by Tom Glenn
Signals intelligence foretold the battle of Dak To.
Was the Tet Offensive Really a Surprise?
VIETNAM '67 NOV. 3, 2017
Paratroopers of the 173rd Airborne Brigade filing past bodies of fellow soldiers killed in the Battle of Dak To. Credit Al Chang/Associated Press
I learned the hard way during the Vietnam War that when intelligence is ignored, people get killed. I spent the better part of the war with the National Security Agency, often undercover. Time after time, I and my colleagues felt like Cassandra, the mythical Trojan princess blessed with foresight but doomed not to be believed. One example was the Battle of Dak To.
By 1967, much of the fighting in South Vietnam was concentrated in the highlands, the mountainous region along the Laos-Cambodia border encompassing Kontum, Pleiku and Darlac Provinces. American military forces were drawn to the region for two reasons. First, it's where the enemy was: The North Vietnamese used the area as a redoubt. The terrain was rough and barren with a sparse population of mostly non-Vietnamese Montagnard tribes, driven there centuries before by the Vietnamese who seized the lowlands for themselves. Second, it was the site of a critical section of the covert infiltration network used by the North Vietnamese to funnel thousands of troops into South Vietnam, known as the Ho Chi Minh Trail by Americans.
That summer and fall I was in the highlands too, working undercover alongside American combat troops. I was a specialist in Vietnamese Communist communications, and I spoke Vietnamese, Chinese and French, the three languages of Vietnam. Unlike many other N.S.A. signals-intelligence specialists, I was willing to go into combat with the American units I was attached to.
I was assigned to coach and assist the Army signals-intelligence team supporting the 4th Infantry Division and the 173rd Airborne Brigade, which was headquartered in the Pleiku area. Our intercept site was at a place called Engineer Hill, high enough so that we could detect enemy transmissions from far away; we also took in intercepts from other teams deployed with Special Forces units operating deep in the highlands.
Throughout September and October 1967, my military counterparts and I kept close track of the communications of the B3 Front, the senior North Vietnamese headquarters for the highlands; its subordinate 1st Division; and two independent regiments, the 24th and the 33rd. Over those months we followed the B3 Front as it deployed a forward command post that established communications with Hanoi -- always a sign of impending combat. Soon it was exchanging a large volume of messages with North Vietnam, mostly sent at night when Vietnamese Communist transmitters normally shut down. The command post and the 24th Regiment moved quickly to Kontum Province. The 33rd Regiment, two provinces to the south in Darlac, initiated combat communications. A new unit, not yet identified, showed up in Pleiku Province, close to our location.
In other words, it looked like the enemy was preparing a highlands-wide offensive.
One moonless night in late October, we located another new North Vietnamese unit, about 20 kilometers from where we were sitting. As if to underline the enemy's proximity, while we were reporting the new unit's appearance, we suddenly came under attack from mortars. The only casualty, it turned out, was a portable outhouse. Still, we were spooked.
Before the end of October, the North Vietnamese 1st Division and its three subordinate regiments had all moved to the Dak To area of Kontum Province -- a region of steep hills and deep jungle valleys. There was a Special Forces base at Dak To, a tempting target, but it was clear to us that it wasn't the only objective. Low-level reconnaissance communications appeared, a sure sign that combat was imminent. Then the division headquarters dispatched a forward command post that took control of the regiments. They were ready.
We informed the American 4th Infantry Division and the 173rd Brigade that an attack on the Dak To area would very likely begin between Oct. 30 and Nov. 4, dates we learned from the reconnaissance unit's messages. But we also warned that units throughout the highlands were preparing for combat. This, we said, was going to be big.
At this point we hit an unexpected obstacle: credibility. Several of us went to brief the commander of the 4th Infantry Division, Maj. Gen. William Peers. I warned him that a multi-division North Vietnamese attack on Dak To would take place at any moment, coincident with attacks throughout the highlands. He shook his head and pointed to our camp on Engineer Hill. "So I'm supposed to believe that some kind of magic allows a bunch of shaky girbs" -- acronym for "G.I. rat bastards" -- "distinguished more for their spit than their polish and abetted by a civilian, to use a tangle of antennas and funny talk to divine the combat plans of the enemy?" He waved us away. The briefing was over.
But we were right. On Nov. 1, a bomb dropped from a B-57 struck somewhere near Dak To. It hit an enemy supply dump, setting off secondary explosions, proof positive that significant numbers of North Vietnamese troops were out there. General Peers sent a unit from the division's 1st Brigade to investigate and make contact with the Special Forces camp at Dak To. Two days later, one of the brigade's battalions landed by helicopter on Hill 978, near Dak To, expecting to meet little resistance. Instead they found thousands of North Vietnamese soldiers, well entrenched and ready. The battalion was decimated. That same day, another American battalion ran into similar trouble on nearby Hill 882. General Peers and the other generals soon recognized the gravity of what we'd been trying to tell them: that the North Vietnamese had moved into the area in force and were itching to fight.
The weekslong series of engagements that followed, collectively known as the Battle of Dak To, was one of the biggest in the war, and one of the conflict's few pitched battles. The North Vietnamese had established defensive positions on several hills, forcing the American and South Vietnamese forces to fight uphill, culminating in a horrifically bloody engagement at Hill 875, from Nov. 19 to Nov. 23. By the end of Dak To, nine American battalions from the 4th Infantry Division and the 173rd Airborne Brigade -- some 16,000 men -- had been committed. American bombers flew more than 2,000 sorties. The Americans eventually won, but at great cost to both sides: More than 2,100 North Vietnamese were killed, as were 376 Americans and 61 South Vietnamese soldiers.
I left the highlands in December when the offensive was all but over. I moved south to work with another team near Bien Hoa, just north of Saigon. When I got there, I saw all the same sorts of communications indicators we were picking up in the highlands before Dak To. We weren't alone; American signals-intelligence units in the northernmost part of South Vietnam were intercepting the same patterns. We realized that an offensive was going to occur throughout the country starting at the end of January.
The N.S.A. pulled together all the evidence, and once again we presented it to the military leadership. And once again, the generals refused to believe us. At the time, Marines in the north were under siege at Khe Sanh, and the top brass in Saigon was convinced that any other North Vietnamese activity was only a diversion, an effort to pull away American forces from what they and Washington believed was Hanoi's plan to reprise the victory over the French at the siege of Dien Bien Phu, 13 years earlier. Despite mounting evidence to the contrary, they did not prepare for an all-out assault, and at the end of January they were thrown back on their heels by the Tet offensive.
Put differently, it's not quite correct to say that Tet caught the Americans and South Vietnamese by surprise. The intelligence was there, and the recent experience leading up to Dak To should have persuaded General William Westmoreland to take it seriously. Instead, he chose not to believe it.
The problem was bigger than Generals Peers and Westmoreland; eight years later, the same mistake was made with the fall of Saigon. By then I was the N.S.A. station chief in the city. I warned Graham Martin, the American ambassador, about overwhelming evidence showing that Saigon was about to be attacked. He refused to believe me and didn't call for an evacuation of the thousands of American civilians still in the city, along with our South Vietnamese counterparts. When the North Vietnamese attacked a few days later, the city descended into panic. I escaped under fire. My South Vietnamese partners, the men I was working with, weren't so lucky. Some 2,700 of them were killed or captured and sent to "re-education" camps.
Was Cassandra blessed or cursed? Those of us who worked in intelligence during Vietnam know the answer.
Tom Glenn is a former National Security Agency employee and the author of the novel "Last of the Annamese."