Toyota Motor Sales, U.S.A., Inc. and its parent, Toyota Motor Corporation, appeal from a judgment of the United States District Court for the Southern District of New York (Edelstein, J.) enjoining them from using LEXUS as the name of their new luxury automobile and the division that manufactures it. The district court held that, under New York's antidilution statute, N.Y.Gen.Bus.Law Sec. 368-d, Toyota's use of LEXUS is likely to dilute the distinctive quality of LEXIS, the mark used by Mead Data Central, Inc. for its computerized legal research service, 702 F. Supp. 1031 (1988). On March 8, 1989, we entered an order of reversal, stating that an opinion would follow. This is the opinion.1

Section 368-d of New York's General Business Law, which has counterparts in at least twenty other states, reads as follows:

Likelihood of injury to business reputation or of dilution of the distinctive quality of a mark or trade name shall be a ground for injunctive relief in cases of infringement of a mark registered or not registered or in cases of unfair competition, notwithstanding the absence of competition between the parties or the absence of confusion as to the source of goods or services.

Mead is a corporation organized under the laws of Delaware with its principal place of business in Miamisburg, Ohio. Since 1972, Mead has provided a computerized legal research service under the trademark LEXIS. Mead introduced evidence that its president in 1972 "came up with the name LEXIS based on Lex which was Latin for law and I S for information systems." In fact, however, the word "lexis" is centuries old. It is found in the language of ancient Greece, where it had the meaning of "phrase", "word", "speaking" or "diction". Pinkerton, Word for Word, 179 (1982). "Lexis" subsequently appeared in the Latin where it had a substantially similar meaning, i.e., "word", "speech", or "language". Oxford Latin Dictionary (1983); Lewis and Short, A Latin Dictionary (1980); Lewis, An Elementary Latin Dictionary (1979).

Like many other Latin words, "lexis" has been incorporated bodily into the English. It can be found today in at least sixty general dictionaries or other English word books, including Webster's Ninth New Collegiate Dictionary and Webster's New World Dictionary. Moreover, its meaning has not changed significantly from that of its Latin and Greek predecessors; e.g., "Vocabulary, the total set of words in a language" (American Heritage Illustrated Encyclopedic Dictionary); "A vocabulary of a language, a particular subject, occupation, or activity" (Funk & Wagnalls Standard Dictionary). The district court's finding that "to establish that LEXIS is an English word required expert testimony at trial" is clearly erroneous. Anyone with a rudimentary knowledge of English can go to a library or bookstore and find the word in one of the above-mentioned standard dictionaries.

Moreover, the record discloses that numerous other companies had adopted "Lexis" in identifying their business or its product, e.g., Lexis Ltd., Lexis Computer Systems Ltd., Lexis Language and Export Information Service, Lexis Corp., Maxwell Labs Lexis 3. In sum, we reject Mead's argument that LEXIS is a coined mark which originated in the mind of its former president and, as such, is entitled per se to the greater protection that a unique mark such as "Kodak" would receive. See Esquire, Inc. v. Esquire Slipper Mfg. Co., 243 F.2d 540, 543 (1st Cir.1957); Intercontinental Mfg. Co. v. Continental Motors Corp., 230 F.2d 621, 623 (C.C.P.A.1956).

Nevertheless, through its extensive sales and advertising in the field of computerized legal research, Mead has made LEXIS a strong mark in that field, and the district court so found. In particular, the district court accepted studies proffered by both parties which revealed that 76 percent of attorneys associated LEXIS with specific attributes of the service provided by Mead. However, among the general adult population, LEXIS is recognized by only one percent of those surveyed, half of this one percent being attorneys or accountants. The district court therefore concluded that LEXIS is strong only within its own market.

As appears in the Addendum to this opinion, the LEXIS mark is printed in block letters with no accompanying logo.

Toyota Motor Corp. has for many years manufactured automobiles, which it markets in the United States through its subsidiary Toyota Motor Sales, U.S.A. On August 24, 1987 Toyota announced a new line of luxury automobiles to be called LEXUS. The cars will be manufactured by a separate LEXUS division of Toyota, and their marketing pitch will be directed to well-educated professional consumers with annual incomes in excess of $50,000. Toyota had planned to spend $18 million to $20 million for this purpose during the first nine months of 1989.

Before adopting the completely artificial name LEXUS for its new automobile, Toyota secured expert legal advice to the effect that "there is absolutely no conflict between 'LEXIS' and 'LEXUS.' " Accordingly, when Mead subsequently objected to Toyota's use of LEXUS, Toyota rejected Mead's complaints. The district court held correctly that Toyota acted without predatory intent in adopting the LEXUS mark.

[T]he absence of predatory intent by the junior user is a relevant factor in assessing a claim under the antidilution statute, ... since relief under the statute is of equitable origin,....

Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 626 (2d Cir.1983) (citations omitted).

However, the district court erred in concluding that Toyota's refusal to acknowledge that its use of LEXUS might harm the LEXIS mark, deprived it of the argument that it acted in good faith. If, as we now hold, Toyota's mark did not dilute Mead's, it would be anomalous indeed to hold Toyota guilty of bad faith in proceeding in reliance on its attorney's correct advice to that effect. See Sweats Fashions, Inc. v. Pannill Knitting Co., 833 F.2d 1560, 1565 (Fed.Cir.1987); E.S. Originals Inc. v. Stride Rite Corp., 656 F. Supp. 484, 490 (S.D.N.Y.1987); Inc. Publishing Corp. v. Manhattan Magazine, Inc., 616 F. Supp. 370, 394-96 (S.D.N.Y.1985), affd, 788 F.2d 3 (2d Cir.1986); Procter & Gamble Co. v. Johnson & Johnson, Inc., 485 F. Supp. 1185, 1201-02 (S.D.N.Y.1979), aff'd, 636 F.2d 1203 (2d Cir.1980). Indeed, even if the attorney's professional advice had been wrong, it does not follow that Toyota's reliance on that advice would have constituted bad faith. Information Clearing House, Inc. v. Find Magazine, 492 F. Supp. 147, 161-62 (S.D.N.Y.1980).

The LEXUS mark is in stylized, almost script-like lettering and is accompanied by a rakish L logo. See Addendum.

The brief legislative history accompanying section 368-d describes the purpose of the statute as preventing "the whittling away of an established trade-mark's selling power and value through its unauthorized use by others upon dissimilar products." 1954 N.Y.Legis.Ann. 49 (emphasis supplied). If we were to interpret literally the italicized word "its", we would limit statutory violations to the unauthorized use of the identical established mark. This is what Frank Schechter, the father of the dilution theory, intended when he wrote The Rational Basis of Trademark Protection, 40 Harv.L.Rev. 813 (1927). See id. at 830-33; see also Shire, Dilution Versus Deception--Are State Antidilution Laws an Appropriate Alternative to the Law of Infringement?, 77 Trademark Rep. 273-76 (1987). However, since the use of obvious simulations or markedly similar marks might have the same diluting effect as would an appropriation of the original mark, the concept of exact identity has been broadened to that of substantial similarity. Community Federal Savings and Loan Ass'n v. Orondorff, 678 F.2d 1034 (11th Cir.1982) (quoting Pro-phy-lac-tic Brush Co. v. Jordan Marsh Co., 165 F.2d 549, 553 (1st Cir.1948)); Dreyfus Fund, Inc. v. Royal Bank of Canada, 525 F. Supp. 1108, 1124 (S.D.N.Y.1981); 2 J. McCarthy, Trademarks and Unfair Competition Sec. 24:13 at 215 (2d ed. 1984). Nevertheless, in keeping with the original intent of the statute, the similarity must be substantial before the doctrine of dilution may be applied. See Alberto-Culver Co. v. Andrea Dumon, Inc., 466 F.2d 705, 709 (7th Cir.1972); Consolidated Cosmetics v. Neilson Chemical Co., 109 F. Supp. 300, 310 (E.D.Mich.1952); Ehrlich, Anti-Dilution Laws Give Plaintiffs Powerful Weapon Against Copiers, Nat'l L.J., May 16, 1983, at 28.

Indeed, some courts have gone so far as to hold that, although violation of an antidilution statute does not require confusion of product or source, the marks in question must be sufficiently similar that confusion may be created as between the marks themselves. See Holiday Inns, Inc. v. Holiday Out in America, 481 F.2d 445, 450 (5th Cir.1973); King Research, Inc. v. Shulton, Inc., 324 F. Supp. 631, 638 (S.D.N.Y.1971), aff'd, 454 F.2d 66 (2d Cir.1972). We need not go that far. We hold only that the marks must be "very" or "substantially" similar and that, absent such similarity, there can be no viable claim of dilution.

The district court's opinion was divided into two sections. The first section dealt with Toyota's alleged violation of the Lanham Act, and the second dealt with the alleged dilution of Mead's mark under New York's antidilution statute. The district court made several findings on the issue of similarity in its Lanham Act discussion; it made none in its discussion of section 368-d. Assuming that the district court's finding of lack of physical similarity in the former discussion was intended to carry over into the latter, we would find ourselves in complete accord with it since we would make the same finding. See Addendum; see also Blue Bell, Inc. v. Jaymar-Ruby, Inc., 497 F.2d 433, 435 (2d Cir.1974). However, if the district court's statement in its Lanham Act discussion that "in everyday spoken English, LEXUS and LEXIS are virtually identical in pronunciation" was intended to be a finding of fact rather than a statement of opinion, we question both its accuracy and its relevance. The word LEXUS is not yet widely enough known that any definitive statement can be made concerning its pronunciation by the American public. However, the two members of this Court who concur in this opinion use "everyday spoken English", and we would not pronounce LEXUS as if it were spelled LEXIS. Although our colleague takes issue with us on this point, he does not contend that if LEXUS and LEXIS are pronounced correctly, they will sound the same. We liken LEXUS to such words as "census", "focus" and "locus", and differentiate it from such words as "axis", "aegis" and "iris".2 If we were to substitute the letter "i" for the letter "u" in "census", we would not pronounce it as we now do. Likewise, if we were to substitute the letter "u" for the letter "i" in "axis", we would not pronounce it as we now do. In short, we agree with the testimony of Toyota's speech expert, who testified:

Of course, anyone can pronounce "lexis" and "lexus" the same, either both with an unstressed I or both with an unstressed U, or schwa--or with some other sound in between. But, properly, the distinction between unstressed I and unstressed U, or schwa, is a standard one in English; the distinction is there to be made in ordinary, reasonably careful speech.

In addition, we do not believe that "everyday spoken English" is the proper test to use in deciding the issue of similarity in the instant case. Under the Constitution, there is a " 'commonsense' distinction between speech proposing a commercial transaction, which occurs in an area traditionally subject to government regulation, and other varieties of speech." Central Hudson Gas & Electric Corp. v. Public Service Comm'n, 447 U.S. 557, 562, 100 S. Ct. 2343, 2349, 65 L. Ed. 2d 341 (1980) (quoting Ohralik v. Ohio State Bar Ass'n, 436 U.S. 447, 455-56, 98 S. Ct. 1912, 1918-19, 56 L. Ed. 2d 444 (1978)). "The legitimate aim of the anti-dilution statute is to prohibit the unauthorized use of another's trademark in order to market incompatible products or services", and this constitutes a "legitimate regulation of commercial speech." L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 32-33 (1st Cir.), cert. denied, 483 U.S. 1013, 107 S. Ct. 3254, 97 L. Ed. 2d 753 (1987). "Advertising is the primary means by which the connection between a name and a company is established ...", Beneficial Corp. v. Beneficial Capital Corp., 529 F. Supp. 445, 448 (S.D.N.Y.1982), and oral advertising is done primarily on radio and television. When Mead's speech expert was asked whether there were instances in which LEXUS and LEXIS would be pronounced differently, he replied "Yes, although a deliberate attempt must be made to do so.... They can be pronounced distinctly but they are not when they are used in common parlance, in everyday language or speech." We take it as a given that television and radio announcers usually are more careful and precise in their diction than is the man on the street. Moreover, it is the rare television commercial that does not contain a visual reference to the mark and product, which in the instant case would be the LEXUS automobile. We conclude that in the field of commercial advertising, which is the field subject to regulation, there is no substantial similarity between Mead's mark and Toyota's.

There are additional factors that militate against a finding of dilution in the instant case. Such a finding must be based on two elements. First, plaintiff's mark must possess a distinctive quality capable of dilution. Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., 42 N.Y.2d 538, 545, 399 N.Y.S.2d 628, 369 N.E.2d 1162 (1977). Second, plaintiff must show a likelihood of dilution, Sally Gee, Inc. v. Myra Hogan, Inc., supra, 699 F.2d at 625. As section 368-d expressly states, a plaintiff need not show either competition between its product or service and that of the defendant or a likelihood of confusion as to the source of the goods or services. Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., supra, 42 N.Y.2d at 543, 399 N.Y.S.2d 628, 369 N.E.2d 1162.

Distinctiveness for dilution purposes often has been equated with the strength of a mark for infringement purposes. P.F. Cosmetique, S.A. v. Minnetonka, Inc., 605 F. Supp. 662, 672 (S.D.N.Y.1985); Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., supra, 42 N.Y.2d at 545, 399 N.Y.S.2d 628, 369 N.E.2d 1162. It also has been defined as uniqueness or as having acquired a secondary meaning. Allied Maintenance, supra, 42 N.Y.2d at 545, 399 N.Y.S.2d 628, 369 N.E.2d 1162. A trademark has a secondary meaning if it "has become so associated in the mind of the public with that entity [Allied] or its product that it identifies the goods sold by that entity and distinguishes them from goods sold by others." Id. In sum, the statute protects a trademark's "selling power." Sally Gee, Inc. v. Myra Hogan, Inc., supra, 699 F.2d at 624-25. However, the fact that a mark has selling power in a limited geographical or commercial area does not endow it with a secondary meaning for the public generally. See Hartman v. Hallmark Cards, Inc., 833 F.2d 117, 121 (8th Cir.1987); Truck Equipment Service Co. v. Fruehauf Corp., 536 F.2d 1210, 1219 (8th Cir.), cert. denied, 429 U.S. 861, 97 S. Ct. 164, 50 L. Ed. 2d 139 (1976) (quoting Shoppers Fair of Arkansas, Inc. v. Sanders Co., 328 F.2d 496, 499 (8th Cir.1964)); Restaurant Lutece, Inc. v. Houbigant, Inc., 593 F. Supp. 588, 596 (D.N.J.1984); Scott v. Mego International, Inc., 519 F. Supp. 1118, 1138 (D.Minn.1981).

The strength and distinctiveness of LEXIS is limited to the market for its services--attorneys and accountants. Outside that market, LEXIS has very little selling power. Because only one percent of the general population associates LEXIS with the attributes of Mead's service, it cannot be said that LEXIS identifies that service to the general public and distinguishes it from others. Moreover, the bulk of Mead's advertising budget is devoted to reaching attorneys through professional journals.

This Court has defined dilution as either the blurring of a mark's product identification or the tarnishment of the affirmative associations a mark has come to convey. Sally Gee, Inc. v. Myra Hogan, Inc., supra, 699 F.2d at 625 (quoting 3A Callman, The Law of Unfair Competition, Trademarks and Monopolies Sec. 84.2 at 954-55). Mead does not claim that Toyota's use of LEXUS would tarnish affirmative associations engendered by LEXIS. The question that remains, therefore, is whether LEXIS is likely to be blurred by LEXUS.

Very little attention has been given to date to the distinction between the confusion necessary for a claim of infringement and the blurring necessary for a claim of dilution. Shire, supra, 77 Trademark Rep. at 293. Although the antidilution statute dispenses with the requirements of competition and confusion, it does not follow that every junior use of a similar mark will dilute the senior mark in the manner contemplated by the New York Legislature.

As already stated, the brief legislative history accompanying section 368-d described the purpose of the statute as preventing "the whittling away of an established trademark's selling power and value through its unauthorized use by others upon dissimilar products." The history disclosed a need for legislation to prevent such "hypothetical anomalies" as "Dupont shoes, Buick aspirin tablets, Schlitz varnish, Kodak pianos, Bulova gowns, and so forth", and cited cases involving similarly famous marks, e.g., Tiffany & Co. v. Tiffany Productions, Inc., 147 Misc. 679, 264 N.Y.S. 459 (1932), aff'd, 237 A.D. 801, 260 N.Y.S. 821, aff'd, 262 N.Y. 482, 188 N.E. 30 (1933); Philadelphia Storage Battery Co. v. Mindlin, 163 Misc. 52, 296 N.Y.S. 176 (1937). 1954 N.Y.Legis.Ann. 49-50.

It is apparent from these references that there must be some mental association between plaintiff's and defendant's marks.

[I]f a reasonable buyer is not at all likely to link the two uses of the trademark in his or her own mind, even subtly or subliminally, then there can be no dilution.... [D]ilution theory presumes some kind of mental association in the reasonable buyer's mind between the two party's [sic] uses of the mark.

2 J. McCarthy, supra, Sec. 24.13 at 213-14.

This mental association may be created where the plaintiff's mark is very famous and therefore has a distinctive quality for a significant percentage of the defendant's market. Sally Gee, Inc. v. Myra Hogan, Inc., supra, 699 F.2d at 625. However, if a mark circulates only in a limited market, it is unlikely to be associated generally with the mark for a dissimilar product circulating elsewhere. See, e.g., Estee Lauder, Inc. v. Cinnabar 2000 Haircutters, Inc., 218 U.S.P.Q. 191 (S.D.N.Y.), aff'd, 714 F.2d 112 (2d Cir.1982); Markel v. Scovill Mfg. Co., 471 F. Supp. 1244 (W.D.N.Y.), aff'd, 610 F.2d 807 (2d Cir.1979). As discussed above, such distinctiveness as LEXIS possesses is limited to the narrow market of attorneys and accountants. Moreover, the process which LEXIS represents is widely disparate from the product represented by LEXUS. For the general public, LEXIS has no distinctive quality that LEXUS will dilute.

The possibility that someday LEXUS may become a famous mark in the mind of the general public has little relevance in the instant dilution analysis since it is quite apparent that the general public associates nothing with LEXIS. On the other hand, the recognized sophistication of attorneys, the principal users of the service, has substantial relevance. See Sally Gee, Inc. v. Myra Hogan, Inc., supra, 699 F.2d at 626. Because of this knowledgeable sophistication, it is unlikely that, even in the market where Mead principally operates, there will be any significant amount of blurring between the LEXIS and LEXUS marks.

For all the foregoing reasons, we hold that Toyota did not violate section 368-d. We see no need therefore to discuss Toyota's remaining arguments for reversal.

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

I concur, but write separately because I disagree with the majority's conclusion that LEXIS is not a strong mark capable of dilution and that LEXIS and LEXUS differ significantly in pronunciation, and I have a different view of the factors that are necessary to a finding of dilution.

It has become talismanic in the New York courts and in this Circuit that a cause of action under section 368-d involves two elements: 1) an extremely strong mark--either because of the mark's distinctive quality or because it has acquired secondary meaning--and 2) a likelihood of dilution. See Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., 42 N.Y.2d 538, 369 N.E.2d 1162, 1165-66, 399 N.Y.S.2d 628, 632 (1977); Miss Universe, Inc. v. Patricelli, 753 F.2d 235, 238 (2d Cir.1985); Universal City Studios, Inc. v. Nintendo Co., 746 F.2d 112, 120 (2d Cir.1984); Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 625 (2d Cir.1983). No showing of competition between the parties or confusion about the source of products is required. See Allied, 42 N.Y.2d 538, 369 N.E.2d at 1165, 399 N.Y.S.2d at 632; Sally Gee, 699 F.2d at 624.

The first element of a dilution cause of action requires the plaintiff to establish that it possesses an extremely strong mark--one "which is 'truly of distinctive quality' or which has 'acquired a secondary meaning in the mind of the public.' " Sally Gee, 699 F.2d at 625 (quoting Allied, 42 N.Y.2d 538, 369 N.E.2d at 1166, 399 N.Y.S.2d at 633) (emphasis in original). A trademark has a distinctive quality if it is "distinctive, arbitrary, fanciful or coined" rather than "generic or descriptive." Allied, 42 N.Y.2d 538, 369 N.E.2d at 1166, 399 N.Y.S.2d at 632. Courts have analyzed distinctiveness for antidilution purposes in much the same way they assess the strength of the mark when evaluating likelihood of confusion. See McDonald's Corp. v. McBagel's, Inc., 649 F. Supp. 1268, 1281 (S.D.N.Y.1986); P.F. Cosmetique, S.A. v. Minnetonka Inc., 605 F. Supp. 662, 672 (S.D.N.Y.1985). "To establish secondary meaning it must be shown that through exclusive use and advertising by one entity, a name or mark has become so associated in the mind of the public with that entity or its product that it identifies the goods sold by that entity and distinguishes them from goods sold by others." Allied, 42 N.Y.2d 538, 369 N.E.2d at 1166, 399 N.Y.S.2d at 633.

The majority concludes that LEXIS is not a strong mark capable of dilution, noting that "the fact that a mark has selling power in a limited geographical or commercial area does not endow it with a secondary meaning for the public generally." Op. at 1030 (citations omitted). The majority adds:

The strength and distinctiveness of LEXIS is limited to the market for its services-attorneys and accountants. Outside that market, LEXIS has very little selling power. Because only one percent of the general population associates LEXIS with the attributes of Mead's service, it cannot be said that LEXIS identifies that service to the general public and distinguishes it from others. Moreover, the bulk of Mead's advertising is devoted to reaching attorneys through professional journals.

Id. at 1031.

This conclusion limits section 368-d's protection to nationally famous marks, because a strong mark capable of dilution is an element of a section 368-d cause of action and a plaintiff can lose on this ground alone. See, e.g., Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., 42 N.Y.2d 538, 369 N.E.2d 1162, 1166, 399 N.Y.S.2d 628, 632-33 (1977).

However, "[t]he interest protected by Sec. 368-d is ... the selling power that a distinctive mark or name with favorable associations has engendered for a product in the mind of the consuming public." Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 624 (2d Cir.1983) (emphasis added). The LEXIS mark has "selling power" among its consuming public-attorneys and accountants. Its lack of selling power among the general public--i.e., the nonconsuming public-should not deprive the company of section 368-d's protection against dilution. See Dreyfus Fund Inc. v. Royal Bank of Canada, 525 F. Supp. 1108, 1125 (S.D.N.Y.1981) ("The statute should not be read to deprive marks from protection against dilution in limited areas of use, since otherwise it would afford protection only to the most notorious of all marks."); Wedgwood Homes, Inc. v. Lund, 294 Or. 493, 659 P.2d 377, 381 (1983) ("We see no reason why marks of national renown should enjoy protection while local marks should not. A small local firm may expend efforts and money proportionately as great as those of a large firm in order to establish its mark's distinctive quality."). The renown of a senior mark is a factor a court should assess when evaluating the likelihood of dilution, not the strength of the mark.

Further, the district court as a matter of fact found LEXIS to be a very strong mark capable of dilution. Because imagination, thought, and perception are required to associate LEXIS with a computerized legal research service, the district court ruled that the mark is at least suggestive or arbitrary. See Dist.Ct.Op. at 10, 26. Moreover, the district court found the mark to possess secondary meaning because it "distinguishes the LEXIS service from other products and it is uniquely associated by the consuming public with the source of the product." Id. at 26. It added:

Mead has expended considerable time, effort, and money to promote the LEXIS mark. In 1987, Mead spent $3.8 million in advertising and promotion. This advertising has proven effective, judging from the LEXIS-related revenues, which exceeded $200 million in 1987. These revenues are even more impressive when viewed in light of their growth since 1984, when revenues totalled $400,000.

Id. at 27. These findings cannot be set aside as arbitrary or capricious and amply support a conclusion that LEXIS is a strong mark capable of dilution.

Several definitions of dilution exist. The legislative history defined the concept as "unlawful injury caused by the whittling away of an established trade-mark's selling power and value through its unauthorized use by others upon dissimilar products." 1954 N.Y.Leg.Ann. 49. The New York Court of Appeals offered the following definition:

The harm that section 398-d is designed to prevent is the gradual whittling away of a firm's distinctive trade-mark or name. It is not difficult to imagine the possible effect which the proliferation of various non-competitive businesses utilizing the name Tiffany's would have upon the public's association of the name Tiffany's solely with fine jewelry. The ultimate effect has been appropriately termed dilution.

Allied, 42 N.Y.2d 538, 369 N.E.2d at 1165-66, 399 N.Y.S.2d at 632. In Sally Gee, this Court refined these general definitions by describing dilution as "an act which 'threatens two separable but related components of advertising value. Junior uses may blur a mark's product identification or they may tarnish the affirmative associations a mark has come to convey.' " 699 F.2d at 625 (quoting 3 R. Callman, The Law of Unfair Competition, Trademarks, and Monopolies Sec. 84.2, at 954-55 (footnote omitted)).

By treating similarity of the marks as a separate element of a dilution cause of action and by evaluating the dilution claim without developing an analytical framework, the majority threatens to muddy the already murky waters of antidilution analysis. See, e.g., Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 625 (2d Cir.1983) (dilution "remains a somewhat nebulous concept"); Home Box Office v. Showtime/The Movie Channel, 665 F. Supp. 1079, 1087 (S.D.N.Y.) ("a claim for dilution differs markedly from a federal trademark claim and appears less well defined"), aff'd in part, vacated in part, 832 F.2d 1311 (2d Cir.1987); see also Shire, Dilution Versus Deception-Are State Antidilution Laws an Appropriate Alternative to the Law of Infringement?, 77 Trademark Rep. 273, 293 (1987) [hereinafter cited as "Shire"] (courts "have consistently failed to analyze how a specific defendant's use has diluted a mark or its selling power when there is no likelihood of confusion"); Greiwe, Antidilution Statutes: A New Attack on Comparative Advertising, 72 Trademark Rep. 178, 183-84 (1982) ("Analysis of the process of dilution has been largely superficial.").

Applying Sally Gee 's definition, the majority finds little likelihood of dilution because LEXIS circulates in a limited market, the products covered by the LEXIS and LEXUS marks differ substantially, and Mead's principal consumers-attorneys-are sophisticated. Although I agree with these findings, I believe the majority has failed adequately to define the likelihood of dilution concept.

Defining likelihood of dilution as "tarnishing" is helpful because that principle can be applied in practice. See, e.g., Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200 (2d Cir.1979) (plaintiff's distinctive uniform diluted by defendant's use of a similar uniform in an X-rated movie); Coca-Cola Co. v. Gemini Rising, Inc., 346 F. Supp. 1183 (E.D.N.Y.1972) (plaintiff's "Coca-Cola" mark diluted by defendant's use of similar lettering in printing "Cocaine" on poster). "Blurring," however, offers practitioners and courts only marginally more guidance than "likelihood of dilution."

There is much to be gained by defining a general concept like "blurring" more specifically. As in this instance, confusion in the doctrine has created problems for trademark attorneys advising their clients about adopting trademarks, for potential litigants assessing their chances of pursuing or defending against dilution claims, and for courts attempting to apply the statute. See Shire, 77 Trademark Rep. at 288. In the trademark infringement context, Judge Friendly defined a similarly broad standard-likelihood of confusion-by articulating a multi-factor balancing test that considers:

the strength of [plaintiff's] mark, the degree of similarity between the two marks, the proximity of the products, the likelihood that the prior owner will bridge the gap, actual confusion, and the reciprocal of defendant's good faith in adopting its own mark, the quality of defendant's product, and the sophistication of the buyers.

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). This test has provided practitioners and district courts a helpful framework for assessing likelihood of confusion. See, e.g., Lobo Enters., Inc. v. Tunnel, Inc., 693 F. Supp. 71, 76-79 (S.D.N.Y.1988); Toys R Us, Inc. v. Canarsie Kiddie Shop, Inc., 559 F. Supp. 1189, 1195-1200 (E.D.N.Y.1983).

Like likelihood of confusion, blurring sufficient to constitute dilution requires a case-by-case factual inquiry. A review of the anti-dilution cases in this Circuit indicates that courts have articulated the following factors in considering the likelihood of dilution caused by blurring:

2) similarity of the products covered by the marks

The application of these factors here requires reversal of the decision below, although on a basis that I believe differs from that stated by the majority.

Dilution is likely only where the junior mark is similar to the senior mark. See McDonald's Corp. v. McBagel's, Inc., 649 F. Supp. 1268, 1281 (S.D.N.Y.1986) ("The similar element that associates defendants' name with plaintiff's family of marks, the use of the 'Mc' prefix with the name of a generic food item, is immediately apparent. New York courts have not hesitated to find 'whittling down' of the identity of the trademark where slogans used by two parties bear such a facial similarity."); Toys R Us, Inc. v. Canarsie Kiddie Shop, Inc., 559 F. Supp. 1189, 1208 (E.D.N.Y.1983) ("although the two names are distinguishable, the identity of the 'R' Us suffixes ... compels me to conclude that the name Kids 'r' Us is likely to blur Toys 'R' Us' product identification").

Exact identity between the marks is not required, see Textron, Inc. v. Spi-Dell Watch & Jewelry Co., 283 F. Supp. 920 (S.D.N.Y.) (granting the manufacturer of "SPEIDEL" watch bands an injunction against defendant's use of trade names "SPI-DELL" and "SPEIDELL" to designate its watch bands), aff'd in part, rev'd in part on other grounds, 406 F.2d 544 (2d Cir.1968), but the more similar the marks the higher the likelihood of dilution. Under some circumstances, of course, lack of similarity alone may be enough to defeat a dilution claim. See, e.g., Universal City Studios, Inc. v. Nintendo Co., 746 F.2d 112 (2d Cir.1984) ("the names and characters in dispute ["King Kong" and "Donkey Kong"] are so different that no reasonable question of fact was raised on the issue of blurring"); Tetley, Inc. v. Topps Chewing Gum, Inc., 556 F. Supp. 785, 794 (E.D.N.Y.1983) ("the broad humor defendant ["Petley Flea Bags"] employs serves to prevent the type of blurring which might result from more subtle or insidious effort at humor at plaintiff's ["Tetley" tea bags] expense."); Warner Bros., Inc. v. American Broadcasting Cos., 530 F. Supp. 1187, 1198 (S.D.N.Y.1982) ("Plaintiff's ["Superman"] and defendants' ["The Greatest American Hero"] are so different, the variations upon some of plaintiff's phrases so distinct, and the use of some of plaintiffs' characters' names so contrasted with that of plaintiffs' works that plaintiffs cannot be said to be in danger of suffering any dilution or damage."), aff'd, 720 F.2d 231 (2d Cir.1983).

The district court's Lanham Act analysis found that LEXIS and LEXUS are "virtually identical" in "everyday spoken English," Dist.Ct.Op. at 12, but that the marks differ physically and will appear in different contexts. The majority reversed the district court's finding regarding the pronunciation of the two marks, arguing that persons with "careful and precise" diction--notably lawyers and television and radio announcers--would pronounce "lexis" and "lexus" differently. Not only do I conclude that the reversal is unwarranted, but with regret I cannot accept the premise on which it is based. I believe the district court correctly found that the marks are similar in pronunciation, but that "differences in context and physical appearance tend to reduce the similarity of the marks...." Dist.Ct.Op. at 13.

2. Similarity of the Products Covered by the Marks

Similarity of the products covered by the marks increases the likelihood of blurring. See, e.g., Toys R Us, 559 F. Supp. at 1208 ("the similarity of the products sold by the stores compels me to conclude that the name Kids 'r' Us is likely to blur Toys 'R' Us' product identification"). Although competition is not required to prevail on a dilution claim, "dilution protects a mark's selling power, [so] relief is available particularly in an area of normal expansion of the trademark owner's business." Id. at 1208. Accordingly, a plaintiff will have a more difficult time establishing blurring when the junior mark appears on a product substantially different from the plaintiff's. See, e.g., Dreyfus Fund Inc. v. Royal Bank of Canada, 525 F. Supp. 1108, 1125 (S.D.N.Y.1981). ("Dreyfus' lion marks are not so unique and arbitrary as to deserve protection in fields totally unrelated to Dreyfus' activities...."); see also Pattishall, Dawning Acceptance of the Dilution Rationale for Trademark-Trade Identity Protection, 74 Trademark Rep. 289, 299 (1984) ("No doubt, however, there can be differences between goods so substantial that the use on the diverse product is too remote to cause dilution of the prior mark's distinctive quality. One is reminded of Learned Hand's 'lipstick and steamshovels' example of no likelihood of confusion....").

I believe the district court correctly concluded that:

the two products at issue, automobiles and computer assisted legal research, are dissimilar. LEXUS is a consumer product, while LEXUS is primarily a commercial product. LEXUS will be sold through independent retail dealerships, while LEXIS is typically sold through its direct sales force. Mead contends that automobile companies, including General Motors, Hyundai, and Mitsubishi, are involved in the computer industry through subsidiaries. Nevertheless, at the moment Toyota is not involved in the computer industry, and has averred that it has no present intent to enter that field. The evidence, therefore, demonstrates that these two products are not similar.

Dist.Ct.Op. at 13-14.

Where the plaintiff's consumers are sophisticated, there is a reduced likelihood that a junior mark will blur the senior mark's selling power. See, e.g., Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 626 (2d Cir.1983) ("Sophisticated retailers and discerning consumers of women's apparel are unlikely to have blurred vision causing them to see 'Sally Gee' upon viewing a Sally Lee label."). The district court found that Mead's market for its LEXIS service contains sophisticated consumers--primarily lawyers and accountants. See Dist.Ct.Op. at 18.

The likelihood of blurring is increased where the junior user acts with predatory intent. Predatory intent involves more than mere knowledge of the senior mark--it requires a showing that the junior user adopted its mark hoping to benefit commercially from association with the senior mark. See, e.g., McDonald's Corp. v. McBagel's, Inc., 649 F. Supp. 1268, 1281 (S.D.N.Y.1986) ("Defendants' print advertisements for the 'Mc (stuffed) Bagel,' as well as the statement by defendants' attorney that McBagel's wished to capitalize on the press attention generated by the McDonald's policing action, establish defendants' attempt to take unfair advantage of plaintiff in the instant case."); Toys R Us, Inc. v. Canarsie Kiddie Shop, Inc., 559 F. Supp. 1189, 1209 (E.D.N.Y.1983) ("After evaluating [defendant's] testimony, I firmly believe that he knew of Toys "R" Us when he adopted the Kids 'r' Us name for his business and that he intended to take advantage of the uniqueness of plaintiff's mark."); Dreyfus Fund Inc. v. Royal Bank of Canada, 525 F. Supp. 1108, 1125 (S.D.N.Y.1981) ("no court has suggested that New York law would not protect the substantial dilution of a strong mark with intent to exploit its favorable impression and associations in the owner's general field of commerce").

The fact that the junior user may have acted in good faith in adopting its mark should not itself be sufficient to relieve it of liability if other factors support a finding of blurring. Predatory intent is a factor for courts to weigh when evaluating the likelihood of blurring, not an independent element of the cause of action under section 368-d. Compare Hoover Co. v. Citicorp Venture Capital Ltd., 674 F. Supp. 460, 462 (S.D.N.Y.1987) (noting that a dilution cause of action involves a strong mark and likelihood of dilution and that "[a] relevant but not conclusive factor is the absence of a predatory intent by the junior user of the mark") (emphasis added) with Lobo Enters., Inc. v. Tunnel, Inc., 693 F. Supp. 71, 79-80 (S.D.N.Y.1988) (holding that plaintiff's dilution claim would fail absent predatory intent, even assuming plaintiff could establish a strong mark and likelihood of dilution).

Although section 368-d makes no reference to the defendant's bad faith, there are sound reasons for assessing predatory intent when considering a likelihood of blurring. Section 368-d is of equitable origin, see Sally Gee, 699 F.2d at 625, and predatory intent is inherent in the statute's legislative history. See 1954 N.Y.Leg.Ann. 49 ("Naturally, the owners of established trade-marks may be expected to welcome protection of their good names against attempts by infringers, with infinite other words to choose from, to obtain a free ride....") (emphasis added). Moreover, the fact that the junior user believes it can benefit commercially from copying the senior user's trademark provides strong evidence of the likelihood of blurring. Cf. My-T Fine Corp. v. Samuels, 69 F.2d 76, 77 (2d Cir.1934) (L. Hand, J.) ("a late comer who deliberately copies the dress of his competitors already in the field raises a presumption that customers will be deceived."). Not surprisingly, numerous courts have recognized predatory intent as relevant to a finding of dilution. See, e.g., Sally Gee, 699 F.2d at 625; Essence Communications, Inc. v. Singh Indus., 703 F. Supp. 261, 270 (S.D.N.Y.1988); Lobo Enters., Inc. v. Tunnel, Inc., 693 F. Supp. 71, 79 (S.D.N.Y.1988); Transamerica Corp. v. Trans America Abstract Serv., Inc., 698 F. Supp. 1067, 1077 (E.D.N.Y.1988); Hoover Co. v. Citicorp. Venture Capital Ltd., 674 F. Supp. 460, 462 (S.D.N.Y.1987); Pan Amer. World Airways, Inc. v. Panamerican School of Travel, Inc., 648 F. Supp. 1026, 1039 (S.D.N.Y.), aff'd, 810 F.2d 1160 (2d Cir.1986); Vitabiotics, Ltd. v. Krupka, 606 F. Supp. 779, 785 (E.D.N.Y.1984); Lord Jeff Knitting Co. v. Warnaco, Inc., 594 F. Supp. 579, 582 (S.D.N.Y.1984).

The district court found that Toyota knew of the LEXIS mark and of Mead's antidilution objections when it adopted the LEXUS mark, see Dist.Ct.Op. at 16, 31, but that "Toyota's behavior falls far short of predatory intent." Id. at 31. In its Lanham Act analysis, the district court stated that "there is no evidence that the LEXIS name would benefit Toyota," Dist.Ct.Op. at 16, and observed that Toyota had offered evidence "that the association of LEXUS with the Mead service was a negative in focus group evaluations of various names." Id. at 17. In its dilution discussion, however, the district court noted: "Toyota contends that LEXUS connotes, among other things, high technology. To that extent at least, LEXUS benefits from the association with LEXIS." Id. at 31. The district court's unequivocal conclusion that Toyota acted without predatory intent, however, resolves any doubt these observations might raise. Id.

As set forth above, a plaintiff must possess an extremely strong mark capable of dilution to prevail under section 368-d. See Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 625 (2d Cir.1983); Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., 42 N.Y.2d 538, 369 N.E.2d 1162, 1165-66, 399 N.Y.S.2d 628, 632 (1977). A court evaluates a mark's strength with reference to the distinctive quality of the mark or its secondary meaning, not its "fame." Thus, a mark can be strong enough to warrant protection from dilution without being "famous" or nationally renowned. See Dreyfus Fund Inc. v. Royal Bank of Canada, 525 F. Supp. 1108, 1125 (S.D.N.Y.1981); Wedgwood Homes, Inc. v. Lund, 294 Or. 493, 659 P.2d 377, 381 (1983).

However, as the majority observes, the extent to which the senior mark is famous or nationally renowned affects the mental associations consumers make between the two marks, which in turn bears upon the likelihood of blurring. See Op. at 1031-1032. This fact is inherent in the examples the legislative history recited--"Dupont shoes, Buick aspirin tablets, Schlitz varnish, Kodak pianos, Bulova gowns, and so forth." 1954 N.Y.Leg.Ann. 49-50. Accordingly, a plaintiff possessing a nationally famous mark needs to focus less on other factors, such as similarity of the products or sophistication of the consumers, to establish blurring. See, e.g., Dreyfus Fund Inc. v. Royal Bank of Canada, 525 F. Supp. 1108, 1125 (S.D.N.Y.1981). ("Dreyfus' lion marks are not so unique and arbitrary as to deserve protection in fields totally unrelated to Dreyfus' activities....")

The district court found that LEXIS is a strong mark among its primary consumers--seventy-six percent of lawyers and twenty-six percent of accountants readily identified LEXIS with attributes of Mead's service. See Dist.Ct.Op. at 11. However, LEXIS is not a nationally famous mark on the scale of Dupont, Buick, Schlitz, Kodak, or Bulova--only two percent of the general public recognized the LEXIS mark as describing Mead's service. See id.

This case raises an issue that is likely to arise rarely in dilution law--the prospect that a junior mark may become so famous that it will overwhelm the senior mark. Dilution under this theory might occur where the senior user's advertising and marketing have established certain associations for its product among a particular consumer group, but the junior mark's subsequent renown causes the senior user's consumers to draw the associations identified with the junior user's mark. Here, for example, Toyota seeks to associate LEXUS with luxury and the carriage trade, which Mead fears may overwhelm LEXIS's association with indispensability and economy.

The district court found dilution based upon the anticipated renown of the LEXUS mark. Survey evidence revealed that seventy-two percent of the general public associated LEXIS either with Mead's service or with nothing at all. The fact that only two percent of the population recognized LEXIS as being associated with Mead's service indicates that the vast majority of the general public associated LEXIS with nothing at all. The district court noted:

The parties have stipulated that in the first nine months of 1989, Toyota expects to spend between $18 million and $23 million on media advertising. Awareness of the LEXUS mark is likely to spread far beyond its potential purchasers through television, radio, billboards and other print advertising.... In short, the LEXUS mark is itself likely to become the sort of "famous" or "celebrated" mark, which Toyota claims is entitled to protection under Sec. 368-d. The effect will surpass "whittling away"--it will dwarf the LEXIS mark.... A hypothetical will perhaps clarify this conclusion. Suppose the telephone surveys conducted in this case were conducted two or three years after the introduction of the LEXUS line of automobiles, with the concomitant blitz of advertising and media attention. The court doubts whether three-quarters of the general population would respond that the word LEXUS brought nothing to mind. Moreover, it is more than likely that, even among Mead customers, the word "lexis" might first bring to mind Toyota's car.

Dist.Ct.Op. at 30-31.

The district court found a likelihood of dilution, reasoning that Toyota's promotional campaign for its LEXUS automobile "will dwarf the LEXIS mark," Dist.Ct.Op. at 30, and that Toyota acted in bad faith--although without predatory intent--by launching its LEXUS line "without any regard for its effect on the LEXIS mark." Id. at 31. The district court's dilution analysis did not refer to the other factors discussed above, nor did it conduct a balancing test.

It is the rule in this Circuit that a district court's findings regarding each of the Polaroid factors are findings of fact to which the clearly erroneous standard applies, but that the balancing of those factors to assess the likelihood of confusion is a legal conclusion subject to de novo appellate review. See Hasbro, Inc. v. Lanard Toys, Ltd., 858 F.2d 70, 75-76 (2d Cir.1988); Plus Prods. v. Plus Discount Foods, Inc., 722 F.2d 999, 1004-05 (2d Cir.1983). Because I discern no legitimate distinction between the district court's role in assessing each of the Polaroid factors and its evaluation of those factors to determine the likelihood of confusion, I believe that the clearly erroneous standard should apply to both findings and that this Court should not employ Plus Product 's approach in its dilution analysis.

Either approach, however, would support a reversal in this case. Although the district court did not use a balancing test, its findings on the various factors demonstrate that it would have found little likelihood of dilution had it done so. LEXIS and LEXUS are pronounced the same, but the marks differ in physical appearance, they will appear in different contexts, and the products bearing the marks are dissimilar. Moreover, Mead's consumers are sophisticated, Toyota adopted the LEXUS mark without predatory intent, and LEXIS enjoys no national renown.

The only finding that supports a likelihood of dilution is the district court's conclusion that LEXUS eventually may become so famous that members of the general public who now associate LEXIS or LEXUS with nothing at all may associate the terms with Toyota's automobiles and that Mead's customers may think first of Toyota's car when they hear LEXIS. See Dist.Ct.Op. at 30-31. This analysis is problematic. First, section 368-d protects a mark's selling power among the consuming public. Allied Maintenance Corp. v. Allied Mechanical Trades, Inc., 42 N.Y.2d 538, 369 N.E.2d 1162, 1163-64, 399 N.Y.S.2d 628, 630 (1977); Sally Gee, Inc. v. Myra Hogan, Inc., 699 F.2d 621, 624-25 (2d Cir.1983). Because the LEXIS mark possesses selling power only among lawyers and accountants, it is irrelevant for dilution analysis that the general public may come to associate LEXIS or LEXUS with Toyota's automobile rather than nothing at all. Second, the district court offered no evidence for its speculation that LEXUS's fame may cause Mead customers to associate "lexis" with Toyota's cars. It seems equally plausible that no blurring will occur--because many lawyers and accountants use Mead's services regularly, their frequent association of LEXIS with those services will enable LEXIS's mark to withstand Toyota's advertising campaign.

Therefore, even if we accept the district court's finding regarding the renown of the LEXUS mark, however, reversal still is required. The differences in the marks and in the products covered by the marks, the sophistication of Mead's consumers, the absence of predatory intent, and the limited renown of the LEXIS mark all indicate that blurring is unlikely.

For the reasons set forth above, I concur.