Arthashastra: Statecraft

Time for our last traipse through the Arthaśāstrawhich started here and continued here.


Congratulations, you’ve become king of a small ancient Indian state. Your first question: how to choose ministers? This is an important enough question that Kauṭilya does a literature review: he summarizes the opinion of various authorities before giving his own. You should not pick your classmates or family retainers (they won’t respect you as a king), nor sycophants (they are devoted by not intelligent), nor “new persons” (who are inexperienced). You should choose men of “high family and possessed of wisdom… ministerial appointments shall purely depend on qualifications.” This sounds hard to disagree with, but it’s worth pointing out that most premodern states were aristocratic and not meritocratic. (And this was long before China’s examination system developed.)

Early governments are often pictured with a small staff. E.g. I was just reading in Mary Beard that Pliny the Younger was appointed governor of a fairly large province and had precisely two  officeholders beneath him. Beyond that, he had to use his own servants, co-opt native (non-Roman) authorities, or use the legions. Chinese magistrates might govern a million citizens with no staff paid by the central government. But the Arthaśāstra describes what sounds like a pretty large and thorough bureaucracy.  Here’s the main offices described:

  • Chamberlain (responsible for treasury and storehouses)
  • Collector general (of taxes)
  • Superintendent of accounts
  • … of the treasury
  • … of (manufacturing) metals
  • … of the mint
  • … of gold
  • … of the storehouse
  • … of commerce
  • … of forest produce
  • … of the armoury
  • … of weights and measures
  • … of tolls
  • … of weaving
  • … of agriculture
  • … of liquor
  • … of the slaughterhouse
  • … of prostitutes
  • … of cows
  • … of horses
  • … of elephants
  • … of chariots
  • … of passports

He also mentions the chief priest, the officer in charge of the harem, the magistrate, the king’s council, and governors of cities, forts, boundaries, and villages.

Kauṭilya writes as if the king could regulate and manage everything. There’s no bright line between public and private. It’s clear there was private activity, but the state also carried on a lot of economic activity on its own. The king also wanted his tax share of everything. There is even a rule that the state should supply dice to gamblers.

A warning on secrecy: counsels have been divulged by parrots, mynah birds, and dogs. (Was this warning literal? But then we say “The walls have ears.”)

The vices of a king are hunting, gambling, women, and drinking. Of these, Kauṭilya concludes that drinking and gambling are the worst. Drinking causes loss of money, corpselike appearance, loss of the Vedas, pain, loss of friends, and addiction to music. For Kauṭilya that’s pretty harsh.

Using confederates, princes should be terrified into avoiding all four. This can be done by drugging his liquor, defrauding him at gambling, accosting his hunting party in the guise of bandits, and showing him “impure women”.

A forest for the king may be set up with wild animals whose claws and teeth have been removed. (This is presumably for relaxation; other forests could be set up for hunting.)

Although there is much advice about how to serve the king, the life of a courtier is described as “living in fire.”

Kings should follow their subjects in dress, customs, language, and religion. (Again, Kauṭilya wrote in a period when kings were often foreign and/or non-Hindu, so this may be a complaint against the times.)

There are suggestions on how a minister can seize power. However, Kauṭilya advises against this; rather, a young prince should be set up as a puppet.


If Kauṭilya has one word for the king, that word is spies. Spies should check on government officials, attempt to corrupt them (so you learn which are corruptible), listen for dissidents, eliminate the seditious. They spread out into neighboring countries to bring information and sow division. Good covers for spies include religious disciple, ascetic, householder, merchant, prostitute, and mendicant woman. Poisoners and assassins are also needed.

If three different spies produce the same story, it can be believed.  If they frequently differ, they are probably making things up and should be dismissed.

Suspicious places to check on: vintners; sellers of cooked rice and meat; gambling houses; houses of heretics. Merchants and physicians are expected to report suspicious clients.

Entrapment is recommended. One neat idea: pretend to have supernatural powers, such as great speed, invisibility, causing sleep, opening locked doors.  See who signs up for lessons. (You can use confederates to pretend to sleep in order to demonstrate your powers.)  Arrest those who then attempt to commit crimes.

A spy can incite the brother of a seditious person to kill him. Then you kill the brother for fratricide.

You can set up traps in a temple, e.g. a wall that falls on your enemy as he enters.

A spy can pretend to be a long-lived ascetic and make friends with an enemy king. The spy claims that he takes a new body every hundred years, and invites the enemy to see the rite. If he shows up, he can be killed.

Spies can pretend to be gods and converse with the king, so the people think the king regularly has divine visitors.


Though kings were expected to rule with wisdom, they were also expected to conquer. “Whoever is superior in power shall wage war.”

It’s presumed that all the king’s neighbors are enemies. But by the same token, the king in back of your enemy might be your friend. The rules for dealing with enemies, friends, and neutrals are pretty complex, and frequently cynical. (If you need to double-cross your enemy, he tells you how to do it.)

Fighting to the end is not wise; better to surrender. Typically a surrendered king was allowed to administer his own territory.

If you have to give children as hostages, it’s best to give princesses, because they “cause troubles” for the court that receives them. Unfortunately he doesn’t explain what troubles!

Is it better to attack a strong but wicked king, or a weak but righteous one?  The wicked king, because his own subjects will refuse to support him.

Is it better to have a small army of bold men, or a large army of effete men?  The latter: there is always work for the weaklings, and numbers terrify the enemy. Besides, you can train the effete men to be more spirited.

You could use an “army of traitors” to look weak and invite attack.

A look-alike for the king should supervise the arrangement of troops.

An untrained army can march one yojana a day (5.5 miles)— the best armies could do twice that. (Other sources on ancient warfare suggests 20 miles a day… but at this period north India still had lots of forest, so Kauṭilya probably knows what he’s talking about.)

Ways to cross a river: a line of elephants; planks spread over pillars; bridges and boats; masses of bamboo; baskets covered with skins.

You shouldn’t harass a defeat army, because it will become reckless and dangerous.

Elephants can be used not only to charge the enemy, but to break into forts, to clear the path, to protect your flank, to ford streams, to quench fires, to carry the treasury.  However, elephants are only good when there’s plenty of water: in dry hot country they become obstinate, or catch leprosy.

The four branches of the army are infantry, elephants, horses, and chariots.  However, it’s clear that the number of chariots is small: a few dozen make up the chariot arm.

Three men can oppose a horse; fifteen are needed to oppose a chariot or an elephant.







All about swords

It’s about time to add one more post to the Weaponry category, for a total of two.

There’s a great series of videos on swords, axes, and other premodern weapons by one Skallagrim. This is just one of them, on why the elaborate swords beloved of video game designers are nonsense.

There’s lots more– reviews of particular swords, tests of swords from various cultures, why lightsabers could be extremely dangerous to the wielder, and perhaps most interesting, advice on swordfighting. Did you know it’s a perfectly valid technique to grasp your own sword by the blade and use it backwards, hitting someone with the pommel?  Did you know you can’t unsheathe a sword or katana sticking up from your shoulder in one swift move, like they do in movies and video games?  (To be precise, you can if it’s a very short sword. If it’s longer than your arm, it won’t clear the scabbard.)

All this is fun to learn about in general, but a particularly valuable resource for fantasy writers and conworlders. It could keep you from writing or drawing something really embarrassing.


Ballad of Mùlán

The original story of 木兰 Mùlán (‘Magnolia’) comes from an 11C anthology— The Ballad of Mùlán (Mùlán cí)— though the actual source is probably centuries earlier.  I had the whole thing in my book but decided that the full Chinese translation was overkill there, so it’s going here instead.


The illustration is of 侯梦瑶 Hóu Mèngyáo in a Chinese production, The Legend of Huā Mùlán— she acquired a surname in a Míng play. Mùlán doesn’t belong to Disney!


Jījī fù jījī, Mùlán dāng hù zhī.

(onomatopoeia) again (onomatopoeia) / Mùlán at door weave

Spin spin, again spin spin, Mùlán, facing the door, weaves.


Bù wén jīzhù shēng, wéi wén nǚ tànxī.

not hear machine-shuttle noise/ only hear girl sigh

The loom’s sound is not heard, only the girl’s sighs.


Wèn nǚ hé suǒ sī? Wèn nǚ hé suǒ yì?

ask girl what SUB think / ask girl what SUB remember

Ask her, what are you thinking, what do you recall?


Nǚ yì wú suǒ sī, nǚ yì wú suǒ yì.

girl also not.have SUB think / girl also not.have SUB remember

She does not think, does not recall anything.


Zuóyè jiàn jūn tiě, Kèhán dà diǎn bīng,

yesterday-night see army notice / Khan big point troops

Last night she saw the army notices:

the Khan is mustering a great army—


Jūn shū shí’èr juàn, juànjuàn yǒu yé míng.

army book 10 2 scroll / scroll scroll exist father name

The muster fills twelve scrolls, Father’s name is in each one.


Āyé wú dà ér, Mùlán wú zhǎng xiōng,

honorific-father not.have big son / Mùlán not.have extended

Father has no grown son, Mùlán has no elder brother.


Yuàn wéi shì ān mǎ, cóngcǐ tì yé zhēng.

will serve city saddle horse / from this replace father levy

I will buy saddle and horse in the market, and take his place.


Dōng shì mǎi jùn mǎ, xī shì mǎi ānjiān,

east market buy spirited horse / west market buy saddle

In the east market she buys a fine horse, in the west a saddle;


Nán shì mǎi pèitóu, běi shì mǎi chángbiān.

south market buy bridle / north market buy whip

In the south market a bridle, in the north a whip.


Dàn cí yéniáng qù, mù sù Huánghé-biān.

dawn from father-mother go / dusk lodge Yellow-River-side

At dawn she leaves her parents,

at dusk she camps at the side of the Yellow River.


Bù wén yéniáng huàn nǚ shēng, dàn wén Huánghé liúshuǐ míng jiānjiān.

not hear father-mother call girl sound / but hear Yellow-River flowing-water sound (onomatopoeia)

She doesn’t hear her parents calling to her,

only the splashing of the Yellow River’s water.


Dàn cí Huánghé qù, mù zhì Hēishān-tóu.

dawn from Yellow-River go / dusk arrive Black-mountain-head

At dawn she leaves the river, reaches the Black Hills at dusk.


Bù wén yéniáng huàn nǚ shēng, dàn wén Yān-shān hú qí shēng jiūjiū.

not hear father-mother call girl sound / but hear Swallow-mountain barbarian-rider sound (onomatopoeia)

She doesn’t hear her parents calling to her,

only the sound of the nomads riding on Mt. Yān.


Wànlǐ fù róng jī, guān shāndù ruò fēi.

10,000-mile go military moment / cut mountain-pass like flying

10,000 miles of riding to battle,

dashing across mountains and passes as if in flight.


Shuò qì chuán jīn tuò, hán guāng zhào tiě yī.

new.moon energy transmit metal watchman.rattle / cold light shine iron armor

The watchman’s clapper rings in the icy wind,

a cold light shines on iron armor.


Jiāngjūn bǎi zhàn sǐ, zhuàngshì shí nián guī.

general hundred battle die / warrior ten year return

Generals fight to the death in a hundred battles;

warriors return after ten years.


Guīlái jiàn tiānzǐ, tiānzǐ zuò míngtáng.

return-come see heaven-son / heaven-son sit bright-hall

She returns to see the Emperor, seated in his bright hall.


Cè xūn shí’èr zhuǎn, shǎngcì bǎi qiān qiáng.

plant merit twelve turn / reward hundred thousand power

He bestows the highest honors on her, and countless sums.


Kèhán wèn suǒ yù.“Mùlán bù yòng shàngshū-láng,

khan ask SUB desire / Mùlán not need honor-book-scholar

The khan asked what she wanted. “Mùlán does not need an appointment to office.


Yuàn jiè míng tuó qiānlǐ zú, sòng ér huán gùxiāng.”

will borrow bright camel thousand-mile sufficient / send child return home

I want only a camel capable of a long journey,

to carry me back to my home town.”


Yéniáng wén nǚ lái, chū guō xiāng fújiāng.

father-mother hear girl come / go.out city.wall mutual support

The parents hear that their daughter is coming;

they wait at the city wall, holding each other up.


Āzǐ wén mèi lái, dāng hù lǐ hóngzhuāng.

honorific-older.sister hear younger.sister come / at door arrange red-adornment

The older sister hears that her younger sister is coming,

she waits at the door, applying red makeup.


Xiǎodì wén zǐ lái, mó dāo huòhuò xiàng zhū yáng. hear older.sister come / grind knife (onomatopoeia) to pig sheep

The younger brother hears that his older sister is coming,

he quickly sharpens the knife for the pig and sheep.


Kāi wǒ dōng gé mén, zuò wǒ xī gé chuáng.

open I east chamber gate / sit I west chamber bed

“I open the door on the east side of the room,

sit on the bed on the west side.


Tuō wǒ zhànshí páo, zhuó wǒ jiùshí cháng.

remove I battle-time robe / dress I former-time skirt

I take off my wartime gear, put on my old clothes.”


Dāng chuāng lǐ yúnbìn, duì jìng tiē huā huáng.

at window arrange cloud-hair / facing mirror stick flower-yellow

At the window she arranges her flowing hair;

before a mirror she applies a flower decoration.


Chū mén kàn huǒbàn, huǒbàn jiē jīnghuáng.

go.out gate see fire-mate / fire-mate all shock-fear

She comes to the gate to meet her fellow soldiers,

who are all utterly shocked:


Tóng xíng shí’èr nián, bù zhī Mùlán shì nǚláng.

with go twelve year / not know Mùlán be girl-youth

Her companions for twelve years didn’t know she was a woman.


“Xióng tù jiǎo pū shuò, cí tù yǎn mílí;

male hare foot push north / female hare eye blurred

“[Held up,] the male rabbit’s foot kicks quickly;

the female rabbit’s eye is nearly closed.


liǎng tù bàng dì zǒu, ān néng biàn wǒ shì xióng cí!”

both rabbit close earth go.along / how can distinguish I be male female

But if both are running on the ground,

how can I tell which is which?”

How the North Won

I just finished How the North Won, by Archer Jones and Herman Hattaway, an extremely thorough military history of the Civil War.  If you are a little hazy on the difference between strategy and tactics, this book will cure that.

And you’ll learn about logistics.  So much logistics.  One of the major figures of the book is a gentleman I’d never heard of– Henry Halleck, who functioned as the northern chief of staff and, in Jones and Hattaway’s telling, something of the architect of the overall Northern strategy.

Henry Halleck

Mid-19th century warfare (i.e. post-Napoleon, pre-WWI) was transformed by two things: the rifle and the railroad, both of which made an infantry army almost impregnable to a frontal assault. 

The rifle made cavalry virtually obsolete on the battlefield: it was fast and accurate enough that it could decimate an oncoming cavalry charge.  And that in turn meant that soldiers didn’t have to stand to resist the charge with bayonets, but could lay down, or better yet take cover behind improvised earthworks, which protected them from artillery fire, the bane of Napoleonic warfare.

A frontal assault was almost always suicide– e.g. the famous charge by Pickett at Gettysburg, one of Robert E. Lee’s few mistakes.  Commanders had better luck with turning movements– i.e. attempts to go right or left round the enemy’s flank, with hopes of moving against his weaker rear.  But even this often failed; even fairly raw troops could be fairly easily turned around to face the threat.  There were very few routs in the Civil War; if you did manage to turn the enemy, he almost always successfully retreated.  (That might be enough to take a major objective; but the typical progress of a Civil War campaign was a series of battles a dozen miles apart, as the defending army fell back to a new defensive position.)

The fastest and cheapest transport and supply lines were still by sea; but the railroad was almost as good, allowing enormous armies to be moved at unprecedented speeds.  Wagon trains, though they had to be resorted to, were a distant third.  William Sherman once estimated that his advance to Atlanta would have required 36,000 wagons and 220,000 mules without the railroad, a simple impossibility.

The North had a bigger army and a huge edge in industrial production, but it faced a huge predicament: a large enemy army was almost impossible to dislodge.  In particular, as one general after another discovered, taking Richmond– a tempting hundred miles from Washington DC– was almost impossible.  Lee’s army was too strong, the city was too well connected by rail and canal to its supply regions, and Lee made no blunders that created openings.  (The North tended to blame its succession of generals– McClellan, Pope, Burnside, Hooker, Meade, Grant– but it was an inherently difficult problem.)

Things were easier in the West, where initially the North had the great advantage of water transport along the great rivers as well as the Gulf of Mexico.  This allowed the conquest of western Tennessee, New Orleans, and ultimately Vicksburg (July 1863).  And that in turn suggested the better strategy: keep Lee busy in Virginia, but slowly conquer the West.  By careful and strong provisioning the North managed to drive through eastern Tennesee to Chattanooga (Sep. 1863).  The Navy was of course blockading the coast; the overall idea was to strangle the South and roll it up from the West– the “anaconda strategy“.

But further progress stalled.  Conquests had to be protected, and the increasingly long supply routes protected; relatively tiny Confederate raiding forces tied up tens of thousands of Union troops.  (This was the one remaining, and still very effective, use of cavalry.)  And the interior railroad lines of the remaining Confederacy allowed armies to be quickly concentrated to resist any large-scale intrusion.

Ulysses Grant

Grant‘s genius was to turn the Confederate raiding strategy against the South.  Instead of small cavalry forces, he’d send a huge infantry force, which would have the usual protection from direct assault.  It could cause immense destruction– tear up railroads, destroy factories, turn slaves producing for the South into soldiers fighting for the North– and best of all, it didn’t need a supply line to protect.  By moving into new territories it would live off the land, and it could simply fight its way into a place where it could be reprovisioned.

Sherman tested out the idea with a raid from Vicksburg to Meridian, and it worked like a charm.  He then fought his way conventionally from Chattanooga to Atlanta (Sep. 1864, just in time to pave the way to Lincoln’s re-election).  He didn’t bother to occupy Atlanta, but just burned it.  Then he cut his way in a wide swath to Savannah.

Grant tried to use the raiding strategy elsewhere, though he was dogged by a few incompetent generals.  But finally some of them got it, notably Sheridan in northern Virginia.  (A subtheme of the book is that both armies functioned as brutal meritocracies.  By 1864 or so the incompetent generals had almost entirely been eliminated)

Grant meanwhile bludgeoned his way to Petersburg, just south of Richmond.  This part of the campaign didn’t look any better than the earlier generals, but at least his attacks kept Lee’s huge army busy.  (One enormous missed opportunity: a clever mining operation blasted a huge hole in the Confederate lines at Petersburg, in July 1864.  At the last minute a well-trained unit of black soldiers was replaced by some less-ready white troops, who didn’t occupy the breach quickly enough.)

Sherman turned north, marching through the Carolinas and reaching Fayetteville in March 1865.  This completed his work of devastation; the remaining railroad links to Richmond were cut.  An ill-advised attempt by Hood to retake Nashville had devastated the Confederates’ last remaining large army besides Lee’s; nothing would prevent Sherman moving north to join Grant.

The end was anticlimactic; Sheridan joined Grant and turned Lee’s army to its right, forcing Lee to abandon Petersburg and Richmond and retreat westward.  Grant followed and Sheridan moved on ahead, trapping Lee’s army.

Sherman’s March to the Sea was brutal, but Archer and Hattaway show that it was a brilliant and necessary military strategy.  Defeating Lee by assault was impossible, and the slower conquest strategy in the west would have taken many more years.  The only way to counter the advantages of the defense in 19C warfare was to destroy the enemy’s means of production.

The book glances only lightly at the political and social situation, besides mentioning Lincoln‘s basic difficulty: the public, the politicians, and the newspapers did not understand the anaconda strategy, but wanted battles won by direct assault… a type of campaign that just was not going to happen.  Lincoln educated himself on military matters, however, and very effectively mediated between the civil and military points of view. 

The story is slow-going at times, but that’s part of the point: warfare at this time was achingly slow.  Some generals exacerbated this– Lincoln described several of them, especially McClellan, as having a case of “the slows”– but even the best generals took months to prepare any movement, and with good reason: maintaining a hundred thousand soldiers was an enormous logistical problem.

A particularly neat feature is the extensive maps– large-scale diagrams explaining the river and railroad network and overall strategy, as well as tactical maps, often several per battle.  The authors also extensively quote diaries and letters of the generals, giving a fascinating glimpse into their mindset, as well as into the strange nature of a civil war– the generals usually knew each other and had often studied under the same professors at West Point.

There’s also any number of interesting facts… e.g., did you know the last Monitor class ironclad was retired only in 1937?

Redoing the military

Pundits are worrying about smaller government, and those with actual brains realize that any actual reduction means a smaller Defense Dept.  Matt Yglesias asks what this might mean, and comes up with “gendarmes”… basically a national police force which can be used in “high-crime jurisdictions” as well as for occupations after an enemy was defeated.

His commenters basically ream him a new one.  1) A national police force wouldn’t fly.  2) Different skill sets.  3) Posse comitatus.  4) That’s the National Guard, dummy.  5) Do you want COIN-trained military assigned to the drug war?

All good points, but I think what Yglesias is really trying to say is twofold: first, the military is almost entirely oriented toward war; we’d be using our tax dollars a hell of a lot more efficiently if some large fraction of what we spend on it was useful domestically too.  Second, we’re not really in a world where we’re about to refight WWII, so we need to radically rethink what kind of forces we need.

Generally, militaries are concerned to fight the last war.  To be really cynical, the problem is that the last wars were mostly insurgencies, and we really don’t know how to win one of those.  Nor does anyone else.  (To be more precise, it’s like good schools: we “know” how to do it in that we can point to a few local success stories.  We don’t know how to scale it up and generalize it so that it works at the national level.)

It’s fair to say that we could use a “nation-building corps”.  The example Yglesias gives is if North Korea is defeated; someone need to rush in from Day One and organize a state so it doesn’t fall into chaos or warlordism.  But it’s also clear that this isn’t something resembling a national police force.

A better sampling of Yglesias is here, where he points out the ending filibusters would be better not only now when Republicans are obstructing Democrats, but later when Democrats would be obstructing Republicans.  Majority rule is a good thing, and filibusters in general have been retrogressive and more annoying than actually protective of anything.  They didn’t prevent either the excesses of the Bush administration, or the substantial accomplishments of Obama.

This horrifying five-foot weapon…

Lethal... or goofy?
Lethal... or goofy?

jwz found an old article featuring British coppers puzzling over a Klingon betleH:

Now, My B.S. detectors go off whenever I see fantasy illustrations of oversized swords with claws and sharpened outgrowths.  I figure that the last 3000 years of combat have refined what makes a good edged weapon, and there’s a reason real weapons don’t look like what art directors come up with.  Though anything with a sharp edge can be dangerous, and I imagine the betleH could parry all right, it looks like it’d be lousy for thrusting.  I’d think a good swordsman would either chop at your legs, or knock it, twisting it into a position more dangerous to the wielder than to the opponent.

IIRC European swords actually got thinner over time– the rapier outperformed the longsword.

But, I’m just speculating.  Surely some of my readers are martial arts geeks and can offer a more informed opinion.