Sun-tzu does the arithmetic out loud, and it's grim. To attack a fortified city you first spend three months building the movable shields and the armored assault wagons. Then another three months raising earthworks against the walls. Six months gone before a man climbs. And if the general can't master his impatience and sends his troops "swarming over the walls like ants," he kills a third of his officers and men — and the city still doesn't fall.1 That is the price tag Sun-tzu hangs on the storming of a walled town, and it is why he files urban assault as the lowest realization of warfare: not forbidden, but the option you take only when there is no other.2
Here is the move this page is about. The wall stops being a thing you climb and becomes a thing you read. Before you spend the six months and the third of your army, you score the city — its strength against yours, what taking it buys, what holding it costs — and you decide whether it is even worth the assault. The city becomes an intelligence object and a cost-benefit problem before it becomes a target.
This is the Chinese strategic tradition's progression from avoiding cities to evaluating them — a centuries-long shift in which the walled town goes from an obstacle you'd rather bypass to the decisive objective you assess, score, and choose. Sun-tzu sits at the start of it, in the late Spring and Autumn period, when siege technology was barely evolving and cities were "formidable targets that required a vast expenditure of time and effort."3 Crucially, in his era cities had not yet become critical targets: armies could move through open countryside and simply leave a stubborn city behind.4 So Sun-tzu's reluctance was rational economics, not dogma. Sawyer is emphatic that Western military writers have mangled this into "Do not attack cities" or "Avoid urban warfare," when Sun-tzu's actual ranking is graded — highest is to attack the enemy's plans, next their alliances, lowest their fortified cities.5 Lowest, not never.
The tomb-recovered Sun Pin Military Methods makes the calculus explicit, in a passage that may have belonged to the Art of War itself: "As for fortified cities that are not assaulted." If our strength is enough to seize it but seizing it gains nothing forward and we can't hold it to the rear — don't assault it. If a forward position would make the city surrender on its own, or if leaving it alone causes us no harm behind — then "even though the city can be assaulted, do not assault it."6 This is the heart of the doctrine in one line: there are cities you can take that you should choose not to take, decided by net gain.
By the time of the Wei Liao-tzu, a military classic composed perhaps a century after Sun Pin, the city has become the primary objective. Its formula is a small theory of the state: "land is the means for nourishing the populace; fortified cities the means for defending the land; and combat the means for defending the cities."7 Take the cities and the whole structure unravels. The reluctance has inverted into priority.
The mechanism is a chain of judgments, each one a filter the city must pass before steel is committed. The first filter is can we even take it — strength against strength. Sun Pin's rule is cold: "If our strength equals theirs, the city certainly will not be taken."8 Parity isn't enough; the attacker needs a decisive margin or the months are wasted. The second filter is does taking it pay — the net-gain test. Seizing a city that gains nothing forward and can't be held to the rear is a loss dressed as a victory. The third filter asks whether the assault is even necessary: a forward position might make the city surrender by itself, and a city that does no harm to your rear can simply be ignored.9
The Wei Liao-tzu turns this into a pre-mobilization intelligence routine. "Attack a country according to its changes. Display riches in order to observe their poverty. Display exhaustion in order to observe their illness."10 You probe the target to read its condition before you decide. Then, before mobilizing at all, you "investigate the strategic balance of power both within and without the borders" — is the army prepared or short, are provisions surplus or scarce, what are the routes of advance and return — and only then move.11 The city's vulnerability is a finding, not an assumption.
The same text even adapts the shape of the operation to the shape of the target, like matching a tool to a fastener. If the territory is vast but the cities small, occupy the land first. If the cities are large but the land narrow, attack the cities first. If the country is vast and the people few, isolate its strategic points. If the land is confined but the people numerous, build high mounds to overlook them.12 Four configurations, four different opening moves — the reading of the target dictates the method, not the other way around.
The final layer of the mechanism is tempo, and the Wei Liao-tzu makes the besiege-or-assault decision itself a function of the intelligence picture. Determining whether to besiege or storm is "simply part of an overall strategic assessment, a basic intelligence operation."13 Encircle the strong city and feed it a "prospect for some minor advantage," causing it to weaken day by day until the defenders reduce their own rations to nothing.14 The siege is not patience for its own sake — it is a slow conversion of a city you can't storm into a city you won't have to.
This page is the general's half of the city question — the cost-benefit and net-gain frame that decides whether a wall is worth attacking at all. It pairs directly with its two siblings to make a complete intelligence picture of urban warfare. Sun Pin's Male and Female Cities supplies the topographic readout — the classification of which walls are physically un-attackable by their setting in marsh, valley, and mountain — that feeds the strength-against-strength filter on this page. And City Defense and Siege Psychology supplies the defender's counter-calculus, the rescue-certainty that determines whether the besieged will exhaust their blood on the walls or weep as they hold them. This page is the hinge between them: it consumes the topographic score and the morale read as inputs to a single decision — assault, besiege, isolate, or bypass.
To the broader corpus, this page contributes the principle that in Chinese strategic thought the most important calculation often precedes the attack and may cancel it. The vault's pages on temple-calculation and net assessment have their natural urban instance here: the city is where the abstract pre-war reckoning becomes a concrete go/no-go on a specific wall.
Set Sun-tzu's two passages side by side and watch the doctrine work as a single argument. The first is the cost: six months of shields and earthworks, and the assault where men "swarm over the walls like ants" and a third of the officers and troops die for nothing.15 This is the image that should sit in a commander's mind — not the glory of the breach but the ledger of the corpses. Sun-tzu deploys it precisely so the reader will do the arithmetic before the assault.
The second passage, from the Sun Pin fragments, is what the arithmetic produces: the fortified city that is not assaulted. Read the logic as a decision tree. We estimate our strength is sufficient to seize it — the first filter passes, we can take it. But: seizing it gains nothing forward, and we couldn't hold it to the rear. So the second filter fails. Conclusion: "even though the city can be assaulted, do not assault it."16 The remarkable thing is that the capability to take the city is treated as irrelevant once the value test fails. A commander who storms a city merely because he can is, by this logic, a fool spending a third of his army to buy a position that gains him nothing.
The case crystallizes the whole doctrine: the swarm-of-ants image sets the price, the not-assaulted city demonstrates the refusal, and between them sits the calculus — the city scored as a cost-benefit object, where the right answer is sometimes to walk away from a wall you could have climbed.
You're standing on a rise a half-day's ride from the enemy's largest town, and your staff wants to know when the assault begins. You don't answer yet. You send men forward to "display riches" — to make a show that draws out a response — and you watch what comes back: how thin the granaries look, how nervous the gate. You run the first number in your head. Can we even take it? Their strength against ours — and if it's parity, you already know the wall will hold and the months will burn. You run the second. If we take it, what does it buy? You picture yourself inside those walls a month from now, supply lines stretched, no advantage gained forward, unable to hold it when their relief column arrives. The picture is ugly, so you keep going. Can we make it surrender without the climb? You imagine seizing the heights and the fords around it, severing its roads, feeding it a small false hope while its rations dwindle day by day. The city would weaken on its own. You feel the pull of the breach — the clean, fast story of storming the wall — and you set it down, because the arithmetic says encircle, not assault. The hardest order you give that day is the one where nobody climbs anything.
The doctrine fails first through impatience — the exact flaw Sun-tzu names. The general "cannot overcome his impatience" and launches the swarm-like-ants assault before the calculus is run, trading a third of his men for a wall that doesn't fall.17 The diagnostic sign is a commander who treats the city as a thing to be conquered rather than evaluated, who has skipped straight to the breach.
The second failure is the capability trap — storming a city simply because you can, with no net-gain test. Sun Pin's whole point is that can and should are different questions, and the commander who collapses them seizes positions that gain nothing forward and can't be held to the rear.18 The sign here is a string of captured cities that the army can't garrison and bleeds men to hold.
The third failure is misreading tempo — choosing a long siege when speed was required, or speed when patience would have won. The late Sung Hundred Unorthodox Strategies draws the line sharply. "Slowness in Warfare" fits the city with high walls and deep moats but many people and few supplies and no prospect of rescue: wait, let them weaken, then take a town that will prove highly defensible once yours.19 "Quickness in Warfare" fits the opposite — ample food, few defenders, but a real danger of relief from outside: strike fast, because if you dawdle the external force arrives and crushes you against the wall while the defenders counterattack from within.20 Get the tempo backward and the city's own configuration destroys you. The sign of this failure is an attacker who has chosen his pace from his own appetite rather than from a reading of the defenders' supplies, numbers, and hope of rescue.
The clearest tension is internal to the tradition and Sawyer preserves it rather than smoothing it: the corpus contains both Sun-tzu's reluctance to assault cities and the Wei Liao-tzu's insistence on making them the primary objective. Sawyer reads this as a genuine historical progression driven by the rise of siege technology (the catapult) and the growth of cities into economic and strategic centers — campaign armies that once flowed around sparse towns were later "thwarted by the heavily fortified strongholds."21 But the texts coexisted in the same received canon, so a commander studying the classics inherited the contradiction as live advice, not as resolved history. The vault keeps both poles: avoid the city / take the city are not errors to reconcile but conditions to read.
A second tension: Sawyer notes that even after the city became the dominant objective, "some reluctance is still seen in the later military writings because they continued to be under the theoretical influence of earlier concepts."22 The later strategists rarely bypassed valuable cities, yet they kept quoting Sun-tzu's "lowest form" line. The doctrine's earliest layer survived as rhetoric even after practice had overturned it — a reminder that a tradition's stated values can lag its actual behavior by centuries.
The cross-dynasty continuity here — Sun-tzu to Sun Pin to Wei Liao-tzu to the Sung Hundred Unorthodox Strategies — is Sawyer's synthetic line, drawn across roughly fifteen hundred years and several text-recoveries (the Sun Pin fragments were tomb finds). The progression is plausible and well-motivated, but the claim that these texts form one continuous conversation about cities is the translator-synthesizer's frame, and the dating of individual passages (especially the Sun Pin material's relation to the Art of War) carries the usual uncertainty of recovered manuscripts.
The pages converge on a single arc with a real reversal inside it. Sun-tzu ranks assaulting cities the lowest art; Wei Liao-tzu and the later strategists make the city the decisive prize — a shift that tracks the rise of siege engineering and the growth of the walled town into the seat of everything worth taking. Sun Pin's male/female typology and Wei Liao-tzu's rescue-certainty psychology then converge on the same practical question from opposite chairs: which walls are worth the blood. The tension is that Sawyer narrates this as one continuous conversation across fifteen hundred years and several tomb-recovered texts. The developmental line is his frame, and whether "cities are lowest" and "cities are decisive" mark an evolution or just different men's emphases is left open — preserved, not resolved.
Handshake with Psychological Resilience and the Survival of Humiliation. The net-gain calculus on this page demands that a commander refuse a winnable assault — that he walk away from a city he could storm because the arithmetic says it gains nothing. That refusal is an act of resilience against a specific humiliation: the public, in-front-of-the-army shame of declining a fight you could win. The resilience page treats the capacity to absorb humiliation as an internal psychological resource; this page shows that resource being spent in a strategic decision, where the cost of looking cautious is paid against the benefit of being solvent. The insight the pairing produces is that Sun Pin's "do not assault it" is not only a calculation but a temperament — it requires a commander who can hold the line against the army's appetite for the breach and his own appetite for the clean victory story. Net-gain reasoning is psychologically expensive; the cost-benefit table only works if someone can stomach the humiliation of obeying it. Neither page alone shows that the rational decision and the resilient self are the same hinge.
Handshake with National Combat Profiles. City evaluation scores a place; combat profiles score a people — and the Wei Liao-tzu's tempo rules show the two scores must be multiplied, not added. Whether to besiege or storm depends on the defenders' supplies and hope of rescue (the place), but also on their resolve — and resolve is exactly what the combat-profiles tradition claims to predict from regional character. The "Quickness in Warfare" lesson, where you must storm before relief arrives, only computes if you've already read whether these particular defenders are the kind who break or the kind who exhaust their blood on the walls. Reading them together produces an insight neither generates alone: the city calculus has a hidden variable — the human one — that it imports wholesale from the combat-profile file. A perfect topographic and supply read of a city is still incomplete until you've scored the people inside it, which is why the defender's psychology (the third sibling) and the regional character file both feed this page's decision.
Sharpest implication. In the mature Chinese tradition, the decisive act of urban warfare is not the assault — it is the evaluation that precedes and can cancel the assault. The city is scored as a cost-benefit object across three filters (can we take it, does taking it pay, is the assault even necessary) and a tempo read, and the highest-skill move is sometimes to refuse a winnable storming and walk away from the wall.
Generative questions.