Once authorized to employ force, Chief of Staff MacArthur refused to allow even President Herbert Hoover to intervene. He considered the military subordinate only until military operations began. Against the wishes of Hoover, MacArthur used massive force to scatter the veterans who he later termed "insurrectionists" animated by the essence of revolution.*

Again the administration, this time Republican, refused to question the surveillance tactics of the Army and again it came under much criticism for the activities of the military in the next election. Criticism of the military drove G-2 underground once more. Intelligence officers were instructed to be discreet in their activities so as to prevent disclosure of the Army's surveillance Again without seeking the approval of their civilian superiors, they continued to gather information on the political activities of civilian groups. In 1936, G-2 began to develop new civil disturbance contingency plans. One of the most elaborate was the intelligence plan for the Sixth Corps Area which encompassed Illinois, Michigan, and Wisconsin. This was an area of strong isolationist sentiment and aggressive labor organizing. The Sixth Corps plan called for the collection and indexing of the names of several thousand groups, such as the American Civil Liberties Union and pacifist student groups which were labeled as communist subsidiaries. Sources of information were to be the Justice Department, the Treasury Department, and the Post Office as well as local state police. In addition, agents were told to contact private intelligence bureaus run by corporations such as General Motors which paid almost a million dollars to the Pinkerton Company between 1934 and 1936 to conduct labor espionage and to sabotage the organizing efforts of the United Auto Workers. Like other G-2 offices at the corps level, those in the Midwest kept watch on pacifist and civil liberties groups because Army intelligence had decided that the existence of these political groups helped more radical groups which might be a direct threat to the government or the military. In other words, the military decided what kind of political activity needed monitoring.

In those days, when the public thought that Army intelligence had been contained, G-2 officers were making confidential reports on the American Federation of Labor and the Congress of Industrial Organizations, and spot reports on communist activities. Ralph Van Dieman, the World War I chief of the MID who had retired to San Diego in 1929, set up his own master anti-communist files and acted as a country-wide clearing house for domestic intelligence. Because the Van Dieman files, now in the hands of the Senate Internal Security Subcommittee, have not been made public, it is not known exactly what his relationship to G-2 was, but it is known that he received reports from G-2 as well as from the FBI, local police, and volunteer groups like the American Defenders. Thus G-2 was able to maintain its structure intact and to continue to define its mission at home without civilian guidance during the inter-war years.

WORLD WAR II

After France surrendered to Germany in the summer of 1940, G-2 urged corps area and departmental commanders to collect domestic intelligence for counter-fifth column plans but agreed to rely primarily on Justice Department information. President Franklin Roosevelt was alert to the dangers of letting Army intelligence expand freely on the crest of espionage and fifth column fears but after the United States officially entered World War II in December 1941 that became increasingly difficult. Although the FBI maintained nominal control over internal security during the war, Army intelligence did not cease to function domestically. G-2 continued to collect political intelligence for War Plans White. It reported on radical labor groups, communists, Nazi sympathizers, and "semi-radical" groups concerned with civil liberties and pacifism. The latter, well intentioned but impractical groups as one corps area intelligence officer labeled them, were playing into the hands

Roger Daniels, The Bonus Marchers: An Episode of the Great Depression (Westport, Conn., Greenwod, 1971), gives an account of the role of Army intelligence and MacArthur. 5 Sixth Corps Area, Emergency Plan-White, December 1936, AG No. 386.

of the more extreme and realistic radical elements. G-2 still believed that it had a right to investigate "semi-radicals" because they undermined adherence to the established order by propaganda through newspapers, periodicals, schools, and churches.

At first the G-2 did not challenge the supremacy of the FBI at home. Nor did it support the mass evacuation of the Japanese from the West Coast, apparently because of the restraint imposed by Chief of Staff George Marshall. But gradually the Counter Intelligence Corps of the Army (CIC) began to take a more active part in surveillance of civilians by including counterespionage, counter-sabotage, loyalty investigations, security education, plant protection, and surveillance for treason, sedition, subversion, and disaffection among its duties. Collection of information on political groups once more became an important part of a preventative security program which used voluntary informants and investigators to collect information. Preventative plans called for recruitment of the American Legion, Veterans of Foreign Wars, and other "patriotic" organizations to help watch the home front. A delimitation agreement set up with the FBI had already given CIC agents wide latitude but even these wide limits were exceeded as the military began to demand more control over industrial labor.

How extensive this military wartime surveillance came to be will not be known until all the files are open. But how far it mingled with civilian security work can be seen from a report later made on Van Dieman's San Diego files. By fall 1944, Van Dieman's agents were reporting on communist meetings, on adult discussion classes of the First Unitarian Church, on activities of aircraft labor unions, on Democratic rallies at a local junior high school, and on private groups within private homes. He was receiving copies of G-2 reports from all parts of the country. The commanding general at the Marine Corps Base in San Diego and the commander of the Eleventh Naval District were sending him reports on communists. He was distributing copies of his reports to military intelligence offices in Los Angeles and to the FBI in San Diego.

COLD WAR ERA

With the end of war, the military intelligence once again refused to demobilize its home front army. Again Van Dieman's files may be taken as a measure of the continuance of military surveillance of civilians. He received reports on the National Urban League, the Youth of All Nations, on labor unions, on scientists, on movie stars. Information from military intelligence reports in the Van Dieman files found its way into the hands of selected politicians and to selected members of the House Un-American Activities Committee in Washington and of the California Un-American Activities Committee in Sacramento. Later G-2 Tokyo, which survived and flourished under General MacArthur in Japan, leaked confidential records to right-wing publicists who used them as the basis of political campaigns against the Democratic policies. General Charles Willoughby, while still chief of G-2 Tokyo, launched his own investigation of American political groups to link them to an international communist conspiracy.

During the Korean War, military intelligence set up a central records facility at Fort Holabird, Maryland, where it began to catalogue domestic and foreign reports from military and civilian investigative agencies. In 1952, when Van Dieman died, a large portion of his vast files with information on 125,000 individuals was also taken over by Army intelligence. In addition, Army intelligence received information from counter-intelligence groups within National Guard units, such as the miniature Counter Intelligence Corps established by the California Adjutant General throughout that state.

Again, we can only estimate the effect of Army surveillance during the Cold War because information is not yet available, but the use of Army

The only information yet available on World War II is in Victor J. Johansen "The Role of the Army in the Civilian Arena, 1920-1970," U.S. Army Intelligence Command Study (1971).

32-996-74-13

intelligence records and the campaigning of former intelligence officers did much to cultivate Americans' fear of surversion and to provide the basis for the growing power of the military at home.

CIVIL RIGHTS ERA

By 1956, the Department of the Army was claiming some continuous peacetime responsibility for collecting information on subversive groups. When President Dwight Eisenhower mobilized troops in 1957, the Counter Intelligence Corps preceded federal troops by two weeks to Little Rock to watch the school and report on local press coverage. The G-2 used the CIC of the 4th Army, intelligence staff officers from the Airborne Division, and FBI agents indiscriminately in the hectic days which followed. Military counter intelligence personnel had charge of surveillance of the nine black youths enrolled at Central High School and of monitoring the Ku Klux Klan and other potential trouble makers. Surveillance in this case continued on the orders of the Secretary of the Army after regular Army troops had been replaced by federalized National Guard. The next year, a Strategic Capabilities Plan restricted the use of intelligence personnel in monitoring civil disturbances until the President judged deployment of troops seemed imminent.

During civil rights activities in Montgomery, Alabama in 1961, the Army developed plans which assumed the Continental Army Command would conduct civilian investigations in domestic emergencies resulting from civil disturbances, and which allowed the Army to employ agents to collect information on civilians when the use of federal troops was "probable." Army agents could only operate within the investigative jurisdiction of civil authorities with specific authority from the President as Commander-in-Chief. The Army was not called into Montgomery in 1961. In 1962, however, members of the 111th Intelligence Corps group conducted covert investigations of civilians, apparently in violation of their directive which called for specific authority from the President. Agents probed "extremist" groups, the reaction of civilians to troop movements, investigated "agitators," and compiled "black, white and gray lists." At the same time, agents failed to assemble adequate reconnaissance information.

A 1963 Continental Army Command plan left surveillance of civilians in the hands of the FBI but specifically authorized the military to file spot reports "as required" on events which might develop prior to the implementation of the plan. Ordinarily civil disturbance information was to be collected mainly through liaison with civilian authorities and through news media reports but an Army commander, if he felt the situation warranted it, could order covert operations if coordinated with the FBI. This plan removed surveillance from central control, allowing the same decentralization which had in the past led to serious invasion of the rights of citizens. In 1965, when this early warning system was transferred to the Army Intelligence Command, the Continental Army surveillance system continued in violation of regulations and without the knowledge of senior Army commanders. Thus the stage was set once more for the expansion of military surveillance of civilians during the next home front crisis.

CONCLUSIONS

In 1967, as I was finishing my book on World War I, it was already evident that Army intelligence had begun a vast surveillance program alarmingly similar to that begun fifty years before. I urged then that the line between civilian and military authority be clearly demarcated. Surveillance again occurred at home with almost no civilian control. In 1970, there were also parallels with 1919 when those who questioned military surveillance of civilians found Army intelligence and its supporters intransigent, willing to use subterfuge, ready to wait out press criticism, and able to plead secrecy to keep its activities beyond scrutiny.

Surveillance of civilians by the military took place during every major home front crisis between 1917 and 1967 regardless of which party was in power. It occurred with almost no civilian control because civilian officials, even those charged directly with the affairs of the War Department or

Defense Department were unable to effectively monitor the operations of military intelligence agents. As the military bureaucracy grew and became more technologically efficient, the difficulty of civilian control became greater. Meanwhile, the practice of the military of defining its mission of civilian surveillance in the broadest possible terms and moving in times of crisis from countering espionage to countering political dissent continued. Without further institutional safeguards to control military surveillance it seems highly likely that this practice will not only continue in the future but become an increasing threat to the constitutional rights of civilians.

"S. 2318 AND THE MILITARY'S LEGITIMATE INTELLIGENCE NEEDS." A STATEMENT BY PROFESSOR CHRISTOPHER H. PYLE FOR THE SENATE SUBCOMMITTEE ON CONSTITUTIONAL RIGHTS, APRIL 9, 1974.

On February 24, 1971, I had the honor of testifying before this subcommittee at its hearings on Federal Data Banks, Computers and the Bill of Rights. At that time I sketched out the history of the Army's surveillance of civilian politics during the late 1960's and spoke of the need to bring it under control.

Since that time, much has happened. The Department of Defense has issued regulations attempting to curb the stateside intelligence operations of all branches of the armed forces. Army intelligence has destroyed files on thousands of politically active, law-abiding citizens, and the Army Intelligence Command, which did most of the unauthorized surveillance, has been scheduled for deactivation.

At the same time, the Department of the Army has not been entirely successful in its cut-back efforts. While its top officials have been attempting to sell their new policy against surveillance to reluctant intelligence officers within the United States, military intelligence units have continued to monitor the political activities of Americans living abroad.

I will not take the time here to describe those activities. They are well documented by the testimony of John H. F. Shattuck of the American Civil Liberties Union and by Andrew Hamilton's recent article in the New Republic ("Shut Up, Soldier: The Watch on the Rhine," March 30, 1974, pp. 13-15). Nor will I recount what I believe to be the distinct threats to liberty and privacy posed by the surveillance and the data banks it generated. They are well discussed in the testimony presented by Barry Mahoney on behalf of the Association of the Bar of the City of New York and by this Subcommittee's past reports.

Instead, I would like to focus my attention on what I believe to be the central, and most difficult question, now facing this subcommittee: What are the military's legitimate domestic intelligence needs?

If it were possible to define with precision each and every interest to be protected, and each and every kind of monitoring to be prohibited in the interest of preserving liberty and privacy, it would not be necessary to address this question. A carefully worded prohibition would suffice. Unfortunately, the problem is not that easy. Even the most ardent civil libertarians must concede that there are occasions when the military has a right to information touching on the politics of persons not affiliated with it. An espionage case involving Communists is an obvious example; attempts by an anti-war group to persuade servicemen to desert is another.

A bill which does not draw careful lines around the military's legitimate and illegitimate informational needs directly threatens the efficiency of government. It also endangers the liberty and property of military personnel, who must make judgments about the law's meaning in the course of carrying out their assignments.

This second danger concerns me most at this time. The government will not go to jail if the law is misinterpreted, but some intelligence agent may, through no fault of his own. Thus, if Congress is to accord due process to government employees while protecting political dissenters, it must state with particularity those investigative and record-keeping activities it does not intend to prohibit.

I. THE MILITARY'S DOMESTIC INTELLIGENCE NEEDS

It seems to me that the military's domestic intelligence needs can be broken down into five basic categories: civil disturbance intelligence, counterintelligence, criminal intelligence, security intelligence, and command intelligence. There is nothing watertight about these compartments; operations begun under one often flow very quickly into the others. But the categories are useful for analytical purposes.

Civil Disturbance Intelligence

During the late 1960's there were three basic kinds of civil disturbances that called for a military response. These were ghetto riots, mass demonstrations, and disruptions of military installations or activities. During the early 1960's there was a fourth category involving interference with the enforcement of federal court orders. It is conceivable that the government might sometime be faced with a fifth: para-military or military resistance to lawful authority of the sort now occurring in Northern Ireland or that occurred in the United States during the Civil War.

Each of these kinds of "disturbances" involves a mixture of political activities, some of which may be constitutionally protected and some of which may be criminal and violent, and each poses a different mixture of problems for the military.

Ghetto riots. The heaviest demand placed upon the military in recent years has been for troops to support civilian law enforcement agencies incapable of controlling mass rioting. In such situations the Army has found a clear need for three kinds of intelligence: reconnaissance of the riot (or potential riot) area, advance (or "early warning") intelligence on escalating violence which threatened to outstrip local and state police and military resources, and tactical intelligence on the targets and activities of rioters.

The Army also assumed that it had a need for personality and organizational intelligence for the purpose of identifying individuals and groups that might incite riots or participate in them. This assumption turned out to be in error. None of the riots of the late 1960's was planned or led by identifiable persons or groups, and few were marked by organized criminal activity directed against the authorities. This is the unanimous conclusion of the military and civilian histories of the era and of the intelligence analysts I have interviewed. The conclusion is confirmed by the Kerner Commission report and echos the findings of every major riot commission for the past half century.

Thus, the Army had no need for mug books, potential round-up lists, or large data banks on persons thought to be prone to rioting. Nor, in fact, were any of these records ever used at command headquarters or by units in the streets. What information military commanders needed to know about racial politics in the riot areas they received from civilian politicians and intelligence analysts who relied primarily on the press.

The major justification for civil disturbance intelligence offered by Army intelligence was that detailed information on incidents, organizations, and individuals was needed for planning purposes. Reports on civil rights demonstrations and ghetto altercations were collected on the theory that examination of their nature and frequency might reveal trends. The idea apparently originated with an after-action report prepared by Cyrus R. Vance following the Detroit riot. Vance, who was the President's representative to Detroit, recommended the analysis of incident reports to see if patterns could be discerned which might predict riots. He also proposed the development of a "normal incident curve" of ghetto crime against which apparent increases in violence could be compared. The need for such information, as he saw it, was at the local level, and the proper people to collect it, he believed, were the police. Unfortunately, his report to the President did not emphasize that point, and Army intelligence responded with a nation-wide incident reporting system. The reports were funneled, among other places, to the Continental Army Command at Fort Monroe, Virginia, where intelligence analysts attempted to construct a weekly "Dow Jones" average of violence in the nation as a whole. The exercise was a waste of time, of course, because a national average could not help the Army to predict where and when the next major riot