The primary issue raised by this appeal is whether one of this nation's major oil companies engaged in deceptive advertising in violation of the Lanham Act, 15 U.S.C. § 1125(a) (1988). The parties to this appeal further call upon this court to interpret the degree to which commercial speech is protected by the First Amendment to the United States Constitution.

Commercial advertising plays a dynamic role in the complex financial and industrial activities of our society, leading author Norman Douglas to go so far as to observe that "[y]ou can tell the ideals of a nation by its advertisements."¹ Because honesty and fair play are prominent arrows in America's quiver of commercial and personal ideals, Congress enacted section 43(a) of the Lanham Act "to stop the kind of unfair competition that consists of lying about goods or services." U-Haul Int'l, Inc. v. Jartran, Inc., 681 F.2d 1159, 1162 (9th Cir.1982). Although "[c]omparative advertising, when truthful and nondeceptive, is a source of important information to consumers and assists them in making rational purchase decision," Triangle Publications v. Knight-Ridder Newspapers, 626 F.2d 1171, 1176 (5th Cir.1980), the consumer called upon to discern the true from the false requires a fair statement of what is true and false.

The plaintiff-appellee in this case, Castrol Inc. (Castrol), a major motor oil manufacturer and distributor of its products, sued in the United States District Court for the District of New Jersey, alleging that Pennzoil Company and Pennzoil Products Company (Pennzoil) advertised its motor oil in violation of section 43(a) of the Lanham Act2 when it claimed that its product "outperforms any leading motor oil against viscosity breakdown." Additionally, Castrol challenged Pennzoil's related secondary claim that Pennzoil's motor oil provides "longer engine life and better engine protection." After a bench trial on the merits, the district court held that Pennzoil's advertisements contained claims of superiority which were "literally false."

Consequently, the court permanently enjoined Pennzoil from "broadcasting, publishing, or disseminating, in any form or in any medium," the challenged advertisements or any "revised or reformulated versions" thereof. This injunction was superseded by a more narrowly tailored Amended Order and Final Judgment, prohibiting only "revised or reformulated false or deceptive versions of the commercials." The district court denied Castrol's request for money damages and attorney's fees and retained jurisdiction for the purpose of enforcing or modifying its judgment. Pennzoil immediately appealed on the grounds that its advertisements did not contain false claims and that the permanent injunction issued by the district court infringed on its right to free speech as protected by the First Amendment.3 We affirm.

The district court opinion thoroughly recites the material facts of this case, which are not in dispute. Castrol Inc. v. Pennzoil Co. & Pennzoil Prods. Co., 799 F. Supp. 424 (D.N.J.1992). We set forth only a distillation thereof essential to an understanding of our disposition of this appeal. Castrol's suit stems from a Pennzoil advertising campaign of its motor oil consisting of print and television commercials. These commercials feature either various members of national race car glitterati, or Arnold Palmer, a professional golf luminary of national repute, asserting that Pennzoil motor oil outperforms any leading motor oil against viscosity breakdown. The court found that the advertisements also implied that Pennzoil's products offered better protection against engine failure than any other leading motor oil.

Motor oils minimize metal-to-metal contact in an engine by providing an optimum protective film between moving parts. The oils are designed to provide sufficient resistance to flow to maintain the oil's thickness and protective film across the wide range of temperatures and stress generated by high speed motor engines. The measure of a motor oil's resistance to flow is called "viscosity." Ideally, viscosity should remain at an adequate level under all types of stress when the oil flows between engine parts in operation, but stress may cause a "shearing" effect which breaks down the viscosity of the oil during engine use. Viscosity breakdown has both a mechanical and chemical effect, the first causing a physical thinning of the lubricant and the latter the formation of sludge, varnish, and deposits on the engine. Because it is critical to motor oil performance, viscosity is the basis on which motor oils are classified and marketed by a grading system known as "SAE J300."

According to SAE J300, the viscosity of unused motor oils is measured by an industry-recognized laboratory test developed by the American Society for Testing and Materials (ASTM). The tests and specifications measure the breakdown in motor oil viscosity caused by the stress of shearing and high temperatures during engine use. The court found that the Committee of Common Market Automobile Constructors (CCMC) has established "the most demanding viscosity breakdown standards." Castrol, 799 F. Supp. at 430. CCMC is an association of major European automobile manufacturers, including Mercedes-Benz, BMW, and Jaguar. CCMC motor oil specifications contain two viscosity breakdown requirements: (1) the Shear Stability or Stay-in-Grade test, and (2) the High Temperature/High Shear (HTHS) test.

The Stay-in-Grade Test requires that the subject motor oil, after being sheared, maintain a minimum kinematic viscosity level in order to remain within its SAE J300 grade. The Stay-in-Grade standard is a "pass/fail" standard, and it does not rank motor oils within each grade. The district court found that all Castrol and Pennzoil passenger car motor oils pass the SAE J300 standards for their specified grades. Castrol, 799 F. Supp. at 429. Both parties to the litigation have stipulated that Pennzoil does not outperform Castrol against the Stay-in-Grade viscosity breakdown standard.

The HTHS test is a more rigorous test, which measures an oil's reduced viscosity during exposure to high temperatures and large shear forces generated by rapidly moving parts similar to the conditions in an operating engine. HTHS standards have been adopted by the Society of Automotive Engineers (SAE), the CCMC, General Motors, Chrysler, and Ford. By this standard's measure, Pennzoil did not outperform Castrol in any way; rather, it was Castrol motor oils which proved superior.4

Pennzoil, however, does not rely on the aforementioned tests to lend credence to its claims of superiority with respect to viscosity breakdown and protection from engine wear. Rather, Pennzoil claimed superiority on the basis of research it conducted utilizing the ASTM D-3945 Test, promulgated by the American Society of Testing and Materials.

Pursuant to this test, motor oil is passed through a diesel injector nozzle at a shear rate that causes a reduction in the kinematic viscosity of the fluid under test. The reduction in kinematic viscosity is reported as a percent loss of the initial kinematic viscosity. These tests showed that Pennzoil motor oil suffered less viscosity loss percentage than Castrol motor oil. Pennzoil contends that percent viscosity loss is one method of measuring viscosity breakdown and therefore asserts that this test substantiates its advertised superiority claims.

The district court, however, found that the ASTM-3945 Test was not a true measure of viscosity breakdown; it therefore relied upon the Stay-in-Grade and the HTHS tests. These tests, along with others conducted by Castrol, led the trial court to find that Pennzoil's claims of superiority for viscosity breakdown and engine protection were literally false. Pennzoil challenges these findings and argues also that the district court's injunction infringes upon Pennzoil's right to freedom of speech.

"We review the district court's conclusions of law in plenary fashion, its factual findings under a clearly erroneous standard, and its decision to [grant or] deny an injunction on an abuse of discretion standard." Sandoz Pharmaceuticals Corp. v. Richardson-Vicks, Inc., 902 F.2d 222, 226 (3d Cir.1990) (citations omitted).

First, Pennzoil asserts that, absent any evidence of consumer confusion in this case, Castrol failed to meet its burden of proving literal falsity by the standard set forth in Sandoz, supra. Likewise, a substantial portion of the dissent is spent urging that Sandoz requires that we consider consumer evidence even in a case where there has been a finding of literal falsity. However, this argument ignores crucial differences between the case sub judice and Sandoz. In Sandoz, we sustained the trial court's findings that the advertisements in question were not literally false and held, where the advertisements are not literally false, a plaintiff bears the burden of proving actual consumer deception. Sandoz, 902 F.2d at 228-29. The Sandoz trial court resorted to proof of consumer confusion only after finding that the challenged claims were not literally false. As that court stated:

Where the advertisements are not literally false, plaintiff bears the burden of proving actual deception by a preponderance of the evidence. Hence, it cannot obtain relief by arguing how consumers could react; it must show how consumers actually do react.

Id. (emphasis supplied) (citations omitted).

In the instant case, however, the trial court found that Pennzoil's advertisements were literally false. Sandoz definitively holds that a plaintiff must prove either literal falsity or consumer confusion, but not both. Sandoz, 902 F.2d at 227; see also McNeil-P.C.C., Inc. v. Bristol-Myers Squibb Co., 938 F.2d 1544, 1549 (2d Cir.1991) (Where the advertisement is shown to be literally false, the court may enjoin it without reference to its impact on the consumer.).

Thus, there are two different theories of recovery for false advertising under section 43(a) of the Lanham Act: "(1) an advertisement may be false on its face; or (2) the advertisement may be literally true, but given the merchandising context, it nevertheless is likely to mislead and confuse consumers." Johnson & Johnson v. GAC Int'l, Inc., 862 F.2d 975, 977 (2d Cir.1988) (Garth, J., sitting by designation).

When a merchandising statement or representation is literally or explicitly false, the court may grant relief without reference to the advertisement's impact on the buying public. When the challenged advertisement is implicitly rather than explicitly false, its tendency to violate the Lanham Act by misleading, confusing or deceiving should be tested by public reaction.

Coca-Cola Co. v. Tropicana Prod., Inc., 690 F.2d 312, 317 (2d Cir.1982) (citations omitted).

Therefore, because the district court properly found that claims in this case were literally false, it did not err in ignoring Pennzoil's superfluous evidence relating to the absence of consumer confusion. In addition, because the advertisement in Sandoz was not literally false, that case's references to consumer confusion, read in context, are completely consistent with the majority's disposition of this matter.

Second, Pennzoil argues that Castrol failed to sustain its burden of proving literal falsity because Castrol never offered affirmative proof to refute Pennzoil's claims, but merely cast doubt upon Pennzoil's research. See, e.g., Castrol, Inc. v. Quaker State Corp., 977 F.2d 57, 62 (2d Cir.1992) ("Where the defendant's advertisement claims that its product is superior plaintiff must affirmatively prove defendant's product equal or inferior.").

Yet Pennzoil's contention is meritless, as the trial record is replete with Castrol's affirmative evidence proving the literal falsity of Pennzoil's claims. For example, between October 25, 1991, and February 26, 1992, the Castrol Technical Center conducted the CCMC, International Lubricant Standardization and Approval Committee, and Chrysler Stay-in-Grade Tests. Both Pennzoil and Castrol met the Stay-in-Grade requirements, thus refuting Pennzoil's contention that it outperforms Castrol with respect to viscosity breakdown.

Between January 1, 1992, and March 25, 1992, the Castrol Technical Center conducted HTHS Tests, and all Castrol motor oils met the HTHS standard established by the CCMC, as well as all the other HTHS specifications. Pennzoil's 5W-30 and 10W-30 motor oils, however, failed to meet this standard, although other Pennzoil motor oils passed this test. Therefore, this test also did not substantiate Pennzoil's claims of superiority; on the contrary, it demonstrated Pennzoil motor oil's inferiority in some respects to Castrol motor oil. Thus, according to the only two industry accepted tests for measuring viscosity breakdown, Pennzoil's claims of superiority were literally false.

Castrol also presented expert testimony and field tests which affirmatively demonstrated that Pennzoil motor oil does not outperform Castrol motor oil with respect to viscosity breakdown. For example, the Southwest Research Institute conducted an automobile fleet test at Castrol's request. Researchers placed Pennzoil motor oil inside a group of three automobiles, each of a different model, and then placed Castrol motor oil inside three cars identical to the first set. The researchers then drove these automobiles through various tests and compared the viscosity breakdown of the two motor oils. According to this test, Castrol motor oil outperformed Pennzoil's product with respect to viscosity breakdown, therefore discrediting Pennzoil's claims.

Moreover, the court found that Pennzoil failed to adequately refute Castrol's affirmative evidence. For example, the fundament of Pennzoil's claim of superior protection against viscosity breakdown was the results of its ASTM D-3945 Test. Castrol's expert, Marvin F. Smith, Jr. (Smith), who was a member of the task force which developed the ASTM D-3945 Test method, testified that the ASTM D-3945 test renders inaccurate results with regard to viscosity breakdown.

Smith explained that the ASTM-3945 Test was developed for manufacturers to measure the quality of one batch of oil against the next batch of the same type of oil. This test was never intended to compare the viscosity breakdown of oils of different polymer classes, and the test cannot perform this function accurately. Pennzoil and Castrol are motor oils of different polymer classes, and thus this test's comparison of the two oils proves nothing relevant. Actually, the test does not measure viscosity breakdown at all; rather, it measures percentage of viscosity loss. As Trial Judge Wolin perceptively observed:

Pennzoil ignores the caveat embodied in the [ASTM D-3945] test as to the significance and use of ASTM D-3945. In § 4.2 it states "[T]hese test methods are not intended to predict viscosity loss in field service for different polymer classes or for different field equipment."

Castrol Inc. v. Pennzoil Co. & Pennzoil Prods. Co., 799 F. Supp. 424, 438 (D.N.J.1992). Thus, it cannot be gainsaid that Castrol presented affirmative evidence to prove the literal falsity of Pennzoil's claims and that Judge Wolin did not find Pennzoil's evidence to rebut Castrol's proof persuasive.

The dissent asserts, however, that a defendant need only establish a reasonable basis to support its claims to render the advertisement literally true. We disagree. Rather, the test for literal falsity is simpler; if a defendant's claim is untrue, it must be deemed literally false.

In this case, Pennzoil made a claim of superiority, and when tested, it proved false. Hence, under this standard, the district court correctly found literal falsity. Therefore, Castrol sustained its burden of proof, especially given this court's narrow scope of review:

If the district court's account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently. Where there are two permissible views of the evidence, the factfinder's choice between them cannot be clearly erroneous.

* * * * * *

[Moreover,] when a trial judge's finding is based on his decision to credit the testimony of one of two or more witnesses, each of whom has told a coherent and facially plausible story that is not contradicted by extrinsic evidence, that finding, if not internally inconsistent, can virtually never be clear error.

Anderson v. Bessemer City, 470 U.S. 564, 573-75, 105 S. Ct. 1504, 1511-12, 84 L. Ed. 2d 518 (1985).5

Pennzoil and the dissent assert that the district court declined to allow Pennzoil to rely on the ASTM-3945 Test to substantiate its claims solely because the test is not an industry standard. They argue, therefore, that the court erred because an advertised product's failure to meet industry standards does not render the advertisement literally false. See ConAgra, Inc. v. Geo. A. Hormel & Co., 784 F. Supp. 700 (D.Neb.1992).

We do not disagree that a test which is not an industry standard can yield accurate results. However, the district court did not enjoin Pennzoil's advertisements merely because the ASTM-3945 was not an industry standard test. It did so because it found that the ASTM-3945 Test did not measure viscosity breakdown at all. On the contrary, that test measured percent viscosity loss.6

First, Pennzoil asserts that its claims regarding superior engine protection constitute common marketplace "puffery," and thus do not violate the Lanham Act. Puffery is an exaggeration or overstatement expressed in broad, vague, and commendatory language.

Such sales talk, or puffing, as it is commonly called, is considered to be offered and understood as an expression of the seller's opinion only, which is to be discounted as such by the buyer.... The "puffing" rule amounts to a seller's privilege to lie his head off, so long as he says nothing specific.

W. Page Keeton, et al., Prosser and Keeton on the Law of Torts § 109, at 756-57 (5th ed. 1984). Puffery is distinguishable from misdescriptions or false representations of specific characteristics of a product. As such, it is not actionable. See Stiffel Co. v. Westwood Lighting Group, 658 F. Supp. 1103, 1115 (D.N.J.1987).

The predicate for Pennzoil's position--that its claims that its motor oil offered better protection against engine wear amounted to mere puffery--is that its advertisements included only general claims of superiority. See Nikkal Indus., Ltd. v. Salton, Inc., 735 F. Supp. 1227, 1234 n. 3 (S.D.N.Y.1990) (General claims that the product was "better" were mere puffery and not actionable as false advertising.); United States Healthcare, Inc. v. Blue Cross of Greater Phila., 898 F.2d 914, 926 (3d Cir.), (In the context of the advertising in the case, the defendant's claim that it was the better health care plan was an innocuous kind of puffery.), cert. denied, 498 U.S. 816, 111 S. Ct. 58, 112 L. Ed. 2d 33 (1990).7

Pennzoil's claim of engine protection by contrast involves more than a mere generality. Here, the claim is both specific and measurable by comparative research. In fact, Pennzoil seeks to substantiate its claims of superiority by reference to testing. This distinguishes the present case from those cited by Pennzoil and defeats Pennzoil's assertion that its claims constitute only puffery. See Gillette Co. v. Wilkinson Sword, Inc., No. 89 Civ. 3686 (KMW), slip op. at 42-43 (S.D.N.Y. Jan. 9, 1991) (performance claim which can be comparatively rated is not puffery); Toro Co. v. Textron, Inc., 499 F. Supp. 241, 249-53 & n. 23 (D.Del.1980) (claims concerning specific product attributes are not puffery.); In re Bristol-Myers Co., 102 F.T.C. 21, 321 (1983), (claims subject to measurement are not puffery). cert. denied, 469 U.S. 1189, 105 S. Ct. 960, 83 L. Ed. 2d 966 (1985)

Pennzoil's failure to specifically mention its competitors in the sentence promoting engine protection also does not render the statement puffery. First, the district court found that the statement compared Pennzoil to its major competitors by necessary implication. Pennzoil stated it is superior to the other brands in protecting against viscosity breakdown, noting that viscosity breakdown leads to engine problems. It left the consumer with the obvious conclusion that Pennzoil is superior to the other leading brands in protection against engine problems. Therefore, Pennzoil did, by implication, compare its effectiveness against engine wear to that of its competitors, and the court's finding did not constitute a rewriting of the commercials, as claimed by Pennzoil.8

Moreover, there need not be a direct comparison to a competitor for a statement to be actionable under the Lanham Act. See American Home Prods. Corp. v. Johnson & Johnson, 654 F. Supp. 568 (S.D.N.Y.1987) (holding that the claim that Tylenol gives unsurpassed relief is not puffery despite the lack of a direct comparison); Gillette, No. 89 Civ 3686 (KMW), slip op. at 42 (the claim "Smoothest, most comfortable shave" is not puffery). Under Pennzoil's logic, as Castrol points out, the Ford Motor Company could claim that its Pinto model offers the best rear end protection as long as no competitor is specifically named. Such a result is impractical and illogical.

Second, Pennzoil asserts that the district court erred in holding that the claims relating to better engine protection were literally false by necessary implication. However, in assessing whether an advertisement is literally false, a court must analyze the message conveyed in full context. Sandoz Pharmaceuticals Corp. v. Richardson-Vicks, Inc., 735 F. Supp. 597, 600 (D.Del.1989), aff'd, 902 F.2d 222 (3d Cir.1990); see also Gillette, No. 89 Civ. 3686 (KMW), slip op. at 9 ("[Defendant's] claim of 'six times smoother,' or any use of the word 'smoother' by [Defendant] in the context of its advertisements for shaving products, necessarily implies that the shave itself is smoother.") (emphasis in original); Cuisinarts, Inc. v. Robot-Coupe Int'l Corp., No. 81 CIV 731 (CSH), 1982 WL 121559, * 2 (S.D.N.Y. June 9, 1982) ("[I]n determining facial falsity the court must view the face of the statement in its entirety, rather than examining the eyes, nose, and mouth separately and in isolation from each other"); Tambrands, Inc. v. Warner-Lambert Co., 673 F. Supp. 1190, 1193-94 (S.D.N.Y.1987) (Representations made by "necessary implication" can constitute false advertising.).

For example, the following message appeared in the advertisement litigated in the Cuisinarts case:

When all 21 of the three-star restaurants in France's Michelin Guide choose the same professional model food processor, somebody knows the score--shouldn't you?

Cuisinarts, 1982 WL 121559, * 2. The court held this advertisement to be literally false, opining:

The basic point is that the ad states, by necessary implication, that Robot-Coupe and Cuisinarts both build professional model food processors, and that French restaurateurs, presented with two existing alternatives, chose the Robot-Coupe model over the Cuisinarts model by the score of 21 to 0. I appreciate that the ad does not make that statement in haec verba. That is not necessary.

Id.

In like fashion, Pennzoil's advertisements claim that viscosity breakdown leads to engine failure. Pennzoil then claims, albeit falsely, that it outperforms any leading motor oil against viscosity breakdown. The implication is that Pennzoil outperforms the other leading brands with respect to protecting against engine failure, because it outperforms them in protecting against viscosity breakdown, the cause of engine failure. Moreover, Pennzoil's Executive Vice President of Marketing, William Welcher, conceded in his deposition that this advertisement conveys the message that a consumer who selects a motor oil other than Pennzoil runs an enhanced risk of engine failure.

Nonetheless, the dissent charges the district court with substituting its judgment of the advertisement's meaning for that of the consumers and further argues that the court's reaction "is at best not determinative and at worst irrelevant." American Brands, Inc. v. R.J. Reynolds Tobacco Co., 413 F. Supp. 1352, 1357 (S.D.N.Y.1976); see also Sandoz, 902 F.2d at 229 ("A Lanham Act plaintiff cannot obtain relief by arguing how consumers could react; it must show how consumers actually do react."). The dissent dismisses the above cited cases which support the necessary implication analysis with the conclusory statement that "[t]hese cases clearly dealt with false representations of fact, not allegedly false implications of a possibly invalid, but scientifically accurate, test." Dissent at 958.

First, as we discuss, the ASTM-3945 Test is not a scientifically accurate test for the purpose for which Pennzoil attempts to utilize it; we repeat it simply does not measure viscosity breakdown.

Second, this case also deals with a representation of fact: whether one oil outperforms another. That is no different than whether one razor gives a smoother shave than its competitor, or any of the other cases cited in the majority opinion.

Despite the cases that we cite above in which courts interpret the meaning of an advertisement without reference to consumer evidence, the dissent baldly asserts that consumer evidence is necessary to interpret the meaning of an advertisement.

However, R.J. Reynolds and Sandoz involved advertisements which were found not to be literally false. See, e.g., R.J. Reynolds, 413 F. Supp. at 1357. In R.J. Reynolds, the court stated that it was asked to determine whether a statement acknowledged to be literally true and grammatically correct nevertheless had a tendency to mislead, confuse, or deceive.

Nor, in reaching that conclusion [of literal falsity], do I tread upon forbidden judicial territory. To use Judge Lasker's phrase ... "the court's reaction is at best not determinative and at worst irrelevant" only in cases where the issue is "whether a statement acknowledged to be literally true and grammatically correct nevertheless has a tendency to mislead, confuse or deceive." Consumer reaction is determinative where an advertisement is ambiguous and, hence, only potentially misleading. In the case at bar, the statement conveyed by the ad is unambiguous.

Cuisinarts, 1982 WL 121559, * 2 (citations omitted); see also McNeil, 938 F.2d at 1549 (where an advertisement is shown to be literally false, the court may enjoin it without reference to consumer confusion). As in Cuisinarts, the district court here found that Pennzoil's advertisements had an unambiguous meaning by necessary implication in their assertions that Pennzoil motor oil protects against engine failure better than its major competitors' motor oils.9

Moreover, Castrol presented copious affirmative evidence to prove that Pennzoil's claim to superior protection from engine failure is false. For example, pursuant to evidence introduced by Castrol, the district court found the testimony of Pennzoil's expert witnesses, Dr. Hault and Dr. Spikes, speculative and unconvincing. The trial court also found Pennzoil's reliance on the ASTM D-3945 Test to be misplaced. Qualified experts testified that percent viscosity loss, as measured by ASTM D-3945, is irrelevant to engine protection. It is the level of viscosity that the engine experiences in operation, not the percentage loss, that is significant to engine performance. Also, the ASTM D-3945 Test measures viscosity loss in terms of low shear viscosity, while the industry consensus teaches that high temperature and high shear conditions most authentically emulate actual engine conditions.

Also, Pennzoil's Vice President of Product Support, Donald M. Johnson, candidly conceded that car owners who utilize Castrol motor oil do not run an enhanced risk of engine failure. In addition, Pennzoil's director of research, James Newsom, testified that under normal operating conditions, consumers would not experience engine wear whether they chose Castrol or Pennzoil. Furthermore, under the industry established test for engine wear protection, both Castrol and Pennzoil met the highest performance requirements of the American Petroleum Institute.

Therefore, there was ample evidence adduced at trial to prove that Pennzoil's claims of superior protection against engine wear were literally false, and thus the district court's finding was not clearly erroneous. See Anderson, 470 U.S. 564, 573-75, 105 S. Ct. 1504, 1511-12 (the appellate court should not disturb the trial judge's findings regarding credibility of witnesses).

Pennzoil also expansively contends that "[i]t is beyond dispute that the district court's order constitutes a plain restraint on Pennzoil's constitutional right to commercial free speech" as recognized by this court in United States Healthcare, Inc. v. Blue Cross of Greater Phila., 898 F.2d 914, 937 (3d Cir.), cert. denied, 498 U.S. 816, 111 S. Ct. 58, 112 L. Ed. 2d 33 (1990). Pennzoil urges that if its motor oil improves in the future and thereby provides better protection from viscosity breakdown than does Castrol motor oil, Pennzoil would nonetheless be restrained from advertising this truism.

The Supreme Court has held that commercial speech is within the ambit of First Amendment protection. See, e.g., Friedman v. Rogers, 440 U.S. 1, 9, 99 S. Ct. 887, 894, 59 L. Ed. 2d 100 (1979); Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 763, 96 S. Ct. 1817, 1826, 48 L. Ed. 2d 346 (1976) (Society and consumers both have a strong interest "in the free flow of commercial information.").

Paragraph (c) of the permanent injunction granted by the district court prohibits Pennzoil from:

(c) broadcasting, publishing or disseminating, in any form or medium, any commercial or advertisement that claims, directly or by clear implication that:

(i) Pennzoil motor oil outperforms any leading motor oil against viscosity breakdown;

(ii) Pennzoil motor oil gives the most protection against viscosity breakdown of any leading motor oil;(iii) Pennzoil motor oil provides better protection against engine failure than any leading motor oil;

(iv) Pennzoil motor oil provides better protection against engine wear than any leading motor oil; or

(v) Pennzoil motor oil provides longer engine life or greater engine durability than any other leading motor oil.

Castrol Inc. v. Pennzoil Co. & Pennzoil Prods. Co., 799 F. Supp. 424, 441 (D.N.J.1992).

In essence, the district court has enjoined Pennzoil only from broadcasting, publishing, or disseminating the very statements which the court found to be literally false. Pennzoil argues that this is a prior restraint, in contravention of the First Amendment of the United States Constitution. At this moment, however, these claims are false, and it is well settled that false commercial speech is not protected by the First Amendment and may be banned entirely. Bates v. State Bar of Ariz., 433 U.S. 350, 383, 97 S. Ct. 2691, 2709, 53 L. Ed. 2d 810 (1977). Moreover, the prior restraint doctrine does not apply in this case because there has been "an adequate determination that [the expression] is unprotected by the First Amendment." Pittsburgh Press Co. v. Pittsburgh Comm'n on Human Relations, 413 U.S. 376, 390, 93 S. Ct. 2553, 2561, 37 L. Ed. 2d 669 (1973). The injunction is also not overbroad because it only reaches the specific claims that the district court found to be literally false. If, in the future, Pennzoil should improve its motor oil to surpass Castrol for viscosity breakdown, Pennzoil can at that time apply for a modification of the present injunction. See F.T.C. v. Brown & Williamson Tobacco Corp., 778 F.2d 35, 44 (D.C.Cir.1985).

The Second Circuit rejected a similar argument regarding a preliminary injunction, stating:

Any time a court issues a preliminary injunction there is some chance that, after the issuance of the order but prior to a full adjudication on the merits, changes in the operative facts will undercut the court's rationale. We will not, however, require the district court to draft a technical and narrow injunction to address the possibility of additional tests which are, at this time, purely hypothetical. If tests supporting its claim do come to light, [Appellant] may move to modify or dissolve the injunction.

Castrol, Inc. v. Quaker State Corp., 977 F.2d 57, 66 (2d Cir.1992); accord Flavor Corp. of Am. v. Kemin Indus., Inc., 503 F.2d 729, 732 (8th Cir.1974); 11 Charles A. Wright & Arthur R. Miller, Federal Practice and Procedure § 2961, at 604 (1973).

Although the dissent correctly points out that we should not supervise "all debate" in a given area, we have a statutory duty to supervise repetition of claims that have been found to be literally false.

We therefore hold that commercial speech that is false when uttered does not enjoy the protection of the First Amendment. The district court thus committed no error in enjoining it.10

In summary, the trial court, in a carefully written and exhaustive opinion, found Pennzoil's superiority claims to be literally false. These findings were based on Castrol's copious affirmative evidence, which included expert testimony, laboratory testing, and field testing. The evidence proved that Pennzoil motor oil does not suffer less viscosity breakdown than Castrol; on the contrary, in that respect, Castrol motor oil is superior. Therefore, ample evidence adduced at trial supported the district court's findings of fact.

Pennzoil's claim that the district court unconstitutionally restrained commercial free speech is also without merit. The only way to insure that Pennzoil will not duplicate its literally false claims in the context of a different advertisement is to enjoin the specific claims. The district court's retention of jurisdiction in the event its order needs modification facilitates adequate recourse if Pennzoil's claims of superiority should in the future become meritorious.

Accordingly, the judgment of the district court will be affirmed. Costs taxed against the appellants.

ROTH, Circuit Judge, dissenting.

Because I think the majority has articulated and applied a standard, for determining "literal falsity" in a "non-establishment" Lanham Act claim for false advertising, that is inconsistent both with this Court's prior rulings and with the statute's primary focus, I respectfully dissent from the conclusions that the majority reaches in Parts II and III.

At the outset, it is helpful to note that, in Lanham Act cases, this Court will "review the district court's conclusions of law in plenary fashion" and "its factual findings under a clearly erroneous standard." Sandoz Pharmaceuticals Corp. v. Richardson-Vicks, Inc., 902 F.2d 222, 225 (3d Cir.1990). Questions of the appropriate burdens of proof for parties in Lanham Act actions are questions of law to be decided under plenary review. See id. at 226-27; McNeil-P.C.C., Inc. v. Bristol-Myers Squibb Co., 938 F.2d 1544, 1548 (2d Cir.1991). Moreover, if the district court articulated incorrect legal criteria to guide its factual analysis, those legal criteria are subject to plenary review.

It is also important to examine exactly what the Lanham Act focuses upon and regulates. The relevant section of the Lanham Act, section 43(a), states:

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--

(1) 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

(2) 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. § 1125(a). This section of the Lanham Act "expressly establishes a private remedy for any violation thereunder." Sandoz, 902 F.2d at 227. "The Lanham Act is primarily intended to protect commercial interests. A competitor in a Lanham Act suit does not act as a vicarious avenger of the public's right to be protected against false advertising. Instead, the statute provides a private remedy to a commercial plaintiff who meets the burden of proving that its commercial interests have been harmed by a competitor's false advertising." Id. at 230 (citations omitted); see also Colligan v. Activities Club, 442 F.2d 686, 692 (2d Cir.1971) ("The Act's purpose, as defined in § 45, is exclusively to protect the interests of a purely commercial class against unscrupulous commercial conduct."), cert. denied, 404 U.S. 1004, 92 S. Ct. 559, 30 L. Ed. 2d 557 (1971); Joseph P. Bauer, A Federal Law of Unfair Competition: What Should be the Reach of Section 43(a) of the Lanham Act?, 31 UCLA L.Rev. 671, 681 (1984) ("[T]he Act leans toward the protection of owners of trademarks from misuse of those marks, rather than the protection of the public from deception.").

Because of the Lanham Act's primary focus upon protection of private parties from false advertising, actual harm to the party alleging a Lanham Act violation is a necessary element of a Lanham Act claim. "The plaintiff must ... show that defendant's misrepresentation is 'material, in that it is likely to influence the purchasing decision.' " U.S. Healthcare, Inc. v. Blue Cross of Greater Philadelphia, 898 F.2d 914, 922 (3d Cir.1990) (quoting Toro Co. v. Textron, Inc., 499 F. Supp. 241, 251 (D.Del.1980) (Stapelton, J.)); see also Sandoz, 902 F.2d at 231 (noting that a Lanham Act plaintiff must prove that "the defendant's promotions contained a material representation or description"). Generally, the preferred evidence of an advertisement's influence over purchasing decisions is actual consumer survey evidence. Even if an advertisement contains inadequately substantiated claims:

a [Lanham Act] plaintiff must produce consumer surveys or some surrogate therefor to prove whether consumers expect an advertising claim to be substantiated and whether they expect the level of substantiation to be greater than that which the defendant has performed. The effect of the advertisement is the critical determination, and it must be demonstrated by a Lanham Act plaintiff regardless of whether the claim is facially ambiguous.

The only exception to the burden to prove that an advertisement is misleading through consumer survey evidence is if the advertisement is "literally false." Sandoz, 902 F.2d at 228. This is so because a court may presume that a "literally false" advertisement will deceive the consuming public and, therefore, excuse the plaintiff from the time consuming and expensive task of conducting consumer surveys. "If advertisements are false on their face, a court may find a 'tendency to deceive' without looking to evidence of actual consumer response." Stiffel Co. v. Westwood Lighting Group, 658 F. Supp. 1103, 1111 (D.N.J.1987). However, because "[t]he effect of the advertisement on the consumer is the critical determination," Sandoz, 902 F.2d at 229, courts must be hesitant to presume that an advertisement will materially deceive consumers. See Lillian R. BeVier, Competitor Suits for False Advertising Under Section 43(a) of the Lanham Act: A Puzzle in the Law of Deception, 78 Va.L.Rev. 1, 29-30 (1992) (explaining that when courts

"strain to find claims literally false, rather than literally true but potentially misleading, thereby short-circuiting any inquiry into consumer injury[,] [t]he result is that, ex post, the decisions tend to permit recovery in cases where the causal chain from the allegedly offensive party of the ad to the supposedly deceived consumer is highly attenuated; in such cases the conclusion that consumer decisionmaking was significantly distorted by the false statement seems unwarranted").

In the present case, the district court, and now the majority, have concluded that Pennzoil's challenged advertisements are literally false. In so doing, the majority concludes that Castrol met its appropriate burden of proving Pennzoil's viscosity breakdown claims literally false, see majority opinion at 944-945, that the district court had not applied an incorrect "industry standards" criterion to evaluate Pennzoil's substantiation for its claims, see id. at 945, and that the district court correctly determined that Pennzoil had, by "necessary implication," made literally false claims of superior protection against engine wear. See id. at 946-948. I disagree with the majority's analysis and conclusions. I would find that the district court applied legally incorrect criteria to guide its factual determinations, and therefore, that its conclusions must be reversed and this case remanded for an appropriate analysis.

I. THE BURDEN OF PROOF FOR A "NON-ESTABLISHMENT" CLAIM OF LITERAL FALSITY

As the majority has implicitly recognized, see id. at 943-944, a different burden of proof faces a Lanham Act plaintiff who challenges an "establishment" advertising claim that states that "tests prove" the asserted proposition than faces a plaintiff who challenges a "non-establishment" claim that makes no statement that "tests prove" the asserted proposition. See, e.g., Castrol, Inc. v. Quaker State Corp., 977 F.2d 57, 62 (2d Cir.1992) ("[P]laintiff bears a different burden in proving literally false an advertised claim that tests prove defendant's product superior, than it does in proving the falsity of a superiority claim which makes no mention of tests."). Such a distinction is quite logical because establishment claims state to the consumer that they are based upon tests and, therefore, provide the consumer with the expectation that tests actually support the claim at issue.

In an establishment case, it may be logical to hold, as has the Court of Appeals for the Second Circuit, that to establish that an advertising claim is literally false, "the plaintiff must demonstrate that such tests are 'not sufficiently reliable to permit one to conclude with reasonable certainty that they established' the claim made." McNeil-P.C.C., 938 F.2d at 1549 (quoting Procter & Gamble Co. v. Chesebrough-Pond's Inc., 747 F.2d 114, 119 (2d Cir.1984)). However, the appropriate burden of proof for literal falsity in a non-establishment case must be different. The crucial and difficult question is what that burden should be, considering that this Court has never expressly defined it.

The majority's analysis of this question is confined to borrowing one piece of dictum from a recent decision of the Court of Appeals for the Second Circuit. The majority states that "[w]here the defendant's advertisement claims that its product is superior plaintiff must affirmatively prove defendant's product equal or inferior." Majority opinion at 943 (quoting Quaker State, 977 F.2d at 63). The majority then applies this standard to hold that, because Castrol cast doubt upon the test Pennzoil offered to support its claims and introduced other tests that purported to show that Castrol was equal to or superior to Pennzoil, Castrol had met its burden of proving Pennzoil's advertisements literally false. See majority opinion at 943-945. The majority did not find it important that the district court "ignor[ed]" a consumer survey that purported to demonstrate that Castrol could not be affected adversely by Pennzoil's advertisements. See id. I find that the majority's analysis has ignored relevant pronouncements of this Court and has obscured the purposes of § 43(a) of the Lanham Act.

There is a crucial fact involved in this case that was ignored by the district court and glossed over by the majority. The record in this case contains a consumer survey and expert testimony explaining that survey that purports to demonstrate that consumers interpreted Pennzoil's claims that its motor oil "outperforms any leading motor oil against viscosity breakdown" and provides "longer engine life and better engine protection" as involving or implying no comparison whatsoever between Pennzoil and Castrol, and in fact involving nothing more than a "dangling comparative" claim that consumers ignored. See App. at 1157-1244.2 The majority stated that the district court "did not err in ignoring [this] ... superfluous evidence." Majority opinion at 943. The majority apparently feels that consumer survey evidence is relevant to Lanham Act analysis only after a court has found that an advertisement is literally true and must determine whether the advertisement misleads or deceives the public. I conclude, however, that the majority is mistaken in taking this approach.

The dictates of logic and of this Court compel the conclusion that consumer survey evidence is relevant to Lanham Act analysis at two analytical stages that precede a determination of whether an advertisement is misleading. First, consumer survey evidence is particularly relevant to the threshold determination of what message an advertisement conveys to consumers. This Court has established that "[w]hen analyzing a challenged advertisement, the court first determines what message is conveyed.... After determining the message conveyed, the court must decide whether it is false or misleading." U.S. Healthcare, 898 F.2d at 922 (emphasis added).3 Indeed, because an advertisement may include a number of implicit and explicit claims and meanings, see BeVier, 78 Va.L.Rev. at 34-35,4 and because "context can often be important in discerning the message conveyed," it may be impossible or improper facially to determine what messages an advertisement will convey to consumers. See U.S. Healthcare, 898 F.2d at 922. "[T]he court's reaction [to an advertisement] is at best not determinative and at worst irrelevant. The question in [Lanham Act] cases is--what does the person to whom the advertisement is addressed find to be the message?" American Brands, Inc. v. R.J. Reynolds Tobacco Co., 413 F. Supp. 1352, 1357 (S.D.N.Y.1976).

Information is required "to determine what message was actually conveyed [by advertisements] to the viewing audience. Consumer surveys supply such information." Johnson & Johnson Merck Consumer Pharmaceuticals Co. v. SmithKline Beecham Corp., 960 F.2d 294, 298 (2d Cir.1992). This Court has explained that the "critical determination" in Lanham Act cases is the "effect of the advertisement on the consumers." Sandoz, 902 F.2d at 229. I find it inconsistent with the framework of Lanham Act analysis previously delineated by this Court to permit a district court to ignore consumer survey evidence regarding what message consumers derive from a challenged advertisement and, instead, to substitute the court's own hypothesis of what message consumers might receive. While in the absence of consumer survey evidence, a court may have to determine what message a challenged advertisement conveys, it is illogical and, I conclude, an error of law in this Circuit, to ignore such evidence when it is presented. Cf. Alpo Petfoods, Inc. v. Ralston Purina Co., 720 F. Supp. 194, 199 (D.D.C.1989) (In determining message conveyed, court considered results of audience focus groups, survey research and testimony of members of audience; court concluded advertising was literally false.), rev'd in part on other grounds, 913 F.2d 958 (D.C.Cir.1990). Indeed, neither the parties in the present case, nor the majority, nor even have I found any Lanham Act case where a court disregarded a consumer survey in favor of its own interpretation of a challenged advertisement. It may occur that a court will find consumer survey evidence to be misdirected or unpersuasive. I conclude, however, that at least discussion and analysis of the survey results should be incorporated in the court's determination of the case.

Second, consumer survey evidence is also highly relevant to a determination of whether an advertisement contains a "material representation or description [that] was false." Sandoz, 902 F.2d at 231. This Court, in Sandoz and in U.S. Healthcare, has recently reemphasized that a Lanham Act plaintiff is entitled to relief only if an advertisement contained a false representation or description that was "material." See Sandoz, 902 F.2d at 231; U.S. Healthcare, 898 F.2d at 922; see also Toro, 499 F. Supp. at 251; Alpo Petfoods, Inc. v. Ralston Purina Co., 913 F.2d 958, 964 (D.C.Cir.1990) (literally false advertisements must be "material in their effects on buying decisions" to be actionable under the Lanham Act). This Court's adherence to a materiality requirement may distinguish its approach from that of the Second Circuit. "[T]he practical effect of the Second Circuit's test is to presume a likelihood of injury whenever the defendant's ad makes a single false statement and there is competition between the plaintiff and the defendant. The effect is, in other words, to presume injury from even the most innocuous of false claims and thus to lighten into virtual nonexistence plaintiff's burden of proving that consumers were in fact deceived by them." BeVier, 78 Va.L.Rev. at 29.5 In light of this Court's adherence to a materiality requirement for Lanham Act relief, I find the majority's reliance upon a Second Circuit standard that eliminates any materiality requirement to be misplaced.

Consumer survey evidence is extremely helpful in determining whether an allegedly false statement is material. Evidence of "consumer reaction" is essential to "the issue of whether or not the ad has actually affected consumer purchasing behavior." Jeffrey P. Singdahlsen, Note, The Risk of Chill: A Cost of the Standards Governing the Regulation of False Advertising Under Section 43(a) of the Lanham Act, 77 Va.L.Rev. 339, 350-51 (1991). Even the Second Circuit has recognized that it is "well established that the truth or falsity of the advertisement usually should be tested by the reactions of the public." American Home Products Corp. v. Johnson & Johnson, 577 F.2d 160, 165 (2d Cir.1978).6 In discussing literal falsity under the Lanham Act, Professor Bauer explains that "[c]onsumer surveys present the most reliable way of establishing the existence of falsity or deception. Indeed, most courts have found this approach dispositive, since the proper issue in these cases is the reaction of the group to whom the claim is directed." Bauer, 31 UCLA L.Rev. at 737. It is only "in the absence of such reliable [consumer survey] evidence, [that] the court must make its own best judgment of the falsity of advertising claims." Id. Therefore, I conclude that the district court committed an error of law in ignoring the consumer survey evidence presented in the present case. Both because that evidence would clarify what message Pennzoil's advertisements conveyed to consumers and because it could demonstrate whether any false message contained in the advertisement would be likely actually to affect consumer purchasing decisions, I would remand this case to the district court for analysis of the consumer survey evidence.

B. The Appropriate Burden of Proof For a "Non-Establishment" Claim of Literal Falsity

I believe that the majority has also incorrectly analyzed the question of what burden of proof is shouldered by a Lanham Act plaintiff pursuing a literal falsity claim in a non-establishment case. The analytical framework articulated by the majority would allow a plaintiff to establish literal falsity in a non-establishment case, even when the defendant has some reasonable basis to support its advertisement, merely by introducing evidence that other "tests" disprove the defendants claim. Such a framework is inconsistent with this Court's previous analysis of section 43(a) of the Lanham Act.

Judge Becker's opinion in Sandoz contains a well-reasoned and thorough discussion of this Court's interpretation of the Lanham Act. The majority dismisses Sandoz because "the court found that the advertisements in question were not literally false." Majority opinion at 943, 944, 947. I feel that this Court's most complete examination of section 43(a) of the Lanham Act cannot be so easily disregarded.

Principally, Sandoz established that consumer reaction evidence is necessary to a non-establishment claim of implied falsity. See Sandoz, 902 F.2d at 229. However, it also included a discussion pertinent to the question of literal falsity. First, the opinion defined literal falsity by referencing PPX Enterprises v. Audiofidelity Enterprises, 818 F.2d 266, 272 (2d Cir.1987). See Sandoz, 902 F.2d at 228. I find this reference significant. In PPX, "Audiofidelity fraudulently represented [on the record labels] that eight albums it marketed contained performances featuring Jimi Hendrix, when they did not." 818 F.2d at 272. The court stated that consumer surveys were unnecessary because "[t]he only possible conclusion to be derived from Audiofidelity's conduct was that consumers actually were deceived by the misrepresentations." Id. This case represents the most narrow, restrictive definition of literal falsity: a bold faced lie about the most crucial element of the product at issue.7 In a case such as PPX, not only was there no evidence introduced that PPX 's claims were true, such evidence would have been impossible because its claim was absolutely false. I believe that the citation to PPX in Sandoz suggests that this Court has adopted a narrow literal falsity standard.

Moreover, Sandoz includes a more comprehensive discussion of whether a claim is "literally false" such that it would "violate the Lanham Act absent proof that consumers are actually misled." Sandoz, 902 F.2d at 228 n. 7. There, the court explains that "[a] Lanham Act plaintiff may be permitted to presume that consumers expect advertisers to have at least some semblance of support for their publicly-disseminated claims." Id. The court also stated that with regard to a "completely unsubstantiated" claim, "there is a plausible argument that the claim is literally false because the advertiser has absolutely no grounds for believing that its claim is true." Id. I believe that this discussion has two implications. First, it implies that Lanham Act defendant may have to produce some evidence that its non-establishment claim is valid. Second, it implies that, once a defendant has advanced such evidence, consumer survey evidence must be introduced to demonstrate that the plaintiff has been materially harmed by a false assertion in the defendant's advertisement. These implications are also consistent with the Sandoz court's reference to PPX. Cf. Alpo, 720 F. Supp. at 199. If a Lanham Act defendant can produce no evidence that its advertised claims are true, and its claims are material, see Sandoz, 902 F.2d at 231, a finding of literal falsity absent consumer reaction evidence may be proper.8

Based upon this Court's analysis in Sandoz and the issues of interpretation and materiality discussed in section I.A., I find that the appropriate burden of production for a Lanham Act defendant in a non-establishment case is to provide some credible substantiation or basis for belief that its claims are accurate. Once the defendant has met this burden, the plaintiff must prove that consumers are materially deceived by introducing consumer reaction evidence.9 This burden analysis is logically compelling when one recalls both the focus of the Lanham Act and the differences between establishment and non-establishment claims. The Lanham Act focuses upon how consumers actually are likely to react, not upon "prevent[ing] misrepresentation in the abstract." BeVier, 78 Va.L.Rev. at 38. In an establishment case, an assertion that "tests prove" a claim may be literally false if "tests" do not prove that claim. Because these establishment claims cite "tests" in general terms, the validity of such a claim can be determined by weighing the credibility of competing tests. But in a non-establishment case, no tests are referred to; there is no explicit warrant for the advertiser's claim. It is therefore inappropriate in a non-establishment case for the court to weigh tests absent evidence that consumers either expect tests to have been performed to substantiate the advertisers' claims or have been deceived by those claims.10

In the present case, Pennzoil did introduce evidence that allegedly supported its advertised claims. The viscosity of motor oil is enhanced by additives, including polymers. Pennzoil contends that its "star" polymer additive provides greater protection against viscosity breakdown. Pennzoil supports this claim with results from the ASTM D-3945 test. This test showed that Pennzoil motor oil suffered less viscosity loss percentage than Castrol motor oil.11 The district court relied upon other tests that it found more credible to hold that Pennzoil's advertisements were literally false. I find that the district court articulated and applied an improper burden of proof. I would remand to the district court to analyze whether the ASTM D-3945 Test possessed the minimal credibility necessary so that Pennzoil had a reasonable basis to believe its claim true and, if so, to examine the consumer survey evidence presented to determine if consumers would likely be misled by Pennzoil's advertisements.12

Apart from the questions of the relevance of consumer survey evidence and the appropriate burden of proof in this case, I believe that the district court committed an error of law by articulating and applying an "industry standard" criterion to judge the scientific tests of motor oil performance introduced by Pennzoil and Castrol. The district court relied upon cases where courts looked to industry standards or custom to determine the meaning of words in advertisements, to find that, as a matter of law, a Lanham Act plaintiff in a non-establishment case can prove an advertisement literally false by demonstrating that it is not proven true by conformity to an industry standard or a test that comprises an industry standard. See district court opinion at 436-37. The district court applied this analysis to the "viscosity breakdown" claim and then, by "necessary implication," to the implied engine wear claim.

This "industry standards" analysis is fallacious. Under the Lanham Act, "[c]ontext can often be important in discerning whether the message conveyed is literally false and this is particularly true where, as here, the target of the advertising is not the consuming public but a more well informed and sophisticated audience." Princeton Graphics Operating, L.P. v. NEC Home Electronics, Inc., 732 F. Supp. 1258, 1266 (S.D.N.Y.1990). The implication of this statement is that Lanham Act standards may be different when an advertisement is targeted to a specific audience than when an advertisement is targeted to the general public. All of the cases cited by the district court to support the "industry standard" test are cases where the target audience of advertisements were knowledgeable members of the industry at issue. See Johnson & Johnson v. GAC Int'l, Inc., 862 F.2d 975 (2d Cir.1988) (advertisement of "polysapphire" dental bridge to orthodontists where product did not conform to industry definition of "sapphire"); Princeton Graphics, 732 F. Supp. at 1262 (advertisement of "compatible" video monitor to computer stores and salesmen where product did not conform to industry definition of "compatible"); American Rockwool, Inc. v. Owens-Corning Fiberglass, 640 F. Supp. 1411 (E.D.N.C.1986) (advertisement of fiberglass that will result in less "quality control" time to construction companies demanded definition of "quality control" used in the industry).

The present case is quite different from those "industry standards" cases. Here, the audience is the general public, not a specific industry. "[T]echnical industry standards are often irrelevant to consumer expectations." W.L. Gore & Assoc. v. Totes, Inc., 788 F. Supp. 800, 807 (D.Del.1992). The general public has no idea what comprises the "industry standard" for oil viscosity breakdown. It is illogical to hold that the general public expects claims of motor oil superiority to be based upon an "industry standard" rather than upon a non-standard, but scientifically accurate, test. Furthermore, merely because a test does not set or comprise an industry standard does not render that test invalid or irrelevant. Just as the ACT is not invalid merely because the SAT is the industry standard for college admissions at eastern U.S. schools, the ASTM should not be invalid just because the "Stay-In-Grade" and "HTHS" tests comprise the industry standards for motor oil viscosity analysis.13

The majority grapples with the district court's "industry standard" analysis by stating that "the district court did not enjoin Pennzoil's advertisements merely because the ASTM-3945 was not an industry standard test." Majority opinion at 945. I disagree. The district court explained that "[t]here is no standard or specification in the motor oil or automotive industries that grades or ranks motor oils based on percent permanent viscosity loss, as measured by ASTM D-3945 or any other method." District court opinion at 438. The district court then explicitly held that "[b]y industry standards the empirical data made available to the Court and the evidence provided through expert testimony renders Pennzoil's viscosity breakdown superiority claim literally false." Id. The district court discounted Pennzoil's test because it did not conform to an industry standard. Therefore, its factual analysis was controlled by an incorrect legal standard. I would remand for reevaluation of the legitimacy of the motor oil viscosity tests without regard to use as an "industry standard" as a controlling criterion.

Additionally, I find that the district court's application of the "necessary implication" doctrine to the engine wear claim was improper. First, this analysis merely extended the "industry standards" analysis made with regard to the "viscosity breakdown" claim to the engine wear claim. Since it was erroneous to perform the "industry standards" analysis in the first place, it is erroneous to extend that errant analysis to the engine wear claims.

Second, the present case is far different from the "necessary implication" cases cited by the majority and the district court. In Cuisinarts, Inc. v. Robot-Coupe Int'l Corp., No. 81 CIV 731, 1982 WL 121559 (S.D.N.Y.) June 9, 1982 (LEXIS GenFed Library, Dist file), the only possible interpretation of the advertisement was that twenty-one top chefs had selected Robot-Coupe professional food processors over Cuisinart professional food processors when, in fact, Cuisinart didn't manufacture professional food processors. Therefore, by "necessary implication," the ad was literally false. In Tambrands, Inc. v. Warner-Lambert Co., 673 F. Supp. 1190 (S.D.N.Y.1987), a pregnancy test that advertised results in "as fast as 10 minutes" was literally false "by necessary implication" because, while it was possible to obtain conclusive positive results in 10 minutes, conclusive negative results required at least 30 minutes. These cases clearly dealt with false representations of fact, not allegedly false implications of a possibly invalid, but scientifically accurate, test. The kind of implication drawn by the district court is not the type of clearly false factual implication required by the Cuisinarts and Tambrands ads.

Moreover, the majority's reliance upon Gillette Co. v. Wilkinson Sword, Inc., No. 89 Civ. 3686 (KMW) (S.D.N.Y. Jan 9, 1991) is misplaced. In Gillette, the defendant's claim that its shaving product was "smoother" was found to be false. The court relied heavily upon consumer survey evidence to determine both what messages the challenged advertisements conveyed to consumers and, ultimately, to hold that the messages were implicitly false. See id., slip op. at 9-23, 27-31, 43. Despite the presence of one sentence stating that "[t]he court finds that the "six times smoother" claims, and all claims using the word "smoother" in Wilkinson's advertising for Ultra Glide, necessarily imply a claim of shaving smoothness superiority," id., slip. op. at 9, the court's analysis and holding in Gillette are entirely consistent with the Lanham Act analysis I advocate and believe properly reflects the law in this Circuit.

While I certainly agree with the majority that "a court must analyze the message conveyed in full context" when assessing whether an advertisement is literally false, see majority opinion at 946, that context is properly determined through consumer reaction evidence, not through the subjective interpretation of the district court. I cannot agree with the majority's finding that, because Pennzoil's Executive Vice President of Marketing "conceded" in his deposition that he thought Pennzoil's advertisement might convey the message that "a consumer who selects a motor oil other than Pennzoil runs an enhanced risk of engine failure," the district court did not err by ignoring consumer survey evidence that demonstrated that not one of 170 consumers tested found that Pennzoil's advertisement conveyed that message. App. at 1203-05. See majority opinion at 947. The "necessary implication" analysis reaffirms my conclusions that it was improper to ignore the consumer survey evidence presented in the record, that the district court and majority articulated and applied an incorrect burden of proof in this case, and that the district court's analysis was fallaciously grounded in an "industry standards" criterion. I would reverse and remand this case for appropriate attention to the consumer survey evidence in the record and for a reevaluation of the tests regarding motor oil viscosity.

Because I would reverse and remand this case based upon the merits of the underlying Lanham Act claim, the district court's injunction would be reversed. I am troubled by the majority's assertion that the district court's injunction, which Pennzoil claims would ban future, properly supported, advertisements that its motor oil provides better protection against viscosity breakdown than does Castrol motor oil, is not subject to prior restraint analysis because Pennzoil's present claims have been found false. See majority opinion at 949. "Any system of prior restraints of expression [bears] a heavy presumption against its constitutional validity." Bantam Books, Inc. v. Sullivan, 372 U.S. 58, 70, 83 S. Ct. 631, 639, 9 L. Ed. 2d 584 (1963). Certainly the subject of prior restraint analysis in the present case is future, truthful advertising that would enjoy first amendment protection. Therefore, prior restraint analysis is appropriate in the present case. See Beneficial Corp. v. FTC, 542 F.2d 611 (3d Cir.1976); Standard Oil Co. of California v. FTC, 577 F.2d 653 (9th Cir.1978).

Moreover, the case the majority cites for the proposition that the district court's injunction was not overbroad, Quaker State, 977 F.2d at 66, involved a "preliminary injunction ... prior to a full adjudication on the merits," not a permanent injunction after a trial as in the present case. While a preliminary injunction may be valid even if it would prohibit a hypothetical advertisement that is extremely unlikely to exist before the trial date, similar reasoning is inapplicable to a permanent injunction after a full trial. This Court has stated that an injunction against deceptive advertising must extend "no further than is necessary for the elimination of the deception." Beneficial Corp, 542 F.2d at 620. Indeed, in finding an injunction overbroad in a Lanham Act case, the Court of Appeals for the D.C. Circuit explained that "[r]edressing the harm that these claims have caused in the puppy food market does not require that a court supervise all future debate on the anion gap theory." Alpo, 913 F.2d at 972.14 I fear that the majority may have failed to heed the warning of the Court of Appeals for the Ninth Circuit in a similar Lanham Act Case:

Nothing is clearer in the emerging law of commercial free speech than that false or misleading commercial speech is clearly "subject to restraint." But the language of the injunction is much broader than the Conclusion of Law and, perhaps unintentionally, raises First Amendment concerns. As currently written, the injunction seems to prohibit comparative advertising even if it is truthful.

U-Haul Int'l, Inc. v. Jartran, Inc., 793 F.2d 1034, 1042 (9th Cir.1986). I find that the language of the injunction, which prohibits Pennzoil for all time from "broadcasting, publishing or disseminating, in any form or medium, any commercial or advertisement that claims" that Pennzoil protects against viscosity breakdown, engine wear or engine failure better than "any leading motor oil," improperly restrains Pennzoil's commercial speech because Pennzoil is enjoined from broadcasting future, even truthful advertisements proclaiming the superiority of its motor oils.

In summation, my position reflects my conviction that in Lanham Act litigation, the courts are not the vicarious avengers of any false statement that may appear in an advertisement. The courts are rather the referees who prevent material misrepresentations that deceive the intended audience of advertising. This role requires that materiality be established. I cannot see how that can be done reliably or effectively under the procedures approved by the majority.

Present: SLOVITER, Chief Judge, BECKER, STAPLETON, MANSMANN, GREENBERG, HUTCHINSON, SCIRICA, COWEN, NYGAARD, ALITO, ROTH, LEWIS and ROSENN,* Circuit Judges.

March 3, 1993.

MANSMANN, Circuit Judge.

The petition for rehearing filed by appellants in the above entitled case having been submitted to the judges who participated in the decision of this court and to all other available circuit judges of the circuit in regular active service, and no judge who concurred in the decision having asked for rehearing, and a majority of the circuit judges of the circuit in regular active service not having voted for rehearing by the court in banc, the petition for rehearing is denied. Judges Becker, Stapleton, Hutchinson and Roth would have granted rehearing.

Norman Douglas, South Wind 63 (Scholarly Press, Inc.1971) (1931).