In Kabul in December, Afghan President Ashraf Ghani told U.S. Vice President Mike Pence that more senior Taliban members had been killed in 2017 than in the previous 15 years combined. “Real progress,” Pence said. The dry language of wire reports does not reveal whether this reply was delivered in the sarcastic tone one might expect, given the utter chaos now reigning in Afghanistan. There are times when one wishes pool reporters used emoticons in their dispatches.

Steve Coll’s “Directorate S: The C.I.A. and America’s Secret Wars in Afghanistan and Pakistan” is an account of a slow-motion military and policy disaster. It is sometimes as affectless as a wire report, but the unadorned facts in its narrative more than suffice to stoke bafflement and despair.

After 17 years of war in Afghanistan, more than 100,000 Afghans are dead, and the Taliban and the Islamic State are competing to inflict wanton violence on civilians in the capital. The only thing U.S. policymakers know for sure is that the situation will degrade fast if we leave. It will probably degrade slowly and expensively if we stay.

Previous attempts at discreet draw-downs have not, Coll notes, been dignified or had positive results. In 2014, at a ceremony marking the end of a phase of U.S. combat in Afghanistan, “the ceremony program noted that attendees should lie down flat on the ground in the event of a rocket attack.”

Coll’s book is chronological, and mostly a catalog of mistakes made and lessons learned far too late, if at all.

He quotes a soldier who summarized his job to Eliot Cohen, then counselor of the State Department: “You walk through a valley until you get into a firefight and then you keep shooting until it stops.” (“That’s a little troubling,” Cohen replies.) Various strategies are attempted – the current one, conceived at the end of the Obama era, involves vigorous use of drones and commando teams – but at no point after 2003 does the United States recover the initiative. Almost every endeavor threatens to be undone in a moment.

By 2012, a quarter of the soldiers killed in the U.S.-led alliance were killed by the very Afghan soldiers they were training.

The mistakes are legion. First, in the heady days after Sept. 11, 2001, U.S. spies were fighting a war with “blood in the mouth” – an attitude that the CIA’s battlefield commander describes as a “burning need for retribution.” This attitude inspired numerous shortsighted policies in the war on terrorism, including the opening of Guantanamo Bay’s prison camp and the policy of making no distinction between al-Qaida militants and those who harbored them.

Most of the Taliban fighters were ornery yokels, with only the vaguest understanding of what America was. They did not require annihilation – of course we slowly discovered that we couldn’t kill them all anyway – and they could, at some point, have been incorporated into the Afghan state rather than hunted in endless war.

Moreover, the Taliban had provided order, and the United States had no plan to install and nurture a similarly orderly government. Afghans quickly grew irritated at an occupier skilled at fighting but uninterested or incompetent at governing.

Coll’s strongest sections detail the relationship not with the Taliban but with Pakistan.

Pakistan is a democracy of 193 million people. But the force that determines its national security and foreign policy is not its elected politicians but its spy agency, Inter-Services Intelligence, or ISI. The agency has a staff of 25,000, and it is not paranoia but good sense to assume that if you are a journalist or politician in Pakistan, its agents are watching you. Foreign government officials treat its director – always a high-ranking army general – all but officially as Pakistan’s leader. Its most secretive division, Directorate S, controls covert operations “in support of the Taliban, Kashmiri guerrillas, and other violent Islamic radicals.”

ISI has been demonized both justly and unjustly; shadowy bureaucracies tend to be spotted in the shadows even when they aren’t there. But Coll’s account of the agency makes it hard to treat it as benign, overall. The Afghan Taliban fights with ISI’s blessing, and its members drop into Pakistani territory to rest and re-equip. (More than one policymaker has concluded that this problem of Pakistani sanctuaries makes defeating the Taliban impossible.) ISI analysts themselves acknowledge the desire to cultivate Taliban fighters for future deployment, especially in Kashmir. According to one estimate, Coll says, 100,000 militants are in Pakistan on ISI’s watch.

Coll reports that the late Richard Holbrooke, tasked by President Barack Obama with fixing the region, considered ISI “obsessed” with India and thought its policy toward Afghanistan was motivated by a desire to curtail perceived Indian influence. ISI viewed the government of Afghan President Hamid Karzai as too India-friendly, and by 2003 – after the United States distracted itself with Iraq – ISI resumed its meddling in Afghanistan, to stave off Indian influence.

Coll himself is, in the venerable tradition of newspaper reporting, absent from the narrative, although his harsh judgment of U.S. policymakers is pervasive. Absolutely nothing works; “the United States and Europe,” Coll writes, “have remade Afghanistan with billions of dollars in humanitarian and construction aid while simultaneously contributing to its violence, corruption, and instability.”

“Directorate S” is one of the most unrelentingly bleak assessments of U.S. policy of recent years, and it shows, regrettably, that American errors have accumulated beyond recovery. The question is less whether Afghanistan can be saved than how its failure will affect the region.

The billion-plus citizens of Pakistan and India have now enjoyed a generation without war, and the fall of Afghanistan could contribute to a premature end to that holiday.

Coll’s previous book on Afghanistan, the Pulitzer Prize winner “Ghost Wars,” is widely considered the best book on U.S. policy in Afghanistan before Sept. 11, 2001. This superlative actually undersells it: If you tore the book into two pieces, the resulting ragged scraps would be the best and second-best books on Afghanistan, respectively. This companion volume is also definitive, if different in effect. “Ghost Wars” struck a tragic tone, with a disastrous conclusion known to the reader.

The conclusion of the policy blunders chronicled in “Directorate S” is not known. But because the errors so often look, in retrospect, unforced, they are just as painful to contemplate, and they should induce shudders as we consider the conclusion to which we might be hurtling this time.