My recent efforts to decipher Agrippa’s fencing manual has me thinking a lot about the types of evidence that are available to us as historical fencing scholars. As students of historical European martial arts, we are faced with the challenge of attempting to recreate techniques that have been out of use for 500 years or more based exclusively on a small number of documents. In doing so, we often must rely on outside information in order to successfully recreate the art, however it is important to consider the types of evidence that can be used, the limitations of these sources, and the order in which we should give them priority. For the vast majority of fencers, the types of sources, in order of importance are as follows:
- The text in the manual
- The images in the manual
- Other related manuals
- Other martial arts principles/ human body mechanics/secondary sources
- Empirical testing
I just read it for the articles…
Certainly the placement of the primary sources themselves at the top of this list should be no surprise. With several centuries between us and historic fencing masters, the closest we can come to them is by reading the words that they wrote in their books. For most modern students, this type of evidence must be viewed through the lens of a translator, however the last 10 years have provided multiple translations of several manuals and a growing ability for scholars to access scans of the original texts themselves which can be combined to help rule out biases introduced by translation. Unfortunately, the treatises themselves tend to be focused largely on “what” to do rather than “how” to do it, assuming a certain level of fencing knowledge on the part of the reader which is not always the case for modern practitioners.
A Picture is Worth 1000 Words
Because of this limitation, we frequently must use other evidence in order to figure out the mechanics of historical fencing actions. The second best evidence we have comes from the pictures found in historic manuals. In my recent attempts to unpack Agrippa’s manual, I’ve needed to rely rather heavily on the images presented in the manual. However, there is a (sometimes big) problem with relying on the images, namely that they are not photographs and can therefore include mistakes made by the artist or may reflect aesthetic ideals (e.g. body proportions, musculature, posture) rather than reality. Likewise, as scholars, it is not always clear what the relevant parts of an image are and therefore we may miss details. On the other hand, images don’t require translation, can show us much more about “how” we must stand/move than the text, and were generally included by the master for a reason. We must be careful, however that we neither over-interpret pictures nor fail to notice nuance that they may present.
… unless he has studied his Agrippa…
We can also use evidence from other, related historical manuals. If our goal is to learn to fence according to a particular manual, we can reasonably apply principles from other, related historic treatises to help fill-in-the-gaps. This is possible to varying degrees depending on the manual(s) that you are interested in studying. For example, if a scholar is interested in fencing according to Capo Ferro’s manual, they will find that other Italian manuals from the same period such as Giganti and Fabris will be highly relevant to helping to supplement Capo Ferro’s manual. Furthermore, that scholar may also find that other Italian manuals are also relevant, and such a scholar would be well-served to explore Agrippa, Manciolino, Marozzo, Alfieri, etc. These manuals describe fencing systems that are further removed from Capo Ferro than say, Giganti’s, but they all roughly fall within the same tradition. In contrast, while studying German manuals may help that scholar develop a better understanding of body mechanics, sword use, etc, their contents will not directly apply to understanding Capo Ferro’s system.
Mind the Gap
Unfortunately for all of us, in many cases the historic manuals don’t tell us enough information about historic fencing systems to definitively declare that we have “gotten it right.” In these cases, modern scholars must 1) Fill in the gaps with something and 2) Determine a method by which to test whether those substitutes are good. There are essentially two methods for doing this, one is to apply principles and techniques that have a solid basis in human biomechanics and/or other martial arts, the other is to apply empirical testing (i.e. try it out and see if it works reliably). The major dangers for using these types of evidence are due to our limitations as scholars. For the most part, western martial arts are the purview of amateurs. Few of us are experts in other martial arts, and even fewer of those can elucidate the fundamental mechanical principles that underlie those arts. Likewise, our empirical testing is significantly limited by sample size.
It’s the Principle of the Thing
While the details of other martial arts may not be directly applicable to the system that we are attempting to study, all martial arts rely on the application of certain physical principles based on the mechanics of the human body, weapons, etc. Once again, some martial arts will be more directly applicable than others, but ultimately what we are seeking to learn at this level is the underlying principles as, once we understand these, we can start to determine the trade-offs and optimizations that are present in our chosen system of study. The pitfalls in using this type of evidence are the possibility of over-generalizing details, mistaking details for principles, and failing to understand the context that determines how to apply the principles.
For example, one area of contention in the interpretation of Italian rapier combat is the placement of weight on the foot. Should fencers place their weight on the ball of their foot or their heels? The manuals themselves do not explain this in the text, and the pictures aren’t clear on this detail either. One way of answering this question is to look at other martial arts, sports, etc to see what they are doing and, in this case, it is likely that if you did so, you would conclude that the weight should be placed on the balls of the feet. However, the conditions of Italian Fencing aren’t necessarily the same as in those other martial arts, so we might be over-generalizing a detail that is present in other martial arts. Seeing several examples of weight placed on the balls of the feet, we might also convince ourselves that this sort of weight placement reflects a broader martial principle, in which case, we would be confusing a detail with a principle. Finally, we might find some examples where weight is placed on the heels, but conclude that we cannot apply this to our current understanding of the Italian rapier guard. This may be due to a failure to recognize other features of the guard that are necessary in order to make placing the weight through the heels useful.
Weird Science
For those of you who know my profession (scientist) it may be surprising that I have listed empirical testing last. However, recall that this ordering is specifically meant for most historical martial artists. However, the problem with empirical testing for historic martial arts is that our disorganized nature and relative amateur status means that few if any practitioners are able to meet the basic requirements of producing good experimental research. Importantly, empirical testing requires that we:
- Constrain what we are testing to falsifiable questions
- Rule out confounding factors
- Demonstrate that our findings are repeatable
Constrain what we are testing to falsifiable questions
The first step is the easiest, however it requires a certain awareness of the body mechanics such that the component pieces of a stance/action can be separated from each other in order to be tested separately. Even many skilled fencers are incapable of elucidating “how” their body moves while fencing. And so, many attempts to fill-in-the-gaps fail by simply asking useless questions. Furthermore, we must determine a metric by which to judge our results. Within the current state of historic fencing scholarship, winning has been treated as the primary metric. However, there are significant limitations to this approach. For starters, the pool of opponents may not be particularly good, which makes this metric prone to reaching false positive conclusions (X worked, therefore X is correct). Similarly, the rulesets for competition may be a poor simulation of historical combat as was discussed in Ruairc’s post last week. Once again, ahistoric techniques may be optimal for winning and it would be a mistake to then conclude that such techniques are actually historical.
Rule out confounding factors
While the first of these steps is hard, the second is far more difficult. Once again, the ability of a scholar to rule out confounding factors relies heavily on their ability to understand how to break actions into parts. The nature of fencing already makes this difficult, and it sometimes is not possible to manipulate small parts of an action without also disrupting other parts. At this stage, the biggest pitfall is that we will reach conclusions that are simply unfounded. A frequent example of problems at this level include the belief that a fencer must swing hard in order to make certain techniques work. These fencers are failing to separate the confounding effect of force from failures in their mechanics. Once again, such mistakes can lead us to conclude that techniques are historical when they are not, but here we are also more likely to make type II errors (false negatives) by concluding that techniques do not work when they actually do (because we have failed to rule out confounding factors and/or are missing a relevant factor).
Demonstrate that our findings are repeatable
The third step of empirical testing is nearly impossible given the current status of historic martial arts scholarship. In order for us to reach a reasonable conclusion that an action is “correct” for a given system based on empirical testing, we must replicate that finding and, this does not only mean that we must, as an individual, be able to perform it reliably, but that it must be able to be performed by other people reliably. It is not enough to be able to conclude that an action is correct simply because you can do it reliably. Any conclusion that has not been put into practice by a reasonable number of people cannot be considered to have been well-supported by empirical testing. This poses a major problem for historic martial artists because there are very few situations where this type of testing is even possible. I would expect that rigorous testing would take the form of several instructors implementing such an experimental technique within their students and then testing 1) whether the technique works for the instructor 2) whether the technique fits well within the instruction of students and 3) whether the technique works for the students. Having this occur in multiple groups helps to control for confounding variables and will also likely be necessary to produce a sufficient number of practitioners in order to generalize any claims based on such an experiment. The worst failures of this part of empirical testing result in echo chambers where strange and unusual techniques are considered correct and good.
In Conclusion
For historical fencing scholars, having a good understanding of the types of available evidence for historical fencing techniques and the limitations of these types of evidence is very important. This knowledge can help fencers to avoid pitfalls, improve their ability to interpret historical texts, help to ascertain the veracity of claims by other historic fencers and, perhaps most importantly, help them to develop as a historical fencer in an efficient manner.
Pingback: Dante di Pietro
Pingback: Ruairc