More than thirty years ago, between 1957 and 1964, CBS provided television stations with a program called "Have Gun--Will Travel." The program starred "Paladin," a fictional cowboy who dressed in black, carried a derringer pistol, and handed out calling cards with a picture of a chess knight. More than forty years ago, in 1947, Victor DeCosta, the plaintiff in this case, began to appear, as a cowboy, at rodeos, hospitals, and charitable events. DeCosta dressed in black, carried a derringer pistol, handed out cards with a picture of a chess knight, and called himself "Paladin." In 1963 DeCosta sued CBS, claiming it had unlawfully copied his idea. Eventually, this court decided that CBS may have copied DeCosta's idea, but, the laws under which DeCosta had sued did not prohibit CBS from doing so. This court held that DeCosta had failed to prove a violation of trademark, or other relevant, laws. Columbia Broadcasting System, Inc. v. DeCosta, 377 F.2d 315 (1st Cir.) [hereinafter DeCosta I ], cert. denied, 389 U.S. 1007, 88 S. Ct. 565, 19 L. Ed. 2d 603 (1967); DeCosta v. Columbia Broadcasting System, Inc., 520 F.2d 499 (1st Cir.1975) [hereinafter DeCosta II ], cert. denied, 423 U.S. 1073, 96 S. Ct. 856, 47 L. Ed. 2d 83 (1976).

DeCosta has now sued again. He has sued Viacom, a company that CBS created, and to which it assigned re-run rights for the old Paladin programs. He again complains that CBS copied his idea; and he says that Viacom, by broadcasting the old CBS programs, has violated federal and state trademark and unfair competition laws. 15 U.S.C. §§ 1114(1), 1125(a). The district court permitted the suit to proceed. DeCosta v. Viacom Int'l, Inc., 758 F. Supp. 807 (D.R.I.1991). A jury found in DeCosta's favor. And, Viacom appeals. In our view, DeCosta's new suit depends for its success upon relitigating issues that this court already has decided against him. And, for that reason, the doctrine of "collateral estoppel" bars his new claims. We therefore reverse the district court and order judgment for the defendant.

* DeCosta's Basic Legal Problems

When Mr. DeCosta first sued, many years ago, CBS claimed that it had not copied his "Paladin" character. Rather, CBS said, both "Paladin's" found their origin, independently, in the same historical sources. A jury, however, rejected CBS's argument. And, ever since, the courts have proceeded on the assumption that CBS, in fact, did copy Mr. DeCosta. Why, then, has Mr. DeCosta not succeeded in obtaining compensation?

The answer to this question ultimately rests upon the fact that the law does not always consider harmful, or always make unlawful, the copying by one person of the creation of another. Free, uncontrolled copying may, of course, prove harmful. It can discourage the creation of new, valuable ideas, works, or products, by diminishing the creator's monetary reward. It can cause commercial confusion, as a copier tries to take advantage of the good will attached to another's name. Free, uncontrolled copying, however, may also prove beneficial. It can promote the widespread dissemination of new works or ideas. "Education ... proceeds from a kind of mimicry, and 'progress,' if it is not entirely an illusion, depends on generous indulgence of copying." Benjamin Kaplan, An Unhurried View of Copyright 2 (1966). Some creators, say, novelists or dramatists, rightly expect compensation from those who buy or use their creations. Other creators, say, academic scientists, teachers, or certain commercial innovators (e.g., the inventor of the supermarket) expect others to copy, and to use, their ideas free of charge.

The result is a need for balance. Courts and legislators have responded to that need with separate, discrete bodies of intellectual property law, each with its own rules. The laws of patents, copyright, trade secrets, trademarks, unfair competition, and misappropriation balance the conflicting interests in protection and dissemination differently in different contexts through specific rules that determine just who will receive protection, of just what kind, under what circumstances, and for how long. See generally WCVB-TV v. Boston Athletic Ass'n, 926 F.2d 42, 45 (1st Cir.1991).

Mr. DeCosta's original legal problem lay in his inability to bring his case within a particular set of protective rules. Copyright law, for example, might in principle have offered protection for his "Have Gun--Will Travel" calling card, but he had brought that card into the "public domain" by distributing it widely, without giving the kind of specific "copyright" notice that federal copyright law requires. DeCosta I, 377 F.2d at 321. Nor was he able to show the type of "confusion" between products essential to success on his trademark, and most of his other, claims. DeCosta II, 520 F.2d at 513-15.

Mr. DeCosta's present legal problem lies in the fact that he previously sued CBS and lost. The traditional legal doctrine of "collateral estoppel" bars relitigation of any issue that, 1) a party had a "full and fair opportunity to litigate" in an earlier action, and that, 2) was finally decided in that action, 3) against that party, and that, 4) was essential to the earlier judgment. See Restatement (Second) of Judgments §§ 27, 29 (1982). Each of Mr. DeCosta's claims now before us depends, for its success, upon his winning an issue now that he lost before, in his litigation against CBS. In particular we held that he had failed to show a "likelihood" of buyer "confusion" between his "Paladin" character and that of CBS. That issue was "essential" to CBS's victory in the earlier action. Mr. DeCosta had a "full and fair opportunity to litigate" that issue in the earlier litigation. And, Viacom, as CBS's successor, here stands in the shoes of CBS.

Mr. DeCosta's argument on this appeal consists of an attempt to escape the bonds of "collateral estoppel" through a claim that legal and factual changes since 1975 (when we decided DeCosta II ) make the "confusion" issue, in essence, a new one. See Restatement (Second) of Judgments § 27, cmt. c; § 28(2)(b), (4) (collateral estoppel does not bar relitigation of an issue transformed by significant factual or legal changes). After considering Mr. DeCosta's arguments in detail, however, we find no legally significant change.

Trademark law seeks to prevent one seller from using the same "mark" as--or one similar to--that used by another in such a way that he confuses the public about who really produced the goods (or service). Confusion may prevent the buyer from obtaining the goods he really wants. It may also jeopardize the commercial reputation of the senior (first) user, which might be tarnished by association with the junior (subsequent) user. To win a trademark case, a plaintiff must show 1) that he uses, and thereby "owns," a mark, 2) that the defendant is using that same or a similar mark, and 3) that the defendant's use is likely to confuse the public, thereby harming the plaintiff. See, e.g., Dieter v. B & H Indus. of Southwest Florida, 880 F.2d 322, 326 (11th Cir.1989), cert. denied, --- U.S. ----, 111 S. Ct. 369, 112 L. Ed. 2d 332 (1990); WCVB-TV, 926 F.2d at 45; Astra Pharmaceutical Prods., Inc. v. Beckman Instruments, Inc., 718 F.2d 1201, 1205, 1209 (1st Cir.1983); Pignons S.A. de Mecanique v. Polaroid Corp., 657 F.2d 482, 486-87 (1st Cir.1981); W.W.W. Pharmaceutical Co. v. Gillette Co., 808 F. Supp. 1013, 23 U.S.P.Q.2d 1609, 1614, 1621 (S.D.N.Y.), reaff'd, amended in other respects, (S.D.N.Y. July 11, 1992); Merritt Forbes & Co. v. Newman Investment Securities, Inc., 604 F. Supp. 943, 956 (S.D.N.Y.1985); but cf. 15 U.S.C. § 1051(b) (creating an exception, not presently relevant, to the use requirement). DeCosta, as we have said, previously failed to show a "likelihood" of public "confusion" between the "mark" (i.e., the "Have Gun--Will Travel" and "Wire Paladin" phrases and the chess knight sign) as he used it, and the same "mark" as used by CBS. DeCosta argues that the legal "confusion" issue in the case before us differs from the issue in his earlier 1975 case because, in 1976, he registered his mark. Columbia Broadcasting System, Inc. v. DeCosta, 192 U.S.P.Q. 453 (T.T.A.B.1976). The fact of registration, he says, changes the legal "burden of proof" rules, making it legally easier for a plaintiff to show "likelihood of confusion." See American Heritage Life Ins. Co. v. Heritage Life Ins. Co., 494 F.2d 3, 10 (5th Cir.1974). That legal change transforms the old legal issue into a new one.

DeCosta is right that a change in relevant "burden of proof" rules can transform a legal issue, permitting relitigation of an issue that "collateral estoppel" otherwise would bar. See Restatement (Second) of Judgments § 28(4). But, we can find no such relevant transformation here. Registration consists of persuading the Trademark Board to issue an appropriate "certificate." 15 U.S.C. §§ 1051, 1057(a). The relevant statute says that the certificate becomes "prima facie evidence" of the mark's "registration," of its "validity," of "the registrant's ownership," and of the "registrant's exclusive right to use the mark ... in connection with the goods or services specified in the certificate." 15 U.S.C. § 1057(b). The statute also says that, in certain circumstances, "registration" makes the "registrant's right to use" the mark "incontestable." 15 U.S.C. § 1065. But, the statute nowhere says that registration makes it easier for a registrant (with a conceded right to use a concededly valid mark that he concededly owns) to prove that a relevant buying public may confuse some other person's mark with his own.

Nor are we aware of any reason why registration here should significantly affect the proof about confusion. To decide whether buyers are likely significantly to confuse two different marks, a court will examine 1) their similarity, 2) the similarity of the underlying goods or services, 3) the relation of the "channels" through which the parties "trade," 4) the relation of the parties' advertising, 5) the kinds of prospective buyers, 6) evidence of actual confusion, 7) the defendant's reasons for using the mark, and 8) the strength of the plaintiff's mark. Boston Athletic Ass'n v. Sullivan, 867 F.2d 22, 29 (1st Cir.1989); Volkswagen Aktiengesellschaft v. Wheeler, 814 F.2d 812, 817 (1st Cir.1987); Astra, 718 F.2d at 1205; Pignons, 657 F.2d at 487; Polaroid Corp. v. Polarad Elecs. Corp., 287 F.2d 492, 495 (2d Cir.), cert. denied, 368 U.S. 820, 82 S. Ct. 36, 7 L. Ed. 2d 25 (1961). The fact of registration, at most, relates to one aspect of the last mentioned factor, the plaintiff's mark's strength.

The particular relation of registration to "strength" concerns what trademark jargon calls "secondary meaning." That term refers to a word's, or a sign's, ability to tell the public that the word or sign serves a special trademark function, namely, that it denotes a product or service that comes from a particular "source." Words and phrases, in ordinary, non-trademark, use normally pick out, or refer to, particular individual items that exhibit the characteristics that the word or phrase connotes (without specific reference to the item's source). The phrase "white eagle," as ordinarily used on a particular occasion, for example, would pick out from a nearby flock of birds, the bird that has white, and eagle-like, characteristics. The phrase "white eagle" in trademark use, however, would denote, or refer to, something special, something other than a white eagle, such as, for example, a beer--"White Eagle Beer"--and, in doing so, it would signify that the item comes from a particular source, say, the White Eagle Company, a Milwaukee brewery. And, insofar as the public takes the word, or sign, to refer to a product or service with a particular source (indeed, a product or service that, for example, might be neither white, nor like an eagle), the word, or sign, has "secondary meaning." Registration relates to strength in that it helps a court conclude that a particular mark, in fact, does possess such "secondary meaning," which is simply to say that it helps a court conclude that a particular set of words or signs does indeed act like a trademark. Dieter, 880 F.2d at 329; Wynn Oil Co. v. Thomas, 839 F.2d 1183, 1187, 1190 (6th Cir.1988); Keds Corp. v. Renee Int'l Corp., 888 F.2d 215, 220-21 (1st Cir.1989); cf. New Kids on the Block v. News America Publishing, Inc., 971 F.2d 302, 305, 23 U.S.P.Q.2d 1534, 1535 (9th Cir.1992) (describing the primary purpose of trademarks as "to identify the source of goods and services").

That "strength" relates to confusion and registration "relates" (in this way) to strength, however, does not help DeCosta. The "strength" of DeCosta's mark was not an issue before us in the earlier cases. This court assumed in its opinions that DeCosta's mark had a secondary meaning. It specifically said that his mark is "distinctive enough so that proof of secondary meaning is not essential," and that, "at least among some people, plaintiff's name and card had come to be associated with him." DeCosta II, 520 F.2d at 513. It went on to find no relevant "confusion," even assuming a mark as strong as registration might have forced it to assume. Thus, the fact of later registration, insofar as it helps establish that the mark has a "secondary meaning," adds nothing significantly new.

We concede that, in one of the cases that DeCosta cites, the court said that "registration is sufficient to establish prima facie (1) the required prior use (2) of a registrable mark (3) which is likely to be confused with another's use of the same or a similar mark." American Heritage, 494 F.2d at 10 (emphasis added). The underlined phrase, however, likely refers simply to the mark's "strength," in which case the phrase is consistent with holdings in other courts. Dieter, 880 F.2d at 329; Wynn, 839 F.2d at 1187, 1190; Keds, 888 F.2d at 220-21. If it means more than that, we do not understand the theory behind it, and we do not follow it. Rather, we agree with the Seventh Circuit, that, in the case before us, "the procedural advantages conferred by registration are [not] substantial, at least in the context of determining the issue of likelihood of confusion. Therefore, we see no inequity in applying collateral effect to the [prior] decision [that likelihood of confusion was not shown]." EZ Loader Boat Trailers, Inc. v. Cox Trailers, Inc., 746 F.2d 375, 379 (7th Cir.1984). In sum, the "new" fact of registration does not warrant relitigating the "likelihood of confusion" issue.

DeCosta next argues that collateral estoppel does not bind him because, since 1975, there has occurred a "modification or growth in legal principles [that] effect a significant change" in the law. See Commissioner of Internal Revenue v. Sunnen, 333 U.S. 591, 600, 68 S. Ct. 715, 720, 92 L. Ed. 898 (1948); Restatement (Second) of Judgments § 28(2)(b). His "new" case, he adds, rests upon a claim that Viacom's present use of "Paladin" would violate this "new" law, not "old," pre-1975, legal doctrine. Hence, the fact that the behavior of CBS (Viacom's predecessor) was lawful before 1975 tells us nothing about Viacom's similar behavior today.

DeCosta finds these "new" legal principles in an area of trademark law called "reverse confusion," an area in which a plaintiff claims that the public will confusedly think that the plaintiff's product emanates in some way from the defendant, rather than the (more ordinary) contrary. An imaginary example may help explain the concept.

Suppose that Tom, in 1970, established a knife company, which used the trademarked name "SupR-Chopper." Later, say, in 1975, Mary established an electric kitchen-blender company, and she used the same "SupR-Chopper" name on her kitchen-blenders. In 1980, Tom sues Mary. To win, Tom must show "confusion." In a traditional trademark confusion case, Tom will claim that Mary's kitchen-blender customers may confusedly think that he, Tom, has expanded into the kitchen-blender business, either directly or by "sponsoring" (i.e., authorizing) Mary to use the "SupR-Chopper" name. If they think that he, Tom, makes (or sponsors) the kitchen-blender, their dissatisfaction with Mary's kitchen-blenders may harm the reputation of Tom's knives; or, even if Mary makes a fine product, insofar as her customers are moved to buy Mary's product because they associate Tom with Mary's product, they thereby permit Mary to take a "free-ride" on the work and investment that Tom made in order to develop a positive image for the name "SupR-Chopper." See, e.g., S.C. Johnson & Son, Inc. v. Johnson, 175 F.2d 176, 180 (2d Cir.), cert. denied, 338 U.S. 860, 70 S. Ct. 103, 94 L. Ed. 527 (1949) (recognizing trademark owner's legitimate claim to protection from "the possibility that the trade practices of the second user may stain the owner's reputation in the minds of his customers"); Triangle Publications, Inc. v. Rohrlich, 167 F.2d 969, 972 (2d Cir.1948) (prohibiting defendant's attempt to profit from "the erroneously supposed sponsorship of the plaintiff").

The less typical, "reverse confusion," case involves somewhat different circumstances. In such a case, Tom is worried that his knife customers will wrongly think that Mary makes, or "sponsors" his, Tom's, kitchen knives (not that he makes Mary's blenders). This "reverse confusion," just like ordinary confusion, may hurt Tom. If Mary's kitchen-blenders work badly, for example, Tom's potential customers may decide that Tom's knives come from a poorly managed company, and they may hesitate to buy them. See, e.g., Plus Products v. Plus Discount Foods, Inc., 722 F.2d 999, 1003-04 (2d Cir.1983) (recording plaintiff's concern that reverse confusion might lead to plaintiff's "reputation for high quality merchandise" becoming "tarnished because of [defendant's] bargain basement, no-frills image"); Banff, Ltd. v. Federated Dep't Stores, Inc., 841 F.2d 486, 490 (2d Cir.1988) (similar).

The problem for DeCosta is that, as illustrated by this simplified example, "reverse confusion" is nothing particularly new. The principal trademark statute does not speak of "ordinary," or "reverse," confusion. It refers simply to copying that is "likely to cause confusion," without dividing confusion into types. 15 U.S.C. § 1114(1). It protects the holder from the harm that confusion might cause, without specifying whether that harm flows from a copier taking advantage of the holder's "good will," or from the copier potentially reducing the value of the mark, say by associating the holder with the copier's own "bad" name. The leading case on the subject, Big O Tire Dealers, Inc. v. Goodyear Tire & Rubber Co., 561 F.2d 1365, 1371 (10th Cir.1977), cert. dismissed, 434 U.S. 1052, 98 S. Ct. 905, 54 L. Ed. 2d 805 (1978), decided two years after our decision in DeCosta II, made absolutely clear that a trademark holder could base a claim on such "reverse confusion." But, in doing so, the court did not suggest that its holding represented a totally new, or novel, principle. The jury instruction which it upheld followed the language of § 1114(1). See id. at 1371. And, Big O reasserted a principle set forth in a case that DeCosta invoked at length in DeCosta I, namely, International News Service v. Associated Press, 248 U.S. 215, 39 S. Ct. 68, 63 L. Ed. 211 (1918). In that case, in addition to deciding that the defendant had unlawfully misappropriated the news product of the plaintiff's investigations, the Court found a wrongful and "significant ... false representation" that that news product was derived from defendant's own work. Id. at 242, 39 S.Ct. at 73. Justice Holmes' concurrence rested entirely on the "false representation" aspect, which he viewed as unfair competition:

The ordinary case [of unfair competition] is palming off the defendant's product as the plaintiff's, but the same evil may follow from the opposite falsehood--from saying, whether in words or by implication, that the plaintiff's product is the defendant's, and that ... is what has happened here.

Id. at 247, 39 S.Ct. at 75; see also Banff, 841 F.2d at 490 (observing that trademark law's traditional objectives, concerned with ensuring that good will remains attached to those who earn it, are equally implicated in "reverse" and "non-reverse" confusion cases). DeCosta cannot therefore claim that the principle (that trademark law protects against a buyer's being led to believe, wrongly and harmfully, that the copier is the source of the holder's product) was unavailable to him or that he was not aware of it in his initial case.

DeCosta's more plausible claim is that the law of "reverse confusion" has itself undergone significant expansion since 1975. Several dicta in the Second Circuit suggest that a plaintiff, claiming reverse confusion, can recover for harm suffered, not because the buying public may wrongly believe that the defendant makes or sponsors the plaintiff's product, but simply because the public wrongly believes that the plaintiff copied the defendant's name. See Banff, 841 F.2d at 490; Lobo Enters., Inc. v. Tunnel, Inc., 693 F. Supp. 71, 77 (S.D.N.Y.1988); PAF S.r.l. v. Lisa Lighting Co., 712 F. Supp. 394, 410 (S.D.N.Y.1989); W.W.W. v. Gillette, 23 U.S.P.Q.2d at 1615. To return to our example, it is as if Tom could win his trademark case even if everyone knows that Tom and Tom's knife company have nothing whatsoever to do with Mary's kitchen-blenders. Tom might still win because the public might wrongly think that Mary thought of the "SupR-Chopper" name first and Tom copied her idea for his trademark. If the public wrongly thought that Tom "pirated" the name, they might think less well of Tom, who would thereby suffer a harm to his good reputation. Were this theory the law, Tom might win a trademark case against Mary, even if Mary used the "SupR-Chopper" name to label a product that nobody thought had anything whatsoever to do with kitchen knives, say, helicopters (called "SupR-Choppers"), for the public still might wrongly believe Tom a "pirate."

We agree with DeCosta that one can find dicta, more recent than 1975, that seem to offer support for such a theory. And, we also agree that such a theory might have offered him a basis for success, had he known of its availability before 1975. The fatal problem for DeCosta in respect to this "change in the law," however, lies in our view that this change is not sound law. We find that it does not correctly state the law of trademarks.

Our reasons for this conclusion are several. First, to adopt this theory would undermine an important limitation central to the law of trademarks, the limitation of trademark protection to the protection of marks as used on particular goods to identify their source or sponsor. See United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90, 97, 39 S. Ct. 48, 50, 63 L. Ed. 141 (1918) (trademark rights are not "right[s] in gross"). As the Supreme Court wrote many years ago, in Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 415, 36 S. Ct. 357, 361, 60 L. Ed. 713 (1916):

[W]here two parties independently are employing the same mark ... in separate markets wholly remote the one from the other, the question of prior appropriation is legally insignificant, unless at least it appear that the second adopter has selected the mark with some design inimical to the interests of the first user, such as to take the benefit of the reputation of his goods, to forestall the extension of his trade, or the like.

Thus, at present the law often permits a person to take a pre-existing name or mark and use it on a different product in a different market. See, e.g., McGregor-Doniger, Inc. v. Drizzle, Inc., 599 F.2d 1126 (2d Cir.1979) (allowing manufacturer of expensive women's coats to use trademark "Drizzle," despite prior registration of "Drizzler" mark for plaintiff's cheaper golf jackets); King Research, Inc. v. Shulton, Inc., 454 F.2d 66 (2d Cir.1972) ("Ship Shape" on hairspray did not infringe registered "Ship Shape" trademark for comb and brush cleaners). If "falsely being thought a pirate" were an actionable harm, no one could safely use a mark ever previously used by another, no matter how different the product, place of sale, or class of buyer. Mary the helicopter maker, for example, would have to make certain that no small company anywhere had used the name "SupR-Chopper" on any product before she attached the name to her helicopter product, lest some of, say, Tom's knife customers believe that she, not Tom, had had the idea first. The specter of resulting lawsuits, inhibitions on the use of names, and a reversal of present presumptions favoring linguistic freedom (in different fields) cautions against what would seem a fairly radical change in the law.

Second, other, non-trademark law offers specifically tailored protection against the most obvious harms that may befall the falsely labeled "pirate." Copyright law, for example, protects the initial users of certain names and phrases against any copier. We have mentioned the possibility that DeCosta might have obtained such protection, at least for his calling card. See DeCosta I, 377 F.2d at 321. But, he did not do so.

A common law tort, the law against "commercial disparagement" (also known as "injurious falsehood"), may also protect a trademark holder against the false implication that he has "pirated" the work of another, where the defendant intends such harm. See, e.g., Public Ledger v. New York Times, 275 F. 562, 565-66 (S.D.N.Y.1921), aff'd, 279 F. 747, cert. denied, 258 U.S. 627, 42 S. Ct. 383, 66 L. Ed. 798 (1922) (allowing relief for defendant's assertion that it had copied "with permission" from the London Times as disparaging plaintiff's rights, if plaintiff could prove that, as defendant knew, plaintiff had--and advertised itself as having--a contract with the London Times guaranteeing it exclusive copying rights); Big O, 561 F.2d at 1373-74.

Both these areas of law, however, contain carefully crafted conditions and limitations, designed to prevent their becoming vehicles for unduly limiting the use of words, phrases, and other forms of speech where no serious harm, in fact, will likely occur. See, e.g., copyright law's "fair use" exemption, 17 U.S.C. § 107; see also Big O, 561 F.2d at 1373 (outlining the special requirements of "commercial disparagement," namely (1) false statement, (2) malice, and (3) special damages). The existence of these other carefully tailored types of protection also cautions strongly against introducing, into trademark law, a kind of overriding concept such as the actionable harm of "falsely being thought a pirate," which concept could well upset the balance between those interests favoring "protection" and those favoring free use and dissemination--a balance carefully developed by legislatures, and slowly by courts, over many years.

Finally, the leading case about trademark "reverse confusion," Big O, supra, suggests that "commercial disparagement" law, not traditional "trademark law," provides proper legal relief for the harm of "falsely being thought a pirate." See id. at 1373-74. Insofar as the court discusses this latter kind of harm, it does so in the context of a "commercial disparagement" type of tort. Insofar as the court discusses trademark "reverse confusion," it does so in the context of confusion about the source of the product, not the source of the name. The court does explain why, in its view, trademark law does not limit recovery to victims of "passing off." And, in doing so, it says that, otherwise, a large firm could simply take someone else's mark and develop a new "secondary meaning" for it. But, nothing in this explanation suggests that "falsely being thought a pirate" automatically produces recovery.

For these reasons, insofar as the doctrine of "reverse confusion" may be thought significantly "new" (reverse confusion about "piracy"), we do not accept it. Insofar as we accept it (reverse confusion involving source or sponsorship), we do not believe it is significantly new. Hence, we do not believe that there are changes in the law here that can overcome the effects of "collateral estoppel."

DeCosta argues that facts have changed since 1975. Hence, the issue of "confusion now" is significantly different than the issue of "confusion then." And, "collateral estoppel" does not bar its litigation. See Restatement (Second) of Judgments § 27, cmt. c. In the earlier cases DeCosta proved that he presented the character "Paladin" at rodeos and through various personal appearances. CBS presented the character "Paladin" in its television programs. This court held that few, if any, buyers of either "product" (rodeo/personal appearances or television programs) would likely believe that either DeCosta or CBS was the "source" of the other's "service." That is to say, few, if any, television viewers were likely to believe that DeCosta produced the TV programs and few, if any, rodeo (or personal service) customers were likely to believe that CBS provided the rodeo, or other personal, appearances. The question is whether the factual changes to which DeCosta now points are such that litigation of the "confusion" question represents a significantly different factual issue. See id.

After reviewing the record, we conclude that DeCosta has not presented evidence of significantly different circumstances for two reasons. First, much of his evidence amounts to no more than added efforts to prove the same "ultimate facts" he failed to prove the first time. Thus, DeCosta found several people (and produced surveys showing other people) who think he has something to do with the CBS television program. He provided four letters from persons who referred to him as "Paladin." He provided a witness who said he thought DeCosta was impersonating the television program character. He introduced a newsletter that says he is "from the T.V. series." He testified that he had met people who thought he was "connected" with the TV series and was an "impersonator" or an "imposter." One of his surveys said that about half of the individuals shown his picture thought he was connected with the TV series or that it was "sponsoring" him. And, he produced a "public relations" expert witness who testified that people would be "confused."

This evidence does not help DeCosta, however, because, in context, it seems designed to prove the same ultimate fact--"confusion"--that he failed to prove before. As the Restatement of Judgments points out, when a party has litigated such an "ultimate fact," and failed, "new evidentiary facts may not be brought forward to obtain a different determination of that ultimate fact." Id. We simply do not see why this kind of evidence could not have been provided the first time. Nothing in the record convincingly explains why those who saw DeCosta when the Paladin television program was current would have been any the less "confused" than those who now see him when "Paladin" is the subject of old television reruns.

Second, DeCosta provided evidence of his having expanded his own activities since 1977. He says, for example, that since that time, he has distributed 60,000 more calling cards (having distributed about 300,000 before 1977); 15,000 more photographs (having distributed about 20,000 before 1977); 15,000 bumper stickers, and 2800 pens with a Paladin legend. He has made more personal appearances at rodeos and ice cream stores, and he appeared on two television talk shows and in one television commercial. A picture of him in costume appeared once in a horsebreeders' magazine. And, he has objected several times to others using slogans such as "Have Cup, Will Travel" (by Dunkin Donuts) and "Have Guns, Will Travel" (by the U.S. Air Force).

This evidence basically shows no more than the same kind of activity in which DeCosta previously engaged. And, we do not see how it can bring him outside the "collateral estoppel" bar. Even were he to have provided evidence of his own, far greater, expansion into, say, the television business, that evidence would do him no good. The litigated holding of "no confusion" in the initial DeCosta cases amounts to a holding that DeCosta had no legal right to exclude others from using his mark in the field of television. Moreover, DeCosta has conceded that CBS/Viacom has used the "Paladin" mark in that field before, and after, he brought his initial cases. Further, the record provides no evidence at all that, since 1977, CBS or Viacom has used the mark in "bad faith," i.e., with an intent or expectation of causing confusion or "forestalling expansion under the mark by the prior user" (in a different field) or harming DeCosta's "reputation or good will." See Restatement (Third) of Unfair Competition § 19(a), cmt. d & illus. 3 (Tent. Draft No. 2, 1990); GTE Corp. v. Williams, 904 F.2d 536, 541 (10th Cir.), cert. denied, --- U.S. ----, 111 S. Ct. 557, 112 L. Ed. 2d 564 (1990) (in assessing whether second user acted in bad faith, "[t]he ultimate focus is on whether the second user had the intent to benefit from the reputation or goodwill of the first user."); Triumph Hosiery Mills, Inc. v. Triumph Int'l Corp., 308 F.2d 196, 200 (2d Cir.1962) (similar); El Chico, Inc. v. El Chico Cafe, 214 F.2d 721, 726 (5th Cir.1954) (similar); James M. Treece, "Security for Federally Registered Mark Owners Against Subsequent Users," 39 Geo.Wash.L.Rev. 1008, 1018 (1971) (mere knowledge of the first user's prior use should not be regarded as bad faith in cases where "the second user [is] in fact remote from the first user's market [and] where consumers are not confused"); cf. Mead Data Central, Inc. v. Toyota Motor Sales, U.S.A., Inc., 875 F.2d 1026, 1037 (2d Cir.1989) (Sweet, J., concurring) (in statutory dilution context, bad faith "requires a showing that the junior user adopted its mark hoping to benefit commercially from association with the senior mark."). Indeed, CBS/Viacom might reasonably have relied upon our prior final judgment as holding that their activities did not cause "confusion" or significantly harm "reputation or good will." Cf. Restatement (Third) of Unfair Competition § 19, cmt. d ("Good faith reliance by the subsequent user on an opinion of counsel is also relevant.").

All this being so, as far as the present case is concerned, it is CBS and Viacom who have the prior right to use the mark in television, not DeCosta. And, for that reason, insofar as DeCosta's expansion into television creates "confusion," he has no legal basis for recovery. See United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90, 39 S. Ct. 48, 63 L. Ed. 141 (1918) (within regional market, defendant first user in that market had priority over plaintiff earlier user in a different region who now sought to enter that market); Value House v. Phillips Mercantile Co., 523 F.2d 424 (10th Cir.1975) (same, where plaintiff registered its mark after defendant had begun to use its); compare Dawn Donut Co. v. Hart's Food Stores, Inc., 267 F.2d 358, 360 (2d Cir.1959) (plaintiff who registered before defendants began to use their mark, and had previously operated in defendants' market and not abandoned its mark in that market, retained priority); see also Scott Paper Co. v. Scott's Liquid Gold, Inc., 589 F.2d 1225, 1231 (3d Cir.1978) ("Priority depends not upon which mark succeeds in first obtaining secondary meaning but upon whether the plaintiff can prove by a preponderance of the evidence that his mark possessed secondary meaning [and, we add, that there was a potential likelihood of confusion] at the time the defendant commenced his use of the mark.") (emphasis added).

Returning to our example, it is as if Tom, the kitchen knife maker, sued Mary the helicopter manufacturer, and a court determined that their use of the same name did not create confusion. Suppose that Mary continues, in good faith, to use the mark on her helicopters, but Tom then expands into the helicopter business. At that point, even if buyers now confuse the source of the two products (Tom's helicopters and Mary's helicopters), Tom cannot recover from Mary, for it is Mary, not Tom, who has the legally prior right to use the name in that field.

DeCosta reminds us that, since findings as to likelihood of confusion can turn on the relation of the parties' uses, a prior finding of no likelihood of confusion will not always bar a subsequent action if circumstances of the parties' uses change. See Sarah Coventry, Inc. v. T. Sardelli & Sons, Inc., 526 F.2d 20, 23 (1st Cir.1975), cert. denied, 426 U.S. 920, 96 S. Ct. 2626, 49 L. Ed. 2d 374 (1976). The problem for DeCosta is that his evidence does not show a significant change. And, in any event, that change would do a plaintiff no good where it consists of his expansion into a field where the record of litigation indicates that the defendant has priority in using the mark.

For these reasons, the judgment of the district court is

Reversed.