes.
By the later Anglo-Saxon period big heavy things like millstones, timber, salt, iron, and bells and fragile things like pots were being transported to and from the remotest parts of England. Building stone was not necessarily local: stone from Barnack (Peterborough) and Box (near Bath) is found in many places up to 80 miles from the quarry.414 People thought little of going overland to Italy.
The supreme test of Anglo-Saxon communications came in the last weeks of the period. On 20 September 1066 King Harold was told in London that Hardrada had invaded the north. The king covered the 200 miles to York probably in 41/2 days. He gathered an army and smote the Danes at Stamford Bridge (note the place-name). Duke William landed at Pevensey, 250 miles from York, on 28 September. Within three days Harold had heard of this second invasion. It took him just four days to get back to London, this time taking his army, which, as we all know, was nearly sufficient to win the Battle of Hastings on 13 October. Few campaigns until the age of helicopters have packed more action into three weeks; it is a tribute not only to Harold's generalship and the endurance of his men but to the effort that his predecessors had put into organizing roads for just such a contingency.
Highways in the Anglo-Saxon charters One in six of the features in English charter boundaries has to do with communications. A total of 1654 roads, ways, and paths are mentioned, 11.6 per cent of all English boundary features; they outnumber every other class of artificial object. Fords and bridges make a further 4.6 per cent of features. In Welsh boundaries, roads etc. form only 4.4 per cent of features and fords a further 2.8 per cent.
Highways are distributed almost evenly throughout England, except in the Fens (Fig. 12.6 View). Charters from Ancient Countryside mention them rather less often (10.4 per cent) than those from future Planned Countryside (12.8 per cent). Highways were probably at least as numerous in the former but less often selected because there were plenty of woods, hedges, and trees which could be used as landmarks instead.
The many Old English words for Highway indicate some thought of classification. The most important highways were called straet, 'street'. These were evidently vehicle roads, often Roman; many of them are still main roads today. Some had names: there are frequent references to Watling Street (already so called throughout its length), the Berkshire Icknield Street, Ermine Street, and 'Buggilde straet' (now called Ryknild Street, a Roman road in Worcestershire). The Foss Way is often mentioned (see quotation at the head of this chapter).
The term herepao occurs 221 times. It literally means 'army path' and is commonly taken to signify a road made for military purposes. When identified on the map, however, herepaos usually turn out to be ordinary roads without apparent military importance; for instance there is one on the Lizard Peninsula (Cornwall). In Latin charters the phrase used is never the expected via militaris but is instead via publica, a phrase reminiscent of Siculus Flaccus and possibly meaning a road maintained at public expense. This suggests that the idea behind herepao (and the occasional herestaet) is that of a Highway wide enough for an army; and that in practice it means a B-class road in contrast to the straet which is an A-class road. Herepa6s rarely have proper names.
There is also a regional difference. In the Midlands, north-east England, and Kent straet is the commoner term and is used for almost all main roads. In Wessex and Devon most main roads were called herepao, the word straet being reserved for a few very important ones.
Nearly half the highways in the charters are called weg, 'way'. This term probably means something less important than a herepao, but it may also have been vague as in later centuries; the same road in Berkshire is called Icenhilda straet and Icenhilda weg. Most wegs are now minor roads, but whether they were all vehicular in Anglo-Saxon times is uncertain; once a 'wheel-way' is specified. Ways are fairly evenly distributed, except that few are designated in the London Basin, Weald, or Essex.
The terms pao 'path' and stig (Modern German Steig, 'path') are the bottom rank of highways, corresponding presumably to modern bridleways and footpaths (occasionally a 'horsepath' is specified). They are less often mentioned than the others; pao is commonest in the Dorset Chalklands and stig in the Woodless Cotswolds, but little can be made of their distribution.
The word 'road' itself is unknown in Anglo-Saxon literature - it is curiously rare before the seventeenth century - but it does occur, as Old English rod, some fifty times in charters. (The exact frequency is uncertain because of confusion with rod, 'rood', ie. crucifix.) The context implies some kind of Highway but the exact meaning cannot be inferred; there is nothing to support the conjecture of some commentators that a track through a wood, the original of the modern 'ride', is meant. R6d is scattered over most of England.
'Lane' (lane) occurs eighteen times; it is thinly scattered but is absent from areas with a strong open-field tradition. It may well mean, as in the Middle Ages and later, a minor road between hedges, but this cannot be proved.
Many highways are described as greenways, stone-streets, broadways, small-ways, sandy-ways, roughways, etc. The commonest term of usage is sealtstraet, which (with sealtr6d, seahpao, etc.) occurs eighteen times, mainly to the south of the great inland salt-deposits of Worcestershire. Other terms are portstr~t, meaning a road to a market town, and cyricpac) or cyricstig 'church-path' (Fig. 9.5). About one way in fifteen was a 'ridgeway' (hiycgweg), a term found all over England but especially in south Berkshire where there are still many named Ridgeways (eg. Fig. 2.3).
All classes of Highway are quite often described as 'old'. Many ways, paths, etc. were already worn down by centuries of usage: the terms 'holloway' (hola weg), 'hollow path', etc. occur thirty-eight times.
River and stream crossings These are mentioned 666 times in English charters, one-sixth being bridges and the rest fords. Their distribution depends mainly on the character of the rivers - Devon, with its countless small streams, has the highest proportion of fords and very few bridges. Bridges are also rare in chalkland regions with their shallow streams. At the other extreme, in the Fens with their presumably deep and muddy watercourses, half the crossings are bridges. Bridges were also common in North-East England, the Thames valley, Kent, Essex, and the Hampshire Basin.
Place-names give us a different insight into Anglo-Saxon river crossings. About 520 settlements in Domesday Book are named after crossings; of these only some forty mention bridges, the rest being fords or wades. The smaller proportion of bridges is very significant, especially as the crossings generating place-names are likely to be more important than those merely chosen as landmarks. Evidently the place-names date on average from a time when bridges were much less common than in the period of the charters. Towns named after bridges (eg. Boroughbridge, Bridgnorth) have a habit of not being mentioned in Domesday: either they did not exist in 1086 or they were then too recent to have administrative status.
Ford and bridge names have a curious geographical distribution. They are, of course, commonest in Devon, where 8 per cent of places in Domesday Book are named after river-crossings; but why should Huntingdonshire, Oxfordshire, and Berkshire have only slightly fewer? And at the other extreme, why should Lancashire, East Yorkshire, and Sussex each have fewer than 1 place in 100 named after a ford or bridge?
We know nothing of how major fords were arranged. The mind boggles at the thought of main roads crossing the unbridged Thames at Oxford or Wallingford. The frequent place-name Stanford suggests a constructed ford of stone like those which still exist on minor roads, but the evidence has usually been dredged away. Occasional mentions of 'earth-bridge' in the charters remind us that in literary Old English brycg is said to mean 'causeway' as well as 'bridge', but this meaning seems to be rare in topography; a causeway is of no use in crossing a river unless it has a bridge in the middle.
Bridge-repair was neglected in the disorganized centuries after the twilight of Rome. Each of the dozen or more Stratford place-names in England tells of a Roman bridge allowed to rot away. I have, however, found thirteen instances in the charters of stan brycg 'stone bridge', one of them an 'old stone bridge'. Some of these were probably Roman bridges still in use; there are several Roman bridges still in use today in southern Europe, albeit mainly in climates where wet masonry is not attacked by frost as it is here. During the Anglo-Saxon period most of the important bridges were reinstated, and a number of unim-portant crossings were bridged. By 1086 place-names such as Bridgeford had become frequent, although Forde (Hants) did not get its name of Fordingbridge until later.
Bridge-work was probably organized by the state. From at least the eighth century it was one of the 'three common dues' (the others being military service and work on fortifications), obligations imposed by the Crown on all landowners. However generous the king might otherwise be, he almost never remitted these duties. The detailed pre-Conquest specification of Rochester bridge happens to survive.415 Anglo-Saxon engineers had somehow contrived to build a timber-framed bridge of ten spans across the deep and fiercely tidal Medway. The landowners of dozens of named places were made responsible for the upkeep of individual piers and spans. This principle of pontage - spreading the repair cost of a bridge among the townships at some distance on either side of it - was to become a normal method of maintaining bridges in the Middle Ages; as late as C. 1720 my college refused to pay towards Magdalene Bridge, Cambridge, as we should have done on account of owning lands beyond it.
A greater achievement still was the bridging implied by North and South Fambridge in south-east Essex (Fig. 12.7 View). The road and field grids of the Rochford and Dengie peninsulas line up across the Crouch estuary that divides them. A principal Roman road would have crossed between the two Fambridges, both so named (Fanbruge, 'fen-bridge') in Domesday Book. But there has been no bridge since the Middle Ages; the tides of the Crouch sweep 1/4 mile wide between the bridgeheads, though the Ordnance Survey remembers a ferry.
Unless we care to swim our horse, like a gallant lover and his lass in 1598,416 we must try 3 miles upstream at Hullbridge. Here we find that the medieval bridge fell down in the seventeenth century and has not been replaced, though the map, optimistic as ever, claims a ford. The lowest twentieth-century bridge, Battlesbridge, is 5 miles above the Anglo-Saxon bridge.
In the Middle Ages the road system of England was rather denser than it is now. The remote dispersed parish of Wimbish (Essex) had at least twenty-five roads known by name417 this not particularly well documented place is typical of Ancient Countryside, in which almost all the modern lanes existed, together with others that have disappeared. Every wood, meadow, house, and barn and most fields and furlongs had vehicle access, and there were also Footpath rights-of-way across fields. Moors and heaths were criss-crossed with tracks linking hamlets and farms (p.279). In open-field areas only the through roads have survived to the present; the countless ways leading to the strips have mostly been swept away by enclosure.
Much less has been written on medieval than on Roman roads, but there can be no doubt of their importance: what of the Canterbury Tales or the ceaseless journeys of King John? Monasteries depended on pilgrims and on the commercial running of distant estates. In the seaward Fens, every particle of the stone, timber, lime, iron and lead used to build the great twelfth and thirteenth-century churches had to be brought from a distance and to be paid for out of the sale of produce at a distance. The East Anglian cloth industry was not located near its raw material or its markets. Any town presupposes arrangements for moving thousands of tons of various materials. Detailed illustration may be sought in M.K. James's study of the wine trade,418 or in L.F. Salzman's chapter on the gongfarmers whose duty was to transport tons of 'dounge' out of urban cesspits into the country.361
The cost of road transport in the period l250-l450, when the value of money was relatively stable, averaged about 1 3d. per ton-mile for goods that were not fragile or perishable (Fig. 12.8 View). This seems expensive to us; for instance the cost of moving ordinary oak timber 50 miles was roughly equal to the original cost of the tree.'54 Water transport, where available, cost about a fifth as much. But there was little economizing in transport, and the documents refuse to support the theory that heavy materials were always of local origin or else were moved by water. When the twelfth-century monks of Abingdon (Berks) wanted timber, they sent twelve-ox wains 120 miles to North Wales,361 passing by on the way the third and fourth largest concentrations of woodland in England at the time.
The medievals were not great builders of new roads. They planned many new towns (eg. Bury St Edmund's) and sometimes made short stretches of new main road in consequence. Parts of the Fens (p.389) seem to have medieval planned rural road systems. New roads presumably also arose from casual shdrt-cuts hardening into highways. But written records of the conscious making of even minor roads are rare:
The Abbot and Convent shall, at their own expense, cause a hedge to be planted lengthwise from the first gate [on 'Magdeleyn lane'] up to the second, so that the place may become a lane (venella) between the two gates . . . taking branches from the trees growing in the said hedge to maintain the hedge.
Colchester Cartulary, 1272419
This establishes what was meant at the time by 'lane' and by its Latin equivalent venella (cf. Modern Scots vennel).
Documentation of roads Roads and paths appear in documents as permanent features of the landscape. Any through road, even between two villages, is dignified as 'the king's way' (regia via); lesser highways are called 'common way', 'lane', 'church-way' (via ecciesiastica), etc. Surveyors used highways to define the location of pieces of land; manorial Courts fined people for misusing or failing to maintain the Highway. A typical survey entry is:
A croft [ie. hedged field] called Woodwardescroft between the Lord's wood called Walramswoode [now Lord's Wood] on the west and the croft called Mellerscroft on the east; one end abuts on the way from Ledenrothyng to Goodleste' [Good Easter] on the north, the other on the Abbess of Berkyng's land.
Leaden Roding (Essex) 1439292
Medieval petty Courts were as much concerned with road offences as their modern successors. The commonest transgression was allowing ditches (or worse) to flood the Highway:
John Unwyne does not clean his ditch towards Sowenewode [Soane Wood, Chapter 7] on the King's way, to the public Nuisance.
Court roll, Co/chester, 1312420
Nicholas Ravensby . has a slowe [slough, a soft place in a lane] and 2 perches of ditch making a Nuisance in Hawkeley lane
Court roll, Great Canfield (Essex) 1507288
William Barbor junior built a latrine on his holding which runs into the King's way to the Nuisance of passers-by.
Court roll, Hatfield Broad-oak (Essex) 1443274
Every court had such cases by the score; at Little Bentley (Essex) in the 1480s the going rate for fines was ld. per perch (5'/2 yards) of offending ditch.421 Barbor did nothing about his latrine and was fined for it at every subsequent court for at least four years.
Almost as common were fines for leaving timber, wood, earth, muckheaps, dead horses, etc. on the Highway or for digging pits in it. The Highway could be obstructed by allowing hedges and trees to overgrow it (p.187).
A general duty to help with the upkeep of local highways and bridges was often an obligation of land tenure. Farmers were also expected to maintain particular bridges and culverts where water from their land ran on to the Highway:
John Dowe ought to make a bridge, called a fotebregge, in the King's way leading from Dunmowe to Ratfeld brodok . on pain of losing 4s.
Great Canfield 1510288
Little positive attention seems to have been given to surfacing roads, but this mattered less than it would now: wheels were designed for soft ground and many roads were wide enough to pick a way round sloughs. A big consignment of timber was once sent in 1309 from Gamlingay to Grantchester (Cambs) 'at Christmas in hard Frost'.422
Ownership of highways; purprestures Minor lanes were occasionally private property:
Lane. Item, pasture in the lane (venella) leading from Rothyng Tye to Albosdonn 1 acre 2¾ roods 0 perches
Leaden Roding survey, 1439292
With this exception, and except also for unfenced footpaths across fields, highways were part of the common-land of the manor. They had definite boundaries (hedges or ditches) and did not belong to the adjacent farmers (cf. p.281). Much business in manorial Courts was concerned with encroachments on common-land, called purprestures.
Purprestures could occur on any common-land, but they most often took the form of narrowing a road, either by a neighbouring farmer pushing out his frontage or by a third party setting up a smallholding within the road itself (Fig.12.9 View). Manorial Courts often condoned purprestures on payment of an annual fme to the lord of the manor:
Robert Cok made a purpresture in the common way, 10 ft long and 1 ft wide . .
Court roll, Redlingfield, Suffolk, 1276423
A purpresture near the Pillory, 54 ft long, 4 ft wide at one end and 1 ft wide at the other end.
Court roll, Earl's Come, Essex, 1428-9424
A purpresture . . . established by Richard Gyva, turner, of Takeleghe
[bounded] by the ditch of the close of the tenant of the said Richard . . . on the north, and the king's way called Stanstret [the Roman road, p.255] on the south; 65 ft long by the said ditch [measured] by a ruler, and in width at both ends and in the middle 3 ft by a ruler. [Gyva paid ld. a year for it.]
Hatfield Broad-oak 1448274
Also at Hatfield, rents were collected in 1328 for 'a certain place of purpresture for a muckheap' and '1 pit of purpresture outside the gate' ;425 in 14467 John Nedeman was fined for appropriating 14 feet by 8 feet of the King's way for 'a
lay-by (diuersorium) to put his cart in'. 274
A more complicated affair came before the royal Courts in 1412. Sir Thomas Hengrave and three others had been granted a licence to divert the road from Mutford to Carlton Colville (north-east Suffolk) where it passed through Mutford Wood; they had stopped up the old road but had not made a new one, and did not own the land on which to do so. The gang were also in trouble for narrowing another road and throwing mud from their ditches on to it, as well as on poaching charges.426 (The diverted road exists to this day around what remains of the wood.)
Main roads Which were the main roads? The Gough Map, made c.1360, shows main roads throughout Great Britain.427 It is also possible to analyze which roads were used on their daily movements by restless kings such as John and Edward 1428
Many medieval main roads are still trunk roads, but there are important differences of detail. The main road into Cornwall did not march across Bodmin Moor and Gossmoor as does the present A30 but went by devious ways round to the north. At the time of the Peasants' Revolt (1381) the main road from London to Norwich crossed the River Lark at Temple Bridge (Icklingham, Suffolk) where there is now but a Breckland cart-track. The present Al the Roman road known from Anglo-Saxon times as Ermine Street - was not the only medieval main road from London through Huntingdonshirc; there was a choice of three parallel routes, of which Ermine Street is the only one still a main road (Taylor 1979). Similar alternative routes existed in many other places; the state of the road surface and bridges doubtless determined which would be used at any one time. This is why many an obscure lane is now called Old London Road.
Bridges When going to Cornwall by road, the discerning traveller leaves England by one of three great bridges within 7 miles of each other. Greyston Bridge, Horsebridge, and Gunnislake New Bridge are all of the fifteenth century and are still perfect; they cross the Tamar in the deep wooded meanders of its gorge. These are remarkable in their survival, but such a proliferation of noble stone bridges was typical of the later Middle Ages. London Bridge, alas, is no more; but we still have the great tidal bridges of Bideford and Barnstaple (Devon) and Wadebridge (Cornwall); the town bridges of Durham (two), Wakefield, Mon-mouth, Abergavenny, and Stratford-on-Avon; the east Midland bridges of St Ives, Bromham, and Great Barford; the two bridges in the bend of the Dart gorge at Holne Chase (Devon); the Auld Brig o'Ayr, the Auld Brig o'Balgownie (Aberdeen) and the bridge of Carrick-on-Suir (Ireland); the list could be extended to more than a page.
Timber-framed bridges in the Rochester manner have disappeared leaving no mark on the landscape - although the last main-road bridge in this tradition, at Selby (Yorkshire), was demolished only in the 1970s, and one still survives on the Wye below Hay. Something is known of the carpentry of the lesser bridges, whose timbers come to light when moats are drained.429
Exeter West Bridge was one of the earliest of the long stone bridges; six-and-a-half of its original round Norman arches are still extant. It was followed in the thirteenth century by the bridge between Huntingdon and Godmanchester. The masons of these towns could not agree on the architectural details, which change halfway across; but they knew their business, for their bridge now withstands the hourly assault of Juggernaut at fifteen times the load that they would have anticipated.
The chronology of bridge-building is best studied in Cornwall. In the 1920s Charles Henderson recorded thirty-two medieval bridges then extant and written evidence of a further thirty-five. He found also fifty place-names involving the Cornish word pons 'bridge'.430 Five out of six of the bridges themselves were in the eastern half of Cornwall, where the big rivers are; but five out of six pons place-names are in the western half of the county (Fig. 12.10 View). (Cornish-language names of settlements are equally common in both halves.)
I interpret this as meaning that pons could be used of even a small bridge or culvert. Bridges began in Cornwall before 930 - the earliest instance of pons -but were still rather few by the twelfth century, when Cornish ceased to be an active language in east Cornwall. Most bridges were built after 1200 and were therefore named in Cornish in west Cornwall, where the language still lived, but in English in the east. Cornwall was a rather poor and backward region, in which bridging probably came later than in England; but for the same reason many ancient bridges survive in the county.
Ordinary bridges were usually maintained by individuals as a manorial custom:
At Churche Ende, Dunmowe, there is a bridge whiche is daungerous, and so evill that neither horse, nor cart may well pass without danger, neither can men convenientlie parse therebi being the Church path without great annoyaunce. The which bridge is to be repaired by the landeholders of the Parsonage.
We find another bridge broke and gone . . . call Stebbeyng Ford Bryge, but who shall make them, we do not know.
Great Dunmow (Essex), Quarter Sessions, 1562431
Henderson drew attention to the variety of ways in which lords, corporations, and local authorities maintained major bridges, sometimes out of philanthropy, sometimes for the income from tolls. The great bridges had a spiritual and not merely a utilitarian significance: they were works of charity and piety, the gift of benefactors or subscribed by the public through indulgences. Many of them had chapels or even resident hermits.
In the wood of the Prior at Brunne there was a certain carpenter named Peter de Burgo, a good and trusty man. He was cutting down the wood and selling it, and had taken two other carpenters to cut up the timber. Re had made himself a shanty in the wood, in which he ate and drank and spent the night with his mates
Two wealthy and well-known merchants from near Stamford and three strangers . . . were travelling together on the way which is called Arningstrete. They came under the wood of the said Prior after sunset . . . and there the strangers made an attack on the merchants, wounded them, and pulled them from their horses and killed them. The cries of the merchants were heard in the church of [Long] Stowe . . . The inquest decided that the said Peter was not guilty of the death of the merchants, because . . . when they were killed he was in church at Brunne.
The news later came to the ears of King E[dward I] of the death of the merchants and of the place where they had been slain. And an edict went out through all the counties of England that all woods through which there was a common right of way should be cut down to a width of sixty feet on either side of the king's road. And the landowners were set a time-limit under heavy penalty. When he beard this, the prior had the trees standing on the woodbank cut down, and the ditch levelled and filled in, and all the bushes rooted up to a depth of 60 feet . . . William Baldewyn [of Longstowe] did the same with his wood on the other side of the king's road, and so passers-by go in greater security than before.
Liber memorandorum ecclesie de Bernewelle, written 1295-6432
Medieval England, like modern Detroit, was dangerous: the traveller could expect to be mugged. Although heathland was the preferred habitat of the top highwaymen from Robin Hood to Dick Turpin, travellers had a strong and persistent fear of woods and wood-pastures. * Main roads avoided the vicinity of woods where possible, but where woodland was unavoidable it was the practice to make trenches, linear clearings each side of the road, to give travellers the appearance of security (Fig. 6.13, Plate III). This was begun before the Conquest in the Chilterns by the piety of Leofstan, Abbot of St Alban's.433 Many trenches were cut in the thirteenth century, and in 1285 they were made compulsory by a statute to which the Barnwell Chronicle, quoted above, was referring:
Commanded . - . that the high roads from merchant towns . be widened, where there are woods, or hedges, or ditches, so that there be no ditch, underwood (suthboys), or bushes, where a man may lurk to do evil near the road, for two hundred feet on one side and for two hundred feet on the other side. Provided that this statute does not apply to oaks, nor to great trees, if they are clear underneath. And if by default of the Lord, who may not want to remove ditch, underwood, or bushes as provided above, robberies are done . . . and murder, the Lord is to be fined at the will of the King. And if the Lord cannot cut down the underwood the country shall help him to do it. And the King wishes that in his demesne lands and woods within Forests or outside, the roads shall be widened as aforesaid.434
As far as I am aware no new trenches were made after 1300.†
The place of the Barnwell Chronicle murder is still instantly recognizable (Fig. 12.11 View). Arningstrete - Ermine Street, now the A14 - runs dark and narrow between Bourn Wood and Longstowe Wood, then as now the only woods on the long stretch from Huntingdon to Royston. The woods have been allowed to grow up to the road again and the scene is once more as it was on that spring evening 700 years ago. Longstowe church stands within earshot. Set back from the road and now within the woods are the thirteenth-century woodbanks (p.98) made to define the edges of the trenches.
Medieval statutes were not obeyed to the letter. The Bourn-Longstowe trenches are in fact 130 feet wide to the middle of the road. Trenches were made haphazardly: along Stane Street, Hatfield Forest has a trench, but I can find no trace of one in Dunmow High Wood. They were not confined to main roads: the village of Leafield, in the middle of the ring-shaped wooded Wychwood Forest, had several trenches leading to it, of which traces remain. As far as I know the statute was ignored in respect of hedges and parks.
The dissolution of the monasteries in the 1530s destroyed the most powerful corporate bodies with responsibilities and interests in long-distance transport. The medieval principle of road and bridge maintenance as an obligation of land tenure was crumbling; much of the kingdom changed hands, and new owners took up the rights more often than the duties of land ownership. There apparently began a steady increase in traffic. Next came the development of specifically passenger-carrying vehicles, of public transport for persons andgoods, and of the habit of thought that regards time spent in travelling as wasted. From the mid-sixteenth century belated attempts were made to reconstitute arrangements for maintaining roads. Successive statutes empowered parish authorities to conscript labour for the purpose. From the late seventeenth century turnpike trusts were set up to maintain particular roads.
* Mr Paul Moxey tells me that even now some east Londoners are afraid to go into Epping Forest.
†The French have similar trenches which they attribute to a road improvement programme of the seventeenth century.
An indicator of the physical decline of roads in the sixteenth and seventeenth centuries is the rarity of bridges built in this period and still extant. I have mentioned a few of the great medieval bridges; I am at a loss to provide a comparable list of bridges built between 1540 and 1740, even though these have had 200 fewer years in which to fall down. Elizabethan and Stuart bridges are as rare as Elizabethan and Stuart churches: piety found its expression in other directions.
The cost of road transport went up from about 1 3d. per ton-mile in the Middle Ages to 15d. per ton-mile by 1700 (Fig. 12.8 View). Not all this rise was due to inflation: in real terms the cost of transport rose 2½-fold between 1540 and 1690. This increase is probably greater than for any other common goods or services, and is the more remarkable because carters' wages are unlikely to have kept up with inflation.89 Its causes must await specialist study, but a plausible guess is that the rise in cost reflects worsening roads. The turnpike trusts had much work to catch up with; not until 1800 did haulage again become as cheap as it had been before 1550.
New roads One of the chief duties of Parliament in the eighteenth century was to pass Acts for making, or more often improving, particular roads. Turnpike trusts covered, approximately, what are now the more important A-class roads. They had powers to acquire land for new roads and to levy tolls on traffic. 'Tumpiking' and other planned improvements to main roads usually involved surfacing, the rounding of some bends, and making cuttings and embankments to spare horses and brakes on steep hills. The turnpike into Cornwall continued to use Gunnislake New Bridge but lessened the breakneck medieval descent to the bridge by 4nserting four hairpin bends. Some completely new main roads were built. The Epping & Ongar Turnpike Trust in c.1830 ruined the seclusion of Epping Forest by making the present All along its length (Fig. 6.13). At about the same time Great Yarmouth was made less isolated by building Acle New Road across the marshes.
The Old and Young Pretenders, like Boadicea, generated new roads. Military roads, attributed to General Wade, were built across Scotland and northern England to prevent a repetition of the Jacobite rebellions of 1715 and 1745. Some of them were taken over for civilian use and are still main roads; many others fell into decay. On the whole the bridges have lasted better than the roads themselves (Taylor 1979).
In the Scottish Highlands, road systems were often reorganized to follow valleys. Many of the older roads still survive in their pre-tumpike state as tracks across moorland; they often pass unnecessarily high over the mountains.
Turnpike and military roads are outnumbered by the new roads made by the enclosure movement. An early example is the curiously ramshackle seventeenth-century road system of the inland Fens (Fig. 12.12 View). Each allotment of 2 or 3 square miles was given its own geometrical grid in ignorance of the grids of neighbouring allotments. The main roads, where not inherited from the medieval Fenland, have been improvised, with many sudden bends at different angles, out of this chaotic geometry
At the parliamentary enclosure of open-fields it was usual to abolish the many ways leading to furlongs, and to set up a planned and greatly reduced set of parish roads. Some of these were old roads unaltered or narrowed (p.199) or slightly realigned; others were wholly new. Through roads, roads past anciently-enclosed crofts, and holloways were less often tampered with than other roads. The new roads run in the straight lines fashionable at the time, but, as C.C. Taylor points out, the Enclosure Commissioners of one parish often forgot to join their minor roads to those of the next parish, resulting in kinks or double-bends at parish boundaries.
Among piecemeal changes in roads the most obvious are diversions around eighteenth-century parks (Fig. 6.4).
Disappearing highways The building of new main roads in the twentieth century is more than balanced by the disappearance of lesser highways. Not only footpaths and minor roads slip out of existence without ceremony: the old Cambridge-Oxford main road via Croydon (Cambs) is now ploughed over for long stretches, though still legally a Bridleway.
The turnpike movement, as well as the enclosure movement, destroyed highways: by improving some roads it caused many parallel unimproved roads to fall into disuse. Of the three Great North Roads in Huntingdonshire, two are now reduced to bridleways and in part lost altogether, their traffic being concentrated on the Al.
The process was hastened in the 1930s when minor roads were tarred. No County Council could afford to tar all its minor roads; the proportion is probably least in Wiltshire, Which still has thousands of miles of chalk road. Roads not favoured with tar often became disused. Long detours are often necessary in Ancient Countryside because the Council forgot to maintain some link in the network of lanes. The late Mr W.H. Palmer pointed out to me the road from Cavendish to Fenstead End, Suffolk, shown on the Bartholomew V2-inch map 1905 (on the advice of the Cyclists' Touring Club) as a 'good secondary road' but now ploughed out. Nearby, three of the four roads to the deserted hamlet of Purton Green, Stansfield, have gone the same way; one of them has a ford beside which, now hidden in the bushes, is a County Council concrete footbridge. The lane beside Hayley Wood (Cambridgeshire) was once quite a main road; the village of East Hatley was strung out along it. In 1862 it was still important enough to be given a full-sized railway level-crossing with keeper's cottage. It has now been reduced by the encroaching wood to the width of a single horse, and after passing the wood it dissolves into unmarked rights-of-way across fields.
Ancient Highways As They Are Now
The survival of an ancient road can mean many things. At one extreme, on Wheeldale Moor near Whitby, one can walk on what are said to be the very flagstones of a Roman road. At the other extreme, the motorist on a dual Carriageway or the walker threading featureless miles of ploughland may each be following the exact line of an ancient Highway of which no material evidence is now visible. Where the course and the boundaries of a road both survive, they may be of very different dates.
Ancient highways are most often preserved on plateaux or at right angles to the contours on slopes. On cross-slopes they require more effort to maintain against the continual creep of soil from above, and tend to disappear when such efforts are not made. In the Dolomites military roads of World War I are already difficult to find on cross-slopes, being at high altitudes where soil creep is fast. RQads get displaced at river-crossings: where a Roman bridge fell down, the road has often been diverted to a ford a mile away, and diverted to a third site when a medieval bridge was built.
Most well-known ancient roads are either ridgeways or run across country regardless of topography. Roads following valleys - and in mountain country generally - are less likely to be preserved, and where preserved are difficult to distinguish from later roads also constrained to follow the same valleys.
Alignments Ancient roads should first be looked for on the map - preferably the last edition of the 1-inch Ordnance Survey, which records both parish boundaries and public footpaths. Maps display alignments - straight or curving
- that are not easily visible on the ground. Ordnance Survey identifications of Roman and prehistoric roads are useful and scholarly, though by no means complete. Besides these, almost any rural road with its own proper name will be of at least medieval antiquity.
Alignments are rarely optically straight, and to identify them calls on the human brain's mysterious capacity for recognizing patterns. The Roman layout of south-east Essex is still visible in the pattern of north-south and east-west roads on the 1-inch map, though more complete and obvious on the 2'/2-inch map. More difficult to spot are those long-distance minor roads that run across country in great sweeping curves (Fig. 12.13 View). Their names - Drove Road, Bullock Road, etc. - often recall their use in the eighteenth and nineteenth centuries for driving cattle on the hoof from pastures in Ireland, Scotland, and Wales to markets in England. This trade, however, probably came too late to have had much influence on the making of roads, and most drove roads are much older in origin: in Norfolk, for instance, the Weeting Drove is probably prehistoric435 and the Fincham Drove is the Roman trunk road that supplied the port of Caister-by-Yarmouth.
Alignments alone are dangerous evidence. The amateur - and occasionally the Ordnance Survey - forgets that Enclosure-Act as well as Roman roads are straight; a straight stretch of road in a medieval landscape often results from the enclosure of a tye or small common. In a famous book, The Old Straight Track, Alfred Watkins drew attention to the fact that certain ancient sites and other objects lie in approximately straight lines across country. Out of this observation has grown the pseudo-science of 'ley-lines' and their magic properties. Watkins's followers have sometimes allowed enthusiasm to prevail over a sober consideration of whether alignments can have arisen by chance, of the distortions introduced by map-projection, of the chronology of the things aligned, or even of whether anything exists along the course of the alignment.384
Alignments need corroborative evidence. A genuine ancient long-distance road is nearly always a parish boundary at least in places. Where there is an apparent gap the road often continues as a public Footpath, hedge, or earthwork (Plate III). Where the alignment is known or suspected but the road itself is missing, traces of the road structure should be sought in aerial photographs (as soil-marks or crop-marks) and on the ground. Often this is a task for the excavator, but sections of road-metal are sometimes revealed by chance in newly-dug ditches. Ancient woods across an alignment should be searched, as they preserve features elsewhere destroyed by ploughing. When Ongar Great Park (south-west Essex), which for 900 years had interrupted the Roman road from London to Great Dunmow (Fig. 6.2) was grubbed out, the road came to light as a conspicuous band of flints across ploughland.
Near Earl's Come (Essex) an alignment of field and parish boundaries, a lane, and the edge of a wood (Fig. 12.14 View) raises suspicions of an ancient Highway. This can be tested nearby in Chalkney Wood, which is crossed by a massive holloway with traces of boundary-banks. Almost certainly this is part of the long-lost course of the Iron Age Cambridge-Chichester road (p.256). The disappearance of the rest of the road is easily explained. The surviving part is on a plateau. Nearly all the missing course would have lain over cross-slopes on which, on these soils, a disused holloway can survive only in woodland. As soon as the road leaves Chalkney Wood it is destroyed by soil-creep leaving only a faint trace.
Road structures and boundaries The straight course of most Roman roads is more obvious on the map than on the ground. Ancient roads, of whatever origin, usually have a course which on the small scale consists of a series of wobbles (Figs. 12.5 View,12.11, Plate III View). Over the centuries, travellers have had to go round fallen trees, sloughs, holes, muckheaps, purprestures, dead horses, and all the things that people put or allow to remain on the Highway; and they have often continued to go round after the Obstruction has disappeared. The resulting small diversions come to look as if indeed 'the rolling English drunkard made the rolling English road'. Occasionally a road has a strong enough structure, like the great agger of Ermine Street in Lincolnshire, to resist these diversions and to preserve the straight line. Sometimes, on major roads, the alignment has been approximately restored by modern roadworks. With these exceptions, Roman roads rarely have the long views ahead which the 1-inch map, on which widths are exaggerated, indicates that they ought to have. An exactly straight road is more likely to be an Enclosure Act road.
I do not know whether the original boundaries of any Roman road are still functional. Medieval and earlier roads are very variable in width, often within a short distance3 and have boundaries which are even more sinuous than the road itself (Fig. 12.5 View). Where one hedge is straight and the other sinuous this generally indicates that a farmer has seized part of the Highway to his own use. In Ancient Countryside a characteristic feature is the sudden narrowing of the Highway where a cottage in a long narrow garden has been built in it. Some formerly wide main roads - eg. north and south of Braintree, or between Birmingham and Stratford-on-Avon, or south of Sherborne (Dorset) - have such 'squatter' houses and gardens, which may themselves be of some antiquity, going on one after another for miles within the original width of the road. Dating the hedges often helps to establish the sequence of these purprestures.
Post-medieval roads, even if not straight, can usually be recognized by their accurately parallel hedges. In the east Midlands, minor roads, often 60 feet wide, made by the earlier Enclosure Acts, contrast with the usually narrower main roads taken over from the Middle Ages. In Cambridgeshire the enclosure commissioners were less generous.
Holloways An expatriate in a new country, where roads roll out prosaically over the ground surface, misses especially the holloways of the English landscape - the lanes mysteriously sunk in deep ravines which protect them from sun and the blasts of winter, lined with great trees whose roots overhang far above, their cavernous shade the home of delicate plants like hart's-tongue fern, shining cranesbill, and moschatel. Holloways are specially typical of parts of England, and have been for more than a thousand years (p.261).
A very few holloways have A)een made by putting a road into an existing natural ravine - a combination called a 'grundle' in East Anglia. Others, on main roads, are cuttings deliberately excavated to reduce gradients. But most holloways are the result of centuries of erosion on unpaved roads. Traffic loosens the surface and prevents vegetation from holding it, and rain washes away the debris. Usually this requires vehicles, but there are a few foot holloways (eg. the man-wide holloway on the coast path north of Cadgwith, Cornwall).
Holloways are widespread, and to select examples can be little more than to give a list of favourite landscapes. They are abundant in the Lizard Peninsula (Cornwall), south-west Wiltshire (especially the Semley country), and the area south and east of Sudbury (Suffolk). Such landscapes of holloways are typical of Ancient Countryside. Well-developed holloways take at least 300 years to form (I have seen an incipient holloway of some 200 years' wear in Massachusetts) and are therefore less usual in Enclosure Act country, although there are many single holloways inherited from earlier periods that enclosure commissioners failed to destroy.
Development of holloways depends partly on topography - they form most easily on slopes - and partly on geology. The grandest I have seen are in the bess of the Kaiserstuhl in Germany, canyon-like lanes up to 80 feet deep with vertical sides and trees meeting over the top. We do not have bess on this scale, but many readers will know the splendid holloways down to Flatford Mill (East Bergholt, Suffolk), formed in an accumulation of bess washed down from elsewhere in prehistory. Upper Greensand forms the dark and intricate rocksided holloways around Midhurst in the Sussex Weald. Lower Greensand forms many in Wiltshire (eg. around Urchfont) and Dorset.
Holloways usually have the sinuous outlines of other ancient roads, especially on Lower Greensand where the sides may collapse after heavy rain. Where a holloway has been widened to serve as a main road, one side is usually straightened and the other left sinuous; examples may be seen at Hitchin and Truro.
Unfenced highways A Highway, being part of the common-land of the manor, is not usually demarcated from any commons which it may happen to cross. Many roads gradually widen into funnels (in Dorset called 'horns') as they pass into greens, heaths, or wood-pasture (Fig. 6.9). A constant feature of the Lizard Peninsula and other parts of Cornwall is the lane from each farmstead funnelling out between fields on to the moorland.
People crossing unfenced land often find existing tracks too rutted or wet for convenience and make new tracks alongside. Bundles of parallel tracks are typical of ancient routes across heath or moorland and can often be seen, at least from the air, centuries afterwards.
Trenches Although the Ermine Street trenches are now overgrown, many others survived to be recorded on the first edition or even the modern Ordnance Survey. Any parallel-sided gap between a road and a wood (Plate III) should be investigated: there are good examples along the roads through the ancient woods that ring Canterbury and in many places on the Al. Other trenches, now partly overgrown but still detectable through their lack of pollards, line the main roads
- other than Epping New Road - through the wood-pasture of Epping Forest (Fig. 6.12). The presence of a trench proves that both the wood and the road are earlier than 1300; however, its absence does not disprove antiquity. Evidence for trenches has often been destroyed or hidden by recent road-widening, grubbing or replanting of woods, ribbon development, and the fashion for 'planting up odd corners'
Bridges These are a neglected aspect of medieval architecture; even the long bridges are seldom appreciated for the first-rank monuments which they are. Within my lifetime two major medieval bridges in Eastern England - Brandon and St Neot's - have been demolished.
The study of minor ancient bridges is still full of the thrills of discovery. It is not a task for the motorist or even for his passenger; it requires cycling remote lanes and wading under culverts. Medieval bridges tend to have pointed arches, parallel stone ribs under the main arch, and parapets with triangular 'refuges' for pedestrians projecting over the piers, but these features were not always present and the parapets may not survive. Southgate Bridge, Bury St Edmund's, was recently rediscovered buried beneath successive roadwidenings. In Cornwall the arch is sometimes replaced by a massive natural flat boulder used as a 'clapper' lintel (but beware of quarried granite slabs used, even in non-granite areas, in more recent times).
Vegetation The verges of highways are old grassland of a peculiar and often rather unstable kind, traditionally grazed and fertilized by the dung of passing beasts and by washings from the road surface. They are not usually among the richest kinds of grassland, but they are important especially in those regions where old grassland of any kind is now rare. In much of England road and railway verges are now the chief home of such general grassland plants as cowslip, knapweed, rock-rose, and hay-rattle (Rhinanthus minor). Few species are confined to verges, but many are commoner there than in other habitats; these include oat-grass (Arrhenatherum elatius - the characteristic roadside grass which farmers occasionally mistake for wild oats), kex or cow-parsley (Anthriscus syivestris - the well-known spring umbellifer, typical of roadsides rather than of hedges between fields) hedge-garlic (Alliaria petiolata), and black horehound (Ba/iota nigra). Chalkland verges have great knapweed and its broomrape parasite (Centaurea scabiosa and Orobanche elatior). A few national rarities, such as the native grape-hyacinth (Muscari atlanticum), are mainly on roadsides.
Most of these plants are not confined to ancient roads. Some, indeed, depend on recent disturbance, such as the (Socratic) hemlock (Conium maculatum) now profuse along motorways and the rare mullein Verbascum puiverulentum sometimes found on road-widenings in Norfolk. A specially rich kind of verge, related to woodland grassland (p.109), is often to be found where an ancient road approaches an ancient wood.
Tracks and earthen road surfaces themselves support certain plants that withstand being stood on. These include the plantains whose pollen marks the beginning of civilization in the Neolithic. Other plants such as the smaller rushes (7uncus articulatus and J. bufonius) and creeping buttercup grow in ruts and sloughs.
Little-used tracks and paths and the trodden ground around pasture gates have a specialized flora. Most of these plants require seasonal moisture. Species of muddy tracks are probably the most severely threatened class of the British flora apart from cornfield weeds. Most of them appear to stay in one place rather than to colonize new sites; they are lost partly through the tendency for traffic to be concentrated on fewer tracks and partly through the modern love of drainage. One of the famous Lizard Peninsula rarities, the tiny rush Juncus mutabilis, grows in cart-tracks across moorland at places where they cross the edges of bess deposits. This rush, a short-lived annual plant, germinates from buried seed where water seeps into a rut. As D.E. Coombe and L. Frost have shown, this very special habitat depends on a vehicle using a centuries-old track just once or twice a year, and is in danger of being lost through the complete disuse of the tracks. The Lizard has many other cart-rut plants including the curious water-fern Pilularia globuhfera and several aquatic buttercups. Another threatened plant is mousetail, Myosurus minimus, a plant of similar habitats including cattle-trodden gateways on Fenland pastures. The fleabane Pulicaria vulgaris was recorded by John Ray in 1660 'in many watery or moist places of the highways' of Cambridgeshire;436 it has long been extinct in the county and is very severely reduced throughout England. A specialized, and also declining, plant of permanent water trickles in Cornish holloways is the delicate Cornish moneywort, Sibthorpia europaea.
The loss of specialized plants is but one of many threats to the course, structure, and vegetation of historic highways. Ploughing and Obstruction of rights-of-way across fields is a familiar offence which used to be committed with less impunity than it is now:
William atteWater senior ploughed the church-way in a field called Warmelee to the grave Nuisance . . . [fined 1d.]
Court roll, Hatfield Broad-oak 1444274
As we have seen, even roads and lanes with boundary hedges are seized and ploughed out by adjoining landowners, though now perhaps less often than earlier this century.
Purpresture, too, is very much alive. The modern version of the classic practice begins with a householder mowing the verge outside his garden, continues with boulders placed to prevent people from driving on the verge, and ends with the ditch filled in and the verge absorbed into the garden. Some farmers grub out hedges, fill in ditches, and cultivate the Highway up to the edge of the asphalt. This too is an ancient abuse:
The Abbot of Sibton ploughed up a certain Royal way in Thorington [Suffolk] in width 3 feet and in length 20 perches.
Hundred Rolls 1272437
Local authorities are curiously reluctant to prevent public land from thus slipping gratis into private hands. They often evade this duty on the pretext that the soil of highways belongs to the owners of the adjacent lands. This notion apparently arises from confusing those highways that are mere rights-of-way across fields with those that are separate parcels of land in their own right. Most roads, all lanes, and some footpaths are the latter. Highways set out by Enclosure Act or Turnpike Act are expressly vested in the Highway authority; and it is perfectly clear from ancient custom that (with rare exceptions) pre-Enclosure Act highways in England, with their verges, are part of the common land of the manor. In neither case do the highways belong to the adjacent landowners, who have no more right to annex them or their verges than does anyone else.
The grassland of roads is neglected and abused. Verges are dug up from time to time. They may be buried by Highway engineers who can think of nowhere better to get rid of excess subsoil (arising through errors in calculating cut-and-fill). Some county councils in the 1960s (and perhaps even now) sprayed verges with weedkillers in the childlike belief that this would prevent weeds from growing on them; weedkillers, alas, kill cowslips more effectively than they kill weeds. At best, mowing depends on the whims and budgets of Highway authorities. Many verges, once mown too often, are now mown too seldom: they turn into tall tussocky grassland and then into woodland.
It is a pity that so much effort and research should have been put into the new vegetation of motorway verges (which, to judge by the precedent of the railways, hardly need special treatment) and so little into the deteriorating plant cover of existing highways. But the latter is not entirely neglected. The work of the Ramblers' Association in upholding rights-of-way is well known. Less publicized is the activity of the county Naturalists' Trusts in identifying and protecting specially important lanes and verges: for example the work of Mrs J. Mummery for Essex Naturalists' Trust.438 This is done in collaboration with Highway authorities, but usually no special action is asked for. The selected verges are given normal management, but are marked by posts so that any destructive treatment can be avoided.
The History of the Countryside
Oliver Rackham 1986
J M Dent & Sons Ltd
Quick links to book pages:-
248, 249, 250, 251,252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275, 276, 277, 278, 279, 280, 281,