In the world of historical fencing, and particularly the fascinating field of research, we sometimes face scholars who express less well-founded hypotheses on certain topics. The question of parrying with the flat instead of the edge, for example, is a recurring topic that causes somewhat curious and, at the same time, interesting debates.

Our  thoughts can be summarized as follows: the parry “par excellence”, as an action seeking to stop the opponent’s blow, done in a way that does not lead to uncertainties in the control of the blade and the point, must necessarily be made with the edge. The flat of the sword may indeed, only in certain situations, be used to deflect a blow or a thrust.

Furthermore, it is a fact that all the Italian treatises from the Middle Ages to today teach only parring with the edge. Only in certain part of the German treatises, after the thirteenth century, appear parries with the flat of the blade, and only executed in particular circumstances.

A clarification, with the help of modern fencing terminology (why re-invent it?): the “flat parry” uses the flat of the blade, NOT “the edge at a certain angle”. Just like it is called “edge parry” not only when the blades clash at a straight 90 degree angle.



One of the largest and most influential supporters of the flat parry is John Clements, Director of the U.S. Association ARMA. The author of interesting articles, essays and videos, defines the parry as ”the deflecting or deviating of the opponent’s blade before it reaches its target”.

Clements’ definition is missing of the most important sense: blocking the opponent’s strike before it reaches the target. Block and deflect (or deviate) are two very different actions, that generate very different results.

Still, even in English texts the differences  between the two types of actions are clearly explained: the terms parry and deflect are often translated with different words, i.e. “to parry” and “to deflect”, which is clear in the treatises on fencing in any historical period.

It can also be said that in English texts technical terms of fencing have often been italianised. For example “to parere” in Pallas Armata (1639, Book I, Cap V):

“To Parere, is to decline, to put by, and to turn off a thrust or blow”

Surely, with this definition taken out of context of earlier and contemporary fencing, we can easily fall into the error of excessive simplification of an action.
Moreover, Alfred Hutton in his Old Swordplay (1892), when referring to the Opera Nova of Marozzo describes the parries of the latter with the generic word “parries” reclassifying in terms of numbers and adopted in full use.

This kind of simplification seems to be the age-old question of the origin of edge or flat parry, which we will attempt to resolve here.

Keeping in mind that the dodge is probably the best way to avoid of being hit, the parry tends to be preferred (as more conservative) when you intend to stay “a misura” to continue the strike action with a response. However, we have to specify what is the meaning of “block”.

This term describes an action made with the weapon to intercept the trajectory of the opponent’s blade thereby nullifing its kinetic energy. The alternative to the block is the “deviation”: the difference lies in the fact that this action is not meant to stop the opponent’s weapon, interrupting its motion, but to divert it by intercepting the path.

In those circumstances it is quite easy to prove the fallacy of the contention that the parry usually takes place with the flat of the blade rather than with the edge. The arguments against the flat parry, and in support of the edge parry (in its general sense) are numerous:

• The biomechanics of the human body structure
• Morphology and chemical composition of the weapon
• Execution of fencing actions that follow a blade binding
• The informations provided by fencing Masters of the past in their treaties.

Let’s proceed with the examination of these aspects (make yourselves comfortable).



Perhaps this is the more intuitive motivation: in the case of a parry, the impact on flat surface causes a strain on the hand and forearm naturally unfavorable to the kinetic chain arm. When you hit with a fist, the kinetic chain expresses the maximum potential of the arm when the shot is taken with the knuckles, or the side of the hand, as that would most effectively with minimal effort.
It ‘s simple to try this by hitting a punching bag, or a nail with a hammer.

For the same reason, an edge parry follows a more natural movement of the body, ensuring the use of greater strength and muscular endurance at the impact. At the same time you get a better control of weapon point compared to the same action with the flat.

As proof, hold in hand a stick and ask a friend to push with a constant pressure (or to hit, also!) on it’s middle, towards his left side . You will notice that according to the different positions of your fist (flat parry or edge parry) you’ll be able to put a different strength opposite to that pressure. It’s natural and the difficulty to resist this kind of pressure between the two positions is verifiable.

Giovanni Antonio Lovino confirms this exposition:

Ad XXXV – Ascanio Attore, Lauso Reo
“…Percioche il suo pugnale si trovò haver poca forza, nel portar fuora la pugnalata di Ascanio: si per la tardezza, con la quale esso andò alla parata; si perche fece la parata col piatto del pugnale: come si è già detto.”



Some scholars and practitioners argue that the flat block prevents the edge of being damaged. As answer, we can say that at same time our edge risks being ruined when an opponent blocks our sword with his own blade, or when we hit a steel armor or a shield.

Filippo Vadi provides an indication about the morphology of the two-handed sword: the Pisan master affirms that the blade sharpening (“razor like”) is long only four fingers close to the point. Now, it is natural that this statement should not be taken literally, since it is clear the futility of a blade longer than a meter that is effective only at the tip: you might as well use a spear. A more rational interpretation is that the blade is sharpened only for the last four fingers because while using the remainder of the edge tends to wear out just because of mutual parries.

Since we hit with the medium-weak of the blade and we parry with the strong, it follows that, although the edge at the strong can be ruined, the sword does not become less effective because the medium-weak will continue to stay sharp.
It should also be said that the dents found on various museum exhibits, mostly found on the strong side of the edge, are very mild and there is no doubt at all that they do not compromise the integrity or the optimal use of the blade.

It ‘s obvious that, in case of dueling (never forget that almost all the treaties we have are about duel fencing) with the sword, the least concern of a duelist was to avoid marks on the blade. The number of blows taken in a duel are hardly enough to affect the sharpening negatively. For example, Fiore de’Liberi in the first game (primo zogho) does a parry “in ponta de spada” (on the point of the sword), with the weak: Fiore does not seem to worry too much about ruining the edge of his blade.

When the sword is used together with the shield, the shield is supposed to protect against the blows, preserving the edge of the sword automatically.

Another essential factor is the hilt (in Italian: elsa or elso) of the sword. The cross-hilt swords of medieval and Renaissance were obviously intended to protect the hand. Already in 1460 King René clearly states it in his famous book on tournaments.

In this quick review of tasks of the hilt, we would like to say a few words to express some concerns on the affirmation of John Clements: according to him the hilt doesn’t protect the hand from the blade of the opponent, but protect fingers from knocking against the opponent’s shield. Specifically, in his interesting Medieval Swordmanship (Paladin Press, 1998), the U.S. researcher says that if the cross guards were really used to avoid the fingers to be hit by the blade, the museum finds should exhibit scratches on the cross because of the impact of the opponent’s blade. It is not obviously so, since the cross guard is not used to direct parry, but only to protect fingers from slipping or a fortuitous bounce of the opponent’s blade (altough it is still relatively uncommon).

John Clements goes on claiming that “the function of wide Medieval cross-guards seems to have been for preventing the user’s hand from slamming into or hitting against the flat of an opponent’s shield.”

This step, thus exposed, may have been misinterpreted by scholars and lovers of historical fencing that formed part of their cultural heritage on his books. Clements’ assertion is, according to me, a mere speculation as we can find no blows delivered with the flat of the shield, or with the boss (except for bucklers with central point) in any fencing treatise. So it is not reasonable to assume that a fencer would normally execute such blows (Do not forget that the shield is a defensive weapon by definition).
The hand could impact against the shield of the opponent, and, in this case, the cross guard is a valuable, but it is wrong to say that this is the primary purpose of this morphology of hilt. If so, we wonder why the Roman legionnaires, who for hundreds of years also used the sword and shield and faced adversaries equipped with a shield, never used the cross guard to protect the fingers from impact against the shields. And yet, surprisingly, the cross has remained in vogue until the XVIII century, when the shield, if not obsolete, had a very marginal use for almost two centuries.

If there is still any doubt, we could seek an authoritative answer in the words of the past Masters. We quote here the words of Docciolini. He specifies in the section on techniques of sword and dagger:
“…e quanto al modo del tenere il pugnale in mano dico che coloro che lo tengono di piatto, con appoggiare il dito grosso nel mezzo del detto pugnale, e si fondono d’andare à parare un taglio che vadia al capo, fanno grande errore; […] il taglio che verrà à correre […] anco che verrà a reggerlo, poniamo che la lo regga, porta grande pericolo, che nel medesimo modo detto di sopra, rimbalzando gli mozzi la mano…”.

In few words, during the parry the opponent’s weapon may accidentally slip or bounce down to the fingers, and the hilt is designed precisely to stop this drift. Parring with the flat means letting the hilt be totally useless, since, it is oriented in a direction not suitable to protect the hand.

Finally, the hilt is often used immediately after the parry to move the opponent’s blade:
Docciolini pg 52 – delle due spade (cap 14)
“…se nò voi venite alla parata, e retto che voi harete il detto taglio, voglio che voi alziate alquanto il vostro pugno destro, che verrete con i fornimenti a mandar la spada avversaria, inverso la vostra parte destra…”

The correct execution of this action described by Docciolini would be impossible from a flat parry.



Francesco Foglia, graduated in industrial chemistry at the University of Rome La Sapienza, who certainly has an authoritative voice on this topic, affirms:

“In consideration of practical use regarding the operation of instruments and tools can be received with the analysis of a few but important structural considerations. First, we briefly illustrate some concepts that will make clear the following considerations. A metal is a solid made up of an internal structure approximated to that of a compact set of all spheres packed together to form a continuous visible macroscopically as a solid object. Imagine a box filled with marbles all the same size, the greater the amount of balls at the same box, the higher the density the same for the same volume. Essentially the mechanical properties of a metal are first and very simplified approximation to this density-dependent. Now imagine you insert inside the same box smaller balls.What will happen is that packing, the density at constant volume, will be increased. Steel differs of the iron for this simple feature, as in the case of box full of balls, the carbon, smaller than iron, is inserted in the mesh formed by iron atoms and increases the mechanical properties due to the “newness” of the addition. The mesh or grid to which reference was made to the microscopic structure of the metal is considered by us. There must be a starting point of the process till a fracture generates itself within the lattice, which develops a macroscopic lesion of the metal. These points are defined defects of the lattice, and are naturally present in a metal. The process of forging, tempering and working of a metal have no other purpose, if not the elimination of many defects as possible, and stabilizing the structure of the metal lattice.

We come now to practical considerations regarding the use of a metal tool as it can be a sword. The mechanical stress to which the tool will be subject to all borne by the aforementioned network, the more the network will have been cured, the more the tool will respond well to the conditions of use, avoiding any unpleasant accidents such as breakage. Some of the most famous physics can be analyzed by a microscopic point of view:

The hardness is the resistance of the wire mesh provides the dislocation of its constituent points, when another body tries to penetrate it.

The elasticity is the ability of the wire mesh to keep the energy of a solicitation, such as a bump, and return very quickly.

The bending resistance is the resistance offered by constituents from the lattice to exchange places with the other, this property together with the ductility is important to prevent such a metal bend. This property, unlike others such as hardness, which are intrinsic to the material, depend on the amount of metal in question. In this case the flexural strength depends on the thickness of such material to be deformed. It is easy to understand from these few hints that the problem of optimization of a metal artifact is not trivial, and that many of the processes are specifically designed for the difference in use that makes the object in question. In particular, in tha case of a sword we have an attitude critically new examining the features of the object for own working. It is easy to see that the hardness of the metal will be an important factor as regards the edge and the pont, whose aim is precisely to overcome the resistance of targets and unleash the power sets the instrument to penetrate the structure.

The elasticity will be especially important for handling: in order to understand this point, let’s image of wielding a sword that responds to an blow with noticeable vibration or impact dissipates part of the force imposed. For flexibility it can seem trivial, but certainly a fencer will not look kindly on the fact that his weapon is deformed each time you come in contact! In the specific problem of the parry by sword against a blow led by a similar tool so you understand that among the pitfalls there are for sure: the breaking, bending deformation and the development of vibrations that affect the hand of wielder. The thickness of the tool is a parameter to be reckoned with in the evaluation of the flexion, and certainly the section cut will be greater than the flat section. Still, about of the hardness we have to consider that in addition to the intrinsic characteristics of the metal come into play during a collision geometric factors. The force impact is to be distributed on the surface of contact between the tools: Who will be benefited with the same force then distributes it on a smaller area. There should be no doubt that the edge of a sword has smaller surface than flat!

Finally we must consider that, with the same elasticity, a stress on the plate induces vibrations on the body of the object greater than on the edge. The reasons are to be found not only in different section solicited, but also in the mechanism by which stress the body travel through the material.The field of forces acting on the sword that receives a blow on the plate is larger than one that responds with the cut, this means that even remote areas are solicited from the point of contact and consequently will be more likely to affect areas with structural defects mentioned above and incur therefore effects on the structural integrity of the material.”

We want to emphasize the importance of this passage, clearly explained by Dr. Foglia, as surely the key-point of the matter on the molecular composition of the steel.

We are proud to present an additional observation of a mechanical nature which has been kindly provided by Prof. Francesco Cirillo, a former fencer, and Professor of Machine Design, Department of Mechanical and Aeronautical Engineering of the Faculty of Engineering at the University “La Sapienza” of Rome:

“Why in fencing is mainly used parrying by cut side and not with the flat of the sword? A purely mechanical response – and it is not the only possible – is the need to minimize the deformation of the weapon and maintain the highest possible frequencies of vibration caused by the impulse induced by collision. These demands stem both from best exercise the art of fencing. While we fight it is imperative to strive for the highest possible speed and accuracy in the execution of orders processed by the brain about the position of the weapon in space. For example: the position of the sword to parry an enemy’s blow. This performance is essentially composed by the determination and decision to place the iron (neurological reflex) and the change of position by moving the weapon sets and rotations of the handle: that is, through the forces that a fencer discharges on the weapon discharged through his arm and his wrist (implementation muscle). From the mechanical point of view, a sword is essentially a cantilevered beam, stuck at one end within a joint (the wrist of the fighter) that moves rapidly in space, with translational and rotational accelerations very high. It has been measured to the tip of a sword Olympic 1100 mm in length, during a match you can reach accelerations equal to 5 G (excluding those due to collisions with the opponent’s tool). Due to the sudden change in position set by the handle, the blade must therefore react to forces of inertia (which are opposed to the change in momentum) of relevant entities.

These forces produce a distortion in the opposite direction to the advancement of motion, that is, to find that the blade in a “backward” or “late” compared to the theoretical position that the blade itself would be in the absence of deformation. The deformation by forces of inertia, therefore, disturbs the action of the fencer’s sword because it does not reach exactly the theoretical position elaborated by the commands of the brain, in this way it reaches that position with a delay that increases with the deformation of the blade. Therefore, for the fencer is preferable the least possible distortion.

Suppose that the section of the sword be a diamond blade very flat, ie with a diagonal lot less than the other. At the same load, the deformation in the plane containing the major diagonal, with load applied on the cutting edge, is much smaller than the deformation that occurs in the plane containing the diagonal lower as the same force is applied to the plate of the weapon. This is due to the fact that the deformation is inversely proportional to the moment of inertia around the neutral axis; in the case of edge load on the Moment of Inertia Jx is much greater than Jy which occurs in the opposite case. In short, the “stiffness” of the sword is directly proportional to the moment of inertia of the section that is opposed to its deformation and is much greater if the weapon has to react to an applied force on the cutting edge.

In the parry, the blow of the sword against the opponent’s blade produces a solicitation of an impulsive nature that sets off vibrations in the weapon itself. The frequencies and amplitudes of these vibrations are complex, however it can break down into the sum of sinusoidal vibrations easily calculable and measurable. Substantial component of these vibrations is the “own way” how blade vibrates subject to impulse, characterized by a “own frequency” and “own wideness”. At equal impulse (exciting energy) higher is the frequency and lower will be the amplitude of vibration. In addition, for the same “exciting energy”, the frequency of vibration is higher if there is a greater “stiffness” of the structure, that is, if there is a greater Moment of Inertia of the section that is opposed to its deformation. Therefore, when we parry with the edge side, the section which opposes the deformation is characterized by increased Moment of Inertia (Jx), which causes vibrations of higher frequency than the parade with the flat side.

At same energy, higher frequency means a lower amplitude and therefore less uncontrolled movements of the blade and of the wrist holding the weapon. It ‘s obvious that even from this point of view it is preferable to parry with the edge instead of the flat of the sword.”

We are sure that the last statement be crucial and indisputable. In other words: it is wrong to claim (as someone does) that flat parry allows the stroke to discharge on the flat side because it’s the flexing part amortize the blow itself: if it is true that every action has an equal and opposite reaction, the spring action of the blade, which bends during the blow and then stretches, produces a rebound, a shift of the sword in the same direction of the blow. In practical terms the result is similar to the fencing action called “botta sul ferro”.



Riccardo Rudilosso, FIS National Instructor and fencing practicer since 2 decades, has an important affirmation to add:

“Even in the fencing sport, with weapons that reach the maximum weight of 800 grams, the flat is never used to deflect or parry. However, in case of further doubts, I have to add the following consideration: the terminology of modern fencing “parata di tasto” identifies a “soft” parry, used for avoid to bouncing off the opponent’s weapon (which leads the “parata di picco”) and cushioning the blow keeping the contact of both blades. With the gained bind it’s possible take advantage to perform actions on the blade’s enemy. The Sforzo is one among these: immediately after the parata di tasto, virtually the only movement of the wrist exerts a strong pushing on the blade of the opponent forward and down. In this way the opponent loses for a fraction of a second the control on the weapon. The Trasporto is another one: immediately after the parata di tasto (such as a parade of 4^) with medium-strong blade I imprint on his weak-medium side a pressure to the blade in order to “carry” to a diametrically opposed location (for example of 2 ^ block position) without interrupting the contact between the two blades. In short, if the parry was made with the flat, it would be absolutely impossible to correctly perform these actions.”



Every lover of historical fencing will have matched his studies with illustrations of medieval Italians and Germans fencing treatises, which are notoriously devoid of perspective. In particular, the sword is always shown with the flat facing the reader. It is not reasonable to assume that the parries were carried out with the flat because otherwise, for the same reason, we could believe that even blows were made with the flat.
However, some German treatises are here presented as examples to support the parade with the flat, and, in addition of certain pictures, could be interpreted in this sense, but very rarely specify which part of the blade should be used to block.

In Talhoffer’s and Dürer’s works, we can easily identify some illustrations of techniques with the Messer that unequivocally describe a parry/deflection with the flat. These techniques are also peculiar because the parry is carried out so as to turn the edge of Messer upward. However these techniques appear to be isolated cases, they seem to be exceptions, not the main rule.
In addition, we have some instruction by other German Masters:

As an example we may quote the last sentence on folio 32v of MS 3227a manuscript (1389), also known as the Hanko Döbringer:

  • “And as you turn aside [Abewendet] all strikes and thrusts with the foremost edge, that is also displacing [Vorsetczen].”

For completeness, we refer to another German author, Sigmund Ringeck: he affirms to parry the enemy’s blow with a “half-hit” against the blow itself. Not much different from what found in Fiore de ‘Liberi and Filippo Vadi: both respond to a blow with a similar one, to impact the opponent’s weapon (incrosar) and eventually proceed to a binding or a response. Surely this action is a parry, and, since it is a half-stroke, it is performed with the edge of the sword and not with the flat.

Countless explanations about the way to make a parry can be found in Renaissance treatises, of which we present here only a few brief but important excerpts:

Achille Marozzo – Opera Nova (1536):

  • “…tragandote lui tal mandritto tu el parerai in su el dritto fillo del pugnale tuo…”
  • “…perché tragando lui mandritto tondo, o fendente, o stocata, o ponta, voglio che tu pari queste botte in fil de spada in atto de guardia de intrare…”

Giovanni Dalle Agocchie – Dell’arte di Scrimia (1572):

  • In 13v: “…in due modi la spada si può parar, ò co’l fil dritto di essa, ò co’l falso…”

Bolognese Anonimous – (sec. XVI):

  • “…et se egli si argomentasse di ferirti di colpo alcuno, tu passarai del piede manco innanzi schifandoti con il dritto filo della spada dal vegnente colpo in guisa propria di guardia di testa…”

Francesco Altoni – Monomachia (sec. XVI):

  • Il giuoco della spada sola – inizio pag 69 verso: “Li colpi di taglio come si debbon fare. Insegnare la parte della spada qual’è da parare, e qual’è da ferire, perché assai importa come il filo buono del mezzo inverso la punta, serve a ferire, ed al mezz’indietro serve a parare. Dire quante sono le difese, che sono secondo me tre, col filo overo fuggita, con l’offesa, e con l’arme.”

Giovanni Antonio Lovino – intorno alla Practica e Theorica… (sec. XVI):

  • Prima fermata di spada: “…voltando sempre il filo buono, contra la spada di quello; ci tiene sicuramente difesi della offesa nimica.”
  • Fine 73 recto: “Overo se la spada nimica sarà fuora di tempo, che vostra signoria non sarà tenuta à ricercare quella spada; ma la basterà solamente serrarlo, come si è detto dinanzi: voltando Però sempre il filo buono, contra la spada contraria; se ben ciò non fosse in tempo; accioche venendosi al ferire, trovi il buon contrasto del filo buono della (73v) Spada. Però che cosi facendo si può con facilità serrare il nimico, e investirlo di punta.”
  • inizio 74 verso: “Vi aggiungo (74v) questo avertimento che convien tenere le braccia ben distese; la spada giusta: col pugno, e il piatto della spada verso terra; accioche il filo buono resti contra l’offesa della spada nimica.”

Federico Ghisliero – Regole di molti cavagliereschi esercitii (1587):

  • Della Theorica pag 31 parlando delle imperfezioni del colpir di taglio: “…la spada per non esser sferica, & perciò di peso non ugualmente grave, molte volte resistendovi l’aere ferisce di piatto: & la spada è soggetta a rompersi.”
  • Della Theorica pag 82 sui 4 modi di parare: “Nel terzo modo di parare si resiste a tutte le ferite col filo dritto; & se bene si potrebbe anchora col falso, nondimeno per esser questa parata debole la tralasciammo: & e mettendosi la spada in linea retta si para, & si ferisce facendosi scorrere i fili della spade…”

Marco Docciolini – Trattato in materia di scherma (1601):

  • Di spada, e pugnale (cap 15): “…tiri l’avversario dove gli piacerà […] non havete da far altro, che presentargli il taglio del detto pugnale…”
  • “…ma poniamo che egli ve la metta (la punta) con il filo dritto, e voi havete à presentare il vostro filo dritto al suo…”

Ridolfo Capoferro – Gran simulacro dell’arte e dell’uso della scherma (1610):

  • Delle Parate: “Si para tanto di filo dritto quanto di fil falso, ben che rade volte, così in linea dritta come in linea obliqua,…”

Erhardus Henning – Kurtze jedoch gründliche Unterrichtung vom Hieb-fechten (1658):

  • Vom Pariren“Positio prima. Alle Hiebe müssen mit der Schneide pariret werden / ratio, weil / wenn man mit der Fläche pariret leichtlich einem die parade weggehauen werden / und alsdann einen Streich erlangen kan.”



Finally, the point is: flat parry does exists, but it has not to be taken as general rule of parrying (as many Clement’s unwise emulators and followers are inclined to do), expanding a priori a concept restricted to a particular historical period, to a particular kind of fencing and to particular fencing actions.