Soviet Historians, Quantitative Methods, and Digital Computers George F. Putnam
D u r i n g the 1960's, quantitative methods began to make a noticeable impact on the work of Soviet historians. Numerous books and articles based on quantitative studies appeared, courses in mathematics and computer data processing were included in graduate history programs, and efforts were made to induce more historians to learn and use the new methodology. Pressure to "quantify" came not only from the small minority of historians employing the new techniques, but also from the Communist Party ideologists. At the Twenty-third Party Congress the Central Committee passed a resolution, "On Measures for the Further Development of the Social Sciences and Increase in Their Rote in the Building of Communism. ''~ The following year, on June 8, 1968, the government newspaper, Izvestiya, carried an article by Aleksei M. Rumyantsev, Feodor M. Burlatskii, and Gennadii V. Osipov on "Concrete Social Research: Tasks and Perspectives." 2 0 s i p o v is the Director and Buflatskii a member of the Institute of Concrete Social Investigation of the Academy of Sciences of the USSR. Burlatskii is also a political reviewer for the Party newspaper, Pravda. Rumyantsev is a member of the University of World Economics ~and International Relations of the Academy of Science. These writers stress the need for social scientists, including historians, to master quantitative techniques and join in interdisciplinary work. The Party seeks a directly useful historical scholarship able to contribute to the solution of pressing social and economic problems facing the USSR in the 1970's. However, they are also the guardians of Marxist-Leninist orthodoxy, and fear that the uncontrolled application of historical reason to concrete problems might undermine the established economic and social system and above all the monopoly of political power held by the Party. They encourage the use of methods intended to produce more reliable, more useful information but they want investigations to be made and results formuiated within the framework of Marxist-Leninist theory. That "bourgeois objectivity" and "bourgeois empiricism" remain the terms of abuse they have been since the end of the 1920's in the USSR is clear from Dzhermen M. Gvishiani's article, "Historical Materialism and Particular Sociologica/ Investigations, ''3 and .that of Rumyantsev and Osipov, "Marxist Sociology and Concrete Social Investigations. ''4 1Kommunist, No. 13 (1967), 3-13. 2Izvestiya, 8 June 1968, pp. 2-3, translated into English in Soviet Studies in Sociology, VoI.VII, No, 3, 3-8.
3Questions o f Philosophy, No. 5 (1965), 47-56. 4 Questions o f Philosophy, No. 6 (1968), 3-13. George 17. Putnam is an associate professor in the Department o f HistoYy at the University o f Missouri.
23
24 Computers and the Humanities/Vol. 6, No. l/September 1971 Important historians also urge greater use of mathematical tools and computers, arguing that if historians do not move in that direction, history will be unable to play its unique and indispensable role in the advancement of knowledge. The question of the role of history is examined b y Iurii A. Polyakov, Vladimir E. Poletaev, and Valentin A. Ustinov in the article "History, Concrete Social Investigations and Cybernetics. ''s At the time the article was written Polyakov was a member-correspondent of the Academy of Sciences and Editor-in-Chief of History of the USSR and the historian Poletaev was a member of the editorial board. The authors point out that using mathematical and computer-oriented tools, social scientists are moving toward a fundamental scientific objective, the development of a common language on the basis of which fruitful interdisciplinary work can be carried on and contributions can be made toward the solution of complex social problems in the USSR. Historians, wedded to narrative description and "common sense" analysis and refusing to learn and apply the new methods, cannot learn this language, cannot join in this work, and run the risk of becoming merely antiquarians. But historical "antiquarianism" is intolerable because no productive insights into social questions can be gained without knowledge of their history. The "laws" of social development, themselves supposedly derived from an empirical study of history, are considered to be forcibly active in every social situation, and historical study shows how they are concretely manifested. Since, according to those "laws," everything develops dialectically, bearing in itself the best of what it has experienced over time, nothing can be understood in statu. A full and useful understanding of any institution, social group, or national group in the USSR, for example, necessarily involves understanding of the way in which it developed. So, the historians point out, even the interdisciplinary efforts of other social scientists cannot succeed without the historical perspective. Historians who refuse to master the new methods and language of unified research and to add their indispensable component to interdisciplinary studies are depriving their country of a contribution which only they can make. Within this context of professional and Party pressure, some historians in the USSR have produced analytic historical studies using mathematical methods and computer data processing. Most of these deal with economic history. Other scholars have introduced what appears to be rigorous training in classical mathematics, statistics, probability, and computer logic and programming into history graduate programs in the USSR. These studies and training programs show how some Soviet historians view quantification and what they would like to do irrespective of Party dogma and historical Establishment promptings. Historical Work Using Mathematical M e t h o d s and C o m p u t e r s
Numerous historical books and articles based on quantitative methods and the use of digital computer systems have been published in the USSR in the past seven years. One of the most prominent historians working in this area is the present Editor-in-Chief of History of the USSR and professor at Moscow University, Ivan D. Koval'chenko, who has produced substantial work, in terms of both the scope of his research and his attempts to explain the use of mathematical techniques and computers to Soviet historians. His most important work is the monograph The Russian Serf Peasantry in the First Half of the 19th Century, a ground-breaking work in Russian agricultural history. In this work Koval'chenko tries to answer the objections of those who doubt the applicability of mathematics to the analysis of historical data. At the same time he warns against an Slurii A. Polyakov, Vladimir E. Poletaev, and Valentin A. Ustinov, "History, Concrete Social Investigations and Cybernetics," History of the USSR, No. 4 (1968), 3-15.
Putnam/Soviet Historians 25 overestimation of what mathematical methods can accomplish and against the everpresent tendency for mathematical manipulations to become ends in themselves: In principle, their application is permissible in all spheres where certain indicators can be expressed quantitatively. On the other hand any sphere of mathematics is only a means of knowledge, one of the methods of the discovery of the internal nature and qualitative uniqueness of the object of study. These methods are most effective insofar as they present a quantitative measure of qualitative indicators, which creates a basis for more exact and objective judgment. (p. 53) Historians should not fear that the use of the universal language of mathematics would bring about an "excessive formalization of the essence of historical phenomena" (p. 53). Historical events in all their variety remain the subject of study; the living persons and the living social situations of history are not deprived of the complexity of their natures and interactions. Only those elements which can be quantified are made the subject of mathematical treatment and then only as a means to explaining their qualitative characteristics. An idea of the substantive side of Koval'chenko's work can most easily be gained from a companion study to The Russian Serf Peasantry which he published as an article. In "On the Application of Mathematical Methods in the Analysis of HistoricoStatistical Data" Koval'chenko attempts to explain to historians the use of one form of elementary statistics and offers a concrete example of an analysis of historical data which could not have been done by "common sense" methods alone. 6 It was the inability to deal with several interrelated variables in his study of agricultural history which sent Koval'chenko to the Laboratory for the Application of Electronic Computers in Humanistic Research of the Institute of Mathematics at the Siberian Section of the Academy of Sciences. There, with the aid of mathematicians and computer scientists, he was able to apply statistical methods (ascertaining of coefficients of correlation and regression) to the problem of determining the dependence of agricultural productivity in nineteenth-century Russia on specific individual variables. At the same time he Was able to analyze the stratification of the peasants: Concretely, the task came down to the explanation of the dependence of the amounts and levels of peasant agricultural production on the working forces provided for the peasants, the intensity of their exploitation and the development amongst them of craft production. Besides that, the study revealed a dependence of the members of different strata among the peasantry (that is, their share in the general number of farmsteads) on the exploitation of the peasants and the degree of development amongst them of crafts production. (p. 16) Of course, traditional historical methods would have focused on these same factors. The additional knowledge provided by the use of statistics is a quite exact indication of the degree of importance of each of the factors with regard to productivity and stratification at different times and in different localities. Another pioneering work in the field of quantification of historical analysis is that of V.A. Ustinov, The Application of Calculating Machines to Historical Science. 7 In this work Ustinov explains calculating techniques, computer mathematics, and mathematical logic; provides an introduction to flow charting and programming problems to be solved 6Ivan D. Koval'chenko, "On the Application of Mathematical Methods in the Analysis of Historico-Statistical Data," pp. 13-20, translated into English in Soviet Studies in History, No. 4 (1964). V. A. Ustinov, The Application o f Calculating Machines to Historical Science (Moscow: "Mysl," 1964).
26 Computers and the Humanities/Vol. 6, No. l/September 1971 on digital computers; and describes five historical projects which were carried out using these techniques. In addition to the work by Koval'chenko and Ustinov,there is by now substantial published historical work in the USSR employing mathematical methods and digital computer processing, some of which is listed in the bibliography at the end of this paper. At the same time serious efforts have begun to spread the use of these techniques by embodying them in the history curricula of Russian universities.
Training Historians in Use of Mathematical Methods and Computers In "Future Historians Master Quantitative Technique, "8 V. A. Ustinov and Leonid M. Goryushkin, Candidates in Historical Sciences, describe the innovative work begun at NovoSibirsk University in 1964-65. There it was recognized that historians must be prepared to do interdisciplinary work and that "different mathematical disciplines and the basis of calculating techniques must be brought into the program of the humanistic disciplines." Specifically they should acquire knowledge of discrete mathematics (the elements of mathematical logic, the theory of numbers, the theory of probability, linear algebra, etc.), the fundamentals of quantitative techniques and mathematics, and also knowledge of classical higher mathematics (elements of the differential and integral calculus). 9 Accordingly, the 'following were included in the history and philology curricula: (1) In the second and third semesters the basis of higher mathematics (100 hrs.). (2) In the fourth and fifth semesters the fundamentals of statistics and the theory of probability (70 hrs.). (3) In the sixth semester the methods of application of electronic calculating machines and mathematical methods in historical research (30 hrs.). To receive his diploma the student must complete a concrete historical research project using these techniques. As a project for the first class pursuing the new curriculum the faculty assigned an analysis of data collected in the Tomsk government by the Imperial census of 1916. This, the most complex census attempted in Russia up to that time, tried to ascertain along with population the amounts of productive resources in Russia. It operated on three levels, the first of which tried to ascertain the following: (1) General facts about all peasant farmers in Tomsk, such as whether they were long-term residents or resettled peasants, and, if the latter, the year of their arrival, what government they came from, etc. (2) Various kinds of information about the population, such as numbers of men and women, whether they were self-employed or hired workers, which of them were prisoners of war and refugees. (3) The quantity of cattle of all kinds and ages, the sizes of crops of different cultures on their own and leased land. The second level touched only five percent of the peasant farmers and took stock of the crop and cattle for 1915-16. The third, carried out in two percent of the peasant households, counted, besides the above, the input and output of basic foodstuffs and fodders and stores of them for that same year. The materials are available at the Tomsk State Archive, on 680,000 cards, containing data on all peasant farmers in the government. The material for the student project was drawn from this source. The students worked on 8,000 cards, containing the data on all the peasant farmers of eleven volosts. Their most general task was to "work out the common and SV. A. Ustinov and Leonid M. Goryushkin, Technique," The University Herald, No. 8 (1967), 83-85. 9Polyakov et al., pp. 10-12,
"Future
Historians
Master
Quantitative
Putnam/Soviet Historians
27
differentiating features of the development of agriculture and class differentiations in separate rayons and amongst peasants of different categories." A second major task was to "establish the main indices of economic prosperity in different regions and social groups." In order to carry out the project the students were given, in addition to the technical courses listed above, a special course about the census of 1916 as a historical source. In addition, they were required to master practical skills, such as the coding and punching necessary to prepare the material for introduction into the machine; preparation of algorithms, flow charts, and programs for the solution of the problem; and design of proper methods of editing and printing out the results of their programs in an efficient and understandable form. The output from this project was to be two sets of punched cards with instructions for their maintenance and use by a variety of researchers.~ o This project and similar ones introduced at Moscow and Leningrad State Universities made it evident to some Soviet historians that such training and practice for historians is practical and useful, indeed necessary, if mass sources are to be utilized. Historians cannot remain isolated from other social scientists and mathematicians if they wish to work on meaningful problems which can be handled only with masses of data. In fact, they concluded that what is needed is a group of specialists of"humanistic profile," well prepared in concrete sociology, mathematics, and quantitative techniques. On the other hand, this and other work done by historians and other social scientists using mathematical methods and computers suggests that computer scientists also have a job to do, particularly in designing the larger, faster memories and much faster, much more flexible input-output devices which the solutions of most problem~ in the social sciences require. Cooperative effort along these lines can lead to the ability to prepare and input masses of information far beyond those now available to historians, to develop ways to analyze this material rapidly in much more complex ways than is now possible and to produce much more useful and challenging conclusions. Within the history profession, quantification is proceeding in the USSR much as it is in the USA. A small but growing number of scholars explore important areas of Russian history, hoping to show that sharper and sometimes previously unattainable conclusions can be reached by use of statistical methods. 11 Some graduate curricula now include training in mathematical techniques and the use of digital computers, apparently on a more rigorous basis than in the USA, where an unfortunate "Ten Easy Lessons" mentality often prevails. Given the attitudes toward science engendered by the Soviet climate of opinion, it is probable that quantitative methods will find more enthusiastic practitioners and fewer "professional humanist" opponents than in the USA. One can expect, within the spoken or unspoken limits set by the guardians of Marxist-Leninist orthodoxy, many productive studies of Russian economic, political, and social history. But the limits are real and presage difficulties for any attempt to pursue objective historical investigation, statistical or otherwise, in the Soviet Union. Particularly is this the case if any serious effort is made to follow the course set out by some Party ideologists and prominent historians, that of integrating history into interdisciplinary studies aimed at analyzing current serious economic-social problems in the USSR. It is likely that "politics" will conflict with "economics" here as it has in other contexts in the past half-century. The apparatus of the Party will continue to resist any attempted "rational" approach to Russian reality which might threaten the Party's monopoly of political power (politics). Pooling the talents of social scientists, including 1 o Ustinov and Goryushkin, p. 84. I t Ivan D. Koval'chenko et al., Quantitative and Machine Methods o f Processing Historical Information, a report presented at the XIII Interna t i ona l Congress of Historical Sciences, Moscow, August 16-23, 1970, and published in English translation by " N a u k a , " Moscow, 1970o
28 Computers and the Humanities/VoL 6, No. 1~September 1971 historians, on the basis of a common methodological base and scientific language sounds like a rational way to approach thorny problems like the continuing divergent interests of city and rural populations and the growing restlessness of the subject nationalities (economics). But it is clear that from the beginning the shadow of "politics" hangs over the whole enterprise. Ideologists warn that Marxism-Leninism as defined by the Party remains the standard by which conclusions from research must be measured. Strictures against unsupported empiricism are not only of the valid kind which point up the inevitable subjectivity of such an approach to the study of man, but are also of the kind which assert that empirical findings must not contradict Marxist dogma. The very Party ideologues and leading historians who are so fervently urging upon historians the use of quantitative methods and digital computers as an aid to the solutions of concrete problems may be conjuring up a force with which they will later have to deal. And historians, in pursuit of the objectivity, rigor, and analytical clarity resulting from the use o f quantitative methods may find themselves confronted with dangers they have managed largely to avoid while working in the narrative-descriptive vein. As Thomas Kuhn has said, "No crisis is so hard to suppress as one that derives from a quantitative anomaly that has resisted all the usual efforts at reconciliation. ''12 12"The Function of Measurement in Modern Physical Science," in Harvey Woolf, ed., QuanffJ~cation: A History of the Meaning of Measurement in the Natural and Social Sciences, (Indianapolis, 1961), pp. 50, 52.
Select Bibliography of Soviet Historical Works Using Quantitative Methods and Computers Iu.L. Bessmertnyi, "Certain Problems of the History of the Nobility in Northern France from the End of the 12th to the Beginning of the 14th Century." French Yearly. Articles and Materials about the History of France, 1966 (Moscow, 1967). V. Z. Drobizhev, "Methods of Statistical Treatment of the Protocols of VSNX (1917-1929)," Journal of Moscow State University, Series IX, History, no. 6, 1965, pp. 70-92. A. V. Gulyga, Iu. A. Levada, eds., Philosophical Problems of Historical Science (Moscow: "Nauka',' 1969). J. Kahk, The Crisis of Feudal Agriculture in the Second Quarter of the 19th Century, Tallin, 1969. Iuo Kakhk, Kh. Ligi, "On the Question of Economic Conditions in the Feudal Obligations of Peasants in the Estlyand Government in the 18th and 19th Centuries (An Attempt at the Application of Electronic Calculating Machines in Historical Research)," Annual on the Agrarian History of Eastern Europe, 1962, (Minsk, 1964). lu. Kakhk, "Application of Electronic Computers in the Research of the Hisforians of the Estonian SSR," History of the USSR, No. 1, 1964, pp. 25-29. Z. G. Karpenko, The Smelting and Metallurgical Industry in Western Siberia, 1700-1860, (Novo Sibirsk, 1963). , "The Study of the History of Manufacturing Establishments of the Socialist Epoch by New Methods," History of the USSR, no. 1, 1964, pp. 20-24. , "On the Application of Calculating Machines for the Study of Manufacturing Cadres," The Party Organization of Western Siberia in the Struggle to Build Socialism and Communism (Materials for'a scientific conference), part 2, (NovoSibirsk, 1963). K. V. Khvostova, Peculiarities in Agrarian-Legal Relations in Late Byzantium (14th-15th Centuries), A Historical-Sociological Sketch. (Moscow, 1968). E. 1. Kolycheva, "An Attempt at the Application of Correlation Analysis for the Solution of Certain Disputed Questions of the History of Serfdom." History o f the USSR, no. 4. 1969, pp. 140-45. I. D. Koval chenko, "On an Attempt at a Mathematical-Statistical Treatment of Selected Statistics on Peasant Agriculture in Russia in the 19th Century," Journal of Moscow State University., Series IX, Histo~ no. 1, 1966, pp. 76-89. , "On the Character and Forms of Differentiation among Landlords' Peasants in Russia in the First Half of the 19th Century," Historical Notes, "Nauka," Vol. 78, 1965, pp. 85-150.
Putnam/Soviet Historians
29
_ L. V. Milov, "On the Principles of Research on the Process of Formation of the All-Russian Agrarian Market (18th-19th Centuries)." History o f the USSR, no. 1, 1969, pp. 27-57. -, V. A. Ustinov, "Rural Life in Russia in the 19th Century: Electronic Calculators Applied to Historical Studies," Annales, Economie, Societds, Ovilizations, Nov.-Dec. 1965, pp. 1128-49. Origin, Questions o f History, no. 5, 1964, pp. 54-67. B.N. Mironov, "The Application of Selected Methods for the Analysis of the Movement of Grain Prices in 18th Century Russia." Annual on the Agrarian History of Eastern Siberia, 1964, (Kishinev, 1966). Iu. Moshkov, "Cybernetics and Methods of Historical Research" (Scientific Conference for the Historical Faculty), Journal o f Moscow State University, Series IX, History, no. 1~ 1966, pp. 90-93. V. V. Nalimov, "Quantitative Methods of the Investigation of the Processes of Scientific Development," Questions o f Philosophy, no. 12, 1966, pp. 38-47. Kh. Palli, "Toward a Method of Treatment of Demographic Materials of Estonia of the 13th-18th Centuries," Theses of Lectures and Reports at the Ninth (Tallin) Session of a Symposium on the Agrarian History o f Eastern Europe (October 1966). Tallin, 1966, pp. 106-8. A. Russman, "From an Attempt at the Application of Fai~tor Analysis in Historical Research," Journal o f Moscow State University, Series IX, History, no. 6, 1966, pp. 78-91. I. Sil'dmyae, L. Vykhandu, "On the Appficafion of Mathematical Methods to the Treatment of Data on Feudal Rent," Scholarly Notes o f Tartu State University, Part 183. Works on Jurisprudence IV, (Tartu, 1966). Source Control, Theoretical and Methodological Problems (Moscow: "Nauka," 1969). V. A. Ustinov, "The Solution of Certain Problems of the History of the USSR by Electronic Computers," History o f the USSR, no. I, 1964, pp. 3-13.
Meetings Announced Continued from page 2. 21-26 August 1972 THE THIRD INTERNATIONAL CONGRESS OF APPLIED LINGUISTICS will be held in Copenhagen, Denmark. President of the conference will be Professor Bertil Malmberg, who is Secretary General of the International Association for Applied Linguistics. The Danish Association for Applied Linguistics (ADLA), under ADLA President Jacques Qvistgaard, is organizing the meeting on behalf of the Scandinavian Associations for Applied Linguistics. The official languages will be English, French, and German. The suggested topics are as follows: applied linguistics (field and scope); quantitative linguistics; contrastive linguistics; translation; lexicography, including terminology; language for special purposes; the syntax of the spoken language; applied phonetics; language planning; goal analysis; curriculum research; examinations, tests, and certificates; criteria for the choice of texts in language teaching; language teaching technology; programmed instruction, including self-instructional material; modern language teaching for adults; the training of language teachers, including the teachers of language for special purposes; special linguistic problems in international cooperation, including the language training of international experts and volunteers; psycho-linguistics and language teaching/learning; child language (first and second language); bilingualism; socio-Iinguistics and language teaching, including language and culture; and relations between communication theory and language teaching. For further information, write to Dr. A. Hood Roberts, Center for Applied Linguistics, 1717 Massachusetts Ave., N.W., Washington, DC 20036.