Lebanon and the Intelligence Community
A case study
Lebanon and the Intelligence Community
David Kennedy and Leslie Brunetta
Since 1957 the Central Intelligence Agency has funded a program with the John F. Kennedy School o/ Government at Harvard University on Intelligence and Policy. Under this program, which is managed by ClA's Center for the Study of Intelligence, the Kennedy School conducts seminars and develops case studies that help to illuminate issues related to the needs of intelligence policymakers. This article is abridged from a case study done in 1988 for the Kennedy School of Government.
When the Reagan administration committed US Marines to Beirut International Airport in September 1982, it had the very highest of hopes. The White House meant to use American leadership and power to achieve great thinks in Lebanon: to end its festering civil war, banish occupying Israeli and Syrian armies, and infuse its battered government and armed forces with the strength they needed to run and protect their country. It meant, along the way, to bolster American influence in the Middle East, win a proxy superpower victory over the Soviet-backed Syrians, and, domestically, banish the "Vietnam syndrome" by demonstrating America's capacity for forceful and resolute action overseas. None of this came to pass. The administration withdrew a year and a half later in near ignominy, with its policy in tatters, Beirut in flames, and more than 250 Americans dead, most of them victims of a devastating and humiliating terrorist bombing.
Throughout America's Lebanon adventure, US intelligence analysts, particularly CIA analysts, were uncommonly convinced that much of the administration's policy was misguided and ill fated. They eagerly awaited the administration's call for a Special National Intelligence Estimate (SNIE), the premier instrument of US intelligence analysis, in which they might spell out the evidence and reasoning behind their reservations. While they waited, they pushed other established channels of communication, and their professional proprieties, to the limit in an attempt to get their message through. But dissatisfaction with Lebanon intelligence was almost universal: policymakers felt increasingly ill served, and analysts felt increasingly ill used. The two sides agreed only, if for different reasons, that intelligence analysis was not playing its proper role. The intelligence process may not, in the end, have offered up many insights about Lebanon, but Lebanon, in retrospect, says a great deal about the intelligence process.
Intelligence and Analysis
In the CIA's Directorate of Intelligence (DI), the Office of Near Fast and South Asian Analysis (NESA) was responsible for Lebanon intelligence. The typical NESA Middle East analyst had a graduate degree in some aspect of Middle Eastern studies and an abiding interest in the region's history, religions, and cultures: many understood at least one of the region's languages and had lived in the Middle Fast at some point.
These analysts were prepared, as was the rest of the DI, both to alert the White House and executive departments to emerging issues of importance and to respond to executive requests for analysis on any particular topic. A constant stream of different and carefully defined intelligence products, known to analysts as "art forms," flowed out of NESA. Most elemental were "talking points." CIA personnel routinely performed dozens of briefings every day (the most important usually being those that always began National Security Council meetings): talking points were typically a few topical items — the intelligence equivalent of a TV headline update — singled out by Agency analysts to be highlighted in those briefings. Talking points generally ran to hard fact, with a minimal interpretive gloss: for instance, that a particular Lebanese militia had just moved heavy weapons within range of the Marines' airport positions. It was then up to those being briefed to decide what they made of it all.
One step — but a giant step — up from talking points were the intelligence community's two morning "newspapers," the President's Daily Brief (PDB) and the National Intelligence Daily (NID). These were very similar, the major difference being that the PDB went only to the president and vice president, the secretaries of state and defense, and the chairman of the Joint Chiefs of Staff, was very secret both in theory and in fact, and thus routinely contained information of very high classification. The NID circulated as well to a large number of senior policymakers and members of Congress, was secret in theory but somewhat less so in fact, and was therefore written more circumspectly. These art forms ran about 15 pages each and were heavily weighted toward topical reporting — of troop movements, terrorist activities, meetings between foreign leaders, weapons tests, and the like — presented in a series of very brief capsule summaries, but also included one or two short (two pages maximum) analytical "feature" pieces. They were written at the CIA but drew on information from throughout the Intelligence Community.
The prestige artforms, and the only ones weighted toward analysis and interpretation rather than reporting, were the National Intelligence Estimates (NIEs) and the SNIEs. These were designed specifically to convey the Intelligence Community's most considered opinions on topics of greatest moment to executive policymakers. NIEs address matters of perennial concern, such as Soviet military capability or Central American political trends. They were long — some tens of pages minimum — prepared according to an annual schedule established by a committee of intelligence and policymaking representatives, and often took months to write. Consensus was highly valued, and each intelligence agency had formal review rights and opportunities to register demurrals_ As a measure of the weight accorded NIEs, they went out over the signature of the Director of Central Intelligence (DCI): formally, as community products, they were the DCI's estimates, although as a matter of convention DCls are expected not to meddle with their content.
SNIEs were estimates that had not been placed on the annual schedule because nobody had anticipated the need. They tended to be written in response to urgent requests from executive policymakers confronted by some sort of crisis, and followed a fast track through the intelligence bureaucracy. They were still supposed to represent the community's consensus, all agencies still had review opportunities, and the DCI still had to sign off, but the drafting process was kept down to a few weeks or, in special cases, a few days. Many analysts see SNIEs as the highest of all art forms, because they offered an opportunity for thorough, sophisticated intelligence to make a concrete, immediate contribution.
William Casey, as President Reagan's new DCI, singled out estimates from the beginning as an area deserving his special attention. He thought, according to people close to him, that estimates, as the distillation of the Intelligence Community's knowledge and wisdom on particular subjects, were very important. He also thought, according to the same people, that they were often not particularly timely or responsive to policymakers' needs. This was not a new thought in intelligence circles; according to analysts and policymakers alike, estimates very often became battlegrounds for extended interagency fights over competing views and fine points of language. To achieve consensus for the final version, the drafters often had to tone down their prose and conclusions with numerous qualifiers and reservations, as well as add footnotes to register different agencies' unresolvable disagreements. The result, too often, was that estimates expressed, mushily, the lowest common denominator of Intelligence Community opinion.
Casey was determined to make estimates more useful. He demanded that turnaround times he cut, language be declarative, and key evidence hi-highlighted. "He'd often look at SNIEs, NIEs, long-term analytical pieces, and mark them up and send them back," a colleague says. "He was proud of his own skills as a writer, and therefore was not the least reluctant to offer editorial judgment." Casey's hands-on approach to estimates served to heighten, at least for a time, CIA analysts' fears that their new boss — who as a Cabinet member was a key administration figure and whose strongly activist conservative views were well known — would try to politicize the estimation process. As time went by, that concern largely faded. The new DCI seemed to respect analysts' intellectual autonomy; in one episode carefully noted in intelligence circles, he sent back a SNIE reporting that the Nicaraguan contras — whom he supported wholeheartedly — had no domestic political base, only to release it when analysts reaffirmed their arguments. At the same time, he reserved the right to hold his own views, on the contras as on other matters, and to advise the president accordingly.
That was, for the most part, fine with analysts. Over the years, analysts and policymakers had come to an elaborate understanding of their respective rights and privileges. It was subtle and largely implicit, but both sides followed it strictly and defended it fiercely. At its core was the right of analysts to say what they thought. Nothing — not the president's policy, not the DCIs preferences, not political implications — was supposed to get in the way of objective analysis. Analysts and policymakers alike believed this (at least in theory), but analysts were often positively combative about it. "If somebody asks me something, and I have information to come to a position, then that's what they're going to hear," says one ranking CIA analyst. "I don't give a damn if it's 'helpful.' If they don't want to know the answer, they shouldn't have asked the question."
The one major and mutually agreed upon limit to analysts right to free expression was an absolute injunction against the tendering of policy advice. Whether in talking points or SNIEs, analysts had to stop on the near side of the line dividing their thoughts about the rest of' the world from their thoughts about implications for American behavior. In Casey's contra SNIE, for example, analysts could say that the contras had no political base, but they could not say further that the US should reassess its Nicaragua policy. The rule, aimed at preserving analysts' objectivity and neutrality, was almost always scrupulously observed.
When it was not, analysts soon heard about it. Policymakers, especially those in the White House who were CIA analysts' premier clients, tended to value above all else their right to make policy decisions without intelligence backchat. The Intelligence Community, in their view, was their servant, not their overseer.
In the experience of a long-time top intelligence official, NSC staff follow a very predictable course, from administration to administration, in their relationship with CIA analysts. "You go through a sort of' honeymoon period, and then a distancing," he says. "In the early days, they're sort of dazzled by all the sources the CIA has to offer. All the classification and secretiveness is very appealing to them, and so they go through an early stage when they're inclined very much to solicit the views of the intelligence agency. You can almost chart when you've passed between a year and a year and a half. At that point, no matter what their background, they become very confident in their own judgment. Their relationship with CIA analysts is superb when they have the same view. When they don't have the same view, increasingly the CIA guys will get cut out of' the picture. Will not even know what's going on."
Managing the tensions created by both sides' insistence on autonomy in their respective spheres had long been a major issue in the structure and management of the Intelligence Community. Since the Ford administration, the key instrument for bridging the gap had been the US National Intelligence Officer (NIO) system. NIOs were experienced analysts, often but not always drawn from the CIA, who were specifically responsible for ensuring that analysis responded to policymakers' needs. They were specialized (there were, for instance, NIOs for the Middle East, for the Soviet Union, for terrorism, and the like), were independent of the individual intelligence agencies, and reported directly to the DCI. Their main job was liaison. NIOs met regularly with NSC officials, assistant-secretary-level officials in State Department and the Pentagon, and the Joint Chiefs of Staff to find out what was happening and how the intelligence community might help, and with the top management and analysts of the CIA, the State Department's Bureau of' Intelligence and Research (INR) and the Defense Intelligence Agency (DIA) to organize the relevant reporting and analysis. They also coordinated and oversaw the production of NIEs and SNIEs. The NIO system was no panacea — tensions between analysts and policymakers still existed — but both sides generally found it a useful and productive link.
The US in Lebanon: The Policy
The story of relations among the White House, the State Department, and the Defense Department (but especially the White House) and the Intelligence Community (especially the CIA) during America', involvement in Lebanon was to be far more a story of tensions than one of links. From the beginning. feelings ran high. "During the summer of 1982, there was a very heady atmosphere in Washington," says one CIA analyst. "Policymakers were envisioning practically a strategic revolution in Middle East policy. They thought if they played Lebanon right, everything — the East-West problem there, the terrorist problem, the internal Lebanon problem — could all be cleared up at once. That was seen even then as naive by the analysts in the trenches. And when intelligence began to pooh-pooh this idea, it wasn't well received..."
Everybody agreed that Lebanon was important, less than in and of itself than as a cockpit of regional tension. Its recent history was one of almost constant internal strife and foreign intervention. Internal politics had been dominated by struggles, often armed, between Christians, especially the strong minority Maronite sect (and its Phalange militia), and Muslims, especially the Druze and Shiite sects (and their militias). According to an informal 1943 compact, the Maronites claimed both the Lebanese presidency and a slim parliamentary majority, but Muslim dissatisfaction with the arrangement was widespread.
Complicating, matters immensely were several hundred thousand displaced Palestinians, including well-armed and well-financed PLO militia members, who had settled in Lebanon because of its proximity to Israel and because no Lebanese Christian faction was strong enough to drive them out. The Palestinians took over a large portion of southern Lebanon, made common, often violent, cause with Lebanon's Muslims against the Christians, and often launched raids over the border into Israel.
In 1975, Lebanon's religious and political tension erupted into a civil war. The fighting was vicious, and atrocities common on all sides. The Christians, unable to handle both Lebanon's Muslims and their Palestinian allies, eventually invited Syria's President Hafez Assad to send forces to stop the fighting. Assad did so; he sympathized mainly with Lebanon's Muslims, but as a regional power broker he was more concerned with preventing Lebanon from becoming an even more powerful Palestinian stronghold.
Once in Lebanon, Assad — who had long aspired to annex Lebanon as part of a historical "Greater Syria" — never fully withdrew, keeping a garrison in Beirut and occupying much of the country's north and east. The Syrian president, who had strong Soviet ties, continued to support Lebanese Muslim parties and militias to keep the Christians weak and off-balance. Beirut itself split into its Muslim western half and its Christian eastern half, each controlled by a bewildering patchwork of sectarian militias.
The Palestinians remained strong and active, and Israel responded to the PLO in 1978 by taking over a belt of Lebanon just north of their common border and placing it under the control of a renegade Lebanese army officer sympathetic to Israel. By the early 1980s, the Lebanese Government, such as it was, and the Lebanese Armed Forces (LAF) held unchallenged sway only over the presidential palace and a few square miles of downtown Beirut.
The situation was a constant worry in Washington. With Syrians and Israelis so uneasily and narrowly separated, another Middle East clash, possibly escalating into a superpower face-off, seemed only a slip away. Nor did the situation seem, in any wise, stable. Israeli Prime Minister Begin and Defense Minister Sharon, the powers in the ruling conservative Likud Party, were known to be looking for a way to break the PLO's back in Lebanon for good and remake Lebanon's political landscape in Israel's favor. They were negotiating terms with a young but rising Lebanese Maronite strongman. Bashir Gemayel: if Israel moved into Lebanon and subdued the Palestinians (and perhaps even the Syrians), and helped install Gemayel as president, would he ensure that the PLO stayed toothless in the long run, and perhaps even sign a peace treaty with Israel? (Lebanon, like all other Arab states except Egypt, had technically been at war with Israel since 1948.) By early 1992, the US Intelligence Community was convinced that Israel would shortly invade Lebanon.
On 6 June, Israel did. The Israeli advance was resisted by the PL0 and the Syrians, but to little avail. The Palestinians were beaten hack, and the Syrians agreed to a cease-fire after Israel, in what was widely regarded as a stunning military tour de force, destroyed the extensive Soviet-supplied Syrian air-defense system in Lebanon and shot down a quarter of its air force. Early pledges to the US to stop the invasion when it reached 40 kilometers into Lebanon were quickly broken, and Israeli columns soon reached and laid devastating siege to Beirut, cutting off power and water to Muslim quarters and pounding, the city, over strong US protests, with heavy artillery and airstrikes.
The US made itself a key player immediately. President Rowan directed veteran diplomat Philip Habib, already in the region as his special envoy, to arrange a general cease-fire. Over the course of the summer, as the Israeli strikes continued, Habib did so. Late in August the bargain was struck: the PLO would evacuate its 15,000 military personnel in Beirut to other Arab countries, Israel would lift its siege and let them go, and the US would guarantee the safety of the Palestinian civilians left behind. Things moved fast thereafter. On August 23, with strong US and Israeli support, the Lebanese parliament elected Bashir Gemayel president. Gemayel, who had a "special relationship" with the CIA's Directorate of Operations, personally assured Habib that when the PLO military was gone their families would he safe. A contingent of US Marines, in company with French and Italian forces, was brought in to supervise the evacuation.
During the evacuation, on 1 September Reagan moved to capitalize on America's sudden Middle East eminence by announcing what came to be known as the Reagan Plan. The plan built on the Camp David accords by proposing that Jordan's King Hussein negotiate with Israel on behalf of displaced Palestinians, probably for some sort of Jordanian-administered entity in the Israeli-occupied West Bank. Although Begin and his Likud Party immediately rejected the idea, Jordan and other Arab countries cautiously endorsed it. The administration was pleased; even Israel's rejection had some positive aspects, as it increased US credibility with the Arab world. Meanwhile, the evacuation went off without a hitch, and the Marines were called home.
Victory, as Washington viewed events thus far, soon turned to ashes. On the afternoon of 16 September, Gemayel was assassinated at his Phalange militia headquarters, almost surely at Syrian direction. Within days, the Israeli Army allowed Phalangist militiamen into two Palestinian camps, Sabra and Shatilla, in Beirut, where they murdered at least 700 civilians in reprisal for Gemayel's death. The United States, France, and Italy, horrified that the US guarantees of Palestinian safety had been so bloodily violated, moved their forces back into Beirut.
This time, Reagan's goals were far more ambitious, as he made clear in a television speech shortly after the Marines returned. Now he meant to cure Lebanon's plight. The Marines, he said, were in Beirut as a peacekeeping force pending the withdrawal of "all foreign forces" and to assist in Lebanese state-building. Spelled out, that meant that the US was going to try to send both the Syrians and the Israelis hack home, ending, their direct involvement in Lebanese affairs, and to strengthen Bashir Gemayel's brother, Amin, who had in turn been elected president, until he could run Lebanon. The Reagan administration was determined, says an NSC official, "to let Lebanon he Lebanon."
The president's policy was, according to a variety of sources, likely that of special presidential envoy Habib. He had Reagan's ear and respect. Moreover, Secretary of State Alexander Haig, who had had strong policy views on the Middle East, had resigned in June, and his replacement, George Shultz, was not exercising the same sort of influence. And although usually powerful Defense Secretary Caspar Weinberger entertained strong doubts about Habib's policy — he believed the Marines lacked a military mission in Lebanon and were at substantial risk there — he had been overruled in bitter debate.
"Habib sort of took the decisionmaking process by storm," a State Department official says. "In Habib, you had a guy who was on the scene, who has an imposing history, who's respected, and who has the confidence of the president and acts accordingly." Habib had such primacy that even much of the administration's foreign-policy bureaucracy was left out of Lebanon matters. "He was handling day-to-day policymaking basically out of his hip pocket in the field," says Bing West, who was then Assistant Secretary of Defense for International Security Affairs. "It was not an interagency process. It was basically Habib reporting back to the White House and to Shultz, and the interagency process trotting along behind."
Habib was confident that he could negotiate speedy Israeli and Syrian withdrawals: so speedy, he thought, that the Marines would be home by Christmas. His strategy was to accord diplomatic priority to brokering an Israeli disengagement agreement with Lebanon. He believed that Syria would withdraw almost as a matter of course once Israel did, a belief "premised," says a senior White House official of the time, "upon Saudi statements, and winks and nods, that yes, we know — we the Saudis — that when you get Israeli withdrawal, you will get Syrian withdrawal, and that Syria has no greater ambitions vis-a-vis Lebanon." Habib and Shultz also believed that the US could count on moderate Arab — chiefly Saudi — pressure to move Assad when the time came. Thus, Syria was kept generally informed of US activities as Habib moved to get Lebanese-Israeli talks under way, but Habib chose not to involve Assad more deeply. Once the Syrians and Israelis were home, Habib thought, Lebanon could be reunited under a Gemayel presidency supported by a revitalized LAF. It would he a great American victory.
The US in Lebanon: The Analysis
The Intelligence Community doubted it. Almost every aspect of Habib's policy was subject to grave reservations among Middle East analysts, who were in uncommon agreement in their views on the matter. "There was a very unusual unanimity of opinion on the realities of Lebanon and the costs of being involved there," says a Defense intelligence official. "Usually there's a tendency in the intelligence coordination process — among CIA, DIA, INR — to soften judgments. With Lebanon, there was a tendency to make it even harder, even alarmist, to point out to policymakers — hard over, up front, early on — that this is a real can of worms."
The doubts began with Syria. All students of the Middle East — not just those with access to classified information — recognized Hafez Assad as one of the region's most determined and fractious leaders. Analysts believed that even if his old ambition to absorb Lebanon were frustrated, he would still regard Lebanon as his back yard, and that he would still be determined to remain a major player in Lebanese events. They feared for any Lebanon policy that slighted or underestimated Syria's interests and power. "What happens in Lebanon is vital to Syria," says one INR analyst. "There was no way Assad was going to let things happen there unless he got something out of it."
The analytic community also believed that the Reagan administration was being overly optimistic about the chances of a quick Israeli withdrawal. Israel's interests, they thought, ran against the existence of another powerful, autonomous Arab nation in the region, particularly one lying along its own borders. And the Reagan Plan, whose aims for the West Bank Israel viewed with dismay, gave Israel a big stake in delaying any resolution in Lebanon, because Jordan's King Hussein had made it clear that while he might negotiate with Israel, he would not do so as long as it occupied Lebanese soil. Prudence, the analysts thought, required skepticism about Israeli pliability.
They also doubted the wisdom of the administration's fundamental goal of unifying Lebanon. "Lebanon is not what it seems," says a Defense intelligence official. "It's not a nation: it has no borders, you can't say what a citizen is. People in the administration went into this thing with blinders on." In particular, long-time close observers of Lebanon's murderous political scene did not recognize Amin Gemayel as either a statesman or a strongman. Nor did they see his family, or the Phalange militia it more or less controlled, as representative of' what passed for the Lebanese polity, or even of its powerful Maronite Christian minority. The Gemayels and the Phalange, most thought, were really just another faction, albeit a strong and important one. "My personal feeling about the Reagan administration is that at that point they tended to see the world in terms of black-and-white, good-or-bad," says a CIA analyst. "This led them to see the Christians in Lebanon as good and the PLO, the Muslims, and the others as bad, when we knew that the Maronites were just as ruthless and manipulative as anyone else."
The analytical community feared, finally, the implications of using Marines as an instrument of US policy in Lebanon. Even before the Marines had returned, INR analysts had explicitly warned that the Marines would not be able to play the neutral role the administration had written for them. "The government in Lebanon is not seen as the Lebanese government, it's seen as the Christians' government, and not even all the Christians recognize it as their government," an INR analyst explains. "Therefore, any move on our part to support that government would be seen as support for the Christians, not as support for a settlement. The presence of' the Marines would be seen as putting off an inevitable distribution of power, and sooner or later they would become targets. I remember someone high up saying, 'Don't write that or the Defense Department won't go in.' But we're independent, and it was in our daily reports." The rest of the analytic community agreed. All in all, analysts had little but foreboding for America's future in Lebanon.
Waiting for the Phone To Ring
The Intelligence Community very much wanted to spell out its case, and waited eagerly for the White House or another part of the administration's policymaking bureaucracy, to request a SNIE. It waited, as time went by, in increasing frustration. "The NSC didn't come calling on us," says one CIA Middle East analyst. "They didn't know enough to know what they didn't know." As a result, analysts had to put off addressing, at length and in one place, the kinds of questions they were eager to answer during a crisis: Who are the major players here? What do they have at stake? What will they settle for? How easily will we be able to influence them? How will US actions be viewed?
There were various theories among analysts about why the call did not come. Decisionmaking, it was fairly clear, had become highly centralized in Habib's inner circle, although exactly what he and his circle were up to was less clear (though analysts had no way of' knowing it, even the president's national security adviser, William Clark, was experiencing some difficulty following and shaping Habib's actions). Habib clearly had strong feelings of his own and, presumably, felt no need for analytic contributions. There were also signs that the inner circle had already settled all of the strategic, and many of the tactical, questions.
One former mid-level Pentagon official recalls that he, and the intelligence representatives he worked with, often emerged from policy sessions wondering why they had even been asked to attend. "I think the Intelligence Community and working-level policymakers were often feeding information up to deaf ears," he says. "Phil Dur [an NSC aid] would come to meetings where we were supposedly discussing possible policies, and referring to the intelligence we had, and say, "That's all tine, but we're going to do this, the decision's already been made." The NSC adviser and the special envoy made all the policies, and it didn't matter what anyone else said." The same official also felt that analysts were rather deliberately disregarded. "Senior policymakers knew that the Intelligence Community was opposed to what was going on," he says. "They just didn't want to hear about it."
All analysts knew for sure, though, was that they weren't getting the chance they wanted to write their SNIE. They were left with lower-level art forms to get their points across. They wrote talking points almost every day for Casey's briefings to the president and the Middle East NIO's briefings at the assistant secretary level, and they reported key bits of intelligence in the PDB and the NID. "These were much more on the tactical day-to-day stuff than big perspective pieces," one State Department official recalls. "What's new with Begin? With Gemayel'? How many men do the Israelis have? What river have they reached'? I don't remember any big warning that getting involved in Lebanon could be hazardous to your health."
Many of the bits and pieces reported this way were not particularly encouraging: before long, for instance, it was known that the Soviet Union had begun a $3-billion military resupply of Syria, that the Lebanese Christian community was not rallying behind Amin Gemayel, and that Israel was changing the road signs in southern Lebanon to Hebrew. Analysts, certainly, were trying to print a consistent and cautionary picture. "I think, in the overall context of the NID and the other smaller stuff, that our reporting was as strong as it would have been in a SNIE," one senior CIA analyst says.
The trouble was, as analysts saw things, that pulling their main arguments from the welter of fact in the lesser art forms was something of an exercise in connecting the dots. "Shultz. complained to Casey at one point that an important analytical line had not shown up in the NID," the CIA analyst says. "I thought he was wrong, and when I checked — read months of reporting — I did find it there, but much more obscurely than I had remembered. It was very much present, on the other hand, in our talking points and some of our own internal work. The lesson is that lots and lots gets said, but in different places in different ways, and you can never be sure just what gets across."
Few in Reagan administration policymaking circles seem to have felt the lack of a comprehensive Intelligence Community estimate oil Lebanon. "I've never been a tan of estimates," says a senior State Department official. "They're usually mushy and cautious. I'm kind of interested when they're on a subject I don't know about, because I'll pick up facts. But estimates often seem just to be instruments of bureaucratic warfare." The same official, on the other hand, welcomed the daily reporting. "A policymaker usually has some expertise of his or her own, after all," lie says. "I use the Intelligence Community as a resource of factual information, but I don't need it for opinions. I have my own."
Many in intelligence, though, thought it was fundamentally incorrect to approach Lebanon — and other similarly complex foreign policy problems — solely through the mechanisms of daily reporting. "The problem is when you're dealing with mysteries rather than secrets," says a senior CIA official. "The Lebanon policy as it evolved was based on mysteries, not just facts." In his widely shared view, the important questions in Lebanon were the intangibles: how far can Assad be pushed? What will the Druze settle for? What is Amin Gemayel's potential for growth? Unless the Intelligence Community stumbled on very unusual information — some kind of definitive proof, for instance, that Assad would not leave Lebanon — these were "mysteries" on which reasonable people could differ. Analysts' only edge in such a debate was superior knowledge and insight. It was an edge they readily claimed. "Lebanon was so complex that it was hard to get the NSC to concentrate on its intricacies," says one CIA analyst. "When you told people there were 40 militias operating in West Beirut alone, you could see their eyes glaze over. So they didn't know who the actors were or what they wanted. But I've been following the civil war in Lebanon since it broke out: that's my career. We couldn't say what we wanted to say in bits and pieces."
Despite the Intelligence Community's frustration at not being asked to write a comprehensive Lebanon estimate, analysts did not try particularly hard either to spark a request for an estimate from somewhere in the administration or to volunteer something comprehensive of their own. It was not that they simply did not care — something that had happened in the past, according to the analysts. "There have been times," one says, "when we've been content just to sit here and be right." Rather, it was that in tandem with the injunction against analysts offering policy advice, there was also a very strong expectation that the analytic community would not speak unless it was spoken to. It was there to provide a service, as defined by its policy masters, and it focused its attention on areas selected for it by those masters, not on other areas it might find of independent interest.
As a practical matter, it was a limitation only on analysis, not on reporting raw data. Analysts were largely free to study, and to think, what they liked, but they were not free to fire off their conclusions at will. Larger and more explicitly political judgments, especially — as about contras' standing — always required a specific request, from, say, an NSC aide.
These restrictions had evolved over the years as the intelligence and policymaking communities sought to ensure that analysts would play an objective, neutral role in the policymaking process. Intelligence and analysis would be sullied if analysts shaped, or were thought to shape, their views to support executive policy factions, likewise, if analysts themselves promoted, or were thought to promote particular policies. "There's sort of an analysts credo," says a senior CIA analyst. "It says, 'I'm ;in analyst; I am objective and pure. I know that the closer I move to policy, the more likely people are not to see me that way anymore.' " Where Lebanon was concerned, therefore, despite the desire of analysts, particularly CIA analysts, for a formal estimate, volunteering what would have been a thinly %cited comprehensive critique of the Reagan administration's major foreign policy initiative was never seriously considered.
Though CIA analysts followed the rules oil Lebanon, many are nonetheless not entirely comfortable with how they handled things. "We could, on our own hitch, have done something, like alternative scenario papers on what Syria would and would not accept," one midlevel analyst says. "We never looked hard at that in any formal way. But even that would have been walking, a fine line. Analysis shouldn't be prescriptive, and much of what we presented on Lebanon was bad news that wouldn't let tile policy go ahead Smoothly. It was very difficult."
There was plenty for analysts to feel uncomfortable about, for Habib was making little diplomatic progress, the Marines did not come home at Christmas, and the first months of 1983 showed little more movement. As the months went by, the analysts' bits and pieces analysis, and its Cassandrish tone, seemed to wear increasingly thin. "Analysts were never penalized for being overly pessimistic, only for being overly optimistic," says one State Department official. "The intelligence people would remind us over and over again of the difficulties. I'd Say, "Your intelligence is true but it's not conclusive. The US has the ability to influence events."
It also grew increasingly sensitive. Caspar Weinberger was resisting the Department of Defense role in Lebanon ever more strongly, and the NSC and the State Department were fighting to hold off his objections. The struggle's glare served to highlight even lesser analytic findings. When two CIA analysts presented a set of talking points outlining unpleasant forces that might be set in motion by a proposed program to use US military advisers to train the LAF, they quickly heard that policymakers were not pleased. "I remember being told, 'This is the president's idea and you're telling him it's a bad idea,' " one says. "And l said, "No, I'm not. I'm just saying that these are the possible consequences you should take into account as you make your decisions. But it was clear they thought I was taking sides."
Langley's Circumscribed View
Most analysts had little notion at the time, however, that sides were being taken. This was not necessarily something they felt they needed to know, because there seemed no good reason why the struggle between State and Defense ought to have affected their analysis. But they were also in the dark about much that clearly did matter. Working-level CIA analysts, for instance, had no idea at the time that special envoy Habib was predicating his entire strategy on the premise that Syria would all but automatically withdraw from Lebanon once Israel did. They had no idea about the terms of the Reagan Plan until they were published in the New York Times, and no idea of what was under discussion in the Israeli-Lebanese disengagement talks Habib was tiled ialing, the keystone of his plan to obtain Syrian and Israeli withdrawals from Lebanon, until an agreement was finally ratified on 17 May 1983.
Though the analytic community's charter was to support the rest of the executive's policy activities, nothing, ensured that they were privy to the actions policymakers took and the choices they faced. Lebanon was not exceptional in this regard. "I conclude, looking back," says a recently retired senior Intelligence Community official, "that I frequently had a far clearer view of what was happening in policy formulation in other countries that I did in my own."
There was a real price paid in the relevance of the Community's analytic work. Syria, for example, rejected the 17 May agreement out of hand as unacceptably advantageous to Israel (Assad also declared Habib persona non grata in Damascus, and Robert McFarlane replaced him as special envoy). When the CIA's analysts learned of the agreement's terms, they were dumbfounded that Habib had let it go through. "Anybody who thought that Assad would buy 17 May," one says bluntly, "was smoking pot." Secretary Shultz, at least, felt ill served. "Shultz thought that the intelligence hadn't adequately warned him that Syria wouldn't accept the agreement," the analyst says. "Well, I could have written a good Syria estimate. But we didn't even know the terms until we saw them in the newspapers."
Because the Intelligence Community is forbidden to monitor US behavior, most of the information that comes its way about US activities, both foreign and domestic, does so at the pleasure of its policymaking masters. The process can be capricious, especially between the CIA and the White House. "There's a fierce independence at the NSC that says, 'Everybody else works for us,' " says a longtime top intelligence official. "There's no systematization that ensures analytic input into NSC decisions. It's always been resisted at the NSC level because it would constitute a check on their autonomy." The common result is a nagging uncertainty in analytic circles about just what the US is up to.
The situation was exacerbated by a number of mundane but still very real factors, chief among them distance and security. "First, you just have the mechanics," says the official. "It's a lengthy trip down, lat least a half-hour drive from Langley to downtown Washington]. It's difficult to get a parking place anywhere near the offices of the State Department or the NSC. You've got the problem of how do you get in: the security badges are no longer given out, as they used to be in the old days, to let you sort of go in and wander and visit. So, unlike the Pentagon, where it's still fairly easy for someone to get a badge and go wander the halls and work their constituency, you can't do that at the NSC. There isn't anything to naturally encourage the informal visit."
The situation was not much different traveling from Washington to Langley. Visitors had to phone ahead for security clearance. stop at drive-through checkpoints, and then wait in the building's lobby for their host to come and escort them through yet another security checkpoint. (In a nice bit of bureaucratic skirmishing, some Pentagon officials have begun to refuse to visit CIA headquarters because CIA officials coming to see them are put through much less rigorous security hoops.)
There was nothing in that process to encourage the informal visit, either. "Every once in a while I'd find time to go talk to the Agency analysts," says a mid-level Pentagon official who worked on Lebanon for the Office of the Secretary of Defense. "Every time I did it was really useful. But it took so long I could hardly ever manage it, it really meant killing halt a day."
These problems most vexed CIA analysts; there was far less distance between policymakers and INR and DIA analysts. For one thing, INR analysts had extensive access to Habib's cable traffic, which meant they were far more keyed in to current events (although they clearly chose not to share all they knew with their CIA colleagues). Both DIA and INR analysts also enjoyed far closer, less formal relationships with the policymakers they served. "Tile INR has direct access to the secretary of state," a top INR official said. "INR analysts deal with policymakers every day. They attend State Department staff meetings, they can knock on doors when they have questions or suggestions." DIA analysts report easy, regular contacts from the Office of the Secretary of Defense and the Joint Chiefs of Staff on down through the defense bureaucracy. "At all levels of the DIA, people were talking to their opposite numbers in those two organizations," a DIA analyst says. Chairman of the Joint Chiefs General John Vessey was particularly noted for reaching down to analytic specialists for personal briefings on issues of the moment.
Relations between the CIA and policymakers were almost always handled at arm's length and on paper, even during crises. And crises, as analysts saw things, were when those methods worked the worst. As one CIA analyst put it, an analyst monitoring a spot not figuring highly in US policy would probably be routinely furnished almost all relevant policy documents because they would be so few, of such low sensitivity, and because there would rarely be any hurry to make decisions. The resulting analysis would therefore be remarkably well grounded (even if hardly anybody in policymaking circles would be very interested in it).
In crises, the tables were turned: much was sensitive, everything was rushed, and analytic access to policy circles became catch-as-catch-can. "Special emissaries and their entourages tend to work around the formal chain of command," says one senior CIA analyst. "That's not wrong, but it's a complicating factor. We get out of the loop, and we don't get to see the factual information and meeting papers that you usually expect to see." Another senior analyst echoes this frustration: "Because you get cut out, even if you're cut in you never know if you're cut in absolutely. We had no access to secure telephone lines, and when Ambassador X calls so-and-so at the State Department, it may never get down on paper. This limits in a needless way the precision of the contribution we can make to policymakers. You know some of what's going on, and you may know all you need to know, but you don't know that. So you're often missing part of the equation."
Part of the equation complicating matters for CIA analysts was the Directorate of Operations' "special relationship" with Bashir Gemayel. Even before the Israeli invasion, a covert network had been in place for dealing with the Gemayels, a network that would have bypassed the analysts completely. Whatever information could be learned from Bashir Gemayel, and later from his brother Amin, and whatever actions the Maronites undertook at the behest of top US administrators would be largely beyond the analysts' ken.
One Intelligence Community senior official suspected that the Gemayel relationship may even have accounted, at least in part, for the decision not to initiate a SNIE. According to this official, that wouldn't have been an unusual outcome: "If a policy matter is drifting toward a policy decision that may end up introducing covert operations, and the DCI is sort of inclined toward that, there will indeed be a couple of people at the CIA who know, they'll be the people in Operations, but in those kind of situations it unhappily is not uncommon for the analytical side to be cut out entirely. And not even know that a topic is under discussion."
Even policymakers, though, report difficulty keeping track of all that was happening around Lebanon. "When the Lebanon crisis hit, it was early in the Reagan administration," a Defense official says. "The administration had not really jelled yet. So when this crisis landed, it was the big game in town. It was almost a sign of how important you were to get in on that act, and to be as much a part of it as possible. Everyone at high levels was trying to get in the rowboat: show that they were in on the action, going on the trip, that their organization was in on it. That's a destructive syndrome. You want your information to be coming in like a Mozart concerto, instead of like Niagara Falls, which is the way it came in."
Few who did get in on the act took care to keep the Intelligence Community up to date on what they learned; many bypassed even their own organizations. Army Gen. Carl Stiner, for example, who traveled as the Joint Chiefs' representative with special envoy Robert McFarlane, reported directly to the chiefs, bypassing not only the Intelligence Community but even the military chain of command. "We'd go to meetings at Defense," one administration official says, "and their intelligence people and uniformed people would be saying one thing, and the chiefs would say, 'Well, that's not what Stiner says.' It was crazy. Everybody's entitled to their own opinions, but we couldn't even get together on the facts."
Analysts, in fact, believed that they were essentially locked in a losing battle for policymakers' attention. "There was so much competing information coming to them from the Israeli leaders, from Mossad [Israeli intelligence], from the Lebanese Christians," says a CIA analyst. They didn't begrudge policymakers those sources (at least not explicitly). They did, however, still feel that they could make a special contribution evaluating the information that came in that way, a contribution they couldn't make without full access.
Less directly, because they knew little of what came to pass in meetings between Habib, Shultz, or their parties and Begin, Assad, the Gemayels, and other Middle Eastern leaders, they felt unable to judge those leaders' actions intelligently. How could they say what Assad was up to when he was surely reacting to what Habib was doing, and Habib's actions were themselves a mystery? "This kind of thing allows policymakers to say, I hope without malice, 'Look, you just don't have the whole picture,' " says a senior CIA analyst. "Sometimes I think it is malicious. The only reason we don't have the whole picture is that they won't tell us."
NIOs: The Missing Link
Policymakers are familiar with the analytic community's complaints about being cut out, on Lebanon and on other matters, and many — if not most — will have none of it. "Whatever the Intelligence Community might say, elements of the Community who had an interest in the Middle East were involved throughout the process at every level, whether it was in interagency working groups at the lowest level or right up to the National Security Planning Group: meetings, conference calls, whatever," says an NSC aide. "Either the DO or his deputy or representative was always there. So they have no reason for saying they didn't know what was going on. Whether those people managed their system well so that their people knew what was going on, I can't speak to, but they were always represented, they always got sensitive materials, and there was very little involvement by the policy community on the operational side of what the CIA was doing."
For the most part, the representative at those meetings, and the only one explicitly charged with ensuring that relations between policymakers and analysts were smooth and productive, was one of the three analysts who in turn served as Middle East NIO over the course of the Reagan administration's involvement in Lebanon. They alone mixed a substantive role, as the Intelligence Community's executive regional analyst, with the procedural responsibility of maintaining effective liaison between the two communities. By all accounts, they enjoyed access, and exercised frankness, out of bounds to even the most senior ordinary analysts. "Graham Fuller [Middle East NIO from the middle of 1983 on] and I had lunch all the time," says a senior State Department policymaker. "He was the Cassandra on this, always saying that Syria would never go along and that the moderate Arabs wouldn't be able to exercise any influence over Assad." But they were not able to correct, and to some extent may have exacerbated, miscommunication and ill will between the two groups they served. NIOs, as Lebanon makes clear, play by complicated rules of their own — rules they to some extent make up as they go alone. Considerable confusion can occur along the way.
The central challenge of an NIO's job is to inform analysts of what policymakers think and need sufficiently to allow them to do timely, relevant work, while protecting policymakers from the exposure of particularly sensitive thoughts and plans. The two parts of' the job inevitably conflict. To get the most out of analysts, NIOs would have to routinely pass on everything they pick up, even in the most rarefied policy circles; to most thoroughly insulate policymakers, they would have to pass on very little. Neither suffices, and NIOs routinely inhabit some middle ground: but only they, usually, know just where that ground ends.
Robert Ames, for instance, a legendary CIA Middle East specialist who was first Middle East NIO, then chief of the CIA's NESA office (a job he held until he was killed in the April 1983 bombing of the US Embassy in Lebanon), enjoyed exceptional closeness to the policy process. George Shultz was so impressed with his depth of knowledge that he made him, first, part of the small team that formulated the September 1982 Reagan Plan, and then part of the team backing up Habib on what became the 17 May agreement.
It was a dream role for an NIO, but the analysts he oversaw nevertheless remained in the dark about such things as Habib's policy toward Syria and the substance of — the disengagement talks — things Arnes certainly knew. "Ames's being on the Shultz's policy planning group was a sanity check to us, but there were limits to what he could tell us," a senior CIA analyst says. "At certain points he just had to leave it at, 'Trust me.' And that's right; NIOs should feel constrained."
Further down the line at the CIA, working-level analysts found themselves swimming in even murkier waters. "Ames was in constant meetings with Shultz and the others, and he would send back for information and analysis," one says. "But we didn't know the context." They would respond as best they could, but they never knew for sure if they were contributing anything useful.
Ames, and the NIOs who followed him, all had good reputations among the analysts under them for being as open as possible. "Some NIOs don't tell you anything because they can't be bothered, or they don't like what they had for lunch, or God knows why," a CIA analyst says. Ames et al. were not seen as arbitrary; their basic rule seemed to be, when uncertain about what it was safe to pass on, to err on the side of caution. The State Department and the White House were not of their own accord telling analysts the substance of the May 17 talks; therefore NIOs would not, either. Only NIOs, however, knew where they drew the lines on any particular matter.
Analysts at least knew that the lines were being drawn. Many policymakers didn't seem to; they just thought that analysts were being unresponsive. Some policymakers didn't seem to recognize that NIOs too subscribed to the analysts' credo. Ames, for instance, although part of Shultz's inner circle, refused to tender policy advice even when asked, according to another member of the 17 May negotiating team. And when Graham Fuller, who reportedly had very strong feelings about what the US was doing in Lebanon, was NIO, he wouldn't say, "You should do this or that," according to a CIA official. "He'd say, 'Shouldn't you be cautious about this or that?' He'd he careful not to cross that line." To many policymakers, analysts — as represented by the NIO and the art forms — didn't seem to be taking into account information that policymakers had given the NIOs; they also seemed to carp chronically without providing any clear-cut warnings or advice about opportunities in Lebanon. "The Agency ran its own show," an NSC aide says. "The hierarchy knew what was going on." The resentment, in some policymaking circles, was palpable.
The October Estimate
By the fall of 1983, events in Lebanon were making it very clear that the administration's policy was not going ahead smoothly. The US was apparently no closer than it had been a year before to getting the Syrians and Israelis home; despite the 17 May accord, or perhaps because of it, both armies were firmly entrenched. Amin Gemayel's presidency seemed to be growing ever more precarious as Syrian-backed Muslim factions pressured him to renounce his treaty with Israel. The Marines suffered their first casualties at the airport, caused by shellfire from Druze turf in the surrounding Shouf' Mountains, late in July, and continued to take intermittent hits from that point on. On 28 August they fired, for the first time, on Druze positions in the hills, and were shortly declared enemy forces by influential Druze leader Walid Jumblatt. Then, in the first week of September, Israel pulled its army out of the Shouf part way back toward its own border. The LAF, backed by the Maronite Phalange militia, tried to take the ground the Israelis had vacated and were beaten back. US vessels off Beirut deployed both F-14 fighters and naval gunfire in their support.
As the center ceased to hold in Beirut, analytic activity in Washington quickened. The call for an estimate still did not come, however. "In August, with the fighting in Beirut's suburbs, the questions picked up," says a CIA analyst. "Policymakers wanted to know what the capabilities of the various factions were, but not what was the overall dynamic. Usually we're asked that question, but not that time." Beirut continued to crumble. Late in September the US battleship New Jersey, which mounted enormous 16-inch guns, took up station off the city. Robert McFarlane negotiated a cease-fire with Assad in Damascus and the fighting in the Shouf and the airport shelling« ceased. It was a good sign, but before Gemayel could even carry out his part of the bargain McFarlane had struck for him — undertaking negotiations with Lebanon's important factional leaders — Jumblatt took a giant step toward partitioning the country by putting the Shouf Mountains under a Druze "civil administration council" and calling for the desertion of all Druze LAF personnel. Over 600 Druze complied.
At this point, the US Intelligence Community finally got a chance to write its Lebanon SNIE. It's not clear just how or why the charge to write the SNIE came down; classification clouds the genesis of particular requests. Analysts speculate that things had gotten so bad that the NSC was moved to ask; that the Pentagon — which was trying hard to convince President Reagan to pull the Marines out of Beirut — might have sought the estimate to use as ammunition in its bureaucratic battle with the State Department; even that Graham Fuller, then the Middle East NIO, might finally have been driven to request it himself. In any case, the time had come, and the analytic community, especially the CIA, was primed.
"The October SNIE was a watershed event," one midlevel Agency analyst said. "It was the first time the entire Intelligence Community was putting its imprimatur on the views that had been present all along in the daily intelligence. At one frozen moment in time, it says, "Here's how all the people in the community feel." Even some who thought that most of what the community wanted to say had probably, over the past year, seeped through in the daily intelligence, relished the chance to write the estimate. " I can't believe that it wasn't known what that SNIE would say," says one CIA analyst. "But there was, maybe, a certain amount of "We've been telling these ... people ... this, and we're going to tell them again, and this time we're going to rub their noses in it."
From the time Fuller called representatives of the various intelligence agencies together to plan the SNIE, it was clear to everyone that it was going to he unusual. To a degree that participants say was totally unprecedented, the CIA, the DIA, and INR agreed on virtually every point the SNIE was to make. The estimate was correspondingly strongly worded. "It was, as we would put it, a 'starchy' estimate," says a senior CIA analyst. "That wasn't surprising; the situation was so hopeless, I don't see how it could be any different. We live in a world of analytic grays, but this was clear-cut. There was no need to try to anticipate and adjust for the usual DIA/INR shadings. Also, Casey and [deputy DCI Robert] Gates had been pushing for more policy-relevant estimates. Still, I don't remember any others — ever — that read like this one did ..."
The SNIE stressed the Community's belief that Syria would not be swayed on Lebanon, that Assad was powerful and able, that he would not walk away without some major prize, that the 17 May agreement offered him nothing, and, that having learned from America's experience in Vietnam, he would wait out any military pressure the administration exerted. "It said there was no easy way to move Assad," one analyst says. "Not that it couldn't be done, but that it couldn't be done on the cheap." It said that the Gemayel government was in serious military trouble, that the LAF would probably never be able to perform any duty more ambitious than maintaining some internal security, and that even that would be beyond it if soldiers were asked to fight against their own religious communities. 'They stressed again the incredible complexity of the Lebanese polity and the unreliability of Gemayel and his minority Maronites. "It may not have been timely," says a CIA analyst. "But I defy anyone to say it wasn't relevant."
Casey was reportedly more than a little taken aback, and the SNIE's editing process was grueling and contentious. "There was lots of' blood on the floor," a CIA analyst recalled. It was a struggle waged almost entirely within the CIA. INR and DIA were aware of the dispute, but they took it a good deal less to heart, mostly because they'd had more of a chance, over the last year, to make their feelings known inside their respective departments. The DCI challenged each key finding in the estimate, analysts think he did so partly because he supported the administration's policy and had to be convinced that his people were right, and partly because he knew the effect the document would have outside the Agency.
"Casey and the other senior management probably anticipated a real backlash from Shultz, so the SNIE got a good scrubbing from them," a CIA analyst says. "They insisted on there being evidence behind the assertions. The feeling was, "General so-and-so doesn't say this about the LAF, they say they're going alone; quite well, and now you say they can't do anything. This is going to carry a lot of freight, and I better be sure we can stand behind it."
Some think that the DCI was impressed by what he heard from his people. "I think, although I obviously don't know, that Casey was converted during this process," a CIA analyst says. "He asked hard questions, but he let the estimate go out." Others think that he simply couldn't force his people to back down. "I think the SNIE finally got through the process because the NIO and the NIO system said, "This is what the analysts really think. If you don't want to know about it, don't ask the question," says another. Some think both factors were at work. In any event, Casey finally put his name on the SNIE. CIA analysts were exultant.
Their feat was a near nonevent in policymaking circles. It was read, in some circles at least, as was evidenced by a degree of backlash from the Joint Chiefs of Staff, who didn't care for its characterization of the LAF, and from the Office of Policy Planning at the State Department, which didn't care for any of it. It had no discernible impact. The administration was occupied more than full time with the worsening situation on the ground in Lebanon and with its own cutthroat bureaucratic struggles (Secretaries Shultz and Weinberger were by then locked in almost literal combat over whether to escalate or withdraw).
Many key policymakers don't remember the SNIE at all, and, many who do, remember it with the greatest frustration. "It was not their job to say Assad couldn't be swayed," says an NSC official. "It was their job to analyze what might sway him, that he has vulnerabilities and here's what they are, and what it will cost: you'll have to kill x number of his soldiers or hurt his economy by this much, or undermine Alawi [the minority Syrian religious group Assad belonged to] control of Syria and change the regime. Was it because they didn't agree with the policy that they could identify no pressure points the US could exert against Syria and Lebanon, or was it because there were none? Well, to my dying day I'll believe it was because the people there who wrote the SNIE didn't believe in putting US pressure on Syria and wanted the US to put pressure on Israel instead. But I never had time to do a postmortem on a SNIE; many of them were useless because they were overtaken by events by the time that you got them."
On 23 October 1983, terrorists, believed to be allied with Iran, blew up the Marine barracks at Beirut International Airport with a truck bomb and killed 241 Marines. Throughout November and December the US directed naval gunfire and air raids against Druze, Shiite, and Syrian positions in the Shouf, losing two planes and one pilot to Syrian antiaircraft fire. The administration remained firm in Gemayel's support, but Congressional pressure for withdrawal mounted. In late January and early February 1984, the LAF was torn apart by internal religious tension, and the White House determined to get out of Beirut. The last Marines were evacuated on 26 February. Within days, Amin Gemayel had gone to Damascus to seek rapprochement with Assad.