The Age of the Unthinkable (17 page)

Shortly after leaving the Ministry of Defense, Farkash took up temporary residence at Israel’s Institute for National Strategic
Studies in Tel Aviv, where he operated from a small, sparsely furnished office on the second floor. He divided his time between
academic pursuits and business, often with companies that had once been a part of his ambit at Unit 8200. One afternoon, as
we sat over cups of bitter Israeli coffee, he began to tell me the story of his time as head of military intelligence. “After
I got the job in March of 2002,” he began, “I asked myself, ‘Where can I fail?’ I decided that there were two areas: Hizb’allah
and Iran. But to deal with this I decided to make an effort to change the way we gathered intelligence, to shift our methods
and the sorts of sources that we used.”

By way of example, he said, in the 1980s the United States moved away from so-called internal sources of intelligence, the
sort of dark black-bag operations reminiscent of a le Carré novel, and toward external sources, such as satellites and signals
intelligence, which means listening to phone calls or reading e-mails. This change had many origins, from Jimmy Carter’s discomfort
with the ethics of a cloak-and-dagger world to Ronald Reagan’s technology-heavy defense budgets. The result was to lock into
place an approach that was profoundly outside-to-inside. As satellite spy photos got better and the ability to pick a single
phone conversation out of the air evolved, spies in the United States and Israel thought they were getting a clearer picture
of what was happening. So they spent more and more money on these sorts of systems, betting that they were moving toward a
keener understanding of their enemies.

Farkash disagreed. In his opinion, such an approach had a lethal flaw: it led intelligence analysts to think that they understood
more than they actually did. Sharper satellite photos did not, in fact, mean a sharper understanding of your enemy, any more
than real-time stock market tickers meant a better understanding of the market. They were tools, not answers. “The problem,”
he explained, “is that in a dynamic environment it is the internal factors that often have the most impact.” And those internal
factors were invisible to satellites and rarely clearly stated in a phone call or a letter. “Sometimes there are things that
seem like they don’t affect you at all,” he said, “but in fact they matter a great deal. You know that famous map of the U.S.
from the cover of the
New Yorker,
where New York is giant and the rest of the country is very small? You can’t look at the world like that map, from the perspective
that you are the center of everything.”

In essence, the leap Farkash was making was to accept that Israel would never be able to fully see the world around it with
anything more than a very dim level of accuracy. There was, he thought, no sense in trying to develop a perfect picture of
the enemy. Appealing as it sounded, it was impossible in practice, and it distracted you, possibly fatally, from what did
matter, which was not what you could see but what you couldn’t. In Israel’s case, Farkash explained to me, the temptation
had been to focus only on issues that seemed to be a direct threat and to try to describe them as perfectly as possible. This
was a natural instinct — after all, with limited resources, it made sense to concentrate most on what appeared to be the greatest
danger. But Farkash was convinced this was a mistake. It was often, he thought, things that didn’t seem to be related to Israel
or that the country had never seen before that led to the most perilous situations and unpleasant surprises. The strategic
landscape Farkash faced, like Bak’s sandpile, was changing every second. It became more complex as time passed.

“You have to constantly ask new questions,” Farkash explained. “This is very difficult for a large organization like the CIA,
but for a small organization like ours it is possible. When you ask different questions, you get different answers.” Spending
years staring at the same problem blinded you to other, more important targets. And you usually (remember those USSR analysts?)
missed the answer to the question you had started with. If “right view” was the key to Zen enlightenment, this insight of
Farkash’s evoked another Zen insight: “The question is the answer.” How you asked and what you asked mattered a lot. And the
problem with the CIA or almost any bureaucracy, Farkash felt, whether it was IBM or the Department of the Treasury, is that
people have the habit of asking the wrong questions, of looking in the wrong places and in the wrong way.

“Look at Assad,” Farkash said, referring to the Syrian president, Bashar al-Assad. From Israel’s perspective, looking at Assad’s
most urgent concerns, none of the top three seemed a direct danger: the survival of his perilously perched government, made
up of Alawite tribesmen, who represent a fraction of the Syrian population; the stability of his economy, which was related
to the ability of Syria to export labor to Lebanon; and the situation in Lebanon, which Syria regarded as a “little-sister
state,” to be ordered about and shaped as Syria pleased. Only when you got to the fourth item — Syria’s thirty-year-old grudge
with Israel about the Golan Heights — did Israel’s security seem threatened. But Farkash bet that it was the balance of all
of these factors — and the injection of problems he couldn’t imagine at the time — that would determine if Syria would attack.

If, for instance, the Lebanese economy slowed down, Assad might attack as a distraction. In 2005, for instance, Hizb’allah
stepped up attacks on Israel’s northern frontier. Did this have a tie to some Lebanese grievance? No. It was coordinated by
Syria and Iran as a reaction to growing pressure on Tehran and Damascus to stop nuclear weapons development. The longer Farkash
looked, and the deeper his human intelligence led him inside Syria, the more convinced he became that these strange interconnections
drove a lot of Syrian policy. “We couldn’t just ask the question of ‘what is the threat?’ ” he told me. You might be tempted
to wait until you saw a buildup of tanks on your border before realizing an attack was imminent. You’d still be staring at
those tanks when you were surprise-attacked by sea, bombed by unplanned airplanes, or attacked in some other way you had never
dreamed of.

So, shortly after coming into office, Farkash started directing his spies to look at details that his predecessors had thought
irrelevant or regarded as second-order concerns: Were people out shopping in Beirut (a sign of the Lebanese economy’s health)?
What was intellectual life like on the streets of Damascus? How were Iraqi refugees inside Syria settling in? What interested
Farkash about such inquiries is that they were new and changing, very different from the “where are the Syrian tanks” questions
he had been told he should ask. Sometimes he would deliberately put some stress on his enemies just to see what he could learn,
and not just in Syria. A friend told me about Farkash’s idea for blowing up meaningless crates going to Iran, for instance,
or shuffling around safe houses, simply to see what would happen. Actions like this might appear to be irrelevant, but they
allowed him to watch his enemies react. And — this was every bit as crucial — Farkash said what he liked best about this approach
was that the questions stressed his system, too. They forced him and his team to look and think differently. Traditional intelligence
briefings came in binders that marked out the military situation, the economic situation, and the political situation as if
they shared only the loosest of connections. These sorts of briefings had been prepared for years. They were an old-style
way of looking. And they were, almost uniformly, wrong.

Starting in 2002, Farkash expanded his innovative approach to Israel’s problem with terrorist attacks. And in the process
he made a discovery that changed almost everything about how Israel (and later the United States) fought terrorists. Using
some of the tools at his disposal, including still-classified technologies from Unit 8200, Farkash began asking different
questions about terror attacks. Instead of the usual “Where will the next attack be?” he began asking “How long does it take
to move $10,000 into Gaza?” and “What is the best way to smuggle bomb parts
into
Ramallah?” He became less interested in why terrorist attacks succeeded and curious instead about how others had failed.
Farkash studied these failures in excruciating detail. This was a shift for Israeli intelligence. But Farkash figured that
the terrorists devoted little time to their successes but obsessed about the failures. He thought he might try to do the same
thing for a little while.

When Farkash stepped back from his surveys of these misfires, he discovered something surprising. Inside the terror webs he
was watching, he saw two different processes at work. On one level — a surface level, the one you could see with the usual
tools of spying — cells were carrying out the day-to-day business of violent fundamentalism: gathering money, recruiting bombers,
and passing talking points to the mullahs in the mosques. But another process was also under way, and it was far more important
and far harder to spot: at a deeper level the terror organizations were changing. In response to Israeli actions they would
adapt. And it was this level, Farkash realized, that was the key to understanding a terror group. It involved very few people,
a hot little core of change and innovation pulsing away inside the larger ball of a terror network. But the operators at that
level were the smartest of Israel’s enemies. The traditional way of attacking terrorism involved tearing down the houses of
people who supported the groups, jailing huge masses of Arabs who had even the lightest connection to terror, trying to cut
off routes into Israel. The goal was to “drain the swamp,” as American anti-insurgent planners in Iraq liked to say, to remove
the support for terror. It was all about attacking the outside of the terror web. Farkash knew this approach was failing;
now he began to suspect it was actually backfiring, making Israel’s enemies
more
dangerous. “Day by day, the tactics we were using,” Farkash told me, “were forcing them to be more innovative.” He paused,
and added, as if slightly incredulous, “
We forced them to evolve.
”

But his new questions had told him, at least, precisely where he should try to strike: at the level of the efficient engine
for adaptation. Doing this required new tools. Farkash and his teams borrowed network-analysis software developed for phone
systems and ecological food webs to discover where the adaptive “nodes” of a terror group were located. Farkash’s algorithms
could look at any network, whether a train system or a telephone exchange, and find the most devastating points an enemy could
attack. (In fact, there was evidence that Al Qaeda was doing this as well. The July 2005 attacks on the London Underground
occurred at three stations that a computerized network analysis might have identified as weak links in the system. London
police, running their own simulation, found that the target combination was the second best among 320 million possibilities.
It seemed unlikely that the bombers had made these choices randomly.)

Farkash found that while terror groups could involve hundreds of people, only a few had the skills that were used to help
the groups change their communications patterns, an essential part of evolution under pressure. So he aimed right at them.
And generally he began an effort to find and then isolate the terrorists who operated at this evolutionary level. What did
this mean in practical terms? “I hate to say it,” one of his colleagues explained when I asked, “and I am glad we don’t have
to decide who. But the answer is targeted killings.” Targeted killing is just what it sounds like: taking out only the most
essential pieces of a network instead of trying to wipe away the whole system. This was a complicated, expensive process.
Israeli intelligence task forces coordinated teams of hundreds of people who would spend millions of dollars for a single
kill. It also presented all sorts of moral questions for the Israeli military: for instance, how many innocent civilians was
it acceptable to kill when trying to take out a keystone terrorist? (The Ministry of Defense asked a group of mathematicians
to work on this problem. They submitted an answer — 3.4 civilians per dead terrorist — but no one was happy with either the
process or the coldness of such a figure.) The tactic, once it had been perfected, seemed to have astonishingly effective
results. The groups were unable to grow and change. Terror attacks plummeted.

There was a lesson here. After all, hundreds of now-defunct Islamic terror groups were closed files in the offices of Israeli
intelligence, having been run out of business. Some of them had looked like sure winners — well financed, stocked with angry
young fundamentalists. What went wrong? If you asked yourself this question, as Farkash did in a moment of philosophical reflection,
you found the answer was that they had failed to adapt. Hizb’allah, Hamas, the Palestine Liberation Organization (PLO) — these
surviving groups were different. They did adjust and evolve. They grew stronger. If you could stop that evolution, he bet,
they would die instantly. If you could understand and even
feel
how they grew and changed, then you could wipe them out very quickly. But first you had to be looking much deeper than traditional
counterterror spies ever had.

* * *

In the old days, many problems in international affairs seemed as if they could be understood from the outside in. If you
knew, say, how many tanks Russia had in East Germany, you probably thought you had a pretty good idea of Russia’s intentions.
Farkash’s argument was that this way of seeing was usually wrong and often deadly. It caused analysts to focus too much on
what they could measure and not enough on what they could not. What he was proposing was a way of seeing that never lingered
on a single variable. This week you might look at how many tanks Syria had, next week at the most popular program on Iranian
television. The goal was to watch for change, to see how the society was moving and understand that the snap of instant change
could come from many surprising places (Gorbachev’s regime-shifting
nomenklatura,
for instance, or the combination of air-bag technology and games). From time to time this way of looking would yield a crucial
insight — like the one about the “evolutionary heart” of terror groups — that you could act on. Then, the next day, you had
to go back to this way of seeing all over again, looking for fresh insights. What you could never do was treat what you needed
to know as something fixed, something that could be managed the way you would run down a list of things to pack for a vacation:
underwear, swim trunks, sunscreen. As the arms inspector David Kay explained to the journalist Bob Woodward about the flaws
in America’s hunt for weapons of mass destruction before and after the Iraq war, “You simply cannot find weapons of mass destruction
using a list. You have to treat this like an intelligence operation. You go after people. You don’t go after physical assets.”
He continued, “You treat it by going after the expertise, the security guards that would have been there, the movers, the
generals that would have seen it, the Special Republican Guard.” It was a suggestion right out of the Farkash playbook: look
deep, focus on things that move and change, never ask the usual questions.