Featured post

More Posts you might have missed on the other site

Here, in order from oldest to most recent are the not-exactly-numerous posts that have appeared on the other site in the past two and a half...

Thursday 4 April 2024

“If anyone wishes to cast off their kindred, let them go to the assembly…” Kinship, community, and identity in Merovingian Gaul

[This is a somewhat rewritten version of a plenary lecture I gave at a conference last year on the dissolvement [sic] of kinship ties, shorn of ad libs and jokes and rendered into slightly more formal acadermic language.

Essentially, I take a clause of the Pactus Legis Salicae - concerning the ritual to be followed by someone who wished to cast themselves off from their kin - as a springboard to explore some ideas about what one might call the 'lived reality' of kinship in the post-imperial West. How rigid were the ties of kimship? How significant were family groups? Were people bound tightly into kindreds and constrained to act in accordance with the obligations of kinship? After some close reading of aspects of the clause itself, I look at why someone might want to cast off their kin, or in what ways this might have been desirable. That in turn raises issues of how we think of kinship, whether (and/or to what extent) it is 'natural' and whether (and/or to what extent) it works in ways differently from social ties that are not - notionally at least - related to blood relationships. What change did following the ritual of PLS clause 60 relly make; could you really cast off your kin, or even specific kinship relationships? I argue that, if anything, following this procedure defined you more closely by your relationships to your kin. The paper ends with some geneal conclusions about the strength of kinship and its relationship to community in Merovingian Gaul.

I am wondering whether it's worth trying to publish. I am not sure it is but if anyone thinks it is worth trying to write up properly and has any suggestions (or offers), they can get in touch. My academic email contact is now guy[dot]halsall[dot]academic[at]gmail[dot]com. Clearly it needs a fair bit of work but I think there might be points here that are worth getting into print.]

In this paper I would like to discuss the practical or lived reality of kinship relations in the post-imperial West, using as a springboard a strange-looking clause of the Pactus Legis Salicae (the Compact of Salic Law). This law-code is generally supposed to have been issued by Clovis, the first king of the Merovingian dynasty about whom we know anything, probably shortly before his death in 511. It applied, it appears to tell us, in the lands between the River Loire and the Carbonnarian Forest – the Ardennes – in other words within what had by the seventh century come to be called the kingdom of Neustria or ‘the new lands’, one of the three kingdoms into which Frankish Gaul was usually divided. Clause 60 of the Pactus concerns what you have to do if you want to cast off your kindred. It reads as follows:

Concerning him who wants to remove himself from his kindred

1: [If anyone wants to remove themselves from their kindred], they should walk [ambulare] into the legal assembly [mallus], before the Thunginus or hundredsman and there they should break four sticks of alder wood over their head and they should throw those sticks into the four corners of the assembly and they should say that they cut themselves off from oath-helping or inheritance or from any reckoning (or calculation) of those people.

2: And if afterwards, if someone from their kin should die or be killed, nothing of the inheritance or compensation will pertain to them.

3: If indeed they should die or be killed their inheritance or compensation will not belong to their relatives but will go to the fisc, or to whomever the fisc wants to give it.

Before progressing, some preliminary points about the text and its translation need to be made. The first is to repeat the well-known fact that the generally-used text of the Pactus Legis Salicae is really the best guess of Karl-Augst Eckhardt, worked out in pretty dubious circumstances in the middle of the twentieth century. The standard edition is something of a Frankenstein’s monster put together from several fundamentally different manuscript traditions, all of which are at least 200 years later than the presumed date of the law’s issue. The last section, about the fisc, has been argued to be a later, though still sixth-century, addition.

Second, there is the gendering of the text. The title of the chapter, and the second and third clauses are cast in the masculine generic. This is actually quite unusual for the Pactus, most of whose clauses begin with a gender-neutral ‘si quis’: ‘if anyone’. On that basis Eckhardt reconstructed the start of the chapter the same way: ‘if anyone wants to remove themselves…’, although the manuscript authority for that is pretty weak – two of the later textual traditions. The earlier text-groups do not have anything; they just begin ‘he should walk into the mallus’.

The 1992 translation of the Pactus translated (as it often does) that si quis as ‘he who’, leading to the curious situation where the sixth-century text used more inclusive language than its modern translation. There were, however, two scholarly reasons for that, both based in old assumptions about post-imperial western law. One was that it was all ‘Germanic’, representing different branches of legal tradition that all ultimately went back to some common body of proto-Germanic custom. That in turn meant that the gaps in one code could be filled in from the others. The other assumption, related to that, was that in ‘Germanic’ law only men were legally competent, all women having to live under the legal guardianship or mundium of men. That was based upon clear statements to that effect in the mid-seventh-century Lombard Edict of Rothari, but that code was promulgated in very particular circumstances and there is no reason at all why its ideas should have applied anywhere at all outside northern Italy. Nor indeed does the general idea find support in the texts of non-Lombard Law, or descriptions of court-cases, all of which envisage or indeed demonstrate women managing legal affairs. Nonetheless, it led to translators up to the end of the last century inserting all sorts of inaccuracies, extra phrases, and jurisprudential pretzel logic, into their translations. The text of Salic Law is, however, usually gender-neutral and makes no statements at all about the legal incompetence of women. Indeed, it usually implies the opposite.

What then might it mean that Clause 60 of the Pactus is, according to most text-classes, cast in the masculine? Once we free ourselves of the assumption that only men were legally competent in post-imperial law, I do not think it is very significant. First there is the simple grammatical point that, provided it was envisaged that at least one man be among those referred to, Latin used the masculine third person, as in many modern languages, such as French – until very recently. Second, an assumption about a norm is not the same as a description. By way of analogy, Geoffrey Elton’s The Practice of History is, in a way that now jars but was at the time completely normal, entirely cast in the masculine generic: when a historian reads his sources, he will, …. But, while Elton might have perceived that most historians were men – and even if he thought this was the correct way of things, or thought that women could not or should not really be historians – you could not argue that he thought that all historians actually were men.  The same probably goes for the masculine generic in the Pactus. It might have been the norm (as I suspect, statistically, it still is even now) for most people engaged in law cases to be male but that did not imply that women were never, or had no right to be, actively involved as principals. It might have been rare; it might have brought about a bit of tut-tutting on the part of some people, but it was perfectly legitimate. The circumstances under which one might want to invoke this procedure might indeed have included those that applied to women specifically. For all these reasons I cast my translation in a gender-neutral expression.

Third, note that the law says that the person should walk (ambulare) to the assembly. Salic Law does at least once use the verb festinare – a plaintiff should hurry to someone’s house – but it uses ambulare several times, always when talking about a ritual where a principal has to go to some sort of public assembly or meeting and either say a particular form of words or undergo an ordeal. There are two points about this. The first is that you are supposed to go to the court slowly, in public, in the gaze of the community. The communal gaze was very important in sixth-century Frankish society and local politics. Related to that point, the second observation is that this specification, like a lot of things in Salic Law, seems to try to stop things being done too hastily. It was, I assume, a big deal to cast off one’s kin so an individual might want to have time to back down before they got before the thunginus and did something from which there was no going back.

Fourth, there is the issue of the four rods. Old scholarship suggested, plausibly enough, that the four rods represented the four shared sets of great grandparents implied in Frankish law’s concentration on people related within three generations. In another clause a person can adopt someone or transfer property to them by throwing a stick into their lap. Staffs or sticks feature regularly in Frankish Law but it is not normally stipulated that they be of a specific wood. This clause says you have to have sticks from a particular type of tree, but why alder? Alder is pretty common throughout Europe but it grows in particular places. It grows in wet areas, by or in streams or marshes; it is a ‘pioneer’ tree species: that is, one of the first species of tree to appear if you let an area of land turn to forest: birch is the pioneer species par excellence. You will not find it so easily in established forest; it gets crowded out by the bigger trees. It is thus a kind of tree found in transitional areas, whether the transition is from dry to wetland or from arable to woodland. I wonder if there is some symbolic value in its use in the transition from being within to being without (or outwith) a family. Another feature of alder is that, being a quite supple wood, it is used for binding things, wattles or hurdles for example, so again there might be symbolic value in breaking a wood usually used for binding. As just alluded to, alder is a strong, pliable wood: alder rods are not easy to break. What this seems to me to imply is that if someone did not want to run the risk of not being able to break the rods and having to take time twisting, bending and tearing them (all over the top of their head, remember) and potentially looking rater foolish, they would either have to go and choose their sticks carefully or they would need to let them dry out and become more brittle. The fact that an individual actor would have to go and find a particular type of tree and then either choose a particular dead bit of the tree, or let the sticks dry out, or somehow prepare them before the assembly all seems, again, to show a concern to stop this undertaking from being carried out hastily or in the heat of the moment. [I’d like to mention here the common use of the tree to symbolise the family. Is the breaking of twigs representative of breaking off a particular branch of the family tree? I would like to thank an attendee of the lecture, whose name I have forgotten but it may have been Jessie, for mentioning this - probably quite obvious - point that I had nevertheless contrived not to think of, in discussion afterwards.]

Did anyone ever do this? The short answer is that we do not know; there are no records. At first sight it seems a strange thing just to make up. Even if we think the law recorded actual procedure, however, at some point a judge or collection of elders must have sat down and invented the ritual. It is true that within a century or two even early medieval people thought Salic Law was a bit weird and poked fun at its provisions but, importantly, in his Feudal Society, Marc Bloch refers to a ritual of breaking twigs and throwing them on the ground to cast off a relationship – in that case, interestingly, a tie of lordship and dependence: ‘A ceremony … in which perhaps was revived the memory of the gestures used by the Salian Frank in times gone by.’ Here I will proceed on the basis that this was not all some flight of fancy on the part of the law’s compilers.

My final preliminary point is that Salic Law is very often concerned with the performance of rituals in front of an audience which, by their very strangeness, were memorable. The memorability of the scene transcended the transience of the ritual itself and thus created a sort of communal record of the event and the legal decision. Documents and written records do not feature. As we will see, this is fairly characteristic of sixth-century northern Gallic society and politics.


After this initial discussion of the details of the clause itself, we can pose a key question: why would anyone want to cast off their kindred? The most recent treatment of this issue in English of which I am aware is forty years old, in Alexander Callander Murray’s 1983 book on Germanic Kinship Structure. Murray devotes a few pages to clause 60 of the Pactus, arguing very cogently that the law surely cannot have meant that someone would actually have cast off all of their kindred. How would that have worked? Would it include his own children? It is important to stress that Frankish kinship was bilateral. The legal rights and obligations referred to in the law technically applied to all people related within three generations on their mother’s or father’s side; if someone shared a great-grandparent on either side with someone else, they had legal obligations to them. Stepping outside the law for a moment to look at the archaeological evidence for the sixth-century northern Gallic countryside, we will see that we are talking about very small settlements grouped into small communities, possibly sharing a cemetery as a ritual focus. In these circumstances, the chances of not being related to someone else within three generations on either the mother’s or father’s side were probably significantly less than sharing a great-grandparent with them. Casting off the whole of one’s kindred would likely leave you with no obligations to, or rights to expect support from, almost everyone in one’s everyday community.  Murray’s conclusion that the law cannot have meant shearing an individual off from their entire kindred but, rather, that it meant that someone cut themselves off from obligations to a specific relative, or a particular group within their kindred, is thus surely correct. We should, then, understand the term parentela in this context not as kindred (though it has this meaning) but as kinship (which it can also mean). The person is going to the mallus and declaring that they cut off their kinship with this (or these) particular individual (or individuals).

This seems to make much more sense. Someone – male or female – might wish to sever their links with a specific kinsperson or group of relatives for a wide range of reasons. A segment of someone’s legally-defined kindred might have been serial wrong-doers and the individual at the mallus had simply had enough of having to help pay their fines. This, broadly, is the circumstance Murray envisages. Alternatively, a Frankish woman might have wanted to remarry but her kin might have been causing problems because of their claims on the inheritance. Someone might just have wanted to cut a person or one side of the family out of their inheritance; remember that this clause comes directly after the famous clause 59.6 about the inheritance of ‘Salic Land’. It might alternatively have been a more positive declaration than that. In a situation of bilateral kindreds and kinship legally defined as within three generations, all sorts of problems might have arisen if there was trouble within the (broadly-defined) family. If some of the demands on kin worked as mechanically as they have often been assumed to have done – even to some extent by Murray – if the ‘feud’ worked in the ways it is supposed to have done one might find oneself in some tricky situations. Not entirely flippantly, what would one do if Baddo, a cousin on one’s mother’s side, beat up or killed Dado, a cousin on one’s father’s side? Did one avenge the killing of cousin Dado by killing one of Baddo’s cousins, i.e. oneself? Did one pay a first cousin’s share of Baddo’s fine and then receive a first cousin’s share of the compensation – by buying oneself a cow? As is well-known, many medieval epics play on the tragedies of the supposedly implacable demands of kinship, but the extent to which such stories ever occurred in reality is debatable. It is possible to see the ritual of clause 60 of the Pactus not simply as a declaration of non-kinship but simultaneously as a declaration of more active kinship with someone else, or with a particular part of your broadly-defined kin-group.

This all raises a raft of really fundamental questions about kinship, kinship ties, and society. Given the stress in the historiography about the importance of the kin-group in early medieval society, why would anyone be mad enough to remove themselves from it? We have just seen that that is probably not what is going on in clause 60, but the issue remains. Just how important was the kindred in sixth-century northern Gaul? Were people bound into their kin-groups by all sorts of weighty ties and obligations that compelled them to undertake particular courses of action, in spite of themselves? Some of these assumptions come ultimately from the idea of the post-Roman centuries as ‘the Dark Ages’, when proper government did not exist and when people were compelled to look after themselves. A individual needed to be part of a kindred to protect themselves, to make sure they could get redress in disputes, and to stop everything descending into anarchy and a Hobbesian bellum omnium. Most early medieval historians now reject that way of seeing things, although some of its basic assumptions and implications remain.

Kinship is a social construction, and it is entirely contingent upon time, place, social context. It is important not to confuse it with biological or genetic relationships. There is nothing natural about it. I, for example, have Greek family, but if they were not Greek they would not be family. They are the in-laws of one of my sisters-in-law. Few English people regard the in-laws of their in-laws as family, but my in-laws’ in-laws absolutely regard me as family and treat me as such. It is not just expectations of hospitality that are under consideration, though; if I had a sister, she would, in Greece, be ruled out from marrying their other son. If we had a huge falling out that would for them be a very real dissolution of kinship ties. If someone fell out with their ‘in-law-in-laws’ in England it would be unusual to regard kinship ties as coming into it at all. The dissolution of kinship ties can thus only be understood in specific social context. Kinship ties cannot simply be understood to have existed on the basis of biological consanguinity.

But, leaving aside ideas of obligations and marriage restrictions one would like to think that there were bonds of affection, at least, between close kin that were natural; mammals are instinctively protective towards their children, at least when they are young. Sadly, however, a few depressing minutes with the grimmer side of the news cycle will disabuse the reader of the notion that feelings of protectiveness and affection are automatic or universal. So, indeed will brief exposure to early medieval history; one will soon find kings killing, maiming, torturing, or blinding their own children. King Chilperic I of Neustria had a hand in the killing of two of his sons, as well as one of his brothers; his wife, the notorious Queen Fredegund, allegedly tried to kill one of their daughters by slamming her head in a treasure chest. So, while there may be no reason to think that medieval people were less emotionally invested in their children, or that the range of relationships and emotions within families was any different then from now, by the same token, there seems no reason to suppose that early medieval people automatically loved their children either. The more important variable is social attitudes and the expectations about how one responded to emotions, how one expressed them, when one was expected to suppress them. Mothers in certain social groups might have been expected to hand their babies over to a wet-nurse, regardless of their own feelings. There is some evidence for early medieval infanticide; I do not think we can extrapolate from that unpleasant, disturbing fact to conclusions about the levels of affect between parents and children. These are areas, tensions, often explored in literature.

There are two sides to the issues I have been setting out. Can kinship-ties be dissolved where people did not think kinship existed? It is important not to confuse kinship with biological or genetic relationships. In many slave-owning societies, slaves are not believed to have kinship. The slave is, in the classic formulation of Orlando Patterson, a ‘genealogical isolate’: the slave has no family; he or she can be sold away from parents, siblings and children; slave marriages are not seen as valid. Salic Law seems to share this world view: ‘if anyone steals someone else’s male or female slave, or horse, or mare’ (PLS clause 10.1). A slave is property and, in this world-view, you are no more dissolving kinship ties by selling a child away from its parents than in selling a weaned foal away from its mother. This statement does not imply that such actions did not involve trauma, that they were not horrific. Horrific inhumanity is the defining characteristic of slavery and we should never forget that. The non-recognition of slave kinship (at least in the sixth century, possibly less so in the seventh) is an illustration of the fact. We do not know what slaves thought about kinship; we might suspect that it was fairly provisional and realist. Were they simply expected to suppress the usual, natural feelings towards their children; did they somehow internalise those expectations? If the ideas of the people who owned slaves and who defined the rules or norms of kinship were that slaves existed in a situation outside kinship, and we do not want to fall into the trap of confusing kinship with bonds of affect, then – except when we are talking of the enslavement of hitherto free people – can we talk, analytically, of dissolving kinship ties among slaves? As intimated, attitudes to slave families it may have been changing in the Merovingian period but, as unpalatable as they are, these are things that need close consideration.

Second, we distinguish certain kinds of non-biological kinship as ‘fictive’ but, as social anthropologists used to say in the ‘80s (and as we have in effect seen), all kinship is fictive or at least constructed. People bolstered biological relationships with other forms of kinship. In 613 Chlothar II of Neustria wiped out what remained of the rival branch of the Merovingian family by killing all his first cousins, twice removed, but the one he spared – Merovech – he spared because he was his godson. Creation of such fictive kinship ties did not always work. In the ‘Revolt of the Dukes’ in the late 580s Queen Brunichildis tried to detach Duke Berthefried from his ally Duke Ursio by reminding him that she was the godmother of his daughter, making them co-parents. But that failed; ‘I will never abandon Ursio’, said Berthefried, ‘until death comes to tear me from him.’ Which, inevitably, it did. The obligations of fictive kinship could dissolve as easily as those of blood ties when push came to shove.

The existence of kinship ties should not be assumed on a priori grounds. Historiographically, there was a tradition of reconstructing all of the links that existed between aristocrats, and creating extended noble families – and reifying the families so created by giving them names: Agilolfings, Widonids, etc. The assumption was that these were ipso facto, real and important political groupings. It is, however, empirically absolutely clear that they were not. Just like the Merovingians, these supposed aristocratic clans took sides against, fought with, and killed their relatives. Who a particular socio-political actor took sides with, or against, within their broad kindred, was clearly contingent. People chose which ‘family’ they were part of at any given moment. Close ties of consanguinity need not have been of paramount importance. One of the reasons that there is a constant theme within political rhetoric, of the need for the ‘strong family unit’ is precisely because it is something that has never existed; it is a fetish.

There is an important caveat to be made here, though. These choices were not necessarily made lightly; we are not talking about some emotion-free ‘rational choice’ model. It is important not to ignore potentially ‘irrational’ affective bonds. Hence the law’s stress on deliberation; taking time to reflect. Whether these affective bonds and the emotional stresses in making a choice were any greater when thinking about family than when thinking about having to break with close friends, old comrades in arms (as with Ursio and Berthefried), and so on, is, however, unknowable and, in my view, unlikely.

Here we return to the Pactus Legis Salicae. Clause 58, two clauses before the one used as the focus of this paper, is the clause ‘Concerning the Chrenecruda’. The chrenecruda is another of the Pactus’ strange public rituals, this time about calling upon an individual’s kin to help them pay a fine if they had no more property to give. This involved the person in question getting witnesses together at their house, taking handfuls of earth from their hearth, and throwing them over the relatives they needed to help them pay – in a designated order, from both sides of their family, and to whom they were related within three generations – and then, shirtless and barefoot, but stick in hand (here it can be any kind of stick) jumping over their fence and symbolically abandoning their house. The sorts of kinship ties or obligations referred to in clause 60 had thus to be activated as well as dissolved through public ritual. The dissolution of kinship ties was thus no more of a problem than the activation of such ties. We return to the assumption that somehow these ties and obligations were natural or that the default situation was that they somehow applied. Mostly, though, they existed in what I would call a zone of potentiality, there to be invoked or to be renounced.


On this basis, let us revisit the law and the process it envisages to think about kinship and identity, within a community and what difference the procedure might make to that. As we have seen, everything has to be done properly, in public. The individual in question walked through the community to the assembly. This is a scene into which they were interpellated, in the term often associated with Louis Althusser: they were called upon to speak from a particular subject position. They were called upon to identify themselves. Now, identity is an overused and under-theorised, under-understood term in medieval studies. What does it mean to adopt or declare an identity, to identify yourself? Identities are categories, they are signs and they raise all sorts of signifieds. They carry with them associated imagery, expectations of correct behaviour.

Our person has walked through the community, gone before the mallus and declared themselves to be related to particular people. They have identified themselves as occupying a particular position within a web of relationships. This is narrowed to some extent by the context but the implications of this law presumably make relevant all sorts of other aspects of their person: age, gender, ethnicity, wealth. The question I want to ask is whether – or in what ways - the procedure actually dissolves kinship ties. The person goes before the community, in a formal setting and, in effect, says ‘here I am’. The other people of the community, seeing them, know them. We know you; you are so-and-so, son of so-and-so and so-and-so. What, then, might it have meant to go before the community, in this context, and declare and create a situation of non-kinship?

John Lennon’s mawkish song ‘Imagine’ contains the lines ‘imagine there’s no heaven / it’s easy if you try’. But it is not at all easy. You cannot imagine there is no heaven without simultaneously bringing to mind the idea of heaven that you want to imagine does not exist. This is the problem that undermines the notion of casting off kinship by public ritual. If you go to court and say, effectively, that by this strange and memorable ritual ‘I want to you to record that I am not related to this or that person, or these people, who, we all know, is/are actually one of my relatives/ are actually my relatives’, that is not something that can be remembered separately from the knowledge that in fact, you are that person’s cousin, or a member of that family. You know my cousin? Imagine I am not their cousin; it is easy if you try. In the circumstances where you might come to me, as a member of that kindred, or as that person’s cousin, to call upon me to do something, on that basis, you should no longer call upon me to act in that way. You have not dissolved the kinship tie at all, as much as fixed it in a particular mode. Henceforth you are that one of the Joneses who cast off his kinship; Baddo’s cousin who did not want to be his cousin. That is ironic given what I said about kinship relations normally existing in a vague zone of potentiality. You might not generally be regarded as a particular person’s second cousin, or first cousin, once removed, in everyday life; henceforth that non-kinship would become a key part of your identity. Even with close kin, whose relationship to you would normally be known, if the problem was that you wanted to cut your links to difficult kin, the problem would hardly be alleviated by this process; if anything it would only make things worse. The spectre of that kin relationship will constantly haunt you.

The only circumstance, really, where this would not effectively be the case would be if you moved to a completely different part of the kingdom, Aquitaine say, where different laws applied anyway. The person is saying ‘I am off and I am not coming back; do not come looking for me in Clermont or Tours or Bordeaux to help with fines; and I promise I have no claim on my kin back here.’ As I wrote this lecture, I began to wonder whether that is really what it is about; it makes sense of the addition of the clause about the fisc too. It is possibly the only circumstance in which kinship and its ties would really be dissolved, at least over time. Apart from with death. I’ll just leave that possibility hanging there.

What makes me think that Clause 60 of the PLS is not really about what to do if you move somewhere else is its context within the code. It comes at the end of a little cluster of clauses: clause 56 is about someone who refuses to come to court; clause 57 is about judges who refuse to give a judgement; clause 58, concerning Chrenecruda, is about the situation where someone does not have the wherewithal to pay a fine and, I assume, his relatives do not automatically choose to help them out; clause 59 is about what happens where someone does not have the usual heirs; and clause 60 concerns what is to be done if someone does not want to be part of a legally-defined kin-group. All four clauses concern situations where the usual procedures do not apply.


As far as I am aware, this is the only post-imperial law about casting off your kin. The corpus of so-called ‘Barbarian’ law from the immediately post-imperial period has nothing else like this. It is difficult to know what to make of such a comparison. The different codes cover different things (it was partly for that reason that the notion of a shared Germanic law was so attractive; earlier historians were tempted to fill in the gaps in one code with clauses from another). Partly this is because the codes are not just records of what they at least suggest are the standard logics and procedures of customary law; some are also collections of legislation relating to very specific circumstances; and some individual clauses are so specific that they can only represent responses to individual cases that had, so to speak, recently crossed the king’s desk. The Lombard Code of Rothari, for instance, has two clauses about people killed by bits falling off buildings being constructed by Master Builders of Como. There are surprising absences even so. There is almost nothing in early Anglo-Saxon law about inheritance. Thus I am not sure that one can extrapolate with any confidence from the fact that the Pactus discusses how to abrogate a kinship relation, whereas other codes do not, to differences in the strength of the kindred between Frankish Gaul and other areas of the West.

Nonetheless, even if we cannot compare with other areas, we can at least ask questions about sixth-century northern Gaul. It is interesting that the lawmakers thought it was worth including a clause on this topic, and might even have known of a ritual that was used. It is, however, impossible to decide whether this was because they thought people were likely to want to renounce or dissolve a kin-relationship, and thus that kin-groupings were pretty fluid in practice, or whether the opposite was true: that the bonds of kinship were so strong that you needed to set out, clearly, a specific, deliberate, memorable public ritual for situations when people really wanted to cast them off. Sixth-century social structure in northern Gaul does seem to have been rather more fluid than it was even a century later but can we extrapolate from that to a corresponding fluidity of kinship, or might we rather argue that the instability we see in other areas might have led to a strengthening of kinship as a unit of social organisation? It is ultimately impossible to choose between these options. What seems clear, though, is the importance of the local community in determining acceptable or unacceptable courses of action. This might have transcended the importance of family.

As I have mentioned, Salic Law is full of public rituals like that in clause 60, carried out in the communal gaze. Like sixth-century burial ritual, they create memorable scenes that record acts and decisions in a situation where written records seem not to have impinged on the concerns of the law’s compilers. They are mostly – though far from entirely – absent from Salic Law’s seventh-century equivalent, Ripuarian Law. But the absence from Ripuarian Law of these rituals is probably related precisely to two things. First: Ripuarian Law frequently states that you can get out of the usual procedures of customary law via the use of documents. We have surviving examples of model documents drawn up precisely to do the sorts of exceptional thing that Salic Law required some kind of public performance to carry out. But access to written documents was socially restricted, so one of the ways in which the increasing rigidity of the social hierarchy manifested itself may have been in the limitation of the opportunities to avoid restrictions of custom to people of higher social classes. Related to that is the second point, which concerns time. Salic Law is very much about the here and now. It has almost no concern with the projection of situations, or how they might change, into the future. What it does, as I have said, is prescribe memorable rituals as records of decisions, but clearly the longevity of such communal memory was limited. Ripuarian Law has much more interest in the future. To some extent that is visible in the way it talks about the preservation of decisions in written documents, records which, potentially, can be kept forever. It also worries about extending the memory of public ritual into the future. If you do not record something in writing, bring small boys to the assembly and box them round the ears so they remember the day that a particular thing happened. Most specifically, there was no need for the ritual of chrenecruda. Where, in Salic Law, someone who could not pay a fine underwent the ritual of throwing earth over living relatives, rippling out to those related within three generations, the equivalent law of Lex Ribvaria simply says if you cannot pay, your family pays over three generations – your children and grandchildren keep paying until the sum is paid off: again we see that concern with the future, and no need for ritual. The sort of decision made via the ritual of PLS clause 60 was recorded by drawing up a written document, and – I assume – was an option only available to people of a certain status.


I have used this odd little clause from sixth-century Frankish law to try and make, or reinforce a number of basic points. Above all, kinship ties are socially constructed; where social norms do not regard relationships as kin relationships, regardless of biological relationship, kinship does not exist; the converse is also true: social norms might regard kinship very much as existing between people who have no genetic connection. In turn, people might have stronger affective ties with people with whom they have no kinship ties of any sort. More importantly, those relationships, ties and obligations are not immanent; they do not exist automatically; people are not automatically obliged to act in predictable, prescribed fashions simply on the basis of the existence of a relationship as, what Bourdieu called ‘official kin’. This view of early medieval people as keyed into extended kinship webs which dominated their everyday lives via systems of familial honour and shame, rather like north-western European views of Mediterranean (or, conversely, ‘Celtic’) societies, is one that ricochets between poles of condescension and romanticism. Clearly, outside the ‘effective’ or practical kin of the immediate nuclear family, early medieval people frequently did choose their family; contingently deciding when, and when not, to make their implication in a particular kinship relationship part of their identity and use it to justify their actions. It seems to me however to be going too far in the opposite direction to see early medieval actors simply as ‘rational actors’, driven purely by dispassionate calculation. People killed and, in the case of Duke Berthefried, who could – possibly – have lived, died for the sake of affective bonds, whether or not based upon kinship. Both of the extreme views make early medieval people into programable two-dimensional cut-outs. The people of the post-imperial centuries in western Europe, as throughout the globe throughout history, were much more unpredictable and interesting than that.

Monday 22 January 2024

Spectres of Marcus: the Roman Empire ‘between two deaths’

This time, ah - ah

Is coming like a ghost time

When I wrote Barbarian Migrations and the Roman West, getting on for 20 years ago, I used a three-part organisation of the text: Part 1: Romans and Barbarians in an Imperial world; Part 2: A world renegotiated; Part 3: Romans and barbarians in a post-imperial world. I used the term ‘post-imperial’ for a couple of reasons. I borrowed it from Andrew Gillet, who had coined it because of problems with the term ‘post-Roman’. People after 476 weren’t in any sense ‘post-Roman’. Many thought they were still Romans; many were still trying to do things to look Roman; many continued to call themselves Romans (not least in the Greek-speaking east). I also thought at the time that the western Empire had ended in 480 and that people at the time knew that it had ended. I thought this partly because of arguments by Jill Harries about Sidonius Apollinaris’ letters at the time, and on the basis of some of the changes that were taking place in material culture in the last quarter of the 5th century. So ‘post-imperial’ seemed like a very good term to use for the period after 480.

I no longer think that the western Empire ended in 480, and I no longer think that people at the time thought that it had ended. About ten years ago I argued that Sidonius’ comments about the Tiber’s dwindling stream compared with the strength of the Moselle, were something that a Gallic poet might equally have written to a third-century Gallic emperor resident at Trier. The work I did on Style 1, about a decade ago (but never published) also stressed the ‘not knowing’, the indeterminacy of the period and the way that the decorative style played with traditional Roman iconography and indeed could not make its point without that.

This leaves me with a terminological problem. Neither ‘post-Roman’ nor ‘post-imperial’ now seem to me to be adequate terms for the period between Nepos’ murder and the wars of Justinian – during which I contend people did begin to realise that they were no longer living in the Roman Empire. So what do I/we call that period? If (or, as I hope, when) I do a 2nd edition of Barbarian Migrations, what do I use as the title for Part 3?

During that period I argue that western European polities (and politics) continued to operate as though the Empire did still exist and they were a part of it. By the earlier 6th century Frankish and Ostrogothic rulers displayed serious imperial pretensions. Indeed it was this that seems to have led the eastern emperors to start to promulgate the idea that the West had ended, been lost, been conquered by barbarians, in the fifth century (in 455 or 476). Politics were oriented towards the notion of an Empire, but an Empire that did not function as such in the west. It was something spectral, or phantasmic: something believed to be there and affecting people’s actions, but not there in reality. It’s in this sense that I like to think of the period between 480 and say 550 as the western Roman Empire ‘between two deaths’. In one sense, as a functioning political organisation, the Empire really did die with Nepos in 480, but in another sense it only died when Justinian declared that it was dead, 50 years later.  When I was interested in the thought of Jacques Lacan, ten to fifteen years ago, I toyed with the idea of trying to think this issue through his writing about Antigone ‘between two deaths’. In – in Lacan’s terms – the ‘symbolic’ register (loosely, the register of language and the world as it is) the Empire died in 480; in his register of the Imaginary (the world as we think it should, or ought to, be) it didn’t die until Justinian’s wars. Now, away from my books, I can’t remember whether the two deaths at stake in Antigone are in the same sequence, first in the Symbolic and then in the Imaginary, but I think there are issues to think about there, in analysing the society and politics of that period.

For now, what I want to stress is that spectral aspect of the period (again, see also my unpublished piece on Style 1 for the 'undead' Roman Empire). So I give you (and, provided no one has come up with this before, claim as my coinage) a new term for the period between c.475 and c.550:


I thank you.

Thursday 30 November 2023

Reflections on the End of Western Antiquity: 4. The supposed ‘Rupture’ of the Ancient Mediterranean , Part 4

Several problems are raised by the economic/political paradigm. As indicated last time, my aim here is not to replace them but to add a new level more concerned with ideas, attitudes or culture. To this end it’s important to note that the broad outlines of economic development sketched in Part 3 match, in general, those of the shifts in culture mapped out in Part 2, of the gradual turning away from each other of west and east (especially of east).

One difficulty is the lack of fit between the political change and the archaeologically-revealed patterns of exchange. The alleged end of the western ‘tax spine’, after the Vandal conquest of Africa, does not seem to me to fit the data very well. One issue is what exactly the Vandal conquest of Byzacena/Proconsularis and the 442 treaty with the imperial court actually meant. This is really for another time, but it seems to me that the treaty is generally reconstructed in terms of a ‘barbarian invaders/conquests’ paradigm that relies on a lot of assumptions that are not necessarily supported by the evidence itself. More importantly for current purposes, though, the patterns of African trade don’t really change after c.440. As we have seen, African traders if anything found new markets after that date and continued to trade with Italy and the south of Gaul as before, even if quantities might have declined. These issues also affect the argument that claims that the end of the western imperial command economy produced a dramatic shift in interregional connections. Logically, this ought to have happened but it does not seem to me to be readily visible in the archaeological data, in which pre-existing patterns continue, even if along (similarly pre-existing) trajectories of reduction in scale. You might not find much ARS or many African amphorae in northern Gaul in the fifth century but then you wouldn’t in the fourth either. Most of the northern Gaulish landscape was harnessed to the provisioning and remuneration of the Rhine army and it seems likely to me that, as James Harland has suggested, Britain might have also played a major role in the supply of the frontier garrisons. Consequently, the breakdown of regular government and the end of the imperial command economy is seen in the well-attested crisis in those areas, not in the patterns of Mediterranean trade.

We might also ask whether the reimposition of imperial rule is in itself sufficient to explain the startling shift in the direction of African trade in the sixth century. This seems superficially attractive as an explanation but closer reflection raises some important problems. One might, first of all, ask why the government would redirect the African annonae to Constantinople, which was amply supplied from Egypt and elsewhere already. Would/should we not see a decline in the market share of Egyptian exports if that were the case? More importantly, why would the imperial government not have reestablished the supposedly-ended (and supposedly crucial) ‘tax-spine’ to the newly reconquered city of Rome, and Byzantine Italian territories? African exports continue to reach Italy, albeit mostly (and increasingly to the detriment of other areas) Rome and in gradually decreasing numbers, until the end of the seventh century.

We might also think more closely about the ways in which wars affect long-distance exchange. Pirenne and others ascribed the breaking of long-distance Mediterranean trade to seaborne piracy, whether Arab or Vandal, but this is an implausible mechanism. Even if one were to assume that no one ever thought to provide merchant vessels with armed escorts, the relationship between piracy and seaborne commerce resembles that – in that early computer programming exercise – between foxes and chickens. If the chicken population [or seaborne merchant traffic] grows, the fox population [or profitable pirate activity] grows with it, because of the increased availability of food [loot]; there comes a point though where the fox population is so great that it is killing chickens faster than the chicken population can reproduce itself; at that point the lack of food leads to a decline, through starvation, in the fox population; eventually, however, the low numbers of predatory foxes allows the chicken population to grow again; and so on. Put more simply, you can’t have pirates without merchant ships for them to prey on. Pirenne’s thesis creates an image of a period during which Arab pirates wearily put to sea in spite of there being no shipping to attack. ‘Well, me hearties, none of us has seen a merchantman for ten years but here we go again; another day another dollar.’ The other strange point about Pirenne’s thesis is that although he thought – correctly – that long-distance trade flowed mostly from east to west, and although he thought this on the basis of texts describing imports from Palestine and Egypt, he still thought that the decisive blow to Mediterranean trade was dealt by the Arab conquests and/or by Arab fleets – creating the image of Arab pirates attacking ships that had sailed from their own ports [‘pirates on the starboard bow, cap’n!’ ‘Don’t worry lad, their ours.’]…

Be all that as it may, we might wonder how warfare affects trade. The key point I want to make is that for it to have seriously detrimental effects it needs to be ongoing for a long time. Even heavy fighting for a year or two, or a few years, is unlikely to bring about any serious shifts in trading patterns. Such events can be factored in to the usual response strategies. Campaigning by even small armies (as were the norm in late antiquity and the early Middle Ages) could produce famine but such could be responded to within the normal patterns of redistribution and relief, where they functioned. They might even produce windfalls for merchants able to sell at famine prices. More serious contenders as causal factors in long-term change are long periods of warfare such as the Gothic wars in Italy, or the wars with the Moors after the reconquest of Africa from the Vandals, or, especially, the long war between the Empire and Persia, and then with the Arabs. Such warfare is important not (or not just, or principally) for disrupting commerce, though it probably did so (though we might also remember that armies need provisioning and can act as markets) but through the disruption of production/supply and demand. Displacement of population or disruption of seasonal agricultural activities through the movement (year in, year out) of armies, associated famine and disease (plague especially from the mid-sixth century) dislocates or prevents the production of the surplus that in some way or other acts as the basis for commerce, as well as (partly in consequence) the ability for craftsmen to make the products that are traded, or the ability for markets to be held regularly with the usual security. It might be that twenty years of insecurity after Belisarius’ conquest had serious effects on some African production, but clearly it didn’t kill it off entirely. Similarly, the Gothic wars didn’t entirely end the market for African products in Italy. The wars in the east from 610 to mid-century do seem to have been fatal for the old eastern Mediterranean complex of commercial networks, production, and distribution, etc. Once again, though, we’re entitled to ask why, when the dust settled, they didn’t reemerge even if the materials being produced took different forms. Certainly, the creation of the Caliphate led to a profound rearrangement of trading networks – greater links between North Africa and the sub-Saharan world being one of the most important developments – but again, one would have to ask whether there was anything economically more logical or natural about those networks than the previously existing ones. Any investigation of that issue will lead you quite quickly to the role of attitudes, ideas, mentalité in shaping networks and connections, commercial and otherwise.

Some of the economic explanations adduced (concerning supply and demand, production costs, etc), while plausible and sophisticated, function mainly as descriptive extrapolations from data. They don’t necessarily emerge from the data themselves. Clearly, too, they are teleological extrapolations backwards from the eventual outcomes. All of this is fair enough, but the issue does need to be flagged up. And again there are some problems that arise, in my opinion anyway. One concerns the demise of the Mediterranean commercial links with the north-west. Some things from the Mediterranean were simply not available locally – olive oil is the obvious commodity here. Given the dominance of African oil production in the late imperial west, this ought to have been an important basis for commerce (and surely was). But what markets were there for this? Western towns generally contracted in size from the third century onwards but Mediterranean wares continue to be found in some north-western regions. The church’s need for oil for lighting and also for anointing, and the growth of Christianity may have provided new markets as, in the fifth century, might the new elite in western Britain wanting to show its prestige and Roman-ness. Alongside the seeming demise of the old Roman elite further to the east, this might explain the shift in the routes via which African exports entered Britain. What seems odd to me, though, is that when the economy in the north-west recovered in the late sixth century (with the revival of towns, greater aristocratic control of surplus, related increase in church foundation, increased monetisation of the economy, growing craft-specialisation, newly established markets, etc) the market share of Mediterranean products, having managed to persist through over 200 years of economic change and (especially) decline, suddenly disappeared. Imports found on sites all around the Irish Sea (east and west, north and south, rather than principally on its southern and western shores, as hitherto) seem to come almost entirely from various parts of Gaul. Descriptively one can say that more local products squeezed out the African imports but one would still need to ask why. Costs, etc, don’t seem to me to present the whole story. The whole point of my discussion thus far is, on the one hand, that normative, supposedly natural, logical, or eternal economic laws of supply and demand are in themselves inadequate to the task of explaining the changes of the fifth to seventh centuries. But so too is the notion that trade can just be ‘turned off’, like a tap, by high-level political events.

It's time, I think, to try to add some elements of ideas and worldviews to the picture to round it out (rather than replace other explanations with them). I have already said that the development of western economic systems runs to some extent parallel to some changes in outlooks and attitudes (though not necessarily to high politics or governmental shifts). This is not in itself news. Long ago it was pointed out (maybe by R.S. Lopez – I will check and correct if necessary when I have my books) that the papacy was quite happy to import Egyptian papyrus until the rulers of Egypt started stamping papyri with (to Christians) unacceptable quotations from the Qur’an. The shifts in trading routes and networks after the creation of the Caliphate are also clearly about something other than laws of economics. The demand for Mediterranean goods in western Britain in the fifth and sixth centuries was surely in large part because of the cultural cachet of commodities produced in, or somehow associated with, the Christian, Imperial, Roman World. As has been mentioned this period represents the high point of Romanitas in the region. Were easterners simply no longer interested in western trade and markets? It’s not as if there were no important commodities to be had there, and – as noted – the market opportunities ought, if anything, to have been improving. Three letters of Gregory the Great (and I am grateful to Helen Foxhall Forbes for drawing my attention to them) are significant here. Gregory is writing to the Patriarch of Alexandria who, he has heard, needs large timbers for Alexandrian vessels. Gregory could, and did, get his hands on just the sort of timber that’s wanted (in Calabria if memory serves), and get it transported up to Rome, but the Egyptians didn’t send ships that were big enough to take them, and cutting them to fit would, obviously, rather defeat the object. Gregory, it seems, is quite keen to help and to provide this timber, which obviously isn’t available in Egypt, but the Patriarch, it appears, can’t even be bothered to write back.

Thus, I would propose that an important reason (not, for clarity the reason, or the most important) for the end of any significant market share for Mediterranean commodities is that final cultural turning away of east and west after the mid-sixth century. This is a conclusion that has some important bearings on the debate upon pre-modern economies, between ‘formalist’ and ‘substantivist’ positions.

Reflections on the End of Western Antiquity: 3. The supposed ‘Rupture’ of the Ancient Mediterranean, Part 3

In the third part of these reflections we finally enter the territory of the Pirenne Thesis, and indeed of my project: the changes of the late sixth and early seventh centuries.

Let’s recap. Around 530, in spite of all the developments discussed last time, in spite of the socio-cultural dislocation that had been going on since the third century, and in spite of the dramatic events of the fifth, western Europe still thought of itself as part of the Roman Empire. Trade and exchange still united the Mediterranean and some areas beyond. Commercial networks reached round the western coasts of the Iberian Peninsula, and Gaul, as far as the eastern shores (mainly) of the Irish Sea. These contacts made the fifth and sixth centuries the period when the Romanitas of the western highlands of Britain was most strongly asserted. An inscribed stone in North Wales shows that those inhabitants of the region who were interested or cared were still aware of who the current consul was.

Obviously, there had been cultural changes. By this date, in most regions of the western Empire, villas had ceased to exist as such. Did this mean an end of Romanitas? Clearly not. The idea that the end of the villas reflected a conscious decision to reject Rome was one of the sillier and most insular ideas (of many) to permeate British archaeology in the ‘90s and ‘00s. Only a few areas still had any significant number of villas (occupied recognisably as villas) by the second quarter of the sixth century (notably the south-west of Gaul, though a revival might have been under way in Italy) but there was heavy investment (including on the site of former villas) in Christianity and in other, new expressions of Roman-ness. As noted, westerners still thought they were living in the Roman Empire. Further, a martial model of masculinity had emerged in the late imperial period which, although ostentatiously setting itself up as rejecting traditional aspects of civic masculinity, was still very much a manifestation of a Roman identity and relied on the existence of traditional ideas for its socio-political cachet. At the start of the sixth century, both Theoderic the Ostrogoth and Clovis the Frank allowed themselves to be addressed as augustus. This turned out to be crucial, as we’ll see.

Let’s pause here, though to look again at the turning away of the east from the west. Jeroen Wijnendaele recently reminded me of the point made by R. Blockley that even in the fifth century, Eastern Roman writers had started to refer to the inhabitants of the Pars Occidentalis as ‘Italians’ or ‘westerners’, while referring to themselves as Romans. Around the middle of the sixth century, Cosmas Indicopleustes wrote his Christian Topography. This, mostly, is concerned with his voyages, early in the century, around the Red Sea and the Indian Ocean, with famously valuable accounts of various regions in and around that sea (whether or not he had actually visited them). What I find interesting, though, is that, although resident in Alexandria and writing about the nature of the world, Cosmas shows pretty much no interest in the western Mediterranean. He knows it exists and goes as far as Cádiz (not actually on the Mediterranean…) but that is about it. One might of course argue that Cosmas’ own experience simply hadn’t taken him that way, and that he was writing a Christian topography based, obviously on the holy land, but that wouldn’t entirely negate my point. After all the Holy Apostles (Peter and Paul) had headed westwards. At around the same time, Procopius wrote his history of Justinian’s wars, the bulk of which concerns the western Mediterranean. What strikes me about Procopius’ work is just how remarkably badly informed about and/or interested in the west he is. His ignorance about the geography of Gaul makes this abundantly clear. Even educated easterners who had been in Italy knew or cared very little about regions to the west. One might argue that Procopius, unlike, say, Ammianus, never went further than Italy but my point is hardly altered by that fact. I don’t think Ammianus ever went to Britain but the information he acquired enabled him to write fairly reliable accounts of what happened there. Procopius’ comments about Brittia (History of the Wars 8.20), by contrast, resemble traditional Graeco-Roman ethnography about the outermost regions of the world rather than things that relate to a former imperial province. Even he thought it sounded more like the sort of thing that you might dream and, for this reason, he seems to have decided that they must relate to a different island completely (Brittia, rather than Britannia[[1]]). Procopius’ ignorance or lack of interest is part of a trend that is only amplified in the Byzantine historians that follow him. While western writers (like Fredegar for example) show an interest in news from the east, however garbled it had become, there seems to be no reciprocation in historical works written in the eastern empire.

Unsurprisingly, if you have been following my thoughts on this stuff over the years, the crucial change seems to have been the new ideology that emerged in Constantinople in around the 520s, usually associated with Justinian (even if it appears earliest under his uncle and predecessor Justin I). Possibly building – to a greater extent than I had realised (see Blockley’s point about fifth-century vocabulary, above) – on attitudes that had developed during the fifth-century, Justin or Justinian added the new – and significantly different – point that the west was no longer even a part of the western Empire, and (by the 530s) needed to be ‘reconquered’ by the Roman Empire. As I have repeatedly argued, this cut away the bases of almost every traditional idea about legitimate power or authority and caused people in the west to try to find new bases of authority. As I see it, this put the cap on over a century of shifting attitudes. Especially from Justinian’s reign, the West became lands of barbarism and heresy, or at least of insufficiently rigorous orthodox religious thinking. It’s a common mistake, however, to think that the West somehow looked up to the eastern Empire in imitatio imperii.[2] After Justinian, in the late sixth to eighth centuries at least, this seems fundamentally mistaken. The west certainly reciprocated the east’s view of the other as sullied by heresy. In Spain, Visigothic writers seem, if anything, to have adopted an attitude of translatio [rather than imitatio] imperii (see Jamie Wood’s work on this, in particular): a diametrically different attitude. Chilperic of Neustria seems to have had a similar idea (Gregory, Histories 6.1). Otherwise, and probably most commonly, the touchstones used to justify various aspects of social organisation unsurprisingly moved away from Rome, especially towards the Old Testament, even if some of the general virtues etc remained the same. By the early seventh century the old trading patterns between the Mediterranean and the Irish Sea had died out, to be replaced by new ones connecting the European Mainland, whether the Rhineland or the Bay of Biscay, with lowland Britain as well as the western highlands, now both shores of the Irish Sea, and the Scottish/Pictish north. Major changes were also under way in the Mediterranean itself. We’ll return to this.

What conclusions can we draw about the end of Mediterranean unity by about 600? (Here there will be a little repetition from the last article.) It seems to me to be important to note that, viewed from c.600, it looks like the culmination of a process that, in social, political, and economic terms, had begun as early as the 3rd century, had picked up pace in the 4th, and taken on new cultural aspects in the 5th. Justinian’s ideology and – especially – his wars can be said to have brought this process to an end. As I said last time, however, it’s important to tread carefully here. On the one hand, I think it’s probably a mistake to see this development as ‘natural’ or ‘inevitable’ and to cite the point that, in long-term perspective, the period of Mediterranean unity is rather more unusual than periods when the east and west form generally separate spheres. Similarly, Chris Wickham made the very good point that long-distance trade round the Mediterranean shouldn’t be assumed to be natural, especially given that most regions around the Mediterranean produce the same principal tradable commodities (grain, oil, wine: the ‘Mediterranean trilogy’). While the difficulties posed to communications or travel by physical geography shouldn’t be underestimated, the idea that, somehow inevitably, they would eventually undermine the features that had brought about unity seems to me to be too crude. After all, the aspects of physical geography that facilitated ‘natural’ connectivity and communication (the sea, rivers, etc) remained just as much as the ones that presented ‘natural’ barriers to the same (mountains – or indeed rivers and the sea…).

Second, and again this is a point that will surprise no one who’s followed my thinking over the past 15 years or so, looking back from c.600 gives a misleading sense of unidirectionality and teleology to the series of events and their final outcome. All of the events I have talked about could have been reversed; none was the only possible response to the situation pertaining at the time; some of the political changes – as noted in the last part of these thoughts – had beneficial aspects as well as, at least with hindsight, negative ones in terms of Mediterranean unity; many had effects, negative or positive, that were not deliberately intended. The narrative arc mentioned at the start of the previous paragraph can be deconstructed (in the proper sense) at every turn.[3]

The main point I want to stress, though this will really be discussed in Part 4, is the importance of ideas, attitudes, and political culture. With that in mind, I return to the core of the Pirenne Thesis: economics. He generally rated Mediterranean unity according to the continuity of the trading patterns that existed in the Roman period and so thought that when (as he thought) Arab conquests and seaborne raids and piracy killed off east-to-west trade that ruptured the unity of the Middle Sea.

Decades of sophisticated study of forms of evidence that weren’t available to Pirenne (principally archaeological evidence of various types) and of more refined study of the evidence that was available to him, has modified some important aspects of his ‘thesis’ while leaving others broadly in place. As we saw in Part 1 of these thoughts, the north-west of Europe formed an economic sphere largely separate from the Roman Mediterranean by the fourth century; on the other hand, the final demise of the eastern Mediterranean economic sphere appears to take place at about the time of the Arab conquests.

A brief summary of my understanding of recent/ish thinking on this might be helpful. There are very good discussions of the problems of this evidence in Simon Loseby’s chapter on the Mediterranean economy in The New Cambridge Medieval History Vol.2 (ed. Fouracre) and in the relevant chapter of Chris Wickham’s The Framing of the Early Middle Ages. One point I would emphasise though concerns the implications of distribution maps. We might find an African amphora in Cornwall, or some African tableware in Marseille. The implication of those places in networks is complex, however. Roman law envisaged that a ship might be away from its home port for at least two years (it enacted that it had to come home within two years), plying different routes perhaps in different ways with different forms of cargo. Material might move from A to B via several shorter hops involving different people rather than just via one long-haul voyage from a place close to the material’s production to one near its consumption. Commodities might be being moved in ships that have little connection to the place where those commodities were produced. It’s possible then that our distributions give misleading impressions of connectivity and mask the nature of the networks, and possibly conceal even more change than they reveal. For example, the general distribution of African finewares between modern Tunisia, Italy, and the south of France might look broadly similar between the fifth and seventh centuries, even if declining in absolute numbers. Theoretically, however, it’s possible that the fifth-century African pottery in Provence came more or less directly from Carthage, in ships that stopped off at various points on the way, whereas sixth-century African wares came via various entrepôts, ultimately being delivered, in short-haul hops, in the ships of Italian merchants. Or vice versa.

There had always been a difference between the eastern and western halves of the Mediterranean (going back to the Bronze Age at least). The Aegean is a sea of many islands facilitating a dense web of routes, while Crete and Cyprus make handy staging posts for sailors travelling in several directions. That said, in the late Roman period, the different regions formed fairly distinct economic regions in many regards. A ‘tax-spine’ connected Egypt and Constantinople. In the West, as intimated in Part 2, the west had also fragmented into a number of economic regions, often quite large. This fragmentation was related to the (generally cultural or socio-political) issues that had eroded the unusual unity of the Mediterranean world by the third century. The Carthage-to-Rome ‘tax spine’ mentioned earlier, certainly dominated exchange between Africa and Italy, the far south of Gaul and the west of the Iberian Peninsula, but it seems to have had rather less effect on other – quite large – regions like central Hispania or the north of Gaul. There were nevertheless general similarities between the two halves of the Mediterranean world. The fundamental dominance of regional over interregional networks is similar in both zones. The Egypt-to-Constantinople ‘tax spine’ shows some similarities to the western Carthage-to-Rome spine, though it has been suggested that it was less dominant (I think that this begs a number of questions). Nonetheless, long-distance commerce around the Mediterranean continued. African Red Slip (ARS) was found (in quantities that varied over time) in the east, and Phocaean Red Slip (PRS) in the west. Other, less archaeologically-visible commodities were also traded from east to west.

Differences seem to be more pronounced in the fifth and sixth centuries. The east continued to be prosperous and the various links between the regions remained. The connections between western regions decline further. The extent of African commerce declines, though it continues to be the most important axis of interregional commerce. On the other hand, as has repeatedly been pointed out, African wares continue to reach the west of Britain, possibly by more direct routes than before. These patterns continue into the sixth century, but after the Eastern Roman reconquest there is a revival of connections between Africa and the East. Trade nevertheless continues between Africa and Italy, Provence and the regions further afield. What at least seems to me to be under-appreciated is the preponderance, within the sphere of African exports, of the imperial territories (not just in the east but in Sicily, Rome, and other Italian regions). Stories suggest that in the late sixth century, the west of Britain was still known in the east as a source of tin. Some luxury products from the east continued to reach Gaul, via Marseille (whether via intermediaries in Carthage and elsewhere is unknown).

All this changes in the seventh century. Some of the change happens pretty quickly. The connections between the Mediterranean and the Irish Sea generally fizzled out between the late sixth and early seventh century. Analyses by Patrick Périn and Thomas Calligaro show that the supply of good quality Sri Lankan garnets to the west also ended around 600. The demise of the prosperous eastern Mediterranean regions and their interconnectivity collapsed in a couple of generations in the mid-seventh century. African commerce to the west struggled on to the end of the seventh century but not beyond. It cannot be claimed that east-west commerce ended in absolute terms by 700 but it was certainly only a shadow of its former self after that, reduced beyond recognition in terms of quantity and the range of goods traded.

Explanations for these changes have understandably concentrated upon issues such as production, supply and demand, and on political change (the Vandals allegedly ending the Carthage-Rome ‘tax spine’; the end of the imperial command economy; the Arab conquests). These have been the result of close analysis and I certainly don’t intend to dismiss them. What I want to ask, though, is whether they are the whole story. There are a number of problems that arise in only looking at this issue either in economic terms or in those of ‘high political’ events. I will look more at these in Part 4.

[1] I open EoWA vol.1 (The Fates of the Late Antique State) with a discussion of this.

[2] I was once at a conference where a non-specialist asked a famous Byzantinist about the relationship between Byzantine culture and the west, and the response was that Byzantine culture was ‘a dominant culture’. This view seems widespread even among some specialists on western European history, but it really lacks any substantive empirical support, other than in the sense that that was indeed what Byzantines thought. In this case I think it was one of those instances where, in the same way that people say that people come to resemble their dogs, historians come to resemble their subjects.

[3] In the last part I promised that I would speculate on whether the circumstances that had brought about Mediterranean unity in the later Republican period could have been reproduced. I am going to break that promise. Maybe at some point I will write some ideas about that but not now. It’s not quite, but it verges on, ‘what if’ history, which I think is mostly ahistorical and intellectually no more than an entertaining parlour game.

Monday 16 October 2023

Reflections on the End of Western Antiquity: 2. The supposed ‘Rupture’ of the Ancient Mediterranean, Part 2.

In the previous post I was arguing, ultimately, that explaining ‘the end of Mediterranean unity’ is not a question of finding an ‘event’ that ruptured Mediterranean unity (the Arab conquests, Vandal Piracy, etc) as much as looking at why the features that had held it together earlier – and which had overcome those features that might militate against unity – came to an end. This post muses rather meanderingly on that issue.

Of course, it might be the case that some decisive event killed off the features that had unified the Mediterranean but there are two points that emerge from that possibility. One is that it seriously recasts the question, and the other is that what we might call structural features do not tend to be killed off by single dramatic events unless they’re already dying. As an example, look at towns ravaged by earthquakes, sacks by enemy armies, great fires, or plagues, but which continued to survive as successful urban centres.

How the Roman World came together

Let’s look first, briefly, at the features that held the Mediterranean World (and indeed the empire, loosely defined, as a whole) together in the earlier Roman period, and then at how these features came to an end. Above all, though the early Roman polity was created by conquest, it was held together by the desire of local communities to be part of the Roman world. This, as far as I can tell, not being a specialist in either Republican or early imperial history, worked differently, in detail at least, in the different parts of the empire (I am going to use that term, all in lower case, to cover the Republican as well as the imperial period). A point often forgotten, at least by us non-specialists in earlier Roman history, is that Rome conquered most of the eastern Mediterranean before it conquered the West. Roman military intervention in Greece began in the last decades of the third century (at the height of the second Punic War) and Greece was effectively conquered when the Romans sacked Corinth in 146 BC (the same year as the destruction of Carthage). By then, Rome controlled much of North Africa and the eastern half of Spain. Some of the Mediterranean coast of Gaul had taken place in 121BC but by the time Caesar began the conquest of further Gaul in 58 BC pretty much all of the Eastern Mediterranean – Macedonia, Greece, Asia Minor, the Levant, Tripolitania – had either been annexed or made into tributary states. The conquest of Marseille didn’t take place until 49 BC, the final conquest of Spain took place after the conquest of Egypt, and of course that of Britain even later.  Even the conquest of northern Italy occurred after Roman claims to hegemony over Greece had been laid down.

There are several key points that emerge from this. Possibly the most important is that in east and west (albeit in different ways) close cultural ties preceded military conquest. Though not a Greek colony, Rome was already a part of the Hellenistic world by the third century. Many of its rivals for domination in Italy were Greek colonies and the Republic had to fight and win a tough war against Pyrrhus, the king of Epirus, a cousin of Alexander the Great. Additionally, Rome bought into the Greek discourse about ‘barbarians’ in its political claims for domination (see, e.g., Emma Dench’s From Barbarians to New Men[1]). All this meant – and this, it seems to me, is a very important point – that Rome looked eastwards rather than westwards. In some ways the Republic was drawn into military action in the eastern Mediterranean and that leads to a third point, which is that Rome exploited regional rivalries to play contenders off against each other. It did this everywhere and even in the late Empire it remained a key strategy beyond the frontiers.

That brings us to the issue of military dominance. I don’t want to dwell too long on this but once upon a time, in military historical circles, there was long discussion about the somehow inherent supremacy of the Roman legionary ‘system’ over the Macedonian/Hellenistic phalanx. This sort of discussion rapidly leaves the realm of historical argument (and indeed, in my view, that of history full stop) and enters that of hypotheticals and counter-factuals – ah, but what if the Macedonians had had a general as good as Alexander? What if this or that factor had not applied? Yeah, what if…? The significant point is that for some reason or other, the Romans do seem to have had a long run of military success against the Hellenistic states (although of course it's worth remembering that the evidence we have is hardly even-handed). The simple fact of being an army that fought regularly and usually won very likely (in my view) had an incalculable effect upon the confidence, morale, and fighting spirit of veteran Roman troops, while repeated defeats possibly had an equal and opposite effect on their enemies. This would be the case regardless of the ‘tactical system’ being used. Certainly, above all, it increased the attraction of allying with, or subscribing to the protection of, Rome. This was the case, a fortiori, in the west, where Roman armies must have outnumbered, ‘out-armoured’, and ‘out-equipped’ their ‘barbarian’ enemies, in addition to having better logistics, heavy siege weapons and so on (Roman accounts of Gallic or Germanic armies numbering many tens of thousands are simply incredible); this fact needs to be internalised when thinking about Roman wars, and indeed the quality of the Roman army, in the West (after all, sometimes they lost…). The ‘bottom line’ was that Roman military success made Roman support or protection worth having and that meant that some communities turned to Rome and drew it further into local rivalries.

In the West, communities in Gaul, Britain and Germania were already linked into Roman cultural orbit before they were conquered. Objects from the Roman world were deployed to display status and prestige and drew people into Rome’s sphere of influence (see Greg Wolf’s Becoming Roman[2]). This continued after conquest when people within local communities competed for standing (after being demilitarised and having their more warlike elements hived off into the auxilia: see Ian Haynes’ work on this, especially: e.g. Blood of the Provinces) by displaying their ‘Roman-ness’ in new Roman-style towns, villas and so on, and above all by seeking status through involvement in local government. These features seem to have been far less significant in the East (where after all it was more a case of the Romans being drawn to Hellenistic culture, something topped up, in the late first and early second century especially, with the Roman attraction to the Greek culture of the ‘second sophistic’) but they were not absent. A few towns even built amphitheatres... Competition between  communities, played out by striving for the advantages of particular legal status, within the Roman system, remained an important element of local or regional politics even into the late imperial period, and even beyond.

The features sketched out created an exceptional situation, as mentioned, where the west and north-west were drawn into a Mediterranean world and where that world was itself unified by constant reference, in local and regional politics, to Rome and its rulers. Eventually I will come back to the issue of whether this situation was repeatable. For now, let’s examine what happened when these circumstances no longer pertained.

The fracturing of the early Roman world

By the third century, if not slightly earlier, many of the factors that had led to unity no longer applied. The products that, in the West at least, had been used to signal participation in the Roman world were by then mostly being manufactured regionally rather than being imported from Mediterranean centres. Economically the western half of the empire went back to being a series of largely independent regional economic networks. Possibly more importantly, the political advantages gained by involvement in, and financial expenditure on, local government, monumental works and so on, were generally no longer brought by this sort of activity. Parts of the west had been over-urbanised in the rush to become Roman. A retreat from this high-point followed. When the Antonine Constitution made all free-born inhabitants citizens, Roman citizenship was no longer something to be competed for. In this situation, in some ways the ‘crisis’ of the third century was always going to be on the cards. However one adds nuance to old views of the ‘third-century crisis’ (it wasn’t that bad everywhere, and not at all in some regions; it didn’t last as long, or occur at the same time everywhere, etc), this was a serious moment. The Palmyrene and Gallic Empires showed that the notion that there might be multiple ‘Roman Empires’ was not seen as entirely alien. With a few different conjunctures the Empire could have fragmented in the third century. One feature that helped ensure that this didn’t happen, as I suggested in Barbarian Migrations and the Roman West, was the continuing hegemony of the notion of Roman ‘civic masculinity’. There were as yet no real alternatives to that in establishing legitimate power at a regional, local, or even familial level. What this meant was a continued relevance of some of the aspects of ‘being Roman’ that had brought the Roman world together. This was a crucial resource and a glue that still held that world together. Again, note that the crucial issue here is one of mentalité.

Responses and problems

As it happened, of course, the Empire did not fall apart and responded effectively to the changes that had threatened fragmentation. Obviously, much of this response was pragmatic and piecemeal and effected over a long time, rather than being the result of the imperial rulers sitting down with their advisors one day and formulating a coherent set of policies. Nonetheless, the Empire as it emerged at the end of the Tetrarchy was a very different place from that which had existed 100 or 150 years earlier.

Various responses, changes, and developments – administrative reform; the emergence of a new civil bureaucracy; the separation of civil and military branches of service; new forms of aristocracy and rewards for service; new capitals; moving the emperor to the frontier – all produced an Empire that was as strong as it had been in the second century and remained so, in the West, for a century (and longer in the East). All these developments, though, had corresponding weaknesses. The picture that follows is broad-brush and (over-)simplified, as well as almost certainly needing updating, but it and – more importantly – the issues it raises still seem to me to be generally valid, in outline at any rate.

The foundation of Constantinople created a focus for the eastern provinces (pinning the Balkans to and Greece to Asia Minor and the Levant, etc) and thus increased the coherence of the East (and continued to do so in some ways well beyond the Arab conquests and into the middle Byzantine period; I am thinking here of John Haldon’s argument that the seventh-century Empire functioned effectively as one huge city state[3]) it created a separate, alternative, eastern pole of attraction. The fact that it was a new foundation had important advantages but it also meant a crucial reorganization of fiscal resources. The Egyptian grain fleet was now diverted to the Bosphorus. As I see it, this made for a more significant rearrangement of existing economic ties and networks than would perhaps have been the case had the emperors decided upon, say, Antioch as the central point of the East.

Moving the western capital to frontier cities, above all Trier, was pragmatically a very effective move. It bound Gaul, most importantly, but also other frontier regions tightly into the imperial state. Older nobilities had to travel to the north to compete for imperial patronage in order to maintain their traditional aristocratic culture of otium and negotium. At the same time, though, it seems to have created a certain resentment among those traditional elites, not used to being sidelined. By early in the fifth century the Gallic and Italian aristocracies largely formed separate networks and this, as I see it, became a crucial feature to be overcome in fifth-century politics. Gratian’s move of the capital back to Milan in 380 was also, in my view, crucial. Though, as I look at it now – 16 years on from Barbarian Migrations – it seems like a potentially imaginative response to the emerging problems of the fourth century western Empire,[4] as it turned out it precipitated crisis. It removed most of the Gallic and Germanic provinces, and Britain, from the close connection with the Emperor to which they had become accustomed; stress and usurpation soon followed.

With the emperors in the west hardly ever resident in Rome, the Empire now had two very separate political centres or foci for political activity. Indeed, the end of the de facto political (rather than ideological) centrality of Rome itself helped unpick the ties that had bound the eastern and western worlds together. The two halves of the Empire began to face in different directions. The social and cultural contacts between east and west began to reduce (which is, obviously, not to say that they ended or became insignificant). I also have an impression (rightly or wrongly) that, after the early fifth century, the direction of those links that remained very much tended to be west to east.

Similarly, though it was an effective response to the problems of the third century, the separation of civil and military branches of imperial service, led to the emergence of an alternative, martial or military model of Roman masculinity, one that stressed things that were antithetical to civic masculinity. This would turn out to provide a political resource for those outside the ambit of the legitimate imperial government in the fifth century: one that hadn’t existed in the third century. Another alternative was found in Christian models of masculinity, not least those stressing asceticism and renunciation. (I have a feeling that the disputes within Christianity also helped divide east and west.)

A key point is that although the new system worked well for a century in the west (and for longer in the east), it was, at least in the west, fundamentally fragile. It worked very well as long as there was an adult emperor able to command armies and manage the distribution and redistribution of patronage (offices etc) between the various interest groups within the Empire. There were numerous groups, especially regionally-focused ones, whose interests needed to be balanced. In Barbarian Migrations I appeared to think that this was a peculiarly late Roman weakness; clearly it wasn’t but the problem does still seem to me to have a distinctive flavour in the late period. Without an active, adult emperor, the focus of politics would turn inwards on the palace itself and efforts to maintain the governing faction’s position. The legitimate western Emperor was a child (or adolescent) for twenty years after 383 leading to internecine struggles for control of the palace and repeated usurpations.

That leads me to my next point. The West was riven by repeated civil wars between 383 and 425. The importance of this can’t be overstressed. The Romans had massive reserves of manpower, of course, but what was lost in these battles was the cutting edge of the Roman army: troops who could be replaced in quantity but not quality. The wars followed at such regularity, moreover, that there was hardly time for a new army to be built up and recover its effectiveness and esprit de corps before it was fighting other Roman armies again and suffering heavy losses even if it won. It was these wars, not the Great Invasion of 406 – which seems not to have involved any serious defeat of a Roman field army – that fatally weakened the Western Empire’s army, leading to the creation of new types of army, based around the groups of barbarian descent that were now within the imperial frontiers.

On the other hand, all of this wasn’t irreversible. After 425, the lesson learnt after 40 years of failed usurpations seemed to be that dynastic succession trumped everything else. For the next decade the western empire had a minor on the throne but the nature of politics changed away from attempted usurpation to struggles to control the court, which could potentially act as a cohesive force.[5]

Nevertheless, and unsurprisingly, when the Valentinianic/Theodosian dynasty came to an end with the assassination of the (like Honorius) possibly underestimated Valentinian III in 455 (a date later given significance by Marcellinus Comes as that of the end of the western Empire) the lack of such legitimacy proved fatal for all the different emperors and their backers. None could defeat the others decisively or otherwise persuade them to submit to their authority.

And yet … two things:

First, people in the west still thought they were part of the Roman world, indeed of the Roman Empire, beyond 480 and on into the 6th century. After 476, if Candidus the Isaurian is to be believed, embassies from Gaul still reached the emperor in Constantinople asking him to resolve western disputes. Western kings still based the legitimacy of their claims to rule on their Roman titles.

Second, through the period cultural connections remained. Traders still sailed the length and breadth of the Mediterranean after the end of the western imperial command economy, demonstrating that, as more recent work has argued, that the latter was not the only force to determine continued commercial and exchange connections. Indeed, as the increasing connections round into the eastern shores of the Irish Sea show, those who were involved in commerce could still adapt to changing circumstances. A key factor here might be the fact that those links became very important to western British leaders responding to the crisis of the fifth century. Like their predecessors, centuries earlier, it mattered to them to be connected to the Roman Mediterranean.

So – where (if anywhere) have we got to? A few key points:

1.       Cultural networks seem to me to be vitally important. Rome looked eastwards because of the cultural world it had become part of; the western expansion of Rome was very much driven by cultural relationships.

2.       The expansion of Roman power relied as much upon local and regional groups buying into Roman protection and or Roman culture as upon simple conquest.

3.       Ideas and culture remained crucial in maintaining the cohesion of the Roman world throughout the period discussed (from say 200 BC to 500 AD).

4.       Political history, economic history, and the history of culture, ideas and mentalité do not always run on parallel tracks. Events in the first do not always have effects in the others; changes in the other areas do not always have political consequences.

5.       Physical geography – seas, tides, currents, the direction of rivers, the location of mountain ranges, high plâteaux, forests, etc – do tend to bind or separate regions but, while extremely important we should not (pace, maybe, Halsall 2007) regard this as naturally, or automatically, determinant, and certainly not as insurmountable. Mountains can be barriers, but passes are links and thoroughfares; seas and rivers connect and divide. None of this is new. We should not assume that the cultural features that overcome certain aspects of physical geographical constraints cannot themselves come to be seen as just as ‘natural’. After all, why would Rome, a city state on the western coast of Italy, look east, especially given the difficulties of navigation between Rome and the eastern Mediterranean?

6.       We might then, equally, suppose that when physical geography does (as I put it – and I am now wondering whether this wasn’t considerably oversimplistic) ‘rear its ugly head’ and connections between regions weaken or end, this might be just as much a cultural response, a decision rather than the inevitable triumph of nature and geography over mentalité (spoiler alert: this will be crucial to my argument next time).

7.       Key events or developments are contingent upon the circumstances that created them. We ought not to see them as automatic, or that the results they had were those that the actors involved had in mind (the piecemeal imperial response to the failings of local government and their overall result might be a case in point).

8.       Nor do we have to assume that the strategies that were adopted, and which worked, or the features that tended in a particular direction, were (even in the case of strategies or policies followed consciously) automatically the best, or the only ones that could have had that outcome. As Roman history shows quite clearly, there are various ways in which the supposedly determinant features of physical geography could be and were overcome. There were always different paths that could have ended up being followed.

9.       Hand-in-hand with that, just as particular effects might not be the result of deliberate policies or strategies achieving their goals, and that even beneficial long-term results might not have been those actually intended (or conceived), none of the developments I have been considering seems to me to have been irreversible. [I think that none of the last three points will be surprising to anyone who has followed my work over the last 25 years at least.]

Having proposed all this, we are – I hope – now in a position to have another look at what did happen in the later sixth and early seventh century and possibly even to suggest a slightly different take on it.


[1] It occurs to me that several of the works I allude to in this piece are 20-30 years old. This is essentially because I haven’t thought much about the issues they discuss for over a decade. That said, they’re good books and the general points they make, and to which I refer, seem to be good ones. Clearly, if I was doing anything more serious, I would need to get up to date.

[2] See note 1.

[3] See note 1.

[4] I might write a separate post speculating on this.

[5] This after all is the argument usually deployed with regard to later seventh-century Francia.