Body of Secrets: Anatomy of the Ultra-Secret National Security Agency (55 page)

But DF
missions were a double-edged sword, as Specialist Davis had discovered. Since
the range of the PRD-1 was only about five miles—on a very good day—the Sigint
soldiers had to be almost in the enemy's camp to locate them. "They were
practically in our lap most of the time," said Parks. "Once, we DF'd
a transmission that was coming from a grass hut not three hundred yards from
me—easy rifle shot if I could have caught him coming out of the hut."

For Parks,
the constant tension took its toll. "It was a rough way to live and work,
and it took a lot out of men even as fit and young as we were," he recalled.
"I'm not talking about the mission—I'm talking about being in that
environment and doing everything it took to try and stay alive. I myself ended
up in the hospital suffering from sheer exhaustion about three-quarters the way
through my one-year tour. Truth is, I awoke
in
the 'hospital' after
passing out cold one fine day. The 'hospital' was actually more like a ward on
the upper floor of a barracks a block from the 856th. They needed to keep an
eye on their own, you know—can't have me giving away any secrets in my
delirium."

By January
1968 NSA had placed Vietnam under a massive electronic microscope. Sigint
specialists even scanned every North Vietnamese newspaper for pictures of
communications equipment. Hardly a signal could escape capture by one of the
agency's antennas, whether in a mud-covered jeep slogging through the Mekong
Delta or in the belly of a Blackbird flying sixteen miles over Hanoi at three
times the speed of sound. Yet the signals were useless without adequate
analysis, and analysis was useless if military commanders ignored it.

A few
years earlier the Joint Chiefs of Staff had calmly approved committing acts of
terrorism against Americans in order to trick them into supporting a war they
wanted against Cuba. Now that they finally had a war, the senior military
leadership once again resorted to deceit— this time to keep that war going.
Somehow they had to convince the public that they were winning when they were
really losing.

"If
SD and SSD [both were Vietcong Self Defense forces—militia] are included in the
overall enemy strength, the figure will total 420,000 to 431,000," General
Creighton Abrams, the deputy U.S. commander in Vietnam, secretly cabled the
chairman of the JCS in August 1967. "This is in sharp contrast to the
current overall strength figure of about 299,000 given to the press here. . . .
We have been projecting an image of success over the recent months. . . . Now,
when we release the figure of 420,000—431,000, the newsmen will . . . [draw] an
erroneous and gloomy conclusion as to the meaning of the increase. ... In our
view the strength figures for the SD and SSD should be omitted entirely from
the enemy strength figures in the forthcoming NIE [CIA National Intelligence
Estimate]."

As
intercept operators trolled for enemy communications, the results flowed back
to NSA, where analysts deciphered, translated, and traffic-analyzed the massive
amounts of data. Reports then went to the CIA and other consumers, including
General Westmoreland's headquarters, the Military Assistance Command, Vietnam
(MACV). Westmoreland's staff included NSA's Sigint reports in the command's
highly classified publications, including the Weekly Intelligence Estimate
Updates and the Daily Intelligence Summaries, both read by Westmoreland. Nevertheless,
MACV refused to include any NSA data in its order-of-battle summaries, claiming
that the information was too highly classified.

There may
have been another reason. NSA's Sigint was making it increasingly clear that
enemy strength was far greater than the military commanders in Vietnam and the
Pentagon were letting on, either publicly or in secret. CIA Director Richard
Helms saw the difference between the estimates and told his top Vietnam
adviser, George Carver, that "the Vietnam numbers game" would be
played "with ever increasing heat and political overtones" during the
year. To help resolve the problem, he asked analysts from the CIA, NSA, and the
Defense Intelligence Agency to travel to Saigon and meet with General
Westmoreland's staff to resolve the differences in numbers.

The
meeting took place in Saigon in September at the U.S. embassy. Over a
conference table strewn with intercepts and secret reports, the Washington
analysts attempted to make their case, but it was useless. Rather than rely on
NSA's Sigint for enemy strength figures, the military instead relied on
questionable prisoner interrogations. "MACV used mainly Confidential-level
documents and prisoner interrogation reports," said a recent CIA study,
"and, in contrast with CIA's practice, did not generally use data derived
from intercepted enemy radio signals, or Sigint."

George
Carver, the lead CIA analyst at the meeting, expressed his anger in an
"eyes-only" cable to Helms, characterizing the mission as
"frustratingly unproductive since MACV stonewalling, obviously under
orders." Despite the evidence, he said, Westmoreland's officers refused to
accept any estimates of enemy forces larger than 298,000, and "the
inescapable conclusion" must be drawn that Westmoreland "has given
instructions tantamount to direct order that VC strength total will not exceed
300,000 ceiling." He added that he was planning to see Westmoreland the
next day and would "endeavor to loosen this strait-jacket. Unless I can,
we are wasting our time."

In the
end, the military refused to budge. Westmoreland's top military intelligence
officer, Major General Phillip Davidson, told Carver to buzz off. "I was
frequently and sometimes tendentiously interrupted by Davidson," Carver
cabled Helms, "[who] angrily accused me of impugning his integrity,"
and who stated that the figures MACV had tabled were its "final offer, not
subject to discussion. We should take or leave it." Eventually, caving in
to the pressure, Carver and the CIA took it, greatly angering many of the other
analysts.

In
November 1967, NSA began reporting that two North Vietnamese Army divisions and
three regiments were heading toward South Vietnam. Follow-up reports continued
over the next several months until the units arrived in South Vietnam, or in
staging areas in the DMZ and Laos, in late 1967 and early 1968.

Other
reports began coming in January 1968 that a major attack was in the works.
William E. Rowe, with the ASA's 856th Radio Research Detachment near Saigon,
picked up intelligence that two Viet-cong regiments were planning to overrun
the U.S. compound at Long Binh, Bien Hoa Air Base, and several other locations
around the Saigon area. In addition to passing the information to NSA, the
Sigint detachment "also told MACV headquarters personnel about reports of
the planned attack on the Bien Hoa Air Base and several sites in Saigon such as
the MACV headquarters building, the U.S. Embassy, the relay station, the radio
station and the Phu Tho racetrack," said Rowe. "MACV headquarters
personnel sloughed off the information. They ignored intelligence reports
indicating the Vietcong were assembling in tunnels, caves, and foxholes."

On January
17, NSA issued the first in a series of intelligence bulletins reviewing recent
Sigint from Vietnam. It was likely, said the report, that NVA units were
preparing to attack cities in Kontum, Pleiku, and Darlac provinces. Other
attacks were being planned against the coastal provinces of Quang Nam, Quang
Tin, Quang Ngai, and Bin Dihn. Still other intercepts indicated that Hue would
be attacked. NSA reported that Sigint had also picked up indications of
increased enemy presence near Saigon.

Despite
all these reports, the mood within Westmoreland's headquarters was upbeat, like
the bridge on the
Titanic.
Although he was being warned that there were
icebergs ahead, Westmoreland knew his massive ship was unsinkable. According to
a recent CIA analysis, "A 'we are winning' consensus pretty much permeated
the Saigon-Washington command circuit; intelligence reports and analyses that
deviated from it tended to be discounted."

Off the
coast of North Korea, the USS
Pueblo
was attacked on January 23,
suddenly turning attention from the growing threat of a North Vietnamese
invasion to the possibility of North Korean invasion. Many in the Johnson
administration saw a connection. "It would seem to us that there is a
relationship," said Westmoreland. Johnson and McNamara agreed.
Nevertheless, there has never been any indication that the two events were in
any way linked.

Incredibly,
despite the fact that NSA's Sigint warnings on Vietnam were becoming more and
more alarming, the USS
Oxford,
NSA's premiere spy ship, was given
permission to leave its station. On January 23, as North Korea captured the
Pueblo
and North Vietnam was on the verge of a major offensive, the
Oxford
sailed
to Bangkok for a week of R&R. It was an enormous gaffe.

The
following day, NSA reaffirmed an earlier report that attacks against cities
were imminent in northern and central South Vietnam. On January 25, NSA issued
another alert, "Coordinated Vietnamese Communist Offensive
Evidenced." The Sigint report gave clear evidence that a major attack was
about to take place, citing an "almost unprecedented volume of urgent
messages . . . passing among major [enemy] commands." The analysis went on
to predict imminent coordinated attacks throughout all of South Vietnam, especially
in the northern half of the country. Tet, the Vietnamese lunar New Year, was
only five days off.

Richard
McCarthy also noticed unusual activity in the days before Tet. He was on a
direction-finding patrol near the Cambodian border in his small RU-6A Beaver.
Nearby was a large rubber plantation, Loc Ninh. "Evening missions were
usually very quiet," he said. "The Americans were all lagging
[sic]
into their night defensive positions, and the VC were preparing for their
night activities. This night was no exception. There was a large component of
the 1st Infantry Division lagging in on the golf course at Loc Ninh, and I
could see the smoke from white phosphorus, as smaller units around the area
were setting in their final protective fires.

"Suddenly
I started picking up a familiar sound. I quickly identified the target as the
reconnaissance element of the VC division that controlled the area. This was
very unusual, because this guy usually didn't come on the scene until the last
phases of planning an attack! When we finished the fix, we
knew that we
had something big. The target was located 300 yards outside of the American
perimeter at Loc Ninh. We tried to contact the [ASA unit] at Loc Ninh, but they
had shut down for the night. I elected to return to base and report that fix,
instead of flying the full four hours that we were scheduled to fly."
McCarthy later learned that his alert had thwarted one of the rehearsal attacks
for the coming offensive.

On January
30, Westmoreland finally saw the iceberg dead ahead. He had just been handed
several warnings, based on Sigint, from the commander of the U.S. forces in the
region around Saigon. The commander, Major General Frederick C. Weyand, had
become convinced, by intercepts, traffic analysis, and DF indications he had
just received, that a major offensive was about to take place. Westmoreland
immediately canceled a previous Tet cease-fire he had issued and ordered that
"effective immediately all forces will resume intensive operations, and troops
will be placed on maximum alert." "These precautionary moves,"
said a recent CIA analysis, "doubtless saved Saigon and the U.S. presence
there from disaster."

That night
Dave Parks noticed something very unusual. "At twelve midnight, the enemy
went on total radio silence," he said. "It was just as if someone had
switched off a light—'Nil More Heard' on any frequency. Now, that spooked the
hell out of me. I had never experienced anything like it. Military units go on
radio silence for only one reason: they're up to something. In this case they
were on the move to their assigned targets." One of his colleagues,
serving a second tour in Vietnam, told Parks, "If anything is going to
happen it will happen at three A.M.— we may as well go and get some sleep."
"He was dead on," said Parks, "we got the hell rocketed out of
us at precisely three A.M. . . . What we didn't expect was the scale and
intensity of the attacks."

About the
same hour, the 856th Radio Research Detachment at Long Binh, which weeks
earlier had attempted to warn Westmoreland of the coming attack, came under
bombardment. "They had been hiding in tunnels and foxholes in the area for
about two weeks, awaiting orders from Hanoi," said William E. Rowe.
"For the next two and a half hours the Vietcong initiated probing attacks
against our bunker line and other positions along our perimeter. . . . Most of
my buddies were in the operations building setting satchel charges and
incendiary grenades to all the filing cabinets, equipment (radios and
receivers), maps and reports—everything that should not fall into the hands of
the enemy."

It was a
ferocious attack. "Each time they attacked," said Rowe, "some
would get hung up in the wire. Each time they attacked, we went crazy, yelling
expletives as we went out to meet them, firing and firing each time they
approached. A mound of enemy dead was forming in front of the concertina, body
upon body. The frontal attacks lasted for another two hours. After each
advance, we would pace up and down the bunker line, nervously anticipating the
next attack. After each attack, the mound of enemy dead got bigger and
bigger."