Daniel Bell, Twelve Modes of Prediction;
A Preliminary Sorting of Approaches in the Social Sciences

Daedalus Vol. 93, No. 3, Population, Prediction, Conflict, Existentialism (Summer 1964), pp. 845-880 Published by: The MIT Press on behalf of American Academy of Arts & Sciences Stable URL: http://www.jstor.org/stable/20026862 ISSN: 0011-5266; OCLC Number: 479215401

In the Cours de Philosophie Positive (1830-1842), perhaps the last individual attempt to write a synoptic account of human knowledge, August Comte cited as an example of the inherently unknowable, the chemical composition of the distant stars and the question of whether there were "organized beings living on their surface."1 Within two decades, Gustave Kirchoff applied spectrum analysis to the stars and provided that very knowledge which Comte had declared to be impossible. And, as space probes reach Mars and Venus, we shall be in a position to answer the second question of the nature of the forms of life, if any, on the solar planets. Few persons today would declare with confidence that some thing is unknowable. So secure is the dominion of science that the obverse attitude rules: today we feel that there are no inherent secrets in the universe, and that all is open; and this is one of the significant changes in the modern moral temper. And yet, every generation now feels that the foundation of its knowledge is inadequate and that the social forms as we know them are bound to change. We expect that science and technology will rework the map of society and no one any longer challenges their claims. One of the hallmarks of "modernity" is the awareness of change and the struggling effort to control the direction and pace of change.

The problem of any science is to understand the sources of change. And in this respect social science is fairly recent. The great intellectual barrier was that men always thought they knew the sources of change, which were also the sources of power, namely the personal will of kings, lawgivers and prophets, those who governed states, drafted laws, and established or reinforced religious beliefs. But only gradually did men realize that behind hese visible sets of acts were such intangible nets as customs, institutions, and cultures, which subtly constrained and set the bound aries of social action. At the same time came the slow realization that there were "social forces" which generated change, whether they be impersonal processes such as demographic pressures (increased size and density of populations), technology, and science, or conscious strivings such as the demands of disadvantaged groups or equality or social mobility.

It is the modern hubris that we can effect the conscious transformation of society. What stands in our favor is that knowledge is cumulative. And, within the open community of science, it is self corrective. We know more of economics than the "Political Arithmetick" of Sir William Petty, who started us out with "Number, Weight and Measure." We have clearer conceptual distinctions than Herbert Spencer and his primitive efforts at establishing social differentiation. We have more complex statistical tools than Pearson or Galton. What is more important, perhaps, is that we have a better appreciation of method. For what method allows us to do is to reformulate insight into consistent explanation. More than two thousand years ago, Plato wrote: ". . . it is in sleep that the wild beast in our nature stands up and walks about naked, and there is no conceivable folly or shame or crime, however unnatural, not excepting incest or parricide, of which such a nature may not be guilty. In all of us, even in good men, there is a latent wild beast which peers out in sleep."2 Yet it was only in The Interpretation of Dreams that such wisdom was assimilated into a coherent theory which explained not only the existence of such impulses but the mechanisms of control through the idea of wishes and repressions.

In this emphasis on method, the function of conjecture is not prediction but explanation.3 Prediction can be derived from experience, such as a farmer's expectation of crop production, without knowledge of the reasons why. Only with adequate explanation – understanding of the relevant variables – can one seek to control or transform a situation. Conjecture, in this sense, stipulates a set of future predicates whose appearance should be explainable from theory. Prediction without explanation is insight, experience, or luck. In this sense, too, the function of prediction is to reduce uncertainty. By stipulating, and thus testing, a hypothesis, one is able to verify the relevant factors which account for a predicted or observed change. For this reason, in distinguishing different kinds of prediction in the social sciences, I am ruling out historical theories of rhythms, periodizations and cycles – in most instances, recurrence is not really identity but, at best, analogy – wherein the investigator seeks for the Pythagorean number which rules the wheel of history or imposes a diachronic pattern on the rise and fall of civilizations.4

In sorting out the different modes of prediction, I shall not be concerned with the relevant methodological tests of adequacy. The sorting is an effort to illustrate the range of approaches in the art of conjecture. The different modes are not "equal" to each other in the level of generality or scope of comprehension. Inevitably there are some overlaps, and a stricter reading might show that one mode subsumes some of the others. But to the extent that one can sort out a distinct number of types, the next step would be to order them in some logical classification and specify which mode of conjecture would be appropriate to what kind of problem.

The Comtean quest for "social laws" in which some basic regularities of human behavior or some major variables such as mass, pressure, gravitational velocity could be synthesized into a set of formulas akin to Newtonian or the later statistical mechanics of Willard Gibbs finds few serious adherents in the social sciences today. Yet one such model, Marx's 'laws of motion" of capitalism, has been one of the most influential ideological doctrines of the past century (in but three years we celebrate the centennial of Das Kapital), and to the extent that it is still an element in the Bolshevik belief-system, it is a factor in assessing Communist policy.

Questions of politics apart, the Marxian model is still one of the most comprehensive efforts to create a large-scale system of prediction based on the interaction of a few crucial variables. Granted that the model has not worked, its 'logic" is worth explication. Marx began, one may recall, with the components of value as constant capital, variable capital, and surplus value (c+v+s), and from these, two ratios become central. One is the rate of surplus value, s/v = s', and the other is the organic composition of capital, the relationship of labor to materials and machinery, c/c+v = q. These two are the primary variables from which the crucial variable, the rate of profit, is derived: p = s' (1-q).

The "law of the falling rate of profit," which Marx derives from these equations, is the key to the system. Each capitalist, trying to increase his own profit, substitutes more constant for variable capital and ends by killing off the total of profits. Out of this tendency, Marx derives the corollary social consequencies: the reserve army of the unemployed, the impoverishment of labor, the centralization (trustification) of capital, a deepening series of crises, the intensification of exploitation, sharpening class struggle, etc.

One can point out, as Paul Samuelson has,5 that the model rests on the difficult assumption of fixed coefficients and single techniques, and that modern tools such as linear programming can allow the substitution, analytically, of many alternative techniques in working out adequate production functions (i.e. the combination of units of labor and capital). Or one can argue the empirical point, as Strachey and others have, that the economic system is not autonomous and that a regulating agency, the State, can readjust the "equations" in response to political pressures.

The crucial element in the Marxian model, which defines it as "social physics", is that these actions took place independent of the will of any single individual or, in the long run, of any groups of individuals. The interesting question remains whether such comprehensive, dynamic models are possible, even in the simplified versions such as Marx employed, and what would be the central variables that one would select to describe social and political interaction.

A different kind of 'law," which has attracted increasing attention in recent years, is that of 'logistics curves," which various authors have fitted to different time-series. The most startling of these are the laws of "exponential growth" or the "doubling rates" of different social phenomena.

The late Louis Ridenour has pointed out that the holdings of libraries have doubled every eleven years since 1870, that long-distance telephone calls have doubled every seven-and a-half years, that since Nellie Bly went around the world in 1889 the time for circumventing the globe decreased exponentially between 1889 and 1928 by a factor of two every quarter-century, and since the introduction of aircraft the rate of change has markedly increased. Derek Price has shown that the number of journals in science has increased by a factor of ten every half-century since 1790, and the number of abstracting journals has followed precisely the same law, multiplying by a factor of ten every half-century. At a certain point in these growths, "critical magnitudes" are reached and curves react to conditions" in different ways.6

Ridenour, in fitting a number of such curves, has argued that the process of growth and saturation in social change follows the so called "autocatalytic" processes of chemistry and biology. Price has sought to identify more differentiated modes of "reaction." He writes: ". . . growths that have long been exponential seem not to relish the idea of being flattened. Before they reach a midpoint they begin to twist and turn, and, like impost spirits, change their shapes and definitions so as not to be exterminated against that terrible ceiling. Or in less anthropomorphic terms, the cybernetic phenomenon of hunting sets in and the curve begins to oscillate wildly.. . . [One] finds two variants of the traditional logistic curve that are more frequent than the plain S-shaped ogive." One variant is the phenomenon, first recognized by Gerald Holton, of "escalation," in which growth curves "pick up" from earlier, related curves in repeated sequence. The other variant is one of violent fluctuation with a logarithmic decline to a stable maximum or to zero. This leads Price to conclude: "All the apparently exponential laws of growth must ultimately be logistic, and this implies a period of crisis extending on either side of the date of midpoint for about a generation. The outcome of the battle at the point of no return is complete reorganization or violent fluctuation or death of the variable."7

If Ridenour's equations hold true and Price's generalizations regarding the "midpoints" of exponential curves are valid, one has here a powerful means of identifying processes of social change and making predictions about the outcomes. And yet, the history of earlier, similar efforts to find underlying 'laws" of variegated social phenomena should give us pause.

In the late 1930s, the Harvard philologist George Kingsley Zipf reported remarkable regularities in such diverse phenomena as the distribution of cities by size, the relation of rank order to frequency of word occurrences, the distribution of the frequencies of publication of scientists, and many others. Zipf tried to bring all these under the roof of a single mathematical relationship, the harmonic law, and devised a pseudo-explanation which he dubbed the "law of least effort." Fifteen years or so later, Herbert Simon was able to show that the similarity of these statistical distributions was not due to any overriding law but was a consequence of the similarity in the structure of the mechanisms.8

Yet this corrective effort has given us an interesting tool of prediction. For in the construction of such probability models we encounter stochastic processes9, where the sequences of outcomes are uncertain and where we are forced to build "conjecturing" into the structure of the model. The development of stochastic reasoning, using probabilistic models, should allow us to do better estimating in problems involving frequency distributions, as well as those involving future outcomes involving the choice of further information.

Some recent efforts to create a "social physics" devolve from the work of the mathematical biologist Nicholas Rashevsky. One of his students, Anatol Rapoport, has written a book, Fights, Games, and Debates10 which sets up mathematical models for mass action and descriptive models for conflict situations. One section of the book deals with arms races, and the equations that are developed seek to describe the development of actual arms races in the same way that the equations of thermodynamics are meant to describe the actual behavior of gases. One intention of the new social physics is to set up general probabilistic laws governing behavior in game-like situations.

One older tradition of social physics is worth mentioning: the efforts, initiated originally by Robert Park, to explain (and predict) the growth of cities and other aspects of social ecology (the spatial distribution of units) as a product of impersonal forces such as competition or technology rather than willed or planned volitional efforts. Among geographers and some sociologists, this effort to create a so-called "biotic" or sub-social framework continues.11

The most familiar form of prediction involves some form of extrapolation from tune-series either as straight-line projections, cyclical turns, or alternative models based on some definition of upper and lower limits. One can say that this mode of prediction differs from "social physics" in that the latter seeks some general principle, or attempts to create a closed system; trend analysis takes some selected area and seeks to make a more limited prediction, ceteris paribus.

The three major kinds of trend analysis have been economic forecasting, demography, and technological change. The assumptions of economic forecasting itself would require a long paper,12 so I restrict myself to the other two.

Demographic prediction, in the large, has always been under the long shadow of Thomas Malthus.13 While the exact relationship between the arithmetical growth rate of food supply and the geometrical tendency of population increase is now suspect, the neo- Malthusians still hold to the general view that the present is in some way an exceptional period and that at some point, either fifty or a hundred years from now, the world "will have begun to go back," as Sir Charles Darwin has put it, "into ... its normal state, the state in which natural selection operates by producing too many people, so that the excess simply cannot survive."14

Sir Charles has made some specific estimates. He feels that by A.D. 2000 the world population will be four billion. He cites the fact that between 1947 and 1953 agriculture production increased by 8 per cent but the world's population by 11 per cent, "so that the world was hungrier at the end than at the beginning. And in fifty years, the four billion will be hungrier than the two and a half billion in 1950."

Leaving aside the question of the definition of hunger, the basic assumptions are threefold: that human population has followed the availability of food supply so consistently that the human animal is responding to biological conditions in no different manner than other animals; that improvements in technology cannot keep pace with the rise of populations; that the voluntary limitations on birth cannot be accepted by the whole of humanity, so that those who limited births would be swallowed up by the others who did not. Each of these is open to question in some manner. As to the first, while this may be true of primitive societies, it is less so of modern industrial countries where the standard of living becomes more controlling of the number of births; the second is still an open question; the third raises vital political questions most notably today about the role of China and secondarily of India in the desires of their governments to control or expand population in these countries and the checks thereon.

What this does point up is that demographic prediction, involving individual and aggregate decisions, the role of custom and the adaptation to standard of living, the influence of education and new class styles, the relationship to economic development and the political role of government, is one of the areas where systematic scrutiny of the conditions of prediction would yield important results.

The area of "technological trends" could lead us very easily, and temptingly, into the alluring field of science fiction. Here one's imagination could quickly climb William James's "faith ladder" and turn possibilities into probabilities and probabilities into certainties. While one would assume that the prediction of invention should be fairly easy, since the antecedent conditions are highly structured, the interesting thing is that a review of such predictions finds them to be singularly inept.15 Reality, it seems, is more recalcitrant than the imagination.

As S. Lilley writes: "The moral for forecasters is: Do not predict individual inventions in detail – that is usually a waste of time." The correct method, he argues, is twofold: extrapolation of present trends, and predictions in the form possibilities. "The predictor need not be concerned with how a technical problem is to be solved, but only whether it will be solved or not." This is based on the principle of "equivalent invention" enunciated by S. L. Gilfallen. Gilfalien, writing in 1937, cites the problem of flying in fog. There were twenty-five means suggested of solving it. To predict which means would be the most successful was hazardous; but one could predict that the problem was solvable. "The point is that we are not concerned with the prediction of inventions, only with their effects. And the fact that there are almost always many possible inventions that could lead to the same desired effect enormously increases the chances of successful prediction," Lilley concludes.

The methods of political forecasting – the work of the Gallup organization, the University of Michigan surveys, the Columbia University studies under Paul Lazarsfeld, the work of Demoskopie in Germany and Sondages in France – would require a separate paper. Strictly speaking, this is not trend analysis, though past trends may be in such crucial as how to allocate "undecided" voters. At bottom the success of the forecast is a function of the accuracy of the sample and the reduction of interviewer bias. Certain presumed disqualifying effects, such as the poll itself creating a "bandwagon," can be discounted. The success of ex post facto explanation is a function of the "process analysis" (i.e. the selection of relevant determinants such as the mass media, or personal influence) employed.16

One form of trend analysis is beautifully illustrated in the Futuribles paper of Colin Clark. It might be described as setting forth the "conditions of fit.17"17 Professor Clark takes as a problem the statement of M. Manshold, the vice-president of the European Community, that for economic equilibrium to be attained in Europe, some eight million persons within the Common Market area dependent on agriculture have to be transferred to other sectors. Clark seeks to establish the rates of transfer which might obtain for each country. He does this by computing two formulas; one, on the production side, deals with the number of persons necessary in the agricultural sector to support the food demands in the country (thus in France, in 1957-58 one agricultural worker provided food for ten non-agricultural workers; in New Zealand, the ratio was one to sixty-two); on the demand side, he calculates the rate at which demand per head for any food product may be expected to increase, based on population growth, income rise, and the elasticity of demand for particular products. At the intersects of the two curves one can chart the rates of reduction of agricultural workers for each country.

Since there is a long history of experience on the nature of income elasticities in the demand for products, Clark has been able to work out his formulas with some exactitude. The theoretical foundation, on the demand side, derives from the observations made less than a hundred years ago by the German statistician Christian Engel: as national incomes rise, a smaller proportion of the increment is spent for food. On the production, or supply, side, one can calculate the increased yields per acre on the basis of the amount of fertilizer, and the types of farming used.

In the Common Market countries, the reduction of eight million persons dependent on agriculture means the transfer of one-third of the farm population to other areas. Professor Clark computes, on the basis of rates of change obtaining since 1950, this change will be achieved in ten years in Germany, the Netherlands, and Belgium. In Italy, where the size of the agricultural population before 1950 had not changed for more than fifty years, the new methods of farming, combined with the rising national income, and reduced marginal increases for food, the agricultural labor force has been declining at a gross rate of 2 per cent per annum, or, if a reduction of one-third in the labor force is the target, it would take thirty years to reach that figure in Italy.

It is rare to have social change pinpointed over time with the exactness that Clark attempts. Since the politics, as well as the economics, of these countries will be affected by these population transfers, it would be useful to have a sociologist or a political scientist "overlay" Clark's predictions with some similar statement of expected social consequences (e.g. how will the voting patterns in these countries change as a result of the reduction of an agricultural labor force by a third, in ten or thirty years). Clark has set up an "independent variable" (a product, it is true, of more general causes). It would be fascinating to see if the dependent consequences could be charted as well.

In his Essai sur l'Art de la Conjecture, Bertrand de Jouvenel describes an order of events that are legally prescribed and traditionally reinforced, which he calls 'les certitudes structurelles." This type of ordering differs logically from trends, because it does not describe processes or time-series that are derivable from aggregate behavior, or which may be immanent, but which are based on custom and law. M. de Jouvenel's example is an interesting one. He describes the problem of a Democrat in 1962 who might like to succeed President Kennedy. He knows that there will be a president of the United States, that elections will take place on the second Tuesday of November in 1964 and 1968, that unless he has a physical accident, John F. Kennedy will be the Democratic candidate, since the party customarily designates the existing office holder. But Kennedy will not be a candidate in 1968 (assuming his re-election in 1964) because a constitutional amendment forbids a third successive term.

Of course M. de Jouvenel, like all of us, had no way of knowing that a physical accident would prevent John F. Kennedy from filling out his term. Yet the "structural certainties" remain. In this instance it is now taken for granted that Lyndon Johnson will be renominated in 1964, and, because he filled out less than half of his predecessor's term, he might well be a candidate for re-election in 1968 as well. M. de Jouvenel had taken this "prosaic" set of facts to illustrate that much of human behavior can be predicted because of such structural certainties. Yet the concept is useful because it is one of the ways of ranking the stability of different kinds of political and social systems. (What are the structural certainties in the Brazilian political system?) This concept is analogous to what sociologists call "institutionalized behavior," and in any established social system, the chief problem is to define the prescribed norms, the modes of conformity, and the limits of legitimate deviation from such institutional norms which are allowed by the system.18

Structural certainties (or institutionalized behavior) are based on a known or open mode of conduct, the rules of which are prescribed and reinforced by legal or moral sanction, and this allows one type of prediction. But there is another form of conduct which is usually implicit, rather than explicit, often unrealized even by the actors, and which has to be inferred and explicated by an analyst. This form is the "operational code" or what might be called the "do's and don't's" of conduct, the implicit rules of the game. In some instances, such as Machiavelli's The Prince, these are normative prescriptions for a ruler. But in other cases these are efforts to discern an underlying pattern of behavior, which is either an adaptive mechanism (or rules of strategy) for a political group, or simply a series of adjustments which permit political survival.

One pioneer in this type of analysis, and prediction, is Nathan Leites. In several books on the Soviet Union and on France, he has sought to codify rules.19 In the case of the Soviet Union, Leites has sought to establish the basic mode of Bolshevik conduct, which he derives from various maxims and precepts of Communist patristic writings; but he roots this, equally, in some psychoanalytic hypotheses regarding Bolshevik character structure. In his study of France, he has sought to delineate the "rules of the game" as observed in parliamentary behavior.

In an analogous but broader fashion, there have been attempts to establish the "national style" of a country. The national style, or the characteristic way of response, is a compound of the values and national character of a country. It is the distinctive way of meeting the problems of order and adaptation, of conflict and consensus, of individual ends and communal welfare, that confront any country. It has been observed, for example, that the American "style" is one that stresses action and achievement, is fundamentally optimistic, believes that life is tractable, the environment manipulable, and that all political problems can be "solved." This does not assume a distribution of such traits among all the persons in the country in any mechanical notion of national character; but it does assume that there is a characteristic way of responding to problems, which is typified in the leadership; and to this extent it can serve as a rough guide to political action.20

5.The Operational System

The "operational code" is an attempt to infer styles of conduct derived from psychological hypotheses or the value patterns of social groups or countries. The "operational system," an older form of analysis, is an effort to specify the underlying source of "renewable power" in a society regardless of the momentary fluctuations of office. Here, too, as in the case of "social physics," the most direct effort derives from Marx, and the classic analysis of this "model" in his The Eighteenth Brumaire of Louis Bonaparte.21

Louis Bonaparte, in Marx's analysis, is an "adventurer" representing no class or social group, although basing himself on the Society of December 10th and the Lumpenproletariat.22 To maintain power, he has to play off one group against another, representing himself at first for the peasantry, and then against them, for the workers, and then against them. Industry and trade prosper in hothouse fashion under the strong government. But the Bonapartist Lumpenproletariat is to enrich itself. "This contradictory task of the man explains the contradictions of his government, the confused groping hither and thither which seeks now to win, now to humiliate first one class and then another and arrays all of them uniformly against him, whose practical uncertainty forms a highly comical contrast to the imperious categorical style of the government decrees, a style which is copied obsequiously from the Uncle." The executive authority has made itself an independent power. But underneath there is still a class system. "Bonaparte feels it to be his mission to safeguard 'civil order.' But the strength of this civil order lies in the middle class. He looks on himself, therefore, as the representative of the middle class and issues decrees in this sense.

Nevertheless, he is somebody solely due to the fact that he has broken the political power of this middle class and daily breaks it anew. Consequently, he looks on himself as the adversary of the political and literary power of the middle class. But by protecting its material power, he generates its political power anew."23 The point here is that a renewable means of power provides continuity for a social system. Whether the specific historical analysis is right or wrong, methodologically, it does sensitize us to look for the institutional sources of power and to specify levels of analysis. Where a system is established (say, property in land), it is "neutral" as to "who" has power. As Schumpeter has said, the rise and fall of social classes is the rise and fall of families, but the mode of defining classes may remain a constant. In specifying levels of analysis, we can try to see what kind of political efforts seek to change the system itself, and which are changes within a system. One of the problems of modern political analysis is that many "operational systems" coexist as modes of power.24 In western democratic countries there is property, transferred through inheritance; technical skill, acquired by education; and political entrepreneurship, whose base is a mass mobilization; and each of these systems provides competing or overlapping routes to power. Yet the identification of such systems and the specification of the levels of analysis is a necessary condition for political prediction.

6.Structural Requisites

The idea of structural requisites focuses not on any underlying system but on the minimal set of concerns any government faces and it tries to identify "strains" or problems on the basis of the government's ability to manage those concerns. The list of what constitutes an invariant set of functions for any political system and the kinds of structures necessary to facilitate performance have varied with different authors. But what the approach does seek is a comprehensive typology which, in the words of one of these system-builders, Gabriel Almond, could allow the political analyst to "make precise comparisons relating the elements of the three sets – functions, structures, and styles – in the form of a series of probability statements."25 In the formulation by Almond, the product of the three sets would yield a matrix with several hundred cells, and an effort to sample frequencies over time by performances of these three sets would create a stupendous problem. Yet the effort to create a typology is the first necessary step for distinguishing different kinds of political problems.26

Au fond, what this approach suggests is what the anthropologist Alexander Goldenweiser once called "the principle of limited possibilities." Goldenweiser was trying to "mediate" the argument between anthropologists on the question of diffusion or independent invention of techniques. Methodologically, he argued that the "relatively fixed features which determined the conditions of effective use" of an object indicated a necessary convergence of forms.27 (One might say, analogously, that Pareto's Tratto of General Sociology is based on the same principle. In it, Pareto disregards the manifest political content of doctrines and seeks to establish a limited number of basic residues, derivations and sentiments, and in the combination of these to establish the basic political forms possible in society.)

Most of these theoretical efforts, if they ever come to fruition, would produce vast sociological "input-out" tables; but a rougher and readier use, implicitly, of the principle of limited possibilities indicates some of its predictive value. Thus, in 1954, Barrington Moore analyzed Soviet society in these terms.28 According to Moore, the recruitment to a ruling group could be handled through traditional (inheritance or nepotism), rational-technical (skill), or political (party loyalty) criteria. The use of any one criterion limits the range of workable alternatives for the solution of problems requisite to the system. Industrialization requires high technical criteria, but the nature of the power struggle dictates that top jobs should go to trusted individuals, while the traditional modes of clique groups serve to protect individuals from the competition of those selected by the other two criteria. By setting up these three "ideal types" Moore posited a number of alternative combinations as predictors of the future direction of Soviet society.

One important consequence of this simplified model was the prediction that terror had gone too far and that some new means of rationalization had to be found. In a different sense, the principle of limited possibilities, by concentrating on the constraints in a system, indicates the limits of change. Edward F. Denison has argued that the institutional economic arrangements in the United States make it difficult to achieve a 5 or even 4 per cent growth rate as hoped for by both President Kennedy and Governor Rockefeller. By analyzing patterns of investment, sources of capital, tax policies, habits of consumption, etc., Denison argues that an upper limit of 3 per cent a year can be reached. Any other efforts would require some drastic changes in the institutional structure of the society, with losses of different kinds of economic choice.29

7. The Overriding Problem

Leicester Webb's essay, "Political Future of Pakistan,"30 provides an interesting illustration of political analysis and prediction which is pitched on the identification of a single overriding problem. While pointing to the obvious problems of unifying a state whose territorial divisions are a thousand miles apart, whose two units speak different languages, whose population is 85 per cent illiterate and which lacks a viable political system, Webb finds that the chief problem which must be solved, before any efforts can be made to deal with the others, is the problem of a social solidarity which could be the foundation for a national identity.

The question of national identity has been singled out by Lucian Pye, for example, as the overriding problem in the creation of a viable political system in Burma.31 Where Pye, though, locates the Burmese problem in the general set of cultural ambiguities (e.g. the contradiction of gentleness and violence in the culture), the Webb essay is more focused in that it points to a fundamental dilemma in Pakistan: the effort to use religious sentiments to create a secular state. The unity of Pakistan lies in its adherence to Islam, yet Islamic thought, and the retrograde attitudes of the sect leaders, hinders the creation of a meaningful political entity. The anomaly, says Webb, is that Pakistan is a "religiously-based state ... at loggerheads with its religious leaders."

From the first, President Ayub has been conscious of this dilemma. In his public addresses he returns again and again to the theme: Pakistan needs an ideology; that ideology must be Islamic. But his two roles are in conflict. As the restorer of order and political innovator, he is of necessity a secularizer; as a solidarity maker he must appeal to religious sentiments. The answer, says Webb, is that there has to be a radical readjustment of Muslim religious thought to bring its precepts in line with new secular needs; and for this reason "Ayub is a religious as well as a political reformer." In his paper, Webb indicates the steps that Ayub has to take to break the power of the local sect leaders and to create his own political institutions. There is a clear set of stipulations by which one can judge whether these steps can be taken, and to this extent the paper serves admirably as a means of predicting the course of Pakistan's political development.

The question whether in any society there is a single overriding problem is largely an empirical one. Yet methodologically it is useful to try and see if such a single problem might emerge, if not in the present, then in the future. In her book New Nations, Lucy Mair establishes a proposition that "the world of technology is one of large political units" and thus raises the question whether some of the new states "are big enough to stand on their own feet."32 In his absorbing study of the breakdown of the Weimar Republic,33 Professor Karl D. Bracher sets up a model (applicable, he feels, to the political breakdown of a number of democratic states) in which the turning point comes when the established regime confesses that some basic problem seems to be "insoluble." In Weimar Germany, Professor Bracher feels that the insoluble problem was unemployment, giving rise to a sense of despair in the regime and a loss of faith in the political system. One could argue that Algeria was one such problem for France in recent years. Or the failure of the Belgians to create an administrative machinery (as the British did in Pakistan, India, Nigeria, etc.) was the cardinal problem in the Congo.

8. The Prime Mover

In Marxian theory, again, the mode of production was the determinant, directly or indirectly, of the political, legal, and ideological forms of the society. There were always two difficulties with the theory. One was the difficulty of any monistic theory – that in explaining everything, it really explains nothing. The other was the ambiguity of the phrase "mode of production." At various times, Marx would talk of the forces of production, the techniques of production, the social relations, etc., but meanings shifted markedly. A number of writers felt that if the term had any meaning, it could only apply to technology. At one point, I identified fourteen different variables that could be included under "mode of production" if one wanted to use it analytically. Yet the general idea of the mode of production has had a powerful influence as the idea of a prime mover of history or prime determinant of social structure.

For analytical reasons, if not always for historical or empirical ones, there may be situations in which a single powerful force can be taken as the independent variable and a whole series of ancillary changes predicted as a consequence of changes in this independent variable. This is the method, for example, that Herman Kahn has adopted in his book On Thermonuclear War and in an unpublished study, "Deterrence and Defense in the Sixties and Seventies."34 As Kahn writes: "In a sense we are adopting an almost Marxian view of the world, with military technology replacing the special role that Marx assigned to the means of production as the major determinant of behavior, and with conflicts between nations replacing the class struggle." It is not, of course, that other elements do not enter; surely political decision is the most decisive and is relatively autonomous, though one can say, after Skybolt, that missile technology has replaced steel production as the indicator of a strength among nations. But for purposes of analysis of future changes in a society, one may want to take a major determinant or prime mover and trace out its effects. The chief point that Kahn is making is that since 1945 there have been three major revolutions in the art of war, with consequent effects on the economy and political strategies of the two major duelists and their allies. By a "revolution" Kahn means a big change, such as the introduction of new sources of energy or a quantum jump in destructive power from a kiloton to a megaton -- changes significant enough to render a prevailing strategic doctrine obsolete. "The year 1951," he writes, "is typical of the new era in which there is the introduction, full procurement, obsolescence, and phasing out of complete weapons systems without their ever having been used in war."

In the "mobilized societies" of the West, military technology, in its impact on the economic budget, in its stimulus of research and development, in its absorption of scientists, in the way it constrains certain political decisions (e.g. the need for overseas bases as a consequence of a weapons system), is in many respects a prime mover and can be studied with profit. Kahn has provided one model by setting up three hypothetical technologies for 1965, 1969 and 1973 (he was writing in 1961), and seeking to assess the alternative strategies and problems created by these changes. It is conjecture as a high art. One can, for example, in examining the emerging states, look to social mobility as a major determinant of the domestic politics of the countries. One may feel that David Apter, for one, has gone too far in saying, "Assuming that no one is ever truly satisfied with the system of social stratification other than conservatives, we find that the basic motive of politics then is a striving motive to expand mobility opportunities, either for some special group or for large segments of the society."35

But one of the most fruitful perspectives adopted by modern sociology is to chart "mobility opportunities" as an index of political change in a society. Modernity, as Edward Shils once remarked, is the entry of masses into society. The extension of citizenship rights, the degree of participation, equality of opportunity – the conditions of unhindered social mobility – are a useful framework to organize the problems of political change.36

9. Sequential Development

Few persons today believe in a theory of social evolution in which each society passes through defined stages in moving from a simple to a complex state. Yet the phrases "economic development," "political development," and even "social development" do suggest that societies go through some sequential phases as they confront greater tasks and have to create specialized mechanisms to handle them. Are there some ordered steps that one can identify, at least as an ideal type? In the construction of the theory of industrial society one notices the lineaments of such a theory. One source, clearly, is Emile Durkheim's The Division of Labor in Society. In accounting for the change from "mechanical" to "organic" solidarity, Durkheim posited a number of steps. Population increases give rise to certain densities. This leads, under certain conditions, to increased interaction and thereby to competition. Social units in competition (e.g. cities, or occupational guilds) can either engage in a war to the death or begin to differentiate into specialized and complementary units. The distinguishing aspect of modern society for Durkheim is the role of the division of labor in creating a more differentiated society.

Talcott Parsons, the American sociologist, has taken Durkheim's theory and tried to account for western institutional development in terms of a theory of structural differentiation.37 Economic development is only possible when there is a separation of the household and the firm. Within the firm there goes on a renewed process of differentiation, separating planning from operations, later ownership from control, etc. In the society, functions such as education, recreation, and welfare that were once lodged within the family begin to be taken over by specialized agencies. In the political arena one finds differentiation in the rise of bureaucracy, etc. To some extent, one can argue, such a theory overlaps the concept of "structural requisites." There are two differences. One is in perspective. The idea of structural requisites is from the viewpoint of the government, i.e. the minimal concerns of any government. The second is that a theory of structural differentiation implies some ordered sequence in the creation of more complex and more special ized agencies for the handling of tasks.

In the new states one has a great laboratory for predicting and testing such hypotheses. In many of these countries, basic functions of the society are concentrated in family and village groups. The onset of industrialization means a transfer of such functions to new and more specialized agencies. Can one say that logically or sociologically such development can be plotted?38

10.Accounting Schemes

By accounting schemes, I mean those efforts to sum up in a single case, or nation, or relatively closed social system, a "trial balance" which arrays all the major factors that play a role within that unit.39 The Futuribles essay of Professor Edmund R. Leach, "The Political Future of Burma," affords an illustration of this method.40 Leach begins with the proposition that the existing situation in Burma ("or for that matter in any other country") may, in principle, be analyzed as the sum of a series of independent factors, a, b, c,. . . n. Social development is then seen as a series of progressions of a' through a to a! ', and b' through b to b' ', etc. But not all factors progress at the same pace, and a profile of the country at time II will have some but not all the factors of the country at time I. In constructing a future profile of the country, some factors which are designated as constant could be predicted with a high degree of certainty; other factors are designated as probable, and still others as pure speculation. If one takes the speculations as a range of alternative possibilities and combines each with the constants, one would have a series of delimited conjectures ordered on some scale of probabilities. In his empirical analysis, Leach distinguishes four groups of factors: A. Factors with a high degree of built-in stability which, for any short-run historical period, can be treated as constant. These include, for Burma, climate and topography, language, religion, the bureaucratic structure of the internal administration, and certain culturally-defined expectations. B. Factors which are subject to more or less linear change, either in increase or decrease. Such rates of change are not immutable, but they do tend to be relatively stable over defined periods. This would include size of population and labor force, the state of communications (roads, railways, waterways), educational level of the population, numbers of trained personnel, capital resources available for investment. C. Factors which are cyclical. Here the idea of a cycle becomes ambiguous. In business cycle theory one assumes a regularity of peaks and downturns, either long-wave (Kondratieff or Juglar) cycles or short-run cycles, based on an identification of specific determinants (rate of innovations, demand for money, etc.). In politics, the idea of a cycle is more a metaphor. This seems to be the case in Leach's examples: foreign policy ("the allies of yesterday are the enemies of today and the friends of tomorrow"); ruling types (using Pareto's metaphors of the "lions" and the "foxes," Leach assumes an alternation of "the politicians" and "the men of decisions"); age of government ("the longer an administration has been in office, the less vigorous its action"); age in government ("young men are vigorous, radical; old men incline to caution and conservatism"). Is there, here, truly a cycle?

D. Factors which are wholly fortuitous and unpredictable. These include, in Leach's itemization: the short-term objectives of active politicians; the coercive pressures of foreign powers; natural calamities, including international wars. The historian, says Leach, tends to emphasize factors of the fourth class. But it may be, he says, that these day-to-day actions have only a "superficial and transient influence upon longer-term developmental sequences." Just as Durkheim sought to predict the rates of suicide in various systems, by seeking general causes (e.g. degree of cohesion) which were independent of any individual case, in a comparable way, says Leach, "those who seek to make political predictions can and should ignore completely the whole class of short-term political events." And Leach seeks to justify such exclusion on the ground that, seen over a pattern of decades, it is the 'long-term developmental sequences" which matter, while the day to-day events are of minor significance. Two broad methodological questions can be raised. Can one dismiss so easily the role of decisions, or what historians call turning points? The annexation of upper Burma by the British in 1885 and the consequent destruction of the existing governmental system down to the village level do not fall, as a factor, under the other classes. It was not wholly fortuitous, yet it was external to, and decisively transformed, the system. One can say that such classes of events, because they are outside the system, cannot be taken into account. But if this is so, then a crucial problem in the area of prediction has no place in the accounting scheme. One would have to exclude, therefore, from the possibility of prediction such events as the occurrence of the October Revolution (many people expected February), the rise of Tito, and similar turning points in the history of the countries of Europe. Secondly, do we want to stop at an "inventory of factors"? Do we not need to specify in some way the skein of relationships so that we know not only classes of factors but their functional dependencies as well? In what way does a joint family system (a constant?) act to constrain certain types of economic development? Under what conditions do the constants change? Is there an ordering principle which can specify the relationship between the groupings as classes of functional relationships?

11. Alternative Futures: the writing of "fictions"

One of the simplest and oldest ways of conceiving, if not predicting, the future was to envisage the possibilities open to man and then create a fiction which in extreme form men call "utopia." (In a somewhat different and systematic sense, the construction of fictions was used by Jeremy Bentham to enlarge the mode of abstractions available to speculative minds, and, quite independently, more than a half-century later, Hans Vahinger was to elaborate this method in his famous book, The Philosophy of "As If." The "as if" was a construct, or fiction, which served a heuristic function. It al lowed us to simplify our assumptions.41 ) In recent years, the writing of "alternative futures" has been a systematic technique of Rand and former Rand theorists, particularly Herman Kahn. (Curiously, and quite independently, apparently, of Bentham and Vahinger, they have called their fictions "scenarios.") What these theorists do is to sketch a paradigm ("an explicitly structured set of assumptions, definitions, typologies, conjectures, analyses, and questions")42 and then construct a number of explicitly "alternative futures" which might come into being under stated conditions. Thus the alternative futures become guides to policy makers in sketching their own responses to the possible worlds that may emerge in the next decade. The writing of a "scenario" is not itself a prediction: it is an explication of possibilities. What it is, in effect, is the step beyond the "accounting scheme" in the construction of a number of plausible profiles, and the explication of the assumptions which underlie each of these alternatives. Herman Kahn, for example, has sketched a number of alternative world futures for the 1970s.43 He has constructed what he has called Alpha, Beta, Gamma, Delta worlds and indicated the kinds of international orders or equilibria which might obtain in each. Working from an "accounting scheme" of constants, relatively predictable sequential developments, constraints, and the like, he tries to assess current political factors (e.g. a degree of U.S.-Soviet detente, the strains in NATO, etc.), the present and future military technology, and possible political factors, and then sketch the alternative results. His Alpha world is one of arms control agreements; the U.S.S.R., Europe, Japan, and the U.S. relaxed and ideologically slack, China maintaining only a defensive military posture, etc. In a variant, Alpha-1, Japan has become the implicit guarantor of South Korea, Formosa, Malaysia and India, and Western Europe is united. In Alpha-2, there is a strong Franco-German Europe, and a rearmed Japan is tending to neutralism. But the basic condition is one of high stability and peace.

The Beta worlds are one of defined structural strains. For example: the U.S.S.R. is losing dominance over the world communist movement; Peking has a low-grade nuclear force; E.E.C. pursues moderately exclusionist trade policies, Japan, Germany and France have independent nuclear deterrents, the Tiers Monde develops hysterical and aggressive political movements, etc. The Gamma worlds, of extensive multipolarity, see the break up of the old alliances: China, France, Germany, Japan, and India develop or procure nuclear weapons, E.E.C. does not develop into a political community, middle-sized countries are on the verge of be coming nuclear powers because of the development of cheap fission (i.e. atom), though not yet expensive fusion (e.g. hydrogen) bombs. In these projections, few sophisticated techniques (game theory, systems analysis, cost-effectiveness ratios) are employed to sketch the future worlds. But what we do have is a systematic identification of relevant factors, and the combination of these, to create a coherent fiction or a set of alternative futures. And, to the extent that these alternative futures are realistic possibilities, one has a surer foundation for policy formulations to meet the various contingencies. A different kind of experiment in conjecture is attempted by the writer in a forthcoming book, The Post-Industrial Society. It is, in essence, not a forecast, but a dissection of the recent past – the "prophetic past," in Chesterton's phrase – in order to identify the structural trends and structural possibilities in the society and to create an "as if" about the future. The study deals with the new role of military technology as constitutive of political decisions, the rise of scientists as a new constituency in the political process, the creation of a new "intellectual technology" (a short-hand term I use for cybernetics, decision theory, simulation, and other intellectual techniques that allow us a new way of dealing with the planning process) and other elements of structural change in the society. These changes are then projected forty years as an "as if" in order to see their impact on the composition of the labor force, class structure, elite groups, and the like. Clearly that world, forty years hence, will not materialize, for there are many unforeseen and "uncontrollable" variables (particularly the political weights of the new nations and their own commitments) which will shape the reality of that time. But if the "as if," as a rational projection of structural possibilities in the society, does have some heuristic validity, by comparing that model with the reality, one might then be able to gain a clearer sense of the actual agencies of social change that were operative in that time.

12. Decision Theory

One includes in decision theory a wide assortment of new techniques: linear programming, utility preference theory, game theory, simulation, etc. Strictly speaking, decision theory is not predictive because it is normative: it seeks to specify probable outcomes if one or another choice is made. The problem, then, is its adequacy as a tool for the policy-maker. M. de Jouvenel has indicated one of the problems in game theory, the problem of agreement on relative values (the assignment of utilities).44 Any adequate discussion would have to go far beyond the length of this paper. The singular point to be noted, however, is that any concern with prediction eventually must explore in detail the formalization procedures of decision theory and seek to assess in what way they are useful in the art of conjecture. In the political realm? where one seeks to assess the intentions as well as capabilities of one's opponents? where does one begin? One technique worth exploring – though not strictly within mathematical decision theory – is "political gaming" or "simulation." In effect, these are political mock wars. One can, using the Rand technique, set up actual teams and allow them to work out a political game as armies work out a war game under simulated conditions of diplomatic negotiations, or one can, with a computer, simulate various situations and work out the alternative strategies and likely outcomes under hypothecated conditions. As with any formalization technique, the added knowledge comes from a specification of the likely variables and an awareness of the range of outcomes, rather than from new "wisdom".45 Most of the difficulties of politics?unlike probability situations, which are based on repeated experiments?is that decisions often have a "once-in-a-lifetime" consequence. A crucial question, there fore, is whether probabilistic methods can be used for political decision-making. In practice, one does this roughly by asking a number of "experts" and then weighing their advice. Is it possible, as L. J. Savage, for example, believes, that one can take expert opinion as a form of experiment and calculate a priori probabilities as a result?

But it is at this point that one runs the risk of the rationalist fallacy of believing that there is a true optimal path for any decision. One of the most heartening developments is the recognition of the "existentialist" element which has entered into modern utility theory, itself the foundation of so much of the work of rational prediction. It used to be that in choosing a strategy in a "game against nature" (i.e. uncontrolled situations), one could follow a maximin path (i.e. go for broke) or a minimax route (i.e. seek to cut one's losses). Now a third strategy has appeared: one can choose, depending on temperament and values, a maximin or a minimax throw, but then, statistically one can hedge one's bet by an added probability which has been termed so neatly, "the criterion of regret." Now clearly, the man who invented that has learned the lessons of love and politics.

II

Why does one seek to predict? This is an era in which society has become "future-oriented" in all dimensions: a government has to anticipate future problems; an enterprise has to plan for future needs; an individual is forced to think of long-range career choices. And all of these are regarded as possible of doing. The government of the United States, for example, makes national estimates in which the intentions and capabilities of opponents are evaluated, and policy is formulated on the basis of these short-run and long-run estimates. Business firms now make regular five-year budgets and even twenty-year projections to anticipate future capital needs, market and product changes, plant location, and the like. Individuals at an early age begin to consider occupational choices and plan for university and later career life. If none of us can wholly predict the future, what we do in these actions, in the felicitous phrase used by Dennis Gabor, is seek to "invent the future."46 In the light of all this, what is remarkable is how little effort has been made, intellectually, to deal with the problems of conjecture. In few of the cases are there genuine predictions of the order: these are the changes that I think will take place; these are the reasons why I think these changes will occur, etc. Most of the works analyzed above seek to specify the problems which countries confront, but only rarely is the further effort made such as: if the problems are solved this way, then the following might happen; or: these are the probabilities that the problem will be handled in this fashion. Yet one should not derogate such efforts. The correct identification of relevant problems is obviously the first step in the conjecture about the future; it is easier to make because it tends to be an extrapolation of the present.

I would like to put forth some problem areas for future investigation which, on the basis of the works surveyed, seem most promising. These are simply some suggestions, necessarily brief, of the productive leads derived from the works analyzed. (Necessarily, too, I cannot within the limits of space and competence detail the scope of these proposals; what I can do is to argue briefly their rationale.

1. The Planning Process

A future-oriented society necessarily commits itself more and more to the idea of planning. This is the chief means of inventing the future. Most of the new states that have come onto the world scene in the last decade have ambitious planning schemes; most of the older societies, to some degree or other, are engaged in planning. One plans, of course, for different ends; one plans in different ways (from centralized administrative to "indicative" planning); one uses different techniques (input-output schemes, systems analysis, shadow prices, simulation). One plans proportions between economic sectors; one does physical planning, as in the layout of cities; one plans for "guided mobility," i.e. the planned transfer from farms to cities. In all these instances, there is an attempt to direct human actions with different kinds of coercions, manipulations, persuasions and cooperations. Can we, with full awareness of the problem of choosing between conflicting values, each of which may be cherished, find some way of choosing the best planning process that is consonant with our be lief in liberty? The function of planning is not only to set forth goals and alternatives and means of achieving these. Equally important, and usually neglected, are the specification of costs and benefits, the reallocation of burdens, and the probable consequences of different kinds of actions. The true function of the planning process is not to designate the most appropriate means for given ends, but to predict the possible consequences to explicate the values of and make people aware of the costs of achieving these.47

Surprisingly, there are few studies extant of the planning process. There are some theoretical studies of how nations should plan and the principles of city planning,48 but few critical studies of how nations and groups actually plan and what can be done to improve both methods and procedures.

2.The Standardization of Social Indicators

Over the past decades, economists have developed different series of indicators to anticipate and evaluate trends in the econ omy. There may be differences in the conclusions drawn from time-series, but by and large there is a consensus of what should be observed. One need not recapitulate here the obvious difficulties in establishing social and political indicators. Some of the difficulty arises from a failure to agree on what should be observed. Most of the writing on the new states, for example, concentrates on such general concepts as "modernization" or "political development" or "new elites," the dimensions, let alone the indicators, of the concept are still to be formulated.49 But one of the consistent themes of the recent writings in modern sociology (cf. Aron, Par sons) is that "industrial society" produces a series of common effects, has an "internal logic" in its creation of a new occupational structure, and with the rise of affluence creates new, presumably common attitudes, despite differences in traditional culture. And the work of Inkeles, Lerner, Pye, and others indicates that the break with tradition, the new patterns of urbanization, the exposure to the mass media and education all tend to create common patterns of thought in the new states. The function of indicators is not to replace analysis or to act as predictors, but to allow comparisons over time within a country, and between countries, and more important, to allow one to anticipate certain likely occurrences.50 It is only when the indicators and the concepts are relatively precise that we could hope that indicators would be predictive of specific events. In rough ways, we tend to use certain indicators as predictive of events. We say that rising unemployment rates presage a swing to radical groups in voting, that migration rates may be coupled with crime and divorce rates, etc. But clearly the present need is for some coherent effort to create sets of social indicators dealing with social and political change.

Some beginning attempts have already been made. Some studies have shown that in the early stages of economic development demo graphic indicators (mortality data particularly) are quite predictive of rates of economic growth.51 Daniel Lerner, in his study of the Middle East, has used such items as literacy, exposure to mass media, and urbanization as indicators of modernization.52 In an ambitious effort, Karl Deutsch has sought to create an inventory of basic trends in international politics, and to set up some indicators for social mobilization. (By social mobilization, Deutsch specifies a cluster of variables to indicate the entry of persons into a political system.)53 Almond and his students have sought through opinion polls and other devices to trace the process whereby individuals in a number of countries move from being "subjects" to be coming "citizens," and to chart some rate of absorption.54 Alex Inkeles and his associates are now engaged in an effort to trace the impact of industrialization on peasants entering the industrial process in a variety of different countries and to set up indicators of attitude change.55 Clearly any useful construction of alternative futures will de pend upon the success of the compilation of such indicators.

3.Models of Political Structures

The most difficult of all proposals is the heart of the social science enterprise itself, the construction of models of political systems. The creation of a model allows us to do two (of a large number of) things. It may allow us to understand the "value-relevance" of a statement; it may allow us to see whether a predicted change is one which simply affects the actors in a system (e.g. a shift of power between groups), or affects the nature of a system itself. To illustrate the two points:

1) From the standpoint of a Mohammedan, all Christians are alike. The involved theological disputes between Catholics and Protestants may have little meaning for him because both are "children of Jesus." To a Catholic, the differences between a "hard-shell Baptist" and a Quaker may have little relevance for him since both are "enthusiasts." In a similar fashion, to a confirmed Marxist, the difference between a Democrat and a Republican in the American political system has little meaning since both are capitalists. But to understand any analysis or criticism, one has to identify the standpoint from which it is made. One function of a model, therefore, is to indicate the value-relevance and level of analysis from the standpoint of the observer.

2) The analysis of power in a society can only be carried on adequately if one has a scheme to identify the relevant actors, the arena, the orientations of the actors, and their relationship fo the underlying system which defines the politics of the society. By a system, I mean here the basis of renewable power independent of any momentary group of actors. Most of political analysis today, I would argue, concentrates on the "intermediate" sectors (e.g. parties, interest groups, the formal structure) or, as in the case of Soviet politics, through Kremlinology to deal with the "small units" (what I have called only half-jokingly the "small c's") of politics,56 but rarely is there an attempt to specify, as a Marxist analysis does, the underlying system of renewable power.

There are few operating models of political systems, on the descriptive or the analytical level, extant. Gabriel Almond and his associates have sought to establish a framework of concepts which would lead to the creation of such a model. Many years earlier, Harold Lasswell and Abraham Kaplan set forth comprehensive definitions of power, but they did not seek to combine these into a system. C. Wright Mills created a mechanistic image of a "power elite." Maurice Duverger, in his book Political Parties, at a lesser level has put forth a useful typology. Most recently, Raymond Aron, in his magisterial book, Paix et Guerre entre les Nations, has formulated certain models of diplomatic systems. But we still lack any comprehensive analyses of different systems of power.57

But here, in this entreaty, one comes full circle. For in the preoccupation with prediction one risks the hubris of the historicist mode of thought which sees the future as "pre-viewed" in some "cunning of reason" or other determinist vision of human affairs. And this is false. One seeks "pre-vision" as much to "halt" a future as help it to come into being, for the function of prediction is not, as often stated, to aid social control, but to widen the spheres of moral choice. Without that normative commitment the social sciences become a mere technology rather than humanistic discipline.

Notes

1. The Positive Philosophy of August Comte, ed. Harriet Martineau (New York: Calvin Blanchard, 1856), pp. 132-133.
2. See The Republic, Book IX, No. 571, in The Dialogues of Plato, Jowett edition (New York: Random House, 1937), p. 829.
3. As Michael Polanyi has written: "Prediction is not a regular attribute of scientific propositions. Kepler's laws and Darwinian theory predicted nothing. At any rate, successful prediction does not fundamentally change the status of a scientific proposition. It only adds a number of observations, the predicted observations, to our series of measurements. . . ." See The Logic of Liberty (London: Routledge, 1951), p. 16.
4. Thus Arnold Toynbee, in tracing western history back to 1494, identifies five great war-and-peace cycles, but one is confounded by the question of what is war (subversion, guerilla forays, rebellions) and what is peace.
5. For one of the most succinct analyses from the basis of modern economic theory, see Paul A. Samuelson, "Wages and Interest: A Modern Dissection of Marxian Economic Models," American Economic Review (December 1957), pp. 884-912.
6. Ridenour's discussion, one of the most provocative of recent times, can be found in Ridenour, Shaw and Hill, Bibliography in an Age of Science (Urbana: University of Illinois Press, 1951). Derek Price's analyses can be found in two small volumes, Science Since Babylon (New Haven: Yale University Press, 1961) and Little Science, Big Science (New York: Columbia University Press, 1963). A seminal source for many of these discussions is the work of Gerald Holton, particularly "Scientific Research and Scholarship: Notes Towards the Design of Proper Scales," Daedalus, (March 1962), pp. 362-399. 7. These quotations are from Little Science, Big Science, pp. 24, 25, 30. 8. See Herbert A. Simon, Models of Man (New York: Wiley, 1957), pp. 97, 145-164. 9. The following description by Paul Levy explains the concept: "The idea of a stochastic process is, at least for a determinist, tied to that of the existence of hidden parameters which do not intervene in the description of the apparent present state of the system studied and which nevertheless influence its future evolution. Our ignorance of their values forces us to speak for the future only of a set of possible evolutions, and in certain cases we can define in that set a law of probability incessantly modified by the knowledge of new data." Cited in Luciene F?lix, The Modern Aspect of Mathematics (New York: Science Editions, 1961), pp. 132-133. 10. Published by The University of Michigan Press in 1960. 11. A representative work is Amos Hawley, Human Ecology (New York: Ronald Press Co., 1950). For a formal treatment of systems of local communities, regions, and societies, see Rutledge Vining, "A Description of Certain Spatial Aspects of an Economic System," Economic Development and Cultural Change, III (January 1955), pp. 147-195. A recent, imaginative effort to use the ecological approach in a holistic way is Clifford Geertz's Agricultural Involution: The Processes of Ecological Change in Indonesia (Berkeley: The University of California Press, 1963).
7. Recently the U. S. Congressional Committee on the Joint Economic Report conducted hearings on the methods employed by different economic forecasters (e.g. Fortune, McGraw-Hill, and others) to make their forecasts. I leave aside here the problem of the failures of short-run predictions or stages theory as involving many complicated considerations. But a study of why many of these predictions went wrong would be fruitful. For example, Frank Notestein posited a stages theory which powerfully influenced David Riesman in the formulation of the theory of character change in the latter's The Lonely Crowd. The Notestein theory, metaphorically, is an S-shaped curve. The bottom horizontal line represents traditional societies with high birth rates but also high death rates; a second stage, that of transitional growth (the sigmoid rise), is one of population explo sion because of a rapidly declining death rate; the third stage is one of incipient population decline based on a new stabilization of low birth rate and low death rate. Riesman associated the traditional type with the first society, the inner-directed type with the second, and the other-directed type with the third. But the stages theory has been proven wrong. Both France and the United States, for example, have recently resumed a steady population rise, but for how long, it is difficult to say. For the original Notestein formulation, see Frank Notestein, "Population – The Long View," in Food for the World, d. T. W. Schultz (Chicago: University of Chicago Press, 1945 ). For Riesman's discussion of his abandonment of the Notestein hypothesis as the basis for his own theory see "The Lonely Crowd: A Reconsideration in 1960," in Culture and Social Character, eds. S. M. Lipset and Leo Lowenthal (Glencoe: The Free Press, 1961), pp. 419-458.
8. See Herbert A. Simon, Models of Man (New York: Wiley, 1957), pp.97, 145-164.
9. The following description by Paul Levy explains the concept: "The idea a stochastic process is, at least for a determinist, tied to that of the existence of hidden parameters which do not intervene in the description the apparent present state of the system studied and which nevertheless influence its future evolution. Our ignorance of their values forces us o speak for the future only of a set of possible evolutions, and in certain ases we can define in that set a law of probability incessantly modified y the knowledge of new data." Cited in Luciene Félix, The Modern Aspect of Mathematics (New York: Science Editions, 1961), pp. 132-133.
10. Published by The University of Michigan Press in 1960.
11. A representative work is Amos Hawley, Human Ecology (New York: Ronald Press Co., 1950). For a formal treatment of systems of local communities, regions, and societies, see Rutledge Vining, "A Description of Certain Spatial Aspects of an Economic System," Economic Development and Cultural Change, III (January 1955), pp. 147-195. A recent, imaginative effort to use the ecological approach in a holistic way is Clifford Geertz's Agricultural Involution: The Processes of Ecological Change in Indonesia (Berkeley: The University of California Press, 1963).
12. Recently the U. S. Congressional Committee on the Joint Economic Report conducted hearings on the methods employed by different economic forecasters (e.g. Fortune, McGraw-Hill, and others) to make their forecasts.
13. I leave aside here the problem of the failures of short-run predictions or stages theory as involving many complicated considerations. But a study of why many of these predictions went wrong would be fruitful. For example, Frank Notestein posited a stages theory which powerfully influenced David Riesman in the formulation of the theory of character change in the latter's The Lonely Crowd. The Notestein theory, etaphorically, is an S-shaped curve. The bottom horizontal line represents traditional societies with high birth rates but also high death rates; a second stage, that of transitional growth (the sigmoid rise), is one of population explosion because of a rapidly declining death rate; the third stage is one of incipient population decline based on a new stabilization of low birth rate and low death rate. Riesman associated the traditional type with the first society, the inner-directed type with the second, and the other-directed type with the third. But the stages theory has been proven wrong. Both France and the United States, for example, have recently resumed a steady population rise, but for how long, it is difficult to say. For the original Notestein formulation, see Frank Notestein, "Population – The Long View," in Food for the World, ed. T. W. Schultz (Chicago: University of Chicago Press, 1945 ). For Riesman's discussion of his abandonment of the Notestein hypothesis as the basis for his own theory see "The Lonely Crowd: A Reconsideration in 1960," in Culture and Social Character, eds. S. M. Lipset and Leo Lowenthal (Glencoe: The Free Press, 1961), pp. 419-458.
14. See Sir Charles Darwin, "Forecasting the Future," in Frontiers in Science, ed. Edward Hutchings, Jr. (New York: Basic Books, 1958), p. 116. See, too, in the same volume, the retort by Fred Hoyle. 17. Colin Clark, "Apropos the Eight Million of M. Manshold," Futuribles papers, Paris, no. 18.
15. For a balanced assessment of this problem, see S. Lilley, "Can Prediction Become a Science?" in Discovery (November 1946), reprinted in The Sociology of Science, ed. Bernard Barber and Walter Hirsch (Glencoe: The Free Press, 1962). A quick review of different kinds of trend predic tions can be found in Hornell Hart, "Predicting Future Trends," in Tech nology and Social Change, ed. Allen, Hart, et al. (New York: Appleton Century-Crofts, 1957), ch. 19.
16. Of the vast literature on the subject, two reviews are of use: G. Dupeux, "Electoral Behaviour," Current Sociology (1954-1955), pp. 318-344, and S. M. Lipset, P. F. Lazarsfeld, et al, "The Psychology of Voting," in Hand book of Social Psychology, ed. Gardner Lindzey (Cambridge: Addison Wesley, 1954). The discussion of "bandwagon effects" in relation to election predictions can be found in Simon, op. cit., ch. 5; the discussion of process analysis in Berelson, Lazarsfeld, and McPhee, Voting (Chicago: University of Chicago Press, 1954), especially ch. 13.
17. Colin Clark, "Apropos the Eight Million of M. Manshold," Futuribles papers, Paris, no. 18.
18. A classic essay in this field is Robert Merton's "Social Structure and Anomie" which seeks to specify under what conditions deviations from such norms or structural certainties obtain. See Social Theory and Social Structure (Glencoe: The Free Press, 1957), ch. IV.
19.. See The Operational Code of the Politburo (New York: McGraw-Hill, 1951); A Study of Bolshevism (Glencoe: The Free Press, 1954); also, Nathan Leites with Constantin Melnik, The House Without Windows: France Selects a President (Evaston: Row Peterson, 1958), and On the Game of Politics in France (Stanford: Stanford, 1959). For a detailed dis cussion of the validity of this approach, see my essay "Ten Theories in Search of Reality: The Prediction of Soviet Behavior," in The End of Ideology (Glencoe: The Free Press, 1960).
20. There have been many efforts to identify an American "style." See, for example, the essay by W. W. Rostow in The American Style, ed. Elting Morison (New York: Harper, 1958), pp. 246-313; D. W. Brogan, "The Illusion of American Omnipotence," Harpers (December 1952); Daniel Bell, "The National Style and the Radical Right," Partisan Review (Fall 1962); Talcott Parsons and Winston White, "The Link Between Character and Society," in Culture and Social Character (Glencoe: The Free Press, 1961), ch. 6. A fruitful comparison of aspects of British and American patterns can be found in Edward A. Shils, The Torment of Secrecy ( Glen coe: The Free Press), ch. 2. For a denial of direct links between cultural values and social structure, see E. R. Leach, The Political System of High land Burma (London: School of Economics and Political Science, 1954).
21. In Karl Marx, Selected Work, (Moscow: Foreign Languages Publishing House, 1935) II. 22. It is difficult to resist Marx's description, an extraordinary piece of political analysis and poetic rhetoric: "This society dates from the year 1849. On the pretext of founding a benevolent society, the Lumpenproletariat of Paris had been organized into secret sections, each section led by Bonapartist agents, with a Bonapartist general at the head of the whole. Along side decayed roués with doubtful means of subsistence and of doubtful origin, alongside ruined and adventurous offshoots of the bourgeoisie, were vagabonds, discharged soldiers, discharged jail birds, escaped galley-slaves, swindlers, mountebanks, lazzaroni, pickpockets, tricksters, gamblers, mcquereaux, brothel keepers, porters, literati, organ-grinders, rag-pickers, knife-grinders, tinkers, beggars, in short the whole indefinite, disintegrated mass thrown hither and thither which the French term la Bohème; from this kindred element, Bonaparte formed the basis of the Society of December 10. A 'benevolent society' – in so far as, like Bonaparte, all of its members felt the need of benefiting themselves at the expense of the working nation." And more of the same! Ibid., pp. 369-370. 23. Ibid., p. 423. Emphasis added.
22. It is difficult to resist Marx's description, an extraordinary piece of political analysis and poetic rhetoric: "This society dates from the year 1849. On the pretext of founding a benevolent society, the Lumpenproletariat of Paris had been organized into secret sections, each section led by Bonapartist agents, with a Bonapartist general at the head of the whole. Along side decayed roués with doubtful means of subsistence and of doubtful origin, alongside ruined and adventurous offshoots of the bourgeoisie, were vagabonds, discharged soldiers, discharged jail-birds, escaped galley-slaves, swindlers, mountebanks, lazzaroni, pickpockets, tricksters, gamblers, mc quereaux, brothel keepers, porters, literati, organ-grinders, rag-pickers, knife-grinders, tinkers, beggars, in short the whole indefinite, disintegrated mass thrown hither and thither which the French term la Bohème; from this kindred element, Bonaparte formed the basis of the Society of December 10. A *benevolent society'? in so far as, like Bonaparte, all of its members felt the need of benefiting themselves at the expense of the working nation." And more of the same! Ibid., pp. 369-370.
23. Ibid., p. 423. Emphasis added.
24. The changes in the features of modern capitalist society is the central point of Professor Postan's essay, "The Economic and Social System in 1970," Futuribles, I, ed. Bertrand de Jouvenel (Geneva: Droz, 1963). In concentrating on the changing character of ownership of capital, he is tracing out a major reorganization of the "operational system."
25. See Almond and Coleman, The Politics of the Developing Areas (Prince ton: Princeton University Press, 1960), p. 59. The introduction by Almond, "A Functional Approach to Comparative Politics," proposes a set of definitions for the "common properties of political systems." A more condensed effort, but almost as comprehensive, is the article by David Apter, "A Comparative Method for the Study of Politics," American lournal of Sociology, LXIV, 3 (November 1958), pp. 221-237. One of the earliest and still one of the most interesting efforts to specify the common requisites of any social system is the essay by Aberle, Cohen, et al., "The Functional Prerequisites of a Society," Ethics, LX, 2 (January 1950) 100-111.
26, One such typology, on which Almond bases part of his system, is that of Professor Edward Shils, from Political Development in the New States (The Hague: Mouton and Company, 1962).
27. "In many technologically indifferent features the range of variability is unlimited: but the requirements of effective use impose limitations, more or less stringent, on the variability of other features – there is a narrowing of the range if the object is to function as intended. . . . What is so patent in technology applies equally to other aspects of culture, where, however, the operation of the principle is less readily discernible." Alexander Goldenweiser, "The Principle of Limited Possibilities in the Development of Culture," in History, Psychology and Culture (New York: A. A. Knopf, 1933), p. 47..
28. See Terror and Progress USSR (Cambridge: Harvard University Press, 1954). 29. Edward F. Denison, The Sources of Economic Growth in the U. S., Committee for Economic Development, 1963
29. Edward F. Denison, The Sources of Economic Growth in the U. S., Committee for Economic Development, 1963.
30. In Futuribles, I (Geneva: Droz, 1963), pp. 295-319.
31. Lucien Pye, Politics, Personality and Nation-Building: Burma's Search for Identity (Princeton: Princeton University Press, 1961).
32. Lucy Mair, New Nations (Chicago: University of Chicago Press, 1963).
33. See K. D. Bracher, Die Auflesung der Weimar Republik (Stuttgart: Ring Verlag, 1957).
34. Herman Kahn, On Thermonuclear War (Princeton: Princeton University Press, 1960), especially chs. IX, X, and "Deterrence and Defense in the Sixties and Seventies," Progress Report for a Study of Crises and Arms Control, Hudson Institute (unpublished), ch. II. 35. Apter, op. cit., p. 221.
35. Apter, op. cit., p. 221.
36. This perspective is central to Professor T. H. Marshall's Citizenship and Social Class (Cambridge: Cambridge University Press, 1950). It has been used by S. M. Lipset to account for variations in the degree of radicalism in western societies (see Political Man, [New York: Doubleday, 1960] chs. 2, 3 ) and by Reinhard Bendix and Stein Rokkan to create comparative models of political development in the West. See "The Extension of Na tional Citizenship to the Lower Classes: A Comparative Perspective," paper for the Fifth World Congress of Sociology, September, 1962.
37. See Talcott Parsons, Structure and Process in Modern Societies (Glencoe: The Free Press, 1960), especially chs. 2 and 3.
38. In an unpublished paper, William Kornhauser, a sociologist at Berkeley, has sought to identify four stages of "political development" which new states have to consolidate in order to achieve "political competence" or the achievement of their defined goals. These four are: national identity, integration of authority, participation or mobilization of the population, constitutionalism. Such tags are largely amorphous in themselves, and with out a comprehensive description of each of the concepts one risks a disservice to Professor Kornhauser. I call attention to it here to indicate that serious thinking about the nature of "sequential development" is taking place among the students of the "new societies." One should point out that M. de Jouvenel's essay on The Evolution of the Forms of Government, though reflecting on past history, does have, in its sketch of the rise and extension of bureaucracy, the basis of a sequential theory. One might say, too, that the concept of "value added," which M. de Jouvenel has called attention to from economic theory, could be the basis of social speculation. One such effort to explain "collective behavior" and particularly revolution is Neil Smelser, Theory of Collective Behavior (Glencoe: The Free Press, 1963). Professor Smelser writes: "As the value-added process develops, it allows for progressively fewer outcomes other than the one we wish to explain. This logic of value added can be applied to episodes of collective behavior such as the panic or the reform movement. Many determinants or necessary conditions must be present for any kind of collective episode to occur. These determinants must combine, however, in a definite pattern. Furthermore, as they combine, the determination of the type of episode in question becomes increasingly specific and alternative behaviours are ruled out as possibilities" (p. 14).
39. In using the phrase "accounting scheme," I am following the conventional language of bookkeeping (comptabilite). However, a more restricted use of the phrase is employed by Paul F. Lazarsfeld to deal with causal in ferences of single decisions. He is attempting, in effect, a methodology for the historian, the clinical psychologist, or the market researcher, who has to account for the most relevant combination of factors in the making of a single decision. For a short statement of Professor Lazarsfeld's intentions, see Lazarsfeld and Rosenberg, The Language of Social Research (Glencoe: The Free Press, 1955), pp. 387-391. An elaboration of this mode of "reason analysis" can be found in Hans Zeisel's Say It With Figures (New York: Harper, 1957), ch. 6.
40. In Futuribles, L, op. cit., pp. 121-154. 41. C. K. Ogden, Bentham*s Theory of Fictions (London: Kegan, Paul, 1932); Hans Vahinger, The Philosophy of "As if (London: Kegan, Paul, un dated).
41. C. K. Ogden, Bentham's Theory of Fictions (London: Kegan, Paul, 1932); Hans Vahinger, The Philosophy of "As if” (London: Kegan, Paul, undated).
42. For the role of paradigms in social research, see R. K. Merton, Social Theory and Social Structure (Glencoe: The Free Press, 1957), pp. 13-16, 50-60.
43. Herman Kahn, "Alternative World Futures," paper HI-342-B IV, Hudson Institute, April, 1964.
44. See Essai sur l'Art de la Conjecture, (Paris: SEDEIS), pp. 92-94.
45. On the techniques of political gaming, see Herbert Goldhamer and Hans Speier, "Some Observations on Political Gaming," World Politics ( October 1959). Harold Guetzkow has edited two volumes on simulation: Simula tion in Social Science: Readings; and Simulation in International Relations, both published by Prentice-Hall, 1962, 1963. A useful volume on decision theory which argues that one can verify ethical judgments is C. West Churchman, Prediction and Optimal Decision (Englewood Cliffs: Prentice Hall, 1961).
46. When individuals and, more particularly, organized movements are committed to rigid invention such as Bolshevik theory, the results can be disastrous. One of the sorriest records of prediction is that of the Bolsheviks. In the 1920s they predicted that the next struggle would be between Great Britain and the United States (as the two largest imperial powers); in the 1930s, that Hitler could not last and that the Communists would succeed him; in the 1940s that the West could not go to war against Hitler; in the early 1950s, that the western economies would stagnate, etc.
47. One of the most interesting studies in recent years, in this regard, is Alexander Erlich's The Soviet Industrialization Debate, 1924-1928 (Cambridge: Harvard University Press, 1960). The reconstruction of this debate makes it clear that the Soviet path was not "inevitable" and that various persons, including those who argued for a course of intensive industrialization, were well aware of the risks and consequences, including the resistance of the peasantry.
48. There is a large and useful literature on planned "organizational change," growing out of the work of such groups as the Tavistock Institute in Lon don, the Institute of Social Research at the University of Michigan, and similar groups.
49. For a useful discussion on the general nature of indicators in relation to concepts, see Paul F. Lazarsfeld, "Evidence and Inference in Social Re search" in Evidence and Inference, ed. Daniel Lerner (Glencoe: The Free Press, 1959), 31 pp.
50. For an enlightening discussion of the need for indicators and their role in anticipations, see the document "Space Efforts and Society," a Document of the Committee on Space Efforts and the Society of the American Academy of Arts and Sciences (Boston, January 1963). In this context, the term anticipation is broader than prediction. Say the authors: "The distinction we would make is that between a reasonably precise knowledge of the probability of a given event happening and a sufficient awareness that something of a general sort might happen, that if it does happen we are not caught by surprise. In planning it is often possible to make provision for handling some consequence even though it can be foreseen only in very general terms." There is another distinction worth emphasizing. Prediction places primary emphasis on the probability of something's happening and is less concerned with the importance of events. Anticipation seeks to weigh the relative importance of likely occurrences.
51. For a representative effort along this line, see Philip Huser, "Demographic Indicators of Economic Development," Economic Development and Cultural Change, Vol. 7 (January 1959).
52. See Daniel Lerner, The Passing of Traditional Society (Glencoe: The Free Press, 1958), especially ch. II.
53. See Karl Deutsch, "Toward an Inventory of Basic Trends and Patterns in Comparative and International Politics," American Political Science Re view (March 1960); and "Social Mobilization and Political Development," Ibid. (September 1961), pp. 463-515.
54. The results of these studies in Germany, Italy, Mexico, Great Britain, and the United States are summed up in the volume The Civic Culture, by Gabriel Almond and Sidney Verba (Princeton: Princeton University Press, 1963).
55. This work is going on in Chile, Nigeria and Pakistan. A preliminary statement can be found in Alex Inkeles, "Industrial Man: The Relation of Status to Experience, Perception and Value," The American Journal of Sociology, LXVI, 1 (July 1960), 1-31.
56. One of the things that has struck me over the years is the way small groups of individuals who have, at an early age, formed a circle or club, move out into the larger political arena. This is a familiar phenomenon in literary or art movements; yet it is equally true in politics. One thinks of the Galileo circle in Hungary in 1919 (for a description, see the essay by Paul Ignotus in the Essays in Honor of Michael Polayni), of the Thakins in Burma, of the "young Turks," who have thus given a cognomen to the phenomenon, etc. While the phenomenon is readily understood as important for political analysis, one finds few studies of the process at all. In thinking about these groups, I was struck by the linguistic singularity (in English) that most of the words describing them begin with a small "c." Thus:

clique club circle cabal cadre crowd clan camarilla claque council curia church cortes cells caucus condomin- c?nacle ium committee conspiracies condottieri college coteries cults

All of which may well be one other illustration of a "Yule distribution" statistically in linguistics.
57. The best recent effort to create a typology of the new States is the monograph by Edward Shils, Political Development in the New States. This study charts the alternative courses of political development (e.g. tutelary democracy, modernizing oligarchies), but does not specify the bases of power, or the social systems of renewable power. The beautiful simplicity of the Marxist scheme is that it took a single variable, wealth in property, or the means of production, as the basis of power. We lack some organizing concepts which will allow us to group the multiple bases of power which now exist in the world into a comprehensive classification, or a system-set.

The following article from the Asahi Shimbun of 2011.2.10 relates Daniel Bell's prophetic role (in Japanese):