The Theory of Democracy Revisited, Part One: The Contemporary Debate [Paperback ed.] 0934540470, 9780934540476

This is part one of a two-volume set. "...[Sartori] reviews the major democratic theories of our time and canvasse

346 50 26MB

English Pages 253 [271] Year 1987

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

The Theory of Democracy Revisited, Part One: The Contemporary Debate [Paperback ed.]
 0934540470, 9780934540476

Citation preview

71 HElllHI EO~Y OF DEM\OCRACY REVSllED

OTHER BOOKS BY GIOVANNI SARTORI

Democrazia e Definizioni (1957) Democratic Theory (1962) II Parlamento Italiano (1963, ed.) Stato e Politica nel Pensiero di B. Croce (1966) Antologia di Scienza Politica (1970, ed.) Tower of Babel (1975, ed.) Parties and Party Systems (1976) La Politica: Logica e Metodo in Scienze Sociali (1979) Teoria dei Partiti e Caso Italiano (1982) Social Science Concepts: A Systematic Analysis (1984, ed.) Elementi di Politica (1987)

The Theory of Democracy Revisited GIOVANNI SARTORI

Albert Schweitzer Professor in the Humanities Columbia University

CHATHAM HOUSE PUBLISHERS, INC.

Chatham, New Jersey

THE THEORY OF DEMOCRACY REVISITED CHATHAM HousE PUBLISHERS, INC. Box One, Chatham, New Jersey 07928 Copyright© 1987 by Chatham House _Publishers, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher.

Publisher: Edward Artinian Design: Quentin Fiore Composition: Chatham Composer Printing and Binding: Hamilton Printing Company LIBRARY OF CONGRESS CATALOGING IN PUBLICATION DATA Sartori, Giovanni, 1924The theory of democracy revisited. Includes indexes. r. Democracy. I. JC423.s274 1987 ISBN 0-934540-46-2 ISBN 0-934540-49-7 (pbk. ISBN 0-934540-47-0 (pbk. ISBN 0-934540-48-9 (pbk.

Title. 32r.8

86-3ro13

: set) : v. 1) : v. 2)

Manufactured in the United States of America

10 9 8 7 6 5 4 3 2 1

Contents PART ONE: THE CONTEMPORARY DEBATE 1. Can Democracy Be Just Anything? I.I I.2 r.3 r.4 r.5

The Age of Confused Democracy Description and Prescription .7 Political and Other Democracies 12 Outline A Coda on Theory 15

3

3 8

2. Etymological Democracy 2.1 2.2 2.3 2.4 2.5

The Meaning of People 21 The People in Mass Society 25 Power of the People and Power over the People Limited Majority Rule 31 The Lincoln Formula 34

2I

28

3. The Limits of Political Realism 3.1 3.2 3.3 3.4 3.5 3.6

What Is Pure Politics? 39 Warlike versus Peacelike Politics 41 Facts and Values in Benedetto Croce 44 Mosca, Pareto, and Michels 46 Realism versus Rationalism 48 Rational Democracy and Empirical Democracy

39

51

4. Perfectionism and Utopia 4.1 4.2 4.3 4.4 4.5 4.6 4. 7

58

The Misunderstanding of Deontology 58 Myth and Utopia Reconsidered 60 Self-Government and the Politically Impossible 64 The Role of Ideals 67 Maximization, Opposite Danger, and Inverted Results The Revolution as Myth 72 Ideals and Evidence 77 V

THE THEORY OF DEMOCRACY REVISITED

5. Governed Democracy and Governing Democracy 5.1 5.2 5.3

5:4 ,5.5 5.6 5.7 5.8

6. Vertical Democracy 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9

86

Public Opinion and Government by Consent 86 The Issue of Consensus 89 The Formation of Opinions 92 Autonomy versus Heteronomy of Public Opinion 96 Electoral Democracy 102 Participatory Democracy III Referendum Democracy and Knowledge u5 Government and Ungovernability 120 131

Majority Principle and Minority Rule 131 The Tyranny of the Majority 133 Election, Selection, and Mal-Selection 137 Minorities and Elites 141 Minority Rule: From Mosca to Dahl 145 The Iron Law of Oligarchy 148 The Competitive Theory of Democracy 152 Anti-Elitism Revisited 156 Polyarchy Defined Normatively 163

7. What Democracy Is Not 7.1 7 .2 7.3 7.4 7.5

Contraries, Contradictories, and Degrees Authoritarianism, Authority, and Power Total State, Democracy, and Absolutism Totalitarianism 193 Dictatorship and Autocracy 203

182 185 190

8. A Decision-Making Theory of Democracy 8.1 8.2 8.3 8.4 8.5 8.6 8. 7 8.8

The Nature of Political Decisions 214 External Risks and Decisional Costs 216 Outcomes and Decisional Contexts 223 Intensity of Preference and Majority Rule 225 Committees and Unanimity 227 Committees, Participation, and Demo-Distribution Consociational Democracy 238 A Coda on the Cost of Idealism 240

vi

214

232

Contents

PART TWO: THE CLASSICAL ISSUES 9. What Is Democracy? Definition, Proof, and Preference

2 57

9.1 9.2 9.3 9.4 9.5

Are Definitions Arbitrary? 257 A Criticism of Conventionalism 260 Words as Experience Carriers 265 267 The Search for Proof A Comparative Evaluation 271

ro.r 10.2 10.3 ro.4 ro.5

Homonymy, Not Homology 278 Direct or Polis Democracy 280 Individualism and Freedom: Old and New The Modern Idea and Ideal 287 A Reversal of Perspectives 290

rr.r u.2 rr.3 rr.4 rr.5 u.6 rr.7

Freedom and Freedoms 298 Political Freedom 301 Liberal Freedom 3.06 The Supremacy of Law in Rousseau 310 Autonomy: A Criticism 315 The Principle of Diminishing Consequences 320 From the Rule of Law to the Rule of Legislators 321

12.r 12.2 12.3 12.4 12.5 12.6 12. 7

A Protest Ideal 337 Justice and Sameness 338 Predemocratic and Democratic Equalities 341 Equal Opportunities and Equal Circumstances Egalitarian Criteria, Treatments, and Outcomes The Maximization of Equality 352 Liberty and Equality 357

10. Greek Democracy and Modern Democracy

2

78

284

11. Liberty and Law

12. Equality

337

344 347

13. Liberism, Liberalism, and Democracy 13.r 13.2 13.3 13.4

Overlaps 367 An Unfortunate Timing 370 Property and Possessive Individualism Liberalism Defined 379

367

376

vu

THE THEORY OF DEMOCRACY REVISITED

13.5 Liberal Democracy 383 13.6 Democracy within Liberalism 13.7 Democracy without Liberalism

386 390

14. Market, Capitalism, Planning, and Technocracy 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8

What Is Planning? 399 What Is the Market? 405 Capitalism, Individualism, Collectivism Market Socialism 417 Democratic Planning 425 Democracy, Power, and Incompetence The Role of the Expert 431 The Government of Science 434

399 410

428

15. Another Democracy? 15.1 15.2 15·3 15·4 15·5 15.6

16. The Poverty of Ideology 16.1 16.2 16.3 16.4 16.5

45°

The Good Society of Rousseau and Marx 450 Democracy and the State in Marx and Lenin 456 Popular Democracy 467 The Theory of Democratic Dictatorship 47° Democracy and Demophily 474 The War of Words 479

49 1

The Exhaustion of Ideals 491 Inevitables and Evitables 497 The Witch-Hunting of Ideas 498 Novitism and Beyondism 5o3 Epilogue 506

Name Index

5n

Subject Index

52 3

viii

Introduction Our ideas are our spectacles. -Alain

Some twenty-five years ago, I published a book on democracy that, in its American translation, carried the title Democratic Theory. The book did well and is still in print in a number of countries. Why, then, write another book on the same subject? Has the theorizing on democracy changed all that much? In particular, have "new" theories of democracy emerged in the meantime? An impressive number of authors have advanced such claims in the r96os and r97os, and even though the claims seldom withstand scrutiny, they do warrant a revisit. In order to assess how the present-day theory of democracy relates to the one in the r95os, this work largely incorporates the earlier one. Even the incorporated part is almost entirely reformulated, however. Why is that? One change's, I suppose, in response to changes-such as the ones that I am about to mention. To begin with, a pervasive change has occurred in the vocabulary of politics. To an unprecedented extent, authors have come to conceive their concepts at whim. This development has been legitimized by the brave new thought that words have arbitrary meanings. If so, we are all entitled to a new freedom, to stipulate freely what words mean. This brave new thought obviously had no impact whatsoever on the hard sciences, but eventually obtained devastating effects in the soft areas of knowledge, particularly in the vocabulary of political theory. Here new theories can be made just out of verbal manipulations. And it is in fact the case that freedom, authority, repression, violence, coercion, tolerance, and many other key terms no longer address-for a sizable public-the problems that the theory of politics has long addressed. Are we still able to communicate intelligibly? Can we still transmit and accumulate knowledge? I would certainly hope so-but not if nothing is done about it. Another, not unrelated change bears on the influence of Marxism. Until the r95os the bulk of the literature was on "democracy;' not on "capitalist democracy." Today Marxists and non-Marxists alike speak of capitalist democracy as a matter of course. The shift is nonetheless a momentous one. IX

THE THEORY OF DEMOCRACY REVISITED

Capitalist democracy is a politico-economic system and indeed, for Marxists, an economic system projected onto a political superstructure. Whether democracy tout court- as a political system- still receives a fair and adequate hearing under the capitalist democracy focus (I personally doubt it), its acceptance does carry crucial built-in assumptions. The same applies to another Marxist-induced change, that is, the current commonplace acceptance of "democratic ideology" as a designator not only of the value beliefs of democracy but also of its theory. To be sure, non-Marxists employ'the word ideology in a neutral, innocent sense. Even so, to concede that "all is ideology" is by no means an innocuous concession of little consequence. A third change stands on a different plane. When the "gospel of negation" swept across the institutions of learning in the mid-196os, the behavioral revolution had won its own revolution, and thus the theory of democracy was largely becoming empirical. This was and is a much needed addition. But the timing was unfortunate, for the empirical theory of democracy was no match for the philosophical, high-flown theorizing of the Frankfurt school and its widespread progeny. This is simply to note that the empirical theory is not, and is not supposed to be, the argumentative theory. It so happened, therefore, that the more democracy became argued-scrutinized, debated, contested- the less the empirical theory of democracy had a say. The decade was carried, not by the behavioral, but by the campus revolution. We have thus been largely left with a theory of democracy-the argumentative one-constructed with much passion but little knowing. As we muddle through the 1980s the question is: Does a mainstream theory of democracy still exist? I think not. Its dismemberment results, somewhat paradoxically, from the joint impact of a decade of negation and a subsequent decade of technicalization. The simplistic, engaged, and enraged literature of the mid-196os has been followed by a sophisticated literature rich in bravura. The bulk of this literature purely and simply ignores the negators of the 1960s; it equally ignores, however, whatever preceded their negations. It starts anew from a tabula rasa. The prisoner's dilemma, the paradox of voting, maximin justice, the logic of collective action, the theory of public goods-all this (and more) attest to a new, exciting creativity. The same applies to the revival of political philosophy. Rawls's original position under a veil of ignorance, Nozick's state of nature, Ackerman's neutrality assumption - these and other "thought experiments" doubtlessly are innovative breakthroughs. The point remains that this new literature does not add up to a full-fledged theory of democracy. If debunking left the theory of democracy in shambles, its recent enrichment is largely single-issue-minded and leaves us with splendid fragments in splendid isolation. But if the theory of democX

Introduction racy no longer obtains a mainstream, it is imperative, I believe, to seek its rebuilding. This requires a bringing together of essentials and, in the process, their tidying up. So, despite its length, this work deals only with bare essentials and their "cleaning;' and may be described summarily as a work on mainstream basics. While the task may appear unexciting, it still is a much needed one because we must have (and give) reasons for the institutions we have, and because democracies are not viable unless their citizens understand them. And my sense is that the "understanding citizen" is on the wane. This is so, I submit, not only because he no longer finds mainstream support (as I have just explicated it) but for another motive as well. Of the many characterizations of democracy, our present-day plight brings to the fore that democracy is "government by discussion." If this characterization is enlarged, it suggests that, as democracies develop, more and more people discuss more and more. This is a good thing-except that as the discussion has grown, correct discussing has in fact declined. This outcome is not surprising, and many reasons can explain it. Whatever the reasons, the conclusion remains that the more we have a "discussion need;' the more it becomes imperative to discuss how to discuss. Hence, I extensively discuss the discussion about democracy. It turns out, then, that this work is, above all, a housecleaning venture, a task of dispelling sloppiness (in argument) and messiness (in conception). This is also, incidentally, my understanding of "theory?' Whatever else the notion may mean, there can be no theory worthy of its name without good arguments; and good arguments result, in turn, from logically correct ways of discussing. What does The Theory of Democracy Revisited actually cover? Since the Table of Contents adequately speaks to which items are sorted out and how they are organized, the question turns on the division of the work into two parts. Volume r deals, in the main, with issues that have surfaced in the last decades. Volume 2 deals, in the main, with the themes whose discussion began in ancient Greece. The two volumes also differ in their respective underlying threads. The title of the first volume is The Contemporary Debate. What is such debate ultimately about? What makes it contemporary? My understanding is that today we characteristically confront fact-versus-value tensions. The classical theory of democracy did not draw, in any systematic manner, a distinction between ideal system and reality. This was so, among other reasons, because until the 1940s no bifurcation existed between normative and empirical theories of democracy. Today, however, we are all highly sensitized to the hiatus between the ideal and the real. By and large, the normative theory applies XI

THE THEORY OF DEMOCRACY REVISITED

to, and elaborates on, the ideals and values of democracy. Whether or not it deliberately prescribes, it certainly ends up with a prescriptive definition of democracy. Conversely, the empirical theory applies to, and generalizes from, the facts: how democracies actually perform and what democracies really are in the real world. Hence, the empirical theory ends up with descriptive definitions of democracy. So far, so good. I, for one, place a strong emphasis on the distinction between prescriptive (normative) and descriptive (empirical) definitions of democracy. But this is the easy part of the arialysis and leaves out its difficult part. Those "facts" that display, to some degree, democratic properties are, in truth, patterns of behavior shaped by ideals. The harsh observer of real-world democracies is observing, in truth, value-molded facts. The crucial problem thus becomes: To what extent and in what manner are ideals realized and realizable? More than ever before, we bear witness to paradises that materialize as infernos, to ideals that not only fail but backfire. The daunting question that we yet have to tackle, let alone resolve, bears on the translation of ideals. This entails that ideals and facts, the ought and the is, have to be related in feedback fashion. And this is very much the thread and the focus of volume r. Here my case is that fact-value tensions are constitutive of democracy, so whatever we are endlessly discussing can be recast in the mold of the debate between idealists and realists, perfectionists and factualists, rationalists and empiricists. This is, I submit, a telling clue not only for organizing an immense literature but also for coming to grips with "tension management" and, in its wake, with the furthering of democracy. The tide of the second volume is The Classical Issues. As the title implies, here the thread becomes more historical, and the focus shifts from factvalue tensions to trial-and-error processes. From the time city democracy was first conceived and practiced in ancient Greece, it has taken humankind more than two thousand years to invent and establish a viable large-scale democracy. Since political systems are human-made and conceived, the invention of present-day democracies hinges on, and is best traced in, the history of ideas and ideals. True enough, much of the machinery of democracy-political parties, for one-came about without being preconceived. Still, no such machinery would be in place had it not been for an endless painstaking thinking about power, coercion, liberty, equality, laws, justice, rights, representation. These words (and many more) are carriers, then, of historical experience in the important sense that their core meanings are largely established by success and failure and thus incorporate historical learning. As a mainstream discourse on politics develops across the centuries it discards the unfit, that is, the meanings (definitions) of power, coercion, liberty, equality, and the like, that have not withstood "trial" and instead have been conducive to "erXU

Introduction ror;' to repeated and patent failures in application. Had no such trial-anderror scanning shaped the vocabulary of politics from Aristotle to, say, Tocqueville, today we would be nowhere, and certainly not even in sight of the democracies that we have been able to build. The fabric of democracy rests, then, on ideas and ideals shaped and sorted out ( upheld or dismissed) by a mainstream of intellectual discourse that began in ancient Greece and was selectively transmitted from generation to generation via words (concepts) that are experience reminders. Today, however, that universe of discourse is in shambles-above all, as we know, on account of a theory of meaning (stipulativism) that inevitably undermines the historical cumulation of knowledge. If meanings are, ultimately, arbitrary, why bother about past arbitrariness? If anything, the past is an encumbrance along the path to the future. Well, no. Ignorance of past experience can only reproduce past errors and horrors. In order to go forward we must avoid falling back. Falling back is but one possibility. Another one is pure and simple falling. Much of the recent theorizing about democracy seemingly takes democracy for granted. Yet democracies have always been, and still are, failure prone. They were short lived, and by all accounts ill suited for survival, in ancient Greece; they quickly succumbed in their medieval communal reincarnations; even today, despite the universal reverence paid to the word, the number of inaugurals soon followed by overthrows and/or of intermittent democracies exceeds the number of long-standing ones. Latin America largely attests to this. In twentieth-century Europe, there have been discontinuous democracies in Italy, Germany, Austria, Spain, Portugal, and Greece. New African states have been democracies for a short spring only-if at all. On this ulterior consideration, let it be asked again: Can we afford a universe of discourse whose terms no longer stand as carriers of experience? I think not. The division of this work into two parts each of which is self-contained should not be construed as a hard-and-fast separation. As is obvious, historical learning also enters volume 1, and fact-value tensions are addressed in volume 2-but with different emphases. And there is no difference between the two volumes as regards their argumentative load, so to speak. From beginning to end, the work is strongly argumentative. Intellectual housecleaning does not come easy and cheap. It bears reiteration that I take a theorizing about democracy to consist, first and above all, of correct arguments correctly connected. If a premise does not sustain a consequence, if logical fallacies are involved, if our basic concepts are obscure or ambiguous, then it is certain that we have a bad theory. I also take it that, of all political systems, democracy is the one that most crucially hinges on clearheadedness. If so, the theory Xlll

THE THEORY OF DEMOCRACY REVISITED

of democracy is also the unraveling of the messy web of arguments that bear on what democracy ought to be, can be, is not, and should not become-lest intended goods evolve into unintended evils. A number of scholars and friends helped me to improve the manuscript at various points. Among them, my gratitude goes especially to Professor S.E. Finer and to my research assistant at Stanford University, Alex Hybel. Since it has taken me almost ten years to bring this work to completion, my muddling through has been sustained, at different times, by the Hoover Institution, the Guggenheim Foundation, the Ford Foundation, and the American Enterprise Institute. Being a Hoover fellow and a visiting scholar at AEI are most pleasant and gratifying experiences; and the leaves afforded by Guggenheim and Ford Foundation grants have been of invaluable help. My indebtedness to these institutions is very great indeed.

xiv

The Theory of Democracy Revisited

PART ONE

The Contetnporary Debate

1. Can Democracy Be Just Anything? It is our way of using the words "democracy" and "democratic government" that brings about the greatest confusion. Unless these words are clearly defined and their definition agreed upon, people will live in an inextricable confusion of ideas, much to the advantage of demagogues and despots.

-Tocqueville r.r

The Age of Confused Democracy

Among the conditions of democracy, the one recalled least is that wrong ideas about democracy make a democracy go wrong. I take this to be a sufficient reason for writing this book. To be sure, the concept of democracy is entitled to be diffuse and multifaceted. This is so, among other reasons, because democracy largely is by now a name for a civilization or, better, for the political end product (to date) of Western civilization. Communism and socialism can be connected to a single major author-Marx-and assessed as deviations from, and implementations or negations of, Marx. Democracy is not amenable to a similar treatment; the towering, single major author on democracy does not exist. The theory of democracy consists, rather, of a mainstream of discourse that goes all the way back to Plato and Aristotle. Yet such a mainstream did provide "democracy" with a basic identity up until the end of World War II. Until then there was little doubt in everybody's mind that fascist and communist regimes were not democracies (they actually did not claim to be such) and that democracies were Western located or Western inspired. In the late 1940s, however, it was authoritatively stated that "for the first time in the history of the world, no doctrines are 'advanced as antidemocratic. The accusation of antidemocratic action or attitude is frequently directed against others, but practical politicians and political theorists agree in stressing the democratic element in the institutions they defend and the theories they advocate?' 1 What are we to make of this? The wishful thinking was that "this acceptance of democracy as the highest form of political or social organization is the sign of a basic agreement in the ultimate aims of modern social and political institutions?' 2 But the fact simply was, and remains, that democracy 3

CAN DEMOCRACY BE JUST ANYTHING?

I.I

has become a universally honorific word. True, the words that we revere are entangling: nomina numina, names are Gods. Nonetheless, we must see to it that democracy is not reduced to a mere trap word. What we have witnessed in the decades that followed World War n is hardly the ascendance of a common ideal that is warming the hearts of humankind; we have witnessed, rather, an unprecedented escalation of terminological. and ideological distortion whose end result eminently is obfuscation. As Orwell neatly presented the case, "In the case of a word like democracy ndt only is there no agreed definition but the attempt to make one is resisted from all sides .... The defenders of any kind of regime claim that it is a democracy, and fear that they might have to stop using the word if it were tied down to any one meaning:' 3 Democracy still has foes; but it is now best evaded in its own name and by means of its own name. Foes aside, the fact is that during the last decades we have progressively lost a mainstream theory of democracy. In part this development can be ascribed to the inclusiveness of the concept. If everybody claims to be democratic, and the more democracy has to be an all-embracing concept, the more we are likely to obtain profusion and, in the aggregate, conceptual confusion. However, a number of intellectual trends have been powerful concomitants in disrupting the mainstream of discourse on democracy. Among these trends I would give first place to the doctrine that words are mere conventions and that all definitions are, ultimately, arbitrary and thus left to our freedom of stipulation. Whatever the fine points of this doctrine-I shall discuss them later 4 -its negative implications and consequences are intuitive: If words mean, in principle, whatever we wish them to mean, then we can only be heading toward Babel. Meanwhile, the applause goes to a Humpty-Dumpty society of word magicians that earn not only a living but a reputation by tampering with language and meanings. A second intellectual trend that has affected negatively the theory of democracy is the Wertfreiheit issue, the freedom-from-value issue. Wertfreiheit lends itself to three different developments. In a first, moderate version it basically consists of "value fairness" and its requirement is that the evaluation and the description be separated. In a second, extreme version Wertfreiheit seeks an actual "value void" that requires, in turn, a sterilized vocabulary, a vocabulary purged of both laudatory and derogatory connotations, of goodbad words. Since the value-vacuum program is difficult to implement, what has actually materialized in its wake is "value avoidance':._ a third version of Wertfreiheit characterized by value shyness or, indeed, by value fear. In this version values are not erased but pushed underground and turned into implicit, covert values. When Wertfreiheit came under heavy fire in the r96os, 4

The Age of Confused Democracy r.r the attack was actually aimed at the third version, or was eminently justified by it. It consisted of two distinguishable charges, namely, that (a) the alleged value-free political science had retreated into irrelevance, and that (b) coveredup values inevitably entail a conservative, status quo bias. The latter charge is of dubious validity. Covert values perform just like overt values (albeit with different efficacy). Regardless of whether the value beliefs of an American Marxist are outspoken or camouflaged, in either case they will not be supportive of the American system. Conversely, a Russian liberal will not (no matter how secretively) uphold the Soviet status quo. Implicit values "conserve" when they happen to be consonant ones (consonant with their realworld referent) and "undermine" when they happen to be dissonant. The first charge is, instead, a valid one. It has indeed been the case that a value fear, or shyness, has brought about-in conjunction with other factors-a retreat of political science into irrelevance. My complaint about Wertfreiheit is, however, of a different nature. Remember that my concern is, here, with "theory;' and specifically with the theory of democracy. In this connection Wertfreiheit has generally been understood in its first version, namely, as requiring the evaluation to be separated from the description. Fine-except that the central problem now becomes how evaluation and description, once divided, relate to each other and, I would add, interact with one another. My complaint thus is that the Wertfreiheit controversy has largely lost sight of the crucial point. In its stead Wertfreiheit has brought about an excessive value avoidance that inevitably calls for retaliation, for excessive value ladenness. The theory of democracy has thus swung from an excess of descriptivism to an excess of value advocacy. A third intellectual trend that has in fact, if unwittingly, enfeebled the mainstream theory of democracy is the behavioral persuasion. The behavioral approach has generated an "empirical theory" of democracy that is indeed a new and most important acquisition. Nevertheless, this development poses problems that have yet to be assessed properly. A first point bears on the dividing line between the empirical and the nonempirical theory. This divide has been found in "normativism;' in the distinction between non-normative theory, which is empirical, and normative theory, which is not. But this divide confuses, to begin with, two very different sorts of norms, namely, a meansto-ends "technical normativism" (largely a Zweckrationalitat, an instrumental rationality), on the one hand, and a "value normativism" (Wertrationalitiit), on the other. If both kinds are expelled from the empirical theory, then this theory covers too little and suffers from shallowness. It is also the case that the normative theory generally picks up for itself value normativism and neglects technical normativism. Now, since it so happens that the theory of de5

CAN DEMOCRACY BE JUST ANYTHING?

I.I

mocracy has labored, over the centuries, precisely on how a value normativism can be keyed into a technical normativism, the perplexing question is: Has everything been done wrong before, or have we wrongly located the watershed? A second problem bears on how the mainstream theory of democracy relates to the empirical theory of democracy. Clearly, in some manner the latter "tests" the former. That is not to say that all empirical theory is supposed to assess whether, and to what extent, the facts correspond to the theorization. Empirical theories may serve other purposes; but when they do abide by a testing purpose, as is most frequently the case, then we are not doing well. The theory of democracy as such is a macrotheory that largely hinges on broad generalizations. Conversely, the research that feeds the empirical theory of democracy produces microevidence, in the sense that the evidence is too small for the generalizations that it purports to test. So, how much microevidence is required to confirm or disconfirm a macrotheory? To make matters worse, the empirical evidence results from operational definitions many of which are only a pale reflection of their theoretical originals. Take, for instance, the generalization that democracy is based on consensus or that democracy is government by consent. Much of the evidence collected on this score seemingly goes to disprove the consensus theory. However, the consensus (operationally defined) being tested is different from the theoretical con~ struct. 5 The problem of finding a fit between macrotheory and microevidence is, admittedly, not an easy one to resolve. Meanwhile, it should be realized that the theory of democracy is being enfeebled by a counter-evidence that is often invalid, that does not really test what it purports to test. There are, then, many reasons that go to explain our current predicament. Bertrand de Jouvenel noted as early as 1945 that "discussions about democracy, arguments for and against it, are intellectually worthless because we do not know what we are talking about.'' 6 His case was, at the time, overstated. Yet de Jouvenel had sensed what was in the offing in the wake of the transformation of democracy into a love word or, as we may also say, into a catchall word. And while many intellectual trends and circumstances contribute to enfeeble the mainstream of the discourse on democracy, the central element of this enfeeblement is, in my diagnosis, the debasement of the vocabulary of politics. Up until the 1940s people knew what democracy was and either liked or rejected it; since then we all claim to like democracy but no longer know (understand, agree) what it is. We characteristically live, then, in an age of confused democracy. That "democracy" obtains several meanings is something we can live with. But if "democracy" can mean just anything, that is too much. 6

Description and Prescription

r .2

r.2

Description and Prescription

If the term democracy can be used to signify antithetical entities and dignify antithetical practices, then it is a meaningless term. What are we talking about? The answer lies, to begin with, in defining. Democracy happens to be a transparent word, that is, a word easily anchored to a literal, original meaning. Hence, it is very easy to define democracy verbally. Literally, democracy means "power of the people;' that the power belongs to the people. But this is nothing more than a word-word definition that renders in a known language the Greek meaning of the term. 7 However, the term democracy stands for something. The question is not only, What does the word mean? It is also, and concurrently, What is the thing? When we try to answer this latter query, we discover that the thing does not correspond to the word or resembles it inadequately. We discover, that is, that there is little fit between the facts and the name. So, although democracy has a precise literal meaning, this does not really help us understand what an actual democracy is. How can we remedy this? At first glance the solution seems to be simple enough. If observation reveals that the term democracy is denotatively misleading, or even that democracy is a high-flown name for something that does not exist, why not seek more fitting labels? In the real world, Dahl points out, democracies are "polyarchies?' If this is so, why not call them that (reserving "democracy" for the ideal system)? But the solution is not so simple. A label can be misleading on descriptive grounds and yet required for prescriptive purposes. And the prescription does not matter less than the description. A democratic system is established as a result of deontological pressures. 8 What democracy is cannot be separated from what democracy should be. A democracy exists only insofar as its ideals and values bring it into being. Polyarchies are not conceived by Dahl literally (i.e., as systems characterized by multiplicity of command and, thereby, by diffusion of power) but, far beyond the sheer literal connotation of the term, as systems "in which power over officials is widely . . . shared" via a relatively high degree of control of the ordinary citizen on leaders. 9 Now, if citizens control their leaders, it can be assumed that the latter are (must be) responsive to the former. Since polyarchy as a structure does not necessarily carry these implications, it is clearly the case that polyarchies are (in part, imperfectly, to varying degrees) the result of democracy as an ideal system. Thus, Dahl's polyarchies are as they are because they embody ideals. No doubt, other political systems too are sustained by imperatives and value goals-but in a very different way. The values of the predemocratic polity (think of feudal society) provided fixity, not dynamics. On the other hand, 7

CAN DEMOCRACY BE JUST ANYTHING?

I.2

the "target polities" exemplified by the communist states entrust the value attainment to a vanguard, to a closed power group that defines and enforces the goal. Democracies are also, in some sense, target polities - but without a vanguard; the targets are established via the democratic process, within democratic procedures, and as a democratic course moves forward. It follows from this that democracy is uniquely open to, and hinged on, a fact-value tension. It can thus be said that only democracy owes its very existence to its ideals. And this is why we need the name democracy. Notwithstanding its descriptive inaccuracy, it helps to keep ever before us the ideal-what democracy ought to be. The term democracy, then, has not only a descriptive or denotative function but also a normative and persuasive function. That a sharp divide between description and prescription is often difficult to draw does not detract from the analytic import of the distinction. That day and night gradually shade into one another does not entail that their difference is only of degree or (still worse) that light and darkness are inseparable. Consequently, the problem of defining democracy is twofold, requiring as it were both a descriptive and a prescriptive definition. One cannot exist without the other and, at the same time, one cannot be replaced by the other. To avoid starting out on the wrong foot we must keep in mind, then, that (a) the democratic ideal does not de.fine the democratic reality and, vice versa, a real democracy is not, and cannot be, the same as an ideal one; and that (b) democracy results from, and is shaped by, the interactions between its ideals and its reality, the pull of an ought and the resistance of an is.

Political and Other Democracies

r.3

The word demokratia was coined in the fifth century B.c., and from that time until roughly a century ago it has been a political concept. That is to say, democracy meant political democracy. Today, however, we also speak of democracy in a nonpolitical or subpolitical sense, as when we hear of social democracy, industrial democracy, and economic democracy. While these are perfectly legitimate meanings, they are also largely responsible for the plight of confused democracy. It is therefore important to clarify these meanings and see from the outset how nonpolitical relates to political democracy. The concept of social democracy-of democracy in a social sense, as a state of society-can be traced back to Tocqueville. In visiting the United States in 1831 Tocqueville was struck by the societal premises-especially equality of status, manners, and customs-of American democracy. In part the social equality that appeared so striking to Tocqueville (as well as to most Euro-

8

Political and Other Democracies

r.3

pean visitors after him) reflected the absence of a feudal past; but it also expressed, undoubtedly, the way in which the "spirit" of democracy permeated the society as a whole. Thus Tocqueville contrasted democracy to aristocracy, and until r848 conceived democracy as a state of society rather than a political form. After Tocqueville, it was Bryce who best rendered the idea of democracy as an ethos and a way of life, as a general state and style of society. To Bryce, American democracy appeared characterized by "equality of estimation;' that is, by an egalitarian ethos that is based on, and resolves itself into, the equal value "which men set upon one another, whatever be the elements that come into this value:'10 A social democracy is, then, a society whose ethos requires of its members to conceive themselves as being socially equal. The label social democracy also applies, by implication, to the network of primary democracies-small communities and voluntary organizationsthat may flourish throughout a society, thus providing the societal backbone and infrastructure of the political superstructure. A multigroup society in which the unit "group" consists of democratically structured groups equally qualifies as a social democracy. On the other hand, a social democracy (not to be confused with Social Democracy as a party name) is a far cry from a "socialist democracy:' The characterizing element of a social democracy is not only that it performs at the societal level but, even more, its spontaneity, its endogenous nature. It is from this latter angle that one appreciates best the extrapolitical nature of a social democracy and the reason that makes it entirely different from a socialist democracy, from a policy enforced by a socialist state upon a society. While a social democracy ascends from the bottom up, a socialist democracy descends from the top down. Social democracy is first of all a way of life, while socialist democracy is above all a way of governing. Industrial democracy is a term launched at the end of the nineteenth century by Sidney and Beatrice Webb. 11 In essence, it is democracy within the industrial plants. True enough, the Webbs went in their design all the way up from the microdemocracies operating at the plant level to an overall system of vocational, "functional" representation, that is, all the way from grass-roots democracy in the industry to a consonant political macrodemocracy.12 Nonetheless, the subsequent theory and practice of industrial democracy have generally dropped the superstructure devised by the Webbs and focused on the industrial plant as the unit of a worker's self-government. It might be said, therefore, that industrial democracy is an adaptation of the Greek direct democracy to an industrial society in which the member of the political community, the polites, is replaced by the member of the economic community, the worker in the workplace. In recent decades the notion of industrial de9

CAN DEMOCRACY BE JUST ANYTHING?

1.3

mocracy has been reformulated and applied either in the form of "codetermination" (Mitbestimmung) or in the Yugoslav type of "self-management" by workers' councils. In the first case the worker has a share in the management, and eventually in the ownership, of the plant; in the second case, as the wording says, he is supposed to realize a self-government of the producers in their workplace.13 In all cases, then, what is intended by "industrial democracy" is clear enough and can be pinned down to precise meanings. Economic democracy is, instead, a multifaceted and, in the end, elusive notion. The logic that leads to this construct is, however, dear enough, and is as follows: Since political democracy restricts itself in the main to political and juridical equality, and since the emphasis of social democracy is on equality of status, it follows from this that economic democracy is, or reflects, the concern for the equalization of wealth. Thus a first definition of economic democracy can be that the label denotes a democracy whose policy goal is the redistribution of wealth and the equalization of economic opportunities and conditions. So conceived, however, economic democracy can well be a complement of political democracy and may simply represent an extension of political democracy. Economic democracy is also used, however, in one of the meanings of industrial democracy. If so, it points less to an equal or near-equal distribution of wealth and more to a worker's control over the economy. In this connection it might thus be said that economic democracy consists of equality of control over the economic productive process. The Marxist understanding of the label differs from both of the foregoing acceptations. In the Marxist literature economic democracy does not presuppose political democracy; it displaces and replaces it. This substitution follows from the materialistic conception of history and what it entails, namely, the negation of the autonomy of politics. In the Marxist view political democracy has no value in itself, no intrinsic reason for being, for it is only the instrument of the domination of the exploiters over the exploited. More precisely, political democracy is a superstructure of capitalistic and bourgeois oppression, and can thus be reduced to capitalistic democracy. What next? When the "capitalistic" and/or bourgeois elements are removed, what happens to the other element, to "democracy?" As we shall see in due course from how Lenin performs on the point, in the perspective cif the withering away of politics the notion of democracy evaporates into a thick mist. If politics as such is a superstructure, and if the real stuff of reality is its economic stuff, the road to communism does not lead to a political system but to an economic system. In the end there will be no need for a noncapitalistic democracy, for we shall not need a polity at all. In the end, then, an "economic democracy" is nothing more and nothing else than a "communist economy." 14 IO

Political and Other Democracies

1.3

To sum up, social democracy and industrial democracy address clearly identifiable structures and connote a nonpolitical democracy in the sense that the two concepts do not address the problem of democracy at the polity level, i.e., as the problem of a statewide and state-managed democracy. On the other hand, economic democracy is by now a highly ambiguous, overly extended label. At one extreme it merely signifies a policy-and, indeed, a policy enacted by political democracy within its structures and via its procedures. At the other extreme, economic democracy is assumed to eliminate and replace political democracy- and becomes a notion that defies pinpointing. Remaining with the "other democracies" that are amenable to a clear identification, the question of how they relate to political democracy obtains a straightforward reply: Democracy in the political sense is a large-scale macrodemocracy, whereas group-centered and plant-centered democracy are smallscale microdemocracies. This is the same as saying that political democracyin the acceptance of the term that has persisted for some twenty-five centuries - is the superordinate sovereign democracy, whereas the other democracies are, inevitably, subordinate democracies. This is, I submit, a simple statement of fact. We may think highly of microdemocracy and have, conversely, a very poor opinion of macrodemocracy. The relationship remains as it is. And the relationship is that if the superordained entity-the polity-is not a democracy, there is little chance, if any, that the subordinate entities can survive and prosper as democracies. Thus nobody denies the importance of social democracy as a vital basis of a democratic polity, nor that a grass-roots, primary democracy may have more value than any other aspect of democracy. Likewise, economic equalization and industrial democracy may matter to us more than anything else. The fact remains that political democracy is the requisite condition, the requisite instrument, for whatever democracy or democratic goal we may cherish. If the master system, the overall political system, is not a democratic system, then social democracy has little worth, industrial democracy little authenticity, and economic equality may not differ from an equality among slaves. This is the reason why "democracy" without qualifier stands for "political democracy;' why democracy is first and foremost a political concept, and also why this book will essentially deal with democracy in the political sense of the term. First things must come first; and political democracy as a method, or as a procedure, must precede whatever substantive achievement we may demand from a democracy. True, "the importance of the democratic political method lies mainly in its nonpolitical by-products:'15 But the "goods" presuppose the machinery, the method that produces them. We are not required to love the machine-but we need it all the same. The criticism leveled against II

CAN DEMOCRACY BE JUST ANYTHING?

I.3

the mainstream theory of democracy as dealing only with political democracy is thus hard to understand and attests, I submit, to an erosion of identity. On all grand subjects, such as democracy, much remains always to be said. But incompleteness is not error. Error lies, rather, with the authors who downgrade or even dismiss the political premises and requisites of whatsoever democracy they may be demanding.

1.4

Outline

In the final analysis our political behavior depends on our idea of what democracy is, can be, and should be. When we declare that one political system is more democratic or less democratic than another, our evaluation depends on what we think of as being a true democracy. Likewise, when we assert, "This is not democratic" or "Here there is no democracy;' the assertion and the resulting behavior presuppose an implicit, if not explicit, definition. Of course, the general public reacts to an image of democracy. Yet these images are the echo of conclusions arrived at in the sphere of theoretical thinking and have been molded by definitions. What ordinary people think reflects, in the long run, what thinkers have previously thought. My basic contention, then, is that if democracy is incorrectly defined (by its definers), in the long run we all are in danger of refusing something that we have not properly identified and getting in exchange something that we would not want at all. Having stressed why we should define democracy, let me return to the how. The basic point is that the thing democracy is not described properly by the word democracy. Yet we can neither escape the word (we can never do without words) nor change it, saying polyarchy instead of democracy, for instancef6 We are thus required to keep in mind that the term democracy leads us to a prescriptive definition and that we shall have to seek a descriptive definition as well. To have two definitions no doubt creates problems; but, as we shall see, the distinction between the descriptive is and the prescriptive ought of democracy actually points to where the solution of our problems lies. Distinguishing between is and ought, real and ideal, brings about complexities but also unravels imbroglios. Take, for instance, the assertion "Socialism is superior to liberal democracy:' It should be obvious that in order to warrant this or, for that matter, any other cross-regime comparative evaluation, we are required to match the real with the real and/or the ideal with the ideal. It will not do to compare socialism as an ideal with democracy in reality. Indeed, this is cheating. Yet this is generally the way of arguing that communist regimes are superior to democracies. What is even more extraordinary is how easily a fallacy of such magnitude goes undetected. It bears I2

Outline

r.4

stressing, therefore, that reference to democracy as it is calls for reference to socialism as it is. Conversely, if reference is made to socialism as an ideal, then reference must be made to democracy as an ideal. Generally stated, the point is that when facts are compared with facts, or when facts are compared with ideals, they must be compared in exactly the same manner across the board. This involves the pairing of descriptions with descriptions, of prescriptions with prescriptions, and the comparing of how well (or badly) each ideal is translated into its corresponding reality. Alertness to the distinction between is and ought immediately shows that it is nonsensical to argue that the ideal of something is superior to the reality of something else. Turning to the complexities, they result from the fact that the distinction between the descriptive is and the prescriptive ought of democracy cannot be left at that. No sooner is the distinction drawn, than the question becomes, What is their connection? The connection is just as important as the distinction. What is and what should be do not proceed on two parallel roads that never meet. On the contrary, they always interfere and collide with each other. So if we do not grasp the nature of the connection, the advantage of establishing a distinction between description and prescription is lost, little by little, along the way. The first part of this work thus deals at length, implicitly when not explicitly, with this relationship. When attention is focused exclusively on what is, the result is a misplaced and misused realism. When we place all the emphasis on what ought to be, we stumble into the pitfall of perfectionism. Misplaced realism and perfectionism are, however, mistakes of a sophisticated kind. Therefore, my argument starts with a criticism of the more simplistic approach, by a discussion of the most elementary and naive type of error, namely, reducing democracy to its name. The product of this simplistic approach is what I call "etymological democracy;' or literal democracy, which I discuss in the next chapter. To be sure, an essential condition for the survival of a d~mocratic system is its intelligibility. While democracy is more intricate than any other political form, paradoxically enough it may not survive if its principles and mechanisms are not within the intellectual reach of the average citizen. Yet when democracy is made to seem very simple, we can be sure that the simplification has gone too far, that we are disserving, rather than serving, democracy. Even though the errors of oversimplification, hyperrealism (if I may call it that), and perfectionism are the snares into which we most often fall when we discuss democracy, our difficulties do not end there. Further difficulties arise from the fact that when democracy was first conceived, it was conceived horizontally. Greek democracy did not pose an altimetric problem, the prob-

13

CAN DEMOCRACY BE JUST ANYTHING?

r.4

lem of a vertical structuring. Modem democracies, or representative democracies, do. But the coining of "elitism" and its success as a word of disparagement go to show that we have yet to come to grips with the problem of how democracy is to be conceived and defined vertically. It behooves me, therefore, to confront this problem squarely. Also, a typical trait of a time of confused democracy is that we no longer know, or no longer attempt to make clear, what democracy is not. And the first part ends with looking at democracy as a web of decision-making processes performed by various units. While my primary intent is not to propose or discover a "new" theory of democracy, if I were to lay a claim about having an original theory of my own, such a theory would be the one expounded in the chapter that concludes the first half of this work. Below the level of generality at which we may speak of a mainstream theory of democracy in the singular, it is true that we have many theories of democracy in the plural. Barry Holden, for instance, organizes his excellent analysis around five types or clusters of theories: radical democratic theory (ROT), new radical democratic theory (NRDT), pluralist democratic theory (PDT), elitist democratic theory (EDT), and liberal-democratic theory (LDT).17 Dahl's classic analysis rests on the tripartition between Madisonian democracy, populistic democracy, and polyarchal democracy! 8 In my account I distinguish between electoral democracy, participatory democracy, referendum democracy, and the competitive theory; and the major contrast is rendered in terms of participatory versus competitive theories of democracy.19 These slicings are all of analytic avail. The important difference remains whether or not the bits and pieces are conceived as parts of a mainstream ensemble. Barry Holden does notice that only one of his theories (LDT) is a full-fledged one, while the others are incomplete or defective theories; yet he seems to give up on the possibility of bringing them together in some selectively incremental fashion. This work is predicated, instead, on the assumption that a mainstream existed in the past and badly needs to be restored for the future. Therefore I see a continuity between classic liberal-democratic theory and subsequent pluralist-competitive developments, whereas the populist and radical theories of democracy (whether in old or new garments) remain peripheral in that they are only normative theories. But all of this can be explicated only as I proceed. One final point. As I have said, I shall deal in the main with political democracy. The first reason for this is that a democratic polity precedes and conditions whatever other democracy we may seek-as I shall stress at length. Another reason for insisting on political democracy is that it is the most difficult of the democratic undertakings. As a political form democracy has to

A Coda on Theory

1.5

reduce the manifold wills of millions, tens or even hundreds of millions, of scattered people to a single authority; and this means that the conditions under which a democratic polity has to perform are but a remote approximation to the optimum conditions found in primary groups and small, integrated communities. Between a face-to-face democracy and a large-scale democratic system, there is a huge, yawning gap. Humankind had to suffer for more than two thousand years in order to build a bridge between the two sides; and in passing from the small democratic communities where all the people can take part to the democracy of the large numbers who cannot take part (at least in the same sense), much of what occurs in the former is bound to disappear on the way. Thus, if we still have democracy in a political sense, we cannot expect of the large-scale, cumbersome political democracy what we can expect of microdemocracies. Actually, the passing from micro to macro involves a change of scale and, indeed, a scale leap in which a tenfold increase in size or units may involve complexities by hundredfolds or even by a factor of thousands. Hence it is highly dubious whether our political macrodemocracies can be correctly conceived and understood as an enlargement of some microprototype, of some primary democracy. The above is not to say that the unit of democracy has been established once and for all. It so happens that the unit that I describe as "macrodemocracy" currently is far more extended than a small city, and far smaller than world size. From where we stand historically-the nation-state unit-we may wish to go back to smaller units or proceed toward larger units. The argument over the optimal unit goes around in circles, as with a set of Chinese boxes: "Any unit you choose smaller than the globe itself ... can be shown to be smaller than the boundaries of an urgent problem. . . . Yet the larger the unit, the greater the cost of uniform rules, the larger the minorities who cannot prevail, and the more watered down is the control of the individual citizen:'20 Since any conceivable unit has its merits and drawbacks, the important thing is to treat them in their distinctiveness- for if the units are mixed up, so is the argument.

1.5

A Coda on Theory

A work on democratic theory should declare how "theory" is understood. I have already noted that it is now customary to distinguish between normative theory and empirical theory, and to employ "normativism" as the dividing criterion. That is to say, if a theory contains norms (prescriptions of any kind), it is a nonempirical and, presumably, a philosophical theory. Conversely, a theory is empirical if, and only if, it is non-normative. I find all of this ob-

CAN DEMOCRACY BE JUST ANYTHING?

r.5

jectionable. As pointed out earlier, "normativism" lumps together, and thus confuses, norms that have very little if anything in common. A first objection is, thus, that ethical and ultimate value norms are one thing, and conditional or instrumental technical norms (that gauge ends to means) quite another thing. There is also a world of difference between prescribing norms and describing norms. In the end, "normativism" turns out to be not only a poor divide on account of its ambiguity but also a wrong divide in principle. But in order to warrant this last conclusion I must .first address th~ notion of theory. Today, Brecht asserts, "political philosophy, political theory and political science are no longer interchangeable terms?' 21 This is correct. But what makes them different? The question ranges too far afield to be taken up here. 22 Let it be taken for granted that a theory of politics is neither a philosophy nor a science of politics, and let us confine the discussion to the notions of "theory" and "empirical theory;' Even so, the issue must be narrowed further. A good way to do so is to note that theorizing is a function of its object, of what it is about. On this consideration the question, What is a theory of democracy supposed to be? hinges on the question, What kind of object is democracy? The Greeks conceived politics as an architectonic science. To modernize and sharpen the classical intuition, it may be said that democracy is an artifact and, indeed, an ongoing macroartifact. This implies, first and foremost, that democracy embodies a project. Hence the theory of democracy may be said to be the outlay of the project of democracy. Of course, this does not imply that we first plan a polity and then build it as planned. We are not that good at pla.nning; and our political artifacts are largely a product of muddling through trial and error. More often than not it is only ex post, when the construction has materialized, that we grasp what it is that we have actually constructed. Thus the project outlined by the theory of democracy largely is a feedback, an analytic reflection on achievements that have not been pre-planned or have not quite worked out as planned. Nonetheless, reference to a project remains telling on two counts. First, democracies are target-striving, not blind, experiments; and it is the theory of democracy qua project that is required co provide orientation and guidance toward those targets. Second, the outco1t1e uncertainty (a project is not a blueprint) does not detract from the fact that all the outcomes that have entered the fabric of democracy have indeed been preceded and promoted by consonant ideas and ideals. The upshot of the foregoing is that where a project is involved, ends are necessarily involved. Hence, ends and goals attainment are a constitutive element of the theory of democracy. If so, even an empirical theory of democracy is required to be, in some sense, normative. But why, then, employ "normativism" as a divide? Why not directly ask what is meant by "empirical"? The

16

A Coda on Theory

1.5

reply to this question is that a theory is all the more empirical the more it is constructed inductively, thereby incorporating what is learned from experience. Conversely, a theory is less and less empirical and, ultimately, nonempirical, the more it disregards experience and is constructed deductively. But empirical theories are not all alike. They differ, at base, in that they either pertain to a theory-research context or a theory-practice context. In the first context we are primarily concerned with how a theory can be tested; in the second context we are interested in whether and how a theory can be applied. When the focus is on the relation between theory and research, we dwell on operationalization, on rendering the theory in terms of operational concepts and operational definitions. When the focus is on the relation between theory and practice, we dwell instead on application, on the means required by the implementation of ends and, more generally, on how the ends are translatable into deeds. Along the first path we seek evidence; along the second path we sort out success from failure. In the end the theory-research track leads to a descriptive theory of existing democracies, while the theorypractice track lands at a theory of feasible democracy. In both cases we obtain, I submit, empirical theories of democracy; but doubtlessly theories of very different sorts. The first one is empirically extracted, whereas the second one is empirically operative. 23 As for the nonempirical theories, for the purpose at hand it will suffice to define them a contrario, that is, as the theories that are neither based on evidence nor on practice. As with the empirical theories, there are many varieties of nonempirical theories of democracy that are variously described as being speculative (or philosophical), ethical, axiomatic, deductive, and rationalistic. Their shared characteristic, however, is that they are not grounded on research nor geared to application and applicability. I still have to reply to the question, How is "theory" conceived here? If we abide by the stipulation that the label "empirical theory of democracy" is to be reserved to a specifically operational and testing-oriented theorizing, then this is not the theory of democracy pursued here. 24 My concern is, rather, with the operative theory of democracy, with how the theory relates to, and passes into, practice. As I have just pointed out, this does not imply that I am expounding a nonempirical theory. But the theory to which I attend certainly is "less empirical" than the one that we specifically call empirical. For one, my theorizing is both prescriptive and descriptive and does deal at length with the norms that constitute the project of democracy. Furthermore, since this work responds to a time or state of confused democracy, it is heavily argumentative and relies very much on conceptual analysis. Perhaps, then, the kind of theory presented in this work comes close to what has always 17

CAN DEMOCRACY BE JUST ANYTHING?

r.5

been called theory without qualifiers, theory pure and simple. 25 In this theory pure and simple, democracy is conceived as a project-though not an executive project, not a blueprint-of an ongoing human artifact that hinges on the set of ideas and ideals that make it, uphold it (in its makings), and that (if misunderstood and mismanaged) will also unmake it. Simply put, once that democracy is in place we may have an "empirical theory" of democracy; but before, and I would say as a preliminary condition, we must have a theory tout court. The artifact "democracy" has to be conceived and constructed before being observed. Democracies exist because we have invented them, because they are in our minds, and insofar as we grasp how to keep them well and alive.

Notes 1. In R. McKeon, ed., Democracy in a World of Tensions: A Symposium Prepared by UNESCO (Chicago: University of Chicago Press, 1951), p. 522. 2. Ibid. 3. "Politics and the English Language;' in Selected Essays (Harmondsworth: Penguin, 1957), p. 149. 4. See chapter 9, herein .. 5. The point is addressed in chapter 5, especially section 2. 6. Du Pouvoir (Geneve: Bourquin, 1947), p. 338. 7. "Word-word definition correlates a word to another word, as having the same meaning. Word-thing definition correlates a word to a thing, as meaning that thing." See R. Robinson, Definition (Oxford: Clarendon, 1954), p. 17, and chap. 2, passim. 8. Deontology means literally "discourse on what must be done;' on dutifulness. The term was introduced by Bentham, who used deontology as a synonym for "the science of morality?' I shall use the term, instead, without specific reference to ethics for any discourse cast in the imperative form. 9. The quote is from Dahl's Modern Political Analysis (Englewood Cliffs, N.J.: Prentice-Hall, 1963), p. 73. But see, more extensively and importantly, R.A. Dahl, A Preface to Democratic Theory (Chicago: University of Chicago Press, 1956), and Polyarchy (New Haven: Yale University Press, 1971), passim. ro. See J. Bryce, The American Commonwealth (1888), abridged 2 vol. ed. (New York: Putnam's, 1959), pt. VI, chap. 6, esp. pp. 514,517,520. Bryce held, however, that democracy was basically a form of government. 11. See Industrial Democracy (London, 1897; new ed. in 2 vols., 1920). 12. See Sidney and Beatrice Webb, A Constitution for the Socialist Commonwealth of Great Britain (London, 1920). See also, in the Webbs' tradition, G.H.D. Cole, Self-Government in Industry (London, 1920). 13. Among the overviews see H.A. Clegg, A New Approach to Industrial Democracy (Oxford: Blackwell, 1960); E. Rhenman, Industrial Democracy and Industrial Management (London: Tavistock, 1968); P. Blumberg, Industrial Democracy: The Sociology of Participation (London: Constable, 1968); D. Jenkins, Job Power: Blue and

18

Notes White Collar Democracy (Garden City, N.Y.: Doubleday, 1973); "Industrial Democracy in International Perspective;' Annals of the American Academy of Political and Social Science, May 1977. The subject of industrial democracy is pursued in chapter 14, section 4. 14. The Marxian and Leninist conceptions of democracy are analyzed in chapter 15, especially sections 1 and 2. 15. Charles Frankel, The Democratic Prospect (New York: Harper & Row, 1962), p. 167. 16. Dahl's argument is that when the same term is used for the "ideal system" and its real-world "imperfect approximation;' then "needless confusion ... get(s) in the way of the analysis" (Polyarchy, p. 9). He is right about the confusion; but different terms are suggestive of and conducive to a separation whose drawbacks are, in my opinion, even greater. In Dilemmas of Pluralist Democracy: Autonomy vs. Control (New Haven: Yale University Press, 1982), Dahl appears to come to this view: "In this book;' he writes, "I use ... interchangeably ... democratic regimes ... [and] polyarchy" (p. n). 17. See B. Holden, The Nature of Democracy (New York: Harper & Row, 1974), passim. RDT roughly applies to the classical, Greek theory of direct democracy (ending with Rousseau); NRDT denotes the new radicalism of the 1960s (from the theory of participatory democracy to, presumably, Marcuse); PDT is centrally Dahl's theory; EDT is essentially Schumpeter's theory; and LDT covers the ground from Locke to Tocqueville and John Stuart Mill. 18. In A Preface to Democratic Theory. Dahl himself is generally identified with the theory of democratic pluralism (even though this concept is only implicit in the Preface). r9. In chapters 5 and 6, herein. 20. R.A. Dahl, "Democracy and the Chinese Boxes;' in H.S. Karie!, ed., Frontiers of Democratic Theory (New York: Random House, 1970), pp. 372-73. 21. Arnold Brecht, Political Theory: The Foundations of Twentieth-Century Political Thought (Princeton: Princeton University Press, 1959), p. 17. 22. My view is outlined in G. Sartori, "Philosophy, Theory, and Science of Politics;' Political Theory, May 1974, and developed in IA Politica: Logica e Metodo in Scienze Sociali (Milano: SugarCo, 1979). The discussion is extensive. See D. Easton, "The Decline of Modern Political Theory;' Journal of Politics, Febuary 1951; A. Cobban, "Ethics and the Decline of Political Theory;' Political Science Quarterly, September 1953; A. Rapoport, "Various Meanings of Theory;' American Political Science Review, December 1958; L. Strauss, What Is Political Philosophy? (Glencoe: Free Press, 1959); E. Weil, "Philosophie Politique, Theorie Politique;' Revue Franr:aise Science Politique 1 (1961); I. Berlin, "Does Political Theory Still Exist?" in P. Laslett and W.G. Runciman, eds., Philosophy, Politics and Society (Oxford: Blackwell, 1962); J.W. Chapman, "Political Theory: Logical Structures and Enduring Types;' in R. Polin, ed., Lldee de Philosophie Politique (Paris: Presses Universitaires de France, 1965); S.S. Wolin, "Political Theory: Trends and Goals;' International Encyclopedia of the Social Sciences (New York: Macmillan & Free Press, 1968), vol. 12, and "Political Theory as a Vocation;' American Political Science Review, Decerriber 1969; J.C. Gunnell, Political Theory: Tradition and Interpretation (Cambridge, Mass.: Winthrop, 1979). 23. Note that I sharply distinguish between "operational" and "operative." I use

CAN DEMOCRACY BE JUST ANYTHING?

operative to say "applicable:' In my understanding, a knowledge is applicable, or operative, when it succeeds in practice, i.e., with respect to the conversion of theory into practice. 24. A good illustration of this kind of empirical theory-specifically focused on assessing performance-is G. Bingham Powell, Contemporary Democracies: Participation, Stability and Violence (Cambridge, Mass.: Harvard University Press, 1982). 25. C.F. Cnudde and Deane E. Neubauer, eds., Empirical Den:xocratic Theory (Chicago: Markham, 1969), distinguish, in their Introduction, between three types of theory, namely, normative, analytic, and empirical (see pp. 1-13). Perhaps, though not quite, my theorizing can be classified as being analytic in their sense of the term, i.e., conceptual analysis.

20

2. Etymological Democracy If we examine not the verbal definitions that most people, including dictionary-makers, give for "democracy," but the way in which they use the word in practical application to affairs of our time, we will discover that it does not have anything to do with self-government. -James Burnham

2.1

The Meaning of People

Etymological democracy is democracy conceived in the original, literal meaning of the term. Thus the etymological definition of democracy is, very simply, that democracy is the rule or power of the people. This is an appropriate start. Not only do words have a history but, invariably, a very telling history. To ignore their reason for being coined, their variations and subsequent departures from the original meanings, is to renounce a compass in a perilous navigation. In particular, the original meaning is never a fancy or fanciful meaning. When we go back to an etymon, we are sure to start on genuine grounds. On the other hand, to inquire about the original meaning of a term is only-as a rule-the first step of an inquiry. This is particularly true of democracy, for it can be easily shown that the premise "democracy is the power of the people" not only affords little mileage but is, from the outset, an unclear premise. In the first place, what is the meaning of demos? Even in Greek the term was not free from ambiguity. In Aristotle's basic framework, democracy is the vis-a-vis and the degraded form of the politeia, of what we might render as the "good citY:' This implies that the demos that enters Aristotle's conception of democracy is made of multiple ingredients. It consists not only of the many but also of the poor; and it furthermore consists of the many and/or the poor as qualified by defects (self-interestedness, unlawfulness, or other) that sustain, in turn, the difference between a democracy and a politeia.1 Since the Aristotelian meaning of demos subserves the purposes of an overall classification of political systems, it is an unnecessary complication to bring it in at this stage. But even if we remain with the pure and simple linguistic usages of demos, we are still confronted with several meanings of the term. 21

ETYMOLOGICAL DEMOCRACY

2.I

Demos in the fifth century B.C. meant the Athenian (or similar) community gathered in the ekklesfa, the popular assembly. However, demos can be assimilated to the entire body; or to the pol/of, the many; or to the plefones, the majority; or to 6chlos, the mob (the degenerative meaning). And the moment demos is translated into the Latin populus, the ambiguities increase. The Roman concept of people is a very peculiar one and can be understood only within the frame of what we call Roman constitutionalism. Yet it cannot be dismissed as irrelevant, since the language of the Middle Ages was Latin. For some fifteen centuries, then, the concept was populus; and this entails that the doctrine of "popular sovereignty" incorporated in our concept of democracy is not Greek and is misunderstood whenever we derive it directly from demos. Finally, a central diversity is entered with the advent of our modern languages, namely, that the Italian term popolo, as well as its French and German equivalents, convey the idea of a singular entity, whereas the English word people indicates a plural: In spite of being a collective noun, it takes a plural verb. In the former case we are easily led to think that popolo, peuple, and Volk denote an organic whole, an "allbody;' which may be expressed by an indivisible general will; whereas in the latter case to say democracy is like saying "polycracy;' a separable multiplicity made up of the unit "each one:• Thus it is no mere coincidence that the holistic interpretations of the concept have come from scholars thinking in German, French, or Italian. In spite of being committed to the maxim that democracy must be made as intelligible as it can be made, the foregoing cannot be reduced to less than six interpretations of "people":

r. People meaning literally everybody People meaning an undetermined large part, a great many 3. People meaning the lower class 4. People as an indivisible entity, as an organic whole 5. People as a greater part expressed by an absolute majority principle 6. People as a greater part expressed by a limited majority principle 2.

The first interpretation is, it seems, the more intuitively obvious one. Yet the people that are citizens of a democracy cannot literally include everybody. On this requirement no democracy has ever existed nor is likely to exist. In the Greek democracies the demos excluded not only women but also those not born free, the slaves (who alone added up to a majority of the town dwellers). Today we still exclude (after the relatively recent inclusion of women) minors, the mentally deranged, criminals serving a sentence, noncitizens, and transients. How inclusive should and can "people" be? No matter how much 22

The Meaning of People

2.1

the issue is deemed important and pressed,2 nobody has yet made the case for total, exceptionless inclusion. That people means "everybody" is a conception that stands, so to speak, on the strength of its denotative vagueness. If everybody without exception is a criterion that would make democracy impossible, the people as a "great many;' or the many, simply cannot be used as a criterion. In no small part democracy is a procedure; and the people as a great many is a notion that poses the impossible procedural requirement of having to determine, on each occasion, how many make a people or suffice to make a people. Furthermore, and again, a great many with respect to which total? So, the first two interpretations create more problems than they solve and may be left at the consideration that their use is a limited one. It may look as if the indeterminacy of (2) is corrected by (3), that is, by establishing that the great many consist of the lower class and, specifically, of the poor and/or the working class. But this interpretation is difficult to defend in principle and increasingly difficult to apply in practice. In Aristotle's time, and for a very long time thereafter, the lower class consisted of the poor, and the poor did largely outnumber the wealthy. However, the more a sizable middle class affirms itself, the more the poor-versus-rich dichotomy breaks down. As a society modernizes, its socioeconomic pyramidal shape gives way to a hexagonal shape. And in the postindustrial societies the "hungry poor" and the proletariat add up to being outnumbered minorities. Furthermore, and aside from its contemporary obsolescence, the identification of the people with the lower class is highly objectionable in principle. While no democracy can literally include everybody, even so the exclusions consistent with democratic principles must either be specifically justified (as in the case of minors) or must be of a procedural nature. For instance, the majority principle does, in some sense, entail exclusions; but flexible and transient ones-for my exclusion today still permits my inclusion tomorrow. Contrariwise, the people as lower class posit a fixed exclusion: Whoever is not lower class is excluded forever. And this is, I take it, either democratically unacceptable or acceptable only if we revert to Aristotle's notion that democracy is a bad polity. The fourth interpretation - the organic and holistic conception - suffers from the central defect of not being conducive to democracy at all or, in any event, of allowing for the justification of any political regime whatsoever. Let us give it a closer look. The preliminary caveat is that "the people" need not be conceived individualistically. Indeed, for a very long time both the Greek demos and the Latin populus have been conceived in a corporate fashion, not as a sum total of separate individuals held fit to decide for themselves. In the corporate conception of life, the individual was nested into authority, church, tradition, and ascriptive aggregates. Even so, the people were not con23

ETYMOLOGICAL DEMOCRACY

2.I

ceived as an "organic fusion" until the Romantic revolution. Rousseau well attests to the difference between the pre-Romantic and Romantic conceptions. Rousseau did call on an indivisible general will; yet it was not entirely indivisible. 3 In Rousseau one still finds the echo of a contract-based vision of society that bears on individual natural rights. With the Romantics this vision is frontally rejected. The Romantics called on a spirit of the people, a Volksgeist or Volksseele, that truly resembled an oversoul. The caveat is, thus, that we must not confuse the corporate with the "organic whole" notion of people. The latter connotes a much stronger fusion than allowed by the natural law tradition, is characteristically Romantic and, in its wake, a product of the philosophy of idealism. And if reference is made to the Romantic, holistic conception of the people, then we are confronted with a notion that legitimizes a tyrannical rule far better than a democracy. From the people as an organic whole it can be easily inferred that each individual counts for nothing; in the name of the whole, each and all can be crushed one at a time; and behind the formula "all as a single one" we glimpse the justification of totalitarian autocracies, not of democracies. A democracy cannot even begin to perform unless it breaks up that formula. We are thus left with the people translated in terms of counting rules, that is, rendered either as an (5) absolute majority or a (6) limited majority principle. Absolute majority means, in this context, that only the majority counts: The greater number of any given population stands for all and has an unlimited (i.e., absolute) right to decide for all. Conversely, the limited majority principle holds that no rights of any majority can be "absolute" (i.e., limitless). The first criterion leads, then, to democracy defined as a system of pure and simple majority rule; the second criterion leads, instead, to democracy defined as a system of majority rule limited by minority rights. In both cases we finally dispose of operative and/or operational criteria. But this is the only characteristic that the procedural interpretations of "people" have in common, for in all other respects they are at odds with each other. At first view it might appear that the solution of rendering "people" as the right of a majority to absolute rule represents the straightforward solution. But this is not so. Establishing the absolute right of the majority to impose its will on the minority, or minorities, amounts to establishing a working rule that works, in the longer run, against the very principle that it extols. If the first winner of a democratic contest acquires unfettered (absolute) power, then the first winner can establish itself as a permanent winner. If so, a democracy has no democratic future and ceases to be a democracy at its inception; for the democratic future of a democracy hinges on the convertibility of majorities into minorities and, conversely, of minorities into majorities.

24

The People in Mass Society

2.2

At a second view, then, the limited majority principle turns out to be the democratic working principle of democracy. This is a most crucial point, to which I shall revert shortly. 4 For the moment I only wish to establish that we have already gone far beyond an etymological definition and explanation of "democracY:' If it is true, as I have just suggested, that the notion of people should be understood as requiring a majority rule to be limited by minority rights, then the question becomes: How can we restrict the power of those who are, in principle, fully entitled to it? This question cannot be answered by having recourse to the sheer "will of the people" notion of democracy, since a people entitled to make decisions in accordance with the principle of majority rule exercises its power within limits only because elements totally extraneous to the popular will come into play. But before entering these complexities, a further scrutiny into the notion of people is in order. A natural outlet of etymological democracy is demolatry, that is, a great deal of talking about the people without ever actually looking at them. As Burdeau remarks, this looking is resented by some as a sacrilege, "a sacrilege because in the Pantheon of political values the people are cloaked by a mystery which cannot be separated from their power." 5 The demolater does speak of "real people"; but he ends, in truth, with making a fetish of an ideal people accurately placed out of sight. Still worse, demolatry is not necessarily demophily, actual love and actual compassion for the poor, the disinherited, the underprivileged. Making a fetish of an ideal people goes hand in hand, often, with a wholesale contempt for actual, existing people. From Robespierre onward we have ample evidence of how easily, in practice, a mystique of the ideal works out as its reverse, as a mystique that prompts bloodshed, a ruthless extermination of people. At the risk of being accused of sacrilege, let us give "the people" some concrete underpinning.

2.2

The People in Mass Society

When the term demokratia was conceived, the people concerned were the demos of a Greek polis, a small, tightly knit community operating on the spot as a collective deciding body. But the larger the polity, the less the concept of people can designate a concrete community and the more it comes to connote a juridical fiction or, in any case, a highly abstract construct. We are no longer living in a polis but in what the Greeks conceived to be its very negation-the megapolis, the political city that has lost all human proportion. The plain fact is, then, that the people of the polis, of the medieval communes, as well as the.third (and fourth) estates of the ancient regime, no long-

ETYMOLOGICAL DEMOCRACY

2.2

er exist. With the collapse of corporative structures and the corporate concep-, tion of life, and with the repudiation of the Thomistic principle that one's existence must conform to one's preordained condition, the people has become always more a Gesellschaft and always less a community, a Gemeinschaft. 6 Thus modern society turns out to be the very opposite of that organic whole that the Romantics in their medievalist dreams deified. Today "the people" stands for an amorphous aggregate, for a highly diffuse, atomized, and eventually anomic society. A new reality requires a new name; and we have come to speak, in this connection, of "mass society?' 7 However, since the theory of mass society and the notion of mass man have come, in the 1960s and 1970s, under heavy criticism, 8 let us not get bogged down into discussing labels and let us begin with asking: Which are the new elements, the new factors, that characterize the present-day state of Western societies? The obvious one is the magnitude, or size, factor. The Athenian citizens that convened in the ekklesfa were, according to Thucydides, a maximum of 5000 and ranged, generally, between 2000 and 3000. No shattering change of scale occurred in the democratic communes of the Middle Ages and up to the time when the Republic of Geneva was idealized and cherished by Rousseau. While some of the implications of the escalation of the size factor demand further probing, the undeniable ones are that the megapolis does not favor a sense of individual efficacy and prompts, instead, atomization and depersonalization into anonymity. A second factor is the dramatic acceleration of history, of the time machine. The present-day world moves at such speed that in the brief span of a lifetime we become entirely estranged from the world we knew as children. Even if the novelty improves on what preceded it, the experience remains a traumatic one of historical uprooting. The third, concomitant factor is the rate and pace of horizontal mobility. In Western societies fewer and fewer people live and die where they are born; and this is the uprooting of community. To be sure, a loss of roots may be a liberation from roots that were chains-yet the dilaceration of connective tissues remains a cost in itself. It is, perhaps, a cost dramatized by a transient aspect, namely, the rupturing nature of the switch from rural to urban life. Nonetheless, since the megapolis does not restore intimacy, the "lonely crowd" 9 is likely to remain with us. In synthesis, the primary-group nesting of our lives is gone; the adjustment to fast and ever-changing environments is a strenuous race littered with maladjusted dropouts; and the emptiness thus resulting nurtures alienation and anomie.10 I was saying that a new reality requires a new name. Why not call it "mass society"? Let us make sure, in the first place, that the critics of the no-

The People in Mass Society

2.2

tion do not misrepresent it; and let us also draw a distinction between the sociological and the political theory of mass society. Thus far I have described the mass society in sociological terms. As such, the theory of mass society points, in essence, to a loss of community, and a loss of community especially provoked by acceleration and uprooting. 11 It does not maintain, on the other hand, that traditional, "communal societies" are a lost paradise. Likewise, it does not imply that the democracies of the past were better or safer democracies. The point simply is to understand and perceive what the concrete referent of "people" is today. Thus far, then, all should be well. The controversy arises when we come to the political theory of mass society and, specifically, when we ask what are the political implications of such a state of affairs. It should not be forgotten that the early political literature on the mass man and the mass society- Ortega y Gasset, Mannheim, Jaspers, up to Hannah Arendt-reflected the shock of the fascist and subsequent Nazi "democratic" seizures of power. How could that seizure have happened? And how does it still happen, to put it in Revel's wording, that the "totalitarian temptation" looms large and lingers over much of Europe?12 Ortega's philosophy was "aristocratic" in the Greek meaning of the term: It extolled excellence. But his distinction between leading minorities and masses cuts across social classes. Ortega's aristoi were conceived ascetically, as exemplar individuals defined by obligations, not by rights. Conversely, the mass man was assimilated by Ortega to the indulged heir of noble families. And despite much debunking, his portrait of a new man who for the first time in history takes everything for granted, who enjoys benefits without being solidary with the conditions that are conducive to such benefits, who refuses to grow up to his responsibilities, and who behaves as a spoiled child, as an ungrateful and undeserving heir, remains a powerful and highly insightful diagnosis.13 With Hannah Arendt it was the masses that made totalitarianism possible; but, like Ortega, she did not identify the masses with the lower or working class, with the underdogs. In her understanding "the masses" are of all-class extraction.14 In Kornhauser's overall recasting of the theory of mass society at the end of the 1950s, an atomized society is easily mobilized and manipulated. The mass man is isolated, exposed, and hence available; his behavior tends to be extreme behavior in which activist modes of response and intervention in the political process are the alternative to apathy. Thus a mass society is open to charismatic domination and total mobilization. In Kornhauser's words: "The psychological type characteristic of mass society provides little support for liberal democratic institutions .... The individual seeks to overcome the anxiety accompanying self-alienation by apathy or activism .... Thus the 27

ETYMOLOGICAL DEMOCRACY

2.2

mass man is vulnerable to the appeal of mass movements which offer him a way of overcoming the pain of self-alienation:'15 These, however, are potentialities. Which is the triggering factor? How does a potentiality pass into a specific actuality? Kornhauser indicates that the crucial variable is to be found in how the elites relate to the non-elites, and vice versa. This is probably right, but the political theory of mass society remains, on this score, largely unfinished. Masses aside, Erich Fromm raised a question that was never raised in former times. In the early 1940s he wondered whether our time was not characterized by a "fear of freedom" and, thus, by a latent "escape from freedom?' 16 His disquieting query was: Is freedom too much of a burden, too much of a responsibility? Totalitarian temptations, fear of freedom, vulnerability, easy manipulability-these certainly are worrying traits. Is the portrait overdone and biased? If so, let us have a corrected portrait and let the bias be amended. It will not be amended, however, by a sheer counter-bias and, in fact, by simply restoring the "mystery" spoken of by Burdeau. The point that bears on my argument is simply this: The man of our time cannot resemble his ancestors; yet the etymologist constructs his edifice on foundations that he refuses to inspect. His demos was buried centuries ago, and he has yet to replace it.

2.3

Power of the People and Power over the People

Up to this point we have analyzed the people, that is, the simpler element of our problem. When we come to the link between the concept of people and the concept of power, between demos and kratos, the difficulties increase. In fact they become insuperable for the etymological approach. While power is analytically a difficult notion, historically and in the history of political thought it is a fairly straightforward one. Power is a political, not an ethical, concept. Also, power is not freedom. Why am I saying this? Am I not stressing the obvious? Not quite. Macpherson has been for some twenty years a highly influential author, and his definition of power is the "ability [of the individual] to use and develop his essentially humane capacities:'17 Well, no. The ability to use and develop our humane capacities has long been a theme of ethics, and enters via this route politics as one of the meanings of "freedom."18 And freedom so conceived stands miles apart from power. That two so utterly different things should be confused attests to our state of confusion; and the ease with which a misnomer (the sheer renaming of a very old concept) is acclaimed as a new concept attests to the thinness of the ice on which we are skating. On both counts, it is not superfluous to

Power of the People and Power over the People

2.3

underscore that in the authors who have labored over the problem of linking demos with kratos, power always is the force and capability of controlling others - including therein the force of disposing of their lives and having them put to death. The crucial distinction in matters of power is the one between titular holders and actual wielders. Power is, ultimately, exercitium: exercising power. How can the people, however understood and defined, be effective wielders of power? As is obvious, the titular right to power does not solve the problem of popular sovereignty. Secular monarchies in their struggle with the church affirmed their independence substituting the formula omnis potestas a Dea, that all power derives from God, with omnis potestas a populo, that all power derives from the people, thus resorting to a democratic justification. This justification was used by Marsilius of Padua, the major medieval proponent of the concept of popular sovereignty, in his defense of Ludwig of Bavaria and of imperial supremacy. But if Marsilius fell back on popular sovereignty to support the empire against the church, the Jesuits resorted, later, to the same principle in order to fight the absolute monarchs and to justify the right to kill the tyrant. And the point is that a titular right to power is little more than a nominal attribution, than a nominal power.19 The medieval doctrine aimed at bridging the gap between nominal power and the exercise of power by the fictio of representation, that is, by having the titular holder delegate the exercise of his power to somebody else. This was indeed a fiction, since the medieval doctrine did not envisage an actual transfer. 20 In particular, in the medieval doctrine of representation the representative- the person or body to whom the exercise of power was delegated -was not required to be an elected representative. During the Middle Ages, and beyond, representation was a praesumptio Juris et de Jure, a presumption allowing no possibility of proving the contrary. Therefore representation could also legitimize-as it did in fact-monarchical absolutism in a position of permanent, irrevocable representation belonging by hereditary right to the sovereign and his descendants. 21 We can understand, then, the intransigent aversion that Rousseau felt for the notion of representation and the reason he reversed the formula, substituting for nonelective representation the principle of election without representation. Rousseau's democracy elected its magistrates but did not give them the chrism of representatives. The people, for Rousseau, do not delegate their power and should not give up the exercise of power. Rousseau saw where the danger lay. Yet Rousseau's solution is hardly a workable solution. He himself admitted that it was realizable only on a small scale, only for very small republics. 22 So we cannot follow his advice. If he who is elected is not regarded

29

ETYMOLOGICAL DEMOCRACY

2.3

as the representative of those who elect him, the election simply creates, per se, an absolute ruler. Throughout history, unbridled rulers have been created by means of an election just as much as by sheer force or hereditary succession. Hence we need both election and representation. An election, as such, does not create a representative. The pope of the Catholic church is elected by the College of Cardinals and yet is under no account a representative of his electors. The linkage is established by a normative expectation (that is also, via removability, a sanctionable expectation) of responsiveness and accountability of the person elected to an electorate. Moreover, and in particular, the election must be a free election. Just as representation without voters has little meaning, voting without free choice cannot result in representative government and becomes nothing more than the people's periodic renunciation of their sovereignty. If presumed representation is insecure, election without choice is fraudulent. 23 In sum, modern democracies hinge on (a) limited majority rule; (b) elective procedures; and (c) the representational transmission of power. This implies that within the people as a whole some people count more, some less; that even the people who constitute a victorious voting majority do not really wield power; and that much of what is called the "will" of the people sounds more like a "consensus" by the people. Once we have recognized this, our discussion shifts from etymological democracy and the will-of-the-people context to the techniques of constitutional democracy. In order to realize democracy the titular attribution of power and its actual exercise do not remain in the same hands. And none of the instrumental and procedural means for achieving democracy as a large-scale political system are either implied or suggested by what the word means. "Power of the people;' it must be emphasized, is simply an elliptical expression. The phrase describes the beginning of a process but leaves it hanging in midair. For power is exercised over somebody, and governing presupposes the existence of the governed. Power of the people over whom? Who are the addressees, the objects, of popular sovereignty? When the formula is completed, it reads thus: Democracy is the power of the people over the people. But the problem then takes on a completely different twist; it does not consist only of the up-going of power but, even more, of its downward descent. If, along this two-way process, the people lose control, then the government over the people is in danger of having nothing to do with government of the people. The crucial issue thus is: How do we maintain and firm up the link between the nominal attribution and the actual exercise of power? While elections and representation are necessary instruments of large-scale democracy, they are also its Achilles' heel. He who delegates his power can

Limited Majority Rule 2.4 also lose it; elections are not necessarily free; and representation is not necessarily genuine. What are the reniedies and the safeguards against such eventualities? The question can certainly be answered, but not on the basis of "literal democracy:' The truth is that a theory of democracy consisting of nothing but the notion of the power of the people is adequate only as long as it combats autocratic power. Once this adversary is defeated, what is automatically transferred to the people is only a titular right. The exercise of power is quite another matter. When this is clearly understood, then we can certainly attend to the ways and means of maximizing the power (the actual power) of the people. However, the more we seek a popular exercitium of power- as in the theory of participatory democracy-the more we should want to find out how the realworld people are. If this is not the case, if the advocates of a direct popular power are precisely those who construct the people as a sacred, inscrutable, and mysterious entity, then we have not moved an inch beyond etymological democracy.

2.4

Limited Majority Rule

While this chapter deals with preliminaries, there is one issue that I wish to press right away. Democracy, I have argued thus far, is not pure and simple popular power. I shall now insist that democracy is not pure and simple majority rule either. Indeed, "majority rule" is only a shorthand formula for limited majority rule, for a restrained majority rule that respects minority rights. Until a few decades ago this was well understood. I doubt that this is still the case today. Throughout the history of humankind, majoritiesethnic, religious, or sheer numerical majorities-have in fact persecuted minorities, at times to the point of exterminating them. Today this is done (conspicuously in Africa, but also in other areas of the world) in the name of majority rule and thereby, as an obvious implication, in the name of democracy. The implication is downright untenable; but I seldom hear that said loud and dear. The current objections to the identification of democracy to sheer (and hence absolute) majority rule apply, more often than not, to the special case of the highly divided and intensely conflictual societies. As Nordlinger puts it, "the democratic model in its orthodox majoritarian form is unsuitable for the regulation of severe conflicts:' 24 I have no quarrel with Nordlinger's conclusion, but I am struck by his premise, that is, by his reference to an "orthodox majoritarian" model of democracy. What model is that? I am not sure. But I am sure that democracy is not unqualified (and thereby limitless) majority rule. 31

ETYMOLOGICAL DEMOCRACY

2.4

That the working principle of democracy is the principle of limited (restrained) majority rule hardly is a novel thought. When Burnham wrote that "the fundamental characteristic of democracy in the sense in which we use the word (regardless of what it meant to the Greeks who invented it) is the concession of the right of political expression to minorities;' 25 he correctly believed to be reporting on a widely accepted conception of.democracy. Lord Acton had put it thus: "The most certain test by which we judge whether a country is really free is the amount of security enjoyed by minorities:' 26 And Guglielmo Ferrero concisely stated the full case as follows: "In democracies the opposition is an organ of popular sovereignty just as vital as the government. To suppress the opposition is to suppress the sovereignty of the people:'27 Why is that so? A first reply might be that the authors just quoted and the tradition they represented had a strong sense and concern for individual liberty. The underlying, supporting argument thus appears to be that "liberty for the individual means nothing if it does not imply the right to pursue a course of conduct and to hold and advocate views which do not have the approval of the majority."28 But suppose that we do not share that concern, that individual freedom is felt to be less important than other values. In such case should we not equate democracy with pure and simple majority rule? The reply still is no. Limited majority rule remains a fundamental characteristic of democracy even without calling on individual freedom. Whether or not we happen to like freedom, the case does not necessarily rest on this consideration. The case rests on the broader consideration that if a majority makes excessive use of its right, the system as such will no longer function as a democracy. Let it be asked again: Why is that so? A first point bears on the very definition of "people:' The people, it was noted earlier, cannot literally consist of everybody; b{.it they cannot be reduced to a greater part of the citizenry either. When the people are translated into a majority criterion, what is being provided is only an "operative definition." That is to say that the people are divided into a majority and a minority by the decision-making process and in order to have decisions made. The fact nonetheless remains that the people consist, overall, of the majority plus the minority. Hence, if the majority criterion is turned (erroneously) into an absolute majority rule, the real-world implication of this switch is that a part of the people (often a very large one) becomes a non-people, an excluded part. Here, then, the argument is that when democracy is assimilated to pure and simple majority rule, by this assimilation a portion of the demos is eo ipso converted into a non-demos. Conversely, democracy conceived as a majority rule limited by minority rights corresponds to the people in full, that 32

Limited Majority Rule

2.4

is, to the sum total of majority plus minority. It is precisely because the rule of the majority is restrained that all the people (all those who are entitled to vote) are always included in the demos. The foregoing is highly abstract. To grasp the implications, assume that a majority is entitled by its very principles to exercise its power without restraints. Inevitably, and almost by definition, such a majority will treat the non-majority unfairly and unequally. This entails that the majority in question can and easily will maintain itself as a permanent majority. But if we have a majority that cannot be turned into a minority, then we are no longer dealing with a democratic majority-that is, with a system whose rule of the game is the majority principle. For a majority principle requires changeable majorities, with the various parts of the body politic being able to alternate in wielding power. Kelsen spelled out the underlying logic of the point with unbeatable perceptiveness. He wrote: "Even he who votes with the majority is no longer entirely subject to this own will. He is aware of this when he changes his opinion [for] ... in order to be free again, he, the individual, would have to find a majority in favor of his new opinion;' 29 Let me comment on this. Clearly, if minorities are not protected, the possibility of finding a majority in favor of the new opinion is unlikely, since he whose opinion changes from that of the majority to that of the minority immediately falls into the ranks of those who have no right to make their opinion heard. Therefore, unless the liberty of minorities is respected, not only would the first electoral test determine, once and for all, those who are free and those who are not; but also the liberty of those who voted for the majority on that occasion would be lost because, in practice, they would not be permitted to change their opinion. Thus the first election would be, in effect, the only true election. And this amounts to saying that such a democracy dies at the moment of its inception. After having pushed the argument to the extreme, its gist is that the citizen of a representative democracy does not lose his freedom, as Rousseau contended, at the instant he votes, precisely because he can decide at any moment to transfer his allegiance from a majority opinion to a minority opinion. It is in this being permitted to change his opinion that not only his enduring liberty is rooted but that democracy maintains itself as an open, self-steering polity. The liberty of each is also the liberty of all; and it acquires its most authentic and concrete meaning when we are in the minority. Pace the slogan on democracy as majority rule, it is the respect and safeguarding of minority rights that sustain the dynamics and the mechanics of democracy. In summary, minority rights are a necessary condition of the democratic process itself. If we are committed to that process, then we must be committed to a

33

ETYMOLOGICAL DEMOCRACY

2.4

majority rule restrained and limited by minority rights. To maintain democracy as an ongoing process requires us to ensure that all citizens (majority plus minority) possess the rights that are necessary to the method by which democracy operates. 30

2.5

The Lincoln Formula

I have held that whenever we judge a democracy by the standard of its literal definition, the faults that emerge are likely to lie in the definition, not in the reality. If a democracy was "power of the people" and nothing else, it would have to follow from this premise that all existing democracies are to b..': perceived, and by implication rejected, as false democracies. Yet political systems exist in which the people are not only protected from power abuses but actually and importantly enter the process of deciding what is to be done and "who gets what." And such polities exist precisely because it was grasped that the problem of how they can be established begins exactly where the etymological understanding of democracy leaves off. The foregoing is by no means intended to suggest that the literal meaning of the word is unimportant. If the etymological definition does not lead us far, it does provide the foundation. That the power belongs to the people establishes a principle concerning the sources and the legitimacy of power. It means that power is legitimate only if it is actually bestowed from below, only if it is an emanation of the popular will, and only if it rests on some expressed, basic consensus. Considered as a theory about the sources of, and the titular right to, power, the literal concept of democracy also indicates that we expect and require from a democratic form that society takes precedence over the state, that demos precedes cracy. Democracy exists, then, when the relation between the governed and the government abides by the principle that the state is at the service of the citizens and not the citizens of the state, that the government exists for the people, and not vice versa. In his Gettysburg Address of 1863 Lincoln provided the most memorable of all the characterizations of democracy: "government of the people, by the people, for the people?' It is symptomatic that this phrase defies exact analysis. If we attempt to dissect it, with respect to the element "government of the people" the caveat is that the preposition of can indicate both the subject or, conversely, the object of an action. Hence, with respect to this feature all the following conjectures are permissible: (a) government of the people meaning a self-governing people, a direct democracy; (b) conversely, that the people are the object of government, that they are governed; (c) that government 34

Notes

emanates from the people in the sense that it derives its legitimacy from the people's consent; (d) that government is chosen by the people; (e) that government is guided by the people. So, the first characterization covers, or can cover, the entire spectrum of politics; not only all the coni:;eivable forms of democracy, but also governments over the people that have nothing to share with democracy. The second element, "government by the people;' suffers from a contrary defect: It is too obscure to allow specific conjectures. By the people in what sense? This formula defies pinpointing. Only the third element, "government for the people;' appears unambiguous: for the people means, clearly, in their interest, for their benefit, to their advantage. But many regimes that never claimed, in the past, to be democracies, did declare themselves governments for the people. And, today, ·communist dictatorships claim to be democracies precisely on that account. 31 This exegetic exercise conveys, and is only meant to convey, that Lincoln's phrase qualifies democracy primarily because it was pronounced by Lincoln. Since we know about Lincoln, we know what he intended. Yet had Stalin been the first to stumble onto the formula, could he not have used it too? Indeed, his government was of the people in the perfectly permissible grammatical sense that he governed them; it is difficult to disprove that his intention was to govern for the people; and by the people could well have meant, in Stalin's mind (since it can mean anything), that he and the people were one, that it was the "general will" that governed through his will. The bottom line is, then, that Lincoln's formula does not suffice to define democracy-its democratic credentials stem from Lincoln. If authored by another source, it could easily mean what Lincoln did not wish or intend it to mean. Under the proviso that the formula has stylistic impetus rather than logical meaning, it should be recognized that this is precisely its purpose and its value. For to use "democracy" in its literal sense opens a prescriptive discourse whose very nature is to remain unfinished, to go on ad infinitum, as well as ad indefinitum.

Notes 1. The basic framework of Aristotle is straightforward: Government of the one, the few, and the many, in either a good or a degraded form. This framework is complicated, however, by an empirical typology that engenders a very untidy array of democracies. See The Politics of Aristotle, trans. and ed. E. Barker (New York: Oxford University Press, 1958), esp. pp. 162-86. This explains why no definition of Aristotle's concept of democracy goes unchallenged. The problem is discussed in chapter ro, section 2. 2. See especially R.A. Dahl, in J. Fishkin and P. Laslett, eds., Philosophy, Politics and Society, 5th series (Oxford: Blackwell, 1979), chap. 6, "Procedural Democracy:'

35

ETYMOLOGICAL DEMOCRACY

3. See chapter n, section 4. 4. See section 4, below. The point will be amplified as we proceed, especially in chapter 6, section 2, and chapter 9, section 4. For a clear and judicious overview, see H.B. Mayo, An Introduction to Democratic Theory (New York: Oxford University Press, 1960), chap. 8, "The Majority Principle and Its Limits:' 5· Traite de Science Politique, vol. IV (Paris: Librairie Generate de Droit, 1952), p. n2. In volumes VI and vn one finds, however, that Burdeau himself pays heed to the mystery. 6. This is F. Tonnies's well-known distinction in Community and Society (1887; trans., New York: American Book Co., 1940). Gesellschaft is society understood as an external, impersonal network of businesslike associations, an exterior mode of coexistence; Gemeinschaft is a community in the symbiotic meaning of the term, that is, a mode of coexistence reaching a maximum of personal interpenetration, a very strong intensity of the "we." These opposite ideal types underpin the passage from primary communities to secondary groups, to the merely formal organizations. 7. The theory of the mass man and society is not to be confused with the earlier studies on the psychology of crowds. For the distinctions between crowds, mobs, audiences, etc., see R.W. Brown, "Mass Phenomena;' in G. Lindzey, ed., Handbook of Social Psychology (Cambridge, Mass.: Addison-Wesley, 1954), 2:833-76. Particularly on "movements;' see N.J. Smelser, Theory of Collective Behavior (New York: Macmillan, 1963); and F. Alberoni, Movement and Institution (New York: Columbia University Press, 1984). 8. See, for all, S. Halebsky, Mass Society and Political Conflict (Cambridge: Cambridge University Press, 1976). The criticism of the theory of mass society belongs to the more general attack on so-called elitism that I criticize, in turn, in chapter 6, especially section 8. 9. Reference is made to D. Riesman (with N. Glazer), The Lonely Crowd (New Haven: Yale University Press, 1950). See also, in this line of enquiry, P. Slater, The Pursuit of Loneliness (Boston: Beacon Press, 1970). 10. Anomie, or normlessness, is the distinctive contribution of E. Durkheimprobably the major classic on the change from traditional to modern society- in his Suicide (1897; Glencoe: Free Press, 1951). n. Loss of community applies to the specific meaning by which community indicates intimacy, a special mode of face-to-face relationship. The term also indicates a merely functionally defined unit (as in the expressions urban community or, even more vaguely, political community). Of course, a real community, a Gemeinschaft, is also functionally significant; but the reverse is not true, for functional significance does not imply any intimate communion. For a comprehensive discussion, see C.J. Friedrich, ed., Community (New York: Liberal Arts Press, 1959); and, specifically, R.A. Nisbet, The Quest for Community (New York: Oxford University Press, 1953). 12. }F. Revel, The Totalitarian Temptation (Garden City, N.Y.: Doubleday, 1977). 13. See especially The Revolt of the Masses (1930; New York: Norton, 1932), chaps. 6 and 12. Camus said of Ortega that he was the greatest European writer after Nietzsche. In the United States Ortega has been little read and largely misread. Few of his radical critics seem to know that he stood for the conjunction of liberalism and socialism. A correct reevaluation is L. Pellicani, Introduzione a Ortega y Gasset (Napoli: Liguori, 1978).

Notes 14. Reference is made to The Origins of Totalitarianism, first published in 1951 (New York: Harcourt, Brace). 15. W. Kornhauser, The Politics of Mass Society (Glencoe: Free Press, 1959), p. II2 and passim. 16. I recall the two titles given to the same work. The Fear of Freedom is the English title (London: Routledge & Kegan Paul, 1942), Escape from Freedom the American one. Both titles are true to the message of the book. 17. C.B. Macpherson, Democratic Theory: Essays in Retrieval (Oxford: Clarendon Press, 1973), p. 50. Since this is Macpherson's central concept, paraphrases of this definition run through his writings. 18. The point is pursued in chapters II and 13, herein. 19. See, in general, E. Crosa, La Sovranita Popolare dal Medioevo alla Rivoluzione Francese (Torino: 1915); and W. Ullmann, Principles of Government and Politics in the Middle Ages (London: Methuen, 1961), pt. 3. For Marsilius in particular, see A. Cecchini and N. Bobbio, eds., Studi Raccolti nel VI Centenario della Marte di Marsilio da Padova (Padova: Cedam, 1942). 20. A.J. Carlyle, Political Liberty (New York: Oxford University Press, 1941) overstates the case of the Spanish representative bodies of the twelfth century, of the 1295 Parliament in England and of the fourteenth-century States General in France. Medieval liberties were hardly sustained by bodies allegedly representing the wilJ of the community, but by the belief (as expressed by Bracton) that the king had two superiors, God and the law- the custom of the community. For the relation between law and liberty, see chapter II, herein. 21. This applies, however, to the thesis of a translatio imperii: that the transfer of power from the people to the emperor was a final one. The contrary argument was that the people only made a concessio, a temporary and revocable transfer of their power. For the inception of the theory of representation, see R.W. and A.J. Carlyle, A History of Medieval Political Theory in the West (6 vols., 1903-36), vol. v, chap. 4, and vol. VI, pt. 2, chap. 6; and M.V. Clarke, Medieval Representation and Consent (1936; New York: Russell, 1964). Mommsen maintained that the principle of representation can be traced back to antiquity, a line of interpretation followed, e.g., by JO.A. Larsen, Representative Government in Greek and Roman History (Berkeley: University of California Press, 1955). I find this thesis misleading, and agree entirely with Rousseau's statement: "The idea of representation is wholly modern; it derives from feudal government.... In the ancient republics the term was unknown" ( Contrat Social, III, 15). 0. Gierke, Political Theories of the Middle Ages, a partial trans. of vol. 3 of Genossenschaftrecht (Cambridge: 1900), p. 64, concurs: "The idea of the exercise of the rights of the people by a representative assembly ... current in the Middle Ages [was] unknown to antiquity." 22. Contrat Social, m, 1, 13, 15. But see chapter II, herein. 23. On the theory of representation, see H.F. Pitkin, The Concept of Representation (Berkeley: University of California Press, 1967); G. Sartori, "Representational Systems;' in International Encyclopedia of the Social Sciences (New York: Macmillan & Free Press, 1968), vol. 13; J.R. Pennock and J.W. Chapman, eds., Representation (New York: Atherton Press, 1968); and A.H. Birch, Representation (New York: Macmillan, 1972). A well-introduced, excellent reader is D. Fisichella, ed., La Rappresentanza Politica (Milano: Giuffre, 1983).

37

ETYMOLOGICAL DEMOCRACY

24. E.A. Nordlinger, Conflict Regulation in Divided Societies (Cambridge, Mass.: Harvard Studies in International Affairs, 1972), p. 33. 25. J. Burnham, The Managerial Revolution (New York: John Day, 1941), p. 162. 26. The History of Freedom in Antiquity (1877). The quotation is from Essays on Freedom and Power (New York: Meridian Books, 1955), p. 56. Even though Acton's concern was religious freedom, it clearly extends to freedom in general. 27. G. Ferrero, II Potere (Milano: Comunita, 1947), p. 217; trans. The Principles of Power (New York: Putnam, 1942). 28. J. Allen Smith, The Growth and Decadence of Constitutional Government (New York: Holt, 1930), p. 280. 29. H. Kelsen, Vom Wesen und Wert der Demokratie (Tubingen: 1929), chap. I. 30. The majority rule issue is taken up again in chapter 8, especially sections 4 to 7. As for the many intricacies of the notions of majority and minority, see chapter 6, sections I to 5. 31. Interestingly, in the Gettysburg Address Lincoln did not avowedly define democracy. When he did, in his message to Congress of April 1861, he qualified democracy as a "government of the people by the same people:' Presumably, "for the people" appeared to Lincoln an implication, not a defining characteristic.

3. The Limits of Political Realism Politics should be realistic; politics should be idealistic: Two principles which are true when they complement one another, wrong when they are kept apart. - M. Bl untschli

3.1

What ls Pure Politics?

It is far easier to know what a democracy should be than to understand what it can be. This is what "political realism" is supposed to find out- if political realism is conceived as taking stock of effectual truth, of Machiavelli's verita effettuale. l purposely recall the wording of Machiavelli because the misunderstanding of political realism can be traced back to him. Since the time of Machiavelli, the realistic approach to politics has been covered up, and covered up along two lines of explication that should be clearly differentiated: (a) as meaning that politics is politics and not something else; or (b) as implying that political realism embodies itself par excellence in a specific type of policy and political behavior called pure politics. In more recent times, _pure politics has been rechristened Machtpolitik, power politics. Pure or power politics generally denotes a type of p01itics not committed to ideals but based entirely on force, fraud, and the ruthless use of power. And it is generally assumed that this is the kind of politics and policy advocated by political realism. Since I object to most of the foregoing, it will be useful to start from the old dispute about what Machiavelli himself said and what his interpreters made him say. The Florentine secretary is considered to be the founding father of political realism inasmuch as he separates politics from ethics and from religion. This is more than to say that he was a dispassionate observer-for Aristotle too was a good observer. The difference between Aristotle's Politics and Machiavelli's Prince is that in the Greek vision of the world, politics, ethics, and religion were fused. Hence we can say in an up-to-date fashion that Machiavelli was value free, whereas Aristotle was not.1 Naturally, Machiavelli was not value free as a follower of Max Weber would be, in the modern meaning of value neutrality. But he happened to 39

THE LIMITS OF POLITICAL REALISM

3.1

live in an era in which the medieval idea of the Respublica Christiana had become meaningless, in which the word "state" acquired its modern meaning, 2 and in which the creation of the state appeared to be the paramount issue. Briefly, Machiavelli was recording the creation of the Renaissance principalities by a new kind of prince. This is not to say that politics formerly had been carried on in moral terms by gentle rulers. But Cesare Borgia was a shocking and unprecedented case even for his disenchanted contemporaries and victims. Thus Machiavelli became a value-free observer because, or partly because, what he observed was an unprecedented political "value void;' 3 This helps to explain why Machiavelli is the first to state bluntly that politics does not bow to moral precepts, thus establishing the so-called autonomy of politics. In interpreting Machiavelli it should always be kept in mind that the Renaissance principalities were unique political microcosms, especially in this respect: At that time politics and the prince were one. It might be argued that the same applies to all tyrannical rule. However, the tyranny described by Aristotle was grounded on popular support: The Greek tyrant started his career as a demagogue. Caesar put an end to the Roman Republic very much in a similar fashion: He was acclaimed ruler. 4 But Cesare Borgia and men like him were neither demagogues leaning on some kind of popular acclamation nor persons calling on a preexisting legitimacy. With respect to the distinction between tyranny ex defectu tituli (because of lack of title, of legitimacy), and tyranny quoad exercitium (on account of how power was exercised), Cesare Borgia was a tyrant in both respects-and thereby a new specimen. The upshot is, then, that Machiavelli observed a peculiar state of affairs in which politics coincided with, and was resolved into, the "nature of the Prince;' In this perspective a most essential distinction goes unheeded, namely, the difference between the politico and politics. Politics is more than the politician, than the politico. We speak of politics and/or a policy as being democratic, socialist, nationalist, and so on. This implies that no matter how powerful the modern prince may be, he is tied to a course of action that is, as such, more powerful than his personal will. Stalin was in a position to do almost anything, but he could hardly have dismissed his identification with communism even if he had wished to. So we must be careful not to equate politics with the politician. Speaking of the politician, we may well construct a typology on the basis of which the pure politico stands at one extreme and the idealistic politico at the other, meaning by this that the former is a cynical policy maker who despises ideals, while the latter pursues them at all costs. Having granted that we may well meet with a pure politician, as well as with an entirely idealistic politician, can we infer from this that "pure politics" and/or wholly "idealistic politics" ex-

Warlike versus Peacelike Politics

3.2

ist? My answer is no. If we refer specifically to the Machiavellian Prince, however, we might be tempted to answer yes. This is why we must refrain from drawing hasty inferences from what Machiavelli actually did say. He did say that politics is not ethics; but he did not say what politics is per se. Therefore, the inference to be drawn from the Machiavellian premises is that politics is "morally impure:' We still have to find out what politics is in itself. 5 On that matter Machiavelli tells us only how a certain type of policy maker behaves. However, even the "pure politician;' if he is shrewd, does not underestimate what he regards as the impure elements that help make his policy succeed, for the true man of politics knows that ideas are forces, that ideals are weapons, and as Machiavelli himself said, that even paternosters are useful bulwarks for a state. The point is, then, that pure politics is as unreal as its opposite, a wholly ideal politics. Every policy is a mixture of idealism and realism; and if either element becomes overwhelming, if too much idealism eliminates realism, or vice versa, then a policy is likely to fail. No one has ever been able to institute successfully either a genuinely pure policy or a strictly ideal and/or moral policy. They both founder for the same reason. What today goes under the name of sheer "power politics" can function only insofar as it is nourished by an ethos. Indeed, Machtpolitik was originally founded on the Hegelian Sittlichkeit, Hegel's higher form of morality. This ethos may appear bar~ baric; in fact, the major bloodlettings and massacres of modernity have occurred and are performed either in the name of a racist ideal of nation or a classless ideal of society. The fact remains that so-called pure politics is also based on ideals and values. One might even go so far as to say that politics becomes more "pure" the more it is Sache, that is, devotion to a task. Machtpolitik is, all in one, Sachepolitik.

3.2

Warlike versus Peacelike Politics

My first conclusion is that the expression "pure politics" is pleonastic when it is intended to mean that politics is not ethics but just politics, and misleading when it is understood as an ideal-less politics. If a politico advocates a pure politics of the latter sort, this shows that he has very little knowledge about the reality of politics- that he is a bad realist. It follows from this that the distinction between ideal-less pure politics, on the one hand, and idealladen impure politics, on the other, draws a wrong boundary in the wrong place. The fundamental distinction is, rather, between (a) a warlike view of politics; and (b) a peace-oriented, !egalitarian view of politics. In the former, force monitors persuasion, might establishes right, and conflict resolution is

THE LIMITS OF POLITICAL REALISM

3.2

sought in terms of the defeat of the enemy-of the "other" looked on as a hostis. 6 In the latter, force is kept in reserve as an ultima ratio, as a last and worse reason, and conflict resolution is sought by means of covenants, courts, and "rightful" procedures. Since the import of this distinction has come to be unduly neglected, let me dwell on it. Clausewitz said that war was the prosecution of politics by other means. The dictum can be reversed to express a generalized warlike posture, as follows: Politics is the prosecution of war by other means. Politics is "like war'~ a war without the arms of war-whenever its central experience is hostility, the perception of the neighbor as an enemy or, in any event, a danger or a menace. When Hobbes depicted the state of nature as a war of each against all, he was also pointing, at least implicitly, to the warlike conception of politics. Note that Hobbes detested it-so much so that he was prepared to pay any price for taking war out of politics. He had yet to see, in effect, how this could be done at a far lesser cost-without surrendering to a Leviathan - and in a far more reassuring way. The first major mitigation of warlike, force-based relations among neighbors had been the Roman construction of civil law and, precisely, the resolution of litigations in courts according to precedent and equity. Roman citizenship was eagerly sought by people who had been conquered because it provided, above all, this precious kind of "peace:' Yet the substitution of force by law was not transplanted from the private plane to the political one, to the power relationship between sovereign and subject. Ultimately, this was the feat of liberal constitutionalism. And since this constitutionalism largely consists of bringing politics under legal processing, of transforming the law of the jungle into the "law of law;' it is appropriate to speak, at this juncture, of a !egalitarian vision of politics-for politics-as-peace stands and falls with legality. It is the security afforded by a rule of law that permits, among other things, open access to, and alternation in, power. If the fall from power endangers the well-being and, eventually, the very life of a power wielder, he will not relinquish power. Conversely, if the security of staying well and alive hinges on "having power;' then power will be sought ferociously, at all costs and with whatever means. Differently put, the peaceful mode of politics assumes that the stakes of politics are not too high; this requires, in turn, that politics be tamed by constitutional legality. The foregoing clarifies two points. The first one is that he who forgets about politics-as-war can hardly appraise politics-as-peace. The latter obtains its due only when it is understood-in the light of this contrast-as a difficult and never-final victory of the laws of law over the law of force. The second, related point is that the best way of bringing back politics-as-war, no matter

42

Warlike versus Peacelike Politics 3.2

how unwittingly, is to be blind not only to the fact that most of humankind has lived, historically, under conditions of might (not of right) but also to the fact that even today the suffering condition of most human beings around the world is the war-type, not the peacelike condition. Let alone that when Marx and his successors speak of "class war;' the word war means what it says. Hence, on Marx's account the warlike conception of politics is being propounded in Western societies as well. Resuming the thread of the argument, the question is: For which sidepolitics-as-war or politics-as-peace-does political realism contend? In the light of the distinction that I have just drawn, the answer is not clear-cut. It is true that the realistic school of politics has, far more often than not, advocated power politics. However, it has not advocated, as a rule, the politics of violence, it has not praised direct action. On the other hand, it is in the name and for the sake of high-flown ideals-today often called democratic ideals-that violence and the politics of violence are best promoted and legitimized. 7 So, who stands where? Historically, the realistic and the democratic schools of politics have long clashed. It is a fact, that is, that the realists generally sneer at "democratic idealism" and that, conversely, the democratic persuasion looks on realism as an incongenial, if not antidemocratic, approach. Yet, if we ask ourselves why this should be the case, the reasons for this enmity generally turn out to be bad reasons. There is no necessary link between a realistic approach and a nondemocratic persuasion; and it is my further contention that the entire quarrel is largely based on taking "realism" for something that it is not. If it is true that there is no such thing, at least in the present-day world, as an ideal-less politics and if, therefore, any policy is committed to, and activated by, ideals, it follows from this that there is a ghost that we should exorcise: the ghost of realism as a distinct, self-sufficient type of politics and policy. What, then, is realism? Or, better, with what kind of realism are we left, once that ghost is exorcised? We are left, I suggest, where Machiavelli started, with his "effectual truth": taking stock of how things really are. That is to say, we are left with cognitive realism. And when the cognitive element and, indeed, the constitutive element of realism are brought to the fore, the argument takes on an entirely new turn. The question, For what or for whom does political realism take sides? becomes a frivolous question. For the answer is, very simply, that a cognitive realism does not side with any side. Any correct descriptive proposition, any empirically verified statement, is a "realistic" statement. Thus, political realism is nothing less, but nothing more, than the factual ingredient of any and all policies. Political realism consists of making us cognizant of the fact basis 43

THE LIMITS OF POLITICAL REALISM

3.2

of politics- period. It cannot reach out at the grand isms of politics. If it does, it cheats. For the grand isms of politics-the policies of racism, nationalism, liberalism, socialism, communism, populism, and so forth- hinge on value options that are not derived from facts but superimposed upon facts. Thus far I have left out of the discussion the charge that political realism is, like Wertfreiheit, tainted by a conservative bias. 8 The reason for the omission will now be apparent. In my argument the point becomes pointless. In my argument realism (or, if one prefers, descriptivism) is not supposed to stand alone, for it cannot, alone, dictate a policy.

3.3

Facts and Values in Benedetto Croce

Italy's fall to fascism in the 1920s provides a most cogent reminder and illustration of the point that when realists and democrats misperceive each other, and thus fight each other at cross purposes, the real loser is democracy. One of the reasons, hardly a minor one, why prefascist democracy turned out to be so fragile when it was put to test in the aftermath of World War r was that its ideals had been worn away and that they had been undermined most effectively, not by the people who would become fascists, but by the heirs of the Risorgimento tradition, by the very people who ought to have been the natural defenders of the democracy that fascism was in the process of burying. The emblematic case is the one of Benedetto Croce, the philosopher who, more than any other, influenced Italian culture during the first half of this century. 9 The case of Croce is not recounted in order to assign responsibility- a matter that is of no interest here-but because it is representative of errors repeated incessantly, today as much as ever, whenever we deal with how facts and ideals, realism and democracy, relate to each other. Croce always advocated Realpolitik. And it was in the name of the "reality of politics" that for almost thirty years he attacked mercilessly the "hypocrisy" of democracy and fought in the breach against what he ironically called the "blandishments of the goddesses Justice and Humanity?'10 Yet, years later, Croce avowed that "it had never seemed to him even remotely possible that Italy could let herself be robbed of the liberty which had cost her so much and that his generation had considered a permanent acquisition?' 11 This avowal implies that from 1896 to 1924 Croce fought a wrong battle on the wrong side; he expounded a philosophy of politics that was untrue to what Croce described, in another writing, as the real "tendency of his feelings;' as his genuine "mental and moral conformation?' 12 Furthermore, since Croce never admitted to this contradiction, the interest of his case resides precisely in how he changed camp-as he did-without retracting his Realpolitik. 44

Facts and Value in Benedetto Croce

3.3

To be sure, had Croce conceived political realism as a cognitive realism, no retraction would have been needed: He would simply have discovered that realism is, per se, neither democratic nor antidemocratic. However, Croce is an eminent representative of the idealistic, post-Hegelian school of philosophy. This implies not only that he cannot accept a cognitive realism but also that he treats the relation between is and ought, between reality and ideals, as idealistic philosophers do. Croce, like Hegel, actually does away with the distinction between is and ought, for the idealistic solution is to merge the sein and the sollen in a dialectical Aufhebung and equivalence. The end result of this merger is that the whole of his political philosophy hovers between too much realism first and too much ideality later, oscillating as it were between a matter-of-fact orientation that leaves no room for ideals and an ethereal approach that leaves no room for reality. In Croce's antidemocratic phase, which ended in 1924, his failure to distinguish between the is and the ought brought him to cancel the sol/en, the deontology. The only thing that mattered, in this period, was politics "as it really is:• Instead of using realism in order to find a solution for the problems of a liberal-democratic order, he made the former the antithesis of the latter, and refuted the norms and values of democracy by adducing facts. In his next phase, Croce switched to a liberalism that was entirely reduced to a moral ideal of Freedom (always written with a capital). So, after 1924, Croce did become aware of the importance of the "ought" dimension, of those ideals that he had previously neglected; but in so doing he went to the other extreme. His liberalism is formulated only in moral terms, is only "ethical liberalism?'13 At the same time Croce never felt that he had to repudiate his former realism. Croce maintained to the end that politics consisted of nothing but expediency, utility, and force. The inevitable consequence is that Croce's ethical liberalism not only remains in midair but also is always undermined by a realism that is indeed inimical to liberalism. Whatever else the liberal civility may be, it surely is the civility that rejects politics-as-war. The liberal person, first and foremost, abhors violence and condemns direct action. Thus, by characterizing all politics as Machtpolitik, as power politics, Croce continued to deny-in his philosophy of Liberty-the typical modus vivendi of liberalism. Croce's solution was, then, to keep liberalism out of politics. In order to preserve his liberalism as a "pure moral ideal;' he refused to taint his ideal of Freedom with the empirical techniques and instruments of the liberal state. To the end Croce was to reject constitutionalism, and with it all the practical means of bringing the liberal ideal down to earth and translating it into reality. All in all, in Croce's philosophy the realistic and ethical elements are, at best, merely juxtaposed; facts and values proceed indepen45

THE LIMITS OF POLITICAL REALISM

3.3

dently of each other, like a world and a superworld. When Croce attempted to carry the political realism of his first period over to the philosophy of Freedom of his next period, his doctrine bifurcated between two positions, each of which by itself was inadequate. On the one side, it is power politics in Bismarck's version; on the other, it is an extrapolitical ethic set in a "higher and different sphere:'14 The result is that his political realism conflicts with his liberalism, and that his ethical liberalism contradicts his realism. Realism and liberalism can very well coexist, but they do not do so in Croce. His political philosophy is vitiated first by an excess of realism and then by an excess of moralism. Having started from an inaccurate distinction between real and ideal, or rather by paying no attention to it, Croce never found a way of connecting what is and what ought to be. The lesson yet to be learned from the exemplary case of Croce is, thus, the centrality of the fact-value relation. The standard wisdom is that facts and values interact. This is, of course, true; but hardly telling. It does not say in what respect the is and the ought must be kept separate, and in what respect we must never forget that the two poles are interdependent and complementary. With reference to the separation, a fact cannot refute a value, nor, vice versa, can a deontology reject a factual statement. On the other hand, in order to cover the whole case of democracy (or liberalism or socialism) we need both the facts and the ideals. It is only when these two preliminary points are kept in mind that we can spell out the proper interactions between "effectual truth;' on the one hand, and our value options in favor of liberalism, democracy, socialism, on the other. Briefly put, no polity can be construed only realistically or only idealistically. If we understand political realism as the search for the factual basis of politics, then the account of the realist stops just where liberalism, democracy, and socialism begin. For liberalism, democracy, and socialism are not the fruit of a Realpolitik but, let us say on symmetrical grounds, of a Phantasiepolitik. They are built on facts, but by fantasy. The same applies to the case of the idealist. His Phantasiepolitik has to take off from factual knowledge - otherwise it will lead nowhere, just to Erewhon. Ideals cannot replace facts, they are superimposed on them.

3.4

Mosca, Pareto, and Michels

The misunderstanding of realism can be condensed in two arguments. The first is this: I do not believe in democracy because reality is in contradiction to it. The second, its retort, is this: Being a democrat, I refuse to be a realist. Neither argument withstands scrutiny. If we do not believe in democracy, it

Mosca, Pareto, and Michels

3.4

is because we believe in different values and not because "reality" need exclude, or be in any way incompatible with, the choice of a democratic deontology. There is no contradiction between a realistic cognition and a democratic creed. If realism is "assessing the facts;' then there can be democratic, just as there is undemocratic, realism. It is bad logic to maintain that we do not believe in democracy because the facts disprove it. Conversely, it is equally illogical to reject a realistic finding because it does not sustain a democratic belief. The way to refute alleged facts is to adduce counterfacts; and a victory over a state-of-fact presupposes that we acknowledge its existence. Descending to a more concrete plane, the authors that have come to represent, somewhat paradigmatically, the realistic school are Mosca, Pareto, and Michels. This trinity was proposed, I believe, by James Burnham, who perceived the three authors as being, at the same time, Machiavellians and "defenders of freedom?' 15 In truth, Croce was far more a Machiavellian than the other three authors. Mosca was not an admirer of Machiavellil 6 and I find it very difficult, furthermore, to perceive Michels as a Machiavellian. However that may be, Burnham's message was that a Machiavellian could also be a defender of freedom. This is correct. Nonetheless, it is equally certain that Mosca was no admirer of democracy, that Pareto did not believe in parliamentary democracy at all (though admiring Swiss, direct democracy), and that Michels abandoned democracy. Thus the interesting question is: Were they antidemocratic because they were realists? And the complementary question is: Should we dismiss their realism because of its antidemocratic bent? Let me first take up the latter question, since it can be disposed of more easily. Gaetano Mosca's law of the "political class" (not of the ruling class) is undoubtedly vague, as his basic concept is much too loose} 7 Yet Mosca's intuition is either correct or false and cannot be dismissed by calling it fascist or a provocation. If a somewhat cohesive political class is found in all political systems, then his law has truth value. Conversely, in order to show that he was wrong we should either find political systems in which there is nothing that resembles what he understood by political class, or challenge Mosca's law on the grounds that he wins his case by definition, that is, by defining political class so broadly as to escape falsification.18 Similarly with Michels. I shall come later to his "iron law of oligarchy" and point to its many weak points.19 What will not do, however, is to simply brush aside his argument as antidemocratic. His law of oligarchy must be tested, to begin with, on the same grounds on which it was built-on a fact-finding basis. As for Pareto, there is nothing inherently undemocratic in his law of the "circulation of elites." Certainly Pareto denounced the hypocrisies of democracy as vigorously as did Croce, and indeed at a higher pitch than Mosca. 20 But this aspect biases

47

THE LIMITS OF POLITICAL REALISM

3.4

his scientific work only to the extent that we can show-as it can be shownthat under the cover of a "pure science" Pareto was delivering, more than anything else, his deep and polemical pessimism. Reverting to the first question, Can it be held that Mosca, Pareto, and Michels were antidemocratic because they were realists? The answer is both no and yes. No in the sense that the nexus is in no way an intrinsic, necessary nexus. The proof is that Mosca remained a realist but stood up, when fascism came to power, as a firm liberal2Lparalleling Croce's experience. Pareto died in 1923, and thus we cannot know what would have been his long-run reaction to fascism, which in 1923 had not yet taken on the outright form of a dictatorship. 22 As for Michels, he was a disappointed socialist who sought self-government and democracy in organizations. Having discovered that intraorganization democracy was impossible, he was ultimately able to align himself even with fascism-disillusion is an erratic state of mind. Not so Croce, the philosopher, or Gaetano Mosca, the professor of constitutional law. Neither felt bound by his own realism to approve an illiberal system. On the other hand, it is equally true that Mosca, Pareto, and Michels (like Croce) did derive and justify their dislike for parliamentary democracy from their findings. They all underplayed, though in different ways and for different reasons, the role of ideals and failed to perceive that a value choice is independent from, and by no means an extrapolation of, a state of fact. But if this is their mistake, then we should separate what they confused. That is to say that the antidemocratic bent of their theories is to be attributed to their "bad realism?' Conversely, insofar as their realism was truly cognitive, they were not in favor of, or against, anybody; they sought and discovered "laws" that we are still discussing, reformulating, and testing. Mosca, Pareto, and Michels were in error, then, when they held, or implied, that reality contradicts democracy. This is a non sequitur. But the theory of democracy shares in the same error when it holds, or implies, that democracy is repugnant to realism. 23 It is precisely by denying citizenship to a "democratic realism" that we help produce, over and over again, a situation in which the realists are excommunicated, and eventually helped to join the antidemocratic ranks. I very much doubt this to be in the best interest of democracy.

3.5 Realism versus Rationalism If the division between realists opposing democracy and democrats who reject realism is ill founded, we are left to wonder how it comes about. John H. Herz points out that political realism "arises inevitably whenever people become fully aware of the failure of repeated attempts to 'reform' political

Realism versus Rationalism 3.5 life, to create a 'better' world, or to oust the 'wicked: ... History, which is the burial ground of such attempted changes, is also the birthplace of realist disillusionment:' 24 Realist disillusionment, exactly. This does not mean, as it is often said, that realism spreads pessimism and engenders disappointment. Disillusion follows from illusion. It is idealism, not realism, that produces disillusionment. Realism would prevent it-if it were effective in time. A first consideration is, then, that realism tends to have an antidemocratic sequel in that it happens to be a retarded realism: It follows the disillusionment brought about by overidealistic policies, instead of preceding disillusionment and helping to prevent it. However, in order to get to the heart of the matter we must relate realism-defined as cognitive realism-to rationalism. For the misinterpretation of realism is characteristic of the rationalistic forma mentis, of the rationalistic mentality. At the source, it is rationalism that fights realism. Therefore, in order to touch bedrock it behooves me to begin by defining rationalism. In the present context, rationalism is best defined as the opposite of empiricism, or in contradistinction to empiricism. Let it also be repeated that both rationalism and empiricism are being conceived as mental orientations, as mentalities, mental patterns, or, perhaps better, mental mechanisms. Finally, it should be understood that "empiricism" is used here as a shorthand for both empiricism and pragmatism. If the empirical mind expresses itself in the cautious "wait and see" motto, the pragmatic mind transforms it in the adventurous, dynamic "try and see" formula. Despite this and other dissimilarities between the progenitor (empiricism) and its progeny (pragmatism), 25 these internal differentiations arise from a common basis and lose salience vis-a-vis rationalism; for the latter is just as distant from empiricism as it is from pragmatism. So, the contrast can be stated, in brief, as the opposition between rationalism and empiricism. A first difference is that the empirical (empirico-pragmatic) mentality stays in medias res, in the middle of things, and thereby close to what can be seen, touched, and tested, whereas the rationalist mentality soars to a far higher level of abstraction, to a level that is far removed from the facts. Thus, while the empiricist is inclined to work back from reality, the rationalist tends to remake reality as a reflection of "reason:' The empirical instinct is to look at how things work out; the rationalist inclination is to rebuild everything ab imis, starting anew from a tabula rasa. While the empirical tenet is that if a program does not succeed in practice there must be something wrong with the theory, the rationalist tenet is that what is true in theory must also be true in practice. Therefore, if or when things go wrong, it is the practice, not the theory, that must be wrong. Hegel's famous equation was "the real

49

THE LIMITS OF POLITICAL REALISM

3.5

is rational" (and vice versa). His disciples developed two interpretations of this equation, namely, (a) it is the rational that must adjust itself to the real; or, conversely, (b) it is the real that must submit to the rational. The latter was the interpretation of the Hegelian left and eminently of Marx. And there is little doubt that it is the "left version" that is congenial, or more congenial, to the rationalistic mind. The victory of Marx and the Hegelian left over the Hegelian right is, at the same time, a victory of rationalism. Perhaps the difference that generates all the other differences is that for the rationalistic Gestalt the criterion is coherence, not applicability. Rationalists are concerned with the construction of orderly logical relationships and do not wish to be asked whether such relationships tell us anything about the real world. Thus, while empiricism tends to be tentative, rationalism tends to be definitive; while the former is eager to learn from experience and to proceed by testing and retesting, the latter goes ahead without tests; while the empiricist pays little heed to rigorous coherence and long chains of demonstration, the rationalist believes only in deductive consistency. In the end, the empiricist turns out to be reasonable rather than rational, while the rationalist puts logical rigor above everything else and hence is rational even at the cost of being unreasonable. It may appear that the contrast is not as clear-cut as I have drawn it. Yet I do not feel that it is overdrawn. Of course, there will be individuals who balance and blend, within themselves, a rationalistic and an empirico-pragmatic orientation. These exceptions do not detract from the fact that when reference is made to cultures and cultural patterns, the two mental mechanisms fall wide apart. Until the end of World War n and the revolution in mass communication, our two mental patterns clearly characterized the AngloAmerican world vis-a-vis continental Europe. I am not implying that everybody who spoke English was an empiricist and that everybody who spoke French or German or Italian was a rationalist. Obviously, we are dealing with prevalences and tendencies; and it is also the case that every culture has its non-conforming, rebellious intellectual minorities. There is a rationalistic undercurrent in Anglo-American thought, just as we find an anti-rationalistic line (which is sometimes empirical but more often plainly irrationalistic) in the history of European culture. Nonetheless, even today American or English "rationalists" are much less such than their European counterparts. Conversely, the continental European "empiricists" still are an impure breed. As reciprocal communication exposure and actual mobility increased, the geographical lines of division have become increasingly blurred. Yet even the hybrids remain, to date, very different hybrids. By and large, Westerners are still marked either by a rationalistic or by an empirico-pragmatic underlying cultural imprint.

50

Rational Democracy and Empirical Democracy 3.6 Reverting to the initial point, How does realism relate, respectively, to rationalism and empiricism? the reply has already been detected, I trust, between the lines. The empirical mind instinctively conceives problems from a practical angle and is characterized by a matter-of-fact anchorage; and this is the same as saying that the empirical mind is characteristically "realistic:• A person nurtured by an empirical culture instinctively tends to use words as a means of arriving at the thing and, therefore, in their observational and descriptive meanings. This is the cognitive orientation of which realism consists. Conversely, rationalism is neither interested nor suited to describing the world as it really is. Rationalists are interested in constructing prototypes and seeking definitive solutions. They want to reconstruct reality in accordance with la raison, which is the same as saying that rationalism is poorly equipped to face practical problems and that an anti-realistic or, at least, an unrealistic attitude is congenial to rationalism. It is, then, within the ambit of a rationalistic culture that the hostility between realists, on the one hand, and democrats, on the other, is deeply rooted. The rejection of "realism" is a recent and still marginal import in the AngloAmerican world; it is, instead, a long-standing feature of the rationalistic democracies. And when a democracy is characterized by an anti-realistic bias, this starts a vicious circle. A democracy deprived of internal realistic correctives becomes more and more an "unreal" democracy in which rhetoric and deeds, ideals and practice, fall more and more apart. Under such circumstances it is only natural that political realism should link with antidemocratic stands. But, let it be repeated, this is not because realism is against democracy as such. It is, rather, the enemy of unrealistic, that is, rationalistic democracy.

3.6

Rational Democracy and Empirical Democracy

The preceding discussion winds up at this conclusion: Empirical democracies are naturally realistic, while rational democracies are apt to be anti-realistic. In the light of this conclusion the argument shifts to, and hinges on, the distinction between rational (rationalistic) democracy and empirical democracy. The former can be identified, historically, as French-type democracy. The latter is, at least historically, a democracy of the Anglo-American type. But the distinction needs buttressing. The question is: What is meant by rationalistic democracies, and what distinguishes them from empirical democracies? A first obvious difference is that whereas the democracies of the French type were born ex nova from a revolutionary rupture, the Anglo-American democracy is the outcome of a gradual and largely continuous process of historical growth. The English Revolutions did not vindicate, politically, a fresh

51

THE LIMITS OF POLITICAL REALISM

3.6

start but the restoration of the Englishman's "birthright;' that is, a pristine Anglo-Saxon constitution (a largely mythical one, to be sure) that had been affirmed against the Norman kings in Magna Charta and had been trampled upon by the Tudor and Stuart usurpations. As for the American Revolution, it was not, in truth, a revolution-it was a secession. The American Declaration of Independence of 1776 was, in essence, a claim for the right to advance along the path of the liberties already existing in England. The French Revolution was, instead, a very deliberate break with, and rejettion of, the past. This original difference implies a sequel of other differences. It would lead me astray to review them, but two points are worth recalling. As Bryce pointed out, the French people adopted democracy "not merely because the rule of the people was deemed the completest remedy for pressing evils ... but also in deference to general abstract principles which were taken for self-evident truths." 26 And Tocqueville seized on another important difference when he noted: "While in England those who wrote about politics and those who engaged in it shared the same life ... in France the political world was sharply divided into two non-communicating provinces .... In one [the politicians] administered; in the other [the writers) formulated abstract principles. Above the real society ... little by little an imaginary society took shape in which everything seemed simple and coordinated, uniform, just, and rational:' 27 The connection of French-type democracy wih rationalism (abstract principles and imaginary society), on the one hand, and of Anglo-American democracy with empiricism, on the other, has long been detected. But the theme still awaits systematic treatment. It should be understood that reference is specifically made, in what follows, to how a people conceive and construct their own democracy and, therefore, to internal (not foreign) politics. It should also be understood that while it can be plausibly argued that realism is nurtured by time, that is, by an adequate span of historical experience, the incidence of the time factor is a complex issue that I cannot afford to discuss. 28 Both empirical and rational democracies have popular sovereignty as their point of departure. Yet the English constitution does not recognize any such entity as "the people" as having a constitutional status. Likewise, the bulk of the American literature on democracy has addressed far less the question, What is democracy? and far more the question, How does democracy work? Until the early r96os, American authors have generally dwelt more on the instrumentalities of democracy than on the concept of popular sovereignty. In the overall, then, the characteristic of the Anglo-American theory of democracy is to be a theory on the ways and means of democratic government, hardly a theory on its ultimate principles and premises. By contrast, in the first article of the Constitution of the Weimar Republic-indeed a prototype

52

Rational Democracy and Empirical Democracy 3.6 of constitutional rationalism-we read that die Staatsgewalt geht vom Volke aus, that the power of the state goes from the people up; and this premise lays the foundation of a tightly knit deductive argument. Again, it is no accident that neither Rousseau's general will nor the Romantic Volksgeist have ever struck roots on Anglo-American soil. It is not only that "people" is, in English, a plural, while it is a singular in French, Italian, and German. The grammatical difference reflects a difference in levels of abstraction. The English people are concrete people; the French peuple (and equivalents) is an abstract entity, a one resulting from the abstractive elaboration that is congenial to the rational Gestalt. The difference, then, is that from the beginning of their respective constructions the rationalistic democracies have leaned heavily on a concept of the People (capitalized to render the difference) that is not even understood, let alone approved, by the empirical mind. The same difference that we fipd at the point of departure exists at the point of arrival. In English-speaking countries it is customary to speak of "government;' while Europeans almost always say "State" (capitalized). Now, there is the same distance between government and State as between the people (plural) and one People (singular). Once again, it is a difference in the level of abstraction. The rationally trained mind is concerned with the State and not with the government-and even less with governments-also because the latter are fluid and changing occurrences, while the State is a fixed structure. And, needless to say, the rationalist (unlike the empiricist) is ill at ease when confronted with muddling through, with shapelessness and fluidity. So, "State" does not mean in the Anglo-American context what it means in the European one. The empiricist, even when he switches from the concept of government to the concept of State, is likely to keep in mind that behind the entity there are still concrete persons. The rationalist, instead, has in mind a depersonalized, impersonal, juridical form. 29 Between the people and the state, the term a quo and the term ad quern, the difference that separates a rational from an empirical democracy can be ultimately reduced to whether the chain of argument is tight or loose, rigid or flexible. That is to say that rational democracies are constructed deductively and rigorously from premise to consequence, joining one link of the chain to the next one as tightly as possible; whereas the construction of empirical democracies largely results from feedbacks and, in this sense, from inductive elements. To illustrate, it is certainly no accident that democratic regimes in a large part of continental Europe evolved in the direction of parliamentary systems, if not assembly systems, whereas a similar development did not take place either in England (where parliamentary government is an inaccurate name 53

THE LIMITS OF POLITICAL REALISM

3.6

for a cabinet system) or in the United States. Likewise, it is surely not fortuitous that all continental European democracies have abandoned (or never adopted) the single-member district system and have basically settled for electoral systems of proportional representation. If the people's power premise is developed with deductive rigor, it follows that (a) true representation is, and can only be, proportional representation; (b) parliament must be the real site of representative sovereignty; and (c) governments should only be (as the wording says) "executives;• i.e., executors of the wills that precede the governmental will. In a deductive chain of reasoning all of the above is a must, a set of necessary logical consequences. How is it, then, that the Anglo-American democracies have not obliged to such "must"? In my argument it is because they are empirical democracies that are not construed deductively but, rather, on the experience that efficient and effective government is important, that assembly systems are malfunctioning systems, and that proportional representation may create more problems than it solves. The point is that the empiricist is impressed neither by "democratic consistency" nor by the display of a "well reasoned" democracy. The foregoing is not intended as an appraisal of the respective merits and demerits of rational and empirical democracies but as a way of buttressing differences. To be sure, it can well be held that the explanation of dysfunctional, so-called nonworking, democracy is precisely that it is a rational democracy. As Goethe said, there is nothing more inconsistent than supreme consistency30 - and the fruit of extreme consistency, at least in politics, is that the rationalistic democracies are always in danger of becoming imaginary democracies, far too removed from reality to be able to master the problems arising in the real world. But this is not the point that I wish to pursue. As I was saying earlier, rationality is not reasonableness, and reasonableness seems to descend from, and corresponds to, the empirical mentality. Probably, the democratic fabric and modus vivendi do require reasonableness more than Cartesian rigor. If this is so, and this is the note on which I wish to conclude, the worldwide appeal of empiricism is not in proportion to its merits. Rousseau, Hegel, and Marx travel throughout the world, and the last is read (and perhaps understood) even in China; whereas no English or American author has been able to gain any comparable influence outside the borders of his own culture. "Rousseau fired a thousand for one whom Benthamism convinced'~ it is Bryce who acknowledged it. 31 Rationalism travels and empiricism does not. Why? In order to be able to circulate a political doctrine must acquire a universality, a level of abstraction and a theoretical backbone to which the empirical mentality pays insufficient heed. While the diffusion and penetration of ideas-at least, in their ideological form-increase all over 54

Notes the world, the empirical mind often manifests itself into a "practicalism" (to use the term of William James) that makes a point of being able to do without ideas. To the extent that this is so, the Anglo-American culture may well train a formidable homo Jaber, but is in danger of not being able to live up to its responsibility to educate the homo sapiens needed to nourish and complement him. Thus, if the rationalist is not trained to solve practical problems, practicalism lacks an adequate intellectual grip and, indeed, intellectual force. It would be much to the advantage of both sides if the rationalistic and empirical approaches could meet halfway. This is the path that I attempt to pursue in this work.

Notes 1. H.D. Lasswell and A. Kaplan measure the difference between Aristotle and Machiavelli as follows: ''A rough classification of a sample of three hundred sentences ... yielded these proportions of political philosophy (demand statements and valuations) to political science (statements of fact and empirical hypotheses): Aristotle's Politics, twenty-five to seventy-five.... Machiavelli's Prince ... consisted entirely (in the sample) of statements of political science in the present sense" (Power and Society: A Framework for Political Inquiry [New Haven: Yale University Press, 1950], p. n8, n. 15). 2.. Until Machiavelli, "state" generally meant status, station. The term designates for the first time a political entity in the opening sentence of The Prince: ''All states and powers that have held and hold rule over men are ... either republics or principalities:' The new meaning reappears in this passage: "When the Cardinal of Roano said that the Italians did not understand war, I replied that the French did not understand the state" (chap. 3). However, both in the Discourses and in the History of Florence Machiavelli used the term state in the medieval sense. See 0. Condorelli, "II Nome 'Stato' in Machiavelli;' Archivio Giuridico 6 (192.3), pp. 77-n2.; F. Chiappelli, Studi sul Linguaggio de{ Machiavelli, and Nuovi Studi sul Linguaggio del Machiavelli (Firenze: Le Monnier, 1952. and 1969 ). 3. This development results from a secularization of politics that differed, in Italy, from what was happening elsewhere in that the Italian prince of the Renaissance did not cloak his power with any sacred or charismatic attributes: for him it was potestas, power, which gave him auctoritas, not vice versa. See F. Chabod, Del "Principe" di Nicolo Machiavelli (Milano: 192.6). 4. So much so that in Franz Neumann's tripartition of dictatorships, the type "caesaristic dictatorship" indicated the mass-approved dictator (this being the difference with respect to "simple dictatorship"). See The Democratic and the Authoritarian State (Glencoe: Free Press, 1957), pp. 2.33-47. 5. This inquiry is pursued in G. Sartori, "What Is Politics;' Political Theory 1

(1973).

6. See Carl Schmitt, The Concept of the Political, trans. G. Schwab (New Brunswick: Rutgers University Press, 1976), who provides the most articulate advocacy of

55

THE LIMITS OF POLITICAL REALISM

this approach. An excellent selection (by G. Miglio and P. Schiera) of his writings is C. Schmitt, Le Categorie de/ Politico (Bologna: II Mulino, 1972). 7. This point is addressed in chapter 4, section 6. 8. The analogy with Wertfreiheit (see chapter r, section r) does not entail that realism and freedom-from-value are assimilable. In themselves, value cancellation and/or value avoidance are not of cognitive import. Wertfreiheit may thus be a necessary but not a sufficient condition of descriptive knowledge. 9. Croce's production is truly enormous: some So volumes (in the final edition of his works) and some 4000 titles. For an assessment of his political philosophy, see N. Bobbio, Politica e Cultura (Torino: Einaudi, 1955), chap. 7 and esp. chap. 13; and G. Sartori, Stato e Politica nel Pensiero di B. Croce (Napoli: Morano, 1966). The essentials of Croce's political philosophy are in the writings collected in Materialismo Storico ed Economia Marxistica (1896-1906), in Etica e Politica (1915-30), and in the 1923-25 issues of the review La Critica. ro. Materialismo Storico ed Economia Marxistica (Bari: Laterza, 1919), Preface to 3d ed., p. xiv. rr. Filosofia, Poesia e Storia (Napoli: Ricciardi, 1951), p. rr72. 12. Pagine Sparse (Napoli: Ricciardi, 1943), 2:382. See also p. 373; and La Critica, 1 9 2 5, p. 347. 13. See in Etica e Politica (Bari: Laterza, 1943), the essay of 1926: "La Concezione Liberale come Concezione della Vita"; and Croce's conclusive article of 1939, "The Roots of Liberty;' in R.N. Anshen, ed., Freedom: Its Meaning (New York: Harcourt, Brace, 1940). Note that in Croce "liberalism" and "liberal" are used in their historical meanings (which are quite different from those in American usage, as explained in chapter 13, herein). 14. Etica e Politica, p. 284. 15. As his title says: The Machiavellians: Defenders of Freedom (New York: John Day, 1943). 16. This is well pointed out by J.H. Meisel, The Myth of the Ruling Class: Gaetano Mosca and the Elite (Ann Arbor: University of Michigan Press, 1958), pp. 246-85, esp. 264-65. 17. Mosca's major work bears in Italian the title, Elements of Political Science (1896), not The Ruling Class, as misleadingly translated in English. His doctrine is thoroughly analyzed by Mario delle Piane, Bibliografia di G. Mosca (Firenze: La Nuova Italia, 1949); idem, G. Mosca, Classe Politica e Liberalismo (Napoli: E.S.I., 1952). See also Ettore Albertoni, ed., Studies on the Political Thought of G. Mosca (MilanoMontreal: Giuffre, 1982). 18. This analysis is pursued in chapter 6, section 5. 19. In chapter 6, section 6. 20. In truth, as Bobbio correctly points out, "the democracy against which he [Mosca] constantly directed his attacks was the pseudo-scientific theory ... according to which the better political societies are those where the majority governs .... But if 'democracy' was understood in the only sense which, according to Mosca, corresponded with the facts, that is, the tendency which brings on a gradual or total renewal of the political class, he . . . was favorable to the development . . . of such a tendency, even if he wanted it to take place with the utmost circumspection" (N. Bobbio, "Gaetano Mosca e la Scienza Politica;' in Saggi su la Scienza Politica in Italia

Notes [Bari: Laterza, 1969], pp. 188-89). Actually Mosca's aversion to democracy was above all a condemnation of the parliamentarianism of his epoch. 21. On Mosca's liberalism (in the European sense), in addition to the works cited of Delle Piane and Bobbio, see P. Piovani, "Il Liberalismo di G. Mosca;' Momenti de/la Filosofia Giuridico-Politica Italiana (Milano: Giuffre, 1951); and A. Passerin d'Entreves, "Gaetano Mosca e la Liberta;' in II Politico 4 (1959). 22. That Pareto too would have resented Mussolini's dictatorship is suggested by G. La Ferla, Vilfredo Pareto Filosofo Volterriano (Firenze: La Nuova Italia, 1954), esp. pp. 160-71 (a work that approaches in a suggestive, though incomplete, manner the complex figure of Pareto). S.E. Finer reaches the same conclusion on the basis of Pareto's correspondence; see "Pareto and Pluto-Democracy;' American Political Science Review, June 1968. 23. This is very much the stance, today, of anti-elitists and, I believe, their error. See chapter 6, section 8. 24. Political Realism and Political Idealism (Chicago: University of Chicago Press, 1951), p. 27. 25. The descendance can be rendered as follows: English empiricism transforms itself into American pragmatism when faced with, and challenged by, a limitless virgin continent to be conquered. It is the "moving frontier" element that goes to explain, though not alone, the diversification. 26. Modern Democracies (New York: Macmillan, 1924), 1:208. 27. I: Ancien Regime et la Revolution (Paris: 1856), n, chap. 13, pp. 222-23. 28. This incidence may be left at noting that European short-lived democracies are unrealistic in internal matters but much less so in their foreign policy, where continuity has been manintained despite the changes of regime. Conversely, American democracy is internally realistic but highly idealistic in foreign policy, a relatively new ambit in its experience. 29. As correctly pointed out by F. W. Maitland, the reluctance of the English to use the word "state" is correlative to their reluctance to accept the doctrine of the juridical personality of the state. See "The Crown as Corporation;' in Collected Papers, vol. 3 (Cambridge: 19n). 30. Maximen und Refiectionen, 899. 31. Modern Democracies, 1:46.

57

4. Perfectionism and Utopia What has always made the State into a hell on earth has been precisely man's attempt to make it into his heaven.

-Holderlin

4.I

The Misunderstanding of Deontology

Thus far my keynote has been that democracy is endangered not by realism but by "bad realism;' by a misplaced and misapplied realism. My keynote now is, symmetrically, that democracy is endangered not by idealism but by "bad idealism;' by perfectionism. The two arguments are complementary: Bad realism is reinforced by bad idealism, and vice versa. At the same time, if perfectionism is abated, hyperrealism is also abated. In the cases of Croce, Pareto, and Mosca, it is clear that these authors reacted, first and foremost, to a bombastic rhetoric of democracy and that a more sober picture of an "imperfect democracy" would have blunted their arrows. In the case of Michels, it was an "impossible democracy" that led him to disavow the very possibility of any democracy.1 Much of the antidemocratic literature either arises from, or is accredited by, similar backgrounds. It is the formula vox populi vox dei, 2 that the voice of the people is the voice of God, that prompts the rebuttal vox populi vox diaboli, that it is the voice of the devil. 3 Just as a bad realism implies that there is a correct realism, so to speak of bad idealism implies that there is a correct idealism. Our problem is thus to underpin the difference between perfectionism (an excessive and misunderstood idealism), on the one hand, and the appropriate use and understanding of ideals, on the other. Every discussion of democracy basically revolves around three concepts: popular sovereignty, equality, and self-government. These concepts are interrelated. The people are sovereign in that they are equally sovereign (unequal sovereignty would imply that some people are sovereign and other people are not). And who is sovereign is not the object but the subject of governmenthence, self-government. These implications can be pressed further. To exemplify, if all are to be sovereign, the implication can well be that self-government

The Misunderstanding of Deontology 4.1 has to replace government (over the people); if all are equally sovereign, this postulates equal power, that is, isocracy; and from equal power one may also infer, albeit with some stretching, equality in general, or equality tout court. The question is: What is the exact nature of these concepts? Certainly they can be conceived descriptively. In such case popular sovereignty is little more than a principle of legitimacy, equality may be narrowed down to equal voting, and self-government will probably be confined to microdemocracies. However, the concepts in question are generally developed and conceived normatively. In the latter case they constitute the basic ingredients of the deontology of democracy. And the present argument assumes that popular sovereignty, equality, and self-government cannot be fruitfully discussed and employed unless we establish (a) that they are normative ideals; and, concurrently, (b) what is the nature, function, and purpose of ideals. The first step-establishing that an ideal is an ideal-appears to be, at first, a trivial one. Not so, as the debate on equality, a debate that has gone on from time immemorial, well goes to show. Jefferson's classic formula was ":All men are created equaI:' 4 For Jefferson this was a "self-evident truth?' Now, whatever the nature of self-evident truths, what is very evident is that his proposition is constructed syntactically as a statement of fact. It can only be defended, however, if reformulated prescriptively, or normatively, as follows: Thou shalt consider all men as if they were created equal. Of course, Jefferson was right (candor apart) in presenting this imperative as he did, since a self-evident truth is more effective than an exhortation. Yet, when the truth value of Jefferson's formula is questioned, then the objection should be met correctly. It is not a fact that all men are equal. Rather, the fact is that they are not equal but very different. Except that equality is, first and foremost, a value principle that commands rules of behavior such as that we must recognize ourselves in the other; or that it is our duty to treat others as our equals, not as our inferiors. There is no need to dwell on how much our civilization owes to these maxims. The point is simply that an elementary correction can cut short a fruitless argument. The statement "men are not equal" can be accepted as a fact by any democrat because it does not constitute an objection. 5 The next step is to inquire into the nature and function of ideals. As the inquiry unfolds it will also appear that the perfectionist is precisely someone who takes ideals for something they are not, who pays little attention to the necessary decalage between ideals and practice and, by the same token, who does not know how to turn, or how to apply, prescriptions to reality. The perfectionist characteristically attends to the maximization of idealshis foot is always on the accelerator-without being in control of the ideals that he propounds. 59

PERFECTIONISM AND UTOPIA

4.2

4.2

Myth and Utopia Reconsidered

While the misuse of realism goes back a long way and has possibly reached a point of exhaustion, the misuse of idealism is of relatively recent vintage and seems to be still striving for fulfillment. Two developments strike me as bearing out the heightening of perfectionism. The first is the passage from a contemplative to an "activistic perfectionism;' The second, and concurrent one, is the withering away of the sense of "the impossible." In the realm of contemplation the longing for, and the 1design of, an ideal world, indeed of an ideal counterworld, is possibly as old as humankind. Plato laid down the blueprint of this kind of mental exercise; and the state of nature doctrine also provided, from the Middle Ages up to Rousseau, an ideal parameter not only for evaluating the real world but also (in its natural rights development) for establishing values in the real world. In both cases, a state of perfection was discovered by, and coincided with, a state of contemplation. It is true that Plato moved to Syracuse in the hope of actualizing his ideal city. But his first visit in Sicily ended with his being sold as a slave in the marketplace of Egina; and his last one might have ended worse. In any event, Plato's solution was the philosopher-king. As for the state of nature, it was conceived, in the main, as a lost paradise. If and when heavenly cities were actively sought and established (as when Munster was proclaimed "New Jerusalem"), they were the children of chiliastic surges, of eschatologic expectations. Throughout the Middle Ages and all the way up to the Puritan saints in arms, the mobilizing force was religion, not Plato's eidos. Actually it was not until Marx that the philosopher-king ( king on account of being philosopher) was upturned into the "revolutionary philosopher;' 6 With the young Marx it is still the philosopher who is in charge because it is the philosopher who knows "what is rational"; but the philosopher must be a revolutionary, for it is revolutionary action that renders "real" what is rational. Thus with Marx perfectionism ceases to be an intellectual perfectionism, and a perfectionistic activism enters the realm of politics. Even so, more than a century had to go by for this process to gain momentum; the revolutionary philosopher obtained mass audiences and became a real mobilizing force only in the r96os. 7 This vintage is indeed a recent one. The other side of the same coin can be detected in the withering away of the sense of the impossible. It might be said that perfectionism advances as "impossibles" retreat. A development symptomatically revealed, I submit, by the alteration in meaning of the concept of utopia. Utopia was a term coined by Thomas More (as a simple transliteration of the Greek form) to connote a fiction that had "no place;' that is, a non-

60

Myth and Utopia Reconsidered

4.2

xistent world. Since More had a critical and constructive purpose in mind in describing his imaginary island, his non-existent was not declared a nonpossible. Yet the word inaugurated a mode of thought on the force, the semantic force, of its prefix, of its no. The utopian who neither calls nor conceives himself as being such-from Plato to Campanella- hopes in his perfect world. The utopian who calls himself such may desire his perfect world but does not believe in its realization. 8 The avowed utopian uses his utopia (for some purpose) but does not will it. Thus, what makes "utopia" distinctive-vis-a-vis "myth;' on the one hand, and "idealism;' on the other-is precisely that a utopian state of affairs is not only in no place but also in no time, in no future. In short, utopia is a non-existent in that it is a non-possible. This is not only the meaning retained by the word in ordinary usage but also the meaning that renders the term useful in our thinking. Karl Mannheim was probably the author who was most influential in destroying the signification that prompted More to coin "utopia?' Utopia, for Mannheim, is simply a state of mind that "transcends" existing reality in a revolutionary direction. But to speak of a state of mind that transcends what exists in a given time is not saying much, for in one sense or another the mind always transcends the existent. If we proceed to the differentiam, the specificity of Mannheim's utopia is to be found in its "revolutionary function?' This is confirmed by the fact that states of mind that transcend the immediate situation in a conservative sense are called by Mannheim "ideologies." In sum, Mannheim makes utopia into nothing more than the counterpart of ideology, placing revolutionary ideologies (rechristened utopias) on one side, and conservative ideologies (ideologies pure and simple) on the other. 9 Now, if the above is what we should understand when saying utopia, utopias are no longer utopias, that is, mental fictions located nowhere.10 And if the word utopia is divested of its specific, original meaning-function, one can well say, with Oscar Wilde, that "progress is the realization of utopias;' 11 with Lamartine that "it is possible that the utopias of today may become the realities of tomorrow;'12 with Mannheim that "utopias are often only premature truths;•u and with Marcuse that "the path to socialism may proceed from science to utopia and not from utopia to science?' 14 By manipulating meanings and definitions one can prove anything. However, the net result of the argument that "everywhere we see utopianism become reality" 15 is the exclusion from our political vocabulary of the term that eminently signifies the unrealizable. If we define utopia in a non-utopian way, we are left without a name to indicate the politically impossible} 6 Do we possess a fall-back position? For instance, can "utopia" be replaced by "myth"? Not quite, I would say. Myth does point to some kind of unreal-

61

PERFECTIONISM AND UTOPIA

4,2

ity, but it makes little sense to say that a myth is "impossible?' This implication falls beyond the semantic horizon of the term, which points, rather, to a make-believe. Note, in this connection, that a myth believer is not the victim of a deceit, since he chooses to believe in it. So, while a myth has no empirically verifiable existence, its kind of unreality is elusive and shades into some kind of belief reality. Even so, if a person asserts, for in~tance, that democracy is a myth or speaks of the myths of democracy, the ordinary listener is entitled to understand that democracy is some kind of imaginary entity. However, just as utopia no longer means utopia, myth no longer means myth (in the sense in which the term has long been used) to authors such as MacIver or Lasswell and Kaplan. To Maciver, myth is a neutral term that "abjures all reference to truth and falsity" and that covers all "value beliefs and notions that men hold"; hence everything is myth-except techniques!? On these premises, democracy too falls under the label "myth of democracy:'18 And in their proposed framework for political inquiry Lasswell and Kaplan follow suit. Under the caveat that "the term myth is not to be interpreted as necessarily imputing a fictional, false, or irrational character of the symbols;• their definition is as follows: "The political myth is the pattern of the basic political symbols current in a society:' Thus, the authors illustrate, "the present concept is close to a number of others ... Marx's 'ideology; Sorel's 'myth; Mosca's 'political formula; Pareto's 'derivations; Mannheim's 'ideology' and 'utopia: "19 The collection of references is dazzling and confirms that myth is, in this usage, just about everything. Inter alia, and certainly, popular sovereignty, equality, and self-government are myths. Does this provide new insights? I think not. To the contrary, the net result of bringing almost everything under "myth" is to create a haze in which analytic tools are of no avail. In particular, and with respect to my central concern, under this treatment there is no way of differentiating the proper from the erroneous use of ideals, that is, of looking into how ideals relate to reality. Be that as it may, the question was whether "myth" provided a fall-back position for conveying the meaning of "no place;' for indicating non-existents not predicated for existence. Having just seen that the destiny of the word myth closely parallels the destiny of the word utopia, 20 the conclusion is rejoined that throughout the current vocabulary of politics we no longer dispose of a term for "the impossible"; and if the impossible is unnamed, it also ceases to delimit "the possibles:• The Promethean myth exits from mythology and enters history. We live in a time in which scientific circumspection and scientistic euphoria both concur, paradoxically enough, in shading off the utopian into the merely improbable. Owing to scientific caution, we have made a point of never

Myth and Utopia Reconsidered 4.2 saying "impossible" and of speaking only of degrees of probability or improbability. And owing to scientistic optimism, we have come to believe that, in principle, nothing is impossible. Having lightly conjoined technical with ethical progress, the perfectibility of science with the perfectibility of humanity, our age reveres limitlessness-we are relying more and more on the expectations of a future of unlimited possibilities. Of course, if everything is possible, we no longer need a term to indicate the impossible. But if we renounce the term and the impossibilities remain, we are apt to drown in utopia before we know it. I take it, therefore, that we still need to use utopia in the core meaning in which the word was conceived. But I shall also propose an adjustment. Let then the definition be: Utopia is the contradictory claim of imagining an "impossible reality" for the purpose of realizing it. The above definition maintains the original connotation of the term; adds, I trust, to its sharpness; and updates the concept in that it reflects the transition from an intellectualistic, contemplative utopianism to an activistic, will-based utopianism. Utopia is no longer an intellectual game and an object of contemplation. The twentieth-century utopian has become a man of action. In this new light, or along this new dimension, we are immediately confronted by the question: How can we know what is utopian? As phrased, the question addresses a purely logical argument. We are in fact asking ourselves, How can we establish a priori what is possible and what is not? Still more pointedly, Is there a logical structure that represents an absolute impossibility, that is, impossibilities that are not contingent on time and place? Let me first take up this formidable question in the setting of life experience, that is, of the personal experiences that defy verbal pyrotechnics and the masqueradings of the verbalist. For instance, has anybody ever succeeded in having his cake and eating it? I believe not. The dictum that you cannot have a cake and eat it points to an absolute impossibility that has never been and will never be falsified. In like manner we can confidently assert that it is impossible to sleep when awake, to spend and save the same money, or to take two roads at the same time. Unquestionably, all these are unconditional impossibilities. Can it be maintained that what applies to the personal experience of "each body" no longer applies to the "allbody;' that is, to history? For all we know, the opposite is far more likely to be true. What is still possible on a small scale often becomes unmanageable, and thus less possible, on a larger scale. Magnitudes add, rather than remove, complications and obstacles. The real difference, when we pass from the circumstances of everyday life to political impossibilities, is that the latter are far more difficult to pin down. There are two reasons for this. First, in our personal experience impossibility is generally revealed immediately and directly to us,

PERFECTIONISM AND UTOPIA

4.2

as an impossibility to do, whereas in politics it is revealed late, indirectly, and as an event-related impossibility to obtain. But the fundamental reason is that the possibles-impossibles of politics escape control because the discourse on politics escapes control. By burying facts under verbal veils, false victories are easily made to seem real victories, and failure is easily vindicated as success. By incessantly repeating that history is a sequence of impossibilities that were made possible, or a cemetery of utopias that were realized, we have persuaded ourselves that this is truly the case. But I have yet to find an author who seriously spells out in which actual instances this has been so. Waiting for such probing, I am struck by how scarcely populated that cemetery remains. The decisive question remains whether there is a logical structure that yields, in politics as well as in any context, an absolute impossibility. I submit that the reply is yes. Utopianism consists of practical contradictions whose logical formulation is this: One cannot obtain more of two things that require contrary actions. I say practical contradictions-this being the crucial clause- because sheer logical contradictions made of sheer words can be construed as easily as we breathe. The contradictions must bear, then, on courses of action. Note, also, that two goals may be declared contradictory and yet afford tradeoffs. In this case, however, the logical structure of the argument is: One can obtain more of one thing if one demands less of its contrary. Thus one can still eat more of a cake and have less of it. The impossibility arises, and is unconditional, when one wants, all in one, more eating and more cake. So, it is perfectly possible to identify a utopia, the non-realizable, in time and ahead of time. The foregoing represents the logical structure of what we often call "dilemmas" of politics, of the grand dilemmas whose solution is compared by Rousseau to the squaring of the circle. We shall encounter many of these grand dilemmas as we proceed. At the moment it should be understood that the preceding argument does not imply that all impossibilities must be conceived and argued in that mold- as I shall go on to show.

4.3

Self-Government and the Politically Impossible

Let us take Marx's utopia, which was to be achieved by the replacement of representative democracy by a literal self-government arising from the withering away of the state. In my opinion, we are justified in considering Marx's self-government an unfounded hypothesis located nowhere, the absolute impossibility of which can be demonstrated a priori. Self-government is conceptually easy to define: It consists of governing ourselves by ourselves. When the notion is applied to the real world, how-

Self-Government and the Politically Impossible

4.3

ever, it must be weighed empirically. This weighing is required to establish the intensity of a self-government, that is, whether and to what extent the word applies (to its referents) in a strong or feeble meaning. For the simplicity of the argument I shall say that a self-government is at maximal intensity, or all the more intense, the more closely a governing corresponds to what the word literally says. Conversely, a self-government is at minimal intensity when the referent bears only a very loose and distant resemblance to what "governing" conveys. Now, it is clearly the case that the intensity of a selfgovernment (as defined) varies in relation to its extension. In the first instance let this extensiveness be identified as a spatial extension, as a size. 21 On these premises I advance the following proposition: The intensity of self-government attainable stands in inverse relation to the extension of self-government demanded.

let us demonstrate, starting from the hypothesis of maximal intensity.

If we think of self-government in the intense, i.e., strong and literal sense of the term, we are referring either to individual inner self-government (selfdetermination) or to the perfect despot. In either case we have maximal intensity-the maximum self-government possible-because the extension is zero (self-government in interiore hominis ). let us next introduce a small extension such as the space of the polis, the ancient Greek city. In this case selfgovernment will display less intensity, that is, it will have to be interpreted in a less exacting and less literal sense than before. By saying self-government we now refer, at best, to a rapid and comprehensive rotation between power holders and power addressees. As Aristotle said, in ancient democracies the citizens governed themselves in the sense that all governed and were governed in turn. Proceeding a step farther, let us assume that self-government is to function over a region or a large city. In this case the extension of self-government sought is already such as to permit only a metaphorical use of the concept. At this point self-government basically means being governed from nearby instead of afar; the term still is meaningful only in contradistinction to centralization. The self-government in question actually denotes autonomous local governments, which can be called self-governing only in that they permit a greater degree of self-government than political systems characterized by centralization. However, the local government of a large city is already a system of indirect government; it is not a self-governing but a representative democracy in which we are governed by intermediaries. So, here self-government ends-when the word means what it says. What do we mean, then, when the term self-government is applied to the extension of sizable nation-states? In this case it is hard to see how the

PERFECTIONISM AND UTOPIA

4.3

citizens could govern themselves in any meaningful sense of the expression 22 unless we mean something else, either control over government or a state of national independence. I have no quarrel, provided that it remains clear that neither independence from other countries nor control over the government is the same as self-government. I do, however, object to any further stretching. In particular, the ultimate design of a world system of billions of people who govern themselves united into a single self-governing system may warm our hearts but simply ignores the problem of intensity, that is, that only a zero intensity of "self" can correspond to the extension "world:' In that imaginary self-government there will be nothing remotely resembling the faintest imitation of selves who govern. 23 Until now I have referred to size. But self-government involves not only a spatial but a temporal extension, a time dimension. The philosopher (the pure theorist) need not concern himself with either the duration or the chronological order of his referents. The reason for this is that a philosophic inquiry concerns atemporal relations, logical and not chronological sequences. But when we come to the empirical level, duration and chronological order do count. Operatively speaking, we cannot do first what has to be done afterward, and there are many things that we can do for a short time but not forever. Let us, therefore, move from extension in space (size) to extension in time (duration). If so, our proposition may be reformulated as follows: The intensity of self-government possible stands in inverse relation to the duration of self-government demanded. This means that a maximal intensity of self-government, such as we have in moments of heroic or revolutionary tension, can correspond only to a minimal duration. In outlining his ideal democracy Marx took the revolutionary episode of the Paris Commune of 1871 as a model.2 4 But can one project a minimal time span into a sort of eternity, into the Hegelian end of history? Can it be overlooked, that is, that the high degree of intensity of the Paris experiment in the self-government of the proletariat ( under emergency, siege conditions) lasted for a very short time? Marx's entire political prophecy-a stateless society with literal self-government-rests not only on "the fallacy of the dramatic instance" but even more on what may be called "the fallacy of timelessness." In Marx, and in the communist theory of society in general, an intensity flash is construed as being continuously and endlessly reproducible over time, in total disregard of the most unfailing of all the experiences of humankind, namely, that duration brings about routine, that honeymoons are not eternal. We can run for hours, but not for years; we can swim for a day, but not endlessly. Marx even failed to suspect-we must assume from

66

The Role of Ideals

4.4

his writings-that the intensity of a self-government covaries negatively with both its extension in space (Paris, after all, is not France) and its duration in time. As a philosopher speculating about a timeless and placeless selfgovernment of the proletariat, Marx is interesting. But as a revolutionary philosopher bent on changing the real world, he is an unalloyed utopian. His political design of replacing governments by self-governing was impossible then as it is now, and will be just as impossible forever. Formerly our ancestors relied too much on argument and too little on testing, and exaggerated the "impossible" vis-a-vis the "improbable?' Nowadays we have gone to the opposite extreme of saying improbable even when we should say impossible. No doubt, if everything must be proved by evidence and experiment and nothing can be demonstrated by argument, then the range of discourse has to be limited to mere probables, to statements and predictions whose truth value can only be that they are likely or unlikely. But evidence and experiment cannot replace argument any more than the latter are a substitute for the former. Impossibilities- I insist- can and must be demonstrated by argument. If we are mistaken, this can again be discovered by argument. But let us not wait for utopia-provoked disasters to occur in order to acknowledge that impossibles exist.

4.4

The Role of Ideals

I have dwelt on the conditions that favor the heightening of perfectionism. If the very prediction of impossibles is declared impossible, inflated and utopian ideals are left to prosper unbridled. This adds to the reasons for inquiring about the nature of deontological discourse, for understanding what ideals are for and thus for mastering the relationship between ideals and the realm of practice. Benjamin Constant, who was deeply aware of this problem because it had been the problem exploded by the French Revolution, sought to solve it by means of intermediary principles, that is, by interposing a mediant term between "first principles" and reality. He wrote: "When we toss into the midst of a society of men a principle divorced from all the intermediary principles that bring it down to us and adapt it to our situation, we create great disorder; because when this principle is torn from all its links and deprived of all its supports ... it destroys and overthrows. But it is the fault not of the first principle but of our ignorance of the intermediary principles?' 25 In the final analysis the problem undoubtedly is, as Constant saw, to adapt the absoluteness of principles to the real world by submitting the deontology to an intermediary reelaboration. And Constant was perfectly right when he stressed

PERFECTIONISM AND UTOPIA

4.4

that "each time a principle . . . seems inapplicable, it is because we do not know the intermediary principle which contains the means of application:' 26 But I would like to take up the argument at an earlier point. Before dealing with the applicative passage, we must be clearheaded about its beginning. Constant said "principles;' and I have spoken somewhat interchangeably of ideals, norms, prescriptions, and values. Norms and prescriptions do not pose problems in the sense that they can be imputed to, and derived from, values and/or ideals. Values ultimately are, instead, intradable entities. But is it necessary for us to get involved in the question, What are values? At this stage I think not. 27 I think not because we can do just as well, if not better, by replacing "values" with "ideals'~ and ideals are much easier to handle and explain. Ideals are born from our dissatisfaction with reality and thus represent, in their genesis, a reaction to what is. 28 If so, ideals can be defined as pictures of a desirable or desired state of affairs that never coincides - by definition -with an existent state of affairs. From their etiology one can equally derive that ideals counteract and/or combat reality. That is, ideals accompany the vicissitudes of history, ever present as their non-acquiescent, countervailing, or adversary element. Are ideals realizable? The question can be answered affirmatively, but can also be answered negatively. An ideal can be realized if we mean that ideals are realizable in part. Conversely, an ideal cannot be realized if we mean that ideals are realizable in full. The assertion that ideals are realizable (in part) is hardly disputable. The assertion is true not only in the obvious sense that ideals have, in general, an efficacy but also in the sense that we do find ideals actually implanted in the real world. Liberal democracy eminently attests to a realization of a set of ideals. The disputation begins, then, when we assert that ideals are not realizable in full, that is, neither literally nor to a fully satisfying extent. Yet this negation is consistent with the preceding affirmation. When we say that ideals never are (in fact) and cannot be (in principle) fully and adequately realized, we are simply saying and assuming that ideals are destined to remain ideals. An ideal is an ideal precisely in that it is not realized, in that it transcends the existent. The point is, in itself, an obvious one-but its implications are less obvious. Furthermore, from the vantage point that ideals cannot be literally realized, we can correctly appreciate how ideals are in fact managed or mismanaged. That ideals are destined to remain what they are implies that ideals are not meant to become facts-exactly in the sense and precisely for the same reason that an ought is not meant to be an is. The ulterior and more provocative implication is that ideals are destined, literally speaking, not to sue-

68

Maximization, Opposite Danger, and Inverted Results

4.5

ceed. At first view, this may seem a paradoxical assertion. Yet it goes to explain what ideals are /or-their reason for being-and why they are employed as they are. Ideals always smack of hybris, they are always excessive. This is as it should be, since ideals are designed to overcome resistances. But if this is so, then to assert that ideals are destined not to succeed is a way of assessing how ideals function and are "functional" to their telos, to their end purpose. To be sure, from the vantage point of effectiveness the actor is required to perceive his ideals as if they were meant for realization. But from the vantage point of the observer the function of ideals is to challenge facts. And the truth of the matter is, I shall contend, that ideals better reality precisely when they are not meant as a reality. That an ideal is not meant to replace what is real amounts also to suggesting that if, as a reduction to the absurd, it were possible to convert ideals exactly into realities, perhaps we would not cherish them anymore. Ideals warm our hearts as they remain at a distance. The rewards are often in the striving more than in the achievement. "ls there anything in life so disenchanting as attainment?" Robert Louis Stevenson's question was rhetorical. To quote him again: "To travel hopefully is a better thing than to arrive:' 29 Literal equality, as exemplified by billiard balls, would be, in all likelihood, as displeasing as the inequalities it opposes. Nothing is more boring, more insipid, than utopia, as one discovers when reading how life passes in utopias. Guns (our ideals) are useful to win battles; but we do not fight a war to win guns. Or do we? So, ideals are realizable but are also non-realizable. Having established in what sense ideals cannot pass into reality, we may now confront the question: How do ideals pass into reality-to the extent that they do-with what end results? We are back to the problem of Benjamin Constant, except that I shall seek a conversion rule rather than specific "intermediary principles:'

4.5

Maximization, Opposite Danger, and Inverted Results

In the preceding discussion the role or function of ideals has been described as reactive, polemical, countervailing, adversary. Admittedly these are very general qualifications. But in order to be more specific we must introduce the contexts, the factual settings, within which ideals perform; and we must distinguish, at the minimum, between the democratic ideal without a democratic system and the democratic ideal within a democracy. In the first setting (outside of democracy) the democratic ideal performs as an adversary ideal; its overriding purpose is negation, the overthrow of the political system that it fights; and the more the ideal is maximized, the

PERFECTIONISM AND UTOPIA

4.5

greater, possibly, its efficacy. In an autocratic context, th·en, democratic perfectionism serves its purpose. A purpose that is "realized" when a democracy takes the place of its defeated adversary. That democracy, no doubt, will be a highly imperfect realization of the ideal; nonetheless, it cannot be disclaimed as a non-democracy, for it was actually set up by a democratic deontology. So, in the new context the deontology no longer confronts an. enemy but the polity that it has in fact generated. This difference cannot be ignored- it makes all the difference-and is actually reflected in a new role ahd configuration of the deontological pressure. The normative or prescriptive definition of democracy characteristically establishes, in the new setting, ideal standards, ideal yardsticks, for evaluating and monitoring the real-world accomplishments of democracies. This layout is distinctive in two major respects. First, the evaluation is "critical" and is not, or no longer, inimical, adversarial, or of "negation?' Second, the ideal standards in question do not idealize other worlds, replacement worlds, but their world. Therefore they are to be construed as "constructive ideals;' in the sense specified by Dennis Thompson: Picture a state of affairs intended to be realizable incrementally, "by trends and non-radical reforms?' Still more pointedly, "to justify a claim that a constructive ideal is realizable, we need evidence to reveal trends which, if continuing, tend to realize the ideal; and to suggest that certain non-radical reforms, if effected, tend to realize the ideal." 30 Thompson perceptively brings out here the point at which ideals interact with, rather than react to, the real world. This does not mean that ideals can ever and entirely be subdued by findings. Nonetheless, an ideal cannot claim to be constructive unless it listens to, and learns from, experience and evidence. This is the test of its constructiveness. In the light of our previous analysis the foregoing demonstrates how ideals can be realized and furthered. What remains to be seen is at which point, or under what form, ideals cannot pass into reality and become self-denying. The setting is now one in which the ideals of democracy deal with their own creature; therefore, the strategy that applies when democracy is only an adversary ideal no longer applies. It is no longer true, that is, that ideals can be safely maximized beyond measure. To the contrary, if the deontology remains unchanged, it begins operating in reverse. When, within a democracy, we retain the democratic ideal in its extreme form, it begins to work against the democracy it has generated; it produces inverted results. That is why, as Herz points out, "political idealism has its time of greatness when it is in opposition to decadent political systems. It degenerates as soon as it attains its final goal; and in victory it dies?' 31 The frequency of that parable does not attest, however, to a historical inevitable. It attests, rather, to an incessant, 70

Maximization, Opposite Danger, and.Inverted Results

4.5

and self-reproducing, mismanagement of ideals. The question thus turns to their management. Since ideals are demands designed to meet resistances, the demand is constructive when gauged to the resistance. In rule form, the logic underlying the successful application of a deontology to the real world can be rendered as follows: To the extent that an ideal is converted into reality, to the same extent it must become feedback monitored. Let this be called the "feedback rule:' Let it also be noted that in the light of such rule the problem no longer is maximization but optimization. The democratic principle in its pure, maximal state calls for "all power to all the people:' But, as we know, this pure principle establishes only a titular right and is of no help whatsoever in implementing the entitlement with the exercise of power. 32 So, here we need the intermediary principles spoken of by Constant. The intermediary principle discovered thus far is representation (as actualized in the garantiste structures of the constitutional state). What does representation accomplish? What does it bring about? First, it reduces power to less power; in a system of representative government nobody is in a position to exercise an absolute (i.e., limitless) power. Second, and concurrently, in a system of representative government the people actually exercise power (political power) by being able to control and change the people in power. With all of this, it is still the case that the original principle is far from being fulfilled: The people still do not exercise power in any full or literal sense of the expression. What do we do next? The temptation, and indeed the easy path, is to repropose the principle in its purity. If so, the intermediary structures (the representative state) are no longer seen as means of implementation and appreciated for what they have achieved; they are perceived, rather, as obstacles and thereby dismissed as impediments along the path of the realization of the ideal. But if we take this course, then the ideal starts operating in the reverse, destructively instead of constructively. Note that the principle as such, in its literal wording, does not point to a limitation of power: ''All power" is a limitless power. This is right in the oppositional setting in which the principle was conceived; but, in victory, "all power to the people" no longer has the effect of limiting power; its actual effect is to affirm an absolutist principle. The rebuttal is that while the absolutism of an autocrat is bad, the absolutism of the people is good. Maybe. The point remains that the principle attributes only a titular right. Therefore a maximization of the ideal that disdains, and rejects as an impediment, the intermediary structures that deal with the exercise can only generate, as such, an absolute power in the name of the people. So, the ideal has, in the end, destroyed its creature. Let us, then, start again from the beginning. 71

PERFECTIONISM AND UTOPIA

4.5

The beginning is, very plainly, that the sine qua non condition for the people to "have power" in any meaningful sense of the expression is that they impede any unlimited power. This is the condition that must be intransigently respected throughout our efforts to maximize the ideal, that is, throughout efforts to add more power to the power that the people actually wield. In terms of our feedback rule this entails that the principle "all power to the people" must be gradually modified, as a democracy develops itself, into the principle all power to nobody. ' It may be argued that the feedback rule overturns the maximization into a minimization of ideals. It seems to me, however, that it is equally true, and perhaps even more appropriate, to speak in this connection of optimization. Also, the transformation of the letter of an ideal may not be at all a transformation of its meaning function, of its "spirit." The democratic principle was conceived as a battle cry, and the battle cry against absolute power is "all power to us:' But when that battle is over, if we stick to the letter of that principle, we lose the concrete benefits that it was truly intended to achieve. If, then, we do not want our ideal to die in victory, we must never lose sight of the "opposite danger?' 33 Since the battle cry "all power to us" is tied to the enemy it was combating, unless it is gradually converted into the principle that nobody should have all power, it will simply bring a reinforced absolutism,_back to life. Its practical effects will be the exact opposite of its intended effects. We endlessly hear talk of ideals being betrayed. If ideals are, by definition (in the definition given), destined not to succeed, then the perfectionist will always and easily discover in anybody else a traitor to ideals. However, the perfectionist too is, in his manner, a traitor to ideals: By sticking to their letter, the perfectionist betrays their intent. Ideals can be betrayed in many ways, and the most insidious one consists of ignoring that the meaning-function of ideals changes as their real-world environment changes. There are not only many ways of betraying ideals but also of being betrayed by ideals. One of them, and probably the surest one, is the way of the perfectionist.

4.6

The Revolution as Myth

Until now I have considered how the ideals of democracy may be either constructive or destructive vis-a-vis their own creature. I have not envisaged, that is, the ideals that propose a replacement world, a better world intended to supersede liberal democracy both in its reality and in its values. To be sure, in a time of democratic confusion a replacement world can always be declared "more democratic:' Yet there is little doubt that not only the real-world de-

72

The Revolution as Myth

4.6

mocracies but also their ideals are fundamentally rejected by the people who at different points in time have variously identified themselves with communisttype replacement worlds. The idols-Stalin, Mao, Castro, Ho Chi Minheventually fall, fail, or are dismissed by their successors. They still attest to a wholesale rejection that has entered our time, that precedes the surge of the 1960s, and that remains with us, if abated, in its aftermath. It would be ludicrous to seek a monocausal explanation of this rejection. Even if the argument is confined, as it is here, to the ambit of the Western democratic world, certainly many factors enter the passage from democracy to, say, post-democracy. Perfectionism is only one of such factors. Still, it is an unduly neglected one-and this consideration prompts me to insist on it. I have been saying that perfectionist attempts at realizing ideals literally are inevitably bound to fail and, indeed, to boomerang. This outcome feeds, in turn, a vicious vortex. To the extent that the perfectionist attributes failure to the intrinsic wickedness of the present-day world, he is easily led to become a preacher, when not a practitioner, of violence. Indeed, in the present-day world the cult of violence finds little support among realists; it is almost entirely the brood of idealists. 34 But the blame can also be put, disjointly or concurrently, on the ideals. Since the attempt to enforce utopian ideals never succeeds, the perfectionist may also come to disbelieve the ideals of democracy. If so, he seeks replacement worlds, either by imagining future ones or attributing imaginary virtues to existing negations of his own world. In either case, his solution becomes "revolution" and revolutionary violence. The word revolution (like the word democracy) is being stretched to almost no end. 35 How is one to distinguish a revolution, on the one hand, from uprisings, rebellions, revolts, insurrections, civil wars, and, on the other hand, from palace revolts, coups d'etat, and military coups? Certainly a revolution is not a mere coup; it is also more than an uprising triggered by famine, poverty, or intolerable oppression, that is to say, more than a revolt or an insurrection. And while a revolution may bring about a civil war (between revolutionaries and counterrevolutionaries), in themselves civil wars need not display revolutionary characteristics and may simply end in secession. In its distinctiveness, then, "revolution" is a violent, mass-supported seizure of power that entails a basic restructuring of the polity. This is, to be sure, a strict political definition of the revolutionary act-not of revolution-and-after, of revolution as an extended or even near-endless process. Whether the revolutionary event also enforces ex post a basic socioeconomic restructuring of society is immaterial to the political definition. 36 The one defining property that distinguishes ex ante a revolution from other mass and violence-based occurrences is a belief basis: A revolution is such in that it is belief-mobilized and affirms

7.3

PERFECTIONISM AND UTOPIA

4.6

a set of counter-beliefs (vis-a-vis the regime it overthrows). What, then, are the characteristics of .the revolutionary subcultures of our time? The question specifically addresses, remember, the point at which our (Western) perfectionists give up the furthering of democracy and seek a better world in the very replacement of democracy. The divide is, of course, in the mode of enactment, in the advocacy of revolutionary violence. But the revolutionary subculture of our time is also characterized by novel and somewhat unique traits. Formerly, revolutions were conceived as rebellions against tyranny. Today, they are conceived by the revolutionary groupuscules as embodying an intrinsic soteriological value. 37 Revolution is beautiful in itself. This type of revolutionary has no need to explain why he revolts; he revolts anyhow, unqualifiedly, on the grounds that revolutionary destruction is in itself creative. If so, it becomes unnecessary to inquire into what the postrevolutionary replacement worlds look like and, concurrently, into their feasibility. Are these replacement worlds impossible or possible, utopian or not? Engels left us with the magniloquent sentence that, after the proletarian revolution, communism "is the ascent of man from the kingdom of necessity to the kingdom of freedom:'38 Marx left us with the utopia derived from the Paris Commune. A long time has since gone by, and what is supposed to come "after the revolution" remains-in its configuration-a well-kept secret. If one wants, Mao added to the picture the "cultural revolution"; but this is only the same thing over and over again: the permanence of revolution. And in Mao's case the Great Leap Forward turned out to be a great leap backward. So, we are left with one discovery: revolution itself in its alleged creativeness. We must, therefore, settle for what we have. But what is it, precisely, that we have? Among the three categories under scrutiny-ideal, utopia, and myththe last, it seems to me, has the best fit for how our revolutionaries conceive revolution. Of course, we may say that revolution is an ideal; but surely not in the sense in which equality or liberty are spoken of as ideals. Ideals describe a desirable end state and are required to have a positive layout, whereas the notion of revolution unfolds itself in the negative. On the other hand, when we speak of revolution as a utopia, it is clear that reference is made not to revolutions themselves (they occur all the time and_ everywhere), but to their aftermath and outcome. With respect, then, to the revolution in itself, the telling category is "myth:' As indicated earlier, the term myth is commonly understood as, and associated with, a make-believe that is also a beliefreality. However, the term also obtains special, ad hoc meanings, such as the one proposed by Georges Sorel. Sorel is the author who present-day eulogists of revolution cite least. This is a pity on two counts. In the first place, Sorel was in many ways the first

74

The Revolution as Myth

4.6

major incarnation, after Marx, of the revolutionary philosopher. 39 In the second place, and even more to the point, he is the one author who does sustain with argument, not just by slogans, the attribution of a constructive value to revolutionary violence. To Sorel, utopia was the last manifestation of faith in reason, whereas myth was the rebellion against reason. Hence he conceived and advocated the general strike-the self-liberating act of the proletariatprecisely as a "myth:' His myth was, then, a global intuition that coincided with an act of the will. And it was to be a truly liberating force because it would enable the proletariat to act without helpers. A myth-driven proletariat would dispose not only of past but also of incumbent masters, namely, of self-proclaimed vanguards. It may be said that here utopia reenters. It does not matter, for the point of interest is that Sorel's political myth is the collective, deliberate, willed expectation of the event that the expectation in fact generates; and this is very much the way in which the "myth" of revolution is employed and deployed by present-day Western revolutionaries. Still, to understand the mythical use, in Sorel's sense, of the notion of revolution does not even begin to demonstrate the alleged creativity of revolution. On what grounds have we come to take for granted, in many intellectual quarters, that revolutions are, as such, "creative"? If we pause to think, this is not at all a self-evident assumption. To the contrary, it is an utterly counterintuitive presumption. Revolutionary action is warlike action; it kills, destroys, and conquers at gunpoint. If anything, it can be argued that revolutions are even worse than wars. When a war is over, it is over. Whoever has survived, whether winner or loser, switches back to a different, much preferred state: a state of peace. Revolutions, especially the total "extended" revolutions of our time, tend to be endless. After ten, twenty, fifty years, people are still thrown into concentration camps, arrested and sentenced "in the name of revolution;' and politics continues to display the naked brutality of a war against enemies. 40 So, where is the creativity, in what does it consist? In Hegel war is compared to a wind that blows over a marsh, substituting clean air for the miasma. That much can also be said-here the analogy holds well-of revolutions. But this is a temporary benefit; the wind falls, and the miasma reappear~. In behalf of the creativity of war the strong argument is that war efforts stimulate technological discovery and its rapid exploitation. However, since revolutions are not won on technological grounds (the insurgents do not win because their weaponry is a more sophisticated one), it is very dubious whether revolutions stimulate technological advance. From this latter vantage point war fares better than revolution. On the other hand, the revolutionaries may have a point if they say that war kills indiscriminately, whereas they kill the right people. Even so, and again, where is the creativity? 75

PERFECTIONISM AND UTOPIA

4.6

The answer ends up by being of this sort: It cannot be doubted that the English and, even more, the French revolutions have created the present-day Western civility,41 and that the Russian and some subsequent communist revolutions have generated new worlds. Nothing of the above can be doubted; but the instances in question do not attest, in truth, to a creativity of the revolutionary act as such. These revolutions exploded the obstacles and removed the impediments to the potentialities that not only gave cause to the conflagration but that subsequently delivered the goods. Per se, the merit of a revolutionary destruction-for destruction it is-is only to permit a reconstruction. But if no reconstructive potential precedes the revolution, no new positive construction (that is, reconstruction) can follow the revolution. Something must preexist the revolutionary act in order to come into existence after it. The benefits that followed from the French Revolution were all spelled out in the demands that "caused" the 1789-94 events, that is, they all derive from seeds that came to maturity during the Enlightenment. There was no utopia in what Locke and Montesquieu had conceived, in their liberal civility; and these (not Terror) became, during the nineteenth century, the realizations of the 1789 revolution. The argument is not dissimilar if we turn to Russia. There is no question that well before 1917 Russian society disposed of, and displayed, outstanding capacities. At the turn of the century Russian mathematicians were in the forefront of mathematics, and if the military capabilities attest, as they generally do, to the state of technology, then the estimate of the experts was, just before the outbreak of World War I, that by 1917 Russia would have attained a military might equal to that of Germany. Estimates aside, by 1914 Russia had become the fourth industrial power of the world. While, then, the Soviet revolution was constructive in destroying the tzarist impediment, it remains an open question whether Russians might have not progressed far more than they actually have without stumbling, with Stalin, into another blocking system. 42 That question will never be solved. But the sure thing is that the Soviet revolution was preceded by a wealth of societal capabilities without which its distinctive creature would simply have been another experience and form of Terror. The point is, then, that revolutions are creative not because of an intrinsic creativeness of the revolutionary act as such-this is the myth- but because they liberate creative forces that are otherwise impeded. In particular, revolutions create a better polity if, and only if, somebody has in mind, before the revolution, what is to be done after the revolution. Their political creativity hinges on having a feasible, or at least credible, replacement world in sight. This is no longer the case with the present-day revolutionary subculture. It is no wonder that the rebel in arms is enchanted with Marcuse's

Ideals and Evidence

4. 7

dictum that the path to socialism may well be from science to utopia. This is all the layout he has. He is a utopian without a utopia. And the reasons for this predicament of his-an entirely novel one-are not far to seek. When liberal democracies became established on the basis of universal or near-universal suffrage, it came to be generally believed that revolutions had lost their reason for being. The "end of revolutions" prophecy has since been devastatingly disappointed; yet the underlying argument has a logical force that is less easily defeated. Nondemocratic systems are rigid systems, that is, they dispose of no built-in mechanisms for either changing themselves or responding to demands for change. Rigid systems, therefore, can only be broken; and this means that they ultimately call for a revolutionary overthrow (unless wars happen to achieve the same purpose). Democracy is, instead, a characteristically flexible system. Above all, or first, democracy is a procedure for processing whatever a society demands, whatever comes up through "voice" and with voice. Even so, democracies do succumb. 43 Yet democracy has taken its revenge on the revolutionaries that sought its overthrow; it has preempted them. The present-day revolutionary subculture is left to believe in a better world mysteriously and miraculously resulting from a creativity that revolutions per se surely do not have. It is small wonder, therefore, that the present-day replacement worlds should remain as empty as they are. The creativity of the Cambodian revolution (despite the ideas absorbed by its leaders at the Sorbonne) consists only of a massacre; when a void preexists, an even greater void follows. Ex nihilo, nihil fit- out of nothing, nothing comes.

4. 7

Ideals and Evidence

It has been shown that ideals are realizable but also non-realizable, that ideals can be betrayed but can also betray, and that ideals can be either constructive or destructive. From now onward I shall deal with the constructiveness of ideals. 44 Implicitly, when not explicitly, I shall ask over and over again whether, and how, ideals are best put to constructive, as against self-defeating, use. The question is: What renders an ideal amenable to successful management, to success in application? It was indicated earlier, in the light of our feedback rule, that the test of the constructiveness of ideals resides in how ideals relate to evidence. But while ideals have been inspected at length, "evidence" has not. It is expedient, as I have done thus far, to counterpoise ideals and facts, just as it is expedient to counterpoise ideas and reality. Under this formulation, we seemingly have a symmetry. Yet these elements are intrinsically very different. Ideals and ideas 45 are in our heads and fully coincide with their mental representations. 77

PERFECTIONISM AND UTOPIA

4.7

Facts and reality are simply out there. In truth, then, so-called facts are experience and evidence about facts - and that is quite another matter. Furthermore, very little of this experience and evidence is based on personal knowledge; 46 most of it consists, instead, on reports about experience and evidence. Even so ideas and facts, ideals and reality, do have-above and beyond their intrinsic differences-one crucial element in common: They are both given meaning, shaped, communicated, and firmed up via words, and indeed through the same words. The word is "democracy" (or liberalism or iocialism) regardless of whether we intend the ideal or the reality. In like manner, freedom, equality, justice, and so on, are, respectively, identical terms (words) for both ideals and facts. Granted that facts are, in truth, reports about facts, and granted, therefore, that to counterpoise ideals and facts in a symmetrical fashion is a grand simplification, in the end the simplification is vindicated. In a fundamental respect that symmetry holds. The quintessential point is that the difficult and precarious linkage between the outer world and the world inside our heads stands and falls on one element: its words and wording. Even the most bare empiricist cannot escape the unescapable, namely, that his data are found and cast within his concepts, and thereby via the wording of his concepts. Putting it the other way around, concepts (as termed or worded) are "data containers" just as much as they are anything else. In the field the researcher employs words (categories, classes, checklists) that (a) select the facts to be searched; (b) prejudge their interpretation; and (c) communicate what has been found. The gist is, then, that if there is a sure way of distorting, and indeed destroying, the process that somehow brings the outer world into our minds, this sure way is to disestablish the vocabulary, to play or cheat with words. The sine qua non condition that permits the assumption that an outer world somehow enters our heads is that throughout this tricky ambulation the vehicle remains constant; this vehicle being the word element and, more precisely, a firmed-up correspondence between a given wording and a given concept. Yet, during the last decades the game in town has been-not always unwittingly- a "word game" that has very effectively destroyed any intersubjective, stabilized correspondence between wordings and concepts. This is what has happened, we have just seen, to the correspondence between the words utopia and myth, on the one hand, and their meaning-function, on the other. A similar kind of disfigurement happened to "violence" when it became fashionable to speak of "structural violence" (i.e., a violence without violence). 47 And throughout this book we shall see that the words consensus, conflict, power, authority, freedom, coercion, and indeed all the central terms of the vocabulary of politics, have been subjected to the same treat-

Ideals and Evidence

4. 7

ment. If so, and to the extent that this is so, much of our evidence becomes invalid; in the aggregate, and over time, it denotes or measures under a same word different things (since it measures highly volatile concepts). Of course, research may, per se, produce valid findings (provided that its word vehicle is kept constant by the researcher). Even so, when all the research findings are added up, the aggregate conversion of "facts" into "apprehended facts" will reflect the messiness of the language basis. That is to say that the greater the interindividual volatility of meanings, the lesser the validity of the pile (of evidence) thus resulting. Under these conditions it becomes very difficult, in the case at hand, to assess which evidence has what bearing on the theory of democracy. The foregoing clarifies-I hope-when it is, and how it is, that so-called evidence becomes (a) valid evidence, in the broad sense that it speaks to the problem that it addresses; and (b) trustworthy evidence, in the sense that it can credibly be assumed to convert (as best as human ways can do it) an "outside" reality "inside" our minds. Let us now revert to the assertion that socalled facts actually are reports of experience and evidence about facts, in order to clarify in what sense experience is, or is not, evidence. If evidence is qualified as "scientific evidence;• or hard evidence, then it is self-evident that it differs from experience. However, this narrow meaning of the term evidence is of little avail to us. This is not only because the social science evidence still happens to be largely insufficient but also, we have just seen, because the disruption of language undermines its cumulability and, therefore, its cumulative validity. 48 Should we conclude from this that it is not evidence (scientific evidence) but experience, notably historical experience, that stands out as the primary sou_rce of our factual knowledge? Tocqueville is ever prophetic, although unassisted by social science. Saint-Simon performed astonishingly well as a prophet of the industrial society, despite the thinness and softness of his evidence. All the classics that outlined the project of our liberal democracies-from Locke to the authors of the Federalist Paperswell understood the world "as it is;' since their project did succeed in the real world; yet their knowledge was almost entirely historical: reports about historical experiences. These considerations notwithstanding, I still believe that the theory of democracy is entitled to make reference to "evidence"; among other reasons because I see no compelling reason for reducing and restricting the notion of evidence to its scientific variety. When used without qualifier, in my usage "evidence" points to whatever can attest credibly to a state of the real world, thereby including the evidence that comes from both historical experience and personal experience. 79

PERFECTIONISM AND UTOPIA

4. 7

Once evidence is defined, we may revert to the question: What renders id~als constructive instead of self-defeating? The test, I said, is how ideals relate to evidence. Ideals that simply reject evidence or that simply negate evidence-as now defined-cannot be, in the long run, constructive. That is to say that when ideals are intended to be constructive, they must interact with, and account for, the pertinent evidence. As spelled out by my feedback rule, this does not imply that ideals ever do, or should, surrender to evidence; it implies, instead, that ideals succeed in their constructive intent when evidence is put to use. Conversely, impermeability and deafness to whatever evidence is a sure sign that an ideal is not being managed constructively, that its real-world implementation will fail. There is still one point that deserves mentioning. Remember that evidence is, in full, reports about evidence. If this is explicated, it means that a report about experience and events actually consists of using words (whether or not supplemented by numerical values), of bringing them together in some coherent and meaningful fashion, of drawing inferences, and also of arguing that the evidence being reported on is not only credible but as credible, or more credible, than previous or contrary evidence. So, to report is already to provide a string of arguments, an argumentation. Nonetheless, I have asserted earlier that evidence cannot replace argument, just as argument cannot displace evidence. Why is that so? The reason is simple, yet important. While a reporting actually unfolds itself as an argument, its inherent limit is to be confined to the ambit of the "until now;' Thus, in a very fundamental sense, not only the impossible, but also the possibles escape the purview of evidence. Evidence attests to past, not to future possibles. It should be well understood, therefore, that while I shall relate, most of the time, ideals to evidence, any investigation of "possibles" (and of their correlative impossibles) requires theoretical argument. A brief recapitulation is in order, not least because we have moved a long way from the beginning. The beginning was that the extant division into realists and democrats is wrongly and ominously drawn, for it divides those who -on fundamentals-belong to the same family and unites those who should indeed be divided. When the chips are down, the divide is violence: whether we praise peacelike or warlike politics, whether we respect the individual and his freedom or whether we disclaim his sacredness, his freedom, and indeed his right to be alive. In truth, the real enemies of liberal democracy are located at the extremes of each camp; they are hyperrealists who negate all ideals, or hyperidealists who negate all facts. Thus, if democracy is attacked from without by mistaken realists, it is also (and at the moment more menacingly) undermined from within by mistaken idealists. The general point is that the 80

Ideals and Evidence

4. 7

excessive disrepute of present-day democracies among their beneficiaries is surely related to our having raised the stakes too high. The ingratitude typical of the man of our time and his disillusionment with democracy are also, and in no small part, the reaction to promised goals that cannot possibly be reached. And to the extent that this is so, the danger threatening a democracy that officially has no enemies left, is, above everything else, perfectionism. Ultimately it is the deluded perfectionist who either resorts to violence and to the maxim that ends justify all means, or who faults ideals that are faultless, thus turning to counter-ideals. It is, instead, my argument that ideals are not made to be "literally" converted into facts but to challenge facts; and that if this is not understood, ideals ultimately become self-denying. If the sights are incessantly raised, then the real world takes its revenge. Nor is there any unassailable reason for assuming that a "good" is optimized the more it is maximized. There is a final item-if only an aside-worthy of mention. In my discussion of perfectionism I have never brought in the notion of demagogy. Even though perfectionists easily become- if they enter politics-demagogues, the two are very different animals. At bottom, the perfectionist is an intellectual animal, brought up in some kind of intellectual hothouse, whereas the demagogue is a born politico, a "natural" political animal. The fact remains- it will be retorted-that democracies are exposed to, and suffer from, demagogic escalations far more than from anything else. Democracy is to politics what a market system is to economics. To pursue the analogy, just as we know no better method of protecting the consumer than forbidding monopolistic concentration of economic power, we do not know a better means for upholding freedom than letting parties (in the plural) compete among themselves. The difference, however, is that competition among economic producers is submitted to the control of consumers who do indeed consume, and are therefore in a position to appraise goods that are offered to them in some tangible form. Competition among political parties, on the contrary, easily escapes the control of the political consumer, for political goods are not tangible and cannot easily be evaluated. The analogy breaks down on a second count as well: that economic competition is legally controllable and controlled, whereas political competition is not. Economic cheats, so to speak, are brought to court and also to prison; political cheats (demagogues) are not. The difference is, then, that unfair, inflationary, and indeed purely demagogic competition hardly has, in politics, effective correctives. In short, economic demagogy does not pay, or pays far less, than political demagogy. 49 This, incidentally, is probably the reason that keeps "demagogy" a specifically political term.

PERFECTIONISM AND UTOPIA

4. 7

There is nothing in the above that I could deny. As the Greeks already found out, the demagogue is as old as politics. The demagogue is the person who tries to fool all the people all the time. Since he is not likely to ever disappear the problem can only be confronted, it seems to me, by having publics that are hard to fool- at least all the time. If demagogy is to be counteracted, the most effective way of doing so is to create, around the political cheater, a credibility vacuum. And since it is the perfectionist who gives credibility to the demagogue, to curb perfectionism is also, if only by indirect implication, to curb the demagogue. A final reason for criticizing democratic perfectionism is, then, not to assist demagogy with our intellectual mistakes. While ideals cannot be fully satisfied, they can be satisfied satis (i.e., optimized). But this is the view, and the task, of the non-perfectionist.

Notes r. See chapter 6, section 6. 2. On the origin of this phrase, see Francis Lieber, On Civil Liberty and SelfGovernment (Philadelphia: 1880), p. 405, and chap. 35, passim. 3. The one towering author whose rejection is unconditional is Nietzsche. His Genealogie der Moral (1887), which is the counterpart (together with the Anti-Christ) of Nietzsche's theory of the Superman and the Herrenklasse, represents the most powerful negation of all the values in which Christianity, liberalism, and democracy rest. 4. In the Preamble of the Declaration of Independence of the United States (4 July 1776). Instead, the French Declaration of the Rights of Man and Citizens of 1789 reads: "Men are born and remain free and equal in rights" (my italics). What makes Jefferson's text characteristic is that it stands unqualified. See, for what he intended, H.A. Myers, Are Men Equal? (New York: Putnam's, 1945), pp. 34-35, 63-64, 136-37. 5. Equality is discussed in full in chapter 12, herein. 6. In the Theses on Feuerbach (1848). 7. Reference is especially made to the "critical philosophy" of the Frankfurt school, well reviewed in L. Kolakowski's monumental Main Currents of Marxism (Oxford: Clarendon Press, 1978), vol. 3, chaps. 10-11. While the school was formed around Horkheimer in the 1920s, its revolutionary mass impact (relatively speaking, of course) occurred some forty years later. Its direct derivation from the Marxian revolutionary philosopher basically stems from the fact that all the authors concerned (including Marcuse) still belong to the dissolution of Hegelian philosophy, albeit variously rejuvenated by sociological or other cross-breedings. 8. More concludes his Utopia (1518) precisely in these terms: "I may rather wish for, than hope after:' 9. See especially Ideology and Utopia (London: Routledge and Kegan Paul, 1936), pp. 173£f. Note that the coupling of utopia with ideology also muddles the latter concept. 10. Mannheim admits that "among ideas which transcend the situation there are, certainly, some which in principle can never be realized" (ibid., p. 177). We should,

Notes

if so, distinguish between realizable and unrealizable utopias- a weird way of restoring what is being negated. n. The Soul of Man under Socialism (Saugatuck: 5x8 Press, 1950), p. 18. 12. Cited by Mannheim, Ideology and Utopia, p. 183. 13. Ibid. 14. H. Marcuse, "The End of Utopia;' Five Lectures (Boston: Beacon Press, 1970), p. 63. He concedes that utopia points to an impossible "when a project for social change contradicts real laws of nature:' However, since there are no known "real laws of nature" in the realm of politics (unless we restore a natural-law philosophy), Marcuse's concession is of little, if any, meaning. In substance Marcuse reduces utopia to "immaturity of the social situation:' 15. J.O. Hertzeler, The History of Utopian Thought (New York: 1923), p. 279. 16. Since my point is strictly conceptual, no reference is made to the contents and variety of utopias (thereby including "distopias;' i.e., counter-utopias). A very interesting substantive analysis is R. Ruyer, J; Utopie et Les Utopies (Paris: Presses Universitaires de France, 1950). See also Frank E. Manuel, ed., Utopias and Utopian Thought (Boston: Houghton Mifflin, 1966). The editor of the symposium significantly admits: "There was an early consensus that we had better not embark upon any attempt to achieve a common definition of the term utopia" (p. xiv). 17. R.M. Madver, The Web of Government (New York: Macmillan, 1947), pp. 5 and 4. Maclver's techniques subsume exact knowledge but are otherwise narrowly defined. 18. Ibid., p. 51. Interestingly, when Madver praises democracy (chap. 8), he never uses "myth:' The myth on which he especially dwells is authority. 19. Power and Society, pp. n7, n6. It is fair to note that Lasswell and Kaplan justify their tour de force with the argument that "we characterize symbols in terms of their functioning, not directly by their properties." 20. The parallelism is so close that Lasswell and Kaplan actually combine their redefinition of myth with the redefinitions of Mannheim, as follows: "The ideology is the political myth functioning to preserve the social structure: the utopia to supplant it" (Power and Society, p. 123). 21. It should be clear that "extension" is used in the ordinary meaning of the word (a smallness or bigness), not in its technical acceptation in logic, i.e., as the denotation (as opposed to the intension or connotation) of a concept. 22. This point is pursued under the focus of "participatory democracy;' chapter 5, section 6. 23. The objection that under conditions of natural harmony the self no longer is a "self" (i.e., that individuals would no longer differ among themselves) does not bear on the point. If the self disappears, the problem disappears; in that case, even the notion of government (a political concept) is abused and becomes meaningless. 24. For the Marxian concept of democracy, see chapter 15, sections r and 2. 25. Des Reactions Politiques (1797), chap. 8. Compare with Kant: "It is obvious that no matter how complete the theory may be, a middle term is required between theory and practice .... For a concept ... which contains a general rule, must be supplemented by an act of judgement whereby the practitioner distinguishes instances when the rule applies from those where it does noC:' (E. Kant, "On the Common Saying: This May Be True in Theory but It Does Not Apply in Practice;' in Political Writings

PERFECTIONISM AND UTOPIA

4.7

[New York: Cambridge University Press, 1970], p. 6r.) Kant was recommending prudence; Constant was saying much more. 26. Ibid. 27. The issue is taken up in chapter 9, section 5. 28. To be sure, ideals can also cover up and camouflage reality, but not at the genetic moment that I envisage. 29. Both quotations can be found in The Oxford Dictionary of Quotations (Oxford: Oxford University Press, 1953). 30. D.F. Thompson, The Democratic Citizen: Social Science and Democratic Theory in the Twentieth Century (New York: Cambridge University Press, 1970), pp. 45, 46. Italics added. 3r. John Herz, Political Realism and Political Idealism (Chicago: University of Chicago Press, 1951), p. 42. 32. See chapter 2, section 3. 33. See again Herz, Political Realism and Political Idealism, p. 168 and pp. 168-89. 34. Realists believe in the force of "force" but do not extol "violence:' As Sergio Cotta pointedly observes, "the truly characteristic fact of our time is the exaltation of violence. Up until the nineteenth century one finds no trace of any substance of such exaltation. Violence was recognized as being inevitable ... but was not praised . . . . If at times men were taught to employ it, this was because violence was believed to be, under given circumstances, a necessary evil, certainly not a good .... What has at times been exalted in the past was force ... not violence" (Perche la Violenza? [L:Aquila: Japadre, 1978], pp. 21-22). 35. For a historical and conceptual analysis, see especially P. Calvert, Revolution (London: Macmillan, 1970); and C. Kotowski, "Revolution;' in G. Sartori, ed., Social Science Concepts: A Systematic Analysis (Beverly Hills: Sage, 1984). The literature is extensive. See T.R. Gurr, Why Men Rebel? (Princeton: Princeton University Press, 1970); and H. Eckstein, "On the Causes of Internal War;' in E.A. Nordlinger, ed., Politics and Society (Englewood Cliffs, N.J.: Prentice-Hall, 1970). See also, in general, C.J. Friedrich, ed., Revolution (New York: Atherton Press, 1966); C. Tilly, "Revolutions and Collective Violence;' in F.I. Greenstein and N.W. Polsby, eds., Handbook of Political Science (Reading: Addison-Wesley, 1975), vol. 3; and A.S. Cohan, Theories of Revolution: An Introduction (New York: Wiley, 1975). For broader sociological considerations, see Chalmers Johnson, Revolutionary Change (Boston: Little Brown, 1966); and especially S.N. Eisenstadt, Revolution and the Transformation of Societies (New York: Free Press, 1978). 36. Marxists conceive revolution as an extended process and thus consider fundamental socioeconomic change a defining characteristic. If so, it can be argued that (a) the first "true revolution" is the Leninist one; and (b) revolutionary violence is a necessary condition of fundamental change. Both implications are too restrictive (and excessively self-serving). An intelligent exposition of the Marxist approach is Theda Skocpol, States and Social Revolutions (New York: Cambridge University Press, 1979). 37. This element is highlighted especially by J. Monnerot, Sociologie de la Revolution: Mythologies Politiques du XX Siecle (Paris: Fayard, 1969). See also L. Pellicani, I Rivoluzionari di Professione: Teoria e Prassi dello Gnosticismo Moderno (Firenze: Vallecchi, 1975). A broad, often insightful coverage of related themes is S. Bialer and S. Sluzar, eds., Radicalism in the Contemporary Age (Boulder: Westview, 1977), 3 vols.

Notes 38. Anti-Duhring (Moscow edition: 1959), p. 39r. 39. Sorel's best-known work is Reflections on Violence (1905; Glencoe: Free Press, 1950). However, his writings are extensive and unquestionably attest to philosophic credentials. 40. This, we have seen (chapter 3, section 2), is how politics is conceived by Carl Schmitt. 4r. Interestingly (as we are reminded by Hannah Arendt, On Revolution [London: Faber & Faber, 1963]), Cromwell's revolution was not called "revolution;' for the notion still maintained in the 1650s its traditional astronomical meaning of circular movement and, thus, of constant orderly recurrence. In this acceptation the "revolution" was the restoration of the Stuart monarchy in 1660. The first revolution to be designated as such by its protagonists was the Glorious Revolution of 1688-an essentially peaceful event. Our understanding of "revolution" (movement without return, rupture) is, thus, the one established in 1789. 42. Note that I am not addressing the question of Alec Nove, Was Stalin Really Necessary? Some Problems of Soviet Political Economy (London: Allen and Unwin, 1964). My conjecture is a preliminary one, namely, that Russia would have done far better than has been the case had its 1917 revolution simply been an unblocking revolution. For a comparative assessment of what the tsarist regime had accomplished in its last decades see Norman Stone, Europe Transformed: 1878-1919 (Cambridge, Mass.: Harvard University Press, 1984), pp. 197-254. 43. For a major, detailed coverage and investigation see J. Linz and A. Stepan, eds., The Breakdown of Democratic Regimes (Baltimore: Johns Hopkins University Press, 1978). Linz's Part 1, "Crisis, Breakdown and Reequilibration" (pp. 1-124), is a truly outstanding analysis. In general, see L. Morlino, Come Cambiano i Regimi Politici (Milano: Angeli, 1980). 44. Thompson, The Democratic Citizen, p. 47, distinguishes between "reconstructive" and "constructive" ideals and qualifies the former as "justifiable with little or no evidence from behavioral social science as it now exists:' I find the distinction, as stated, unconvincing. 45. Since ideals are expressed via ideas, at the source they need not be distinguished. 46. Its importance is well explained, almost to a point of overstatement, by M. Polanyi, Personal Knowledge (Chicago: University of Chicago Press, 1958). 47. See J. Galtung, Essays in Peace Research (Copenhagen: Ejlers, 1975), 1:5, "Structural and Direct Violence:' To be sure, "structural violence" can only mean that structures are "constraints:' But if constraints are called violence, either violence is nothing much or everything is violence. In either case the concept is disfigured. 48. Within the context of the theory of democracy, the most notable exception to this broad generalization is the evidence on public opinion. See chapter 5, sections 1-4. 49. For party competition and the economic analogy see G. Sartori, Parties and Party Systems: A Framework for Analysis (New York: Cambridge University Press, 1976), esp. chaps. 6.1 and 7.r. More generally, and for other disanalogies between economics and politics, see the perceptive analysis of W.C. Mitchell, "Efficiency, Responsibility, and Democratic Politics;• in J.R. Pennock and J.W. Chapman, eds., Liberal Democracy (New York: New York University Press, 1983), chap. 14.

85

5. Governed Democracy and Governing Democracy It is contrary to the order of nature for the large number to govern and for the small number to be governed.

-Rousseau

5.r

Public Opinion and Government by Consent

Politics ultimately hinges on the relationship between the governors and the governed. It has been argued that this dichotomy holds for all political systems except democracy. However, the fact that a democratic decision-making process blurs the dividing line between the governors and the governed does not entail that governing and being governed become mingled. To say that a precise boundary cannot be drawn is not the same as saying that there is no boundary. A boundary can be diffuse and nonetheless exist. Since in order to have democracy we must have, to some degree, a government of the people, let us immediately ask: When do we find a "governing people;' the demos in the act or the role of governing? The answer is: at elections. This is no belittlement, for the democratic process is indeed encapsulated in elections and electing. For one, and first of all, elections verify consensus and do away with presumed or fraudulent consensus. However, we still must remember that elections are a discontinuous and elementary performance. Between elections the people's power remains largely quiescent; and there is also a wide margin of discretion (let alone discrepancy) between broad electoral choices and concrete governmental decisions. Further, elections register the voter's decisions; but how are these decisions arrived at? Elections compute opinions; but where do these opinions come from and how are they formed? 1 What is the genesis of the will and opinion that elections limit themselves to recording? Voting has a prevoting background. While we must not downgrade the importance of elections, we cannot isolate the electoral event from the whole circuit of the opinion-forming process. Electoral power per se is the mechanical guarantee of democracy; but the substantive guarantee is given by the conditions under which the citizen gets the informa-

86

Public Opinion and Government by Consent 5.1

tion and is exposed to the pressure of opinion makers. Ultimately, "the opinion of the governed is the real foundation of all government:' 2 If this is so, elections are the means to an end-the end being a "government of opinion;' that is, a government responsive to, and responsible toward, public opinion. When all is said, we say that elections must be free. This is true, but it is not enough; for opinion too must be, in some basic sense, free. Free elections with unfree opinion express nothing. We say that the people must be sovereign. But an empty sovereign who has nothing to say, without opinions of his own, is a mere ratifier, a sovereign of nothing. The foregoing assertions touch a number of troublesome issues. A preliminary question is whether there is any reason for saying opinion (of the public) instead of employing some other term. The second question is in what sense an opinion is deemed public. Only at this juncture can we assess when an opinion disseminated among the publics is free - to what extent and in what sense. Furthermore, the notion of a government of opinion relates to the notion of "government by consent:' If so, we shall also have to enquire about consent, and this entails that the consensus theory of democracy must confront the thesis that democracy rests, instead, on conflict. While it may be argued that a public opinion has always existed and necessarily exists in every society, the interesting fact is that the expression was coined only during the decades that preceded the French Revolution. "Public opinion" was preceded by the late-Roman vox populi, by the medieval doctrine of consent and, if you wish, by Rousseau's general will. 3 Had these former conceptions all spoken to the same referent, there would have been little reason for coining "public opinion:' When compared to Rousseau's general will and, more important, to the medieval doctrine of consensus, a first novelty is that the notion of public opinion, unlike the other two, is eminently divisible. 4 The point of major interest is, however, the choice of the term opinion. Since the writers of the Enlightenment knew the classics well, they can hardly be suspected of not knowing what they were saying. The presumption is strong that they said "opinion" because they intended doxa, not episteme, not knowledge. A third noteworthy point is that in the expression public opinion the term "public" does not indicate only the subject (of the opinion) but also the nature and domain of the opinions in question. In its primary meaning an opinion is called public not only because it is diffused among publics but also because it pertains to "public things;' to the res publica. In short, public opinion is first and foremost a political concept. This entails that an opinion on public things is- must be- an opinion exposed to information about things public. All told, public opinion can thus be defined as follows: a public, or a multiplicity of publics, whose diffuse states

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.1

of mind (opinions) interact with flows of information concerning the state of the res publica. 5 To be sure, a state of mind, or of opinion, is composed of manifold ingredients: needs, desires, preferences, attitudes, an overall belief system, and still more, as we shall see. Yet in order to be politically relevant the opinions in question must be politically sensitized, that is, exposed to information concerning polity-related events. With these qualifications, then, public opinion is a modern label for modern developments. Indeed a Jama popularis, a set of rumores (hearsay), cannot meaningfully become a public opinion until technology permits the appearance of a press, of "news" and news leaflets. The conditions that allow for a free public opinion belong to the overall process of opinion formation and will be analyzed in that context. Let us, therefore, deal first with the problem of consensus, beginning with how the notion of a government of opinion links with "government by consent:' The argument is straightforward: Governments put in office by elections that reflect the opinions of the electorate and that are, furthermore, made to be responsive (by the recurrence of free elections) to their electorates are governments that may be called-without excessive forcing of meaning- consented-to governments. 6 If this assertion has become controversial, it is so because we mix the electoral theory of consensus with an overall, grand theory of consensus. Furthermore, in this latter connection we also tend to confuse the argument that bears on the optimal, or facilitating, conditions of democracy, with the argument that bears, instead, on what is the nature of society. The first confusion is easily handled by underscoring that electoral consensus (consent) is only electoral consensus.7 That is to say, above all, that it neither confronts nor res·olves the problem par excellence of the political philosopher, namely, what is the nature and what are the foundations of the "political obligation." 8 Clearly, electoral consent does not explain why people obey, and even less why they should obey, why they feel and should feel obliged to comply with orders and commands. But all of this has little, if anything, to do with the proposition that democracy is, characteristically, government by consensus-a government that rests on, and responds to, the opinions expressed at elections. The above might appear as a restrictive interpretation of "government by consent" on the grounds that opinions are not expressed uniquely in elections. Indeed, special groups such as the media, economic interest groups, and idea groups express opinions at all times, and surely more forcefully than electorates. In particular, electoral democracy may be integrated or acted for by "demonstration democracY:' 9 But these are all reasons, it seems to me, for pressing the point that the democratic test is the electoral test; for it is only 88

The Issue of Consensus

5.2

elections that manifest a "general consensus;' that is, the opinions of all the people (who care to manifest opinions). Contrariwise, the voices that make themselves heard above and beyond elections are elite or minority voices; they are voices of a slice, usually a very small slice, of the people. And even millions of demonstrators are not the people (as long as many more millions of citizens are left voiceless). Mussolini's march on Rome in 1922 certainly mobilized many people; it was less frightening than other comparable European marches of recent decades; and in fact it was peaceful in that it succeeded unopposed. Was it "demonstration democracy"? If the eulogists of demonstrations as superior to elections were consistent, by their criteria they would have to reply that it was. By my criteria, the overthrow of a government, if not a regime, by, say, one million demonstrators who oblige tens of millions of other citizens to take refuge and safety in their homes has absolutely nothing to do with democracy. It bears stressing, therefore, that demonstration democracy has little to share with public opinion and with the resulting notions of consented government and governing by consent. If democracy grants-as it does-the right to decide their destiny to all the people, then the opinions that indicate a general consensus or, conversely, a generalized dissensus about governing are the opinions delivered by the voters at large at elections, and via elections only.10

5.2

The Issue of Consensus

We may now address the concept of consensus as such, and in general. As mentioned, our theory of consensus often mingles two very different issues. A first query is, What does consensus explain about society? The other query is, What does consensus explain about democracy? Let us quickly dispose of the first question. Since real-world political societies are found to be more or less (a) consensual or conflictual, (b) integrated, segmented, or disintegrated, (c) homogeneous or heterogeneous, it is hardly surprising that consensual theories of society (homo homini socius) should be challenged by conflictual theories of society (homo homini lupus), and vice versa. 11 The fact nonetheless remains that democracies are political forms superimposed on both agitated societies consisting of anti-social animals and tranquil societies made up of social-minded animals. Therefore, from the vantage point of the theory of democracy a given state of society-whether less or more consensualconflictual- simply represents a set of facilitating or, conversely, complicating conditions of a democratic performance. And this consideration introduces the pertinent issue, namely, what does consensus explain about democracy (not about the nature of society)?

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.2

Consensus per se is not particularly hard to define. First, consensus is not actual consent: It does not imply the active consenting of each to something. Second, even though much of what we call consensus may simply be acceptance (i.e., consensus in a weak and basically passive sense), in all cases the general defining property of consensus-acceptance is a "sharing" that somehow binds.12 But a sharing of what? With respect to the theory of democracy at least three objects of possible sharing, of possible agreement, need to be neatly distinguished: (a) ultimate values (such as liberty and equality), which structure the belief system; (b) rules of the game, or procedures; (c) specific governments and governmental policies. These objects of consensusdissensus can be transformed, respectively, and following Easton, into three levels of consensus: (a). consensus at the community level, or basic consensus; (b) consensus at the regime level, or procedural consensus; (c) consensus at the policy level, or policy consensus.13 The first object or level of consensus-which we may identify as basic consensus-establishes whether a given society shares in its entirety the same value beliefs and value goals. When this is the case, we have, following Almond, a "homogeneous political culture:' 14 To be sure, this is a communitylevel, societal consensus; but the important qualification is that the society in question is the political society, not society in other respects.1 5 When this is not the case, we obtain, in Almond's conceptualization, a fragmented, heterogeneous political culture. And the relation of cultural homogeneity (basic consensus) or heterogeneity to democracy are by now pretty well established. Democratic forms are in fact superimposed on both homogeneous and heterogeneous political cultures. It cannot be asserted that consensus at the community belief level is a necessary condition of democracy.16 On the other hand, there is overwhelming evidence that unless a democracy succeeds in creating, over time, a consonant basic consensus, it performs as a difficult and fragile democracy.17 It can thus be safely asserted that consensus on fundamentals is a facilitating condition of democracy. While basic consensus is not a prerequisite of democracy, it certainly is a helpful condition. For one thing, it helps to establish its legitimacy.18 Furthermore, a good indicator of "successful democracy" is whether, over time, it acquires basic consensus; whereas a lack or loss of basic consensus attests to failings and failures of democracy. The second object or level of consensus may be identified as procedural consensus in that it establishes the so-called rules of the game. These are many-as one can easily infer from reading constitutions (whose norms largely spell out the procedures that regulate the exercise of power). Yet there is one paramount rule of the game that must precede all the others, namely, the rule that establishes how conflicts are to be resolved. If a political society does

The Issue of Consensus

5.2

not share a conflict-solving rule, it will conflict over each conflict-and this is civil war, or paves the way to civil war. Actually, civil wars and revolutions end precisely when the winner establishes which rule (if only by which ruler) will peacefully solve conflicts. In a democracy, this conflict-solving rule is majority rule.19 That is to say that unless and until the majority "rule of the game;' or principle, is generally accepted, a democracy has no rule for processing within-conflicts and can hardly begin, therefore, to perform as a democracy. Clearly, then, procedural consensus, and specifically consensus on the majority conflict-solving rule, is the sine qua non condition of democracy. It is very appropriate, therefore, to speak of procedural consensus as a consensus that bears on the regime. If the majority principle is not accepted, or at least acquiesced in, it is democracy as a regime, as a political form, that is not accepted. The saying that in democracy we agree to disagree actually means, when spelled out, that (a) we must first agree on the rules for disagreeing and for processing disagreements; and (b) disagreement within such rules is the disagreement that democracy protects and furthers. It is, then, the third object or level of consensus-consensus over policies and governments-that brings to the fore consensus-as-dissensus and sustains the view forcefully stated by E. Barker that "the basis and essence of all democracy" is "government by discussion;' 20 This is indeed the context in which dissent, dissensus, and opposition emerge as characterizing elements of democracy. The point hardly requires buttressing. Let it simply be recalled that dissent over policies and opposition to governments is dissensus vis-a-vis the personnel in government, not over the form of government. If the latter is intended, what is at stake is basic consensus or procedural consensus, and probably both. In resume, basic consensus, or consensus on fundamentals (the value beliefs that structure our belief system), is a facilitating condition, though not a necessary condition of democracy. This is a consensus that democracy may acquire as an end product. By contrast, procedural consensus, and preeminently consensus on the conflict-solving rule and rules, is a necessary condition, indeed, the prerequisite of democracy. This consensus is the beginning of democracy. 21 Therefore, it is only when consensus addresses policies and the personnel in government that the emphasis shifts on discussion, dissensus, and the crucial role of opposition. But even here the argument is not that consensus is irrelevant or (still worse) that it is undesirable for democracy; it is, rather, that dissensus is assumed and required to bring about changes in consensus, that is, a new consensus or new consenters on different things. And since these changes in consensus are reflected in changes of governments, how can it be denied that democratic governments (in office) are consented

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.2

governments? The recent discovery that conflict, not consensus, is the basis and the essence of democracy represents, in my view, an ill-judged thesis. 22 To be sure, the conflict theory of democracy can always be explained as a polemical thesis designed to stir the waters. Nonetheless, it not only overshoots but largely misses the mark. If one is to judge by the fact that Burke was the first author who spoke in defense and praise of parties, then it would appear that until the second half of the eighteenth century the theory of consensus was b~sically a unanimity theory of consensus. Until then it had been firmly held and believed that a divided, dissensual polity was a doomed polity unfit for survival. Rousseau still postulated (indeed, more than his contemporaries) a monolithic unity and unanimity. Only as parties and party systems developed in the nineteenth century did it come to be acknowledged and realized that a pluralistic consensus, or (depending on the emphasis) a pluralistic dissensus, was not only compatible with, but also beneficial to, the good polity. The crucial point is, therefore, that dissent, opposition, adversary politics, and contestation are all notions that acquire a positive value, and a positive role, within the context of pluralism, that is, within the pluralistic conception of society and history. Prior to whatever else it may be, pluralism is the belief in the value of diversity. 23 And believing in diversity-in a dialectics of diversity-is antipodal to believing in conflict. Hence, what the theory of democracy derives from its pluralistic matrix is not, and cannot be, a praise of "conflict" but, instead, a dynamic processing of consensus based on the principle that whatever claims to be rightful, or true, must hold its own against, and be revitalized by, criticism and dissent. 24 The simplest way of making the point is that, for the problem at issue, conflict is a wrong, i.e., misleading word. True, conflict has come to be used in a very weak sense, as when we say conflict of interests, and the like. Yet, we equally and still use conflict to mean "war;' that is, with reference to warlike behavior. Given the potential deadliness of the matter, why use an inevitably confusing term when dissent and dissensus provide the correct focus and are indeed the correct complements of consensus? The question-let it be reasserted- is not whether society is intrinsically characterized by conflictual relations; the question is whether consensus, and which consensus, facilitates democracy, constructs democracy, or is, instead, irrelevant to, if not undesirable for, democracy.

5.3

The Formation of Opinions

Opinions are not innate, nor do they arise from nothing. The question, What

92

The Formation of Opinions 5.3 is public opinion? is best answered by sorting out three processes: (a) the trickling down of opinions from elite levels; (b) the bubbling up of opinions from the grass roots; and (c) reference group identifications-in that order. The elite-induced formation of opinions is well described by the cascade model formulated by Deutsch. 25 In this telling imagery, opinions flow downward in a multi-step fashion, as in a cascade broken by a series of pools. The top pool consists, for Deutsch, of the economic and social elites. It is followed by the pools of political and governmental elites, of mass media, of opinion leaders and, finally, of the mass public. Let us leave aside, for the moment, who really triggers the process. Above all, the explanatory value of the cascade model lies in bringing out the extent to which the opinion-forming processes are, as they descend, interrupted and regenerated at each level horizontally, that is, by intra-basin or intra-pool interactions. Naturally the assumption is that we are looking at a free polity, at a free society. It is under this assumption that each level affords, in principle, a full interplay of influences against influences, of resources against resources, of stimuli against stimuli. That is to say, each level opens up anew a dialectic of opinions and counter-opinions engendered, if by nothing else, by competitive temptations and rewards. In practice, of course, we should not expect all the pools to be as turbulent or as competitive as they can potentially be. Likewise, at each point in space and time we shall discover that resources and influences are unequal, that they do not quite counterbalance one another. But, over time, the pendulum of the prevailing influences does swing, and across the board (of the various levels) the aggregate outcome is never as desired and designed by any one single source of influence. 26 Of the five pools of "opinion interplay" singled out by Deutsch, two are of particular salience: that of mass media and that of opinion leaders. In present-day democracies, it is the mass media that play the greatest and central role in forming public opinion. The notions of gatekeeping, of agenda setting, of watchdog function, of prismatic deflection and/or distortion, and the like, centrally apply to the performance of the media and its impact. With reference to Deutsch's layout of the cascade it could be argued that the media pool is either very close, altimetrically, to the two preceding basins, or that it incessantly spills back into them. Putting it all in one blunt sentence: The world is-for the public at large-the media message. If the media are not, in democracies, as "persuasive" as it is sometimes feared, this is in no small part because the next, and truly big jump of the cascade hits on, and is mediated by, local opinion leaders.27 These opinion leaders consist of the politically attentive stratum of a population - the 5 to ro percent of a public that follows public affairs. 28 To be sure, members of the informed (interested and 93

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.3

attentive) public do not automatically, or in their totality, become opinion leaders. Yet they do constitute a reference group for their respective communities of friends and neighbors, for the audiences to which they speak. Thus local opinion leaders may either block or reinforce, deflect or amplify, and do in any case select the importance of, and attribute credibility to, the messages of the media. This implies, first, that another overall re~huffiing occurs between the message as emitted and the message as received and, second, that the mass public is not, in its way of absorbing the media messages, an atomized, wholly unstructured public. How does the bubbling up of public opinion relate to the cascade model? If it were only a matter of bubblings, then it could be argued that these gurgles are accounted for by the cascade model, for Deutsch amply allows for feedbacks and feedback loops. Yet the welling up of opinions takes, at times, the outright form of an irresistible swell, of a massive tide. And, clearly, tides of opinion are left unexplained by the cascade model. In fairness to Deutsch, it should be pointed out that his focus is on international affairs; he therefore addresses the same public that Walter Lippmann had harshly qualified as an absent and silent "phantom public?' 29 International affairs, until it comes to actual war, war fears, or to a cutting off of vital resources, are generally very remote from the ordinary citizen. His lasting attention is aroused and his grumblings are persistently vigorous when it is his shoe that pinches; and this is far more often the case in domestic affairs and in bread-and-butter matters. Even so, the inherent limit of the cascade model is that it pays too little heed to "idea groups" and their locations. As with international resource interdependencies, this is a development that attains its momentum with the postindustrial societies, and eminently with the centrality attributed by these societies-if only for technological needs-to higher learning institutions. 30 In concurrence with other factors, the massive expansion of higher education produces, in fact overproduces, a very sizable intellectual population that can be accommodated less and less where it fits or where it feels entitled to be. Unemployed or misemployed intellectuals by now loom large, thus concentrating more and more in the residual basin - the bottom one. And since opinions originate, ultimately, from ideas, the proliferation· of idea groups, of intellectual nuclei that remain at the lower ends of the social stratification scale, is tantamount to a proliferation and intensification of the bubbling-up processes. All in all, I would say that our democracies have attained a stage in which upgoing processes of opinion are best seen as a self-contained, independent process. This is also because the bubbling up swells, at times, into powerful tides. Tides of opinion may well be intermittent, and come at very irregular intervals. Yet, when they come, they overtake and take the upper levels (poli94

The Formation of Opinions 5.3 ticians, the so-called military-industrial complex, and even the media) by surprise; and they do leave a long-lasting mark. While it has long been true that a welling up of consequence was never really originated at the grass-roots level, the growth and diffusion of a proletarianized intelligentsia, of the "massified intellectual;' speak to a different pattern. To be sure, counter-elites are still elites, but they no longer coincide with an elite level; they are mass-level elites. Neither the cascade nor the bubbling-up processes exhaust, however, the "what is" of public opinion. The opinions of the individuals who compose, when added up, a public, are also largely derived from identifications with a variety of concrete and/or reference groups: the family, peer groups, work groups, religious groups, ethnic groups, party identifications, and class identifications. And the identifications in question have little to do with being informed and acquiring information. In effect, we speak of identifications precisely because we are now looking at opinions that definitely do not result from an exposure to information. Even when so exposed, individuals have opinions without information, and indeed opinions that may well fly in the face of the evidence carried by information. In sum, individuals have opinions that are not linked to information, that precede it, and that indeed negate or refuse information. This is the aspect of public opinion that Berelson brought to the fore in a classic passage: For many voters political preferences may be considered analogous to cultural tastes. . . . Both have their origin in ethnic, sectional, class, and family traditions. Both exhibit stability and resistance to change for individuals, but flexibility and adjustment over generations for the society as a whole. Both seem to be matters of sentiment and disposition rather than "reasoned preferences:' While both are responsive to changed conditions and unusual stimuli, they are relatively invulnerable to direct argumentation. 31

It is not indispensable for my argument to assess in general the relative weight and proportions of the three ingredients that enter the overall composition of a public opinion. The important point is how the aforesaid ingredients relate to the information element. A public opinion is such, I noted at the outset, not only in the sense that "the public" is the subject, but also in the sense that it relates to "public objects;' that is, that it bears on opinions that result from, and react to, information on the state of the thing public. By this latter criterion, our three ingredients can be appraised as follows. The cascade model of opinion formation is the crucial one with respect to the information element, that is, to the amount and nature of the information that happens to be contained in a public opinion. On the other hand, the downgoing processes of opinion formation also constitute the critical juncture, the strain point of the entire edifice. The importance of this ingredient

95

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.3

is matched by the risk potential that goes with it. For the "informed" will of the people may also be its least authentic will. The more one "receives" from the media, the more one is potentially exposed to media manipulation. The identification-based ingredient of public opinion is, instead, its noninformed element. Here we characteristically find opinions that either ignore or resist "the news;' opinions that precede and prejudge the information. On the other hand, as Berelson's analogy with a cultural taste well highlights, the opinions anchored to identifications constitute the resilient and least vulnerable aspect of public opinion. We may thus say that this is the non-manipulated, non-manufactured, untampered will of the people. Finally, the bubbling-up mode of opinion expression stands-with rtspect to the information element-somewhere in-between. Some tides of opinion are clearly related to policies about which the public is informed. But other tides of opinion seemingly arise from identifications, religious creeds, ideological beliefs, ethnic sentiments, and the like, which have nothing to do with information and are actually intensified by distorted, if not utterly disinforming, messages.

5.4

Autonomy versus Heteronomy of Public Opinion

An invariably recurring assertion is that every modern or modernized society, be it democratic or not, exhibits a public opinion. If the importance of a point is to be judged by its frequency of repetition, the above statement must have an import that escapes me-for I find it either trivial or positively misleading. That every modernized society contains a public opinion is true only in the banal meaning that an opinion is public because it is located in the public. As a rule, however, the assertion that every society has a public opinion is plugged into the implication that every regime finds an antagonist in the force of public opinion. This I believe to be false; and the assertion is certainly misleading in that it covers up the crucial distinction between (a) an opinion that is public merely in the sense that it is disseminated among the public, and (b) an opinion that the public to some degree has formed by itself. In the first sense we have an opinion made public but in no way produced by the public: therefore public in a trivial and largely mystifying meaning. In the second sense we have instead an opinion of the public, meaning that the public is the subject. In the first sense any society can be credited by definition with a public opinion- and with this the crux of the matter is beheaded in one blow. For the problem is, When is it that a public opinion represents an autonomous force? Conversely, How is it that a public opinion is turned into a rubber stamp?

Autonomy versus Heteronomy of Public Opinion

5.4

Until the 1920s there was no need to distinguish between opinions in, and opinions of, the people. 32 Until the advent of totalitarian controls and mass media, what the people had in mind had also to be, in some meaningful sense, expression of their minds. The literature of the 1940s had to reckon, instead, with the fact that a popular, near-unanimous consensus had materialized also in dictatorships that had formerly been free countries, and thus with the reality of a public opinion that actually was a uniform package of statemade opinions enforced on the public. 33 This literature has since been largely dismissed. In an overview of the mass communication literature Wilbur Schramm denounces a "somewhat pathological scare over propaganda" in the studies up until the 1950s, contends that the subsequent thirty years of research have "demolished" the former views, and merely concedes to such views that if "a particular point of view were to monopolize the media channels ... the influence of propaganda is likely to be much greater'c_ much greater, to be sure, than very little influence. 34 The foregoing is representative, I believe, of the current thinking of the mass media specialist, whose case basically rests on the consideration that we now dispose of a scientific evidence that in part corrects and in part contradicts the impressionistic evidence of the earlier literature. The "impressionistic" can be granted- and yet, does our "scientific evidence" prove what it is assumed to prove? Notice, to begin with, that the research evidence, on which the communication specialist currently builds his case and from which Schramm concludes that it demolishes the fear about propaganda effectiveness, is almost exclusively an evidence confined to the United States, and largely an evidence that bears on advertising or on a propaganda assimilated to advertisement. Now, research findings speak to where they are found, not to everywhere. That propaganda and advertisement are assimilable in the United States cannot disprove that they are highly distinguishable elsewhere. And where this is the case, nothing is "demolished" by an evidence that does not apply. Be that as it may, there is a paradoxical sideline to the scientific persuasion of our time. The tenet of such persuasion is that the only reliable evidence is the scientific evidence, which is, and can only be, research-based evidence. Fine- but what about the countries in which research is impeded and cannot even enter? This question is seldom confronted. We are able to shun it, I suspect, because in the back of our minds we have become "degreeists;' that is, addicts of the principle that all differences are differences in degree. If so, the world closed to science can and indeed must be assumed to be "more" or "less" like the world that we know (i.e., open to science). But this is a far too facile solution. The net effect of degreeism is that never before have we been as parochial and Western centered, vis-a-vis the non-researchable worlds,

97

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.4

as we are today. With all its merits, hard evidence is myopic; it demands a close look from nearby. And if evidence must be confined exclusively to research evidence, then without research there is ignorance, and the serious scientist must simply acknowledge it. Instead we have entered a path of blinding and blindness. In the name of hard evidence we dismiss the soft evidence; but since most countries are researched inadequately, if at all, we are left with neither. The overall irony is that our scientific rigor ends up rewarding the banishers of science. Whoever does not allow social science research to enter has a far easier life, and gets off the hook far more easily, than one who does. The secrecy of the secret world pays off. Bearing the above caveats in mind, How does an autonomous public opinion compare with a heteronomous one? Let us begin with the notion of a~tonomy, and precisely with this question: What is it that makes a public opinion- the opinions of the public-a free and independent state of opinion? Doubtless all these notions- autonomy, self-sufficiency, freedom, independence, and the like - are relative notions.· They are relative, let me hasten to add (lest I be confused with the degreeist I have just criticized), and yet find a neat cutoff point in the following alternative: whether or not the state controls all the instruments of socialization and all the media _:'control" meaning that one voice only, the voice of the state, is allowed to speak. 35 The conditions that allow for a relatively autonomous public opinion can be summarized under two headings: (a) a system of education that is not a system of indoctrination; and (b) an overall structure of plural and diverse centers of influence and information. I shall have to leave the first condition as a reminder. 36 The second set of conditions is already well explained in the cascade model of opinion formation discussed earlier37 and can thus be dealt with briefly. The essence of the argument is that a free public opinion derives from, and is sustained by, a polycentric structuring of the media and by their competitive interplay. In short, the autonomy of public opinion presupposes market-type conditions. Note that the argument does not assume that a polycentric, competitive structure of the media finds audiences that play off one source against the other and that make up their mind after having compared them. When this is the case, so much the better. But the benefits of mass media decentralization and competition are - so runs the argument-largely mechanical, and of two sorts. First, a multiplicity of persuaders reflects itself in a plurality of publics; this goes to produce, in turn, a pluralistic society. Second, a market-type system of information is a self-checking, watchdog system, for each channel is exposed to the vigilance of the other channels. The merits or the adequacy of the polycentric structure are often challenged, but are vindicated as soon as we take a comparative perspective and

Autonomy versus Heteronomy of Public Opinion

5.4

turn to unicentric opinion-making systems, that is, to the opposite structural arrangement. 38 The potency of the mass media instrument does not impress Western specialists, for they happen to observe a mass media system neutralized by its dispersion and its pluralistic counterbalancings. But the potency of the instrument per se, in its potentiality, cannot be doubted. Orwell's 1984 is a nightmare, but by no means an impossible one. Since our pursuit of utopian "impossibles" neglects the nightmarish "possibles" that do loom ahead, it bears stressing that the world conjectured by Orwell is, technically, altogether feasible; our know-how has reached a point at which the ultimate sanctuary, the human mind, can be conquered and is in fact being conquered. Aside from the terrifying potentialities of brainwashing, subliminal persuasion, and drug-based control, what has already become a full-fledged reality is a totalitarian, unicentric system of opinion making. Let us succinctly describe it in ideal-type fashion. It will be recalled that the virtue, from a democratic vantage point, of the cascade mode of formation of public opinion crucially resides in the fact that each pool adversarially reshuffles, independently from the other basins and in its own peculiar manner, the messages that enter it. I also pointed out that the hierarchical ordering of the cascade outlined by Deutsch somehow misrepresents the process, for the upper three basins are, in our democracies, largely intercommunicating, though distinct, basins. Moreover, it is highly dubious that trends of opinion are really conceived and originated by socioeconomic or political elites; they are generally ignited by idea groups. The major concentration of idea groups is, naturally, in the learning institutions and the media (thereby including publishers); yet these groups are by now very effectively spread across all the levels of the cascade, thereby including the bottom one. This implies that our democracies have increasingly flattened out the cascade. A second and, indeed, democratic virtue of democracies is thus that their opinion-formation processes no longer are characteristically hierarchical; not only are the altimetric differences at most points of the cascade very small but the trickling down has no fixed hierarchy to it. At times, and frequently enough, the cascade actually begins with the media. And, occasionally, the cascade itself is swollen by an upgoing tide of opinion. In a totalitarian system of opinion making all of the foregoing disappears. 3 9 Its central characteristic is to establish a neatly hierarchical cascade in which each basin has only a reinforcing effect, an amplifying effect. There is only one voice, the sole voice of the regime; and it flows down undisturbed, without interruptions, because the pools are now still, muted into sounding boards, into resonance boxes. In totalitarian regimes this is accomplished in two ways: via a vault of terror along the hierarchical line, and via an inces99

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.4

sant mobilization at the grass roots. A vault of terror functions smoothlyprovided that the sanctions are, as they in fact are, severe enough-on the principle that everybody passes the bucket to the next person along the line (and this even if everybody along that line hates it). Concurrently, the masslevel mobilization by party activists serves the twin purposes of destroying the bubbling-up nuclei and the spontaneous opinion leaders. The ulterior characteristic of such a system is that it extends far beyond the opinion-making processes; it accompanies its members from birth to grave 'throughout all the sites of socialization. In particular, and above all, the process of education becomes a process of indoctrination in which one doctrine only exists, and all other doctrines are banned. In all the areas that are not strictly technical, education is displaced and replaced by a propaganda fidei, by the official, exclusive, faith of state. From womb to tomb the bombardment is both incessant and unopposed, for the artillery is all on one side and commanded by the same general. To be sure, a unicentric system of totalitarian propaganda implies a "closed system;' More precisely put, the greater the sealing off and the impermeability, the greater the efficacy. The obstruction of exits, together with censorship over all the messages from outer worlds, is an essential condition because outside parameters and external bases of comparison not only undermine the "sole truth" but impede total lying, or total silence (the two are complementary); whereas blackouts or unfettered lying are perfectly possible, and daily practices, in closed systems. Clearly, then, unicentric propaganda systems are bound to obtain a highly uniform spread of opinions in the public. The surprise is not that totalitarian systems, if adequately and truly totalitarian, are consented to. The point is whether, and in what sense, this consensus can be ascribed to opinions of the public. Let us ask: Under a monochromatic, unicentric system of relentless indoctrination, to what extent can the opinions in the public be also opinions of the public? The answer is not far to seek: only with respect to identifications, to the identification basis of opinions. And even this ingredient of public opinion shrinks, and is effectively atrophied, by totalitarianisms. As Lenin knew well, the secret of control systems is to sever the horizontal lines of communication: to begin with, in the party; but also in the society at large. Throughout a totalitarian control system, all the channels of communication tend to be vertical. With respect to the opinions rooted in identifications, this means that they cannot join forces, that they are left to survive in isolated, silent, and thereby inoffensive niches. So, whatever survives of a genuine public opinion is only its more primitive and static ingredient- and at bare survival levels. Conversely, the efficient and major part of the opinions in the IOO

Autonomy versus Heteronomy of Public Opinion 5.4 public consists of heteronomous, state-made opinions. If it is true, as it trivially is, that all modernized societies have a public opinion, it is more true that in unfree societies there is no public opinion such as exists in democracies. The above describes the limiting case and does not entail that unicentric systems are eminently invulnerable. It is very hard to keep them as closed and as neatly sealed as is optimally required. Therefore, as long as counter, better worlds-democracies-exist, they represent, at least in the long run, a potent undermining factor. Orwell correctly hypothesizes that perfect tyranny requires outer enemies of the same sort. Furthermore, the monotony of the propaganda bombardment may well produce saturation, apathy, and, in the longer run, reactions of outright rejection. Nonetheless, even in their failings and eventual failure unicentric systems of total propaganda do achieve one sure victory: They create totally misinformed and disinformed publics. With respect to the information element of public opinion, this element is destroyed in that it is totally corrupted. So, totalitarian systems are not likely to be short lived. Short of military defeat, the endogenous backing out from totalitarianisms is a difficult and slow process; it has to wait, inter alia and through the turnover of generations, for the putrefaction of the initial ideological impetus and purity. And the point remains that in one respect at least totalitarianisms are unfailingly successful: in being systems of wholesale, systematic misinformation. This conclusion raises the issue of how good the information is in the polycentric systems. If information means "true" or "correct" information, the fact that no such thing exists in the unicentric propaganda systems does not settle the question of how, and to what extent, a competitive structure of multiple informers ensures correctness of information. While it is plausible that a multiplication of channels adds to the quantity and completeness of information, it is less clear why it should also add to its correctness or objectivity. Much of the present-day criticism of the information processes in democracies addresses the point that the "power to inform" is a power of the few and, in particular, that it is distributed unequally. True; but it still remains to be shown what benefits would accrue from a system of equal voice to all. If the point is that the consumer of information is powerless, then he lacks power in the same sense and extent to which the consumer in general, the economic consumer, does. The analogy is that the consumer of information relates to an oligopoly of information producers (media), just as the economic consumer relates to an oligopoly of producers of goods. So, also in the latter context it could be argued that a "just" economic system requires the power of the consumer to be equal to the power of the producer. Hence, the equivaIOI

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.4

lent of the formula "equal voice to all" is that each and all should be "equal producer-consumers:' However, any cost-benefit analysis bears out that while the benefits resulting from the latter formula would be very dubious, its costs would be staggering. The analogy suggests, then, that while the net benefits of a competitive multiplicity of media are well established, what is not established is that further net benefits would follow from multiplying a multiplicity to no end. Remember, to maximize is not necessarily to optimize. A counterbalancing pluralism of the media addresses, and to some extent resolves, the problem of the autonomy of public opinion. Equality in the media or, at any rate, of the media is something else, and it has yet to be demonstrated whether it would solve whichever problem it addresses. My sense is that while in matters of information an incessant vigilance is of the utmost importance, the current agenda of grievances and therapies is largely misdirected. If, as I believe, the whole edifice of democracy ultimately rests on the relative fairness, impartiality, or correctness of the information delivered to the public, then I find little evidence that this is the real concern of our protesters or the likely end result of their demands. Structures cannot do what their concrete operators will not do. And when it comes to the objectivity, fairness, and correctness of the information, we are left, in the final analysis, with a professional ethic: the ethic of the respect for truth. We may acknowledge that Truth (in capitals) is unattainable and yet pursue the ideal of truth. Ultimately, then, everything hinges on the value belief in truth-in the worth of truth. But we now live in a world replete with ideologized persuaders for whom "the cause" takes priority over truth. And if we fail to recognize this, we fail to recognize and tackle the most.

5.5

Electoral Democracy

I said at the outset that free elections with unfree opinion express nothing. Enough has been said on what free or unfree opinion means. Let us turn, therefore, to how public opinion performs at elections (in democracies), to what elections themselves express, and thus to what may be identified as "electoral democracy:' If there is one area in which the theory of democracy disposes of hard and plentiful evidence, it is the field of public opinion and voting behavior. 40 It is here also that the gap between macrotheory and microevidence is as small as it can be. 41 Even so, bridging the gap is not easy. When opinions are actually being researched (by the pollster or by an interviewer), the concept of opinion must be operationalized- as in the following definition of Lane and Sears: ''An opinion is an 'answer' that is given to a 'question' in a given situa102

Electoral Democracy 5.5 tion?' The immediate implication is that "when the question or situation varies somewhat, a somewhat different response can be expected?' 42 And the import of the point is that the assessment of the internal consistency or inconsistency of opinions becomes a tricky affair. For instance, critics of the consensus theory of democracy cite the finding that while respondents generally agree on abstract principles, they are no longer in agreement when it comes to concrete implications of the principles. 43 Is this inconsistency? Is it hypocrisy? As Lane and Sears point out, it need not be. Aside from the fact that concrete responses relate very much- as they should- to concrete settings, the interpretation of the aforesaid finding confronts three ulterior problems. First, it cannot be assumed (as we have seen at length) that ideals are meant to be implemented literally. Second, there is no single nor simple way of deriving concrete implications from abstract principles. Third, verbal behavior can be almost meaningless if its intensity is not weighed. However that may be, let us come to the findings that unquestionably and directly speak to the crucial issue: How much does the public know, know wrongly, or not know at all about public affairs? In short: What is the information base of public opinion? Here the answer is crushingly, throughout mountains of evidence, of a similar tenor: The state of inattention, non-interest, sub-information, perceptive distortion, and, finally, plain ignorance of the average citizen never ceases to surprise the observer. The percentages vary, for they hinge on how loosely we measure "sufficient information:• But across all democracies, the picture resulting from voting studies and polls invariably is a sad picture indeed of the information base-let alone all the rest-of a large majority of the citizenry. 44 It is a safe roundabout generalization that apathy or depolitization is widespread, that the ordinary citizen has little interest in politics, that citizen participation is minimal if not subminimal, and that in many respects and instances the public has no opinion but, rather, inarticulate feelings made up of moods and drifts of sentiment. Two major problems arise with respect to this conclusion. The first one bears on the causes and remedies of such a state of affairs. The second one bears on how well, or how badly, the theory of democracy fares on these findings. We shall discuss them in that order. So, and first, how do we account for the apathy-lack of interest, high level of ignorance, minimal participation-of a large majority of citizens? Is this a somewhat physiological state of affairs? Or is it a state of affairs due to impediments that can be removed, to causal factors that can be altered? Since these questions have been incessantly posed and since remedies have been incessantly proposed for over a century, maybe some answers have been provided by the very length of the trial. When the debate was at its peak, 103

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.5

that is, when universal suffrage was being fought for, the winning argument was that people would learn how to vote by voting. When this learning process did not live up to its expectations, poverty and illiteracy took the blame. Both points are well taken. Practice is a learning process; nor can we expect poverty stricken and illiterate, poorly educated citizens to be, in any meaningful sense, capable and interested citizens. In a substantial number of countries, however, voting has been practiced at more than sufficient length. In addition, the ratios of poverty and illiteracy have been dra'matically altered and reduced. Yet no heartening improvements have followed; apathy still looms large, with no detectable, long-range trend of change for the better. Where do we go next? A first diagnosis, and related therapy, centrally addresses the problem of information. A second one puts high hopes, instead, on increased levels of education. A third response places its bets on a new rallying symbol: participatory democracy. The revival of the ideal of participatory democracy will be given a separate treatment. The other two arguments can be looked into right away. The information processes are blamed on three counts: (a) quantitative insufficiency; (b) bias; and (c) qualitative poverty. The first charge is not very convincing. Quantitatively speaking, the problem may well be too much, not too little, information: The ordinary citizen is now swamped by an excess of messages that he cannot possibly digest and that turn him off. The second charge-that the information is biased-is more to the point, but hardly in its current version, namely, that it is capitalistically biased. Early in the game Plamenatz well responded to this allegation: The workers, when they vote, understand the issues about as much and as little as their employers. That they do not do so, that they are kept in ignorance, that the political vocabulary current among them is one evolved in the interest of the rich- all these statements appear to me to be false. The political vocabulary in current use is much larger than it was two hundred years ago, and most of the words and phrases added to it were invented by radicals and socialists. Indeed, many of them were either coined by Marx or else made popular by him. The language of politics, as it is spoken in Western Europe, is as much "proletarian" as it is anything else. 45

Since then, it is the anticapitalistic, not the capitalistic, bias that has grown mightily. Throughout continental Europe also non-Marxists have come to adopt, if often unwittingly, a pidgin Marxism and a vocabulary of debunking and witch hunting. And as long as the denouncers of "tendentiousness;' of a mobilization of bias, simply advocate their own one-sidedness-often a much more thorough one-the malady is likely to increase, not diminish. Here we do confront an extremely serious problem: the problem of the de-

Electoral Democracy

5.5

creasing standards of the professional ethic of "respect for truth." But this is hardly the concern of the present-day decriers of the tendentiousness of the information. We are thus left with the third charge: the poor quality of the information. This poverty can be readily granted (I find it appalling) and yet be difficult to cure. In particular, when we specifically come to the visual message we come to an instrument that is inherently truth-disserving. In most free societies most video newsmakers do not lie or do not intend to lie; yet television per se, as a technique, does several things which, combined, imply that what is shown is often so incompletely true as to be altogether false. The cure can only consist, here, in having more responsible (and, by the same token, less responsive) media makers. On the other hand, the fact that the media are largely monitored by marketing and poll responses might well be seen as a tribute to democracy. Furthermore, if the purpose is to reach out at larger audiences, a better or higher quality of the media may not serve that purpose. The second diagnosis and therapy argues that better informed and more interested citizens will result from higher levels of education and from their spread. This prediction is sustained- it is claimed- by the finding that the politically informed public correlates positively with the more educated public. This argument is in turn sustained by the concomitant finding that the wellto-do (who are also, as a rule, the better educated) vote in a higher proportion than the have-nots. However, the correlation between higher education and better information is either largely tautological or of dubious validity (when it ceases to be circular). As for the finding that the rich participate in politics more than the poor, it is above all an American finding. Let me take up first the thesis that participation in politics is a function of wealth or, at any rate, of the spread of wealth beyond poverty levels. While the evidence on the American voter does speak in this sense, 46 it should not be forgotten that the United States is highly anomalous with respect to all other democracies in three interconnected respects: a truly enormous pool of nonvoters; a very low mobilizational capacity of the parties; and an equally low degree, at the level of mass electorates, of ideological polarization.47 When we come to the more polarized democracies, or to democracies whose electoral turnout ranges from 80 to 90 percent levels, the evidence often speaks to entirely different hypotheses: that the crucial variables (for electoral participation and beyond) are status polarization, class consciousness, the mobilizational capability of parties and mobilizing networks in general and, of course, how much is at stake. Thus the reason that makes wealth salient in the American scene may well be that no other variable-happens to be salient. In any event, the argument that the well-to-do should be expected to be more

105

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.5

politically participant than the have-nots lacks a convincing rationale. The contrary argument is that the worker is, or can be, more participant (in the full sense) than the bourgeois, whenever he finds in the union or in the party activity a compensation for the monotony and estrangement inherent in postcraftsmanship, in merely repetitive work. Indeed, why should the low income strata of a population be less motivated and gratified by the rewards of political activity than a leisure class that can afford better distractions and does not need politics for furthering its well-being? We may now revert to the thesis that political information is a function of higher levels of education. In many respects this is true almost by definition. For instance, and clearly, one who is illiterate can have firm opinions but hardly informed opinions. Similarly, if education includes education in or about politics, then it goes by definition that the two things increase together. But one can be highly literate and yet remain politically illiterate. There is no compelling reason why a general growth of the levels of instruction should specifically reflect itself in a growth of politically informed publics. There is, instead, a very good reason that explains why this may not be the case. Information is a "cost" that draws on scarce individual resources in time and attention. Hence, acquiring information in one sector requires neglecting other sectors. It might be argued that since politics affects everybody, politics makes exception to this rule. However, the cost of becoming and staying informed becomes rewarding only beyond a given threshold, only when a critical mass of accumulation has been attained. Thus the distribution of political information turns out to be as uneven and discontinuous, across a population, as distributions in other fields of interest. In Converse's wording, "it is not hard to understand why electorates are so extremely heterogeneous in political information;' for "there is a dramatic inversion of the relative weights of costs and benefits of paying attention to politics somewhere between the upper and lower edges of mass electorates." 48 Pushing the case to the extremes, even if everybody had, say, five years of university education, is there any reason for expecting a dramatic conversion to politics of the interests of that population? I think not. That education in general is not likely to bring about a significantly improved politically educated public is confirmed by an additional consideration. Let "politically educated" be defined not only as a state of being informed but also as a state of cognitive competence (an ulterior element on which I dwell later). Now, with respect to the cognitive parameter Schumpeter made the very relevant point that "the typical citizen drops down to a lower level of mental performance as soon as he enters the political field. He argues and analyzes in a way which he would readily recognize as infantile within the 106

Electoral Democracy 5.5 sphere of his real interests. He becomes a primitive again. His thinking becomes associative and affective ... ?' 49 Is this an overstatement? Probably not, because drops in mental performance are the rule whenever we exit from the territory in which we operate. The reason for this is obvious enough, namely, that the matters we truly understand are those with which we have personal experience, and that the ideas we really master are those that we are capable of formulating by ourselves. An astronomer who discusses philosophy, a chemist who passes judgment on music, or a poet who talks about mathematics will not utter less nonsense than the average citizen interviewed by a pollster. The difference is that a specialist will generally plead ignorance in other fields of specialization, whereas the citizen (everybody) is asked to concern himself with politics and in the midst of the general incompetence no longer realizes his own. The difference is, then, that in the other zones of ignorance we are discouraged from trespassing, while in the political realm we are encouraged to do so. Schumpeter's point implies that the political field is not the "sphere of real interests" of the typical citizen. But why should that be so? And cannot this state of affairs be changed? Converse, I have just recalled, assesses the problem in cost-benefit terms, though in somewhat narrow fashion. A concomitant consideration of Brittan points to "the lack of incentive for the ordinary citizen ... to become even moderately well informed. In private lives, the cash constraint is an influence for rationality. People know in household budgeting that more of one thing means less of something else?' 50 Brittan compares here homo politicus to homo oeconomicus and, on such basis, actually addresses two points. One is that politics does not and cannot provide, for everybody, rewards and incentives as economics does - and this, Brittan points out, is "a major obstacle to trendy ideas about 'participation? " The other point does not bear on motivating people to be informed but on explaining how, and in what sense, the "rationality" imposed by the laws of the household (i.e., economics) does not apply to the citizen. The explanation is, quite simply, that the citizen does not "pay" and that on most occasions the consequences of his actions escape internalization. All in all, and resuming our thread, it seems to me that in the course of removing the obstacles that disfavor the "outs" vis-a-vis the "ins" we have paid too little attention to the difference between necessary and sufficient conditions. Literacy presumably is a necessary condition for authentic citizenship; yet one can be well learned and politically ill learned. Likewise, a decent standard of living is a necessary condition; yet political participation does not increase significantly-either in quantity or in quality-as wealth spreads and increases. Participation was generally higher around the turn of 107

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.5

the century when enfranchisement was being extended or fully granted; it tends to decline with habituation to voting; and the single factor that best explains its variance across time and countries is whether the stakes of politics are perceived to be high. Thus it does not appear that we have meaningfully succeeded, despite one century of attempts, in changing for the better the ratios between inert and active citizens. It would also be hazardous to assert that our therapies have brought about an improved performance of the ordinary voter. Today, just as in the past, the democratic citizen in most instances does not know what the issues are, what solutiom, have been proposed, what consequences are likely to follow, and even what the candidates running for office (let alone the parties) stand for. So much for what the findings tell. The next question is: How does the theory of democracy fare in the face of such findings? Since the theory in question is the theory of electoral democracy, before discussing what electors do, let us first recall what elections do. Succinctly stated, elections do not enact policies; elections establish, rather, who will enact them. Elections do not decide issues; they decide, rather, who will decide issues. As Dahl neatly puts it, strictly speaking "all an election reveals is the first preferences of some citizens among the candidates standing for office;' for "we can rarely interpret a majority of first choices among candidates in a national election as being equivalent to a majority of first choices for a specific policY:' 51 If anything, elections may reveal even less than Dahl suggests, for often enough they do not even express "first preferences" or first choices (as we shall see in a moment). Furthermore, and from a different angle, Arrow's "paradox of voting" implies that majority decisions may well be spurious in the sense that they do not reflect the preference of the majority and/or that elections fail to reflect the "social preference" (i.e., the overall preference rankings) of the voters. 52 If this is what elections reveal, or fail to reveal, what does the elector do when he votes, that is, how does he vote, on what grounds and criteria? Since voting behavior varies over time, across individuals, and also very much across countries, the overall picture is best rendered by two models. In the first model, that may be called the issue voting model, the sequence is: (a) issue preferences; (b) issue perceptions; (c) vote for the candidate or party that appears closer in issue stands. In the second, alternative model, which may be called the party identification model, the sequence is: (a) self-placement on a left-right, or progressive-conservative, or cleavage-based spectrum; (b) corresponding party images; (c) vote for the party one identifies with, i.e., closer on the relevant spectrum. It goes without saying that the two models represent limiting cases, within which the actual behavior of actual voters ro8

Electoral Democracy 5.5 may well display blendings of various criteria. In addition, electorates may vote/or (on positive grounds and prospectively), or vote against (on negative or retrospective punishment grounds); and voters may not express their first or most preferred preference but their last or least disliked preference (for they would otherwise simply waste their vote). As a rule of thumb, I would say that issue voting finds its most favorable soil in two-party, low-polarized polities but becomes all the more unlikely and difficult to practice in the multiparty democracies, especially as the ideological distance and polarization of a polity grows. 53 In any event, it is clear that the elector is-and is eventually forced to be - a great simplifier. I have noted that it is only in a weak and vague sense that elections tell how to govern; primarily they establish who shall govern. This is true, let it be added, both for the single-member district system and for systems of proportional representation. Should we search, therefore, for more sensitive electoral instrumentalities that might reveal better than the ones in use the will of the majority in regard to specific policy issues? Or is the machinery, in all its existing varieties, as imperfect as its utilizers? My feeling is that the electoral instruments say just about as much as the voters have to say. To probe this matter let us turn to the issue of "rational voting:• So far, and mildly put, the notion of a rational voter has been handled poorly. As Converse points out, voting behavior studies either fail to define rationality or tend to define it as a choice that "maximizes perceived (or expected) utility" for the choice maker. But this definitio~ "can only push us in a tautology: any behavior the actor chooses to engage in must maximize his or her perceived utility or otherwise the actor would make another choice:' 54 Aside from the fact that by the above definition of rationality any elector is by definition rational, it is also appropriate to wonder why a maximization of utility should be rational at all. Suppose that I vote for being paid without working. Surely such a proposal maximizes my immediate, perceived utility; but would I be rational, in any sensible sense of the word, by having that proposal enacted-as would be the case-for the whole collectivity? The truth of the matter is that "perceived utility" definitions of rationality are appropriate only for individual and individually bounded decisions processed by market mechanisms-not for collectivized decisions enforced upon all, as is the case with political decisions. 55 A vote that enacted a salary without work for all would be disastrously irrational not only for the collectivity but also for myself as a member of that collectivity. Reverting to the specifics, the voting behavior studies that employ the parameter of rationality equate the rational voter with the issue voter. But without adequate proof. Aside from the preliminary consideration that issues 109

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.5

are often wrapped into issue packages (party platforms) that are not easily unwrapped, the issue voter would be "more rational" (in some intuitive sense) than the party identifier under two essential conditions: the correct perception of the issue(s), and the correct understanding of which issue solution produces which consequences. Since these conditions are seldom met, why should an issue voter who gets everything wrong and has no notion of how ends relate to means be rationally superior to the despised identifier? Thus Converse is perfectly right in concluding that the question of rationality in voting behavior is "impossibly formulated?' 56 I would go a step further: I wonder whether it should be formulated at all. If, in fact, elections decide about who will decide, the implication is that the burden of rationality does not rest-in the electoral theory of democracyon electorates: It is shifted on to their representatives and, thereby, on the theory of represeqtative democracy. The first point is, thus, that in awaiting a rational public opinion (that expresses itself in rational voting) that neither appears definable nor to exist, what is lost sight of is the crucial requirement of an autonomous public opinion that expresses itself freely. The retort might be that this is to remain content with too little. I have gone to some pains, however, to show that this requirement is in no way an easy given. Be that as it may, a "rational parameter" either misrepresents reality (as when the issue voter is deemed to be rational) or puts impossible demands on the demos and, thereby, on the entire edifice of democracy. The last consideration brings me back to the question: How does the theory of democracy fare with respect to the findings on the poverty of public opinion and on the grand simplifications that occur in voting? My reply is that, despite all our misgivings about the practice, still the theory has a good fit for it. The electoral theory of democracy argues-let it be recalled-that (a) democracy postulates an autonomous public opinion, (b) which sustains, via elections, consented governments, (c) which are in. turn responsive to the opinions of the public. Nothing of the above is contradicted or falsified by the facts, by our evidence. But, of course, this conclusion holds within the limits of the assumption that the people exercise their power as electors, that is, in terms of electoral power. If we maintain, instead, that the power of the people should not be the mere power of deciding who should decide the issues but the power of actually deciding them, then we confront an entirely different argument. This different argument has long been associated with the democracy of the ancients and with the theory of direct democracy. Currently, however, it has been revived under the name and in the perspective of a "participatory democracy'~ to which I now turn.

IIO

Participatory Democracy

5.6

5.6

Participatory Democracy

Since the notion of participatory democracy remains, to date, fuzzy, let us attempt to pin it down with respect to the neighboring and clearer notions of (a) direct democracy; (b) referendum democracy; (c) electoral democracy; and (d) representative democracy. Electoral democracy has just been discussed. Representative democracy can simply be defined, for present purposes, as an indirect democracy in which the people do not themselves govern but elect representatives who govern them. As for the relation between electoral and representative democracy, it will suffice to note that the former is a necessary, though not sufficient, condition of the latter. This is also to say that the concept of representative democracy comprehends electoral democracy; but the reverse is not true. While modern democracies are jointly electoral and representative, an electoral democracy that does not elect representatives is a distinct possibility. The other two notions-direct democracy and referendum democracy-require a somewhat more detailed explication. Although a direct democracy can be simply defined a contrario, that is, as a democracy without representatives and without representational transmission belts, yet this specimen displays subspecies that command separate recognition. Any direct democracy is, in some sense, a self-governing democracy. But we know that the meaning of self-government crucially hinges on the size factor; 57 so does the meaning and reality of direct democracy. A literal, authentically self-governing direct democracy can be said to exist only with reference to relatively small groups-say, up to assembly-size groups. Beyond the assembly size, the most meaningful distinction is, I suggest, between observable direct democracy, and a direct democracy whose size escapes direct observability, that is, a greater than observable one. To illustrate, the democracy of the ancients qualifies as observable, for it resolved itself not only in the gathering of the citizens in a single place but in an observable behavior of the participants. But the magnitude that allows for such observability is in the order of a few thousands. And that order of magnitude already brings about a great deal of "indirectness" in a direct democracy. In reality, the gathering of the demos in the ekklesfa represented the spectacular, hardly the efficient part of the Greek politefa. Thus, only in part was Greek democracy truly and authentically "direct." And when we come to a greater-than-observable direct democracy, what makes it direct is only, or basically, the inference that it is not a representative kind of democracy. The notion of referendum democracy will help to qualify this last assertion. A referendum democracy is, as the wording says, a democracy in which the demos decides directly the issues, though not by gathering together, but III

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.6

discretely, via the referendum instrument. Even so, referendum democracy can be conceived as one of the subspecies of direct democracy. It can also be said that referendum democracy represents the overcoming-afforded by technology - of the size and space limitations of direct democracy. This having been conceded, two reasons justify treating referendum democracy as a species in its own right. The first reason is that referendum democracy is ."direct" in that it does away with intermediaries; but it loses the other characteristic of direct democracy: directness of interactions. It is, so to speak, a clirect democracy of isolated, discrete individuals-not of interacting participants. This is an important difference. The second reason is that the referendum instrument can also be entered into the theory and practice of representative democracy. This being so, the case could be made that referendum democracy brings together and actually merges direct and representative democracy. But this would be misleading. It will not be necessary to enlarge on the point that when referenda are included in representative democracy, they are also subsumed under it. Therefore, what I intend by referendum democracy is not the mere utilization of the instrument but its establishment as the mechanism of a democracy. It will be necessary, instead, to enlarge on the difference between direct and referendum democracy. One such difference, we have just seen, is that the referendum actor is like the electoral actor: He performs in solitude, by himself, without a "debating participation:' His deliberation is not preceded by a dialogue that goes to shape the deliberation. Thus the "enlightenment of discussion" -crucial, for Ernest Barker, to the bettering of decisions 58 -is ruled out. Certainly the issues submitted to referendum will be debated on the media; but the referendum-type decider remains, if he listens, a passive listener who does not contribute, not even minimally, in that debate. Another difference bears on the question: Who sets the agenda and the actual formulation of the issues submitted to referendum? This question reencounters my former point that the greater the size of a direct democracy, the lesser its authenticity. Clearly, in referendum democracy the agenda setting is crucial; yet it is hard to believe that it will ever be set by direct democracy criteria, i.e., by the people themselves. Having outlined the map of the well-identifiable and definable species of democracy, where should we place on such map a participatory democracy? It is fair to reply: nowhere in particular and, to differing extents, everywhere. 59 In the main, the argument of most propounders of the notion is not that participatory democracy should do away with elections or utterly dismiss representation; and even though the "participationist" cherishes the ideal of direct democracy, he seldom brings his theory to the point of coinciding with it. II2

Participatory Democracy 5.6

What is definite, and unanimously held by the participationist, is that "electoral participation" is neither real participation nor the appropriate site of participation. I agree. To speak of the she.er act of voting as participation is little more than a manner of speech and certainly leaves us with a weak and overly diluted meaning of the term. But the point has been made many times, it is not much of a discovery, and certainly does not suffice to warrant the claim that the theory of participatory democracy is either new or, by itself, a theory. The theory's status and its novelty derive, then, from the centrality it assigns to the concept of participation and, thereby, to participation understood in a strong and undiluted meaning of the word. Again, I agree. Properly and meaningfully understood, participation is taking part in person, and a selfactivated, willed taking part. That is, participation is not a mere "being part of" (a mere being involved in some occurrence), and even less an involuntary "made-to-be part of." Participation is self-motion and thus the exact reverse of being put into motion (by another will), i.e., the opposite of mobilization. That this is what the participationist must mean is underscored by the fact that all the virtues he attributes to participation -self-mastery, self-realization and self-education - accrue to the strong, certainly not to the weak, meaning of the word. When dealing with self-government, it was shown that the intensity of the concept is inversely related to its extension (in space and time). The rule applies even more simply and intuitively to participation defined as a voluntary taking part in person; its intensity- namely, authenticity and effectiveness- is inversely related to the number of the participants. In a group of five, the part of each partaker is one fifth; in a group of fifty, the part of each participant is one fiftieth; and so forth. Thus, participation can be neatly operationalized as a ratio expressed by a fraction. As the denominator grows, the "part" (share, weight, import) of each partaker diminishes in the same amount. Clearly, few concepts are as easy to pinpoint and define as participation.60 It is equally clear that when the concept is operatively and operationally defined, it does not afford the mileage that the participationist tries to extract from it. It should be recalled that the mainstream theory of democracy has never neglected participation conceived as personal, active involvement. The praise of voluntary associations, the theory of the multigroup society, the themes of intra-party and intra-union democracy 6Lall add up to a large literature that does extol the centrality of participation to democracy. So, it was never denied that participation is the essence of microdemocracies or that it provides a vital infrastructure for the overall superstructure, i.e., for the demo-

u3

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.6

cratic polity. What was affirmed was that the magnitude increases, and as we move from small groups all the way up to the level of the political system, participation neither explains nor suffices to sustain the edifice of representative democracy. Therefore, if the indictment of the participationist is that prior to the 1960s participation was a neglected part of the overall theory of democracy, this indictment is, as a matter of record, incorrect. If his argument is, instead, that participation plays no important role in the specific theory of the democratic state, this is correct- but is this a fault? The question thus turns to whether the participationist does better than the so-called elitist on the same grounds on which the latter performs: political macrodemocracy. This is the question that I confront in the next chapter. For the time being, let me simply point out that the more the participationist presses his point, the more he is in danger, albeit unwittingly, of being more "elitist" than the one he is attacking. If we are serious about participation, as the participationist should be, then it can hardly be denied that a taking part is meaningful, authentic, real, only within the ambit of small groups. The participationist is right when he disparages electoral participation, for a ratio of taking part of one to tens of thousands or, in the aggregate, of one to tens of millions, renders that partaking meaningless. It follows from this that one of the two routes open to the participatory theory of democracy (the other one being referendum democracy) is to place the emphasis on, and attribute an important role to, small and intense groups. Such groups must be intense because this is the ground on which they come into existence and persist over time; and they must be small if the requirement is, as it is, to provide the optimal size for optimal participation. Fine. But is this not one of the meanings of "elitism;' namely, that the few do better, and count for more, than the passive, inert, apathetic, nonparticipant many? It is apparent that I still have not replied to the question, What is it that makes the participatory theory of democracy distinctive? The attempt to sort out this theory with respect to the neighboring theories-especially direct and representative democracy-was left at the conclusion that the participationist variously cuts across them. And the attempt to reconstruct the theory of participatory democracy on the basis of the concept of participation is either frustrated by the fact that "participation" is never defined and incessantly slips through our fingers, or leaves us with small-group democracy and eventually with a vanguard theory (and practice) of the active and intense small group. In order to salvage the participatory theory of democracy as a sui generis, distinctively new theory, we are thus left to explore a third path, namely, the path of pushing it forward in the direction of referendum democracy.

Referendum Democracy and Knowledge

5.7

Thus far I have not made recourse to the notion of populism, or of populistic democracy. Whether employed appreciatively or negatively, in either case "populist democracy" has largely remained, over the decades, a rhetorical label with little substantive content. 62 It did not seem to me, therefore, that there was anything to be gained, in my initial·charting, by comparing participatory democracy with populist democracy. One obscurity does not illuminate another obscurity. However, the point that could be made now is that either the theory of participatory democracy firms up positively where it stands with regard to which participation it endorses in which sites, or the participationist is not faring better than the earlier populist.

5.7

Referendum Democracy and Knowledge

Referendum democracy is here understood as a macrodemocracy that replaces representative democracy. While no such democracy currently exists, it is, by now, technologically feasible. Each voter would dispose of a video terminal in which the issues and their proposed solutions are displayed, let us say weekly, and would simply have to press yes-no-abstain buttons. Aside from the (important) question about who would actually set the agenda and formulate the issues, it is apparent that a referemium democracy so conceived would fulfill the central demand of the direct and, in its wake, of the populist and participationist democrat, namely, that the people themselves decide the issues. True, the size of the voting population would be such as to reduce to meaninglessness the part (weight or role) of each partaker. Thus the sense of inefficacy experienced by the present-day elector would reproduce itself in the future issue decider. But there is no way of getting around the drawbacks of size. Referendum democracy is open to a major objection, namely, that it sets up an outright zero-sum mechanism of decision making, that is, a literal majority rule system that rules out minority rights. On each issue, the winning majority takes all, the minority loses all. Not only majority rule would become absolute or unlimited, but no tradeoffs, no compensations, could occur among issues either. Since every referendum-type decision is a discrete and self-contained decision, it cannot be tempered by "exchanges;• by crossissue adjustments or corrections (not even if required on the grounds of consistency). In short, the objection is that referendum democracy is a conflict maximizing structure that represents not only the perfect but also the most unintelligent (since it would be purely mechanical) incarnation of a systematic "majority tyrannY:' 63 However, neither the theory of direct democracy nor the coherent participationist has ever been sensitive to the "limited" majority rule requirement. Indeed, it is hard to see how one can advocate direct de-

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.7

mocracy if one perceives the importance of limiting the majority rule principle. Therefore, the objection that referendum democracy becomes a zero-sum democracy should not be expected to shake the confidence of the populist or the participationist in the overriding goodness of a referendum-based polity. It cannot be doubted that referendum democracy as described would realize the ideal of governing democracy. This does not mean, it must be underscored, that the citizens would, each and all, come close to the fulfillment of their wishes. What each individual wants for himself can become, in the aggregate, an outcome no one wanted. Indeed, when a large universe is involved, the pure and simple adding up of individual preferences may engender aggregate consequences that negate the individual intentions. So, a governing or self-governing democracy is not the self-governing of each over himself. But if we do seek a governing democracy, then referendum democracy would provide it-and the technical feasibility of the latter is currently beyond doubt. So, how is it that the electronic solution is so seldom, if at all, proposed by the people who incessantly denounce representative "governed" democracy as a false democracy? 64 The moment of truth-the point at which the ideal of a literal democracy is realized-has finally come. At the moment of truth, it is the hidden truth-what remains untold or buried deep down-that surfaces. And I submit that the untold truth is that even the most extreme participationist realizes that he demands more than he actually desires, and indeed far more than is desirable. Why should he not demand referendum democracy, unless he too senses that it would be an overkill? Let me pause on where the argument stands. My main focus has been on the participatory theory of democracy; and the winding up of the ground covered thus far is that there are at least three kinds of participationists: (a) moderates; (b) elitists in disguise; and (c) pure. 65 The first group does not propound a new theory, but rather a new or renewed emphasis (on participation). The second group covers up, under a fashionable and most convenient disguise, the displacement by counter-elites of preexisting elites. The third group, the one that remains relevant at this stage of the argument, represents the most recent incarnation of the general category of the perfectionist. And here we have a nice test case of the principle of inverted results, of the thesis that ideals pushed to their extremes operate in the reverse. 66 The formula of referendum democracy attests that the target of a literal, self-governing democracy is by now within actual reach. But how do we know that referendum democracy is or would be an overkill and, indeed, a self-killing? It is dearly the case that a referendum democracy as defined would place on public opinion a burden that would be incommensurably greater than it is in the representative democracies; this implies, in turn, that the reality of

rr6

Referendum Democracy and Knowledge

5.7

public opinion must now be reappraised in an entirely new light. It will be remembered that the theory of electoral democracy is satisfied by the requirement of the "autonomy" of public opinion, for the ulterior requirement of some kind of "rationality" is passed on, by the electors, to those they elect. But if the elector becomes the decider, then the burden of some kind of rationality comes to rest entirely with him. And hie Rhodus, hie salta-it is here, Rhodus, that you must jump. It may have not passed unnoticed that throughout my analysis of public opinion reference has been carefully made to its information base. As pointed out at the outset, when the expression was coined, its coiners well knew the difference between doxa and episteme, between opinion and knowledge. So, information is not knowledge. To be sure, knowledge presupposes information; but it does not go by definition that he who is informed is knowledgeable (if so, knowledgeable simply means informed; the problem is beheaded by a tautology). At a minimum, knowledge implies a grasp of, and a mental control over, the information that is in no way supplied by the information itself. Weber defined rationality as the capability of gauging means to ends; 67 and this is the concept of rationality that nicely goes to define knowledge. On the other hand, while knowledge must be neatly separated from information, it can be conjoined with competence. Thus, given an equal amount of information, a person may either be competent or remain incompetent, depending on whether he perceives correctly which means are appropriate to what ends, and thereby which consequences derive from which decision or action. When the chips are down, it is above all the problem of knowledge, of competent understanding, that must be squarely confronted. We can circumvent it, or fudge it, as long as electors simply elect; but not if voters are issue deciders. It is, then, from the standpoint of "cognitive mastery" that we can and must appreciate the quality jump, indeed, the quality leap, implied by the passage from electoral democracy to referendum democracy. In the first formula, the requirement is only that people's opinions be exposed to information (and even non-informed opinions will do). In the second formula, the requirement becomes that adequate information is transformed into adequate knowledge, into an understanding of the problems, of their intricate interdependencies, of the overall effects of allocations and reallocations of resources. Again, in an electoral democracy the question as to whether we learn how to vote by voting can be left undecided or to wishful thinking; it is not crucial. In referendum democracy, instead, everything hinges on whether people learn how to participate (in decision making) by participating (in pressing the buttons on the video). If a quality jump is required, as it surely is, then it must be demanded of participation itself. The question is: Does parrr7

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5. 7

ticipation (as described) teach knowledge (as defined)? To this question, the answer surely is no. There is neither a grain of evidence nor of plausibility that would sustain a different answer. The objection could be that the participationist has an authentic "full participation" in mind, and that his contention that participation is a learning experience applies to the personal, active experience of partaking. In such case, the argument must be taken up at its beginning, that is, at the consideration that participation in the full sense assumes "intensity:' The full participator is such because his reward is in the activity itself. Whatever the prior motivations, the party activist, the incessant demonstrator, the engaged member of a grouping, feels intensely about politics. In this connection the question becomes: How does intensity unfold and where does it lead to? The virtuous sequence would be something like this: intensity .......+ interest .......+ attention .......+ information -+ knowledge. But this turns out to be, statistically, a very infrequent sequencing. As Lane and Sears attest: One of the best established patterns of public opinion is the U-shaped relationship between extremity on some substantive issue ... and intensity of feeling. The more extreme the stand, the more intensely do people feel about it.... This is so well established, in fact, that in searching for a "zero" position on an issue scale, researchers will choose the position on which people feel less strongly.... There is something congenial between extremity and intensity.... The forces nourishing extremity seem also to nourish the intensity. 68

So, issue intensity tends to correspond to extremism: a very strong, very definite, very self-assured "siding" in a two-sided world-either all white or all black. This is tantamount to saying that intensity-high affect and strong passion - is the least congenial soil for knowledge. 69 The extremist is such because he has no doubts; he already knows, and is sure of what he knows. Conversely stated, if the extremist did see the pro's and con's, the complexity and the multisidedness of issues, he would not be an extremist. Therefore, the highly intense participant does not qualify as being correctly informed and even less as pursuing knowledge, for he seeks only reinforcing information, only confirming knowledge. Certainly intense participation (as defined) is not without exceptions. Furthermore, it may be corrected by actual experience and responsibility in office. Remember, however, that we are envisaging ex hypothesi the mass level participation of populations for which no such corrective exists. As a matter of statistical frequency, therefore, participation does not turn out to be a learning process in the sense required by referendum democracy, that is, in the direction of producing an informed and knowledgeable state of opin-

n8

Referendum Democracy and Democracy 5. 7 ion, to wit, a citizenry that would cognitively and "rationally" master meansends interconnections. If anything, the participant idealized by the participatory theory of democracy goes to produce the opposite state of opinion. The intense, extremized participant usefully challenges an excess of inertia of the inert citizen, thus performing a positive role within the context of representative democracy. But as a decision maker the extremist would wreck a democratic polity far more thoroughly, and surely much faster, than his foe, the apathetic citizen. The foregoing entails that the pure or advanced theory of participation misunderstands the intensity factor not only at its utmost pitch-in its association with extremist behavior and cognitive blindness-but also, and correlatively, at its lowest one, that is, in its association with apathy. It was Berelson who first wondered how a mass democracy could work at all "if all the people were deeply involved in politics:' Having noted that "extreme interest goes with extreme partisanship and might culminate in rigid fanaticism that could destroy the democratic process if generalized throughout the community;' he concluded that it was the "heterogeneity of the electorate;' a balancing "between action motivated by strong sentiments and action with little passion behind it;' that made democracy work. 7 °Clearly, Berelson's recognition of the positive role of apathy was relative, or comparative, that is, weighed against the opposite excess. That this defense of apathy should have become the great villain of the subsequent literature, the emblematic confession of the "elitist fear" of mass participation, is a misreading and a caricature. A society may be too torpid, but may also be too agitated. To be sure, one can object that Berelson's point is too generic, that one should distinguish between non-extremism and non-partisanship 71 and, as I myself think, that the argument needs more analytic premises. Even so, while Berelson's praise of apathy was'relative and qualified, the participationist praise of intense participation is absolute and unqualified; a small error, if you wish, is thus replaced by a grand error. 72 We have performed a mental experiment that warrants, I trust, two straightforward conclusions. The general one is that public opinion-as it exists and in its reality-can and does sustain the edifice of representative democracy, for electoral democracy requires of public opinion just about what public opinion gives and probably can give. Conversely, the superseders of representative democracy have not even begun to scratch the surface of the problems raised by their alleged superseding. The second conclusion bears specifically on the moment of truth brought out by our mental experiment. At the moment of truth we can no longer remain blind to the fact that information is not knowledge and, furthermore, that knowledge-cognitive competence and control-

rr9

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.7

becomes more and more the problem as politics becomes more and more complicated. The growing complexity of the world of politics can hardly be doubted; it results not only from increasing and global interdependencies but from the very expansion of the sphere of politics. The more the visible hand and political engineering displace the invisible hand of automatic adjustments (or maladjustments), and the more politics enters everywhere, the less we are in control of what we are doing. Thus the problem of knowledge comes to the fore also in that we are running into a "knowledge crisis:' If these premises are correct, they imply that a referendum democracy would quickly and disastrously founder on the reefs of cognitive incompetence. Electoral democracy defers the problem, for it does not require of the electorate to be knowledgeable, to be competent. But to defer a problem is to encounter it at some other point or occasion. It is strange that our searchlights should be so much trained on the "rational voter'~ an unnecessary and phantomlike entity-rather than on the "rational decision maker." I attempt to redress this misdirected focusing in the next chapter. Meanwhile one should never tire of repeating that knowledge - cognitive competence and means-ends rationality-is not a problem that the theory of democracy can afford to ignore from beginning to end.

5.8

Government and Ungovernability

In this chapter I have explored the foundations of democracy and the soil on which real-world democracies actually rest. Foundations require a solid ground. This is why this chapter relies heavily on evidence- indeed, and for once, on social science evidence. The foundations of democracy are the legs of democracy. It is only after having measured these legs and their pace that ideals come to prominence and give stimulation and direction to our walking. Of course, if we so prefer, we may reverse the itinerary, starting from the ideals and working back, as it were, from the head to the legs. What I find impermissible is to propose refoundings of democracy without ever coming to, or even mentioning, the evidence. Since this has become a widespread practice, it is well to recapitulate, by way of conclusion, on what masses of converging evidence go to confirm or, conversely, to disprove. First, that democracies are "governed" does not detract from the fact that they are democracies. This is so because, in the Western democratic polities, public opinion is an autonomous force, and because electoral power is an effective power. Let it be added that electoral power is far from being a merely intermittent power, as it is well explained by the principle of "anticipated reactions" and abundantly confirmed by the media and opinion poll inBu120

Government and Ungovernability

5.8

encing of the behavior of the power wielders. On the other hand, our democracies must be declared governed on the irrefutable ground that they are representative democracies. But this is not the question. The question is: What is the balance, and how does the balance shift, between the governed and the governors? More specifically, Is it true, and in what sense is it true, that in our democracies the demos increasingly is "less governed" and "more governing?" With respect to whether the trend of liberal democracies attests to a maximization of the power of the people, the interpretations loom large. In 1957 C. Wright Mills wrote: "Surely those who have supposed the masses to be ... on their way to triumph, are wrong. In our time ... the influence of autonomous collectivities ... is in fact diminishing. Furthermore, such influence as they do have is guided; they must now be seen not as publics acting autonomously, but as masses manipulated at focal points into crowds of demonstrators. For as publics become masses, masses sometimes become crowds." 73 Mills feared, then, that democracy was being minimized. When he expressed this fear, Georges Burdeau had just published three volumes on "governing democracy" whose gist was that democracies start as governed but that they have now become, or are in the process of becoming, governing democracies in that an omnipotent popular will imposes itself on the state. 74 As I see it, in almost all democratic countries government is indeed exposed to growing pressures from below, and the reality of mass democracy as well as the rhetoric of populistic democracy press more and more on the structures of the liberal-democratic state. Yet, when Burdeau asserted that some countries, notably the Eastern so-called popular democracies, had actually achieved the stage of a governing democracy, he was clearly guilty of confusing appearance with substance, more democracy with sham democracy. On the other hand, while Mills was keenly aware of the difference between appearance and substance, he propounded a conspiracy theory of history that I find very difficult to believe. In order to gain a more balanced perspective, two caveats are in order. A first caveat is that we had better not take demagogy for a paideia and end by believing that by fostering activist modes of intervention and response, we are approaching the ideals of democracy. Demagogy only shifts popular sovereignty from the locuses where it maintains a capacity for judgment and reasonableness to settings where it loses it. A crowd of thousands will approve enthusiastically a proposal that would undoubtedly be rejected if it were presented to the same people divided into small groups. Michels reduced the argument to its gist: "It is easier to dominate a mass than a small audience:' 75 By this path, then, we easily arrive at democracy by acclaim, at a massive 121

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

5.8

manipulation of popular sovereignty that shouts out the actual will of the actual people. The lesson to be kept in mind here is that not much is needed to drain a nominal maximization of popular sovereignty of all real substance. Let us also not forget that increasingly effective weapons have been invented to attack the human mind, against which our armor remains as feeble as it has always been. The second caveat is that less power of the governors does not imply, by any necessity, more power of the governed. The game need not be zerosum; it may well be a minus-sum game in which both parties lose, in which the power lost by the governors is not gained by the governed. This is the development that has received growing attention and is described as a state of overload and ungovernability. 76 And it seems to me that this is the more telling diagnosis. Aside from the eventual rise or return of charismatic leaders, the general trend of Western democracies has been thus far in the direction of a diffuse powerlessness, of impotence and paralysis; for quite some time the power in ascendance has been a multiple, cross-cutting "veto power;' the power of blocking action. So much with governed democracy. As for "governing democracy;' clearly the expression indicates an ideal. And my discussion suggests that this ideal has attained-or is nearing~the point of fulfillment beyond which ideals become counterfulfilling and start operating in the reverse. In fact, much of what is being acclaimed as a maximization of governing democracy is pure and simple deceit: illusions that have produced delusions-or worse. The nondeceitful realization of the ideal would be referendum democracy. Why is it in so little demand? Unless we assume that our maximizers are deliberate cheaters, we must assume that they too understand that referendum democracy would only precipitate a suicide. So, what are we supposed to do with the "governing democracy" ideal? In accord with the rules that govern the management of ideals, I would say that we are still perfectly entitled to pursue the ideal of a governing democracy in the meaning of more self-government. But in so doing we do have to reckon with the facts. Let us not delude ourselves: A maximization of democracy in input, and precisely along the path that makes the people less governed in that they become more self-governing, crucially hinges on the performance of the average citizen. This performance will not be improved by activist modes of participation, by an escalation of affect, intensity, or ideological heating. It may be improved, under the qualifications and caveats recalled earlier, by education; but this is surely a slow process, and by no means a miracle-producing recipe. Meanwhile, it is well to be clearheaded as to what public opinion is, about what it either does or cannot do. 122

Notes

There is little point in asking from the ordinary citizen who constitutes the bulk of public opinion to express, issue by issue, articulate, informed and "rational" judgments. 77 Man is not a winged creature; and since the beginning of time, whoever has ignored this simple truth has always brought us to the brink of a precipice only to explain, after we have fallen over, that we should have known how to fly. But there is an altogether different way of looking at public opinion: as a pattern of attitudes and a cluster of basic demands. From this standpoint, the public is by no means a "phantom:' As long as the public is allowed autonomy, public opinion is a protagonist that does carry weight. It would be entirely mistaken to infer from the poor quality of the ordinary citizen that he amounts to an absentee. He may well be politically illiterate, but he is there- and potently conditions decision making. While policy making does not spring from Berelson's "cultural taste" any more than music from the people who attend a concert, or literature from readers, public opinion assures the success or failure of a policy just as it establishes the success or failure of a writer. There is, then, no public opinion impotence. But neither is there a public opinion omnipotence. If public sentiment, or opinion, accounts for the success or failure of a policy, it seldom initiates a policy. 78 The average voter seldom acts, he reacts. Political decisions are seldom generated by the sovereign people, they are submitted to them. And processes of opinion formation do not start from the people, they pass through them. Even when opinion tides do occur, the triggering factor can hardly be said to reside in the people as a whole. When exerting an influence, the people are also influenced. Before they want something, they are often made to want it. 79 When we are told that the people themselves govern, let us make sure that what is on display is not a sheer facade democracy, a pure and simple sham democracy.

Notes 1. It is often said that elections compute preferences. However, unless "preference" is understood as a sheer volition, voting preferences result from, or are conditioned by, opinions (as they will be defined). The two theses are complementary. 2. A. V. Dicey, Lectures on the Relation between Law and Public Opinion in England during the Nineteenth Century (London: 1905), p. 3. 3. For these precedents, see W. Bauer, Die Oeffentliche Meinung und ihre Geschichtliche Grundlagen (Tubingen: 1914). See also N. Matteucci, "Opinione Pubblica;' in Enciclopedia def Diritto (Milano: Giuffre, 1980), 30:420£f. 4. It should be stressed that the medieval "consent" is very distant from our concept. See especially M. V. Clarke, Medieval Representation and Consent. 5. This characterization includes opinions about candidates and parties; it ex-

123

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

eludes, however, the opinions about "private objects" currently covered by students of mass communication and pollsters. 6. There is, admittedly, a difference between consensus and consent, as pointed out by the perceptive analysis of P.H. Partridge, Consent and Consensus (London: Macmillan, I97I). Semantically, "consensus" does not imply any active involvement (it may consist of a sheer, totally passive acquiescence), whereas "consent" points to an activity of consenting. We may either say, therefore, that consensus is the broader category of which consent is a species or that consensus-consent constitutes a continuum that ranges from a weak to a stronger meaning of the co,icept. While my argument is, basically, on consensus, "consented government" is not, in the case in point, inappropriate, for a specific consent is also involved. 7. J.P. Plamenatz, Consent, Freedom and Political Obligation, 2d ed. (Oxford: Oxford University Press, 1968), defines it as follows: "Where there is an established process of election to an office, then, provided the election is free, anyone who takes part in the process consents to the authority of whoever is elected to the office" (p. 170). On a more empirical plane, electoral consensus may either be "supportive" or simply "permissive?' On this distinction see V.0. Key, Public Opinion and American Democracy (New York: Knopf, 1961), pp. 28-35. 8. This is to agree with A. Passerin d'Entreves: "political philosophy ... appeared to all those who throughout the centuries have practiced it, as a critical inquiry on the grounds of political obligation?' (See "On the Notion of Political Philosophy;' in K. von Beyme, ed., Theory and Politics [Den Haag: Nijhoff, 1971], p. 312 and passim.) Since this inquiry is not part of my brief, the reader is referred to J.R. Pennock and J.W. Chapman, eds., Political and Legal Obligation (New York: Atherton, 1970); R.E. Flathman, Political Obligation (New York: Atheneum, 1972); Carole Pateman, The Problem of Political Obligation: A Critical Analysis of Liberal Theory (New York: Wiley, 1979); A. J. Simmons, Moral Principles and Political Obligations (Princeton: Princeton University Press, 1979); John Dunn, Political Obligation in Its Historical Context: Essays in Political Obligation (New York: Cambridge University Press, 1980 ); J.S. Fishkin, The Limits of Obligation (New Haven: Yale University Press, 1982). 9. See A. Etzioni, Demonstration Democracy (New York: Gordon and Breach, 1970). IO. This is not to dismiss the weight and role of opinion polls but to firm up that elections, and elections only, are the institutionalized procedure for registering a state of opinion. II. Hobbes typically represents the conflict interpretation. Durkheim's concept of solidarity and T. Parsons's stress on integration are illustrative of the consensus perspective. To be sure, a number of authors stand on a middle ground. I would say, e.g., that R. Dahrendorfs view that "in a free society conflict ... is still there, and it is there to stay" (Class and Class Conflict in Contemporary Society [Stanford: Stanford University Press, 1959], p. 318) belongs to a pluralistic just as much as to a conflictual interpretation of society. 12. Characteristics such as support, approval, deliberation, decision, are accompanying properties that qualify specific forms of consent. For an overall conceptual analysis, see Partridge, Consent and Consensus; and G.J. Graham, "Consensus;' in Sartori, ed., Social Science Concepts: A Systematic Analysis. For a sociological analysis, see A. Etzioni, The Active Society (New York: Free Press, 1968), pt. 4. See also G. 124

Notes Parry, "Trust, Distrust and Consensus;' British Journal of Political Science, April r976. r3. See D. Easton, especially A Systems Analysis of Political Life (New York: Wiley, 1965), chap. r8. Easton's specific terminology is "support" for the regime and for the authorities. r4. See G. Almond, especially "Comparative Political Systems;' now in Political Development (Boston: Little Brown, 1970), chap. r. To speak, in this connection, of "shared ideology" can either be misleading or untelling. r5. That is to say that in other realms-beliefs about God, the family, virginity, abortion, etc.-a society may be highly dissensual. r6. As H.J. McClosky puts it, "a democratic society can survive despite widespread popular misunderstanding and disagreement about basic democratic and constitutional values:' (See "Consensus and Ideology in American Politics;' American Political Science Review, June r964, p. 376 and passim.) McClosky's conclusions are questioned, however, by S. Eldersveld, Political Parties: A Behavioral Analysis (Chicago: Rand McNally, 1964), pp. 183-219. 17. If the indicator for dissensual societies at the belief level is polarization (defined and measured, e.g., as a left-right ideological distance), then the polarized democracies are the least functioning and most breakdown-prone democracies. This is my argument in Parties and Party Systems, esp. chaps. 6 and 10,4, But see esp. G. Sani and G. Sartori, "Polarization, Fragmentation and Competition in Western Democracies;' in H. Daalder and Peter Mair, eds., Western European Party Systems: Continuity and Change (Beverly Hills: Sage, r982). r8. In this connection the ulterior point is that as long as a crisis of legitimacy is not resolved, all the other "crises" cannot be resolved and democracy eventually succumbs to a crisis overload. See L. Pye, "The Legitimacy Crisis;' in L. Binder et al., Crises and Sequences in Political Development (Princeton: Princeton University Press, 1971). On how legitimacy is reinforced or enfeebled see esp. S.M. Lipset, Political Man (Garden City, N.Y.: Doubleday, 1960), pp. 77-83. r9. This statement is qualified in chapter 6, sections I and 2. Here the point simply is that procedural consensus is well separable from substantive consensus. 20. Reflections on Government (Oxford: Oxford University Press, 1942), pp. 67 and 67-72. 2r. This implies that it is not basic consensus, but procedural consensus that needs, if the need arises, to be enforced (and was in fact enforced on non-consenters when Western democracies were established). 22. The best defense of this view is D.A. Rustow, "Agreement, Dissent and Democratic Fundamentals;' in von Beyme, ed., Theory and Politics, pp. 328-42. Here Rustow leans very much on Friedrich's The New Belief in the Common Man (Brattleboro: Vermont Publishing Co., 1942), where one finds the startling thought that requiring agreement upon fundamentals might be a "first step in the direction of totalitarianism:' However, this suggestion crucially hinges on the premise that totalitarian regimes "made frantic efforts ... to secure uniformity of opinion and belief" (p. r57). The distinction between autonomy and heteronomy of public opinion (section 4, below) implies that Friedrich's inference is impermissible: The two states of opinion have nothing in common. Rustow himself held a more balanced view in A World of Nations (Washington, D.C.: Brookings Institution, 1967), pp. 233-35. 23. This is a most condensed summary of my analysis of pluralism in Parties

125

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

and Party Systems, chap. 1. It should be understood that I am not referring here to the various, specific schools of pluralism: the early English pluralists (J.N. Figgis, F.W. Maitland, G.D.H. Cole, S.G. Hobson, and the early Laski), the American "group approach" pluralists (from A. Bentley to David Truman), or the current polyarchal theory of pluralism (examined in chapter 6, especially sections 5 and 7). For a succinct overview, see D. Nicholls, Three Varieties of Pluralism (New York: St. Martin's Press, 1974). My point also precedes the criticism of "democratic pluralism" illustrated, e.g., in W.E. Connolly, ed., The Bias of Pluralism (New York: Atherton, 1969). 24. This case was well made by J. Stuart Mill, On Liberty (1859), esp. chap. 2. 25. See Karl Deutsch, The Analysis of International Relations (Englewood Cliffs, N.J.: Prentice-Hall, 1968), pp. 101-10. 26. On this and other counts the cascade model does inflict a mortal blow to the single ruling class theory discussed in chapter 6, section 5. The cascade model equally challenges, by implication, the Marcusian speculations on the "one-dimensional man." 27. The other major reason is (we shall see shortly) that opinions may not derive at all from media exposure and information. 28. The importance of this stratum was discovered by the early Columbia studies (see P.F. Lazarsfeld, B. Berelson, H. Gaudet, The People's Choice [New York: Columbia University Press, 1944, 1948]; and Berelson, Lazarsfeld, W.N. McPhee, Voting [Chicago: University of Chicago Press, 1954]), and becomes theoretically incorporated in the so-called "two-step flow" model (see, e.g., E. Katz, "The Two-Step Flow of Communication;' Public Opinion Quarterly, Spring 1957). Deutsch's insight is to amplify and reformulate this partial model into an overall, multi-step flow. 29. The Phantom Public (New York: 1925). Also Lippmann's Public Opinion (New York: 1922) remains a classic analysis. 30. This is, in D. Bell's scenario of the post-industrial society ( The Coming of the Post-Industrial Society: A Venture in Social Forecasting [New York: Basic Books, 1973]), one of its characterizing features. 31. Voting, p. 3rr. The French electoral sociology finds historically based electoral predispositions (still rooted in the memories of the 1789 revolution) that add a further, potent element of viscosity to Berelson's imagery. See F. Goguel, ed., Nouvelles Etudes de Sociologie Electorale (Paris: Colin, 1954), and, in general, the tradition of electoral geography inspired by Andre Siegfried. 32. The first systematic work on public opinion is A.L. Lowell, Public Opinion and Popular Government (New York: 1913). Lowell was concerned with a very different distinction, namely, the difference between a true "public" and a mere "majoritY:' 33. For a general overview, three useful readers are D. Katz et al., eds., Public Opinion and Propaganda (New York: Dryden Press, 1954); W. Schramm, ed., The Process and Effects of Mass Communications (Urbana: University of Illinois Press, 1954); B. Berelson and M. Janowitz, eds., Reader in Public Opinion and Communication, 2d ed. (Glencoe: Free Press, 1966). 34. I retranslate from W. Schramm's encyclopedia article on "Mass Communications;' in Enciclopedia def Novecento (Roma: Istituto Enciclopedia ltaliana, 1975), 1:913. 35. In what follows I largely draw from my article on "Public Opinion;' in Enciclopedia def Novecento (1979), 4:937-49. 36. "Degreeists" have come to assimilate also propaganda, indoctrination, and

126

Notes education; yet, at least one line of demarcation cannot be canceled, namely, the difference between an exposure to many, alternative views (education) and the exclusion of all views but one (indoctrination). 37. I do not recall the bubbling-up model and the identification base of public opinion, because these ingredients already attest, eo ipso, to its independence. 38. This is what J. Habermas eminently fails to do in Strukturwandel der O/fentlichkeit (Neuwied: Luchterhand, 1962), whose English title is History and Critique of Public Opinion; his denunciation of the Western debasement of public opinion is never compared to its fate under unicentric state monopolies. Yet, if Western public opinion reflects a "false consciousness;' this must be even more the case in the single ideology, single ruling class societies. 39. For the employability of the concept of totalitarianism, see chapter 7, section 4. My argument is confined here to the totalitarian case because it is the opposite limiting case. 40. See, in general, P.E. Converse, "Public Opinion and Voting Behavior;' in F. Greenstein and N. Polsby, eds., Handbook of Political Science (Reading: AddisonWesley, 1975), 4:75-169. More specific findings are analyzed in depth by N.H. Nie, S. Verba, J.R. Petrocik, The Changing American Voter (Cambridge, Mass.: Harvard University Press, 1976). A basic text of the Michigan research school remains A. Campbell, P.E. Converse, and W.E. Miller, eds., Elections and the Political Order (New York: Wiley, 1966), which was preceded by Campbell et al., The Voter Decides (Evanston: Row, Peterson, 1954), and The American Voter (New York: Wiley, 1960). For the earlier Columbia studies, see n. 28 above. 4r. Even though the studies quoted in the note above shy away, as a rule, from assessing the bearing of the findings on the theory of democracy. See, however, Berelson's "Democratic Theory and Public Opinion;' Public Opinion Quarterly, Fall 1952, as well as Voting, chap. 14; and the volume edited by E. Burdick and A.J. Brodbeck, American Voting Behavior (Glencoe: Free Press, 1959), in which Burdick and a number of contributors attempt generalizations from the electoral data to political theory. 42. R.E. Lane and D.O. Sears, Public Opinion (Englewood Cliffs, NJ.: PrenticeHall, 1964), p. 13. 43. See, for all, J. W. Prothro and C.M. Grigg, "Fundamental Principles of Democracy: Bases of Agreement and Disagreement;' Journal of Politics, May 1960, which finds a consensus at the abstract level that is not followed up, and appears inconsistent with the specific responses obtained at the concrete level. Reference is made to the concept of consensus discussed above in section 2. 44. The large-scale direct evidence to this effect is provided by the polls. See especially two articles of H.G. Erskine in Public Opinion Quarterly: "The Polls: The Informed Public" (1962, pp. 669-97), and "The Polls: Exposure to Domestic Information" (1963, pp. 491-500). Most textbooks on public opinion illustrate and buttress the point at length. In addition to the works of Converse and of Lane and Sears, see: D.S. Ippolito and T.G. Walker, Public Opinion and Responsible Democracy (Englewood Cliffs, NJ.: Prentice-Hall, 1976); and R. Weissberg, Public Opinion and Popular Government (Englewood Cliffs, N.J.: Prentice-Hall, 1976). 45. In McKeon ed., Democracy in a World of Tensions, pp. 318-19. See, in general, N. Abercrombie, S. Hill, and B.S. Turner, The Dominant Ideology Thesis (London: Allen & Unwin, 1980). I27

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

46. This is not to imply that American scholars have not been sufficiently comparative. The literature on development and modernization (reviewed, e.g., by S.P. Huntington and JI. Dominguez, "Political Development;' in Greenstein and Polsby, Handbook of Political Science, 3:33-47) attests to the contrary. The text implies, however, that the research designs of the cross-national inquiries are heavily influenced by the American assumptions. 47. To illustrate, in the index of polarization developed by Sa~i and Sartori, the United States obtains the value .08 compared to .27 for Switzerland, .28 for Germany, .3I for the United Kingdom, and .64 for Italy and Finland. Set; Daalder and Mair, Western European Party Systems, p. 324. 48. In Greenstein and Polsby, Handbook of Political Science, 4:98. 49. ].A. Schumpeter, Capitalism, Socialism and Democracy, 2d ed. (New York: Harper, I942), p. 262. 50. S. Brittan, Participation Without Politics (London: Institute of Economic Affairs, I975), pp. 37-38. 5I. A Preface to Democratic Theory, pp. I25, I27. 52. W.G. Runciman, Social Science and Political Theory (Cambridge: Cambridge University Press, 1963), "crudely" asserts that "what Arrow has done is to show that strict democracy is impossible" (p. 133). I would not go that far; see chapter 6, n. 18, herein. 53. The paragraph is a condensed summary of my argument in Parties and Party Systems, esp. pp. 328-42. 54. In Greenstein and Polsby, Handbook of Political Science, 4:u9; but see 4:u825, passim. 55. A collectivized decision is a decision that binds the collectivity and is, by the same token, taken away from the individual. It should not be confused, therefore, with collective decisions, which are simply decisions taken by a given collectivity. "Collectivized" bears on the nature of a decision, not on who takes it. That the distinction is crucial is shown in chapter 8, herein. 56. In Handbook of Political Science, 4:n5. 57. See chapter 4, section 3. 58. See n. 20, above. 59. On this consideration it is apparent that the bibliography on participation differs from the bibliography on participatory democracy. Concerning the former, see, generally, L. W. Milbrath, Political Participation (Chicago: Rand McNally, 1965), 2d ed. with M.L. Goel, 1977; G. Di Palma, Apathy and Participation (New York: Free Press, I970); and N. Nie and S. Verba, "Political Participation;' in Greenstein and Polsby, Handbook of Political Science, vol. 4. Concerning the latter, specific authors will be discussed in chapter 6, section 8. Three good symposia are Geraint Parry, ed., Participation in Politics (Manchester: Manchester University Press, 1972); T.E. Cook and P.M. Morgan, eds., Participatory Democracy (San Francisco: Canfield Press, 1971); J.R. Pennock and J.W. Chapman, eds., Participation in Politics (New York: Atherton, 1975). A recent proposal is B. Barber, Strong Democracy: Participatory Politics for a New Age (Berkeley: University of California Press, 1984). 60. This is not to deny that a historical and conceptual analysis reveals complexities, as shown by L.A. Scaff, Participation in the Western Political Tradition (Tucson: University of Arizona Press, 1975). See also Maurizio Cotta, "II Concetto di Parteci-

128

Notes pazione Politica;' Rivista Italiana di Scienza Politica, 1 (1979). My point addresses, as the text makes clear, the ease with which "participation" can be operationalized. 61. Just to recall a few names, the importance of voluntary associations was first signaled by Tocqueville; the theory of multigroup society spans from the early English pluralists to Merriam and David Truman; and participation has always been called upon as the antidote of Michels's law of oligarchy (see chapter 6, section 6). 62. "Populism" is a term of Russian origin, and Franco Venturi's magisterial IL Populismo Russo (Torino: Einaudi, 1972) conceives it as the predecessor, until 1881, of Russian socialism. At another extreme Dahl dignifies ~nd substantiates the notion of "populistic democracy" by bringing under this label the prescriptive theory of democracy (see A Preface to Democratic Theory, chap. 2). This is not, however, the usual understanding of "populism" to which I make reference in the text. Whether a democratic prescriptivism becomes populistically employed depends on how it is employed, and specifically on how the function of ideals is understood (see chapter 4, especially sections 4 and 5). 63. In one of the many meanings of the expression discussed in chapter 6, section 2. The zero-sum characteristic of referenda is explicated in chapter 8, sections 3 and 6. 64. The notable exception is R.P. Wolff, who does outline an "instant direct democracy" that corresponds to what I call referendum democracy. See In Defense of Anarchism (New York: Harper & Row, 1970), pp. 34-37. While Wolff's argument is feeble, he deserves credit for his consistency. 65. For the pure or "radical" viewpoint, see C.G. Benello and D. Roussopoulos, eds., The Case for Participatory Democracy (New York: Viking Press, 1971). 66. See chapter 4, sections 4 and 5. 67. More exactly, Weber fundamentally distinguished between a value-related and an end-related rationality in social action (see esp. Wirtschaft und Gesellschaft, chap. 1, ii, 2-3, and passim). Reference is made, in the text, to the latter one. See, for an analysis of Weber's understanding of rationality, Pietro Rossi, "La Teoria della Razionalita in Max Weber;' in G. Duso, ed., Weber: Razionalita e Politica (Venezia: Arsenale Cooperativa Edi trice, 1980 ). 68. Public Opinion, pp. I05-6. 69. In and by itself, extremism can be defined as "going to an extreme in zealous attachment to a particular value;' thus involving "disregarding the rules which hamper the achievement of the extreme" (E.A. Shils, The Tonnent of Secrecy [Glencoe: Free Press, 1956], p. 231). In this connection the important point of Shils is that the "really crucial dividing line in politics is between pluralistic moderation and monomaniac extremism" (p. 227), and that it has been the grievous mistake of both conservatives and liberals to believe that they had "more in common with those who claimed to represent their values by extreme and accentuated methods than ... with those who opposed them within a framework of moderation" (p. 228). My focus is, however, on the cognitive, not the behavioral, implications of extremism. 70. In Berelson et al., Voting, pp. 314-15. See also W.H. Morris-Jones, "In Defense of Apathy: Some Doubts on the Duty to Vote;' Political Studies, February 1954. 71. As W.G. Runciman puts it: "It could, for instance, be argued that non-extremism is a good thing but non-partisanship a bad one" (in Laslett and Runciman, Philosophy, Politics and Society, 1962, p. 42). 129

GOVERNED DEMOCRACY AND GOVERNING DEMOCRACY

72. The discussion is pursued in chapter 8, herein, where the problem of intensity is centrally addressed. 73. The Power Elite (New York: Oxford University Press, I957), p. 309. 74. See vols. 5, 6, 7 of Burdeau's Traite de Science Politique, esp. the last two, which are specifically dedicated to a democratie gouvernante. 75. See R. Michels, La Sociologia de! Partito Politico (Torino: I9I2), pp. 24-27. 76. The point is pursued in chapter 6, section 9, where the bibliography is given inn. rn8. 77. This point is reinf~rced by the distinction between "popular" and "public" opinion forcefully pinpointed by Robert Nisbet ("Public Opinion versus Popular Opinion;' Public Interest, Fall I975), in whose view the opinion "that can be had from any person on any subject, however complex or remote, merely by asking" (p. I85) is a wholesale distortion of what public opinion meant to the authors who labored on this concept. 78. Morris P. Fiorina, Retrospective Voting in American National Elections (New Haven: Yale University Press, I98I), well develops the point into a reward-punishment theory of retrospective voting. 79. These considerations are qualified by my previous discussion of the cascade model of opinion formation, section 3, above.

6. Vertical Democracy A leaderless society is not a society at all, for whenever two or more men form a society and live together there is no such thing as uncontrolled, unrestricted, uninfluenced behavior. -F.S. Haiman

6.I

Majority Principle and Minority Rule

Politics has to do, in the main and most of the time, with subordination, superordination, and coordination - in essence, with the hierarchical structuring of collectivities. I shall call this the vertical dimension of politics. Politics also has a horizontal dimension, but this dimension becomes salient only in democracies and shares, historically, their destiny; it ceased to characterize politics with the downfall of the Greek polis and did not significantly reenter the theory and practice of politics until the nineteenth century. This is well shown by the development of the vocabulary of politics from the time of Aristotle onward. The only Latin term that maintains the Greek, horizontal vision of politics is respublica, whose closest English rendering is "common weal;' All the other terms developed during the Middle Ages- such as principatus, regnum, dominium, gubernaculum - refer to the verticality of politics. The same is true for the distinctive contributions of Machiavelli and Bodin, respectively, the terms state and sovereignty. And it is still the case that the vocabulary of politics that applies to any and all polities- power, rule, command, coercion, government, state-characteristically addresses the vertical, not the horizontal, dimension.1 In the light of the distinction between vertical and horizontal politics, it is apparent that in the preceding chapter our focus was on horizontal democracy. Public opinion, electoral democracy, participatory democracy, referendum democracy- all represent a horizontal implementation and diffusion of democracy. This is a correct start, for the uniqueness of democracy resides precisely in establishing, or reestablishing, the horizontal dimension of politics. Yet democracy is not anarchy-lack or absence of command. Public opinion, elections, participation, and a deciding demos (in some form or other) represent the foundations of the edifice; but foundations, essential as they 131

VERTICAL DEMOCRACY

6.1

are, are something that support a superadded construction. It is time, then, to look to democracy as a system of government and, more generally, to the vertical structuring of democracy-vertical democracy, for short. If electoral democracy typically sums up the horizontal layout of democracy, its vertical follow-up or conversion is representative democracy. But the vertical structuring of democracy raises issues that cannot be disposed of by the theory of representation. The nasty question is: How is it that majority rule winds up, in the end, with minority rule? The merit of this question (as framed) is that it prompts us to take a close look into the terms rule, majority, and minority. Let us begin by analyzing "rule;' which is a very tricky term indeed. While it is clear that rule may either be intended as a principle (criterion or method) or as a substantive rulers hip, what is often not clear is which one of the two meanings is intended. A same sentence may make perfect sense in either or both meanings. To dispel ambiguity, when rule is used in the first meaning we are well advised to say, in full, rule of the game, or otherwise principle or criterion. As regards the second meaning-rule as rulership-the caveat is that there is ruling and ruling. The rule of Stalin has little, if anything, in common with the rule of F.D. Roosevelt, that is, there is a world of difference between dictatorial rulership and democratic leadership-a difference that can be rendered (I have just suggested) by saying "leadership" whenever rule is intended in a weak sense, as "direction" more than outright command. 2 So, the seeming contradiction between majority rule and minority rule may simply result from the wording. If the relation under consideration happens to be one between a criterion (regula) and who rules (regulator), then the contradiction disappears. In the first place, not only is a majority principle not a substantive majority rule, but the overall architectonics of democracy makes the derivation of the latter from the former highly implausible. In the second place, minority rule may be descriptively misleading and may be challenged on two grounds: (a) that democracy produces minorities (in the plural), not a minority in the singular; (b) that democracy does not permit "rule" in the strong sense but only in the weak sense conveyed by the term leadership. Taking a step back, we should keep in mind that what is being investigated is a highly complex, overall interplay between the governed and the governing. This interplay consists of a multistage and multifaceted process in which concrete majorities and concrete minorities materialize (and eventually dissolve) in various ways at various levels. We shall attempt to unravel this complicated interplay as we proceed. At the outset, only two points stand firm: where the process starts, and what it purports to avoid. The process starts with the rule (of the game) that establishes how conflicts are to be solved; and this rule 132

The Tyranny of the Majority

6.2

(method) is the majority principle. As for the intent, it was eloquently prefaced by Hamilton: "Give all the power to the many, they will oppress the few. Give all the power to the few, they will oppress the many?' 3 The intent, then, is to avoid giving "all the power" either to the many or to the few by distributing it in turn and/or concurrently to majorities and minorities.

6.2

The Tyranny of the Majority

We must now disentangle the contexts or settings to which the terms majority and minority are variously applied. The settings that require separate consideration are, at a minimum, three: (a) constitutional structures and processes, (b) electoral and voting arenas, and (c) society at large. Therefore let us say that the majority-minority relation obtains three major, context-embedded, clusters of meaning: constitutional electoral (voting) societal In the constitutional context the concern is about minorities, not majorities. More precisely, the problem that comes to the fore in this context is that the minority or minorities must have the right to oppose, the right of opposition. 4 It is here where the expression "majority rule and minority rights" acquires its most precise meaning and a particular prominence. If the opposition is hampered, harassed, or stamped on, we may thus speak of "tyranny of the majority" in the constitutional meaning of the expression. There is, however, a second constitutional meaning of tyranny of the majority, that which Madison and Jefferson called "elective despotism." What they feared was the despotism of an assembly government unrestricted by division of power: an elective body (a parliament, but specifically its lower house) that would concentrate in its hands an unlimited and, by this token, a tyrannical power. 5 That their fear was well justified was shown soon after by the French revolutionary gouvernement conventionnel, indeed a perfect incarnation of elective despotism. However, the elective despotism envisaged by Madison and Jefferson does not really bear on the majority-minority relationship but on the principle that undivided power is always an excessive and dangerous power. So, the tyranny of the majority that acquires prominence in the constitutional perspective is the one that bears on minority rights, and especially on whether the right of opposition is respected or not. In the electoral-voting context the argument takes an entirely different twist. Here the focus is exclusively on the majority principle, that is, on "ma1 33

VERTICAL DEMOCRACY

6.2

jority" understood as a rule of the game. And the argument is, very simply, that whoever votes with the majority (i.e., as most other voters do) is on the winning side. Conversely, whoever votes with the minority (thus failing to join a plurality) is on the losing side: His vote counts for nothing. In voting, then, "minority" simply denotes those who must submit to the will of the majority (even if a simple plurality). The point therefore is that, in voting, the minority has no rights: It consists of those whose vote was lost- period. 6 The implication is that in electoral-voting contexts the expression "majority tyranny" is inapplicable and meaningless. To be sure, while each act of voting is a finite, single-shot act, when we cast a vote we may well initiate a process, as is especially the case with electoral voting. And while the single voting act should not be confused with the process that it eventually triggers, it should not pass unnoticed that the process not only affects the notion of "winning" but is bound to multiply the number of losers. The winning elector is an elector who gets his candidate elected (or who votes for a party list that obtains, in his constituency, at least one seat). However, if we pass on to consider the ensuing sequel of events, which is a multi-stage process, it is apparent that the elector may win at the constituency level and yet lose at other levels: at the parliamentary level (where his representative may belong to a minority party), and again at the governmental level (when his party is excluded from government). The bearing of the foregoing is twofold. First, it brings out the importance of treating separately the voting act from the voting process. Second, it bears out the earlier point that we cannot derive a majority rule (rulership) from the majority principle. One reason for this is that if the majority principle applies at successive levels, a multi-stage processing may well eliminate, one at a time, a sequel of minorities that add up to a majority of the initial voting population, of the demos at large. Let us now turn to the meaning of "majority" in the third setting, that is, in a society-wide or societal context. The social meaning of the term majority is the meaning characteristically attributed to the expression "tyranny of the majority" by Tocqueville and by John Stuart Mill. What troubled Tocqueville and, subsequently, Mill, was the danger of a spiritual tyranny, that is, of an extreme and suffocating social conformity. Here the majority-minority relationship is no longer important in itself but in its bearing on, and for, the individual. Consequently the focus shifts onto the relationship of society to the individual. The antithesis is between majority versus freedom of the individual, or between majority versus intellectual independence. "The democratic republics make despotism superfluous;' said Tocqueville, "because the majority itself draws a formidable ring around thought?' 7 And Mill wrote: 1 34

The Tyranny of the Majority 6.2 When society itself is the tyrant-society collectively over the separate individuals who compose it . . . it practices a social tyranny more formidable than many kinds of political oppression, since ... it leaves fewer means of escape, penetrating much more deeply into the details of life, and enslaving the soul itself. Protection, therefore, against the tyranny of the magistrate is not enough: there needs protection also against the tyranny of prevailing opinion and feeling, against the tendency of society to impose ... its own ideas and practices as rules of conduct on those who dissent from them . . . and to compel all characters to fashion themselves on the model of its own. 8

In the above, reference is made to some kind of substantive majority, hardly to the majority criterion or principle. Also, what Mill described has little to do with a majoritarian tyranny, but rather (in his own wording) with "social tyranny?' On either count, his case is not very convincing. I am not saying that collectivities do not control and eventually oppress their individual component members, but that Mill's social tyranny long precedes democracy and does not appear to have much to do with democracy. Small-town and village communities easily suffocate individuality precisely in the way he described; yet this is simply the unpleasant side of "commonality;' the drawback overlooked by the idealizations of communes and community life. A puritan community of earlier times was far worse-with respect to what Mill feared and lamented-than anything that he ever saw in his time. Should we, therefore, dismiss Mill's argument? And should we, by implication, dismiss also Tocqueville's case? I would not go that far. For Tocqueville did have a point when he spoke of tyranny of the majority-one that is missed when his concept is translated as "social tyranny:' The point is that the majority principle (note that I have switched to the principle) adds an element of legitimacy, a right, to what is otherwise a sheer fact, namely, that social conformity exists and entails costs and excesses. This is not quite the argument of Tocqueville. 9 It suggests, however, that there is a reason for bringing the notion of majority to bear on our misgivings about social tyranny. This reason is that "the tendency of society to impose its own ideas and practices;' i.e., to impose conformity, finds in the majority principle a principle of legitimization. And if this is our reading (between the lines) of Tocqueville and Mill, then also the social meaning of "tyranny of the majority" deserves to be kept in mind. Before proceeding further, it is well to recapitulate. First, the majority principle poses the problem of protecting minorities. This is, first and foremost, a constitutional problem. In this context we seek, then, a limited majority principle, that is, we seek to delimit and temper its application. Conversely put, if the majority principle is unlimited or absolute, then we have a "tyranny of the majority" in the constitutional meaning of the expression. This qual1 35

VERTICAL DEMOCRACY

6.2

ification does not imply that the issue is confined to constitutional provisions. When we put our mind to the problem of empowering without overpowering, we soon discover that juridical checks alone cannot solve it. This is, however, an additional reason for stressing that a governing majority that crushes the rights of opposition does indeed embody a tyranny of the majority. In this context, then, the mark is well hit. Second, the majority principle confronts the problem of manufacturing a governmental majority. This is the case when the majority principle applies to the electoral-voting process. In this context, each majority (plurality) test eliminates-level by level-its corresponding minority. In order to fabricate a governmental majority the majority criterion can only perform, on each occasion, as a winner-take-all principle. It follows that the expression "tyranny of the majority" is meaningless in the electoral-voting context. At the end of the process the case may also be that a numerical minority of the citizenry emerges-at the governmental level-as the winning majority. Here, then, a majority is often only the largest minority. Third, the majority principle may aggravate social tyranny (as characterized by John Stuart Mill) by legitimizing it. In this respect the tyranny of the majority feared by Tocqueville and Mill-tyranny over the individual-remains a matter of concern. At first sight it may seem odd that the founding authors of liberal democracy- Madison, Jefferson, Tocqueville, Mill-were so much more concerned with a tyranny of the majority (in one sense or another) than with a tyranny of the minority. But reflection shows otherwise. Since democracy replaces a tyranny by some minority, and since its rule (principle) is that the majority is always right, it attests to their farsightedness that their attention should shift to the "opposite danger;' to the danger inherent in the new principle. Another noteworthy point is that the analysis of the principle does not bear out a substantive majority rule of the many, of the multitudes or, if you like, of the masses. The reason for this is not far to seek. When reference is made to an institutionalized body (a government, a parliament, a party), the referent of "majority" is some kind of cohesive and identifiable operating unit. But when reference is made, as in the case at hand, to large-scale, dispersed collectivities, the referent of "majority" generally is a set of ephemeral aggregations. An electoral majority is largely an artifact of an electoral occasion and thereby largely an artifact of the party system qua system of channelment.10 As for issue majorities, they tend to dissolve and recompose themselves issue by issue. Bluntly put, the majority of a citizenry- a "mass majority'~ is a process of endless amalgamation and dissolution of myriad groups and individuals. This is not to deny that even a mass majority may coalesce and per-

Election, Selection, and Mal-Selection

6.3

form, over time, as an operating unit. However, a mass majority becomes an "operating majority" if, and only if, it acquires some kind of fixity, of over time cohesiveness. For this to happen, a majority of a population must consist of strong party or class or racial identifiers. In Western democracies this has seldom been the case. Most of the time concrete mass majorities are also intermittent and mobile majorities that cannot sustain or produce a "majority rule" (rulership) in any strong sense of the expression. So, with reference to mass majorities there is little reason to fear a tyranny of the majority; and the point is that from the majority method for making decisions one cannot derive that there is any one group that constitutes the majority and makes the decisions. The majority method connotes only a mathematical majority; it does not denote an enduring major part of a collectivity. 11 I noted that the overall design is to ayoid giving "all power" to either majorities or minorities. Our analysis goes to show that this is precisely what happens. In particular, along the electoral-voting process concrete majorities produce concrete minorities, which are in turn submitted to the majority criterion - and this all the way from a mass electorate up to a government.12

6.3

Election, Selection, and Mal-Selection

The vertical construction of democracy hinges on electing and elections free, recurrent, and competitive elections. And it is at elections that we have not only the decisive but also the full application of "majority rule" both as a concrete ruling and as a rule of the game. Surprising as this may seem to us, the idea of majority rule and, specifically, of the majority criterion can hardly be antedated to Locke.13 When Aristotle spoke of the "rule of the many;' he described a state of affairs, not a standardized criterion for solving controversies and making decisions. How little the Greek rule of the many was a criterion is attested to by the fact that Greek democracies generally elected their officials by lot, by a random mechanism. The medieval author who revived the idea of popular sovereignty, Marsilius of Padua, used the expression major et valentior pars: the greater and "most able" part. Throughout the Middle Ages and beyond, the notion of major pars, of the greater part, was never separated from the notion of melior pars, of the better part. And while it may be said that from Locke onward the majority criterion begins to emerge as we now know it, that is, as a quantitative criterion divorced from qualitative attributes, yet it was not until the French Revolution that this divorce became apparent. Rousseau, for one, had no part in this development, for his general will was "general" by virtue of having the quality of an objective, general interest. 137

VERTICAL DEMOCRACY

6.3

Why did the majority principle gain acceptance as slowly and as reluctantly as it did? The answer is, very simply, that the right of the majority is not the "rightness" of the majority. In 1801, in his first Inaugural Address, Jefferson said: ''Although the will of the majority is in all cases to prevail, that will to be rightful must be reasonable?' Jefferson thus took the step of giving priority to the majority principle, but in full and declared awareness that "rightfulness" cannot reside in numbers alone. The objection was trenchantly formulated a few decades later by a French doctrinaire, Royer-Collard: "The will of a single person, the will of many, the will of all, is only a force that can be more or less powerful. Neither obedience nor the slightest respect is due to these wills purely and simply because they are wills?' 14 And it was Taine, in 1875, who forcefully summed up in a famous line the crucial grievance: "Ten million ignorances do not make up one knowledge?'15 The problem, or the perplexity, may be phrased thus: What is the right of a sheer quantity? Why does the greater number have a greater value? I personally doubt that there is a knock-down answer to this question. However we may go about in replying, the answer hinges on a quantitative factor. For instance, we might construe the majority principle as a logically necessary implication of voting equality and argue, on this premise, that each will is "equal in value" to any other and therefore that the greater the number of wills gathered together, the greater their collective "weight of value?' But this and similar arguments are far from being unassailable, for the retort can still be that numbers create might, not right. A majority is a quantity-and a quantity cannot engender a quality. Recently, the defense of the majority criterion has been given a new twist and sought, by Douglas Rae and Michael Taylor, on probabilistic grounds. Rae assumes, in Rawlsian fashion, an original "veil of ignorance?' In the original position individuals are uncertain about what the agenda will be and how other individuals will act on it. Hence the individual will prefer a voting rule that minimizes the probability of his supporting an issue that is defeated, or of opposing an issue that wins-and majority rule is such a rule, for it minimizes the probability of his being outvoted. In Rae's words the unfolding of the argument is as follows: "Under majority rule the expected (summed) frequency of events A and B is at a minimum, and majority is therefore optimal under this criterion. This in turn suggests ... that majority rule maximizes the probability that our (anonymous) individual will 'have his way' with respect to a given proposal. And this, in the long run, suggests that majority rule will optimize the correspondence between individual values and collective policies?'16 I bow to the ingenuity of the above and yet feel that the argument strains plausibility and probabilities to a vanishing point. The Rae-Taylor

Election, Selection, and Mal-Selection

6.3

theorem assumes randomness and probabilistically significant frequencies. Neither conditions obtain. No ordinary voter is ever likely to feel that majority rule is "best" in that it gives him better chances of being less frequently outvoted or, conversely, of being more frequently on the winning side. Were this his calculus, he would be wrong, for the voter does not dispose, ·during his adult lifetime, of enough shots for such frequencies to occur. But why construe the justification of the majority principle either on winning-chances grounds or on grounds of rightfulness? Why not defend the majority principle like Churchill defended democracy, that is, on the argument that it is a ghastly criterion, except that others are worse? After all, what is being considered is a technique, an instrument. Any society needs procedural, conflict-solving, and decisional rules; and majority rule is that procedure or method that best fits the requirements of democracy. This is indeed my view. Yet, instruments are instruments for-for something. And the moment has come for taking a closer look at what the instrument is for and how it performs. I have just said that majority rule is the procedural rule that best suits democracy. But is that so? Why, for example, is it better than unanimity rule? The straightforward reply is that majority rule avoids stalemate while allowing large collectivities to have a say.17 But majority rule is also spoken of as a decision-making criterion; and here its "betterness" cannot be defended unqualifiedly. Decision making is a very broad category that covers disparate phenomena. Electors, we say, decide; but electoral decisions are very different from decision making as a process of issue deliberation. Thus, what an electorate decides is not what a committee decides; indeed, electoral deciding has little in common with deliberative deciding.18 In particular, electoral decisions are, as decisions, very thin; they merely or mostly "decide the deciders!' Let us therefore reserve the notion of decision making to how "deciders decide" and simply speak, in the electoral context, of election and/or selection. So our referent here is majority rule as an electoral instrument (not as a decisionmaking instrument). The question now is: How does majority rule perform as an electoral-selective instrument? Even though few are aware of it, the way of conducting elections, the secret vote, the simple versus the qualified majority, all of this and more were virtually invented anew by, and arrived to us from, the monks. There is nothing surprising in this fact. As early as the eighth century the monastic orders were faced with the problem of choosing their superiors. Since the monks could not resort to the principle of heredity or to force, they had to find a way of selecting their "heads" by electing them. As a result of experience and experiments carried on over centuries, the electoral constitutionalism of the 1 39

VERTICAL DEMOCRACY

6.3

various religious orders achieved a refinement and a complexity that is unsurpassed.19 And the constant, central element of this constitutionalism is precisely how the major pars must relate to, and must be conditioned by, the better part: the sanior pars, or the melior pars. Notwithstanding the fact that the case of religious orders relates to optimal conditions, monks knew well that even they were not angels, and they therefore never cea1;,ed to labor on how to choose the ablest and the fittest, on how to make sure that a majority of the worst would not overcome the minority of the best. 20 As we move to the lay world, the ancient regime crumbled, first and foremost, because people no longer accepted a society whose organizing principle was inherited privilege-privilege divorced from capacity and merit. Our liberal-democratic world is born from the vindication of the principle that the unjust rule of the nonelected-those who exercise power by right of heredity or conquest-be replaced by the rule of the "selected:' Men wanted to choose who was to rule them, and demanded the right to substitute their perception of merit and ability to an ascribed or self-declared merit and ability. No society that I know of is interested in having a rule of the worst. Thus the distinctiveness of democracy resides in the principle that no one can proclaim himself, by himself, as "better" than any other; this must be decided by others. Elections have been conceived, then, as an instrument of selection in the qualitative meaning of the term. John Stuart Mill still attests to the fact that, at the outset, also in our democracies elections were advocated and instituted as a quantitative instrument designed to make a qualitative choice, for the thread of his major political work is precisely to design a system of representative government in which parliament would contain "the very elite of the country:' 21 Designs often fail to work out as intended. As time went by, the quantitative emphasis usurped the place of the qualitative. While the original intention was to count in order to select, in present-day democracies the instrument has taken control over its purpose. As the indictment goes, majority rule has become a sheer "quantity rule" governed by the maxim: Catch as many votes as you can, in any way you can. We thus obtain a facsimile of Gresham's law. Just as the bad apples contaminate the good ones, so the law of quantity devalues quality. If elections were meant to select, they actually misselect or mal-select, that is, they select in reverse. Under the law of numbers, those worthy of choice are far more often than not squeezed out by those unworthy of choice. In the end, "valuable leadership" is replaced by poor leadership, by unworthy leadership. 22 It will be argued that no instrument has virtues without defects and that, in any event, the usurpation of the qualitative by a quantitative emphasis is 140

Minorities and Elites

6.4

an inescapable development, an unavoidable fact. Maybe. Yet it is equally true that democracy must, in order to exist, be value sustained. How is it, then, that a value chorus that has of late been so remarkably vociferous has remained so remarkably silent in this matter? After all, elections are the crucial starting point of the vertical construction of democracy. If it is true that elections misselect, and indeed contra-select, it follows that much of what democracy is about starts on a wrong footing. If values are important-as they are to the value chorus and to me-then a value preoccupation should be important at all points of importance. If we are to judge, however, by the current debate on electoral systems, it is apparent that the problem of selection and, in this connection, of the quality of leadership, has entirely disappeared from view. Our attention has been increasingly monopolized by the issue of "exact representation;' of how well or how inaccurately votes are translated into seats. I have no quarrel with this concern, except that it cannot displace other concerns. Whether or not electoral systems produce parliaments that represent their electorates "in proportion" is one issue. Whether electoral systems are selective processes that select for the worse, or not, is another issue. That elections should comply to a mirroring function is a legitimate demand. That elections should comply to a selective function is, I submit, an equally legitimate demand. Yet, to the best of my knowledge Ernest Barker was just about the last major author to stress (in 1942) that "we cannot abandon the idea of value; we cannot enthrone the majority just because it is a majority and superior in quantity. We have to find some way of linking value and quantity together as things inseparably connected:'23 The 1960s were characterized by a powerful reentry of the value concern. Even so, it is precisely those people who complain the most today about how little the values of democracy are revered, who most avoid seeking "some way of linking value and quantity." I take to task this facile Daltonism later. 24 At the moment I wish to underscore that majority rule is not a better instrument than others if we allow it to become a sheer rule of quantitative might. If the law of numbers is, by now, a fact, it needs even more than many other facts to be counteracted by a value pressure. This chapter is about this. For a democracy that surrenders to the inevitability of a worthless leadership, of mal-selection, is a democracy that the demos itself, in the long run, feels is not worth upholding.

6.4

Minorities and Elites

Enough of "majority" in its many meanings. It is now time to confront "minority;' not only in its many meanings, but also in its overabundant set of appel-

141

VERTICAL DEMOCRACY

6.4

lations: political class, ruling (dominant) class, elite(s), power elite, ruling elite, leading minorities, leadership, and more. This richness of denominations by no means implies that "minority" enjoys the advantage (over the term majority) of disposing of one name for each meaning. To the contrary, the wealth of denominations that refer to minority has only added confusion to profusion: "Different labels are used to refer to the same concept and different concepts are covered by the same label;' 25 Let us begin to clear the ground by noting that all the' above recalled expressions refer to some kind of concrete minor part, not to "minority" as an artifact of democratic procedures (such as the part of a voting population defeated at elections or the lesser part of a parliament). In the second place, when the political analyst addresses vertical democracy, he is not interested in any conceivable kind of substantive minority but only in the minorities that add up to some kind of controlling group. To be sure, religious, ethnic, linguistic, and other minorities play an important role in politics; but they do not enter the vertical argument unless they perform as a political controlling group. We may thus circumscribe our object of enquiry as follows: the extent and modality of the political controlling power of groups smaller than half the universe over which such power is exercised. Needless to say, the resources of political power may be nonpolitical (economic or other). It should be understood, therefore, that a controlling power is political when its resource base is political office, and/or whenever it operates via the channels of politics and affects the decisions of policy makers. 26 Another preliminary caution bears on the difference between the question, What is a controlling minority? and the question, Who belongs to a controlling minority? The first query addresses a conceptual problem, the second an empirical problem. The conceptual task is to define "controlling groups" with respect to their characteristics and to identify them, if they differ, by different labels. The empirical problem is to ascertain whether controlling groups actually exist and who is in control of what. From the conceptual analysis we demand a framework and/or a typology, while the empirical task is to verify which controlling group exists as defined, that is, as qualified by the characteristics imputed to it. Failure to perceive the conceptual and empirical inquiries in their distinctiveness or to address them in the appropriate order-conceptually before empirically-has led to a "theoretical literature dealing with elites and power ... [that] wallows in a sea of ineffable confusion."27 The criteria for singling out a controlling minority are numerous. Among them, two are of paramount importance. The first criterion is altimetric: A controlling group is such because it is located, along the vertical structuring

Minorities and Elites

6.4

of societies, "at the top?' Accordingly, we may say that in every society power resides in a top power class. The altimetric criterion assumes that he who is at the top "has power"; an assumption sustained by the rationale that having power brings you to the top, and indeed that one has power precisely because he is at the top. The retort could be that our criterion fares well as long as we confront a power pyramid; but far less well when we confront a "stratarchy;'28 that is, a power configuration without an apex (an untidy and truncated pyramid). The altimetric criterion, however, also applies to a stratarchy, provided that we specify that each stratum will have its own top (and that the ensuing complexities are accounted for). Let us say, then, that in the altimetric perspective every society is a stratarchy and that the resulting stratocracy may either be concentrated at one apex or distributed among various apexes. The altimetric criterion leaves us with the sanctification of a state of fact: Who is at the top is at the top, and who is there is "powerful;' wields and owns power. Can that be all? The Middle Ages and feudal-type societies rested on the principle that each person should live in accord with his own station; yet, it was already during the Middle Ages that the principles of the valentior, melior, and sanior pars were construed. And the ancient regime was overthrown precisely in the name of the value criterion that the vertical fabric of society should be entrusted to eminence (as recognized). From this vantage point, one is not at the top because one has power but, quite to the contrary, a person has power and is at the top because a person deserves it. The second criterion is, then, a merit criterion. How are these two criteria translated in, and expressed by, the current nomenclature? Since in the end it is Pareto's term-elite-that has prevailed,29 it is important to understand why Pareto adopted this term and how he conceived it. In his Treatise it is said very dearly that "elite" refers to the people who have the highest ranking in "capacity" in their field of activity. 30 But it is in an earlier work that Pareto comes the closest to providing a full definition of his concept, which reads as follows: These classes ["men ranked in accord with their degree of influence and political and social power;' and/or "the so called superior classes"] constitute an elite, an "aristocracy" (in the etymological meaning: aristos = the best). So long as a social equilibrium is stable, the majority of their component members appears to be eminently endowed with certain qualities, no matter whether good or bad ones, which ensure power. 31

There is little doubt, then, that Pareto picked out "elite" because it carried into French and Italian - his two connatural languages - the Latin connotation of eligere (a selective choice) and thereby, albeit indirectly, the original meaning of the Greek aristoi-the best by excellence (not by birth). Thus, 1 43

VERTICAL DEMOCRACY

6.4

Pareto's concept is, first, qualitative, and becomes altimetric by implication. To be sure, this implication provides the key to the Paretian "circulation of elites:' To wit, when desert and power are conjoined, then we have a stable state of social equilibrium; when they become disjoined, then we have a disequilibrium that brings about a circulation: The altimetric, de facto elites are displaced by the "capable;' i.e., true elites. While it may be said, therefore, that Pareto's concept was both meritocratic and altimetric, nonetheless the two criteria are linked in that order, and Pareto's ultimate' winner always is, in history, the capable elite, not the elite in power. More than anybody else it was, I believe, Lasswell who established "elite" as the general category for discussing what we have come to call, in his wake, the "ruling elite model:' But Lasswell adopted the word, not the concept, of Pareto. In Lasswell the qualitative connotation of the term elite disappears. One of his typical definitions reads: "The political elite is the top power class:' 32 This is a purely altimetric connotation. Elsewhere, elite simply coincides with "having power;' as in the following definition: "The elites are those with most power in a group:' 33 As anyone can see, this is a radical transformation of Pareto's conception; a transformation whose merit is overtaken by a demerit. The merit is analytic, namely, it is of analytic avail to separate the altimetric characteristic (or the de facto power characterization) from the qualitative characterization. The demerit is semantic: Why say "elite" when nothing is intended of what the term means, i.e., conveys by virtue of its semantic force? Furthermore, if "elite" no longer indicates qualitative features (capability, competence, talent), what term shall we employ when these characteristics are intended? Thus the semantic distortion generates, in turn, a roundabout conceptual distortion - as we shall see. It would lead me far afield to construct, or reconstruct, an overall framework in which the wordings render the conceptual focus. Let me simply firm up in summary form the points covered thus far. First, our subject is the controlling power of controlling groups. That is to suggest that this wording focuses the issue better than any other. Second, if we are to improve on Pareto with Lasswell and, conversely, to amend Lasswell with Pareto, then we should distinguish, both terminologically and conceptually, between power structure and elite structure. Not all controlling groups are, by definition or by any necessity, "elite minorities" (in Pareto's primary sense); they may simply be "power minorities" (in Lasswell's sense). If so, let us call them that; for either we differentiate the labels or the two things will inevitably get mixed up. There remains a third point. We have specified, at least along one crucial dimension, the notion of minority; but have yet to specify, in some basic sense, the notion of controlling power. 1 44

Minority Rule: From Mosca to Dahl

6.5

6.5

Minority Rule: From Mosca to Dahl

When we come to the heart of the matter, we are currently apt to refer to what Dahl calls the ruling elite model. Actually, what is intended and discussed under this label is, above all, Mosca's law of the political class. Hence my preference is for saying "ruling class model;' In any event, what matters is the notion of rule. As noted when analyzing majority rule, whenever the concept of rule is conceived substantively as an actual ruling, all rulerships are not alike. And when we come to minority rule, of the two connotations of "rule" the one that comes to the fore no longer is rule-as-criterion but ruleas-ruling. So it is now that we must look into what is implied by the notions of (a) controlling power, (b) leadership, and (c) rule. As I understand it, controlling power is the more general category, and this on two grounds. When we say controlling groups and power, the focus is on the actual ways and means of exercising power, regardless of whether they are formalized or institutionalized, and regardless of their intensity. The notion of leadership, and thereby of leading minority or minorities, applies more specifically to a political class; but its major semantic indication is that "command" is understood in a weak sense, as having low intensity. Finally, the notion of rule (ruling) definitely applies to a political class, generally involves the use of the state power apparatus, and points to the strong sense of "command;' that is, to the more intense kind of exercise of power. The above specified usages allow for the controversy on the ruling elite model (or, as I prefer to say, on the law of the ruling class) to be formulated as follows: first, whether or not control groups (power minorities) add up, in any given setting, to a singular or to a plural; second, and more precisely, whether or not such groups are (as in Meisel's three Cs formula) characterized by group consciousness, coherence, and conspiracy; third, and conclusively, whether or not in any given setting power minorities can be concretely nailed down. Depending on which side of these alternatives passes the test of verification, in one case the finding is appropriately summarized by speaking of ruling class in the singular, and in the other case by saying leadership and leading minorities in the plural. Ultimately, the basic distinction that our findings are called on to warrant is between (a) government by rulership and (b) government by leadership. 34 We have thus reached the point at which the conceptual analysis can fruitfully sustain an empirical analysis, and indeed requires empirical operationalization and testing. It was C. Wright Mills who, above all, influentially held that the United States was in fact dominated by a ruling class (Mills called it "power elite") characterized as the military-industrial complex: a cohesive clique of the big rich, the corporate bosses, the military brass, and a small 145

VERTICAL DEMOCRACY

6.5

group of key politicians. 35 And it is Dahl, above all, who responds to Mills with a critique of the ruling class model. 36 Dahl's model is central to all debates on the actual power configurations of political societies. For he finally sets down test conditions for ascertaining whether a polity is or is not controlled by a single, cohesive power group, by one ruling elite (as Dahl has it), or (as I prefer) by one ruling class. In the first place, if we assume that there is one ruling class, that a society is actually controlled by a power clique, then such an assumption cannot escape the burden of proof. For in such case a political class is not a mere category, a "class" resulting from a classification; we are imputing to it, instead, an operative, concrete existence-and an existent of this sort must be found to exist. The case is such, then, as to make it both feasible and imperative to trace down and identify who is the ruling class. Conversely, if the tracing fails, the assumption is disconfirmed. In the second place, the notions of power and control must be meaningfully operationalized. Here Dahl's point is that power is exercised, or that a power issue truly arises, only when a decision is controversial. Hence the test condition established by Dahl is that in order to prove that a ruling class exists we must prove that, with respect to a sequel of controversial decisions, one identifiable and durable group (the ruling or power group) regularly prevails. But why confine "power" to "controversial decisions"? Dahl convincingly explains that otherwise we are likely to impute power (actual power) to power resources or potentials-not only a confusion, but a confusion that opens up an endless regression toward a power that is so ultimate as never to be grasped. Dahl's conditions have been counteracted by the argument of Bachrach and Baratz that power is also power to suppress issues, to eliminate them from visibility;37 an argument that draws on Schattschneider's statement that "some issues are organized into politics while others are organized out:' 38 This is a statement of fact. However, Schattschneider interprets it as a "mobilization of bias':.__and this is wholly gratuitous. Not that I deny bias. But the most biased view is to call everything bias. Perhaps functional prerequisites for the existence of any society do exist. 39 Perhaps, also, politics would instantly become unmanageable if all conceivable issues were simultaneously brought to the fore and to resolution. So, perhaps, this is not bias. Awaiting a serious imputation of bias to be qualified and sustained, we are left with the assertion that "values ... are built into the political system:' 40 So they are. But what has this truism to do with the point at issue, which is (by now I myself have to be reminded) whether a single conglomerate of sinister interests masterminds all of this? The argument of Bachrach and Baratz is, as stated, difficult to prove or disprove, and also drifts away from the issue. 41 146

Minority Rule: From Mosca to Dahl 6.5 All in all, it seems to me that Dahl's two demands-that a ruling class (if assumed to be such) be concretely identified, and that power imputations be verifiable-well resist assaults. Furthermore, it is not correct to argue that his conditions are so stringent that he wins his case almost by definition. Actually I would contend that his conditions are perfectly applicable to a number of dictatorial polities and to all communist regimes, where one would indeed find (if research were permitted) a ruling group that passes Dahl's tests. On the other hand, Dahl's premises should not automatically lead to the inference that if no "ruling class" is found, then there is nothing to be found. Dahl confronts one model of power structure, not all models of power structure. What he convincingly demonstrates is that the labels "ruling elite" and "power elite" (in the singular) are misnomers, that both misrepresent and misconceive the kind of power structure that exists in the United States and in democracies in general. Yet democracies too have a power structure. If Mosca's law means that every political society shapes up, in its vertical structuring, as a pyramid or-if a stratarchy-as a squat figure, then Mosca's law would verge on meaninglessness, for it is a platitude to assert that there will always be rulers and ruled, governors and governed. In order to escape triviality Mosca's law must be assumed to imply something more, to wit, that all political societies are ultimately controlled by one minority- and the question thus hinges on how "one" is to be intended. In Meisel's interpretation the minority in question is required to be self-conscious (of its group interest), cohesive, and characterized by solidarity in pursuing a common course of action. While these conditions could be met, they were not put forward by Mosca for the obvious reason that he well realized that they can hardly be generalized. 42 The generalizeable condition must be, therefore, a lenient condition, namely, that one is "one" as long as it is not a fragmented multiplicity of antagonistic power groups. Empirically put, if one is to be "one;' it must be identifiable and identified. If, conversely, this "one" cannot be concretely identified, the implication is that Mosca's law is falsified. And I would say that Dahl does provide the empirical tools for making sure that Mosca does not win his case by definition. We are left with the question: What is the vertical power structure of democracies? Under Dahl's criteria and (I would add) under all the testable criteria devised thus far, democracies are characterized by diffusion of powerindeed by an extent of diffusion that baffles the "ruling class model." Clearly, therefore, the model that applies to democracies is another model. We shall probe into this different model in due course. At the moment I wish to suggest that it is appropriate to describe it-in contradistinction to the otheras a leadership by minorities model characterized by a multiplicity of criss1 47

VERTICAL DEMOCRACY

6.5

crossing power groups engaged in coalitional maneuverings. To be sure, the fact that the processes of leading and influencing become, in democracies, enormously complex and elusive-as we have seen with respect to the cascade model of opinion formation - should not be taken to imply that power minorities have little power or that they cancel themselves out among themselves. Maybe; maybe not. At certain points in time we may find, as Riesman finds, a "veto group system:' 43 But at other points in time, let alone in space, we may find winning coalitions. 44

6.6

The Iron La,w of Oligarchy

The inapplicability of the ruling class model to democracy vindicates, against Mosca's law, that democracy is not a minority cover-up. But another path of inquiry leads to the outright, drastic conclusion that democracy is "impossible": the path of Michels. Indeed, Michels's law-the iron law of oligarchy- questions the very possibility of democracy. It is true that Michels did not propound a general theory of democracy; he concentrated on the political party. The original title of his most important work was The Sociology of the Political Party in Modern Democracy. 45 Nonetheless, the conclusions that one can draw from his analysis are crucial to democracy per se, for two reasons. The first is that a democratic system is largely, in actual operation, a party system. As Kelsen forcibly put it: "Modern democracy is founded entirely on political parties; the greater the application of the democratic principle the more important the parties:' 46 With the sole major exception, perhaps, of the United States, political parties have become such an essential hinge that a number of authors perceive democracy not simply as a party system but as a "partycracy" (partitocrazia), as a party tyranny in which the actual locus of power is shifted and concentrated from government and parliament to party directorates. 47 The second reason is that the phenomenology of parties has a paradigmatic significance. If the democratic way of life springs from the voluntary creation of small and free communities inter pares-as it doesparties too are formed as voluntary associations and are, in fact, their typical political expression in large-scale democratic polities. From this point of view, then, parties are the type of political organism that most closely resembles, or should resemble, the prototype of every authentic political democracy. Unquestionably Michels put his finger on a strategic point. Moreover, he dealt with the question of organization - and we live in an increasingly organized (even when disorganized) world. No field of human endeavor, nowadays, does not seek organization, or does not seek to enlarge and/or perfect

148

The Iron Law of Oligarchy 6.6 its organization. From all points of view, therefore, we cannot downgrade the importance of Michels's message, which is, in a nutshell, that organization destroys democracy and turns it into oligarchy. In his words: "He who says organization, says tendency to oligarchy.... The machinery of organization ... completely inverts the position of the leader in respect to the masses . . . . Wherever organization is stronger, we observe a smaller degree of applied democracY:' 48 According to Michels, this is an "iron law;' a process that can be neither averted nor stopped. It is inevitable that every party seeks the greatest possible number of members; it is inevitable, therefore, that "opinion parties" gradually turn into "organizational parties;' And since the power of the leader increases as the need for organization grows, all party organization tends to become oligarchical. Michels ended his classic study with the following assertion: "The existence of headship is an inherent phenomenon of all forms of social life. It is not incumbent on science to find out if it be a good or an evil. . . . However, there is great scientific as well as practical value in establishing the fact that every system of leadership/rulership is incompatible with the most essential postulates of democracY:' 49 Many criticisms can be made of Michels's diagnosis and prognosis. In the first place, he speaks of oligarchy and organization without ever clearly defining these concepts. In this connection I have argued elsewhere that Michels may well have an iron law of bureaucracy, but only a "bronze law" (by no means iron clad) of oligarchy. 50 The gist of this line of criticism is that since there are many different types of organizations, we cannot conclude, without qualification, that all are necessarily oligarchies incompatible with democracy. In the second place, Michels's field of observation is too limited, being chiefly restricted to the German Social-Democratic party of his time. In the third place, he is not justified in passing from the premise "parties are not democratic" to the conclusion "democracy is not democratic;' The proof he adduces is too narrow for the breadth of his conclusion. Notwithstanding all this, Michels's law by and large still holds, if only as a "bronze Jaw." 51 For the first objection can be met by observing that the basic argument about organization is a generalization that, vague as it is, does tap a persistent and persisting trend. The second objection can be answered by saying that Michels's case, the pre-1914 German Social-Democratic party, is always relevant to the large mass parties of Europe, which still are, as a rule, hardly more democratic in origin and performance. And the third objection has been handled by pointing out that if we extend the investigation to other organized sectors of political activity-especially trade unions-we are unlikely to find in any large-scale organization more internal democracy than Michels found in political parties. If so, the thesis that "democracy leads to oligarchy" 52 stands. 149

VERTICAL DEMOCRACY

6.6

Let it be added, parenthetically, that if the antidote to Michels's law is "more participation;' then our present-day understanding of the logic of collective action adds weight, if anything, to the Michelsian pessimism. In the argument of Mancur Olson, parties and unions provide, for their members and followers at large, collective (indivisible) goods, that is, benefits that accrue to each member of the group whether or not he is a participator and contributes to their achievement. Therefore, the individual "has no incentive voluntarily to sacrifice his time or money to help an org~nization obtain a collective good; he alone cannot be decisive in determining whether or not this collective good will be obtained, but if it is obtained because of the efforts of others he will inevitably be able to enjoy it in any case:' 53 It appears, then, that the larger an organization, the less it is "rational" for its actual or latent members to share its burdens. If so, the very nature of collective, indivisible benefits goes to justify and motivate apathy. 54 Hence there are more grounds today than at the beginning of the century for fearing that Michels's predictions were well founded. It can be seen that I take Michels's argument seriously. Nonetheless, I consider it exemplary of a fundamental error in premise, that is, of how we may seek democracy without ever finding it. If we agree to judge a democratic state by contrasting its organizational structures with prototypical voluntary associations, on this yardstick it will be hard to prove that Michels is mistaken. But can we proceed from a face-to-face democracy to a nationwide democratic form as if the two things were comparable and belonged to the same continuum? This is the question. Michels conceived democracy a la Rousseau. 55 In his framing of the problem Michels is no different from Proudhon, Marx, or Bakunin; they all refer to the matrix of voluntary associations, and using this as a yardstick, come to the conclusion that the political democracy under which they lived and we live has no organized form that corresponds to that prototype. At this point, the prophecies run counter to one another. Along one path democracy possesses the future, but its advent is postponed to the day that all the organized superstructures that repress it-above all, the state-have been dismantled. Along the other path the superstructures are, if anything, destined to grow, and therefore democracy is forever unrealizable. In the first case, we consider it possible to enlarge to infinity the prototype of voluntary associations and to convert it into that gigantic self-operating collective entity of which Marx and the anarchists dreamed. In the second case, we recognize that in the process of enlargement the prototype is distorted, and so we conclude that large-scale democracy is utopian. However, while the prophecies are at odds their underlying diagnosis (insofar as it impinges on the present)

The Iron Law of Oligarchy

6.6

is the same: Our so-called democracies are apocryphal. The two camps join forces, in their practical impact, in the same negation because both Michels and the perfectionist, the pessimist and the optimist, are looking for democracy with the same lantern. And the trouble lies in the lantern. That the light of the lantern does not illuminate, or illuminate enough, is blatantly revealed by the fact that it sheds no light on the difference between our allegedly false democracies and actual nondemocracies-between living under the rule, say, of Churchill or of Hitler. Neither Marx nor Michels perceive and explain this difference. They are unable to explain it, very simply, because they never grasped how a large-scale democracy is actually produced. They seek democracy in structures, not in interactions. They want to find it immobilized in, inside or within something, instead of seeking it as a dynamic among groups and organizations. To be sure, structures are important. But their critical importance lies, with respect to how a macrodemocracy comes about, in their interplay. If this preliminary point is missed, then we are likely always to land at where democracy is either dead or cannot be, and never to arrive at where democracy is alive and exists. Michels sought democracy inside the large organization. But organization is, after all, a response to a bigness, to something that otherwise gets out of hand. So, we organize or have recourse to organizations not in order to create a democratic form but primarily in order to dispose of a body that is orderly and efficientwhich is an entirely different thing. It follows from this obvious consideration that our problem-understanding how a polity can be democratic- begins exactly at the point where Michels left off. Instead of looking inside organizations, let us look outward, at the interplay between antagonistic and competing organizations-political organizations, to be sure. Why and for whom do they compete? Clearly, they compete for supporters and because their strength comes from the numbers that follow them. How do they compete? Evidently, and overtly, by promising benefits and advantages to their followers. The implication is that the unorganized majority of the politically inactive becomes the arbiter- and eventually the tertium gaudens- in the contest among the organized minorities of the politically active. So, no matter how oligarchic the organization of each minority turns out to be when examined from within, even then the result of the competition between them is, in the aggregate, democracy. This is not, let it be stressed, an optimal state of affairs. Even so, and, if you wish, despite everything, an overall macrodemocracy still results from the sheer fact that the power of deciding between the competitors is in the hands of the demos. This is what not only Michels but the Marxists in general and many of the anti-elitists of the moment fail to see. The one who did see into this more clearly than

VERTICAL DEMOCRACY

6.6

anybody before him was Schumpeter. We thus arrive at the competitive theory of democracy-our next subject.

6. 7 The Competitive Theory of Democracy In the classic theory of democracy- so runs the argument of Schumpeterthe selection of the representative "is made secondary to the primary purpose ... to vest the power of deciding political issues in the electorate"; but the truth of the matter is that the deciding of issues by the electorate is "secondary to the election of the men who are to do the deciding:' On this premise Schumpeter puts forward the classic definition of what is now called the competitive theory of democracy: "The democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people's vote." 56 As anyone can see, Schumpeter's definition is strictly procedural: Democracy is resolved into a method. The noteworthy point is that Schumpeter confines his argument to the input side, or moment, of the overall process of democracy. Hence, it must be asked: How do we move on from the method to its democratic consequences, that is, from democracy in input to democracy in output? The reply is afforded, I suggest, by Friedrich's principle or rule of "anticipated reactions:' 57 In the case at hand the rule can be spelled out as follows: Elected officials seeking reelection (in a competitive setting) are conditioned, in their deciding, by the anticipation (expectation) of how electorates will react to what they decide. The rule of anticipated reactions thus provides the linkage between input and output, between the procedure (as stated by Schumpeter) and its consequences. Let then the completed definition be: Democracy is the by-product of a competitive method of leadership recruitment. This is so because the power to elect also results, in feedback fashion, in the heeding of those elected to the power of their electors. In short, competitive elections produce democracy. And let the above be called-within the ambit of the competitive theory-the feedback theory of democracy. The foregoing will be explicated and implemented as we proceed. For the moment the important point is that it establishes, by actually explaining how democracies work, that large-scale democracy is not an enlargement nor a sheer adding up of many "little democracies." Thus the analogy, or the yardstick, is no longer the small voluntary group; it is, or lies in, a system of feedbacks, of chain reactions. And I submit that this, despite a subsequent plethora of alleged supersessions, is the new theory of democracy or, better, what is new in the mainstream of our theory.

The Competitive Theory of Democracy 6. 7 The competitive or, more fully, the competitive-feedback theory of democracy is often referred to as a "model." Let it be granted that it is such in one of the permissible meanings of the term. If so, the caveat is that a model, at least in the social sciences, is never a key to everything. If the claim has ever been that all of democracy resides in inter-elite competition, then the model in question warrants no such claim. Similarly, the competitive theory cannot and in effect does not imply that the democracy we have is the best one we can have. Nor does the model address the perfectibility of democracy. Clearly, then, the competitive theory of democracy is not the whole of the theory. However, it endows the descriptive theory with a set of necessary and sufficient conditions for a political democracy to exist. It is correct to say that the competitive theory defines democracy "minimally." The fact remains that we now possess a model that explains the hitherto unexplainedand what is otherwise misunderstood. Schumpeter presented his theory as "another theory:' This is right; except that another theory is easily overworked and converted into an alternative theory. Indeed, today the widespread view is that we afford two alternative theories of democracy-the classic one and the competitive one-among which we are free, or even required, to choose. But my argument implies that this is an utterly misleading optic and option. The theory of democracy started out, as it was supposed to do, with a preeminent concern with foundations. When research came to the fore, the empirical focus of the theory was thus very much on the individual voter and on his aggregation into majorities. However, the more we pursued that path, the more it became apparent that the foundations are not the complete building and that the building (vertical democracy) can neither be explained nor represented as a polity monitored by the will of the majority. At base, the objection to the "first" theory of liberal democracy is, then, that it is incomplete, and that its incompleteness leaves us with loose and also very naive ends. In retrospect it can now be seen that the incompleteness of the preSchumpeter theory is (was) rescued by an excess of prescription over description: Whenever the theory is at a loss, it has recourse to an ought. At some of the crucial junctures, the ideals are either displayed in a vacuum of facts or smuggled in as facts themselves. As a result we are left to wonder how democracies can perform at all. This is so because the initial theory fails to grasp the part, the very crucial part, played by the mechanisms of the system, by the fact that its operators are compelled to compete vis-a-vis the consumer market. And here enters, with its additions, the competitive theory. What is then the basis for the notion that we have two contrary and alternative theories of democracy? To the best of my scrutiny, this notion owes 1 53

VERTICAL DEMOCRACY

6.7

its plausibility to an untenable historical reconstruction. To be sure, one can always speak of a "classic theory" of democracy that precedes the presentday one; but the classic theory construed by the anti-elitists is a wholly gratuitous entity. I come to that in the next section. At the moment I simply note that the theory (so to speak) of the "two theories" of democracy is very convenient indeed. If we dispose of two alternative theories among which-we are told-we are entitled to a free pick, then the competitive theory of democracy can be dismissed, without any further ado, with one line: "I choose, between the two, the classic theory:' But there is no such alternative. However, since the anti-elitist attack addresses simultaneously (and, to my mind, confusingly) two targets, namely, the competitive theory of democracy and the polyarchal theory of pluralism, it is necessary, before responding to this attack, to move from Schumpeter to Dahl. While Dahl abides by the competitive theory, his emphasis is very different from Schumpeter's. Dahl takes on where Schumpeter leaves off; that is, Dahl seeks a pluralistic diffusion and reinforcement, throughout the society as a whole, of inter-elite competition. If Schumpeter's problem is to understand the functioning of democracy, Dahl's problem is, in addition, to further democracy. Bearing this difference in mind, let me recast Dahl in my own language, thus alerting the reader to differences in strategy. Dahl's basic strategy is to reserve the word democracy for the "ideal system" and to use "polyarchy" as its real-world approximation. 58 I accept, instead, "democracy" for the real world, but divide its meaning in two halves: the prescriptive (normative) and the descriptive (denotative). 59 Clearly, these are parallel ways of attacking the same problem. Yet strategies make a difference. A major difference is that my own emphasis is, all along, on how ideals affect the real world and, reciprocally, on how the real world receives, but also frustrates and disappoints, ideals. Along this path the focus is, then, on ideals management; and my complaint is that, in the management of ideals, we are still in a bare infancy. All is well (I hope) until I arrive at the final knot: defining democracy along the vertical dimension. Since it behooves me to do so both descriptively and normatively, it is at this juncture that my strategy leads me, it will be seen, into a sensitive area. Descriptively, the "feedback model" can be rendered by saying that democracy is an electoral polyarchy. For Dahl this label would be redundant, for his concept of polyarchy includes, by definition, free and competitive elections (and other properties as well). Yet, when labels succeed, they acquire a life of their own; their semantic inertia largely outweighs the conceptualization of their inventor. Now, semantically, "polyarchy" stands in cont~adistinction to "oligarchy:' Therefore the term polyarchy, in and by itself, conveys only

The Competitive Theory of Democracy

6.7

that an oligarchy is broken up, that it is transformed into a multiple, diffuse, and, at best, open constellation of power groups. That "electoral polyarchy" is not a redundancy is also borne out by the fact that a nonelective polyarchy is perfectly conceivable. For instance, the medieval world could be fitted into a polyarchal description but for one property: openness. However, openness is not, semantically, a necessary property or characteristic of polyarchy. Openness associates with polyarchy because we intend, even when we do not say it, "elective" polyarchy. It is the recurrence of elections that in fact implies openness. The point is, then, that a political system can be polyarchal without being based on popular suffrage. In this case it will still be very different from an oligarchy, and even from an oligopoly, on account of its diffusion: Each power unit will be left with small (in extension) or little (in intensity) power. Therefore, a nonelective polyarchy will meaningfully afford a reciprocal delimitation and control among leaders, at least in the sense that its dispersion defies cartelization. However, this is still a far cry from democracy. Democracies too avail themselves of a reciprocal control among leaders; but after having established first the control of leaders, upon leaders. The crux of the matter is, thus, that to restrain, control, and influence leaders, the demos must have the full and unfettered power to choose them-regular elections must regularly occur. Furthermore, and equally important, it is the notion of election that establishes the association with competition and competitiveness. In and by themselves, neither oligarchy nor polyarchy imply competition in the sense that gives meaning to, and is meaningful for, the competitive theory of democracy. Hence, "electoral polyarchy." To make this shorthand even shorter is, in my view, to say too little. The above is only the descriptive upshot. Also, as I have underscored, it does not address the perfectibility, but only the feasibility, of democracy. Yet, before exploring the maximizations, let us make sure that we do not miss how much our minimal characterization already contains. When we call democracy a polyarchy, we are not simply saying that many leaders take the place of one. If that were all the difference, there would not be much over which to rejoice. In like manner, when talking of elective polyarchy, we are not saying that we are simply allowed to choose among a set of leaders. If that were all, one might again conclude, in a disillusioned vein, that the leaders change but the domination remains. This view, when held, actually attests to the fact that we have neither grasped the feedback model nor, specifically, the central role played, within this model, by the rule of anticipated reactions. I.et it thus be repeated that this is indeed the rule that connects and keeps in tune the voting act with the representative process. If it is true-and it is 1 55

VERTICAL DEMOCRACY

6.7

most of the time-that the leader subject to periodical electoral removability is concerned with how the voters will react to his actions, it follows that he will be monitored (at least, when sensitive issues come up) by the anticipation of what that reaction, whether positive or negative, might be. With the addition of Friedrich's principle, then, the model becomes self-contained. My earlier descriptive definition was: Democracy is the by-product of a competitive method of leadership recruitment. It may now be explicated in full as follows: Large-scale democracy is a procedure arid/or a mechanism that (a) generates an open polyarchy whose competition on the electoral market (b) attributes power to the people and (c) specifically enforces the responsiveness of the leaders to the led. 60 We are ready, at this point, for the next problem: the furthering in democratization of democracy- and precisely its furthering beyond that amount of real (not fictitious) democracy with which the competitive machinery leaves us. The democratization of polyarchy, as Dahl would have it, occurs jointly, though not simultaneously, along two directions: (a) liberalization and/or public contestation; and (b) inclusiveness and/or participation. 61 Dahl conceives this mapping and its categories as the "directions" along which any and all regimes change and are eventually transformed into one another. If the problem is narrowed-as it is here-to polyarchies, it is my reading that "participation" applies better than "inclusiveness" 62 and, correlatively, that "liberalization" applies better than "contestation?' 63 However that may be, the sketch suffices to indicate the directions along which Dahl confronts, within his theory of polyarchy, the agenda that Schumpeter did not confront. Before proceeding further, let the following two points be underlined: First, what makes democracy possible should not be mixed up with what makes democracy more democratic; second, unless the two problems are treated in this order, the oxen may well wreck the cart rather than pull it. Let us now come to the anti-elitist critique of so-called elitism.

6.8

Anti-Elitism Revisited

We have come across the notions of anti-elitist and elitist a number of times. It is high time to inspect them in that order, since it is "anti-elitism" that currently defines "elitism?' The anti-elitist persuasion can be said to display two distinguishable faces, two sides. 64 One is plainly polemical and essentially consists of establishing "elitism" as an epithet, as a derogatory term. The other side, the positive one, consists of extolling a participatory theory of democracy in opposition to the theory of representative democracy and, in particular, to the competitive theory.

Anti-Elitism Revisited 6.8

In order to sustain the disparagement, Schumpeter and the SchumpeterDahl line of authors are made to appear as a continuation of the Mosca-Pareto tradition of thought. 65 The innuendo is that since Mosca and Pareto were wicked, 66 so also are their descendants (among whom I have the privilege of being included). But there is no truth to this genealogical reconstruction. It cannot be held that the competitive theory of democracy is "elitist" because it links, in spirit or otherwise, with Mosca or Pareto. The point is not that Schumpeter never cited Mosca and mentioned Pareto only as an economist (with one peripheral exception). The point is that any reading-let alone a fair reading-of Schumpeter shows that the sources of his theory lie simply in his dissatisfaction with the theory of democracy (not of anti-democracy) as it stood in his time. Let it be added that this mistaken genealogical reconstruction, serviceable as it may be to debunking, seriously muddles the concept of elitist. For the sense in which Mosca is declared elitist has nothing in common with the sense in which Dahl's theory of polyarchy is called elitist. 67 It is, however, when we come to the positive statement of anti-elitism the theory of participatory democracy-that genealogical reconstructions become an even more crucial part of the argument. This is so because the antielitists wish to convey the message that they restore the "classic theory" of democracy, and this against the "revisionism" of the elitists- a revisionism inaugurated, according to Karie!, by none other than Tocqueville. 68 Here, then, we encounter the theory of the "two theories" of democracy on which the anti-elitist centrally builds his case. It is important, therefore, to clear up this matter. When reference is made to a theory of democracy, the understanding generally is that the theory in question is pro-democratic, not anti-democratic. However, under this understanding there is no Greek "classical" theory of the democracy of the ancients, for Plato was strongly critical of Greek democracy and Aristotle classified the specimen among the degenerate forms of the polity. 69 We have to come, along the millennia, all the way up to Rousseau to find the first idealization of the democracy of the ancients. But not quite, for Rousseau neither gave prominence to the term 70 nor conceived democracy as either his predecessors or the moderns have. To make the point bluntly, by anti-elitist standards Rousseau was a thoroughgoing elitist. He disliked the Athenians, praised the dubiously democratic Romans and, worse still, the Spartans. He equally praised the Republic of Venice (surely a full-fledged aristocratic oligarchy) and looked at Geneva as the real-world possible embodiment of his ideal city. 71 Now, the Geneva observed by Rousseau was an estate-based aristocratic republic, with office reserved only to the first order 1 57

VERTICAL DEMOCRACY

6.8

(the "citizens") out of five, and with the last three estates (by all counts more than 90 percent of the population) excluded not only from access to office but also from the right to vote. If one considers, in addition, that Rousseau rendered the good life in terms of a purely static polity, 72 one can only be amazed at the anti-elitist exaltation of Rousseau. 73 The lethal flaw of the participationist construct is that it fails to notice that the democracy of the ancients is not the democracy of the moderns. The former was a city-confined, direct, and individual-disregarding (illiberal) democracy; the latter is a nation-wide, representative, and individual-regarding (liberal) democracy; and two referents that are as diverse as these call for two different theories. 74 Retrospectively, it can always be said that participation characterized the democracy of the ancients; but, if so, it cannot be left unsaid that the theory of that democracy was critical of participation. For whenever the classics (up until Tocqueville and including Rousseau) made the inference from popular power to an activist kind of participation, they were unanimous in pointing to this element as the ruinous and self-destructive element of democracy. Conversely, when the referent becomes (from Tocqueville onward) liberal democracy, then participation no longer centrally characterizes either the reality or the theory of representative democracy. So, where is the "classical" participatory theory of democracy that the anti-elitists claim to represent and revive? Most anti-elitists seemingly take it as a self-evident truth that an entity called "the classics" is on their side. In the rare cases in which this contention is sustained with names and argument, the best they can do is to bring together Rousseau, James Mill, John Stuart Mill, and G.D.H. Cole as founding fathers of the theory of participatory democracy. 75 This is, indeed, a short list of classics; and nothing much is lost if we make it even shorter by setting aside (as minor, secondary classics) James Mill and Cole. So, is there anything that Rousseau and John Stuart Mill can father or generate in common? I think not. Rousseau is the last spokesman for a sui generis idealization of the democracy of the ancients; John Stuart Mill is a major spokesman for liberal and representative democracy. Furthermore, it is not only very hard to construe Rousseau as a participationist of the contemporary variety; it is even harder, and it requires heroic omissions, to build up John Stuart Mill into a father figure of participatory democracy. Mill did write that "the only government which can fully satisfy all the exigencies . . . is one in which the whole people participate"; 76 but the conditions that surrounded and qualified this sentence are, by anti-elitist standards, outrageously elitist. Not only did Mill advocate a plural vote for the more qualified and educated on the principle that "it is not useful, but hurtful, that the constitution ... should declare

Anti-Elitism Revisited 6.8 ignorance to be entitled to as much political power as knowledge"; 77 he also required a literacy test and held that paupers who did not pay taxes should be excluded from voting. 78 The point is, thus, that far above and beyond being (or not being) a founder of the participatory theory of democracy, Mill's central worry was "the danger of a low grade of intelligence in the representative body and in the popular opinion which controls it;' that the constitution might not "secure an adequate amount of intelligence and knowledge in the representative assembly?' 79 All told, the pitting of a participatory theory against the competitive theory of democracy is not sustained, and cannot rest, on counterpoising a virtuous against a wicked genealogy of authors. The virtuous assemblage (Rousseau with Mill plus minor adjuncts) adds up to being a non-group that even cursory scrutiny rends in tatters. Likewise, the wicked lineage (from Mosca to Dahl) is by no means a lineage; it is only a straw target. But such genealogies- both of alleged supporters and enemies- as are made by the antielitists may dissolve, and yet our authors may still have something to say of their own. That is, a theory may be founded anew-without the help or support of founding fathers. We are thus prompted to inquire about the kind of theory that the anti-elitists have in mind. Is it, for instance, a purely normative and idealistic theory, that is, a theory solely concerned with ideals? Yes, but not altogether; and this on three grounds. First, if we are set upon a purely normative theory, then it should be clear to us that such a theory provides, and is only meant to provide, a yardstick-whereas anti-elitism also and eminently portrays itself as a program. Second, a revival of ideals should produce some refinement, or at least a somewhat precise elaboration of the ideals being cherished-and this is by no means the case. Third, even though participation may be construed as an ideal, it still is an intermediary ideal, hardly an end in itself. Perhaps some of our authors feel that people should participate just for the sake of participating. Even so, the overriding ideals of liberal democracy are and remain liberty and equality. Now, on political liberty as such (in its difference from other dimensions of freedom), the anti-elitists are remarkably lukewarm. 80 They brandish, instead, the ideal of equality-but in a purely routine fashion. Equality has been, of late, the subject of an enriching and stimulating scrutiny;81 but not from anti-elitist quarters. Since the anti-elitists do not seem interested in, or articulate on, the distinction between idealistic and realistic theories, 82 let us simply ask a most simple question: Which concept of participation sustains their theory? But this, it turns out, is not a simple question at all. The remarkable feat of our authors is, in effect, the care with which they avoid defining their central con159

VERTICAL DEMOCRACY

6.8

cept. The reason for this is not far to seek. It will be recalled that I have defined participation as a taking part in person, whose authenticity and efficacy is expressed by a ratio: r to ro, r to 100, and all the way to r to billions; 83 and since the participatory theory does stress that participation must be "real;' my definition would appear to be very much to the point. Not so. It is not so because the participatory theory does not wish to confron.t the size problem. While much of it deals, in substance, with industrial democracy and other kinds of microdemocracy, 84 in principle the claim of th'e participationist is universal and extends, therefore, to macrodemocracy. But if participation in a group of, say, ten is the same as participation in a collectivity of, say, ten million, then the price exacted for this ubiquity is that "particip«tion" is a word without conceptual content-strong in sound and fury but evanescent in meaning. It is appropriate to note at this point that one of the major charges leveled against elite theories is that they either equivocate between a broad and narrow definition of politics, or that their definition of politics is too narrow. There is some truth in the first indictment, as in the point made by Bottomore that in Mosca and Raymond Aron the argument oscillates somewhat ambiguously from "a concept of plurality of elites to the quite different concept of a multiplicity of voluntary associations:' 85 But this oscillation does not entail that the alleged elitists downgrade the vital importance of voluntary groups, of Tocqueville's intermediary structures, in short, of societal pluralism. And the further point is that when it comes to drawing the boundaries of politics, the attackers-the anti-elitists-equivocate far more, and far more grievously, than those under attack. To wit, Bachrach's case rests, in the main, on rejecting the narrowing of democracy to a "political method" and therefore on enlarging "the political scope to include the more powerful private institutions" and, more generally, generalized participation. 86 Now the boundaries of politics can indeed be drawn at very different points; but there is nothing inherently wrong in conceiving politics either narrowly or diffusely. The mistake, when there is one, lies in adopting borderlines unsuited to the problems under investigation. Therefore there is absolutely no mistake (pace Bachrach) in defining politics narrowly whenever the object of concern is the polity and, ultimately, the legal monopoly of force. This is as it should be. Since the competitive theory of democracy centrally addresses the master system, the democratic state, it correctly refuses to confuse (as the anti-elitists do all along) the master system with its subsystems, the macro with the micro. It is not that the alleged elitists deny that political democracy needs the implementation of participation, social democracy, or even industrial democracy. The difference is that the so-called 160

Anti-Elitism Revisited 6.8 elitist realizes what the anti-elitist stubbornly refuses to see, namely, that if the enforcer of force, the state, is not a democracy, then all the infrastructures, all the microdemocracies cherished by the latter, are in mortal danger. Bachrach is perfectly entitled to adopt a diffuse definition of politics. By so doing, however, he fails to address what is crucial to the vertical dimension of politics. If there is a great deal of confusion as to what politics is, and as to where it ends, those who are by far the most confused in this matter are the anti-elitists. To return to the main thread, the theory of participatory democracy rests on "participation" as a word that is hardly amenable to conceptual underpinning. What the anti-elitist means by participation is nothing more than that the people should rule. Since this is deja vue, and since he does not even attempt to show how the people can be made to rule, he certainly has no ground for assailing those who do. Not only does anti-elitism have little, if anything, to display in the positive; it also has little to display in the negative. We thus come to a last and most intriguing question, namely, by what criterion is an elitist declared to be such? Who is an elitist? Once again, I am afraid, I am at a loss. If the criterion for sorting out non-elitists from elitists resides in whether or not an author accepts as a fact that relations of supraordination and, conversely, subordination (i.e., power structures) unavoidably exist in any society, then by this token most anti-elitists would also be elitists. If the criterion resides in the intention, and precisely in whether the theory of elites is brandished to challenge the reality and the very possibility of democracy, then Pareto, Mosca, and Michels were elitists, whereas the Schumpeter-Dahl line of authors is anti-elitist, since it is characterized by the contrary intent. Is the criterion, then, participation? In this case participation must be construed (however defined) as a watershed; the divide is not how much participation we can have (an empirical question), but whether participatory democracy can replace, at the polity level, representative democracy- and by this yardstick most anti-elitists would, once again, be elitists. We are seemingly left, then, with how an author handles the "masses" (in the elite-mass relationship). To illustrate, an elitist is supposed to see "the chief function of the elite as holding the masses back, to restrain them from the temptations ... of perfectionism and the pitfalls of demagoguery"; to assume that the masses are "degenerate"; that the "purpose of election is not to enhance democracy" and that "the corrective does not rest in educating the electorate in an attempt to inculcate higher standards conducive to the selection of better qualified leaders?' 87 But who is, currently, such a wicked elitist? 88 Since I cannot find any, I have run out of cases. 161

VERTICAL DEMOCRACY

6.8

If we do not know who is an elitist or why he is such, by the same token we equally do not know who is an anti-elitist-aside from one unmistakable characteristic, namely, his aptness to employ "elitism" as an object of disparagement. And this leaves us, at best, with an embarras de pauvrete- as my concluding reflections will underscore. It should be clear, at this stage of the argument, that the competitivepolyarchal theory of democracy is, in the main, a descriptive theory that actually explains how democracies work and perform. The next step in the argument is to point out that a descriptive theory is not enough, that we also need a prescriptive theory of democracy. Strange as it may seem, if this is the complaint of the anti-elitist, it is also my complaint. From here onward, our respective paths certainly diverge. However, since I have no monopoly on how the normative implementation of the theory of democracy is to be pursued, my objection to anti-elitism is by no means an objection in principle -concerning the principle of a normative need- but an objection in method, or of method. First, I do not understand why the normative theory should reject the descriptive one. In what way are we better off without a theory of democracy that explains its functioning? Second, it is even more obscure how the ills of real-world democracy denounced by the anti-elitists relate to, and would be corrected by, their participationist therapy. Not only is this essential connection nowhere explained, but the anti-elitists' way of handling their case is, by all standards, as poor as can be. Let us leave aside the systematic, wholesale misreading of the authors that happen to be on the anti-elitist reading list of foes. The distortion goes far deeper than that because it denatures the very notion and existence of evidence. I, for one, am by no means a believer in the sanctity of evidence, nor do I hold that ideals should submit to it; I hold only that if we want our ideals to be constructive, they must relate to factual verification. But the antielitist method (so to speak) is simply to transform the evidence into a sort of Cheshire cat. The real world is profusely cited, but only when it suggests negation, never when it permits testing. Indeed, any disconfirming evidence is either simply erased or becomes a fabrication by elitists. For instance, the fact that participation is low is transformed into an "elitist argument"; the fact that the average citizen is poorly informed is called a "presumption"; and so forth. 89 Thus the essence of the anti-elitist strategy is to surmount all difficulties by simply ignoring them. In dealing, in the previous chapter, with referendum democracy, we have encountered the finding that participation correlates with intensity and that intensity correlates, in turn, with extremism. If this be true, it would imply that the participatory ideal may have nothing to do with the enrichment and the full development of human capacities. We 162

Polyarchy Defined Normatively

6.9

should thus expect from the participationist counter-findings, or at least counter-arguments, on "intensity"; but we only encounter, on the matter, his untarnished silence. It has also been seen that generations of scholars have labored on Michels's iron law of oligarchy and how it could be defeated (e.g., by using Schumpeter as an antidote). The solution of the anti-elitist is not to respond to Michels at all. To make things worse, the expansion of public goods sustains, if anything, the expectation that incentives for mass participation are decreasing. Surely this is a problem that the participationist should confront. But not a word. Finally, we would want to know from the participationist how his recipe meets the need for competence, rationality and, ultimately, knowledge. To all of the above Kariel replies, in a fairly representative summation, that "the idealized conception of democracy assumes that all men are basically disposed to participate ... that they are capable of weighing the consequences of their actions . . . that the economic resources exist to enable men to become rational, responsible citizens?' 90 Now, I have stressed all along that without an "idealized conception of democracy" there would be no real-world democracy-but the above really is too easy. If we can get away with disposing of problems in such manner, then we can get away with anything.

6.9

Polyarchy Defined Normatively

It is time to come to the prescriptive definition of democracy as a system of government. While this problem is seldom confronted directly, it underlies our appraisal of leadership. Are elites and leading minorities a necessary (or even unnecessary) evil, or are they a vital and beneficial asset? Ultimately, the alternative is: Should we downgrade or upgrade leadership? The line of authors who speak in favor of the latter view is impressive, both over time and in eminence. For the ancients, it is Thucydides who reminds us that the greatness of Athens reached its height with Pericles precisely because "by his rank, ability, and known integrity he was enabled to exercise an independent control over the multitude?' 91 After our having begun anew, more than two millennia later, Bryce reviewed the experience of his time in this concise sentence: "Perhaps no form of government needs great leaders so much as democracy does?' 92 Some fifty years later, in 1937, after the downfall of democracy in Italy, Germany, and Spain, de Madariaga wrote: "Despite appearances, liberal democracies are dependent on leadership even more so perhaps than other, more authoritarian forms of government; for ... their natural tendency to weaken the springs of political authority must be counterbalanced by a higher level of ... authority on the part of their

VERTICAL DEMOCRACY

6.9

leaders:' 93 In the same years Karl Mannheim reached the same conclusion: "The lack of leadership in the late liberal mass society can be ... diagnosed as the result of the change for the worse in selecting the elite .... It is this general lack of direction that gives the opportunity to groups with dictatorial ambitions:' 94 As World War n was approaching its end, in a classic text of the r94os Lindsay's reflection was: "If democracy is to survive it will have to employ and use every bit of skill and knowledge and leadership it can get hold of. This complicated interdependent world in which we are living cannot be run without knowledge and skill, foresight and leadership. Any cult of incompetence can only lead to disaster:' 95 The theme recurs in V.O. Key, the author who best defends the wisdom of electorates: "The critical element for the health of a democratic order consists in the beliefs, standards and competence of those who constitute the influentials, the opinion leaders, the political activists in the order.... If a democracy tends toward indecision, decay, and disaster, the responsibility rests here ... :' 96 And Daniel Bell puts it thus: ''Any estimate of the ability of a society to meet its problems depends . . . on the quality of its leadership and the character of the people:' 97 All these authors (and one could continue citing at length) command respect. It can hardly be doubted that their words sincerely reflect the lesson of experience and indeed of a long, thoughtful, and often painful process of learning. Yet, if one looks into the theory of democracy in general, it can hardly be said that the message conveyed by the above quotations has been incorporated in terms of theoretical status; the praise of leadership remains a sideline of little theoretical consequence. And this quite apart from the anti-elitist attack. I have long dwelt throughout the preceding chapters on the interactions between the inertia of facts and the tension of ideals, and my keynote has been that democracy needs both realism (awareness of the facts) and idealism (value pressure upon the facts). Accordingly, and on behalf of consistency, my question now is: How does the vertical problem of democracy fare on these tests? The answer bluntly is that it does not fare at all. The plain fact is that the ideals of democracy have remained very much what they were in the fourth century B.c. 98 And if the ideals of democracy are still, in the main, its Greek ideals, this means that they address a direct, not a representative, democracy. This implies that even today the deontology and the value pressure of democracy speak only to the horizontal dimension of politics. To be sure, even the Greek polis had magistrates and some minimal verticality. Yet the vertical dimension of an ancient town democracy compares to the vertical dimension of a nation-wide representative democracy as the campanile of Venice compares to Mount Everest. Thus the astounding fact is that we have created a representative democracy-performing a near-

Polyarchy Defined Normatively

6.9

miracle that Rousseau still declared impossible-without value support. And not only has the vertical construction of large-scale democracy not been accompanied by a consonant ideal, but the ideals at hand can be converted overnight-as we rediscovered in the 196os-into a battle cry against representative democracy. Saying the least, the vertical dimension of democracy remains to this day ideal-less; and the predicament is that it easily finds, in our ideals, hostile ideals. It is evident that direct democracy (whether past or present) has no need or place for a vertical value pressure. But it should be equally evident that we have long gone beyond, and irreversibly so, the Greek formula. No matter how much we may succeed in regenerating small democracies of the direct variety, the fact remains that face-to-face democracies can only be parts of larger units and, ultimately, microparts of an overall unit that is always an indirect democracy hinged on vertical processes. If this be the case, should we abandon these processes to their inertial, "natural" wickedness? Correlatively, can our future be handled simply by revitalizing the ideals of the past, that is, ideals that are alien to the problems of representative democracy? This is what the anti-elitists and, more generally, the New Left seemingly believe, for their message and medicine consist of a pure and simple return to, and expansion of, horizontal politics. It seems to me, instead, that while this message leaves the altimetric construction unaffected-exactly as altimetric as it is-it can only render it more and more dysfunctional. Let it be stressed, in the attempt, however vain, of not being misunderstood, that if there is- as there is- a literature on democracy that is too complacent, that conveys that we are doing as well as human imperfection permits, this is a view that I do not share. Were I not unhappy with how our democracies perform, I would stay content with democracy defined descriptively, to wit, as a diffuse, open system of controlling groups in electoral competition with one another-thus sparing myself the trouble of pushing forward the normative concern to where I am now pushing it. Descriptively, I have said, democracy is an elective polyarchy. But what ought it to be? Since polyarchy is a state of fact, what is its corresponding deontology, its corresponding state of norms? At base, the issue is not only whether in the long run representative democracy can perform and, one hopes, improve its performance without a value pressure of its own; the issue is, even more pressingly, how it can go on performing in the face of a value pressure that ever more devalues the vertical dimension. This devaluation is patently attested by the current state of our vocabulary. The set of words that characteristically addresses the vertical dimension is "election;' "elite" and "selection;' These terms were all conceived as evalua165

VERTICAL DEMOCRACY

6.9

tive screens. Election connoted for some fifteen centuries a qualitative choosing- as in "the elect;' those called by God, of the language of Protestantism. Elite derives from the same root and was coined-when aristocracy lost its original meaning and simply came to denote an estate - precisely to connote "the best;' the aristoi, the choice part (this being, as we know, the meaning in which Pareto picked up the term). Selection originates, instead, from seligere, but gradually rejoins eligere (when election specifically denotes the act of voting) in conveying an identical meaning: choosing on account of excellence or fitness. In the present-day language of politics all these connotations are either lost or under attack. Election is reduced to one meaning only: the mere act of voting. Selection means little more than a mere preference of the will-when it is not distorted and debased as "discrimination." Consequently, in our language "the elected" are simply the persons voted into office; and to say that the elected should be "selected" strikes us more as a redundancy than as an axiological specification. Finally, elite is first transformed by Lasswell into a neutral word and, with the anti-elitists of the day, into a boo word. In either case, the term elite is associated, in total oblivion of its reason for being, with the powerful and/or the privileged. The foregoing transformations can be seen as part and parcel of the growing ideological distortion of language that Orwell so well detected and illustrated. 99 But while this is true for the recent excesses, the transformations in question reflect, in the main, the horizontal value-vision of politics. In either case we are left with a vicious circle that is being accelerated into a vicious vortex. Without value connotations we obtain a value void; and when a good word is inverted into a bad word, we are left with a vocabulary that can only express bias. At the close, we end up with a reality that remains, for lack of usable words, axiologically unheeded. This vicious vortex cannot be escaped unless the axiological element is brought again to the fore. In turn, this can be done only by restoring the language that expressed it. Let me start with "selection." Here the cause is far from being lost, since it is only in politics-very revealingly-that the term has been neutralized, when not distorted.100 Many who use election and selection interchangeably automatically switch, in all the nonpolitical fields, to the evaluative meaning of the term. A scientific enterprise is required, in order to be scientific, to "select" its personnel. In academia, the "selection" of a candidate for a position is assumed to mean that the chosen is the best. When a firm recruits, it either "selects" or it is likely to go out of business. Is it the case that the politics of democracy is so simple, and so intrinsically different from all other societal processes, that here a selection is either superfluous or even sinful? If this is not the case, then let my first axiological definition be: Democracy 166

Polyarchy Defined Normatively

6.9

should be a selective system of competing elected minorities. Let it equally be, still more briefly and in symmetry with the descriptive definition: Democracy should be a selective polyarchy. If one pauses to think, "selective polyarchy" is in itself a telling, highly meaningful expression. Yet its meaningfulness cannot be expected to get easily across. Swimming against the tide of an impoverished language calls for a sustained effort. So I now choose to attack the problem from another anglethe angle of equality. Clearly, our way of handling the vertical problem of democracy crucially hinges on our way of handling the concept of equality. This is less clearly the case if one argues that while equality is the central value of horizontal democracy, by the same token it neither is nor can be the central value of vertical democracy (whose central value is, instead, liberty). Even so, lwish to ask whether and how the concept of equality can be entered into the vertical dimension. Montesquieu was repeating the lessons of Plato and Aristotle when he wrote that "the principle of democracy is corrupted not only when the spirit of equality is lost, but also when the spirit of extreme equality is assumed, and everyone wants to be equal to those whom he chooses to govern him?' Thus, Montesquieu added, "democracy must avoid two excesses: the spirit of inequality, which leads ... to the government of one person; and the spirit of extreme equality, which leads to the despotism of one person?'101 Rousseau, although his antagonist, spoke to a parallel point when he asserted that "it is manifestly against the laws of nature, no matter what way we define them, that ... an imbecile lead a wise man?'102 Now, and regardless of how we define the imbecile and the wise man, the fact is that we incessantly appraise the next man as not being "worth m:ich" or, conversely, as being "first rate:' The criteria are irrelevant to the point (I need not spell out the criteria of beauty in order to sustain the assertion of fact that people are so rank-ordered). Since human beings come in all shapes and kinds, each individual may well appraise the other with a yardstick of his own. Nonetheless, we do judge and evaluate others all the time as having superior, equal, or inferior qualities. Does the foregoing imply that we have entered a realm from which equality is ousted? No, not necessarily. Rather, I would put it thus: We are now confronted with whether to equalize upward or downward. When it comes to this option, the anti-elitists actually press, even if unwittingly, the accelerator of a downward leveling, since their case explicitly rests on valuing only a horizontal conception of democracy. But is it really the case that the "elitists" sustain in some adequate way the contrary option, the equalizing upward? This is an intriguing question indeed. Since the socalled elitists are not a group that hangs in any way together, and since we

VERTICAL DEMOCRACY

6.9

do not even know by what criterion an elitist is such, we are left with only one thread to follow, namely, who uses the term elite in what meaning(s). I have already indicated, in this connection, that Lasswell altered the meaning of Pareto by transforming elite into a purely altimetric concept defined as those with most power, the top power class, or "the power holders of a body politic;'103 and that this is a semantic alteration of consequence. In all likelihood Lasswell reduced elite to mean "those in power" in order to conform to Wertfreiheit, to the freedom-from-value concern; even so, by giving "elite" a neutral connotation Lasswell went to unnecessary trouble and, by so doing, generated trouble. He went to unnecessary trouble because he could have picked up a neutral term already in existence. Mosca's "political class" is perfectly value free. Power group, controlling minority, power holders, equally are value free terms. And he created trouble-at least for the theory of politics- because his redefinition destroyed the only surviving _value term in the set. If elites, specifically political elites, are defined on sheer power or altimetric grounds, we are precluded by this very definition from looking into the discrepancy between elite qualities (and standards), on the one hand, and power positions (unduly assimilated to elite positions), on the other. As a consequence, elite research misses what is fundamentally at stake-which is not that the powerful exist, and not only whether power elite(s) is a plural, but ultimately whether the powerful represent authentic or apocryphal elites.104 Thus the inevitable net implication of the Lasswellian approach is either gratuitously to impute elite value to whatever power structure happens to exist or to devalue whatever may be of value in such a power structure - or both things in unholy combination. From this we can arrive at the sheer sanctification of the status quo or, conversely, at its wholesale desecration. In the first case, the anti-elitist attack finds here its best justification; in the second, its natural ancestor. In either case, we sin by confusing/act with legitimacy and, in principle, by having drained the value content out of a value issue. Reverting to the option between equalizing upward or downward, the argument was left at the consideration that whereas the anti-elitist message can only encourage (whatever the intentions) a downward leveling, the intriguing question was whether the contrary option of equalizing upward has in fact been a concern of the alleged elitists. The question has in part been answered: The Lasswellian school, given its concept of elite, displays no such concern. And the remaining part of the answer is the same, since in the last decades the emphasis of the theory of democracy has largely been, from all quarters, on horizontal democracy; and the more we conceive democracy along the horizontal dimension alone, the more (to paraphrase Marcuse) we have a

168

Polyarchy Defined Normatively

6.9

unidimensional democracy that corresponds to a highly impoverished unidimensional equality. To be sure, authors do move on from "equality of power" conceived as a horizontal equality (the equal power of the demos) to "equality of opportunity;' that is, to an equality that assumes vertical processes. Yet equal opportunities point to a beginning, to a start, more than to an end state. If one is concerned with the value implication of each kind of equality, then it would seem that equality of opportunity justifies a moving up, an emerging, but need not attribute value to this emerging. It would seem, therefore, that the equality that centrally qualifies the vertical processes and processing of democracy is "equality of merit;' that is, Aristotle's proportionate equality. In order to conceive equality as an upgrading value the maxim that applies is: The same to sames, that is, to each in proportion to his desert, ability, or talent.105 The earlier normative definition was this: Democracy should be a selective polyarchy. It can now be implemented by the following definition: Democracy should be a polyarchy of merit. The argument that equalizing unequal talents is not a just but an unjust equality may leave us cold. Even so, the argument difficult to assail is that equality in merit (in proportion to capacity) benefits the society as a whole, whereas equality in demerit (equal rank to unequals) is a disserving equality, a collectively harmful equality. In Rawlsian terms, "social and economic inequalities are to be arranged so that they are both (a) reasonably expected to be to everyone's advantage, and (b) attached to positions and office open to all:'106 Now, if things were so arranged, they would fittingly describe a merit-based polyarchy. One objection, among others, will be that the above soars at an overly abstract level. In order to bring some concreteness into the argument the question that especially needs to be asked is equal with respect to whom? And in order to reply I propose to call on the notion of reference group and, more precisely, on elite (in the original meaning of the term) construed as a reference group. The connection is that in conveying the idea of "worthy of choice" the term elite points to a reference group-and precisely to a value reference group.107 The question, Equal with respect to whom? can thus be answered: with respect to elite value parameters. The implication is that the concrete elites by no means consist of, and coincide with, those who happen to be powerful (the actual political class). To the contrary, in the reference-group perspective concrete groups are under constant scrutiny; they provide "reference" on the basis of their virtues if, and only if, they have virtues. We may summarize the point thus: Equality concretely elic;its an upgrading, a value lift, when linked with "elite'~ provided that the term is construed as a reference group and entered into a reference theory of elites.

VERTICAL DEMOCRACY

6.9

Since, in the course of this chapter, we have come a long way, let me recall the thread of the argument. When we confront democracy as a system of government, we have to confront the problem of controlling groups and of leadership. One way of doing this is to acknowledge that power is unevenly distributed, that power groups exist, and that they will in all likelihood continue to exist. This may be called the realistic stance; and here my objection is not that any of this is empirically false but that it leaves matters exactly as they are. The opposite way of confronting the problem is exemplified by the anti-elitist stance; and here my objection is that the short-run polemical gains of the anti-elitist message are largely outweighed by its damage in depth. Let there be no mistake on the fact that by disvaluing meritocracy, we simply obtain immeritocracy; that by disvaluing selection, we simply obtain disselection; and that by disvaluing equality in merit, we simply obtain equality in demerit. As the argument is placed in perspective, the timely question is: Where do the present and imminent dangers lie for democracy as a political form? In some kind of "minority rule"? I think not. For the plain fact is that democratic governments are all- some more, some less - in loss of authority and clogged by too many demands that they are unable to process. Note that overload is not big government. While big government may be said to facilitate overload, certainly one can exist without the other. So we live in a trafficjam, cross-pressured democracy characterized by low governing capability, to wit, by low resistance to demands and low ability to take and carry through decisions. More often than not, the pattern of the 1960s and 1970s has been one of indecision, shortsightedness, inefficiency, and overspending.108 Not all of this is displeasing. Actually it forcefully attests - against the contrary claims of perfectionists, participationists, and populists-that representative democracy is by no means a sham, a polity in which the people are deprived of their power. For all of this confirms the extent to which the representative linkage has maximized responsiveness. However, responsiveness is but one of the elements of representative government. A government that simply yields to demands, that simply gives in, turns out to be a highly irresponsible government, a government that does not live up to its responsibilities. A representative is not only responsible to, but also responsible for. This is the same as saying that representation intrinsically consists of two ingredients: responsiveness and independent responsibility.109 And the more governments become responsive to at the detriment of being responsible for, the more we are likely to be misgoverned and/or ungoverned. Which is equally to say that the more we have indulged in responsiveness, the greater the need for independent responsibility-what leadership is really about.

Notes We thus return to the question from which we started, namely, whether leadership is or is not an integral element of democracy. The old but now rejuvenated view is that leadership is needed only to the extent that the role of the people remains secondary. This view easily gets the applause. Yet if its propounders really believed in it, why not have leaders replaced by "administrators" appointed by lot? Waiting for this alternative to be put to the test, let me bring my own brief to a close. With democracy defined as an elective polyarchy we do not address the "good" functioning of the system, for electoral competition does not assure the quality of the results but only their democratic character. The rest-the worth of the output-depends on the quality (not only the responsiveness) of leadership. But while the vital role of leadership is frequently acknowledged, nonetheless it obtains only a negligible status within the theory of democracy. My search for a vertical normative definition attacks this problem. I have proposed to this effect a reference theory of elites and two shorthand definitions conceived as mutually reinforcing, namely, that democracy ought to be (a) a selective polyarchy, and (b) a polyarchy of merit. As John Stuart Mill put it, "when we desire to have a good school, we do not eliminate the teacher:' 110 Whatever the worth of these proposals, the point remains that the theory of democracy in general has failed to pursue the projection of the value of equality into the vertical dimension. If a headless society were at all possible, we could indeed rejoice; on this score we have lately been doing very well. But if headlessness or leaderlessness is no solution, then our current downgrading or fear of elites is an anachronism that blinds us to the problems and perils that confront us. The more we lose sight of democracy as a system of government, the more our predicaments are aggravated-and the more they are with us.

Notes I. For a historical analysis of the concept of politics and the underpinning of the distinction between its vertical and horizontal dimensions, see Sartori, "What Is Politics;' Political Theory, February 1973, pp. 5-26. 2. My point, here, is basically semantic. It does sustain, however, a definition of leadership as "leaders inducing followers to act for certain goals that represent the values and the motivations ... of both leaders and followers" (J. MacGregor Burns, Leadership [New York: Harper & Row, 1978], p. 19). While Burns's connotation may be excessively democratic specific, I certainly subscribe to the implication that dictators should neither be called nor considered leaders. J.R. Pennock, "Democracy and Leadership;' in W.N. Chambers and R.H. Salisbury, eds., Democracy Today: Problems and Prospects (New York: Collier Books, 1962), pp. 122-58, is an insightful discussion of the many facets of leadership. See also Robert C. Tucker, Politics as Leader-

I7l

VERTICAL DEMOCRACY

ship (Columbia: University of Missouri Press, r98r). However, none of the aforesaid authors neatly distinguish, as I suggest, between democratic leadership and dictatorial rulership. 3. In J. Elliot, ed., Debates on the Adoption of the Federal Constitution (Philadelphia: Lippincott, r94r), 5:203. 4. The centrality of opposition for democracies is forcefully highlighted by R.A. Dahl, esp. in Political Oppositions in Western Democracies (New Haven: Yale University Press, 1966). See also G. Ionescu and I. de Madariaga, Opposition (London: Watts, r968); and Rodney Barker, ed., Studies in Opposition (London: Macmillan, r97r). In these writings the notion of opposition extends, as it should, far beyond the constitutional sphere. Yet, unless the right of opposition is constitutionally protected, nothing else may follow. The notion of minority rights (see chapter 2, section 4) is a broader one, for it also implies that minorities should be "respected" in their vital concerns. I discuss this further in chapter 8, section 7, under the "consociational democracy" heading. 5. See The Federalist, no. 48, and Jefferson's Notes on the State of Virginia. I am overlooking the other meanings of "majority" in Madison, both that which we might call federal (the possible tyranny of the majority of the larger and stronger states over the smaller ones), and that which refers to the problem of factions; the first because it simply extends the general principle to the specific case of the United States, and the second because the problem of factions is too complex to be discussed here (I discuss it in Parties and Party Systems, chaps. r and 4). 6. It should be underlined that my argument is strictly confined to electoral voting. "Decision making" properly called is taken up in chapter 8, herein. 7. De la Democratie en Amerique, vol. 1, pt. 2, chap. 7: "De !'Omnipotence de la Majorite aux Etats-Unis, et de ses Effets" (pp. 265, 266 in Gallimard ed., Paris, r951). 8. On Liberty (r859), chap. r (p. 6 in the annotated text edited by D. Spitz [New York: Norton, 1975]). For his indebtedness to Tocqueville, see Mill's Dissertations and Discussions (Boston: 1864), 2:79-161. For the differences between the two, and other finessings, see the analysis of Dennis F. Thompson, John Stuart Mill and Representative Government (Princeton: Princeton University Press, 1976), pp. 69-77. 9. Actually Tocqueville builds his case on the leveling and depersonalizing effects of equality. Interestingly, while Tocqueville's concern is underplayed by present-day pluralists, it is taken up by authors who, on most matters, are hardly of Tocquevillian descent. See, e.g., Barrington Moore's concern about a "popular totalitarianism" (Political Power and Social Theory [Cambridge, Mass.: Harvard University Press, 1958], pp. 83-87, 182-83). ro. The channeling function of parties is underscored in my Parties and Party Systems, chap. 3.1. Without parties, 50 percent plus one would come close to being a random outcome. rr. Note that my point is not the technical one that the majority method need not, and often does not, reflect the preferences of a substantive majority, of a majority will. 12. For additional elements and the current uses of"majority rule" see, in general, Elias Berg, Democracy and the Majority Principle (Stockholm: Akademiforlaget, 1965); and].R. Pennock, "Responsiveness, Responsibility and Majority Rule;' American Political Science Review, September 1952, and his article, "Majority Rule;' International

Notes Encyclopedia of the Social Sciences, vol. 9. See also H. McClosky, "The Fallacy of Absolute Majority Rule;' Journal of Politics u (1949): 637-54. 13. The standard treatment is W. Kendall, John Locke and the Doctrine of Majority Rule (Urbana: University of Illinois Press, 1941). However, Kendall misrepresents Locke as a strong defender of majority rule. In truth, Locke accepts it on grounds of expediency. 14. Cited in de Jouvenel, Du Pouvoir, p. 3ro. 15. H. Taine, Origines de la France Contemporaine (1875), Preface. 16. See D. Rae, "Decision Rules and Individual Values in Constitutional Choice;' American Political Science Review, March 1969 (the quotation is at p. 42); and M. Taylor, "Proof of a Theorem on Majority Rule;' Behavioral Scientist, May 1969. Since Rae often speaks of "the committee;' I suspect that he actually has in mind a small group. If so, his argument gains in plausibility; but Rae actually endorses Black's allinclusive definition of committee (that I criticize in chap. 8, n. 21). An ingenious defense of majority rule is also provided by Bruce A. Ackerman, Social Justice in the Liberal State (New Haven: Yale University Press, 1980), chap. 9, esp. pp. 274-93. 17. These are joint characteristics. On grounds of stalemate avoidance alone, majority rule enjoys no superiority over dictatorial or authoritarian rule. 18. I elaborate on these points (and specifically on the committee type of decision making) in chapter 8, section 5. The distinction between single-shot electoral choices and continuous issue deliberations also explains why I do not dwell on the problem of the intransitivity of preferences as stated in Arrow's "paradox of voting:' Arrow actually arrives at a "general impossibility theorem" that demonstrates, given certain conditions, the impossibility of attaining valid social choices in the sense that majority decisions cannot claim to reflect the first preferences of the majority. See Kenneth J. Arrow, Social Choice and Individual Values (New York: Wiley, 1951, 1963); but esp. his summation, "Values and Collective Decision Making;' in P. Laslett and W. Runciman, eds., Philosophy, Politics and Society, 3d Series (New York: Barnes & Noble, 1967). While Arrow's paradox of voting opens up a fascinating discussion, with regard to electoral behavior the issue has grown out of proportion; and I concur with W.H. Riker's point that the "minimal liberal conception" neither requires nor assumes that voting reveal a coherent popular will (see Liberalism Against Populism [San Francisco: Freeman, 1982], p. 244). 19. Only the evolution of the constitutionalism of the Republic of Venice presents a comparable and perhaps even greater degree of sophistication. The fundamental study is G. Maranini, La Costituzione di Venezia, 2 vols. (Venezia: 1927, 1934). 20. The importance of the constitutionalism of religious orders and of the evolution of their electoral techniques is highlighted by Leo Moulin, "Sanior et Major Pars;' Revue Historique de Droit Fran~ais et Etranger, vols. 3 and 4 (1958), pp. 368-97 and 491-529. See also his Le Monde Vivant des Religieux: Dominicains, Jesuites, Benedictins (Paris: Calmann-Levy, 1964). 2r. Considerations on Representative Government (1861), chap. 7 (p. 113 in the Liberal Art Press edition, New York, 1958). It is noteworthy that Mill already used "elite" (in the meaning subsequently taken up by Pareto). The quotation belongs to a context in praise of Hare's system of proportional representation; it also renders, however, a persistent theme of his Considerations. Mill was so concerned to ensure the influence of the "best" that he advocated (see chap. 8, pp. 135-43 in 1958 ed.) a

173

VERTICAL DEMOCRACY

plural voting based on occupational and educational criteria. See also nn. 77-79 below. 22. The complaint was already stated by Mill as follows: "At present, by universal admission, it is becoming more and more difficult for anyone who has only talents and character to gain admission into the House of Commons" (Considerations, chap. 7, p. n2 in 1958 edition). In the United States this grievance (a commonplace in continental EUiope) applies mostly to machine and local politics. Yet C. Wright Mills noted: "Knowledge and power are not truly united inside the ruling class" (,The Power Elite, p. 35 1 ). 23. Reflections on Government, p. 66. Among the few subsequent exceptions is Luigi Einaudi's essay, "Major et Sanior Pars;' in Jl Buongoverno (Bari: Laterza, 1954), esp. pp. 92-93. 24. In sections 8 and 9, below. 25. A. Zuckerman, "The Concept 'Political Elite': Lessons from Mosca and Pareto;' Journal of Politics 39 (1977): 327. 26. It is unnecessary for my present purpose to enter the many intricacies of the concept of power. See Dahl, Modern Political Analysis, chap. 5; J.G. March, "The Power of Power;' in D. Easton, ed., Varieties of Political Theory (Englewood Cliffs, N.J.: Prentice-Hall, 1966); Jack H. Nagel, The Descriptive Analysis of Power (New Haven: Yale University Press, 1975); Jan-Erik Lane and H. Stenlund, "Power;' in Sartori, ed., Social Science Concepts. For how "power" and "authority" relate to one another see chapter 7, section 2. 27. A. Giddens, in P. Stanworth and A. Giddens, eds., Elites and Power in British Society (Cambridge: Cambridge University Press, 1974), p. ix. 28. The notion of stratarchy is developed by S. Eldersveld, Political Parties: A Behavioral Analysis. Note that my argument does not bear on class structure and is not altered, therefore, by whether a structure is pyramidal, hexagonal, overlapping or, as suggested by G. Lenski, Power and Privilege: A Theory of Social Stratification (New York: McGraw-Hill, 1966), reciprocal. 29. Via Lasswell, as we shall see. Thus C. Wright Mills says "power elite;' Dahl "ruling elite;' and so forth. Mosca's term was, instead, "political class:' Mosca also employed "leading class" (classe dirigente); but the latter label is peripheral to his thinking. According to A. Lombardo, La Struttura del Potere (Roma: Bulzoni, 1972), pp. 27-34 and passim, classe dirigente applies to the sociological aspect of Mosca's thinking, which Lombardo neatly distinguishes from his political science. The basic difference between Mosca and Pareto is that the political class of Mosca is a political concept referred to the problem of the organization and exercise of political power; the elites of Pareto are a concept of social dynamics (with a much wider scope) focused on the qualities that are necessary to become part of that class (the theory of residues), and on the causes of their origin and decadence (the theory of the circulation of the elites). 30. See Trattato di Sociologia Generate (1916), nos. 2027 and 2031. 31. I Sistemi Socialisti (1902), Italian ed. (Torino: UTET, 1954), p. 21. (My translation, italics in the original.) The best guide in English into the Paretian maze is the anthology selected and introduced by S.E. Finer, V: Pareto: Sociological Writings (Oxford: Pall Mall Press, 1975). 32. "Agenda for the Study of Political Elites;' in D. Marvick, ed., Political DecisionMakers (Glencoe: Free Press, 1961), p. 66. 33. Lasswell and Kaplan, Power and Society, p. 201. 1 74

Notes

34. This formulation is of H.A. Hermens, The Representative Republic (South Bend: University of Notre Dame Press, 1958), pp. 21-25. 35. The Power Elite, passim. Mills conceded that "the power elite today involves the often uneasy coincidence of economic, military, and political power" (in A. Kornhauser, ed., Problems of Power in American Democracy [Detroit: Wayne State University Press, 1957], p. 166). This "uneasy coincidence" does not quite fit the image of one power elite; yet Mills did hypothesize a ruling class in the singular whose cohesiveness, when found wanting, was upheld by a conspiracy view of history. A defense of Mills is K. Prewitt and A. Stone, The Ruling Elites (New York: Harper & Row, 1973), chap. 4. Among the criticisms, see D. Bell, "The Power Elite Reconsidered;' in H. Girvetz, ed., Democracy and Elitism (New York: Scribner's, 1967), pp. 32off. 36. See ''A Critique of the Ruling Elite Model;' American Political Science Review, June 1958, pp. 463-69; and, for the empirical testing, Dahl's community power study, Who Governs? (New Haven: Yale University Press, 1961). 37. See P. Bachrach and M.S. Baratz, "Two Faces of Power;' American Political Science Review, December 1962. In a subsequent work, Power and Poverty (New York: Oxford University Press, 1970), the second face of power is defined, very differently, as the "decision" of non-deciding, of impeding decisions. While this redefinition is perfectly acceptable, it no longer constitutes a criticism of Dahl's approach. 38. Quoted by Bachrach and Baratz, "Two Faces of Power;' p. 949. 39. As concisely set forth, e.g., by D.F. Aberle et al., "The Functional Prerequisites of a Society;' Ethics, January 1950. But see R.K. Merton, Social Theory and Social Structure (Glencoe: Free Press, 1957), p. 33 and passim. 40. Bachrach and Baratz, "Two Faces of Power;' p. 950. 4r. For ulterior criticisms see the debate, "Nondecisions in the Study of Local Politics;' American Political Science Review, December 1971, pp. 1063-no4, esp. F. W. Frey's "Comments"; and the rebuttal of N. W. Polsby, "Empirical Investigation of the Mobilization of Bias in Community Power Research;' British Journal of Political Studies, June 1979. An attempt to rescue the nondecision thesis is S. Lukes, Power: A Radical View (London: Macmillan, 1975). Lukes's argument is criticized by B. Hindness, "On Three-Dimensional Power;' Political Studies, September 1976. It should be clear that my discussion does not deal with the middle but with the top level of power. 42. See Meisel, The Myth of the Ruling Class: G. Mosca and the Elite, p. 4 and passim. That neither Mosca nor Pareto propounded the three Cs formula-consciousness, coherence, and conspiracy- is pointed out, among others, by Zuckerman, "The Concept 'Political Elite; " pp. 332-34. 43. See The Lonely Crowd. Riesman joins the view that "the ruling class theories applied to contemporary America seem to be spectral survivals" (p. 238; but see his chap. n, passim). 44. The literature on elites and democracy is extensive, and extends far beyond the preliminaries discussed in this section. See, in general, S. Keller, Beyond the Ruling Class (New York: Random House, 1963); T.B. Bottomore, Elites and Society (London: Watts, 1964); Geraint Parry, Political Elites (London: Allen & Unwin, 1969), which also reviews the works of F. Hunter, R. Dahl, N. Polsby, and R. Presthus; A.S. McFarland, Power and Leadership in Pluralist Systems (Stanford: Stanford University Press, 1969); D. Rustow, ed., Philosophers and Kings: Studies in Leadership (New York: Braziller, 1970); P. Stanworth and A. Giddens, eds., Elites and Power in British Society;

r75

VER TI CAL DEMOCRACY

R.D. Putnam, The Comparative Study of Political Elites (Englewood Cliffs, N.J.: Prentice-Hall, 1976). See also n. 2., above. 45. The book appeared in German in 19rr and in Italian in 1912.. Since Michels was bilingual both texts may be considered originals. It was translated shortly after under the title Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (1915; Glencoe: Free Press, 1966). 46. Vom Wesen and Wert der Demokratie, chap. 2.. The linkage between democracy and parties (in the plural) is discussed extensively in my Parties and Party Systems, esp. chaps. 1-3. 47. See, for all, G. Maranini, Miti e Realta delta Democrazia (Milano: Comunita, 1958). In parallel fashion, the interposition, between the citizen and the state, of the party system also creates difficulties to the theory of representation. Along the representational process, it is the party, as Herman Finer vividly put it, that becomes "king:' Cited in S.H. Beer, British Politics in the Collectivist Age (New York: Vintage Books, 1965), p. 88. 48. La Sociologia del Partito Politico (Torino: 1912), p. 33. For a summary of Michels's theses "on the oligarchic tendencies of political organizations" see his Studi sulla Democrazia e sull'Autorita (Firenze: La Nuova Italia, 1933), pp. 58-59, where one finds the following passage written in 1909: "If there is a sociological law which political parties follow ... this law, when reduced to its most concise formula, must sound like this: the organization is the mother of the rule of the elected over the electors" (p. 49). 49. La Sociologia de! Partito Politico, p. 419. The German text reads Fiihrertum, and the Italian text sistema di capi; therefore, simply to translate "leadership" (as in the English version) ill renders the text. I add, however, "leadership" to remind that Michels's concept also extends, for lack of the distinction, to the latter notion. 50. In "Democrazia, Burocrazia e Oligarchia nei Partiti;' Rassegna Italiana di Sociologia 3 (1960), pp. rr9-36, where I also point to the difference between Michels's and Max Weber's approaches to bureaucratization. See, more generally, Lipset's Introduction to the Collier Books edition of Michels's book (New York, 1962.); and J. Linz, "R. Michels;' International Encyclopedia of the Social Sciences, vol. 10, and esp. his book-length Introduction to the Italian 1966 revised edition (Bologna: II Mulino) of La Sociologia de! Partito Politico. An overall assessment is Giorgio Sola, Organizzazione, Partito, Classe Politica e Legge Ferrea dell' Oligarchia in Roberto Michels (Genova: E.C.I.G., 1972.). 51. As authoritatively acknowledged, among others, by M. Duverger: in Michels's work "the oligarchic tendencies of mass organizations are still described in terms of the contemporary situation" (I.es Partis Politiques [Paris: Colin, 1951], p. x). In a similar vein S.M. Lipset writes: "The obvious conclusions of this analysis are that the functional requirements for democracy cannot be met most of the time in most unions" (Political Man [Garden City, N.Y.: Doubleday, 1960], p. 394). Michels is often confirmed unwittingly or indirectly. See, e.g., H. Karie!: "The voluntary organizations or associations which the early theorists of pluralism relied upon ... have themselves become oligarchically governed hierarchies" (The Decline of American Pluralism [Stanford: Stanford University Press, 1961], p. 2). 52. La Sociologia de/ Partito Politico, Preface, p. xiii. 53. The Logic of Collective Action: Public Goods and the Theory of Groups

Notes (Cambridge, Mass.: Harvard University Press, 1965), p. 134. Specifically, "the average person will not make a significant sacrifice for the party he favors, since a victory for his party provides a collective good" (p. 164). 54. As I will point out in section 8, below, the theory of participatory democracy simply ignores the theory of collective goods (let alone Michels). A.O. Hirschman counters Olson by pointing out that "public-oriented action belongs ... to a group of human activities that ... carry their own reward;' that the striving is not a cost but, rather, "part of the benefit;' if not the reward itself. (Shifting Involvements: Private Interest and Public Action [Princeton: Princeton University Press, 1979], pp. 85-86). This adds to Olson's "selective incentives" and explains what Olson leaves poorly explained. It does not seem to me, however, that Hirschman's argument substitutes the other one: on a larger scale, and over longer time periods, the "Olson law" still holds. 55. Michels stated, in effect, that representative systems are impossible recalling the Rousseauian postulate that the exercise of the will cannot be alienated (La Sociologia de! Partito Politico, p. 37). 56. Capitalism, Socialism and Democracy, p. 269. While Schumpeter's competitive theory does draw on a central economic analogy, it is not an economic theory as is, instead, A. Downs, An Economic Theory of Democracy (New York: Harper, 1957). I address the importance of the Downsian analysis in Parties and Party Systems, chap. 10. For a general appraisal of the economic theories of democracy, see Brian Barry, Sociologists, Economists and Democracy (London: Collier-Macmillan, 1970), especially chap. 5. 57. This rule was formulated by Carl Friedrich in the 2d ed. of his Constitutional Government and Democracy (Boston: Ginn, 1941), chap. 25. The chapter was omitted in the 1950 edition, but the rule was restated in Friedrich's subsequent work, Man and His Government (New York: McGraw-Hill, 1963). 58. Over time Dahl provides a variety of versions of this basic stand. In A Preface to Democratic Theory Dahl associates "polyarchal democracy" with "egalitarian polyarchies"; in Polyarchy the democratization of "full polyarchies" is measured by their "liberalization" and "inclusiveness" (see below nn. 61-63); in Dilemmas of Pluralist Democracy a country is declared "a pluralist democracy if (a) it is a democracy in the sense of polyarchy and (b) important organizations [subsystems] are relatively autonomous" (p. 5). All told, Dahl can be said to conceive polyarchies as "relatively (but incompletely) democratized regimes" (Polyarchy, p. 9). 59. See chapter 1, section 2. 60. This last point is completely missed by the anti-elitists. Their argument generally is that since the majority of the people cannot organize themselves into pressure groups, the majority remain voiceless. The fact that the majorities have voice, and often a winning one over pressure groups, precisely as electoral majorities, is consistently glossed over. 61. See Polyarchy, pp. 4-8. 62. This reading is plausible on two counts. First, it is Dahl who points out that inclusiveness alone leads only to "inclusive hegemony;' to wit, plebiscitarian and mobilizational regimes. Second, when Dahl addresses specifically the "good society" (in After the Revolution? [New Haven: Yale University Press, 19701), he does address, in the main, the problem of participation. 63. My difficulty with "contestation" is that it does not pay heed to the principle

177

VERTICAL DEMOCRACY

of the opposite danger (see chapter 4, section 5). Contestation may, but may not, be a democratizing force. 64. The representative writing is P. Bachrach, The Theory of Democratic Elitism: A Critique (Boston: Little, Brown, r967). Besides Bachrach, the gist and (in my opinion) the remaining best of the anti-elitist literature largely is in the articles assembled by Karie!, ed., Frontiers of Democratic Theory, pt. 3, pp. 95-323. For ease of consultation and brevity, reference will hereinafter be made to Frontiers. , 65. For instance, Bachrach's Theory of Democratic Elitism brings together Mosca and Schumpeter as "precursors" of present-day elitists. Bachrach1 could and, indeed, should have added to the group Max Weber, who was as "elitist" as Mosca and Pareto. Weber condemned the fuhrerlose Demokratie and praised a plebiscitary democracy that elects a charismatic leader. This thread is accurately documented by Luciano Cavalli, Il Capo Carismatico (Bologna: II Mulino, r982). 66. If for no other reason, because (it is at times alleged) they were put to use by fascism. But this happens to be false. As N. Bobbio correctly points out: "In the two major doctrinaires and creators of the doctrine of fascism, the philosopher Gentile and the jurist Rocco, the theory of elites had no part, not even peripheral. ... The actual followers of the theory of the political class have not been fascist writers, but anti-fascist and democratic writers .... The only serious attempt ... to apply and to refine Mosca's ideas ... has been made by the demo-radical pupil of Gobetti, Guido Dorso; and the only reelaboration of Pareto's ideas ... has been undertaken by the demo-liberal Paretian Filippo Burzio" (Saggi sulla Scienza Politica in Italia, pp. 247-48). 67. This is buttressed by noting that Dahl can also be interpreted as a "populist." In L. G. Sharpe's view, regardless of whether Dahl's emphasis is on "the processes by which ordinary citizens exert a relatively high degree of control over leaders" (as in A Preface), or on "the continuing responsiveness of the government to the preferences of its citizens" (as in Polyarchy), in either case "he remains a firm populist;' for Dahl allows "no place whatsoever for the capacity of government to respond:' (See "American Democracy Reconsidered;' British Journal of Political Science, r973, p. r32 and passim.) 68. In Frontiers, p. 32. 69. This implies that "participation" (as conceived by the participationist) resulted in a liability, not an asset. The point is, however, that Aristotle's polftes qualifies an anthropology, that is, an ontological being part (of the community), not our taking part as a mode of activity. 70. Rousseau's general category for the good polity is "republic;' which is "any government guided by the general will, which is the law" (Contrat Social, 11, 6, note); and "democracy" is but one of the forms of a republic, with respect to which Rousseau is characteristically ambivalent. On the one hand, he is quite critical. Democracies, he points out, are highly exposed to "civil wars and internal agitation" (Contrat, III, 4); and a republic in which the people themselves administer the laws, i.e., a Greektype direct democracy, is "necessarily ill-governed" (Second Discourse, Dedication). On the other hand, whenever Rousseau comes close to saying that democracy is the best form of republic, his immediate and repeated disclaimer is that a rigorously understood "true democracy has never existed and will never exist" (Contrat, III, 4 and IV, 3). Since Rousseau also asserts that the "best" political institution is an "elective" aris-

Notes tocratic republic, for "it is the better and more natural arrangement that the wiser govern the multitude, when it is assured that they will govern it in the multitude's interest" (Contrat, m, 4, 5), on balance his message appears to be that there is no best government in the absolute (Contrat, m, 9), and that a same recipe cannot be applied to all circumstances (Contrat, m, 8). 71. See Rousseau's Dedication to the Second Discourse (the Discours sur l'lnegalite) of r755 where he emphatically describes Geneva as the best regime to be wished for under contemporary conditions. It was not until r764, in his Letters Written from the ·Mountain (and after the authorities of Geneva had turned against his Emile and his Contrat Social), that Rousseau changed his praise in a harsh condemnation of Geneva as being a republic in name but a despotism in fact (esp. Letters 7 and 9). The idyllic relation between Rousseau and the Geneva republic in r754 is well recounted by M. Cranston, Jean-Jacques: The Early Life and Work of Jean-Jacques Rousseau, 1712-1754 (New York: Norton, r983), chap. r7. 72. Thus J.N. Shklar correctly interprets Rousseau as the last of the traditional utopians in that his ideal society condemns change. See "Rousseau's Two Models: Sparta and the Age of Gold;' Political Science Quarterly, March r966. 73. Rousseau's theory of democracy is examined in detail in chapter rr, sections 4 and 5. 74. See chapter 10, herein. 75. See, for all, G. Duncan and S. Lukes, "The New Democracy;' reprinted in Karie!, Frontiers (with the new title "Democracy Restated"). Also C. Pateman, Participation and Democratic Theory (Cambridge: Cambridge University Press, r970), abides by this pedigree (minus James Mill). 76. This is the quotation (drawn from the end of chap. 3 of Mill's Considerations) on which Duncan and Lukes, who consider Mill the "central democratic theorist;' build their case (in Frontiers, p. r91). See contra John R. Lucas, Democracy and Participation (Harmondsworth: Penguin, 1976), esp. chap. 8, which constitutes a judicious reappraisal of Mill's stance. See also n. 79, below. 77. Considerations, chap. 8 (p. 142 in 1958 ed.). See also n. 21, above. 78. For the literacy test, see Considerations, chap. 8 (pp. 131-33 in r958 ed.); for the exclusion of the paupers on the grounds that they do not have a stake in keeping down the public expenditures, see pp. 133-35. 79. Considerations, chap. 7 (p. 102 in 1958 ed.) and chap. 6 (p. 92). That, for Mill, "competence" takes priority over participation is well documented by Thompson, John Stuart Mill and Representative Government, esp. pp. 54-90. 80. Political freedom is defined and differentiated from other freedoms in chapter rr, sections 1-3. 81. See chapter 12, especially section 5. 82. Reference is made, for realism, to chapter 3, and, for idealism, to chapter 4, herein. Compare my treatment with how the relation is cavalierly handled, e.g., by J.L. Walker, "A Critique of the Elitist Theory of Democracy;' reprinted in Frontiers with the title: "Normative Consequences of Democratic Theory." 83. See chapter 5, section 6. 84. A book that well illustrates, if unwittingly, these vagaries is C.G. Benello and D. Roussopoulos, eds., The Case for Participatory Democracy. Also Pateman's Participation and Democratic Theory mainly deals, despite the breadth of its title, 1 79

VERTICAL DEMOCRACY

with democracy in industries. 85. Elites and Society, p. n8. Bottomore, although a critic of elite theories, does not belong to the anti-elitist group as identified and criticized in this section. Reference is made to Aron's "Social Structure and the Ruling Class;' British Journal of Sociology 1 (1950). But see, better, R. Aron, "Classe Sociale, Classe Politique, Classe Dominante;' Archives Europeennes de Sociologie 2 (1960). 86. The Theory of Democratic Elitism, p. 97. 87. Bachrach, pp. 40-41. 88. Actually Bachrach attributes the views that I quote to me. The caricature and disparagement are so evident that the editor of the Italian translation of Bachrach's book writes that his misreading of my Democratic Theory is "a patent case of polemic distortion" (M. Stoppino, "Presentazione;' La Teoria dell'Elitismo Democratico [Napoli: Guida, 1974], pp. xvii-xviii); and a more careful reader of my work wonders whether my understanding of democratic leadership differs "very much'; from the views of]. Stuart Mill (Parry, Political Elites, p. 152). 89. Note that the argument is not that participation is low and information poor; these are statistics. The "argument" is that opportunity costs are involved. Why is this "elitist"? Is it wrong? No argument is provided on the argument. 90. Frontiers, p. 32 (my italics). 9r. History of the Peloponnesian War, trans. Richard Crawley (New York: Oxford University Press, 1950), bk. II, chap. VII, pp. 142-43. 92. The American Commonwealth (1888), p. 432 in 1959 ed. 93. S. de Madariaga, Anarchie ou Hierarchie (Paris: Gallimard, 1936), p. 56. 94. Man and Society in an Age of Reconstruction, rev. English ed. (London: Routledge & Kegan Paul, 1940), p. 87. 95. A.D. Lindsay, The Modern Democratic State (London: Oxford University Press, 1943), p. 261. 96. Public Opinion and American Democracy, p. 558. Reference is made to Key's The Responsible Electorate (Cambridge, Mass.: Belknap Press, 1966). 97. The Cultural Contradictions of Capitalism (New York: Basic Books, 1976), p. 204. 98. With one major exception, the "valuation" of the individual person (see chapter 10, section 3). But this exception has no bearing on the present argument. 99. See esp. Orwell's masterful Appendix to 1984; but also "Politics and the English Language;' now in Selected Essays. 100. For instance, H. Eckstein and T.R. Gurr define "selection" as follows: "procedures by which incumbent supers themselves choose their successors and/or peers" (Patterns of Authority: A Structural Basis for Political Inquiry [New York: Wiley, 1975], p. 372). This stipulation (a) loses the qualitative connotation of the term; (b) alters its meaning (selection means what "cooptation" means); and thus (c) increases, quite unnecessarily, its ambiguity. ror. J;Esprit des Lois, bk. VIII, chap. II. 102. Discourse on Inequality, closing sentence. 103. See above section 4, and nn. 32 and 33 for the first two definitions. The third is in Lasswell and D. Lerner, The Comparative Study of Elites (Stanford: Stanford University Press, 1952), p. 13. In like manner G.L. Field and John Higley define "elites" as "the persons who occupy strategic positions" and underscore "that 'elite' 180

Notes is not used . . . to designate persons allegedly distinguished by 'superior' personal traits or skills" (Elitism [London: Routledge & Kegan Paul, 1980], p. 20). 104. It is interesting to note that while all of this is missed in equal manner by elitists and anti-elitists (no matter how arranged), it was instead a major concern of C. Wright Mills, who contrasted the power elite to the intellectual elite and sought to render the former accountable to the latter. See The Causes of World War Three (New York: Simon & Schuster, 1958), chap. 7. 105. The analysis of equality, and the underpinning of the notions of equality of opportunity and merit, must await chapter 12, esp. sections 4 and 5. 106. This is the "first statement" of the "original position" of J. Rawls's second principle of justice. See A Theory of Justice (Cambridge, Mass.: Harvard University Press, 1971), p. 60. 107. Note that, as in the case of selection, the term elite has been devalued only in the sphere of politics. When we speak, e.g., of intellectual elites, the original connotation remains. 108. See M.J. Crozier, S.P. Huntington, and J. Watanuki, The Crisis of Democracy: Report on the Governability of Democracies to the Trilateral Commission (New York: New York University Press, 1975); and A. Lombardo, La Crisi delle Democrazie Industriali (Firenze: Vallecchi, 1977). An overall assessment is Richard Rose, ed., Challenge to Governance: Studies in Overloaded Polities (Beverly Hills: Sage, r980). The economic aspect is emphasized in R. Rose and B. Guy Peters, Can Government Go Bankrupt? (New York: Basic Books, r978). But see, more specifically, S. Brittan, "The Economic Contradictions of Democracy;' British Journal of Political Science r (1975); and J.M. Buchanan and R.E. Wagner, Democracy in Deficit: The Political Legacy of Lord Keynes (New York: Academic Press, 1977). 109. I portray independent responsibility as the characterizing element of modern representation (in contrast to the medieval mandate theory of representation) in my article, "Representational Systems;' International Encyclopedia of the Social Sciences, where I make the full argument. For the essential bibliography, see chapter 2, n. 23, herein. no. Considerations, chap. 15 (p. 228 in 1958 ed.).

181

7. What Democracy Is Not What is totalitarianism? An ancient tyranny under new conditions, or a new phenomenon? - Milovan Djilas

7.r

Contraries, Contradictories, and Degrees

A definition must embrace the whole of what it defines, but no more. Thus, to define is, first of all, to assign limits, to delimit. An undefined concept is, by the same token, a boundless concept. The standard manner of delimiting a concept is to define it a contrario, by contrast, i.e., by establishing its opposite, contrary, or contradictory. Hence, in order to establish what democracy is we must also establish what democracy is not, that is, what is the opposite of democracy. Generally, definitions a contrario are the easiest. What is white? It is the contrary of black. What is good? It is the contrary of bad. And so forth. The problem simply is to find-when seeking to delimit a concept-a "good opposite." But while most terms have antonyms, a good oppositeness of meaning may not be found easily. For example, when we attempt to define politics, we often say that politics is neither economics nor ethics, thus revealing that the concept of politics has no established and even less exhaustive opposite. The main point to note, however, is that the logical status of two contrasting terms may be of two sorts: They may either be or not be contradictories. In what follows I use opposite as the generic term for when the specification is unnecessary, i.e., for both cases; contrary for opposites that are not contradictories; and contradictory for two terms that are not only mutually exclusive but also exhaustively exclusive. The logical point is this: The principle of the excluded middle applies to contradictories-not to contraries. It is only with contradictories that tertium non datur, that no third possibility exists (e.g., either one is alive or dead, married or single, two-legged or quadruped). Contraries are mutually exclusive but not exhaustively so; they admit, therefore, third, intermediate possibilities (e.g., neither big or small, neither hot or cold, neither rich or poor).

Contraries, Contradictories, and Degrees

7.1

While the mishandling of contraries and contradictories is a contributing factor, the major difficulties in determining what democracy is not currently stem from the methodological orientation promoted by the quantitative bent of the social sciences. Under the quantitative impetus we are no longer prompted to ask what is (or is not) a democracy; we are, instead, required to ask to what degree, if any, a political system is a democracy. Does the second question supersede the first one? Is it a better question? This appears to be the prevailing view. I take it, instead, that both are equally legitimate and, indeed, mutually complementing questions. Let us assess first the logical status of asking: What is democracy? Under this formulation democracy is construed as an entity, as an object concept, and, more precisely, as a specific class (type) of political system. Therefore the logical treatment entailed is classificatory, that is, a binary, dichotomous or disjunctive treatment: We are required to decide whether a given polity is either democratic or not. This also entails that the differences sorted out by such treatment are of kind (not of degree). And since we are dealing with a logical treatment, it is no objection at all to object that "what is" questions are vitiated by ontological and essentialist assumptions. If people misconstrue a logical treatment, this is a fault in its users, not of the logic that they misemploy. When the question is, To what degree is a polity more or less democratic? then we are no longer identifying an entity but predicating something about it. This is the same as saying that "democracy" is now conceived as a property concept, as a property (characteristic, attribute) of political objects. Under this formulation the logical treatment is no longer binary (yes-no), but continuous (greater-lesser), usually described as the transformation of dichotomous (discrete) characteristics into continuous characteristics. It also follows - logically, not ontologically-that under this treatment differences are of degree. It must be noted, however, that How much? questions can be formulated in two fundamentally different ways. In the case at hand we may either ask to what extent a democracy is democratic, or ask to what extent any polity is democratic. If the first question is being asked, we shall see that the problem is easily handled. The difficulties arise with the second question, as I shall immediately explain. When we predicate that a political system is democratic to some extent or degree, the preliminary question is: Democratic with respect to which characteristic( s)? For instance, Douglas Rae selects the property "majority rule"; 1 but democracy is majority rule cum minority rights-quite a different thing. Felix Oppenheim selects the characteristic "participation"; 2 but this concept is by now so ill-defined that it might even lead to the finding that (on a par-

WHAT DEMOCRACY IS NOT

7.r

ticipation measure) the fullest democracy ever to exist was China at the time of its so-called cultural revolution. The difficulties along this path are at least three. First, there are hosts of characteristics or properties eligible for selection; not only majority rule and participation, but also equality, freedom, consensus, coercion, competition, pluralism, constitutional rule, and more. Second, these characteristics are so interrelated that any single measure of any selected category is likely to produce highly erratic rank orderings. Third, and to my mind foremost, if we start, as we in effect do, from a conceptual morass, operationalizations (remember that we are presumably seeking measures) can only add extravaganza to fuzziness. 3 But let us assume, for the sake of the argument, that all these difficulties have been overcome. We still have to confront the question of the heuristic value of this approach. We are asking, let it be recalled, how much (or little) of a given democratic characteristic, or of a set (index) of democratic characteristics, can be found across all political systems. Now, it is very unlikely that any measure of any category would ever yield a zero value, i.e., indicate a total absence of the characteristic being measured. If so, we would have to conclude that al] the existing political systems are democracies, albeit to a lesser and lesser degree; or, conversely, that all existing polities are nondemocracies, except that;' say, Cambodia or Albania are more nondemocratic than Britain. Quite aside from the stultifying nature of conclusions of the sort, what is completely missed by this kind of degreeism, or continuism, is that political systems are systems, that is, bounded wholes characterized by constitutive mechanisms and principles that are either present (albeit imperfectly) or absent (albeit imperfectly). Could Stalin or Hitler have been ousted from power by a free election? No. Can a President of the United States be impeached? Yes. Do parties (in the plural) compete among themselves in the Soviet Union? No. Do they compete in West Germany? Yes. It can now be easily seen why the question, To what degree is a democracy democratic? is an altogether different question. When this is the query, then the scholar assumes-even if unaware-a prior classificatory breakdown. If so, there is no incompatibility between the classificatory and the degree treatments; within democracy as a class (type), one can assess as many variations in degree (of more-and-less democracy) as one sees fit. The point that plagues our discussions bears, then, on the sloganistic abuse of the phrase: All differences are differences in degree. If one held, today, that all differences are in kind, most people would take this to be a logical stupidity. Yet the .first maxim is just as stupid as the second one. There is nothing in the intrinsic nature of human artifacts that establishes that all differences are of degree-exactly as there is nothing in the intrinsic nature

Authoritarianism, Authority, and Power 7.2 of things that establishes that they are of kind. Differences are of degree if so treated (logically). Similarly, differences are of kind under the classificatory (per genus et differentiam) treatment. Whether differences are qualitative or quantitative, in kind or degree, discontinuous or continuous, is a matter of logical treatment and therefore a matter of deciding which logical treatment is appropriate for what purpose. 4 In my view, What is (democracy)? and How much (democracy)? are both rightful and complementary, not mutually exclusive, questions. It is also my view that they should be asked in that order. This is so because without establishing first what a thing is (and is not), we cannot establish its degree of being whatever it is declared to be. My stand is, thus, that variations within democracy or of democracy (relating to more and less democracy) require that we first establish to what they apply, that is, that we first decide what is, and what is not, a democracy. Only a very sloppy logic can dispose of all problems by declaring that everything is a matter of "more and less." This sloppy logic has prompted, in turn, the neglect of a contrario definitions- a neglect that needs to be redressed. Real people do live under regimes and political forms that they either wish to escape or long to enter. Indeed, in recent times millions of people have fled their homelands at the risk of their lives. They have not done so for a mere increment, for a lesser or greater "degree" of what they already had-they have sought what they did not have.

7.2

Authoritarianism, Authority, and Power

let us revert to what, or which, is a "good opposite" of democracy. The choice is abundant: tyranny, despotism, autocracy, absolutism, dictatorship, authoritarianism, and totalitarianism. Tyrant and despot are ancient Greek names. Absolutism and autocracy have entered the vocabulary of politics (despite their respective Latin and Greek roots) only since the eighteenth century. Dictatorship is a Roman term, but acquired its current meaning in the twentieth century. 5 Authoritarianism and totalitarianism are the more recent coinages; they emerged in the aftermath of World War I. Thus, when today the question is asked, What is the opposite of democracy? we tend to reply: totalitarianism or authoritarianism. And it is totalitarianism that is assumed to represent, at least in common parlance, the fullest negation of democracy. It is appropriate, then, to begin with the more recent entries in the series and, first, with "authoritarianism:' Authoritarianism derives from "authority;' and only a suffix, only an ism, separates meanings that are, or can be, astronomically distant. Authority is a very old Latin term, and has never been (at least, until few decades ago) a pejorative word. Across the centuries, au-

WHAT DEMOCRACY IS NOT

7.2

thority has invariably been good, an appreciative term. Today, however, authoritarianism is a derogatory word; it indicates an excess and an abuse of authority, indeed, an oppressive authority that crushes liberty. I have no quarrel with this transformation provided that we save "authority" from being contaminated (if not ultimately swallowed) by "authoritarianism?' That is to say that it is important to maintain authority as a distinctive concept that we cannot afford to spoil. The original meaning of "authority" is probably related'to the Latin verb augere, to augment. 6 In its earliest meaning it indicates that those in authority reinforce, confirm, and sanction a course of action or thought. For the Romans the auctoritas of the living depended on and stemmed from the authority of the city's founders. Thus, what is "augmented" by those in authority is the foundation. In its original meaning, then, authority belonged to a traditionalistic conception of history, to history conceived as a growth of the past, as the confirmation and expansion of an Ursprung, of an inception. This meaning fades away when history begins to be conceived, at the end of the eighteenth century, as the adventure of man breaking with the past, as the invention and creation of a future rather than as a reinforcement of origins. Yet, as the meaning changes, the value connotation remains unaltered. 7 It would lead me far afield to follow, from the Middle Ages and beyond, the elaborations of the concept. Let me simply underscore that auctoritas, though related to potestas, to power, was always neatly distinguished from it. Authority was seldom, if ever, defined in terms of power until Max Weber and, more accurately, until he was mistranslated into English. Weber's central term was Herrschaft- indeed a difficult term to render. Of all its permissible translations, "authority" is the worse one- and it became, unfortunately, the standard one in English. Originally Herrscha/t typically connoted seigniory (feudal lordship) and, thereby, domination and mastery. It does not have the harsh overtones of Macht (which is both power and force), and this suggests that "rulership" and "rule" might be its current best English rendering. But no harm would have followed had Herrschaft been translated "power?' 8 Let it be added that German does possess the word Autoritat. Therefore, had Weber intended "authority;' he did have the word for saying so. But he intended the power to impose and receive unquestioning and instant obedience. The mistranslation of Herrschaft turned out to be a devastating event for the semantic field covered by the terms authority, influence, power, force, and coercion. In particular, "authority" became for the sociologist what "power" was for the political scientist; and as the word authority also entered organization theory and the vocabulary of psychology, the concept began to drift, dismembered, in all directions. Even in the field of politics "authority"

186

Authoritarianism, Authority, and Power 7.2

seldom maintains, today, trace of its ancestry. To wit, in political theory the concept of authority has always been linked with legitimacy, not with legality. But when Lasswell and Kaplan cavalierly asserted that "authority is formal power;' 9 this momentous reduction of legitimacy to legality went largely unchallenged and was largely followed.10 On this and other counts we have thus lost sight of how and why authority contributes to the understanding of politics. When a concept becomes mishandled beyond repair by the specialists, one way out is to look at ordinary usage. This is not because ordinary usage possesses any special wisdom - it does not- but because the viscosity of ordinary usage may retain prior wisdom. A terminological study conducted under UNESCO auspices indicates the "common usage" of the term authority as "power that is accepted, respected, recognized, legitimate:' 11 Since the orbit of meaning captured by this report of usage does reflect an impressive, if too often forgotten, "authoritative" literature, I shall abide by it. Power or, at least, political power is generally associated with coercion and coercibility, sanctions and sanctionability. To be sure, potes-tas comes from a verb that means "capacity;' being able to, as is still transparent in French (where pouvoir means, as a verb, "I can"). However, the etymology of the word power expresses only part of its political meaning, which is not capacity to do, but puissance, "capacity of having things done" (pouvoir de faire faire). This shift is very clear in the German Macht, which is both power and force. Who is in power has the power to bestow rewards; but especially, and more distinctively, the power to inflict deprivations. Thus, power orders; and state power issues commands sustained by the legal monopoly of force.12 But when power is so conceived, it is immediately apparent that it does not suffice to explain how a political society hangs together and why its members comply. Intra-societal processes aside, political processes cannot be said to be only power processes (as defined). Philosophers have long insisted that the cement of political societies is not provided by commands but by the very different thing that they call "political obligation:'13 Alternatively, but also concurrently, we generally say that political regimes are upheld by their "legitimacy" and undermined, if not ultimately destroyed, by a crisis of legitimacy.14 In turn, a crisis of legitimacy is generally signified by, and detected as, a crisis of "authority!' Though it cannot be said that the concepts of obligation, legitimacy, and authority are interchangeable, they are sufficiently interconnected to permit the conclusion that "authority" explains what "power" leaves unexplained. If political processes are not only power processes, the needed implementation is that political processes are also authority processes (under the defini-

WHAT DEMOCRACY IS NOT

7.2

tion of authority given above). And if this reasoning is correct, let the explication of "authority" be pursued in the light of the distinction between power processes and authority processes.15 While power orders and is backed, when necessary, by enforcement, authority "appeals"; it is not a function of rewards and ceases to be such if enforced. Authority is, then, the form of power, or the form of influence;6 that arises from spontaneous investiture and that draws its efficacy from being listened to, from acknowledgment. We can equally say that authority is a power based on prestige, on deference. Ultimately, authority reflects eminence, and the most succinct meaning of auctoritas can be rendered as "moral influence?'17 In any event, authority does not coerce us into doing something, for an authority relationship does not make us unfree to act otherwise. As Flathman correctly points out, authority is consonant with autonomy.18 Correlatively, and as already noted, authority involves legitimacy. Indeed, authority goes to explain legitimacy, and legitimacy authority; the two concepts are so intertwined as to represent two faces of the same coin. Authority gets things done (or not done) not by commanding but by "rightfully" asking or suggesting. This is why we associate authority with leadership that receives spontaneous support. And this is also why-or the meaning when-the crisis of our democracies is described as a crisis of authority. Pausing on this last remark, when power and authority are conceived as the symmetric vis-a-vis of each other, power without authority is either oppressive power (a situation in which naked coercion surrogates for, and ultimately destroys, authority), or impotent power (a situation of powerlessness). From this one can easily perceive the crucial importance of authority for democracy. In order to minimize oppression (coercion) without falling into impotence, democracies require power backed by authority. Analytically, this is not to say that authority is something that accompanies power but to say that authority replaces power. The point can be pressed, normatively, as follows: Democracy should aim at transforming power (a vis coactiva) into authority (a vis directiva). This is so because the greater the role and extent of authority-of asking instead of commanding-the lesser the role and extent of power, i.e., the lesser the power of power. So, the assertion that the crisis of democracy is a crisis of authority has a special pointedness. Far from being repugnant to democracy, authority can be said to be its power formula par excellence. The ideal that enhances democracy is not the conquest of power but, on the contrary, its minimization, and therefore the replacement of "power holders" by "authority holders?' Possibly this is one of the reasons that prompts us, in democracies, to speak of the authorities, state authority, and the structure of authority, thus 188

Authoritarianism, Authority, and Power 7.2 leading a number of authors to select authority as their central concept. The drawback of this formulation is that the pointedness of the concept is lost when authority is simply conceived as the generic term for any and all asymmetric relations "among hierarchically ordered members of a social unit that involves the direction of that unit:'19 More on this pi;esently. Let us first press the issue of concern, which is how authority generates authoritarianism and the feedback of the new term on the old one. The genesis was simple, and there was a good logic in it. Authoritarianism as a name for a political system was coined by fascism and was intended as a laudatory term; it was indeed meant to carry over to a dictatorial state the favorable attributes or associations of authority. As the label was received in the democratic camp, the value connotation was inverted, and the meaning was readjusted accordingly. For its propounders, authoritarianism was a regime in which "true authority" was restored- as against the putrefaction of the decadent, authority-less "pluto-democracies:' For the democrats, authoritarianism is instead a regime that counterfeits and abuses authority. As anyone can see, the conceptual border is thin. The discrimen resides in what is "true" authority and/or in which is the "good" authority and, eventually, in how much authority is not too much authority. Of course, a divide can be thin and yet be drawn. But not when authority is defined as "formal power;' or institutionalized power. Under this definition (and similar ones) it goes by definition that since all regimes have a formalized structure of power, they all have a structure of authority. If so, we are precluded from dividing the so-called authority of authoritarianism from the authority that sustains democracy. It looks as though "authoritarianism" cannot be a good opposite of democracy unless we do better with "authority." An overly extended and diluted concept of authority blurs the divide between democracies and authoritarianisms. If "state authority" applies to any and all states, and if all politics, everywhere, consists of authority patterns and of "authoritative allocations':._ as Easton would have it 20 -then a no-difference suggestion is being conveyed that makes authoritarianisms nicer than they are or, conversely, democracies uglier than they deserve. The fact that Easton's adjective is "authoritative" does not alter the problem, for his predication applies to politics per se and subsumes, therefore, both authority and coercion. Hence, Easton cannot avoid recognizing a "coercive authority" and arguing that a thief ordering at the point of a gun still represents a case of authority. 21 However, Easton can be put to helpful use by distinguishing between authoritarian and authoritative. Thus, if we wish to convey the idea that democracy typically requires power as al!-thority, we can resort to distinguishing between authoritarian (nondemo-

WHAT DEMOCRACY IS NOT

7.2

cratic) authority, and authoritative (democratic) authority. Along the same lines, and for the same reason, the "authoritarian personality" should be distinguished from the "authoritative personality" in order to make sure that the personality type that can best serve democracy is not assumed to be, without further ado, authority-less. 2 i But it needs little reflection to see that these finessings leave us on thin ice-for we have long ceased to be yirtuosi of language. Thus far my analysis has been confined to the conceptual give and take (and reciprocal corruption) between authority and authoritarianism. But authority can be construed in a very different association, namely, in its relation to liberty. Until the 1920s, authority was understood and perceived, almost invariably, as a vis-a-vis or a correlative of freedom. In this long-standing perspective, the argument is that true freedom accepts authority, just as true authority recognizes freedom. Freedom that does not acknowledge authority is arbitrary freedom, licentia not libertas. Vice versa, authority that does not recognize freedom is authoritarianism. This is much better, that is, much clearer, for it shows nicely in what way democracy needs authority and yet is not authoritarian. When we look at authoritarianism from the vantage point of liberty, then it does clearly follow that authoritarianism negates authority (since authority goes to define freedom). As Friedrich accurately asserts, "in a totalitarian society true authority is altogether destroyed:' 23 By the same token it can be said that the more a regime is authoritarian, the less it rests on authority. To conclude, it is the assertion that authoritarianism is a political system that leaves little, if any, room for freedom that best underpins the notion. This is also to say that it is via the authority-liberty detour that "authoritarianism" is best pinpointed as an opposite of democracy. As the argument proceeds I shall suggest contraries that are, I believe, more pointed ones. For the time being, I simply note that while authoritarianism certainly can define democracy a contrario, one should be wary of construing the notion as a contradictory of democracy.

7.3

Total State, Democracy, and Absolutism

Let us get on to "totafitarianism:' Totalitarianism is a redundancy of totality and simply conveys, as such, the idea of something that embraces and comprises everything. If so, or to the extent that the word totalitarianism sets our thoughts on the idea of totality, our reactions are likely to be ambivalent. While we still may value privacy and the "small state;• we are also aware that in some form a trend toward "more totality" characterizes our time. Demo-

Total State, Democracy, and Absolutism

7.3

graphic pressure, the continually increasing interdependence of all aspects of the industrial and service society, the demand for welfare and for more rational, functional, and planned societal forms, to mention only the more conspicuous reasons, lead present-day governments, whether they like it or not, to concern themselves increasingly with everybody and everything. In this connection we could well say, therefore, that all modern states are somehow developing in the direction of a "total state:' The democratic state is no exception. If anything, it is more legitimized to become an all-pervading, total state than any other regime. For, in principle, no political formula can justify a total expansion of political power as easily as democracy. The power that originates in everyone is authorized, ex hypothesi, to do everything. "The democratic fiction;' de Jouvenel remarks, "lends the leaders the authority of the whole. It is the whole that wills, it is the whole that acts." 24 These reflections perhaps help explain why a number of authors stress that totalitarianism does not indicate an extension - the total state- but rather an authoritarian and/or absolute exercise of power. But this line of explanation can lead only to the dilution of the concept of totalitarianism. That is, neither authoritarianism nor absolutism can be the explanans of totalitarianism. Even if we possessed (though I do not grant it) a solid theory of authoritarianism, it would still be the case that while any totalitarian system is also authoritarian, the reverse is not true. The only necessary property of an authoritarian exercise of power is that the freedom of the subjects will be trampled on. That this should take place in a totalitarian manner is a further condition. The same is true when totalitarianism is linked with, and explained through, "absolutism:' Absolutus is a term used by the Latin commentators of the Aristotelian concept of panbasileia (translated as rex absolutus) and did not have a derogatory meaning. Potestas absoluta, absolute power, merely indicated a supreme, perfect, complete, and (in this reference) intangible power. It is only with Machiavelli, and especially with Guicciardini, that a potesta assoluta is linked with virtual tyranny (and this only in a peripheral way) and that the Latin root absolvere, which conveys the meaning of "setting free;' begins to be associated with an imperfection. 25 Bodin's De la Republique of 1576 is generally considered the first systematic statement of the secular theory of absolutism. But his maiestas, while a power that stands above positive law, still is subject to natural and divine law. Thus Bodin's theory of the sovereignty of the state does not call for an absolute monarch that no longer submits to any law. For this we have to await Hobbes. And it was not until the eighteenth century that "absolutism'' was coined (as a noun) and used in its modern sense. In the modern meaning, then, the absolute power of absolutism denotes

WHAT DEMOCRACY IS NOT

7.3

a power set free from controls, released from restraints: a limitless, discretional exercise of power. However, the concept owes much of its definiteness to the fact that we associate it with the absolute seventeenth- and eighteenth-century monarchies -which were still patrimonial states, that is, a state conceived as the monarch's estate. It follows that as "absolutism" is applied to the present we must ask ourselves which are the other conditions-aside from the patrimonial conception of the state-that facilitate or indicate an absolutism. In a first sense, a power that is not opposed, in fact, by adequate countervailing powers is an absolute power. In this connotation, absolutism is related to the concentration of power. The more a society loses its pluralistic structure and the more its intermediary forces are weakened, the more easily are created the conditions that make absolutism possible. Thus any state that concentrates all power in itself is, potentially, a state that can exercise absolute power. This implies also that an overcentralized democracy 26 - one that replaces the spontaneous interplay of a multigroup society with its own unicentric will- is in a position to exercise an absolutist kind of power. In a second sense, we have absolute power when it is not disciplined and limited by law. In this meaning the absolute state is the nonconstitutional state, that is, a state legibus solutus, unbounded by law, in which the power holders are not restrained by, or have become released from, constitutional checks and limits. In this connection the query is: Is the democratic state necessarily a liberal-constitutional state, and more precisely a Rechtsstaat of the garantiste type in which the sovereign demos (no less than the former prince) is legibus obligatus, obligated by, and submitted to, higher laws? In the last two centuries this has been, in fact, the case. But if we say "necessarily;' then the answer is that democracy and constitutionalism are not indivisible partners. The Athenian democracy foundered, among other reasons, on the reefs of a demos that claimed to be above the law. And the maximization of democracy seeks, or at any rate entails, a minimization of the garantiste aspect of liberal constitutionalism. So, a democracy may become absolute also in the constitutional meaning of the expression.27 On all counts, then, absolutism is a poor opposite of democracy. It is obvious that a democracy cannot be absolute in the sense that absolute monarchies were; but a democratic absolutism is well within the range of possibilities. Democratic legitimacy per se does not call for the limitation of power. A limited power while fighting or resisting another power, popular sovereignty becomes in victory (in the absence of counterpowers) a source of unlimited power. More specifically, if we contend that democracy is achieved by taking all the power from the despot to give it to the people, all this operation ac-

Totalitarianism

7 .4

complishes is to present us with an absolutism in reverse. 28 And in this case one is perfectly justified in maintaining that it is precisely democratic legitimacy that gives absolute sanction to power. For there is no appeal when sovereignty is exercised in the name of the people; we are already in the final court of appeal. As to whether absolutism helps in determining the meaning of totalitarianism, it is clear that any totalitarian system is also absolute. But the reverse is not true: Absolutism need not be totalitarian. Absolutism implies only that power is exercised discretionally. Therefore the concept of totalitarianism must be confronted in its own terms, without leaning, for its explanation, on the preexisting nomenclature of political forms.

7 .4

Totalitarianism

The word totalitarianism first appeared in 1925, and, like authoritarianism, it was invented by fascism. To Mussolini, the "total state" sounded impressive, important; and the label "totalitarian state" appealed to his vanity and rhetoric. But this was mostly and largely rhetoric. Fascism, in Italy, was a clear-cut case of authoritarian dictatorship-far more than a "simple dictatorship" but far less than a totalitarian state. The case was very different in Germany under Hitler. 29 Indeed, the early authors who had personal knowledge of totalitarianism invariably made reference to what happened during Nazism. Contrariwise, no Italian who lived the experience of fascism has ever written a scholarly book taking fascism seriously as a totalitarianism. 30 It should be clear, therefore, that the democratic literature on totalitarianism launched by the pioneering volume of Hannah Arendt and epitomized in the classic 1956 volume of Friedrich and Brzezinski-let us say the literature of the 195oshad two paramount referents: Nazism and Stalinism. 31 As new generations came on the scene, fascism gradually became, together with Nazism, the number two case (it was easy, it was dead), and communism (which was, instead, well and alive) became more and more a debated case, and a case that made "totalitarianism" debatable. Before entering these developments, a good starting point is a 1953 discussion subsequently edited by Friedrich in the symposium Totalitarianism. 32 The discussion is excellent, and brings out the intrinsic difficulties of the concept. The disagreements center on a major alternative: whether the word should indicate a trait that is common to many different state-societies, or whether it should designate a particular type of political system characterized by its own special syndrome of features. A closely related alternative is: Should "totalitarianism" be used for all epochs, including ancient times, or 1 93

WHAT DEMOCRACY IS NOT

7.4

should it be reserved for our time, to indicate an experience whose novelty we wish to emphasize? These alternatives reflect broad strategies of inquiry, with respect to which my firm view is, here as elsewhere, that we should both promote the economy of language and combat the diseconomies of ambiguity. I therefore take the view that totalitarianism is a new name for the hitherto unnamed. If, instead, we dress totalitarianism in old cloth~s, the chances are that the category is being wasted or that confusion is reinforced. These are precisely the results, I submit, of applying "totalitarianism" retrospectively to ancient societies and/or applying it loosely to a variety of different polities. Certainly, if we wish, we can hold that the ancient polis was totalitarian or, with Franz Neumann, that the government of Sparta and the regime of Diocletian are examples of totalitarianism. 33 Except that the analogy is superficial. As has been well pointed out by Karl Loewenstein, "We are not entitled to qualify certain ancient autocracies as totalitarian because the telos of the state-society ... rarely articulated in secular terms, was accepted without disagreement by the power holders and the power addressees alike, and so deeply ingrained in tradition that it required neither ideological formulation nor enforcement:'34 The caution applies equally to the extension of "totalitarianism" to preliterate societies and to Oriental despotisms-as in Wittfogel's study of hydraulic societies. 35 The surprise element of the total domination of our times is that it can be superimposed not only on societies traditionally conditioned by despotic rule (such as Russia and China) but also on societies historically nurtured by, and exiting from, the Christian, liberal, and liberaldemocratic tradition. Furthermore, as Aron pointedly puts it, ''Asiatic despotism did not entail the creation of a new man nor the wait for the end of prehistorY:' 36 If these elements are discounted, or underplayed, we miss the very key to the matter. 37 We should also make clear whether we wish "totalitarianism" to denote a type of polity or something else. In Edgar Hallet Carr's view, totalitarianism is "the belief that some organized group or institution, whether church or government or party, has a special access to truth:' 38 In turn, this definition allows Carr to argue that totalitarianism is as old as the world, that it has always existed and always will (except for the brief parenthesis, he asserts, of the "age of individualism"), thus providing the excusatio of the Soviet variety of totalitarianism. 39 But this does not follow. The referent of Carr's definition is not totalitarianism as a political regime but the totalitarian mind (belief system, ideology, mystique, religion, or vision of the world). This is also the referent of J.L. Talmon 40 and other authors, and represents a legitimate and fascinating line of speculation. Yet, and at most, the totalitarian mind can 1 94

Totalitarianism

7.4

enter the definition of the entity (totalitarianism as a political system) only as one of its component elements. Given the mind, the system is not given. So, the question that we still have to answer is: Does the noun totalitarianism enable us to identify a specific political system or regime? This is the question that Friedrich, more than anybody else, centrally addressed as early as 1953. In his contribution to Totalitarianism, Friedrich lists five requisites for a totalitarian system: (a) an official ideology; (b) a single mass party controlled by an oligarchy; (c) government monopoly of arms; (d) government monopoly of mass media; and (e) a terrorist police system. Shortly after, a sixth requisite was added: (f) a centrally directed economy. 41 It is symptomatic, if hardly surprising, that the initial list is varied even by those contributors who agree with Friedrich that totalitarianism is a specific political system characterized by several traits. 42 It is hardly surprising because we are scarcely doing better, for instance, with the characterizing traits of authoritarianism. 43 However, the thing to note is that the characteristics in question are displayed as a syndrome. That is to say that a totalitarian system results from their interconnectedness and mutual reinforcement. Now, the "syndrome solution" can hardly be credited with a convincing logical status. We have taken recourse to this medical analogy because we are unable to decide which are the defining, and which the contingent and accompanying, properties of a given concept. In the case in point, and for the sake of illustration, the government monopoly of arms is surely not a defining property (it is shared also by democracies), and an official ideology also goes to characterize authoritarian systems. Furthermore, when Friedrich and Brzezinski added to the list the feature of a centrally directed economy, they undermined their case. Nazism (let alone Italian fascism) did not have a centrally directed economy; so, if "totalitarianism" applies to both Hitler's and Stalin's regimes, we have a feature that is strongly present in one case (the Soviet one) and utterly absent in another. This will not do even within the loose context of a syndrome. But the attack on the notion of totalitarianism has gone far beyond these and similar feeblenesses. While feeblenesses can be cured, from the middle of the r96os onward an important mood has been of outright dismissal; the concept was not in need of mending but had, instead, to be rejected. 44 The dismissal of the concept is recommended on a variety of grounds. The two most recurrent arguments are that totalitarianism is a cold-war propaganda tool biased by polemical overtones, and that communist regimes are by now different among themselves and from what they were in the 1950s. The first argument leaves me cold. If abuse and propaganda exploitation were 1 95

WHAT DEMOCRACY IS NOT

7.4

a criterion for dismissing concepts, on such grounds democracy and fascism should be dismissed even more; and in the end we would be left, in politics, with almost no usable label. That the word totalitarianism is misused (as it doubtlessly is) cannot sustain the conclusion that it should not be used. It is also untrue to the facts that totalitarianism was coined during the cold war as an ideological weapon. "Friedrich and Brzezinski's clear intention was to summarize ... the substance of the literature devoted to fascism, national socialism, and bolshevism produced over a thirty-year peribd .... The concept 'totalitarianism' was not intended to serve 'counterideological' purposes ... it reflected antecedent work:' 45 Overall, the first argument attests, more than anything else, to the highly politicized and ideologized nature of the debate. The second argument has, instead, merits. As formulated by Michael Curtis, its gist is that while the concept of totalitarianism was useful at the outset, it loses its utility now that Nazism and fascism are defunct, that Russian communism is no longer Stalinist, and that the communist world is highly diversified. 46 It will be seen shortly that these points confront us with the nature and appropriate use of typological constructs; they also impose a revision of the initial formulations; but they cannot uphold the dismissal thesis. By the criterion of Curtis we would also be required to dismiss "feudalism" (which is extinct and was highly diversified), "economic liberism" (by now highly impure), and, even more to the point, "anarchy:' Indeed, take anarchy. Has any polity ever existed that any theory of anarchy would accept as an approximation - no matter how pale, imperfect, and insufficient-of its realization? The reply is certainly no. Yet we all use the concept (with profit, I believe), and no crusade has ever been launched against it. All in all, the reasons given for doing away with totalitarianism are insufficient. The good reason for rejecting totalitarianism is to replace it with a better concept-not with nothing. As with many other concepts (notably ideology), we first spend a decade or more in muddling them or rendering them meaningless and then come up with the bright thought that since they have been mishandled beyond repair, we cannot use them anymore. However, at some point we shall have to confront the fact that instead of letting concepts go to pieces, we had better keep them in good shape. Reverting to totalitarianism, I certainly do not deny its many original weaknesses nor the need to revisit the concept in the light of its new real-world denotations. In particular, it is true that unless the definition of totalitarianism is made more precise, we shall know less and less to which systems it applies.47 The preliminary issue remains whether we are heuristically better off by dismissing, rather than improving, the concept.

Totalitarianism

7.4

The major complicating factor with totalitarianism is, of course, that the word is particularly disliked by those to whom it may apply (just as, at the other end, democracy is liked by those to whom it does not apply). We shall have to live, I am afraid, with this factor. On more scholarly grounds it appears that the fuzziness of the concept is especially caused not only by applying it, indiscriminately, across time and space but also by extravagant connotations that bear little or no connection to the idea of totality. It would seem to follow that "totalitarianism" can be pinned down if, and only if, (a) the term is strictly allocated to a new, contemporary phenomenon; and (b) we remain within the orbit of the semantics, of what "totality" brings into focus. The modernity of totalitarianism is, to begin with, the modernity of the technology that jointly sustains and permits a totalitarian extension and penetration of power. The point that most traits of a totalitarian regime are "technologically conditioned" 48 cannot be disposed of with degreeism, with the degree slogan. Let me recall, in this connection, our earlier examination of the present-day vulnerability of public opinion, 49 which is, in turn, a function of technological factors. The passage from the written vehicle to the radio and from the radio to television consists of technological leaps, not increments. Similarly, that 6 billion human beings can be instantly killed en masse by radiation is not "more advanced" technology; it simply could not be done by chopping off heads, one by one, with swords. A concurrent, central aspect of the modernity of totalitarianism resides in the ideologization of politics. Since "ideology" has been diluted to a point of meaninglessness (another concept ripe for dismissal?), let the ideology in question be qualified as a political religion both with regard to its all-embracing scope and its millenarian grounding. So the modernity of totalitarianism parallels the modernity of a religionlike ideologization. Totalitarianism is both engendered and legitimized by the peculiar ideological temper of modern politics, which first manifested itself during the course of the French Revolution. The Puritans were saints in arms. The Jacobins and their subsequent reincarnations sanctified arms. As for my second recommendation, namely, that we should stay within the semantic orbit of the word, let me simply note that a semantic reason is invariably our best reason for selecting a given word. 50 If we say totalitarianism without feeling that it is a telling label, this is a more than sufficient reason for not saying it. The question thus turns on what the semantic projection of the term leads us to perceive. As we pass from "total" and totality to "totalitarianism;' what is implied is that we are alerted to, and wish to point at, the unprecedented intensity, pervasiveness, and penetration - both in breadth and depth-that political domination can assume. Hobbes's l.evia1 97

WHAT DEMOCRACY IS NOT

7.4

than is a baby monster when confronted with Orwell's, and the tyrannies of the past seem innocent and innocuous compared to what totalitarian dictatorships afe or can become. A term was needed to indicate this difference; and totalitarianism was and remains, I submit, effective and telling to this end. 51 Semantically, then, totalitarianism denotes the imprisonment of the whole of society within the state, an all-pervasive political domination over the extrapolitical life of man. As Finer describes this state of affairs: "The entire society is politicized; if private areas of life still survive, they do so ... on sufferance, as it were, from the government which at any time and for any reason may control, invade or take them over?' 52 When in Italy the saying was "Everything in the State, nothing outside the State, nothing against the State;' this was little more than a boast. But if the phrase everything in the state is taken seriously and applied to its limit by the instruments of coercion afforded by our technological know-how, we indeed find ourselves confronted with the "ultimate invasion of human privacy;' 53 with the destruction of everything that is spontaneous, independent, diversified, and autonomous in the life of human collectivities;54 in short, with a huge political garrison where a mass society has been incapsulated into the state. This is precisely what distinguishes totalitarianism from absolutism, authoritarianism, and other varieties of dictatorship: "The destruction of the line between state and society and the total politicization of society. . . . This is not merely a question of more or less political power. The difference is one of quality, not quantity?' 55 Trotzsky, who knew what he was talking about, said it vividly: "The totalitarian state goes far beyond Cesaro-Papism ... Stalin can justly say ... La societe cest moi."56 The next question might be: How can a garrison state of such kind be achieved and maintained? When Montesquieu produced his mapping of political forms, each of them was explained by a sustaining, specific constitutive "principle": civic virtue in democracy; moderation in aristocracy; honor in monarchy; and fear, crainte, in despotism. 57 That despotism is feared and characteristically governs by fear is hardly a disputed point. However, totalitarianism is an ultimate escalation of despotism, the strongest of all despotisms. Does it follow that it is also characterized by the highest degree of fear, that is, by terror? If reference is made to Hitler and Stalin, the answer is affirmative - and this goes to explain why government by terror was the keynote of Hannah Arendt's interpretation and of much of the literature of the 1950s. As communist regimes multiplied and stabilized themselves, it became apparent that terror was a variable rather than a defining element of totalitarian systems. This is not to predict that we have seen its end with Stalin but to note that closed polities legitimized by an ideological eschatology and based

Totalitarianism

7.4

upon total, monopolistic indoctrination need not have recourse to terror. 58 To be sure, shakeups of Thermidorian or Stalin-like heights of terror are always possible. It had all too hastily been proclaimed that totalitarianism had come to an end with the death of Stalin when Mao's so-called cultural revolution (in fact, a revolution against culture with some 400,000 dead casualties and an estimated 3 million people purged, harmed, or imprisoned) brought about an experience of massive fear of unprecedented scale. Even so, heights ofterror are likely to be intermittent and can be said to belong to the pathology, not the physiology, of totalitarianism. This having been granted, totalitarian regimes do remain fear-based regimes that handle with enormously high sanctions and brutal suppression whatever cannot be handled by means of ideological indoctrination. If terror is understood, therefore, as fear at its highest degree of pervasiveness, or potentiality, under this milder definition it can well be maintained that there is "more terror" in totalitarianism than in authoritarianism and, a fortiori, in pure and simple dictatorships. Yet, when all of the above is stated we are still left with a "nightmare" 59 more than with a definite and specific type of power structure. As has also been said, totalitarianism points to a tendency, or to a "temptation;' 60 but far less well to a definite type of state. The semantic projection of the word is evocative, or allusive, rather than descriptive. We are thus brought back to asking: How do we identify, empirically, a regime as totalitarian? As noted, the enumeration of its characteristics does not easily lend itself to agreement; moreover, the characteristics themselves are not easy to pin down. While it is unjust to forget that much of the same applies to many of the concepts of politics, nonetheless this is hardly a reason for not attempting to do something about it. Since the "syndrome solution" amounts to little more than a sheer listing of characteristics wrapped together by a medical analogy, in order to improve the empirical applicability of the concept one is wont to separate its defining characteristics (the characteristics that decide whether the category is to be used or not) from its shared characteristics, that is, from the ones that apply to dictatorships in general. Furthermore, and still more difficult, all these characteristics need to be firmed up in their degrees of intensity and assessed in their variance over time and across space. I cannot pursue this path- except for one quick observation on how beginnings relate to duration. Over the long haul totalitarianisms lose (as all new regimes do) their initial fervor, the impetus on which they were founded, and become, as the decades go by, routinized. Democracies too become routinized, and of democracies also one may say (as is said of totalitarianisms) that their routine is a loss of authenticity akin to a process of corruption and decay. Since routin1 99

WHAT DEMOCRACY IS NOT

7.4

ization is a constant feature of life and history, it cannot be a sufficient reason for asserting that a democracy no longer is a democracy or that a totalitarianism no longer is a totalitarianism. Routinization does bring about changes -but these are within-regime (within-class) variations. Outright terror, the permanent purge, 61 concentration and extermination camps-all of these are likely to be pre-routine characteristics. On the other hand, the routinization of totalitarianism may reinforce, rather than enfeeble, other traits. Kolakowski, for instance, singles out the total lie as the characterizing feature of totalitarianism. In his diagnosis the triumph of later Stalinism "consisted not simply in that virtually everything was either falsified and suppressed- statistics, historical events, current events ... books (occasionally Lenin's texts)-but that ... the borderline between what is [politically] 'correct' and what is 'true' seems really to have become blurred; by repeating the same absurdities time and again, they [the functionaries] began to believe or half-believe in them themselves:' 62 However that may be, the issue remains whether there is something that we can do with the "syndrome" with which the literature on totalitarianism leaves us. I believe there is. An uncomplicated manner of putting totalitarianism to efficient use is, in effect, to construe the concept, very deliberately, as an ideal type. Regardless of how Max Weber conceived the notion, there is, so to speak, one type of "ideal type" that is very fitting to the purpose at hand: the polar type. A polar type (of ideal type) is only the extreme or polar end of a continuum, whose logical function is simply to define the continuum. So conceived, totalitarianism simply consists of all the characteristics of oppressive regimes at their highest point of conceivable perfection. Thus, totalitarianism now stands at, and stands for, the end of a continuum (a discontinuous one, to be sure) whose opposite defining end is democracy (equally conceived as a polar type). This implies that no concrete system is expected to be "purely" totalitarian, just as no concrete democracy is expected to be a pure democracy. The assumption is merely that real-world totalitarianisms-just as realworld democracies-simply approximate, more or less closely, their respective polar parameters. Under this treatment, therefore, there is no point in accusing totalitarianism of being an overworked concept, for this is true, by definition, of all ideal types. The polar formulation of totalitarianism also beheads the familiar accusation that the category calls for a static approach. To the contrary, at each different point in time countries can be placed at different points along our continuum; and, over time, some countries will exit from the orbit of totalitarianism while other countries will enter it or be moved closer to the totalitarian end of the continuum. 200

Totalitarianism

7.4

Lest I convey the impression that we have made a near-miraculous move, let me immediately point out that problems do remain. By conceiving of totalitarianism as a polar type we no longer need to decide which are the defining characteristics nor to labor on whether all the selected properties have to be present. Nonetheless it is clear that the characteristics being selected as primary ones will alter the rank ordering of the concrete cases vis-a-vis their polar parameter. 63 Furthermore, the polar-type treatment does not afford cutoff points. Finally, the polar opposition between totalitarianism and democracy leaves a large in-between space of countries that are definitely neither totalitarian nor democratic. How is this vacant space to be filled? It has generally been filled by the concept of authoritarianism. But here we again have an issue. Authors who dismiss the notion of totalitarianism sustain their case by reclassifying the formerly called (or wrongly called) totalitarian systems as being, at least nowadays, "authoritarian:' When a concrete case is reclassified, the reasons for doing so have to be factual- and this is an empirical assessment. But moving the cases from one class to another does not affect, per se, the system of classification. Had tyrannies ceased to exist in the ancient Greek cities, Aristotle's system of classification would have remained intact. That no cases fall, at a given point in time, into the box of the "rule of the one" does not entail that the box itself disappears. So, even if all totalitarianisms were reclassifiable-as a matter of empirical fact-as authoritarianisms, it does not necessarily follow- as a matter of theory-that the class "totalitarianism" loses its reason for being. If or when the distinction between totalitarianism and authoritarianism is under attack, the best reason given for the attack is that it might "provide an apologia for authoritarian politics:' 64 While this is largely an argument ad hominem, it brings out the important point that it is wrong to construe the difference between totalitarianism and authoritarianism in terms of greater versus lesser actual misdeeds. Indeed, a "simple dictator" can do worse, on grounds of cruelty, killings, torture, police brutality, and similar inhumane characteristics, than a totalitarian dictatorship. The crucial difference between totalitarianism and authoritarianism is not in what they actually do (they do very different things over time and across countries) but in their respective potentialities. Remember, "totalitarianism" does not convey, semantically, the idea of greater coercion but of greater reach. To be sure, equal coercion in more respects, in more areas, adds up to greater coercion-but this is an inference. An inference that can be countered by observing that the more a totalitarianism succeeds in the pervasiveness of its control, the less it needs to have recourse to naked coercion. The valid reason for doing away with the distinction between totalitarianism and authoritarianism is to show that this is a distinction that carries no 201

WHAT DEMOCRACY IS NOT

7.4

differences-and this has not been shown. To say that Mussolini and Stalin, Salazar and Hitler, Franco and Cambodia's Pol Pot, Nasser and Albania's Enver Hoxa, all equally impersonate(d) a same type of political system is not a gain in, but a loss of, analytic clarity. In particular, it does make a difference whether dictatorial regimes have, or have not, a succession problem; and while simple and authoritarian dictatorships seldom outlast the life span of the person-dictator, totalitarian dictatorships have in fact been more durable and are capable of handling succession crises. It also make~ a great deal of difference whether and to what extent subsystem autonomy and subgroup independencies (in areas other than politics) are permitted or tolerated; and while nontotalitarian dictatorships are left to pursue, vis-a-vis outer groups, exclusionary policies, totalitarian dictatorships have both the force and the motivation for pursuing the utter destruction of all subsystems. Finally, it does make an enormous difference whether dictatorial regimes have, or have not, ultimate legitimacy claims. In this latter respect the divide does not lie in having an "official ideology" (even African dictators have learned, by now, how to concoct one) but in the nature and scope of such ideology. Nontotalitarian dictatorships are such precisely in that they fail to provide a religion-like official ideology. A simple dictator is where he is simply by having the force to be and stay there. An authoritarian dictator does proclaim an official ideology, but either on limited grounds or with little belief penetration. Conversely, totalitarian dictatorships rest, in their justification, on some philosophy of human nature related to a scheme of perfecting the world, of redeeming it from past and present sins, and thus of serving an ultimate higher and indeed absolute good. This does not require, let it be underscored, intensity of belief. In the long run, religions only require a few true believers and may not need fanatics at all. When I say, then, religious-like ideology, I account for the routinization of ideological fervor. What remains, and may long last, is the "perceiving of the world" resulting from the life-long indoctrination and message control that totalitarianisms secure for themselves. In quick summary, when totalitarianism is construed as a polar type (an ideal type of the polar variety), it does contribute, I submit, to the mapping of political systems. On the other hand, the theory of totalitarianism is all the more convincing the more it remains within the semantic orbit of the term it employs. If so, the major defining characteristic of the totalitarian polity is total penetration and extension. On this view, characterizations such as a "centrally directed economy" cannot be entered as defining characteristics but as a measure or indicator of the extensiveness of a totalitarian reach. Similarly, on this view what matters is not the official imposition of one ideology, nor the intensity with which it is held, but its all-remaking, all-inclusive, and 202

Dictatorship and Autocracy 7.5 millenaristic nature. The feeble point of the construct is that it is not structural and does not lend itself to structural underpinning. From this vantage point my sense, or response, is to transform totalitarianism as an object concept (the designator of a system) into "totalitarian" as a predicate, as a property concept. In such a case, the structure is dictatorship, and one of its variants (its most extreme, if not ideal-typic one) is the totalitarian dictatorship. However, that "totalitarianism" is not a good empirical-structural type does not entail in the least that the concept is dismissible.

7.5

Dictatorship and Autocracy

We may now review the remaining set of possible opposites of democracy and/or of the ways of indicating its boundaries. To this end I shall quickly dispose of tyranny and despotism; I shall dwell instead on dictatorship and, finally, on autocracy. Tyranny and despotism are concepts that survive in our present-day vocabulary mostly as labels. Tyranny, it is true, has a long and elaborate record in the history of political thought. For the Greeks it was only prior to the fifth century B.C. that tyrant and king (basileus) were both used to signify the rule of the one. From then onward tyranny came to denote the lawless, illegitimate, and degenerate form of single-person rule. Since Greek times, the term was used continuously, and was further refined during the Middle Ages. It is sufficient to recall the distinction of Bartolo da Sassoferrato and Coluccio Salutati between tyranny quoad exercitium (relating to the way of exercising power) and tyranny ex defectu tituli (relating to the illegitimate or violent acquisition of power) that was followed by the literature of the Monarchomachs on the right to kill the tyrant. 65 Nevertheless, it is hard to see how these distinctions, and for that matter all the Renaissance and seventeenth-century literature on the subject, apply to modern conditions and the present-day camouflages. Our plebiscitary dictators can hardly be considered tyrants ex defectu tituli, and our criteria for spelling out a tyranny exercitio are too different from the Greek and the Renaissance moral and natural law criteria. On the other hand, the notion of despotism has never been elaborated in depth. 66 The reason for this is that, until Montesquieu, despotism did not enter a classification of the systems of government. For the Greeks, despotism was the system of the "barbarians" eminently represented, in their perception, by the Persian empire. In their view, Asians were slaves by nature and hence naturally submitted to despotism. On this account, despotism had nothing to share with tyranny. Tyrants ruled over Greeks (i.e., a demos capable of freedom) and were thus short-lived. Despots ruled unchallenged and un203

WHAT DEMOCRACY IS NOT

7.5

endingly over people incapable of being free. Therefore despotism has long remained a label "for others;' As for absolutism, it has been seen that the term does not afford much mileage. All in all, these terms basically yield the general idea of a displeasing and oppressive political power. If we want to know more about the tyrannical, despotic, and absolute ways of exercising power- and especially about the form of state that permits this, exercitiumwe are well advised to refer to the more modern concepts. Dictatorship is actually the term that has come to repla~e tyranny; and this because we have come to use it very much, in the last fifty years, in the meaning formerly expressed by tyranny. In many ways, the transformation of the idea of dictatorship parallels, in its logic, the transformation of the idea of absolutism. Absolutism could not be understood in a defective sense until the idea of potestas amplissima (of discretionary, i.e., absolute power) was confronted with the counter-idea of the princeps legibus obligatus, the prince bound by law (the constitutional state). This is why the negative meaning of absolutism is not fully affirmed until the eighteenth century; and never without exceptions, as is evidenced by the notion of enlightened absolutism. The same applies to dictatorship. It is only after an adequate and successful experience of "government by consent" that we perceive that dictare, to dictate (the root meaning of the word), can stand for a distinct type of political system; and it is only at this point that dictatorship can take on a negative connotation - as the negation of government by consent. The Roman dictator was a short-lived (six months) extraordinary magistracy strictly designed to meet military emergencies. The institution degenerated in the third century B.C. and ultimately with Caesar, thus allowing ambivalent memories of it. Yet both Machiavelli and Rousseau praised the Roman dictatorship. That the term maintained, until a century or so ago, its association with the Roman magistracy is well attested by the fact that during the course of the Italian Risorgimento, Farini in 1859 (in Emilia) and Garibaldi in 1860 (in Sicily) proclaimed themselves "dictators;' thus ostensibly implying that the word had a favorable aura (and thrill). The first important departure from this meaning was, apparently, in Blanqui, who presumably inspired Marx. But aside from the fact that Marx himself hardly used "dictatorship of the proletariat;' his usage implied only, in utter simplicity, the idea of recourse to force. 67 Aside, then, from scanty references to the "Bonapartist dictatorship" of Louis Napoleon, only with fascism did "dictatorship" come to be definitely perceived as a distinct and sui generis type of state. Thus it is not surprising that a theory of dictatorship (in the post-Roman sense) arises only in the 1920s and 1930s. What is more surprising is how little it has progressed since. 68 204

Dictatorship and Autocracy 7.5 Up to the present moment there is no satisfactory theory of dictatorship that establishes its varieties in relation to a specific kind of power seizure and legitimation, to the problem of duration and succession (indeed the soft underbelly of dictatorships), and, in sum, to a definite form and exercitium of state power. Perhaps it is also the case that the theory of dictatorship is hindered by the Marxist sidetracking of the concept into the societal sphere- as when Western democracies are declared bourgeois and capitalistic dictatorships. 69 To be sure, what Marxists mean by "dictatorship" is what Gramsci more appropriately called "hegemony:' Yet, had there been a theory of dictatorship worthy of the name, it would have been less easy to propagate the self-contradictory notion of acephalous dictatorships ruled by such amorphous entities as the bourgeoisie or the proletariat. Notwithstanding the fact that our theory of dictatorship is still highly unsettled and incomplete, the term has the merit of lending itself nicely to structural elaboration. Simply put, a dictatorship is a nonconstitutional rule, either because the rulers make a sham of a preexisting constitution or because they write a constitution that empowers them, in practice, to do whatever they wish. 7 From this vantage point, the structure of dictatorships is, and can be neatly worked out as, the reverse of constitutional structures. It is not simply that we revert from a princeps obligated by laws to a prince legibus solutus, unrestrained by laws. By now the organizational chart of any state is a highly complex one; and this implies that even dictatorships display highly complex power structures. To wit, a dictatorship no longer is a party-less state: Instead of prohibiting all parties, it generally establishes itself as a singleparty state. But there are many uncertainties, today, as to where the singleparty structure ends (or begins). 71 So, the structural underpinning of dictatorships remains very much in need of implementation. Let it also be remembered that Marxism has popularized the loose notion of a "class dictatorship;' in the wake of which we do hear talk of democracies as dictatorships. Since we live in a time in which no absurdity is left unsaid, misuses of the term need not discourage us from using dictatorship as a contradictory of democracy. Yet, and once again, that democracy and dictatorship are mutually exclusive political forms is not an assertion beyond assault. Is there no term at all, then, that stands as an undisputed and hardly disputable good opposite of democracy? I submit that "autocracy" is such a term. The advantage of the concept of autocracy over the ones reviewed thus far is that it points directly to a constitutive "principle" (in Montesquieu's sense): the method of creation of the power holders with respect to the legitimacy basis of power. Therefore, when the dichotomy, or the alternative, is "democracy or autocracy;' the symmetry is very neat and hard to fudge. This

°

205

WHAT DEMOCRACY IS NOT

7.5

is so, first and foremost, because we now have a definiens (autocracy) that is clearer, simpler, and far easier to handle than the definiendum (democracy). If the condition that makes for a good definiens is that it lessens the obscurity or complexity of the definiendum, of the term that is to be defined, surely autocracy well satisfies this condition. Let, then, the thesis be that democracy is non-autocracy, the exact contrary and indeed the precise contradictory of autocracy. This means that democracy denotes a political system characterized by the refusal of personalized power, of a power over the citizens that belongs to somebody. Power is nobody's "property;' More specifically, democracy stands for a system that hinges on the principle that no one can proclaim himself ruler, that no one can hold power irrevocably in his own name. Precisely because the autocratic principle is repudiated, the democratic axiom is that man's power over man can only be granted by others-and this always and only on a revocable basis (for otherwise the grantors of power would, at the same time, renounce their power). Henceforth, leaders must result from a free, unfettered designation of those who are to be led. That is equally to say that whenever this empowering others to designate you is being tampered with or counterfeited- either because dissent is impeded or alternatives are not offered-democracy is killed at its inception. Whatever else democracy may be, or should be, if it is not this-the exact reversal of autocracy-it is not. From the non-autocracy perspective we also understand better how the power of the people is substantiated and, as a preliminary condition, preserved by liberal-democratic institutions. The constitutional devices by which the exercise of power is made to hinge on its investiture, and by which public officials are bounded by preordained juridical structures, are the sine qua non conditions that allow for the replacing of leaders, for limiting their time in office, for holding them responsible to the people, and for averting abuses of power. Since defining democracy is a central concern in this book, let the above be restated as a definition, to wit: Democracy is a system in which no one can choose himself, no one can invest himself with the power to rule and, therefore, no one can arrogate to himself unconditional and unlimited power. To be sure, this is a definition that states the characteristics of democracy in the negative. It cannot be, in any sense, an exhaustive definition. Within this limitation, however, it displays a peculiar strength. For one thing, it sets forth defining properties (not contingent or accompanying properties). That is to say that any social or political system based on different principles is not, by definition, a democracy. In the second place, our definition seems to lie at the intersection at which, and indeed before which, the description and 206

Notes the prescription depart from one another. When we say that only others can empower you, this is both a descriptive statement (democracies are in fact constructed on this principle) and a prescriptive declaration (democracies should be so constructed). We are left to ask how the a contrario definition - democracy as nonautocracy- relates, on the one side, to democracy defined as an electoral polyarchy (the competitive-descriptive theory) and, on the other, defined prescriptively as a selective polyarchy. I understand the three definitions in question as being not only consistent with one another but as cumulating with one another. In effect, the a contrario focus sustains the importance of the electoral-competitive definition, which may be viewed as its positive and procedural implementation. Likewise, if democracy is not auto-cracy, it still is cracy; and this points to the fact that our ideals should not be misdirected in the quest for acephaly, for leaderlessness. Let it be added that the ex adverso path need not confine itself, as I have done here, to the strict orbit of the signification of "non-autocracy?' For instance, the implications of the autocratic principle are that power will be uncontrolled, unlimited, and as concentrated (monocratic) as circumstances (other than the will of the power addressees) permit. By the same token, and conversely, the implications of non-autocracy (i.e., democracy) are that power will be limited, controlled, and as polycratic as required for its taming. What is not cannot ad~quately and conclusively tell us what is. Yet, when democracy is defined in contradistinction to autocracy, we immediately obtain a precise demarcation of where democracy begins (or ends) and a clear focus on where its positive identity rests.

Notes 1. See "Political Democracy as a Property of Political Institutions;' American Political Science Review, March 1971. 2. See "Democracy: Characteristics Included and Excluded;' The Monist, January 1 97I. 3. An extreme but telling illustration of the arbitrariness that can enter this transformation is the following operationalization of party "competition" by A. Przeworski and J. Sprague: "a knock at the door by a party voter" ("Concepts in Search of Explicit Formulation: A Study in Measurement;' Midwest Journal of Political Science, May 1971, p. 208 and passim.) 4. This discussion is pursued in G. Sanori, F.W. Riggs, and H. Teune, The Tower of Babel (Pittsburgh: International Studies Association, University of Pittsburgh, 1975), pp. 22-25; and also in my Parties and Party Systems, pp. 295-98. 5. The Roman dictatorship was an extraordinary magistracy limited in time and functions, not a political form. Therefore, the meaning of dictatorship that concerns us is a recent one. See section 5, below.

207

WHAT DEMOCRACY IS NOT

6. I largely follow the interpretation of H. Arendt, "What Was Authority?" in

C. Friedrich, ed., Authority (Cambridge, Mass.: Harvard University Press, 1958). The essay is reproduced in revised form in Between Past and Future (New York: Meridian Books, 1968), chap. 3. On the etymological point see also S. de Grazia, "What Authority Is Not;' American Political Science Review, June 1959, pp. 322-23. 7. For the history of the idea of authority, see G. Quadri, II Problema dell' Autorita, 2 vols. (Firenze: La Nuova Italia, 1940); and T. Eschenburg, Uber Autoritat (Frankfurt: Suhrkamp, 1965). B. de Jouvenel, Du Pouvoir (Geneve: Editions Cheval Aile, 1947), is less to the point, but insightful. Very relevant essays are in the symposium ed. by Friedrich, Authority; and in L. Bryson, L. Finkelstein, R.M. Maciver, and R. McKeon, eds., Freedom and Authority in Our Time (New York: Harper, 1953). See also F. Bourricaud, Esquisse d'une Theorie de l' Authorite, 2d rev. ed. (Paris: Pion, 1969); M. Stoppino, Le Forme de! Potere (Napoli: Guida, 1974), chap. 3; and R.E. Flathman, The Practice of Political Authority (Chicago: University of Chicago Press, 1980). 8. Domination is the rendering of R. Bendix; rulership of C.J. Friedrich. The Italian translators of Weber have generally settled for "power" (even though Herrschaft is less powerful, so to speak, than Macht). To be sure, when Herrschaft is conveyed as authority, Weber's translators have in mind "legitimate Herrschaft," i.e., a subclass of rte general category. However, that Weber misleads his translators does not salvage the mistranslation. Legitimate authority conforming to law is, in German, Regierung. 9. Power and Society, p. 133. They were inspired and preceded by Merriam. Note that the distinction between formal (legal) and real (de facto) power already served the purpose that Lasswell and Kaplan had in mind. 10. See, for all, R. Bierstedt, in M. Berger et al., eds., Freedom and Control in Modern Society (New York: Octagon Books, 1954). His essay "The Problem of Authority" illustrates how a hitherto well-circumscribed issue can become a field of vagaries. 1r. In Bulletin International des Sciences Socia/es 4 (1955): 718. 12. Most current definitions of power are broader, so much so that they could be read as definitions of "influence:' Remember, though, that my focus is, in the main, on state power, not on power in general. See M. Stoppino, Potere Politico e Stato (Milano: Giuffre, 1968). 13. The classic reference is T.H. Green, Lectures on the Principles of Political Obligation (London, 1895). See also J. Plamenatz, Consent, Freedom and Political Obligation; and E. Barker, Principles of Social and Political Theory, chap. 5. But see, for the bibliography, chapter 5, n. 8, herein. 14. A classic forewarning on the disappearance of "the sense of legitimacy;' and on the crisis thus resulting, is G. Ferrero, The Principles of Power: The Great Political Crises of History (New York: Putnam, 1942). 15. To be sure, the distinction is analytical. In most concrete instances individuals comply to orders or appeals for mixed and variously weighed motives (and, to begin with, on grounds of routine). But to grant that pure power or pure authority relations are rare occurrences does not detract from the analytic import of the distinction. 16. It is immaterial to my argument whether the broader category (the genus) is declared to be power or influence. My point is only that a loss of analytic clarity results from assimilating authority to either. 17. The latter is the rendering of Reinhard Bendix, Kings or People (Berkeley: University of California Press, 1980), p. 669. 208

Notes 18. The Practice of Political Authority, p. 191 and passim. Here Flathman contradicts R. Paul Wolff's argument that to accept authority is to renounce autonomy. 19. This is the definition of H. Eckstein and T. Gurr, Patterns of Authority, p. 27. Despite differences, their work follows Easton's track (see their pp. 33-35). 20. See esp. D. Easton, The Political System (New York: Knopf, 1953), chap. 5. 21. A Systems Analysis of Political Life, p. 208. Since the example is given in a footnote, I suspect that Easton himself feels uncomfortable with the ultimate, but inevitable, implications of his focus. To be sure, the thiefs authority is "illegitimate;' i.e., perceived as such by the victim because it is coercive authority. The objection remains: Since "coercion" goes with "power;' to involve authority in the matter is to spoil the concept. 22. The classic work is T. W. Adorno et al., The Authoritarian Personality (New York: Harper, 1950). An important follow-up is R. Christie and M. Jahoda, eds., Studies in the Scope and Method of "the Authoritarian Personality" (Glencoe: Free Press, 1954). A further caveat bears on the difficulty of drawing inferences from the individual to aggregates and, ultimately, to the polity. As F.I. Greenstein remarks in his judicious overview of this literature, "authoritarianism" applies ro many different units and levels of research, and "authoritarianism at any one of these levels is not necessarily accompanied by authoritarianism at other levels" (Personality and Politics [Chicago: Markham, 1969], p. 98). 23. In C.J. Friedrich, ed., Totalitarianism (Cambridge, Mass.: Harvard University Press, 1954), p. 274. The point is elaborated in his contribution to Authority, pp. 28££. 24. Du Pouvoir, p. 316. Actually, de Jouvenel precedes Talmon (seen. 40, below) in coining the expression "totalitarian democracy." Schapiro equally stresses that totalitarianism is more closely linked to democracy than any other previous despotism (in Leonard Schapiro, ed., Political Opposition in One-Party States [London: Macmillan, 1972], pp. 275-76). 25. See R. De Mattei, "Assolutismo;' Enciclopedia de/ Diritto (Milano: Giuffre, 1958), 3:917ff. On the history of the origins of the absolute state, see esp. Perry Anderson, Lineages of the Absolutist State (London: New Left Books, 1975). 26. Lasswell and Kaplan distinguish between centralization and concentration of power, and rightly point out that the latter, not the former, is a characteristic of despotic rule (Power and Society, pp. 224-25). Accordingly, "overcentralization" may indicate the passing from centralization to concentration. 27. All these points are developed in chapters 13 and 15, herein. 28. See chapter 4, section 5. 29. The theory of totalitarianism in the 1930s, eminently represented by Giovanni Gentile, Carl Schmitt, and Alfred Rosenberg, is covered by Martin Janicke, Totalitiire Herrschaft: Anatomie eines Politischen Begriffes (Berlin: Duncker & Humblot, 1971). 30. A judicious appraisal of the facts (not of the words) is Alberto Acquarone, L'Organizzazione dello Stato Totalitario (Torino: Einaudi, 1965), whose final chapter is reprinted in E.A. Menze, ed., Totalitarianism Reconsidered (Port Washington: Kennikat Press, 1981). An illuminating collection of interpretations is C. Casucci, ed., II Fascismo: Antologia di Scritti Critici (Bologna: II Mulino, 1982). See also Renzo de Felice, ed., II Fascismo e i Partiti Italiani (Bologna: Cappelli, 1966), and Le Interpretazioni de! Fascismo (Bari: Laterza, 1974). In fact, Mussolini closed the conflict with the Vatican (granting to the Catholic church an interference in Italian affairs that the

209

WHAT DEMOCRACY IS NOT

liberal-democratic regime had considered unacceptable), maintained the Monarchy, and never seriously attempted to infiltrate the army (whose career officers remained loyal, throughout the fascist regime, to the symbols of the Monarchy). If such a state of affairs is deemed totalitarian, then there is no point in employing the notion. 31. See H. Arendt, The Origins of Totalitarianism (I95I), rev. eds. (New York: Meridian Books, I958 and I966); and C.J. Friedrich and Z. Brzezinski, Totalitarian Dictatorship and Autocracy (Cambridge, Mass.: Harvard University P,ress, I956). Reference will be made to this edition, not to the one revised by Friedrich in I965. Arendt stated very clearly that fascism was not a totalitarianism; and while her treatment of Stalinism is feeble, the one of Nazism remains (if overly demonological) very insightful. The Friedrich-Brzezinski volume is far superior on Stalinism, but includes fascism (albeit peripherally) under the totalitarian categorization. While this inclusion establishes an unfortunate precedent, the book is basically on the Nazi and Soviet totalitarianisms. 32. See n. 23, above. Totalitarianism was published in 1954. 33. The Democratic and the Authoritarian State (Glencoe: Free Press, I957), pp. 246-47. 34. Political Power and the Governmental Process (Chicago: University of Chicago Press, I957), pp. 59-60. In the light of this consideration one can hardly accept Karl Popper's understanding of Plato as "totalitarian:' See The Open Society and Its Enemies (1945) rev. ed. (London: Routledge & Kegan Paul, I952), 1:87 and passim. 35. See K.A. Wittfogel, Oriental Despotism: A Comparative Study of Total Power (New Haven: Yale University Press, I957). My caveat does not detract from the stature

of this classic work. 36. R. Aron, Democratie et Totalitarisme (Paris: Gallimard, I965), p. 3I9. 37. I am not persuaded, therefore, by Barrington Moore's excursions into the "totalitarian elements in pre-industrial societies:' In Political Power and Social Theory, chap. 2, his referents are traditional India, China under the Ch'in dynasty, and Calvin's theocracy in Geneva. In his Social Origins of Dictatorship and Democracy (Boston: Beacon Press, 1966) Moore adds to the list Meiji Japan and a set of developing regimes. The analysis, brilliant as it is, lacks theoretical power. 38. The Soviet Impact on the Western World (New York: Macmillan, 1949), p. no. This pamphlet provides the underlying, central thread of Carr's subsequent monumental History of Soviet Russia (New York: Macmillan, 1951-64), in seven volumes. 39. The symmetrically opposite thesis, namely, that it is democracy that is as old as the world and is the "natural" result for the human being, is developed at length by Jean Baechler, Democraties (Paris: Calmann-Levy, 1985). 40. See his classic The Origins of Totalitarian Democracy (London: Secker & Warburg, 1952). Note, however, that Talmon's point that "the totalitarian democratic school is based upon the assumption of a sole and exclusive truth in politics" (p. 1) has little in common with Carr's. 41. See pp. 52-53. The last requisite was entered in Friedrich's volume with Brzezinski, Totalitarian Dictatorship and Autocracy. 42. See esp. the contributions (in Totalitarianism) of A. Inkeles and K. Deutsch. The enumeration of characteristics lends itself not only to additions but also to reductions, as in G. Almond's synthesis: "Totalitarianism is tyranny with a rational bureaucracy, a monopoly of the modern technology of communication, and a monopoly 210

Notes of the modern technology of violence" (in H. Eulau, S. Eldersveld, M. Janowitz, eds., Political Behavior-A Reader [Glencoe: Free Press, 1956], p. 39). 43. Thus far the major, serious attempt at developing the theory and the characterizing features of authoritarianism is the one of Juan Linz. See esp. his "Totalitarian and Authoritarian Regimes;' in Polsby and Greenstein, Handbook of Political Science, vol. 3. 44. See esp. Alfred Meyer, The Soviet Political System (New York: Random House, 1965); H.J. Spiro, "Totalitarianism;' International Encyclopedia of the Social Sciences 14:n2 (where he questions the scientific usefulness of the concept); B.R. Barber and M. Curtis, in Totalitarianism in Perspective: Three Views (New York: Praeger, 1969). For a strong, overall criticism of this literature, see D. Fisichella, Analisi de/ Totalitarismo (Messina: D½nna, 1976), chap. 1 and Conclusion. See also the discussion of Linz, Handbook, 3:241-47; and G. Hermet et al., eds., Totalitarismes (Paris: Economica, 1985). 45. A. James Gregor, in Menze, Totalitarianism Reconsidered, p. 143. 46. In Totalitarianism in Perspective, pp. 54, 63, 93-94, 105. 47. This is the constructive line of criticism of F.J. Fleron. See esp. "Soviet Area

Studies and the Social Sciences: Some Methodological Problems in Soviet Studies;' Soviet Studies, January 1968. The constructive, as against the dismissal, line of criticism is also taken by Leonard Schapiro, Totalitarianism (New York: Praeger, 1972); and by a number of contributors to Totalitarianism Reconsidered, notably Karl D. Bracher, A.J. Gregor and, somewhat more surprisingly, Ernst Nolte. Of Bracher see, more fully, Zeitgeschichtliche Kontroversen: Um Fascismus, Totalitarismus, Democratie (Miinchen: Piper, 1976). 48. Totalitarian Dictatorship and Autocracy, pp. 10-13. 49. See chapter 5, esp. section 4. 50. It should be understood that I use "semantics" in the strong sense of the word, that is, to mean that words are in themselves "suggestive" and thought-molding: they are carriers of a way of perceiving and conceiving the world. I explicate this stance esp. in Sartori, ed., Social Science Concepts: A Systematic Analysis, pp. 15-23. 51. Perhaps totalitarianism is the word that Tocqueville was seeking when, in one of his prophetic passages, he wrote: "I believe that the kind of oppression which threatens democratic nations will not resemble any other form previously experienced in the world; our contemporaries would not be able to find anything like it in their memories. Try as I may I cannot find an expression that can reproduce the exact idea I have formed of it; the ancient words despotism and tyranny are entirely inadequate. It is something new; I must therefore attempt to define it, for I cannot name it" (De la Democratie en Amerique, vol. II, pt. IV, chap. 4; p. 2.34, in 1951 ed.). 52. S.E. Finer, Comparative Government (London: Allen Lane, Penguin Press, 1 970), p. 75. 53. Nisbet, The Quest for Community, p. 202. 54. See Z. Brzezinski, Ideology and Power in Soviet Politics (New York: Praeger, 1962), pp. 15-20. 55. F. Neumann, The Democratic and the Authoritarian State, p. 245. 56. Stalin (New York: Grosset & Dunlop, 1946), p. 421. 57. See L'Esprit des Lois, vol. 1, bk. m. 58. This is explained in my earlier discussion of consensus, public opinion, and 2II

WHAT DEMOCRACY IS NOT

propaganda, chapter 5, sections 1-4. As Jeane Kirkpatrick points out: "If a chief task of the totalitarian is to translate ideology into culture . . . as new norms are internalized, new beliefs accepted, new habits established, the need to coerce conformity through crash 'thought reform' programs and punishment of dissenters should decline;' See Dictatorships and Double Standards (New York: Simon & Schuster, 1982), pp. n4-15. 59. This is George F. Kennan's perception: "When I try to picture totalitarianism to myself as a general phenomenon what comes into my mind most prominently is neither the Soviet picture nor the Nazi picture as I have known them in the flesh, but rather the fictional and symbolic images created by such people as Orwell or Kafka or Koestler or the early Soviet satirists. The purest expression of the phenomenon, in other words, seems to me to have been rendered not in its physical reality but in its power as a dream, or a nightmare" (in Totalitarianism, pp. 19-20). 60. This is notably the perspective of Revel's The Totalitarian Temptation. 61. See esp. Z. Brzezinski, The Permanent Purge in Soviet Totalitarianism (Cambridge, Mass.: Harvard University Press, 1956). 62.. "Totalitarianism and the Virtue of the Lie;' in Irwing Howe, ed., 1984 Revisited: Totalitarianism in Our Century (New York: Harper & Row, 1983), p. 129. 63. As mentioned above, for Kolakowski the primary characteristic would appear to be the "total lie;' while for other authors (of the revisionist, not the dismissal persuasion) it may well be a different one. E.g., in Djilas's view the "very essence" of the current Soviet rule is that the "triumph of the party bureaucracy occurs simultaneously with the transformation of the state into an ideological military empire" (in 1984 Revisited, pp. 138-39). As different characterizations are deemed to be central, different rankings willl follow. 64. Michael Walzer, in 1984 Revisited, p. 121. Walzer does note that "it was never the intention of the theories of the fifties to celebrate authoritarianism" (p. 108); but "contemporary conservatives" are not equally innocent. In their argument, "Since authoritarian rulers aren't even touched by messianism ... since all they want is to hold on to their power ... they don't produce the terrible upheavals of totalitarian politics" (pp. 108-9). 65. See G. Nicoletti, Sul Diritto al/a Resistenza (Milano: Giuffre, 1960). The constitutional complexity of this development is highlighted by Julian H. Franklin in the Introduction to Constitutionalism and Resistance in the Sixteenth Century (New York: Pegasus, 1969), and in John Locke and the Theory of Sovereignty (Cambridge: Cambridge University Press, 1978). 66. The best analysis of the concept is R. Koebner, "Despot and Despotism: Vicissitudes of a Political Term;' Journal of the Warburg and Courtauld Institutes, December 1951, pp. 275-302. 67. The Marxian meaning in analyzed in chapter 15, sections 2 and 3. 68. The first major work is Carl Schmitt, Die Diktatur (1921, rev. ed., Miinchen, 1928), largely an advocacy and also largely juridical in its orientation. It was followed, especially in Italy, by a strictly juridical literature. From the opposite camp, and in a historical perspective, Alfred Cobban, Dictatorship: Its History and Theory (London: Cape, 1939) remains a standard text. An unfinished, perceptive typology of dictatorships was left by Franz Neumann, "Notes on the Theory of Dictatorship;' and was included in his The Democratic and Authoritarian State. The pamphlet of M. 212

Notes Duverger, De la Dictature (Paris: Julliard, 1961), is hasty. We are, more or less, left at that. See my "Appunti per una Teoria Generale della Dittatura;' in K. von Beyme, ed., Theory and Politics. Subsequent developments basically focus on military dictatorships and hardly address the theory of dictatorship as such. 69. Marcuse repeats the operation with the concept of totalitarianism, diluting and extending it to the societal manipulation of the advanced industrial societies (see One-Dimensional Man [Boston: Beacon Press, 1964], p. 3). Thus, for Marcuse, the Gulag concentrationary world and being manipulated into "consuming" fall under the same category. 70. The constitutional meaning of "constitution" is clarified in chapter II, esp. section 7. See also G. Sartori, "Constitutionalism: A Preliminary Discussion;' American Political Science Review, December 1962. 71. This is a point discussed in my Parties and Party Systems, chaps. 2 and 7, where I also pursue (pp. 225-28) a typological analysis of dictatorships (totalitarian, authoritarian, and pragmatic) across other characteristics.

2r3

8. A Decision-Making Theory of Democracy We can hope to understand existing democracies only to the degree that we are prepared to take into account the dimensions of the intensity problem and the complexities it introduces. -Willmoore Kendall

8.r

The Nature of Political Decisions

The preceding chapters have been largely devoted to conceptual housecleaning. I now wish to propose something less than a "new theory" of democracy and yet take a fresh look into the matter from a decision-making perspective. Let me immediately begin by distinguishing between four kinds of decisions: (a) individual; (b) group; (c) collective; and (d) collectivized.1 Individual decisions are taken by each individual for himself, regardless of whether he is inner or outer-directed. Group decisions imply that decisions are taken by a concrete group, that is, face-to-face, interacting individuals who meaningfully partake in the making of such decisions. Collective decisions are hardly amenable to a precise definition; generally they are understood to mean decisions taken by "the many:' Contrasted (as my distinction implies) to group decisions, collective decisions assume a large body that does not and cannot perform-on account of its size-as concrete groups do. It should also be underscored that a collective decision should not be confused with a collective preference; the former need not generate the latter, to wit, an outcome that meaningfully expresses the social preference. We then have collectivized decisions. Collective and collectivized decisions may be said to share the property of not being, in any meaningful sense, individual decisions. Even so, collectivized decisions are very different from all the other kinds. Individual, group, and collective decisions all make reference to a subject, to who makes the decisions. Collectivized decisions are, instead, decisions that apply to, and are enforced on, a collectivity regardless of whether they are taken by the one, the few, or the many. The defining criterion no longer is who makes the decisions, but their scope: Whoever does the deciding, decides for all. 2

The Nature of Political Decisions

8.1

The notion of collectivized decisions permits, to begin with, the assertion that politics consists of collectivized decisions. 3 Note that collective and collectivized decisions correspond to one another only when the universe issuing the decisions coincides with the universe receiving them. This coinciding is of great theoretical interest and may in fact occur. It occurs less and less, however, as the size of political units increases. At the macrolevel, therefore, it can be said that politics ultimately consists of decisions (enacted decisions) that are removed from the competence of each individual as such and are made by somebody for someone else. This does not imply in the least that a collectivized decision is also a decision on behalf of its addressees-this may, but equally may not, be the case. The deciders decide for all only in the sense that their deciding falls on everybody's head. Naturally, while all the decisions of a political nature are collectivized decisions, the obverse is not true: Not all collectivized decisions are political. For instance, when we speak of economic power we again refer to coHectivized decisions, to the fact that somebody (the capitalist, the corporation, etc.) takes decisions for, and imposes them upon, wage earners and consumer publics. The difference between political power, economic power, and other powers as well, cannot be found, then, in the notion of collectivized decisions. Their difference is, rather, a hierarchical one. That is to say that collectivized decisions are political in that they are (a) sovereign; (b) without exit; and (c) sanctionable. 4 Sovereign in the sense that they can overrule any other rule; without exit, as Hirschman would put it, because they reach out to the frontiers that territorially define citizenship; and sanctionable in the sense that they are sustained by the legal monopoly of force. If politics is perceived as consisting of those collectivized decisions that are both overarching and of greater consequence to the well-being (or ill) of each and all, is it appropriate to begin with the libertarian ideal of Marx or with the question of the anarchist: Why have politics at all? The question is not trivial. After all, why should we like decisions taken for us (in our place) by others, especially when-as in the case of politics-they can even affect life and liberty? The answer has been given thousands of times, but another time will not hurt. In a hypothetical state of nature all decisions are individual decisions. On the other hand, any organized collectivity submits to rules of collectivization at least in the sense that it accepts collectivized decisionsthis being the condition of its organization. Yet the respective ambits of individual versus collectivized decisions vary enormously across contemporary societies, even under equal technological and environmental conditions. For instance, the area of collectivized decisions is incommensurably greater in socialist (communist) than in non-socialist countries. The basic reason for 215

A DECISION-MAKING THEORY OF DEMOCRACY

8.r

this difference is ideological and need not be labored upon. Let us make sure, however, that the point on the ideological factor is made in a meaningful way. It is often said that we are confronted with two ideologies-one individualistic and one collectivistic- and, thereby, with two intractables that must be left at that. However, this way of disposing of the problem overstates the deadlock. The so-called ideology of individualism largely yields to collectivization whenever the utility or necessity of the latter is reasonably demonstrated. The obverse is not true. The ideology of collectivizati'on is unyielding, for it perceives private or individual decisions as intrinsic evils - both because individualism is bad in itself and because it entails private property, private capital accumulation, and all the wrongs thus resulting. The "two ideologies" argument applies, then, to only one of the two. This allows us to distinguish between the "ideology" and the "utility" of collectivizing decisions, and to note that aside from ideological dogmatisms the matter can and is in fact assessed in cost-benefit terms. The reasons given for collectivizing decisions formerly left to individual choice are generally related to technological imperatives and to the service and collective good needs of contemporary societies. In a number of cases, however, it is an open question whether the benefits of collectivizing a given decisional area (schooling, housing, transportation, utilities, and so forth) are not offset, at least in the longer run and in terms of cumulative effects, by the costs. Hence it is both useful and important to ask: When is it either necessary or convenient to collectivize an area of decisions? The ulterior and related question is: How should we proceed in collectivizing decisions?

8.2

External Risks and Decisional Costs

The aforesaid questions can be tackled on the basis of two very simple analytical tools: (a) the costs of deciding; and (b) the risks resulting from collectivized decisions. Axiomatically stated, my premises are as follows:

Axiom r: All group or collective decisions have internal costs, that is, costs for the decision makers themselves, generally spoken of as decisionmaking costs. Axiom 2: All collectivized decisions involve external risks, that is, risks for the addressees, for those receiving the decisions from the outside, ab extra. While the notions of cost and risk have a familiar ring, the axioms show that they have been subjected to pruning and readjustment. 5 To begin with, in my understanding decisional costs are in-group costs; they refer only to 216

External Risks and Decisional Costs

8.2

who decides. Conversely, external risks are out-group risks; they refer only to the collectivity for whom decisions are taken. In the second place, when I say internal costs, reference is made exclusively to the costs of the process of deciding (not to the losses or gains of the members of the deciding body). Internal costs are only (as defined) costs in time, energy, futility, and the like. In the third place, when I say external risks, I mean risk: I say so very deliberately. A cost is determinable and is generally determined (at least, ex post). A risk is, instead, an ex ante indeterminacy. Furthermore, a risk is a particular kind of uncertainty, namely, a potentiality perceived in its dangerousness. It is losing (not winning) that is spoken of as a risk. The full argument is, then, that (a) collectivized decisions involve external risks; that (b) external risks may not result in harm; but that (c) the problem is precisely to increase the probability of "benefit outcomes" and minimize the likelihood of "harm outcomes:' This is the reason for laying emphasis not only on the uncertainty factor but also on the danger element involved. It is true that the collectivity that receives the decisions ab extra may either benefit or suffer from them. But the matter of paramount concern is that a collectivity may not be benefited. Hence a collectivity that receives decisions that are not of its own making is always exposed to a risk. Since our focus is on political risks, these are, concretely, of two sorts: primarily risks-harms of oppression but also (even though I shall have to consider them only tangentially) risks-harms resulting from incompetence, stupidity, or sinister interests. Axiom 2 implies that external risks come about only when, and only if, an area of decision is collectivized. That is to say that if a decision is taken by the same group or collectivity to which it applies, no external risk is involved. To be sure, the group will contain winners and losers and may also decide stupidly for itself. Nonetheless, the losers or the self-harmers have participated in the decision. It cannot be said, therefore, that they were exposed to an external risk in the sense in which nonparticipants are so exposed. Whatever happens-in terms of losses and gains-within the group that actually partakes in a decision is immaterial to the point that whenever an area of decision is collectivized, we thereby create (a) a deciding body; and (b) a risk exposed out-group, which is such because it cannot decide for itself. Reverting to the internal costs, that is, to the decisional costs, Axiom r assumes that only group (or collective) decisions involve decision-making costs. A "one man" decision such as, in politics, the dictatorial decision, has a zero decisional cost. The dictator may well have to bear, when deciding alone, high psychological costs; but these are immaterial to the problem at hand. The first point is, then, that decisions have costs only with more than one decider. In the second place, it bears stressing that the costs in question 217

A DECISION-MAKING THEORY OF DEMOCRACY

8.2

are procedural and, basically, costs in time and fatigue. By and large, decisions that take up a lot of time can be said to cost more than fast decisions. However, if time is a good measure, the measure must be extended to the "time lost;' meaning by this the time wasted on decisions that are not made or e°i-idlessly postponed. low productivity, inefficiency, immobilism, and paralysis can all be brought back to decisional costs. In short, the body that makes the decisions has "costs" (whatever else may be at stake), while the collectivity that receives them faces "risks" (whatever these may turn out to be). Both notions are narrowly defined. Costs are only internal and procedural; risks are only external and harm related. On these premises we may now proceed. With reference to the decisional costs it is intuitive that the crucial variable is the number of persons participating in the decision. As a rule of thumb, the greater their number, the higher the decision-making costs. We may thus say that the cost of decisions is a function of the size of the deciding body. To be sure, this holds good only under the assumption that the decision makers are independent, self-monitoring individuals who are free to express themselves. A thousand people crowding together and proceeding by acclamation do not fall under the rule because there is nothing they really decide; they simply ratify decisions already taken. The rule can thus be restated as follows: Provided that each participant has an independent say, the number of deciders stands in direct relation to the costs of decisions - they increase together. If this be true, it follows that it is irrational to enlarge a decisional body (i.e., to augment the costs of decisions) without reason. Consequently, the question turns on what the reason could be. Among the many possible answers, the most forceful one is, I submit, that a decisional body is enlarged in order to afford greater protection to third parties, that is, in order to reduce the external risks. If this be correct, we may set forth a second rule: The number of deciders stands in inverse relation to the external risks - as the deciding body grows, the external risks diminish. Let us make sure that the argument has no hidden flaw. Assuming, for the sake of simplicity, a collectivity of roo people, the first case is that r person decides (collectivizes the decisions) for 99. In such a hypothesis the external risks are highest, or maximal, while the decisional costs are zero. Second case: The decisions are taken by 10 people. Surely the decision-making costs will rise. Will the external risks diminish? Yes and no. Yes in the trivial sense that they will affect 90 instead of 99 persons. But we cannot say for sure that the 90 are likely to face lesser risks. If we say so, we must have some other factor in mind. So here we have a problem. Third case: All 100 people decide for themselves. Clearly, the external risks will be zero, while 218

External Risks and Decisional Costs

8.2

Max

Decisionmaking

MinL----~=========:::::::::==-100

10

Number of Persons FIGURE

8.r

EXTERNAL RISKS AND DECISION-MAKING COSTS

the decision-making costs will be maximal. Here the argument is again flawless. We have sensed some incompleteness, but in the main the two rules seem to hold and to make sense. If so, we are seemingly approaching a stalemate. To the extent that decision-making costs and external risks are inversely related, or covary negatively, and that both are conceived as monotonic functions, there is little to be said in favor of collectivizing areas of decision. We may have to do so on grounds of sheer necessity, but this would be all, and not at all pleasing; for either the external risks are too great or the decisional costs too high- and whatever is unloaded from one of these elements is reloaded on the other one. To be sure, the monotonic assumption is an unrealistic assumption. Even so, we still have no satisfactory solution. For the solution clearly lies in reducing the external risks much faster than, and before the point at which, the decision-making costs escalate-as shown, illustratively, in figure 8.r above. The figure conveys that in order to have a solution the curve of the external risks must plummet and cross the curve of the decisional costs before an 219

A DECISION-MAKING THEORY OF DEMOCRACY

8.2

accelerated stepping up of the latter. With respect to our fictitious collectivity of 100 persons, the point of least ordinate (IO people) strikes the optimal balance between external risks and decision-making costs-an optimal balance that indicates not only when it is convenient to collectivize an area of decision but also how to do it conveniently. The query is: In what way can we obtain, in the real world, curves like those hypothesized in t.he figure? The problem would be insoluble if the number of the participants in the decision was the only variable. But two supplementary variables come to our rescue and should now be entered: I.

2.

The method of forming the deciding body: how it is recruited or appointed and what is its composition or nature. The rule with which decisions are made: the decision-making principles and procedures.

The first intervening variable is central to the object of reducing external risks. Instead, the second intervening variable taps, primarily though not exclusively, the costs of decision making. The first intervening variable assumes-on account of the expression "deciding body''._ that the decisions in question are group or collective decisions. This assumption simplifies the overall argument; but it is well to note, for the completeness of the argument, that the point also applies to the rule of the one. There are many kinds of single ruler. He may be (a) a same person for indefinite time (the monarch, the head of a church, the dictator); or (b) any person for a definite time (i.e., a different person at short points in time). This major difference hinges, in turn, on how a single ruler establishes himself, that is, on the method of appointment or self-appointment. For the external risks vary as the method of appointment varies, that is, depending on whether the "one" acquires office by heredity (the monarch), election (e.g., the pope of the Catholic church), force (the dictator), lot (anyone), and, in this connection, indefinitely or temporarily. To exemplify, with the absolute monarch or the dictator the external risks are extremely high, especially in terms of risks of oppression. Even so, hereditary succession in office is likely to produce rulers whose personality traits are very different from the traits displayed by a person (dictator) who seizes office by force. The head of the Catholic church equally enjoys office without term; but the fact that he is elected (if only by a very restricted college) implies that he is "selected." On this count alone the external risks are, in all likelihood, strongly curtailed. If the election of an absolute ruler for as long as life lasts increasingly becomes a risky affair, this is so on the grounds of longevity. Nonetheless, the difference between a nonelected and an elected single ruler importantly affects the risk 220

External Risks and Decisional Costs

8.2

factor. Finally, a single ruler appointed by lot on a short-term basis need not be feared in terms of risks of oppression but deserves to be highly feared on grounds of risks of incompetence. So, the method of creation of the decider(s) is central to the calculus of the external risks, regardless of whether the decider( s) is (are) one or many. And it is only because I especially wish to tackle the case of democracy that my argument assumes that the deciders are more than one. The second intervening variable-the rules of decision making-essentially bears, I have said, on the costs of deciding. When discussing the rules that govern decisions, the baseline is unanimity rule. The propounders of unanimity point out that this is the only decision-making rule that attributes equal weight to each decider. This is not always so. While it is true that each member of a deciding body holds, under this rule, a veto power, it does not necessarily follow that it is an equal power. If, for instance, a roll-call procedure is adopted, the first voter would have a disproportionate power; for either everybody also obliges, or there will be no decision. On the other hand, the case can be made that unanimity involves not only an equal but also an excessive, blackmail type of power. Fine points aside, it is clear that the unanimity rule works, or can work, only within the ambit of small groups. This is not a reason for brushing it aside on the consideration that its decisional costs are always too high to bear. As a matter of fact, we shall see that unanimous decisions are by no means as infrequent as they are deemed to be. The point remains that unanimity rule cannot work when large groups are involved. Majority rule or, better, majoritarian rules enter whenever unanimity rule turns out to be impractical. 6 Under majoritarian rules, a given proportion of the deciding body or collectivity carries no weight at all; but the chances of resolution, of arriving at a decision, increase. There are at least three magnitudes subsumed, often confusedly, under the majority rule heading: (a) qualified majorities (often a two-thirds majority); (b) simple or absolute majority (50.or percent); (c) relative majority, or plurality, that is, the major minority (a less than 50 percent majority). Since each of the aforesaid majorities can be measured either with respect to the universe (the entire collectivity entitled to decide), or to those actually present and/or actually voting (a person may be present and abstain), it follows that we can count majorities in at least six ways. There would be no confusion if the simple majority of the universe were always called "absolute majority" and the simple majority of the showers were always called "relative majority:' But this is not in fact the case. Thus, absolute majority, relative majority, and plurality easily get mixed up. On what grounds are majoritarian criteria chosen? Clearly, as we pass from a qualified to a simple majority and, ultimately, to any majority (plurali221

A DECISION-MAKING THEORY OF DEMOCRACY

8.2

ty), the decisional costs are lessened. When a qualified majority is required, decisions are hard to pass, and many decisions will be blocked. When any majority will do, there will always be a decision. The reason for lowering the majoritarian ceilings is, then, to ease decisions and, ultimately, to make sure that a matter cannot be left undecided. If the lessening of decision-making costs were the only criterion, however, we would always settle f9r pluralities. We do not; the reason for this being that we are also concerned, we say, about minority rights, about protecting minorities. More precisely stated, the ulterior criterion is the reduction of external risks. 7 This is the reason that constitutional revisions generally require qualified majorities and that important decisions require an absolute majority (of the universe). All in all, then, in choosing a particular majoritarian rule we have to strike a balance between expediency (reducing the decisional costs) and safety (reducing the external risks); and this balance is struck at different majoritarian thresholds as a function of the greater or lesser import of the things to be decided. So far, so good. The point of controversy resides in the argument that unanimity rule and qualified majorities require the major part to acquire the support of a minor part and thus amount to establishing (albeit to a different extent) a "minority rule?' But this argument is unconvincing on two grounds. In the first place, it is not clear why it should be confined to unanimity and qualified majorities; for pluralities too add up to enthroning, in a far stronger sense (and with far greater frequency), a minority rule. Hence, the minority rule indictment must lead, as stated, to the conclusion that majority rule is "real" only under the absolute majority criterion (the simple majority of the universe). If so, most electoral verdicts would be delegitimized-for not many of those elected receive a greater than 50 percent share of the vote of the electorate (not merely of the actual voters). In the second place, and more important, it is simply wrong to equate any majority that is not an absolute majority to a minority rule. The power of blocking action is incommensurably different from the power of taking action. Thus a minority that impedes action is not the same as "minority rule?' Ruling consists of decisions, not nondecisions. Putting it the other way around, nondecisions (impeding or blocking action) surely restrain, limit, or even obstruct "ruling;' but are not equal to ruling. Indeed, the minorities in question are not empowered to decide. They cannot impose their preferences-they can only protect them. Pulling the threads together, we now dispose of three variables: (a) the number of deciders; (b) the way of selecting (appointing) them; and (c) the rules of deciding. We also need to remember that the problem is to minimize the external risks in relation to the costs of decision making. More precisely, we must obtain, on the one hand, a more than proportionate reduction (in222.

Outcomes and Decisional Contexts

8.3

deed, an exponential reduction) of the external risks and, on the other hand, a less than proportionate increase in the decisional costs. In my analysis this problem is amenable to solution on two grounds. The first one is that the external risks are not so much a function of the number of participants in a decision, but primarily a function of the method of formation of the deciding group and, thereby, of its composition and nature. This is so, first and foremost, because the method of formation of a deciding body includes, among its possibilities, the representative method. And what makes all the difference in the world- from the vantage point of the external risks-is whether or not a deciding group is a group of representatives. 8 Since the theory and practice of political representation are all too easily criticized, let it be forcefully stressed that apart from the representative techniques of controlled transmission of power there is no other known technique for coping with the external risks. Reverting to our earlier example, the reason that allows a collectivity to entrust the power of deciding for the whole to 10 persons only is that this deciding group is assumed to consist of representatives. Otherwise, the collectivity is protected only by all its members entering the decision-making body. This is fine with 100 people; but it quickly becomes unfeasible as the numbers grow. The foregoing considerations imply that the problem of minimizing the external risks without clogging the decision-making process is more importantly resolved on the ground of how the deciding body is formed than on the ground of how it decides. In effect, the representative method of forming the deciding body permits a vertiginous fall of the curve of the external risks (p. 219), whereas the rules of decision making allow only for a lessening, or a smoothening, of the decisional costs. If you like, the two curves display a very different elasticity. Consequently, the key is representation: for only the drastic reduction of the universe of those represented to a small group of representatives permits a momentous reduction of the external risks (of oppression) without aggravating the decisional costs. If the only way of minimizing the external risks were to augment the number of deciders, it would almost never be convenient to collectivize an area of decisions.

8.3

Outcomes and Decisional Contexts

Thus far my focus has been, in the main, on the external risks and how to cope with the problems thus resulting. I now turn to the decision making itself, not only with respect to the rules according to which decisions are made but, in the overall, with respect to their nature in outcome. The question now is: how one decides, with what end result(s). In order to cover this new and 223

A DECISION-MAKING THEORY OF DEMOCRACY

8.3

also more extended ground, new elements need to be brought into the picture: type of outcome decisional context intensity of preference The type of outcome will have to be reduced to its most abstract formulation, namely, whether it is generally beneficial in a positive-sum modality, or non-beneficial (to all) in the zero-sum modality. 9 Since these notions are drawn from the theory of games, I need not explicate them in any detail. Indeed, for my purposes I need to define them in the simplest possible way. A game is said to be zero-sum when one player gains exactly what another player loses. The problem here is simply to win. The adventurous player will try to inflict the maximum damage on his adversary(ies), if at a greater risk for himself. The cautious player will choose, instead, a strategy that minimizes his own possible losses. In every case, when a game is zero-sum, the alternative is simply to win or to lose. Contrariwise, a game is said to be positive-sum when every player can win. If so, the problem ultimately becomes how to share and slice the gains. In game theory, positive-sum games are explicated as cooperative and bargaining games. However, as we move from these analytic constructs to politics-to politics conceived as a game-it should be understood that positive-sum politics need not be cooperative and may well result from a blending of cooperation and conflict. Bearing this proviso in mind, I would put it thus: To the extent that we move away from "politics as war"10 and come near to "politics as bargaining;' to the same extent it is fitting to say that we are shifting from zero-sum to positive-sum politics. 11 The decisional context can be dichotomized into discontinuous or continuous. The context is discontinuous when we are faced with separate, discrete issues. This is the case in referendums and in elections. Regardless of the frequency, and even when a referendum submits to the voter a cluster of issues, the voter necessarily responds with discrete decisions. Likewise, elections are, for the electors, a one-shot decision. On the other hand,. we also find groups (concrete groups) in charge of a stream or flow of decisions. In such a case the decisional context is continuous, for they are no longer faced with discrete issues-at least in the sense that the deciding body need not treat them discretely. Note that here the decisions "flow" in a theoretically endless stream and that nobody can decide in isolation. On both counts the group is likely to link the various issues, regardless of their intrinsic affinities, by engaging in reciprocal exchanges. A group that so performs is generally identified as a committee. But more on this shortly. The immediate point is 224

Intensity of Preference and Majority Rule

8.4

that a decisional context is continuous when a stream of issues is handled by linking them, that is, when the issues are not treated discretely. As the foregoing implies, a discontinuous and a continuous decisional context are not separated by a fixed or "natural" borderline. Whether an issue is dealt with discretely or not can be said to depend on the context in which the issue is placed. On the other hand, whenever large numbers are involvedas in general elections or referendums - one-shot decisions on discrete issues are unavoidable. In this sense, discontinuous contexts can be said to be imposed by the circumstances.

8.4 Intensity of Preference and Majority Rule The intensity of preference-our third element-requires a more extended treatment than the other two. The intensity factor confronts us with the fact that every issue elicits a different degree of affect, of involvement or disinterest. We are thus confronted with the unequal intensity of individual preferences. Preferences vary not only in being diverse but also in that they are strongly or feebly held.12 The fact is well known - it is part and parcel of our daily life experience-but its political implications often escape us.13 A first set of implications bears on majority rule, that is, on the majoritarian decision-making criteria.14 We have recourse to majority rules on the assumption that people have contrary preferences15 and that, if deadlocks are to be overcome, what is preferred by a greater number (be it an absolute majority or any plurality) must prevail over the preference of the lesser number. Fine. But it should not escape us that the rule totally ignores the fact that these preferences have a different intensity. Majority rules weigh individuals as individuals; this means that they equalize unequal intensities. So majority rules rest on a fiction, indeed, on a very thin and unrealistic stipulation: Let us pretend that preferences are equal in their intensity. Let us pretend it; but let us also realize that they are not. With this I am not suggesting in the least that we should attribute equal weight to equal intensities, instead of equal weight to individuals declared to be equal. In Kendall's phrasing, I am not suggesting that we substitute decisions "preferred by most members" with decisions "most preferred by members:'16 The intensity criterion cannot establish a workable rule-and our first need is to have regulae.17 I wish only to explain why the majority principle is never accepted in full, why its application often falls short of the mark, and especially why intense minorities dispute the principle and definitely refuse to submit to it. The real facts of life are that one strong No regularly overcomes two weak Ayes and, conversely, that one obstinate Yes usually beats down two feeble

A DECISION-MAKING THEORY OF DEMOCRACY

8.4

Noes. Thus, all collectivities probably contain- at any given moment, on given issues-an intense subgroup that stands a chance to win over the apathetic (least intense or indifferent) subgroups even if the intense subgroup constitutes a minority of the collectivity in question. More fully stated, a 5r percent majority is unbeatable if it also consists of intense members; but a 51 percent (or even greater) majority is not unbeatable if it is a feebly intense majority. It is important to note that nothing would be changed in this state of affairs if the material power resources (e.g., wealth) of each and a'll were perfectly equal. This is to note, then, that "intensity" is the most overlooked, yet fully independent and most powerful, base of power. However that may be, the indisputable fact is that intense minorities carry an extra weight in the decisiontnaking processes; their intensity compensates for their numerical inferiority. Hence the crucial question becomes: Do we find the intensity factor randomly associated with majorities and minorities? Or is it the case that intense minorities are far more frequent and far more durable than intense majorities? The evidence on the basis of which this matter has been tentatively explored thus far does not confront this question, or does so very partially.18 But many of us have witnessed the resurrection of direct democracy practices and have had ample opportunity to rediscover that direct democracy (as practiced) turns out to be a true paradise for active, usually very small, minorities. In the literature on the so-called campus revolution, a recurring estimate is that 5 to ro percent of the student population pulled the wires and had its way. Why is this so? In the case at hand the reply is not far-fetched. Intensity is the element that brings concrete groups together, that activates them, and that accounts for their impact and force of attraction. The broad question remains: Why should intensity be, as a rule, an attribute of small minorities and not also of majorities? A major explanation is, I submit, that an intense majority can well materialize but, in all likelihood, on a single issue, or on a set of issues revolving around one core. Therefore, over a broad range and sequel of issues we can only expect to find, at best, different intense majorities each of which dissolves as the issues change. Thus, an intense majority amounts to an occasional majority. Instead, small groups can be lastingly and equally intense on a global set of issues. Indeed, this is how and why they come into being.19 The difference is, then, that intense minorities are real groups, while intense majorities are ephemeral aggregates. If they are not, then it will be discovered that they are mobilized by intense minorities-and we thus come a full circle. It has surely not escaped the reader that the above considerations move beyond the majoritarian rules and go a long way toward explaining and underpinning the law first formulated by Mosca, that is, the sense in which it is

Committees and Unanimity

8.5

always true that minorities rule. In our earlier discussion I have challenged the "single ruling class" hypothesis but have acknowledged that "controlling minorities" always exist and always steer the course of history. 20 I trust that we now see why this is so. In the final analysis, and all other conditions having been made equal, who is intense is active; who is active wins over the inactive; and only small groups are likely to be durably intense and active, over time, on global sets and sequences of issues. That small, intense groups may equally be highly factious, seditious, and, in the end, self-destructive does not prove a different point; rather, it goes to explain why the success of intense minorities is far less frequent than one might otherwise expect. Reverting specifically to the majority principle, we have noted that it disregards the unequal intensity of individual preferences and that, in turn, the intensity factor takes its revenge over the majority principle. We are thus prompted to wonder whether some other decisional rule might avoid this fix. For instance, is it not the case that unanimity rule handles the intensity problem better than majority rule? As regards the principle, it can hardly be disputed that unanimity does justice to the intensity of preferences; indeed, it overdoes justice, for it legitimizes the veto of one sole intense dissident. But while unanimity can be upheld-on this and other grounds-in the abstract, it seemingly fails us in practice. The objection is well known and appears formidable: The decisional costs of unanimity rule are self-negating, for they add up to paralysis. This conclusion sounds pretty final and doubtless applies whenever it applies. But does this conclusion apply to all cases? Is it always the case, that is, that the cost of unanimity is paralysis? This is the question to which I now turn.

8.5

Committees and Unanimity

Up to this point we have looked at the unequal intensity of preferences as an obstacle to majority rule. But the fact that intensities differ and obtain different distributions can also be an advantage. If everybody's preferences were always equally and strongly intense on all issues, how could any deciding body ever reach an agreement? In effect, agreements are reached precisely because its members are not equally intense on all issues. In short, the mechanism of group agreement largely consists of the non-intense giving in to the intense. And this is particularly the case with those decision-making groups spoken of as committees. Since the notion of committee is largely left to intuitive understanding, it will be necessary to provide an exact meaning and denotation for it. 21 I shall define "committee" with respect to three characteristics, as follows. 227

A DECISION-MAKING THEORY OF DEMOCRACY

8.5

First, a committee is a small, interacting, face-to-face group. Being an interacting group, it cannot consist of less than three members, for interactions begin to be meaningful when they are triadic. 22 But how large can "small" be? The face-to-face requirement does give a first delimitation; but an assembly is still a face-to-face group, yet an assembly largely outnumbers a committee. 23 The maximal efficient size of a committee is established, in effect, by its operational code. In practice, this means that committees generally range from three to, say, thirty members. To be sure, thirty is a very loose approximation. A committee may perform poorly with ten members, but manage to perform nicely with forty-it all depends on the extent to which its members fully comply with the mode (or code) of operation I am about to describe. Second, a committee is a durable and institutionalized group. It is institutionalized in the sense that its existence is recognized, whether legally or informally, by the fact that certain things have to be done by, and through, a specified group. A committee is institutionalized, we may say, by the tasks assigned to it. However, a group cannot be, or become, institutionalized without being durable. This latter characteristic is not related to the actual permanence or stability of the members of the group. As we know from roletaking theory, whoever enters an institutionalized group is likely to assume the time perspective of the institution. Thus a group is durable-regardless of its actual rate of turnover-when its members act as if they were permanent. What counts is the expectation. 24 Third, a committee is a decision-making group confronted with a flow of decisions. We may equally say that whenever decisions come in streams, or strings, they "naturally" call for a committee-like counterpart. So, when we speak of committees we make reference to a continuous decisional context25 in its difference (seen earlier) from discrete decisions on discrete issues. The importance of pinning down the notion of committee is borne out by two ulterior considerations. The first one is that committees, and the committee system thus resuJting, largely escape not only visibility but awareness. This is not only because a committee system actually performs in a penumbra under conditions of low visibility (a very important accompanying characteristic, as we shall see) but also because of its dispersion, its fragmentation. It follows that the crucial role played by the "committee subsystem" within any political system is-far more often than not-either grossly underscored or grossly disparaged, and generally both. My second reason for dwelling on the definition is, thus, that the committee system is, in one, the most pervasive, crucial, and misunderstood part of the real "stuff'' of politics. 26 All decisions enacted by any polity are previously examined, discussed, and actually drafted by one or more committees. And since a government is (under

Committees and Unanimity

8.5

my definition) a committee, often enough it is a committee that also decides in the final instance. 27 How do committees actually work? Hardly ever on the basis of majority rule. Decisions are usually not brought to a vote. If they are, the vote is generally pro forma, a show of hands for the sake of the record. Most of the time decisions in committees are unanimous. But this does not imply in the least that committees abide by unanimity rule. The rule entails that each member possesses a veto power- and no such power, or principle, enters the committee practice. On the other hand, if committee decisions are so often unanimous, this is not because their members are of the same mind-they are not. Committees generally end up with unanimous agreement because each component of the group expects that what he concedes on one issue will be given back, or reciprocated, on some other issue. Since this is a tacit understanding, it can be called an operational code. The essence of this operational code is do ut des, I give to get in return. The notions of logrolling, bargaining, settlement by compromise, and mutual adjustment also tap this modus operandi. We also speak very much, nowadays, of vote trading and "payments:• Yet none of the foregoing expressions bring out what makes for the distinctiveness of the committee-type "compromising;' When saying compromise, we generally understand that the parties to a settlement meet somewhere halfway on the specific matter to be settled. Now, even in committees each resolution may be long fought and end in a middle-course solution. But there is more to it. What is peculiar to committees is that their members engage in exchanges over time and having especially in view a future time. As I was saying, each member of the group tacitly "expects;' In order to bring out the time element, I shall speak of a principle, or mechanism, of deferred reciprocal compensation. As we probe further into the operational code of committees it is very important to keep in mind that committee members are not empowered to veto decisions. Therefore, if a member takes the course of endlessly obstructing the proceedings, he just ends up with making a nuisance of himself; and this behavior calls for retaliation. That is, a majority of members will see that the obstructionist gets his punishment, or learns the lesson, in due course; he will be cut off from reciprocation when his turn would otherwise come. It can also be easily seen that a mechanism of deferred reciprocal compensation presupposes two intertwined conditions with which we are by now familiar: (a) unequal intensity of preferences; and (b) a flow of forthcoming decisions. The group can be unanimous inasmuch as the distribution of the intensity of preference tends to change from issue to issue, so that at every moment the members who feel less intensely about a problem are disposed 229

A DECISION-MAKING THEORY OF DEMOCRACY

8.5

to give in to the members who feel strongly about it. But this disposition needs, in turn, to be lubricated and reinforced by a return in due course, that is, on future decisions: Whoever concedes today expects to be paid back some other day. Prima facie, the committee type of decision making may strike us as being fragile and precarious, as being exposed to too many conditions. In reality, it turns out to be a widely practiced and efficient decision-making system and this because it hinges on very realistic incentives and rewards. As will already have been understood, decisions in and by committees are positivesum. The essence of a decisional system based on deferred reciprocal compensation is, in effect, that all the members of the group stand to gain and, moreover, that this positive-sum game is a continuing one. Therefore, committees are far from proceeding on feet of clay. Quite to the contrary. Given the "transacting disposition" afforded by unequal intensities, committees sustain a mere disposition with a concrete interest in over time, overall, positive-sum returns. To be sure, every committee has its internal conflicts for two reasons at least: that no preestablished harmony presides over the distribution of preferences, and that the gains are seldom evenly distributed. Therefore, from time to time even committees have showdowns, that is, have really to decide by a majority vote. However, if the recourse to the majority principle is not an exception but becomes the rule, this means that a committee no longer works. Hence, we may say that a committee ceases to be a committee. In any event, what matters is to understand why the majority principle represents the watershed between committees and non-committees or, if you wish, between functioning and malfunctioning committees. While committees put to efficient use the unequal intensity of preference, we already know that majority rule has no use for intensities and, if anything, counteracts them. This is, then, the first respect in which decisions by committees and decisions by majorities are at odds. But they are also at odds in an ulterior, major respect. Decision by committees are positive-sum. The same cannot be said for decisions by majorities: The majority principle is zerosum, that is, it produces zero-sum outcomes. When we have recourse to this principle, the majority wins all, the minority loses all, and the majority can be said to gain, with respect to whatever is at stake, what the minority loses. Decisions by committees and decisions by majority principle also differ in a third respect. The majority criterion imposes a dichotomous structure of choice such that voters and decision makers are somewhat forced into expressing their first preference, and their first preference only. Contrariwise, decisions by committees clearly allow for preference orderings and indeed encourage agreements based on second or third preferences. 28 230

Committees and Unanimity

8.5

Reverting to the major difference, the one between zero-sum versus positive-sum outcomes, I should immediately put forward the caveat that what is true in the abstract may not be true in all concrete situations. As we shall see shortly, the "majority game" can be played (in parliaments) in ways that do not lead, over time, to zero-sum outcomes. Nonetheless, it can hardly be disputed that the majority principle as such is a zero-sum principle. A group or collectivity that regularly and really decides by majority vote (a) treats each issue as a discrete issue; (b) resulting, issue by issue, in a zero-sum outcome; which in turn (c) puts a premium on the formation and/or stabilization of a "winner take all" majority. Let us pause on the last implication and ask: Why is it that stable winning coalitions are encouraged by majority voting and not by the committee type of decision making? After all, also in a committee a coalitional majority will find it rewarding for itself to persist and to take all, all the time. Certainly, when this is the case, a committee ceases to function as a committee. But this is seldom the case because the committee members have more to lose than to gain by destroying their own operational code. Temporary alliances, yes; stable coalitions, no. 29 A winning majority that always wins builds up, within a committee type of face-to-face small group, an intense, frustrated minority of regular losers, which is likely, in turn, to slow down the proceedings, to engage in obstructionist rather than cooperative tactics, and in the end to heighten the overall decisional costs to an unbearable point. In the end, then, the "committee advantages" would be lost. So, why not play the game smoothly, each for himself, in reciprocity of give and take? Up to now I have looked at committees in isolation. But every committee is inserted into a web of other committees: the committee system. How does the system operate or, more parsimoniously, how do the committees belonging to a same iter (path) or a same network interact and coordinate? The coordination occurs via a second mechanism: side payments. While I borrow the notion of side payments from game theory, it should be clear I use it in contrast to the notion of internal payment. The mechanism of deferred reciprocal compensations to which I have made reference thus far can also be described as a mechanism of deferred internal payments, that is, of intra-group exchanges. It follows that the side payments that occur among and across a set of committees are external payments, payments made to outer groups. It is appropriate, however, to identify them as side payments because they are made to other committees, to groups of the same kind that play (and reciprocate) the same game. I thus define "side payment" as the concession(s) each committee has to make to other committees (in the same order of business). 231

A DECISION-MAKING THEORY OF DEMOCRACY

8.5

It should also be clear that side payments do not need to be explicitly negotiated. It is quite possible, of course, that side payments among committees take the form of laborious vote-trading negotiations. But this is hardly the distinctive aspect of side payments conceived- as they are here conceived - as a mechanism of coordination. The greater the complexity of the network (which does become very complex in democracies), the greater the need for automatic, or quasi-automatic adjustments. This means that most side payments will not be explicitly bargained but will simply o~cur in terms of anticipated reactions. Committees, like individuals, discount in advance, implicitly, the likely reactions of the third parties affected by their decisions. Thus the anticipation of a negative reaction feeds back into a state of indecision, or else in transforming a coherent decision into a "please all" incoherent decision. I do not assume, therefore, that the inter-committee mechanism of coordination works to anybody's satisfaction. I am simply trying to explain how it can work at all. Coordination is the telos; but much of the actual process is best described as a muddling through. In the summing up, a committee system performs, within each committee, on the basis of deferred reciprocal compensations (or exchanges) and, qua system, on the basis of side payments largely guided by anticipated reactions. The internal payments (reciprocal compensations) are conducive to unanimous decisions and positive-sum outcomes. The external side payments represent the costs, but also the sine qua non condition of a spontaneous (as opposed to imposed) process of adjustment and coordination. What needs to be added is that external payments are not necessarily circumscribed by, and within, the boundaries of a committee system. To the extent to which a committee system includes "representative committees" that are responsive to the popular will, to a related extent the external payments reach out to this broader world. But to this we shall return.

8.6

Committees, Participation, and Demo-Distribution

Committees are, at a minimum, the decision-forming and often, in the final analysis, the decision-making bodies in whatever polity and under whatever regime. Nonetheless, a committee system (subsystem) is molded by the political system to which it belongs. Thus a comn.iittee system that operates within a democracy acquires characteristics of its own. In order to make comparisons that are relevant to the purpose at hand, we may set aside the committees that are part and parcel of the bureaucratic machinery (also because their number varies as a function of the extent of bureaucratization) and focus on the committees that primarily address, at manifold levels, the formulation of 232

Committees, Participation, and Demo-Distribution

8.6

policies. Under this sorting, a first feature is that in democracies committees are in the process of proliferating far more than in autocracies. 30 One reason for this development is that whenever a deciding body becomes too large, it usually generates, within its ambit, a smaller group, i.e., a committee, which actually drafts the decisions subsequently enacted by the larger body. This multiplication lends itself to two contrary interpretations: either it amounts to a development of anti-bodies and represents, therefore, a counterdemocratic development; or it is perfectly congruent to the pluralistic development of democracy. If one adopts the latter view, then the proliferation of committees can be said to maximize participatory democracy by affording ulterior sites for "real participation:' The latter contention is not without merit. Since participation has no "real" meaning other than taking part in person, if we are serious (as the theory of participatory democracy ought to be) on the matter, then we have here a most simple and most easily operationalized concept: Participation is a ratio that can be expressed as a fraction and related to a frequency. 31 In a group of 10 each takes part, i.e., participates, as r/ro, and takes more part the more frequently the group meets. In a group of roo each takes part as r/roo-and so forth. As the denominator grows, it measures the diminishing share or weight of each partaker. Concurrently, though less surely and importantly, as the frequency diminishes, so does the import of participation. Hence it is unquestionably the case that participation is a significant, authentic, and effective taking part only in small groups (and hardly beyond assembly-size groups). When we speak of electoral participation and, in general, of mass participation, the concept is overstretched and points, more than to anything else, to "symbolic participation;' to the feeling of being included. There is little question, then, that committees represent the optimal unit for real participation (above and beyond sheer demonstrating). It does not follow, however, that the demand for participatory democracy can really be met on these grounds. To increase the occasions of participation by increasing the number of committees resolves the problem of who is on them. But what about the excluded? Clearly, their problem is not settled by the participation of others in their place; it can be solved only in terms of control, by the extent to which they-the citizenry of a democracy-are in control of the decision-making bodies. We are thus referred back to the representative techniques of controlled transmission of power as the means for minimizing the external risks. We have also arrived, at the same time, at the distinguishing characteristic of committee systems in democracies: the existence of committees that are responsive and accountable to the citizenry at large and, in this sense, of representative committees. In democracies, committees remain and 2

33

A DECISION-MAKING THEORY OF DEMOCRACY

8.6

indeed augment. But this first trait is far less distinctive and far less momentous than the fact that democracies transform the method of formation-the recruitment and composition - of committees. This is not to say that all committees are composed of representatives resulting from free elections and electoral procedures. In effect, even in democracies most committees are not so recruited; they are special-purpose groups "representative of" talents and technical competencies. But a decisional system resembles a traffic system in that it can be controlled at a few strategic crossing or merging poirits. Therefore, the control function can be satisfied by relatively few representative committees -eminently the government and the standing commissions in parliamentplaced at the appropriate junctures. We are now in a better position to confront the grand question: How does a committee system stand vis-a-vis a democratic system? If the question is dramatized, it amounts to asking whether committees and democracy are inimical to each other. In less earthshaking terms the question becomes whether committees slow down the furthering of democracy or whether they sustain any stage of democracy. As we have long seen, assessments of democracy -whether it exists, and to what degree-depend on the parameter employed. If democracy is transliterated into demo-power, that is, understood as a literal power of the people, then nothing will ever suit. When the chips are down, power rests on its exercise, not on the titular attribution. Hence literal democracy must be a literal self-government. 32 And literal self-government can be operationalized and measured just like participation: It is the ratio between the governing of each over the others and, conversely, of the "all-body" over "each-bodY:' We are thus seemingly confronted, at the macrolevel, with a squaring of the circle. Instead of raising our arms in despair, however, I propose to change perspective. While microdemocracies can still be conceived solely in input and thereby resolved into the notion of demo-power, I submit that this is no longer true for macrodemocracies, which are best conceived and furthered in output, that is, in terms of demo-distribution. 33 I take it, then, that what can still be importantly improved is not the power end of the problem-more power to the people-but its end result: more equal benefits, or less unequal deprivations, to the people. Even though scholars are somewhat reluctant to acknowledge it, they are in fact dealing less and less with who has power and growing more and more interested in payoffs and allocations, that is, with the effects of power decisions: who gets what. This, let it be added, is also the way in which democracies are generally perceived by their demos. For the public at large, popular rule hardly means that the demos should actually take power in its own hands; rather, it means the fulfillment of popular wants and needs. 2 34

Committees, Participation, and Demo-Distribution

8.6

Now, if democracy is assessed in output, then it can be seen that a committee system is not an anti-body that counteracts the furthering of demopower, but a decision-making system that sustains demo-distributions. This is so, let it be repeated, on the crucial premise that democratic regimes create representative (accountable and responsive) committees placed at strategic junctures. Under this condition the side payments trespass the committee boundary and become external payments in general, that is, payments that extend to the universe of the represented. Hence a positive-sum decisional system linked to the people by the umbilical cord of representation is positive-sum also on behalf of the people. The contention cannot be pushed so far as to imply that by this route we approach equidistributions, or a Pareto-optimum, let alone a Rawls-preferred solution. 34 On account of the magnitudes at stake, my contention only is that a positive-sum outcome benefits all or, better, all as a generalized aggregate-but not each one, nor each one in similar amounts or at the same time. 35 I say "demo-distribution" because this is what I mean this and not more. While my claim - positive-sum allocations- is modest, its importance should not be downgraded, especially if one takes into account that the argument does not bear solely on the committee system but also on decision-making systems in general and ultimately, by this route, on politics (defined as the master set of collectivized decisions). Let us, therefore, extend the argument to the theory of decisions. Thus far we have seen that decision-making systems can be divided into positive-sum (both within the deciding group and for the out-group) and zero-sum. 36 I have also stressed that larger-than-committee-size groups are generally compelled to employ majoritarian rules and that these are, in principle, zero-sum. However, my earlier warning was that the majority principle did not necessarily lead, in practice, to zero-sum politics. We must now clarify this matter and underpin the extent to which majority rules coincide with a zero-sum decisional system and inevitably result in zero-sum wins or losses. The majority principle does entail zero-sum outcomes in the following cases: (a) elections (voting for office); (b) referenda; and (c) whenever a concrete majority is relatively stable and crystallized. This is so on two different grounds: the one-shot, discrete nature of the decision (in the case of elections and referenda), and the nature of the majority in question (it must be "concrete" and "crystallized"). The implication is that the majority principle will not produce zero-sum overall outcomes under two joint conditions: (a) a continuous flow of decisions submitted to (b) cyclical, relatively fluid or fluctuating concrete majorities (unstable coalitions, if you wish). Even under these circumstances each decision is, no doubt, zero-sum; but the process is likely 2 35

A DECISION-MAKING THEORY OF DEMOCRACY

8.6

to produce, in the aggregate, positive-sum compensations among the changing majorities. 37 Let it also be pointed out that the complement of a continuous flow of decisions must not only be a concrete but also an institutionalized group (as previously defined): in substance, either a committee-size group or an assembly such as a parliament. Occasional, crowd-like, or unruly assemblies that can be attended, each time, by different publics will not do. Under special circumstances, then, majority rule is not zero-sum rule. In practice, and in the realm of democratic politics, this adds up to saying' that while parliaments abide, of necessity, by the majority principle, nevertheless parliamentary decision making can result, over time, positive-sum, (a) if its majorities are cyclical; or (b) if a parliamentary majority is permeable (to the demands of the opposition); and/or if (c) it lacks discipline and displays low coalescence. On the other hand, when the aforesaid circumstances do not apply, or whenever a decisional context is discrete, majority rule is zero-sum rule. It is time to recapitulate and bring our strands to a conclusion. It is clear, I trust, that an ideal decision-making system would have to satisfy the following requirements: (a) Each individual should be given the same weight; (b) equal intensities (of preference) should be given the same weight; (c) positivesum and zero-sum outcomes should be appropriately balanced; (d) external risks should be minimized; (e) decision-making costs should be minimized. As the enumeration suffices in itself to show, there is no one principle, rule, or decision-making system that can even begin to meet all these requirements. 38 What does happen is that each unit applies the decision-making rules that are feasible and connatural to it. These units can be reduced to: committees institutionalized assemblies any dispersed voting collectivity 39 Committees shun majority rule, seek unanimous agreements via internal deferred payments, and adjust to the outer world, or incorporate its demands, via side payments. Institutionalized assemblies, instead, must abide by majority rule but may or may not end up, over time, with zero-sum outcomes, depending on the fixity of their majorities. Dispersed voting collectivities are characterized- regardless of their size- by the fact that a dispersed universe is unable to interact and to enter vote trading: Each choice maker is left to choose discretely. The unit is clearly residual, yet unified by the following traits: Each actor can only vote; his vote is necessarily expressed issue by issue; counts only if it adds up to some winning majority; and the outcomes are always and necessarily zero-sum. Restated in the essential: Dispersed voting collectivities can neither bargain nor enter into agreements.

Committees, Participation, and Demo-Distribution

8.6

Two points need pinpointing. First, the fact that a dispersed voting collectivity is not defined by the numbers involved does not imply that size thresholds are irrelevant. In effect, if the numbers do not surpass the assembly size, a dispersed collectivity can be gathered; whereas beyond the assembly size a collectivity is necessarily dispersed (in terms of its characteristics). The second point bears on the difference between the two grand cases of the category: general elections and referenda. In the former case, electorates at large choose a person or party that becomes, in turn, entitled to make the decisions for them. Thus, while the electoral results are in themselves zero-sum, the voting act is a voting into office that projects itself into processes that may become (in parliament, but especially in their committees) positive-sum. In short, the voting act is not a final, self-contained act. Instead, referenda are final. In this case, voters at large do not choose choosers but decide, and thereby close, an issue. Thus referenda are definitely zero-sum and cannot lead, in any sense, to a cooperative game. 40 Since my major focus has been on the first unit-the committee-and since a committee system is seldom adequately appraised, it is appropriate to attempt such an appraisal. There is much to be said in favor of committees. For one thing, only face-to-face smaJl groups with a weJl-established but highly flexible operational code (reciprocal compensations can be deferred) allow for a "reasoned" and discussed elaboration of decisions. For one thing, then, (a) committees can well claim to be the optimal decision-forming unit. Moreover, (b) committees not only account for the unequal intensity of preferences but put it to efficient use. And if "committees of representatives" are entered, then a committee system can be credited with these additional merits: (c) allowing for a drastic reduction of the external risks (of oppression) at no or minimal increase of the decisional costs (as compared to assembly costs); and (d) producing positive-sum outcomes for the collectivity at large (demodistribution). Last, but not least, substantive minorities (ethnic, religious, or other) that are inexorably beaten when decisions come to a majority vote find in committees the setting in which (e) their more intensely preferred claims stand a good chance of being consented to. This high praise should not lose sight of the limits. In essence, the other side of the coin is that positive-sum outcomes basically lead to incremental change. 41 Rapid or decisive change confronts clear cut yes-or-no alternatives and hence demands zero-sum decisions. It should be well understood, therefore, that I am not implying that positive-sum politics should be preferred, whenever possible, to zero-sum politics. On the other hand, given that the majority principle cannot account for the unequal intensity of individual preferences, and given the extent to which its implementation is disturbed and 2 37

A DECISION-MAKING THEORY OF DEMOCRACY

8.6

deviated by the intensity factor, it would seem to follow that (a) majority rules should be employed with a clear cognizance of their shortcomings; and that (b) such rules are best employed either faute de mieux, for lack of anything better, or when a turn of events needs, at whatever cost, to be enforced. Since no single optimal system of choice making appears to exist, the next step is to appraise at which points too much committee rule, or too much majority rule, become counterproductive.

8.7

Consociational Democracy

The question as to where lies an appropriate balancing between positive-sum and zero-sum decisional procedures is difficult to answer in the abstract. It is more easily answered by bringing into the picture the distinction of Arend Lijphart between "majority rule democracy;' on the one hand, and "consociational democracy;' on the other. 42 Majoritarian democracy is, according to Lijphart, only one type of democracy, the one inspired by the Westminster model; but there is another type of democracy, consociational democracy, in which "majority rule is replaced by joint consensual rule;' and whose working principles ("grand coalitions, mutual veto, proportionality, and segmental autonomy") doubtlessly depart from the majoritarian principle. 43 There is no question, it seems to me, that Lijphart's "two models" of democracy represent an important contribution to the empirical theory of democracy and that his argument in favor of the consociational type of democracy is both forceful and faultless in its claim that segmented and deeply divided societies cannot be based on majority rule; their option is to be either consociational or not to be democracies at all. It is apparent that my decision-making approach parallels and goes to underpin the thesis of Lijphart. It also suggests, however, some corrections to it. But I must first bring forward my own argument to the point at which it matches the one of Lijphart. So far I have spoken of intense minorities in general, implying that such minorities can indeed be very small ones, for their defining condition is intensity, and intensity only. The theory of consociational democracy addresses, instead, ethnic, linguistic, and religious minorities that are not only large minorities but also cleavage minorities, I mean, minorities defined (and upheld) by a particular structure of cleavages, namely, by cumulative, reinforcing, and, specifically, "isolative" cleavages. 44 These cleavage-based minorities do not disturb my scheme of analysis but represent a special case that must be explicated on its own grounds. Cleavage minorities do not disturb my scheme of analysis because they must still meet the condition of being "intense" (if they are not, they are only a reconstruc-

Consociational Democracy 8. 7

tion of the observer of no relevant real-world consequence). The difference is, then, that cleavage-based minorities are intense-when they are such der-ivatively, that is, because they are, in the first place, ethnic, linguistic, or religious minorities that identify themselves on these criteria and also perceive themselves as disfavored or menaced in their cherished identity. This difference brings about the ulterior difference that the things about which cleavage minorities are intense are foreknown and, at least in their core, well deter, mined. This is also to note that while cleavage minorities are often large minorities, their behavior is far more predictable than that of the small minorities brought together only by their intensity. And at this point my argument joins the one of Lijphart. A first observation is that I would definitely stress the ideal-type, indeed the polar-type, nature of the distinction between majoritarian and consociational democracy. Lijphart stresses, instead, the empirical and empirically extracted nature of his types and, by so doing, is in danger of overstating his case. The contrast is empirically overdrawn, to begin with, in that no realworld democracy abides by absolute majority rule. The English constitutional practice (the conventions of the constitution) hinges on the respect of minority rights surely as much, and eventually more, than a number of democracies not inspired by the Westminster model. It must be forcefully stated, therefore, that to equate any democracy purely and simply with majority rule is only a shorthand expression or, otherwise, an error. The majority rule in question is always, if a democracy is to survive as such, a limited majority rule. And my analysis of the operational code of committees goes to show that the limitation of majority rule is not only a principle upheld by the conventions of the constitution but also, and largely, a by-product of the committeetype modus decidendi. So, it is in fact the case that in all democracies most decisions are not majoritarian, zero-sum decisions. The contrast is somewhat overworked, empirically, also at the other end, that is, with respect to the consociational democracy type. Lijphart speaks of a "mutual or minority veto" and asserts that "majority rule is here replaced by negative minority rule7' 45 These dictions might also include the liberum vetum practiced and indeed abused in the Polish Diets of the seventeenth and eighteenth centuries. It is true that Lijphart immediately qualifies the veto practice he has in mind as being identical with Calhoun's "concurrent majority" principle. But then Calhoun's formulation -in which the veto is an implication, not the standing principle-is perhaps a better one. 46 Lijphart also pushes his consociational point a bit too far in his claim that "the greater speed and decisiveness of majoritarian government are more apparent than real:' 47 If this is so, and I do grant that it may be so, it is either because ma2 39

A DECISION-MAKING THEORY OF DEMOCRACY

8.7

joritarian government is not as majoritarian as it appears on the surface, or because consociationalism is not as thoroughly consociational and veto based as we are told. To pursue the argument in my own terms, I would put it as follows. First, in some countries majoritarian rules can cope with "intensity;' whereas in other countries the decisional rules must submit and adapt themselves to intensities. Second, the types "majoritarian democracy" and "consociational democracy" represent two typical ways of striking the balance Hetween zerosum and positive-sum politics. Third, along a continuum whose polar ends are "always majoritarianism" or "never majoritarianism;' the concrete democracies are likely to be all the more majoritarian the more they are consensual, homogeneous (culturally), and non-segmented (in their structure of cleavages), and all the less majoritarian (i.e., consociational) the less these characteristics obtain. Somewhat differently restated, while we always find a blend of majoritarian and non-majoritarian decisions, the proportions vary, and vary in response to this rule of thumb: The greater the presence of intense minorities, the less a zero-sum governing is advisable and democratically feasible. All in all, and to conclude, many or even most decisions reflect neither majority rule (even when votes are cast and counted) nor unanimity rule (even when decisions are unanimous). They do not really constitute an enforcement of majority rule, literally and strictly understood, because, in the face of intense opposition and intense minorities, majorities usually make concessions, smooth out the angles, and enforce, when they do, a "tempered" majority will. On the other hand, the unanimous or near-unanimous decisions of committees do not result in any way from unanimity rule, for its characterizing element, the veto, plays no part in the operational code of deferred reciprocal compensations. We may put it thus: The intensity variable creates a large inbetween area of decision making that is not strongly or strictly majoritarian, and yet cannot be said to reverse majority rule into a substantive minority rule. Rather, what can be said is that the greater the incidence of the intensity factor, the greater the number of nondecisions and/or the greater the number of committee-type decisions.

8.8 A Coda on the Cost of Idealism It may have been noted that in this chapter I no longer deploy the thread of the preceding ones, namely, the counterpoising of the "ideal" and the "real;' of prescriptive democracy and descriptive democracy. This is because I have been looking into the actual mechanisms of decision making, regardless of the personality traits and motivations-idealistic or not-of the concrete de-

A Coda on the Cost of Idealism

8.8

cision makers. I have also been looking into "invisible" rather than "visible" politics-which is also to say into deeds rather than words-and therefore, presumably, into an arena where the rhetoric of politics matters little. Authentic idealists are also found, to be sure, around a committee table; but most verbal idealists, the demagogues and populists that we hear at the hustings, become remarkably businesslike when they perform off the record and under conditions of committee invisibility. So, the closer we come to how actual decisions are concretely made, the closer to one another we find idealist and realist politicians to be. Does the foregoing imply that I intend to conclude on a realistic keynote? In a way, yes-but on ad hoc grounds. Having dwelt on the prescriptive element and on the role of ideals about as much as on the realistic understanding of democracy, at the end of this journey it is appropriate for me to appraise whether the current theory of democracy suffers more from idealistic or realistic defects and neglects. The prevailing mood of the 1960s and 1970s is well rendered, I believe, by a writing entitled The Cost of Realism. 48 Actually, the title says it all: It conveys the message that our theory of democracy is suffering from an overdose of realism and, thereby, from a poverty of idealism. As the writer concisely puts it, "the cost of realism has been the practical abandonment of what has been the distinctive moral function of democratic politics and government."49 I first wish to discuss the entry of morality into the matter and, subsequently, to raise the complementary question: What is the extant cost of idealism? In point of clarification, it must be well understood that norms, prescriptions, and ideals need not be, and do not coincide with, the realm of ethics. Moral ideals are a subset, and a very specific subset, of ideals in general. It is also uncalled for to associate "classical democracy" with moral purpose. The classical democracy in question is only Greek democracy; and since the ancient Greeks had yet to separate ethics from politics, religion, economics, and what not, to qualify the democracy of the ancients as a morally ordained polity can be quite misleading. In the third place, one is also brought to wonder whether the democrats who speak of a "moral function" of government, or of politics, grasp the full implication of such stand. It took endless sufferings and bloodsheds to establish the separation between the realm of God and the realm of Caesar; a separation that also implies that the realm of ethics is not the realm of politics and, concretely, that political persecution, torture, and slaying cannot be legitimized nor redeemed by moral purpose. Thus, to bring morality into politics is akin to playing with fire- as we have only too well rediscovered since Hegel theorized a "political ethos;' Sittlichkeit.

A DECISION-MAKING THEORY OF DEMOCRACY

8.8

Both fascism and Nazism found the Hegelian Sittlichkeit highly convenient; and when Marxist totalitarianisms claim for themselves the mission of creating a new man, the "good man;' at whatever cost, their source is not the ethics of Kant but the ethical state of Hegel. God relieve us, then, from the moral functions of government. Of course, democracy has, or should have, moral foundations. But when the case is so phrased, it is an altogether different case. 50 And the fact that these foundations have become thinner and thinner is hardly ~ fault of the realist. Unfortunately, the "loss of ethics" re.Beets a long-standing historical trend that present-day generations have simply inherited. The moral foundations of any free polity have to do with the sense of dutifulness, with the understanding that rights involve obligations and that there is a value and a gratification in doing things "for nothing;' gratis. There is no such thing as a free lunch, but there is such a thing as free giving, as acting amore Dei. But dutifulness, doing in exchange for nothing, and the like, have long been eroded by an economic-like vision of politics. And if this is the argument, then I do subscribe to the view that the present-day crisis of democracy is very much a crisis of ethical foundations. Let there be no misunderstanding, therefore, on where I stand on the matter of how politics relates to ethics. As I was saying, when we talk about the normative aspect of politics, we are not necessarily talking about moral prescriptions. Since all norms are not ethical norms, using the form ought does not automatically place us in the sphere of ethics. But, of course, the divide between politics and ethics depends very much on how we define morality. In order to do so, manners of speech, such as "political morality;' social morality, professional ethics, and the like, are better set aside. Since I have already explained my dislike of the Hegelian Sittlichkeit (though I do not blame Hegel for the misuse and the unintended consequences of his moral philosophy), we are left, in essence, with the ethics of Kant-and I would indeed draw the line between politics and ethics on Kant's criteria: The realm of morality is the realm of "disinterested actions;' 51 However, it is the Weberian distinction between Gesinnungsethik and Verantwortungsethik, between ethical principles pursued regardless of consequences and, conversely, an ethic answerable for consequences, 52 that best underpins the point at which ethics and politics depart from one another. Kant left us with this distressing question: Should we tell the truth to a dying person? Under Weber's criteria, the Gesinnungsethik answer is that we should, while the Verantwortungsethik answer is that we may not. Admittedly, when I say that Weber's criteria well divide ethics from politics, I am forcing the letter (though perhaps less the spirit) of what Weber said; for he spoke of two kinds of "morality." I am suggesting, instead, that there

A Coda on the Cost of Idealism

8.8

is only one ethic, only the Gesinnungsethik. When we are faced with the Verantwortung, that is, with being responsible for consequences, we are entering another sphere in which we have to reckon with the problem of means and of how we should behave as political and not solely as moral animals, as individuals living in a polis and not solely in themselves. In Kantian-Weberian terms, then,.it is very clear that our democracies suffer from a lack of moral basis and are indeed in "moral crisis:' 53 The citizen of present-day democracies is told by all quarters that he should be "rational;' meaning by this that he should perceive and calculate rationally his own interest and that his political behavior should be, if rational, exactly like his individual economic behavior. Can a truly ethic-minded person accept this? What has an image of man that portrays him-not in fact, but normatively-as a well-calculating egoist to do with ethics? I leave the question at noting that it does not appear to concern the present-day idealists and "evaluativists" (the propounders of evaluative political science) any more than the realist. I am not questioning their good faith when they claim to express ethical concerns; but the question remains: On what ethics is this claim based?54 In any event, here the point is whether the "idealist" persuasion of politics can ascribe to its merits the one of seriously attempting to restore ethical foundations worthy of the name. In my view, this would be a very great merit indeed; but it cannot be awarded. So, if it is fair to ask, Which is the cost of realism? it is equally fair to ask, symmetrically, Which is the cost of idealism? While I have addressed these questions, from a variety of angles, all along, 55 I now propose to draw the nets ashore on the basis of the analytic framework employed in this chapter. On such basis the trends that I would consider distinctive of our time can be itemized as follows: A wholly unjustified and indeed dangerous neglect of the problem of external risks 2. A very unclear perception of the size thresholds that affect both the costs of decision making and the switch from committee rule (so to speak) to majority rule 3. An emphasis on more visible politics, as against its low visibility areas, with no clear understanding of what is involved 4. A hypertrophy of the arenas that are, first, entered by politics and, second, politicized 5. Conclusively, a very naive democratic primitivism that pitches direct and participatory democracy against control and representation I.

It can be easily seen that the first two charges address matters that should be handled by the realistic, i.e., descriptive, theory of democracy. The inade2 43

A DECISION-MAKING THEORY OF DEMOCRACY

8.8

quate alertness to the external risks is well exemplified by the decay of garantiste (protective) constitutionalism in the face of an exponential growth of the potential of power. The behavioral attack on institutionalism and so-ca11ed formalism has gone too far. We may bicker on how effective constitutions can be-yet it is better to have a good constitution than a bad one, and I dare add that I prefer to have the constitution on my side. 56 As for the second indictment, I simply note that our reckless way of handling size thresholds without accounting either for the decisional costs or the drawbacks of m·ajority rule, is well illustrated by the devastatingly simple manner in which the participatory demand was accommodated, in most of Europe, as the 1960s entered the 1970s: by transforming committees into quasi-parliaments, into poorly performing over-sized bodies. The third point, that is, the visibility problem, is missed or overlooked by the realist camp and, at the other end, misunderstood and aggravated by the idealistic camp. It is also a point that needs some elaboration, for I have confronted it thus far only tangentially, by implying that low visibility is very important to the operational code of committees. No doubt, democracy seemingly demands transparence, that the house of power be a house of glass. The rational, as distinguished from the moral, basis of this stand is that high visibility a11ows for better control and thereby reduces the external risks. This is indeed so; and it implies that even if visibility entails higher decisional costs, these may weJJ be justified. But the coin has another side. As we well know from personal experience, the same person behaves very differently as he or she switches from low to highly visible contexts; and this means that the visibility element may improve but also distort behavior. For instance, visibility distorts when it imposes "image selling" to the detriment of "responsible behavior." Furthermore, visibility can well enhance, if not create, conflicts; so much so that removal from visibility is the most practiced and practical way of lessening tensions. In conflictual, polarized, or deeply divided societies, it is indeed the case that paralysis is averted precisely to the extent that bargains are struck under conditions of rigorous invisibility. As we turn from internal to foreign policy, the dictum "open covenants openly arrived at" fares even worse. To be sure, we can legitimately ask that poker be transformed into another game; but we cannot ask of one player only to uncover his cards, while the others are still playing poker. All in all, if "more visibility" is displayed, as it is being displayed, as a universal panacea, it is likely to produce far more ills than it cures. To the extent that visibility hampers responsible behavior, instigates image selling and demagogy, intensifies conflicts, leads to decisional paralysis or, in international politics, to defeat, to the same extent external risks are best looked after by other means and ways of control. 244

A Coda on the Cost of Idealism

8.8

Let alone that the efficacy of a searchlight diminishes with its diffusion. Too much visibility, on too many things, jams visibility. The last two indictments mostly pertain, instead, to the "costs of idealism?' Let me explicate them swiftly. How should we understand, to begin with, the expression hypertrophy of politics? As noted at the outset, decisions are collectivized in many spheres and arenas, and politics enters only when decisions are collectivized by the "sovereign" decision makers entrusted with the legal monopoly of force. I have dwelt, in the main, on the risks and costs of collectivized decisions. But the notion also helps pinpoint the difference between the extension of politics and politicization. More and more decisions can be collectivized and brought-in democracy-under the authority of parliaments, yet their implementation can be "depoliticized?' 57 Politics may well enter everything-as the contention goes-but, if so, not all politics is politicized.58 There is wide disagreement as to the point at which the expansion of politics is necessary and beneficial. There is less disagreement, however, on the respective merits and demerits of politicization; and my point is confined, in effect, to the politicization of politics. When politics enters a given arena, this may be for the better- as when we refer to an apathetic citizen who becomes interested, informed, and participant- or for the worse. When is it for the worse? In two cases: (a) when politics becomes overheated, that is, the politics of violence, intimidation, intolerance, and ideological discrimination; and (b) when it enters in one of these forms (even its mildest one) the judiciary, the army, civil service, and institutions of learning. And while my scheme of analysis has little to do with what causes the overheating of politics, it does suggest a number of caveats with respect to its containment. The preliminary caveat is that new electorates for creating new electorally appointed bodies should not be created as lightheartedly as they are. Quite apart from whether the decisional costs are worthwhile and appropriately assessed, the ulterior question is whether what enters-via the electoral route-is only politics in its innocuous and/or positive sense, or politicization in its most undesirable aspects and cumulative drawbacks. The final point is, in essence, that if participatory democracy is conceived as being inimical to representative democracy- and if the former actually undermines the latter-then I very much fear that both are in deep water. But this needs some elaboration. As we know, the notion of participatory democracy can cover a number of different things, to wit: (a) participation in terms of interest, attention, information, and competence; (b) participation in support of "voice;' i.e., pursued in terms of demonstration democracy; (c) power sharing, that is, real and effective participation in decision making; (d) a par2 45

A DECISION-MAKING THEORY OF DEMOCRACY

8.8

ticipation that amounts to a true direct democracy. So, what bodes ill for what? With respect to the first item, we all agree that we badly need more interested and better-informed citizens. But the participationist pushes the point far beyond this recommendation. And while it is fair to say that nobody has adequately followed up how "intensity" relates to information and, beyond that, to knowledge and competence, the blame for this grand omission is especially deserved by the idealist. 59 Blames aside, the assumption '"that anyone who becomes involved has a positive contribution to make ... is naive and unworkable .... There are three general kinds of people: those who become involved and make contributions; those ... who contribute little or nothing; and those who, with an unerring drive toward the wrong thing . . . cause the rest of society much trouble, expense, and sorrow:' 60 Dahrendorf makes an equally forceful point: "I submit that the permanent participation of all in everything is in fact a definition of total immobility . . . it would mean a mixture of permanent theoretical debate and permanent practical inaction." In his view (with which I concur), under the unqualified extolling of participation, "the Citizen is about to overreach himself, to create conditions which would make useless the very principle which he tries to establish:' 61 With respect to the second item, demonstration democracy, we may simply say that more "voice" is beneficial, provided that voice does not become "violence:' We must again pause, instead, on the third item, which is also the central one for the participationist. Since the "reality" of participation is expressed by a fraction, and since it can hardly be disputed that its efficacy is optimal in committees, one way to increase real participation is to increase the number of committees. But as this path is in fact pursued, it appears selfdenying and rapidly leading to a dead end. The more numerous the committees, the more a decisional iter is slowed down (time costs) and the greater the incidence of side payments (to the limit of utter discoordination). Thus, the proliferation of committees quickly reaches a ceiling beyond which what is gained in terms of power sharing is disproportionately lost in terms of efficacy and efficiency. If the opportunities of real participation offered by committees were matched against the number of claimants, the whole system would rapidly collapse under the weight of diseconomies of scale. We are left, then, with what I have called "referendum democracy;' conceived as a daily direct democracy in which the citizenry sits before a video and allegedly self-governs itself by responding to the issues in the air by pressing a button. How nice-and how deadly. Since referenda do not choose the choosers but decide the issues without any further ado, its virtues inextricably hinge, to begin with, on the state of information and the level of competence

A Coda on the Cost of Idealism

8.8

of large publics. As Rousseau put it, the people want a good that they often fail to see. And the world of Rousseau was incommensurably more simple and intelligible than our present world. So much so that even the expertspolitical scientists, but also economists- appear more and more unable to grasp it. Thus the idea that the government of our fantastically complex, interconnected, and fragile societies could be entrusted to millions of discrete wills that are bound to decide at random, with a zero-sum instrument-this idea is indeed a monumental proof of the abyss of under-comprehension that is menacing us. I conclude. Aside from the above reduction to the absurd, in the overall it seems to me that we are pursuing targets that are out of proportion, unduly isolated, and pursued blindly, and that we are, therefore, in the process of creating- at the minimum - a wholly unmanageable and ominous overload. The most distressing part of the trends recalled under my five headings is precisely the mental fog in which they occur. Can we understand politics without any understanding of risks and costs, rule by committee and rule by majority, the relevance of size and the nature of outcomes? I think not; yet very little is said and done about it. We are beginning to realize-in the prosperous democracies - that we are living above our means. But we are equally, and even more grievously, living above and beyond our intelligence, above our grasp of what we are doing. The more we engage in remaking the body politic, the more I am struck by the uneasy feeling that we are apprentice sorcerers who are turning politics into a gigantic negative-sum, or minus-sum, game - a game in which we are all bound to lose.

2

47

A DECISION-MAKING THEORY OF DEMOCRACY

Notes 1. This chapter draws, but modifies and largely expands, on the article "Will Democracy Kill Democracy? Decision-Making by Majorities and by Committees;' Government and Opposition, Spring 1975. 2. The distinction between collective and collectivized decisions is generally missed by the political economy and/or the public choice literature. Two good surveys of the aforesaid literature are Dennis C. Mueller, Public Choice (Cambridge: Cambridge University Press, 1979); and Norman Frohlich and Joe A. Oppenheimer, Modern Political Economy (Englewood Cliffs, N.J.: Prentice-Hall, 1978). 3. This is but one characteristic, not an exhaustive definition. For a detailed analysis, see my "What Is Politics;' Political Theory, February 1973. 4. The hierarchical difference between political and economic power assumes market-type conditions. In a state-owned, centrally planned economy the difference becomes minimal for, in practice, the wage earner does not have an exit option; furthermore, the economic sanction (being left to starve) is even more formidable, in and by itself, than most politically enforced sanctions. 5. The basic reference is J.M. Buchanan and G. Tullock, The Calculus of Consent (Ann Arbor: University of Michigan Press, 1965). While I am greatly indebted to this brilliant work, it will be seen that my conceptualization fundamentally departs from it. 6. In this chapter the expression majority rule applies only to the principle; it is a decisional criterion, a regula, not a substantive rulership. For other meanings, see chapter 6, sections 1 and 2. 7. This formulation is more precise because in the argument on minority protection it is often unclear whether reference is made to in-group or out-group minorities. 8. I say advisedly "group of representatives;' for the expression "representative group" may also be referred to sociological representation, i.e., to some kind of resemblance. I intend, instead, responsiveness to and responsibility for. 9. For the simplicity of the argument, I neglect the "minus-sum" or negative-sum outcome, in which all the parties stand to lose. It should also be understood that here reference is made to in-group outcomes. How these affect third parties (out-groups) is an ulterior problem to which we shall come later. ro. See chapter 3, section 2. W.H. Riker, The Theory of Political Coalitions (New Haven: Yale University Press, 1962), postulates that all politics is zero-sum. If so, all politics is like war; a view to which I very much object. 11. The specialist will note that, in my definition, zero-sum and positive-sum are rendered as ways of playing or as ways of perceiving the game (not as fixed structures), and that I deliberately avoid reference to whether the game is, in itself, constantsum or not. 12. Here it is unnecessary to distinguish between "preference" and "choice;' and to ask what motivates a choice. Without entering the complexities of choice theories, I shall use preference as the most generic term of its semantic field. Also, the argument does not assume that each individual actually has preferences on everything. 13. To the best of my knowledge, Dahl's A Preface to Democratic Theory, esp. pp. 90-rr9, is the first work that systematically builds the intensity factor into the theory of politics. 14. The fullest discussion of the issue is W. Kendall, "The 'Intensity' Problem

Notes and Democratic Theory" (with G.W. Carey), in N.D. Kendall, ed., Willmoore Kendall Contra Mundum (New Rochelle: Arlington House, 1971), pp. 469-506. For a bibliography on majority rule in general, see chapter 6, n. 12, herein. 15. In themselves, preferences may simply be different, not contrary, ones; but majoritarian decisional rules transform a spectrum of preferences into a yes-no structure of alternatives. 16. Willmoore Kendall Contra Mundum, p. 472. 17. Reference is made to electoral voting and mass electorates. Within small groups, intensities can be expressed and weighed, e.g., by a system of "point voting" in which each voter is given an equal stock of vote points that he can divide or concentrate at his discretion and therefore, presumably, in accord with his intensity rankings. I cannot see, however, how a point voting system could be made to work with mass electorates. Here its closest approximation is the "cumulative voting" system with multimember constituencies; a system that permits the elector to concentrate all his votes on a single candidate. But cumulative voting has not truly expressed, when tested, intensities; and its actual drawbacks are greater than its potential assets. 18. See, e.g., the discussion of the point by Lane and Sears, Public Opinion, pp. 108-13. Aside from the fact that their evidence is exclusively American, their argument is confined to the universe of the voters, whereas my question addresses a variety of universes and assumes (even though I must leave it at its most abstract formulation) that the game of politics is played across different-size units and all the way to statistically insignificant concrete groups. 19. One might add that small groups are more easily interacting groups; a condition that renders them not only self-sustaining but also capable of self-reinforcement. It may also be the case that individuals identify more fully and more easily with small rather than large groups. These are only, however, conjectures. 20. See chapter 6, sections 4 and esp. 5. It will also be noted that my argument is far broader, even though it winds up at a similar conclusion, than the one of 01-_ son's Logic of Collective Action (see chapter 6, section 6, herein). · 21. Actually, in the technical literature inspired by Duncan Black the definition of "committee" is utterly counterintuitive, for in his influential The Theory of Com-· mittees and Elections (Cambridge: Cambridge University Press, 1958), ·Black's stipula0 tion is: "By a committee we will mean any group of people who arrive at a decision by means of voting" (p. 1). Under this stipulation electorates at large are, when vot• ing, committees. 22. Let alone the consideration that a committee of two would be unable to solve deadlocks. Other characteristics of "group;' such as shared goals or a sense of belong~ · ing, should not be included among the defining characteristics of committees. 23. In the English constitutional parlance the House of Commons can conveiid. as a "committee:' This is a terminological anomaly related, however, to ad hoc tulesL ; of procedure. See K. C. Wheare, Government by Committee (Oxford: Clarendon Pte'ss/t 1955), pp. 7-9, who otherwise adheres to common usage and grants that "if a commjt-r . tee of one is thought of as a contradiction in terms, so is also a committee ofait~~u¥/>, sand and one:' >; }X: 24. My argument does not assume that a committee should be perceived,as,iJn(K. ending in time (duration is not eternity), nor that it will cease to perform as,li,.;c;p:Wi:J. mittee when its life span is preestablished and comes close to ending. Once"'J~ffimit/· ·/_::_! ·,, .

+

A DECISION-MAKING THEORY OF DEMOCRACY

tee members have taken on their role, the role performance is likely to continue to the last day. Note, incidentally, that the reputation of being a good committee member paves the way to a multi-committee career. 25. This property satisfies the requirement technically known, in game theory, as the "supergame" of prisoner's dilemma games. On the prisoner's dilemma, see R.D. Luce and H. Raiffa, Games and Decisions (New York: Wiley, 1957), pp. 94-II3. For the possibility of a cooperative solution emerging in a prisoner's dilemma supergame, see Michael Taylor, Anarchy and Cooperation (New York: Wiley, 1976); pp. 28-97. 26. This is the central intuition, on a philosophical plane, of B. ,de Jouvenel's The Pure Theory of Politics (Cambridge: Cambridge University Press, 1963). 27. It is obvious that these considerations apply to all large organizations. The argument is confined to political systems for brevity. 28. As understood here, a first preference need not be the most preferred one, but merely the one preference for which we settle. To be sure, preference orderings can be accounted for in a variety of ways. Among the electoral systems, the single transferable vote (the Hare system) and the alternative ballot do allow for the expression of second-third preferences; and the same can be said of double ballot runoffs. A basic technique for obtaining preference orderings is the "Condorcet criterion": Choose that candidate or that proposal that defeats all others in pairwise counts. However, preferences need not be construed in pairs; they may also be "funneled;' i.e., discarded one after the other until the two solutions preferred by most are put to the final vote. Significantly, the ad hoc literature hardly mentions, in connection with preference maneuverability, the committee system. 29. I am here contradicting Riker's generalization that the process of group decisions is "invariably ... a process of making coalitions" (The Theory of Political Coalitions, p. 12). A coalition certainly is an "alliance;' but it must be characterized by some degree of stability, for it would make little sense to say that a coalition is unmade and remade issue by issue. Hence, in my argument coalitions are stable alliances, and one possibility; temporary alliances are another possibility; and atomized behavior cannot be ruled out either. 30. Wheare recalls Churchill's outburst, "We are overrun by them [committees], like the Australians were by rabbits"; and his Government by Committee well attests (if only with reference to the "committee forming part" of the British machinery of government) to this proliferation. 31. See the earlier discussion of participatory democracy, chapter 5, section 6. 32. Whose impossibility is discussed at length in chapter 4, section 3. 33. The distinction bears on the extremes of the size continuum, that is, I understand "microdemocracy" as not surpassing the size of a small city, whereas "macrodemocracy" refers to large-scale states with a territorially dispersed population. Between these extremes, it is difficult to establish how size relates to democracy. See R. Dahl and E.R. Tufte, Size and Democracy (Stanford: Stanford University Press, 1973). 34. A solution is Pareto-optimal when at least one person is better off and nobody else is worse off. A solution is Rawls-preferred when it gives as much as possible to those who have least. I am not impressed by these "solutions." With regard to the Pareto principle, the objection is that no change of any import could ever be made without damaging somebody. As regards the Rawls's principle, I concur with the criticism of Dan Usher: "The doctrine that our notions of what is just and unjust can 250

Notes be rationalized with reference to the worst off person in society ... is preposterous . . . . Taken literally, it means that a slight deterioration in the condition of the worst off person is considered more important than a major improvement in the well-being of millions of other people" (The Economic Prerequisite to Democracy [New York: Columbia University Press, r981], p. 149). 35. This formulation complies with the caveat of Douglas Rae that "political democracy sets numerical limits on dissatisfaction, but makes no stronger promise than that" (American Political Science Review, March 1971, p. II9). In effect, my case does not rest on popular "want-satisfaction" assumptions. 36. It should be clear, by now, why my distinction is not tied (as in game theory) to the one between variable-sum and constant-sum games. Over time, and as we shift back and forth from in-groups to out-groups, a priori constant-sum assumptions detract from the real-world fit of the zero and positive-sum constructs. 37. This point is probed, with reference to the distinction between "predominant" and "countervailing" majorities, by R. D'Alimonte, "Regola di Maggioranza, Stabilita e Equidistribuzione;' Rivista Italiana di Scienza Politica 1 (1974). 38. See, contra, Buchanan and Tullock: "The logrolling process [full side-payments] provides the general model of analyzing the various choice-making rules" (The Calculus of Consent, p. 123). My conclusion is, instead, that no such general model exists. 39. I omit other identifiable units, such as small groups in general or occasional (non-institutionalized) assemblies, because the former are not amenable to any precise operational code, and the latter are utterly rule-less. 40. I cannot enter into the typology of referenda and even less into the actual performance. See, for the latter, D. Butler and A. Ranney, eds., Referendums: A Comparative Study of Practice and Theory (Washington, D.C.: American Enterprise Institute, 1978). My objection in principle to "referendum democracy" is stated in chapter 5, section 7. 41. On incrementalism, C. E. Lindblom, The Intelligence of Democracy: Decision Making Through Mutual Adjustment (New York: Free Press, r965), remains a classic analysis. 42. See A. Lijphart, The Politics of Accommodation: Pluralism and Democracy in the Netherlands (1968; 2d ed., Berkeley: University of California Press, 1975); and "Typologies of Democratic Systems;' now reproduced in part in Democracy in Plural Societies (New Haven: Yale University Press, 1977). In his most recent volume, Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries (New Haven: Yale University Press, 1984), Lijphart changes "consociational democracy" into "consensus democracf' I shall not follow this renaming since it may convey, if unwittingly, the idea that the Westminster, majoritarian model is not consensus-based. See also the criticism of majority rule of Nordlinger, Conflict Regulation in Divided Societies. 43. Lijphart, "Majority Rule versus Democracy in Deeply Divided Societies;' Politikon, December 1977, p. n8. This article well summarizes his case. 44. The "insulation" characteristic is, I submit, very important to the theory of consociational democracy, since cumulative cleavages may be conflict maximizing just as much as conflict-minimizing. The point is made in Sani and Sartori, "Polarization, Fragmentation and Competition in Western Democracies;' in Daalder and Mair, eds.,

A DECISION-MAKING THEORY OF DEMOCRACY

Western European Party Systems, esp. pp. 331 and 337. Note, also, that my argument does not cover "class minorities;' for which I must refer to my chapter, "From the Sociology of Politics to Political Sociology;' in S.M. Lipset, ed., Politics and the Social Sciences (New York: Oxford University Press, 1969), esp. pp. 75-87. 45. "Majority Rule versus Democracy in Deeply Divided Societies;' Politikon, p. n8. 46. See J.C. Calhoun, A Disquisition on Government (1851). Calhoun's "concurrent majority" consists of giving to "each interest or portion of the community a negative [my emphasis] on the others.... It is this negative power, the power of preventing or arresting the action of the government- be it called by what term it may, veto, interposition, nullification, check, or balance of power-which in fact forms the constitution. . . . It is, indeed, the negative power which makes the constitution - and the positive which makes the government. The one is the power of acting; and the other the power of preventing or arresting action. The two, combined, make constitutional governments:' (See in J.M. Anderson, ed., Calhoun: Basic Documents [Carroltown: Bald Eagle Press, 1952], p. 51.) While Lijphart recalls Calhoun, the quotation shows that the argument of the latter is largely constitutional. 47. Politikon, p. 123. However, Lijphart qualifies this statement as applying to the long run (see Democracy in Plural Societies, pp. 50-52). 48. Lane Davis, "The Cost of Realism: Contemporary Restatements of Democracy;' Western Political Quarterly, March 1964. 49. Ibid., p. 38. The emphasis is mine; but Davis repeats at p. 43: "The heart of classical democracy is moral purpose." 50. See, e.g., J.H. Hallowell, The Moral Foundations of Democracy (Chicago: University of Chicago Press, 1954). See also, and importantly, the writings of Reinhold Niehbur. 51. The Kantian meaning is lost when "interest" is made synonymous to any "motivation" whatsoever. Kant did not intend "actions without motivation;' but that moral actions are not motivated by utilitarian, selfish, or ego-regarding criteria. On this view, utilitarian ethics fathers the non-ethics of our time. 52. See Max Weber, Politik als Beruf, 2d ed. (Munchen, 1926), pp. 55ff, trans. in From Max Weber (New York: Oxford University Press, 1946). 53. This is the sense of Solzhenitsyn's criticism of the Western civility (in his Harvard Commencement Speech, 1978), and very much the perception that Eastern European refugees have of our democracies. I return to this crisis and the loss of ethics in chapter 16, section r. 54. The writings of C.B. Macpherson (Democratic Theory, but also The Political Theory of Possessive Individualism [Oxford: Clarendon Press, 1962]) well illustrate a moral impetus and approach; yet I am at loss in detecting his ethics. 55. Reference is especially made to the management-mismanagement of ideals (chapter 4, sections 4-5 and 7), to the discussion of participatory democracy (chapter 5, section 7), and to the anti-elitist polemic (chapter 6, sections 7-9). 56. For constitutionalism and the protection of law, see chapter n, herein. 57. The prime example is the independence of the judiciary, designed to separate rule formulation (of parliaments) from rule adjudication. The demand for a "neutral;' professional civil service protected from party colonization addresses the same problem. 58. I only concede arguendo that politics enters everything. As Heinz Eulau vig-

Notes orously puts it with reference to the assertion that "the teaching of political science is itself a political act;' this proposition is "theoretically trivial, empirically false ... morally wrong." See "The Politicization of Everything;' in V. Van Dyke, ed., Teaching Political Science (Atlantic Highlands: Humanities Press, 1977), p. 55. This is so, inter alia, because "to be theoretically valid and empirically acceptable the Proposition must be falsifiable by evidence to the contrary;' whereas "the premise of the politicization of everything ... is always true and its denial is always false" (pp. 56-57). Eulau also points out that "if everything is political, nothing is political. ... Although pretending to be 'theoretical' the conception of politics as something immanent and universal in all human action is only a monstrous tautology ... devoid of substance" (p. 58). I believe this to be a correct criticism. 59. See chapter 5, esp. section 7. 60. F. MacKinnon, Postures and Politics: Some Observations on Participatory Democracy (Toronto: University of Toronto Press, 1973), p. 8. Note that the author "believes in participatory democracy" (p. ix). 61. R. Dahrendorf, "Citizenship and Beyond: The Social Dynamics of an Idea;' Social Research, Winter 1974, pp. 691 and 692.

2 53

·This Is a splendid update of Sartori 's classic work. in which he reviews the maior democratic theories of our time and canvasses astutely the salient issues among them . Sartori synthesizes a theory of his own which he proffers as a new main stream view to his readers His trenchant and swift-moving argument moves deftly among competing schools of thought. The book"s greatest strength lies in Sartori's demonstration that prcscrip1Ivc and doscr1p11vo theories (the ideal and the real) must be blended, to be valid. in an integral whole-in a theory of the democratically possible. The clarity and dramatic power of this erudite work render it very accessible to undergraduate students. William T. Bluhm, The University of Rochester "Everyone seriously interested in democratic theory will welcome the appearance ol The Theory of Democracy Revisited. His earlier Democratic Theory, of which the new book is a revision, -nade a vigorous. vital, and distinguished contribution to our contemporary understanding of democratic theory. In the new book Sartori takes tater discussions Into account where they are appropriate, but he has not made the mistake of excesslvely reworking an already outstanding work. I expect to roadors a gonorat,on from now the rev,s,on wilt still be a fresh and lively interpretation of democratic theory. Robert A. Dahl, Yale University "This Is a splendid book which represents a major contribution to the study ol pofitics- arguably the most important theoretical analysis of democracy since World War II. It is beautifully written In a style whose luc1d1ty is pierced every now and then by the author's mordant wit and sense of irony. As for its scholarship. it is rare 1f not unique to find an author as familiar with the empmcal findings of µollt1cal science as he is with the classical texts of political philosophy, or who marries phifosoph1cal analysis with h1stoncaf perspective. S. E. Finer, Un,versity of Oxford ·This is lhR kine! of ;in;ilysis th;it IS S;irtori"