John Keegan, a much-admired military historian for three decades, writes in a way that sets him apart from most authors who began their working lives as academics. He brings to the literature of war a deep affection for revealing detail, and it's clear that he loves to be surprised by what he learns. His pleasure animates the material for his readers.
Keegan theorizes expertly, he summarizes brilliantly, but he quickens to specific facts and images. A fine journalist (as readers of the Post's Comment pages know), he never forgets that history remains news until we know it, and becomes news again whenever we see it in an unexpected way.
In the first of his well-known books, The Face of Battle (1976) Keegan, typically, tells us that the 15th-century British soldiers at Agincourt fought with passion partly because they were fuelled by alcohol and partly because they planned to hold prisoners for ransom and loot battlefield corpses. In his new book, Intelligence in War: Knowledge of the Enemy From Napoleon to Al-Qaeda (Key Porter), he mentions in passing that the Royal Navy, long before telegraphy and radio, was yearning desperately for speed in communication. In the 18th century it devised a relay system of semaphore stations that could transmit a message from London to the English coast in two minutes.
Keegan begins Intelligence in War with a question: How much does military intelligence matter to victory? Examining case studies from Nelson's pursuit of Napoleon's fleet across the Mediterranean in 1798 to the Battle of the Atlantic in the 1940s, he concludes that intelligence systems, while essential, mean less than we imagine. He endorses the opinion of David Kahn, a leading historian in this field, that military intelligence is a secondary factor in war.
Wars are usually won by strength; espionage is less vital, and so are behind-the-lines special operations. Libraries have been written on the Cambridge University traitors, for instance, but they seem to have helped the Soviet Union very little.
The French Resistance during the Second World War, vigorously encouraged by the British, was far less effective than Winston Churchill hoped at the time and popular culture later suggested. Movies of the 1940s convinced me that Resistance fighters blew up just about every bridge and train in France, but the record shows that the Germans were neither impressed nor inconvenienced.
Still, in one of his most effective studies Keegan leaves us to weigh for ourselves the balance between intelligence services and fighting men. This is when he describes the way code-breaking led to the U.S. victory at Midway in June, 1942, the battle that reversed the course of the Pacific War only six months after Pearl Harbor.
Japan had more carriers and better planes, but U.S. intelligence helped offset those advantages with the element of surprise. American code-breakers, scoring the greatest single coup in the history of naval intelligence, guessed where the Japanese carriers could be found. But even then, the Americans had to fight desperately and needed luck and one shrewd airman to succeed. Five of the six American squadrons that attacked the carriers were shot out of the skies, and up to the last few minutes Midway appeared lost.
It seemed the battle was finished and the Japanese fleet had escaped over the horizon. But it happened that a Japanese destroyer had briefly held back to fight off a U.S. submarine. Then it raced to rejoin the other ships, leaving behind a vivid white wake. The leader of a fleet of dive bombers, Lieutenant-Commander Clarence McCluskey, couldn't see the destroyer but spotted the white streak in the water beneath him, guessed what it meant, and followed. The destroyer led him to three carriers, all of them unprepared for battle, their planes (having battled with the other American squadrons) waiting on deck for repairs and reloading, trailing fuel hoses.
McCluskey's planes destroyed all three carriers in five minutes. So on the morning of June 4, 1942, between 10:25 and 10:30, "Japan's plan to conquer the Pacific was reduced to ruins."
That white wake in the Pacific water stays in the reader's mind; Keegan cites it three times. He doesn't forget to include one of the odd Japanese blunders that also helped the Americans, a classic mistake that appears several times in the history of encryption. In the spring of 1942 the Japanese had a new code book, but they had trouble distributing it because their ships were spread across the Pacific. So they transmitted some messages in both the old and the new codes, to ensure accuracy of reception. The Americans already knew the old code, so the two versions worked like a Rosetta Stone. By comparing the old ciphers with the new, the Americans could understand the changes in code the Japanese were making.
Keegan suggests that intelligence services, in the new war on terrorism, will need to revive traditional spying. Al-Qaeda, being a loose confederacy, can't be fought with modern techniques. What the anti-Islamist campaign requires is the old-fashioned one-man operation, the spy celebrated in literature of long ago -- like Kipling's Kim, with "his ability to shed his European identity and to pass convincingly as Muslim message-carrier, Hindu gallant and Buddhist holy man's hanger-on," or a John Buchan character abandoning and adopting new disguises. It will be ironic, Keegan suggests, "if the literature of imagination supplies firmer suggestions as to how the war against terrorism should be fought than academic training courses in intelligence technique."
American and European analysts, having tapped millions of phone calls to catch Al-Qaeda messages, are currently dealing with the greatest flaw in electronic surveillance: It produces far too much data. This isn't a new problem. In 1941 the Americans had many more transcripts of Japanese naval signals than their small staff of decryptors and translators could handle. Some of those messages would probably have warned of the attack on Pearl Harbor, but they weren't decoded until 1945, when they were of only historical interest.
As Keegan says, intelligence material tends to be muddled, partial, contradictory. It makes me think of the fossil record, which can never be more than a collection of fragments. To make any sense of what they gather, intelligence analysts need as much imagination and intuition as paleontologists.