Chapter 2: central reference services
In stacks of paper mounting to the sky
For wanted needles we searched low and high.
For wanted needles we searched low and high.
NOTE: The text covering material on the Central Intelligence Agency, Pilgrim in the CIA, was cleared for publication by the Publications Review Board of that Agency pursuant to CIA Public Affairs Regulation HR 6-2: Non-official Publications and Oral Presentations by Employees and Former Employees (Revised 7 August 1984). Such clearance for security reasons does not, of course, imply any official Agency endorsement of this manuscript.
PAB
Fall 1986
PAB
Fall 1986
Thus in August 1957 I became assistant director of CIA in charge of its central reference services.
It was one of CIA's largest offices. To us came much of the traffic from other parts of government in response to CIA's indicated general and special needs. We in turn read and sent on to CIA components the materials they had expressed an interest in receiving. We also decided what to keep for future reference, thus serving as CIA's institutional memory. This involved a decision of how to store the material so as to be able to retrieve it on demand.
All manner of security classifications were handled, from unclassified through restricted, confidential, secret and top- secret, each requiring its own degree of safeguards. The format of the material varied widely. There were books, magazines, newspapers. These were handled by our library in a manner consistent with normal library practices. There were aide-memoires, telegrams, cables, film, photographs, maps, charts, reports from CIA overseas stations and from foreign service officers, military, naval and air attaches in State's overseas embassies. Multiple copies arrived by the thousands.
Physically, we were located in the old skating rink (Riverside Stadium), in parts of tempo M, Q and #13, more or less in close proximity to each other.
Heretofore, I had been involved in one of CIA's main functions, the production of national intelligence; I was now to be involved in another of its main functions, performing services of common concern. In this category things were done less because they could not be done without collaboration, but more because they could be performed better in one place and thus avoid duplication of effort. The leading such services were biographic intelligence, industrial intelligence and photography (both still shots and film) and collecting information of intelligence interest from departments of government other than State and Defense.
In producing estimates the main problems had to do with what to say in a timely way that would be unmistakenly understood by those who should read it. In central reference we had to identify what was needed by whom, how much time to spend on it at intake to assure recovery from store when needed versus spending little time on an item that might never be called for, but, if called for, would require manpower at output.
Some of our reference tools we produced ourselves; for example, intelligence subject and area codes. Some we contracted out once volume got beyond us.
We were early to adapt machines to help us cope with our problems. Initially the IBM accounting machines were used, punched cards for indexes and aperture cards (with eight 1" frames of film per card) our micro document file.
Then when we got a request for information on a particular subject, area, source, classification, date, and document number, the search by IBM machines through the index yielded the desired references. Having these, we could then retrieve the film strips and from these make enlargements (hard copy) for our customer. Thus using millions of index cards we controlled millions of reports. From the EAM equipment we moved on to conversion to computers. Eventually a new office of computer services was created, centralizing in the Agency these and other computer-based systems. Today, I understand, much is available to the analyst directly through desk top terminals.
My relationship with the Community's Intelligence Advisory Committee (later the U.S. Intelligence Board) was now as chairman of one of its subcommittees, the Committee on Documentation, whose task it was to oversee the development of information processing systems in the Community in an effort to achieve compatibility between systems. I served as chairman for ten years.
Just as career changes came to its employees over the years, so did time bring physical changes in CIA's surroundings. The Eisenhower administration backed Allen Dulles' plan for new housing for his Agency. He wanted a campus environment for his people – a grouping of buildings in a sylvan setting, removed from the bustle of Washington, yet close enough to facilitate our work with the White House, State, and the Pentagon. Finally, a portion of the extensive holdings of the Bureau of Public Roads at Langley was made available. The firm of Harrison and Abramowitz, architects of New York's Rockefeller Center, did the design.
But interminable delays in getting approval by endless bureaucracies, including the Congress, and rising costs necessitated an abandonment of the campus concept. Instead, we got a single large building. In addition a power plant building and, later, a building for our printing services were built close at hand.
Access to the site required an extension of the George Washington Memorial Parkway (which later, once the beltway was built, was further extended to connect). The setting itself was, and remains, ideal. The severe concrete structure is relieved by pebble-encrusted window frames and three inner courts. A free-flowing first floor base supports an additional six stories, the opening of the U-shape holds hostage a single-storied cafeteria whose generous panes marry the interior and out-of-doors.
President Eisenhower laid the cornerstone for our new building in November 1959. At the ceremony, I found myself sitting next to a bubbly matron who asked me what was going on. I told her, and she was much surprised, for she had merely followed the traffic in, assuming that where so many were gathering something interesting must have been planned!
Major Agency components were moved in as construction progressed. My office was among the early units moved in during 1961, six years after passage by Congress of the bill authorizing the building's construction. We worked out a weekend move following much planning and preparatory work and so could boast providing uninterrupted service during regular working hours.
But Agency expansion was such that even with the completion of the new building thousands of employees remained in other housing spread across the Washington area.
In the interest of economy, the interior of the new building was uniformly painted battleship gray. There was one exception. I had succeeded in getting Mr. Dulles to approve special funds for the library. We made it a showpiece. The librarian, Alexander Toth, and I took the train to New York to select contemporary furniture, which we did after having spent the previous evening dining at Luchow's and seeing "My Fair Lady." A very successful trip.
We were engaged in many interesting things. One, for example, was the attempt to use electronic means to translate a foreign language into English, referred to as mechanical translation. Russian was the source language most in need of being translated. Scientific material was abundant, and Russian linguists were limited. CIA had a project, as did the Air Force, Navy, Army, National Science Foundation, Bureau of Standards, to mention only a few. Many universities, including Harvard, MIT, Georgetown, were involved as were companies like IBM and ITEK. Over time, government expenditures mounted. My interagency committee monitored the progress made and sought to restore sanity to this field of scientific endeavor insofar as government funds were involved. We achieved this by getting the National Academy of Science to conduct an independent study and to provide guidelines for the expenditure of such funds as we might wish to devote to MT. This the Academy did in its report, "Language and Machines: Computers in Translation and Linguistics," publication 1416, (1966). There is still no MT system today. There are many ways, however, in which machines greatly facilitate the work of the translator – rapid search of merged dictionaries for the correct equivalent to name only one.
A broad project undertaken was the study of community information processing systems. This was aimed at ascertaining the flow of intelligence information from its initial source through various processing agencies on through production offices, and its ultimate dissemination in some form to users. The idea was to learn what was going on and thereby have a rational basis for improving the effectiveness of our operations. We learned a great deal, but actual results were less than hoped for two basic reasons: No man can serve two masters; and, it is easier to build a better mousetrap if you go it alone. In short, the intelligence arm of a service, say Navy, is an integral part of the Naval establishment and must be in tune with its methods of operation, which means that there are limits to which it can alter its procedures in order to cooperate with another agency. Moreover, some managements believe that if you build the best system, others will follow you and thus, over time, differences between systems will narrow. Progress made eventually owed much to advances in technology making conversion and transfer of data easier and less expensive.
It was one of CIA's largest offices. To us came much of the traffic from other parts of government in response to CIA's indicated general and special needs. We in turn read and sent on to CIA components the materials they had expressed an interest in receiving. We also decided what to keep for future reference, thus serving as CIA's institutional memory. This involved a decision of how to store the material so as to be able to retrieve it on demand.
All manner of security classifications were handled, from unclassified through restricted, confidential, secret and top- secret, each requiring its own degree of safeguards. The format of the material varied widely. There were books, magazines, newspapers. These were handled by our library in a manner consistent with normal library practices. There were aide-memoires, telegrams, cables, film, photographs, maps, charts, reports from CIA overseas stations and from foreign service officers, military, naval and air attaches in State's overseas embassies. Multiple copies arrived by the thousands.
Physically, we were located in the old skating rink (Riverside Stadium), in parts of tempo M, Q and #13, more or less in close proximity to each other.
Heretofore, I had been involved in one of CIA's main functions, the production of national intelligence; I was now to be involved in another of its main functions, performing services of common concern. In this category things were done less because they could not be done without collaboration, but more because they could be performed better in one place and thus avoid duplication of effort. The leading such services were biographic intelligence, industrial intelligence and photography (both still shots and film) and collecting information of intelligence interest from departments of government other than State and Defense.
In producing estimates the main problems had to do with what to say in a timely way that would be unmistakenly understood by those who should read it. In central reference we had to identify what was needed by whom, how much time to spend on it at intake to assure recovery from store when needed versus spending little time on an item that might never be called for, but, if called for, would require manpower at output.
Some of our reference tools we produced ourselves; for example, intelligence subject and area codes. Some we contracted out once volume got beyond us.
We were early to adapt machines to help us cope with our problems. Initially the IBM accounting machines were used, punched cards for indexes and aperture cards (with eight 1" frames of film per card) our micro document file.
Then when we got a request for information on a particular subject, area, source, classification, date, and document number, the search by IBM machines through the index yielded the desired references. Having these, we could then retrieve the film strips and from these make enlargements (hard copy) for our customer. Thus using millions of index cards we controlled millions of reports. From the EAM equipment we moved on to conversion to computers. Eventually a new office of computer services was created, centralizing in the Agency these and other computer-based systems. Today, I understand, much is available to the analyst directly through desk top terminals.
My relationship with the Community's Intelligence Advisory Committee (later the U.S. Intelligence Board) was now as chairman of one of its subcommittees, the Committee on Documentation, whose task it was to oversee the development of information processing systems in the Community in an effort to achieve compatibility between systems. I served as chairman for ten years.
Just as career changes came to its employees over the years, so did time bring physical changes in CIA's surroundings. The Eisenhower administration backed Allen Dulles' plan for new housing for his Agency. He wanted a campus environment for his people – a grouping of buildings in a sylvan setting, removed from the bustle of Washington, yet close enough to facilitate our work with the White House, State, and the Pentagon. Finally, a portion of the extensive holdings of the Bureau of Public Roads at Langley was made available. The firm of Harrison and Abramowitz, architects of New York's Rockefeller Center, did the design.
But interminable delays in getting approval by endless bureaucracies, including the Congress, and rising costs necessitated an abandonment of the campus concept. Instead, we got a single large building. In addition a power plant building and, later, a building for our printing services were built close at hand.
Access to the site required an extension of the George Washington Memorial Parkway (which later, once the beltway was built, was further extended to connect). The setting itself was, and remains, ideal. The severe concrete structure is relieved by pebble-encrusted window frames and three inner courts. A free-flowing first floor base supports an additional six stories, the opening of the U-shape holds hostage a single-storied cafeteria whose generous panes marry the interior and out-of-doors.
President Eisenhower laid the cornerstone for our new building in November 1959. At the ceremony, I found myself sitting next to a bubbly matron who asked me what was going on. I told her, and she was much surprised, for she had merely followed the traffic in, assuming that where so many were gathering something interesting must have been planned!
Major Agency components were moved in as construction progressed. My office was among the early units moved in during 1961, six years after passage by Congress of the bill authorizing the building's construction. We worked out a weekend move following much planning and preparatory work and so could boast providing uninterrupted service during regular working hours.
But Agency expansion was such that even with the completion of the new building thousands of employees remained in other housing spread across the Washington area.
In the interest of economy, the interior of the new building was uniformly painted battleship gray. There was one exception. I had succeeded in getting Mr. Dulles to approve special funds for the library. We made it a showpiece. The librarian, Alexander Toth, and I took the train to New York to select contemporary furniture, which we did after having spent the previous evening dining at Luchow's and seeing "My Fair Lady." A very successful trip.
We were engaged in many interesting things. One, for example, was the attempt to use electronic means to translate a foreign language into English, referred to as mechanical translation. Russian was the source language most in need of being translated. Scientific material was abundant, and Russian linguists were limited. CIA had a project, as did the Air Force, Navy, Army, National Science Foundation, Bureau of Standards, to mention only a few. Many universities, including Harvard, MIT, Georgetown, were involved as were companies like IBM and ITEK. Over time, government expenditures mounted. My interagency committee monitored the progress made and sought to restore sanity to this field of scientific endeavor insofar as government funds were involved. We achieved this by getting the National Academy of Science to conduct an independent study and to provide guidelines for the expenditure of such funds as we might wish to devote to MT. This the Academy did in its report, "Language and Machines: Computers in Translation and Linguistics," publication 1416, (1966). There is still no MT system today. There are many ways, however, in which machines greatly facilitate the work of the translator – rapid search of merged dictionaries for the correct equivalent to name only one.
A broad project undertaken was the study of community information processing systems. This was aimed at ascertaining the flow of intelligence information from its initial source through various processing agencies on through production offices, and its ultimate dissemination in some form to users. The idea was to learn what was going on and thereby have a rational basis for improving the effectiveness of our operations. We learned a great deal, but actual results were less than hoped for two basic reasons: No man can serve two masters; and, it is easier to build a better mousetrap if you go it alone. In short, the intelligence arm of a service, say Navy, is an integral part of the Naval establishment and must be in tune with its methods of operation, which means that there are limits to which it can alter its procedures in order to cooperate with another agency. Moreover, some managements believe that if you build the best system, others will follow you and thus, over time, differences between systems will narrow. Progress made eventually owed much to advances in technology making conversion and transfer of data easier and less expensive.
All materials on this website are © copyrighted, with all rights reserved. Nothing may be reproduced without written permission from Julia J. Borel. See copyright page for more information.
