What’s Medieval about the Medieval Frontier?

As I mentioned a couple of posts ago, in early July I had the genuine pleasure of attending Leeds International Medieval Congress. Good academic conferences are identified not just by the questions they answer, but by the questions they don’t; that is to say, the problems they raise in your mind that you realise you don’t have the solution to. It was on the evening of the first day of the IMC that I crossed paths with one such problem. I was lucky enough to be taking part in a roundtable on ‘Rethinking the Medieval Frontier’ organised by the ever-wonderful Jonathan Jarrett (of A Corner of Tenth-Century Europe fame). It was a productive discussion, much of which was concerned with the geography of the medieval frontier, with an unexpectedly large dollop of Islamic cartography thrown in for good measure.

It was only as the session was wrapping up in anticipation of much needed sustenance that the question of periodisation suddenly came to my mind. We had been talking about medieval frontiers. What exactly was medieval about them? What characteristics could we use to identify a medieval frontier, beyond the raw fact of chronology? Sadly, we were heading out the door at this point, so I didn’t have the opportunity to throw this possibly pedantic puzzler to the room. But the question stayed with me.

On one level, this might be a slightly pointless problem. We could say that a medieval frontier is defined by being a frontier that existed in the medieval period. They form part of the medieval historian’s beat because they are accessed through our familiarity with the context, languages, sources, previous scholarship and allied disciplines. In that way and to that extent they can be ‘medieval’ without having to be otherwise distinguished. Further, insisting that everything from the era be in some essential way distinctive betrays a naïve understanding of the vagaries of our periodisation, that they reflect some deep property rather than being terms of art employed for convenience. This is particularly the case with scholarship of the medieval frontier, which has always drawn heavily on case studies from different eras, deriving ideas and insights from the Roman limes, Chinese frontiers and, most famously, the American West.

On another level however, I think it might be genuinely helpful to at least try to consider what might be distinctive about frontiers in the Middle Ages in order to inject a little more reflexivity in thinking about the subject. Partly I would like to have a means of sanity checking comparisons we may want to draw from reading about, say, European interactions with the peoples of the Americas or Qing relations with steppe nomads, by having a set of models about medieval frontiers. Doing this will also aid medieval scholars of the frontier in talking to each other. It is hard for any one historian alone to understand them in aggregate. But all too frequently, when medievalists gather together to discuss frontiers what we get is a ‘show-and-tell’ session where everyone shares their favourite frontier and nothing more coherent than a book with a collection of disparate case studies emerges. Having some sort of model of what makes a medieval frontier medieval might actually give us a means of holding more meaningful conversations.

Unfortunately, I don’t have a model of the medieval frontier, at least not today, and I suspect that if such a model can be made it will have to be by others. What instead I want to do now is propose a couple of hypotheses that I think can be used to distinguish medieval frontiers in Western Eurasia from those that came before and after. These will not be exhaustive and I’d welcome suggestions of other distinguishing marks. I would also call them tendencies rather than features. My sense is that all of the categories I’m about to discuss appear in all of the periods, but that they do so to a greater or lesser extent, and that variation in significance and strength is what I think is distinctive. 

Frontiers of Faith

The first tendency that comes to my mind is one that I think separates the medieval from the earlier Hellenistic/Greco-Roman world, which is the role of faith in the making of the frontier. Roman borders were infused by the numinous and divine. Lares and other deities patrolled and protected boundaries both private and public. The city of Rome itself was defined by its sacred pomerium. Fustel de Coulanges famously argued that the classical city was constituted by its civic cults. Political authorities legislated on matters of piety and morality.

Nonetheless, the classical Roman world had a much more flexible relationship with the divine than in many later eras. Provided you followed the law, paid homage to the imperial cult and didn’t offend public decency, you were more or less left alone. It wasn’t perfect toleration, but it enabled religions from across the empire to spread and build a following. Gods from outside the Roman pantheon could be assimilated. Punic Melqart, Egyptian Amun and Celtic Sulis were discovered to be Hercules, Jupiter and Minerva in disguise. New cults were adopted at the imperial centre. There were limits to this freedom, such as the suppression of the druids in Gaul or the oppression of Christians who refused to acknowledge the imperial cult. The overall picture however is of a fluid world, where religion was not a particularly important dividing characteristic on the frontier.

The embrace of more exclusive faiths by imperial authorities such as Sasanian Zoroastrianism, Christianity in the Roman empire and its successor kingdoms and of course Islam and the Caliphate changes this. We can start seeing ‘religious’ frontiers. Sometimes these were spaces of conflict, where the pious would travel to serve the divine with the sword such as the rabats of ninth-century Cilicia, or thirteenth-century Prussia on the border of Lithuania. At other moments such frontiers could be spaces where people from different religions could meet and talk. Faith could become a marker of difference and you might know whether you had crossed a Christian-Muslim frontier by whether you heard the ringing of bells or the call of the muezzin.

While any student of the Middle Ages could tell you that inter-faith relations were complex, that almost all medieval polities had a mix of religions within their borders, and that people communicated, conducted business, allied and otherwise connected with the religious other, I think this does matter, albeit in complicated ways. Religion could harden the frontier, making the people on the other side more alien, less comprehensible and less moral, while creating reasons for conflict. It could also create contact across the frontier, through pilgrims travelling to holy sites or missionaries come to spread the truth of their faith. Elsewhere these faiths might undermine political boundaries by encouraging people to see themselves as members of a wider ummah or Christendom.

Technological Termini

Modernity inherited much of the medieval world’s relationship with religion. A more interesting distinction that separates it from the medieval world to my mind is the absence of technology as a force for creating types of frontiers. To give you an example, the rise of effective gunpowder siege artillery in the middle of the fifteenth century created a period of about a century where European frontiers became very fluid very quickly, as demonstrated by Mehmed II blasting his way into Constantinople in 1453 or Charles VIII tearing into Italy in 1494. A number of frontier empires appear at this point, such as the Habsburgs and Ottomans dividing up Hungary between them in 1526, or the rapid expansion of Muscovy.

For another example, this time from outside of Europe, the arrival of horses and firearms in the eighteenth century dramatically changed the Great Plains, turning a population of sedentary agriculturalists into highly mobile peoples who depended upon buffalo for subsistence. The result was a much more fluid frontier, where Indigenous groups such as the Comanche and the Lakota built empires that depended upon the exploitation of agriculturalists and interaction with European traders and settlers for goods and markets.

These are distinctive frontier zones that emerged rapidly with the widespread adoption of a set of technologies which allowed people to interact with their environment or existing power structures in a different way in a manner perceptible over the course of a lifetime. I struggle to think of a good medieval analogy. The Vikings, with their shipborne mobility that allowed them to arrive in and connect new spaces may be a possibility, but I’m not sure that’s necessarily a new technology. The use of castles to control frontier zones comes close, particularly something like Edward I’s Ring of Iron across northern Wales in the late thirteenth century, but even here I’m not certain that this represented a dramatic escalation of pre-existing practices of fortification, or was particularly distinct from the use of castles elsewhere away from the frontier.*

Rhuddlan Castle, erected by Edward I in 1277 as part of the last serious infrastructure spending in Wales by a government based in England.

I suspect that there are a couple of reasons for the change. Medieval Eurasia more or less encountered technological change at a similar pace, with ideas and innovations diffusing slowly across the continent. That means that you’re very unlikely to be hit by an entirely new technological package all at once, and your society and political organisation had time to adapt to it. By contrast, improvements in communication meant that people like the Apache and the Comanche might encounter a whole range of new technologies very quickly, with dramatic consequences, as they adapted to a world where they could get horses from the Spanish and guns from the British and create entirely new types of empires on the Plains.

A second factor is that everything counts in large amounts. One large cannon may have an impact, until it explodes or you run out of ammunition for it. But the likes of Mehmed II and Charles VIII could mobilise siege trains of artillery, with multiple big guns and the capacity to build more. This is long before the Industrial Revolution, but the combination of increasingly sophisticated manufacturing meeting states with growing resources allows technology to become more revolutionary because it was appearing on a much larger scale than before.  Factors like these enabled technology to shape the frontier in the modern world in a way that it doesn’t in the medieval world.

State Limits

This third and final tendency is the one I’ve hesitated most over, primarily because it involves piling into a whole range of historiographical landmines I’d prefer not to, but here we go. By the standards of the present, the Roman state and states in early modern Europe were adorably, laughably feeble. Even after Diocletian increased administration, the former was run on a tiny staff of which the emperor and his personal household were not a small percentage. The latter were often ramshackle affairs dependent upon independent contractors and private companies to project power. These states had limited information about their subjects and had to work with local elites on the ground to get anywhere at all.

I suspect that in terms of formal state organisation both of these examples comfortably outclass the administrations of most of western Europe until the thirteenth century. Over the course of the early medieval period, the capacity of states in western Europe to raise tax revenue waned. This had knock-on effects on their ability to maintain a standing army or fortifications. As I’ve specified, this is a process most pronounced in western Europe, but we also see analogous developments in Byzantium and the Caliphate on a slower timescale. It should be noted that there is nothing intrinsically bad about this. Strong states are not inherently good, nor are weak states inherently bad. Even from the perspective of a state, the good state is one that works and acclimatises to the resources and demands of the time. Just as a species of fish that has started inhabiting a lightless cave may lose its eyesight, so a state adapting to new realities may lose traits that were previously adaptive.

The relevance of all this for our purposes is the way this changes the frontier. A frontier staffed by a permanent standing army of professionals whose wages, food, accommodation and supplies are paid for and shipped in by taxes and which is run by career officers who may have served on other frontiers is going to look very different to one run by a military aristocrat whose legitimacy may come from royal appointment, but whose power is based on being the head of a family that has dominated the locality for generations, and who is supporting a retinue on the basis of their own land and grants of land to key subordinates.

The business of supplying the former is going to do exciting things for the economy as it redirects communication and commerce towards the frontier zone. The latter are much more likely to be embedded in the local landscape, unlike the deracinated professional forces. The military aristocrat probably has considerably more autonomy and less oversight. This might lead to them seizing opportunities to expand or plunder over the border. It might instead encourage them to identify with the people on the other side of the frontier with whom they have more in common. There is obviously huge variance here, and exactly how that cashes out I’m not sure, but I would be surprised if this wasn’t a factor.


Frontiers are where easy answers and neat categories go to die. The tendencies I’ve proposed here in this post are not exhaustive. They may not even be accurate. At the moment they represent the best answer I’ve got for this problem. This question came to me as a result of speaking with others and I suspect that that will be where any real answers ultimately emerge. Any attempt to define the medieval frontier will demand a deep understanding of frontiers across an enormous geography and chronology, one that can only be achieved through collaboration with other scholars. Nonetheless, I hope that this represents a contribution to the conversation, even if it only puts the same annoying question that has been buzzing around my head into those of others.

* [Ed.: I don’t usually comment on Sam’s posts, but reading this it occurred to me: if we separate thinking about frontiers from thinking about regnal borders, isn’t this pretty much incastallemento? That is, the model whereby in a few decades around the year 1000 a profusion of castle-building across Frankish Europe transforms the nature of sub-royal polities to create units of power such as, say, the counts of Grignon, dependent on fortresses divorced from Carolingian administrative units. This isn’t to endorse incastellemento as an accurate model of eleventh-century politics – but it’s interesting to think with in this regard…] 


Did India have an Early Middle Ages?

When writers in Europe started dividing history into the ancient-medieval-modern periodisation system familiar today, they did so in light of a past and future which they believed to be Roman. The end of the Western Roman Empire marked the beginning of their medieval period. Likewise, the revival of classical culture and city life that figures such as Petrarch and Bruni perceived in their own times signalled the onset of modernity, a new era that rescued the spirit of the ancient from the medieval caesura. Historians can (and do) question the usefulness of this periodisation for Western Eurasia, but it looks even worse when applied to the majority of the human world for whom Rome was at most a name. While the Roman Emperor Valens was dying in the aftermath of the Battle of Adrianople in 378, Teotihuacán was expanding its political and cultural influence among the cities of the Maya. As Alaric sacked Rome in 410, the Gupta empire was enjoying a year of unexpected peace, following the successful conclusion of a series campaigns by Chandragupta II. And 476, the year that Romulus Augustulus began his career as a former emperor, was the same year that Emperor Xianwen of Northern Wei ended his by being assassinated; but was otherwise just another year in the age of the Northern and Southern dynasties. In this light, attempting to fit these places into a periodisation synchronised to that of the western half of the Roman empire seems perverse.

Worse still, such efforts come with their own history. This is particularly the case with India, the region I’ve been thinking a lot about recently. James Mill’s periodisation of Indian history into Hindu-Muslim-British (The History of British India, 1817) was quickly conflated with ancient-medieval-modern. This naturally suited colonial administrators, as it made modernity in India coterminous with British rule. It was also surprisingly popular with certain groups of Indian nationalists, because it reified their vision of Hindu culture as the output of a pure ‘classical’ period that represented the true India, while implicitly making the ‘Muslim age’, generally identified with the establishment of the Delhi Sultanate in 1206, the bad ‘medieval’ period.

The Ghateshwara Mahadeva temple, dedicated to Shiva, in Rajasthan. Tenth century, but is it early medieval? (source)

All of this context is intended to signal a note of caution for the rest of this post. Over the past couple of years, I’ve spent more time than I previously expected reading about the history of India in the eighth and ninth centuries, initially as I tried to track down the origins of Charlemagne’s elephant, Abu al-Abbas, and later out of straightforward fascination. One of the things that has struck me are the parallels between what I found in India and Western Europe in this period, in ways that I don’t think exist in the Caliphate or Tang China, to give two examples. What I’d like to do today is consider some of those parallels and ask whether we can meaningfully talk about an ‘early medieval India’.

I am grossly beyond the limits of my competence here, with my expertise merely stretching to having read a few books. In the interests of brevity, I have removed most of the qualifications and admissions of ignorance that should follow nearly every sentence. The Indian subcontinent is enormous and fantastically diverse, and most of what I have to say applies to Northern India (the Imperial Cholas are fascinating but seem to me to be doing their own thing). Everything I say should be taken with extreme caution, and I would welcome comments and corrections on this post even more than usual.

I’m also by no means the first person to talk about ‘early medieval’ India. In the 1950s and 60s Marxist historians such as Damodar Dharmananda Kosambi and Ram Sharan Sharma began applying that label to the period between about 550 and 1206. These scholars used the term for these centuries because although they pre-dated the Delhi Sultanate, they identified the rise of feudalism in this period, making it medieval in their eyes. I have spent a career avoiding talking about feudalism in Europe (and most historians who deal with it would probably say it develops after the early medieval phase) so I’m certainly not going to engage with it now in India, except to say that’s not what I’m talking about. Nor am I necessarily talking about the implied chronological sequence inherent in the terminology. My ‘early medieval’ India does not need to be followed by a period resembling the high or later Middle Ages in Europe. Rather, it seems to me that early medieval Europe is defined by a number of features that also appear in India at roughly the same time.

Feature 1

The region used to be dominated by a single large empire with pretensions to universality, run in large part by appointed officials. It is now divided between a number of competing successor states which are each still large entities, and which explicitly claim the legacy of empire, but which are increasingly devolving administration to local landed elites. There is nonetheless considerable continuity in political institutions, as well as in language and culture.

This is fairly obviously the Roman empire, whose legacy carried on most straightforwardly in Byzantium, but also in the ‘barbarian’ kingdoms to the west, seen through continuities in political elites and structures, and the prestige of Latin and Classical culture, right down to Charlemagne having himself crowned Emperor in Rome.

But it’s also the Gupta empire, which dominated northern India from the late third to the sixth centuries. After its decline, large kingdoms emerged, all influenced by the Sanskrit language and culture fostered by the Guptas. The most successful of these competing rulers, Harsha (r.606-647), briefly united most of the region, but this was short-lived. The empires that came after all fought for the same sites, including Harsha’s capital of Kannauj. Particularly interesting is the tendency for tax collection to be farmed out and gradually diminish. Imperial administration increasingly depended on a network of maharajas and rajas embedded in the locality whose status was inherited, rather than the court bureaucrats who had previously been more important.

Possible Objections – The successor kingdoms to the Gupta are generally much less explicit about their imperial inheritance and there isn’t anywhere that resembles Byzantium in terms of prestige and continuity. Although many of the texts that these monarchs drew upon took their form in the Gupta period, the actual legends they’re interested in were set in much earlier centuries. 

Feature 2

A class of religious and cultural specialists spreads across the region. Although they and their beliefs existed before, this period sees the standardisation of their spiritual texts and status. They acquire new prominence, legitimising secular rulers who give them land and the resources to build large temples and other cult structures. A lot of the really interesting intellectual ideas emerges in the form of commentary on scriptural texts, and older literature gets repurposed to fit the spiritual demands of the time.

This is Christianity in Western Eurasia, manifesting itself in the network of church buildings, church officials and institutions such as monasteries, employing large numbers of people and getting resources in the form of land, tithes and influence. Lay and ecclesiastical leaders combined in order to try to determine correct beliefs and practices. Moments like the Carolingian Renaissance get started with the preservation and reinterpretation of Classical work and Patristic writings, with an explosion in expositions and encyclopaedic writings. Classical texts such as Virgil’s Aeneid were reread to reveal Christian truths.

This manifests itself in India as the codification of Hinduism in the post-Gupta world. Key texts such as the Puranas are compiled or standardised, and commentaries start being produced on classical texts. I particularly enjoy the coincidence that Bede (672/3-735, best known in his own time for Biblical commentary) and Adi Shankar (likely early eighth century, the great commentator of the Upanishads and the Bhagavad Ghita among others) probably chronologically overlapped. Brahmins start appearing across the region, acquiring large landholdings from the rulers they advise on correct practice and morality. Kings also found large temples which dominate the landscape. The Ramayana and the Mahabharata start acquiring a much more normative and sacred quality than they previously had.

Possible Objections – There is no equivalent to the Ecumenical Councils, or a Pope or a Patriarch trying to ensure unity and coordinating believers from above. As a consequence, you get much more variety in practice (see for example the Upapuranas which adapt the Puranas with Bengalese local customs). The religious landscape was much more eclectic, with Buddhism (see the Pala empire) and Jainism (the Chalukya empire) continuing to receive royal support, with a little bit of everything showing up everywhere. 

Feature 3

A decline in the size of urban centres and in the volume and importance of long-distance trade, as power moves to the rural countryside.

Cities across the Roman world, but particularly in the western half, become smaller and less dense, and trade, particularly in bulk goods such as grain and pottery, declines. Economies become simpler and more local, with fewer specialists. Similar patterns are described in India in the period.

Possible Objections – This is the one I’m least happy with. Most of the literature I can find makes this case for India, but I’ve also read some convincing work that suggests that the archaeology hasn’t yet been done to really say for certain, and that our understanding of long-distance trade in this period is based on some dubious assumptions, so possibly treat this with even more caution than the rest of this post.

Comparative Thoughts

Assuming that these parallels are real, it’s hard to point to a shared cause. Both the Guptas and the Romans had to deal with invasions by steppe nomads from Central Asia in the fifth century (the Hunas and the Huns respectively, and whether or not these are the same people is an argument I am very happy to stay clear of). But beyond their contributions to the collapse of said empires, it’s hard to see the Hunas/Huns as particularly important to shaping what followed. Likewise, while these regions represented respectively the eastern and the western frontiers of the early Caliphate, it’s difficult to see that commonality as being significant for other shared features. Arguments based on pan-Eurasian or global phenomena such as climate would need to explain the different paths taken in China or the Caliphate at the same time.  

At the moment, I’m inclined to see the parallels as the symptom of a shared pattern – what happens when a unipolar world with a complex literary culture and urban economy starts breaking up. Being a specialist in a different region, what I’m most interested in is how this type of comparison can help me get new perspectives on the places I work on. For me these fall into two categories: things which strike me as very different, and things that are the same but which I hadn’t previously noticed.

An example of the former is discussion of forest-people, now often known as Adivasi. Indian monarchs had a complicated relationship with the inhabitants of the forests, which could be found across the subcontinent, needing the products of their home (including elephants) and their expertise at extracting them, but finding them hard to control. Both parties benefitted from exchange, but viewed each other with suspicion, with the forest-people often being depicted as not quite human.

I can’t really think of an equivalent to this in the western early Middle Ages, where multiple lifestyles based on dramatically different ecological niches were entwined in quite the same way. Mountainous regions like the Alps, or marsh like the Fens just aren’t large enough or essential enough. The closest I can come to are places like the Hungarian Plain, or the homes of the Sámi, but these are fairly geographically contained. Instead, it puts me more in mind of the mutual dependence of pastoralist steppe nomads and their sedentary cereal-agriculture-practising neighbours along the Silk Road. One of the things this suggests to me is just how ecologically specialised the Roman world was. (Another point here might be to compare the thin strip of wheat-producing coastal North Africa ruled by the Roman with the much deeper reach of the Caliphate, which does a better job of engaging with the pastoralist peoples in the interior).

But there are also things that come which I recognise but the significance of which had previously escaped me. Indian scholarship puts a great deal of emphasis on this period as the age when monarchy becomes the default political system, eclipsing the oligarchical republics (gana-sanghas) that had previously spread across India. This is of course something that also happens in the Mediterranean, as the myriad city-states of the Mediterranean with their varied constitutions were first brought together under Roman rule and gradually subsumed into a political world shaped by emperors and kings (the new urban republics of Italy, Flanders and the Hansa would emerge after the early medieval period). This is a process that was obscured to me by my tendency to begin somewhere in the fourth or fifth century, in the more centralised administrative regimes of late antique Rome and the Sasanians, but is something I’d like to keep in mind in the future.

Another point that leapt out to me is the emphasis that many of these studies (particularly those of B.D. Chattopadhyaya and Hermann Kulke) put on the spread of the state to new places and the intensification of the presence of the state in old regions. Competing empires expand their reach into previously peripheral regions while in other places new states emerge in conversation with those from outside (the Deccan being a prime example). Elsewhere, the growing number of sub-kings and temples means that the land is being more intensively governed than it previously was, particularly in the countryside.

I was initially inclined to put this down as something that was different from Europe in this period. However, further thought suggested a number of places where state structures start emerging in early medieval Europe where there hadn’t been before, such as Saxony or Scandinavia. Elsewhere, by the tenth century places such as lowland Britain or northern Iberia are far more intensively governed than they were in the Roman period, through state systems that were less dependent on cities and more on palaces and religious institutions than previously.

I don’t have a grand thesis to conclude this post with. Historians are by necessity specialists in something, if only because the alternative becomes extremely shallow and loses all purchase in historical reality really quickly. But what I hope I’ve suggested are some of the possible benefits of looking for situations and environments that resemble one’s own specialisms in order to get a sense of what similarities and differences we find.

[Editor’s note: after reading this post, there were a lot of points in it I wanted to explore further. I am therefore pleased to say that – schedules of all involved allowing – a specialist in the history of the Indian Ocean has agreed to help us delve deeper into this comparison sometime in Q1 next year, so keep an eye out for that!]

[Further Editor’s note twelve months down the line: I’m sad the previous note didn’t work out, but schedules of all really didn’t allow: all three of us involved changed institutions at least once; in two cases that involved international moves and in one case multiple international moves. I’m still hoping we can work something out but academic precarity being what it is, it’s a difficult thing to co-ordinate…]