dissenting.
I respectfully dissent. In my view, the trial court’s finding that there exists a likelihood of confusion is clearly erroneous. Moreover, the majority’s holding sanctions a violation of Novak’s first amendment rights. The T-shirts simply expressed a political message which irritated the officers of Mutual, who decided to swat this pesky fly buzzing around in their backyard with a sledge hammer (a federal court injunction). We should not be a parly to this effort.
In regard to the majority’s interpretation of the six SquirtCo factors indicating a likelihood of confusion: The Mutual trademark is no doubt a strong one. There also is a superficial “similarity” between Mutual’s mark and the Novak design. Clearly for a parody to be successful, some similarity must be present, but the differences here are unmistakable: an emaciated human replaces the virile “Indianhead” and “Mutant” replaces “Mutual.” Moreover, the products on which the design and mark are used are simply not “competitive,” indirectly or otherwise. Mutual sells insurance; Novak sells T-shirts — probably to peace activists. Further, Novak had no intent to pass the T-shirts as Mutual’s product.
The only SquirtCo factor really in dispute in this case is whether consumers could be “actually confused” as to who sponsored or produced the T-shirts.
*404The evidence on this issue consisted solely of a survey by a nationally recognized firm. Four hundred persons over the age of twenty-one were interviewed at shopping malls in New York, Denver, Chicago and San Francisco. Neither T-shirts nor insurance were featured at the shopping malls where the interviews were conducted. No one was asked whether he or she was interested in buying a T-shirt or insurance.
The interviewees were shown the T-shirts described in the majority opinion and asked: “Does anything come to your mind when you look at this tee shirt?” Ninety-two percent said yes. Plaintiffs Exhibit No. 49, p. 21. Those answering yes, were then asked: “What comes to your mind when you see this tee shirt?” Twenty-one percent answered Mutual of Omaha, or the TV program Wild Kingdom sponsored by Mutual. Plaintiff’s Exhibit No. 49, p. 22. Interviewees were then asked, “anything else,” and an additional eight percent then mentioned Mutual of Omaha. Plaintiff’s Exhibit No. 49, p. 22.
As recognized by the majority, the only possible evidence of actual confusion is contained in the following question: “Would you say that Mutual of Omaha goes along with or does not go along with these tee shirts in order to make people aware of the nuclear war problem?” Plaintiff’s Exhibit No. 49, p. 25. Twelve percent of those who mentioned Mutual of Omaha in the earlier questions gave an affirmative answer to this question.
This survey question, however, is fundamentally flawed and should be given little evidentiary weight. The question is blatantly suggestive. It plants the idea of nuclear war in the mind of the interviewee. Thus, an interviewee who only casually glances at the T-shirt or who doesn’t understand the message in the T-shirt is tipped off by the question that the T-shirt has something to do with nuclear war. But, if the interviewer had not tipped these people off about the message on the T-shirts, many would never have even come to the misconception that Mutual sponsored the message on the T-shirt.
The only class of interviewees that really matter in this case are those who, on their own, recognized the T-shirt contained a message about nuclear war and also believed that Mutual sponsored it. This survey gives us no firm data on how many of these interviewees there were, except that the percentage is somewhere from zero to twelve percent. That is not compelling evidence. See Jordache, 828 F.2d at 1487-88. It is at most an inference of the possibility of confusion, and such a showing is not sufficient to prove infringement. Jordache, 828 F.2d at 1482.
The survey suffers from additional flaws. The sample was miniscule and taken in areas that Novak’s message had not reached and had no realistic chance of reaching. His “mom and pop” operation reached a few thousand people, nearly all of whom lived in the Omaha area. Mutual, on the other hand, operates throughout the United States and sells nearly two billion dollars of insurance each year to millions of men and women.
There is not a whit of evidence that either Mutual’s sales or its image suffered to any degree, or more importantly that No-vak did or could do anything that would be likely to cause confusion among the millions of policyholders of Mutual or those in the pool likely to purchase insurance.1
There can be no doubt that Novak used the “Mutant of Omaha” design in a satiri*405cal manner. He intended to expose what he considered the folly of nuclear war. He did not intend to harm Mutual’s reputation, and there is no evidence that he did so or was likely to do so.
Most importantly, the majority opinion cannot be squared with the first amendment. The Supreme Court has consistently held that prior restraints on publication are presumptively invalid and may be sustained only under the most extraordinary circumstances. Lowe v. SEC, 472 U.S. 181, 204-05, 105 S.Ct. 2557, 2570-71, 86 L.Ed.2d 130 (1985).2 These are not extraordinary circumstances.
The majority justifies its assault on No-vak’s first amendment rights by stating that “protection afforded by the first amendment does not give Novak license to violate the [property] rights of Mutual where adequate alternative avenues of communication exist.” Unfortunately, neither the district court nor the majority identifies how Mutual’s property rights were harmed by the message on the T-shirts — its feelings, yes, but its reputation, highly unlikely.
Scholarly opinion rejects the loose definition of property damage taken by the majority. One commentator has said:
By extending protection to cases outside the existing legal framework and imposing liability for “satiric appropriation,” courts have stifled creativity without protecting any legally recognized rights. Courts must realize that satire is a special form of expression and evaluate it under established rules rather than according to personal reactions. Strict application of traditional rules of law will protect all parties. A good satire will likely offend someone. Good law, however, must allow the offense.
H. Dorsen, Satine Appropriation and the Law of Libel, Trademark, and Copyright: Remedies Without Wrongs, 65 B.U.L.Rev. 923, 964 (1985).
Scholarly opinion also rejects the notion that product parodists must use “adequate alternative avenues of communication.” One writer considers many of the cases relied on by the majority, see, e.g., Mutual of Omaha Ins. Co. v. Novak, 775 F.2d 247, 249 (8th Cir.1985); Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., 604 F.2d 200, 206 (2d Cir.1979); L.L. Bean, Inc. v. Drake Publications, Inc., 625 F.Supp. 1531, 1537-38 (D.Me.1986); Reddy Communications, Inc. v. Environmental Action Found., 477 F.Supp. 936 (D.D.C. 1979), and he cogently explains why “trademark parody” must be distinguished from other kinds of intrusions on another’s property:
[T]he courts have uncritically relied on the characterization of trademarks as a type of “private property,” leading them to the conclusion that, even if the first amendment does apply to trademark parody, the property rights of the trademark owner must take precedence over the parodist’s right of free expression because the latter can always communicate his ideas in ways that do not require the use of the trademark itself. Here the property metaphor, useful in other contexts, completely distorts the analysis, causing courts to hold, in essence, that the parodist’s rights must yield to property rights whenever the parodist has adequate alternative avenues of communication. This notion derives from the Supreme Court’s decision in Lloyd Corp. v. Tanner. There the Court held that the owner of real property has the right to exclude unwelcome speakers from that property. But in the Lloyd situation, the owner’s enforcement of its property rights restricts only the place where the speaker’s expression may occur. Giving precedence to the owner’s property rights comports with the doctrine that a restriction on the time, place, or manner of expression is valid so long as it leaves the speaker with an alternative means of delivering his message. Application of this rule is distinctly inappropriate, however, when the property *406involved is not real estate but a trademark — a form of intangible property that itself conveys or symbolizes ideas. Because a parodist expresses ideas through the use of another’s trademark, the owner’s attempt to enjoin the parody goes to the content of the speech and not merely to the time, place, or manner of its delivery.
Note, Trademark Parody: A Fair Use and First Amendment Analysis, 72 Va.L. Rev. 1079, 1111-1112 (1986) (footnotes omitted).
I agree with Professor Robert Denicola that the general approach to trademark parodies should be this:
If the injury to the plaintiff’s trademark arises from the ideas that have been expressed through the defendant’s use, it is no answer to cite the “shopping center” cases [allowing restrictions on time, place and manner] to justify suppression on the ground that “adequate alternative avenues of communication exist.” The issue is not where the defendant may speak, but rather what he may say. The first amendment will not permit the trademark owner the power to dictate the form, and thus the effectiveness, of another’s speech simply because his trademark has been used to express ideas that he would prefer to exclude from the public dialogue.
R. Denicola, Trademarks as Speech: Constitutional Implications of the Emerging Rationales for the Protection of Trade Symbols, 1982 Wis.L.Rev. 158, 206.
In sum, the evidence in this case shows at most a possibility of trademark confusion, no real evidence of actual harm to property, and, most importantly, a significant intrusion upon the defendant’s first amendment rights. For these reasons, I dissent.
. I simply cannot go along with the majority's acceptance of the district court's finding that "consumers are not likely to exercise the ‘high degree of care or thought’ necessary to ‘distinguish Mutual from the message on [Novak’s] products.” Supra at 401. As I stated previously, we do not know how many of the survey interviewees, on their own, recognized the message about nuclear war, and believed Mutual sponsored it. But, if there are some in this category, I simply cannot believe that any significant number would base an important decision such as buying insurance on their notion that Mutual took a stance on nuclear war. Cf. Jordache, 828 F.2d at 1487 (district court’s finding that customers are likely to exercise a "high degree of care in purchasing clothing that costs between fifteen and sixty dollars" is not clearly erroneous). Moreover, of those who would be influenced by the misconception, it is entirely likely that a fair number would view a stance against nuclear war quite favorably.
. See also J.T. McCarthy, Trademarks and Unfair Competition § 31.38 (2d ed. 1984); R. Deni-cola, Trademarks as Speech: Constitutional Implications of the Emerging Rationales for the Protection of Trade Symbols, 1982 Wis.L.Rev. 158, 202-207.