126 Chapter Six
Michels' iron law of oligarchy, the second is a much more classically abstract theory, functionalism; the third is that much admired and much vilified school, rational choice theory. The basic principle underlying the discussion is thai it is not useful to set up lists of what "good" theory should be, because lists by their nature tend to be disconnected, thin, sterile, and artificial. It is much more useful to study actual theories, to learn to evaluate these theories kindly and intelligently, and to use these theories as guides to intellectual analysis. And for these purposes, a picture of a real (heory is worth a thousand definitions.
IRON-CLAD THEORY IN POLITICAL SCIENCE
Robert Micheis' theory has a paradoxical position in American political science, because while virtually every member of the discipline is familiar with "the iron law of oligarchy," the substance of the theory is largely ignored. This is clearly a result of Michels' unpalatable conclusion; if, as the theory of oligarchy holds, small elites invariably control social groups, then democracy is impossible. Michels' own response to this problem is both cheerful and unsatisfying. He offers the story of an old peasant who, on his deathbed, tells his sons there is a fortune buried in the fields, but fails to tell them exactly where. The sons therefore vigorously dig up the land for years, vastly improving the soil quality, the harvests, and iheir wealth; but never finding the supposedly buried goal. Democracy is like this ephemeral gold, Michels says,
a treasure which no one wifl ever discover bv deliberate search. But in continuing our search, in laboring indefaligably to discover the indiscoverable. we shall perform a work which will have fertile results in the democratic sense (Michels. 1959. p. 405).
While economic development and the spread of education will increase the strength of democratic movements, Michels argues, in general the oligarchic struggles-the battles between different elites to control the people-constitute a cruet game, "that wi!l continue without end" (Michels, 1959, pp. 406, 408). It is interesting here to recall one of the "behavioralist" works discussed earlier, Dahl's study of New Haven elites, since Who Governs? (1961) wholly supports the thesis of elite rule. Dahl, however, called it polyarchy, which put a better face on the matter without substantially changing the thrust of the argument.2
Many people feel that the very phrase, "the iron law of oligarchy," tells all one needs to know about Michels' theory, yet there is much more to be discovered in this classic example of "real theory" in the social sciences. Michels' book first of all represents i\ neat example of what is mcani by the level-s of theory, according to whic theory has a hierarchy of applications from the most specific to the most genera . undoubtedly accunile to s;iy, indeed, thai a good theory must have this full 3rray^

Michels* book encapsulales first a theory of European socialist parties, then a theory of similar ideological parties (including, for instance, syndicalists), then a theory of European parties generally (various conservative or catch-all parlies) with an occasional reference (o U.S. or Latin American parties; then rises to other working class organizations such as labor unions. At further higher levels of abstraction, Michels shows thai his theory reaches not just parties but to the government as a whole, lo the state itself; finally, at the highest and broadest analytic level, the theory applies to ail organizations of whatever type-thus the "iron law".3
THE LAW OF OLIGARCHY
As a socialist of (he old-fashioned warm-hearted kind, whose only goal is a true democracy in which all social classes equally share, Michels'often-repeated purpose in the analysis is to find out how democracy may be achieved and preserved. The text is therefore infused with his lack of neutrality, yet this bias does not diminish the clarity of his vision, as the behavioralists assumed such bias inevitably would do. Instead, his hopes for democracy increase the sharpness with which Michets faces the paradoxes his study reveals, as he inquires "into the causal infrastructure that produces the government of human organizations.
The iron law of oligarchy, which Michels calls a "universally applicable social law," is finally stated in the conclusion of the book after having been implied, discussed, and illustrated for the previous almost 400 pages:
. . . every' organ of the collectivity, brought into existence through the need for the division of labour, creates for itself as soon as it becomes consolidated, interests peculiar to itself. The existence of these special interests involves a necessary con/lid with the interests of the collectivity (Michels,' !959. p. 3S9).
Special attention should be directed to the word "every" in this statement, because it brings Michels' law into exact correspondence with (he Popper-Hempel school of positivist thought, which required that any scientific law or hypothesis should be universally applicable- Very few social theories even attempt to meet such a stringent criterion; even natural scientists have some difficulty achieving it.'1
Theories in political science, especially, are often so cluttered with qualifications that they make no claim (o universality at all-and are therefore equivalent, as some Wtics have remarked balcfully, to saying thai water runs downhill except on Sundays sod in Argentina, But Michels not only claims to present a universally applicable law. "e makes a strongly documented case for ils validity. Since his brief definition is per-hsps unduly succinct to give Ifie lull flavor o(' the theory, it is useful to look at it in

THE CAUSAL MODEL
Michels bases his theory of oligarchy on two quite separate foundations, first, the technical side, involving the necessity of leadership in human affairs, and secondly, ihe psychological side, involving men's love of domination. The technical argument is based on the premise that the "sovereign masses arc altogether incapable of undertaking the most necessary resolutions" (Michels, 1959, p. 25). This results in part from sheer numbers, but also on the masses' disinterest in anything but their own immediate affairs and their resulting, ignorance of important issues. That this fact still liolds even in modern educated societies is supported by works in the Olson school, which document the difficulty of interesting anyone in collective action; as well as by any public opinion poll. where citizens show themselves innocent of the most basic facts about politics and government policy.'1
Because of the impossibility of direct democracy. Michels continues, it is necessary to delegate decision-making power to representatives who will carry out the will of the masses; and while initially all members of the mass are equal, certain talents are necessary for such delegates, and these skills differentiate them from nondelegates or nonieaders (Michets, 1959, pp. 27-28). "Organization implies the tendency to oligarchy," and as a result of organization "every party or professional union becomes divided into a minority of directors and a majority of directed," until "a rigorously defined and hierarchical bureaucracy" is created (pp. 32, 34). This leads Michels to (he conclusion that mass voting, usually thought to be the mark of true democracy, is actually its opposite, for in the act of voting one renounces one's right to rule, one's sovereignty, by bestowing it on someone else. and representative government is a fraud, disguising the wills of the leaders as the will of the mass (pp. 38,40).
Intimately joined with this view of the bitter fruits of organization is Michels' recognition of the even more bitter results ofnonorganization, for he repeatedly refers to the incompetence of the masses, and is equally emphatic that "no undertaking can succeed without leaders, without managers," and Iheir authority must necessarily increase for the good of the whole (Michels, i959. p, 89). Leaders therefore must have oratorical skills, supplencss of mind. physical presence, force of will, breadth of knowledge, self-sufficiency (even though it may appear as arrogance), and perhaps even "goodness of heart and disintercstcdness" (pp. 71-72). While Michels briefly suggests thai democratic leaders may thus form a new aristocracy, the rule of the lew but also the rule of the best. he turns shortly to the second piilarofthe theory, the leaders' motivations, iheir practices, and the resulting tendency to autocracy.
EMPIRICAL OBSERVATIONS
Michels noted in respect lo leaders a phenomenon lhal political scientists and citizens lend lo notice even today, that once in office, leaders tend to stay there for long periods. I

Certain individuals, simply for the reason thai they have been invested with determinate functions, become irremovable, or at least difficult to replace. Every democratic organization rest.1;, by its very nature, upon a division of labour. But whenever a division of labour prevails, there is necessarily specialization, and the specialist.': become indispensable (Michels. 1959. p. 10!).
In addition, Michels argues, leaders would not be willing to serve at all if (hey were "likely to be dismissed at any moment"; to justify their acquisition of complex skills, they must be assured of job stability (Michels, 1959, p. 102). The masses are grateful to leaders, even to the point of idolatry, Michels contends, and (he masses fear chaos if the leaders are lost. The masses even acquiesce when leaders line their own pockets, since power and wealth "fascinate" (he masses and stimulate the ambition "of alt the more talented elements (o enter the privileged bureaucracy of the labour movement" or beyond (p. 161). "The desire to dominate, for good or evil, is universal" and "every human power seeks to enlarge its prerogatives" (pp. 206-207).
Because leaders have. once in position, their own interests independent of the masses (although this is clothed in democratic rhetoric), and have many instruments at their command to manipulate the masses, most challenges to the leaders' positions come not from the masses but from contending elites, according to Michels. These challenges are opportunistic, he says, and occur when the incumbents show signs of weakness by failing to maintain touch with their supporlers; or may be based on personality conflicts between "great men"; or on age-cohort differences, on differences in social background, or on regional stratification (Michels, 1959. p. 167). The result of such struggles should by now be predictable to Michels' readers: The conflicting parties do not defeat each other but instead accommodate one another, they "surrender a share of the spoil" to each other (p. 169).
EVALUATING THEORY
To the original question posed at the beginning of the chapter. What is a theory?, we now have one answer, Michels' iron law of oligarchy, which links specified variables (the division of labor, attitudes of masses and leaders, government outcomes) into a coherent, logically linked pattern of relations that explains some part of human activity. Beginning with democracy. Michels adds the incompetence of masses, as masses. to act; brings in delegation thai creates representatives who become specialists in the public division of labor; shows lhal representatives develop their own interests independent of those of the masses, and these leaders arc able to control the masses. Thus. democracy leads to oligarchy, and (his happens in all cases. Notice you need not agree with it to accept thai // i.v 11 theory, a connected set of propositions that say something definite about the real wortd.

Michels* iron law explains only certain phenomena, however. !t does nol explain (I) how people govern, whether we!! or ill, (2) what policies they put into effect; (3) which groups are favored or disfavored; (4) why leaders respond differently to similar problems; and so on. So having a universal theory does not imply you have a complete theory; there is always new work to be done.
IS THERE A USE .FOR LARGE FUZZY THEORIES?
If Michels' iron law is one answer to the question, What is a theory?, there are other answers that take an entirely different approach. Among these is functional theory, which directs its attention not to specific individuals and their class-related organizational activities, but to the operation of social wholes as social wholes. The political science versions of functionalism, sometimes called structural-functionalism (Almond and Coleman, I960), or sometimes not called functionalism at all, as in Easton's refusal to call his obviously functional systems theory by that name (1965) have now fallen on hard times. Both existed in the climate of the behavioral period, and except where incorporated in textbooks have both fallen out of favor as the behavioral movement fell out of favor, or perhaps more so.
Students of comparative politics should not, however, delude themselves that change in political science is linear, or that what is dead and buried today will not turn up again in the future, when everyone has grown tired of the currently stylish approaches and is looking for something that seems new and different. Indeed, where many political scientists think functionalism should be written off as a failed attempt to grasp political wholes such as societies and nations, other political scientists think functionalism is one oflhe best and brightest of theoretic hopes (Dogan and Pelassy, 1984). American political scientists especially, who are prone to accept only individualistic theories, for no better reason than that American culture trains us all in individualism as a social norm, need (o learn lo appreciate alternatives such as functional ism that throw events into a light entirely different from microanalytic theories.
Structural-functionalism as put forth by Almond and his several collaborators provides a place to reconsider funclionalism, because it amply illustrated the major flaw in most "grand" funclional theories. This flaw was their positive or optimistic bias, their assumption that social and political systems could be assumed to be carrying out all the functions or systemic necessities that were required to keep the system in good order. Thus the sociologist Parsons postulated functions such a "adaptation" and "pattern maintenance." which seemed to enshrine the status quo as the best of all worlds; and Almond and Coleman defined the political functions ot rule making and execution, interest articulation and aggregation, which seemed to r fcr primarily to an optimally functioning, democratic, political system.
[o:aslons systems theory look a similar lack: Readers of Raston's systems tn ogy

interest rather than in the system's interest, and that indeed leaders may be hazardous to the health of their followers; and yet he brushes these facts aside and allows theoretic access only to behavior that expresses the "public interest" or the needs of the whole system.''These strategies of thought led justly to criticisms that functionalism was conservative, concerned primarily with not rocking the social boat; idealistic, based on Ihe most unrealistic and frequently indefinable assumptions, such as social equilibrium; and nontestable, not applicable to events in the real world. But are these faults inherent in the nature of functionalism or just in the work of the particular political scientists who have practiced it up to this time?
MIDDLE-RANGE FUNCTIONAL THEORY
There is an alternative functional tradition to which political scientists often refer but never seem actually to practice: that so-called "middle range" functional theory associated with the name of Robert Merlon, a sociologist who took a quite different approach to functionalism than did Parsons- Merlon's approach can be illustrated briefly by one of his best known studies, which inquired into the nature of the political machines and bossism so prevalent in American cities earlier in this century (Merton, 1957,pp.71-82).
Where many political scientists of the period reacted to machines and political corruption in the cities by initialing or advocating reform movements, the use of professional city managers, and so on, Merton took a middle-level functional approach, asking why the machines existed, or, in other words, what function they served. His answer was that bosses and machine organizations served a social welfare function in limes and places where there was no other source of such help. When a man was out of work he could go to the local machine representative and ask for, and probably get, either a job or something to tide the family over; when people got in trouble of various sorts, the machine had ways of untangling things; if people were too poor 10 celebrate holidays, the machine turned up with a lurkey or children's toys. In return, of course, those who had been aided (or hoped to be) were expected to carry out the wishes oflhe organization when it called upon them. Functionaiism of ibis sort puts cily machines in a different light. The problem is not so much corruption but poverty and ignorance. You don't need a professional city manager so much as an improvement in economic conditions and educalion, because when these change, the machines naturally lose their hold.
THEORY AS A GUIDE TO INQUIRY
This version of functionalism is quite dilTcrcnl from the "grand" son of functional-ism, iind this middle-ran^c funclionalism is <i fine and subtle instrument waiting to be

ing that all working socielies must. by definition, be fulfilling some highly ;>bslracl functional necessities, middle-range functionalism sets researchers thinking about ihe kinds of lower level functions thai may be relevant to social behavior: about the kinds of people who may or may not fulfill the functions: and about what happens when no one fulfills the functions. For instance, what happens when the social welfare function (the need for help of poor and marginal individuals who are unable to survive on their own) is not met? One outcome is that the people in question starve-read Dickens on the disastrous conditions in the English slums oflhe nineteenth century. Generally. starving people do not rebel, so that is not a likely outcome, although one may wish to reconsult Scott and Popkin on the issue-
Anothcr possible outcome oftlic failure oflhe welfare function is a deep. abiding discontent that creates a functional opportunity for someone to fan the discontent and achieve political prominence on the basis of fulfilling this unmet "functional necessity." This suggests thai the functionalists were right about the existence of certain social needs that are larger than the needs of isolated individuals; insofar as groups of people in a society are in similar circumstances, and feel lha( something is not "right" about their lives, then a social need exists-there is a functional "necessity" ripe for exploitation.7
Middle-range functionaiism is useful in directing attention beyond the actual observable operation of societies to their "invisible" aspects, the opportunities or dangers presented by certain configurations of events- A society in which there is excessive crime or violence offers opportunities for certain types of political behavior. and for politicians, in office or candidates for office, who advocate certain solutions. A large-scale comparative framework might be built on the type of functions different kinds of societies meet or fail to meet. By their problems, and their non-problems. societies might be compared almost independently of their actual government systems, except so far as those systems themselves create needs. Notice too that Michels' theory can be recast in middle-range functional terms.
THE POLITICAL ECONOMY APPROACH
There is a third major answer to the question. What is a theory?, that is different from either oflhe foregoing approaches, and this is rational choice theory, encountered earlier in the discussion of Downs and more recently with Popkin's peasants. Rational choice theory is frequently placed at the apex of the theoretic continuum, among those who practice this variety of theory, but it is preferable to see rational choice theory rather as one option among equals, for despite its admitted strengths, they come with such severe associated weaknesses that tlic theory, like all others, must be treated with skepticism and a willingness to substantially expand ils premises.
Kalional choice theory has never achieved in comparative politics anything like the popularity it lias enjoyed among specialists in U.S. politics, or even among s u-denl.s of international relations. I his difference is revealing of many of the problems

of rational choice theory as a genera! theory of politics, it may be quite effective when it is possible to hold constant a whole variety of institutional and cultural variables. But when rational choice theory faces the wealth of variation found in the field of comparative politics, it quickly falls short.
In criticizing rational choice theory, however, one must be careful to remind oneself of its heroic role in the development of the political science discipline, a role that newcomers may overlook because they did not live through the behavioral revolution personally. If the "old" political science was swept aside during those years, rational choice theory was in the vanguard of (he attacking forces, and was much more formidable than simple empiricism because it used all the weapons of theory, logic, and economic analysis. Why was it so destructive? It was destructive primarily because its approach undercut any sacredness that political institutions might have had for traditional thinkers, the ideas of democracy as an especially public philosophy, the belief in humankind's ability to construct for itself institutions that would be worthy of its higher nature,
THE ICONOCLASTIC FUNCTION
In the place of these hoary ideas, rational choice theorists such as Downs (1957), Simon (1965). Riker (1962) and Buchanan andTullock (1962) put a model of man as individual, rational, greedy, and calculating; and the model applied to leaders and followers alike. The power of this idea was increased by its being presented with many of the most sophisticated tools oflhe economic profession, for the reputation of which political scientists have always had considerable reverence. But equally important was that rational choice theory so plainly predicted just the kind of behavior one could see by looking around at the existing practices of government and politicians. Men did wheel and deal, and public institutions often seemed to operate more for the private than the public good.
This said, however, the rational choice schools'originality was not maintained as new generations succeeded the founding fathers. Because so much of the credibility of rational choice theory depended upon its being an absolutely universal theory, more attention was devoted to ingenious efforts to prove its applicability to situations where it seemed inapplicable, than on the more sensible course of considering how the foundations of the theory might be enlarged beyond narrow economic behavior to wider issues of cognitive, intellectual, and norm-laden behavior. A recent suggestion of how this goat might be achieved is Aaron Wildavskys suggestion of a "cultural rationality, which allows an expansion of human goals beyond monetary self-interest to a far wider range of culturally conditioned goals such as duly. service, salvation, or power(Wildavsky. IW}. This is un old idea. tracing back at least in modern political science to Ihc work of 1 larold I.asswell (Lasswell and Kaplan, 1950), but it has never previously been incorporated into empirical research models. The modification is especially relevant to students <if comparative politics because it builds into the deduc-

live base of rational theory a model of variant idea systems on which choice can be based, thus combining logical rigor with a wider appreciation of the many goals- ethical, idealistic, or social, as well as economic-upon which people act in their political lives.*
THE CULTURAL RATIONALITY MODEL
Comparative politics is not without resources to meet Wildavsky's suggestion, and the source of these resources illustrates why political scientists may profit from reading the works oftheir predecessors. A book about political culture, published as long ago as 1965, contained a series of studies on various countries in the developing world that encapsulated iheir operating principles in a manner fully appropriate to insertion in a rational decision model (Pye and Verba, 1965). Decision rules for members of Turkish village society as suggested by the analysis of Dankwart Rustow, might, for instance, be formalized as follows (Pye and Verba, l965,pp. 171-198; Lane, 1992).
/. The group s welfare is more important than the individual s.
2. But where' fhere is no coufncs, she individual is free to act on his own (to
pursue personal wealth, for instance).
3. The group is a hierarchy based upon age (which measures "wisdom " or
experience).
4. The group s welfare is defined by the elders; and therefore their directions are So be loyally obeyed.
5. These directions typically include:
a. Islamic aiislerifv h. Courage in protecting the group (martial bravery, stoicism, etc.)
While ordinary people will not of course work out their beliefs with such form;ilily, this cultural idea system serves to make explicit the foundations upon which village behavior seems to proceed. Although it defines a pattern of behavior and not a conscious logic, the belief model is obvious, simple, and with minor modifications applicable to many similar societies or groups.
Leonard Binder s description of middle-class Egyptian political culture descn s
another "rational" choice model, and shows how a norm system can be boln tional" and "modem" (Pye and Verba, 1965. pp. 396-449).

3. Where the hierarchical rotes are inapplicable, the individual is expected to behave competitively (Lane, 1992. p. 371)
Since mis model is taken from rural Egypt, it is curious to note how similar it is to the bureaucratic model developed, for instance, by Gordon Tullock (1965) and others.
These examples, with their diversity of belief, of logic, of strategy, and ofpol-icy, vividly illustrate how narrow is the self-interest axiom of the rational choice theorists, but also how easy it is to open it out to include widely diverse attitudes- Ali are, in a sense, self-interested, but the manifestations are so sweepingly unlike simple economic rationality, and so unlike each other, that the cultural model's necessity and its vitality become clear.
METHODS OF THEORY BUILDING
A constant theme throughout this discussion of the recent history of the field of comparative politics has been the centrality of theory in political research. Behavioral-ism was a leap into an unknown future where theory would preside; development research proceeded under the field's firsi attempts at theory; the statists revived old European theories in the hope of making their ideas useful for comparative politics;
and state-society theory and the new institutional theories have adopted yet other theoretic interests. But this brief review reveals why nothing definitive has yet been said, in the present essay, about the nature of theory.
The reason for this reticence is that theory has, over recent decades, changed its meaning more than once. To retain a belief that there is one "best" type of theory is to ignore the diversity of the theoretic enterprise. That reality-the multiplicity of types of theory-is a major reason for the title's being the "arl" of comparative politics. There are of course rules for theory building, but Ihe rules are in themselves inadequate for comparative research because they are incomplete. Science must therefore proceed with the assistance of art-open-minded creativity in working with the materials that make up the field in which one has chosen to work. Those who do research
in the physical sciences have always known this; social scientists are beginning 10 catch on.
The do-it-yourself nature of the sciences is the argument made in one of the classic works in social science theory. Lave and March's Introduction to Models in the Soda/Sciences (1975). Since the lesson is one that manycomparativists have learned on their own bul have never paused to articulate, it is helpful to use this workofmethod-°logy to clarify some of the fundamenl;il techniques involved in building theory or '"odels (the authors use the terms interchangeably). Lave and March emphasize a way 01 looking at the world thai is innovative. Instead of asking the scientist to collect fcamsofdata and allowing the compuler to do the analysis. Lave and March advocate
shriost quaintly old-fashioned approach to inquiry-thinking about facts, in a °minon-sensc, probing, intclligeni way. The basic rule thai guides inquiry, according

to these authors, is to start with a known fact, often a perplexing fact. and to ;isk oneself "how it might have come about." The answers to this question become the hypotheses of the model or theory, and can then be tested against observed reality.
"DUMB" THEORIZING
Lave and March delight in the invention of close-to-home, simple examples of how the theory-building process works, in order to puncture the common notion that "theory" is an esoteric exercise suitable only to geniuses- As an example of this approach, take their case of the "dumb" athletes, or. to put it more politely, the question of why it is in school situations that persons active in athletics are so often stereotyped as less
academically acute than their nonalhlelic schoolmates.
The question with which Lave and March begin is not "Why are athletes dumb?"
b'ul (and this is a major shift in the style of inquiry) why are they "perceived" to be so? Of course, this is quite a different matter, and it forces the inquirer to use his or her head in an active, positive, imaginative way. Lave and March's suggested theories for this small but not uninteresting case are easy enough-one might easily have thought them up oneself, had one put one's mind to it. First, athletes' academic weakness may simply be a result of lack of time; they spend so many hours on the track or field, or in the gym, that they do not have time left for study. To test such a theory the inquiring theorist need merely observe athletes in their off season and see whether
their performance goes up.
But one theory alone will not do the job. so the authors explore further possibilities, such as that human beings generally need to excel in only one field in order to maintain their self-esteem, and athletes-being satisfied in their athletic work thai provides esteem-do not bother to seek further glories. Another theory proposed by Lave and March, and one that demonstrates a truly scientific objectivity, is that athletes are not dumb at all but that their peers are jealous of their success and systematically denigrate their intelligence even when they do well in schoolwork (Lave and March, 1975. pp. 58-60). Thus, Lave and March illustrate in simple cases that theory is not mystical but practical. This is a vital lesson, and one that comparativists have
learned, in their own way, over recent decades.
Lave and March describe several characteristics of this pattern of theory-construction that help lo clarify what such theory entails. They emphasize first thcsiu > of "processes" underlying observable events; second, the use of "implications as dynamic way of tying together sequences of facts; third, the effort to generalise that ones conclusions are carefully raised lo higher levels of scientific sigm ica and finally, "explanation" as the criterion of good theory (Lave and March. -_r^ 4(M2). Examples from recent comparative politics research put flesh on these
relic bones. .
of Barrington Moore's analysis of democracy, fascism, and commurnsn^

interest is centered almost wholly not on (he governmental forms but on the underlying forces thai produced them. "Process" means the identification of the basic relevant actors in the Field of study (aristocracy, peasants, bourgeoisie), and (he delineadon of how (hey interacted, each according to its own resources, goals, and circumstances, to create the different government outcomes.
THE LOGIC OF IMPLICATIONS
Lave and March's second point-Ihe use of implications in order to build a rounded model of a political process-is especially evident in Huntington's (1968) analysis of the development process, because Huntington "fills in" the matrix of known facts by providing general connectors that show the political logic of various situations, a logic that is not always directly observable. In modeling the predicament of a king forced to pursue development policies lest his competitors make backwardness an issue, Hunlington credits him with a least some ability to peer into the future, making specific implications from known regularities ofpolitical life. "The bourgeoisie is with me today because they are hungry for economic opportunity, but when they get richer and more powerful they will become my worst enemies."
Lave and March show how strong such a theoretical approach can be, despite its differences with simple empiricism which only collects facts and would think it presumptuous to go "beyond" those immediately observable facts. Familiarity with the comparativists discussed in earlier chapters shows, however, how dynamic implications can reveal underlying causes and enrich inquiry.
The issue ofgeneralizability. Lave and March's third point, is essential to science, which seeks to discover "universal" facts. Positivism, (he philosophy of science lhat was adopted by the behavioral ists in political science, argued that there was only one way to achieve universality: by discovering propositions of the Form "All x are y," and not allowing any qualifying adjectives to weaken the force of the "all." It was nol considered proper, for instance, to say that economically motivated persons would do x, and politically motivated persons would do y, and so on. The qualifying adjectival phrases made such statements equivalent to saying all persons are mortal except ^metimes they are nol, according to this viewpoint.
Generalizability takes on a new meaning in Lave and March's model-building ^proach. Instead of collecting variables inio grand statistical summaries, the new approach carves its way inio the center of political processes and shows that, underlying e "'fTerent surface facts, a single core process is involved. It is ihis identification a core political process lhat makes comparative models universal. A fine example 01 this pattern is Migdais (1988) "survival" theory, which begins wilh observations n peasant villagers who must adapt their life strategies lo the exigencies around em Physical exigencies such as land, climale and crops; and social exigencies such
e we^ to placate the local strongman, whose assistance (hey may need when methmg goes awry-bad harvests, (rouble wilh the law, illness.

But Migdal's theory does nol stop with this first level, it rises to notice thai strongmen also have their problems of survival (rebellions clients, insufficient resources); as do local officials (rebellions strongmen, bureaucratic superiors), and so on up to the lop of the political pile where presidents and other high politicians also struggle for survival against their own particular odds. In this way, clarification of a "process" at one level may be generalized to many levels, and may create a model with wide-ranging applicability.
FIND YOURSELF AN EIGHT-YEAR-OLD
What is a "good theory"? For Lave and March it is "all of the above" plus one final defining characteristic, the quality ofexpianation. Explanation is obviously necessary, but is slightly ambiguous as a criterion because different people may find different explanations appropriate, depending on their level of information and background knowledge, and on their purposes. One workable solution to this problem is to specify to whom explanations should be directed; and one plausible audience may be defined as bright eight-year-old kids. People who use this criterion argue that children of this age are sufficiently grown up that they are curious and critical-they are, in fact, the ones who notice when emperors have no clothes-and they are good natural judges of what "makes sense." Older children, according to this argument, have already begun to take on the quality of adults, and have a tendency to be satisfied with conventional theories, or to be impressed by empty and pretentious ones.
Whatever audience one chooses, "explanation" involves presenting all the major elements in a political process, showing how they interact to produce observed outcomes, and how these outcomes restructure the future interactive situation by the rules and resources and sometimes even the participants. As this makes plain, the explanatory emphasis lies in the logic of the causal connections. The older method of evaluating theory was its ability merely to predict, and this hardly expanded ones knowledge- Prediction would have been content to say that if you slept overnight in the marshes you would catch malaria. This might have been sufficient for the health-conscious, but not for science, which was only satisfied when it identified the mosquito's role in carrying the plasmodium that caused the disease.
EPILOGUE
r ' *^ ll
What then is theory? Or, more specifically, what is theory in comparative politic!*. is a quest, as the history here recounted has shown- But it is not a fruitless quest, o conclusions have been reached. Comparalivists have, over a couple score o _^ reached an opening in the woods, where the visibility is better than it is in the 'o'o^ parts of the forest. In this "clearing" comparativists find themselves increasing,y r^mtni-i;.hl.' with whole-svsten-i neneralizing that cannot be analyzed operali

into concrete individuals or unified groups. "Progress" or "development" or "demw racy" are no longer seen as forces working themselves out in the everyday world
Activity instead is sited in individual human beings who may act to brine about progress or democracy, if they define their goals as being forwarded by such behavior; but individuals also who may not be useful to the system- Many tyrants prefer their people (o live in poverty, ignorance, and fear. Even popular democratic politicians may prefer that their activities not be too transparently clear to the people they govern. The state-society perspective on political systems has brought comparative
polilics to this realistic appraisal of political behavior, and it will be difficult to retreat from such an obviously practical viewpoint.
To summarize the argument of this review of comparative history, it is useful to bring together the elements of this model thai various comparativists, working on their own but within the larger context of the comparative field, have over the years hammered into shape. The answer might be left as an exercise, to be "read off" from earlier chapters (hat showed that a "politics" model kept turning up as a way of providing deep political explanations for a variety of individual problems and topics. But
the model is compact and easy to summarize-as models indeed should be-and it serves as a conclusion to the present analysis.
The politics model defines individual actors as engaged in a dynamic process within institutions that their behavior serves to reinforce, to modify, or to overthrow. This political process centers on the practical confrontation of individual persons as they debate or negotiate issues that involve not only the distribution of the goods and evils of life, but of the rules controlling both distribution and overall governance. The key to this political confrontation lies in the type and amount of resources the participants can call upon; one recalls here Huntington's memorable phrase about the rich using bribery, the students using riots, and the military using coups (1968).
Political struggles are not always as clearly observable as one might expect- In many societies the politics seems silept and there appears to be perfect tranquility- until someone imposes the straw that breaks some camel's back, and the peace is shattered. Only depth analysis can recognize the possibilities inherent in some situation before they actually rise to the surface,
When the comparativist goes forth into the field, to study some country, town, village, organization, some group of any kind, does the comparative politics model
provide research guidance? Study of those who have practiced it suggests that the model does provide guidance.
o The first step is 10 determine who are the major participants, those who are actively engaged in, or appear as if they may come to engage in, the business of the group. These activists are distinguished from the "audience" in order to simplify
the analysis, but it is remembered thai audiences may, if conditions change, break into the action arena.
For all participants must be defined the idea systems, the beliefs, attitudes, goals, 3nd values that structure their view nfihp world

the arena. Important here are both individual ideas and group ideas, the latter studied under the heading of political culture.
o The next step is the discovery of the "rules," both explicit and implicit, under which the group operates. Who has the right to take what kinds of actions-who pays, who spends, who obeys, who rules? And, of course, who disdains following rules?
" The realistic estimation of the participants' resources is the next step. Resources here include not only physical resources but the ideas that may give some members precedence over others; the allies lhat may be called upon; the numbers of persons in some affected groups; and even obstinacy, the triumph of will over mere facts.
o The nature of the interaction in any comparative politics model will depend on the underlying structure of the situation, which defines what the opportunities and constraints on behavior will be, on the basis of the types of participants, their number, their resources, and their ideas. A field in which all actors are equally endowed, both cognitivety and physically, wilt lead to equitable, light government. A field in which one man is aggressive and ninety-nine others are timid will surely become an autocracy.
o The structure of the situation will, next, create certain functional opportunities, physical and social realities that qualify as needs to be met and invite members or outsiders to fulfill them. Thus a famine creates a need for food for the starving, and a political opportunity to those who provide it. Or an incompetent leader creates a functional opportunily for a coup.
The beauty of the politics model as practiced by many of the comparative theorists discussed in the course of this essay is the logical structure within which these elements are combined. One comes across the pattern again and again in the works of Huntington, Migdal, Wallerstein. Moore, Skocpol, Evans, Scott and Popkin: Because of a given situation (inadequate room for national expansion, or a weak bourgeoisie, or oversized turbulent urban populations, or sandy soil), then a certain situation arose, which changed the political situations of the participants, so that a new situation ensued and so on until one reaches the conclusion that this interlocked set of processes has provided an explanation.
It Is a simple model, and it works. That has been the moral of our tale.
NOTES
1. Ball (1987) has made this point forcefully in his discussion of the relation between the philosophy of science and political science. He suggests thai political scientists must become less dogmatic about falsifying theories, and more charitable in evaluating them (pp. 34--t 'o
2. Dahl as well as Michels has sli-ong social democratic sympathies (Baer, Jewell. Sigelman, 1991). Note thai Dahl's refulalion of'a single coherent elite docs not really nice Michels'argument, since Michels never said there was such a unified clile. only elites.

3. Stinchcombe (I968) provides an excellent discussion of ihe several levels of theoretical construclion.
4. The issue is a complex one. Sludenis should begin with Hempel (l°65) and Popper (1959); and can bring themselves into the current level of discussion wilh Giere (1988).
5. Sec the discussion in Chapters 2 and 4 above.
6. Easton raises ihc issueofnonresponscbyeliies late in his discussion of the syslems approach, and admits that authorities may seek to pursue their own purposes even against the objectives of other members. But he concludes: "we do not have to pursue this line of discussion further" (1965a, p. 441). This refusal to look at negative systems is perplexing in a theorist who sought to build a general theory,
7. Terry Moe touches on these issues in his organization theory, which emphasizes the role of entrepreneurs who invent new interest groups in areas where old interest groups fail to cover issues relevant to some sectors of the public (1980).
8. See Lane (1992) fora fuller discussion of cultural rationality.
REFERENCES
Almond, Gabriel A., and James S. Coleman (Eds.). The Politics of the Developing Areas. Princeton: Princelon University Press, I960.
Baer, Michael A., Malcolm E, Jewel!, and Lee Sigelman (Eds.). Political Science in America:
Oral Histories of u Discipline. Lexinglon: University of Kentucky, 1991.
Ball, Terence- (Ed.) Idioms of Inquiry: Critique and Renewal in Political Science. Albany:
SUNY Press, 1987.
Buchanan. James M. and Gordon Tullock. The Calculus of Consent {\9kl). Ann Arbor: University of Michigan Press, 1965.
Dahl, Robert A. Who Govern.'.'? New Haven: Yale University Press, 1961.
Dogan, Maltei, and Domimquc Pelassy. How lo Compare Nations: Strategies in Comparative Politics. Chatham. Chatham House, 1984.
Downs. Anthony- An Economic Theon-of Democracy. Mew York: Harper and Row, 1957.
Easton. David. The I'olnic-.il Sy.wm: An Imfiiiry into the Stale of Political Science. New York:
Alfred A. Knopf, 1953.
Easton, David. A Framework for Politico! Analysis. Englewood Cliffs: Prentice Hall, !965a.
Easton, David. A Systems Analysis of Political Life. New York: John Witey and Sons, 1965b.
Giere. Ronald N. Explaining Science: A Co^nitiw Approach. Chicago: University of Chicago Press. 1988,
Hempel. Carl Gustav. Ax/wcis of Scientific Explanation And Other Essays in the Philosophy of Science. New York: The Free Press. 1965.
Huntington. Samuel P. Political Onh'r in Changing Societies. New Haven: Yale University Press, 1968.
Lane, Ruth. "Political Culture: Residual Category or General Theory?" Comparative Political Slwlies 25. no. 3 (October 1992): 362- 87.
Lasswcll. ilamltl D, and Abraham K.ipl.m. Power ami Society: A Framework for Political Inquiry New Haven- Yale Universily Press. [950.
Lave. Cimrles. and James March, An Ininnluiiinn to Modeli in lln' Social Science?.. New York:
l!arpcr& Knw. 1975

• [ На главную ]
• [ Библиотека ]
• [ Reader ]
• [ Программы ]
• [ Абитуриенту ]
• [ Фотоальбом ]
• [ О ФП ]
• [ Гостевая книга ]
• [ Ссылки ]
• [ Обратная связь ]