LIBMAN COMPANY, Plaintiff–Appellee, v. VINING INDUSTRIES, INCORPORATED, Defendant–Appellant.
The Libman Company brought suit against Vining Industries for infringement of a federally registered trademark on a broom. After a bench trial, the district judge enjoined Vining from selling the infringing line of brooms and in addition awarded Libman almost $1.2 million in monetary relief, representing Vining's profits from that sale. The main ground of the appeal, and the only one we need discuss, is that the district judge committed clear error in finding that consumers were likely to mistake Vining's broom for Libman's.
A broom has, of course, a head of bristles. In 1993, after being twice turned down, Libman succeeded in registering with the U.S. Patent and Trademark Office a trademark that consists of a color scheme in which one vertical band or segment of bristles is a different color (a “contrasting” color, in the language of the trademark registration) from the remaining bristles. The particular choice of contrasting colors is not part of the trademark, however. The contrasting-color band was sometimes red and sometimes green or black, the rest of the bristles being either a very dark gray, verging on black, with the red band, or a lighter gray with the green or black band. Libman had begun marketing these brooms in 1990. They sold well. In 1993 Vining began marketing its own contrasting-color broom, the contrasting colors being light and medium gray.
The parties agree that Libman cannot prove infringement of its trademark without proving that consumers of brooms are likely to be confused about the source of Vining's brooms—and to think that they are Libman's. 15 U.S.C. § 1114(1)(b); AHP Subsidiary Holding Co. v. Stuart Hale Co., 1 F.3d 611, 615 (7th Cir.1993). A trademark is not a property right, but an identifier; so, provided no one is likely to be confused by the alleged infringer, there is no impairment of the interest that the trademark statute protects. The evidence of likelihood of confusion in this case is vanishingly thin. Vining sold several hundred thousand of the allegedly infringing brooms, yet there is no evidence that any consumer ever made such an error; if confusion were likely, one would expect at least one person out of this vast multitude to be confused, or more precisely one would expect Libman to have been able to find one such confused person. Nor was any survey conducted in an effort to determine the likelihood of confusion. The district court pointed out, moreover, that “from a distance, with the cardboard cover in place, the [Vining] broom doesn't have the appearance of a Libman broom.” The head of each broom is sold with a plastic wrapper around it, but the opaque label affixed to Vining's wrapper, unlike the label on Libman's wrapper, is so large that it hides the contrasting colors of the bristles. It does not hide them completely, but, especially since they are merely different shades of gray, you have to peer pretty closely to notice this feature of the broom. The labels are not similar and of course the brand names are different. Consistent with the different style of packaging, Libman's advertising (at least the advertising that is in the record, but that is all we have to go on) shows the undressed broom, its contrasting colors boldly displayed. Vining's does not.
But “with their covers removed,” the judge went on to say, “the two brooms are quite similar in appearance.” The photographs in the record do not support this characterization, as the only thing the brooms have in common, besides being brooms, is that their bristles are in contrasting color bands rather than being all of one color. But this is an element of similarity and we shall not quibble over the district judge's use of the word “quite.” The brooms, however, are sold in their wrappers. There is no evidence that the wrapper is ever removed before a sale to the consumer. (There is evidence that at trade shows the wrapper is sometimes removed, but those are promotions to the trade, not to consumers.) A consumer who is curious about the strength or stiffness or other tactile properties of the bristles can feel them through the plastic wrapper, which is very thin; she would have no occasion to ask the salesman to remove the wrapper.
So why is the undressed state of the broom relevant? Because the consumer might, upon removing the opaque wrapper from the Vining broom when she brought the broom home, think that she had bought a Libman rather than a Vining. Here is how her confusion might hurt Libman: The two brands of broom, though sold through the same type of outlet (supermarkets and mass-market retailers), are rarely sold by the same outlet, the reason being that retailers prefer to stock only one brand of broom. It's a cheap but bulky item; they don't want to fill up the store with different brands of it. Because a broom is so cheap (under $10), consumers don't spend a lot of time mulling over their decision whether to buy. If their old broom is wearing out, which usually happens after a year or so, they'll look for a new one the next time they find themselves in a store that sells brooms. Suppose that in 1993 you buy a Libman broom. You like it; you think it's a great broom; and you associate the contrasting color bands with the name “Libman.” Eventually the broom wears out and you have to buy a new one. You find yourself in a store that does not stock the Libman broom (you don't know this), but only the Vining broom. If you saw them side to side you would know that the Libman, and not the Vining, broom was the one you had had a good experience with. But you don't see them side to side. All you see is a broom that has contrasting color bands. You think it's a Libman broom, and buy it. Had you known it was not a Libman broom you would have waited to replace your old broom until you found yourself in a store that stocked the Libman broom.
This is a plausible narrative, one consistent not only with the objectives of trademark law but also with a large number of cases which hold that where, as in this case, the public does not encounter the parties' trademarks together, the existence of minor differences that would clearly distinguish them in a side-by-side comparison does not refute an inference of likely confusion. E.g., International Kennel Club v. Mighty Star, Inc., 846 F.2d 1079, 1088 (7th Cir.1988); James Burrough Ltd. v. Sign of the Beefeater, Inc., 540 F.2d 266, 275 (7th Cir.1976); Sun–Fun Products, Inc. v. Suntan Research & Development Inc., 656 F.2d 186, 189 (5th Cir.1981); see 3 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition ¶ 23.17 (3d ed. 1995). As we put it years ago in a case involving an alleged confusion between “Dramamine” and “Bonamine,” “we must determine the purchasing public's state of mind when confronted by somewhat similar trade names singly presented.” G.D. Searle & Co. v. Chas. Pfizer & Co., 265 F.2d 385, 388 (7th Cir.1959) (emphasis in original).
It would have been nice had the district judge mentioned this theory of confusion rather than just moving without transition or explanation from the fact that the parties' brooms are not confusingly similar when seen side by side to the conclusion that one infringed the trademark of the other. Libman did not object when at argument we judges looked at the brooms side by side and remarked their dissimilar appearance. It further muddied the waters at the argument by describing this case, as its brief had done as well, as a case of “reverse confusion.” The term refers in trademark law to an entirely different practice, that of the alleged infringer's so saturating the market with advertising for his trademark that the public comes to believe that he, rather than the plaintiff, is the source of the plaintiff's trademarked product. Sands, Taylor & Wood Co. v. Quaker Oats Co., 978 F.2d 947, 957 (7th Cir.1992); 3 McCarthy, supra, ¶ 23.01.
In view of the conceptual confusion it comes as no great surprise that Libman neglected to present evidence to support its misnamed theory that satisfied consumers of its broom might be fooled when they shopped for a replacement in a store that sold only Vining's broom. To insist on evidence might seem to be to commit the error of thinking that proof of actual confusion is required in a trademark-infringement case, and of course it is not unless damages are sought. Web Printing Controls Co. v. Oxy–Dry Corp., 906 F.2d 1202, 1204–05 (7th Cir.1990); Resource Developers, Inc. v. Statue of Liberty–Ellis Island Foundation, Inc., 926 F.2d 134, 139 (2d Cir.1991). Libman sought injunctive relief and Vining's profits, not damages. Taco Cabana Int'l, Inc. v. Two Pesos, Inc., 932 F.2d 1113, 1126 (5th Cir.1991), aff'd on other grounds, 505 U.S. 763, 112 S.Ct. 2753, 120 L.Ed.2d 615 (1992). But our point is only that a finding of likely confusion can no more be based on pure conjecture or a fetching narrative alone than any other finding on an issue on which the proponent bears the burden of proof.
To this Libman might reply that deliberate copying (“bad faith”) is one of the factors that courts rely on in determining the likelihood of confusion. It is. Computer Care v. Service Systems Enterprises, Inc., 982 F.2d 1063, 1070 (7th Cir.1992); Paddington Corp. v. Attiki Importers & Distributors, Inc., 996 F.2d 577, 586–87 (2d Cir.1993). But in context, “bad faith” in the district judge's opinion does not appear to mean that Vining was trying to confuse consumers. So far as appears—and it is all that the record supports—Vining noticed that Libman's brooms were selling briskly, inferred that consumers like brooms with contrasting color bands, and decided to climb on the bandwagon. We call that competition, not bad faith, provided there is no intention to confuse, and, so far as appears, there was none. Cf. M–F–G Corp. v. EMRA Corp., 817 F.2d 410, 412 (7th Cir.1987). The distinction between copying “features whose value to consumers is intrinsic and not exclusively as a signifier of source” and features whose only significance is in identifying the source of the product is emphasized in Thomas & Betts Corp. v. Panduit Corp., 65 F.3d 654, 663 (7th Cir.1995). See also Duraco Products, Inc. v. Joy Plastic Enterprises, Ltd., 40 F.3d 1431, 1453 (3d Cir.1994). The line is a fine one. W.T. Rogers Co. v. Keene, 778 F.2d 334, 339–40 (7th Cir.1985). But there is neither evidence nor a finding (as we read the district judge's opinion) that it was deliberately crossed.
The two brooms are not packaged alike and do not have similar names; while the “undressed” brooms are similar, they are not identical; the appearance of the brooms in advertising was dissimilar; and the trademark is a thin one, since adding a colored stripe is hardly a distinctive way of marking a product. We now know, of course, that a color can be a trademark. Qualitex Co. v. Jacobson Products Co., 514 U.S. 159, 115 S.Ct. 1300, 131 L.Ed.2d 248 (1995). And we do not hold that Libman's contrasting-color trademark was insufficiently distinctive to be registrable. Cf. Application of Data Packaging Corp., 453 F.2d 1300 (C.C.P.A.1972). (Vining argues this, but we need not and do not reach the issue.) Still, we're in the gray area (pardon the pun) where a rather commonplace design used for a trademark may be taken by consumers as a form of decoration—a way, here, of jazzing up the humblest of utilitarian products. 1 McCarthy, supra, ¶ 7.17, p. 7–77. We pointed out that Vining may have taken it so in deciding to produce a similar broom.
We do not want to make a fetish of testimony, expert or otherwise. Sometimes it is obvious just from comparing the products that consumers are likely to be confused as to their source. But this is not such a case. If the record were limited to the brooms themselves and the advertising for them, no reasonable person would think that there was a substantial danger of confusion. We take the district judge to have acknowledged this in the passage we quoted earlier from his opinion. Libman's narrative of possible confusion cannot be regarded as better than a hypothesis, and a hypothesis that has not been tested. It should not have been very hard for Libman to find some satisfied owners of its brooms and confront them with the Vining broom and see whether they thought it was the same brand of broom. Without such evidence it would be pure speculation to conclude that anyone, let alone a significant fraction of the broom-buying public, could have been misled into believing that the Vining broom and the Libman broom were one and the same brand. Restatement (Third) of Unfair Competition § 20, comment g (1995).
Either consumers are confused at the point of sale, or they become confused later and this carries over to the next time they are in the market for the product. The district court rightly disparaged the first theory and failed to discuss the second. The evidence supports neither, and the judgment for the plaintiff must therefore be reversed with instructions to enter judgment for the defendant.
I am unable to join in the majority opinion because I believe it is contrary to the law of this circuit regarding “likelihood of confusion” and because I am concerned that my colleagues have disregarded the “clearly erroneous” standard of review that governs a district court's findings in this area. Given the centrality of the “likelihood of confusion” concept in the law of trademark infringement, I must respectfully dissent.
I. The Elements of “Likelihood of Confusion”
At the core of Libman's trademark infringement claim is the assertion that broom consumers are likely to be confused about the source of Vining's brooms, and are likely to think that Vining's brooms are Libman's. The brooms marketed by Vining used color-contrasted bristles in a fashion resembling the design registered by Libman as a protected trademark in 1993. The gist of Libman's argument is that satisfied customers of its broom could be misled by this similarity when shopping for a replacement broom in a store that carries only the Vining brand. At trial, Libman presented no evidence of actual confusion among consumers (i.e., no survey data or anecdotal evidence showing that consumers were in fact confused by Vining's use of color-contrasted broom bristles), nor was it required to do so by any of the cases in which this court has defined the elements of “likelihood of confusion.”
The majority opinion disclaims any intention to “make a fetish out of testimony, expert or otherwise” and it concedes that proof of actual confusion is not required in a trademark infringement case unless damages are sought. Nevertheless, my colleagues lay great emphasis on the absence of actual-confusion evidence in making their case that “[t]he evidence of likelihood of confusion in this case is vanishingly thin.” The majority opinion notes that:
Vining sold several hundred thousand of the allegedly infringing brooms, yet there is no evidence that any consumer ever made such an error; if confusion were likely, one would expect at least one person out of this vast multitude to be confused, or more precisely one would expect Libman to have been able to find one such confused person.
This analysis, in my view, departs from well-established precedent in this circuit, which holds that “[a] variety of factors may be material in assessing the likelihood of confusion” and that “[n]one of these factors by itself is dispositive of the likelihood of confusion question.” McGraw–Edison Co. v. Walt Disney Productions, 787 F.2d 1163, 1167 (7th Cir.1986) (citations omitted). The relevant factors, as they have been delineated by this court, are:
(1) similarity between the marks in appearance and suggestion;
(2) similarity of the products;
(3) area and manner of concurrent use;
(4) degree of care likely to be exercised by consumers;
(5) strength of complainant's mark;
(6) actual confusion; and,
(7) intent of defendant to ‘palm-off his product as that of another.’
Smith Fiberglass Prods. v. Ameron, Inc., 7 F.3d 1327, 1329 (7th Cir.1993); AHP Subsidiary Holding Co. v. Stuart Hale Co., 1 F.3d 611, 615 (7th Cir.1993); International Kennel Club, Inc. v. Mighty Star, Inc., 846 F.2d 1079, 1087 (7th Cir.1988).
The majority's approach to the likelihood-of-confusion issue “sweeps under the rug” our well-established seven-factor test, and instead dwells on Libman's failure to produce evidence of actual confusion:
Without such evidence [of actual confusion] it would be pure speculation to conclude that anyone, let alone a significant fraction of the broom-buying public, could have been misled into believing that the Vining broom and the Libman broom were one and the same brand.
This fetishizing of actual confusion evidence flatly contradicts this court's oft-stated holding that “the plaintiff need not show actual confusion in order to establish likelihood of confusion.” Sands, Taylor & Wood Co. v. Quaker Oats Co., 978 F.2d 947, 960 (7th Cir.1992), cert. denied 507 U.S. 1042, 113 S.Ct. 1879, 123 L.Ed.2d 497 (1993) (citations omitted); Computer Care v. Service Systems Enterprises, Inc., 982 F.2d 1063, 1069 (7th Cir.1992) ( “actual confusion need not be demonstrated; the test is likelihood of confusion and actual confusion is only one element of the test”); Schwinn Bicycle Co. v. Ross Bicycles, Inc., 870 F.2d 1176, 1187 (7th Cir.1989) (“[T]he plaintiff need not prove every factor in order to prevail․ The weight and totality of the most important factors in each case will ultimately be determinative of the likelihood of confusion, not whether the majority of the factors tilt the scale in favor of one side or the other.”).
Obviously, evidence of actual confusion can be highly probative of the likelihood of confusion. For this reason, our cases recognize that such evidence, when available, is “entitled to substantial weight.” Helene Curtis Industries, Inc. v. Church & Dwight Co., 560 F.2d 1325 (7th Cir.1977), cert. denied 434 U.S. 1070, 98 S.Ct. 1252, 55 L.Ed.2d 772 (1978). Nevertheless, the nature of a particular product can make it unlikely that this kind of evidence will ever surface. Under such circumstances, as one commentator has observed, the absence of actual confusion evidence is not particularly significant:
[C]ourts generally acknowledge that a plaintiff need not prove actual confusion in order to establish trademark infringement. This ․ is especially true where the unavailability of such evidence is expected, for example, where the competing items are low-priced goods usually purchased on impulse and consumers would not be expected to complain about purchasing the wrong product․ In such instances, courts generally treat the actual confusion factor as a non-factor, with the absence of such evidence having no adverse impact on the plaintiff's ability to prove infringement.
It must [also] be remembered that in many instances the fact that a plaintiff has not adduced actual confusion evidence does not necessarily mean that actual confusion has not occurred.
Michael J. Allen, The Role of Actual Confusion Evidence in Federal Trademark Infringement Litigation, 16 Campbell L.Rev. 19, 23–25 (Winter 1994) (emphasis added). Libman introduced testimony, apparently undisputed by Vining, that brooms are low-cost items generally purchased on impulse. This testimony is “relevant to the likelihood of confusion because presumably consumers take less time purchasing low-cost items, and haste increases the possibility of confusion.” AHP Subsidiary, 1 F.3d at 616; Maxim's Ltd. v. Badonsky, 772 F.2d 388, 393 (7th Cir.1985) (risk of confusion is greater with low-cost items because purchasers are unlikely to complain when dissatisfied, thus bringing to light confusion).
As discussed in more detail below, it is the trial court, as trier of fact, and not the court of appeals, that evaluates and weighs the factors. Schwinn Bicycle, 870 F.2d at 1184 (“the district court has discretion to determine the evidentiary weight to be given any single factor.”). We have held that “[i]t is paramount for the trier of fact to look to the totality of the circumstances and evidence and then, after careful consideration and reflection, in its own sound judgement assess the likelihood of confusion.” Id. (emphasis added).
I am satisfied, as my colleagues apparently are not, that the district court engaged in a careful and reasoned weighing of the evidence. The court analyzed the factors as follows:
The two brooms are similarly designed [factor # 1], are similar products [factor # 2], and have the same sale outlets and purchasers [factor # 3]. The two products are advertised through the same media [factor # 3]. Libman has no evidence of actual confusion [factor # 6]. But as the witness Robert Libman pointed out, the purchase of a broom is not a large expense item and is not an event that is likely to cause a buyer to call up and complain about being confused about which product he or she bought [factor # 4]. There is some evidence in the record that purchasers recognize the Libman mark and occasionally call ․ asking where they may purchase the broom they saw advertised [factor # 5]․ While a market study of likelihood of confusion would have been helpful evidence to the court, the elements of a prima facie showing of likelihood of confusion are present and Vining has not rebutted that showing. The court finds, therefore, that it is more probably true than not that [Vining's] use of a contrast color band on its brooms is likely to cause confusion among ordinary purchasers as to the source of the brooms.
With respect to the question of Vining's bad faith (i.e., its intent to “palm-off”), the court found that “while the evidence is not strong ․ there was certainly no effort on its part to avoid infringement. In fact, at best it appears that [Vining] just didn't care.” [factor # 7].
The trial judge (as the trier of fact) thus considered each and every one of the required factors and concluded that a likelihood of confusion existed, notwithstanding the absence of actual confusion evidence. I am unable to join in the majority opinion because I believe that it elevates one of these factors—actual confusion—to a status that is clearly inconsistent with our precedent in this area. If the court wishes to embark on a new course and give actual confusion a more prominent role than the other analytical elements,1 it should do so explicitly.
II. Appellate Review of Likelihood–of–Confusion Determinations
“Whether or not there is a likelihood of confusion is a question of fact as to the probable or actual actions and reactions of prospective purchasers of the goods or services of the parties.” McGraw–Edison, 787 F.2d at 1167 (quotation omitted, emphasis added). We must therefore uphold a trial court's likelihood-of-confusion determination unless there has been a showing of clear error. Sunmark, Inc. v. Ocean Spray Cranberries, Inc., No. 95–1017, slip op. at 7 (7th Cir. August 29, 1995); Smith Fiberglass, 7 F.3d at 1329; AHP Subsidiary, 1 F.3d at 616; Forum Corp. of North America v. Forum, Ltd., 903 F.2d 434, 438 (7th Cir.1990); Scandia Down Corp. v. Euroquilt, Inc., 772 F.2d 1423, 1428 (7th Cir.1985), cert. denied 475 U.S. 1147, 106 S.Ct. 1801, 90 L.Ed.2d 346 (1986) (question of likelihood of confusion is “all fact and no law”).
The Supreme Court recently described the clearly erroneous standard as “significantly deferential,” noting that an appeals court may reverse the factual findings of a district court only if it is left with “a definite and firm conviction that a mistake has been committed.” Concrete Pipe & Prods., Inc. v. Construction Laborers Pension Trust, 508 U.S. 602, –––– – ––––, 113 S.Ct. 2264, 2279–80, 124 L.Ed.2d 539 (1993) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 541–42, 92 L.Ed. 746 (1948)).
I am concerned that the majority has not applied the “clearly erroneous” standard of review in this case, but rather has engaged in de novo fact finding. In one part of its opinion, the majority paraphrases Libman's argument concerning the likelihood of confusion, namely, that satisfied Libman customers will be confused as to the source of Vining's color-contrasted brooms when they encounter them in stores that carry only the Vining broom. The majority initially characterizes this as a “plausible narrative” (emphasis added). Elsewhere in the opinion, however, the majority contradicts itself by deriding Libman's likelihood-of-confusion argument as “pure conjecture,” “fetching narrative,” and “no better than a hypothesis.” The majority's ultimate conclusion, which appears to be based largely on its own inspection of the brooms side by side,2 is that “no reasonable person would think there was a substantial danger of confusion.”
In my considered opinion, the majority too easily “brushes aside” the Supreme Court's admonition in Anderson v. Bessemer City that a reviewing court may not reverse a factual finding simply because it would have decided the issue differently:
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.
470 U.S. 564, 573–74, 105 S.Ct. 1504, 1511–12, 84 L.Ed.2d 518 (1985) (emphasis added); see also Dugan v. United States, 18 F.3d 460, 463 (7th Cir.1994) (the function of the appeals court is not to reweigh evidence or to substitute its own judgement for that of the trier of fact). Clear error is the standard of review, as Anderson teaches, even when a district court's findings rest not on credibility determinations, but on physical or documentary evidence. Id. at 574, 105 S.Ct. at 1512; see also Fed.R.Civ.P. 52(a). It is important for us to apply the clearly erroneous standard properly and consistently when we are called upon to review factual findings. As Chief Judge Posner has himself reminded us, “district judges are specialists in finding facts, as we are not, and ․ our primary function, which is to maintain the uniformity and coherence of the law, is not engaged by a judgement so dependent on the specific circumstances of each case.” Brock v. TIC International Corp., 785 F.2d 168, 171 (7th Cir.1986) (emphasis added). When appellate courts intrude on the fact-finding role of the district courts, they undermine the legitimacy of trial courts in the eyes of litigants, multiply appeals by encouraging appellate re-trial of factual issues, and needlessly reallocate judicial authority. Fed.R.Civ.P. 52(a) Advisory Committee's Note (1985 amendment).
These observations apply with particular force in the area of trademark infringement, which is highly fact-specific. See Patricia J. Kaeding, Clearly Erroneous Review of Mixed Questions of Law and Fact: The Likelihood of Confusion Determination in Trademark Law, 59 U.Chi.L.Rev. 1291, 1309–1310 (Summer 1992) (deferential review promotes judicial accuracy and the efficient use of appellate resources because “fact-specific issues ․ can be determined with the same, or nearly the same, degree of consistency at the trial level.”). In the Scandia case, this court considered and rejected an appellant's argument that we should review de novo a district court's conclusion that two business logos were confusingly similar. We noted that:
The district judge is manager of the entire swirl of facts, and the clearly erroneous rule is based in substantial measure on a belief that because appellate courts never are in a better position than the district court [to evaluate factual issues], and often are in a worse one, a substitution of judgement would increase the randomness of the process without increasing accuracy over the run of cases. ‘Duplication of the trial judge's efforts in the court of appeals would very likely contribute only negligibly to the accuracy of fact determination at a huge cost in diversion of judicial resources.’
Scandia, 772 F.2d at 1428 (quoting Anderson ).
It seems to me that the majority, contrary to our own precedent, has evaluated the likelihood of confusion afresh and substituted its own judgement for that of the district court. My colleagues are too willing to revisit factual issues: not just the ultimate factual issue in this case (likelihood of confusion), but subsidiary findings on issues such as the strength of Libman's mark and the similarity of the broom designs. The district court found that Libman's use of contrasting colors was the “distinguishing feature” of its trademark and that this mark was, at a minimum, strong enough to be protectable under the trademark law. The majority apparently disagrees, and (oblivious to the large amount of money Libman spent to develop, market, and advertise its color-contrasted brooms) disparages the mark as “a thin one,” “a rather commonplace design,” and merely “a way ․ of jazzing up the humblest of utilitarian products.”
The majority also disagrees with the district court's finding that Vining's adoption of a color-contrasted design made its broom similar to Libman's, even though Vining used a different pair of contrasting colors. However, as the majority concedes (at least in one part of its opinion), this is a “plausible” finding. It is not as though the trial judge concluded that Lake Michigan and the Sahara Desert are confusingly similar because they are both large. In other words, although I happen to agree with the trial court's finding, my agreement or disagreement is beside the point. The court's clearly-stated finding of similarity is not so tenuous that it should leave any of the members of this panel with “a definite and firm conviction” that a mistake has been committed. The same is true of the district judge's ultimate determination that ordinary purchasers were likely to be confused as to the source of Vining's brooms.
The clearly erroneous standard, as Chief Judge Posner has observed, “requires us appellate judges to distinguish between the situation in which we think that if we had been the trier of fact we would have decided the case differently and the situation in which we are firmly convinced that we would have done so.” Carr v. Allison Gas Turbine Div., General Motors Corp., 32 F.3d 1007 (7th Cir.1994). Lacking the firm conviction required by the clearly erroneous standard, I must respectfully dissent.
In my view, the majority opinion is “likely to confuse” the law of trademark infringement in this circuit by giving too much analytical weight to the actual confusion factor and by encroaching on the fact-finding role of the district court. I therefore dissent and state for the record that I would affirm the district court.
1. See Edwin S. Clark, Finding Likelihood of Confusion with Actual Confusion: An Analysis of the Federal Courts' Approach, 22 Golden Gate U.L.Rev. 393 (Spring 1992) (recommending that courts rely more heavily on actual confusion and that they explicitly acknowledge the importance of this factor).
2. The value of such a comparison is questionable. Trademark infringement can exist when, as in this case, the goods or services at issue are not typically seen by consumers side by side. An allegedly infringing mark:must ․ be evaluated on the basis of its likely effect upon consumers who do not have [both marks before them] but who may have a general, vague, or even hazy, impression or recollection of the [plaintiff's] mark[ ] such as may trigger a belief that a relationship exists.James Burrough, Ltd. v. Sign of the Beefeater, Inc., 540 F.2d 266 (7th Cir.1976).
POSNER, Chief Judge.