Intelligence Isn't Always Decisive: Rommel and Fellers - A Reappraisal

Intelligence isn't always decisive. It can be, but there are cases where it is claimed to have done so when it didn't.   A case in point is German general Erwin Rommel and the 1942 phase of the Western Desert Campaign.

Rommel's best source of communications intelligence was the reading of the deciphered reports of an American military attache in Cairo, Major Bonner Fellers.

Major Bonner Fellers served as attache from October 1940 until late June 1942. Diligent in his duties, Fellers passed along all he could learn to his superiors in Washington, D.C.

Fellers transmitted his reports using the State Department's “Black Code” [Military Intelligence Code No. 11.]  Unbeknownst to him, the Italians had stolen the code in 1941 and may or may not have passed it along to the Germans. The Hungarians reportedly gave the Germans a set of substitution table for the code in December 1941. With the table in hand, the Germans were reading the messages starting January 1942.

Some well-placed participants and writers say that that Fellers was responsible for Rommel's victories in 1942. Once he lost Fellers, he started suffering a chain of defeats. Was Fellers really that responsible for either?  Are their descriptions of the intelligence Rommel received from Fellers accurate?  Did Rommel use this intelligence?

The Germans described Fellers as the “Good Source.” Hans~Otto Behrendt, a senior member of Rommel's intelligence staff, described the value and subsequent loss of Fellers' information as: “We now no longer had this incomparable source of authentic and reliable information, which contributed so decisively during the first half of 1942 to our victories in North Africa.”

Herbert Schaedel ran the radio intercept station at Lauf, Germany, at this time. He concluded: “Rommel, each day at lunch, knew exactly where the Allied troops were standing the evening before.”

Wilhelm Flicke, who worked in a variety of positions in German intelligence and penned a history of its role during the Second World War, wrote about Fellers:  “He [Rommel] missed the Fellers telegrams.  Rommel's operations came to a standstill. . . The period of Rommel's great victories was over.”

Some writers have also drawn the same conclusions. Bruce Norman, author of Secret Warfare - The Battle of Codes and Ciphers, characterizes the value of Fellers' information as:  “For Rommel, the leak had provided the clearest picture of enemy intentions ever available to a military commander. . . . the British advance began – perhaps coincidentally, perhaps not – as Fellers went off the air.”

C. J. Jenner in his article and quoting Sir Stewart Mitchell describes Fellers as:  “Given access to everything in the 8th Army and he sent back detailed reports on it all to Washington.”  Seemingly per Mitchell, Jenner adds that Rommel had all of this information within eight hours and concludes:  “Largely on the basis of Gute Quelle's [Fellers] decisive reports, Field Marshal Rommel came close to capturing Cairo and Alexandria, and thereby the Middle East.”

Author Gershom Gorenberg makes the same point:  “Rommel would still have had his source [Fellers] and the battle [El Alamein] could have gone differently.  If that had happened, there is no knowing what course the war would have taken — or what the Middle East would look like today.  Indeed, if that had happened, there is no knowing to which victors history would belong.”

Fellers was of great value to Rommel and the Axis, but the claims above are neither accurate nor explain Rommel's fate. Here are some of the problems:

Attributing an effect to a single cause is very tempting, especially when using qualifiers such as “perhaps” or “could have.”  It can make for a dramatic, catchy title and it offers a neat, tidy story that is easy for readers to understand.  It is also rarely the case, such as here.

In the first half of 1942, Rommel was at times on the defensive. There were times he was on the offensive and advancing.  How can this be?  The obvious answer is that there are many more factors than just Fellers that explain Rommel's outcomes. John Ferris, a historian of the campaign, addresses this problem at length and sums it up well:  “The range of these outcomes shows the need for care in assessing how intelligence affected Rommel’s operations.”

American and British archives document what Fellers sent and it can be very tempting to assume what Fellers sent is exactly what Rommel received and used.  This is not true.  Rommel received a numerical subset that was somewhat distorted in its contents.

Flicke admits that the Germans did not intercept all of Fellers messages in spite of the resources devoted to it.  Some pieces were lost due to poor reception or interference.  In spite of his initial claim above, Flicke elsewhere wrote that they could not necessarily decipher all the traffic.

It is not an uncommon practice for intelligence personnel to mistakenly not pass information that they see as irrelevant.

Intelligence has to be timely to have the most value.  Flicke cannot make up his mind on this matter.  Within the same document he writes that they had no delay in deciphering the messages and that they had a two to three week delay from February until late May 1942.

The problem was that the substitution table was apparently changed at least once at some point.  This led to delays in deciphering Feller's traffic until the new table was figured out.

Numerically, I would estimate the subset at about 75% of Fellers messages (-2% never intercepted, -2% never deciphered, -1% seen as irrelevant, -20% deciphered too late to have real intelligence value, rounding = 75%) One might disagree with my estimate, but the fact remains that it is a subset.  Every message was not intercepted, deciphered, passed along, etc.

The 75% subset was then corrupted in its contents.   According to David Kahn in How I Discovered World War II's Greatest Spy and Other Stories of Intelligence and Code the Germans edited the information before passing it on.  Editing creates distortions.

Was Fellers information always accurate?  Did Rommel really always know what the British intentions were?  Their groupings?  Ferris writes that the British kept Fellers ignorant of “their intentions, plans and dispositions,” even by the fall of 1941.  He adds that Fellers grossly underestimated British armor by hundreds of tanks in the spring of 1942.  F.W. Von Mellenthin, who served on Rommel's staff, admits they may not have launched their offensive at this time had they known actual British armor strength.

But larger than a corrupted, numeric subset that was inaccurate at times is the problem of Rommel himself.  He was well-known for performing his own reconnaissance, leading from the front at times, and being constantly on the move.  According to Ferris, this approach led to Rommel, “abandoning signals touch with all of his headquarters for days at a time.”  Obviously, he couldn't receive Fellers information if no one could reach him.

Berhendt admits that he isn't sure how much of the intelligence Rommel used.  Ferris states that Rommel was not a good consumer of intelligence.  He concludes, “The 'Good Source' [Fellers] did not matter much because Rommel did not use it well.”

Many factors went into Rommel's victories and defeats in the Western Desert Campaign.  Bold claims and dramatic statements attributing them to a single source such as Fellers are enticing but inaccurate.  While Fellers messages played a role, it was not a decisive one.