[These excerpts from Fuller's article reveal both his method and his basic thesis.]

It may be said that the essence of adjudication lies not in the manner in which the affected party participates in the decision but in the office of judge. If there is a judge and a chance to appear before him, it is a matter of indifference whether the litigant chooses to present proofs or reasoned arguments. He may, if he sees fit, offer no argument at all, or pitch his appeal entirely on an emotional level, or even indicate his willingness that the judge decide the case by a throw of the dice. It might seem, then, that our analysis should take as its point of departure the office of judge. From this office certain requirements might be deduced, for example, that of impartiality, since a judge to be "truly" such must be impartial. Then, as the next step, if he is to be impartial he must be willing to hear both sides, etc.

The trouble with this is that there are people who are called "judges" holding official positions and expected to be impartial who nevertheless do not participate in an adjudication in any sense directly relevant to the subject of this paper. Judges at an agricultural fair or an art exhibition may serve as examples. Again, a baseball umpire, though he is not called a judge, is expected to make impartial rulings. What distinguishes these functionaries is not that they do not hold governmental office, for the duties of a judge at a livestock fair would scarcely be changed if he were an official of the Department of Agriculture. What distinguishes them from courts, administrative tribunals, and boards of arbitration is that their decisions are not reached within an institutional framework that is intended to assure to the disputants an opportunity for the presentation of proofs and reasoned arguments. The judge of livestock may or may not permit such a presentation; it is not an integral part of his office to permit and to attend to it.

If, on the other hand, we start with the notion of a process of decision in which the affected party's participation consists in an opportunity to present proofs and reasoned arguments, the office of judge or arbitrator and the requirement of impartiality follow as necessary implications. The logician Frege once took the expression "I accuse" as exemplifying the complex implications contained in the most ordinary language. We may say that the verb "to accuse" presupposes five elements: (1) an accuser, (2) a person accused, (3) a person before whom the accusation is presented, (4) an act charged against the accused, and (5) a principle by which the act may be condemned. The similarity to the analysis here presented is apparent; the fifth element, it should be noted, corresponds to the notion of a reasoned argument. Of course, Frege was concerned merely to spell out the implications contained in a phrase, not, as we are here, with the problem of creating and maintaining a social institution that will give effect to those implications.

It may be objected at this point that "reasoned argument" is, after all, not a monopoly of forensic proceedings. A political speech may take the form of a reasoned appeal to the electorate; to be sure, it often take other forms, but the same thing may be said of speeches in court. This objection fails to take account of a conception that underlies the whole analysis being presented here, the conception, namely, of a form of participating in a decision that is institutionally defined and assured.

When I am entering a contract with another person I may present proofs and arguments to him, but there is generally no formal assurance that I will be given this opportunity or that he will listen to my arguments if I make them. (Perhaps the only exception to this generalization lies in the somewhat anomalous legal obligation "to bargain in good faith" in labor relations.) During an election I may actively campaign for one side and may present what I consider to be "reasoned arguments" to the electorate. If I am an effective campaigner this participation in the decision ultimately reached may geatly outweigh in importance the casting of my own single vote. At the same time, it is only the latter form of participation that is the subject of an affirmative institutional guarantee. The protection accorded my right to present arguments to the electorate is almost entirely indirect and negative. The way will be held clear for me, but I shall have to pave it myself. Even if I am given an affirmative right (for example, under the "equal time" rule of the FCC) I am given no formal assurance that anyone will listen to my appeal. The voter who goes to sleep before his television set is surely not subject to the same condemnation as the judge who sleeps through the arguments of counsel.

Adjudication is, then, a device which gives formal and institutional expression to the influence of reasoned argument in human affairs. As such it assumes a burden of rationality not borne by any other form of social ordering. A decision which is the product of reasoned argument must be prepared itself to meet the test of reason. We demand of an adjudicative decision a kind of rationality we do not expect of the results of contract or of voting. This higher responsibility toward rationality is at once the strength and the weakness of adjudication as a form of social ordering.

In entering contracts, men are of course in some measure guided by rational considerations. The subsistence farmer who has a surfeit of potatoes and only a handful of onions acts reasonably when he trades potatoes for onions. But there is no test of rationality that can be applied to the result of the trade considered in abstraction from the interests of the parties. Indeed, the trade of potatoes for onions, which is a rational act by one trader, might be considered irrational if indulged in by his opposite number, who has a storehouse full of onions and only a bushel of potatoes. If we asked one party to the contract, "Can you defend that contract?" he might answer, "Why, yes. It was good for me and it was good for him," If we then said, "But that is not what we meant. We meant, can you defend it on general grounds?" he might well reply that he did not know what we were talking about. Yet this is precisely that kind of question we normally direct toward the decision of a judge or arbitrator. The results that emerge from adjudication are subject, then, to a standard of rationality that is different from that imposed on the results of an exchange.

I believe that the same observation holds true when adjudication is compared with elections. The key to the difference lies again in the mode in which the affected party participates in a decision. If, as in adjudication, the only mode of participation consists in the opportunity to present proofs and arguments, the purpose of this participation is frustrated, and the whole proceeding becomes a farce, should the decision that emerges make no pretense whatever to rationality. The same cannot be said of the mode of participation called voting. We may assume that the preferences of voters are ultimately emotional, inarticulate, and not subject to rational defense. At the same time there is a need for social order, and it may be assumed that this need is best met when order rests on the broadest possible base of popular support. On this ground, a negative defense of democracy is possible; the will of the majority controls, not because it is right, but - well, because it is the will of the majority. This is surely an impoverished conception of democracy, but it expresses at least one ingredient of any philosophy of democracy, and it suggests a reason why we demand of adjudication a kind of rationality that we do not expect of elections.

This problem can be approached somewhat obliquely from a different direction by asking what is implied by "a right," or by "a claim of right." If I say to someone, "Give me that!" I do not necessarily assert a right. I may be begging for an act of charity, or I may be threatening to take by force something to which I admittedly have no right. On the other hand, if I say, "Give that to me, I have a right to it," I necessarily assert the existence of some principle or standard by which my "right" can be tested.

Attention is now directed to the question, What kinds of tasks are inherently unsuited to adjudication? The test here will be that used throughout. If a given task is assigned to adjudicative treatment, will it be possible to preserve the meaning of the affected party's participation through proofs and arguments?

This section introduces a concept - that of the "polycentric" task - which has been derived from Michael Polanyi's book The Logic of Liberty. In approaching that concept it will be well to begin with a few examples.

Some months ago a wealthy lady by the name of Timken died in New York leaving a valuable, but somewhat miscellaneous, collection of paintings to the Metropolitan Museum and the National Gallery "in equal shares," her will indicating no particular apportionment. When the will was probated the judge remarked something to the effect that the parties seemed to be confronted with a real problem. The attorney for one of the museums spoke up and said, "We are good friends. We will work it out somehow or other." What makes this problem of effecting an equal division of the paintings a polycentric task? It lies in the fact that the disposition of any single painting has implications for the proper disposition of every other painting. If it gets the Renoir, the Gallery may be less eager for the Cezanne but all the more eager for the Bellows, etc. If the proper apportionment were set for argument, there would be no clear issue to which either side could direct its proof and contentions. Any judge assigned to hear such an argument would be tempted to assume the role of mediator or to adopt the classical solution: Let the older brother (here the Metropolitan) divide the estate into what he regards as equal shares, let the younger brother (the National Gallery) take his pick.

As a second illustration suppose in a socialist regime it were decided to have all wages and prices set by courts which would proceed after the usual forms of adjudication. It is, I assume, obvious that here is a task that could not successfully be undertaken by the adjudicative method. The point that comes first to mind is that courts move too slowly to keep up with a rapidly changing economic scene. The more fundamental point is that the forms of adjudication cannot encompass and take into account the complex repercussions that may result from any change in prices or wages. A rise in the price of aluminum may affect in varying degrees the demand for, and therefore the proper price of, thirty kinds of steel, twenty kinds of plastics, an infinitude of woods, other metals, etc. Each of these separate effects may have its own complex repercussions in the economy. In such a case it is simply impossible to afford each affected party a meaningful participation through proofs and arguments. It is a matter of capital importance to note that it is not merely a question of the huge number of possibly affected parties, significant as that aspect of the thing may be. A more fundamental point is that each of the various forms that award might take (say, a three-cent increase per pound, a four-cent increase, a five-cent increase, etc.) would have a different set of repercussions and might require in each instance a redefinition of the "parties affected."

We may visualize this kind of situation by thinking of a spider web. A pull on one strand will distribute tensions after a complicated pattern throughout the web as a whole. Doubling the original pull will, in all likelihood, not simply double each of the resulting tensions but will rather create a different complicated pattern of tensions. This would certainly occur, for example, if the doubled pull caused one or more of the weaker strands to snap. This is a "polycentric" situation because it is "many centered" - each crossing of strands is a distinct center for distributing tensions.

Suppose, again, it were decided to assign players on a football team to their positions by a process of adjudication. I assume that we would agree that this is also an unwise application of adjudication. It is not merely a matter of eleven different men being possibly affected; each shift of any one player might have a different set of repercussions on the remaining players: putting Jones in as quarterback would have one set of carryover effects, putting him in as left end, another. Here, again, we are dealing with a situation of interacting points of influence and therefore with a polycentric problem beyond the proper limits of adjudication.

Now, if it is important to see clearly what a polycentric problem is, it is equally important to realize that the distinction involved is often a matter of degree. There are polycentric elements in almost all problems submitted to adjudication. A decision may act as a precedent, often an awkward one, in some situation not foreseen by the arbiter. Again, suppose a court in a suit between one litigant and a railway holds that it is an act of negligence for the railway not to construct an underpass at a particular crossing. There may be nothing to distinguish this crossing from other crossings on the line. As a matter of statistical probability it may be clear that constructing underpasses along the whole line would cost more lives (through accidents in blasting, for example) than would be lost if the only safety measure were the familiar "Stop, Look & Listen" sign. If so, then what seems to be a decision simply declaring the rights and duties of two parties is in fact an inept solution for a polycentric problem, some elements of which cannot be brought before the court in a simple suit by one injured party against a defendant railway. In lesser measure, concealed polycentric elements are probably present in almost all problems resolved by adjudication. It is not, then, a question of distinguishing black from white. It is a question of knowing when the polycentric elements have become so significant and predominant that the proper limits of adjudication have been reached.

In speaking of the covert polycentric elements almost always present in even the most simple-appearing cases, it should be noted that the efficacy of adjudication as a whole is strongly affected by the manner in which the doctrine of stare decisis is applied. If judicial precedents are liberally interpreted and are subject to reformulation and clarification as problems not originally foreseen arise, the judicial process as a whole is enabled to absorb these covert polycentric elements. By considering the process of decision as a collaborative one projected through time, an accommodation of legal doctrine to the complex aspects of a problem can be made as these aspects reveal themselves in successive cases. On the other hand, if a strict or "literal" interpretation is made of precedents, the limits of adjudication must perforce be more strictly drawn, for its power of accommodation has been reduced. 

If problems sufficiently polycentric are unsuited to solution by adjudication, how may they in fact be solved? So far as I can see, there are only two suitable methods: managerial direction and contract (reciprocity).