Cornelis H.A. Koster
Remembering Andrei Petrovich Ershov 1
Andrei Ershov had been a member of Working Group 2.1 on ALGOL of the International Federation for Information Processing since 1966 2. I joined the same Working Group in July 1967 3. At that time, I knew him only from his contributions to the development of the new Algorithmic Language which was later called ALGOL 68. We were both involved in the titanic attempt of the computer scientists of that time to develop a scientifically well-founded universal programming language. Its predecessor ALGOL 60 had shown to the world that such an undertaking was possible, but it had been mostly ignored or even rejected by the major commercial companies. I will try to share with you my recollections of Andrei Ershov against the background of those times.
First contact
The ALGOL 68 project was an international effort at standardization, so international that it even crossed the barriers of the Iron Curtain which for a long time divided the world. The project was a coup by the research community to transfuse the life of all programmers with the fruits of research, to provide the best concepts and notations to the workers toiling in the darkness of the commercial world. It ended in a split across the whole programming world, between those people inspired and formed by ALGOL 68, convinced of its clarity and simplicity, and those who hated it, convinced of its obscurity and complexity. This split started in the bosom of the Working Group itself.
At its meeting in Zandvoort in May 1967, IFIP WG2.1 had charged Aad van Wijngaarden, my unforgettable boss, to write a proposal for a successor to ALGOL 60. He developed a new two-level grammar formalism and applied it relentlessly to formalize the syntax of the new language. He orthogonalized the concepts and foundations of programming languages, and described the semantics of the resulting language in an informal but extremely precise way. In the course of this work, he gathered Barry James Mailloux, John L. Peck and myself as co-editors to support him in this superhuman editing task. At the Working Group meetings he duly discussed the progress, but he also kept up by surface mail a growing exchange of ideas with groups and individuals all over the world who commented succeeding drafts and scrutinized every part of the syntax and semantics, often providing large and detailed revisions.
A group at Philips/MBLE in Brussels contributed the Brussels Brainstorms, people at TU Munich the Munich Meditations, various individuals from all over Europe, USA and Japan became regular correspondents. Prom the beginning, these included groups and individuals from the Soviet Union: the Editors became accustomed to the perceptive and constructive criticisms and contributions from Novosibirsk 4, Leningrad and a few other places. The names and opinions of Ershov, Tseytin, Lavrov, Levinson became well known to us.
The first meetings
In August 1968 in North Berwick, just after a triumphant presentation on ALGOL 68 by Aad van Wijngaarden at the IFIP World Congress in Edinburgh and a few months before the deadline of the Final Report, I met Andrei Ershov for the first time in person. Serious and thoughtful in behaviour, conciliatory and cooperative in his interaction with people, charming and humorous, he was a pillar of strength and support in a Working Group which was turning ever more critical of van Wijngaarden's work.
But this is not the place to go into the glorious tragedy that was taking place. What counts is Ershov's wholehearted desire for scientific collaboration, for international standards uniting the academic world against the overwhelming might of the commercial world.
In August/September 1971 he organized a meeting of IFIP WG2.1 in Novosibirsk, giving me a first glimpse of Russia and the Soviet Union. He came to welcome us in Moscow with his people, transferring us by small and prehistoric busses to Domodedovo Airport. After a very late night flight we arrived with our jetlag in Novosibirsk, where more busses and some derelict official Wolga's transferred us to Akademgorodok.
I remember the delight of first seeing Akademgorodok, the blissful absence of cars in the town center, the houses free-standing in the woods, the Golden Valley and its hotel, with dejournaia's guarding every floor; the House of Scientists, the hospitality of the people.
Ershov was clearly the boss of the beehive called Computing Centre 5, with a large office and a villa in the woods 6 suitable for a corresponding Academician. It was a pleasure to finally meet his collaborators, Rar, Baehrs, many others. Inside the hotel our hosts behaved stiffly and formally, outside in the woods they opened up and we could freely discuss. Although the eyes of the Secret Service were definitely on us, I got the impression that Akademgorodok was far enough from the Russian border to allow those dangerous intellectuals somewhat more freedom than usual.
The years of joy
This was for me the first of many visits to the Little Town of Academics, traveling under the shield of the Siberian Branch of the Academy of Sciences which had its own hotel in Moscow 7. Since Russian scientists got little opportunity to travel abroad, Ershov brought western computer scientists to Russia for conferences, short presentations and longer stays. In this way, Russian scientists could increasingly take part in international research.
There were meetings of various IFIP Working Groups, WG2.1 on Algol and WG2.4 on Systems Implementation Languages. In September 1975 he organized a Symposium on Methods for the Implementation of Algorithmic Languages, with most speakers coming from various countries behind the Iron Curtain. Together with Andrei Ershov I edited the proceedings which were published by Springer (as LNCS 47).
The conference was attended by Russians from all over the Union, many of whom at that time spoke little English (unfortunately my knowledge of Russian remained rudimentary). It was a wonderful time, with wonderful experiences. I remember walking along the Ob estuary with two gentlemen, Mr. Markov who was associated with the Riad 8 copy of systems 360/370 and Mr. Marx (what's in a name 9) from IBM, discussing whether it would be sensible for Riad to have a maintenance contract on the IBM OS software (in the end they did not dare).
A high point for me came when Ershov requested me to present a course about the CDL compiler-compiler technology. During these lectures he gave a simultaneous translation, line by line. After every line I had the time to think, and to listen to his lively rendering. He mimicked even my gestures. He had a deep understanding of the subject, sometimes explaining something in Russian that I had not even said yet. With such an intelligent and emphatic translator, lecturing was an exhilarating experience.
At this time the Novosibirsk group wanted to do a simultaneous implementation of ALGOL 68, PL/1 and Simula 67. When I asked him how he would realize this extremely ambitious project, Ershov said: "We are going to take the ten best mathematicians of the Soviet Union who are graduating this year". And he looked like he could do it. Andrei Ershov continued to contribute to the work on ALGOL 68, for instance by the excellent Russian translation of the ALGOL 68 Report made by his people, but he also moved with the times. His work on mixed computation fitted well with the later work of WG2.1 on Transformational Programming.
The last meeting
In 1988 I managed to organize a trip to Russia for about 30 students and a few staff members from the University of Nijmegen. Transported by "Intourist", we visited Leningrad, Novosibirsk and Moscow. My Russian friends received us warmly. Presentations were made about our work and theirs. Our students had ample opportunity to meet Russian students, discussing issues of teaching and other matters. Prom their later reactions I know that for all of them this was an unforgettable and eye-opening experience. Although they saw the Soviet Union in its declining days, they also learned something about the spirit of Russia. Russia was Zen for us, I can tell you.
Unfortunately Andrei Ershov could not meet us in person in Novosibirsk. He was in Moscow, in the best oncological hospital of the country. Arriving with the group in Moscow, I went to visit him in the hospital and found him charming and inspiring as always, but very tired. Although the cancer was destroying him, he proposed that we should go for dinner in a newly-opened private restaurant in town. We had a good Ukrainian meal and toasted to the future. I sang for him one or two Ukrainian songs. We discussed the past and he presented me a little book of his poems 10. Then we each went our way.
Reflection
Andrei Ershov's death came just a few years before the end of the Soviet Era, which had overshadowed all his life and scientific work. He was very sad about the growing dereliction and stagnation of his country. Although I did not want to press him on this subject, he must have known that the Soviet Empire could not hold out. He would have been happy to know that its end was peaceful (thanks to Gorbatchov) and that the shocks of embracing capitalism have readily been absorbed.
Andrei Ershov believed in the brotherhood of scientists, in their responsibilities in Society. Together with Aad van Wijngaarden, Fritz Bauer and many other western scientists he seems to have grimly decided not to accept the prevailing division of the world in two irreconcilable political camps. He wanted to emancipate Russian science, to open the doors to the world.
How pure, in retrospect, were the motives of those scientists who, like me, came to the Soviet Union on special invitations to present western research, in particular their own? Did they see themselves as bringers of the light? Were they secretly tickled by being glorified as experts? Were they "fellow travelers" who supported the regime?
Speaking for myself and the colleagues I know, we had no illusions about the regime and were well aware how badly it was treating its critical citizens and those it considered as criminals. We were informed about the Gulag and recognized the heavy propaganda. But we could not see the Russian people as enemies, more as victims. We felt the eager desire for knowledge and collaboration, the shared humanity. We too believed in the brotherhood of science, the need for openness, collaboration and respect that would free us all from totalitarian repression.
I count myself very lucky to have known Andrei Ershov. He was an example for me and a father figure. He was missed deeply by his friends and collaborators after his untimely death. What is it that makes Andrei Ershov such an unforgettable person?
Cornelis H.A. Koster
Dept. Comp. Sci., University of Nijmegen, The Netherlands
kees@cs.kun.nl
draft June 22, 2006
Delivered on June, 26, 2006 at the memorial session of the International Conference "Perspectives of System Informatics" held to honour the 75th anniversary of academician Andrei Ershov.
1 Delivered on June, 26, 2006 at the memorial session of the International Conference "Perspectives of System Informatics" held to honour the 75th anniversary of academician Andrei Ershov. Commented by I. Kraineva and A. Rar.
2 Officially, A. Ershov’s status as a USSR representative in IFIP Working group 2.1 was confirmed in October, 1963. But the proceedings of WG 2.1 refer to him as the Group member, starting from its foundation in 1962.
3 For the first time Cornelis Koster attended the WG meeting as an observer in Tirrenia, Italy, in June 1968, and as its member in Munich, in December 16-20, 1968.
4 Following "ВВ" and "ММ", those from Novosibirsk were called Novosibirsk Newcomers – NN.
5 A. Ershov was only the head of the Programming department of the Computing Center.
6 The author must have mistaken the Small House of Scientists, where the guests were welcomed, for Ershov’s villa, but he’d never lived in a separate cottage.
7 It’s the Hotel of the USSR Academy of Sciences.
8 Later "Riad" was called the ES computer series.
9 This is a very popular citation from "Romeo and Juliet".
10 Probably, Andrei Petrovich recited his poetry to the author – the book was published in 1991.