The Invisible Cryptologists: African-Americans, WWII to 1956

The Invisible Cryptologists: African-Americans, WWII to 1956 Jeannette Williams with Yolande Dickerson, researcher (2001). Series V, Volume 5 from the United States Cryptologic History publications.

Jeannette Williams was an intelligence analyst during the 60’s and 70’s, reporting on missile and space activities. Later, she was promoted to the Agency’s senior operational officer, and as an assistant inspector general. She retired in 1998 after 35 years of service, and has since worked as a principal research analyst with Logicon DPC Technologies, under contract to NSA. She’s a graduate of the Federal Executive Institute and the National Senior Cryptologic Course.

Yolande Dickerson began working at NSA in 1967 in computer operations, Quote: “After holding numerous analytic and administrative positions, Ms. Dickerson worked in the National Cryptologic School as an education and training officer and adjunct faculty member.” She then worked in the Center for Cryptologic History as a historian before retiring in 1998. She has since worked under contract as a senior systems analyst in the NSA Technology Support Division, and senior research analyst in the CCH.

The U.S. Army, and the government itself, started out with a really poor track record for race relations. While there had been a small, segregated black American group in the Army’s Signals Intelligence Service during WW II, their existence had been almost entirely forgotten soon after. In 1996, the Center for Cryptologic History received a book of photographs of civilian employees, and only two of the photos including pictures of a Black American – one of the man receiving an award, and the other of the same man with his family. Coupled with a few vague stories of the segregated unit from interviews with older members of the NSA, it became clear that the CCH had missed a significant part of the Agency’s past, which Director David A. Hatch wanted corrected. Jeannette Williams was assigned to the writing tasks, and Yolande assisted as “appointment setter, record keeper, transcriber, and researcher.”

A fair amount of this book is going to make certain whites very “uncomfortable,” which may be a good incentive for reading it. It’s short, at only 52 pages, but does include text from interviews with the few people still alive from that SIS unit in the 1940’s.

The history starts somewhere around 1944. Colonel W. Preston Corderman was chief of the Signal Security Agency, and Earle F. Cooke was chief of the cryptanalysis effort (B Branch). According one story, Eleanor Roosevelt had put pressure on the Signal Corps to make sure that 12% to 15% of their personnel “had to be black and gainfully employed” (in Cooke’s words). An editorial note states that there’s no evidence of Eleanor being involved, but that there must have been pressure from “someone in a high place.”

Cooke only knew one Black at the time, a messenger working at Arlington Hall named William (Bill) Coffee. Bill had been born in Abingdon, Virginia, in 1917, and studied English at Knoxville College, Tennessee. “During the closing years of the Depression, from 1937 to 1940, he was enrolled in the Civilian Conservation Corps.” Coffee’s job title officially changed to cryptographic clerk in June, 1944, and the org chart listed him as Assistant Civilian In Charge of B-3-b, with nineteen subordinate civilians.”

There’s no indication that Bill received any formal cryptographic training, and at the beginning he worked alone in B-2a-8, the Commercial Unit in the Code Recovery Section of the Cryptanalytic Branch. This group’s function was to take intercepted traffic in foreign languages, using commercial code books (used for transmissions between companies) to determine the flow of goods and people that could conceivably be used for analyzing troop build-ups or movements. Bill started in Jan. 1944, and he was joined in Feb. 1944 by Annie Briggs, who had worked with him in the messenger unit. In one of the many re-orgs, this one in November 1944, “the unit became part of the General Cryptanalytic Branch, which was headed by Lieutenant Colonel Frank B. Rowlett, one of the four original cryptanalysts hired by William Friedman. Its designator became B-3-b. The organizational chart reflects Lieutenant Benson K. Buffham as the chief and William D. Coffee, assistant civilian in charge.”

B-3-b then “exploited nongovernmental commercial code messages originating from Australia, Great Britain, Germany, Sweden, Spain, Portugal, Bulgaria, Turkey, Afghanistan, Russia, China, Indochina, Thailand, Japan, Egypt, South Africa, Ecuador, Uruguay, Paraguay, Brazil, Peru, and Argentina.” While there are no specific personnel numbers at this point, there were enough people in the group to split tasks up into three categories. Annie Briggs led the biggest group – Production – which identified codes, decoded messages and provided clerical support. A woman named Ethel Just headed “a small group of translators in the Language unit,” and Herman Phynes ran B-3-b technical, which solved encipherments. Quote: “Twenty-eight years into the future, Herman Phynes would be a GG-16, army flag officer equivalent, and NSA’s first African-American office chief in the Operations Directorate.” Herman graduated with a B.A. degree (1941) from Howard University, but his major is not specified.

One other Black important to the group was Bernie Pryor, a messenger for the SIS, who was reassigned to the Personnel Branch as a clerk in April 1944. He became the human resources unit for everyone in Coffee’s group, providing services such as “orientation briefings, information on housing and recreational facilities, and counseling on work, family, and personal financial matters.”

By April, 1945, Coffee was directing 30 people across six sections, plus a secretary. Quote: “Most were engaged in the major activities of code identification and decoding; researching and analyzing unknown codes; and translating.”

According to Benson K. Buffham, the B-3-b administrative chief, “They [B-3-b analysts] were responsible for detecting anything that would be transpiring which wasn’t routine trade. Of course there was a great deal of traffic, because we were monitoring all the international communications, particularly from Tokyo and Berlin – all the enemy traffic. But, it all had to be gone through, because you had to be sure that we weren’t missing something that would be a violation of the international embargoes. Although item for item, it wasn’t as important as diplomatic traffic, they performed an invaluable service by going through all that material and making sure there wasn’t anything in it that would have been useful for us in the wartime effort.”

Coffee was transferred in Feb. 1946 to the Intercept Control Branch. There, he supervised a new typing unit formed to assist the automatic Morse transcription section at Vint Hill Farms, Warrenton, Virginia. Herman Phynes replaced him as assistant O.I.C. (officer in charge). On April 3, 1946, General W. Preston Corderman, chief, Army Security Agency, presented William Coffee with the prestigious Commendation for Meritorious Civilian Service. And in Feb. 1947, Phynes was promoted to Officer in Charge of the Commercial Code unit, reporting directly to Frank Rowlett.

After WW II, the U.S.’s security focus turned to the Soviet Union. While the booklet doesn’t specifically say what happened to the Commercial Code unit, it was most likely disbanded. However, there was increased hiring from the end of the 40’s to the mid-50’s in the GG-2/GG-3 level, with most Black Americans working in two sections of the Operations organization – machine processing (equipment operations and keypunch), and Russian plaintext processing.

Initially though, the head of SIS, William Friedman, was credited with introducing IBM equipment to the SIS in 1935. At first, they used a key punch to record plain text and code groups in order to randomize them in the creation of code books. Soon after, the machines were put to work in searches, frequency counts, and statistical computations for breaking enemy codes. The first machine section chief of staff was a former tabulating equipment operator, Ulrich Kropfl, starting in Dec. 1939.

It’s not clear when the first Blacks started working in the machine section. Alton (“Tony”) B. Dunkinson was already there when a guy named David Shepard (who was interviewed later) started in 1944. Alton had been a signal man for the New York City subway system, and at SIS worked as a technician helping in the development of specialized hardware to be connected to the IBM equipment. He left the SIS in the early 50’s to become an engineer at a systems development company formed by Shepard. More Blacks began entering the machine section (then a part of the Army Security Agency (ASA)) in 1947, including Geneva Arthur, who worked in key punch. By the mid-fifties, the nonsupervisory positions in both the tabulating equipment and keypunch units were filled by civilian Black-Americans, while most of the supervisors were White.

With the focus on Russian traffic, the ASA developed a group for machine processing translated decrypts (the Russian plaintext traffic processing unit). Fifteen Blacks transferred from the Census Bureau in 1947 to ASA, including Dave Bryant. Then in August 1948, Russian transmission of messages goes to landlines only, and the intercept of encrypted radio traffic drops to zero. From this point, the only available messages were Russian plaintext, which could still tell the U.S. government “about Soviet intentions to make war, [and] Soviet capabilities to make war.” By 1950, the ASA had become the Armed Forces Security Agency (AFSA), and by July they were receiving over one million messages a month. The Russian Plaintext Traffic Processing unit (AFSA-213, also known as “the snakepit,” “the plantation,” and “the black hole of Calcutta,”) now had 170 people, and was projected to need another 350 by April 1952. One of the other men working in AFSA-213 from 1951-55 was William Jones. From a later Oral History interview:

“After I was hired and cleared, I was marched down to the first wing, first floor of A Building and escorted into what was a huge wing, and in there was nothing but black people, except there was an Air Force major in charge of that operation. When you walked in the door, there were long tables, at least twenty feet long, perhaps three and a half/four feet wide, lined up against the wall on both sides of the wing with an aisle down the middle. On each of those tables there must have been eight, possibly ten machines that were like typewriters, except they were printers. They had a little device on the side of them that would read a tape; this was five-level punch paper tape. You put the tape in the reader, start the machine, and the machine would type what was on the tape in hard copy. We only had to type a heading on the page which consisted of basic data like the TO and FROM, the intercept station designator, and the date and time of intercept.

“Well, it didn’t take long to pick that operation up. I knew that somewhere people were selecting messages, and I began to wonder how did they pick some messages and throw some away. So as a result, I enrolled in a Russian course at the Department of Agriculture. Soon, I was off the machines and pulling tapes, based on keywords. Once we pulled the tapes, we bundled them in categories and put numbers on them that designated subject areas. But it was a little more complicated than that. Many times the print on the tape was not clear, so we had to read the punched holes. I think what we did was critical because we threw away what we thought wasn’t any good. If there was anything good in there, it was lost – it went in the burn bag. We were called ‘scanners’.

“We heard nothing about career choices or moving to any other place. I went to the Department of Agriculture to take Russian because I didn’t know they would teach it here. There were people in that place [AFSA-213] who had degrees, had teaching experience, and a bunch of them had advanced degrees. It was kind of revealing, I think, to ultimately find out that most of the black people who came to the Agency, no matter the kind of experience they had, wound up there.”

The founder of AFSA-213, Jack Gurin, was a Russian-American. Seeing the mind-numbing work the department was doing, Gurin decided to try to make some improvements as early as 1948. Quote:

“There was this outfit that took these paper tapes where you couldn’t read them unless you could read these little holes. You put them into this machine. You make sure it’s all lined up properly, and you press the ‘on’ button. It starts typing and sheets of printed paper would come out. The equipment was called ‘Cxco’. I don’t remember what that stands for. These people, as I remembered, were all college graduates; all black and all college graduates. Their job was to sit there and watch the machine and make sure it didn’t jam. If it jammed, you stopped the machine and pulled out the keys or fixed the paper. Then you started it again and waited for the next jam. That was their job.

“I looked at them and said this is ridiculous. They were college graduates. They all have some kind of brains. Since the traffic consisted of official messages, they always had the address. I took all the people in this outfit and taught them the elements of Russian, at least the alphabet. I gave them a sheet that said, this [unit] is number 27, this [unit] is number 29, and so forth. So what they did, as the stuff was coming off, they looked at it and looked at the address and put down the correct number. It wasn’t terribly challenging, but it was a lot better than what they were doing before.”

From here, the author backtracks a little and looks at several other groups that also hired Blacks starting from about 1948, including MPRO (machine processing, which consisted of keypunch and tabulating machine operators), and R&D. AFSA’s first Black engineer was Carroll Robinson, hired around 1948, who was tasked with working on the team building ABNER-1, the Agency’s first in-house digital computer. Two other technicians, graduates of the Hilltop Radio-Electronics Institute, were Mitchell Brown and Charles Matthews.

In an interview with Mitchell Brown, 24 June 1999, he said, “[After the war] I had to find a job. They didn’t give electrical type work to blacks here, so I went on unemployment. As a veteran, I got 52-20. You got $20 for 52 months or $52 a month for 20 months, something like that. I would go to the VA every month to sign for my check, and eventually I met a guy there, Mr. Hollywood, and told him that I was looking for work and my classification was electrician. He said that he doubted that I could find any work as an electrician, but there was a guy who repaired radios who might accept an apprentice. This was a black engineer, Mr. Gresham, who worked at the Bureau of Standards on very classified stuff. He was very well qualified and he took me on. When I went back to VA for my 52-20 check and told Mr. Hollywood that I had a job as an apprentice, he said that if this man would open a school, he could send him all the personnel that he needs. This was the founding of Hilltop Radio, and I was the first student.”

In 1952, the National Security Agency (NSA) gets signed into existence and takes the place of the AFSA.

One more pioneer in the AFSA was Raymond Weir, Jr., a former D.C. schoolteacher and Army soldier serving under Captain Fred Hazard. Hazard entered the Agency in the Security division at about the time they were implementing polygraph (lie detector) screening during the job interview process. Hazard brought Raymond in, in 1951, to conduct the screenings of Black applicants. Eventually, his tasks widened, and he finally retired as the chief of the Investigations Division (M54, NSA). Additionally, he became the first African-American president of the American Polygraph Association, and “testified before the United States Senate Ethics Committee in its 1979 financial misconduct investigation of Senator Herman Talmadge (D-Ga).”

For additional success stories:
James Bostic left the Census Bureau in 1952 and “joined NSA as a tabulating equipment operator. A gifted, largely self-taught programmer and systems analyst, he became known as the “the Optimizer” during a career that paralleled the agency’s advancements in computer technology. An early tour in ABNER 1 operations was followed by assignment as a programmer on ABNER 2 and as a software designer for a mass file storage and retrieval system. Before retiring in 1986, he led the terminal subsystem development team for a UNIX-based system.

And, Vera Shoffner Russell, an African-American mathematician, reported to the Agency in 1951 and “was assigned as a programmer on the early computers, ABNER 2, ATLAS 1, and ATLAS 2.”

The booklet concludes with a recap of the working conditions during WW II and on up to 1956, promising that things get better after this, as described in other Agency histories.