Ask the Audience: Retreating in Italian Rapier

So I’ve come across a couple different mechanics for the retreat, in a historic Italian context:

Through the Heel – peel up the toe (dorsiflexion) and straighten the front leg, pushing the front heel into the ground to drive the body back.

Through the Ball – flex the foot in the opposite direction (plantarflexion), using both the ankle flexion and knee extension to push yourself back.

A student of mine is interested in this question, and when I couldn’t give a satisfactory answer I promised to seek a wider array of input. Which do you think more correct for the system (or are both acceptable? Or is something else better?), and why? Please support your answer with text from the primary sources, better-known principles of Italian fencing, anatomy/kinesiology, or physics. Personal anecdote is weak but acceptable.

42 comments to Ask the Audience: Retreating in Italian Rapier

  • Dante di Pietro

    Through the heel. If you initiate the recovery by bending the rear leg first, the heel, as a roll and push, is best.

    Toes are better if your rear leg isn’t doing its job.

  • Gawin

    Through the heel. From a motion perspective, moving through the heel allows you to form a single, straight-line vector of motion whereas pushing through the toe applies the force along two separate vectors, creating non-linear motion.

    Also what Dante said about a weight shift. It may not be the whole body, but you’re at least shifting the weight of the right leg forward before moving backwards.

    • Gawin

      I should probably clarify. In the heel method, the single vector is formed on an angle from the hip towards the heel. The leg, while bent, remains rigid.

      If we use our toe, we can accomplish the same mechanics, but only if we keep both our ankle and knee joints rigid. If we do so, we can form a single force vector from our hip to our toe. This is, however, difficult to manage because we now must maintain rigidity through 2 joints rather than one. Furthermore, the ankle is a relatively small joint and without significant strength training, will not actually be able to do this. Finally, the change in angle from the shift to the toes makes it more difficult to keep the knee rigid.

      However, these aren’t the mechanics that you specified and indeed, if we use both the knee and ankle joints in the motion, we’ll encounter several problems. First, we will find that our weight must come forward in order to actually support ourselves using the ball of our foot. If we are holding our guard on the ball of our foot, then we will be slightly more forward, if we must transfer from heel to toe in order to retreat, we add a tempo to the action. Setting that aside, however, we also find that the mechanics underlying how we carry out the motion also change. When we use the knee and ankle, we form a compound lever, which may sound good, but it is important to remember that levers conserve work such that lifting a heavy object requires a trade-off in distance, so while we might consider the use of a compound lever here to be advantageous for helping us amplify the amount of force we are applying, the reality is that each lever provides a multiplicative increase in the distance that must be moved. Furthermore, this system also results in non-linear motion of the relevant parts, as the action of compound levers is to transcribe curved paths. We also run into a problem if we consider the angles at which we are using our levers to apply force into the ground. If we drive through our heels, you will note that we are applying force roughly parralel to the tibia which is near perpendicular with the ground whereas if we drive through the ball of our heel/toes, we are applying force roughly parallel to our metatarsals, which is far closer to parallel with the ground.

      • Gawin

        I spent some more time thinking about how to best explain this to your student.

        Given your description, I see three possibilities for how the motion is being carried out. In one, the fencer starts by flexing the ankle, then by extending the knee. In the second, they start by extending the knee and then flex the ankle. In the third, they sorta try to do both at once.

        Case 1: The fencer is only moving themself with their quadriceps. They are also only indirectly applying force to the ground as the vector of the applied force is first parallel to the femur and is redirected by the bend in the knee (which may also be bad for the knee)

        Case 2: The fencer is only moving themself with the gastrocnemius. The knee extension negates the involvement of hip muscles because they can’t apply force towards the goal of moving the knee. Likewise, because the ankle flexion occurs after the knee extension, the quadriceps can’t provide much help because the ankle joint is the one involved in the final application of force.

        Case 3: If both joints move at the same time, both the quadriceps and gastrocnemius will be contributing effort towards the motion, but neither will be able to do so optimally, as the motion of each joint will be “cushioned” by the bend in the other. In other words, this method doesn’t actually give you anything rigid to push against.

        In contrast, if you drive the retreat by keeping the knee and ankle rigid, the muscles involved in the retreat are larger and stronger and the core, gluts, and quadriceps. Here what matters is the rigidity of the knee and ankle, so in theory you could retreat from the ball of your foot by keeping those joints rigid, however this is more difficult than simply pushing through the heel, so there’s no reason to do so.

  • Cailin

    Even in strip fencing, which keeps a far more center weighted stance, the energy is channeled through the heel of the foot. There are a few ways that the action can start from the ball of the foot, but these actions quickly transfers to heel, and they have very limited application. A few things need to be said: (1) the techniques I’m referring to are modern adaptation found nowhere in the manuscripts; (2) the techniques take the better part of a lesson to teach mechanics and proper implementation into a bout per technique ; and (3) the techniques all end by driving off of the heal.
    To the point at hand, go with the heel. As Gawin has pointed out, motions from the heel utilize larger muscles. The ultimate goal of foot work is to be able to win the distance game with your opponent, and the smaller amount of ground each step covers, less time each step takes, and smother your foot work is the better. Larger muscles can move a large amount of weight with more control and efficiency.

    • Ruairc

      I’ve been observing Youtube videos of Olympic athletes to try to get a better idea of how the best perform a similar action, and if it might cross over to a historical context. It’s not definitive, of course, but almost nothing is in HMA research.

      Counterarguments include the freakin’ enormous calves of the fencers in the plates, which suggest pressing the body up through the toe (plantarflexion). Counter-counter-arguments, of course, remind that the plates show odd proportions and taking details from them is always suspect. Counter-counter-counter-arguments fall back to the idea that a single plate tells little, but a detail that is consistent across several tends to be reliable. And so it goes.

      I will submit all this to my student and see what is made of it.

      • Dante di Pietro

        “Big calves” is a counterargument? By all means, I invite anyone to fence with a comparably porous counterguard.

        • Ruairc

          Presumably, one of the reasons the fencers are illustrated nude is so that anatomical details and geometries are more apparent. A great many elements of motion, strength, and stability can be accomplished in multiple ways by the human body. (Some of them are demonstrably worse, which is why form is important when you lift weights.) If fencers appear to be using certain muscles instead of others, this is relevant.

          I’m not 100% behind this yet, but I can’t deny it as a valid line of inquiry.

          • Dante di Pietro

            If drawings appear…

            I’m not going to put much stock in that when compared with things like mechanical analysis and anatomy.

          • Ruairc

            Assuming the plates suggest one thing but some other thing seems better, I see several possibilities. Broadly, they are:

            1 – The models screwed up
            2 – The artists screwed up
            3 – The printer screwed up
            4 – The modern reader/interpreter screwed up
            5 – Renaissance fencers used different body mechanics (possibly even less efficient ones!) than the ones we favor
            6 – Science is wrong

            I feel safe ruling out the last one. To the extent that the hypothesized actions are unnatural or vastly inferior to other options, I feel safe ruling out #5 and perhaps #1.

            But I’m also not going to pretend that my knowledge of anatomy and kinesiology is so advanced that I can assert #4. It’s worth investigating, if simply so I’ll have a solid, scientific reason to say “no, this other way is better”.

            I don’t think HMA ever suffers for having less dogma and more experimentation.

          • Dante di Pietro

            The experiment, in this case, can be safely limited to: “try one way, then try the other, make sure you’re not doing either wrong, then stop doing the one that’s worse.” How many months will go to waste going down every rabbit hole there might be?

          • Ruairc

            Going down rabbit holes should be a significant part of our research. Sure, not every one merits deep investigation, but I’d think “is there another way to retreat? When/how/why is it better or worse?” would be more fundamental, and the answer more enlightening, than “what exactly is going on in this one part of this one plate, which details a response to a situation I’ve literally never seen in free fighting?”

            None of this suggests we neglect work on the other core elements of fencing. This is a research question, and is to be treated as such.

          • Dante di Pietro

            I don’t think either of those questions provide any great depth of enlightenment unless you can physically perform the actions under stress. Academic understanding is of limited utility when not coupled with practical application; mental clutter does not strike me as beneficial.

          • Ruairc

            Duh?

            Achieve academic understanding first, then move to practical application.

          • Dante di Pietro

            That seems like a faulty approach to me. Academic understanding is very limited and lacks the elements necessary to claim to *know* a physical skill in a useful, practical way.

      • Gawin

        The calves aren’t particularly large compared to say, the rest of the leg?

  • Ràðólfr

    The arguments for heel are interesting. From personal athletic experience as well as from various papers on body mechanics, I would say toe for 1 reason in particular. You are engaging 3 muscles instead of 2. If you drive through the heel you use the flutes and quads. If you use the toe, you add the calf. In the various running and jumping sports that I have participated, most notably parkour, utilization of as many muscle groups simultaneously as possible is key to power generation. Just to prove the point, stand up straight and jump straight up, then try it without using your toes.

    • Wistric

      Is a retreat supposed to be a jump?
      What would you say are the qualities of a good retreat?

      • Ruairc

        Eh, are you familiar with the Bolognese salto? Sure, different context, different weapons, different guard, and the rapier masters don’t even use the same word (near as I can tell – haven’t read everything in the original Italian). But there’s some retreating, jumping footwork for you.

        • Wistric

          A salto is a salto (and a rare duck at that) and a retreat is a retreat. We’re discussing the mechanics of a retreat.

          • Ruairc

            What makes you think a retreat should be different?

            I can find nothing in the primary sources detailing the mechanics of a retreat; furthermore, the Italian ritirare (the word usually translated as “retreat”) is also rendered “withdraw”, “pull back”, “retire”, etc, and can refer to the body, the sword, or the feet … and it’s not always explicit as to which.

            These sorts of ambiguities are usually resolved by trying out all the options until the most effective option emerges or other textual references begin to make sense. But this can be a dangerous process, rife with confirmation biases and compounding errors.

            I’m a bit concerned that we’ve got too much (unsubstantiated) modern footwork in our historical fencing as it is. Fabris talks about the “advance” as something to be done in one or two very, very specific situations. We use it all the time.

            Despite my misgivings, I’m trying to keep an open mind.

          • Dante di Pietro

            A retreat should be different because it’s a “different context, different weapons, different guard.”

            Your physical position basically dictates the optimal and available paths of movement, no matter who you are and what you’re doing.

            Insofar as the types of steps, they’re all contextually bound. Those contexts are very apparent in application; if a gathering step is better than an advance at a particular time, it’s very easy to tell.

            I think that thinking critically about our actions and interpretations, and then testing them, solves any misgivings worth having.

          • Ruairc

            Context includes your physical fitness and training, as well as your opponent’s. It’s possible that the “right” thing to do only becomes right when you’re strong and quick enough to actually do it (e.g. pretty much everything Fabris).

            At any rate, if it’s so very easy to tell, why are there so many who do it so very wrong? And why is it that the best fighters can be very successful using footwork the masters explicitly recommend against?

            I’m just not as confident as you are that our experimentation and rigor have been sufficient to declare the question settled. There’s nothing wrong with saying “we’re not sure because we don’t have sufficient information from primary sources – indeed, some of the things we’ve found to be successful are in apparent contradiction. This is what we think, and why, but we could well be wrong.”

          • Gawin

            “And why is it that the best fighters can be very successful using footwork the masters explicitly recommend against?”

            Most SCAdians aren’t very good. The primary method of defense in the SCA is measure. The manuals are explicit about footwork that is used in or near measure, but what you do when you’re out of measure is largely irrelevant to whether you’ll get stabbed.

            Retreating isn’t artful because you cannot kill your opponent in a single tempo.
            Parries aren’t artful because you cannot kill your opponent in a single tempo.

            Both of them will keep you alive at wide measure and most SCA fights happen at that measure.

            Likewise, the vast majority of advances in the SCA occur out of measure or are attempts to enter measure that are accompanied by the opponent retreating, thus they end out of measure.

            Why do we use so much footwork in the SCA? Because our darn opponents won’t stand still and let us enter measure.

          • Dante di Pietro

            Who is successful using what footwork?

          • Ruairc

            In general, the most successful SCAdians use ahistorical footwork, along with ahistorical blades and ahistorical tactics. So do most HEMA rapierists, for that matter. My goal is not to recreate an ahistorical style of fencing, so I don’t see this as a good lead, except inasmuch as it suggests patterns of sound martial movement. Evidence, strong evidence even, but far from proof.

            “This is the best we have. This is why. But we could be wrong.”

          • Dante di Pietro

            “In general…”

            Provide some evidence.

            I’ll spare you your own standard of “proof.” Evidence will do.

          • Ruairc

            You dispute this claim? After the talk of “cheater blades” and “C&T is great, because it stops ahistorical HR tactics”? Do I need to (once again) look up my old data showing that rapiers in 1600 weighed about 3 lb, and had around an inch of width at the ricasso, compared to our <2 lb colichemardes and Zen Warriors? Want me to find videos of SCAdians winning with very, very modern footwork?

          • Ruairc

            I must say, your response is curious. Wouldn’t you rather encourage an enthusiastic student to pursue some new line of inquiry? Isn’t this how we learn, and overturn dogma?

            I am not so bad an instructor that I am challenged on a whim, and I can answer most questions quite satisfactorily. If I haven’t been able to convince this student (using some of the arguments presented here, and another couple besides), maybe they know something I don’t. Why not let them build a case? It won’t greatly stunt their training or make them injury-prone in the meantime …

          • Dante di Pietro

            Well, you make a claim about very successful SCAdians using ahistorical footwork and offer zero evidence during a conversation where an actual biomechanical explanation for one point wasn’t enough to refute your concerns about calf size in drawings.

            The blades, etc., is a digression, so I’ll ignore that.

            In any case, you’ve moved from refusing to consider substantial evidence as “proving” the point to making a much less founded assertion that modern people use “ahistorical” footwork “that the best fighters can be very successful using footwork the masters explicitly recommend against” without any evidence, and have become belligerent when asked to provide examples.

            Of course, if we grant these other movement patterns as “ahistorical” instead of “not Italian rapier as we know it,” then it still only means that “potentially successful” has a range of actions and contexts that includes things that aren’t found in historical fencing.

            And then being a successful SCA fencer isn’t automatically synonymous with being a good historical fencer, since the rules involved allow for some strategies to prosper that wouldn’t under different conditions; this in no way invalidates any historical style.

            At some point, “we can’t ever *really* be sure” becomes a waste of time and effort, no longer good skepticism and critical thinking, but a hindrance. One must even be skeptical of skepticism, and know when to accept that a point has been sufficiently well proven and isn’t unseated because it is not a 100% certainty.

            What I find most puzzling is that you seem unable to accept that there is a wealth of information that can be brought to the study of HEMA that is not contained within the manuals, but is nonetheless as applicable as the words of Fabris et al.

          • Dante di Pietro

            No, I wouldn’t, any more than I would encourage an aspiring chemist to mix random household chemicals together in an unventilated room. A sufficient amount of ignorance is harmful.

            There’s a reason why I’ve been fencing for 15 years without much time off and have none of the injuries that plague my counterparts.

            I was also a teacher for 11 years, and I had plenty of kids who were smarter than they were knowledgeable, and they all heard the same thing: “Do what I say, try it out for a while, and if it’s not working right, we’ll revisit this.” No one ever did, because I had enough expertise to be right and we saved a lot of time not reinventing the wheel over and over. By the time they had enough knowledge to understand what they were questioning, they stopped questioning it.

          • Ruairc

            I guess “belligerent” would be one way to describe it. I might opt for “frustrated”. What evidence do you want? Citations from the manuals, juxtaposed with video? I can’t criticize such a standard, but I’m also not actually willing to put in that much time to provide evidence for such a minor point – and the point might not even matter.

            (You asked me: what footwork do successful people use? What population should I consider as successful? SCAdians, despite the admittedly different context? Olympians, despite the even-more-different context? If not them, then who?)

            I take your point – that successful fencing suggests that one particular methodology is better. But I still see reason for some doubt, because I can only tell a student “no, you’re wrong” if they’re contradicted by extant historical evidence, or if what they’re doing is blatantly harmful to themselves. Everything else is subject to questioning and degrees of confidence, if only because I am not a Ph.D. in sport science and don’t really know what I’m talking about.

            I am in a situation where a student has, after trying it out my way, nonetheless asserted a predilection for something else. Maybe they’re wrong, and if I let them experiment, they’ll come to realize it. But maybe this is because I don’t have enough expertise to be right. HEMA is still young, and it’d be the height of arrogance to assert that I’ve got it all figured out.

            A central tenet of my classes is to always consider that we might be wrong, to be seeking and open to new ideas, specifically so we don’t end up like the HEMA groups you describe here. I wouldn’t be bringing this question up unless I thought it had some value. If you think my approach devolves too readily into sophistry and pettifogging, we’ll just have to disagree; I’ve seen it work, and I trust this process more than I trust your knowledge.

            The footwork I describe has no additional potential for injury. If you think it does, either you’ve misunderstood, or one of us has a deeply flawed understanding of human anatomy.

          • Dante di Pietro

            The standard I’m asking for is the one you’re holding us to, and over something that is at least comparably settled as true.

            Did you read what Gawin wrote? If so, that’s why I’m bewildered that this conversation is still ongoing. He literally explained the mechanics at play in relative completeness. The only answer is that if your stance is correct and you retreat using your toes, you’re probably giving Dante a clean, clear tempo to take your kneecap. I feel comfortable assuming that the masters optimized, even if that’s not true, so when there’s a full structural breakdown explaining an optimization there’s really no reason to continue discussing it because of “big calves.”

            For reference:

            Gawin
            August 24, 2015 at 4:45 pm · Reply

            I should probably clarify. In the heel method, the single vector is formed on an angle from the hip towards the heel. The leg, while bent, remains rigid.

            If we use our toe, we can accomplish the same mechanics, but only if we keep both our ankle and knee joints rigid. If we do so, we can form a single force vector from our hip to our toe. This is, however, difficult to manage because we now must maintain rigidity through 2 joints rather than one. Furthermore, the ankle is a relatively small joint and without significant strength training, will not actually be able to do this. Finally, the change in angle from the shift to the toes makes it more difficult to keep the knee rigid.

            However, these aren’t the mechanics that you specified and indeed, if we use both the knee and ankle joints in the motion, we’ll encounter several problems. First, we will find that our weight must come forward in order to actually support ourselves using the ball of our foot. If we are holding our guard on the ball of our foot, then we will be slightly more forward, if we must transfer from heel to toe in order to retreat, we add a tempo to the action. Setting that aside, however, we also find that the mechanics underlying how we carry out the motion also change. When we use the knee and ankle, we form a compound lever, which may sound good, but it is important to remember that levers conserve work such that lifting a heavy object requires a trade-off in distance, so while we might consider the use of a compound lever here to be advantageous for helping us amplify the amount of force we are applying, the reality is that each lever provides a multiplicative increase in the distance that must be moved. Furthermore, this system also results in non-linear motion of the relevant parts, as the action of compound levers is to transcribe curved paths. We also run into a problem if we consider the angles at which we are using our levers to apply force into the ground. If we drive through our heels, you will note that we are applying force roughly parralel to the tibia which is near perpendicular with the ground whereas if we drive through the ball of our heel/toes, we are applying force roughly parallel to our metatarsals, which is far closer to parallel with the ground.

            Gawin
            August 24, 2015 at 7:18 pm · Reply

            I spent some more time thinking about how to best explain this to your student.

            Given your description, I see three possibilities for how the motion is being carried out. In one, the fencer starts by flexing the ankle, then by extending the knee. In the second, they start by extending the knee and then flex the ankle. In the third, they sorta try to do both at once.

            Case 1: The fencer is only moving themself with their quadriceps. They are also only indirectly applying force to the ground as the vector of the applied force is first parallel to the femur and is redirected by the bend in the knee (which may also be bad for the knee)

            Case 2: The fencer is only moving themself with the gastrocnemius. The knee extension negates the involvement of hip muscles because they can’t apply force towards the goal of moving the knee. Likewise, because the ankle flexion occurs after the knee extension, the quadriceps can’t provide much help because the ankle joint is the one involved in the final application of force.

            Case 3: If both joints move at the same time, both the quadriceps and gastrocnemius will be contributing effort towards the motion, but neither will be able to do so optimally, as the motion of each joint will be “cushioned” by the bend in the other. In other words, this method doesn’t actually give you anything rigid to push against.

            In contrast, if you drive the retreat by keeping the knee and ankle rigid, the muscles involved in the retreat are larger and stronger and the core, gluts, and quadriceps. Here what matters is the rigidity of the knee and ankle, so in theory you could retreat from the ball of your foot by keeping those joints rigid, however this is more difficult than simply pushing through the heel, so there’s no reason to do so.”

    • Ruairc

      The typical on-guard position in Italian rapier is back-weighted, and the front foot (which we assume is powering the retreat) is already somewhat plantarflexed. This is not the case when running or jumping, and anecdotally it seems to make a significant difference to power generation.

    • Dante di Pietro

      I agree that using the toes would be optimal given a different starting point and purpose.

  • Dante di Pietro

    Oh, one thing: if toes feels better for the student, they’re probably not bending their rear leg to start the recovery. That’s the mistake, and if you make that mistake, toes will work better.

    So toes are OK if you’ve already made an error and need to compensate.

  • Gawin

    Could you make a video of what your student is doing?

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>