As soon as a beginning fencer has developed an acceptable level of proficiency with simple attacks, we start to teach and learn compound attacks. The common definition of a compound attack is an attack of two or more blade tempos, with the final tempo being the actual attack, and all previous tempos being feints. We could also define it as an attack of two or more simple actions, with the final simple action being the actual attack, and all previous simple actions being feints. The standard simple attacks, one direct, the straight thrust, and three indirect, the disengage, counterdisengage, and coupe, thus form the building blocks of compound attacks (I have left out the rather rare and difficult to perform countercoupe). But, is this all there is to compound attacks?
Well, in a word, no. First, what is our objective for this family of actions? The compound attack has one fundamental objective – opening the line, and keeping it open for long enough, for the final attack to hit. It achieves this objective by presenting the opponent a realistic enough threat by feint, that he must react. This reaction moves the blade away from the final line of the attack. By committing to movement, the opponent is forced into the disadvantageous position of having to recognize the developing threat the final action represents, stop the movement against the line of the feint, and attempt to form a parry. In doing so she is forced into a new decision (OODA) loop, is behind the initiative of the attacker, and has to deal with the inertia of the parry against what has become empty space.
Second, we have to consider the flexibility of the compound attack. If you use the short or long tactical wheels as the basis for decision making, you are left with the idea that the compound attack is a single thing. Parry one compound attack and you can parry them all. That is simply not so. With four simple attacks, you have 4 times 4 = 16 possible combinations of the two tempo simple attack. Then consider that the low-hi-low and hi-low-hi combinations exist for at least the disengage based attacks, and that attacks from the inside to the outside and the outside to the inside present two different recognition and reaction problems. If you have never seen a feint of coupe-coupe attack from 4 to 6 to 4 (4 to 3 to 4 for sabre fencers), it is likely that the opponent could then hit you with 6 to 4 to 6 (3 to 4 to 3) on the next action, and you would be unable to tell what the last two actions were. In order to be able to react, we have to be able to recognize what is happening (the observe and orient steps of OODA). This reality makes the compound attack much more dangerous than it would seem in the tactical wheel or by its characterization as a two tempo attack.
Third, feints work because opponents recognize and react to the threat they appear to represent. How do we establish that threat? The obvious answer is that we hit the opponent first with a simple attack. Now the opponent has to react to what may be another simple attack in the same line. Notice that I said the same line – we want the opponent to identify the feint as the action that hit previously. Compound attacks become more dangerous if we use previous successful, or even nearly successful, simple attacks as the feint that sets up the next action. This means that the compound attack is not an isolated event – it is part of the evolving narrative of your tactical plan for the bout.
Fourth, the compound attack, as a prepared attack, is very suitable for tempo changes and the combination of footwork, blade, torso rotation, and finger action in accelerating to reach maximum point speed at what Harmenberg calls the critical gap for scoring, the last 8 or so inches to target. Combined with footwork you can slow down the opponent in the feint and then accelerate to maximum speed in the lunge. Or you can execute both a full speed in the lunge.
A note – in the Classical Period, four part compound attacks were commonly taught, although most masters admitted that three parts were all that you could hope to make work in a bout. When I learned in the 1960s, three part actions were occasionally still taught. Now, anything more than a two part action is rare.