A guest article, kindly contributed by Ross Bailey.
It’s not often that I get asked to put together something that covers my actual profession (archaeology) and my main activity outside my profession (HEMA), so I may have gotten a little carried away here. Initially this was intended to be a brief piece on experimental archaeology as applied to historical weapon use and reconstructions. This has grown to become a look at the kind of synthesis and approach I would personally attempt if reconstructing a historical fighting style for which we have no surviving technical material.
There’s a relatively high chance that this “brief look” will end up on the hefty side of brief. People may also disagree with it. That’s entirely fine! I don’t purport to be a global authority with all the answers, but if arguing with the points I’ve made here gets even one person thinking a bit more critically about things and doing the research to properly give a differing opinion, I’ll be a very happy person.
Experimental archaeology should be a solid recursive experimental process that links in a more than superficial way to the archaeological record. ‘I’ve picked up a historical pattern sword and sort of waved it around a bit and hit some things with it and think, therefore, they used it like this’ really does not count. It can cover a wide gamut; from experiments intended to better inform us of how stratigraphic elements were deposited on site, allowing us to better interpret archaeological sites; through use-experiments of flint tools on a variety of materials to try and identify use-wear patterns that will allow a better interpretation of how a flint tool was used; to entirely experiential, adding nothing new to the body of archaeological knowledge, but allowing the public to experience what it was like to knap flint, create a clay pot, or use a saddle quern, drawing from existing archaeological knowledge and the existing archaeological record. The one key element of all of them is that they meaningfully engage with the archaeological record. In a similar way, any attempt at research to recreate a fighting style should meaningfully engage with as many different sources of historical information as possible.
In this article, for historical fighting styles, I am referring to those that would fall into Michael Chidester’s category of Type III HEMA/WMA:
- Type I being still living traditions like Jogo de Pau and Savate;
- Type II being the ‘usual’ HEMA suspects, working from a source that gives useful descriptions of actual techniques to create a reconstruction of the style detailed in the manuscript;
- Type III being reconstructed from a combination of by-proxy historical sources (sagas, historical biographies, archaeological evidence, illustrations, etc. rather than period fighting manuals).
From my own point of view, and this is heavily my own opinion, I’m only really happy to consider Type III reconstruction attempts as valid attempt if the corpus of evidence backing up the techniques being suggested is sizeable enough (for each technique) for it to have a significant chance of being a more-or-less decent chance of it being reflective of a historical technique.
A workflow for reconstruction
For myself, a baseline workflow for “a reconstruction of historical fighting with the [insert weapon here]” would be something like this (with a few examples to illustrate some pitfalls and give food for thought):
1) choose a specific, tight, time-period and geographical area for the weapon and stick with it.
Think of the differences we see even within a century or two in surviving treatises, or within the same century but in different parts of the world. If we choose something tight and local we hopefully minimize muddling the issue with these from the outset.
Example: The Irish Ring-hilt. This is a fairly recognisable form of sword, fairly tightly dated. Unfortunately, we have no solid documentary evidence on the methods of its use. For the purposes of some of these examples I’m going to pretend I currently know nothing about these ring hilts, other than having seen the giant one in the illustration Thus go the soldiers in Ireland beyond England by Albrecht Durer. Perhaps I’ve seen this and thought “That’s a really cool sword, and it’s from my neck of the woods, that’s the very thing I want to learn”. So, I’m looking at around 16th century Ireland.
2) look at documentary sources
With said time-period and geographical area chosen, start with any documentary sources contemporary to that period from that area and go through everything possible; don’t forget legal records, and don’t forget pictorial evidence – marginalia doodles in manuscripts, period artwork, relief carvings on tombs, graffiti, etc. Compile every quote / passage / depiction / whatever that touches on the weapon that we can lay our hands on. We catalogue it, categorise it, see if there are any patterns coming up.
Example: Irish ring-hilts. We don’t have a huge amount that mentions them directly from that period, not that can be distinguished from generic mention of “any kind of sword because we just called it a sword”, and nothing of a technical nature at all. Not the most useful start. But we do have other depictions of them – Drawn After the Qvicke and so on. So, let’s catalogue them as reference material to compare to anything else we find, and now we move on…
Be aware of the nature of the sources and be critical of them. John Derricke’s Image of Ireland might give us some nice illustrations of various aspects of Irish life, and since we’re fairly sure he was in Drogheda in the late 1500s, he was almost certainly on hand to see the material culture. At the same time since his book was mostly to make his father’s English victories over the Irish look good, and was quite disparaging of the Irish in view of this, we need to view anything he records or depicts with that in mind. This holds true even more for earlier periods.
Giraldus Cambrensis, for another example, gives a nice account of Irish axe use in a single hand that seems fairly reliable. And then we read his accounts of Barnacle geese and other common things and start to wonder if something as straightforward as “they held the axe this way” from him is so reliable after all when used without further corroboration. Common sense can be applied, but what may seem common sense to us may not have been common sense to the people of the time.
3) look at material culture
For the time-period and area in question, we look at the material culture (archaeological record) – are there surviving examples of the weapon? How many? In what proportion to other weapons from that place/time? Are we seeing survival bias? Is what we’re finding in the material culture matching up to what we’re seeing in the fragments of documentary evidence? Compile which bits corroborate each other and, importantly, which bits seem to contradict each other.
Example: Irish ring-hilts again. We have a few surviving examples. Mostly currently languishing in display cabinets in the National Museum in Dublin and a couple in the Ulster Museum in Belfast. We have several publications that give useful details, notably Irish Medieval Swords, 1170-1600, (Figure 2) and other works that follow and hang off it. Oakeshott gives us some handy comparative material to other swords and a typological framework, and the recent Sword: Form and Thought gives us some interesting ideas on methods we might want to investigate in terms of mapping handling characteristics and visual representations for comparison.
And now we may be hitting some interesting problems here: the initial survey of extant Irish ring-hilts isn’t showing much that’s a good match for the huge one in the Durer illustration. The most common seem to be single-handers with fairly short hilts, much closer to an arming sword in that respect (and to those shown in Drawn After The Quicke). Now we have a decision – do I want to adjust my aim for the most common type of ring-hilt and or stick with something that’s now very possibly an outlier? Have I uncovered anything in my documentary / pictorial research that suggests that single-handers should have been in the minority and that I’m looking at some form of survival bias? On the same subject, how do ring-hilts stack up against other swords found in Ireland from this period? Are there more? Less? Is the ring-hilt a minority group in the corpus of 15th-16th century Irish swords? Are we just seeing ring-hilts appear more often in publications because they’re more unusual? Where were they found? Is the ring-hilt a regional style predominantly found within the territories of the Gaelic lordships or is it found in all territories? How were they found – is there some kind of retrieval bias? Were the blades imported? If so where were they imported from (and is it possible techniques travelled with them)?
After 2 and 3 we should hopefully be starting to get a reasonable feel for the culture in question at the period of use. Even if we haven’t found much of direct applicable use to the weapon in question this is fine; nothing exists within isolation and social and cultural pressures can have effects on the use of the weapons. All the little bits may add up to something useful. It also gives us a backdrop for the weapon we’re researching which should colour our understanding of the use of the weapon and point towards any further research we need to do to get a proper understanding of the context the weapon was used in. For example, for the ring-hilts, do I need to look at the Irish legal material to see whether there were restrictions on who could/would use these swords? What was the general culture and learning like – do I need to know my Aristotle? In legal issues again, we know the Brehon laws rated fines for injuries, do we need to search these in detail to try and get an idea of what kinds of violent injuries were common – did this have an effect on fighting styles where controlling techniques may have been trained as much as battlefield/killing techniques?
4) widen the search geographically
For the time-period in question, we widen our search geographically. Do we have any technical sources that address similar weapons from neighbouring areas or areas with significant links? Or do we have surviving weapons from the archaeological record of those areas that closely match those we’re trying to reconstruct techniques for? Even accounts like “whilst in Ireland we saw that they […]” can provide useful by-proxy corroboration. All this material from outside the geographic region is our next category of ancillary evidence – the material culture and chronology suggest that these similar weapons may have been used in a similar way, but this is less reliable evidence as cultural differences may be a factor (think of both the similarities and differences between Fiore and early Liechtenauer longsword).
Example: Irish ring-hilts. We have technical material for similar weapons from Europe. If we’ve done our legwork from 3 we should have a relatively informed idea of things like handling characteristics, closest matches in the Oakeshott typology and beyond, maybe even the most common areas that the blades were being made in and imported from. All of these can help us focus on certain regions to narrow which sources we should be looking at.
We may also have found out about the likes of Eoghan Ruadh Ó Néil, a Irish noble who left Ireland and spent 40 years in the early 17th century in the Irish regiment of the Spanish army in the Spanish Netherlands who then returned to Ireland in the 1640s. Since we have an established link to the Spanish military, we can investigate the roots of the link, and try and ascertain how much in the way of technical swordsmanship might have passed back and forth. This might suggest Iberian sources such as Pietro Monte orFigueyredo (particularly for the large Durer style ring-hilt) might be worth studying for some technical material.
5) widen the search to other time periods
We repeat steps 2, 3, and 4, but for other time periods than the period of study, looking for anything relating to your weapon / culture / use. Bear in mind that as the time period shifts the material gets less reliable as a window on the use of the weapon in the chosen period.
Example: The Ulster Cycle (we’re breaking away from ring-hilts for the moment). This is an epic cycle set in Iron Age Ulster and gives plentiful accounts of fighting, battles, and material culture. Unfortunately, it is a very unreliable source for the material culture of the Iron Age in which it was set – very little of what is described in it matches known archaeological material from the Iron Age. The descriptions have a lot more in common with a cross between the material culture of the Early Medieval period that it was written down in and the other epic poetry (the Iliad etc.) that the scribes were familiar with. Similar holds true for a lot of other sources which are a century or more removed from the period they’re discussing.
6) look at any other ancillary evidence
We look for any other ancillary evidence to tie in. Maybe we’ve thought about looking at the osteological record to see how perimortem weapon trauma patterns stack up against the historical accounts in terms of where people are actually being hit – if there are big differences, that’s something to think about. Maybe we’ve thought about looking for signs of damage on the surviving weapons themselves to see if that can tell us anything about their use. Huge care needs to be taken with this. Lots of things can cause problems for us here.
Example. Bronze Age weapons trauma. It would be lovely to try and reconstruct a style of fighting with Late Bronze Age swords from Ireland. They’re truly beautiful weapons. We do start to run into problems fairly quickly though. First, the predominant method of disposing with bodies in the Bronze Age was cremation, and a lot of the cremations we find are token deposits; not so great for finding trauma on bone (even assuming any violent strikes with a blade connected with bone rather than soft tissue). Since the predominant method was cremation, even when we do find some intact skeletal remains we’re left with the problem of “Why were they buried rather than cremated?” They’re very much in the minority and thus quite likely have a set of special circumstances behind them and thus cannot be taken as representative of the general population. So, we have a small sample group which is already biased in some way that we’ll never know the circumstances of.
Unfortunately, we can say something similar for the swords. Many have been recovered from river dredging and bog sites. Are we looking at structured deposition – votive offerings, some form of formal decommissioning, discarded as a punishment, something else? Or material from fights over territorial landmarks such as fords? Some of the intentional damage on some of the weapons would tend to suggest the former. In that case, can we expect them to show useful signs of having been used in combat? How do we tell the less-spectacularly decommissioned votive offerings from the lost and discarded swords?
If we’re looking at damage to weapons, even if we set aside structured deposition and damage that may come with that, can you reliably tell that the damage was caused during combat? If there’s damage, can you tell if it’s all from the same incident? Has repair or reworking removed any signs of earlier damage so that we’re only looking at the most recent use before deposition?
7) spend time with martial arts
Throughout this, we spend a lot of hands-on practical time with martial arts. We learn how bodies move. We learn how force is generated in a variety of ways. We learn which movements and methods of movement work, and which ones will actually cause us harm if we use them repeatedly (looking at collapsing knees especially here…) We become familiar with all of this. This gives us an experiential background of how fighting systems hang together in a real way and how the body moves through them. It helps to avoid “This is our interpretation based on evidence X, Y, and Z” hitting the unfortunate circumstance where it matches all available evidence, but just isn’t physically practical.
8) compile straight-forward implementations where the evidence is greatest
Given the evidence compiled in 1-7, we work on the simplest most straightforward implementations of techniques that fit the most solid evidence that we have. These are most likely to be the closest to accurate. We reference the body of evidence for this speculative interpretation, especially if we’re presenting it to others. If we don’t back our interpretations up with evidence, we’re effectively falling into the school of “I just reckon they probably did this” and might as well cite Ancient Aliens as our source.
9) compile straight-forward implementations where there is less evidence
We repeat step 8, working down the list of evidence we have for the less and less solid evidence, stopping when the evidence gets too tenuous for the technique to be anything other than complete speculation.
10) consider the body of techniques
We look at the body of techniques, see how it hangs together, compare it to the overall “framework” for the style that seems to be suggested by the most reliable sources (i.e. if the sources suggest a weapon/style was fought in close ordered ranks and the speculative interpretations require a good amount of space on all sides and seem more optimal for single combat, there’s a contradiction that needs to be addressed). We reference the sources for the framework as well as for the techniques themselves when presenting this to others.
11) revisit and re-evaluate
Just like “normal” HEMA, and all good archaeology (experimental or not), we continue to revisit and re-evaluate the interpretations based on the most up to date evidence, and we never stop looking for more evidence.
12) engage with other people
Most importantly, we get out there and engage with other researchers, scholars, and martial artists. They’ll point us in directions we’d never even thought of, knock down our ideas when there are holes we never knew were there (and more importantly when there are holes we don’t want to admit are there), and will generally help keep things grounded and sane.
Obviously depending on the weapon and period we may find ourselves leaning more heavily on some of these points than others. With something 9th century, for example, we might end up with step 1, something from 2 in terms of brief mentions of raids, some little marginalia, and some effigies, a fair amount from 3 that gives us some physical data and corroborating evidence but little martial insight, some corroborating elements from 4, and a lot of by-proxy evidence from 5 in terms of sagas and historical accounts that we need to be very critical of. Also maybe a few much later manuscripts dealing with very broadly similar weapon sets that require a lot of reverse-engineering and understanding of everything we’ve uncovered in 1-5 to try and highlight the likely similarities; 6 might reinforce things that we’re thinking or show up some promising leads as dead-ends; 7 will do some critical error checking in terms of whether things actually are viable, and then 8, 9, and 10 may come up with a rather reduced set of basics that hang together stylistically in a way that, whilst not being verifiably 100% accurate, at least isn’t inconsistent with the characteristics, cultural context, and available evidence, for that weapon set.
In summary, for a Type III to be a realistic attempt to reproduce a period fighting style, for me, it would need to be heavily backed up with:
- a properly critically evaluated combination of consistent and reinforcing material from contemporary sources (documentary, artefactual, pictographic…) of the period of use of the weapon,
- which is further backed up with corroborating evidence from the archaeological record, reverse engineering from more recent sources, and a solid physical grounding in martial arts, and as much other ancillary material as possible.
Just like “normal” HEMA, this is one giant triangulation, just on a much broader scale, that will allow as solid as possible a reconstruction of the “best fit” elements.
There’s nothing to say that everything on this list should be the efforts of a single individual. Particularly specialized elements dealing with artefact analysis, osteology, translations, and so on. Check the likes of EXARC for a broad swathe of experimental archaeology covering a huge range of topics from community outreach and education to highly technical experiments designed to expand the information we can usefully recover from the archaeological record. Keep an eye on the Society for Combat Archaeology. Go read the Bulletin of Experimental Archaeology, and Acta Periodica Duellatorum, trawl through JSTOR; go have fun exploring all of the potential sources of information and have fun following rabbit holes.
 McClintock, H. F., 1950, Old Irish and Highland dress, Dundalgan Press. p 30-31
 McClintock, H. F., 1950, Old Irish and Highland Dress, Dundalgan Press. p 31-34
 Halpin A., 1986, ‘Irish Swords c1170-1600’ in Proceedings of the Royal Irish Academy. Section C: Archaeology, Celtic Studies, History, Linguistics, Literature, Vol. 86C (1986), pp. 183-230
 Oakeshott, E., 1991, Records of the Medieval Sword, Boydell Press
 Grotkamp-Schepers et al, 2015, The Sword: Form & Thought, Deutsches Klingenmuseum
 Kinsella, T., 2002, The Tain: Translated from the Irish Epic Táin Bó Cuailnge, Oxford University Press
 Mallory, J. 1992, The World of Cú Chulainn: The Archaeology of Táin Bó Cúailnge, in Aspects of the Táin, December Publications, p. 103–153
 Cahill, M. and Sikora, M. (eds), 2011, Breaking ground, finding graves — reports on the excavations of burials undertaken by the National Museum of Ireland 1927-2006. National Museum of Ireland Monograph Series 4, Wordwell
Ross Bailey is one of the instructors at Medieval Combat Group in Belfast, Northern Ireland, primarily researching and teaching Lecküchner’s messer. Ross has been a professional archaeologist in the commercial sector for 12 years and is currently a site director with Northern Archaeological Consultancy Ltd. As well as playing with sharp bits of steel in his spare time, he also creates sharp bits of flint, producing occasional lithic specialist reports, and experimental knapping sequences to aid in interpretation.