This document was scanned
from hard copy, and may contain uncorrected scanning errors.In Defense of AmbiguityChristopher Honeyman This article was first published in Negotiation Journal, January 1987.Experienced mediators are familiar with the negotiator who takes a strong and seemingly unalterable stand on a point of principle—only to exchange it for money when the price is right. But one reason why agreements can be difficult to get is that a matter of principle often is at the root of the dispute. An explicit confrontation over such a question can result in a total inability to work out a settlement; so it should not be surprising that negotiators and mediators sometimes paper over these cracks with calculated or innocent ambiguities. This paper will examine the function of ambiguity in agreements, and defend its deliberate use under certain circumstances. There will always be those who value the effect of ambiguities on their earnings, such as a full-time labor arbitrator of my acquaintance who once declared happily that "There's no such thing as clear contract language." But most neutrals seem to have a vague disapproval of ambiguity. This is excusable. Most neutrals practice a specific profession—mediation, arbitration, adjudication and so forth—and proceed from the "settlement" of one case to another of like kind. A general preference for wrapping up loose ends fits with the notion of "settlement," and a neat and tidy job seems consistent with professionalism. The parties to the dispute generally have a different perspective. Even while negotiating an agreement, they are looking down the road to the later interpretation of that agreement; and in the case of permanent relationships, they are quite likely to be engaged in the arbitration of one dispute, the litigation of another, and the negotiation of a third at the same time. The sense of the liabilities and limits of a written agreement which this generates is heightened by the internal politics of any party which consists of a group. Much of what happens in negotiation can be seen in terms of a struggle between radical and moderate elements within each party. In a multi-faceted negotiation, the fact that the moderate element on one issue may be the radical element on another obscures, but does not change, the essential relationship between moderate and radical. Bear with me if I encapsulate a negotiating group as consisting of a radical minority and a moderate majority. (Where the situation is reversed, effective negotiation or mediation is unlikely). In such a group, the radicals can be expected to emphasize philosophical and ideological purposes, partly out of conviction, but also because this gives them a platform in the continuing attempt to garner public support, and perhaps become the dominant faction. The moderates, meanwhile, are likely to emphasize the practical results of accommodation as opposed to confrontation. I have elsewhere commented (Honeyman, 1985) that a mediator's tendency, or even function, is to help the moderate out-argue the radical within a given party. When possible, it is of course preferable to divert an argument over "principle" into a "pragmatic" channel, so that an explicit agreement can be reached. Since this cannot always be accomplished, the best alternative may be to leave a deliberate gap or other ambiguity in the agreement. In my opinion, clarity heightens the differences, while ambiguity can serve to let each faction serve its position in principle, if not necessarily in practice. The traditional view of ambiguity in an agreement is that it implies the presence of either an unconsidered point or a deliberate failure to come to grips with the problem. In either case, that view amounts to a belief that where there is ambiguity there is no agreement. But as every arbitrator knows, agreements are to be read as a whole. I take the view that the interpretation of an ambiguity depends partly on the enforcement mechanism specified in the agreement. Every agreement contains at least one express or implied means of securing compliance with its terms. Different mechanisms of enforcement generate different results when exposed to ambiguity, and this can be predicted by anyone who possesses a working knowledge of the several mechanisms. For this reason, an agreement, read as a whole, can in fact give an interpretation to the ambiguity. Here a thumbnail sketch of the characteristics of the various enforcement mechanisms is in order, because the choice of negotiation, mediation, arbitration, fact-finding, litigation or unilateral action implies something about the balance of power within each party as well as that between the parties. Since each of these processes is reasonably well understood, at least by sophisticated negotiators, the choice can be an index to the interpretation of other parts of the agreement. In contrast to most of the other processes, negotiation can be summarized as relatively
creative. In a subsidiary dispute arising from an ambiguity in the larger agreement,
negotiation is apt to result in a specially-tailored solution. This may involve elements
never discussed in the original negotiations, and may even add something to the original
agreement explicitly. Though I am most familiar with the operation of labor agreements, I
doubt that they are a special case in the common practice of reaching grievance
settlements based on the facts of the given situation, and including the formula that the
settlement is "without precedent" to either party's subsequent assertion of the
meaning of the original agreement in a different case. This "custom-tailoring"
characteristic derives from the responsibility for resolving the matter being squarely on
the parties themselves: They have the same power to remake their world that they had when
the original dispute was being settled. But as the enforcement mechanism for an agreement,
negotiation clearly has one drawback—it is not binding. Therefore, any agreement
which allows negotiation also provides for an express or implied secondary method of
resolving a subsidiary dispute.
Mediation operates in much the same environment as negotiation. Like negotiation, it implies the existence of a fall-back process. Fact-finding tends to lose the "creativity" aspect, while it introduces the element of a call for public support of a particular solution, made under the neutral's auspices. The choice of fact-finding therefore implies an intent to subject the parties, to some degree, to the opinion—or prejudice—of the larger community. The choice of litigation explicitly adopts as the schema of the parties' settlement mechanism the rules and mores of the society-at-large. Two of these, the burden of proof and the interest in equity, are particularly relevant here. The burden of proof implies that whichever party is accused of violating the agreement by its actions has the advantage. So if one party can be expected to be consistently the actor, and the other the reactor, one might postulate that a radical element on the first side would be happier to see litigation adopted as the enforcement mechanism rather than, say, arbitration, where the burden of proof is less clearly on the complaining party. Yet the general tendency of courts to consider equity principles in doubtful cases may be similar to the arbitrators' habits discussed below; and other factors too complex to discuss here also enter into the equation, so that for a given "reactor" litigation could prove a more advantageous choice of mechanism than this quick overview would indicate. Unilateral action (such as the right to strike as the last step of a union/employer grievance procedure) is clearly a tool more desirable to a radical than to a moderate. It negates the value of the concessions made in the original agreement, and also provides radicals with an excellent opportunity to "pump up" their constituency. Both of these tendencies are likely to allow the radical element to assert itself to a greater degree where unilateral action is permitted than in the course of the other dispute settlement mechanisms. Its existence as part of an agreement, whether express or implied, thus gives a clue to which faction wound up on top in the internal struggle. I have left arbitration until last because, in some ways, it is the most revealing process in this context. Arbitration alone combines the terminating force of an imposed decision with a value system theoretically drawn from the agreement itself. If in fact ambiguity implied the absence of an agreement, arbitrators' pronouncements on ambiguous clauses should reflect that. The fact that they typically do not is instructive as to the underlying intent of the amorphous mass of contending factions which we, oversimplifying, call "the two parties." Some arbitration agreements contain as their terms of reference the external statutes. These are, to all intents and proposes, merely cheaper forms of litigation. But many arbitration proceedings are governed by principles that combine an injunction to the arbitrator to arrive at an award which "draws its essence" from the parties' agreement with an underlying set of nonlegal expectations, expressed in labor cases as "the law of the shop." Labor arbitrators in particular can regularly be heard to denounce the notion that they should apply external law in their decisions, and the system of labor arbitration has continued with surprising stability of attitudes for many years. It is significant, therefore, that the bulk of arbitration awards concerned with ambiguous contract language seem to apply a general rule that ambiguity implies moderation. The long line of subcontracting cases supplies an excellent illustration. In theory, nothing could be simpler than for the parties to a labor agreement to write such language as "The employer may subcontract work." or "The company may not subcontract work." However, either phrase will conflict with a basic ideological principle of one of the parties. In the union's case, the principle is that the union represents those who perform certain work, and by strenuous efforts it has managed to raise their wages. The last thing a union can accept is the notion that an employer can, without restriction, give away the employees' work to the cheapest labor it can find through a subcontractor. At the same time, the question of "make or buy" is basic to the manufacturing process of any complex product. Industries have a long history, predating any organization of the employees, of routine decision-making as to the best and most efficient method of producing components of a product or service. Management therefore finds an ideological obstacle to agreeing to give up the right to make such decisions. Some parties have been able to agree on specific language providing for the right to subcontract under certain circumstances and not under others. But of greater interest here is the plethora of labor agreements which are silent as to subcontracting, even though both parties have been aware for years of the potential or actual issue. These contracts will often contain a seniority clause, a recognition clause and a general management rights clause. Typically in subcontracting cases, the union will argue that the recognition clause, combined with language protecting seniority rights and the specified wage scale for the performance of certain work, shows clearly that in the absence of language specifically allowing subcontracting, the agreement must be read as preserving work to the employees represented by the union. The company, in turn, will argue that the management rights clause (generally using a formula something like "the company shall have all rights of management except as limited by this agreement") clearly shows that, since the union has not managed to negotiate language expressly restricting subcontracting, the company has plenary rights to subcontract any work it chooses in the best interest of the business. And then both parties will knock it off and proceed to try their case based on the ambiguity customarily recognized to arise from these conflicting clauses. The standard desk reference of the labor arbitration trade is How Arbitration Works (Elkouri and Elkouri, 1952 and 1985). The Elkouris accurately describe the various tests used by different arbitrators where no specific contract language (or conclusive evidence of bargaining history) exists concerning subcontracting. Those general tests, all of which discuss the common clauses favoring the union and favoring the company, are less interesting than the fact that they converge on the same practical considerations. The Elkouris summarize these standards as follows: past practice; justification of the present instance; effect on the union or the bargaining unit; effect on individual employees; the type of work involved and its relation to employees' usual work; whether suitable employees or equipment are available in-house; whether the subcontracting will be regular or long-lasting; and special circumstances, such as an emergency. What is significant to this discussion about the standards cited above is that every one of them tends to secure a moderate answer. The resulting fact-driven awards have, of course, regularly prompted the "we was robbed" reaction almost expected of a losing party to a labor arbitration, but have in the larger sense found acceptance. This is proven by the simple fact of the longevity of these standards, and by the parties' bilateral failure to renegotiate the underlying contract language over a long period of time. A second example, even more widely practiced than the first, is the common clause in labor contracts providing that an employee may be discharged (or disciplined) for "just cause." This clause is not ubiquitous; some parties have been able to agree on a laundry list of circumstances that do or do not warrant discharge, and have written that into their agreements. But such an effort generally runs into roadblocks of principle that tend to exacerbate the dispute. Consider, for instance, even so "obvious" a standard for discharge as proven theft: The company says "theft is theft," the union says "what if it's just a pencil?" Or lateness: the union says "you can't fire someone for being late," the company replies "then we'll get a few employees who just keep doing it." In the practical result of countless labor negotiations, the phrase "just cause" is preferred to the divisive thrashing-out of the possible permutations and combinations of circumstance. The phrase is ambiguous in a different sense than the mutually conflicting clauses of the subcontracting example—here there is a single expression, ambiguous because of its vagueness./2 But the purpose and effect are the same. The company agrees to "just cause" because it is hard to maintain a claim that it should have the right to discharge employees for unjust reasons or no reason at all. The union agrees to it because it is hard to maintain that employees who by definition are getting what they deserve should be kept on the job. Even a cursory review of the vast profusion of cases decided under this standard shows that the awards, as in the subcontracting example, are generally moderate in tone and fact-driven. And the parties routinely complain about the result, but rarely change the underlying standard. In either example, the negotiators and mediator who constructed the agreement and the arbitrator who interprets it all show their relationship to the moderate and radical elements on both sides, by the practical—and predictable—result of their actions. The arbitrator in particular—a creature of the agreement and, one hopes, uniquely sensitive to its nuances, serves the moderates' goals by distinguishing between tolerable and intolerable incursions into each party's "principles." The arbitrator couches his or her decision in terms of the agreement, but strains to avoid using the term "equity" in order to escape the accusation of the radicals on the losing side that the award does not "draw its essence" from the agreement. And in turn, the courts refrain from second-guessing the arbitrator by applying general legal principles, adopting instead the "Steelworkers' Trilogy"/3 standards for deferral to arbitration's results. From the point of view of the moderate on either side who desired a workable if not ideal agreement, the various processes of dispute resolution thus form an intricate ecology, in which each depends on the others for the success of the whole. Some may object that this is instead an indication of the alleged tendency of arbitrators to compromise. But an award that does not clearly give the whole issue to either party is not proof of the arbitrator's lack of integrity, nor does it demonstrate that the arbitrator was prejudiced. In a "compromise" decision, the arbitrator presumably is motivated by a desire not to offend anyone, but in the awards under discussion here the arbitrators are properly performing their function, which is to interpret conflicting or vague provisions so as to give meaning to the whole agreement. The net result for negotiators and mediators is that ambiguity can be employed as a tool in achieving an adequate, if imperfect, settlement of a dispute. Provided that the enforcement mechanism is appropriate to this end, leaving ambiguities can imply that the disposition of subsidiary disputes will be fact-driven and moderate in overall effect. This enables the moderates in either group to compel discourse on terms acceptable to them, and to retain control of their party by avoiding reliance on rhetoric and ideology. Ambiguity therefore becomes a sophisticated means for ensuring that the general philosophy of the party does not have to be compromised explicitly (which would allow radicals to claim that the moderates had "sold them down the river"), while allowing enough putative "wins" by the opposing party to make agreement possible. Mediators are institutionally in favor of settlements and should not sneer at an ambiguity when that is a necessary element in obtaining an overall agreement. A possible exception, worthy of ethical discussion, is the situation where the mediator, in either suggesting or allowing ambiguous language, suspects that an ambiguous agreement will not be interpreted according to the standards applied above, but will instead be subject to unilateral action. That involves a highly difficult calculation of the "mutual best interest" of conflicting parties. But where a "moderating" system obtains for disposition of subsidiary
disputes, there is nothing inherently wrong with gracefully admitting the impossibility of
reaching a complete "meeting of the minds." Allowing an ambiguity to pass into
the agreement, in the expectation that it will later be interpreted in terms not likely to
cause a wider dispute, is just another way to skin the cat. NOTES 1. But it is often argued that arbitrators have a tendency to "split the baby" in order to remain acceptable to both sides. I doubt the truth of that assertion, but must note that the fact that it is so often made implies that it is at least widely believed by parties using arbitration. From this it could be inferred that these parties enter into their agreements "knowing," and implicitly tolerating, this expected outcome. 2. William Empson (1947) distinguished a series of different types of ambiguity, though Empson, writing about poetry, identified some types not relevant to disputes. (A "fortunate confusion" is a concept of ambiguity attributable, in the context of an agreement, only to someone who hopes to make a living from it.) Of Empson's list, these seem applicable to disputes: "when a detail is effective several ways at once"; "simultaneous unconnected meanings"; "a contradictory or irrelevant statement, forcing the reader to invent interpretation"; and "full contradiction, marking a division in the author's mind." All of these are consistent with a condition not normally present in poetry—the opposing purposes of plural authors. Varying circumstances can cause negotiators to accept any of these types of ambiguity; study might reveal that certain forms best fit particular negotiating situations, but at present all I am prepared to say is that deliberate ambiguity in an agreement has the same origin and overall purpose no matter which of these forms it may take. 3. These three U.S. Supreme Court decisions set the standards for federal courts to use
in determining both arbitrability of a given dispute and whether or not an award should be
overturned on the merits. Particularly relevant are the Court's statements that the courts
"have no business overruling (the arbitrator) because their interpretation of the
contract is different from his," provided that the award "draws its
essence" from the agreement, and that the award should be upheld unless it is
"clear" that an arbitrator has exceeded his or her authority. See United
Steelworkers of America v. Enterprise Wheel & Car Corp., 80 S. Ct. 1358.
REFERENCES Elkouri, F. and Elkouri, E. A. (1952). How arbitration works. Washington: BNA Books; fourth ed., 1985, pp. 540-542. Empson, W. (1947). Seven types of ambiguity. New York: New Directions. Honeyman, C. (1985). "Patterns of bias in mediation." Journal of Dispute Resolution: vol. 1985, pp. 141-149.
|
|
|