Summary
discussing the fuzzy line between legal and factual determinations in contract interpretation
Summary of this case from La. Generating, L.L.C. v. Ill. Union Ins. Co.Opinion
Nos. 266, 366, Dockets 83-7511, 83-7557.
Argued November 3, 1983.
Decided April 5, 1984.
Joseph J. Magrath, 3rd, New York City (Bigham, Englar, Jones Houston, New York City, of counsel), for defendants-appellants-cross-appellees.
Donald M. Waesche, Jr., New York City (Waesche, Sheinbaum O'Regan, P.C., John R. Foster, New York City, of counsel), for plaintiff-appellee-cross-appellant.
Appeal from the United States District Court for the Southern District of New York.
Before TIMBERS, NEWMAN and CARDAMONE, Circuit Judges.
An explosion aboard an ocean-going tanker sailing from the United States to Antwerp, Belgium caused extensive damage to the ship, rupturing several of its cargo tanks loaded with liquid chemicals. As a result some of the liquid cargo spilled out and later solidified. We are asked on this appeal to construe the scope and coverage of a marine hull and machinery insurance policy. The specific question is which of the costs incurred by the shipowner to remove chemical debris from the cargo spaces and to clean and repair the damaged tanks are chargeable to the hull insurer. There is no American judicial precedent concerning this kind of coverage.
FACTS
Antilles Steamship Company Ltd. (Antilles), plaintiff-appellee, engaged in the business of ocean transport of chemicals, is the owner of the M/V "Alchemist". The defendants-appellants, 62 foreign and domestic insurance companies of which 63 are members of the American Hull Insurance Syndicate (Syndicate), are the underwriters that insured the "Alchemist's" hull.
The "Alchemist" is a steel tanker fitted with 34 tanks (18 of which are stainless steel) designed to carry liquid chemicals. She loaded liquid chemicals in Taft, Louisiana, pursuant to a contract of affreightment between Antilles and Union Carbide Corporation for carriage to Antwerp, Belgium. Among the chemicals taken aboard were over 200 tons of glacial acrylic acid (GAA). Another chemical aboard, loaded at a prior port, was ethylene norbonene (ENB). The only characteristic of ENB relevant to this appeal is its unbelievably penetrating, noxious odor. The two chemicals were placed in adjacent stainless steel tanks sharing a common wall. No. 6 starboard wing tank aft (No. 6 aft) was loaded with GAA, and No. 6 starboard wing tank forward (No. 6 forward) with ENB. On September 14, 1975 two days after taking the chemicals aboard, the parties entered into a one year contract of insurance on the "Alchemist's" hull and machinery. Seventy percent of the risk was insured in the American market with defendants and the remaining 30 percent was underwritten in the London market. The identical British and American policies contained the standard American Hull Institute clauses, including the "Inchmaree" clause involved here.
About an hour before midnight, on September 16 while the "Alchemist" was sailing smoothly at sea, the GAA in No. 6 aft suddenly and explosively polymerized for reasons which still remain unknown. The resulting sudden high pressure caused that aft tank to bulge, the common wall between it and No. 6 tank forward to rupture, and seams to open in No. 6 aft, permitting material to escape into the cofferdam, i.e., the empty space between the wall of the tanks and the skin of the ship. The skin or shell plating of the ship in the vicinity of No. 6 tank aft also bulged. The thunderous explosion accompanying the polymerization of the GAA caused the ship to vibrate severely and produced intense heat, 260 degrees F. or higher (the readings taken at the time were at the top of the thermometer) near the two No. 6 tanks. The "Alchemist's" crew reacted quickly during the emergency and successfully reduced the danger of fire by hosing down the deck and pumping seawater into the cofferdam nearest the No. 6 tanks. As a result, a mixture of GAA, ENB and seawater filled these tanks and the cofferdam. The tanker was able to complete the voyage to Antwerp under its own power, and there discharged its sound cargo.
Port officials in Antwerp would not permit the seawater-chemical mixture to be unloaded because of fear that its extremely noxious odor might create an air pollution problem. Later, the "Alchemist" sailed to a tank cleaning facility located near Rotterdam. Upon her departure from Antwerp without even having been permitted to open her cargo tanks, Masefield's evocative words might have been on the crew's mind: "I must go down to the seas again, to the lonely sea and the sky." More likely though — since on the run to Rotterdam the "Alchemist" hoped to avoid the noxious odor by sailing against the wind — this later line from "Sea Fever" was offered as a fervent prayer: "And all I ask is a windy day with the white clouds flying."
J. Masefield, "Sea Fever," reprinted in Twelve Centuries of English Poetry and Prose 838 (rev. ed. 1928).
On October 21 the tank cleaning company commenced removal of the putrid cargo still aboard. This operation proved to be infinitely more arduous and costly than anticipated because GAA, a water-like liquid in its monomer state, hardens when polymerized. The mixture at the top of No. 6 aft and No. 6 forward tanks posed little problem and vacuum trucks removed it in short order. Most of the one and a half feet of semi-solid polymer at the bottom of No. 6 forward tank was removed in the same manner. What remained was shoveled out by hand. It only took a few days to clean the No. 6 forward tank. Similar methods were used to clean the cofferdam with high pressure washing equipment, vacuum hoses and, finally, shovels.
Meanwhile attention focused on the site of the bulk of the GAA in No. 6 aft tank. For the next 30 days this tank had to be cleaned by hand on an around-the-clock basis. The mixture was about 20 feet deep in the tank. At first near the top it was shovelable. As the work progressed deeper into the tank the polymer became less rubbery and much harder. Near the bottom it became necessary to use heavy-duty compressed air jackhammers to break up the granite-like mass which extended across the bottom of the tank from bulkhead to bulkhead, and adhered to them.
In the course of the work substantial damage was done to the stainless steel skin of the tank. After the "Alchemist" was repaired at a nearby ship yard, Antilles claimed the cost of removal of all of the GAA-ENB-seawater mixture and the cost of repairs from both the British insurers and from defendants by submitting a statement of adjustment in particular average for over a million dollars. The London market underwriters responsible for 30 percent of the risk paid their portion of the claim without protest. Defendants, the American market underwriters, at first paid only for the actual damage to the walls of both tanks and the skin or shell of the ship; later, defendants agreed to pay in addition those costs associated with removal of the mixture from the cofferdam and removal of the last residue of the mixture adhering to the bulkheads of the No. 6 aft and forward tanks, a sum stipulated to be $51,858.76. Defendants refused to pay for the cost of removing the bulk of the hardened mass from the cargo tanks which, it was stipulated, cost Antilles $365,339.72.
In this case, after the completion of all cleanup and repairs of the "Alchemist," plaintiff had a statement of particular average prepared by its insurance broker, Alexander Alexander, Inc., in order to make a claim under the "Inchmaree" clause of the policy. In English law under section 56(1) of the Marine Insurance Law "a loss may be either total or partial. Any loss other than a total loss . . . is a partial loss." Section 64(1) of the same law provides that "a particular average loss is a partial loss of the subject-matter insured, caused by a peril insured against, and which is not a general average [___] loss." The loss must be accidentally caused by a peril insured against and under section 69(1) the assured is entitled to "reasonable cost of repairs, less customary deductions," but not exceeding the sum insured. See R.J. Lambeth, Templeman on Marine Insurance, 256-57 (5th ed. 1981) (Templeman). Thus, the particular average in this case was a statement or claim for a partial loss. The statement was a detailed compendium of the casualty, various surveys, repairs at the shipyard, clean-up at Rotterdam and itemization of all expenses connected with these activities.
Antilles filed this action in the United States District Court for the Southern District of New York seeking recovery for the removal of the hardened polymer from the ruptured No. 6 aft. The parties tried the case before District Judge Conner on March 16, 1982. In the brief bench trial where most of the facts were stipulated, plaintiff called only two witnesses. The first, James W. Kingston, was Marine Superintendent for Steuber Company, the manager and operator of the "Alchemist" at the time when the explosion occurred. Kingston testified to the accident and its aftermath. Plaintiff's second witness was Raymond M. Hicks, Jr., claims manager for the defendant American Hull Insurance Syndicate. He testified as an expert regarding the custom and practice of average adjusters and marine insurers in cases of solidified, difficult to remove cargo. Defendants called no witnesses. As part of the case both parties submitted post-trial briefs.
The district court issued two opinions and orders. In its opinion dated May 24, 1982, 539 F.Supp. 572, the court found that the defendants were obligated to indemnify plaintiff for removal of all the hardened polymer. In a supplemental opinion and order dated August 5, 1982 the court clarified its earlier opinion. After analyzing and distinguishing the somewhat arcane authority in what he recognized as a case of first impression, Judge Conner fashioned a result that permitted plaintiff to recover for the cost of removing the rockhard polymer, but not for the costs of removing from the two No. 6 tanks the "shovelable" material, i.e., the liquid and rubbery substances. Defendants appeal the resulting judgment awarding plaintiff $271,179.01 plus costs and interest. Antilles cross-appeals claiming it was entitled to the entire cost of removal of all substances.
DISCUSSION I. Choice of Law
Before beginning an analysis of the law we explore briefly the choice of applicable law. After some dispute below, the parties have now agreed that New York law applies in the first instance. See also Wilburn Boat Co. v. Fireman's Fund Insurance Co., 348 U.S. 310, 321, 75 S.Ct. 368, 374, 99 L.Ed. 337 (1955) (state law controlling on marine insurance issues). Such issues as burden of proof and general rules of insurance policy interpretation are substantive rules governed by settled New York law.
To the extent that no New York precedent exists, the maritime nature of this insurance contract dictates that we anticipate that New York courts would look to English law in view of the "special reasons for keeping in harmony with the marine insurance laws of England, the great field of this business." Queen Insurance Co. v. Globe Rutgers Fire Insurance Co., 263 U.S. 487, 493, 44 S.Ct. 175, 177, 68 L.Ed. 402 (1924) (Holmes, J.); see Lenfest v. Coldwell, 525 F.2d 717, 724 n. 15 (2d Cir. 1975). These sophisticated contracting parties should be presumed familiar with basic English law principles that have shaped custom and practice in the hull insurance field.
The claim made was under the "Inchmaree" clause of the hull and machinery policy issued by defendants. The term "Inchmaree Clause" derives from an 1887 case involving the steamer "Inchmaree". It arose in the case of Thames and Mersey Marine Insurance Co. v. Hamilton, Fraser Co., 12 App.Cas. 484 (1887), when a pump valve closed accidentally and caused water, intended for a boiler, to go instead into an air chamber where it split the pump. The House of Lords held that this was not a loss caused by a "peril of the sea", nor a loss within the words "all other perils, losses" contained in a hull insurance policy. To counter this restrictive judicial view, a clause covering additional perils or negligence came into being and has evolved and broadened the scope of coverage afforded insureds. That clause in marine policies is now called the "Inchmaree" clause. See Templeman at 444-452. The relevant portion of the "Inchmaree" clause in the policy before us is: "[T]his insurance also covers loss of or damage to the vessel directly caused by the following: . . . Explosions on shipboard or elsewhere."
II. Intent of the Parties
The district court's task in this case was to ascertain whether the statement of particular average filed by Antilles came within the scope and coverage of the "Inchmaree" clause of the defendants' policy. This is a matter of intent. Blain Richards Co. v. Marine Indemnity Insurance Co. of America, 635 F.2d 1051, 1054 (2d Cir. 1980). The sophisticated and knowledgeable parties here engaged in a classic maritime transaction — insuring a vessel's hull and machinery. The issue was their "reasonable understanding . . . as to the meaning of their insurance agreements." Id. In marine insurance law there is interplay and tension between a hull underwriter's duty to pay the reasonable cost of repairing "damage to the hull" and a shipowner's duty to pay for discharge of his ship's cargo. In fixing these respective obligations, a line must be found where the shipowner's responsibility for discharge of its cargo ends and the hull underwriter's duty to pay for damage to the hull begins.
The district court apparently did not view the policy coverage issue as one of fact and did not explicitly make a finding as to the parties' intent. We believe the central question here is at least a mixed question of law and fact. See 3 Corbin on Contracts § 554 (1951 and 1980 Supp.). Because the contracting parties here were free to agree on coverage for this type of loss, the emphasis of plaintiff's proof, as well as the district court's analysis, should have been on the parties' mutual understanding. In addition to the language of the policy and the circumstances surrounding its execution, custom and usage, as well as the established law, all bear on the ultimate "fact" of the parties' "reasonable understanding." While the basis for the trial court's holding is not free from doubt, we note that his opinion refers to certain facts which influenced his interpretation of the clause ( e.g., the English underwriters' settlement of the claims). Such suggests that Judge Conner implicitly found, as a matter of fact, that plaintiff had proved the parties contemplated this sort of coverage.
III. Antilles' Proof at Trial
Antilles concedes that it was required to prove by a preponderance of the evidence three elements: (1) it was insured on a marine insurance policy underwritten by defendants; (2) a loss occurred; and (3) the loss was an event within the terms of the policy. The first two presented Antilles with no difficulty. Defendants conceded the first element in their answer and stipulated the second by agreeing that an explosion had occurred. Antilles' final hurdle was to prove that there was a loss within the terms of the policy, i.e., that the shipboard explosion, which concededly damaged the vessel, required the underwriters to pay for the cost of removing cargo or cargo debris from the ship's cargo spaces after the "Alchemist" had reached harbor safely in Antwerp. It was this last hurdle that plaintiff failed to clear.
Plaintiff has the burden of showing that both the loss and the peril which caused it fall within the meaning of the policy terms. See Northwestern Mutual Life Insurance Co. v. Linard, 498 F.2d 556, 561 (2d Cir. 1974); S. Felicione Sons Fish Co. v. Citizens Casualty Co., 430 F.2d 136, 138 (5th Cir. 1970), cert. denied, 401 U.S. 939, 91 S.Ct. 936, 28 L.Ed.2d 219 (1971).
Antilles points to the testimony at trial of the expert witness Hicks who testified concerning customs and practices of the marine insurance industry. We examine this proof as it bears on the intent of the parties, construing it in a light most favorable to plaintiff. The testimony at trial demonstrated: the English Marine Insurance Act of 1906 provides, and all parties to hull policies apparently contemplate, that the underwriter is to bear the "reasonable cost of repair" of a vessel and machinery damaged by a covered peril, and underwriters generally assume responsibility under the "Inchmaree" clause for removal of cargo in two cases, i.e., removal of cargo from the cofferdam area and gasfreeing the vessel's cargo tanks to permit repairs. Hicks noted his familiarity with these principles of law and custom and also recognized the competing legal principle that a shipowner bears the cost of removing cargo at the port of destination. On cross-examination, when asked whether in his opinion there was coverage in this case, he answered "no." No other evidence is found in the record to refute that opinion.
In essence, Hicks was of the view that the general practice of hull underwriters is to pay for the cost of removing only that portion of solidified or adhering cargo which remains in the cargo areas beginning with the layer where the process of removal begins to cause damage to the bulkhead. The witness acknowledged that he had stated publicly that the "`reasonable cost of repair' refers solely to the costs necessary to restore the vessel's `physical being.'" And, in response to a leading question, he affirmed that removal of the hardened polymer was necessary to restore the vessel's "physical being." Antilles repeatedly cites this testimony as proof supporting its argument that the loss falls within the "Inchmaree" clause. Quite the contrary, even when viewed most favorably to plaintiff, Hicks' testimony related solely to the limits of the hull policy's coverage, not to whether the cost of cargo removal, the loss here claimed, is covered. Moreover, Hicks elsewhere noted that even sound cargo would affect the "physical being" of the vessel for purposes of effecting repairs. Thus, we conclude that Antilles failed in its burden of producing sufficient evidence to prove a prima facie case that removal of the bulk of the hardened cargo fell within the intended scope of the "Inchmaree" clause.
Antilles presented various letters and memoranda prepared over the years by experts in the marine insurance area. None of these documents supports plaintiff's argument that coverage here was contemplated. At most, these exhibits proved that hull underwriters customarily pay for removal of cargo or its derivatives from void spaces, e.g., a cofferdam, and for removing the last layer of residue from cargo spaces. Because these documents uniformly recognize the general rules established by Field v. Burr, infra, and the other authorities, Antilles' documents were at best equivocal and at worst detrimental to its claim of coverage.
The district court also noted that "the fact that the London market paid the claim undercuts defendants' argument that the law, especially English law, requires rejection of Antilles' claim." The settlement achieved between Antilles and the London market has little or no probative value regarding the American underwriters' liability, and arguably contravenes the spirit, if not the literal terms, of Fed.R.Evid. 408. Cf. Reichenbach v. Smith, 528 F.2d 1072, 1074 (5th Cir. 1976); Sun Oil Co. v. Govostes, 474 F.2d 1048, 1049 (2d Cir. 1973) ( per curiam); Hawthorne v. Eckerson Co., 77 F.2d 844, 847 (2d Cir. 1935). Although the English underwriters' decision to pay cannot be viewed as a "settlement" within Rule 408 because there was apparently no real dispute, the probative weight of its payment on the question of the American insurers' liability is conjectural. A desire to avoid substantial litigation expenses, the need to maintain good client relations or even a mistaken view of the facts (underwriters suggest that plaintiff's brokers successfully passed off the disputed expense as "cleaning required solely in connection with damage repairs") or law might have motivated the London market to pay. It sheds little light on the present parties' intentions in entering into the hull policy.
We entertain some doubt as to whether the evidence taken as a whole was sufficient to make out plaintiff's prima facie case. Even if plaintiff produced enough proof to survive a motion for a directed verdict, a review of all the evidence in the case, including particularly the customs and practice in the field of marine insurance, convinces us that Antilles failed to sustain its burden of proof by a preponderance of the evidence. The trial court's findings to the contrary were therefore clearly erroneous.
IV. Parties' Knowledge of Marine Custom and Practice
In eliciting proof of custom and usage from its expert witness, plaintiff specifically developed the general practice of particular average adjusters regarding solidified cargo. In the course of testifying, Hicks clearly evidenced his familiarity with the authorities in this field and described a practice, the existence of which those authorities confirm. The language used in the subject marine policy consists of certain words of art and phrases that over generations have acquired content and meaning, revealed through an examination of customs and practices in the field of marine insurance. When knowledgeable parties enter into a contract of marine insurance containing such specific terminology, they should therefore be presumed to have intended that its terms be applied in accordance with such general understanding, absent an indication of contrary intent. In that connection underwriters introduced into evidence the deposition of an insurance adjustor representing Antilles that indicated complete familiarity with the history and customs of marine insurance law. Thus, charging both of these parties with knowledge of the English custom and practice of marine hull insurance, we turn to examine it.
V. Analysis of Law
The seminal — and perhaps only — judicial decision on the scope of a hull policy in a damaged cargo situation is Field v. Burr, 1 Q.B. 679. In that case an appeal panel of the Queen's Bench affirmed a trial court's conclusion that a hull underwriter was not liable for expenses incurred in removing cottonseed cargo which, having mixed with seawater after an accident where the vessel was holed, had become a "putrid mass." The appellate court held for the underwriters, finding only "damage to the shipowner in his business as carrier," not damage to the hull. Id. at 679. The Field court also discussed the hypothetical case of solidified cargo such as cement. Even as to this, the appellate court stated that only the last layer adhering to the fabric of the ship properly would be recoverable in particular average. Removal of the bulk of the cement would remain the shipowner's responsibility. Shipowners, underwriters and adjusters are all familiar with the Field case. Marine insurance commentators discuss it at length and view its holding as stating the basic contemporary practice in the field. See 2 Arnould on Marine Insurance § 1126 (16th ed. 1981); L. Buglass, Marine Insurance General Average in the United States, 166 (2d ed. 1981).
Leslie Buglass, a leading American commentator, describes the American adjusters' practice as follows:
In practice if, for example, the damaged cargo has solidified and has adhered to the structure of the vessel so that in effect the vessel has sustained damage, the cost of removing that part which has adhered to the vessel is recoverable from hull underwriters as part of the reasonable cost of repairs. This is necessarily an arbitrary division, but usually a percentage is allowed, depending on the width of the solidified mass. Similarly, if the cargo debris has entered places where cargo would not normally be (such as the engine room, pump room or double-bottom tanks) the cost of cleaning and removing such cargo is considered "damage" and therefore part of the reasonable cost of repairs. In all other circumstances, the cost of removing cargo debris from a vessel (whether or not such removal is necessary to effect repairs) is not considered to be part of the cost of repairs but rather a consequential loss or enhanced voyage expense. Such extra expenses may be recoverable from freight underwriters or Protection and Indemnity underwriters, depending on the circumstances in each individual case.
Apparently no court has confronted the problem of removal of solidified or adhering cargo since Field v. Burr. In 1968 the question whether solidified cargo can be "damage to the vessel" arose in an arbitration in San Francisco before Martin P. Detels, former chairman of the Association of Average Adjusters, as arbitrator. Barge J. Whitney, A.M.C. 996. In that case the owners of a barge contracted to carry asphalt from California to Alaska. The barge was equipped with special heating coils through which hot oil or steam could be forced to liquify the cargo at the port of discharge. During the process of loading, the heating coils fractured because of condensate which had remained in them; as the hot asphalt flowed into the coils it solidified and rendered the coil system inoperative. Not surprisingly, removal of the solidified asphalt was difficult and expensive, and the shipowner claimed that expense from the hull underwriter in particular average. The arbitrator resolved the "difficult question" ( id. at 1003) in favor of the underwriters, and found Field v. Burr "nearly indistinguishable." He ruled therefore that the underwriters were responsible only for damage done to the vessel during loading, for repairing the deck which had to be penetrated for the installation of new heating coils during discharge and for the last layer of asphalt adhering to the tanks after removal of the bulk of the cargo.
We conclude that cargo debris in a cargo tank at the port of destination does not constitute "damage to the vessel" within the parties' intended meaning of the "Inchmaree" Clause, and that the lower court's finding on this issue was clearly erroneous. Every merchant vessel requires repair work at some time or other and cargo often must be discharged to do that work. Once a voyage is completed, hull underwriters are not concerned with the cost of cargo discharge, no matter what its form. This view on the coverage of blue water hull policies has been prevalent in marine insurance law for 85 years. We think it must be well understood by now. The hardship to a shipowner arising from the extra expense for cargo removal may be protected against by insurance with Protection and Indemnity underwriters, but the parties to this hull and machinery policy obviously did not contemplate such added expense.
CONCLUSION
Accordingly we reverse the $271,179.01 judgment in favor of Antilles. The insurers were required under the "Inchmaree" clause to pay only for those expenses incurred in removal of the mixture from the cofferdam and in removing the last residue of material clinging to the bulkhead of No. 6 tanks aft and forward. Hull insurers customarily pay the cost of cleaning cargo bulkheads of cargo residue where it is a prerequisite to repair work, e.g., an oil tanker must be "gas freed" to permit welding and the like. Of course, what is a "residue" will vary and should be a matter for the trier of fact to determine in each case recognizing that the word means remnant, vestige or trace. This definition accords with the view that only that portion of the cargo which adheres to a bulkhead is chargeable to the hull insurer. Fortunately, here, that amount has been stipulated by the parties to be $51,868.76, and that amount has apparently been paid since the trial. The bulk of the hardened polymer in the tanks was cargo, and the expense of its removal must be borne by the shipowner, not the hull insurers.
Since the defendants have paid the sums for which they are liable, and plaintiff is entitled to neither the additional sum awarded by the district court nor the further sum it seeks on its cross-appeal, the judgment is reversed and the complaint is dismissed.
I agree with the majority that the hull insurance policy covering "damage to the vessel" renders the insurers liable only for the costs of removing debris from non-cargo areas and removing the residue of adhering cargo from the bulkheads of the cargo areas. I write separately because of my differing views as to the analytical framework in which we reach an interpretation of the contract contrary to that reached by the able District Judge. These differences concern the venerable dispute as to whether an issue is one of law or of fact.
The majority proceeds from the premise that the question of interpreting the contract is "at least a mixed question of law and fact." P. 199. The majority then observes that "the emphasis of plaintiff's proof, as well as the district court's analysis, should have been on the parties' mutual understanding." Id. The majority maintains that Judge Conner "implicitly found, as a matter of fact, that plaintiff had proved the parties contemplated [coverage for the cost of removing all of the rock-hard debris from the cargo areas]." Id. Finally, the majority concludes that Judge Conner's findings are clearly erroneous. I believe that Judge Conner made no finding of fact in the traditional sense, that he interpreted the insurance contract as a matter of law, that the question of interpretation is a close question on which reasonable minds may and obviously do differ, and that, as appellate judges, we may reverse the District Judge's conclusion because we disagree with it, not because it is clearly erroneous.
Commentators and courts differ as to whether interpretation of a contract is an issue of law or of fact. The matter has relevance for determining whether the issue is to be decided by the trial judge or, when available, the jury, whether the determination, if made by the trial judge, is subject upon appellate review to the "clearly erroneous" standard of Rule 52(a) of the Federal Rules of Civil Procedure, and whether the construction is to have stare decisis effect. Professors Corbin and Williston express the view that interpretation of a contract is always a question of fact. 3 Corbin on Contracts § 554 (1960); 4 Williston on Contracts § 616 (W. Jaeger 3d ed. 1961). However, courts have traditionally stated that, where evidence extrinsic to the undisputed terms of a contract is not presented, the interpretation of the contract is "a pure question of law." Hamilton v. Liverpool, London and Globe Insurance Co., 136 U.S. 242, 255, 10 S.Ct. 945, 950, 34 L.Ed. 419 (1890); see Pennsylvania Avenue Development Corp. v. One Parcel of Land, 670 F.2d 289, 292 (D.C.Cir. 1981); Official Creditors Committee of Stratford of Texas, Inc. v. Stratford of Texas, Inc. (In re Stratford of Texas, Inc.), 635 F.2d 365, 368 (5th Cir. 1981); Hoffman v. American Mills Co., 288 F. 768, 772 (2d Cir. 1923); see also Parsons v. Bristol Development Co., 62 Cal.2d 861, 865, 44 Cal. Rptr. 767, 770, 402 P.2d 839, 842 (1965) (Traynor, C.J.) ("It is . . . solely a judicial function to interpret a written instrument unless the interpretation turns upon the credibility of extrinsic evidence."); but see Dobson v. Masonite Corp., 359 F.2d 921, 923 (5th Cir. 1966) ("Interpretation is always a question of fact"). It may be helpful to characterize such cases as involving only "textual interpretation" to distinguish them from the fact-specific cases involving extrinsic evidence of the parties' actual intent offered to aid in the interpretation of ambiguous contract terms. Professor Williston suggests that courts are mislabeling the textual interpretation question as one of law. He recalls Professor Thayer's observation that "judges have always answered a multitude of questions of ultimate fact, of fact which forms part of the issue. It is true that this is often disguised by calling them questions of law." J.B. Thayer, A Preliminary Treatise on Evidence at the Common Law 202 (1898), quoted in Williston, supra, § 616 at 649.
The dichotomy between issue of fact and issue of law has no bearing on the selection of the fact-finder in this case, since the court, sitting in admiralty, was the adjudicator of both the law and the facts. However, the choice is pertinent to the precedential force of our ruling and to our standard of review, since findings of fact, but not conclusions of law, may not be disturbed on appeal unless "clearly erroneous." Fed.R.Civ.P. 52(a). In this latter connection, the choice could also be implicated by the proposed amendment to Rule 52(a), which would make findings of fact subject to the "clearly erroneous" standard of review, "whether based on oral or documentary evidence." Preliminary Draft of Proposed Amendments to the Federal Rules of Civil Procedure, 98 F.R.D. 337, 359 (1983). I assume that the proposed change is intended to apply to instances where documentary evidence, whether in the form of affidavits, depositions, or otherwise, is offered to establish facts, and not to a case like this where a written contract, whose meaning is unaided by any evidence, oral or documentary, must be interpreted as a matter of law.
Professors Corbin and Williston both recognize, however, that where the terms of an agreement are definite and undisputed and there is no doubt concerning the surrounding circumstances, interpretation is thought to be a matter for the court. This is so, they note, not only when the terms are so clear that only one reasonable interpretation is available, but also when the terms are subject to either of two reasonable interpretations but no extrinsic evidence is presented that aids in ascertaining the parties' intent. Corbin, supra, § 554; Williston, supra, § 616. As would be expected, courts that describe the issue of textual interpretation as a question of law also hold that the issue is for the court, not the jury. E.g., Hamilton v. Liverpool, London and Globe Insurance Co., supra, 136 U.S. at 255, 10 S.Ct. at 949; Pennsylvania Avenue Development Corp. v. One Parcel of Land, supra, 670 F.2d at 292; Official Creditors Committee of Stratford of Texas, Inc. v. Stratford of Texas, Inc. (In re Stratford of Texas, Inc.), supra, 635 F.2d at 368; Hoffman v. American Mills Co., supra, 288 F.2d at 772.
One may wonder how it can be that courts, not juries, decide issues of textual interpretation in contract cases if they are really reasonably disputable issues of fact. I think the answer is, with all respect to the acknowledged giants of contract law, that issues of textual interpretation, in the absence of extrinsic evidence of the parties' intent, are really not issues of fact in any meaningful sense of that term. In asserting the contrary, Professor Corbin helpfully identifies the following contexts in which interpretive questions arise:
searching for the meaning of the two contracting parties, or for the meaning given to words by the one person who used them, or for the meaning that was given by words by another person who heard or read them and acted in reliance on them, or for the meaning that a reasonable man or an intelligent user of English or an average resident of the community would have given to them.
Corbin, supra, § 665 at 219. The first three contexts are specific to the case to be adjudicated. The fourth, involving interpretation by a reasonable person, is not case specific; it is the interpretation that will apply in all cases involving the meaning of the contract terms at issue, whenever extrinsic evidence of intent is lacking to aid in interpreting those terms. If such extrinsic evidence is available, the issue becomes one of fact because it is entirely realistic to examine as a question of fact the actual intent of the parties in the particular case. See, e.g., United Truck Bus Service Co. v. Piggott, 543 F.2d 949, 950 (1st Cir. 1976); Palmer v. Howard, 493 F.2d 830, 835 (10th Cir. 1974); Hoffman v. American Mills Co., supra, 288 F.2d at 772; see generally Judge Friendly's helpful opinion in Meyers v. Selznick Co., 373 F.2d 218, 221-23 (2d Cir. 1966). In the absence of such extrinsic evidence, however, it is a strained and unrealistic use of the term "question of fact" to apply it to the inquiry as to what meaning a reasonable person would give to the contract term at issue. Of course, if the law's hypothetical reasonable person were to appear in the courtroom as a deus ex machina and respond to an inquiry as to what he or she thought the terms meant, the answer would be a matter of fact in the traditional sense — an observable, recordable historical event. But the law's reasonable person has not yet appeared on the witness stand of any courtroom. That person remains a legal construct.
I do not mean to suggest that an inquiry concerning a hypothetical question cannot be an issue of fact. An estimation of future profits or earnings had a breach of contract not occurred or a tortious injury not been inflicted is a hypothetical inquiry, appropriate for the factfinder upon evidence of sufficient quality to make the estimate. That inquiry, however, is fact-specific: it asks what the profits or earnings would have been had specific facts transpired in the future. An inquiry as to what a reasonable person would think the contract terms mean is of a different nature. We have no expectation that but for the event giving rise to the lawsuit, the law's hypothetical "reasonable person would have ever materialized to express a view as to the meaning of the contract. We are not predicting what a future fact would likely be; we are ascertaining what the law says should be the consequences of the existing uncertainty as to the meaning of the terms.
When we say that a contract has a particular meaning only because a reasonable person would so interpret it, we are leaving the realm that Professor Corbin has so carefully defined as "interpretation" and entering the area he classifies as "construction." Corbin, supra, § 534 at 9:
By "interpretation of language" we determine what ideas that language induces in other persons. By "construction of the contract," as that term will be used here, we determine its legal operation — its effect upon the action of courts and administrative officials [footnote omitted].
I agree with Professor Corbin that we are interpreting language when we ask what it means to the parties who made a contract, or to the person who used the language, or to the person who heard or read the words and relied upon them. In those contexts, we want to know what ideas the words induced in the pertinent individuals, and if there is extrinsic evidence that will aid in knowing the workings of their minds, we receive it and use it to determine their meaning as a matter of fact. But in many cases the meaning of words must be ascertained in a context to which the parties gave no thought at all. Since the words induced no idea pertinent to the case at hand, it is unrealistic to pretend to be searching for the idea that the words induced. Cf. Corbin, supra, § 534 at 9 ("When a court is filling in gaps in the terms of an agreement, with respect to matters that the parties did not have in contemplation and as to which they had no intention to be expressed, the judicial process should not be called interpretation."). Instead we are searching for the legal operation of the words, the consequence that courts will ascribe to the use of the words as applied to a particular situation as to which the parties entertained no specific intention at all. In Professor Corbin's terms, that task is one of "construction," and it presents an issue of law.
The point may be illuminated by considering the somewhat comparable problem of identifying the roles of court and jury in determining the existence of negligence. Dean Prosser observes that the issue of negligence is often said to be "a mixed question of law and fact," which means "not only that both the court and the jury have an important part to play in the determination of the issue, and that separate functions are assigned to each, but further, that these functions to some extent overlap, and that it is not easy to fix any definite line of demarcation [footnote omitted]." W.L. Prosser, Law of Torts 205 (4th ed. 1971). He identifies as questions for the court determining the sufficiency of the evidence to permit a finding of the facts, determining the existence of a duty, and determining the general standard of conduct; that leaves for the jury determining the weight of the evidence as establishing the facts and determining the particular standard of conduct. Id. at 205-07. It is the distinction between the general and the particular standards of conduct that is relevant to the issue of contract interpretation. In tort law the court decides that, in most cases, the general standard is the conduct of a reasonable person under similar circumstances; in determining the particular standard, the jury decides what the reasonable person would have done under similar circumstances. This latter question, Dean Prosser observes, "is said to be one of fact, but it should be apparent that the function of the jury in fixing the standard differs from that of the judge only in that it cannot be reduced to anything approaching a definite rule." Id. at 207 (emphasis added). Thus, in deciding how the reasonable person would have acted under the circumstances, the jury is not answering a traditional inquiry as to fact; it is formulating a standard of law. It is a more refined standard than the general standard of reasonable care applicable to most circumstances, but it is nonetheless a standard, a particular standard formulated for the precise circumstances of a given case.
Professor Bohlen has observed that if the "reasonable man" were the "average man," his conduct under the circumstances disclosed by the evidence would be a question of fact. But he is not. His use, Professor Bohlen notes, enables the decision-maker to determine what ought to be done. Bohlen, Mixed Questions of Law and Fact, 72 Univ.Pa.L.Rev. 111, 113 (1924).
Ascertainment of the particular standard of care applicable to the circumstances of a case could be viewed as a question of law. We leave it to the jury in doubtful cases, Dean Prosser suggests, "because the public insists that its conduct be judged in part by the man in the street rather than by lawyers, and the jury serves as a shock-absorber to cushion the impact of the law" [footnote omitted]. Id. We do so also for the practical reason that since the circumstances of the accident are almost always in some dispute, a judge endeavoring to instruct the jury as to the particular standard of care applicable to the circumstances would have to formulate a variety of standards to fit with all of the permutations of facts that the jury was entitled to find.
In our Circuit Learned Hand noted that in cases tried to a jury the question of the proper standard of care is "treated" as a question of fact "although obviously it is not a question of fact, for it measures the duty and the liability which the law imposes; upon appeal in a cause tried to a judge, we think it should not be so regarded." Sidney Blumenthal Co. v. Atlantic Coast Line R. Co., 139 F.2d 288, 290 (2d Cir.), cert. denied, 321 U.S. 795, 64 S.Ct. 848, 88 L.Ed. 1084 (1943). Judge Friendly has observed that application of a legal standard to undisputed facts is "not exactly" either a question of fact or law and that once an appellate court recognizes that the issue is not one of fact, "there is no necessity for going further and endeavoring to establish that it is one of `law' in the usual meaning of that term." Baranow v. Gibraltar Factors Corp. (In re Hygrade Envelope Corp.). 366 F.2d 584, 588 (2d Cir. 1966). Holmes expressed the view that specific standards of conduct should increasingly be set by courts so that people would know in advance what they are supposed to do in a given set of circumstances. O.W. Holmes, Jr., The Common Law 110-12 (1881).
Professor Bohlen has suggested that determination of a particular standard of conduct is neither a question of fact or law but what he calls an "administrative" decision because it reflects a standard promulgated after the fact to determine liability in the case at hand and not primarily to guide conduct in the future. He observes, however, that the decision "is more nearly akin to a declaration of law than to a finding of fact . . . ." Bohlen, Mixed Questions of Law and Fact, supra, at 114.
The possibility of doing so has been long recognized. "[I]t is entirely possible to give a series of hypothetical instructions adapted to every state of facts which it is open to the jury to find." O.W. Holmes, Jr., The Common Law, supra, at 122. "Now it is perfectly clear that rules of law could be so formulated and so administered as to exclude the jury from making these evaluations [concerning the particular standard of conduct]." James, Functions of Judge and Jury in Negligence Cases, 58 Yale L.J. 667, 676 (1949).
In contract, as in tort, it strikes me as making sense to describe the inquiry concerning the response of the hypothetical reasonable person as more akin to an issue of law. The inquiry as to what the reasonable person would have thought the contract terms mean, like the inquiry as to how the reasonable person would have acted under the circumstances, is the ascertainment of a governing standard that is then applied to the facts of the case. In tort we prefer, as a matter of both policy and practicality, to let the jury select the particular standard of care; since it would be unsettling to say that the jury is deciding an issue of law, we prefer to say that the issue is one of fact, or sometimes, without identifying the different subquestions involved, we simply say, unhelpfully, that the overall issue of negligence is a mixed question of law and fact. In contract, however, the policy and, frequently, the practical considerations do not support jury consideration of the meaning of terms in the absence of extrinsic evidence. On the contrary, rather than wish for the leavening influence of a jury to decide in tort cases under what circumstances a person's conduct should render him liable for another's injuries, we should deem it in the public interest in contract cases to have as much certainty as possible as to the meaning of contracts, especially those involving terms used by others beyond the contracting parties. And, unlike most tort cases, the facts of a contract breach are frequently undisputed, requiring determination only of the applicable standard to be derived from the words of the contract. I therefore agree with those courts that view the textual interpretation of a contract, in the absence of extrinsic evidence, as essentially an issue of law. It is of special importance to do 80 in the instant case involving a standard form of contract for marine insurance, the uniform interpretation of which is significant for the entire shipping industry.
That policy considerations can and should determine the appropriate roles of court and jury has long been noted:
Logic and neatness of legal theory have always called loud, at least in recent centuries, for special verdicts, so that the true significance of ascertained facts might be ascertained and declared by the one tribunal fitted to do this finally and with authority. But considerations of policy have called louder for leaving the jury a freer hand. The working out of the jury system has never been shaped merely by legal or theoretical considerations. That body always represented the people, and came to stand as the guardian of their liberties; so that whether the court or the jury should decide a point could not be settled on merely legal grounds; it was a question deeply tinged with political considerations.
Thayer, Preliminary Treatise, supra, at 218.
Judge Friendly has suggested that the law-fact dichotomy in the context of contract interpretation is to be avoided. Meyers v. Selznick Co., supra, 373 F.2d at 222 n. 2. By whatever label one characterizes the issue of textual interpretation, it is important to recognize that its determination by a trial judge is not insulated upon review by the "clearly erroneous" standard, see Eddy v. Prudence Bonds Corp., 165 F.2d 157, 163 (2d Cir. 1947) (L. Hand, J.), cert. denied, 333 U.S. 845, 68 S.Ct. 664, 92 L.Ed. 1128 (1948).
In this case, though some brief testimony concerning custom was presented by the plaintiff, it is apparent that this testimony provides no guidance whatever as to the intent of the parties, and it was not so relied upon by the District Judge. If anything, that evidence, as the majority notes, would support the interpretation of the defendants. Thus, for me there is only the question of law as to what is a proper interpretation of the pertinent terms of the contract. I do not fault the plaintiff for not offering proof of the parties' understanding, nor the District Judge for not directing his analysis to that topic. In truth there was no evidence of the parties' understanding; they never contemplated the episode that gives rise to this suit, and they had no specific intention concerning coverage for the costs of removing cargo solidified by an explosion and adhering to bulkheads. Nor do I agree that Judge Conner implicitly found as a fact that the parties had contemplated coverage for these costs. Judge Conner, in my view, decided the case by applying the law's standard of what a reasonable person would think the terms mean. By doing so, he construed the contract as a matter of law. In disagreeing with Judge Conner, I do not conclude that he made a clearly erroneous interpretation. In this case, as in many other contract cases, there are two reasonable views as to what the law's hypothetical reasonable person would say the terms mean. Judge Conner chose one. We choose the other. If we were reviewing a finding of fact under the "clearly erroneous" standard, I would affirm. Because I believe we are reviewing a decision on a question of law on which my views coincide with the views of the majority, I join the majority in voting to reverse.
I disagree with the majority's suggestion that Judge Conner made a factual interpretation of the contract by pointing to the fact that the English underwriters paid their share of the disputed claim. That circumstance sheds no light on the factual inquiry as to what was in the minds of the parties to this lawsuit at the time they entered into their contract. At most it provides a circumstance arguably supporting Judge Conner's view that his legal interpretation of the contract was shared by those familiar with English law.