Chapter 27
recordings are in the public domain. For more information, nor to volunteer, please visit LibriVox.org. Recording by Marianne. Final Report of the Senate Select Committee on Presidential Campaign Activities, Volume 3. Chapter 10. The Select Committee's Use of Computer Technology. Roman 1. Introduction and Overview. To accomplish its investigative task, the staff of the select committee had to analyze a considerable amount of information available
concerning the Watergate incident and related activities. Included among this data was the transcript of the Watergate trial held in January 1973 and the various newspaper articles published since the break-in on June 17, 1972. In addition, many interviews, investigative reports, and much documentary evidence had to be obtained for the committee to conduct a thorough investigation. In early March, the committee's information was processed by hand.
An elaborate file card indexing technique was employed patterned after one used by military intelligence agencies. Trained researchers would read testimony and documentary evidence and abstract the facts contained therein. References to each person discussed, along with the date of the activities testified about and the location of the activities, were included with the paragraph or abstract. This data, after being taped on cassette tape records, was transcribed onto 3x5 cards. One card was made for each person, date, and location discussed and placed in three files.
Thus, staff members could utilize any of the files. By the middle of March, over 10,000 file cards had been produced, yet the backlog of information available had scarcely been processed. Other problems developed. Since there was only one complete file card system, only one person could employ the system at a time. The size of the file made reproduction of the system for each individual's use impracticable.
To prepare reports about individuals, dates, or general events, the card file had to be totally reviewed by hand, consuming valuable time, and given the time frame within which the committee had to operate, it was believed that this procedure would seriously affect the total investigation. Therefore, in late March, the select committee decided to explore the possibility of using
automation to facilitate organization and analysis of the information to be collected. Samuel Dash and members of the Select Committee staff met with Paul Reimers, Coordinator of Information Systems, and other persons from the Library of Congress Information Systems Office. The requirements of the system were established. A generalized file system was needed with full text processing and all Boolean search capability.
The Select Committee had to be able to formulate its search request with both positive and negative qualifiers. An updating capability had to be a major feature of the system since new information would be continually obtained by the select committee. Lastly, but probably most important, the select committee's security procedures precluded use of an online system. Therefore, the proposed system would have to allow for batch application processing. After a few meetings in early April, it became apparent that a computer system did exist that was well-suited to fulfill the needs of the select committee.
It was BIBSYS. BIBSIS. The Bibliographic System, or BIBSIS, is part of the Library of Congress Information Network designed to enable the user to establish and maintain a machine-readable information file. Using a modified machine-readable catalog, Mark II, internal processing format, it can produce a variety of reports from the database. For example, some of those envisioned during the meetings were
cross-reference reports on individuals mentioned, special reports on specific individuals mentioned, chronological date of event reports, and KWIC, keyword in context, indexes, all of which are discussed below. Tags were used to identify fields of data at the time of data entry, but without compromising the use of the data in later processing. The technique required a minimum of machine identification data to be input in conjunction with the textual information.
The subsystems of BIBSIS, developed for the select committee's use, were the file maintenance cycle, the retrieval system, the field inversion cycle, and the report cycle. The Library of Congress made available for the select committees use the following hardware. One IBM 370 Mod 145, 512k, three 3330 disk packs, and one 2314 disk pack, two 2400 tape drivers using nine track tapes, 1600 BPI, two 1403 line printers, one 2501 card reader, and a Digidata converter for MTST input.
The software consisted of OS VS1 operating system, sort and utilities, ANS COBOL, compiler version 2, 12 application programs written in ANS COBOL, and two ALC application programs. With the approval of the select committee, the computer operation was inaugurated. Within two weeks, sample programs had been adapted for the committee's use and tested. Sample records were processed in early May, approximately one week before the public hearings began.
The tests were successful. On May 15, 1973, using a borrowed staff of key-to-tape operators, the first actual records were input. Roman II. Processing the Data. The input procedures were simple yet comprehensive. Each bit of testimony, either public or confidential, was read by a staff of trained researchers
who abstracted out the essential facts. These researchers dictated the source of the information or the witness, the date the testimony was given, and the geographic location at which the testimony was taken. The title of the document and the page number from which the facts were abstracted was also included.
Following this, a complete summary of the information was dictated onto the tape. To ensure a chronological sorting capability, researchers were instructed to produce one abstract for testimony concerning activities on a certain date. Through this method, as will be discussed later, the data could be sorted in a day-by-day chronological order.
The remaining fields in the records contained references to persons quoted by the witness and the date of the quote, persons acting in the testimony along with the date and location of the activity. Lastly, each record was categorized into 44 general subject topics, delineating the general areas under investigation.
Once this was dictated by the researchers, the cassette tapes were given to a secretarial staff which transcribed the information onto input sheets. These sheets were returned to the computer research staff for a series of editorial changes, both substantive and technical. Each sheet was reviewed for accuracy three times before it was ready for computer input.
the editorial process was complete, the records were given to two key-to-tape machine operators, who transferred the information onto magnetic tape. Once on tape, the information was printed out from the computer tapes or dumped to ensure that no other errors existed. These preliminary printouts or dumps were then edited for substantive and technical errors. The tape, with corrections, was returned to the key-to-tape machine operators who corrected the errors on the tape. The tapes
were then ready for the actual input into the machine. This initial process was later altered with the advent of Magnetic Tape Selectric Typewriters, MTSTs. The use of these machines enabled the computer staff to eliminate two steps in the input process. Since a skilled secretary could format the data by adding the appropriate computer codes directly from a cassette tape, there was no longer any need to type up input sheets. Secondly, the MTSD machine simultaneously
stored the data on magnetic tape cartridge and provided a readable copy of the data on tape. Therefore, the staff no longer needed to dump or print out the tapes created. The computer used by the select committee was an IBM 370-145 located at the Library of Congress Annex. Elaborate security measures were implemented designed to ensure that no unauthorized individual organization could obtain information from the computer database. It was decided that no data
would be stored at the library itself. Instead, the master tapes containing all the data collected were kept secure within the select committee's files. When the actual processing took place on Monday, Wednesday, and Friday from midnight to 6 a.m., the tapes were delivered to the computer room by a select committee staff courier along with a police escort. Once in the machine room, the courier was not permitted to leave the tapes alone, and only he, a systems analyst,
and an operator were permitted in the room during the processing of the data. When select committee data was processed, no non-committee applications were run simultaneously. A completely dedicated machine was used to ensure no spillover of data from or to the select committee's computer file. Upon completion of the processing, the courier, under police guard, returned to the select committee's offices with the tapes and any readable data. Roman 3. Format of the Data
Each bit of information or record consisted of eight fields of data. The first field contained a number uniquely identifying the record. The second field contained four subfields, the name of the person or organization that was the source of the abstract, the date the testimony or data was given, a computerized transfiguration of that same date listing the year, month, and day in a six-digit number,
and the city, state, or location at which the information was taken. The next field contained the title of the document from which the abstract was taken. The microfilm number, indicating on which roll and frame the document abstracted from was filmed, was included in this field. After the third field, the abstract was included.
These were summaries of the factual testimony given. During the public hearings, only facts relating to the events under investigation were abstracted. Discussions of legal principles, select committee procedure, and anecdotes were not abstracted. In most cases, the abstracts were based upon testimony given concerning a certain date or event. They were not verbatim accounts of the actual testimony.
After the abstract, there existed in most every record a field which contained comments by the abstractor. These comments ranged from identifying the date of the transaction to pointing out that the abstract was inconsistent with previously given testimony or identifying a person involved in the transaction who was not otherwise identified during the specific testimony abstracted. After the comment, the abstractor who inserted the comment included his or her name and the date the comment was inserted into the record. The sixth field, which followed, contained the names of any individuals quoted in the abstract, along with the date the quote was made.
As mentioned before, each record contained a cross-reference list of all persons or organizations acting in the abstract. This was included in the seventh field. There were four subfields to this field. The first contained the name of the person or organization.
The second contained the date that person acted. The third listed the six-digit transfiguration of the computer date. And the fourth was the location that the person or organization acted. Each record contained a variable number of these fields based mainly on the nature of the abstract.
After the cross-reference field, each record contained at least one subject-topic field. These fields contained three-digit number designations of specific subject topics and were designed to fit the various areas under investigation. This record format permitted the select committee to analyze the stored data in the following methods.
Name discussed, date of event, subject matter, persons quoted, witness, and source document. Part of the record input into the machine was a reference to each person or organization about which the abstract dealt. In addition, the date this person or organization acted was included. Therefore, requests could be and were structured to instruct the computer to retrieve all records with an individual's name in this reference field, and then to sort all those
records in chronological order based upon the date associated with that name. The result, a chronological report of every fact about which testimony had been given concerning any individual or organization mentioned. Since reconstructing a sequence of events was vital to the Select Committee's investigation, this was also a major feature of the computer's capability. Through proper programming, a complete chronological listing of events could be
reconstructed based upon all the record's input into the database. This enabled the staff to retrieve testimony from a multitude of witnesses concerning a particular date or series of dates, and then analyze, given the nature or the source of the information, that is, whether it was from a newspaper article or given under oath, what actually happened. Each record, as mentioned before, was categorized into various subject areas. If the testimony dealt with the Watergate
phase of the investigation or campaign financing phase, it was appropriately coded as such. These subject codes enabled the staff of the select committee to retrieve and analyze data concerning general events within the overall investigation. They could also be added as positive or negative qualifiers to other retrieval requests to limit the topical scope of a computer run. For example, a request could be made for all testimony about a certain individual, but only if
that information dealt with the general Watergate phase of the investigation or selected areas within the other phases of the investigation. Additionally, the format of the records permitted the staff to obtain information from any document where a particular person was the source or where a specific individual was quoted. This request could be limited so that information was retrieved from certain documents, either as a group, to wit, only executive sessions or interviews or public
statements under oath or individually, such as a particular person's interview, an executive session, or public testimony before the select committee. Thus, once testimony was placed on computer tapes, the problems previously experienced with the card file system were alleviated. The references could be rearranged on the machine's disk packs and then printed out in a desired format. Since it was done through automation rather than manual review of a card
file, the time necessary to complete the request was minimal. Roman IV. Investigative use of the computer. Since time constraints precluded training each staff member computer request methodology, the select committee created a specially trained staff to act as a liaison between the investigative staff and the computer system's analyst. Requests from the investigative staff were thus translated into the necessary computer terminology by the computer staff. When a particular witness or
series of witnesses was to be interrogated, a request was made to retrieve from the computer all the information the select committee had collected. For security reasons, no staff member could obtain a computer printout without first receiving written authorization from either the chief counsel or chief minority counsel. This ensured that the heads of the staff knew who was getting what information. During the public hearings, printouts on the upcoming witnesses
were run regularly. Given the inputting system employed, it was possible to have one week's testimony in the computer ready for retrieval and analysis before the beginning of hearings the following week. Thus, to maintain a complete file, new testimony was continually being integrated with data obtained from prior witnesses. Printouts of information on each new witness were distributed to the Chief Counsel and the Chief Minority Counsel. The printouts listed all information
in the computer file about the witness that had been stored in the machine up to that date. Once provided, the printouts served as an additional source upon which interrogation were based. The same procedure was followed for confidential interviews and testimony taken in executive session. Printouts were produced, integrated into the staff member's personal notes, and the interrogation undertaken. With the computer's facility to not forget, and a properly
entered name, date, or subject matter, the interrogator had at his fingertips all information about the person to be interrogated. Since data was input into the machine at least three times per week, the printouts produced were very much up to date. A source document title along with the pages from which the facts were abstracted was included with each abstract. This enabled the user, if necessary, to go to the original document to obtain the verbatim account.
In addition, the computer was employed to sort and analyze other data. The select committee subpoenaed telephone records from a number of individuals. These records were presented to the staff of the committee in the form of the monthly bill statements received by the person under subpoena from the telephone company. A separate computer file was established listing the telephone number from which the call was made, the number to which the call was placed,
the date of the phone call, the locations, city, state, of both phone numbers, the time of day of the call, designated in terms of a 2400 hour clock, 1630, 0843, etc., as well as the duration of the call. Collect calls and those made from phones other than the business or home phone of the caller, yet nonetheless billed to either the home or business phone of the caller, were also included in the file. Approximately 12,000 telephone calls were input into this file. Printouts were prepared,
upon request, that retrieved and sorted phone calls from one number to another, all calls to one specific number or series of numbers owned by an individual, phone calls made during a specific time period, etc. Thus, the staff was able to review thousands of telephone calls quickly and efficiently. Through the use of a master list of telephone numbers and corresponding owners, the staff could isolate contacts between particular persons either generally or on specific dates
from data that was accurately collected by an independent entity, the telephone company. Roman 5. Cooperation with other investigative bodies. Special Prosecutor Archibald Cox, shortly after his appointment on May 18, 1973, held a series of meetings with the select Committee's chief counsel. Mr. Cox realized that his office, too, could alleviate many of the information retrieval problems facing it through the use of a computer. After a few meetings between
the members of his staff and the Select Committee computer staff, Special Prosecutor Cox requested the Select Committee provide his office with all public information from its computer tapes. In addition, since the system was a totally new one, it was agreed that the Select Committee would train personnel from his staff to abstract testimony and input it into the computer. For three weeks, personnel from the Special Prosecutor's Office were trained by the computer
staff of the Select Committee. From September 14, 1973, to the end of January 1974, the Select Committee made available to the Special Prosecutor over 10,000 records containing public information. By doing so, it saved the Special Prosecutor's Office the almost 20,000 man-hours it had taken the Select Committee to computerize the information. On March 5, 1974, Senator Irvin, responding for the entire committee to a request by then-Special Prosecutor Jaworski,
agreed to turn over all materials in the select committee's files that would aid him in his investigation. This agreement included all information on computer tapes, over 25,000 records. In January 1974, the House Judiciary Committee, investigating the possible impeachment of the president was, just as the Special Prosecutor, faced with the problem of devising an information retrieval system. Mr. John Doerr, special counsel to the Judiciary Committee,
requested information from the Select Committee's computerized files. On February 7, 1974, the Select Committee met and agreed to furnish the House Judiciary Committee with all information on its computer tapes. One week later, duplicates of all the Select Committee's computer tapes were turned over to the House Judiciary Committee, a transfer of information saving them thousands of man hours. After receiving many requests from various state and local bar associations for
information concerning possible unethical conduct on the part of attorneys in the Watergate incident, the committee decided that it would be compatible with its investigation to supply public information in printout form. Using as a conduit for requests the American Bar Association's Center for Professional Discipline, the committee made available to those bar associations public information in printout form on individuals identified by the associations as being under
investigation. Requests for public information printouts were also made to the committee by attorneys representing individuals indicted for their participation in the Watergate incident. Knowing the tremendous value of computerized printouts and that these trials were to determine ultimate individual criminal liability, the committee felt, out of fairness, that these printouts should be made available to the defendants. On May 9, 1974, the committee in
executive session unanimously voted to make available to these individuals, at their expense, printouts of public information and any confidential information that the person making the request had given to the select committee. Roman 6. Microfilm Procedure All documentary evidence obtained by the committee and maintained in the security area
was microfilm to allow for greater accessibility of this material to the staff. Because of security procedures, all filming and processing of the film was done on the premises of the committee. Two microfilm viewers were installed to allow for immediate viewing and analysis by any staff member. The computer record format had been adopted to provide for the microfilm number, indicating the roll and frame number where the document could be found.
It was included in all printouts. Roman 7, Disposition. Pursuant to Senate Resolution 369, the Select Committee, at the termination of its activities, transferred its records, including computer tapes, to the Library of Congress. The Senate Committee on Rules and administration was empowered to exercise exclusive control of and access to the use of such records. In so doing, the Senate Rules Committee was directed to make parts of such records or certified
copies of the same available to courts, the special prosecutor, congressional committees, and subcommittees, federal departments or agencies, or individuals satisfying the Senate Committee on Rules and Administration of their legitimate need for parts of such records. Thus, the information retrieval system that was responsible for the efficient and effective way in which the Select Committee conducted its investigative and reporting functions,
was made available to the above mentioned for future application. Moreover, the Senate Committee on Foreign Relations became the first standing committee of Congress to utilize the system in almost the same form used by the Select Committee. It is a fitting commendation for the system that was conceived so rapidly, yet worked so well, that its use be perpetuated by the Congress of the United States. End of Section 27.
