In an attempt to illustrate further how one might puzzle out difficult passages I’m including another question that Christian Olbrich (Fechten Passau) and I discussed. This is the “jump back” or salto indietro covered by so many Radaellian masters (cf. p. 169 Rossi, section 43). This is footwork somewhat larger than the standard retreat and can be confusing not only because of how some authors describe it, but also because it’s often discussed in conjunction with various types of double-measure (doppia misura) and the cross-step (passo doppio indietro).
The question we discussed was:
The jump back in the stop hit (cut) to the arm: Barbasetti describes the first movement in the jump back as “bending the body backward, throwing back the head also”. Rossi, and a few other sources describe the cut as being done with the setting back of the right foot, and any parry following with the re-setting of the left foot. But if I’m in a distance where my opponent could, with a lunge, just hit my head, then crossing the right foot backwards will effectively break measure enough to get me out of reach even to his arm, and if I don’t throw back my head but leave it where it is, I will have my weight shifted dramatically forward so that any farther backward movement is either impossible or very, very slow. I also cannot expect a lunge of my opponent to carry his arm towards me, since I would mainly try the stop hit into my opponent’s feint. Any thoughts on the correct execution here?
There is a lot to consider in this one. When do we attack the arm normally? Only when it’s exposed. So, perhaps on a feint, or, against any clumsy attack where the opponent somehow exposes their arm or starts out of distance and holds the attack long enough for us to attempt the counter. I look at this two ways: if we are in critical/lunge distance, and they attack in such a way that we can make a stop-cut, we might need to use the jump-back, but if they are in advance-lunge distance, then we may need only retreat a step or half-step. Naturally we are not going to try a stop-cut against a decent attack or closed-line. 
Barbasetti probably should have started with the last clause in his section 12:
[German] Diese Sprungbewegung ist unerlässlich, sobald es sich darum handelt, Hiebe gegen den Vorderarm des Gegners zu führen, während sich derselbe gerade zu einem Hiebe anschickt. [p. 39 in Das Säbelfecthen, 1899]
[English] Nevertheless, this action is very useful. It enables you to touch your opponent in the arm at the start of his attack. [p. 18 in the 1936 English edition]
The German edition explains this better to my mind—“useful” is not the same as “essential/unerlässlich.” It IS essential to use the jump-back IF they are faster than we are on the attack or have longer reach.
[English] 12. The Jump Backward
It’s not just an issue of translation, but choice of description. In writing “bending the body backward, throwing back the head also” the master confuses things. The passage in full reads:
The jump backward is executed as follows:
Bend the body backward, throwing back the head also, while passing the right foot behind the left one to a distance of about 20 inches; finally retaking the guard, crouching upon your legs.
This exercise requires a great deal of study. In order to execute it successfully, it is necessary to repeat it until the legs are trained and the body accustomed to maintain a perfect equilibrium.
Nevertheless, this action is very useful. It enables you to touch your opponent in the arm at the start of his attack. 
[German] 12. Sprung rückwärts.
Der Sprung rückwärts wird auf folgende Weise vollzogen: Der Körper wird zunächst dadurch nach rückwärts gebracht, dass der Kopf mit Energie in den Nacken geworfen wird. Zu gleicher Zeit wird der rechte Fuss mindestens 50 Centimeter hinter den linken gestellt, schliesslich mit einem Sprung die Fechtstellung mit gebeugten Knien wieder eingenommen.
Es ist nothwendig, diesen Sprung oft zu üben, um den Beinen die nöthige Kraft und Sprungfertigkeit, dem Körper aber das Vermögen zu verleihen, das Gleichgewicht zu erhalten.
Diese Sprungbewegung ist unerlässlich, sobald es sich darum handelt, Hiebe gegen den Vorderarm des Gegners zu führen, während sich derselbe gerade zu einem Hiebe anschickt. [38-39]
Here again the German reads more usefully. It reads “the body (Der Körper) is first brought backward (wird… rückwärts gebracht).” “Brought” implies movement different than English “bend.” With reference to moving the head the language is much the same—“throwing the head back” conveys much the same meaning as having the head (der Kopf) thrown back (Nacken geworfen wird) with energy (mit Energie).  Taken together the sense is that one is shifting the weight backward as one strikes to the arm, but in this case one makes a short jump (einem Sprung) instead of merely taking a step back. This is important to note as lifting the right leg back past the left need not be a jump; it can also be a cross-step, a variety of retreat. Jump connotes speed and urgency, both critical considerations in effecting a counter-attack made while moving backwards.
Barbasetti does not cover double-measure per se. In later sections, where he covers the stop-hit and cut to the arm, he again refers to the jump-back (cf. Sections 60-62, pp.101-109; pp. 126-131 in Das Säbelfecthen). This said, the advance-lunge and rapid advances are present—in fact, he remarks that “in sabre fencing the adversaries keep out of reach, and therefore the advance before the lunge is almost a normal condition” (p. 70; so far as I can tell, this line is an addition to the English text or perhaps from another section in Das Säbelfecthen).
Other Radaellian masters do refer specifically to double-measure. Rossi discusses the cross-step back and jump back in the same section (43 Passo doppio indietro e salto indietro). The difference is in how the feet move and how fast. The cross-step he recommends has the right foot brought past the left at about 40 cm (compare Barbasetti’s 50cm), then left back past the right to reassume guard. The jump-back, on the other hand, follows the same pattern but has both feet leave the ground, is executed more quickly, and is used against a fast attack. Rossi explains that advance-lunge and double retreats reflect a similar measure. He writes:
La doppia misura pei colpi diretti al corpo è la misura giusta pei colpi diretti al braccio. Bisogna quindi, per la doppia misura al braccio, un altro passo indietro, come si vedrà dale linee del piano regolatore. [149-150 in Rossi’s Manuale Teorico-Pratico per la Scherma di Spada e Sciabola]
Double measure [his advance-lunge] for blows directed at the body is the correct measure for blows directed at the arm. For double measure to the arm, another retreat is necessary, as seen by the lines of the marked piste. [p. 161 in Seager’s translation]
This is a good rule, as it keeps one safe, but again reflects the reality of fencing when the duel was still a reality. Modern fencing, in contrast, breaks this down into two measures, a stop-cut followed by a simple retreat and a stop-cut followed by what Rossi describes.
The goal, in either case, is to create more space to act. Of note, Rossi advises that “In all the marches, make sure to deviate the body as little as possible from the guard position” (In tutte le Marcie si osservi, per regola, di scomporre il meno possible il corpo dalla posizione di guardia, 169; Seager, 179). I believe that Rossi and Barbasetti, despite appearances, agree—however described, both are referring to a controlled, rapid retreat to allow one space and time to achieve a counter-attack to the arm. Barbasetti’s description of throwing one’s head and body back I understand, but taken literally it would have one move in a very unbalanced way. The sense of it is that one reaches to cut while simultaneously preparing to move backward.
Measure, however, is only one consideration: we’re also making the stop-cut in a particular tempo. For example, with the more modern stop-cut, we often describe it as reaching out to cut as the opponent starts their attack, and then immediately taking a step back to parry and riposte. Assuming an open line to the attacker’s arm, we are making an attack into their tempo to do this and as such must regain measure to remain safe (if we hit them, great, but we must also be prepared should our counter fail or not stop the opponent).
In terms of measure with the example I just gave, we’re reaching out to hit that arm at lunging distance/critical measure, not from close or advance-lunge measure. To maintain that safe distance we stop-cut (in lunge distance) and then recover backward as they come at us to keep that same lunging distance. If we don’t, we’ll be in close measure and far less capable of acting just as you explain. Measure and time are bedmates—they’re intimately associated. As a final note, we do not normally cross the legs to do this, we just retreat.
The jump-back is a little bit larger a movement in my mind, less a step than almost a ballestra in reverse (unless one is making the jump-back by crossing the legs). It’s still made in the same tempo, that is, into the opponent’s attack, but for whatever reason requires us to take more distance on the retreat (Rossi is clearer on this point). Passing backward, so right foot passing past the left and then resetting into guard, can cover more ground securely—as I conceive of it, it’s the difference between making two retreats and one cross-step back. So much of this depends on one’s reach, timing, and the opponent’s ability to navigate distance as well as your own: at times a simple half-step might suffice, at others a jump-back is necessary.
The only time I normally use the jump-back is against opponents much faster or taller than I am. I’m around 6’/1.8m tall, but often fight people 6’4-5”/1.95-1.99m tall. Their reach is such that I can make the stop-cut well-enough, but unless I jump back I will not be able to maintain distance well enough to be safe. If I use the jump-back as described against someone who doesn’t have that reach, then I end up out of measure. Sure, I’m safe that way, but we’ve just reset. Better to attempt the counter-attack and move just enough to cover and riposte.
Most masters are quick to say that one needs to practice these maneuvers seriously. They require a lot of coordination, a keen sense of measure and timing, and can go spectacularly wrong if we get any one aspect incorrect. I know footwork is not everyone’s favorite thing to do, but it’s the foundation for everything. To develop effective use of actions like the jump back, practice it; include it in the mix of advances, retreats, advance-lunges, reverse-lunges, and cross-steps. This will give the mechanical aspects exercise, but you will need a partner to hone your stop-cut, and, perfect how to make this important counter-attack in various tempi.
 Like many rules, we follow them until we discover appropriate times not to. There are instances in which one might attack a strong position as a species of second-intention attack. This is a complicated topic and too much to go into detail here. Because there are exceptions, I wanted to say so.
 The German supplies in den Nacken, “in/with the Neck,” which makes sense, but which in English reads less eloquently. The sense is that one moves the head back to prompt the body backward. This may also reflect the natural void one no doubt made with certain actions to avoid being hit in the face. We see similar shifts of the head, often to the side, in works like that of Girard (1740).
Cav. Luigi Barbasetti, Das Säbefechten, übersetzt von K.u.K. Linienschiffs-Lieutenant Rudolf Brosch und Oberlieutenant Heinrich Tenner (Wien: Verlag der “Allegemeinen Sport-Zeitung,” 1899.
Luigi Barbasetti, The Art of the Sabre and Epée (New York, NY: E.P. Dutton & Co., Inc., 1936).
Giordano Rossi, Manuale Teorico-Practico per la Scherma di Spada e Sciabola (Milano: Fratelli Dumolard Editori, 1885).
Giordano Rossi, Sword and Sabre Fencing, translated by Sebastian Seager (Melbourne: Melbourne Fencing Society, 2021