Under Secretary of Defense for Policy Douglas Feith visits Pakistan in 2005.Faisal Mahmood / Reuters

To all those struggling to please their boss—to decipher what exactly needs to be “FIXED ASAP” or why exactly the TPS reports need cover sheets at all—allow me to introduce you to Douglas Feith.

Feith is best known as the top policy adviser to Defense Secretary Donald Rumsfeld during George W. Bush’s administration, and a divisive architect of the Iraq War and war on terrorism. But he is also the man who, according to documents released this past week by the National Security Archive, once received a message from Rumsfeld very simply entitled “Lists,” which read, in full, “We ought to think through what are the bad things that could happen, and what are the good things that could happen that we need to be ready for in both respects. Please give me a list of each.”

He is the man who at one point got a note from the defense secretary with the subject line “Oil” and the text: “We ought to have on our radar screen the subject of oil—Venezuela, the Caucauses [sic], Indonesia—anywhere we think it may exist and how it fits into our strategies.”

And he is the man who—as my colleague Alexis Madrigal pointed out in 2011, when Rumsfeld himself released a trove of materials from his time at the Pentagon—was once bombarded with what may qualify as the most intimidating Monday-morning ask of all time. Under the subject “Issues w/Various Countries,” Rumsfeld wrote, “We need more coercive diplomacy with respect to Syria and Libya, and we need it fast. If they mess up Iraq, it will delay bringing our troops home. We also need to solve the Pakistan problem. And Korea doesn’t seem to be going well. Are you coming up with proposals for me to send around?” (For some perspective, The Atlantic has described the dysfunctional, terrorist-harboring, nuclear-armed Pakistani government as “the ally from hell” and North Korea’s nuclear-weapons program as an “intractable problem” with virtually “no solution”; given the opportunity, this magazine would have responded to the defense secretary with a 9,000-word essay.)

The Rumsfeld Papers

Rumsfeld and his aides famously referred to these mini-memos as “snowflakes,” but they seem more like itty-bitty earthquakes designed to send recipients scrambling for cover under their desks. How do you possibly respond to these prompts? Do you drop everything and spend your waking hours trying to solve the Pakistan problem? Do you write back “Thanks for flagging this!” and then archive the message into oblivion?

On behalf of confused workers everywhere, I went ahead and asked Feith.

At the peak we probably hit two dozen [snowflakes] a day,he recalled. The snowflakes typically landed not as emails, but as hard copies accumulating in Feith’s office. “If an interesting thought came into [Rumsfeld’s] mind or he heard a notable remark from someone, he took out a little leather portfolio that held index cards and he would jot a note. Over the course of the day, he would jot notes. And then as soon as he got back to his office he would dictate snowflakes based on those notes.”

“I would take home five or six inches of paper every night”—a portion of which were “snowflakes [and] answers to snowflakes”—and then do paperwork at home from 4:45 to 6:55 a.m. the next morning, he said.

The blizzard was so routine, in fact, that the “Issues w/Various Countries” memo wasn’t especially memorable; Feith wasn’t sure of the circumstances surrounding it. But he told me how things probably played out. “Once you got to know [Rumsfeld] and the way his mind worked, it wasn’t bizarre to get [a note] like that,” Feith explained. “He understood that he had global national-security responsibilities. And so, in the course of the day, as he would have different meetings … things would pop into his head and he would say, ‘I haven’t addressed this in a while. Where are we on Pakistan? Where are we on North Korea?’”

“What he wanted first and foremost is: Do we have an understanding of what are the three or four or at most five things that are the main U.S. national interests with respect to that subject?” Feith continued. “The beginning of any discussion of what we should do has to be what should we be trying to achieve.”

“The most dramatic example of this was right after 9/11 when a large part of the government operated on the assumption that the goal of the United States after 9/11 was the same goal that we had after every terrorist attack of recent decades, which is: Find out who did it and hit them,” Feith noted. “Rumsfeld said: ‘That is really not the goal.’ 9/11 was such a major event that he said the goal of the United States should be to prevent the next attack … so you don’t have a series of 9/11-style attacks, because it could change the nature of the country forever.” Bush ultimately chose that more expansive goal, and “the difference was gigantic”: “The whole idea of the war on terrorism was … we have to make sure not only that the group that did 9/11 doesn’t do another 9/11-type attack, but neither does anybody else.”

When a snowflake drifted Feith’s way—like the “Lists” one, which referred to the early stages of the war in Afghanistan, or the “Oil” one, which referred to previous work Feith had done on energy security—he would often call his team into his office and brainstorm a response that would then be turned into a carefully crafted, and thus rather unsnowflakian, memo for Rumsfeld. “People used to think that I was crazy about editing,” Feith said. (Critics claimed that Feith botched policy decisions while fixating on the precise wording of memos.) “But they didn’t understand that if you gave [Rumsfeld] a memo that was not edited to a T, he would sometimes just toss it away and say ‘This isn’t helpful.’” In the case of the “Issues w/Various Countries” request, Feith imagines that he did a series of briefings with Rumsfeld on said issues with said countries in which he began by specifying U.S. goals and the assumptions behind those goals, and then listed the pros and cons of several proposed courses of action.

Initially, as with the items that the National Security Archive just published, the snowflakes appeared without due dates, which proved problematic in an intense line of work where hours could feel like days and days like weeks. “It was almost like a crisis for me,” Feith said. Snowflake deadlines and other reforms were soon implemented: “We wound up setting up a system of notebooks relating to snowflakes and answers to snowflakes so that we got beyond this unbelievably painful business of the secretary saying ‘I asked you for that three days ago and I haven’t heard from you’ when I say ‘I got the snowflake yesterday and we’re going to have an answer for you this afternoon.’ ... We would stamp things when they came in and they would go into notebooks and then we kept records [in] snowflake folders.”

“Every once in a while the secretary would do a snowflake review,” Feith continued. “He would call somebody into his office and go through his own stock of snowflakes and say, ‘How did you respond to this one? How did you respond to this one?’ That was pretty unpleasant. That was worse than the dentist. When that happened once or twice, I created a system where instead of going down there and trying to remember how I answered each one of these snowflakes and when, I would come down there with a book and he would say, ‘What happened to this snowflake?’ And I’d say to him, ‘What’s the date?’ And he would tell me the date and I’d open it up and say ‘Here’s the snowflake. Here’s when it came into my office. Here’s what I sent back to you.’ … Once he understood that his snowflakes were being taken seriously and he was in fact getting answers, that calmed him down a lot.”

Snowflake-style communication “wasn’t just a management technique. It was a way of living, a way of thinking,” Feith told me. “Snowflakes are in some cases requests, demands, and prods.” But many of them were also “little thought pieces”: Rumsfeld “had thousands of ideas that weren’t even half-baked that he was willing to put down in these snowflakes … and he felt completely confident that ... if somebody in the future saw it and tied it to that date, they would say that that’s a reasonable thing to have thought or asked on that date. … He was not afraid of being wrong. He was not afraid of raising something that he would later repudiate. And he didn’t mind having a record of all his thoughts.” (Those thoughts have indeed come under renewed scrutiny in recent days as a result of the latest document dump. Observers have pointed out, for example, that Rumsfeld was railing against the threat of Defense Department bureaucracy in the days before the 9/11 attacks and inquiring about the languages spoken in Afghanistan nearly two months after the United States invaded the country.)

Korea obviously still isn’t going well, and the Bush administration’s war in Iraq produced the very mess that Rumsfeld feared Syria and Libya would bring about in 2003. I asked Feith whether he ever solved the Pakistan problem. “No, of course not,” he responded.

Then he got philosophical: “Technology advances. You can say polio got solved. But if you’re talking about a problem with a country—even if you talk about something like Nazism: Nazism to a large extent got solved by the defeat of the Axis in World War II. But as we see, fascistic Nazi-style thinking can make a comeback. And while it may get solved in one country, it could pop up in another. It could even pop up in the same one under different circumstances.”

“Problems that relate to human nature never actually get solved,” Feith argued. “They simply change form.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.