By the next day four distinct levels of air strikes had been identified: (1) against known MRBM installations; (2) plus airfields; (3) plus SAM sites and coastal missile sites; (4) plus other significant sites. The most limited air strike conceivable involved a ‘‘50 sortie, 1 swoop air strike limited to the missile complex, followed by open surveillance and announcement that future missile sites would be similarly struck.’’ The stronger action would aim to ‘‘eliminate all Cuban air power or other retaliatory capacity’’ and would involve up to two hundred sorties per day. These options combined in a number of permutations, in part depending on whether they would come with or without warning. A trade began to emerge between military and political moderation. A small strike might be manageable without warning, but warning mandated a big strike. The risk of not attacking the SAM sites was reduced so long as there was no opportunity for them to be readied for action. Kennedy had not attended the meeting, but McCone’s view, after briefing him, was that the president was ‘‘inclined to act promptly if at all, without warning, targeting on MRBM’s and possibly airfields.’’

At the time this was Rusk’s position. His colleagues in the State Department had more qualms about surprise attacks and no diplomatic initiatives. They had heard talk of swift and decisive action limiting political damage before and were unconvinced. It was an ‘‘illusion’’ that would probably lead to full war, judged Charles Bohlen, about to leave for Paris. He wrote in haste to Rusk to urge that diplomatic channels be fully ‘‘tested out’’ prior to military action, if only to establish a case for the action. George Ball was becoming concerned about the irreversibility of any air strike, as well as the large number of Russian casualties that would result from a full strike, and the consequent risk of a violent Soviet reaction. In a note on 17 October 1962 he stressed the importance of a moral line, given the stance adopted by the United States on Pearl Harbor, Hungary, and Suez. ‘‘It is my strongly held view,’’ he now observed, ‘‘that we cannot launch a surprise attack against Cuba without destroying our moral position and alienating our friends and allies.’’

Robert Kennedy, crucially, was now of this opinion. During the previous day’s discussion he had been in a transitional mode. One part of him was still in Mongoose, searching for ways to get at Castro and even at one point musing about the possibility of fabricating an incident to provide a pretext for intervention. His few interventions had a hard-nosed ring to them, though he never obviously argued for one course of action. By the day’s end, sensing his brother’s line, he grasped the set of political concerns that he held on to for the rest of the crisis. By the second day he saw how the European allies, having been ‘‘under the gun’’ for years, would now take the position that this reaction to a few missiles in America’s backyard was ‘‘hysterical.’’ The president was aware of the same risk: The allies thought the United States ‘‘fixated . . . slightly demented on this subject.’’ When Ball spoke of ‘‘carrying the mark of Cain on your brow for the rest of your life,’’ Robert again backed him up. The Americans had resisted a Russian first strike. ‘‘Now, in the interest of time, we do that to a small country. I think it’s a hell of a burden to carry.’’

Even civilian hawks such as Dillon and McCone were clearly uncomfortable with the idea of launching a surprise attack. On 18 October Robert Lovett, another prominent and generally hard-line figure in the foreign policy establishment, told the president that an air strike that resulted in great bloodshed would be deemed excessive when the insertion of missiles into Cuba was not a self-evidently aggressive act, ‘‘a sledgehammer . . . to kill a fly.’’ Also working against an air strike was evidence of the difficulty of keeping it limited. More intelligence came in on the Soviet offensive capability that would have to be destroyed. The discussion had started, on the Tuesday, with three possible missile sites. By Thursday, 18 October, the previous day’s U-2 imagery indicated a possible five sites, two of which might take the longer-range SS-5. Some sixteen of the shorter-range SS-4 were deemed operational and could be launched eighteen hours after a decision was made to do so. The SS-5s, by contrast, would take longer, possibly until December. There was no hard evidence that nuclear warheads had arrived.

The Joint Chiefs continued to oppose restricted strikes. In part this was because they were more politically ambitious. On 17 October they wrote to McNamara still proposing a comprehensive strike combined with a naval blockade, although the actual elimination of the Castro regime was now only an option (requiring an invasion) rather than a prime objective. Having failed to consider possible Soviet reactions the previous day, they now addressed this problem at some length and concluded that they would be manageable.43 Robert records his brother’s skepticism: ‘‘They, no more than we, can let these things go by without doing something. They can’t, after all their statements, permit us to take out their missiles, kill a lot of Russians, and then do nothing. If they don’t take action in Cuba, they certainly will in Berlin.’’

For the U.S. Air Force, dominated by a strategic bombing philosophy, a limited strike appeared as an unnatural act. No merit was seen in doing less than possible, and certainly not less than necessary. If there was a risk of failure, then the response was to add to, not subtract from, the scale and intensity of the strikes. In a transparent reference to Chief of Air Staff Curtis LeMay, Robert Kennedy later recalled ‘‘the many times that I had heard the military take positions which, if wrong, had the advantage that no one would be around at the end to know.’’ LeMay’s own immediate comment on possibly the same exchange was somewhat more coarse. He had just returned from Europe and was infuriated that he had not been consulted about the developing crisis. His conversation with Robert Kennedy was cut short after the latter had been informed that B- 52s could only use nuclear weapons. LeMay, angry at being denied the opportunity to brief the White House, observed, ‘‘What a Dumb Shit!’’

The Joint Chiefs never deviated in their advocacy of a comprehensive military strike. Moreover, their assessments of the number of aircraft required tended to grow, as they not only found new targets but also identified new tasks. Taylor at one point warned his colleagues, ‘‘You are defeating yourselves with your own cleverness, Gentlemen.’’ This was exasperating enough to the doves, anxious about the military’s insensitivity to the dynamics of the crisis. If anything, it was even more so to civilian supporters of an air strike, who found themselves being asked to subscribe to something far more substantial than necessary. This was particularly galling for Dean Acheson, one of the most forceful proponents of a surprise limited air strike aimed at taking out missiles. He observed later that when developing plans the military became so anxious to remove all possibilities of doubt that ‘‘the proposals are apt to be at least as dangerous as the original danger.”

On 18 October Sorensen reported that the tide was running against a limited strike, and the next day that this was ‘‘no longer recommended even by those who first proposed it because of the dangers presented by a surviving and substantial air capability. This build-up should be hit as a whole complex or not at all.’’ Yet Kennedy remained unconvinced. He never accepted the arguments for a comprehensive air strike. When the full NSC met on 20 October he ‘‘directed that air strike plans include only missiles and missile sites.’’

The same process occurred with air strikes, the main military option to be discussed. A variety of versions were discussed, from one euphemistically described as ‘‘surgical,’’ directed at the offending missiles and nothing else, and others with extended targets, from the air defense network to air bases. Kennedy initially inclined toward an air strike, although even then he was struggling to identify the most limited option available.

When the deliberations began just before noon on 16 October 1962, the practicalities of an air strike soon came to the fore. During the day all the major issues relating to an air strike were aired, though the discussion was less than systematic. At issue was not whether an air strike should take place but when, how, and against what. How important was it to strike before the missiles became operational? Could the strike be limited to missile bases, or would it also be necessary to attack the SAM sites, airfields, and possible nuclear storage sites? Would an air strike achieve what was asked of it? Dare the United States risk the loss of surprise by warning that an attack was imminent? Should air strikes be followed by an invasion? What would be the Soviet response?

Of these the question of effectiveness was, surprisingly, one of the least explored. When the president asked about the extent of the ‘‘take-out,’’ Taylor cautioned him that it would ‘‘never be 100 percent . . . we hope to take out a vast majority in the first strike.’’ Taylor anticipated that any discovery of a target would prompt continuous air attack for a number of days. He also warned against targeting the missile bases alone while neglecting the surface-to-air missiles lest these intercept the incoming air craft. This followed a tentative query from Bundy as to whether the strike could be limited ‘‘in surgical terms to the thing that is the cause of action,’’ suggesting that this would have a substantial political advantage. Rusk and McNamara both accepted that a strike would have to be comprehensive, covering airfields and possible nuclear storage sites. The more targets added, the more substantial the operation and the greater the number of civilian casualties likely.

In his first remarks McNamara insisted that any air strike must be undertaken before the missiles became operational. When Rusk queried this on the grounds that the Soviet Union knew that any nuclear weapons launched against the United States, from whatever source, would mean ‘‘general nuclear war,’’ McNamara cautioned that there could be no confidence in the quality of Soviet control over the warheads. McNamara also raised the question of whether to ‘‘precede the military action with political action,’’ answering in the affirmative and apparently envisaging direct contact with Khrushchev. By contrast, Dillon suggested that the best way to avoid a Russian reaction was a quick strike ‘‘with a statement at the time saying this is all there is to it.’’ Once the issue was in the open, he maintained, Khrushchev would be obliged to threaten retaliation to any U.S. action.

Summing up toward the end of the midday discussions, the president was reconciled to the likelihood of an air strike at least against the missiles.

We’re certainly going to do number one; we’re going to take out these missiles. . . . The questions will be whether, which, what I describe as number two, which would be a general air strike. That we’re not ready to say, but we should be in preparation for it. The third is the general invasion. At least we’re going to do number one.

What is important about this statement is not so much that he was prepared to authorize an air strike but that he was looking for a minimalist version. There had been no clear recommendation to limit the attack in this way; if anything, the weight of argument had been against this, especially from the military.

Late that afternoon the Joint Chiefs of Staff met to consider a series of questions posed by McNamara. They rejected all limited strikes as being ‘‘unsound,’’ risking a loss of surprise, and inviting retaliatory attacks on the United States and the Guantánamo base. No strike could proceed without taking out the air defenses. They also looked to striking ‘‘tanks and all other significant military targets.’’ But this assumed greater ambition: ‘‘Objectives—eliminate threat to US and liberate Cuba.’’ They were also convinced that the threat from the missiles was such that they should be attacked even after they became operational. To the question posed by McNamara, ‘‘What Soviet reaction may be anticipated?’’ the answer was ‘‘Soviet reaction unknown.’’

Just as the Joint Chiefs’ meeting was ending, ExComm began another session. It took a long time to move beyond the problems that had preoccupied the group earlier. The State Department had done some work on the possibility of direct messages to Castro, with the idea that the Cuban leader might be convinced that he was becoming the victim of Moscow’s policies. The president soon moved the discussion back to the military options and a limited air strike. McNamara warned that even a limited air strike would involve ‘‘several hundred sorties’’ (there was not the intelligence at that time to be more precise). The judgment of the Joint Chiefs, reported Taylor, was that such a strike would be a mistake. They argued for two or three days to get full intelligence and then, if the ‘‘target system . . . really threatens the United States, then take it right out with one hard crack.’’

Kennedy resisted this conclusion, reluctant to ‘‘abandon just knocking out these missile bases.’’ This was a much more ‘‘defensible, explicable, politically or satisfactory-in-every way action than the general strike which takes us . . . into the city of Havana.’’ Once again McGeorge Bundy provided the only support for going solely for the missiles: ‘‘The punishment fits the crime in political terms . . . we are doing only what we warned repeatedly and publicly we would have to do. We are not generalizing the attack.’’ The advantages of a ‘‘surgical’’ strike were that it was in some way proportionate, and that it avoided expanding the crisis into one deciding the whole future of Cuba.

Kennedy also inclined toward a prior public statement. Taylor was worried that any warning would mean that the missiles could be pulled ‘‘in under trees and forest and disappear almost at once.’’ McNamara added that the missiles could be readied ‘‘between the time we, in effect, say we’re going to come in and the time we do come in,’’ and argued against an announcement if a strike was chosen. This was where the diplomats began to feel uneasy. Almost as the meeting ended George Ball brought in the comparison with Pearl Harbor and expressed his fear as to where things were leading: ‘‘You go in there with a surprise attack. You put out all the missiles. This isn’t the end. This is the beginning, I think.’’ Robert Kennedy had already been struck by the same analogy, passing his famous note to Sorensen: ‘‘I know how Tojo felt when he was planning Pearl Harbor.’’

As assessments of the balance of power were being traded at one level and private correspondence at a higher one, a minor crisis erupted in the middle of Berlin. On 19 September General Clay had reappeared in Berlin, now as a special adviser to Kennedy, but clearly expecting to call the shots. He was a tough hard-liner, popular with the Berliners as one committed to their cause. His relations with all the other American generals in Europe with responsibility for Berlin, from Norstad down, were poor, and those with the State Department were no better. Clay did admire Acheson, and Kennedy may have wanted to use him in the same way, to demonstrate the participation of all shades of opinion, but like all his personal appointments that cut across formal bureaucratic and military lines of command, it sowed confusion and complicated policy making.

Clay also had a different attitude to risk taking than Kennedy. He outlined his theory to Rusk:

Harassments not in direct conflict with our basic rights do not worry me although I think they should be checked as they occur to the extent possible without show of force. To show force when we do not intend to use it is one thing we must avoid. However, I cannot accept the escalation theory that any reaction on our part leads to further actions and possible war. The complete acceptance of this policy would lead to the continuing erosion of an already eroded position.

On his return to Berlin Clay decided to prepare for a big confrontation by ordering the U.S. military commandant, General Watson, to get his engineers to build a replica section of the wall in a secluded part of West Berlin in order to experiment on various ways to break it down. When the Commander in Chief (CINC) Europe, General Bruce Clarke, heard of this, he demanded that the experiment be stopped and the replica dismantled, but not before it had been photographed by Soviet intelligence. The news reached Moscow on 20 October. It was in this context that the communists viewed Clay’s moves. Washington was ignorant of this episode.

Clay wanted to probe the limits of communist tolerance, just as Ulbricht wanted to do with the West. Khrushchev was blustering to obscure what was in effect a retreat, but the East German leader was still bent on eroding the Western position in Berlin. This was a gradual campaign, independent of Soviet direction, and the target of its next stage was the rights of allied civilian personnel to move in and out of East Berlin.

On 22 October the State Department’s Allen Lightner was stopped going to the opera in East Berlin with his wife. He refused to show identification. To assert his rights, Lightner returned, this time accompanied by military police, and went into East Berlin. Ulbricht then issued a decree stating that allied civilians would have to identify themselves. Though the British and French had already allowed their officials to show passports to East German guards, the Americans continued as before, with an armed patrol. On 25 October, as the latest American probe was following its customary procedure, Clay asked General Watson to place U.S. tanks on alert near the sector boundary at Checkpoint Charlie. This appeared in Moscow as potential confirmation of Clay’s intentions to move against the wall. Coupled with Gilpatric’s speech, this all began to look sinister in Moscow. Had easing the deadline emboldened Kennedy?

Civilians showing identification was not an issue upon which many others than Clay wanted to make a stand, yet he was allowed to continue. On 27 October American tanks once again drew up to back a patrol, but this time Marshal Ivan Konev, like Clay taken out of retirement to add weight to the Soviet military command, decided to respond by pushing his own tanks up to face the Americans. The East Germans announced that the checkpoint would be closed until the American tanks were removed. Clay responded by moving some tanks directly onto the demarcation line, to be followed by Soviet tanks moving to a hundred yards from the Americans. So on 28 October at Checkpoint Charlie, the new flash point of the cold war, American and Russian tanks pointed their guns at each other.

Kennedy rang Clay to compliment him on his nerve. The general responded that he had not lost his but he was worried about those in Washington. ‘‘I don’t know about those of my associates,’’ replied the president, ‘‘but mine are all right.’’ This was highly unlikely. This was the sort of situation Kennedy dreaded: a contrived incident over a secondary issue that could lead to a tank battle in the middle of Berlin with who knew what consequences to follow. It is likely that Clay was told that an opportunity to withdraw would soon come. The previous day Robert had got in touch with Bolshakov to pass on the message to Khrushchev that American tanks would withdraw if Soviet tanks did so, and that it would be possible for the two leaders to have a productive exchange of opinions.  Soon Russian tanks were pulled back a bit, and American tanks followed suit. The communists did not, however, back down on their demand for identification.

Clay’s view was that the Soviet Union was not going to let the East Germans lead them to war barely fifteen years after the last one, and so he was looking for ways to limit the ability of the East Germans to operate unilaterally. The issue of who was in charge in Berlin was a sensitive one for the Russians, and in this incident they attempted to hide their role by obscuring unit markings on tanks and dressing the crews in overalls rather than uniforms. Reporters still noted that officers directing the tanks were wearing Soviet uniforms. In this way the exercise proved a point and may have sown distrust between the East Germans and the Russians. For their part the Russians may have felt that they had thwarted an American escalation of the crisis. Both leaders had noted the utility of ‘‘back-door diplomacy.’’

The episode was an ‘‘exercise in military theatricality.’’ Ausland reports that American officials in Berlin believed that the conflict would have been contained but notes that ‘‘Kennedy and Khrushchev proceeded thereafter with even more caution regarding Berlin.’’ A lasting legacy of the dramatic images of confrontation left over from this episode was that Checkpoint Charlie became the starting point for war-gamers, novelists, and documentary makers when they were trying to work out how to start a third world war.