Parries are overrated.
They are difficult to execute, unreliable even when executed perfectly, and leave you wide open if anything goes wrong. They cede the initiative to your opponent — by definition, you can only parry after you see the attack — so you always risk being hit first. They are to a sword fight as fortresses are to war: not to be relied upon, even if they are useful in a pinch.
Because parries are useful. Sometimes you will need to parry, or at least, project the threat of being able to parry. Perhaps the opponent has besieged you on your back line, or overtaken you in the open with a surprise attack, or they are relentlessly lobbing attacks at you in the box.
Those are the times to form up, circle the wagons, raise shields. You cannot counterattack or make their attack miss or take over the attack. You have no choice but to parry.
The definition in the FIE Rulebook (English Edition 2019) states that “the parry is the defensive action made with the weapon to prevent an offensive action arriving.”
Translation: If you block an attack with your sabre, you can call it a parry.
Anything goes. Any blade actions that intercept the opponent’s attack. Anything that deflects the attack at least momentarily so the defender has enough time to riposte. Some fencers, like Aron Szilagyi, push this definition to the extremes. Others, like Veniamin Reshetnikov, stop the attack in their tracks (in his case, because he learned to parry on the three-day journeys of the Trans-Siberian railway between Novosibirsk and Moscow, in a cramped cabin with his coach and some rolled up newspapers). The rules don’t care either way.
I find it useful, though, to break down the vast range of possibilities for parries into a few different approaches. There are three ways to “prevent an offensive action arriving” with the blade.
‘Guard parries’, or guard ‘positions’, block the incoming attack by just getting in the way. You guess or see where your opponent is going to hit you on your body, and you put your sabre in front. Think Roman legionaries raising shields against withering arrow fire. Guard parries are fast to execute and cover a lot of target area, but cannot stop really strong attacks.
‘Punch parries’ go out to meet the opponent’s blade with a direct oppositional blow; a punch. Think of it as a shield bash rather than a cut: you take the same position as a guard parry but you slam it into the attack, rather than waiting for it to come to you. Punch parries are stronger than guard parries but cover less area; they are best used in close quarters when the opponent has overtaken you or when you attempt to forward parry.
‘Beat parries’ also move to meet the attack, but with a cut intended to deflect the path of the incoming blade rather than stop it completely. They sweep the space in front of you clear of incoming attacks to all targets, for a moment, unlike other parries which stop attacks to a specific target. Think flak cannons versus missiles. Beat parries work best when the attack is launched from far away, making it weak and easy to track.
Within each of these types, it can be useful to further differentiate between ‘linear parries’ that move in a straight line from start to finish, and ‘circle parries’ that loop around the attack trajectory before intercepting it. The former are faster but cover less target area; the latter are the opposite.
A complete treatment of every type of parry is beyond the scope of what I can cover in this article. Instead, we will focus on guard parries and use them to explore the underlying mechanisms that make parries work.
Guard parries are the most basic and instinctive type of parry. They provide a foundation for the other two types. They are the ones you see depicted by line drawings or photos in fencing textbooks. They are the only ones you depict with static pictures; I suspect this explains why some fencers believe that they are the only type of parry.
There are 8 standard guard positions, labelled numerically from 1-8 (traditionally in old Italian). There is huge variety in how actual, flesh-and-blood fencers actually adopt these positions; partly due to individual body mechanics, and partly due to the influence of what the top athletes in the circuit are doing, or what your country’s national “style” of fencing dictates, or the particular interpretation of the positions used in your club, or your coach’s own idiosyncrasies. Just use a style that works for you.
There isn’t even necessarily agreement as to how many guard positions there are, or which numbers denote which positions, except for the first five which seem more or less universal. One of them is required by convention as the starting position for each exchange: position 3, tierce, which blocks attacks to the sword-arm, flank, and the cheek on that side.
Position 4, quarte, a reflection of position 3 across the body, blocks attacks to the belly, chest, and the other cheek. Position 5, quinte, held horizontally over the head, blocks hits from above. Position 2, seconde, held upside down beneath a raised sword arm, blocks hits to the underarm, low flank, and belly that are too low to be stopped by either the 3 or 4 positions.
Position 1, prime, is rarely used. It is apocryphally named because of its resemblance to the position of the sabre immediately after being drawn from a waist scabbard. It covers much the same target area as position 4, but is stronger against downward diagonal strikes like through-cuts while being weaker against hits to low belly, like barrel-rolls or uppercuts.
The rest of the guard positions — 6, 7, and 8 — are rarely used to defend in sabre, but are used as preparation positions before a parry or hit. Position 6 is like position 3, only forwards and further on the outside. Position 7 puts the blade low across the body. Position 8 puts the blade low on your sword-arm side. There are other, more exotic positions, like the “hanging prime” behind your back, and the “reverse five” above your head which coaches and counter-parry aficionados use. There are as many positions as people can contort their bodies and sabres into.
The positions themselves are only a guide in any case. A good guard parry uses the middle of your sabre blade — halfway between tip and pommel — to catch the opponent’s hit. You only make guard parries in exactly the same position as in the diagrams if your opponent is kind enough to hit to exactly the middle of those positions; anywhere else, and your position will shift.
Guard positions are also designed to flow into each other without clear distinctions between them: parry 1 to parry 5 to parry 3; “reverse five” to parry 4 to parry 2. Is a diagonal parry against a diagonal cut to your neck a parry 3 or parry 5? Who cares, as long as you block the cut.
Guard parries are fast. Their motions are simple. Once you know, or guess, where the opponent wants to hit you on your body, put the middle of your sabre in front of it. Angle your sabre so that it is perpendicular to the path of the incoming hit. Now you have error bars on either side of your guess. Angle your sabre so that its front edge will make impact against the attack. This braces the sabre against your entire body, via your thumb and arm.
Put as much distance between your sabre and your body as you can. The more distance, the less likely that your opponent will crush through or whip around your parry. Put as much distance between you and your opponent as you can. There is no shame in making their attack miss.
You can use whatever footwork you like to do so: retreats, jump backs, and crossover retreats. Guard parries work especially well with the crossover retreat. I think of this combination as being fundamentally a fall-short in which the guard parry has been added as an insurance policy.
Guard parries are the only type of parry that works well with the fall-short. They don’t require you to be braced against the ground, unlike punch and beat parries, so you can use them while off-balance in the fall-short. They are fast, so you can wait until the last moment to move into position. This makes it harder for your opponent land feint attacks; good defenders can make multiple guard parries in quick succession.
Against slow or remising opponents, you can even see where their attack will hit before you react with the appropriate parry, rather than having to guess and setup. There are tricks to help you do so, using your peripheral vision and choosing between parries that flow together like 1, 3, 5 rather than between ones that don’t, like 2 and 3. Practicing these reactive parries is the bread-and-butter of most one-on-one sabre lessons.
Against strong opponents though, it is not possible to parry by pure reaction no matter how fast you are. The Paladin illustrated this with a drill he picked up from the Maestro: two fencers, attacker and defender, standing at lunge distance apart. The attacker tells the defender where they will hit, then does so when they feel ready. If done correctly, the defender can never parry in time.
We dubbed the separation distance the “PONE” — the point of no escape — because the defender couldn’t even pull away and parry. The hit would travel to the target in less time than it takes for a human to react. And that was when they already knew where the attack would hit, what path it would take, and were primed to parry.
But of course in real life, people parry all the time. Sometimes it is because the attacker attempts to hit from too far away, or they slam into a prepared guard position, or they telegraph their hits ahead of time — if you know where and when a hit will land, parrying it is easy.
That’s how parries work the rest of the time, when the defender tricks the attacker into hitting into a target and at a time of the defender’s choosing.
Parries work best in the box. There, you have a good chance of guessing when your opponent wants to attack and where they want to hit. They don’t have too many choices to pick from. If they do want to attack — as opposed to chasing you down — they have to attack fast, at least as fast they think you can attack, and usually in their favourite rhythm. They usually have a favourite target they like to hit as well.
So a parrying defender just has to observe their opponent, pick their favourite rhythm and target, then rush forwards in a fake long attack while exposing that target. Then when the defender sees a hint of an attack, they can parry. It doesn’t work every time, but works far more often than not.
Parries are less successful and more difficult to set up outside of the box. While you can be confident that the attacker will eventually attack, they have much more leeway in when they actually do so. Priority also allows them to get much closer and hit along a more confusing path than they could get away with in the box. They often don’t even have a target in mind; classically-trained fencers are trained to hit “open-eyes” to whatever target they see open at the last possible moment, at the PONE distance.
The defender must do three things to succeed with the parry. They must prevent the attacker from getting to PONE distance. They must entice the attacker into hitting into a target of the defender’s choosing. And, finally, they must compel the attacker into hitting when the defender wants.
Preventing the attacker from getting to the PONE distance is just a case of keeping away far enough and using beats to hold them away.
The beats also serve to close off some targets while exposing others. Cunning defenders restrict their beats to certain areas while keeping a single obvious target open. Witness them clearing a plane near the attacker’s eyes, while leaving their underarm exposed for a parry 2, or sweeping clean the space in front of their belly while leaving their sword arm side curiously unprotected for a parry 3. Such deceptions work best against “open-eyes” attackers, but they work well enough against attackers who have favourite targets as well, provided you know what they are.
Beats alone are often sufficient to compel the attacker to attack — they get nervous every time a beat comes close to taking priority away. Some attackers, though, require additional motivation. Fake counterattacks work well. Real stop cuts work better. Just remember to parry afterwards.
Of course, if the defender has the space and time to keep distance and set up traps, they should be attempting safer actions like the fall-short or beat rather than the parry. Only when those don’t work should the defender be thinking of using parries, and then as a supplement to another action like the fall-short or stop-cut, rather than a goal unto itself.
Then there are the occasions where the defender has no other options but the parry. Such occasions rarely provide the defender with any opportunity to set a trap. In these moments, the defender just has to guess where and when the hit will come, and put their shield up.