Home Page

Oral History and its Digital Preservation for Libraries:

Brown University Project on

Nikita Khrushchevs Dictations, 1966-1971

Irina Lynden and Allynn Wilkinson

Brown University, Providence, R.I., U.S.A.

ABSTRACT: Digitization of sound materials is another step in library efforts to use new methods for collection preservation. Digital sound documents have unique features which enable them to be used for both preservation and access.

In 2000-2002 Brown University worked on a project to organize and digitize the oral memoirs of Nikita Khrushchev, which he dictated from 1966 to 1971. The last dictation on the Soviet Intelligencia and the Thaw was finished five days before Nikita Sergeyevich died.

The history of Khrushchevs Oral Memoirs is quite intricate. Under the threat of confiscation, his son Sergei made two copies of the original tapes. Each set contained thirty-nine tapes with over 300 hours of dictations, recorded in four track mono at the slowest speed. One copy was hidden in Russia and the other was smuggled to the United States in 1960s. Until 2000 this second copy had been stored at the Columbia University, Department of Oral History. Based on transcripts from these tapes, Strobe Talbott prepared the first edition of the memoirs Khrushchev Remembers, published in English in 1970. In 1999 a complete version of the memoirs, compiled and edited by Sergei Khrushchev, was published in Russia (Vremia. Liudi. Vlast. Moskovskie Novosti, 1999. 4 volumes).

The original tapes were indeed taken by the Communist Party Central Committee, and have never been accessible. The copies of the tapes could not be easily used by historians because, during copying, the original sequence was destroyed. Also, their age and fragility made playing them potentially destructive. The goal of the Brown University Project was to digitize Nikita Khrushchevs Dictations for use and preservation and return them to their original order. This paper gives a comprehensive analysis of the Project, with special attention to the technology and software used, and the search tools created for the Memoirs.


In November 2000 the new Multimedia Lab of Media Services, Brown University, took responsibility for 39, eight-inch, reel-to-reel audiotapes, which held Nikita Khrushchevs oral memoirs. These tapes were not the originals, but copies made by Khrushchevs son Sergei in 1968-1971, and conveyed to the United States for Time/Life publishers. These tapes are a valuable historical document for studying Soviet History, as well as the history of many other countries at the time of Khrushchev. The tapes had been stored for years at the Harrison Institute, Columbia University and went virtually unused because they were fragmented and disorganized.

History of Nikita Khrushchevs Memoirs.

As Sergei Khrushchev wrote in a special chapter of the complete four-volume edition of his fathers memoirs, . . . (Vremia. Liudi. Vlast.) ( / Moscovskie Novosti, 1999)1, conversations about dictating the memoirs began in 1966. That summer a pensioner of the state level and forcefully retired prominent political leader, Nikita Khrushchev, was recovering from an illness. The family tried to distract his attention from sickness to some useful and interesting activities. Others, who had known Khrushchev for years, were also constantly asking him to write his memoirs.

Thus in August 1966 a tape recorder was brought to Petrovo-Dalneye, to the Khrushchevs dacha. In the very beginning there was not even a rough plan of the Memoirs, and, as the first theme, Nikita Sergeyevich chose the Caribbean Missile Crisis of 1962. Lev Petrov, his son in law, took the finished tape home, typed a transcript and edited it for Khrushchevs review. Unfortunately, Khrushchev did not like the final text at all, and his enthusiasm was lost. Dictating was stopped for nearly a year. Meanwhile, the policy of the new Brezhnev administration, especially its idea to revive Stalins glory, troubled Khrushchev more and more, and motivated him to get back to dictating to tell the truth.

In the summer 1967 the work on Memoirs was begun again in earnest. In the beginning, his wife Nina Petrovna tried to type daily dictations and edit them, but very soon it became clear that, to follow Khrushchevs pace, a professional typist was needed. From then on two people worked on the dictations: Leonora Finogenova, as typist, and his son Sergei as editor of the transcripts. A regular schedule of work developed. During the day Nikita dictated and in the evening Sergei would pick up a finished tape. Later, at his apartment, Leonora Finogenova typed the text listening through ear-phones. After editing Sergei returned the tapes and transcripts to Petrovo-Dalneye for Khrushchevs review. Normally, Khrushchev dictated 3-5 hours a day, in the morning and in the afternoon. According to Sergei, his father did not use any sources and impressed his family and friends with his excellent memory. In some cases, when Khrushchev could not remember a name or a fact, he would note it right in the dictations, asking his son to check or correct and add the information in the transcript.

Copying of original tapes.

With the growth of the amount of tapes (kilometers of them!), and pages of transcripts, Khrushchev developed a growing concern about the safety of his Memoirs. He knew his former comrades too well and realized that they would like to control his work, and, if not, confiscate the dictations. Thus Khrushchev and his son devised a plan to copy the original tapes and save them in a separate place. The place was soon found. It was a storage closet in the apartment of Sergeis friend near the metro station Sokol. Time proved Khrushchevs concerns right. The Central Party Committee learned about dictations from informers and, in April of 1968, he was called to the special meeting with Kirilenko, Pelshe and Demichev. They demanded that he cease his dictations and bring the tapes and transcripts to the Central Committee. Khrushchev refused. This meeting depressed Khrushchev deeply, and he almost stopped his work until the end of 1968.

In 1969 Khrushchev returned to his dictation and started to work regularly until his first heart attack in 1970. After recovery, he dictated again. Nikita Sergeyevich was not feeling well, he sdal, and his voice on the last tapes is weak and unstable. The last record was made 5 days before Khrushchev died. This last story about the Soviet Intelligencia was especially difficult for him. It was entitled as I am not the Judge ( / Ya nie sudia). When Sergei, as usual, brought the transcript for review, Khrushchev did not like it. He asked Sergei to destroy the transcript and erase the tapes. Sergei never did, and so the last testament of a very ill Khrushchev, explaining his decisions, and, in some case expressing apologies to people, still exists.

Copying audio tapes in the U.S.S.R. in 1960s was not an easy task. There was a shortage of tapes in retail stores and the Khrushchev family was being watched by KGB. Sergei copied the tapes in secret and haste, with very limited resources. He would copy dictations using smaller tapes at maximum speed on two tracks simultaneously, using his home Grundik machine. Through this process the original records from longer tapes were cut and the original order of dictation on the copied tapes lost.

Moving tapes abroad.

Back in 1968, the call from the Central Committee resulted in the search for a safe place abroad to protect the Memoirs until the time for making them public arose. The story of transferring Khrushchevs Memoirs abroad would make quite a mystery novel or movie, should there be a writer or a film director interested. It reveals many sicknesses of the old regime, and thus is very instructive. As was mentioned before, the history of the Memoirs can be found in Russian in the second volume of . . . (Vremia. Liudi. Vlast.), and in English in Sergei Khrushchevs Khrushchev on Krushchev (Little, Brown and Company, 1990)2. Here it will be mentioned only that the plan succeeded and the first portion of tapes and transcripts was smuggled abroad to a secret place. From then on, every new portion of tapes was copied twice: once for the storage closet in Moscow and once for the foreign site.

Publication of the Memoirs abroad.

Initially, smuggling the tapes from the U.S.S.R. was not done with the goal of publishing the Memoirs abroad. Nikita Khrushchev was too devoted to the Soviet Union and Soviet people. His dream was to publish his Memoirs there and in Russian. But pressure from the Party and KGB was growing. In July 1970 the original tapes and transcripts were confiscated by the KGB while Khrushchev was in the hospital. This made Khrushchev agree to a foreign publication. Interestingly enough, the agreement with the publisher, Little, Brown and Company, was signed by the smuggler of the tapes and not by Khrushchev or his family. Consequently, the former obtained the copyright, which is valid still. Not one of the Khrushchevs ever received royalties for the book. The first volume of the English edition Khrushchev Remembers was prepared by Strobe Talbott, and published in 1970 by Little, Brown and Company3. It was not a complete edition, since Khrushchev was still dictating. That same year the book was translated into Russian and published by Progress in Moscow, but with the note DSP ( / For Restricted Use). The second volume of the English edition, (Khrushchev Remembers: The Last Testament), was published in 1974 after the last tapes arrived in the West4.

Contraversies about the Memoirs.

Even published, Nikita Khrushchevs Memoirs carry controversy. The controversy began in 1970, when Khrushchev was forced to sign a letter stating that a publication announced by Little, Brown and Company was a fabrication (), and that he never sent his Memoirs abroad. Sergei, his son, also wrote and signed a letter stating that he never made copies of the Memoirs. He did this to protect the copies from possible disclosure. In 1974, after Khrushchevs death, when the second volume of the Memoirs was published in the West, officials from the KGB took action. They forced Sergei to sign a letter, which was sent to a famous American publisher, stating that the Last Testament was a fabrication. The controversies continued after the Institute of Marxism-Leninism received transcripts from the Central Committee. Historians started to use these as a primary source in spite of changes and cuts that may have been made in the prints. Because of these controversies it is very important to return Khrushchevs dictations to their original order and make them available for researchers to study the history and resolve disputed issues.

Moving the tapes and transcripts from Columbia to Brown.

After the book was published, Time/Life found a place for the tapes and their Russian transcripts in the Harriman Institute, Department of Oral History, at Columbia University. That Institute has many similar oral documents of important political leaders which are open for public use.

In 1988 transcripts and tapes from Columbia were copied and sent back to Sergei Khrushchev in Moscow upon his request. Sergei started his work on a complete Russian edition of the Memoirs. It was published in 1999, when Sergei Nikitovich was living in Providence, RI and working as a Senior Fellow at the Watson Institute, Brown University. He interested Brown University in the archival project to digitize the tapes and return them to the order of original dictation. Thus, the tapes were brought to Brown in 2000.

Digitizing and organizing the dictations: The Brown University project.

As mentioned previously, Brown University Library received 39, eight-inch, reel-to-reel audiotapes. The audio was recorded in four track mono at the slowest speed (1 7/8 feet per second). This meant that one tape held eight hours of information. The goal of the project was to preserve and enhance the audio by putting it in a digital format and also to recreate (to the best extent possible) the original order of dictation. Since the transcripts of the tapes had been rearranged and published into a four volume work, it was also possible to reference the audio to these indexed books.

First stage of the project, November 2000-February 2001

Since the audio would have to be digitally processed anyway, it was decided to digitize it initially at a faster speed (3 3/4 fps). This allowed digitization of 300 hours of material in only 150 hours. The digital format was then processed to remove noises caused by deterioration and copying. Speed and, in some cases, reverse recording was also corrected. This audio engineering was done by Michael Haumasser of Brown University Library, Media Services. 

A standard reel-to-reel tape deck connected to a Macintosh G4 with an 866 megahertz processor and 512 megabytes of RAM was used in the project. The software package was DigiDesigns Pro Tools LE. Pro Tools was chosen because of its simple interface and the difficulty in accidentally erasing information. Since several people would be working on the project it was important to use a software package with built in redundancies.

The audio was captured onto 80 gigabyte, external, firewire hard drives. Once the audio was pitch shifted to the proper speed and processed it took up most of the free space on three of these drives.

A naming convention for the audio was an important consideration. It was necessary to reflect the original place, track and direction of the audio. The tapes were originally arranged into two parts. Part I consisted of 13 tapes labeled 1 to 12 with a tape 2 and a tape 2a. Part II consisted of 26 tapes numbered sequentially. The first section of the naming convention was either 01 or 02 and referred to the part to which the tape was associated. The second section was the number of the tape. Thus 01-08 was tape 8 from the first section. Leading zeros were used throughout in order to ensure proper automatic sorting.

The last section of the naming convention specified which of the four mono tracks, the tape direction (forward or reverse) and the stereo channel (right or left). On the forward pass the right channel was track one and the left channel was track three. When the tape was turned over the right channel was track two and the left channel was track four. The only possible choices for the final naming section were: FL1 (forward, left, track 1); FR3 (forward, reverse, track 3); RR2 (reverse, right, track 2) and RL4 (reverse, left, track 4). A track named 02-18-FR3 indicates the audio comes from the second part, tape 18 and the right track 3 of the forward tape direction.

With the naming convention in place a small database was begun in Microsoft Access to track the exact, original position of each section of audio. This database began with just five fields; original tape; start time; end time; length and region name. This last field was the name assigned to an individual story at the second stage of the Project.

The second stage of the Project

With all 39 tapes digitized and a naming convention in place, work was begun on the content of the dictations. The digital audio was played to find discrete fragments and separate and save them as regions. To define the fragments not only sound but also its digital image was used. The 350 hours appeared to be cut into approximately 300 pieces, from 1 to 130 minutes each. When the first ten tapes were finished, the fragments (regions) were put together in the sequence of original dictation. The first and last sentences of each fragment were especially important and were taken down verbatim and written on index cards. Sergei Khrushchev would always rewind the tape slightly when he started a new reel so as not to lose information. Thus, if the phrases matched, that meant the fragments should be placed one after another. This worked well when the repetitions did not exceed a sentence or two but in some cases the overlap might be 20 minutes. In such cases it was difficult to match fragments.

Transcripts of the first 10 tapes were very helpful tools. They were made from the original dictations and transported abroad. Therefore the order of fragments could be checked against them. Tapes from the second part were much more difficult to work with. The original transcripts of the second part had been cut up, and the only Ariadnas thread was the 4 volumes of the complete memoirs, published in 1999. These could only be used with restrictions, however. Working with the transcripts, Strobe Talbott, and, later, Sergei Khrushcev, would cut and paste the pages of stories into chronological order. The goal of the Brown University Project was to restore the order of original dictation which was not chronological. With the sound files notes taken while listening to the audio were the only way to restore the order of the stories.

In the process of the work sound records were compared with two documents: the transcripts of the memoirs for the tapes in the first part, and the four-volume edition for those in the second part. Sound regions were also cross-referenced with pages in the both documents. Later this proved to be very useful for search purposes.

Each identified region was saved into the database. The database proved to be a very useful tool throughout the project. Regions could be easily added and rearranged. Every second on the tapes could be accounted for by resorting the database and overlapping regions could be detected. Each record for the regions contained the following fields: original place, start time, end time, and length. Later fields were added to reflect the general topic and restored order. Once volume, chapter and page number information were added the database became the basis of the projects search engine. By the time all of the audio was examined and referenced the database had grown to 34 fields with both Latin and Cyrillic characters and 453 entries.

In the process of listening and arranging the audio, sections were identified with particular audio problems. These regions were then processed separately to make the audio as clear as possible. The sound quality ranged from very good to quite poor. Remarkably however, out of all 250 hours of unique material, only 30 minutes could not be understood. Many sections had a hum caused by the original duplication. The range of this hum made it problematic to remove without losing Khrushchevs voice as well. Notch filters created in Pro Tools were used to selectively remove ranges containing the worst of the hum. This sometimes causes Khrushchevs voice to lose its deeper tones but at least his words can be better understood. In many cases it was necessary to leave in some background noise to preserve the audibility of the material.

As the sound was being processed and prepared to be burned on CDs, the referencing work and database were nearly completed. It was time to create the end-user access points.

Third stage of the project

It was decided to make the search feature based in HTML since it is plain text and forward compatible. The CDs were audio CDs (not data CDs) because it is thought to be a more archival format. In addition, however, the audio data was also processed into the .aiff format (audio interchange file format) to make it more universally readable for the finished, digital product. To make the search engine forward compatible as well, it had to be created in a non-compiled programming language that did not require a live network connection. The search feature would also be included on a data CD with the final, archival output of two sets of 261 audio CDs. A javascript search was decided on because it uses a simple, text based, amorphous database that can be examined without any proprietary programs.

HTML pages that would serve as a table of contents for each CD were created using Macromedia DreamWeaver Ultra Dev. This program is designed for an online connection to a database via active server pages (.asp) and several other, similar formats. Unfortunately, UltraDev does not allow all entries of a database to be output as individual web pages. In order to get the 261 individual pages, one long page containing all entries was output and each entry was extracted manually. Another drawback of UltraDev was that it did not support the output of Cyrillic characters and these had to be pasted in from a list of all the chapter titles.

Once the 261 pages had been created the shareware program, Atrise Everyfind, was used to spider the pages and create the javascript search. Everyfind was selected because it was written in Russia by Andew Revvo and it supports Cyrillic searching using the Win-1251 keyboard layout. Other HTML pages were added for explanatory material, a list of topics and technical help. At this point the two archival sets of audio CDs were completed. The final phase of the project was to create a more portable form of access that could be made available to other archives.

Final stage of the project
External hard drives had been used throughout the project because they were compact, secure and relatively inexpensive. Although hard drive storage is not ideal due to its fragility, as a means of distribution, it was acceptable. All of the audio files (comprising over 45 gigabytes) as well the HTML pages and search engine fit on one 60 gigabyte drive costing less than $250. The audio is embedded into the HTML page so that it can be played within the page itself. There is no need to locate and insert the correct CD. Everything is self contained within the drive and accessible from the HTML pages.

The drives can connect by USB, USB 2 or FireWire and a drive was found that allows for all of those types of connections making it very compatible. In practice, the standard USB interface is a little slow to load the audio and it is recommend to upgrade to USB 2 or FireWire by inserting an inexpensive card in the host machine. It is hoped that these drives will give good service for several years after which the audio can be ported over to the newest, best storage medium. Constant migration to new media is essential to keep digital projects accessible into the future.

Copyright © 2003. International Library Information and Analytical Center.
All rights reserved.