Whether "LA" is a protectible trademark for a low alcohol beer marketed by Anheuser-Busch Incorporated is the issue before us. Following Anheuser-Busch's introduction of its product, the Stroh Brewery Company introduced "Schaefer LA," also a low alcohol beer. An action to enjoin Stroh's use of LA followed. The district court, 587 F. Supp. 330,1 finding that the term LA was suggestive in that it required some imagination to connect it with the product and, accordingly, was a protectible trademark, issued a permanent injunction. On appeal, Stroh argues that as a matter of law LA is generic or descriptive in nature, since the term is comprised of the initials of the phrase "low alcohol." Stroh also argues that should the issue be determined to be a factual one, the findings of the district court are clearly erroneous. We conclude that whether LA is suggestive is a factual issue and that the findings of the district court are not clearly erroneous, and we affirm the judgment of the district court.

For several years Anheuser-Busch had been interested in producing a low alcohol beer both because of its need to develop new markets for growth and because of growing public concern over the health aspects of beer consumption and the problem of drunken driving. Other breweries had similar interests and, in fact, two smaller companies, Hudepohl Brewery and Christian Schmidt Brewing Company, had marketed reduced alcohol beers under the brand names of Pace and Break. Anheuser-Busch accelerated its plans in the fall of 1983, obtained approval from the Bureau of Alcohol, Tobacco and Firearms (BATF) for an early version of its LA label on November 21, 1983, and began using the LA mark on beer shipped in interstate commerce on December 5, 1983. On January 20, 1984, Anheuser-Busch issued a press release informing the industry and public of its plans to sell LA beer. On the same date, application was made for registration of the mark LA. In March, Anheuser-Busch commenced marketing the beer. At the time of the commencement of the trial on May 7, 1984, it was marketing the product in ten test markets preparatory to distribution on a national basis.

On March 5, 1984, Stroh issued a press release announcing its plans to introduce a low alcohol beer. On March 7, 1984, Stroh obtained BATF approval of two separate beer labels using LA, one with the product name Schaefer and the other with the product name Old Milwaukee. On March 26, 1984, after receiving objections from Anheuser-Busch, Stroh issued a press release stating that it planned to market Schaefer LA beer. This action, brought on April 18, 1984, against Stroh for trademark infringement, trademark dilution and unfair competition under 15 U.S.C. Sec. 1125(a) (1982), the Lanham Act, and for violation of the Missouri antidilution statute, Mo.Rev.Stat. Sec. 417.061 (1978), followed.

During the trial, marketing executives of both parties testified, as did Yoram (Jerry) Wind, a Professor of Marketing and Management at the Wharton School, University of Pennsylvania, who testified as to the results of marketing surveys he had conducted and Herbert E. Krugman, a consultant in public opinion and consumer research. There was testimony concerning low alcohol beers in Australia and the predominance in that market of Tooth LA beer. There was detailed evidence of Anheuser-Busch's studies concerning the new product, particularly those dealing with the selection of a name which stressed the desire of Anheuser-Busch to select a brand that would allow it to preempt the new market. Stroh representatives testified that they believed LA would become the generic name for this segment of the beer market. There were frequent references to the use of the words "Light" and "Lite" in the light beer market, the predominance in that market that the Miller Brewing Company had achieved, and Anheuser-Busch's admitted desire to emulate Miller with respect to LA in the low alcohol market. Insofar as other evidence must be considered in resolving the issues in the case, we will discuss it in detail at that time.

In its opinion, the district court began its analysis by recognizing the following principles:

A term for which trademark protection is claimed will fall in one of four categories: (1) generic, (2) descriptive, (3) suggestive, or (4) arbitrary or fanciful. These categories, like colors in a spectrum, tend to blur at the edges and merge together, making it difficult to apply the appropriate label. * * * The correct categorization of a given term is * * * a factual issue.

WSM, Inc. v. Hilton, 724 F.2d 1320, 1325-26 (8th Cir.1984) (citations omitted). After careful analysis, the court concluded that LA was not generic, descriptive, arbitrary or fanciful2 but was suggestive and therefore was protectible.3 The district court considered consumer perception of the term LA as demonstrated by public opinion surveys in reaching its conclusion. The district court then considered whether there had been infringement under the Lanham Act and Missouri's antidilution statute. It found that Anheuser-Busch was the first to appropriate and utilize the LA mark and that Stroh had appropriated LA in order to dilute its distinctiveness in the marketplace and create confusion between the products.

On appeal Stroh argues that the district court erred as a matter of law since the initials of a generic or descriptive term such as low alcohol cannot assume trademark status, since the court found itself that LA is a "category descriptor," since consumer perception was employed as the principal test for determining whether LA was generic or descriptive, and since Anheuser-Busch had the burden of proof to establish that it was entitled to a trademark for LA. Stroh makes other arguments, including ones based on a report to Anheuser-Busch from a consulting firm stating that LA would not be protectible as a trademark, and on a failure to ask in the opinion survey the question, "What, if anything, do the initials LA mean to you in relation to this product?" It also argues that the injunction is overly broad in that it extends to "any other acts which are likely to cause confusion, mistake or deception with plaintiff." The arguments made are sufficiently general to cause us to conclude that Stroh is also suggesting that the findings of the district court are clearly erroneous.

The threshold issue in this case is whether the district court's determination that LA is suggestive and therefore protectible is a factual finding to which the clearly erroneous standard must be applied or is erroneous as a matter of law. As we have held previously, the categorization of a term for which trademark protection is claimed is considered to be a factual issue, WSM, Inc., 724 F.2d at 1325-26, and thus is to be reviewed under the clearly erroneous standard of Fed.R.Civ.P. 52(a). Stroh, admitting as much, argues on appeal that nonetheless the finding of fact was predicated on an error of law not subject to the clearly erroneous standard.

Stroh first argues that the initials of a generic term such as "low alcohol" cannot assume trademark status, citing National Conference of Bar Examiners v. Multistate Legal Studies, 692 F.2d 478, 488 (7th Cir.1982) ("Under settled trademark law if the components of a trade name are common descriptive terms, a combination of such terms retains that quality. * * * Abbreviations for generic or common descriptive phrases must be treated similarly."), cert. denied, --- U.S. ----, 104 S. Ct. 69, 78 L. Ed. 2d 83 (1983). In National Conference, however, the plaintiffs sought to protect both the phrase "Multistate Bar Examination" and the initials "MBE." As the district court noted in its discussion of the case:

[National Conference ] is distinguishable from the facts here because plaintiff is not seeking to protect an accompanying phrase, such as low alcohol, light alcohol, less alcohol, light alternative, nor any of a number of word combinations for which the mark LA could stand. And, of course, because of the BATF regulations at the time of its initial label-making decision, plaintiff has not designated on the label what LA does stand for.4

Rather, the question in the instant case is whether, when a party seeks to protect initials alone which are also the initials of a generic or descriptive phrase, the initials are to be equated with that phrase. The key issue, as the cases demonstrate, is the nature of the relationship or tie between the initials and the generic or descriptive phrase and the relationship or tie between the initials and the product. As the district court properly held, if some operation of the imagination is required to connect the initials with the product, the initials cannot be equated with the generic phrase but are suggestive in nature, thereby rendering them protectible.

In Modern Optics v. Univis Lens Co., 234 F.2d 504, 43 C.C.P.A. 970, (1956), the court concluded that "CV" was a valid trademark even though it was composed of the initials of the words "continuous vision" which were alleged to be descriptive of trifocal lenses. As did the district court, we find the reasoning of Modern Optics persuasive:

[A]lthough we have considered appellant's contentions as to the descriptiveness of the words "continuous vision," it is not necessary to determine whether those words are merely descriptive of trifocal lenses, since appellee is not seeking registration of those words, but a mark whose dominant feature is the letters "CV." The letters "CV" are, of course, the initial letters of the words "continuous vision," and it is possible for initial letters to become so associated with descriptive words as to become descriptive themselves. * * * It does not follow, however, that all initials of combinations of descriptive words are ipso facto unregistrable. While each case must be decided on the basis of the particular facts involved, it would seem that, as a general rule, initials cannot be considered descriptive unless they have become so generally understood as representing descriptive words as to be accepted as substantially synonymous therewith.

Id. at 506 (citations omitted). The question before the Modern Optics court then became one of whether the record had established that the initials in question were a generally recognized term for the generic or descriptive term from which they were derived. Id. The district court correctly realized it faced the same question in its consideration of the meaning of the term LA.

The statement in National Conference which Stroh urges that we adopt as a rule of law cites as sole authority FS Services v. Custom Farm Services, 471 F.2d 671, 674 (7th Cir.1972). In that case, plaintiff was denied protection for the letters FS, where such were used in colors and motif totally dissimilar to its mark. Id. at 673. The specific rule stated in FS Services, however, is that "[a]bbreviations for generic terms where they are generally recognized must be treated similarly." Id. at 674 (emphasis added). As the court in Kampgrounds of America v. North Delaware A-OK Campground, 415 F. Supp. 1288, 1292 (D.Del.1976), pointed out, the FS Services court placed heavy "reliance on its determination that the abbreviation FS had itself become a descriptive or generic term representing the descriptive words farm service or farm supply." We believe that the sentence from National Conference upon which Stroh relies must similarly be so qualified.5 In addition, we do not believe that the presence of a very brief statement in a case from another circuit made in an entirely different factual context, which demanded no particular consideration of its import, should by its very presence cause its elevation to the status of an established rule of law.

Stroh cites as authority 3 R. Callmann, The Law of Unfair Competition, Trademarks, and Monopolies Sec. 18.23 (4th ed. 1983) and asserts that "the author states that numerals and letters are not valid trademarks if they are not used solely for the purpose of indicating origin." The section of the treatise on which the assertion is based discusses numerals and letters that either stand alone (normally as grade marks to indicate the nature, quality or size of a product or service) or are commonly understood abbreviations. The treatise discusses elsewhere the issue in the instant case and there states that the issue of the descriptiveness of an acronym "depends on whether the initials are recognized as equivalent to the full phrase." Id. Sec. 18.10. Stroh also refers us to some thirteen cases in support of their contention. We have examined these cases carefully and find that in none is it stated as a matter of law that initials of generic or descriptive phrases are unprotectible.6 We conclude that the district court did not err as a matter of law in engaging in a factual determination of the proper classification to be accorded the initials LA.

Stroh next argues that the district court erred as a matter of law in enjoining Stroh's use of the mark "Schaefer LA" after finding that LA itself was a "category descriptor." We do not believe, however, that the court so found. Stroh's assertion depends on the following passage from the district court's opinion:

The facts reveal that Anheuser-Busch presented the industry with sufficient notice of its selection of a brand name. At that time, defendant, on the basis of marketing considerations, decided to use its established brand name with a category descriptor consisting of the same mark plaintiff was using as a brand name--LA (Defendant's post-trial brief, p. 18) (emphasis added).

A "category descriptor" is presumably a term descriptive of a category. Whether the term LA was descriptive in nature was discussed at length by the district court. It specifically found that it was not. We believe that the district court in the passage above simply meant what it said: Stroh had used Anheuser-Busch's brand name LA as a category descriptor together with its own established brand. Stroh may have used LA as a category descriptor and hoped it would become such generally,7 but the district court does not say that it found the term to be such. To read the sentence otherwise would grossly distort the specific findings of the court. Since we do not believe the district court found LA to be a category descriptor, we find no error as a matter of law in the court's issuance of its injunction.

Stroh next argues that the district court erred as a matter of law in relying primarily on consumer perceptions in its classification of LA. It asserts that consumer perceptions should not be used as the principal test for "terms that have started out as generic category descriptors." Stroh's assertion assumes that LA is a "generic category descriptor."8 As discussed above, the district court specifically found LA to be neither generic or descriptive. Thus Miller Brewing Co. v. Jos. Schlitz Brewing Co., 605 F.2d 990, 995 (7th Cir.1979), cert. denied, 444 U.S. 1102, 100 S. Ct. 1067, 62 L. Ed. 2d 787 (1980), the case Stroh cites as authority, is not on point: that case deals with the irrelevance of a survey of the responses of 988 beer drinkers to "the meaning of a familiar word of Anglo-Saxon heritage" (i.e., "light"). The mark LA in a beer-drinking context obviously is not so familiar. As the Miller case itself points out, when "a coined word for a commercial product" is involved, the proper test becomes "what 'buyers understand by the word.' " Id. (quoting Bayer Co. v. United Drug Co., 272 Fed. 505, 509 (S.D.N.Y.1921)).

In WSM, Inc. we specifically stated that "[t]he test for deciding whether a word has become a generic title of a product or service is one of buyer understanding: 'What do the buyers understand by the word for whose use the parties are contending.' " 724 F.2d at 1325 (citations omitted). The same test applies when deciding whether a word is descriptive. In re Abcor Development Corp., 588 F.2d 811, 814 (C.C.P.A.1978).

Stroh argues that in addition to consumer perceptions, other factors such as the understanding of those familiar with low alcohol beer (presumably those in the beer trade, since consumers are being distinguished) and the intent of Anheuser-Busch in adopting the mark LA should have been considered. The test, however, is what consumers, not persons in the trade, understand the term to be. WSM, Inc., 724 F.2d at 1328 n. 3; Big O Tire Dealers v. Goodyear Tire & Rubber Co., 561 F.2d 1365, 1369 (10th Cir.1977), cert. denied, 434 U.S. 1052, 98 S. Ct. 905, 54 L. Ed. 2d 805 (1978); Nature's Bounty, Inc. v. SuperX Drugs Corp., 490 F. Supp. 50, 54 n. 3 (E.D.N.Y.1980); Salton Inc. v. Cornwall Corp., 477 F. Supp. 975, 985 (D.N.J.1979); Blisscraft v. United Plastics Co., 294 F.2d 694, 699 (2d Cir.1961) ("In determining whether a claimed trademark is descriptive at the time of its adoption, its meaning to a nonpurchasing segment of the population is not important. The critical question is whether the mark is descriptive to the prospective purchasers of the article.").

Even if LA has been used in some descriptive manner by persons in the trade, it is immaterial if consumers do not perceive LA to be a generic or descriptive term for beer. We therefore conclude that the district court did not err as a matter of law in relying primarily upon consumer perceptions in classifying the term LA.

Stroh finally asserts that the district court erred as a matter of law since the burden of proving the trademark status of LA fell on Anheuser-Busch. We agree that since the initials LA were not registered, Anheuser-Busch had the burden of establishing trademark status. National Conference, 692 F.2d at 488; Reese Publishing Co. v. Hampton International Communications, 620 F.2d 7, 11 (2d Cir.1980). Stroh does not indicate where in the record it finds indications that the district court placed the burden of proof on Stroh. We see no indication from the record that the district court so acted. Rather, our examination convinces us that the burden of proof was properly allocated.9 Hence, we find no error of law.

Thus, we find no error of law by the district court on the particular points urged by Stroh. Since we do not believe that we should abandon our rule that the question of the proper category in which a term is to be placed for trademark purposes is a question of fact, we shall follow our precedent in WSM, Inc. and review the district court record under clearly erroneous standard of Rule 52(a).

Stroh argues that since the term LA was merely generic or descriptive, the district court improperly relied on the consumer survey offered into evidence. Stroh further attacks the methodology of the survey.

As we have noted above, Stroh's argument would have required the district court to assume what it had to find, for in determining if a term is generic or descriptive, a court is to view the term from the standpoint of the average prospective purchaser. In determining the viewpoint of the prospective purchasers in trademark and unfair competition cases, substantial weight may be accorded the result of a properly conducted survey. See Squirtco v. Seven-Up Co., 628 F.2d 1086, 1091 (8th Cir.1980); Humble Oil & Refining Co. v. American Oil Co., 405 F.2d 803, 817 (8th Cir.), cert. denied 395 U.S. 905, 89 S. Ct. 1745, 23 L. Ed. 2d 218 (1969); Union Carbide Corp. v. Ever-Ready, Inc., 531 F.2d 366 (7th Cir.), cert. denied, 429 U.S. 830, 97 S. Ct. 91, 50 L. Ed. 2d 94 (1976); James Burrough Ltd. v. Sign of the Beefeater, Inc., 540 F.2d 266, 278 (7th Cir.1976). Courts have frequently relied upon properly conducted surveys in determining whether marks are generic or descriptive. Indeed, survey evidence may be the most practical manner of approaching the evaluation of public reaction in such cases. E.I. DuPont de Nemours & Co. v. Yoshida International, 393 F. Supp. 502, 518 (E.D.N.Y.1975). For instance, in DuPont v. Yoshida, the court rejected the argument that the mark TEFLON was generic on the basis of two surveys which showed that 48% and 68% of consumers recognized TEFLON as a brand name, even though one of the surveys showed 31% thought TEFLON was a common descriptive name. Id. at 526. See also Coca-Cola Co. v. Overland, Inc., 692 F.2d 1250, 1254 n. 11 (9th Cir.1982); Stix Products v. United Merchants & Manufacturers, 295 F. Supp. 479, 490-92 (S.D.N.Y.1968). The survey offered into evidence was properly central to the district court's findings of fact.

Stroh also argues that the consumer survey was improperly designed. The survey was designed by Dr. Yoram Wind, Professor of Marketing and Management, Wharton School, University of Pennsylvania.10 The district court found that the consumer survey was fairly and scientifically conducted by qualified experts and impartial interviewers, that the study drew responses from a sample of a relevant portion of potential consumers, that the questions, upon which the results relied, did not appear to be misleading or biased, and that the recordation of responses was handled in a completely unbiased manner. These findings were not challenged on appeal. Stroh, however, argues the survey was defective because respondents who were familiar with low alcohol beer were not asked what LA meant.11

This issue was addressed at trial. On cross-examination Dr. Wind was asked why the question, "What does the designation LA suggest to you?" was not asked. Dr. Wind testified:

If I will ask them after all these questions what does it mean or what does the designation of this, * * * first of all I'm directing them, I'm leading here, it's a leading question. I am directing them to start scratching their head what does it mean and then they can go into all types of puzzles and games especially in an experimental setting. They will tell me Los Angeles or try to guess what else.

What I'm trying to do is try to get the true meaning and the best way of getting the true meaning is open ended questions, not directive questions that ask the respondent to tell me. What can be better than asking what it is, describe it, or what type of beer is it, or what is the brand name. Very simple non-biasing type questions.

Dr. Wind also stated that if a student of his had included such a question in "a research design he would have gotten an F in the course." We believe the district court did not err in finding Dr. Wind's defense of his research design persuasive. Further, an opposing expert, Dr. Herbert E. Krugman, testified on direct examination that to ask such a question would be to engage in "a fishing expedition." He further stated that respondents would "start guessing" if so asked and that some might guess "Los Angeles," some "Light Anheuser" and some "low alcohol." That such a variety of answers was posited as probable responses is further indication of the suggestiveness of the term LA.

Dr. Wind was cross-examined at great length. The district court also heard the views of Dr. Krugman. It is clear from the district court's opinion that the Wind surveys were carefully scrutinized before it reached its determination that Stroh's evidence did not contradict the essence of Dr. Wind's research findings. While differing views were expressed, we cannot say that the trial court was clearly in error in reaching its conclusions about the surveys.12 Stroh also argues that the omission of its proposed question means Anheuser-Busch has not met its burden of proof. Given our conclusions regarding the surveys, we must also conclude that the district court did not clearly err in finding that the weight of credible evidence supported the conclusions it reached and, hence, that Anheuser-Busch's burden of proof was met.13

As we observed above, the district court grounded its analysis of the protectibility of LA on four mutually exclusive categories: the generic, descriptive, suggestive, and arbitrary or fanciful.14 See WSM, Inc., 724 F.2d at 1325. The district court found the term LA to be suggestive in nature.

Stroh argues that the term is generic. We have addressed above and found wanting Stroh's contention that initials of generic phrases must similarly be so classified. We have also addressed and rejected Stroh's contention that a "marketplace survey can serve little purpose" in determining if a term is generic. Since we have found that the district court was not clearly erroneous in its evaluation of the consumer survey offered into evidence and in according it the weight it did, we must also conclude that the court was not clearly in error in finding as a matter of fact that LA was not generic.15

Stroh also argues that LA is descriptive, not suggestive. The district court adopted the distinction between the two stated in Abcor Development, 588 F.2d at 814: "Generally speaking, if the mark imparts information directly, it is descriptive. If it stands for an idea which requires some operation of the imagination to connect it with the goods, it is suggestive." See Union Carbide Corp. v. Ever-Ready, Inc., 531 F.2d at 379; Abercrombie & Fitch Co. v. Hunting World, 537 F.2d 4, 11 (2d Cir.1976). As with genericness, the descriptiveness of a mark is to be determined from the standpoint of the average prospective purchasers. Abcor Development, 588 F.2d at 814. The district court therefore properly looked to the consumer survey in reaching its conclusion that "LA * * * stands for an idea which requires some operation of the imagination to connect it with the product."

Stroh argues that only when one is unaware that the product is low alcohol beer could any imagination be involved in assessing the nature of the product from the initials LA. However, the survey demonstrated that of those respondents who had seen, heard of, or tried LA beer, when asked "what type of beer is it?" only 24.4% answered "low alcohol," "light alcohol," "half alcohol" or "less alcohol." We cannot say that the trial court was in error in finding, after examining such results, "that consumers do not generally recognize the term LA to immediately connote low alcohol when they see such on plaintiff's product."

Stroh argues that the commercial context in which the term LA is used, including advertising for the product, must be considered, citing Abcor Development, 588 F.2d at 814 ("Evidence of the context in which a mark is used on labels, packages, or in advertising material directed to the goods is probative of the reaction of prospective purchasers to the mark."). The statement in Abcor Development was made in response to a claim that the proper test was the meaning of a term to an "ultimate consumer, who has never seen the product and does not know what it is." Id. at 812. In the instant case, the meaning of the term LA to consumers who had seen or heard of or tried the product was surveyed. As the district court noted, "the relevant time frame for determining a mark's protectibility is the present, not the future." It may well be that at some time hence Anheuser-Busch may more forcefully link the term LA to the phrase "low alcohol." It is clear, however, that the commercial context in which LA beer was introduced was not ignored in the design of the survey.

Further, at trial, a witness called by Stroh, James R. Buell, Jr., a former Stroh employee employed at the time of trial by Anheuser-Busch, stated that he felt LA was suggestive of the phrases "low alcohol," "less alcohol," or "light alcohol," and that one of the subjects under discussion at Anheuser-Busch had been "ways to speed up or strengthen the association in the buying public, the consumers between LA and the fact that it's a low alcohol product." If the connection were as obvious as Stroh now suggests, such need to strengthen the association would have been unnecessary. Again, we do not believe the district court was clearly erroneous in finding that LA is suggestive in nature.

Stroh also contends that the district court failed to properly consider in reaching its findings the Australian experience with low alcohol beer, the lessons to be learned from Miller's marketing of light beer, and the intentions of Anheuser-Busch as discerned from internal documents, including a preemptive move by Anheuser-Busch in the low alcohol market.

Tooth LA, a low alcohol beer, has been marketed in Australia for about five years. Stroh argues that the "bar call" for low alcohol beer in Australia has become "Give me an L.A.," that the term LA is generic or descriptive there, and that a word that is generic or descriptive in a foreign country should be accorded the same status in the United States. The cases and treatise cited by Stroh in support of the latter proposition, however, deal with the question of whether words from modern languages should be treated as equivalent to their English translations (such as French "cafe" for "coffee"). A number of cases hold that a term may be generic in one country and suggestive in another. In Carcione v. The Greengrocer, Inc., 205 U.S.P.Q. (BNA) 1075 (E.D.Cal.1979), the defendant argued that the trademark Greengrocer, which is a generic term in Britain for a retailer of fruits and vegetables, was not entitled to protection as a trademark in the United States. The court rejected this argument on the ground that it is irrelevant how a term is used outside the United States: "The parties agree that the term is generic in Britain. Since we deal here with American trademark law, and thus American consumers, neither British usage nor the dictionary definition indicating such usage are determinative." Id. at 1077. See also Seiko Sporting Goods U.S.A., Inc. v. Kabushiki Kaisha Hattori Tokeiten, 545 F. Supp. 221, 226 (S.D.N.Y.1982) (efforts to prove the mark Seiko was generic in Japan rejected since only issue was consumer recognition in the United States); Abercrombie & Fitch, 537 F.2d at 10. We believe these holdings are correct. Therefore, the district court was correct in not ascribing weight to the Australian experience in its determination of consumer perceptions or its ultimate classification.

Stroh discusses in its brief the marketing by Miller of light (or, as Miller spells it, "Lite") beer. Miller initially claimed trademark rights in "Lite"; extensive litigation ensued.16 In the meantime, states Stroh, "Lite" became the bar call, and Miller gained a preemptive stranglehold on the light beer market.

We find the Miller cases inapposite to the instant case. Those cases established that "light" was a term that had been used for a great many years in the beer industry and by consumers to describe a beer's color, flavor, body, alcohol content, or a combination of these and therefore was generic or descriptive. We believe the term LA is sufficiently dissimilar from "light" that these cases afford little precedential value.

Stroh argues that internal documents of Anheuser-Busch reveal both a descriptive or generic use of the term LA and an intention to preempt the low alcohol beer market with a mark similar to "Lite." Stroh in particular refers to a report from Interbrand Corporation, an independent marketing consultant firm which speaks of preemption and also renders the opinion that LA is not protectible.

We believe the evidence does indicate that Anheuser-Busch likely hoped to gain a significant share of the low alcohol beer market and to establish a mark that might become the "bar call."17 We also believe the evidence does establish that LA was, at times, used in a generic or descriptive sense internally. However, such matters are not relevant to the trademark status of LA, given the rules of law we have discussed above.

Stroh also argues that the injunction issued by the district court was unnecessarily broad in that it was not limited solely to beer and enjoined "doing any other acts * * * likely to cause confusion, mistake or deception with plaintiff." After careful consideration of the district court's order, however, we conclude that the injunction was properly drawn so as to afford Anheuser-Busch effective relief against future infringement of its trademark LA.

We affirm the issuance of the injunction by the district court.

BRIGHT, Circuit Judge, dissenting.

In this case, a major beer brewer, Anheuser-Busch, invokes federal jurisdiction seeking trademark protection for its brand name "LA". I would deny protection because its claim to the mark "LA" reflects a preemptive and anticompetitive intent to capture the emerging market for low alcoholic content beer by converting the initials of "low alcohol" (l.a.) into a trademark.

Counsel for Anheuser-Busch commented that this case is not complex. I agree. An artificial complexity arises only when pragmatic and flexible interpretations of the trademark laws become enmeshed in the rigid categorization of terms. Here, for example, the district court found LA to be a suggestive mark and, on that basis, granted Anheuser-Busch trademark protection. That finding, however, rests upon conclusions drawn from a legally irrelevant market survey. Furthermore, granting Anheuser-Busch a trademark on the term "LA" would subvert the objectives of the law of trademarks. For these reasons, I respectfully dissent.

I. ANHEUSER-BUSCH HAS FAILED TO ESTABLISH LA AS A TRADEMARK.

At the outset, I observe that because the mark "LA" (for "low alcohol")1 has not yet been registered, Anheuser-Busch bears the burden of proof in establishing that it can be trademarked. National Conference of Bar Examiners v. Multistate Legal Studies, Inc., 692 F.2d 478, 488 (7th Cir.1982); Reese Publishing Co. v. Hampton International Communications, Inc., 620 F.2d 7, 11 (2d Cir.1980); see supra at 638. The objectives of the trademark laws are: (1) to prevent confusion among the consuming public as to the source of goods or services and, (2) to indicate ownership and permit the trademark owner to control the product's reputation. See James Burrough Ltd. v. Sign of the Beefeater, Inc., 540 F.2d 266, 274 (7th Cir.1976); Sweetarts v. Sunline, Inc., 380 F.2d 923, 927 (8th Cir.1967); 3 R. Callmann, The Law of Unfair Competition, Trademarks, and Monopolies Secs. 17.01, at 2 & 17.13, at 58 (4th ed. 1983) (hereinafter cited as 3 R. Callmann). To realize these objectives, a trademark must distinguish a product from similar products and identify its owner or producer. Consequently, only those marks characterized as distinctive may receive trademark protection. 3 R. Callmann Sec. 18.01, at 1. Thus, to gain trademark protection for the mark "LA", Anheuser-Busch must demonstrate by clear and convincing evidence, see Application of Meyer & Wenthe, Inc., 267 F.2d 945, 949, 46 C.C.P.A. 919 (1959), that LA primarily serves to designate its source (from Anheuser-Busch) rather than to describe or indicate a characteristic that it shares with other beers of low alcoholic content produced by other brewers. Anti-Monopoly, Inc. v. General Mills Fun Group, 611 F.2d 296, 301 (9th Cir.1979); Roselux Chemical Co. v. Parsons Ammonia Co., 299 F.2d at 862; 3 R. Callmann Secs. 18.01, 18.03.

I do not take issue with the district court's decision to commence its analysis of whether LA is distinctive, and therefore entitled to trademark status, under the categories set forth in WSM, Inc. v. Hilton, 724 F.2d 1320, 1325 (8th Cir.1984),2 nor do I disagree that a factual issue arises in the categorization of a term. Id. at 1326. The clearly erroneous standard provided in Rule 52(a) of the Federal Rules of Civil Procedure for appellate review of a trial court's findings of fact does not, however, relegate the courts of appeals to the task of merely rubber stamping those findings. An appellate court may set aside even a finding of fact with some evidence to support it if, after reviewing " 'the entire evidence[, it] is left with the definite and firm conviction that a mistake has been committed,' " Anti-Monopoly, Inc. v. General Mills Fun Group, Inc., 684 F.2d 1316, 1318 (9th Cir.1982) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948)), or if the fact finding was based upon an erroneous legal standard. Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844, 855 n. 15, 102 S. Ct. 2182, 2189 n. 15, 72 L. Ed. 2d 606 (1982); Trout v. Lehman, 702 F.2d 1094, 1100 (D.C.Cir.1983); Save Lake Washington v. Frank, 641 F.2d 1330, 1334 (9th Cir.1981).

A. The Market Research Survey.

The majority relies heavily on the district court's determination that the initials "LA" do not describe but merely suggest the beer's low alcoholic content. This finding rested upon a market survey, conducted at the behest of Anheuser-Busch in contemplation of this litigation. The survey concluded that consumers perceived LA as the brand name of the product and not as a description for low alcohol beer.

I agree with the majority that the proper test for determining whether a mark is "generic", "merely descriptive", or "suggestive" looks to the beliefs of the average prospective consumers, which may be determined through properly conducted surveys, and, therefore, in the circumstances of this case, reject Stroh's contention that the proper test looks to the opinions of experts or the actual usage in the industry itself. See WSM, Inc., 724 F.2d at 1328 n. 3; Big O Tire Dealers, Inc. v. Goodyear Tire & Rubber Co., 561 F.2d 1365, 1369 (10th Cir.1977), cert. denied, 434 U.S. 1052, 98 S. Ct. 905, 54 L. Ed. 2d 805 (1978); Application of Abcor Development Corp., 588 F.2d 811, 814 & n. 15 (C.C.P.A.1978); 3 R. Callmann Secs. 18.01, at 2 & 18.04, at 13, 15 n. 8. However, the market survey introduced by Anheuser-Busch in the district court suffers several fatal defects, and cannot properly serve as a basis for the district court's finding that LA is suggestive, rather than generic or merely descriptive, of low alcohol beers.

Dr. Yoram Wind conducted the market survey for Anheuser-Busch. He testified that in order to determine whether "the LA designation (on the can of beer) connote[s] a light alcohol or low alcohol beer", the study focused only on the responses of those consumers who had never seen, heard of, or tried the beer, and who were exposed only to the LA label on a can. In essence, then, Dr. Wind sought to answer this question by "isolat[ing] the effect of the can from that of other advertising, promotion material and word of mouth communication." Not surprisingly, only 4.4% of the 226 consumers surveyed who fell into this category described the product as a low alcohol beer, and only 5.8% responded "low alcohol" to the follow-up question "what type of beer is it?"3 These responses, considered in conjunction with the responses of 76% of these participants that LA was the brand name of the product, obviously would support the district court's determination that consumers recognize LA as the brand name of the product and that it does not connote low alcohol to them.

The perceptions of a group of wholly uninformed consumers are, however, simply legally irrelevant. Dr. Wind's approach essentially measures consumer perception in an abstract, hypothetical context: what does a mark convey immediately to a consumer who has never before seen the product and who might not even know that such a product category exists. As the United States Court of Customs and Patent Appeals observed in Application of Abcor Development Corp., such an approach would "separate the concept of the average prospective purchaser from the world of reality." 588 F.2d at 814. Thus, the court rejected that approach and held, that in determining the reaction of prospective consumers to a mark, "[e]vidence of the context in which a mark is used on labels, packages, or in advertising material directed to the goods is probative * * *." Id. The Patent and Trademark Office Trademark Trial and Appeal Board, in its decision in the same case, stated that consideration should also be given to "the possible significance of the term in relation to the goods, and the likely reaction thereto of the average purchaser as he encounters the goods in the marketplace" in determining consumer perception. 197 U.S.P.Q. (BNA) 547, 548 (T.T.A.P.1978); see also 3 R. Callmann Sec. 18.04.

Thus, a legally relevant test of consumer perceptions must reflect the real world and a real market. In this case, then, the appropriate group of consumers to survey would have been an informed group of consumers, familiar with low alcohol beer.4 Certainly, trademarks are used in the real world in a commercial context through advertising. Anheuser-Busch presumably introduced its low alcohol beer with the intent of communicating to the public the low alcohol character of the beer. Although Dr. Wind based his conclusion that LA did not connote low alcohol beer solely on the responses of consumers with no prior exposure to the product or its advertising, the survey did include 369 consumers who had seen, heard of, or tried LA by Anheuser-Busch and who were exposed either to a can of LA or its supporting advertising. The responses of these consumers shed a different light on the question of whether LA connotes low alcohol beer.5

Of these respondents, 120 were exposed only to a can of LA: 23% described it as a low alcohol beer product and 24% responded "low alcohol" when asked what type of beer it was. The results are even more significant when one examines the responses of the other 249 respondents who were exposed only to the Anheuser-Busch advertising for LA. Sixty-two percent of this group described the product as a low alcohol beer. If we focus only on those who gave some answer to the question "describe the product", fully 76%6 described the product as a low alcohol beer.7 Even Dr. Wind acknowledged that consumers exposed to the advertising responded by talking about low alcohol.

In his testimony, Dr. Wind sought to minimize the impact of the responses describing the product as a low alcohol beer, noting that even among the consumers exposed to the advertising, the overwhelming majority identified LA as the brand name of the product in response to the question "what is the brand name of this beer?". Dr. Wind testified that he could not conclude that LA connoted low alcohol even to those consumers who responded "low alcohol" to the questions "describe the product" and "what type of beer is it?". Indeed, he points to the overwhelming number of these participants who responded LA when asked to state the brand name as evidence that consumers understand LA to be a brand name and that it does not connote to them the low alcoholic content of the beer.

In essence, Dr. Wind concluded that if a respondent recognizes LA as the brand name, he or she cannot also recognize that it stands for low alcohol, even in those instances in which the respondent had previously described the product as a low alcohol beer. This conclusion appears disingenuous. It rests on the false premise that LA can have only one meaning to consumers. Any term, after all, can mean more than one thing. It requires very little reflection to recognize that the average consumer, exposed to only one brand of beer and advertising designed to emphasize the brand name rather than the alcoholic content of that beer, might think both that LA stands for low alcohol and that it is the brand name for the product. Although Dr. Wind drew his conclusions from the study on the assumption that the respondents perceived LA as referring either to the product's brand name or to a description of its alcoholic content, the questions were not presented to the respondents as either/or propositions and it seems highly unlikely that they perceived them as such. Dr. Wind's expressed concern about avoiding a bias in the testing result is certainly appropriate. That the structure and type of questions employed avoided a bias in one direction does not, however, preclude the possibility that they created a bias in another.

The average consumer presumably has no conception of what is legally required for a brand name to receive trademark protection. Thus, just because a majority of the consumers surveyed thought, after being exposed to a can prominently labeled LA and advertising that stressed LA as the brand name, that LA was the brand name may not establish that the mark is entitled to trademark status. What is equally as impressive as the number of consumers who thought LA was the brand name, and more noteworthy for purposes of trademark analysis, is that, notwithstanding the focus in the advertising on emphasizing LA as the brand name, over 60% of those consumers who had previously been exposed to LA and who were exposed to the Anheuser-Busch advertising described LA as a low alcohol beer.

One could, by focusing on the responses of those consumers who had seen, heard of, or tried LA and who were exposed to the advertising, conclude that LA connotes low alcohol and, thus, is merely descriptive and not registrable. But, it is not necessary, in view of the record and the allocation of the burden of proof in this case, to go that far.8 Examination of the market survey reveals serious defects--namely, that the conclusion that LA does not connote low alcohol beer was drawn from the legally irrelevant sample group of uninformed consumers asked to draw meaning from an isolated product label, and that the survey was otherwise insufficient because of the questionable structure and make-up of the questions asked. Thus, upon analysis, this survey does not, in my view, support the district court's finding that LA is suggestive.

B. Industry Expert Opinion Testimony.

The majority also cites, as support for the district court's finding that LA is suggestive, the testimony of James Buell, an Anheuser-Busch executive, who stated that he felt LA was suggestive of low alcohol. See supra at 642. However, as noted supra at 645, the opinions of industry experts are irrelevant; the relevant test looks to the perceptions of informed prospective consumers. Thus, Buell's testimony cannot, as a matter of law, provide support for the district court's finding of suggestiveness.

My review of the record discloses no other evidence offered by Anheuser-Busch, or relied upon by the district court, to support the finding that LA is suggestive. Thus, I conclude that Anheuser-Busch has failed to meet its burden of proof in showing that the mark "LA" is distinctive and entitled to trademark status. Consequently, the district court's finding that LA is suggestive, rather than descriptive or generic, constitutes clear error and should be set aside.

C. Effect of Initials "LA".

On the basis of this record, Anheuser-Busch has failed to carry its burden of proof of establishing the registrability of LA. As a legal matter, Anheuser-Busch should not be granted a trademark in LA because those initials do not sufficiently distinguish its brand from the product--low alcohol beer. In determining whether a mark may become a trademark, the ultimate criterion focuses on the mark's distinctiveness: does it primarily indicate origin and ownership. The four WSM, Inc. categories--"generic", "merely descriptive", "suggestive", and "arbitrary and fanciful"--were propounded to demonstrate the continuum along which brand names lie. Under that scheme, "generic" and "arbitrary and fanciful" represent the two extremes--the former which can never constitute a trademark and the latter which clearly does. "Merely descriptive" and "suggestive" reflect the middle of the continuum where distinctiveness is a close question. Although these categories are useful conceptually, we need be mindful that the ultimate focus in analyzing trademark claims is always on the distinctiveness and source-designating capacity of a mark.9

In assessing the distinctiveness of the mark "LA", the key question is whether it substantially differs from terms in the public domain so that consumers will associate it exclusively with Anheuser-Busch. See 3 R. Callmann Sec. 18.01. An acronym is really no different than a mark consisting of a word or words; it cannot be a valid trademark if it describes or refers to the product or any of its characteristics or if it is not used solely to indicate origin. See Nature's Bounty, Inc. v. Basic Organics, 432 F. Supp. 546, 551 (E.D.N.Y.1977); 3 R. Callmann Sec. 18.23. Thus, initials recognized by prospective consumers as equivalent to the phrase for which they stand are descriptive and may not be trademarked if the full phrase cannot be trademarked. See FS Services, Inc. v. Custom Farm Services, Inc., 471 F.2d 671, 674 (7th Cir.1972); Modern Optics, Inc. v. Univis Lens Co., 234 F.2d 504, 506, 43 C.C.P.A. 970 (1956); 3 R. Callmann Sec. 18.10; accord National Conference of Bar Examiners, 692 F.2d at 488.

Several factors lead me to conclude that consumers will associate LA with "low alcohol." First, I would note the penchant of the American public to adopt nicknames and other shortened appellations, including acronyms. This tendency has been duly noted elsewhere. See Application of Abcor Development Co., 588 F.2d at 815 (Rich, J., concurring); 3 R. Callmann Sec. 18.04. It seems highly likely that LA will become the "bar call" for low alcohol beer in the United States, just as it did in Australia. See infra Part II(b). Indeed, this is precisely what Anheuser-Busch intended to capitalize on when it selected LA as the name for its low alcohol beer. See infra Part II(a).

Second, "[i]t is the use of the mark in a going trade or business to distinguish the product of the user from that of others and not the choice of it which creates what becomes the subject of property in the mark." Standard Brands, Inc. v. Smidler, 51 F.2d 34, 36 (2d Cir.1945) (emphasis added). Thus, I think it crucial to consider the initials "LA" in the context of the Anheuser-Busch advertising and the probable course of development in the low alcohol beer segment.10

In Grove Laboratories Inc. v. Brewer & Co., 103 F.2d 175, 179 (1st Cir.1939), the court noted that although the initials "LBQ" are arbitrary and fanciful, use of them on a label along with the words "Laxative Bromo Quinine" might render the initials descriptive.11 Accord Horlick's Malted Milk Co. v. Borden Co., 295 Fed. 232 (D.C.Cir.1924) (initials "MM" used on label along with "malted milk" descriptive of product); cf. Kampgrounds of America, Inc. v. North Delaware A-OK Campground, Inc., 415 F. Supp. 1288, 1291-92 (D.Del.1976):

If it were apparent, however, from the manner in which the KOA mark * * * were displayed that KOA did indeed stand for Kampgrounds of America [for example, if the KOA mark always appeared alongside of the name Kampgrounds of America], then I might be more convinced [that KOA is descriptive because it stands for a descriptive phrase].

The record reveals that consumers may not now associate LA with low alcohol only because this category of beer has only recently come on the market and much of the general public does not even know of its existence. Although Anheuser-Busch does not display the term "low alcohol" on the LA label, and primarily promotes LA as the brand name, the accompanying advertising does note the low alcoholic content of the beer. Little doubt exists, in my view, that LA will come to be synonomous with low alcohol in the average consumer's mind after a few months of heavy advertising. This belief is reinforced by the appearance of numerous articles in newspapers and magazines since the beginning of 1984 explaining that the term "LA" stands for low alcohol.12 This kind of association tends to confirm that the informed consumer exposed to the new product will more likely than not equate LA primarily with low alcohol, rather than associate the letters as an Anheuser-Busch trademark.

In sum, because I think that Anheuser-Busch has failed to marshall sufficient evidence to establish the registrability of the mark "LA," and that, on the record before us, the initials "LA", as a matter of law, represent a descriptive designation of the low alcohol product, I would set aside as clearly erroneous the district court's finding that LA is suggestive. Because the district court granted trademark protection on the basis of that finding, its judgment, in my view, lacks support. Accordingly, I now turn to an independent consideration of whether Anheuser-Busch should receive trademark protection for LA.

II. GRANTING TRADEMARK STATUS TO LA WOULD FRUSTRATE THE FUNDAMENTAL INTENT OF TRADEMARK PROTECTION.

The categories originally devised to aid the courts in analyzing trademark claims have assumed a life all their own and a tremendous body of doctrine has evolved that has become so obscured by semantic distinctions and unwarranted rigid categorization, mutually exclusive boundaries and hierarchical rankings, that it has choked off the vitality of the categories and their continuing efficacy in the analysis of trademark claims. "Arbitrary and fanciful", "suggestive", "descriptive", and "generic" have achieved talismanic significance: the adverse parties and the courts attempt to pigeonhole the disputed mark into one of the categories with a consequent outcome-determinative effect, instead of addressing the underlying equities to ensure that the objectives of the trademark laws are carried out. Courts must be mindful of the need to apply the law of trademarks flexibly and pragmatically in light of the existing circumstances.

Resolution of trademark issues under the subtle nuances of categorization often fails to consider the relationship between the mark, the nature of the product, and the applicant's objectives in seeking registration of a mark. Moreover, the factual patterns in trademark cases are so unique that the relationship between mark, product, and producer demands careful case-by-case analysis rather than extensive analysis of prior trademark decisions. See Application of Abcor Development Corp., 588 F.2d at 815 (Rich, J., concurring). With this as background, and on the basis of my conclusion in Part I(c) that the initials "LA" are merely descriptive of the generic phrase "low alcohol", I turn to a consideration of the equities in this case and the proper application of the law of trademarks to LA.

Descriptive or generic marks are denied the protection of the trademark laws for two reasons:

(1) to prevent the owner of a mark from inhibiting competition in the sale of particular goods; and (2) to maintain freedom of the public to use the language involved, thus avoiding the possibility of harassing infringement suits by the registrant against others who use the mark when advertising or describing their own products.

Id. at 813. It would be quite anomalous if the courts were to apply the law of trademarks, which is but a part of the law of unfair competition, in a manner which would permit companies to use the trademark laws as a sword to unfairly restrict competition rather than merely as a shield to protect their own products and their investment in promoting those products from unfair infringement by others. Yet, this is precisely what the majority today permits.

A. Anheuser-Busch Internal Documents and the Lite Beer Experience.

Sales of beer in this country have remained relatively flat for some time. The best prospects for sales growth in the beer industry occur in the development of new market segments. The prime example of this, and one not lost on any of the major breweries, including Anheuser-Busch, was the development of the "light" beer segment in the 1970's. Miller was the first major company to market a light beer and enjoyed roughly a two-year period during which it had the sole entry in the light beer market. Miller called its brand "Lite" and sought trademark protection. The Seventh Circuit ultimately determined that "light" or "lite", when used in the beer context, was generic and thus not registrable as a trademark. See Miller Brewing Co. v. G. Heileman Brewing Co., 561 F.2d 75 (7th Cir.1977), cert. denied, 434 U.S. 1025, 98 S. Ct. 75, 54 L. Ed. 2d 772 (1978).13 That decision, however, only denied to Miller the exclusive rights to the use of the term "lite" and Miller, through extensive advertising, established "lite" as the "bar call" for the low calorie beer segment. Thus, anyone who enters a bar and asks for a "light", gets "Lite" beer from Miller. According to the parties to this suit, Miller's current 57% share of the light beer market is largely attributable to its initial position as the sole market entry and the evolution of "lite" as the "bar call" for the category.

Between 1982 and 1983 many of the major brewers14 became interested in low alcohol beer as a possible new market segment. Numerous Anheuser-Busch internal documents and reports prepared for Anheuser-Busch during the planning stages in late 1983 and early 1984 indicate Anheuser-Busch's desire to "preempt" the low alcohol beer market "a la Miller and 'lite' beer." For example, the Product Naming Strategy Report, dated November 14, 1983, prepared by the Interbrand Corporation for Anheuser-Busch contains a penciled notation "preemptive" under the heading "Brand Name Objectives" and a recommendation that "Anheuser-Busch has an opportunity to 'write the book' on the reduced alcohol segment in exactly the same way Miller did in the light beer segment * * *." A November 1983 briefing report on reduced alcohol beers produced by James Buell, an Anheuser-Busch executive, lists as the first brand name objective for the Anheuser-Busch low alcohol beer that it be preemptive like "Lite." In another internal briefing document dated December 20, 1983, Charles Wirtel, another Anheuser-Busch executive, states that "it is our charge to be preemptive in the category * * * to be the major beneficiary of its emergence." Finally, and most tellingly, in the "Marketing Guide" furnished to its wholesalers on March 16, 1984, Anheuser-Busch states, under section II labelled "LA--The Bar Call of the 80s":

As the first product in this new market, LA has an excellent opportunity to become THE bar call for light alcohol beers. People will quickly learn to ask, not for a "light alcohol beer," but simply for an "LA." We should be able to establish a bar call that is virtually unbreakable.

Marketing Guide at 12. Further on, in the section addressing sales promotion, the Guide observes that "[o]ur new product has a natural bar call and should preempt the category during this introductory period." Id. at 21.15

These internal documents are relevant to show Anheuser-Busch's intent to use LA in a descriptive manner to preempt the low alcohol beer market. The trademark laws, however, are designed to protect names selected by a company and promoted with the intent of indicating the ownership and source of the goods. It is inconsistent with the objectives of the trademark laws to suggest that the protection of those laws should be extended to cases in which a company attempts to arrogate for business-promotional reasons a mark that is essentially descriptive. These documents disclose just such an intent, see Nature's Bounty, Inc. v. Basic Organics, 432 F. Supp. at 552 (considering intent and purpose of company in determining claimed mark to be descriptive), and that intent runs directly into the prohibition against granting trademark protection to a mark that would inhibit competition. Application of Abcor Development Corp., 588 F.2d at 813.

B. The Australian Experience.

In deciding to move into the low alcohol beer segment, both Anheuser-Busch and Stroh carefully studied the Australian experience with low alcohol beer since the market developed there in late 1979. Three companies in Australia use the designation LA on the label of their low alcohol beer. The evidence of the Australian experience indicates that LA quickly became the "bar call" or generic term for the low alcohol beer category, and that the Tooth Brewing Company, by advertising "Tooth LA" very heavily, had obtained a 70% share of the low alcohol beer market.

Stroh argued that the evidence that LA had become the "bar call" in Australia and that Tooth LA had captured a huge share of the market reveals that LA is in fact generic and thus not registrable for trademark protection. While I agree with the majority, see supra at 642, that the Australian experience is not relevant to a determination of consumer perceptions or the ultimate classification of LA in the United States, I do think that that evidence is relevant to show that Anheuser-Busch exercised no creative effort in arriving at the decision to call its low alcohol beer "LA". Judge Frank, concurring in Standard Brands, Inc. v. Smidler, stated that "[s]ince plaintiff did not originate the symbol--which, as heretofore noted, had previously been utilized by others * * * [in different contexts]--the scope of its monopoly should be peculiarly restricted." 151 F.2d at 43. Similarly, in this case I believe that Anheuser-Busch should not be granted trademark protection for a term it simply borrowed from other companies in the same industry--albeit in a different country--with the intent to preempt the market.

The unescapable conclusion to be drawn from these facts is that Anheuser-Busch sought to obtain a trademark on a term that it thought would become generic, or at least the "bar call", for this segment of the beer market in order to gain a preemptive hold on the low alcohol beer market just as Miller did with the light beer market. I do not in any way mean to denigrate Anheuser-Busch for that attempt. Its innovative strategy in opening up a market for a beer with a reduced alcoholic content deserves credit and it is entitled to exploit fairly this market. It may not, however, exploit this market by appropriating for its product a term that inherently belongs to the public, including its competitors. The trademark laws of this country were designed to protect brand names used to signify the source of a product and thereby enable a company to protect the reputation of its goods and to prevent confusion among the consuming public. It is untenable that the courts should extend the protection of these laws to sanction anticompetitive, preemptive objectives.16 Accordingly, I would refuse trademark status to Anheuser-Busch for its use of LA on its low alcohol beer.

III. THE INJUNCTION IS OVERBROAD.

Even if I agreed with the majority that the district court's finding that LA is suggestive, and thus entitled to trademark protection, was not clearly erroneous, I would modify the injunctive order by striking that portion which enjoins Stroh from "doing any other acts which are likely to cause confusion, mistake or deception with plaintiff." Rule 65(d) of the Federal Rules of Civil Procedure requires that injunctive orders be specific in terms and describe in reasonable detail the acts sought to be restrained. Injunctions must be tailored narrowly to remedy the specific harm shown rather than to enjoin all possible breaches of the law. Hartford-Empire Co. v. United States, 323 U.S. 386, 409-10, 65 S. Ct. 373, 385-86, 89 L. Ed. 322 (1945); Aviation Consumer Action Project v. Washburn, 535 F.2d 101, 108 (D.C.Cir.1976); Davis v. Romney, 490 F.2d 1360, 1370 (3d Cir.1974).

In this case, Anheuser-Busch complained that Stroh's use of LA in conjunction with one of Stroh's brand names--"Schaefer" or "Old Milwaukee"--would dilute the effectiveness of Anheuser-Busch's LA brand and confuse consumers. The injunction specifically prohibits Stroh from using LA or any colorable imitation. That part of the order is appropriate and should suffice to rectify the harm in this case. The problem with the "or any other acts" provision of the order is that it does not "prevent uncertainty and confusion on the part of * * * [Stroh, and it does not] avoid * * * [a potential contempt citation] on a decree too vague to be understood." Schmidt v. Lessard, 414 U.S. 473, 476, 94 S. Ct. 713, 715, 38 L. Ed. 2d 661 (1974); Ideal Toy Corp. v. Plawner Toy Manufacturing Corp., 685 F.2d 78, 83 (3d Cir.1982). This provision simply does not give Stroh "fair and precisely drawn notice of what the injunction actually prohibits." Granny Goose Foods, Inc. v. Brotherhood of Teamsters, 415 U.S. 423, 444, 94 S. Ct. 1113, 1126, 39 L. Ed. 2d 435 (1974).

It is uncertain, for example, whether this provision means that Stroh would be subject to contempt if it named its low alcohol beer "Loal", "Lial," or some other acronym for low or light alcohol. Anheuser-Busch is entitled to protect only its brand name, not the suggestive nature of it. Thus, even if the majority is correct in granting Anheuser-Busch trademark protection for LA, the injunction order is either too broad or too vague. Accordingly, I would strike the "any other acts" provision in the order.