What do you want from a historical label?

This Autumn there has been a lot of talk about whether the term ‘Anglo-Saxon’ should, on the whole, be put aside. [There are different recent takes here Mary Rambaran-Olm and Michael Wood]. Surely: yes. The history of the modern English word is firmly rooted in early modern ways of thinking about the world divided into races, and such thinking has continued into ways some people frame racist worldviews. Because words evolve and diversify in meaning across different contexts, the label also came to have less immediately problematic connotations, as it became a standard way to classify preconquest English history. Some of the issues were raised in a seminal article in 1985 by Susan Reynolds.[1] Reynolds noted the tension between the ‘ethnological’ and chronological uses of the word, but she also argued that both can do more to obscure our understanding of the past than to enlighten it. Nearly 35 years later, this still needs some serious consideration.

Variations of ‘Anglo-Saxon’ are certainly attested in the early Middle Ages – Paul the Deacon in the eighth century, for instance, calling the English Anglisaxones, or the ninth-century author of the Life of Alcuin in Tours calling the priest Aigulf Engelsaxo. (For both, there was probably an implicit contrast with the Altsaxones of Germania). The problem is that such terms are rare, inconsistently applied, and more often other words were used as identifiers. Ethnographic – and, relatedly, political – labels have to be understood in their diversity and in the contexts in which they were deployed. They are tied to their own moments, reflect changing power structures, and are parts of different arguments and ideologies across time and space. These are things that need to be understood. And indeed, for me, these are often the interesting bits of history anyway.

The worlds people have wanted to describe as ‘Anglo-Saxon’ have varied wildly too. The fifth-century world of migrating non-Christian warriors from Scandinavia and Germania was not very much like the immediate preconquest, Christian, unified and bureaucratic kingdom of Edward the Confessor 600 years later. The way it is used can cover so much ground that, as a description, it doesn’t really mean much analytically concrete. Reynolds concluded her analysis by saying, if people wanted to use the term, they really needed to think harder about what they meant by using it. It is not always clear that people have done that.

For sure, there will probably long be grey areas of public usage. Last year’s smash hit exhibition ‘The Anglo-Saxon Kingdoms’ at the British Library is probably a good case in point. The name, I am sure, owed more to brand recognition, focus groups, and the need to get people through the door, than the propping up any dubious agendas. Maybe they would choose a different name if there were a ‘next time’. But the important thing for the organisers, in 2018, after the Brexit vote, at a relatively politically conservative public institution in the UK, was that they were able to put together a powerfully-articulated statement about how the worlds of the early English kingdoms were international and (relatively) culturally diverse. Underneath the headline, there was a positive and timely story to be told. I hope that’s what it is remembered for.

In historical scholarship, it should not be too hard to avoid ‘Anglo-Saxon’ and in fact to use labels more useful for analysis.[2] I may have called my first book Anglo-Saxons in a Frankish World, but my use of ‘Anglo-Saxon’ was mostly to flag that it might be relevant to people who worked in what might then have been called ‘Anglo-Saxon studies’. (And yes, I would have picked a different title in 2019 than 2004). Past the title, use of the term was limited (if not systematically so): I discussed some of the problems raised by Reynolds and tried to understand what early medieval writers meant by their different ethnographic and political terms. Sources connected to or about English religious figures on the continent talked about English (Angli), Saxons, Mercians, Northumbrians, West Saxons, East Anglians – all sorts. And they didn’t always mean the same thing by the same word every time. This was useful to acknowledge in the face of modern English and German historiographies that did (and sometimes still do) tend to perceive a simpler English-German binary, especially as that binary situation was often defined by things like World Wars. That part of the study was, at worst, a good starting point for understanding what the sources and scholarship were trying to achieve and how authors understood the role of nations, ethnicities, and cultural groups in history.

One of the common responses to the recent debate has been ‘well, if we don’t use “Anglo-Saxon”, what do we use?’ (I am going to sidestep the issue of the name of the learned society, to which I don’t belong now anyway, but I am glad it decided to change). The first response should be ‘well, what are you attempting to describe or analyse?’ Confronted with many things and many historical labels for them, we can probably come up with more than a single, anachronistic, and vague label to assess them under. As one of my colleagues recently opined as an example, it could be much more useful and precise anyway to describe the south coast port of Hamwic as ‘West Saxon’ than ‘Anglo-Saxon’. As a trading port, it could even be more revealing to study it in relation to continental ports rather than, say, the Northumbrian kingdom. There are probably even better labels that aren’t attached to either ethnic or political labels, but I don’t work on Hamwic so I would be guessing. The point is: there are often more precise labels that might illuminate evidence better, and it always helps when we talk about why we label things the way we do.

We should not, of course, think that ‘Anglo-Saxon’ is a unique thought-space in which diversity of labels, concepts, and peoples get squished together in problematic ways. The thought-habits are endemic.

Studies of the Frankish kingdoms can get quite messy as people try to align modern expectations and medieval realities. The dynastic names ‘Merovingians’ and ‘Carolingians’ often get used as synonyms for the subjects of those dynasties when they never are in the sources (and barely ever, even, for the dynasties themselves). Gallia and Galli get translated as ‘Francia’ and ‘Franks’ on a curiously regular basis rather than ‘Gaul’ and ‘Gauls’, reinforcing a conflation of gens Francorum with regnum/a Francorum that cares little for diversity of ethnic, political, or local identities. (Maybe the idea of ‘Gaul’ just seems more obvious to me because I grew up reading Asterix books). Historians often refer to ‘the Frankish Church’ – the title, of course, of a classic book by Michael Wallace-Hadrill. But again, it’s not a term people at the time ever really used – it is a way for modern historians to make sense of a regionalised but mostly-transnationally-minded set of institutions by understanding them within a simpler national framework.

With that last example, it may not always be inappropriate to describe the churches under the Merovingians or Carolingians as operating as a ‘Frankish Church’ – i.e. as a Church defined at least in part by the geographical parameters and political concerns of the Frankish king(s). Where one believes that it is meaningful to describe it in such a way, it is worth explaining why as part of the analysis. This is how ‘responsible labelling’ works.

Similar problems about how we square modern and premodern labels have also just been raised by Naomi Standen with regards to China.[3] It is again very much our word – an early modern creation of disputed origin that vaguely maps onto Zhōngguó 中國, but another term that a) is not actually used very much compared to other labels and b) runs the risk of treating as stable and homogeneous a culturally (very) diverse and geographically-changing political world. Like ‘Anglo-Saxon’, it has advantages in convenience and brand recognition, but as a label used lazily in historical analysis, it oversimplifies – and sometimes to nationalist ends that are not always innocent or used innocently by readers later. Again, Standen advocates paying more attention to regional variations and the situational use of language and labels. Without interrogating our assumptions or those of the sources, we understand neither very well. And let’s be fair: this is not really a new or radical suggestion.

And as an aside: if we want to have more conversations across and between cultures – as many medievalists now do with comparative and world histories – it is useful to go into things aware that problems with labels are widespread and, often, highly political.

The consequences of interrogating labels can lead to some uncomfortable discussions, sure. There has, for instance, been some fierce recent exchanges about Irish influence in the early Middle Ages. There has been a wealth of good scholarship over the past 70 years that has identified the significance of Irish figures and learning outside Ireland itself, from Northumbria to Italy. Without denying that exactly, some recent scholarship has noted that ‘Irishness’ is not always particularly explicitly labelled or discussed as such in the early medieval continental sources.[4] But there have been objections too, and some scholars have argued that some efforts to recalibrate narratives of Irish influence against the evidence has amounted to a systematic attempt to denigrate Ireland. This is not what was intended, certainly, and it will be interesting to see how such discussions proceed from here.

In early medieval computus – my pet favourite example – there are certainly some interesting challenges in getting the balance right. Some ideas and models common in continental European centres clearly had travelled via Ireland, but this includes:

  • Texts not from Ireland that were still transmitted through it, but which might also have been available in copies from other sources.
  • Texts developed in Ireland, but to different ends by people fighting different battles in time and space.
  • Texts developed by people from Ireland or trained by people from Ireland, but who wrote somewhere else altogether.
  • Texts developed by people with no direct Irish connections only indirectly engaging with material from Ireland.

Unpacking the ‘cultural identity’ of such knowledge cannot be done easily – it needs multiple labels, some precision, and a lot of explaining of what any of it means, either for them or us. We don’t get far in studying the past by shying away from complexity.

Responsible labelling is about finding languages to analyse the past effectively and in ways that don’t distort it… and then finding ways to discuss that analysis in ways that are meaningful to us without distorting it again. It is not just about picking a couple of appropriate labels, but also about what we then seek to do with them in the stories we tell. It is hard, and even trying very hard people will make ‘mistakes’ in the eyes of other people over and over again. Language is rarely ever used by two speakers in quite the same way. (I am married to an American academic – I feel this keenly on a daily basis). Many problems are also generated by unintentionally loose rhetoric and the Humpty Dumpty Defence (‘the words mean what I want them to mean’). It is even harder to navigate all this when media such as Twitter mean that people from different conceptual universes of thought are operating in the same place. But that’s life in 2019.

And maybe some of the debates coming up in 2016-2019 is crucial. What is the history that we want and need now? The study of history is often political and the discipline engages with and reflects the world around it. We should not expect to see the past in the same terms as people have always done, because people have never always seen the past on the same terms. As Reynolds said, we have to think about what we mean by our words and what others may think we mean. The labels we choose will define the kind of discipline we are trying to create. And I, for one, would like to see a medieval studies that is supportive, diverse, inclusive, and engaged – and always prepared to interrogate its assumptions. But let’s also remember that labels are only one small thing: there are bigger and more serious structural issues to confront in society and in academia if we want anything like that kind of medieval studies.

We might not always agree. We might not always be right. We should be open to listening and asking if we could do better.


[1] S. Reynolds, ‘What do we mean by “Anglo-Saxon” and “Anglo-Saxons”?’, Journal of British Studies, 24. 4 (1985), 395-414.

[2] FYI in addition to what I am about to say, Apocalypse in the Early Middle Ages contains zero mentions of ‘Anglo-Saxon’ and Early Medieval Hagiography two (once summarising something written by Colgrave, once summarising something written by Rollason, in both cases precisely because they had used ‘Anglo-Saxon’).

[3] N. Standen, ‘Colouring outside the lines: methods for a global history of eastern Eurasia 600-1350’, TRHS sixth series 29 (2019).

[4] E.g. S. Meeder, The Irish Scholarly Presence at St Gall: Networks of Knowledge in the Early Middle Ages (2018), esp. ch. 1.