In August 1839 the people of Kabul watched a procession come into their city and could not work out what they were looking at.
There was Shah Soojah, a ruler most of them could barely remember, forced out of power twenty-five years earlier, brought back looking old and submissive. And riding beside him — visibly the more important of the two — was a forty-five-year-old English civil servant in a bright-coloured uniform topped by a cocked hat fringed with ostrich feathers.1
Why had these people come? What were they doing here?1
William Macnaghten, chief secretary of the Calcutta government and now the Queen's envoy to Kabul, could have answered that question with total confidence. He was there to liberate Afghanistan.
Two and a half years later his head and limbs were paraded through the streets of that city and his torso was hung from a meat hook in the bazaar.1
The decision was made at a meeting in June 1838. Lord Auckland, the British governor-general of India, was worried about Russian influence: Persia was already an ally, Afghanistan was being courted next, and if that succeeded the British in India would be cut off to the west and exposed.1
The obvious move was to outbid the Russians and negotiate with the Afghan ruler, Dost Mahomed. Auckland proposed a surer solution: invade, and install Shah Soojah, who would then be indebted to England.1
Macnaghten thought it brilliant. His reasoning, laid out in Greene's account, is a list of assumptions each of which is false:
The British army would sweep away the primitive Afghan tribesmen without difficulty. They would present themselves as liberators, freeing Afghans from Russian tyranny. They would bring the support and civilising influence of England. And once the shah was installed the army would leave, so British influence — though powerful — would be invisible to the Afghan public.1
His support was so sound and enthusiastic that Auckland not only approved the invasion but made him the top British representative in the country.
What followed has a specific shape worth watching, because every adjustment Macnaghten made was locally reasonable and pushed the situation further toward the cliff.
The army stays. Dost Mahomed fled north and began building an army. To the south, chieftains whose lands had been plundered for food during the invasion were stirring up trouble. The shah was so unpopular that he could not be left unprotected. Reluctantly, Macnaghten ordered most of the British army to remain until things stabilised.1
The families arrive. Time passed, and Macnaghten decided the officers and men of this increasingly permanent occupying force should be allowed to send for their wives and children, so that life would be less harsh. The families came, with their Indian servants.1
He expected this to have a humanising, civilising effect. It alarmed the Afghans instead. Were the British planning a permanent occupation? Everywhere they looked there were Englishmen talking loudly in the streets, drinking wine, attending theatres and horse races — imported pleasures they had introduced to the country — and now their families were settling in.1
A hatred of everything English began to take root.1
The stipend is halved. London complained about the cost of the occupation. Macnaghten knew where to cut. The mountain passes carrying Afghanistan's main trade routes were held by the Ghilzye tribes, who for many years and under many different rulers had been paid a stipend to keep them open. He halved it.1
The Ghilzyes blocked the passes. Sympathetic tribes rebelled elsewhere. Officers who urged a vigorous response were rebuked for overreacting. Now the army would have to stay indefinitely.1
Two people told Macnaghten what was actually happening, and his answers to both are the diagnostic material.
To everyone warning him about the growing hatred, he had the same reply: everything would be forgotten and forgiven when the army left. The Afghans were childlike, emotional people; once they felt the benefits of English civilisation they would be more than grateful.1
The second warning came from Shah Soojah himself, who for months begged to be allowed to capture and kill his main rivals — an Afghan ruler's traditional method of securing his position. Macnaghten told him that a civilised country does not use murder to solve political problems.1
The shah knew what his own people respected, and it was not "civilised" values. His failure to deal with his enemies made him look weak and unrulerlike, and left him surrounded by them.
Macnaghten would not listen.1
By October a mob had killed a British official in his home. Local chiefs were conspiring to expel their overlords. The rebellion spread until Macnaghten had to admit he lacked the manpower to put down a general uprising.1
But why panic? The Afghans and their leaders were naïve. He would regain the upper hand through intrigue.
So he ran two negotiations at once. Publicly, an agreement for British troops and citizens to leave in exchange for food on the retreat. Privately, an offer to a few key chiefs: whichever one put down the rebellion and let the English stay would be made vizier, and loaded with money.1
Akbar Khan, chief of the eastern Ghilzyes, responded. On 23 December, Macnaghten rode out to a private meeting to seal the bargain.
After the greetings, Akbar asked whether he wanted to go ahead with the treachery they were planning.
Thrilled to have turned the situation around, Macnaghten cheerily answered that he did.1
Without a word of explanation, Akbar signalled his men. He had never had any intention of betraying the other chiefs. Macnaghten was seized, and on the way to prison a mob formed and — with a fury built up over years of humiliation — literally tore him to pieces.1
Everything unravelled in days. Some 4,500 troops and 12,000 camp followers were forced into an immediate retreat in bitter winter weather. The Afghans, contracted to supply them, did not — certain the British would never leave unless forced, they harassed the column relentlessly. Civilians and soldiers perished in the snow.1
On 13 January 1842, British forces at the fort in Jalalabad saw a single horse struggling toward the gates. Its half-dead rider, Dr William Brydon, was the sole survivor.1
Greene's verdict is blunt, and it is not about intelligence gathering:
The knowledge that would have averted the catastrophe was at Macnaghten's fingertips long before he launched the expedition.1
Englishmen and Indians who had lived in Afghanistan could have told him that the Afghan people were among the proudest and most independent on the planet, that foreign troops marching into Kabul would constitute an unforgivable humiliation, and that these were not people yearning for peace and reconciliation — they saw strife and confrontation as a healthy way of life.1
Macnaghten had the information but refused to see it. Instead he projected onto the Afghans the values of an Englishman, which he mistakenly assumed were universal. Blinded by narcissism, he misread every signal along the way.1
And so each strategic move — leaving the army in Kabul, halving the stipend, refusing to overreact to the rebellions — was exactly the opposite of what was needed.1
The final miscalculation was the same error at its purest: imagining that money and an appeal to self-interest would buy loyalty from the very people he had spent three years humiliating.
The version of this you will actually meet is not an invasion. It is a good-faith intervention into a group you do not belong to.
You arrive somewhere — a new team, an acquired company, a client's organisation, a community — with a genuine improvement. Better process, better tools, a fairer arrangement. You expect gratitude, and something else happens instead.
The first tell is the phrase Macnaghten used. Once they feel the benefits, they'll be grateful. Whenever you find yourself deferring someone's objection into a future in which they will have come round, you have stopped modelling them and started modelling a version of yourself who has already agreed with you.
The second tell is the humanising gesture that lands as a threat. He brought the families in to make things gentler; it read as permanent occupation. Before any softening move — the social event, the open-door policy, the "we're all one team now" — ask what it looks like from a position you do not occupy. Warmth is not self-interpreting.
The third is the cost cut against a payment nobody remembers the reason for. The Ghilzye stipend had been paid by rulers for generations because it purchased the passes. It appeared on the ledger as an expense with no visible service attached. Any long-standing payment that looks like waste is a bill for something invisible; find out what before you halve it.
The fourth is the local expert whose advice offends you. Shah Soojah's counsel was that he needed to kill his rivals. Macnaghten's refusal was morally right and strategically fatal, and the chapter does not resolve that — but you should notice that he did not treat it as information about how power worked there. Rejecting someone's proposed action does not oblige you to discard their reading of the terrain.
And the last: when you find yourself thinking they're naïve, I'll manage this with cleverness, you are Macnaghten riding out to meet Akbar Khan. Condescension is what he had instead of a threat model.
The First Anglo-Afghan War is well documented and Brydon's arrival at Jalalabad is one of the most recorded moments in British imperial history. [POPULAR SOURCE] for Greene's psychological framing. The vault's The First Afghan War carries the campaign-level account and a larger figure for the column — approximately 16,500 including camp followers, against Greene's 4,500 troops plus 12,000 followers; the two are roughly consistent and the discrepancy is not load-bearing.
Greene's diagnosis may be too flattering to the individual. "Blinded by narcissism" locates the failure in one man's psychology. But Macnaghten's assumptions — that the natives were childlike, that English civilisation was self-evidently desirable, that gratitude was the natural response to occupation — were not personal quirks. They were the standard operating beliefs of the enterprise that employed him, and Auckland shared them, and so did the meeting that approved the invasion. Treating this as a reading error by an unusually narcissistic envoy makes it a solvable individual failing. Treating it as the ideology functioning normally makes it structural, and much harder to guard against, since the whole point of a shared assumption is that no one in the room flags it.
The chapter also cannot decide whether Macnaghten's refusal to sanction murder was a mistake. It is presented inside a list of misreadings, immediately after the shah's warning that Afghans respected strength — which frames the moral position as another failure of perception. Greene does not say outright that Macnaghten should have let the killings proceed, and he does not say the opposite either. The instrumental logic of the chapter points one way and the sentence declines to follow it, which is this book's characteristic move: state the principle, let the case carry the cost, and don't price it.
Put this next to The Soviet War in Afghanistan — built from Strategy 22 in this same ingest — and the pairing is almost unbearable.
Same country. A hundred and forty years apart. An outside power installs a client ruler it distrusts, keeps an army in to protect him, watches its presence generate the hatred that requires more army, cannot leave without conceding a catastrophic loss of prestige, and eventually gets out having achieved nothing at enormous cost.
Greene uses the first as a lesson in reading people and the second as a lesson in exit strategy, and never connects them. But the connection is the more useful fact: the British error was psychological (projecting English values) and the Soviet error was structural (no exit from a gamble), and both produced the identical campaign. Which suggests neither diagnosis is complete. Something about occupying that place produces the same shape regardless of which mistake you arrive with.
Against Metternich and the Close Embrace, the chapter's other case, the contrast is exact and deliberate: Metternich succeeded by inhabiting Napoleon's mind, Macnaghten failed by furnishing the Afghan mind with his own.
Psychology → Projection, and Why Information Doesn't Cure It. Macnaghten is a clean case of projection in the ordinary sense: attributing your own contents to someone else and then reacting to your own contents as though they came from outside. The clinical literature treats projection as a defence — it protects the self from something by relocating it.
The insight the pairing produces, which neither field states alone: the reason projection is strategically lethal rather than merely inaccurate is that it is self-repairing. Macnaghten received disconfirming evidence constantly for three years — the blocked passes, the murdered official, the officers' warnings — and each piece was metabolised into the existing model rather than testing it. The rebellions were overreactions by excitable children. The hatred would evaporate on departure. A model that explains its own counter-evidence does not degrade under contact with reality; it gets more detailed. Which means the corrective cannot be more information, and this is precisely where an intelligence apparatus fails: it is built to supply the one thing that doesn't help. The only thing that breaks a projection is being wrong in a way that costs you something immediately, and by the time Afghanistan supplied that, it was a mob in a Kabul street.
Business → Market Entry and the Grateful-Customer Fallacy. Firms enter unfamiliar markets on Macnaghten's exact reasoning: our product is better, their alternatives are primitive, they will be glad we came, and our presence will be experienced as help. The graveyard of Western retail and tech ventures in markets they did not understand is built on it.
The insight neither domain gives alone: the failure is almost never in the assessment of the product and almost always in the assessment of what the incumbent arrangement was for. The Ghilzye stipend is the perfect object here. It looked like a bribe on a British ledger; it was the price of the trade routes, and it had been paid by every previous ruler because every previous ruler knew that. Local arrangements that look irrational to an outsider are usually paying for something the outsider hasn't noticed is being purchased — a relationship, a tolerance, a suppressed conflict. The entrant who "removes inefficiency" is frequently cancelling a payment and then discovering, at full price, what it bought.
Macnaghten did not lack information. He had it, from people who had lived there, before he ever set out. What he lacked was any capacity to receive it, because his model of Afghans was made of himself — and a model made of yourself absorbs every contradiction as a detail. This is why intelligence failures are so rarely fixed by intelligence: the constraint is not what reaches you, it's what you are able to let be true.