VORNADO AIR CIRCULATION SYSTEMS, INC., a Kansas corporation, Plaintiff–Appellee, v. DURACRAFT CORPORATION, Defendant–Appellant.
This case presents an issue of first impression in our circuit concerning the intersection of the Patent Act and the Lanham Trade–Mark Act. We must decide whether a product configuration is entitled to trade dress protection when it is or has been a significant inventive component of an invention covered by a utility patent.
After expiration of any patents or copyrights on an invention, that invention normally passes into the public domain and can be freely copied by anyone. The district court found, however, that because the spiral structure of the household fan grill in question is “nonfunctional,” a status largely determined by the availability of enough alternative grill designs so that other fan manufacturers can effectively compete without it, the grill can serve as trade dress.1 The court held that the grill could be protected under Lanham Act section 43(a) against copying by competitors, because that copying was likely to confuse consumers.
The court's injunction effectively prevents defendant Duracraft Corp. from ever practicing the full invention embodied in the patented fans of plaintiff Vornado Air Circulation Systems, Inc., after Vornado's utility patents expire.2 For the reasons discussed below, we find this result to be untenable. We hold that although a product configuration must be nonfunctional in order to be protected as trade dress under section 43(a), not every nonfunctional configuration is eligible for that protection. Where a product configuration is a significant inventive component of an invention covered by a utility patent, so that without it the invention cannot fairly be said to be the same invention, patent policy dictates that it enter into the public domain when the utility patents on the fans expire. To ensure that result, it cannot receive trade dress protection under section 43(a). The district court's order is reversed.
History of the spiral grill designs
The product configurations at issue in this case are two household fan grills with spiral—or arcuate—vanes, produced by the plaintiff, Vornado, and the defendant, Duracraft.
The idea of using a spiral grill on a fan is not new. An arcuate vane structure for propellers “applicable to ventilators and the like” was reflected in expired U.S. Patent No. 1,062,258, a utility patent issued May 20, 1913, to G.A. Schlotter, and arcuate vanes were incorporated into a household fan guard as early as 1936, as shown by expired U.S. Patent No. 2,110,994, a utility patent issued to J.H. Cohen.
Vornado began selling its fans with spiral grills in November 1988, at a time when it was the only fan company using that type of grill. On January 9, 1989, Vornado's founders, Donald J. Moore and Michael C. Coup, applied for a utility patent on their ducted fan with a spiral grill. They asserted, among other things, that their spiral grill produced an optimum air flow, although their own tests had shown that it performed about the same as the more common straight radial grill, and later tests suggested that some other grills worked better in some respects.
Their patent application claimed a fan with multiple features, including the spiral grill. The inventive aspect of Vornado's spiral grill was that the point of maximum lateral spacing between the curved vanes was moved inboard from the grill's outer radius, so that it was at the impeller blade's point of maximum power. Vornado emphasizes that its fan grill was not patentable by itself because a spiral grill per se was already in the public domain as “prior art,” a patent law term for what was already known from previous patents or other sources.
On May 22, 1990, Messrs. Moore and Coup were issued a utility patent. They subsequently applied for and on February 22, 1994, were granted a reissue patent expanding their claims, including those that involved the arcuate-shaped grill vane structure.
Vornado advertised its grill as the “Patented AirTensity TM Grill,” although the company had no separate patent on the grill. Between January 1989 and August 1990, Vornado sold about 135,000 fans. In its advertising, the company touted the grill as a “true achievement in aerodynamic efficiency,” “the result of determinant ergonomic design,” with “[u]nique AirTensity TM vortex action,” accomplishing “a high degree of safety and functionality.” See Appellant's Br. at 8–10; Appellant's App. at 1831, 1977.
In August 1990, Duracraft began offering an inexpensive electric household fan called the Model DT–7 “Turbo Fan.” The grill on Duracraft's Turbo Fan incorporated a spiral vane structure that was copied from Vornado's considerably more expensive fan models but was purposely designed not to infringe Vornado's patent. Apart from its look-alike grill and some aspects of the fan blade design, the Turbo Fan differed significantly from Vornado's fans in its overall configuration, its base and duct structure, its center knob, neon colors, packaging, labeling, and price. See Appellant's App. at 1190 (Mem. & Order). The box in which the Turbo Fan came had a circle cut out of the front so that the grill design showed through and was emphasized when the fan was displayed in its box.
By November 1992, Duracraft had sold nearly one million Turbo Fans in the United States. The Turbo Fan was the company's second-largest-selling household fan product.
District court's findings
Vornado sued, alleging that Duracraft had intentionally copied Vornado's grill design, but both sides agreed that the Turbo Fan did not infringe Vornado's patents. Vornado argued during the bench trial that the curved vanes in the “Patented AirTensity Grill” were legally nonfunctional, which they had to be in order to be protected as trade dress under section 43(a).
The district court found that the spiral grill was functional in a lay sense but not in a legal sense, based on our definition of trade dress functionality in terms of the competitive need to use a feature. See Brunswick Corp. v. Spinit Reel Co., 832 F.2d 513, 519 (10th Cir.1987).3 The district court found that Vornado's grill did in fact perform a unique function in the way that it shaped the flow of air coming from the fan, but the difference in air flow produced by it as compared with other grill designs was not great enough for a customer to perceive, so it made no competitive difference. The court also noted that other feasible grill structures could easily do as well on other relevant performance tests, and the spiral grill was not shown to be cheaper to manufacture.4
The district court did not find enough evidence to support a finding of aesthetic functionality, a type of functionality based on decorativeness or attractiveness, which we have previously recognized. Brunswick, 832 F.2d at 519. Nor did the district court find that Duracraft would suffer a marketing disadvantage if it could not use the spiral grill.5 The court found that the grill's value lay not in its operational attributes but primarily in its appearance, which the court said suggests something about the fan's performance and creatively suggests Vornado's identity.6
The court found that the grill design was nonfunctional and held that trade dress protection of nonfunctional product configurations under the Lanham Act was not incompatible with patent law. The court further found that the grill design was a suggestive symbol combined with a device, and thus inherently distinctive, so that no showing of secondary meaning was required.7 The court found that consumers were likely to be confused by Duracraft's use of a similar grill, and granted Vornado an injunction but no damages on the section 43(a) claim.8
Duracraft contends on appeal that the district court committed legal error in: 1) concluding that Vornado's trade dress claim was not barred by federal patent law; 2) rejecting Duracraft's statutory estoppel argument; 3) concluding that the spiral grill could be a “symbol or device” within the meaning of section 43(a); 4) enjoining the use of the spiral grill without proof of secondary meaning in any relevant market; 5) concluding that Duracraft's continued use of the spiral grill would be likely to confuse an appreciable number of persons; and in 6) issuing a nationwide injunction against continued sales of the Turbo Fan, regardless of how it might be packaged or labeled.
We have considered all of the issues raised by Duracraft, but, finding the patent law argument to be dispositive, we do not decide the other questions.
Protection of trade dress under section 43(a)
Section 43(a) of the Lanham Act, 15 U.S.C. § 1125(a), provides a federal cause of action for unprivileged imitation, including trade dress infringement. Hartford House, Ltd. v. Hallmark Cards, Inc., 846 F.2d 1268, 1271 (10th Cir.), cert. denied, 488 U.S. 908, 109 S.Ct. 260, 102 L.Ed.2d 248 (1988). Trade dress features are those comprising a product's look or image. Brunswick, 832 F.2d at 517. In Hartford and Brunswick, we held that although in the past, trade dress infringement had consisted of copying a product's packaging, “trade dress” modernly could also refer to the appearance of the product itself. As those two cases demonstrate, trade dress analysis may be applied to a single feature or a combination of features.9
A plaintiff in a trade dress infringement case must make two showings. First, the plaintiff must show either (a) that its product's trade dress features (or feature) are inherently distinctive because their intrinsic nature is such as to “almost automatically tell a customer that they refer to a brand,” or (b) that the trade dress has become distinctive through acquisition of secondary meaning, so that its primary significance in the minds of potential consumers is no longer as an indicator of something about the product itself but as an indicator of its source or brand. See Qualitex, 514 U.S. at ––––, 115 S.Ct. at 1303; Two Pesos, 505 U.S. at ––––, 112 S.Ct. at 2757–60; Kellogg, 305 U.S. at 118, 59 S.Ct. at 113. Second, a plaintiff must show that potential customers are likely to be confused by the defendant's trade dress into thinking that the defendant is affiliated, connected or associated with the plaintiff or that the defendant's goods originated with, or are sponsored or approved by the plaintiff. See 15 U.S.C. § 1125(a).
The producer of an allegedly infringing product may defend 10 by showing that what the plaintiff is claiming as its trade dress is functional, and therefore that all competitors must be permitted to copy it in their own products, regardless of any producer-identifying capacity it may possess. See Brunswick, 832 F.2d at 517, 520.
The conflict before us
Subsequent to Hartford and Brunswick, the Supreme Court ruled in Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 109 S.Ct. 971, 103 L.Ed.2d 118 (1989), clarifying that patent law creates a federal right to copy and use product features that are in the public domain, whether under an expired patent or for lack of patentability in the first place. Id. at 165, 109 S.Ct. at 985; see also Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 229–33, 84 S.Ct. 784, 787–89, 11 L.Ed.2d 661 (1964); Compco Corp. v. Day–Brite Lighting, Inc., 376 U.S. 234, 237–39, 84 S.Ct. 779, 781–83, 11 L.Ed.2d 669 (1964). Duracraft argues that Bonito Boats means that useful product features, which comprise utility patent subject matter, may not be protected as trademarks or trade dress and thereby be permanently monopolized by a single producer.
Vornado replies that there is no problem or inconsistency in its ability to obtain both patent protection for its fan in toto and trade dress protection for its spiral grill. Vornado argues 1) that of the main Supreme Court cases on which Duracraft relies, Sears, Compco, and Bonito Boats are all distinguishable, and Kellogg predates the Lanham Act; 2) that the Supreme Court and Congress both have said section 43(a) applies to product shapes; and 3) that the functionality doctrine properly reconciles the Patent Act with the Lanham Act, for if a product feature is not necessary to competition, no patent law purpose is served by allowing it to be copied.
We find each of these arguments wanting. At the same time, we need not rule as broadly as Duracraft would have us do either. We need not deal with whether every useful or potentially patentable product configuration is excluded from trade dress protection.11 Vornado does not argue that its grill was not a significant inventive component of its patented fans.12 Without that particular grill, the Vornado fan would not be the same invention that it is. We focus, therefore, on the law with regard to product configurations that are patented inventions or significant components thereof, and whether these product configurations can serve as trade dress.
The statute itself
The wording of section 43(a) does not help us answer the question before us. Neither the latest version of the statute nor the wording prior to the 1988 revision makes any mention of product shapes or configurations.13 We cannot assume by the absence of such reference, however, that section 43(a) was not meant to apply to them, because most of the law of section 43(a) is not reflected in the statute. Justice Stevens has described in great detail the way in which the courts, through “judicial legislation,” have expanded the scope of section 43(a) far beyond its original wording as a ban on false designations of geographic origin, creating at least a partial federal law of unfair competition. See Two Pesos, 505 U.S. at –––– – ––––, 112 S.Ct. at 2761–66 (Stevens, J., concurring in the judgment); see also J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition §§ 1.09, 27.03[b] (3d ed.1994).
Supreme Court precedents
When asked to balance the concerns of patent law against those of unfair competition law with respect to the copying of product shapes, the Supreme Court has ruled repeatedly over the years that the right to copy must prevail. See Bonito Boats, 489 U.S. at 167–68, 109 S.Ct. at 986; Sears, 376 U.S. at 232–33, 84 S.Ct. at 789; Compco, 376 U.S. at 235, 238, 84 S.Ct. at 780–81, 782; Kellogg, 305 U.S. at 119–22, 59 S.Ct. at 113–15; Singer Mfg. Co. v. June Mfg. Co., 163 U.S. 169, 185, 16 S.Ct. 1002, 1008, 41 L.Ed. 118 (1896).
Applying the common law of unfair competition, the Court held in Kellogg that it was not unfair competition for Kellogg Co. to copy National Biscuit Co.'s pillow-shaped shredded wheat cereal after invalidation of the design patent for the cereal shape and expiration of the utility patents for the machines to make it, where Kellogg had made reasonable efforts to distinguish its product by using a different carton, label, company name, and biscuit size. See Kellogg, 305 U.S. at 119–22, 59 S.Ct. at 113–15. In Singer, the Court reached the same conclusion regarding the defendant's copying of Singer sewing machines after their patents had expired. See Singer, 163 U.S. at 185–202, 16 S.Ct. at 1008–15.
Sears and Compco pitted patent law's public domain principles against state unfair competition statutes with respect to product copying where patents had been invalidated. In both cases, the Court again held that patent law's public domain concept must prevail over unfair competition concerns about consumer confusion where those concerns arose solely from the product copying. See Sears, 376 U.S. at 232, 84 S.Ct. at 789; Compco, 376 U.S. at 235, 238, 84 S.Ct. at 780–81, 782.
In 1989, in Bonito Boats, the Court yet again addressed the copying of product shapes, and, because of the same patent law public domain concerns, struck down another state statute, this one prohibiting the use of the direct-molding process to copy unpatented boat hulls. Bonito Boats, 489 U.S. at 167–68, 109 S.Ct. at 986.
Vornado would have us ignore these holdings, and we must acknowledge that distinguishing Sears, Compco, and/or Bonito Boats has become a veritable jurisprudential art form in recent years, in which we as well as many other courts have engaged. See Brunswick, 832 F.2d at 526 n. 7; McCarthy, supra, §§ 7.24–7.25 (discussing cases); Ralph S. Brown, Design Protection: An Overview, 34 UCLA L.Rev. 1341, 1360–62 (1987); Jay Dratler, Jr., Trademark Protection for Industrial Designs, 1988 U.Ill.L.Rev. 887, 916–24 (discussing courts' use of section 43(a) to make “an end run around Sears/Compco”). It may come as no surprise that we are able to identify distinctions between those cases and the one before us, the most salient, of course, being that in none of the prior cases did the Court apply federal unfair competition law under section 43(a) of the Lanham Act.14
To say we find distinctions is not the end of the story, however. Although we may not be strictly bound by the Court's holdings in these cases, we find it impossible to ignore the clear and continuing trend they collectively manifest in favor of the public's right to copy. We turn, then, to the legislative history to see whether Congress has reversed this direction.15
Ambiguous legislative history
Congress acknowledged and, to some extent, ratified the judicial expansion of section 43(a) in 1988, when it enacted its first comprehensive revision of the Lanham Act. Among other things, Congress broadened the wording of section 43(a), but, as noted above, that wording itself does not help us.
The Senate report accompanying the bill explained that the section was reworded “to codify the interpretation it has been given by the courts. Because Section 43(a) of the Act fills an important gap in federal unfair competition law, the committee expects the courts to continue to interpret the section.” S.Rep. No. 515, 100th Cong., 2d Sess. 40 (1988), reprinted in 1988 U.S.C.C.A.N. 5577, 5603.
The district court relied heavily on the Senate report in concluding that Congress intended to grant trademark and trade dress protection to nonfunctional product configurations. That report observed that the courts have applied section 43(a) to cases involving “infringement of unregistered marks, violations of trade dress and certain nonfunctional configurations of goods.” Id. (emphasis added). It also stated that in the revised definition of “trademark,” the words “symbol or device” were retained “so as not to preclude the registration of colors, shapes, sounds or configurations where they function as trademarks.” Id. at 44 (1988 U.S.C.C.A.N. at 5607) (emphasis added).
The district court read too much into this report. We cannot conclude that because at the time of the report, “certain nonfunctional configurations of goods” had received trade dress protection, Congress was saying that in the future, all such configurations should.16 As of 1988, when the Senate report made its observation, we are unaware of any legal precedents upholding protection for those nonfunctional configurations covered by utility patents. But there was case law denying protection to a closely related type of nonfunctional design.
In In re Shakespeare Co., 289 F.2d 506, 508, 48 CCPA 969 (1961), the Court of Customs and Patent Appeals denied trademark registration to a recognizable spiral marking on fishing rods that resulted from the applicant's patented method for making the rods. The court held, as a matter of policy, that even though the spiral mark itself was nonfunctional, protecting it as a trademark would impermissibly interfere with the patent law. Other competitors either would not be able to use the patented manufacturing process after expiration of Shakespeare's patents, or they would have to go to the trouble of removing the marking. Id.
The Senate report never even mentioned these patent law concerns, let alone purported to address or resolve them.
Functionality doctrine: An incomplete answer
To interpret the Senate report as broadly as Vornado would have us do, we would also have to ignore the strong possibility that if Congress thought at all about patent policies in 1988, it assumed that any product qualifying for a utility patent would automatically be functional. It would have been understandable for Congress to assume that a nonfunctionality requirement would eliminate any possible conflicts between the Lanham Act and the Patent Act—at least as to utility patents—given the repeated statements by various courts and commentators that functionality doctrine has precisely that purpose and effect. See Rogers, 778 F.2d at 337; Sylvania Elec. Prods., Inc. v. Dura Elec. Lamp Co., 247 F.2d 730, 732 (3d Cir.1957); McCarthy, supra, § 7.26 ; Dratler, supra, 1988 U.Ill.L.Rev. at 928, 938. But see Kohler, 12 F.3d at 646–50 (Cudahy, J., dissenting); Ferrari S.P.A. Esercizio Fabriche Automobili E Corse v. Roberts, 944 F.2d 1235, 1252–53 (6th Cir.1991), cert. denied, 505 U.S. 1219, 112 S.Ct. 3028, 120 L.Ed.2d 899 (1992) (Kennedy, J., dissenting). It even appears that at times the Supreme Court may have made the same assumption. See Qualitex, 514 U.S. at ––––, 115 S.Ct. at 1304; Bonito Boats, 489 U.S. at 166–67, 109 S.Ct. at 985–86; Inwood, 456 U.S. at 863, 102 S.Ct. at 2193 (White, J., concurring in result).
Despite what appears to be a widespread perception that product configurations covered by utility patents are automatically functional for Lanham Act purposes, the district court in our case ably demonstrated that this is not so. Configurations can simultaneously be patentably useful, novel, and nonobvious and also nonfunctional, in trade dress parlance.
This is the case because to meet patent law's usefulness requirement, a product need not be better than other alternatives or essential to competition. Stiftung v. Renishaw PLC, 945 F.2d 1173, 1180 (Fed.Cir.1991); see also Kohler, 12 F.3d at 649 (Cudahy, J., dissenting). To obtain a utility patent, an inventor need only show that an invention is 1) useful in the sense of serving some identified, beneficial purpose, and then—much more difficult to prove—that it is 2) novel, i.e., not previously known, and 3) nonobvious, or sufficiently inventive, in light of prior art. See 35 U.S.C. §§ 101–103; Donald S. Chisum, Patents §§ 3.01, 4.01, 5.01 (1994).
Functionality, by contrast, has been defined both by our circuit, and more recently by the Supreme Court, in terms of competitive need. See Qualitex, 514 U.S. at –––– – ––––, 115 S.Ct. at 1304–07; Hartford, 846 F.2d at 1272–74; Brunswick, 832 F.2d at 519; see also Restatement (Third) of Unfair Competition § 17 & cmts. a & b (1995). If competitors need to be able to use a particular configuration in order to make an equally competitive product, it is functional, but if they do not, it may be nonfunctional. The availability of equally satisfactory alternatives for a particular feature, and not its inherent usefulness, is often the fulcrum on which Lanham Act functionality analysis turns.
As some courts have explained the competitive need test, it conceivably could allow one producer to permanently appropriate any distinctive patented invention for exclusive trademark or trade dress use as soon as its patent expired and sufficient alternatives became available to make the invention no longer one of a few superior designs. See Clamp Mfg. Co. v. Enco Mfg. Co., 870 F.2d 512, 516–17 (9th Cir.), cert. denied, 493 U.S. 872, 110 S.Ct. 202, 107 L.Ed.2d 155 (1989); In re Bose Corp., 772 F.2d 866, 872 (Fed.Cir.1985); In re Morton–Norwich Prods., Inc., 671 F.2d 1332, 1339 (C.C.P.A.1982); In re Cabot Corp., 15 U.S.P.Q.2d 1224, 1228 (T.T.A.B.1990); A. Samuel Oddi, The Functions of ‘Functionality’ in Trademark Law, 22 Hous.L.Rev. 925, 942–43 (1985). But see McCarthy, supra, § 7.29 (“Functional patent protection and trademark protection are mutually exclusive.”).
Reconciling two federal statutes
Given that the functionality doctrine does not eliminate overlap between the Patent Act and the Lanham Act, we must decide whether Vornado is right that this doctrine nevertheless should be used to limit patent law's public domain.
Except to the extent that Congress has clearly indicated which of two statutes it wishes to prevail in the event of a conflict, we must interpret and apply them in a way that preserves the purposes of both and fosters harmony between them. See Digital Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, ––––, 114 S.Ct. 1992, 2002, 128 L.Ed.2d 842 (1994); Brotherhood of R.R. Trainmen v. Chicago River & Indiana R.R., 353 U.S. 30, 40–42, 77 S.Ct. 635, 640–41, 1 L.Ed.2d 622 (1957). Where, as here, both cannot apply, we look to their fundamental purposes to choose which one must give way.
Purposes of the Patent Act
First, patent law seeks to foster and reward invention; second, it promotes disclosure of inventions to stimulate further innovation and to permit the public to practice the invention once the patent expires; third, the stringent requirements for patent protection seek to assure that ideas in the public domain remain there for the free use of the public.
Aronson v. Quick Point Pencil Co., 440 U.S. 257, 262, 99 S.Ct. 1096, 1099, 59 L.Ed.2d 296 (1979). The “centerpiece of federal patent policy” is its “ultimate goal of public disclosure and use.” Bonito Boats, 489 U.S. at 157, 109 S.Ct. at 980; see also id. at 146, 151, 109 S.Ct. at 975, 977–78; Goldstein v. California, 412 U.S. 546, 569, 93 S.Ct. 2303, 2316, 37 L.Ed.2d 163 (1973); Scott Paper Co. v. Marcalus Mfg. Co., 326 U.S. 249, 255–56, 66 S.Ct. 101, 104–05, 90 L.Ed. 47 (1945); Singer, 163 U.S. at 185, 16 S.Ct. at 1008.
Vornado suggests that no patent law purpose is served by allowing copying of product configurations that are not necessary to competition. We cannot agree. We find no support in the Patent Act itself or its application for the proposition that the patent goals are limited to enhancing competition, at least in the direct sense. To the contrary, patents operate by temporarily reducing competition. They create monopolies to reward inventors who invent “ ‘things which are worth to the public the embarrassment of an exclusive patent.’ ” Graham v. John Deere Co., 383 U.S. 1, 9, 86 S.Ct. 684, 689, 15 L.Ed.2d 545 (1966) (quoting Thomas Jefferson, author of the 1793 Patent Act). Although competition ultimately may be enhanced by the increased product supply that results from operation of the patent law, the system's more obvious objective is to give the public the benefits of technological progress.
In this respect, it is significant that the framers of the patent system did not require an inventor to demonstrate an invention's superiority to existing products in order to qualify for a patent. That they did not do so tells us that the patent system seeks not only superior inventions but also a multiplicity of inventions. A variety of choices is more likely to satisfy the desires of a greater number of consumers than is a single set of products deemed “optimal” in some average sense by patent examiners and/or judges. And the ability to intermingle and extrapolate from many inventors' solutions to the same problem is more likely to lead to further technological advances than is a single, linear approach seeking to advance one “superior” line of research and development. We conclude that patent law seeks the invention and the passing into the public domain of even what trade dress law would consider nonfunctional inventions.
Allowing an inventor both patent and trade dress protection in a configuration would not necessarily inhibit invention directly. Quite the opposite, this double benefit would probably increase an inventor's direct incentives to pursue an idea. But the inventor's supply of ideas itself and freedom to experiment with them might diminish if the inventor had to do a competitive market analysis before adopting useful features from others' inventions once their patents expired. See Bonito Boats, 489 U.S. at 161–62, 109 S.Ct. at 982–83 (stating that federal patent scheme allows public to ascertain status of intellectual property embodied in a manufacture or design, and “[t]he public may rely upon the lack of notice in exploiting shapes and designs accessible to all”).17
As to the second patent law objective, encouraging public disclosure of inventions, it is not immediately apparent what effect, if any, the trade dress protection in question would have. But this case clearly shows that trade dress protection can directly interfere with the public's ability to practice patented inventions after the patents have expired, and that it undermines the principle that ideas in the public domain should stay there. We conclude that the inability freely to copy significant features of patented products after the patents expire impinges seriously upon the patent system's core goals, even when those features are not necessary to competition.
Purposes of section 43(a) of the Lanham Act
The core concepts of trademark protection are that consumers not be confused, misled, or deceived as to whose product they are buying, that sellers' goodwill—or investment in their reputation for quality—be protected, and that competition thereby be enhanced. See Park 'N Fly, Inc. v. Dollar Park and Fly, Inc., 469 U.S. 189, 198, 105 S.Ct. 658, 663–64, 83 L.Ed.2d 582 (1985). “[T]he protection of trademarks and trade dress under § 43(a) serves the same statutory purpose of preventing deception and unfair competition.” Two Pesos, 505 U.S. at ––––, 112 S.Ct. at 2760. Because trademarks promote competition and product quality, “Congress determined that ‘a sound public policy requires that trademarks should receive nationally the greatest protection that can be given them.’ ” Park 'N Fly, 469 U.S. at 193, 105 S.Ct. at 661 (quoting S.Rep. No. 1333, 79th Cong., 2d Sess. 6 (1946), reprinted in 1946 U.S.C.C.S. 1274, 1277)); see also 15 U.S.C. § 1127.
Clearly, any limitation on the use of product designs as protected trade dress will give trade dress less than “the greatest protection that can be given.” Park 'N Fly, 469 U.S. at 193, 105 S.Ct. at 661. But this statement by the Court quotes from the 1946 Senate report, and we cannot assume that Congress in 1946 intended to be as expansive in its protection of product configurations as in its protection of traditional word or picture trademarks. At that time, section 43(a) was not viewed as a broad federal unfair competition provision covering product configurations under the rubric of trade dress; that gloss was added later by judges. See Two Pesos, 505 U.S. at –––– – ––––, 112 S.Ct. at 2762–63 (Stevens, J., concurring in the judgment).
The degree to which a producer's goodwill will be harmed by the copying of product configurations correlates with the degree of consumer confusion as to source or sponsorship that is likely to result from the copying.18 We do not doubt that at least some consumers are likely to ignore product labels, names, and packaging and look only to the design of product features to tell one brand from another. These consumers are likely to be confused by similar product designs, and to the degree that this confusion is tolerated, the goals of the Lanham Act will be undermined.
But the Lanham Act, like common-law unfair competition law and most state unfair competition statutes, has never provided absolute protection against all consumer confusion as to source or sponsorship. For its first fifteen to twenty years, the act was not even applied to the shapes of products or their containers. See McCarthy, supra, §§ 1.09, 7.31, 27.03[b]. And even after it was, courts consistently denied protection against that degree of confusion caused by the copying of functional configurations.
As a practical matter, the fate of nonfunctional configurations within patented products has rarely been the subject of legal analysis, because courts often have found such designs to be functional. See, e.g., In re Bose Corp., 772 F.2d at 872–73; New England Butt Co. v. Int'l Trade Comm'n, 756 F.2d 874, 877–79 (Fed.Cir.1985); In re Hollaender Mfg. Co., 511 F.2d 1186, 1187–89 (C.C.P.A.1975); Best Lock Corp. v. Schlage Lock Co., 413 F.2d 1195, 1199, 56 CCPA 1472 (1969); In re Shenango Ceramics, Inc., 362 F.2d 287, 291–92, 53 CCPA 1268 (1966); In re Deister Concentrator Co., 289 F.2d 496, 503–05, 48 CCPA 952 (1961).
In two appellate decisions where courts have considered nonfunctional designs that were part of utility-patented products, we find an even split. In In re Shakespeare, 289 F.2d at 508, the Court of Customs and Patent Appeals upheld patent law principles and refused trademark registration. In Clamp, 870 F.2d at 516–17, the Ninth Circuit upheld the district court's finding of trademark status for the nonfunctional design of a “C” clamp. The Ninth Circuit appears to have assumed without discussion that all nonfunctional product configurations can be used as trademarks or trade dress—an assumption which, as noted above, we do not make.
We recognize also that consumer confusion resulting from the copying of product features is, in some measure, a self-fulfilling prophecy. To the degree that useful product configurations are protected as identifiers, consumers will come to rely on them for that purpose, but if copying is allowed, they will depend less on product shapes and more on labels and packaging.
We conclude that protecting against that degree of consumer confusion that may arise from the copying of configurations that are significant parts of patented inventions is, at best, a peripheral concern of section 43(a) of the Lanham Act.
Conclusion: Balancing competing policies
Given, then, that core patent principles will be significantly undermined if we do not allow the copying in question, and peripheral Lanham Act protections will be denied if we do, our answer seems clear. Much has been said in this and other section 43(a) cases about whether a second competitor needs to use a particular product design to compete effectively. But where Lanham Act goals are not the only ones at stake, we must also examine the degree to which a first competitor needs to use a useful product feature instead of something else—a name, a label, a package—to establish its brand identity in the first place.
It would defy logic to assume that there are not almost always many more ways to identify a product than there are ways to make it.19 See In re Water Gremlin Co., 635 F.2d 841, 844 (C.C.P.A.1980) (“[A] merchant who wishes to set himself apart has no dearth of means to do so.”). And if one of the ways to configure the product itself has been deemed important enough to the advance of technology for the government to grant a utility patent, we must find its value as a product feature to exceed its value as a brand identifier in all but the most unusual cases.
We hold that where a disputed product configuration is part of a claim in a utility patent, and the configuration is a described, significant inventive aspect of the invention, see 35 U.S.C. § 112, so that without it the invention could not fairly be said to be the same invention, patent law prevents its protection as trade dress, even if the configuration is nonfunctional.20
In future cases, the contribution of a particular configuration to the inventiveness of a patented product may not always be clear, and we do not wish to rule out the possibility that a court may appropriately conduct a factual inquiry to supplement its reading of the patent's claims and descriptions.
But in this case, we do not find it necessary to remand for such an inquiry. Vornado included the arcuate grill vane structure as an element of its patent claims and described the configuration as providing “an optimum air flow.” Then, after the first patent issued and Vornado subsequently found evidence that other grill structures worked as well as or better than the spiral grill, Vornado did not repudiate or disclaim in any way the grill element of its patent. Instead, Vornado sought and received a reissue patent that expanded its claims with respect to the grill.
Even if we discount entirely Vornado's extensive advertising campaign emphasizing the importance of the “AirTensity Grill,” this patent history on its face obviates any need for a remand on the question of inventive significance. We simply take Vornado at its word. Because the “Patented AirTensity Grill” is a significant inventive element of Vornado's patented fans, it cannot be protected as trade dress. The district court's order is REVERSED.
1. Under the Lanham Act, trademark owners can register their marks federally and thereby benefit from certain statutory advantages and presumptions. See, e.g., 15 U.S.C. §§ 1065, 1072, 1111. One does not register a product's trade dress—its overall look or image—but trade dress is protected under § 43(a) of the Lanham Act, 15 U.S.C. § 1125(a).
2. It may also interfere with Duracraft's ability to practice other inventors' earlier spiral-vane inventions covered by previous, expired utility patents. But we need not rest our holding on this further interference, because the conflict with the ability to practice Vornado's invention in the future is clear.
3. In Brunswick, we examined at length the requirement that a product feature's design be nonfunctional before it can qualify for trademark or trade dress protection. We found that the cone-shaped cover of a spin-cast fishing reel was nonfunctional, because competitors did not need to use it in order to make an equally competitive product. Our definition of functionality relied heavily on the existence of a sufficient number of alternatives, or substitutes, for the design in question. See id. at 518–20. The question of potential conflict with patent law was not before us in Brunswick, however, because the evidence concerning an expired utility patent had not been made part of the record on appeal. See id. at 521 n. 4.
4. Duracraft contends on appeal that the court either overlooked or misunderstood some of Duracraft's evidence regarding the grill's superior efficiency. Appellant's Br. at 37 & n. 9. We need not decide whether the court's factual finding was clearly erroneous, because we decide this case on a point of law.
5. The court did not discuss the evidence suggesting that, regardless of whether the Patented AirTensity Grill offered any real operational advantages, Vornado's advertising to that effect had succeeded in convincing at least some customers that it did. See Appellant's App. at 1854–72, 2676, 2678–81, 2683–85, 2701, 2703–05, 2707–09. If Vornado stimulated enough consumer demand to create a submarket for fans with spiral grills, Duracraft would be entitled to compete in that submarket, as well as in the broader fan market. See J.C. Penney Co. v. H.D. Lee Mercantile Co., 120 F.2d 949, 954 (8th Cir.1941) (“If ․ the public believes generally that a certain feature adds a utilitarian value to the goods—whether it actually does or not—and will be materially influenced to purchase them on that basis, over other competitive goods in the market, it will be held to be functional.”).
6. The name “Vornado” is a combination of the words “vortex” and “tornado,” and the court found that the spiral design evoked a mental picture of these two concepts.
7. Vornado insists in its brief that the district court made an alternative finding of secondary meaning, but what the court actually found was just an association between the grill design and Vornado, not that the primary significance of the design in the consumer's mind was as a brand identifier, rather than as a grill type. Some of our previous cases may have led Vornado to believe that mere association is sufficient. See, e.g., Marker Int'l v. DeBruler, 844 F.2d 763, 764 (10th Cir.1988) (stating that a mark has acquired secondary meaning if because of long association with a product or firm it has come to stand for that product or firm in the minds of the public). But the Supreme Court has made clear that the test is the more stringent “primary significance” standard. See Kellogg Co. v. National Biscuit Co., 305 U.S. 111, 118, 59 S.Ct. 109, 113, 83 L.Ed. 73 (1938) (holding that plaintiff company could not monopolize name “shredded wheat” because it had shown only an association in consumers' minds of the name with the company, not that the term's primary significance to the consuming public was the producer, and not the product). For other cases also citing the “primary significance” test, see Qualitex Co. v. Jacobson Prods. Co., 514 U.S. 159, ––––, 115 S.Ct. 1300, 1303, 131 L.Ed.2d 248 (1995); Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, –––– n. 4, 112 S.Ct. 2753, 2756 n. 4, 120 L.Ed.2d 615 (1992); Inwood Labs., Inc. v. Ives Labs., Inc., 456 U.S. 844, 851 n. 11, 102 S.Ct. 2182, 2187 n. 11, 72 L.Ed.2d 606 (1982).
8. Vornado did not appeal the court's treatment of its registered trademark and common law unfair competition claims.
9. In Brunswick, we found protected trade dress in the cone-shaped cover of a spin-cast fishing reel. See 832 F.2d at 517–20. In Hartford, the protected trade dress was a combination of features in a series of greeting cards. See 846 F.2d at 1270–76.
10. In its motion for emergency stay pending appeal and in its reply brief on appeal, Duracraft questioned whether the Supreme Court's language in Two Pesos, 505 U.S. at ––––, 112 S.Ct. at 2758, requires us to change our holding in Brunswick, 832 F.2d at 520, that the burden of proof regarding functionality is on the defendant in an unregistered trade dress infringement case. Duracraft failed to raise the burden of proof issue in its opening brief on appeal, and thus it has waived the argument, so we do not reach the question. State Farm Fire & Casualty Co. v. Mhoon, 31 F.3d 979, 984 n. 7 (10th Cir.1994) (citing Headrick v. Rockwell Int'l Corp., 24 F.3d 1272, 1277–78 (10th Cir.1994)); Abercrombie v. City of Catoosa, Okla., 896 F.2d 1228, 1231 (10th Cir.1990).
11. Nor do we need to take a position on whether utility patents should be viewed differently than design patents, as some courts have held. See Kohler Co. v. Moen Inc., 12 F.3d 632, 636–43 (7th Cir.1993); L.A. Gear, Inc. v. Thom McAn Shoe Co., 988 F.2d 1117, 1130–32 (Fed.Cir.), cert. denied, 510 U.S. 908, 114 S.Ct. 291, 126 L.Ed.2d 240 (1993); W.T. Rogers Co. v. Keene, 778 F.2d 334, 337 (7th Cir.1985); In re Honeywell, Inc., 497 F.2d 1344, 1348–49 (C.C.P.A.), cert. denied, 419 U.S. 1080, 95 S.Ct. 669, 42 L.Ed.2d 674 (1974).
12. Indeed, the fact that Vornado applied for and received a reissue patent broadening its claims with respect to the grill would belie any insignificance argument the company might make.The district court noted that neither party had asked the court to rule on the Vornado patents' validity, Appellant's App. at 1152 (Mem. & Order), and it did not do so. Our analysis applies the statutory presumption of patent validity to all of the extant and expired patents involved in this case. See 35 U.S.C. § 282.
13. Prior to revision, the relevant portion read:(a) Any person who shall affix, apply, or annex, or use in connection with any goods or services, or any container or containers for goods, a false designation of origin, or any false description or representation, including words or other symbols tending falsely to describe or represent the same, and shall cause such goods or services to enter into commerce, and any person who shall with knowledge of the falsity of such designation of origin or description or representation cause or procure the same to be transported or used in commerce or deliver the same to any carrier to be transported or used, shall be liable to a civil action by any person doing business in the locality falsely indicated as that of origin or in the region in which said locality is situated, or by any person who believes that he is or is likely to be damaged by the use of any such false description or representation.15 U.S.C.A. § 1125(a) (West 1982).After the revision, the section read as follows:(a)(1) Any person who, on or in connection with any goods or services, or any container for goods, uses in commerce any word, term, name, symbol, or device, or any combination thereof, or any false designation of origin, false or misleading description of fact, or false or misleading representation of fact, which—(A) is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person, or(B) in commercial advertising or promotion, misrepresents the nature, characteristics, qualities, or geographic origin of his or her or another person's goods, services, or commercial activities,shall be liable in a civil action by any person who believes that he or she is or is likely to be damaged by such act.15 U.S.C.A. § 1125(a) (West Supp.1995) (This citation also reflects paragraphing changes from a 1992 amendment, but that amendment made no substantive changes to the wording relevant here).
14. We also can distinguish Singer because it was decided at a time when the tort of unfair competition was generally perceived to require fraudulent passing off, and unfair competition law has since evolved and expanded, so that it may be violated without active deceit, wherever a product shape has acquired secondary meaning and a likelihood of consumer confusion is found. See McCarthy, supra, § 5.02. The force of this particular distinction is reduced, though, by the fact that both the Supreme Court and our own circuit have also preserved defendants' right to copy even where defendants were found to have deliberately palmed off their goods in an effort to deceive consumers. See Warner & Co. v. Eli Lilly & Co., 265 U.S. 526, 531–33, 44 S.Ct. 615, 617–18, 68 L.Ed. 1161 (1924); Midwest Plastics Corp. v. Protective Closures Co., 285 F.2d 747, 750 (10th Cir.1960) (quoting Reynolds & Reynolds Co. v. Norick, 114 F.2d 278, 281 (10th Cir.1940)).Kellogg, Sears, and Compco are distinguishable on the basis that neither secondary meaning nor inherent distinctiveness was found in those cases, whereas in our case the district court found inherent distinctiveness. And, finally, it is possible to distinguish Bonito Boats on the grounds that the statute struck down in that case offered broader protection to product shapes than does § 43(a). See Kohler, 12 F.3d at 641–42.
15. Duracraft bases its argument on both the Patent Clause, U.S. Const. art. I, § 8, cl. 8, and the Patent Act of 1952, codified as amended at Title 35 of the U.S.Code. We do not address the constitutional argument, however, unless this case cannot be resolved as a matter of statutory law. Jean v. Nelson, 472 U.S. 846, 854, 105 S.Ct. 2992, 2996–97, 86 L.Ed.2d 664 (1985).
16. Vornado similarly draws too much from the language of two Supreme Court cases it cites. It points to Two Pesos, 505 U.S. at ––––, n. 1, 112 S.Ct. at 2755 n. 1, where the Court adopted a definition of trade dress that included product shapes, and to Bonito Boats, 489 U.S. at 166, 109 S.Ct. at 985, where the Court noted that in § 43(a), “Congress has thus given federal recognition to many of the concerns that underlie the state tort of unfair competition, and the application of Sears and Compco to nonfunctional aspects of a product which have been shown to identify source must take account of competing federal policies in this regard.” Although both cases support the notion that § 43(a) will protect some nonfunctional configurations, neither requires protection for all such configurations. In determining whether the type of configuration before us should receive trade dress protection, we will follow the Court's instruction and “take account of competing federal policies in this regard.”
17. While the Principal Register of trademarks provides similar notice to the public for registered trademarks, 15 U.S.C. § 1072, trade dress is a much more nebulous concept, not subject to registration, and an inventor may have difficulty determining whether someone is claiming any given product shape or configuration as its protected trade dress.
18. This is so because unfair competition law protects only the goodwill a producer has attained by virtue of its reputation for quality, not any goodwill that inheres in the product itself. If people like a particular product regardless of who produces it, that type of “product goodwill” is something all competitors are entitled to appropriate, once any applicable patents expire. See Kellogg, 305 U.S. at 121, 59 S.Ct. at 114–15; Versa Prods. Co. v. Bifold Co. (Mfg.), 50 F.3d 189, 207 (3d Cir.1995).
19. Although a producer may find efficiencies in combining the brand-identifying function with a product's utilitarian function by using a useful product feature as a trademark or trade dress, we accord this type of efficiency little weight. Although the efficient combining of form and function is at the heart of good industrial design, promoting it is not a Lanham Act objective. For discussion of the repeated, unsuccessful attempts over the years to persuade Congress to pass a general industrial design protection bill, see Bonito Boats, 489 U.S. at 167–68, 109 S.Ct. at 986; Duraco Prods., Inc. v. Joy Plastic Enters., Ltd., 40 F.3d 1431, 1446 (3d. Cir.1994); Brown, supra, 34 UCLA L.Rev. at 1395–99; Dratler, supra, 1988 U.Ill.L.Rev. at 888, 904–05.
20. We note that our resolution of this issue will avoid any confusion as to the right to copy a patented invention which may arise by virtue of the fact that a feature can switch back and forth between being functional and nonfunctional with the vagaries of the marketplace. Compare In re Honeywell, Inc., 8 U.S.P.Q.2d 1600, 1604–05 (T.T.A.B.1988), with In re Honeywell, Inc., 532 F.2d 180, 182–83 (C.C.P.A.1976).It also avoids potential problem scenarios in which a commercially unimportant but patented configuration is deemed nonfunctional and registered as a trademark, and then later, because of a change in the direction of research and development, inventors wish to use the old technology taught by the expired patent but cannot, because of the trademark. The Fourth Circuit has held that once a trademark becomes incontestable, it may not be cancelled on functionality grounds. Shakespeare Co. v. Silstar Corp. of America, Inc., 9 F.3d 1091, 1099 (4th Cir.1993), cert. denied, 511 U.S. 1127, 114 S.Ct. 2134, 128 L.Ed.2d 864 (1994). But see William M. Landes & Richard A. Posner, The Economics of Trademark Law, 78 Trademark Rep. 267, 295–96 (1988).
STEPHEN H. ANDERSON, Circuit Judge.