Birds taking flight, a column of dust, or the words of an envoy can provide information, but they do not speak for themselves. Chapter 9 combines observations of the terrain and of the opposing army with criteria for assessing one’s own position.
The terrain is also part of the information
Chapter 9 begins with troop placement and observation of the enemy. In Lionel Giles’s 1910 version, mountains, rivers, marshes, and plains pose different problems. The choice of location affects movement, access to resources, and troop exposure.
Within terrain and adaptation: chapters 8 to 11, this section deals with concrete clues. Before interpreting what the rival is doing, the observer must take into account the space in which they are seen. Remaining still in a protected position and doing so in an exposed place do not necessarily suggest the same thing.
The text links, for example, a swollen, foaming river with rainfall upstream and advises waiting before crossing it. Observation changes a marching decision. There is no need to know first-hand everything that happened in the watershed to recognise that the crossing conditions have changed.
It also points to places where vegetation and the lay of the land can hide people. That distinguishes a pending inspection from a conclusion already confirmed: a space suitable for concealment does not prove that it is occupied. The six terrains of chapter 10 develop another classification, which should not replace these specific observations.
Birds, dust, and movement: from clue to hypothesis
In Giles’s section 22, birds rising into the air are linked to an ambush. His comment reproduces Zhang Yu’s explanation: birds that suddenly take flight may have been disturbed by soldiers hidden below. The commentator’s explanation expands the brief wording of the treatise; it is not a field observation documented by Giles.
The same passage links startled animals with a sudden attack. Both formulations assign a possible human cause to a disturbance in the environment. Reading them today as an infallible code would erase the work of interpretation itself: an animal’s movement does not by itself identify the one who caused it.
The next section distinguishes forms of dust. A tall, narrow column is associated with moving chariots; a low, spread-out cloud, with infantry. The description sometimes summarised as cone-shaped dust comes from that reference to a high, pointed rise. It does not provide an angle or measurement for recognising vehicles with precision.
Dust that branches out is linked to groups gathering firewood. Small clouds that come and go are interpreted as camp activity. What matters is that the text compares shapes and movements rather than assigning the same meaning to any visible cloud.
It is helpful to separate three levels. The signal is what is observed: a low cloud over an area. The hypothesis proposes a cause: it could correspond to people on the march. The inference relates that hypothesis to other conditions and to a decision. This distinction explains the reasoning; it does not reproduce the chapter’s formal terminology.
A conclusion improves if it can be checked against clues that do not all depend on the same observation. Describing the same dust again as a sign of movement and as a sign of presence does not provide two independent checks. Nor does the lack of visibility prove that there is no activity.
The chapter includes chariot movements and changes in formation among military signs. These are observations closer to an army’s functioning than the startle of a bird, though they still need context. Counting a movement does not amount to knowing the plan that produced it.
What the words and the troops’ behaviour say
The case of the envoys shows why a single word must not be isolated. Section 24 links humble language with an increase in military preparations and interprets this as a sign of advance. Humility, on its own, does not contain that conclusion: the reading depends on the relation between speech and activity.
Later, section 38 links envoys who offer compliments with a desire for a truce. It would not be accurate to summarise both passages with the rule, ‘whoever speaks kindly is deceiving.’ The way one addresses the enemy is interpreted differently depending on the conditions that accompany the message.
Observation also extends to material needs. The treatise links soldiers leaning on their spears with hunger, and men who drink before carrying water with thirst. These are associations presented by the source, not methods for diagnosing a person’s physical state with certainty from an isolated gesture.
In the end, attention returns to one’s own force: treatment, training, and the carrying out of orders all play a part in its coordination. Reading the other side does not replace organisation. That relationship connects with the flexibility of command in chapter 8: recognising a signal is of little use if one then clings to an interpretation that new data contradict.
A hypothetical example in a strategy video game makes it possible to carry over the distinction without attributing modern mechanics to Sun Tzu. A player no longer sees rival units in one area of the map. They may think the units have moved, but also that they have simply fallen outside their vision. Both explanations require different responses. This comparison belongs to macrostrategy in esports, not to the literal content of the treatise.
External observation leaves questions open. The five kinds of spies in chapter 13 introduce another source of information, with their own problems of access and reliability.
View all articles in Terrain in Sun Tzu: adaptation and situations in Chapters 8 to 11
