Chapter 3: attack priorities, command, and intelligence

Chapter 3 ranks military options by what they let you achieve and preserve. The hierarchy begins with the adversary’s plans and places siege last. The argument goes on to cover the balance of forces, the relationship between general and ruler, and the knowledge needed to decide.

Why attack options are not interchangeable

In Lionel Giles’s 1910 translation, Chapter 3 begins by preferring to leave the enemy state and forces intact rather than destroy them. That criterion helps interpret the later hierarchy: achieving an objective and laying waste to what you intend to act upon do not produce the same result.

The idea does not amount to a general renunciation of the use of force. The treatise continues to examine how to bring the adversary under control. The meaning of supreme military excellence deepens that ideal; to read the chapter in full, you have to continue on to the criteria that guide the choice of means.

The order can be represented by this outline. Each level acts on a different object, so it is not enough to rename an operation in order to place it in a higher category.

  1. Acting on plans: frustrating the intention or the conditions on which the adversary’s project depends.
  2. Acting on alliances: preventing the adversary from obtaining or bringing together the support that strengthens their position.
  3. Confronting the army: resolving opposition by clashing with its deployed forces.
  4. Besieging cities: committing resources against a fortified position, an option the text advises avoiding whenever possible.

The first priority requires understanding what the adversary aims to achieve. An action may inconvenience them without changing their plan. The difference between causing a nuisance and altering the conditions of their project explains what it means to attack strategy, without reducing that concept to any deception or show of force.

At the second level, Giles translates the passage as preventing the joining of enemy forces and explains it through separation from their allies. The role of alliances in Sun Tzu allows this dimension to be examined. The brief statement does not list bribes or a repertoire of diplomatic manoeuvres that can simply be added as a literal translation.

Siege receives a material explanation: equipment preparation, prolonged labour, and the risk that impatience will lead the general to order an expensive assault. Its lower place in the hierarchy is linked to the cost of war in Chapter 2. Destruction and time consumed may reduce the value of what is obtained.

Available force and the responsibilities of command

The chapter also distinguishes situations of superiority, equality, and inferiority. It states numerical proportions and links different possibilities of action to them. It does not offer a formula that lets you enter two quantities and obtain the outcome of a battle. The comparison must preserve the other conditions the treatise itself examines.

Giles also includes Zhang Yu’s clarification that a numerical difference may be offset by other qualities. This is an interpreter’s comment, not an additional number in the text. Its purpose is to prevent the list from being read as if all fighters and situations were equivalent.

A hypothetical example makes the limit clear. Two forces with the same number of members may be in different states of readiness, dispersion, or supply. Equality of numbers answers one question, but leaves others open. The chapter does not compel battle merely because a numerical comparison seems favourable.

The general’s capability is linked to the strength of the state. The text then examines three ways in which the ruler can harm their army. The first is to order advance or retreat without knowing that the force cannot do so. The problem is not the existence of an order, but its disconnect from real possibilities.

The second warning concerns governing the army while ignoring its particular conditions. Giles interprets the third as appointing officers without discernment or understanding of the circumstances. Both place part of the risk within the organisation itself: a force may lose coherence before the adversary defeats it.

This criticism of interference does not establish that the general has unlimited authority over any political matter. The issue under examination is the competence to decide in a military situation. Confusing operational autonomy with a lack of accountability would go beyond what the passage explains.

What information does the decision-maker need?

Towards the end, five conditions appear for recognising the possibility of victory: knowing when to fight and when to refrain, handling forces of different sizes, maintaining a shared will, being prepared against an unprepared adversary, and having a capable general without harmful interference from the ruler.

This list belongs to Chapter 3. It does not repeat the five factors from Chapter 1, although some topics are related. Its function here is to gather the capabilities needed to choose and carry out an option; it does not grant an automatic guarantee to anyone who claims to possess them.

The relationship between preparation and opportunity continues in the arrangements of Chapter 4. Being organised makes it possible to avoid certain weaknesses of one’s own, but it does not force the adversary to leave an opportunity available. That is why the chapter’s analysis does not end with a general order to attack.

The final maxim about knowing the other and knowing oneself must be read in that context. Knowledge of oneself helps identify limits; knowledge of the adversary allows one to assess what they are trying to do and the means at their disposal. The passage does not assign statistical percentages to each combination of knowledge.

The hierarchy provides direction, but it needs diagnosis. Without understanding the adversary’s objective, their support, and the capability of one’s own force, it is impossible to justify why one option would be preferable to another. A list of priorities does not replace the information needed to apply it.

Before assessing how effective an action is, it is worth specifying what it aims to change: a plan, a support relationship, a deployed force, or a fortified position. That distinction avoids calling any favourable outcome strategic.

View all articles in Sun Tzu's Strategic Preparation: Chapters 1 to 3