This is Part One of the special on Hadrian's Wall. You can read Part Two here: Backs against the wall, 117-113.

It is 1,900 years since the Roman emperor Hadrian (r. 117-138) made landfall in Britain. His presence marks a departure from business as usual, as the island was not a standard destination for imperial inspections. Instead, Hadrian was only the second reigning emperor to make the trip, following Claudius (r. 41-54), who came to claim the glory for his invasion in AD 43. The sparse surviving Roman accounts tell us little about Hadrian’s activities in Britain, and nothing at all about his motive for visiting in 122. It is certain, though, that the island was convulsed by unrest during his tenure.

One ancient document known as the Historia Augusta bluntly declares that ‘the Britons could not be kept under Roman sway’ following Hadrian’s accession. Sadly, the nature of these disturbances passes without comment, but we do know that decades later the events in Britain were remembered as an example of severe Roman troop losses. These cannot be explained by a misfiring imperial campaign, as Britain was not a target for expansion under Hadrian. Instead, the emperor was something of a maverick for fixating on securing the edges of Roman territory rather than conquering new lands. So, what caused these casualties, and can they help shed light on the greatest legacy of the emperor’s entanglement with Britain: Hadrian’s Wall?

Today, we are so familiar with the existence of the Wall that its construction can appear inevitable. It would not have seemed that way when work on the monument began. In its finished form, the Wall ran for 80 Roman miles (117km) from Wallsend near Newcastle to Bowness-on-Solway beyond Carlisle. For most of its course, there were three linear barriers: a wide ditch to the north, the curtain wall itself, and a huge earthwork called the Vallum to the south. The curtain bristled with more than 160 towers known as turrets and 81 small posts called milecastles. A series of Wall forts placed either on or directly behind the curtain held thousands of soldiers. Nothing even remotely comparable had been seen in Britain before. The result is surely still the greatest military fortification ever fashioned within its shores.

But to what end? The purpose of Hadrian’s Wall remains mysterious. The only Roman statement on the subject, again in the Historia Augusta, declares a desire ‘to separate the barbarians from the Romans’. But what did this mean in practice? Today, few would argue that all north–south movement across Hadrian’s Wall was blocked. Instead, modern scholarly views include the Wall controlling and taxing the peaceful movement of people, or repulsing full-blown barbarian armies. As we will see in the following features, examining earlier clashes between Britons and Romans raises a new possibility: that the Roman army had become enmired in a form of conflict that it was poorly equipped to resolve. Hadrian’s Wall promised a way out.

Battling for Britain: 43-117

It is easy to see the Roman conquest of Britain as a one-sided affair. When Claudius sent an invasion force in AD 43, he was consciously emulating the exploits of Julius Caesar, who led expeditions to the island in 55 and again in 54 BC. The ancient sources make it clear that Claudius’ attack was motivated by a desire to strengthen his position in Rome with a military victory, rather than any threat Britain posed to imperial interests. In this regard, Claudius was doubtless pleased. He duly received his triumph, while an inscription found in Rome states that the initial phase of the conquest was achieved ‘without any reverse’. The surviving ancient historical sources agree that the invasion got off to a promising start, but it was not to last. Ultimately, the advance ran out of steam and Rome failed to complete its conquest of Britain. This was partly a product of imperial politics and more pressing military priorities arising elsewhere. But it also suggests that the Britons were doing some- thing right when it came to resisting Roman aggression. Examining a handful of clashes in the 79 years between the arrivals of the emperors Claudius and Hadrian offers a sense of where the strengths and weaknesses of the opposing forces lay.

Island struggle

It is generally accepted that the core of Claudius’ invasion taskforce was provided by four legions, amounting to roughly 20,000 men. These were the crack citizen soldiers trained to fight as heavy infantry, which acted as Rome’s main instrument of military might, and a modest complement of cavalry. The legions were probably supported by a roughly equal number of auxilia. This term translates as ‘helpers’, and was applied to units of soldiers recruited or conscripted from within Rome’s occupied territories. Such manpower was simultaneously one of the spoils of war and the engine powering further expansion. An ability to draw on specialist fighting styles developed by different groups was a source of considerable Roman pride. While ancient stereotypes – Batavians were natural riders, Syrians skilled archers, and so on – should be treated with as much caution as modern ones, there is no doubt that auxiliary soldiers could help round out Rome’s military capabilities.

Both legions and auxiliary units combined to take the field for Rome’s preferred method of winning wars: in a pitched battle between two massed armies. Set-piece clashes of this kind tend to monopolise discussion of Roman campaigning, and opportunities for martial glory were certainly found in the immediate aftermath of the Claudian invasion. Twice the Britons held rivers in force against the Roman advance, and twice Celtic auxiliaries swum them in full armour to bring the fight to their enemy. The taskforce also halted on the approaches to Camulodunum – Colchester – and summoned reinforcements, supposedly after being checked by fears of staunch opposition. This conceit had been cooked up in advance, though, to ensure Claudius could personally administer the coup de grâce at what was then the pre-eminent settlement in Britain. Breaking British resistance at Camulodunum certainly did not take long once the fighting commenced, as the ancient author Suetonius says that Claudius only spent 16 days on the island. Other than ensuring he reaped the rewards from this audacious military feat, Claudius clearly found little to detain him. The emperor duly ordered his commander to conquer the rest of the island.

There were more battles to be fought as Roman forces spread out and advanced. One legionary commander was the future emperor Vespasian, and Suetonius records his involvement in 30 battles, the subjugation of two tribes, and capture of over 20 towns, as well as the Isle of Wight. The two most famous clashes, though, came later in the conquest period. The denouement of the Boudican revolt in AD 61 and the Battle of Mons Graupius in AD 84 offer examples of set-piece battles that we have unusually detailed information about. In both cases these accounts were written from a Roman perspective – something that is true of all known ancient descriptions of conflict between Roman forces and Britons. The lack of a counter-perspective is keenly felt by scholars of Roman Britain, but despite the certainty that our surviving sources are tainted by bias and blindspots, these battles repay examination.

In the first case, the rebel queen Boudica successfully put Colchester, London, and St Albans to the torch, before setting a numerically vastly superior horde of warriors on a Roman force that was drawn up for battle, but only about 10,000 strong. The upshot was a slaughter: almost 80,000 Britons were reportedly slain, for the loss of just 400 Roman soldiers, with a similar number wounded. Even allowing for considerable Roman exaggeration, the kill ratio of 200:1 implies a catastrophic collapse in effective resistance by Boudica’s force. It was a similar story 23 years later at Mons Graupius, somewhere on the Highland fringe in Scotland. There, a Roman force commanded by the celebrated governor Agricola fielded 8,000 auxiliary infantry and 3,000 cavalry, with the legionaries kept in reserve, against a Caledonian host of 30,000 warriors. Agricola was able to achieve victory without even committing his legionaries to the struggle, with the total casualties claimed to be 10,000 Caledonians for 360 auxiliaries, a kill ratio of 28:1. To put this in some kind of perspective, the figures that Livy gives for the devastating defeat inflicted on Rome by Hannibal at the Battle of Cannae, in 216 BC, indicate a kill ratio of just 6:1. Given that the Britons sustained such horrifying losses, it is little wonder that they, and indeed northern Europeans more generally, have been accused of ‘military ineptitude’ and practising ‘primitive warfare’ during this era.

In the manner of banditry

Perhaps, though, one of the reasons for the appalling casualty rate among the Britons is that pitched battles of this kind were not their usual way of war. There are certainly hints of this in the ancient literature. It is observed, for instance, that the Caledonians assembled at Mons Graupius brandished ‘shields [that] were small and swords too long; for the British swords, without points, did not admit of locked lines and fighting at close quarters’. To put it another way, their kit was poorly suited to this kind of combat. Looking at surviving Iron Age martial equipment from Britain more generally tells a similar story. This has recently been studied by Yvonne Inall, who notes that the armour and helmets protecting Roman soldiers are rarely found associated with Britons. Some exceptions, like the flamboyant headgear sported by the North Bersted warrior, who was buried in West Sussex, seemingly hint at Continental connections. In most cases, protection for British warriors came in the form of a shield alone. While elite warriors could be equipped with swords and heavy spears, allowing them to engage in dramatic duels, throwing spears were commonly wielded by less august combatants. Widespread use of spears makes perfect sense, as they are cheaper and easier to manufacture than swords, but also formidable weapons in skilled hands. Archaeology has revealed victims who were buried with spearheads still lodged in their bodies, providing powerful testimony to the effectiveness of this weapon.

Inall concludes that the surviving kit points squarely towards mobile, loosely formed combat, with most warriors keeping their distance, while ‘throwing spears and hurling insults’. The Britons, she argues, were skilled at ‘raiding and guerrilla warfare’, with the majority of combatants acting as skirmishers. Today, we are familiar with the devastating impact that guerrilla fighters can have on conventionally superior regular forces. Tactics take many forms, but guerrillas frequently specialise in springing surprise attacks to strike soldiers while they are vulnerable, and then making a clean getaway before the regular forces respond in strength. Fighting mobile and elusive enemies can be a major challenge for soldiers trained to better their opponents on the battlefield. A much more modern example of the level of adaptation that it can demand of regular forces is found in a 1959 presidential committee report on the situation in Vietnam, which noted that defeating guerrillas ‘requires widespread deployment rather than concentration. It requires small, mobile, lightly equipped units… It requires different weapons, command systems, communications, logistics…’. Just as the Britons were poorly equipped to face Roman forces in pitched battle, then, so too countering the Britons’ preferred methods may have taken Roman commanders out of their comfort zone.

Although a fascination with the ‘proper’ battles fought between Romans and Britons generally eclipses accounts of other, seemingly lesser, engagements, the ancient histories are seasoned with plenty of examples of guerrilla-style combat. One of the most striking descriptions is provided by Julius Caesar, who offers an eye-witness account of an attack mounted during his expedition in 55 BC. On that occasion, a legion sent to gather corn became vulnerable after the Britons anticipated its movements: ‘when the corn had been cut from the rest of the neighbourhood one part remained, and the enemy, supposing that our troops would come hither, had hidden by night in the woods: then when the men were scattered and, having grounded arms, were engaged cutting corn, they had suddenly attacked them. They had killed a few, throwing the rest into confusion before they could form up, and at the same time surrounding them with horsemen and chariots’. Luck alerted Caesar to the danger, and he arrived to find ‘the legion crowded together, while missiles were being hurled at it from all sides’. Following the arrival of Roman reinforcements, the legionaries were rescued and both sides withdrew.

Broad reruns of this tactic were encountered on numerous occasions following the AD 43 Claudian invasion. Indeed, one of the first challenges facing the Roman force was rooting out an enemy to defeat. We are told that there was ‘a good deal of trouble in searching them out’, with the Britons taking ‘refuge in the swamps and the forests hoping to wear out the invaders in fruitless effort’. This hints at a reliance on low-level surprise attacks prior to the two opposed river crossings. If so, these fleeting encounters provided a foretaste of the far more concerted dose of such resistance that the Roman army would face in Wales and Scotland. In the former case, the Silures – a group living in south-east Wales – proved to be particularly skilled at targeting vulnerable soldiers, allowing them to thwart Roman attempts at conquest for a quarter of a century from AD 49. An account of the struggle written by the Roman historian Tacitus describes frequent attacks conducted ‘in the manner of banditry’, conveying palpable scorn for such desultory combat.

On another occasion, a sizable portion of a legion was targeted in a fashion that presents a broad sequel to the ambush sprung on Caesar’s soldiers: ‘A camp-prefect and some legionary cohorts, left behind to construct garrison-posts in Silurian territory, were attacked from all quarters; and, if relief had not quickly reached the invested troops from the neighbouring forts… they must have perished to the last man.’ Here, then, soldiers were targeted after being detailed to construction duties, and presumably not expecting combat. Tellingly, perhaps, Tacitus demurs from providing Roman or Silurian casualty figures, but he concedes that ‘the prefect fell, with eight centurions and the boldest numbers of the rank and file’. On that basis, the total number of Roman soldiers slain probably exceeded the losses at either the final battle of the Boudican revolt or Mons Graupius. The number of Britons killed, by contrast, is likely to have been far less, as they proved adept at breaking off combat when reinforcements shifted the odds in favour of their opponents. This is spelled out by an incident when a Roman foraging party blundered into a trap. After legionary support arrived on the scene, ‘the enemy escaped with trivial losses’. In almost all cases where we have sufficient details to make a judgement, successes by Britons came hand in hand with surprise attacks.

Chasing shadows

After such spirited resistance, the subjugation of the Silures comes as something of a literary anti-climax. Rather than narrating a dramatic final battle, Tacitus simply states that in the AD 70s the governor Julius Frontinus ‘surmounted not only the valour of the enemy, but also the physical difficulties of the land’. Attacks fizzling out, rather than events coming to a head in a decisive clash, is often the reality of a victory against guerrillas, but if that was the case in Wales, how did the Roman army achieve it? Archaeology can help here, as it was around this time that a web of Roman military posts, combining forts interspersed with much smaller fortlets and occasional towers, was spun through Wales. Such an intricate network appears to be unprecedented in Britain, but is a good fit with what the 19th- and early 20th-century military strategist C E Callwell advocates as the best bet for victory over guerrillas: ‘the sub-division of the theatre of war into sections, each with its commander, its chain of posts, and its mobile columns’. He notes that a combination of larger and smaller posts is essential to safeguard supplies and communications. Frontinus’ decision to spread his forces more widely seemingly marks a step towards embracing the tactics necessary to defeat elusive enemies.

It is a similar story in Scotland. There, Tacitus chronicles especially bold adversaries, who were willing to risk direct assaults on forts, perhaps indicating larger warrior bands. In AD 83, they found another way to catch a legion unawares, when they ‘attacked by night with their combined forces the Ninth (and weakest) Legion: they cut down the pickets and burst in on a scene of somnolent confusion.’ Once again, when Roman reinforcements responded in strength, the Caledonians melted away. Tacitus attempted to put a positive spin on this by declaring that ‘had not the marshes and forests covered the fugitives that victory would have ended the war’. The crucial point, though, is that they did get away. And once again, a Roman strategy in Scotland was to establish a chain of forts interspersed with fortlets, mimicking the approach in Wales.

Why then, if the Caledonians were achieving successes using guerrilla tactics, did they give battle the following year at Mons Graupius? A calculated Roman contrivance must be suspected. Tacitus has Agricola declare in a speech that the enemy ‘have been dragged from their coverts’, suggesting a certain lack of enthusiasm to take the field. Finding a way to force a set-piece battle is recommended by Callwell, who observes that it will greatly benefit the regular forces. He proposes that threatening a settlement or shrine may oblige an enemy to fight. When these options are unavailable, Callwell candidly discloses that ‘regular troops are forced to resort to cattle lifting and village burning and the war assumes an aspect which may shock the humanitarian’. It is noticeable that the Battle of Mons Graupius came at the end of a campaigning season, but Tacitus is uncharacteristically coy about the nature of events leading up to it. Some scholars suspect that this time was spent committing chilling provocations of the kind outlined by Callwell. Whatever the truth, as we have seen, the Caledonian stand at Mons Graupius ended in a cataclysmic defeat. As far as Tacitus was concerned, this marked the end of the affair: Britain was conquered. But events elsewhere were about to intervene.

Just a few years after victory at Mons Graupius, trouble on the Danube prompted the withdrawal of about a quarter of the Roman army in Britain. A force poised for advance found itself plunged into retreat. Although an attempt was made to hold southern Scotland, by AD 105, Roman forces were being reconfigured on the Tyne–Solway isthmus in northern England. This is a place where Britain narrows abruptly and a series of valleys offer an admirable east–west corridor for cross-country movement. The isthmus had been home to Roman military bases since the early AD 70s. At some point, an engineered highway was built between forts at Corbridge in the east and Carlisle in the west. Today, this road is known as the Stanegate, which has given its name to the system of forts, fortlets, and towers that was established on the isthmus. There is ongoing scholarly debate about whether this Stanegate system should be seen as a means to tighten security on an outlying highway, or a frontier in its own right. What is certain, is that the technique of interspersing forts and fortlets, which overlaps with accounts of guerrilla-style resistance in Wales and Scotland, was replicated along the Stanegate in 105 or thereabouts.

That adopting such a configuration was no coincidence is indicated by a document found at Vindolanda, one of the Stanegate forts. A report written around the end of the 1st century AD seems to offer an assessment of the tactics used by local groups: ‘the Britons are unprotected by armour. There are very many cavalry. The cavalry do not use swords nor do the wretched Britons mount [or possibly this should be translated as ‘take up fixed positions’] in order to throw javelins’. Such a description is an excellent fit with the picture of mobile, unarmoured, spear-throwing Britons already built up from the ancient literature and archaeological evidence. It seems likely, then, that the withdrawal from Scotland failed to free Rome from the trials of tackling hit-and-run attacks mounted by elusive adversaries. Indeed, Callwell cautions that the worst mistake a regular army can make when facing guerrillas is to abandon the initiative by ceasing to advance. Perhaps the spectacle of a Roman occupying force surrendering territory did indeed galvanise resistance, or maybe long-simmering tensions were simply brought to the boil. Whatever the cause, as we will see in the following feature, the military situation in Britain was set to deteriorate. •

Matthew Symonds is the author of Protecting the Roman Empire: fortlets, frontiers, and the quest for post-conquest security (Cambridge University Press, 2017).

See also: C E Callwell (1906) Small Wars: their principles and practice.

Yvonne Inall (2019) ‘New light on Iron Age warfare in Britain’, in P Halkon (ed.) The Arras Culture of Eastern Yorkshire: celebrating the Iron Age (Oxbow Books), pp.67-84.

All images: M Symonds, unless otherwise stated.