This appeal in a suit over a collective bargaining agreement presents a fundamental issue of contract law, that of drawing the line between an ambiguous contract, requiring interpretation, and a contract that, because it cannot be said to represent the agreement of the parties at all, cannot be interpreted, can only be rescinded and the parties left to go their own ways. Colfax, the plaintiff, is a manufacturer of envelopes. It does some printing of its envelopes, and the seventeen employees who do the printing are represented by the defendant union. Colfax has two printing presses. One prints 78-inch-wide sheets in four colors. The other prints 78-inch-wide sheets in five colors, but most of the time Colfax prints only four-color sheets on it.

Colfax has so few printing employees that it does not bother to participate in the collective bargaining negotiations between the union and the Chicago Lithographers Association, an association for collective bargaining of the other Chicago printing companies whose employees are represented by this union. Instead, whenever the union and the CLA sign a new collective bargaining agreement, the union sends Colfax a summary of the changes that the new agreement has made in the old one. If Colfax is content with the changes, the union sends it a copy of the complete new agreement, which Colfax signs and returns. If Colfax doesn't like the terms negotiated by the CLA, it is free to do its own bargaining with the union.

The collective bargaining agreements specify minimum manning requirements for each type of press used by the printers. The agreement in force between 1987 and 1991 fixed those minima as three men for four-color presses printing sheets 45 to 50 inches wide and four men for four-color presses printing sheets wider than 50 inches. Five-color presses printing sheets more than 55 inches wide required five men unless only four colors were printed, in which event only four men were required. The upshot was that under these agreements, all of which Colfax had signed, Colfax had to man each of its presses (which were 78-inch presses) with four men except on the rare occasions when it printed five-color sheets on its second press, and then it had to add a man.

In 1991 the union negotiated a new agreement with the CLA and sent a summary of the changes to Colfax. The letter enclosing the summary asked Colfax to indicate whether it agreed to the terms in the summary. (This may have been a departure from past practice, in which Colfax signed the complete agreement rather than the summary, but if so neither party makes anything of it.) In a section on manning requirements, the summary lists "4C 60" Press--3 Men" and "5C 78" Press--4 Men." Believing (in part because union members who claimed to be familiar with the new agreement had told Colfax that Colfax would really like the changes in it) that this meant that all presses operated as four-color presses would now require only three men to man them, Colfax's president and majority shareholder, Charles Patten, signed the union's letter, indicating acceptance of the terms in the summary. Later a copy of the actual agreement arrived, but it contained a crucial typo, which supported Patten's understanding of the summary. When a corrected copy of the agreement finally arrived, the manning requirements stated in it were different from what Patten had understood from the summary. Four-color presses between 45 and 60 inches required three men, but all four-color presses over 60 inches required four men. The changes had not benefited Colfax at all, and because it was under competitive pressure, it would have liked to negotiate better terms. Patten refused to sign the agreement but the union took the position that Colfax was bound to it by its acceptance of the summary.

Colfax brought this suit under section 301 of the Taft-Hartley Act, 29 U.S.C. Sec. 185, for a declaration that it has no collective bargaining contract with the union because the parties never agreed on an essential term--the manning requirements for Colfax's printing presses. The union counterclaimed for an order to arbitrate. The union's position was that Colfax had accepted the new agreement, which requires arbitration of all disputes "arising out of the application or interpretation of this contract." The district judge granted summary judgment for the union, concluding that the reference to the new manning requirement for a four-color 60-inch press in the summary of changes that Colfax had accepted referred unambiguously to 60-inch presses and had no application to any other presses, such as Colfax's 78-inch presses. Colfax has appealed.

One way to describe the issue that divides the parties is that they disagree about the meaning of the term "4C 60" Press--3 Men." Colfax believes that it means four-color presses printing sheets 60 inches and over, while the union believes that it means four-color presses 60 inches and under (down to 45 inches). Remember that the previous agreement had allowed the use of three-man crews on four-color presses between 45 and 50 inches. The union interprets the change as extending the upper bound of the three-man range to 60 inches. Ordinarily a dispute over the meaning of a contractual term is, if the contract contains an arbitration clause, for the arbitrator to decide. But sometimes the difference between the parties goes so deep that it is impossible to say that they ever agreed--that they even have a contract that a court or arbitrator might interpret. In the famous though enigmatic and possibly misunderstood case of Raffles v. Wichelhaus, 2 H. & C. 906, 159 Eng.Rep. 375 (Ex. 1864), the parties made a contract for the delivery of a shipment of cotton from Bombay to England on the ship Peerless. Unbeknownst to either party, there were two ships of that name sailing from Bombay on different dates. One party thought the contract referred to one of the ships, and the other to the other. The court held that there was no contract; there had been no "meeting of the minds." See generally A.W. Brian Simpson, "Contracts for Cotton to Arrive: The Case of the Two Ships Peerless," 11 Cardozo L.Rev. 287 (1989).

The premise--that a "meeting of the minds" is required for a binding contract--obviously is strained. 2 E. Allan Farnsworth, Contracts Sec. 7.9, at p. 251 (1990). Most contract disputes arise because the parties did not foresee and provide for some contingency that has now materialized--so there was no meeting of minds on the matter at issue--yet such disputes are treated as disputes over contractual meaning, not as grounds for rescinding the contract and thus putting the parties back where they were before they signed it. So a literal meeting of the minds is not required for an enforceable contract, which is fortunate, since courts are not renowned as mind readers. Let us set the concept to one side, therefore, and ask how (else) to explain Raffles v. Wichelhaus and cases like it. It seems to us as it has to other courts that a contract ought to be terminable without liability and the parties thus allowed to go their own ways when there is "no sensible basis for choosing between conflicting understandings" of the contractual language, as the court said in an American Raffles -like case, Oswald v. Allen, 417 F.2d 43, 45 (2d Cir.1969), quoting William F. Young, Jr., "Equivocation in the Making of Agreements," 64 Colum.L.Rev. 619, 647 (1964). In Oswald the misunderstanding arose because the parties did not speak the same language (literally). In Balistreri v. Nevada Livestock Production Credit Association, 214 Cal. App. 3d 635, 262 Cal. Rptr. 862 (1989), the parents of an aspiring farmer thought they had pledged property they owned in Sebastopol to secure a loan to their son, and indeed the lender's cover letter described the property as "your Sebastopol residence." But the actual deed of trust listed the parents' home in Petaluma as the collateral. The court held that there had been no meeting of the minds.

Raffles and Oswald were cases in which neither party was blameable for the mistake; Balistreri a case in which both were equally blameable, the parents for having failed to read the deed of trust, the lender for having drafted a misleading cover letter. It is all the same. Restatement (Second) of Contracts Secs. 20(1)(a), (b) (1981). If neither party can be assigned the greater blame for the misunderstanding, there is no nonarbitrary basis for deciding which party's understanding to enforce, so the parties are allowed to abandon the contract without liability. Neel v. Lang, 236 Mass. 61, 127 N.E. 512 (1920); Konic International Corp. v. Spokane Computer Services, Inc., 109 Idaho 527, 529, 708 P.2d 932, 934 (App.1985). These are not cases in which one party's understanding is more reasonable than the other's. Compare Restatement, supra, Sec. 20(2)(b). If rescission were permitted in that kind of case, the enforcement of every contract would be at the mercy of a jury, which might be persuaded that one of the parties had genuinely held an idiosyncratic idea of its meaning, so that there had been, in fact, no meeting of the minds. Cf. Young, supra, at 646. Intersubjectivity is not the test of an enforceable contract.

The clearest cases for rescission on the ground that there was "no meeting of the minds" (or, better, that there was a "latent ambiguity" in the sense that neither party knew that the contract was ambiguous) are ones in which an offer is garbled in transmission. The cases we have cited are all of that character, if "transmission" is broadly construed. Vickery v. Ritchie, 202 Mass. 247, 88 N.E. 835 (1909), provides a further illustration. A landowner and a contractor signed what they believed to be duplicate copies of a contract for the construction of a Turkish bath house. Because of a fraud by the architect for which neither the contractor nor the landowner could be blamed, the copy signed by the landowner stated the price as $23,000 and the copy signed by the contractor stated it as $34,000. Through no fault of their own, the parties had signed different contracts. Or consider Konic International Corp. v. Spokane Computer Services, Inc., supra. The seller quoted a price of "fifty-six twenty," which the buyer thought meant $56.20. In fact the seller had meant $5,620. In both cases rescission was permitted, the first being a case in which neither party was at fault, the second one in which both were equally at fault, being careless in their utterance and interpretation, respectively, of an ambiguous oral formula.

Our case is superficially similar. The actual terms of the 1991 agreement were muddied in the summary that the union gave Colfax and that Colfax signed, making it possible that the parties had different understandings. The difference between this case and the others is that Colfax, unlike the hapless promisors in the cases we have cited, should have realized that the contract was unclear. The buyer in Konic thought--really thought--that he was being quoted a price of $56.20, and no doubt fell off his stool when he discovered that the price was a hundred times greater than he thought. But the expression "4C 60" Press" does not on its face speak to the minimum manning requirement for a 4C 78" Press. The union's interpretation, that the phrase merely extended the upper bound of the old range for three-man four-color presses from 50 to 60 inches, may or may not be correct. The fact that the union restated and clarified the interpretation in the corrected agreement that it sent Colfax is not decisive on the question, because it is the summary rather than the corrected full agreement that is the contract between these parties. But Colfax, if reasonable, could not have doubted from reading the summary that interpretations of the kind that the union and the district judge later placed upon it would be entirely plausible. Colfax had a right to hope that its interpretation would prevail but it had no right to accept the offer constituted by the summary on the premise that either its interpretation was correct or it could walk away from the contract. "Heads I win, tails you lose," is not the spirit that animates the principle that latent ambiguity is a ground for rescission of a contract.

It is common for contracting parties to agree--that is, to signify agreement--to a term to which each party attaches a different meaning. It is just a gamble on a favorable interpretation by the authorized tribunal should a dispute arise. Parties often prefer to gamble in this way rather than to take the time to try to iron out all their possible disagreements, most of which may never have any consequence. Colfax gambled on persuading an arbitrator that the reference in the summary to the four-color 60-inch press meant what Colfax believes it means. The union gambled on the arbitrator's adopting the meaning that the union later made clear in the full agreement--but, to repeat, if there is a contract it is (the parties agree) the summary, read in light of the collective bargaining agreement that was being modified, that is the contract between these parties.

When parties agree to a patently ambiguous term, they submit to have any dispute over it resolved by interpretation. That is what courts and arbitrators are for in contract cases--to resolve interpretive questions founded on ambiguity. It is when parties agree to terms that reasonably appear to each of them to be unequivocal but are not, cases like that of the ship Peerless where the ambiguity is buried, that the possibility of rescission on grounds of mutual misunderstanding, or, the term we prefer, latent ambiguity, arises. A reasonable person in Colfax's position would have realized that its interpretation of the term "4C 60" Press--3 Men" might not coincide with that of the other party or of the tribunal to which a dispute over the meaning of the term would be submitted. It threw the dice, and lost, and that is the end of the case. It cannot gamble on a favorable interpretation and, if that fails, repudiate the contract with no liability. Cf. Prudential Ins. Co. v. Miller Brewing Co., 789 F.2d 1269, 1278 (7th Cir.1986).

We would have a different case if the ambiguity were over whether the parties had agreed to arbitrate their disputes. The duty to arbitrate is contractual, and the interpretation of the contract that creates the duty is for the court. AT & T Technologies, Inc. v. Communications Workers of America, 475 U.S. 643, 106 S. Ct. 1415, 89 L. Ed. 2d 648 (1986). But courts will not allow a party to unravel a contractual arbitration clause by arguing that the clause was part of a contract that is voidable, perhaps because fraudulently induced. Prima Paint Corp. v. Flood & Conklin Mfg. Co., 388 U.S. 395, 406, 87 S. Ct. 1801, 1807, 18 L. Ed. 2d 1270 (1967); Matterhorn, Inc. v. NCR Corp., 763 F.2d 866 (7th Cir.1985). The party must show that the arbitration clause itself, which is to say the parties' agreement to arbitrate any disputes over the contract that might arise, is vitiated by fraud, or lack of consideration or assent, as in Three Valleys Municipal Water District v. E.F. Hutton & Co., 925 F.2d 1136, 1140 (9th Cir.1991); that in short the parties never agreed to arbitrate their disputes. Wheat, First Securities, Inc. v. Green, 993 F.2d 814, 818-19 (11th Cir.1993). Colfax and the union had in a long course of dealing always agreed to submit their contractual disputes to arbitration, so the only question is whether their dispute in this case was a dispute over the meaning of their contract. It was, so it had to be arbitrated, for that was the parties' chosen method of resolving disagreements, whether or not Colfax would have signed on to the 1991 agreement had it realized what the agreement actually meant or could be interpreted to mean. Matterhorn, Inc. v. NCR Corp., supra, 763 F.2d at 873.

We go further: Even if, contrary to our earlier analysis, there was no "meeting of the minds" (in the artificial sense in which the law of contracts uses the term) on the manning requirements in the 1991 agreement, there was a meeting of the minds on the mode of arbitrating disputes between the parties arising from any collective bargaining contract (including a summary of changes in a previous contract) that Colfax signed. Under the Supreme Court's decision in Prima Paint, a contract dispute is arbitrable even if one party argues that the contract should be rescinded because it does not express an actual agreement of the parties, for example because it was induced by fraud. All that is important is that the parties have agreed that arbitration rather than adjudication would be the mode of resolving their disputes. A different view would in many cases deprive the arbitrator of an important contract remedy--rescission. This point has implications for the scope of the arbitrator's responsibilities, of which more presently.

We thus affirm the district judge's decision, but point out that her conclusion that the disputed term unequivocally bears the meaning that she assigned to it (which incidentally is not identical to the union's interpretation, for she thought it a point--60-inch presses, period--while the union thought it a range--60-inch presses and down) does not bind the arbitrator. His is the responsibility, subject to the excruciatingly limited right of judicial review of arbitral decisions, to interpret the agreement. It will therefore be open to Colfax to argue to the arbitrator that, under a proper interpretation of the contract, there really was no meeting of the minds over the manning requirements and therefore that the contract should be rescinded after all. The only essential point at this stage of the litigation is that whether or not there was (as we believe, without meaning to bind the arbitrator) such a meeting of minds, there was sufficient mutual understanding to create an enforceable contract to submit the issue to arbitration.

AFFIRMED.

I accept without reservation Judge Posner's lucid and admirable explanation of the various gradations and nuances of the term "meeting of the minds" and the widely misunderstood case of Raffles v. Wichelhaus, 2 H. & C. 906, 159 Eng.Rep. 376 (Ex. 1864). As Professor Farnsworth notes, our understanding of these issues "would be improved if [the] much-abused metaphor [of 'meeting of the minds'] were abandoned." 1 E. Allan Farnsworth, Contracts Sec. 3.5, at 168 n. 2 (1990).

Nonetheless, this seems to me in some ways a much easier case than either of my colleagues' opinions would let on. Interpreting the contract is the arbitrator's responsibility. Only questions of contract formation are for the court. By framing this question as one of a "meeting of the minds," Colfax tries to turn an ordinary question of interpretation into one of formation, and thereby get out from under the arbitration clause. Judge Posner's opinion correctly dismisses this attempt.

I am inclined here, however, to affirm the judgment of the district court on its own terms and on the rationale supplied by that court. The summary settlement document to which Patten assented purported to be a summary of the manning changes in the prior (1987-1991) agreement made by the parties in the course of the current negotiations. These are all changes on their face addressed to the requirements of specific presses of specific sizes and having specific color capabilities. Thus the following relevant entries appear:

1st Pressman, 2nd Pressman, 1st Feeder, 2nd Feeder or through attrition 65% Helper

There was no entry for the 4-color 78" press in which Colfax was interested. The inescapable inference is that the manning for that press had not been changed.

The parties in arguing the issue have made various points about "ranges", e.g., presses of 60" or less or presses of 60" and more. "Ranges" have apparently figured in various agreements of the parties in the past and present, but there is nothing in the summary settlement document that suggests "ranges." This document refers, as I have indicated, to the manning requirements of specific presses having specific characteristics. These were the only presses with respect to which the collective bargaining agreement changed the manning requirements.

If deductions about "ranges" were to be drawn from these specific changes (and I see no reason to do this), they might be valid if drawn in a size-downward direction. Thus, if a 60" press required a 3-man crew, it would seem reasonable to infer that a smaller 40" press would not require more than a 3-man crew. The Union has suggested this as a valid inference. On the other hand Colfax suggests that the change in the 60" press manning requirement should extend upward to include a larger 78" press. This inference could not gain support from the common sense of larger presses requiring larger crews.

In any event, the district court took a clear position in this matter:

The settlement agreement's designation of a specific press size stands in marked contrast to the prior agreements that described press sizes in terms of ranges; it does not support the inference that a range of press sizes was referenced by the 60" designation. Because there is no ambiguity in the terms of the settlement agreement, the court is precluded from considering extrinsic evidence to interpret the meaning of the terms used.

Op. at 12, 1993 WL 291894.

The district court thereupon granted summary judgment on the Union's counterclaim to compel arbitration. It is clear enough that Colfax's acceptance of the summary settlement agreement unambiguously supports the Union's position that there is a collective bargaining agreement containing an arbitration clause and that the meaning of the contract terms should therefore be submitted to arbitration.

Whatever difficulty this case has caused comes from the fact that the "doctrine" of "meeting of the minds" can be seen in two different ways. The first way is to see it as an element of contract formation. We typically say that a contract is formed where there is an offer and acceptance. But where the offeror offers one thing, and the offeree accepts another, we sometimes say that no contract has been formed because there is no consensus ad idem--no agreement on the same thing. On this view, without a "meeting of the minds," there is "simply no agreement to which the parties could be bound." Marvin A. Chirelstein, Concepts and Case Analysis in the Law of Contracts 35 (1990). Judge McDade's concurrence seems to me to embrace this view ("There either was a 'meeting of the minds' and a contract formed, or there was not a 'meeting of the minds' and therefore no contract between the parties.").

The doctrine could alternatively be understood simply as a metaphor. This is the position I understand Judge Posner to take. On this understanding, where a court declares that there is no meeting of the minds, it is not really saying that there is no contractual relation between the parties, but simply declaring that it cannot discern a non-arbitrary way to decide whose interpretation is best. It therefore will not enforce either version, but will instead allow remedies of rescission and restitution, and send the parties their separate ways. This is exactly what the court does where it allows a party out of a contractual obligation on the grounds of impossibility or of mutual mistake. As in those situations, it is as if there were no contract at all, though the suggestion that "no contract was formed" is just a metaphor. We should not literally believe, as Professor Chirelstein would have us, that a contract does not and never did exist.

Whether the requirement that you need a meeting of the minds to form a contract is literal or metaphoric usually doesn't matter. The result is typically the same either way. The parties go off as if no contract were ever formed. But in cases like this one it is dispositive. As noted, the court decides only whether a contract exists. Interpreting the contract is the arbitrator's task. If no meeting of the minds literally means no contract, the arbitrator would not be at liberty--under AT & T Technologies, Inc. v. Communications Workers of America, 475 U.S. 643, 106 S. Ct. 1415, 89 L. Ed. 2d 648 (1986)--to find there has been no meeting of the minds. If there were no meeting of the minds, Judge McDade argues, there would be no contract for it to arbitrate. The arbitrator--whose job is only to interpret--would have interpreted himself out of a job. Judge Posner avoids that catch-22 by seeing the doctrine as metaphor. I am ultimately persuaded that this view is more suitable, and therefore agree with Judge Posner that the arbitrator could still find there to be no meeting of the minds, and therefore allow rescission and restitution.1

I also note, though it is surely dictum at this point, that the Union has clearly conceded in its briefs that the court has only determined that there is a collective bargaining agreement binding Colfax to arbitration and that arbitration should proceed. I should think the Union would now have difficulty asserting that an arbitral award was precluded by a prior judicial determination.

McDADE, District Judge, concurring.

I concur in the rather novel, but realistic, approach of the majority in resolving this issue of contract formation by finding that the parties agreed or had a "meeting of the minds" to execute a contract which contained an ambiguous term relating to a change in manning requirements for the four-color presses. A more traditional analysis would have required remandment to the district court for the admission of extrinsic evidence on the issue of contract formation once it was determined that the language of the summary setting forth the change in manning requirements was ambiguous.

The whole point of the lawsuit is the issue of whether a contract existed between the parties vel non, or as argued by Colfax, whether there was a "meeting of the minds" on an essential term, a requisite of contract formation. The majority finds that the requisite "meeting of the minds" exists because the parties "agree[d] to an obviously ambiguous term," acknowledging that "the actual terms of the 1991 agreement were muddied in the summary that the Union gave Colfax and that Colfax signed, making it possible that the parties had different understandings." Rather than finding that these "different understandings" resulted in an ambiguity necessitating the admission of extrinsic evidence on the issue of contract formation, the majority observes that "it is common for contracting parties to agree--that is, to signify agreement--to a term to which each party attaches a different meaning." Since the language of the manning change is obviously ambiguous (or more preferably, "patently" ambiguous, in that the ambiguity arises from the language itself) and Colfax did nothing to clarify it before signifying his assent, a contract was formed despite the ambiguous term. To hold otherwise is found by the majority to be inconsistent with contract law principles awarding recision for "latent ambiguity" as distinct from the patent ambiguity involved in the case before us. I share the majority's view that Colfax had to know its interpretation was not the only plausible interpretation of the language of the manning change and that Colfax accepted the risk of being wrong in the meaning it assigned to the term. Having taken the risk, Colfax must now face the consequences of a dispute with the Union on this issue. I also share the majority's view that it was the explicit purpose of the arbitration clause to require arbitration of all such disputes "arising out of the application or interpretation" of the contract.

Having concluded that there was a contract which should be submitted to arbitration pursuant to its terms, I cannot agree with the majority that the arbitrator has the right or authority to consider anew the decision which this court has already made. The majority cannot have it both ways. There either was a "meeting of the minds" and a contract formed, or there was not a "meeting of the minds" and therefore no contract between the parties. This court having decided the former, it is beyond the purview of the arbitrator's function to decide the existence of a contract--her function is to apply and interpret the contract. To hold otherwise would be to license the arbitrator to reverse the finding of this court as to contract formation, a duty belonging to the court and not the arbitrator.