Opinion and Order

LINDSAY, District Judge.

Of a thousand shavers,

two do not shave so much

alike as not to be distinguished.

Background

Norelco has recently released its latest innovation in the dry-shaving industry, the Reflex Action Razor (the "Reflex Action"). According to Norelco, this product is an improvement over prior dry-shaving art, because of the Reflex Action's "contour reflex action," which, Norelco says, adjusts to the contours of the user's face. In August of this year, Norelco launched the largest advertising campaign in that company's history, to promote the Reflex Action in the United States. The campaign is an aggressive one, and much of it is focused on the Christmas 1996 buying season. A Norelco press release describes Norelco's promotional efforts for the Reflex Action as an "in your face" campaign (appropriately, it seems, given the product). Norelco hopes to convert users of wet shavers, age 25 and above, to the Reflex Action. To that end, the primary message that Norelco seeks to convey to consumers is that the Reflex Action is less irritating than wet-shaving.

Norelco's advertisements appear on television and radio, in the print media and in direct mailings to three million households in the United States. The advertisements on television and in the print media often depict a wet shaver, not identified by name, as an unpleasant being or as doing unpleasant things. For example, in one advertisement, produced in both print and television formats, the wet shaver spits out flames like those that might be produced by a flamethrower or a dragon. In another, also produced for television and in print, the handle of the wet shaver turns into the coiled body of a snake, and a forked-tongue flicks outward from between the twin blades of the shaver. In yet another advertisement, an animated wet shaver transforms into an unsightly creature with small, but very sharp, teeth, snapping at the viewer. The radio advertisements feature sizzling noises or the sounds of scraping sandpaper and peeling velcro, which the advertisement associates with wet-shaving, and more joyful sounds like the Hallelujah Chorus from Handel's Messiah, which the advertisement associates with a shave with the Reflex Action. Most, but not all, of the advertisements contain a comparison between the degree of irritation caused by shaving with the Reflex Action and the degree of irritation from wet-shaving. The comparison, of course, is favorable to the Reflex Action. Almost all of the advertisements include Norelco's tag line for the campaign: "Anything Closer Could Be Too Close for Comfort."

Gillette takes issue with both the underlying message, that is, that the Reflex Action shaves with less irritation (stated in some advertisements as "clinically proven" to shave with less irritation), and with the presentation of that message. As to presentation, Gillette claims that the visual images and sound effects improperly disparage wet-shaving in general and Gillette's products in particular. Gillette contends that the advertisements are false or misleading under section 43(a) of the Lanham Act ("section 43(a)"), and that they constitute unfair and deceptive trade practices in violation of Mass. Gen.L. ch. 93A ("chapter 93A"). Gillette has applied for a preliminary injunction that would prohibit Norelco from continuing its advertising campaign, during the pendency of this litigation, to the extent that the campaign violates section 43(a) or chapter 93A, or both. It is to that application that this opinion is addressed.

Analysis

I. Can Gillette Satisfy the Requirements for a Preliminary Injunction?

There are four things Gillette must show to prevail on its application for a preliminary injunction: "(1) that [Gillette] will suffer irreparable injury if the injunction is not granted; (2) that such injury outweighs any harm which granting injunctive relief would inflict on [Norelco]; (3) that [Gillette] has exhibited a likelihood of success on the merits; and (4) that the public interest will not be adversely affected by the granting of the injunction." Camel Hair Cashmere Inst. v. Associated Dry Goods Corp., 799 F.2d 6, 12 (1st Cir. 1986); accord Polar Corp. v. Coca-Cola Co., 871 F. Supp. 1520, 1521 (D.Mass. 1994) (in context of false advertising claim).

In section 43(a) cases, the complicated balancing involved in analyzing considerations one, two and four above need not be undertaken if the court concludes that some challenged statement made by a defendant is likely to be found "literally false." See Associated Dry Goods, 799 F.2d at 16 ("[T]he district court's finding that it was probable that defendants' labels were literally false in itself warranted the grant of the injunction sought.").

Because the likelihood of success is the most crucial of the requirements for the granting of an injunction, it is appropriate to begin the analysis of this case with that issue.

A. Likelihood of Success on the Merits.

As a preliminary matter, the court must review the state of the law in the First Circuit with respect to whether false advertising claims are actionable under section 43(a). Historically, the First Circuit has read section 43(a) restrictively. In Samson Crane Co. v. Union Nat'l. Sales, Inc., 87 F. Supp. 218 (D.Mass. 1949), aff'd 180 F.2d 896 (1st Cir. 1950), the district court concluded that section 43(a) was very narrow in scope, applying only to claims of "passing off." The court held that section 43(a) "must . . . be construed to refer not to any competitive practice which in the broad meaning of the words might be called unfair, but to that `unfair competition' which has been closely associated with the misuse of trade-marks, i.e., the passing off of one's own goods as those of a competitor." Id. at 222. This interpretation has been most recently followed (albeit with little brio) in Clamp-All Corp. v. Cast Iron Soil Pipe Inst., 851 F.2d 478, 491 (1st Cir. 1988) ("After examining the record, however, we have concluded that this is not an appropriate case in which to reconsider the validity of Samson Crane."), cert. denied, 488 U.S. 1007, 109 S.Ct. 789, 102 L.Ed.2d 780 (1989).

In 1988, Congress amended section 43(a). It now reads, in relevant part:

(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 —

* * * * * *

(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.

As both parties appear to assume, the better argument is that the narrow reading given section 43(a) in Samson Crane cannot survive the 1988 amendment. The legislative history of the amendment makes clear that the new language of the Act is intended to codify the broad interpretation given section 43(a), before the 1988 amendment, in circuits other than the First. The amendment

revises Section 43(a) of the Act to codify the interpretation it has been given by the courts. . . . As written [before the amendment], Section 43(a) appears to deal only with false descriptions or representations and false designations of geographic origin. Since its enactment in 1946, however, it has been widely interpreted as creating, in essence, a federal law of unfair competition. For example, it has been applied to cases involving . . . actionable false advertising. . . .

S.Rep. No. 515, 100th Cong., 2nd Sess. 40 (1988), reprinted in 1988 U.S. Code Cong. Admin. News 5577, 5603 (citations omitted); see also 3 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 27.02[4] (3rd ed. 1995).

District courts in the First Circuit have concluded, or at least have assumed, that the 1988 amendment "consigned Samson Crane and its progeny to the dustbin of First Circuit jurisprudence." Kasco Corp. v. General Services, Inc., 905 F. Supp. 29, 34 (D.Mass. 1995); accord Pacamor Bearings, Inc. v. Minebea Co., 918 F. Supp. 491, 497-98 (D.N.H. 1996) (holding Samson Crane to be no longer controlling in the First Circuit); Polar Corp. v. Coca-Cola Co., 871 F. Supp. 1520 (D.Mass. 1994) (granting preliminary injunction under section 43(a) for product disparagement); but see General Electric Co. v. Iljin Corp, No. 89 Civ. 40094, 1993 WL 41752 at *7 n. 8 (D.Mass. Feb. 12, 1993). In light of the purpose of the 1988 amendment, as expressed in its legislative history, and the predominant case law in this circuit after the amendment, the court concludes that the restrictive reading given section 43(a) in Samson Crane is no longer law in the First Circuit.

With preliminary matters put to rest, let us turn now to the claims Gillette makes in this case and the likelihood of a successful prosecution of those claims. Gillette makes a number of contentions on the basis of which it asserts that it is likely to succeed in proving at trial that Norelco's advertising for the Reflex Action violates section 43(a). Each contention will be analyzed in turn.

1. The Challenge to Norelco's Superiority Claim

Gillette challenges the substantive claim of superiority made by Norelco that the Reflex Action shaves with less irritation than wet shavers, and that the Reflex Action is "clinically proven" to shave with less irritation than wet shavers.

Although these two statements are similar at first glance, the law developed under section 43(a) makes clear that they are, in fact, quite distinct. Case law under section 43(a) separates comparative claims into two categories, "establishment claims" and "non-establishment claims."

An establishment claim is one that says, in substance, that "tests or studies prove" a certain fact. A non-establishment claim is a general claim of superiority. A plaintiff contesting an establishment claim bears a different burden from that to be borne by a plaintiff contesting a non-establishment claim.

A plaintiff's burden in proving literal falsity thus varies depending on the nature of the challenged advertisement. Where the defendant's advertisement claims that its product is superior, plaintiff must affirmatively prove defendant's product equal or inferior. Where . . . defendant's ad explicitly or implicitly represents that tests or studies prove its product superior, plaintiff satisfies its burden by showing that the tests did not establish the proposition for which they were cited. We have held that a plaintiff can meet this burden by demonstrating that the tests were not sufficiently reliable to permit a conclusion that the product was superior.

In a non-establishment claim, the plaintiff's burden is greater: the plaintiff must prove actual falsity of the challenged claim, "not merely that it is unsubstantiated by acceptable tests or other proof." Chesebrough-Pond's Inc., 747 F.2d at 119. Because proof of the non-establishment claim places a greater burden on Gillette, the analysis of Gillette's likelihood of success will turn first to Gillette's case on that issue.

a. The Non-Establishment Claim

Gillette challenges Norelco's non-establishment claim, that is, the unadorned claim that the Reflex Action shaves with less irritation than wet shavers. As noted above, to succeed on a challenge to a non-establishment claim, Gillette bears the burden of proving that Norelco's product is not superior to wet shavers, as Norelco claims it to be. In an effort to meet this burden Gillette has produced a study, which it says, demonstrates that the Reflex Action is not superior to Gillette's Sensor or SensorExcel, with respect to the irritation characteristic.

Gillette's study consisted of an in-plant, "split-face" comparison of its top-of-the-line shaving product, the SensorExcel, with the Reflex Action. The study was conducted after this lawsuit was commenced. The participants were thirty Gillette employees who regularly participate in comparative product studies. The customary shaver used by each of them was the SensorExcel.

Each participant was given a Reflex Action and told to shave one side of his face with that shaver and the other side of his face with the SensorExcel for a period of two weeks. Each day, the participants recorded their subjective evaluations of the two shavers as to six characteristics, a number of them measuring irritation. The conclusion of the study was that the SensorExcel, not the Reflex Action, was less irritating.

This study has a number of disabling flaws. First and foremost is the fact that the participants in the study were all Gillette employees. The only criteria for screening the participants were that their regular shaver was the SensorExcel and that they were available, that is, they were not then engaged in any other product testing. Thus, for example, Gillette did no screening to exclude those employees who might be investors in Gillette's securities, or those who might be aware of this litigation, or those who might have seen or heard Norelco's advertisements.

But more than all of this, the use of Gillette employees in a comparative study of a Gillette product and a competitor's product is inherently biasing. The bias arises, in part, from the natural feelings of loyalty to the employer that can be expected in a test of the employer's product against that of another. Moreover, it is not far-fetched to suppose that some participants in the test might see an unbroken chain running from a test unfavorable to Gillette, to advancement by the competitor in the shaving products industry, to a threat to the job security of Gillette employees.

Gillette responds that it has used employees in product tests for upwards of ten years. But Gillette also admits, grudgingly, that most of these tests have compared Gillette products then on the market to Gillette products then in development. In such a situation bias of the employees is a much less troubling problem, if a problem at all.

Furthermore, even Gillette's own experts acknowledge that using employees in a study that compares the employer's product to the product of a competitor is undesirable. Dr. Jacob Jacoby, who produced a consumer reaction survey for Gillette in this case, specifically "excluded those . . . who worked for any business that makes, distributes or sells shaving products" because "such people might possess special familiarity with the issues that might yield responses atypical and not representative of the remainder of the universe." Jacob Jacoby, "Do Consumers Extract Disparaging Meanings From the Advertising Campaign for the Norelco Reflex Action Razor?" at 11 ("Jacoby Study"); see also Deposition of Sharon Keith at 19, Exhibit Q to Declaration of Ronald Drozdenko (employees of Gillette excluded from a 1992 consumer use test pool because of "bias which they might introduce"); Jacoby Study at 6 ("It is generally considered good research practice to exclude from the universe people who live in households where someone works . . . in a relevant industry (in this case, a company that makes, distributes or sells shaving products).")

Given the biasing of the in-plant study, resulting from the use of Gillette employees as the test subjects, and the subjective nature of the measurements used, the study was enfeebled from the start.

But there are other flaws as well. One of these has to do with the concept of acclimation — a period during which shaving results with a new shaving system are not optimal, because of the adjustment of the consumer to the new product. Acclimation is more fully discussed in later sections of this opinion. It is enough to say for present purposes that the instructions for use of the Reflex Action tell consumers that they will not likely find the Reflex Action less irritating until they have used it for a period of at least twenty-one days. Gillette too appears to recognize acclimation as a phenomenon. In its usual comparative testing of shavers, Gillette has made it "standard company policy to conduct such tests over a three-week period." Deposition of Sharon Keith at 16, Exhibit X to Declaration of Ronald Drozdenko. In the in-plant study, however, Gillette conducted its tests over a fourteen-day period only, and took subjective measurements from the very beginning of the test.

Moreover, Gillette compared only its top-of-the-line product, the SensorExcel, to the Reflex Action. It is not clear to the court that the statement that the Reflex Action shaves with less irritation than wet shavers would be understood by consumers to mean that the Reflex Action shaves with less irritation than all wet shavers, or most wet shavers, or the average of all wet shavers. Gillette has produced no consumer survey evidence demonstrating that Norelco's claim is understood to be a claim of superiority over all wet shavers. Without such a survey, the court cannot say whether a study that compares the Reflex Action only against the SensorExcel is an appropriate test of the truth of Norelco's general superiority claim.

The court need not reach the question of whether Gillette's failure to conduct its in-plant study over a longer period is a fatal defect in the study, or the question of whether a study that matches the Reflex Action against the SensorExcel exclusively would yield useful results. The difficulty with the in-plant study, that makes unnecessary any further evaluation of it, is that the study simply cannot overcome its connate bias. It is therefore unlikely to persuade a factfinder of the falsity of Norelco's general superiority claim. Because Gillette can point to nothing else that belies Norelco's non-establishment claim, the court concludes that Gillette is not likely to succeed in disproving that claim.

b. The Establishment Claim

Norelco relies upon a study done by Pharmaceutical and Cosmetics Evaluations ("PACE") as the basis for its establishment claim that the Reflex Action has been "clinically proven" to shave with less irritation than wet shavers. PACE, an independent testing organization, actually performed two studies designed to produce comparative data on the degree of irritation from wet-shaving as compared to the degree of irritation from shaving with the Reflex Action. A brief summary of the methodology and results of the two studies follows.

In the first PACE study, labeled a "pilot" study, PACE conducted a split-face shave test. The participants were twenty-one men who normally shave with a wet shaver. They were told to shave one side of their faces with their normal shavers and the other side of their faces with the Reflex Action for two weeks. During this two-week period the participants were told to, and did, record their perceptions about a number of attributes of the different shavers. PACE evaluated these responses for days eleven through fourteen in an attempt to determine which shaver was perceived by the participants to be less irritating. The participants then returned to the PACE offices, where they performed a final, split-face shave. Objective and perceptual measurements of irritation were then recorded.

The pilot study produced results which were, for the most part, favorable to the Reflex Action. Gillette, however, points to two results, which were less favorable to Norelco, and contends that these results vitiate the reliability of PACE's final study, the study on which Norelco relies as support for its "clinically-proven" claim.

As one objective test of irritation, PACE measured the skin temperature on three parts of the face, the cheek, jawline and neck, using an electronic infrared thermometer. PACE concluded that, "[f]or each site, and for the maximum measure of the three sites, the blade showed a lower temperature than the Reflex." Gillette Exhibit No. 19, Pilot Study of Irritation at N00080.

In addition, in the pilot study, both objective and perceptual ratings were recorded for irritation at only three test sites. At one of these sites, the neck, the Reflex Action did not compare favorably with the wet shavers. As the study notes, "the Reflex did not show an advantage in reducing irritation on the neck. . . . [A number of objective tests] tended to favor the blade on the neck." Id. at N000076.

In its recommendation for the final study, PACE urged the elimination or modification of the temperature measurements and the enlargement of the number of facial sites at which measurements would be made. With respect to the temperature test, the report notes that the test "[m]ay go in opposite direction unless evaporative effects can be controlled. Record as last measure. Use as an `internal' measure." Id. at N00088. To de-emphasize the Reflex Action's less favorable results in the neck area, PACE recommended adding the lips and the chin as testing sites in the final study, because the addition of these sites "[s]hould favor the Reflex [and] [m]inimize impact of Neck in a `Combined Irritation Measure.'" The changes recommended by PACE were adopted for the final study.

Gillette also challenges the conclusion of the pilot study that Norelco "won" four of the five perceptual ratings in the pilot study. The pilot study asked the participants to record their subjective interpretations of five "irritation ratings": stinging, redness, burning, irritation signs and overall irritation. Gillette argues that, because Norelco has made a superiority claim in its advertisements, it is proper to combine the results from those participants who favored wet-shaving with the results from those who expressed no preference as between the two types of shavers, and then compare that combined result to those who favored Norelco. When the results of the perceptual ratings are compared, in the manner Gillette suggests, three of the perceptual ratings called "wins" for the Reflex Action actually become "wins" for the wet shavers. Over fifty percent of the participants in the pilot study favored the Reflex Action as to irritation signs and overall irritation, but as to the other three perceptual ratings (stinging, burning and redness), a majority of the participants favored wet-shaving or expressed no preference. On these ratings, therefore, the pilot shows that Reflex Action was not superior to wet shaving.

This same analysis applies to one of the final questions put to participants in the pilot study — a question asking the participants to compare the Reflex Action (after the ten day acclimation period) to their normal wet shaver as to degree of irritation. Only 42.8% of the respondents replied that the Reflex Action was less irritating; 14.3% said the two shaving systems were about the same; and 42.9% said the wet shaver was less irritating. Thus, as to this question, as well, the wet shavers achieved no worse than parity with the Reflex Action. Indeed, for section 43(a) purposes, the wet shavers may be said to have "won" on this final question. (The wet-shaver results, combined with the no preference results, totaled 57.2%.)

A few months after completing the pilot study, PACE submitted to Norelco its second study — the final study. As noted above, it is the final study upon which Norelco relies for support of its statement that the Reflex Action is "clinically proven" to shave with less irritation than wet shavers. The final study differed in significant ways from the pilot study. The final study involved fifty-two men, who normally shave with a wet shaving system. In contrast to the pilot study, the final study utilized a "full face" design, not a "split-face" design. The participants went to PACE's headquarters on day one, where they shaved with their normal wet shavers. Objective and perceptual measurements of irritation were then taken. The participants were then given a Reflex Action and told to shave exclusively with that shaver for three weeks. After that period, the participants returned to PACE, where they shaved with the Reflex Action. Objective and perceptual measurements of irritation were again taken. Finally, the participants returned the next day, where they again shaved with their wet shavers. Objective and perceptual measurements were again taken. The two sets of wet-shaving measurements (one recorded just before the three weeks of shaving with the Reflex Action and one recorded just after that period) were then averaged, and this average was compared to the results for the Reflex Action. The final study concluded that "8 of the 9 irritation measures independently reached statistical significance" in favor of the Reflex Action. Gillette Exhibit No. 20, Final Study of Irritation at N00056.

Gillette criticizes PACE's final study on a number of grounds, and argues that, in the end, the final study is wholly unreliable. Gillette's most strenuously-urged argument is that "the so-called `pilot study' showed PACE and Norelco how to modify the final study so that the data obtained in that study would be skewed unfairly in favor of Norelco's product." Gallerani Affirmation, ¶ 43.

It is true, as Gillette claims, that PACE did not include the temperature test and that it measured irritation in six, as opposed to three, parts of the face in the final study. Norelco has offered both general and specific explanations for these variations.

The specific explanations may be quickly stated. As to the temperature test, Norelco argues that the shaving cream and water used in wet shaving affect the measurement of this objective test. As to the addition of test sites in the final study, Norelco responds that by adding test sites, a better reflection of overall irritation could be obtained.

Norelco's general explanation is that Gillette misapprehends the purpose of a pilot study. A pilot study, according to Norelco, is not part of a testing protocol, but is meant to assist in the creation of a final, meaningful test. In other words, a pilot study, according to Norelco, is a device "to examine methods, procedures, problems and the feasibility that a hypothesis would be supported with more extensive (and expensive) research." Declaration of Ronald Drozdenko, ¶ 90.

Gillette has made a number of other criticisms of the final study, beyond those criticisms that relate to differences between the pilot and the final study. Most of these criticisms have been answered by Norelco to the court's satisfaction. Gillette, however, does make certain criticisms of the final study that do require some discussion here.

First, Gillette argues that the fact that PACE took measurements of irritation caused by the wet shavers at only two points in time (at the beginning and at the end of a three-week period during which the participants shaved exclusively with the Reflex Action) is a methodological error. Both of the PACE studies acknowledge, either explicitly or implicitly, that an acclimation period is required for a person who converts from a wet shaver to the Reflex Action. That such an acclimation period is required prompts Gillette to argue that PACE's final study is flawed insofar as it is based on measurements of irritation from the wet shaver taken after the participants had presumably become acclimated to the Reflex Action.

Norelco responds that the protocol for the final study called for the participants in the study to shave first with their customary shavers. No acclimation period was necessary for an accurate measurement of irritation on the first wet-shave, because that shave was performed with the same shaving system the test subjects had always used. They were for that reason, Norelco says, already acclimated to the wet shaver used for the first measurement.

Moreover, Norelco contends, there was no error in averaging the results of the shaves performed before the participants used the Reflex Action for three weeks with the results taken from the shave performed after that period. Norelco, with support in the record, says that when the data are broken down and the individual measurements are taken, the results of the measurements of irritation for the Reflex Action, after three weeks of use, were better than the results for wet shavers measured either before or after the three weeks of Reflex Action use. Thus, whether the Reflex Action is compared with the separate results of each of the two wet-shaves or with the average of these results, the Reflex Action fares better.

Gillette also argues that the fact that PACE took measurements of irritation from wet shavers only at the beginning and at the end of a three-week period of Reflex Action usage raises the problem of the "naive participant." According to Gillette, as participants in the study became accustomed to rating their shaves with the Reflex Action over the course of the three-week period, they subjectively rated the Reflex Action as less irritating. For proof, Gillette points to the pilot study in which the irritation reported for the participants' normal wet shaver was progressively lower as the study continued. Gillette contends that the wet-shaving rating should have remained constant, because each participant used his customary wet shaver throughout the period of the study. The lessening of the subjective measures of irritation, Gillette posits, results from the participants' losing their naivete and becoming experienced in rating shaving performance.

Norelco's response is three-fold. First, Norelco says that, whether the phenomenon of "naive participants" existed in the final study is nothing more than speculation. Second, Norelco says that the phenomenon of "naive participants" would not necessarily give an advantage to Norelco. Norelco's final argument (and, in the end, its most convincing argument) is that, to the extent that there was a "naive participant" problem, it affected only the subjective results of the final study.

This last argument provides a convenient segue to Gillette's criticisms of the objective measurements of irritation used in the final study.

First, Gillette says, PACE failed to take steps to eliminate the possibility of conscious, subconscious, or unconscious investigator bias in its final study. That bias might exist, according to Gillette, because the investigators must have known that they were conducting a test for Norelco, and therefore they might well have taken their objective measurements from parts of the face that appeared most irritated when they measured the irritation from wet-shaving, and from parts of the face that appeared less irritated when they measured irritation from the Reflex Action.

Norelco has two responses. First, it says it was not always possible for the tester who took the objective measures to know what product he or she was testing, because the participants' schedules overlapped. Second, Norelco says the test sites for the objective measurements were predetermined, based on anatomical coordinates, and could not be moved by the investigators to take readings from what might have appeared to them to be more or less irritated areas.

Gillette next argues that it was inappropriate to use a blotter to measure "nicks and cuts" because some nicks and cuts show up, when pressed with a blotter, that are not visible to the naked eye. Norelco's response is a terse, but pointed one: the blotter test was not a test for the appearance of irritation, but for irritation itself; and nicks and cuts may exist even if they cannot be easily seen.

After reviewing the design, execution and results of the final study, the court finds that Gillette has raised questions which suggest that the final study may be flawed in several respects. But the flaws do not appear to the court to be so severe that Gillette is likely to succeed in demonstrating that the study as a whole is unreliable. The criticisms of the final study which the court finds most convincing relate to the reliability of the perceptual measurements. The reliability of the results of the objective measurements have been left largely, if not completely, unimpaired by Gillette's challenges. Accordingly, the court concludes that justification is lacking for an injunction that would prohibit Norelco from making its establishment claim that the Reflex Action has been clinically proven to be less irritating than wet shavers.

2. Visual Images as Disparaging Statements

Gillette argues that the visual images in many of Norelco's print and television advertisements disparage Gillette's wet-shaving products, particularly Gillette's Sensor and its SensorExcel. Gillette contends that Norelco's advertisements for the Reflex Action are explicit messages that equate wet shaving with things like having one's face burned by a flame, bitten by a sharp-toothed animal or stung by a snake or a swarm of bees. In short, Gillette says, Norelco's advertisements portray wet-shaving as a dangerous or painful experience, and because they do, the advertisements are literally false.

An initial, but important, point must be made here. Section 43(a) makes actionable statements that are either false or misleading. The method of proof varies, however, depending on whether the charge is that a statement is false on the one hand, or that it is misleading, on the other. If the assertion is made that a statement is literally false, then there is no need for the court to consider consumer survey evidence. See Associated Dry Goods, 799 F.2d at 15 ("[A] court may grant relief on the basis of its own finding without reference to consumer reaction to the product when the defendant's representations are actually false."); McNeil-P.C.C., Inc. v. Bristol-Myers Squibb Co., 938 F.2d 1544, 1549 (2d Cir. 1991) ("Where the advertising claim is shown to be literally false, the court may enjoin the use of the claim without reference to the advertisement's impact on the buying public.") (quotation and citation omitted).

However, "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." Sandoz Pharmaceuticals Corp. v. Richardson-Vicks, Inc., 902 F.2d 222, 228-29 (3d Cir. 1990) (emphasis in original) (citation omitted). The showing is usually made by means of a consumer reaction survey. See Johnson Johnson * Merck Consumer Pharmaceuticals Co. v. Smithkline Beecham Corp., 960 F.2d 294, 298 (2d Cir. 1992) ("[T]he success of a plaintiff's implied falsity claim usually turns on the persuasiveness of a consumer survey."); L F Products v. Procter Gamble, Co., 845 F. Supp. 984, 995-1000 (S.D.N.Y. 1994) (analyzing surveys), aff'd 45 F.3d 709 (2d Cir. 1995). A plaintiff has carried its burden to show that consumers are confused if a substantial, or not insubstantial, number of consumers are shown to hold the confused view. See, e.g., Johnson Johnson-Merck Consumer Pharmaceuticals Co. v. Rhone-Poulenc Rorer Pharmaceuticals, Inc., 19 F.3d 125, 134 n. 14 (3rd Cir. 1994) (twenty percent confused is sufficient); Smithkline Beecham, 960 F.2d at 298 (plaintiff must "demonstrate that a statistically significant part of the commercial audience holds the false belief allegedly communicated"); Coca-Cola Co. v. Tropicana Products, Inc., 690 F.2d 312, 317 (2d Cir. 1982) (sufficient "that a not insubstantial number of consumers were clearly misled").

Gillette contends that the visual images in Norelco's advertisements send a message that is "false on its face" (another interesting phrase in the context of this case), so that consumer reaction evidence is not required. In the alternative, Gillette argues that, at the very least, the gross exaggerations made in the visual images in Norelco's advertisements are misleading. In connection with this latter point, Gillette has submitted a consumer reaction survey concerning the challenged advertisements.

In general, Gillette argues that Norelco's visual images "disparage" wet shaving. A comment concerning a product disparagement claim under section 43(a) is therefore required here. Product disparagement involves more than a simple negative statement about a competitor's product. The 1988 amendment to section 43(a) explicitly makes actionable false or misleading statements about one's own products or the products of another. See S.Rep. No. 515, 100th Cong., 2nd Sess. 40 (1988), reprinted in 1988 U.S. Code Cong. Admin. News 5577, 5603 ("The committee agrees . . . that the public policy of deterring acts of unfair competition will be served if Section 43(a) is amended to make clear that misrepresentations about another's products are as actionable as misrepresentations about one's own."). The relevant question under section 43(a), therefore, is whether Norelco has made false or misleading representations of fact, about Gillette's products, and not merely negative representations about those products.

The messages conveyed in visual images are to be considered in section 43(a) cases. See Tropicana, 690 F.2d at 317-18 ("The visual component of the ad makes an explicit representation that Premium Pack [orange juice] is produced by squeezing oranges and pouring the freshly-squeezed juice directly into the carton. This is not a true representation of how the product is prepared."); Clorox Co., 930 F. Supp. at 781 ("In assessing the meaning of the Commercial, the Court is not limited to the spoken words: the visual images can be equally important."); L F Products, 845 F. Supp. at 1000 (same).

If, as Gillette contends, Norelco's advertisements are meant to convey the message that wet-shaving is a dangerous or painful experience, the advertisements are not literally false. It cannot be gainsaid, as a matter of common sense, that stroking one's face with the blade of a razor is to some degree, however small, a dangerous activity. It is an activity, moreover, that can be painful. Gillette's own research bears this out. In a study called "Frequency of Irritation", Gillette found that 56.8% of the users of wet shavers in that study had experienced "stinging" or "burning" at some time; 48.5% had experienced "tenderness" at some time and 49.5% had experienced "redness" at some time. Gillette Exhibit No. 46, "Frequency of Irritation." In addition, complaints received by Gillette provide substantial anecdotal evidence of the danger and pain sometimes experienced by those who shave with wet shavers. For example, one customer, in December, 1994, complained to Gillette that, using the SensorExcel he "took a chunk of skin out of [his] chin." Summary, Consumer Complaint Data at G000544, Exhibit L to Declaration of Michael H. Steinberg. Another complained, in January, 1995, that the Sensor cartridge "gives him a bad razor burn" and "bites him or eats him alive." Id. Still another complained in March, 1995, that, after using Sensor Cartridges for five days his face became "red and puffy," requiring him to visit a physician. Id. at G000546. It may be, as other consumer research studies submitted by Gillette appear to show, that there is a high degree of satisfaction with respect to the comfort experienced by those who shave with wet shavers generally and with the Sensor and SensorExcel in particular. But none of these studies and nothing else Gillette has submitted shows that Gillette is likely to be able to establish at trial the literal falsity of any message that associates at least some danger and pain with wet shaving.

The question now becomes whether the Norelco visual images are misleading. As evidence that these images do mislead consumers, Gillette relies on the Jacoby Study. Gillette argues that the Jacoby Study demonstrates that the visual images mislead consumers by distorting the degree and likelihood of the danger and pain associated with wet shavers. To put it another way, Gillette claims that consumers receive from the Norelco advertisements the misleading message that the danger and pain of using a wet shaver are equivalent to the danger and pain of a snake bite, bee stings, or burns from a flamethrower.

In the Jacoby Study, two different advertisements were shown to two different groups of approximately 150 men who regularly shave with a wet shaver. The first advertisement was the so-called "Flamethrower" print media advertisement, as it appears in printed material. The second was a control advertisement showing the text of the Flamethrower advertisement, and the picture of Norelco's product, but omitting the visual image of the wet shaver emitting fire. Jacoby Study at 9. The group that saw the advertisement as published more often responded to questions with answers referring to a relatively high degree burning. From this, the Jacoby Study concludes:

Disparagement caused by the graphics is obvious from the fact that, while those exposed to the modified print ad from which the allegedly disparaging graphics had been excised believed that the advertisers were claiming that shaving with a bladed razor would produce a certain (near average) amount of irritation, those exposed to the as published ad were significantly more likely to extract the meaning that the advertisers were claiming that shaving with a blade razor would produce a greater amount of irritation.

Jacoby Study at 32-33.

The testers for the Jacoby Study also asked the test subjects to identify the brand of wet shaver being compared to the Reflex Action. Fourteen percent answered "Gillette" upon seeing the published advertisement. Jacoby Study at 22 (Table 2).

The legal question, however, as noted above and as admitted by Gillette at the hearing, is not only whether a negative message about wet-shaving is received by consumers, but whether this message is misleading. It bears repeating here that Gillette's own studies and consumer complaints demonstrate that the metaphors played on by Norelco — burning, stinging and biting — are colloquialisms for sensations actually experienced by users of wet shavers. Given this predicate and the court's finding that Gillette is not likely to succeed in demonstrating the falsity of Norelco's superiority claim, the question is whether the use of exaggerated images to describe the kinds of irritation actually experienced by users of wet shavers violates section 43(a).

The job of marketing specialists involves the communication of feelings, sensations and ideas to consumers. Often this is accomplished through the creative use of metaphors, and metaphorical images, which are designed to impart abstractions to the consuming public. It is not unusual for these images to exaggerate their underlying meaning, beyond the point of believability, in order to ensure that that underlying message is conveyed. In this case, Norelco has done just that. Gillette acknowledges that consumers are not likely to believe that their wet shavers will turn into a flamethrower or that a snake's tongue will slither from the razor's cartridge. The Norelco visual images, Gillette admits, are exaggerations. What they exaggerate are realities not uncommon to wet-shaving.

The case law under section 43(a) has developed the concept of "puffery." "Puffery is an exaggeration or overstatement expressed in broad, vague, and commendatory language." Castrol Inc., 987 F.2d at 945. The Castrol court, quoting colorful language from Prosser and Keeton, continued: "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." Id. (quoting W. Page Keeton, et. al. Prosser and Keeton on the Law of Torts § 109, at 756-57 (5th ed. 1984)). "Mere puffing, advertising `that is not deceptive for no one would rely on its exaggerated claims,' is not actionable under § 43(a)." U.S. Healthcare, Inc. v. Blue Cross, 898 F.2d 914, 922 (3rd Cir.) (quoting Toro Co. v. Textron, Inc., 499 F. Supp. 241, 253 n. 23 (D.Del. 1980)), cert. denied, 498 U.S. 816, 111 S.Ct. 58, 112 L.Ed.2d 33 (1990); see generally U-Haul Int'l, Inc. v. Jartran, Inc., 522 F. Supp. 1238, 1244-45 (D.Ariz. 1981) (discussing puffery), aff'd 681 F.2d 1159 (9th Cir. 1982).

Generally, puffery has been applied to positive comments made about one's own products, rather than to negative comments made about the products of another. This is mainly because of those interpretations of section 43(a) which, before the 1988 amendment, held that negative comments about another's products were not actionable at all under section 43(a). Since the 1988 amendment, a number of courts have logically extended the puffery concept. For example, in U.S. Healthcare, the Third Circuit concluded that commercials that featured actors expressing their preference for Blue Cross over a competing HMO and ending with the tag-line, "Better than HMO. So good, it's Blue Cross and Blue Shield," were "the most innocuous kind of `puffing,' common to advertising and presenting no danger of misleading the consuming public." U.S. Healthcare, 898 F.2d at 926. In American Express Travel Related Services Co., Inc. v. Master Card Int'l Inc., the court concluded that exaggerations about the difficulty of finding an ATM machine which accepted American Express was puffery. 776 F. Supp. 787, 790 (S.D.N.Y. 1991) ("The exaggeration used by defendants to convey the message that it is more difficult for an American Express Card holder to find a location where he can get cash with his card is mere `puffing.'").

The Gillette corporate family itself has employed puffery for at least one of its products. At the hearing on the present application, Norelco brought to the attention of the court that Braun, a subsidiary of Gillette that manufactures dry shavers, like Norelco, has "puffed" the danger and pain of wet-shaving. A Braun advertisement for a women's shaver, appearing in the Holiday 1996 Sharper Image Catalog, begins with the plea, "Ladies, we beseech you: Lay down your blades." The advertisement then continues the theme of exaggerated negative statements about blade shaving: "There's no need to resort to the kind of violence inflicted by razors or depilatories," and closes with the statement, "Never feel the harsh, cold blades of steel again." Gillette responds that the advertisement was created by the Sharper Image, and that Braun did not authorize it. Colliander Reply Affirmation, ¶ 3. Gillette's response is like applying tissue paper to a razor nick on the face: it may stop the bleeding, but it does not completely heal the wound. The well-taken point of Norelco's argument is that Gillette can hardly complain about Norelco's puffing when a product of Gillette's own subsidiary is advertised in a similar way, in a catalog that surely cannot have escaped Gillette's attention, even if the advertisement were not specifically authorized.

The court concludes that the Jacoby Study does not demonstrate actionable disparagement by Norelco of Gillette's products. At best, the Jacoby Study demonstrates that the visual images are effective in conveying to the consuming public the theme that wet-shaving is more irritating than shaving with the Reflex Action. Because Gillette has not demonstrated the likelihood of succeeding in proving the falsity of Norelco's general superiority claim, perforce it has not demonstrated likelihood of success in proving the visual images misleading.

3. Gillette's Assertion that Norelco's Representations as to the Reflex Action are False or Misleading by Omission

Gillette next contends that Norelco's claim that the Reflex Action is less irritating is false or misleading because it fails to include a material caveat — that the shaver is less irritating only after an acclimation period of at least twenty-one days. A failure to disclose a material fact can support a claim under section 43(a). The Third Circuit, quoting McCarthy on Trademarks and Unfair Competition, has noted: "While it has been stated that a failure to disclose facts is not actionable under § 43(a), it is equally true that a statement is actionable under § 43(a) if it is affirmatively misleading, partially incorrect, or untrue as a result of failure to disclose a material fact." U.S. Healthcare, 898 F.2d at 921 (quoting 2 J. McCarthy, Trademarks and Unfair Competition § 27:713 (2d ed. 1984)); Energy Four, Inc. v. Dornier Medical Systems, Inc., 765 F. Supp. 724, 731 (N.D.Ga. 1991) (same). In connection with the 1988 amendment of section 43(a), Congress considered including a section that explicitly made actionable a failure to disclose. The Senate report explains that the section was not included

to respond to concerns that it could be misread to require that all facts material to a consumer's decision to purchase a product or service be contained in each advertisement. . . . The committee does not through the deletion indicate that it condones deceptive advertising, whether by affirmative misrepresentation or material omission, and leaves to the courts the task of further developing an [sic] applying this principle under section 43(a).

S.Rep. No. 515, 100th Cong., 2nd Sess. 41 (1988), reprinted in 1988 U.S. Code Cong. Admin. News 5577, 5603-04.

As discussed earlier, Norelco has offered various explanations for the purpose of the acclimation period for the Reflex Action. Norelco does not seriously dispute, however, that an acclimation period of at least twenty-one days is important to any success of the Reflex Action as a less irritating shaver than wet shavers. The evidence on this point abounds in the record of this case. The PACE pilot study reported, for example, that "an appropriate acclimation period is critical to the reduced irritation with the Reflex." Gillette Exhibit No. 19 at N00076 (emphasis added). PACE's final study compared the Reflex Action with wet shavers only after participants in the study used the Reflex Action for a three-week acclimation period. Moreover, one of Norelco's criticisms of one of Gillette's closeness tests (discussed infra) is that the protocol for the test did not provide for an acclimation period for the Reflex Action. Declaration of Ronald Drozdenko ¶ 55. When a consumer opens the packaging for the Reflex Action, the first thing he or she will likely see upon removing the product is a red sticker affixed to and sealing a plastic bag enclosing the shaver. The seal warns:

Stop! Before You Shave . . . If you've been shaving with a blade, it takes 30 days for your skin to gradually adjust to Norelco's closest shave. Once you do, you'll never go back!

The instructions which accompany the Reflex Action, under a heading captioned in bold letters, "Important" devotes an entire page to the acclimation period, telling the purchaser, among other things:

At first you may not get as close a shave as you expect, or your face may even become slightly irritated. This is normal since your beard and skin will need time to adjust.

Even some of Norelco's print advertisements make reference to the acclimation period, even if it is in type that renders the actual wording of the reference nearly indecipherable to eyes unaided by a magnifier. Gillette Exhibit No. 60A.

Norelco began to develop its advertising strategy for the Reflex Action in November, 1995. One of its advertising consultants, Calle and Co. ("Calle") advised:

Few consumers go through the two-three week trial period to get used to an electric, primarily because they have not been supplied with a motivating objective to do so. They assume the objective is to achieve a blade close shave. Expectations are for immediate results. Rationale that the razor `teaches' or `trains' the hair to grow in the right direction for electric cutting is difficult to perceive and has low credibility.

Gillette Exh. No. 24, Report of Calle and Co. for Norelco at N00194.

It appears that Calle's purpose in giving Norelco this advice was to suggest that Norelco's advertising give "positive trial motivation," for the acclimation period: "The majority of target men note a strong interest and willingness to use an electric shaver for the full trial period when supplied with the more motivating and clearly defined objective of getting closeness with less irritation." Id. Thus Calle, in its recommended "concept statement," suggested that the message about the acclimation period be presented positively: "To get less irritation with every shave, use [the Reflex Action] for three weeks. You'll see irritation lessen with every shave. Prove it to yourself or get your money back." Id. at N00199.

It appears that Norelco has largely rejected the idea of casting the acclimation period in a positive light. Norelco has decided instead to cast no light at all on acclimation in much, if not most, of its advertisements — this despite abundant evidence, known to Norelco, that acclimation is essential to favorable results for the Reflex Action, with respect to the lessening of skin irritation.

Gillette has presented strong evidence that Norelco, in omitting from most of its advertisements any reference to an acclimation period, has shaved the truth, as it were. The court finds that Gillette is likely to succeed in establishing at trial that, without a statement that a user of the Reflex Action can expect a less irritating shave only after an acclimation period of at least twenty-one days, Norelco's advertisements are misleading.

4. Norelco's Closeness Claim

In its initial papers in this case, Gillette challenged Norelco's claim, made in the package insert accompanying the Reflex Action, that the Reflex Action provides a "closer, more comfortable shave." At the hearing, Gillette broadened its attack to include Norelco's mailings to potential customers that contain the claim by Norelco that the Reflex Action gives a "smoother, closer, much more comfortable shave every time." Gillette contends that the claim that the Reflex Action provides a closer shave is false.

a. Closeness Claim Made in the Package Insert.

Section 43(a) applies only to claims made "in commercial advertising or promotion." Thus, the first question raised by Gillette's challenge of the package insert is whether section 43(a) applies at all. A growing number of cases discuss the analytical framework for determining whether a claim is made in "commercial advertising or promotion." A four-factor test seems to be gaining acceptance:

In order for representations to constitute "commercial advertising or promotion" under Section 43(a)(1)(B), they must be: (1) commercial speech; (2) by a defendant who is in commercial competition with plaintiff; (3) for the purpose of influencing consumers to buy defendant's goods or services. While the representations need not be made in a "classic advertising campaign," but may consist instead of more informal types of "promotion," the representations (4) must be disseminated sufficiently to the relevant purchasing public to constitute "advertising" or "promotion" within that industry.

In most situations, this test is easy to meet. "Generally, a plaintiff can easily satisfy its burden . . . by pointing to paid advertisements by a commercial defendant on television or radio, or in newspapers or magazines." Gordon Breach, 859 F. Supp. at 1532. Most of the claims in the case at bar fall into this category, and are thus clearly made "in commercial advertising or promotion." However, the claims that are contained solely in the package inserts are not so easily categorized.

Section 43(a) is a remedial statute and should therefore be broadly construed. See Seven-Up, 86 F.3d at 1383; Gordon Breach, 859 F. Supp. at 1532. The statute is not so broad, however, that it includes all statements made by one competitor about its or another competitor's product. See Garland, 895 F. Supp. at 279 ("[T]his court has found no indication that Congress, through its use of the language `commercial advertising or promotion,' intended to extend Lanham Act coverage to every isolated alleged misrepresentation made to a potential customer by a business competitor."); Mobius Mgmt., 880 F. Supp. at 1021 (recognizing that § 43(a) "does not have boundless application") (quoting Alfred Dunhill Ltd. v. Interstate Cigar Co., 499 F.2d 232, 237 (2d Cir. 1974)); American Needle Novelty, Inc. v. Drew Pearson Marketing, Inc., 820 F. Supp. 1072, 1078 (N.D.Ill. 1993) (refusing to read § 43(a) so broadly so as to "sweep within the ambit of the Act any disparaging comment made in the context of a commercial transaction").

In the present case, the difficult question is not whether the statement is "commercial," but rather whether the commercial statement is made in "advertising or promotion." See, e.g., Seven-Up, 86 F.3d at 1384 n. 6 ("The presentation is clearly `commercial' in nature, and we therefore need only address whether it is commercial `advertising' or `promotion' under the Lanham Act."). To state the matter differently, the issue raised by Gillette's challenge to the package inserts is whether the package inserts are within the cincture of the third prong of the Gordon Breach test; that is, whether they are statements made "for the purpose of influencing consumers to buy defendant's goods."

In defining "advertising and promotion," courts have given the words their "plain meaning." See Seven-Up, 86 F.3d at 1384. This has meant that courts may rely on dictionary definitions of the terms. For example, in American Needle, the court relied on Webster's to define advertising as "the action of calling something to the attention of the public [especially] by paid announcements" and to define promotion as "the furtherance of the acceptance and sale of merchandise through advertising, publicity, or discounting." 820 F. Supp. at 1077 (emphasis and alteration in original) (quoting Webster's Ninth New Collegiate Dictionary 59, 942 (1986)); see also Garland Co., 895 F. Supp. at 276 (relying on Webster's Third New International Dictionary). Courts have repeatedly stressed that advertising and promotion must be aimed at the consuming public and intended to influence buying decisions. See Seven-Up, 86 F.3d at 1386 (presentation was advertising or promotion, because it "was specifically intended to influence consumers to buy [the] product"); Mobius Mgmt., 880 F. Supp. at 1020 ("[O]nly promotional representations that are directed at the purchasing public can be reached by § 43(a).") (emphasis added); EventMedia Int'l Inc. v. Time, Inc., No. 92 Civ. 0502, 1992 WL 321629 at *3 (S.D.N.Y. Oct. 26, 1992) ("Hence, only commercial speech that a competitor employs for the express purpose of influencing consumers to buy the competitor's goods or services is actionable under section 43(a).") (emphasis added).

No court, since the enactment of the 1988 amendment to section 43(a), has addressed whether the timing of the release of the statements (i.e., whether they are made before, versus after, the purchase of the product) affects whether they can be considered advertising, and thus within the purview of section 43(a). One court, however, did address this question before the enactment of the amendment. In Marcyan v. Nissen Corp., 578 F. Supp. 485 (N.D.Ind. 1982), aff'd 725 F.2d 687 (7th Cir. 1983), the court considered whether a statement in an instruction manual was actionable.

This manual is not advertising material, nor is it distributed to the general public for the purpose of promoting plaintiffs' products: it is a user's manual and is provided to a purchaser of defendants' equipment together with the equipment in order to describe its proper use. . . .

. . . . .

. . . Furthermore, since the statement was made in a manual which, typically, would not be available until after the purchase was made . . . there is no likelihood of influencing the purchasing decision.

The court agrees with the conclusion of the Marcyan court. Advertising or promotion implies that the statements are made to influence a consumer in his or her choice to purchase a product. Statements made inside the product's packaging, available to consumers only after the purchase has been made, do not affect the choice to purchase, that choice having been made at an earlier point. The court thus concludes that Norelco's package inserts are not "commercial advertising or promotion" as that phrase is used in section 43(a).

b. Closeness Claim Made in Mass. Mailing

Superior closeness is mentioned twice in Norelco's mass mailing. The two statements come after Norelco has specifically explained the necessity of an acclimation period. The mailing reads:

Keep in mind that, like any finely-tuned machine, [the Reflex Action] requires a short break-in period. You've trained your beard by the way you shave it, but all it takes is three weeks to retrain it. During that time, your face will gradually adjust to the new Reflex Action shaving system, until, finally, you will be enjoying a smoother, closer, much more comfortable shave every time. . . . So stop suffering — start shaving closer and more comfortably than ever before with the new Norelco Reflex Action razor.

Gillette Exhibit No. 59 (emphasis added).

Gillette argues that Norelco has no support for this claim. This is true enough; but the burden is upon Gillette to establish the falsity of the claim, not upon Norelco to prove the truth of the claim. See Quaker State, 977 F.2d at 63. Gillette has proffered studies which, it says, demonstrate that the Reflex Action does not shave with superior closeness.

The first study is the previously-discussed, in-plant study among Gillette's employees. This study, in addition to rating perceptual measures of irritation, also took perceptual measurements of closeness. The result indicated that the Gillette SensorExcel gave a closer shave. The unreliability of the Gillette in-plant study has been discussed in connection with the claims made by Norelco regarding irritation. The study fails as a basis for Gillette's contention that Norelco's closeness claim is false for the same reason that it fails as a basis for disproving Norelco's irritation claims.

Gillette's second closeness study was a five-day study of a panel of fourteen men, ages twenty-one to sixty-five years, conducted by the Gillette Research Institute ("GRI"). Each participant was asked to shave one half of his face with the SensorExcel and the other half with the Reflex Action. A perceptual test consisting of visual and tactile inspections by the participants was measured at four points: immediately after the shave, one hour after the shave, two hours after the shave and four hours after the shave. The perceptual testing was conducted on each of four days of the study. An objective test was conducted on the second day. Approximately twenty-four hours after each participant shaved his face, the length of the beard on each side of the face was measured using a videomicroscope. This procedure was repeated on days three through five of the study. The data from the four days of the perceptual testing were averaged as to each participant, for each shaver and for each of the four points in time, and then averaged across participants for each shaver and for each of the four points in time. The results of the objective test were averaged as to each participant for each shaver on the basis of cheek sites, neck sites and combined sites. These data were then averaged across participants for each product per site. The averaged results for both the perceptual and objective tests favored the SensorExcel in terms of the closeness of the shave.

Norelco attacks the second closeness study on a number of grounds.

First Norelco contends that this second closeness study did not exclude Gillette's employees. That is not quite accurate, but in making the point, Norelco does raise questions about the pool from which the test participants were chosen. The participants in the second closeness study were not Gillette employees. They were, however, members of a larger pool of panelists that GRI regularly uses for closeness studies. Affirmation of Thomas L. Gallerani, ¶ 72. It is unclear from the record whether these test participants are paid, but it is reasonable to assume that they are, and that they were aware that they were involved in testing on Gillette's behalf. Furthermore, it is unclear whether the process by which the participants were selected was such that it excluded those who were knowledgeable of this litigation and those who had seen Norelco's advertisements.

Norelco also contends that the universe of Gillette shaving instruments was limited so that the participants in the study were men who preferred Gillette's SensorExcel. Again, the criticism is not quite accurate. It does not appear from the testing protocol that only men who prefer the SensorExcel were used as test subjects. Gillette Exhibit No. 48. It is true, however, that the Reflex Action was tested only against the SensorExcel. Affirmation of Thomas L. Gallerani, ¶ 73.

The most serious criticism by Norelco of the GRI closeness study — and the one that is most persuasive to the court — is that there was no acclimation period for the Reflex Action. The study lasted only one workweek, and Gillette took measurements of closeness after just one shave with the Reflex Action. Indeed, little or no acclimation was possible in this study because, it appears that the study required that the participants switch the side of the face on which each shaver was used, on each successive day. Gillette Exhibit No. 48. ("On one side of the face was the test product and the other side was the control product. The products were presented in a random balanced scheme with daily crossover.") Gillette Exhibit No. 48 at 1 (emphasis added); Affirmation of Thomas L. Gallerani, ¶ 73 (same).

Insofar as its closeness claim is concerned, Norelco has specifically qualified the claim in its advertisements, with a notation about the acclimation period. In light of this explicit limitation, the GRI study, which measures closeness in one week only, does not suffice to demonstrate the falsity of Norelco's claim that the Reflex Action shaves closer than wet shavers after an acclimation period.

The court is concerned about the absence of any evidence from Norelco to support its closeness claim. To be sure, as the court has noted several times in this opinion, with respect to a non-establishment claim, Gillette bears the burden of disproof. Even so, Norelco would be wise, in preparation for further proceedings in this matter, to offer its own evidence as to the legitimacy of its closeness claim, and not merely to rest on the impeachment of Gillette's studies.

5. The "Tag Line" — "Anything Closer Could Be Too Close for Comfort"

Gillette next argues that the tag line, with which Norelco ends many of its print and television advertisements, is false or misleading in violation of section 43(a). Section 43(a), however, applies only to false or misleading statements of fact. See, e.g., Eventmedia Int'l Inc., 1992 WL 321629, at *4 (statement not actionable because "[i]t is a statement of opinion").

The law concerning whether a statement is one of fact, rather than a statement of opinion, has been discussed explicitly in the law of defamation. In deciding whether Norelco's tag line is a statement of fact or opinion, the court therefore turns to defamation jurisprudence. The District of Columbia Circuit has set forth the leading test for whether a statement is fact or opinion. The test lays out four steps: "first . . . analyze the common usage or meaning of the allegedly defamatory words themselves. . . . [S]econdly, consider the degree to which the statements are verifiable — is the statement objectively capable of proof or disproof? . . . [T]hirdly, examine the context in which the statement occurs. . . . [E]xamine, finally, the broader social context into which the statement fits." Ollman v. Evans, 750 F.2d 970, 979-83 (D.C. Cir. 1984) (en banc), cert. denied, 471 U.S. 1127, 105 S.Ct. 2662, 86 L.Ed.2d 278 (1985). The Supreme Court has rejected a constitutional protection for all statements of opinion, see Milkovich v. Lorain Journal Co., 497 U.S. 1, 110 S.Ct. 2695, 111 L.Ed.2d 1 (1990), but has done so, at least partially, because in some situations "expressions of `opinion' may often imply an assertion of objective fact." Id. at 18, 110 S.Ct. at 2705. In holding the statement in Milkovich actionable, the Court came close to applying the four considerations of Ollman. See Milkovich, 497 U.S. at 21, 110 S.Ct. at 2707 (discussing common interpretation of language, general tenor of article and verifiability of implication). Important in the Supreme Court's analysis was whether "the connotation . . . is sufficiently factual to be susceptible of being proved true or false." Id. The lesson to be drawn from Ollman and Milkovich is that capacity for verification is the most important question in determining whether a statement is one of fact.

It is difficult to imagine how one might verify whether something "could be too close for comfort." What are the standards by which "could be too close for comfort" objectively can be measured for truth? And even as to the subjective, there seem to be no criteria by which a generalization could be made about the phrase. Gillette has not explained in what respect the statement is false. Nor has Gillette offered standards of either objective or subjective verification of the phrase. That comes as no surprise, because the conditional "could" is denotative of only a possibility; and things that are possible can occur, but they may not.

All of this leads to the conclusion that the tag line is at worst a statement of opinion or of innocuous puffery, not actionable as a false statement under section 43(a).

Gillette also argues that the tag line is misleading. It offers, in support of the argument, findings from the Jacoby Study to the effect that 36% of those surveyed believed that the tag line conveyed a specific product benefit, and that 19% thought if another shaver gave a closer shave, that shaver would cut or tear the skin. As to the first point, Gillette has not shown how the tag line misleads, even if it does convey a specific product benefit. Norelco asserts — and the court has already decided that Gillette is not likely to be successful in disproving — that the Reflex Action (after the acclimation period) is a less irritating shaving system than a wet shaver. That indeed is a specific product benefit. As to the latter point, the court has already discussed the commonly-experienced manifestations of irritation associated with wet-shaving. That discussion suffices to lay to rest any argument that the tag line misleads as to the question of the capacity of wet shavers for irritation.

The court determines therefore that Gillette is not likely to succeed in demonstrating, on the merits, that the tag line violates section 43(a).

6. The Celebrity Study

Finally, Gillette challenges the appropriateness of any reliance by Norelco in its advertisements on a so-called Celebrity Study conducted by Ketchum Public Relations ("Ketchum"). Norelco's representations, based on the Celebrity Study are a kind of establishment claim, in that Norelco asserts that a certain percentage of celebrities who took a "twenty-one day test drive" with the Reflex Action (i.e., used the Reflex Action for twenty-one days) now favor that product over wet shavers. Gillette therefore need only demonstrate a likelihood of proving that the Celebrity Study is "not sufficiently reliable to permit a conclusion," Castrol, 977 F.2d at 63.

There is hardly a question that the Celebrity Study is not "scientific." Its participants are a random selection of celebrities, including athletes, media personalities and "movers and shakers." W. Graham Hueber, assistant director of research at Ketchum, designed and coordinated the Celebrity Study and prepared the report which embodies its results. In deposition testimony, given in connection with this case, Mr. Hueber pointed out that the Celebrity Study could only be used in very limited circumstances, because it was not a generalizable, representative study. "This is not a representative sample of the people under study, so you cannot project these numbers to all professional athletes, all people in the media, all business leaders, all Ketchum employees." Deposition of W. Graham Hueber, October 31, 1996, at 77, Exhibit Q to Declaration of Michael H. Steinberg.

Moreover, the study contains an inherent bias which, in the view of the court, makes it unreliable, even as a measure of the reactions of those actually tested. First, included with the packet of materials distributed to each test subject was a letter on Norelco stationery. The letter said:

Welcome to the Norelco Reflex Action 21-Day Test Drive Team. You are one of an elite group of opinion leaders selected to give the Norelco Reflex Action a test drive before it is available to the general public. You will be among the first to use the latest in shaving technology. . . . After 21 days of use, Reflex Action promises to give you a closer, more comfortable, more convenient shaving option.

Gillette Exhibit No. 22 at N00027 (boldface in original). The letter thus discloses that Norelco is the sponsor of the study and sets out Norelco's expectations for the results of the study.

At the outset, then, a bias in favor of Norelco is injected. That bias is insinuated into the text of the survey. For example, one question reads: "After week two you really start to see the difference, as your skin and beard adjust to this new way of shaving! How does the performance of the Reflex Action compare now to your blade shave?" Id. at N00034 (emphasis in original). Finally, as Gillette notes, Norelco did not screen to exclude those participants who may have had an eye toward a possible endorsement contract with Norelco.

Given the biasing of the Celebrity Study, it is of virtually no value as support for Norelco's claim that a large percentage of celebrities, after using the Reflex Action for twenty-one days, preferred it over their previously-used wet shavers. Gillette, therefore, is likely to succeed in its challenge to any claims based on the Celebrity Study.

B. Consideration of the Other Factors Necessary for Preliminary Injunctive Relief

Gillette has demonstrated a likelihood of success in proving that there are two aspects of Norelco's overall advertising for the Reflex Action that violate section 43(a). Specifically, the court has found that Gillette is likely to prove that all claims by Norelco that the Reflex Action shaves with less irritation than blades are misleading, to the extent that they omit reference to the acclimation period. The court has found also that Gillette is likely to succeed in proving that Norelco's claims based on the Celebrity Study are false, because the Celebrity Study is unreliable and thus affords no justification for the claims made.

In the case of the ruling on the Celebrity Study, the court need not consider the other factors necessary for the grant of a preliminary injunction. With respect to that issue, the court has found that Gillette is likely to succeed in establishing the literal falsity of Norelco's claims. For that reason, Gillette is relieved of the obligation to show that its application for injunctive relief meets the standards of irreparability of harm, favorable balance of the hardships of injunctive relief and benign — or at least no malevolent — effect of injunctive relief on the public interest. Associated Dry Goods, 799 F.2d at 16.

In the case of the finding that Gillette is likely to succeed in establishing that Norelco's advertisements are misleading, in the absence of a reference to the acclimation period, the situation is different. As to that issue, Gillette must meet all the customary standards for injunctive relief. Id. at 12. Gillette has shown that it can meet those standards.

As to the questions of irreparability of harm and the balance of harms, the court begins with the proposition, well-settled in the First Circuit, that any harm to the defendant from the granting of an injunction is to be discounted by the degree that the plaintiff can show likelihood of success. EEOC v. Astra USA, Inc. 94 F.3d 738, 743 (1st Cir. 1996); Securities and Exchange Commission v. World Radio Mission, Inc., 544 F.2d 535, 541 (1st Cir. 1976). Here, Gillette has made a very strong showing of its likelihood of success in proving misleading those claims, by Norelco, of superiority as to irritation that fail to include reference to the acclimation period. The evidence is substantial that Norelco knew that acclimation is critical to the success of the Reflex Action, but resolutely omitted mention of it in much of its advertising.

To the scale, already tipped in favor of Gillette by its strong case, one must add a consideration of the nature and objective of Norelco's advertising campaign. It was Norelco's avowed purpose to convert as many men as possible, twenty-five and older, from wet shavers to the Reflex Action and wrest from the wet-shaving industry as much of the market for shaving products as possible. Gillette, as the world leader in sales of wet-shaving products, Affirmation of Sharon Keith, ¶ 10, was an obvious target of Norelco's offensive. To that end, Norelco mounted its largest-ever promotion, an acknowledged militant crusade on all advertising fronts. What Gillette stands to suffer from Norelco's campaign, in the absence of some relief, are an irreparable loss of sales to consumers, otherwise loyal to wet-shaving technology, and irreparable diminution of its good name and standing in the shaving products industry. What Norelco stands to lose if an injunction is granted is a share of the market that it never had, but seeks to obtain. Taking into account Gillette's strong case and its probable hardship should an injunction not issue, the court finds that the balance of harms weighs in favor of Gillette.

As to the question of how the public interest will be affected should an injunction issue, all that need be said is that consumers are obviously better served if advertisers are made to refrain from misleading them. "The public has a right not to be deceived or confused." Church Dwight Co. v. S.C. Johnson Son, 873 F. Supp. 893, 912 (D.N.J. 1994) (quoting W.L. Gore Assoc., Inc. v. Totes, Inc., 788 F. Supp. 800, 813 (D.Del. 1992)).

C. Conclusions as to Injunctive Relief

Accordingly, an injunction will issue. It will not be the broad injunction sought by Gillette, but a much narrower one, fashioned to address the precise wrongs that Gillette is likely to prove at trial. The injunction will prohibit Norelco from making any claim in its advertising that the Reflex Action shaves with less irritation unless Norelco also includes, in plainly perceptible language, a statement to the effect that less irritation from the Reflex Action will be experienced only after an acclimation period of at least twenty-one days. The injunction will also prohibit Norelco from making any claim concerning the Reflex Action, based on the Celebrity Study.

D. Bond

Rule 65(c) of the Federal Rules of Civil Procedure provides: "No . . . preliminary injunction shall issue except upon the giving of security by the applicant, in such sum as the court deems proper, for the payment of such costs and damages as may be incurred or suffered by any party who is found to have been wrongfully enjoined or restrained." The purpose of a bond is to "cover `costs and damages' due directly to the injunction itself; the bond does not provide a means of recovery for the party who happens to succeed in the underlying action." Flag Fables, Inc. v. Jean Ann's Country Flags and Crafts, Inc., 753 F. Supp. 1007, 1019 (D.Mass. 1990). But the posting of a bond "is not a jurisdictional prerequisite to the validity of a preliminary injunction." Aoude v. Mobil Oil Corp., 862 F.2d 890, 896 (1st Cir. 1988).

Given these standards, the court believes that a bond should be posted in a sufficient amount to cover the costs of the changes Norelco must undertake in its advertising campaign, as a consequence of the injunction issued today, and any damages resulting from the interruption in the advertising campaign caused by the necessity to make those changes. Norelco has asked that a bond of $20 million be ordered. That request was premised upon the issuance of the broad injunction sought by Gillette — an injunction that would have precluded Norelco from making any superiority claims for the Reflex Action as to irritation and closeness, in any form. The evidence now before the court is insufficient to guide it in determining the size of the bond appropriate for the actual injunction issued. Accordingly, the court will order that Norelco submit by affidavit, on or before December 6, 1996, such evidence as it considers appropriate to the determination of the size of any bond to be ordered. Gillette may respond to Norelco's submission on or before December 11, 1996. The injunction, however, will be effective immediately.

II. The Injunction

After consideration of Gillette's application for a preliminary injunction in this matter and for the foregoing reasons the court orders as follows.

1. The application of the plaintiff, The Gillette Company ("Gillette"), for a preliminary injunction is hereby GRANTED in part and DENIED in part.

2. The defendant, Norelco Consumer Products Company ("Norelco"), its officers, agents, servants, employees, and attorneys, pending a final determination of the issues raised in this litigation, are enjoined and restrained from making any statement or claim, in any advertisement or promotion, to the effect that the Reflex Action shaver (the "Reflex Action") shaves with less irritation than wet shavers, without also including a sufficiently conspicuous and contemporaneous statement to the effect that a user of the Reflex Action should not expect to experience less irritation until he or she uses the Reflex Action for a twenty-one day (or three-week) period of acclimation (the "disclaimer").

(a) In printed advertisements, the disclaimer shall be displayed in type size at least as large as that in which the principal portion of the advertisement appears.

(b) In television advertisements, the disclaimer may be presented in visual form only, or orally only, or both visually and orally. Any visual-only presentation, however, shall be made at the same time as the presentation of the superiority claim and shall be of sufficient size and duration that there can be a reasonable expectation that it will be noticed.

(c) In radio advertisements, the disclaimer shall be made contemporaneously and directly in connection with the claim of superiority.

3. Norelco, its officers, agents, servants, employees, and attorneys, pending a final determination of the issues raised in this litigation, are enjoined and restrained from making any statement or claim, in any of its advertisements or promotions, to the effect that the Reflex Action shaves with less irritation than wet shavers "starting from day one."

4. Norelco, its officers, agents, servants, employees, and attorneys, pending a final determination of the issues raised in this litigation, are enjoined and restrained from making any statement or claim in any of its advertisements or promotions based upon the so-called Celebrity Study. This limitation shall not extend to individualized product endorsements, irrespective of whether the individual endorser was identified through the Celebrity Study.

5. Norelco shall file with the court and serve on The Gillette Company ("Gillette"), on or before December 6, 1996, an affidavit setting forth any evidence deemed by Norelco to be appropriate for the court's consideration in determining the size of any bond to be ordered in connection with the injunction hereby issued. Gillette may respond to Norelco's submission or before December 11, 1996. Such response may include any argument by Gillette, with supporting authority, that no bond need be ordered.

6. The preliminary injunction hereby issued shall be effective immediately. Norelco shall take immediate steps to remove from all advertising media any advertising for the Reflex Action that is inconsistent with the terms of this order. Nothing in this order is intended, however, to require Norelco to recall any advertisements the distribution, dissemination or circulation of which is now beyond the control of Norelco. Norelco shall file with the court, on or before December 6, 1996, a statement setting forth the steps it has taken to comply with the terms of this order, including the terms of this paragraph 6.

SO ORDERED.