In the late​ 160s ad, Roman recruiters fanned out across the islands east of Jutland, which had long sent its young men to serve on the empire’s Rhine frontier. The armies of Marcus Aurelius needed soldiers, lots of them. Among the several hundred new recruits was a tribesman called Harigast. Like the others, he spoke a form of what scholars now call Proto-Germanic, but the young men came from lots of different places and mutual comprehension was limited, so they soon began to pick up elementary Latin from the recruiters. After making landfall on Jutland, it was five weeks’ march to the imperial frontier, the limes. When they arrived at a fort on the river Main, in the province of Upper Germany, they were subjected to vigorous inspection by a Roman centurion and distributed among the auxiliary camps of the Middle Rhine, where they were taught to march in step, dress a line and use swords and heavy shields, rather than the spears and javelins they knew from home. The Roman army of the second century, somewhat under half a million men all told, was made up of two classes of soldier: legions recruited from Roman citizens and auxiliary units of various shapes and sizes, recruited from non-citizens who could look forward to enfranchisement and a handsome retirement package if they survived 25 years of service. The auxiliaries might envy the better pay and shorter service of the legionaries, but legionaries did a lot of ditch-digging, wall-building and road-mending between campaigns, while auxiliaries got to see the world and enjoy the thrills of the battlefield whether or not there was a big war on.

Marcus Aurelius had been campaigning for half a decade, in Moravia and Slovakia against the Marcomanni and Quadi, and in the Hungarian Puszta against the Sarmatians. These wars chewed up men in formidable numbers, but that was nothing compared to the plague that arrived in Europe in 166, travelling back from Mesopotamia with the victorious legions of Marcus’ adoptive brother and co-emperor, Lucius Verus. The pathogen in question was probably smallpox, and it put whole legions hors de combat. Morbidity was high, thanks to the excellence of Roman road networks and the close quarters in camp, and mortality may have reached 30 per cent in the worst-hit units. For want of troops, Marcus had to call off the entire campaigning season of 168. It was this manpower crisis that sent the Roman recruiters so deep into Scandinavia. There was no evading the disease, everyone caught it eventually; it killed Lucius Verus in 169 and probably Marcus himself eleven years later.

Harigast must have caught it too, but survived, and at some point was promoted out of the ranks to become a junior centurion, with eighty men under his command. Eventually, he found himself the most senior centurion still standing. As primus pilus Harigast was the de facto commander of his unit: auxiliary cohorts were officered by an ever-changing parade of military tribunes, well-born young Romans taking their first steps in imperial service, and by equestrian prefects, junior officers in their first serious posting. As the senior NCO, the primus pilus was the unit’s collective memory and key to its morale – the hoped-for honour was for the entire cohort to be granted citizenship while still in service, after a battlefield performance so heroic that the emperor himself heard of it. But Marcus died in 180 and Commodus, his son and successor, made peace rather than follow up his father’s plan to create two vast new provinces, Marcomannia and Sarmatia, which would have linked the Rhine to the Elbe and Tisza and rationalised the empire’s northern frontier from the North Sea to the Black Sea coast. Ground had already been broken on the big legionary forts that were to form the provinces’ backbone when Commodus called everything back to the Danube line. Tens of thousands of men like Harigast had their promising Roman futures snatched from their hands.

The most recent enlistees went home. The short-timers, those with lives and families on the limes, decided to stay put. But for many of the men, their cohort was their family, their primus pilus its patriarch, and they had every reason to keep on soldiering. About four hundred men decided to follow Harigast and they headed north across the Danube, picking up the Elbe east of Prague. They travelled through countryside still chastened by years of Roman campaigns, but stayed disciplined, marching to schedule, pitching camp and keeping watch. Near what is now Hamburg, they left the Elbe and tracked north, aiming for Jutland. By the time they passed the Flensburg fjord, they had fighting to do. The tribesmen here knew how to use the ponds and marshes of the flat landscape as weapons, and would sink the armour and weapons of defeated invaders in certain special lakes to thank the gods for victory. Harigast’s men suffered some losses, but this only bound them more closely together, as they left the mainland in boats they had commandeered.

On Funen, just east of Jutland, they found that another discharged cohort had settled at Gudme, a few kilometres inland, where they could levy a tax on lucrative traffic between Funen and Zealand. Harigast’s cohort carried on, crossing to Zealand and then to Scania in southernmost Sweden. There, between modern Malmö and Lund, they found good, lush farmland, and did a rough and ready job of surveying a site on which to build a permanent base camp. They squared it up and laid down the two perpendicular main streets that formed a parade ground at the central intersection. At the north-east corner of the parade ground, they put up a commander’s residence for Harigast, a smaller hall for the conduct of business and administration of justice, and a chamber to house their standards and their gods: the praetorium, basilica and sacellum they’d grown used to. The rest of the settlement was shared out by rank, senior officers near the centre, junior officers and men further out. Some of the natives fled, others acquiesced in working for new masters. The cohort kept up its routines, ready to deploy at short notice. Every week, officers and men gathered at the praetorium to feast on the abundant local fish and game and, by their first spring in the new settlement, the seeds they’d carried with them from the empire had sprouted: they had mustard on their pork again, and parsley sauce on their fish. They made a good living, as warriors and traders and lords of their new land. The settlement endured for several centuries. The praetorium, basilica and sacellum dominated the site and were rebuilt every few decades in the face of wind and weather. The founding generation’s descendants followed their ancestors into Roman service, but on very different terms: no longer lone recruits, they travelled to the limes in trained warbands, ready to earn a salary in foreign or civil war.

Harigast’s story is one of many that can now be told thanks to Dagfinn Skre’s extraordinary book, which gives Scandinavia an ancient history for the very first time. Harigast is a composite figure of my own imagining, his story a fabulation of the kind sometimes used to bring to life the stories of subalterns and the historically voiceless. But everything in it did happen, and not just once or to only one individual. Scandinavia has long been a blank for historians of the Roman era: the scholarship is in languages classicists tend not to learn and it is hard to read past the usual framing, which gives separate prehistories of Denmark, Sweden and Norway. Catalogues of Roman artefacts found in the northern barbaricum and distribution maps of Roman coinage make for arid reading. Over the past two generations, the boundaries of the classical world have been stretched to the steppe frontier of China, the coasts of Kerala and Sri Lanka, the Gulf of Aden and Ethiopia. But Scandinavia and the Baltic have remained stubbornly invisible. Skre’s book is too long and too discursive, but it achieves what had seemed impossible: it gives Scandinavia a fully realised place in the history of antiquity, as a dynamic periphery of the Roman world.

Skre reminds us that the impression of Scandinavia’s immense distance from Mediterranean Europe is a legacy of medieval parochialism and the shock of the Vikings. In reality, continental Denmark is only a four-week march from the Roman frontier at twenty kilometres a day (Roman armies did thirty on better roads). Hålogaland in what is now Norway, which lies at 70º north and is, thanks to the Gulf Stream, the northernmost place on earth where arable farming is possible, was settled in this period; it’s a month’s sail to Denmark along the Norvegr, the Norwegian coastal route which is protected by a network of islands and sandbanks. From central Sweden, the mouth of the Vistula is also a month away, and Rome’s Danube frontier a five-week march. These distances are not trivial, but nor are they remarkable by Roman military standards, and Scandinavian men had, on a small scale, been travelling to the empire since the late first century bc, mainly to take service as individuals in the Rhineland.

Then the Marcomannic Wars of Marcus Aurelius, which began in 166, created an enormous demand for new soldiers that Scandinavia helped meet. There are no literary sources at all for this event. But every type of proxy evidence converges on the existence of a late second-century watershed that produced a complete transformation in Scandinavian society and came about in response to interaction with Rome. First, there is the coinage. The Roman silver denarius was first struck during the Second Punic War, but the vast majority of the tens of thousands of denarii found in Scandinavia date to the two decades before 195. A few of these probably circulated within a limited monetary economy, but most were hoarded. Their distribution coincides quite closely with a sudden increase in the number of warrior graves in the century after 180. Still more eloquent are the mass burials of arms and armour preserved in the anaerobic conditions of Jutland bogs – lakes when these spoils of defeated enemies were sunk in them. One of the largest depositions, known as Illerup Ådal, shows how large these defeated armies could be. The bog contains the armour and weapons of at least four hundred warriors; its excavators posit an army of about a thousand men, mainly infantry, with perhaps forty mounted officers. The swords are all of Roman make, while the armour and other gear show three distinct levels of quality and decoration, presumably corresponding to senior and junior officers and regular soldiers respectively. Personal kit, which discloses regional origins more clearly than arms and armour, shows that these armies arrived in Jutland from the north. They came across the Skagerrak and Kattegat from southern Norway and Sweden in the earlier period, and along the Gulf of Bothnia and through the Baltic from central Scandinavia somewhat later. The uniformity of their equipment, and the fact that their numbers match those of either one or two Roman cohorts, suggests they were mercenary units en route to the limes who fell foul of the local population.

These early third-century armies had probably mustered in one of many new settlements that had upended existing patterns of land tenure. The evidence for this disruption again converges on the late second century and again suggests that units returning from the Marcomannic Wars carved out new landholdings, led by the men who had been their unit commanders in the Roman auxiliaries. Buildings and settlement layout at sites like Gudme on Funen, Sorte Muld on Bornholm, and Uppåkra – the model for Harigast’s fortress – in southernmost Sweden show the imperial template. Uppåkra, in particular, resembles a Roman army camp, more or less rectilinear, with adaptations of a praetorium, basilica and sacellum north of the intersection of its main thoroughfares. Such structures have no local precedents in Scandinavia and, where they were built, displaced existing farms and imposed a new order on the landscape. Further north, in southern Norway and central Sweden, analogous settlements were more geographically dispersed but similarly disruptive of existing communities and revealing of Roman inspirations. At Hove in south-east Norway, a sixty-metre-long great hall had to be approached up a flight of stone steps, themselves an imposing rarity in the north, which rose from a rectangular depression that had been excavated and levelled to resemble a Roman forum: this single edifice displaced five or six existing farms.

Whatever the proxy evidence, the Roman imprint is present. Dill, poppy, parsley, cress, carrot, rape and mustard, all staples further south, are first attested in Scandinavia at these new settlements. Upright stones, with legends in the runic futhark script naming both a deceased man and his commemorator, also appear for the first time in this period; on Rome’s northern frontier, epigraphic commemorations peaked in the late second century – this type of epitaph was almost exclusively used for soldiers. Even the archetypally northern custom of ring-giving by lords to retainers was possibly an imitation of the emperor’s distribution of decorations after a victory. These dona militaria frequently took the form of torques and arm-rings and, while still common during the reign of Marcus and his immediate successors, were supplanted by cash disbursements during the first decades of the third century. Again, there is that chronological convergence. Scandinavian warlords revolutionised their cultural world by adapting the lessons they had learned in Roman service, and they did so in a window of less than a generation at the end of the second century.

This new landscape of lordly spaces, Skre demonstrates, was maintained for the next couple of centuries, which would not have been possible in an economy built solely on plunder and conquest. Trade is well documented, along several defined routes: the Norvegr and routes from the south of Norway and west of Sweden, which headed to Jutland and then overland to the Rhine. Baltic routes touched the islands of Gotland and Bornholm before landing at the mouth of the Vistula in Poland, the start of the venerable amber trail to the Carpathian basin and the Danube. A coastal route from Jutland to the estuaries of the Rhine, Meuse and Scheldt was slow to develop, but island-hopping in the Baltic was a habit of long standing. Both commodities and high-value products from the Roman world were filtered through seasonal market sites in the south, such as Lundeborg on Funen. The finest objects generally stayed in the south, but everyday goods – mainly copper cooking equipment from the Rhineland and all sorts of glassware from the Danube – were taken north by the trade routes.

Avariety​ of animal products travelled in the other direction, some of them collected by Sámi intermediaries in the far north, where huge pits for rendering seal, walrus and whale blubber into oil have been discovered. Pitfall traps for elk and fur-bearing mammals were also common, and a Roman wildlife trade which could manage Egyptian hippopotamuses and sub-Saharan giraffes coped with northern species easily enough: an eclogue of Calpurnius Siculus offers the tantalising possibility that a couple of polar bears might have drifted south on the pack ice from Svalbard and ended their lives in the third-century Colosseum. More certain is the trade in iron, which came from north-central Scandinavia where the subarctic pine forest of the taiga begins. Iron of different origin has different properties and thus different uses, and each kind was smelted into a distinctive shape to indicate its provenance: for example, spade-shaped ingots produced in Swedish Jämtland, with thin tips advertising their capacity to hold an edge, have been found in the southern trading hubs on Bornholm and Funen. Slaves were another regional commodity. Roman-style shackles have been found throughout Scandinavia, though not enough for us to guess what role the north played in keeping the empire supplied with human chattel.

Scandinavia’s main export must always have been its fighting men. Demand came and went, tied to the vagaries of imperial politics and the entrepreneurship of particular lords. By the third century, Scandinavian armies seem to have been serving alongside the empire’s enemies too: large numbers of coins, minted in Ukraine and imitating Roman models, made their way back to eastern and central Scandinavia, suggesting that warbands went to fight for the various Gothic leaders who were then establishing a loose hegemony over the region. There was also intra-Scandinavian warfare, both between coalitions of lords and against places like Öland and Gotland that had preserved tribal patterns of landholding in the years after the Marcomannic Wars. The evidence for all this is plentiful and differentiated enough that one could plot many stories like Harigast’s – not as historical fiction, but as the extrapolation of real, if unvoiced, historical activity that has left explicit traces.

Writing protohistory – the narrative of times and places that don’t have a written history of their own, but are within the orbit of somewhere else’s written history – provides opportunities and pitfalls. Scandinavia, before the fifth century, is a perfect example of the opportunities: on the very furthest horizon of which the Graeco-Roman world was aware, while lacking any actual knowledge of it. At the same time, and almost uniquely, a sufficient quantity of legible proxy evidence exists to allow dynamic stories to be told despite the absence of any written records. Both the quantity and the variety are important, because they put real constraints on possible interpretations and forestall choose-your-own-adventure speculation. At the same time, where several interpretations of the proxy evidence are possible, there is no written evidence to nudge one towards a conclusion that might well be wrong, no text to impose its tyranny of words.

It’s insufficiently recognised that the vast majority of the evidence we have for continental Europe in the Roman era is of the same sort that Skre has collected for Scandinavia: proxy traces of human activity. Closer to the imperial frontier, however, the written word intervenes: Greek and Roman sources notice events in the continental barbaricum just often enough to cast a spell, and the non-textual proxy evidence is subverted by the mere existence of the Graeco-Roman literary voice. We are simply so accustomed to putting the written word ahead of any other source of information that the outcome is unavoidable. Even if we consciously try to privilege the proxy evidence of archaeology, and take it as our interpretative guide, it will usually have been excavated with an unconscious awareness of whatever written evidence exists. The same is true of more recent types of proxy evidence, from palaeogenomics to markers of climate change.

Skre demonstrates, with great skill, how to understand history in the absence of written sources, and exposes the extent to which many histories of the fall of Rome bring in archaeological, or climatological, or DNA evidence to justify, or contradict, conclusions predicated on the scant written record. Skre’s last hundred pages feel more familiar because they treat fifth and sixth-century events that are directly related to the collapse of the Roman military presence in the Rhineland and the severing of the eastern and western Roman empires. This is the period when the Svíar, Gautar and Danir appear in the written sources (eponyms for Swedes, Geats and Danes), the era recalled in the genealogies of Beowulf. Others have also told the story of these first evanescent Scandinavian kingdoms and Skre does it as well as any of them. For many readers, these will be the clearest and most conventionally convincing chapters in the book. But they lack the magic of his revelations.