War of Numbers (23 page)

Over the next ten days, there was a good deal of back-and-forth over the station cable. Its “last cut” came in, and I reread it for further details. It said that the Winter-Spring campaign’s next phase was planned for “January to March 1968”; that the communists had set for themselves
“the task of occupying and holding some urban centers in South Vietnam, and isolating many others”; and that their object was “to inflict unacceptable military and political losses on the Allies regardless of VC casualties during a U.S. election year, in the hopes that the United States will be forced to yield to resulting domestic and international political pressure and withdraw from South Vietnam.” “In sum,” it concluded “the war is probably nearing a turning point … The outcome of the … campaign will in all likelihood determine the future direction of the war.”
⁴⁴

This was heady stuff to let loose on the White House without comment, so Carver sent it around Langley first to solicit opinions. The first to come back was from the Office of Current Intelligence, the one that produced the Sitrep. The station was blowing smoke, it said; “We question whether communist strategy … is in a state of flux,” and OCI had seen nothing “to suggest that the communists think they can really mount a decisive campaign.”
⁴⁵
Much more likely, the VC would do what they’d been doing all along, which was to wage “protracted warfare.” Protracted warfare doesn’t include all-out offensives.

The next opinion was mine. At this point my main feeling about such an attack—other than it seemed to make sense—was that if one was coming it would be enormous. Therefore, I used the opportunity to take another swipe at the order of battle. I listed the usual omissions plus a couple of new ones.
⁴⁶
In the last few days several documents had turned up that mentioned regular VC infantry battalions listed in the order of battle, but with strengths much higher than the OB acknowledged.

Carver sent the station assessment to the White House on 15 December, but with a cover sheet of his own, written on Office-of-the-Director stationery. Doubtless mindful of OCI’s pooh-poohing, he pointed out that the station cable was “a field study and should not be read as the considered opinion of the agency.”
⁴⁷
The field—whose assessment was based on captured documents—didn’t get all the high-level intelligence available in Washington, he continued, and the VC’s “current activity patterns” were not unlike those which prevailed
in 1965 and 1966. He didn’t mention my comments on the order of battle. When I saw this omission, I felt the CIA was digging itself into a deeper and deeper hole. Only a few days before it had sent another big “Will To Persist” memo to McNamara, using all the numbers from Fourteen Three.
⁴⁸
I’d tried to change them upwards, but had been told that unfortunately the manpower account had gone downstairs to Paul Walsh, the one who had recognized in late October “the apparent obligation for the estimate to be consistent with the figures agreed to as Saigon.” At that time we’d been prisoners of the Saigon agreement; now we were prisoners of Fourteen Three.

Christmas came and went, so did New Year’s, and about the only person I talked business to regularly any more was George Allen. From time to time I brought him “extraordinary documents”—which he’d asked for concerning the offensive, and which still pointed to a big one—and we jawboned about the evils of doctoring enemy strength estimates. On the fourth of January, MACV’s year-end Order of Battle came in. It had actually
declined
, another 9,600 guerrillas having bitten the dust, and with some additional seepage from the regulars.
⁴⁹
I looked at the fine print to discover, for example, that the R20 Local Force Battalion near Danang had dropped 50 men, that the 32nd North Vietnamese Regiment had gone from 1,600 to 1,100 men, and so on.
⁵⁰
The decline was damn peculiar for an enemy that was supposed to be priming for a big attack. On the following morning, Helms’ New Year’s briefing package for Congress arrived on my desk using Fourteen Three’s faked numbers,
⁵¹
and that afternoon, Theresa told me that my transfer had come through to leave the director’s office. She said I was to report to the DDI’s new Vietcong Branch down on the fourth floor at the end of the month.

Meanwhile in Vietnam, VC activity was picking up. The communists were closing in on a U.S. Marine Corps redoubt called Khe Sanh near the DMZ, and there were reports that as many as three North Vietnamese divisions were involved. One was the 304th. “The Fightin’ Three Oh Four,” George Allen had cried a couple of weeks earlier; “It fought at Dien Bien Phu!” This was the first mention of Dien Bien
Phu I’d heard for some time, but it had since become the main topic of conversation. Its surrender in May 1954 had precipitated the French decision to leave Indo-China, and the theory had developed that the communists wanted to pull a repeat performance on us at Khe Sanh. In discussing the 1954 fracas George knew what he was talking about. He had been a Vietnam analyst for the Army back then and could describe the fall of the French base bastion by bastion. To keep abreast of the latest battle, Major Blascik—who had just opened a large situation room at the other end of the hall from Carver—was negotiating with the Cartography Section for a giant chart of the area around Khe Sanh.

On the morning of 12 January, a Friday, I was watching the installation of the map, which was truly gigantic, when Theresa hollered in from the hall that I was wanted on the telephone. I picked it up and the caller said: “That you, Sam? This is Colonel Hawkins.”

“Colonel Hawkins!” I said. “Where are you?”

“Up in Baltimore at Fort Holabird. I run a course in intelligence up here, and this afternoon I’m coming down to visit Langley.”

“Can we talk about the OB?”

“That’s one of the reasons I’m coming,” he said. We made an appointment for 2:00
P.M.

It was the best news I’d had in months. For several weeks now I’d been chipping away at the enemy’s regular army, finding companies here, platoons there, never more than a few hundred men a day. Here was the man who probably knew more about VC regulars than anyone else in the United States government. Furthermore I could ask if he still felt as he had earlier about the rest of the OB, particularly the guerrillas. As for MACV’s chicanery, it was best left alone, at least for the time being. The colonel was taking a big risk as it was, and my object was not to rake muck, but to find enemy soldiers. When he entered the situation room at two, we must have shaken hands for a solid minute, all that time smiling broadly.

“Those were the worst three months of my entire life,” he said, obviously referring to July, August and September.

“They were none too pleasant for me either, Colonel,” I replied. “Now let’s get down to business.” We talked for two hours. He confirmed his earlier opinion that there were as many as 120,000 guerrillas, and identified a number of service units MACV had scaled down. We ended up on regulars. When he’d left Saigon in mid-September, his regular estimate was “conservative,” but not outrageously so. Certain types of formations needed work, such as sappers (I knew about these), city units (one for each of the forty-four province capitals), and oddball formations such as
vung
(area) units. A good place to look for missing battalions was among the divisions, he suggested, of which the communists had nine. As for the recent decline in regular strengths in the OB, he couldn’t explain it, not having seen any documents since his departure from Vietnam. As we were breaking up, he invited me to Holabird to give some lectures on the VC. I said I’d be delighted to help him out, and meant it.

After the weekend, Theresa typed up my notes in a formal memorandum of conversation,
⁵²
which I sent around the building with the notation “This is completely off-the-record, so please be careful in using it. That is, protect Colonel Hawkins.” Among those to get it were George Carver and Paul Walsh, the new manpower chief, but neither of them did anything about it, such as demand a higher strength estimate. Thanks to Helms, we were still tied to Fourteen Three.

Having been pointed in the right direction by Colonel Hawkins, I got to work, starting with the divisions. Now Vietnamese communist divisions are like ours in that they have three infantry regiments, plus a number of support units, such as artillery. So I got out my MACV Order of Battle and looked up the Vietcong Ninth Division, which operated not far from Saigon. The OB listed the three regiments all right (the 271st, the 272nd, and the 273rd), but it had
no support units whatsoever.
⁵³
Christ! I dove into my files and quickly came up with a VC document about the Ninth.
⁵⁴
The document listed
ten
support units, including four companies and six battalions, among them an artillery battalion, a mortar battalion, an anti-aircraft battalion, and a battalion of
scouts, arguably the best troops in the entire division. I looked up the other eight divisions, and similar omissions, normally including all the divisional artillery.

Next I investigated city units. As a rule, city units didn’t operate in the cities themselves, but stayed just beyond the suburbs, waiting for an opening. In urban assaults, like the one forecast by the station cable, such units would be expected to lead the way. I began with Danang. Danang didn’t have
one
city unit, it had several, including the T89 and T87 battalions,
⁵⁵
and a whole mess of commando platoons, none carried in the OB. What surprised me most was the number of VC documents on the T89, maybe twenty. How could the OB Section possibly have missed it? It looked fishy. On the other hand, most things looked fishy to me nowadays, and I telephoned everyone I could think of to tell them about omitted regulars. I might as well have been talking about hot dog buns. By Monday, 22 January, it was clear that the number of unlisted regulars was at least 50,000—the equivalent of seven extra divisions—and conceivably a lot more.

Monday was an active day. Over the weekend, the enemy had loosed a big artillery barrage at Khe Sanh, and the Marines had captured a North Vietnamese lieutenant who claimed that the communists intended to take the place in the not-too-distant future. To prepare for the attack General Westmoreland was plastering the adjacent landscape with B-52 raids, and Major Blascik was marking up his enormous map. Blascik’s map was incredible to behold. He had two helpers now, and with red and blue pins and various kinds of stickers they were plotting not only the last-known locations of all friendly and enemy units, but also the precise whereabouts of every skirmish, no matter how small, and even the strike pattern for each stick of B-52 bombs. Word of the map spread quickly through the building, and CIA people began to arrive in the situation room to take a look. George Allen was in his element. “You should have seen what they did to Gabrielle,” I heard him tell a caller, “Gabrielle” being the name of a French bastion which fell to the Vietminh in the early stages of Dien Bien Phu. As if taking their
cue from George, the communists that afternoon overran a Special Forces camp a few miles west of Khe Sanh.

By Thursday, the map had become a lodestone for official Washington. Dignitaries came from the Pentagon, the State Department, and even the White House to see it.
^*
A Blascik helper had penciled in the trenches, which the visitors eagerly traced with their fingers. Communist divisions in the general vicinity now included the 304th, the 325C, the 324B, and the 320th. As he was leaving the situation room, I heard a dignitary tell George Allen: “My God they’ve got a lot of troops up there.”

When the visitor had gone out the door, I took Allen aside. “George, you know as well as I do, it’s not just at Khe Sanh. The VC have troops all over the place. They could launch an attack everywhere in the country and still have plenty of men to spare.”

George said: “I know, I know. I tell them that, but they won’t listen. They aren’t interested in the OB. They want to hear about Khe Sanh.” Later in the day I read a newly arrived study from NSA. It was called “Coordinated Vietnamese Communist Offensive Evidenced in South Vietnam,” and it noted “an almost unprecedented volume of urgent messages … passing among major commands.” Most of the messages concerned the northern half of the country, but there was some “possibly related activity” in “Nam Bo.” Nam Bo was the south, including Saigon. The study said the precise timing of the offensive was unclear, but that several VC messages referred to it as “N-Day.”
⁵⁶

I passed the next day, Friday, quietly at work, spent the weekend at home, and returned to Carver’s office on Monday morning to pack. It was 29 January and my transfer orders to the DDI’s new Vietcong Branch on the fourth floor read “end of the month.” I guessed it would take the entire day to ready my files for the big move, since they now occupied the better part of four stand-up safes. The safes held an almost complete set of MACV’s nine-thousand-odd bulletins, and a variety of
manila folders, including one marked “VC Winter-Spring Campaign.” It was funny how little mention there’d been of the station cable since it had gone to the White House in mid-December. I presumed the NSA study that had predicted a “coordinated offensive” was talking about the same thing.