the beginning of the Cold War, was another attempt to ensure centralized coordination and routing of intelligence. But, as the 9/11 commission and others have found, over the years, it, too, succumbed to turf battles that hindered effective sharing of intelligence across bureaucratic lines of authority. One of the chief recommendations now under debate is the creation of a new post of a single “director of national intelligence” or “intelligence czar,” to make sure that information from all sources—spies, informers, satellites, communications intercepts, and so on—is brought together and analyzed as a whole.

Long before Pearl Harbor or September 11, there was another instance when a few officials recognized that the key to making an intelligence system work lay in being able to “connect the dots.” It, too, was a lesson that had been learned the hard way—and then was promptly forgotten once the crisis was past. It, too, was the brainchild of an extremely tough-minded civilian lawyer-turned-Army colonel, who brought an outsider’s perspective and a skill at dealing with complex, often messy problems to what had customarily been the narrow purview of military men.

Not many people today have heard of George H. Sharpe. I hadn’t heard of him until I happened to read a short CIA historical report that—in little more than a passing reference—credited him with having established “the first ‘all-source intelligence’ organization in U.S. history.” Sharpe was the director of the Bureau of Military Information of the Union’s Army of the Potomac during the Civil War. In contrast to all the dubious and romantic tales of femme-fatale spies and daring moonlight dashes on horseback that still choke the literature of Civil War espionage, the story of Sharpe’s unheralded reign as an intelligence chief is one of quiet, meticulous analysis—which nonetheless produced a series of coups that helped change the course of the Civil War from Gettysburg to the final siege of Richmond.

Although he never had more than about 70 men on his full-time payroll, and though the intelligence challenge he faced was but a fraction of what a modern spymaster must confront, Sharpe nevertheless pioneered the fundamental principle of modern intelligence analysis and organization that remains valid today. As the CIA historical report notes, he “obtained, collated, analyzed, and provided reports based on scouting, spying behind enemy lines, interrogations, cavalry- reconnaissance, balloon observation, . . . flag-signal and telegraph intercepts, captured Confederate documents and mail, southern newspapers, and intelligence reporting from subordinate military units. This structured approach, which ended with the Confederate surrender, was not reinstitutionalized until 1947, when the CIA was created.”

George Henry Sharpe was born in 1828 in Kingston, New York on the Hudson River. His father, who died when George was two, was a wealthy merchant, and the family was left well-off. Sharpe graduated from Rutgers at 19—delivering the salutatory address, in Latin—then went to Yale Law School and breezed through the New York bar exam at the age of 21, worked in New York City for the law firm

of Bidwell and Strong, then traveled to Europe and worked for the U.S. legations in Vienna and Rome, where he acquired diplomatic and linguistic skills that his contemporaries would remark on throughout his life, at last returning to his hometown to set up his own law practice.

With his receding hairline and large drooping mustache, he looked more like a prosperous small-town merchant than a future spy chief. But, to those who knew him, there was no doubting either his intellectual formidability or his native vigor, determination, and charisma. When Fort Sumter was attacked, he hastened back home from a business trip and, without bothering to consult the colonel of the 20th New York militia (and apparently overlooking the fact that he had resigned his own captain’s commission in the regiment earlier that year), immediately began enrolling volunteers and assumed command of the regiment’s Company B. A year later, in response to President Lincoln’s call for 300,000 volunteers, he used several thousand dollars of his own money to raise a new regiment, the 120th New York, of which he took command as colonel. The unit played a small but steady part in the fighting around Fredericksburg in November and December 1862. Fiercely loyal to his regiment, Sharpe refused an offer of promotion and the command of a brigade.

In January 1863, Major General Joseph Hooker became the third general named by Lincoln to command the Army of the Potomac, succeeding the irresolute George McClellan and the ineffectual Ambrose Burnside. Hooker was, in many ways, an unsavory character. He had schemed behind Burnside’s back to get his job; he had brayed to the press that what the country needed was a dictator, and it was pretty clear whom he had in mind for that position. As for Hooker’s headquarters, it was described as “a place which no self-respecting man liked to go, and no decent woman could go. It was a combination of barroom and brothel.”

But, as a leader of men and—more unusually for a man of action, as an administrator, also—Hooker well deserved his epithet “Fighting Joe.” Upon taking command, he immediately cracked down on corruption in the supply services, ordered hospitals and food to be improved, and began demanding results everywhere. This was exactly what Lincoln had been hoping for. When the president appointed Hooker, he had handed him a gently chiding letter in which he admonished the general for his dishonorable conduct toward Bumside and for his remark about the country’s needing a dictator. “Of course it was not for this, but in spite of it, that I have given you the command,” Lincoln wrote. “Only those generals who gain successes, can set up dictators. What I ask of you is military success, and I will risk the dictatorship.”

Among the other deficiencies that Hooker found in his new command was the absence of a single document about the Confederate forces that he now faced across the Rappahannock River. “There was no means,

no organization, and no apparent effort, to obtain such information,” he fumed. McClellan had created an intelligence service, of sorts, in the Army of the Potomac; it had been headed by the famous Chicago detective Allan Pinkerton, who ran the private detective agency that bore his name. (Before the war, when McClellan was vice president of the Illinois Central Railroad, that company had been one of Pinkerton’s largest clients.) As McClellan’s spy chief, Pinkerton had proved reasonably successful in breaking up some none-too-professional Confederate spy rings in Washington. He also had done a conscientious job of gathering information from interrogations of Confederate prisoners and deserters.

But his isolation from the rest of McClellan’s headquarters led to a hopelessly splintered authority over the flow of intelligence and a repeated failure to synthesize and cross-check information against other sources. Scouting of enemy positions was left to individual division commanders. The cavalry, traditionally the eyes and ears of the Army, reported to McClellan directly, bypassing Pinkerton, and, at this stage of the war, was also notoriously lackadaisical in pursuing aggressive reconnaissance of enemy positions. Balloonists and Signal Corps observers sent their reports to the nearest division headquarters or to McClellan, again leaving Pinkerton out of the picture. Other key sources, such as southern newspapers, were ignored altogether.

Befitting the worldview of a civilian detective, Pinkerton sent what spies he did infiltrate behind enemy lines almost exclusively to Richmond, the enemy capital—”probably,” writes Edwin C. Fishel, author of The Secret War for the Union, “because a city could be penetrated much more readily than a field army’s headquarters.” Pinkerton’s spies in Richmond provided ample fodder for dramatic tales of derring-do published after the war, but they actually obtained little information of importance.

Worst of all, Pinkerton’s reports to McClellan almost always consisted of raw, undigested intelligence, without any context or analysis. This left to McClellan himself the crucial job of interpretation, and that was where the real mischief took place. Although Pinkerton has been ridiculed for years by historians for providing McClellan with grossly inflated estimates of Confederate troop strength, in fact, this was but a symptom of the unprofessional way in which McClellan’s whole intelligence system was organized. Initially, Pinkerton actually put together a surprisingly accurate picture of the Confederate “order of battle"—its organization into corps, divisions, brigades, and regiments. But, before he could take the next step and use this organizational chart to build up an estimate of Robert E. Lee’s total troop strength, McClellan was already whining to Washington that he needed more men before he could attack the Rebel army, claiming that Lee outnumbered his forces by three to one.

When Pinkerton reported that the Confederates numbered 98,400—double the number now known to be the actual figure, but still far less than the 170,000 McClellan had claimed—McClellan started leaning on him to adjust his figures, “to be sure

and cover the entire number of the Enemy.” And Pinkerton did his best to defer to his commander’s wishes.

Under a commander more supportive of objective intelligence analysis, Pinkerton would almost surely have produced a much more accurate estimate. But, as Fishel observes, the much-maligned detective “worked for a general who did not really use intelligence except to justify his own dislike of fighting.”

When McClellan was relieved of his command, Pinkerton folded his shop, taking all his files and his team of civilian detectives with him. Under McClellan’s successor, Burnside, the remnants of the Army of the Potomac’s “secret service” consisted of a single private. Hooker, when he came in, was determined to remedy that at once. He ordered his provost marshal general, Marsena Patrick, to “organize and perfect a system for collecting information as speedily as possible.”

It was a tall order for the provost marshal general, whose major job was arresting deserters. Patrick complained in his diary: “I am trying to make up a system of Secret Service, but find it hard to organize where there is so little good material. . . . I do not fancy the class of men & think they do not fancy me.” But a few days later, Patrick recorded a stroke of luck; he had had a “long conversation” with Colonel Sharpe about the job of chief of the “Secret Service Department.” Patrick noted that Sharpe “appears well, & I think would be a pleasant man to be associated with.”

Patrick had been a brigade commander in a division that included Sharpe’s regiment, and so was at least generally acquainted with him. Besides being a lawyer, a point Patrick commented on favorably, Sharpe had some powerful political connections that no doubt recommended him. He had been active in local Republican politics in upstate New York, and his law partner was a U.S. congressman well connected with Secretary of State William Seward.

Sharpe was reluctant to leave his regiment, but, in the end, was talked into taking the headquarters assignment. He immediately threw himself into the task, organizing the “Secret Service Department” into an efficient, businesslike organization, reflected in the new low-key name it soon acquired, the Bureau of Military Information (BMI). Almost immediately, there began to flow into Hooker’s headquarters a steady stream of understated, but confident reports on Lee’s army, its movements, disposition, and intentions. Many of them in Sharpe’s own hurried handwriting, others filling hundreds upon hundreds of message blanks from the U.S. Military Telegraph service, they today fill box upon box in the National Archives, and they cover everything from reports of individual interrogations of Confederate prisoners of war to comprehensive assessments of the enemy’s organization and maps reconstructing the locations of enemy units and fortifications.

Within a few months, Sharpe’s bureau was able to produce what

would prove a deadly accurate tally of Lee’s army just before the Battle of Chancellorsville. By correctly identifying the Confederate order of battle down to individual regiments, and correlating hundreds of interrogations of prisoners, deserters, and refugees with other scraps of intelligence, the BMI produced, on April 28,1863, an estimate that placed Lee’s strength at 55,300, a mere 2 percent less than today’s generally accepted figure of 56,500. A distinguished visitor who inspected Hooker’s headquarters at the time wrote that “we have a moral certainty of all that is necessary to know in regard to the enemy, every regiment and brigade, division etc., all the latest arrivals and departures, etc., all collated, compared from many sources and fully confirmed. The secret service of Gen. Hooker is far superior to anything that has ever been here before. . . . Nothing transpires in the enemy’s camp that he is not speedily informed of.” Besides indicating the strength and location of enemy units, Sharpe’s reports regularly supplied information on the whereabouts of artillery batteries, the state of repair of railroads, local topography, the locations of mines, the quantities of enemy supplies, troop morale, recent orders to the troops, and impending movements.

For the first time, the Army of the Potomac knew more of what was happening on Lee’s side of the Rappahannock than vice-versa. And it was Sharpe’s work that directly made possible Hooker’s brilliant march on Lee’s rear at Chancellorsville a few days after his April 28 intelligence estimate, a coup undone only by Hooker’s subsequent sudden loss of nerve, which prevented the Union corps from following up their huge advantage. (Another intelligence coup at Chancellorsville was the discovery that the Confederates had broken the Union Army’s flag-signal code; Union signalmen then signaled a fake message that sent Jeb Stuart’s cavalry off in pursuit of a non-existent Union cavalry raid.)

Constantly updating the enemy order of battle was Sharpe’s most potent weapon, for not only was it valuable intelligence in its own right, but it served as a continual check on the reliability of information received from sources. As the historian William Feis notes in his book Grant’s Secret Service, every Confederate prisoner captured was asked by Sharpe’s interrogators to identify his regiment, brigade, division, and corps, when and where he had entered the front lines, and how and where he had been captured or why he had deserted. Sharpe later explained that “the state of our information” built up in this manner “has been such as to form a standard of credibility by which these men were gauged, while each was adding to the general sum.”

Sharpe’s interrogation reports are also notable for the swift and keen assessments of their sources’ credibility that almost always accompanied them: “From a reliable source"; “Honest but not intelligent"; “Informants are intelligent and not disposed to state more than they know to be true"; “All this is rumor, but it comes from so many sources, it is worth attention"; “Informant

is very ignorant of the route he has taken or names of places he has been, so much so as to raise a doubt as to whether he has made the trip at all.”

Supplementing the intelligence derived from interrogations was a mass of reports from Sharpe’s own spies and scouts, from cavalry reconnaissance, and (less frequently, but regularly) from intercepted enemy flag-signal messages, purloined letters, and “open sources,” such as Southern newspapers. Although Sharpe did not directly command cavalry or balloonists or Signal Corps observation posts, he was vigorous in obtaining their reports and fitting them into his growing matrix of knowledge about the enemy. On occasion, officers from his bureau were informally attached to cavalry units, to help steer their intelligence-gathering efforts where they were most needed.

All of this data flowed into Sharpe’s headquarters, which initially had a permanent staff of about 18 men who traveled with the Army of the Potomac’s headquarters and reported directly to Sharpe. Over time, the range of topics covered was immense, encompassing not just tactical military intelligence, but also political, economic, and psychological intelligence about the Confederacy: the price of flour in Richmond, the schedule of train departures, the state of civilian morale.

As its success and efficiency became known, Sharpe’s bureau grew, setting up “branch offices” throughout the Eastern theater, each with a team of some 30 to 40 scouts. Within his headquarters, Sharpe was soon ably assisted by several deputies, all men after his own intelligent and careful stamp. His chief assistant, John C. Babcock, an architect in civilian life, had charge of the critical order-of-battle assessments, keeping meticulous records, preparing reports, and constantly updating maps and organizational charts of the Confederate Army. Captain John McEntee, a former neighbor of Sharpe from upstate New York, oversaw scouting operations and interrogation reports. At its core, the BMI remained a small operation, but it was now a strikingly professional one that knew its business and produced a highly refined product that went directly to the top generals. “As a result of the BMI’s all-source capability,” writes Feis, the Union commanders “received not an assemblage of undigested bits of news seemingly of equal weight, but true intelligence, the finished product of systematic information analysis.”

Unlike Pinkerton, Sharpe sent his scouts — most of them non-commissioned officers and enlisted men —right into the enemy military camps. Some masqueraded as smugglers or federal deserters and hung about a rebel camp for a few days before vanishing back across the lines; others, even more daringly, donned Confederate uniforms and posed as soldiers separated from their units or as members of irregular Confederate forces like John S. Mosby’s rangers.

One especially daring BMI scout, Sergeant Milton W. Cline, managed to attach himself to a Confederate cavalry captain and rode the entire length of Lee’s lines a few days before the Battle of Chancellorsville. Among Sharpe’s papers are requests to federal military authorities for tens of thousands of dollars in captured Confederate currency for him to

give to his military scouts and civilian spies to use.

Another frequently valuable source of direct news from behind the Southern lines was “contrabands,” escaped slaves who had often been in a position to gather a great deal of information without arousing suspicion. Believing that slaves were too ignorant and docile either to understand military matters or to risk their lives aiding the Union, southerners often spoke freely in front of them about recent or planned troop movements and military activity. “Three contrabands have just been sent in bringing important information,” Sharpe reported on November 14, 1863.“They heard from their master in conversation that the Army was moving toward Louisa C[ourt] H[ouse].”

It was information from an escaped slave that would provide the crucial advance warning in June 1863 of Lee’s sudden movement north into Maryland and Pennsylvania, the move that culminated in the decisive Southern defeat at Gettysburg. After Chancellorsville, Lee’s army had undergone a rapid series of reorganization and reinforcement maneuvers that left Sharpe scrambling to pinpoint its strength and exact whereabouts. The locations of two entire divisions under General James Longstreet were a mystery; some reports had them in Richmond, others on their way back to Fredericksburg.

By May, a combination of interrogations of deserters, reports of spies and informants in Fredericksburg, visual observations from Signal Corps flagmen (who could often spot enemy movements at long distances from their tall signal towers), and balloon observations enabled the piecing together of a more solid picture. On May 27, Sharpe was able to report to Hooker in considerable detail on the locations of Lee’s various divisions. He also reported that something big was afoot: “The Confederate Army is under marching orders, and an order from General Lee was very lately read to the troops, announcing a campaign of long marches and hard fighting, in a part of the country where they would have no railroad transportation.

“All the deserters say the idea is very prevalent in the ranks that they are about to move forward upon or above our right flank.” But still, there was no precise information on Lee’s intended objective.

That suddenly changed late on June 12, when a telegram arrived at Sharpe’s headquarters from his assistant Captain McEntee, who was deployed at the moment with the Union cavalry forces: “A contraband captured last Tuesday states that he had been living at Culpepper C. H. for some time past. Saw Ewells (Jackson's) corps pass through that place destined for the Valley & Maryland. That Ewells corps had passed the day previous to the fight [the cavalry battle at Brandy Station, which had just taken place on June 9] & that Longstreet was then coming up.” Later that evening, a second wire arrived from McEntee: “Continuation of last statement. Gen. R. E. Lee Hd Qrs were at Culpepper C. H. on Tuesday

last. Ewell arrived there with command night of 6th. Cooked 4 days rations. Marched morning of 7th. Column half day passing through town. Longstreet arrived night of 8th & marched 9th. This boy knew many of the troops. I think statement reliable.”

Once again, it was Sharpe and Babcock’s meticulous attention to the Confederate order of battle that provided the vital check on the reliability of the information supplied by the young contraband, whose name was Charlie Wright, who had been able to name more than a dozen individual regiments in Ewell’s and Longstreet’s corps, and all of them agreed with Sharpe’s data.

For years, conventional cavalry reconnaissance was credited for Hooker’s order to set his troops on the move, shadowing on the eastern side of the Blue Ridge Lee’s movement into Maryland and Pennsylvania. But Edwin Fishel’s research in the Army of the Potomac records in the National Archives, which turned up McEntee’s telegrams, established conclusively that it was the BMI’s reports based on Wright’s information, combined with the bureau’s painstakingly assembled order-of-battle data, that provided the crucial warning. Hooker’s timely movement both shielded Washington from Lee’s forces and allowed the Union Army to arrive at Gettysburg in time to seize the advantageous high ground.

Sharpe’s bureau continued to follow Lee’s movement into Maryland and Pennsylvania, with reports from local citizens and other sources. A telegram wired to Sharpe from his assistant McEntee in Baltimore reported on June 24: “The students of St. James College Hagerstown came down from there today—They report . . . Gen. Ewells corps with 70 pieces of artillery as having passed through Hagerstown the last passing yesterday going towards Chambersburg.…They report no force between Hagerstown and Frederick on the Boonsboro road, except one cavalry camp 4 miles from Hagerstown.”

Still, fittingly, it was once again Sharpe’s emphasis on the patient, unspectacular amassing of data, rather than any dashing cloak-and-dagger exploits, that provided the final key to the Union victory at Gettysburg. Two days of desperate Rebel attacks had been beaten back, but Major General George Meade, the new Union commander, was uncertain whether his own troops could take much more if Lee decided to renew his attack the next day.

However, Sharpe’s order of-battle expert, Babcock, had discovered a fact of the greatest importance from a careful tabulation of the 1360 Confederate prisoners taken that day: The prisoners came from nearly a hundred different regiments, representing every single one of Lee’s brigades, save only the four brigades of Pickett’s division, which was still coming up from the rear. Lee, in other words, had no fresh troops left to throw into the fight, except for Pickett’s lone division.

At 9:00 P.M. on July 2, Meade called a council of war to decide whether the Union Army would stand and fight for another day or withdraw. Sharpe was summoned, and he presented his finding that Lee had already used almost everything he had. When he was finished speaking, one of the generals exclaimed to Meade,

“General, we got them nicked!” A moment of silence then settled over the room. Sharpe, who had scarcely eaten all day, had been glancing forlornly at a small plate of crackers and a half-pint of whiskey sitting untouched on the table in Meade’s headquarters. Finally, another of the generals in attendance spoke up: “General Meade, don’t you think Sharpe deserves a cracker and a drink?”

The council unanimously voted to stand and fight. The next day, Pickett secured a name for himself in history synonymous with futile courage by sending his men charging against Union artillery—which mowed down two-thirds of them in half an hour.

Sharpe’s bureau had its share of failures. The worst arguably occurred in June 1864 when the BMI failed to detect General Jubal Early’s II Corps slipping out of the line to launch a lightning raid through the Shenandoah Valley, into Maryland, and to within five miles of the White House. In response, Sharpe built up his intelligence sources in the Valley, recruiting three Virginia Unionists who lived near key railroad depots and who could report on troop movements. He also strengthened his ties with two spy networks in Richmond, one led by the superintendent of the Richmond, Fredericksburg & Potomac Railroad, the other by the subsequently (and, for once, deservedly so) famous Union spy Elizabeth Van Lew. The BMI was now producing a series of “information maps” that plotted the locations of all the Confederate forces. They were distributed on a regular basis to Union commanders.

By being able to watch for any shift of troops between the Richmond-Petersburg area and the Valley, Sharpe provided the crucial advantage that allowed Ulysses S. Grant (who had assumed command of all the Union armies and to whose staff Sharpe was now attached) to “take from Lee his last remaining weapon: strategic mobility,” as the historian William Feis put it. Grant himself said that, with Sharpe’s intelligence system, Lee “could not send off any large body without my knowing it.”

Present at the Confederate surrender at Appomattox, Sharpe, now brevetted a brigadier general, had one last opportunity to show off his mastery of his order-of-battle intelligence about the enemy. Under the terms of the surrender, Lee’s officers and men were required to give their vow not to take up arms again against the government of the United States. A small printing press was set up, and some 28,000 parole forms were run off and distributed to Lee’s men. General Sharpe supervised the operation for a week—and frequently was able to astonish Southern soldiers by telling them which brigade and division they belonged to when they themselves did not know.

After the war, Sharpe was sent by Secretary of State Seward to Europe on a “secret mission” to try to track down Americans living there who were believed to have been involved in the Lincoln assassination. In 1870, he was appointed by his old boss, now President Grant, as United States marshal for the Southern District of New

Yor,k and at once went after the Boss Tweed ring. “Many times, General Sharpe’s life was threatened by the followers of Tweed,” his obituary in The New York Times would later note, but Sharpe nonetheless succeeded in securing convictions against two of the most notorious perpetrators of election fraud and in overseeing an honest census that purged the election rolls of fraudulent voters. He later served as speaker of the New York State Assembly. He died in 1900.

Unlike the dozens of self-proclaimed spies and secret agents and spymasters (Pinkerton among them) who wrote flamboyant, self-dramatizing, or even wholly invented memoirs of their Civil War exploits, George Henry Sharpe and his aides kept their accomplishments to themselves. His lengthy obituary in The Times devoted several paragraphs to his “distinguished army career,” but referred to his assignment as chief of the Bureau of Military Information in but a single sentence, noting that it was “a dangerous and delicate mission”—and providing no further enlightenment.

Yet, as a recent U.S. Army study on the BMI notes, a fair reading of history would unquestionably place Sharpe among the ranks of America’s most important intelligence directors. He established the first truly modern intelligence organization, pioneering the practice of “comparing intelligence from a number of sources and evaluating it before passing it along.” He also clearly saw his job as being one of telling it like it is, rather than telling it like the commander wanted to hear it.

That, too, is a lesson for our times.