Click here for footnotes

I. INTRODUCTION
TRADEMARKS are used to identify the sources of products or services. Regardless of what form trademarks take, they enable consumers to select favored products or services. [1] Well-established trademarks are thus valuable marketing tools, and companies spend substantial sums of money to develop, promote, and preserve them. [2]
As a result of these expenditures and promotional efforts, some trademarks have assumed a pervasive presence in our lives. [3] In the process, trademark owners have experienced a consequence they neither intended nor desired: their marks have been used by others in parodies and satires. Both parody and satire "conjure up" a recognizable "original" but alter it in a humorous and sometimes caustic fashion. [4] This is done at the least to entertain and often to engage in commentary on a public issue or an aspect of life. [5] The familiarity of certain trademarks makes them natural targets for the work of parodists and satirists. [6]
Trademark parodists are sometimes direct competitors of the trademark owners. [7] More often, however, they fall in one of two other categories: (1) commercial entities that sell goods or provide services of a different type than those of the trademark owners (and hence are, at most, indirect competitors); [8] or (2) magazine publishers or other similar parties who seek to entertain or make an editorial comment. [9]
Sometimes the parodies are humorous, though not necessarily to the trademark owner, [10] and generally in good taste. For example, McDonald's Corporation has sued other companies that used "Mc"-prefixed trade names in order to capitalize on McDonald's substantial goodwill. [11] Similarly, Toys R Us, Inc. has adopted an aggressive enforcement policy concerning other parties' uses of "R Us" in business names. [12] Other times, the parodies' humor and comment appear in a context that the trademark owner finds distasteful or scandalous. Consider, for instance, Coca-Cola Company's successful suit against the producer of an "Enjoy Cocaine" poster, [13] and L.L. Bean, Inc.'s unsuccessful suit against an adult magazine whose parody used the plaintiff's trademark in a setting L.L. Bean considered unwholesome and offensive. [14]
When such an alteration of the original trademark appears in a parody or satire, the mark's owner may not be content simply to grin and bear it. Trademark owners have become increasingly inclined in recent years to seek relief against these unconsented uses of their marks through "trademark parody" suits. [15] Pursued on various legal theories, these suits are designed to stop actions that, according to trademark owners, threaten to destroy not only the strength and distinctiveness of their marks, but also the positive connotations the marks previously produced in the minds of consumers. [16] Because weakened or tarnished trademarks are likely to be less effective in the marketplace, the suits also seem to be motivated by the owners' desire to continue receiving reasonable returns on the investments they have made in their trademarks. [17]
The rights of trademark owners, however, are not the only interests at stake in these cases. Because parody and satire necessarily involve expression, trademark parody litigation pits the first amendment's guarantees of freedom of speech and press against trademark owners' property rights. The resolution of cases in which trademark parodists have raised first amendment defenses has been especially difficult for courts. [18] Judicial treatment of the free expression aspects of these cases has generally been unprincipled and less than insightful. [19] There is a need for a useful means of assessing first amendment defenses in trademark parody litigation.
This article examines the proper role of the first amendment in the resolution of trademark parody cases. Section II discusses the legal theories used by trademark owners when they seek to enforce their property rights against parodists. [20] To facilitate a proper evaluation of the free expression arguments sometimes made by parodists, section III discusses the levels of first amendment protection given to different sorts of speech. Section IV analyzes the courts' attempts to address these first amendment arguments of parodists. [21]
Section V proposes an approach designed to take adequate account of trademark parody plaintiffs' strong interests in trademark preservation and defendants' competing constitutional interests in free expression. The suggested approach depends heavily upon distinguishing between parodies that are commercially motivated and parodies that cannot credibly be considered commercial in nature. This article proposes that commercial parodies should receive a minimal level of first amendment protection, and indeed no first amendment protection if they are likely to cause confusion concerning the source of a product or service. Noncommercial parodies, however, should receive substantial protection under an approach that restricts the application of basic trademark rights theories and balances freedom of expression interests against potential harm to the trademark owner. [22] The approach urged in section V stems from the Supreme Court's longstanding distinction between commercial and noncommercial speech--a distinction so far applied chiefly in cases other than trademark rights cases. [23] Moreover, as will be demonstrated, the suggested approach is consistent with major first amendment decisions of the Supreme Court during the past three terms. Although none of these decisions involved a trademark parody, taken together they emphasize the high level of first amendment protection given to noncommercial speech (including noncommercial speech of an offensive nature), and demonstrate a widening gulf between the respective levels of protection given commercial and noncommercial speech. [24]

II. TRADEMARK OWNERS' RIGHTS AND CLAIMS ASSERTED AGAINST PARODISTS

A. Nature and Sources of Legal Protection for Trademarks
Trademark law is part of unfair competition law, which brings together under a single heading a wide range of legal doctrines that deal with relations between business competitors. [25] Nevertheless, trademark law also possesses strong consumer protection elements, as demonstrated by the notion that trademarks enable consumers to identify favored products without fear of confusion concerning the source of those products. [26] Traditional trademark law thus hinges liability for an allegedly unfair competitive act upon whether harm to consumers is likely to result from the act. [27]
Unlike patent and copyright law, trademark law is not exclusively a creature of federal statute. [28] Trademark law began as common law, a still-existing source of protection. [29] Trademark law acquired significant federal stature in the Lanham Act of 1946, recently refurbished by the Trademark Law Revision Act of 1988. [30] Although federal law is now the preeminent source of trademark protection, the federal system of trademark registration is permissive rather than mandatory. [31] Common law protection for trademarks thus persists without being preempted by federal law. [32] Some states extend protection to trademarks in statutes whose provisions are similar to those of the Lanham Act or the common law, [33] or in anti-dilution statutes. [34]
Under these various sources of trademark law, trademarks are classified as intangible property, and trademark owners acquire significant property rights in their marks. [35] These rights, however, are not absolute. Instead of giving trademark owners monopoly protection and a legal ability to stop all unconsented uses of their trademarks, state and federal laws permit trademark owners to enjoin unconsented uses and recover damages only when those uses contravene certain recognized standards set forth in the available theories of recovery. [36] Therefore, the elements of proof imposed by these theories of recovery must be examined.

B. Infringement

1. Elements of Infringement
Trademark infringement is the leading theory relied upon by mark owners in suits for unconsented uses of the owners' marks. [37] Regardless of whether the trademark is entitled to protection under federal law, [38] the common law, [39] or a state trademark statute, the plaintiff must prove two essential elements in order to establish that there has been infringement of his mark. First, the plaintiff must demonstrate that, without his consent, the defendant used the plaintiff's protected trademark or something substantially similar thereto. [40] Second, the plaintiff must prove that this use by the defendant created a likelihood of confusion among consumers. [41] Proof of actual confusion is not required. [42]
The necessary likelihood of confusion may be as to the source or origin of the product or service. [43] In other words, if the public is likely to believe that the plaintiff, rather than the defendant, provided the defendant's goods or services, then the likelihood of confusion element would be satisfied. [44] The likelihood of confusion element, however, is not restricted to probable confusion concerning the narrow source or origin question. It may be satisfied by proof of a probability that consumers could become confused as to whether the plaintiff sponsored or endorsed the defendant, its goods, or its services, or as to whether the plaintiff was affiliated or associated with the defendant or its business activities. [45]
Known as the "keystone" of trademark law, the likelihood of confusion element usually becomes the critical issue in a trademark infringement case. [46] This requirement imposes a significant limitation on the property rights enjoyed by trademark owners because a defendant's unconsented use of the plaintiff's trademark is not an infringement if the use does not create a likelihood of confusion. Thus, for example, a florists association's use of "THIS BUD'S FOR YOU" in advertisements for rosebuds and fresh-cut flowers was held not to infringe Anheuser-Busch, Inc.'s beer slogan "This Bud's for you" because the requisite likelihood of confusion was lacking. [47]
Various federal appellate courts have developed lists of factors for determining whether a likelihood of confusion was created by a defendant's use of the plaintiff's trademark. [48] A typical list [49] called for consideration of the following factors: (1) the strength of the plaintiff's trademark; [50] (2) the degree of similarity between the plaintiff's mark and the defendant's use; [51] (3) the "competitive proximity" of the products or services with which the plaintiff and defendant have used the mark; [52] (4) the intent, if any, of the defendant to pass off his goods or services as those of the plaintiff; [53] (5) the amount of actual confusion, if any, among consumers as a result of the defendant's actions; [54] and (6) the degree of care consumers are likely to exercise in deciding whether to purchase goods or services provided by the plaintiff and defendant. [55] It is not necessary that all of the factors point toward the same conclusion. When the factors do not unanimously suggest a particular conclusion, the court must identify those factors that weigh most heavily in light of all the relevant circumstances, and must determine whether those factors point toward a likelihood of confusion. [56]

2. Trademark Parody Cases and Likelihood of Confusion Issue
The likelihood of confusion issue is central to a trademark infringement claim regardless of whether the defendant's use of the plaintiff's mark was in the context of a parody. The cases in which the alleged infringement amounted to a parody are of particular interest, however, because they have led to seemingly irreconcilable results on arguably similar facts. [57] Concededly, the likelihood of confusion determination is largely fact- specific. Consequently, one might sometimes expect divergent results in cases involving arguably similar but somewhat different facts. One might also assume, however, that the context in which the defendant used the plaintiff's mark should normally prevent a finding of likelihood of confusion, so long as the public would reasonably identify the defendant's use as a parody. In other words, the nature of parody, which involves the taking of "shots" at a recognized original, ordinarily would seem likely to cause the public to realize that the plaintiff was not the source of the parody, that the plaintiff did not sponsor or endorse the defendant's use of his trademark, and that the plaintiff and the defendant were not affiliated in business. [58] This assumption is sometimes, but clearly not always, borne out in the trademark parody cases.
Tetley, Inc. v. Topps Chewing Gum, Inc. [59] provides a useful illustration of the expected operation of the assumption that a trademark parody would not ordinarily create confusion. In Tetley, the court denied Tetley's request for injunctive relief against parodies of its trademarks for "Tetley Tea" and the "Tiny Little Tea Leaf Tea" slogan. [60] A Topps sticker, designed for children and sold across the country, had used the name "Petley" on a picture of a product box that depicted a dog scratching fleas. [61] The words "40 Flea Bags" and "Tiny Little Dog Fleas" also appeared on the picture of the box. [62] In concluding that there was no likelihood of confusion concerning source, sponsorship, endorsement, or affiliation, the court stressed the "heavy handedness" of the defendant's parody. [63] The public's probable recognition of the sticker as a parody would lead to a related recognition that the plaintiff trademark owner was not responsible for the sticker. As the court recognized in Tetley, the fact that the parody causes consumers to think of the original does not by itself mean that consumers are likely to be confused concerning source, sponsorship, and the like. [64]
The obviousness of the defendant's parody has constituted a major factor in judicial findings of no likelihood of confusion in both commercial and non-commercial settings. In the commercial setting, for example, a defendant's use of the name "Lardashe" on its large-size designer jeans was held not to cause a likelihood of confusion, and thus did not infringe the plaintiff's familiar "Jordache" trademark. [65] "Miami Mice" t-shirts parodying "Miami Vice" television show trademarks [66] and the previously- discussed parody of the "This bud's for you" slogan by a florists' association were also held not to cause a likelihood of confusion. [67]
In the non-commercial cases, obvious parodies are used as vehicles for the expression of viewpoints or commentary of an editorial nature rather than to sell a product or service. [68] For instance, in Reddy Communications, Inc. v. Environmental Action Foundation, Inc., [69] defendant Environmental Action Foundation, Inc. (EAF) was a nonprofit corporation whose purpose was to provide information about environmental issues. In various pamphlets and books, EAF used caricatures of the plaintiff's trademark, the Reddy Kilowatt character, to express adverse comments about the electric utility industry. These uses of the caricatures were clearly contrary to the purposes underlying the plaintiff's longstanding use of the Reddy Kilowatt trademark in utility companies' advertisements and promotional materials. [70] Nevertheless, the parody was obvious, not only because EAF's versions of Reddy Kilowatt were caricatures, but also because of the unmistakably negative commentary in which the caricatures played a role. [71] The court concluded that there was no trademark infringement because reasonable readers of the EAF literature would be unlikely to believe that the plaintiff and the utility companies with which the plaintiff was allied--the very objects of EAF's criticism--produced, sponsored, or endorsed the EAF materials. [72]
Similarly, in Stop the Olympic Prison v. United States Olympic Committee, [73] STOP, a nonprofit organization, used USOC's trademarks on a poster that protested a plan to turn the Olympic Village at Lake Placid into a federal prison at the conclusion of the 1980 Winter Olympics. Three trademarks were used: the word "Olympic" was part of the "Stop the Olympic Prison" heading at the top of the poster; the familiar interlocking Olympic rings were superimposed on a drawing of vertical steel-gray bars; and the Olympic torch was depicted as being thrust through the rings and bars by a silhouetted forearm. [74] STOP's clearly noncommercial use [75] was not technically a parody because it was not even arguably humorous, but it did involve adverse commentary sufficiently similar to a parody to warrant discussion here. The court rejected USOC's claim of trademark infringement [76] because the necessary likelihood of confusion was not established. [77] In view of the context and content of STOP's poster, reasonable persons who viewed the poster would not be inclined to believe that USOC produced, sponsored, or endorsed it. [78]
In still other cases, the defendant's parody may have straddled the commercial/noncommercial line but was nonetheless so obvious that there was no likelihood of confusion. For example, in Pillsbury Co. v. Milky Way Productions, Inc., [79] Pillsbury lost a trademark infringement claim against the publisher of an adult magazine that had depicted a likeness of Pillsbury's trademarked "Poppin' Fresh" character engaging in various sexual activities. [80] Because the parody appeared in the context of an adult magazine and contained material that was hardly consistent with anything Pillsbury would ever authorize, the court found that even the most gullible reader would not regard the parody as the work of Pillsbury. [81] A similar approach was recently taken in Cliffs Notes, Inc. v. Bantam Doubleday Dell Publishing Group, Inc. [82] In Cliffs Notes, the United States Court of Appeals for the Second Circuit rejected an argument that the defendant's Spy Notes, parodies of the familiar Cliffs Notes summaries of well-known works of literature, were confusingly similar to Cliffs Notes. [83] The court concluded that the obviousness of the parodies [84] would likely prevent any public confusion as to whether the publisher of Cliffs Notes also published or sponsored Spy Notes. [85] Various other decisions involving parodies that were either noncommercial or close to the commercial/noncommercial line have held that the defendant's use of the plaintiff's trademark did not create a likelihood of confusion. [86]
Not all of the parody cases litigated on a trademark infringement theory are won on the defendant's argument that the obviousness of the parody prevented the necessary likelihood of confusion. In fact, trademark owners have won a high percentage of the infringement claims against parodists. [87] Some of the cases decided in favor of the trademark owners have involved parodies that might seem obvious but were nonetheless shown by competent evidence to create a likelihood of confusion. For example, fast food giant McDonald's Corp. prevailed on separate infringement claims against parties that employed the names "McBagel's" for a bakery and restaurant [88] and "McSleep" for a proposed chain of motels. [89] In each case, consumer survey evidence demonstrated that any parodying of the McDonald's name [90] was not so obvious as to prevent a likelihood of confusion among consumers. The surveys demonstrated a rather high level of actual confusion among consumers, many of whom believed that McDonald's was involved in some respect in the McSleep and McBagel's ventures. [91]
Survey evidence showing actual confusion also played a key role in Mutual of Omaha Insurance Co. v. Novak. [92] The court held that Mutual of Omaha's trademark was infringed by the defendant's use of "Mutant of Omaha" on t-shirts and other items. The court concluded that the defendant's parody of the Mutual of Omaha mark [93] would not be readily apparent to consumers, and could thus cause confusion as to source, sponsorship, or endorsement. [94]
In other cases, however, it seems that the court's distaste for the defendant's parody has caused the court to summarily conclude that there was a likelihood of confusion without engaging in meaningful analysis of the issue. [95] For instance, in one of the early trademark parody cases, Coca-Cola Co. v. Gemini Rising, Inc., [96] the court granted Coca-Cola a preliminary injunction against the producer of a poster entitled "Enjoy Cocaine." [97] The words appeared in the familiar cursive script used by Coca-Cola for its brand name. [98] The court's opinion reflects a preoccupation with the objectionable nature and subject matter of the defendant's poster. One gets the impression that the court was eager to enjoin what it considered a distasteful poster, [99] even though careful analysis of the likelihood of confusion question might have revealed a roadblock to the desired result. [100]
Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd. [101] is perhaps an even stronger example of a court's giving short shrift to the likelihood of confusion issue and finding trademark infringement primarily on the basis of objectionable speech or actions by the defendant. [102] The plaintiff claimed trademark rights in the physical appearance of its cheerleaders' uniform. [103] It objected to the use of a likeness of the uniform in the sexually-explicit movie "Debbie Does Dallas." The conclusion of the movie featured a cheerleader who engaged in various sexual exploits while clad in parts of the uniform. In affirming the trial court's finding of trademark infringement, [104] the United States Court of Appeals for the Second Circuit disapprovingly referred to the content of the film as "gross and revolting" and "depraved." [105]
Neither the Second Circuit nor the district court in Dallas Cowboys conducted an insightful analysis of the likelihood of confusion issue. The district court simply concluded that likelihood of confusion could be inferred from the moviemaker's apparent intent to capitalize on the popularity of the Dallas Cowboys Cheerleaders. [106] The drawing of such an inference was erroneous, however, because the intent to use a recognized trademark is not necessarily the same as an intent to confuse the public. [107] An inference of likelihood of confusion should only be drawn from the latter sort of intent. [108]
The Second Circuit also seemed to misapply this basic concept. It noted that the uniform shown in the movie "unquestionably brings to mind the Dallas Cowboy Cheerleaders. Indeed, it is hard to believe that anyone who had seen defendants' sexually depraved film could ever thereafter disassociate it from plaintiff's cheerleaders. This association results in confusion . . . ." [109] The Second Circuit failed to appreciate the difference between a defendant's use that merely calls to mind the original trademark and a use that calls to mind the original in such a way that there is a likelihood of confusion concerning source, sponsorship, or endorsement. The former and the latter are different sorts of uses. [110] Indeed, in Dallas Cowboys, the objectionable way in which the trademark was used--the primary object of the court's concern and the major reason it found a likelihood of confusion-- should have led the court to conclude that there was no likelihood of confusion. The nature of the movie was such that the public would not be inclined to believe the trademark owner had any involvement with it.
In addition to its role as an essential element of trademark infringement, the likelihood of confusion element has special significance in the infringement cases that involve parodies. [111] As will be explained later, when the likelihood of confusion requirement is properly applied, it serves as a check on liability and usually keeps the infringement theory from posing serious first amendment difficulties. [112]

C. Section 43(a) of Lanham Act of 1946 and Trademark Law Revision Act of 1988
Federal law sets forth not only the basis for infringement claims concerning registered marks, [113] but also provides another theory often relied upon by plaintiffs in trademark-related litigation. Section 43(a) of the Lanham Act of 1946, as recently amended by the Trademark Law Revision Act of 1988 (TLRA), creates a civil right of action in favor of persons likely to be harmed by another party's use, in connection with the providing of goods or services, of false or misleading designations of origin, or other false or misleading representations. [114] This section may be invoked regardless of whether the plaintiff has a federally registered trademark. [115] In many respects, section 43(a) is a federal unfair competition provision whose broad language may be read to cover a wide range of behaviors considered competitive torts under the common law. [116]
In some section 43(a) cases, the allegedly false or misleading representation usually is an implication that the plaintiff sponsors or endorses, or is affiliated or associated with the defendant or the defendant's business activities. [117] Such a representation arguably arises from the defendant's use of a mark confusingly similar to the plaintiff's. [118] Thus, section 43(a) claims of this sort are essentially the same as trademark infringement actions. [119] This is true regardless of whether the section 43(a) claim pertains to a parody of the plaintiff's mark [120] or to another allegedly infringing sort of use. Accordingly, the dual elements of substantial similarity and likelihood of confusion, which are required in common law infringement claims and in claims for infringement of federally registered marks, are also required in section 43(a) suits of the infringement variety. [121]
In addition to its use as an infringement theory in trademark parody cases, the recently amended version of section 43(a) may also be used by trademark parody plaintiffs as a federal vehicle for what is effectively an injurious falsehood claim. [122] The tort of injurious falsehood addresses a defendant's disparagement or casting of aspersions on a plaintiff's goods, property, services, or business. [123] The amended version of section 43(a) authorizes civil suits for a defendant's false or misleading representation which, "in commercial advertising or promotion, misrepresents the nature, characteristics, qualities, or geographic origin of his or her or another person's goods, services, or commercial activities." [124] The quoted language should allow a suit based on the defendant's representation concerning the plaintiff's goods, services, or business, because the plaintiff would be "another person" within the meaning of the statute. Such a conclusion is consistent with the apparent intent underlying the TLRA's amendment of section 43(a): to alter the prevailing judicial interpretation that section 43(a) reached only the defendant's representations concerning his own goods or services. [125]
Therefore, certain disparagement and other injurious falsehood claims that formerly would have been actionable only under the common law may now be pursued under the new section 43(a). [126] Bringing these claims under section 43(a) may be advantageous to plaintiffs because section 43(a)'s language does not impose the stern proof requirements normally encountered by plaintiffs in injurious falsehood cases. [127] Finally, the TLRA version of section 43(a) requires that the false or misleading representation must have been made "in commercial advertising or promotion." [128] All disparagement or other injurious falsehood claims occurring outside the advertising or promotion setting should remain actionable only under the common law and not under section 43(a). [129]
Because the amended version of section 43(a) is so new, it remains to be seen to what extent plaintiffs will rely on it as a federal vehicle for injurious falsehood claims. It is clear, however, that if trademark parody plaintiffs attempt to use the statute as a basis for a disparagement claim, they must be able to establish that any falsehoods asserted by the defendant concerning the plaintiffs' trademarks necessarily were falsehoods concerning their "goods, services, or commercial activities." [130]

D. Dilution
As demonstrated earlier, neither federal law nor the common law provides a trademark owner with complete protection because unconsented use of a mark will not entitle its owner to relief unless the use created a likelihood of confusion. [131] The likelihood of confusion requirement thus creates a gap in the protection a trademark owner receives.
Approximately one-half of the states have attempted to close this gap, at least partially, by enacting what are known as anti-dilution statutes. [132] These statutes set forth the dilution doctrine, which does not require the trademark owner to prove likelihood of confusion. [133] Unequivocally geared toward furthering trademark owners' economic interests in preserving the value of their marks, the dilution doctrine recognizes that the owners' interests may be endangered by other parties' uses of the marks even when those uses are not likely to create consumer confusion. Anti-dilution statutes therefore allow mark owners to obtain relief against parties whose actions violated neither federal law nor the common law. [134]
Dilution claims based on these statutes have become popular among plaintiffs in trademark rights litigation regardless of whether a parody is involved. [135] Most anti-dilution statutes are identical to or patterned after the Model State Trademark Bill, [136] which provides: Likelihood of injury to business reputation or of dilution of the distinctive quality of a mark registered under this Act, or a mark valid at common law, or a trade name valid at common law, shall be a ground for injunctive relief notwithstanding the absence of competition between the parties or the absence of confusion as to the source of goods or services. [137]
As indicated by the quoted language, the owner is entitled to injunctive relief for unconsented use of his trademark if he can show either dilution of the distinctive quality of the mark or likelihood of injury to business reputation. [138]

1. Dilution by Loss of Distinctiveness of Mark
The first theory for relief, from which the dilution doctrine derives its name, requires the trademark owner to demonstrate that the third party's use would be likely to dilute the distinctive quality of the owner's mark. [139] This theory is based on the concern that a gradual "whittling away" of a mark's usefulness and value may result from the public's being exposed to uses of the mark, or substantially similar versions thereof that were not intended by the mark's owner. [140] Even if the public realizes that the party making the use is not connected with the mark's owner, or that the respective uses of the mark's owner and the other party pertain to dissimilar products or services, a loss of usefulness or value may occur. [141] The mark's distinctiveness will be lost because the mark will no longer exclusively identify the mark owner as the source of certain goods or services. Instead, the mark will also come to identify the other party as a provider of certain goods or services. When that occurs, the usefulness and economic value of the mark to its original owner will be severely diminished or even destroyed. [142]
For example, the distinctive quality of the "Polaroid" mark, used for many years by Polaroid Corp. in connection with its cameras and related products, was held to be in danger of being diluted by another firm's subsequent use of "Polaraid" as the name of its heating and refrigeration business. [143] This dilution occurred regardless of whether consumers were likely to believe that the nationally-known producer of cameras had begun providing refrigeration services. [144] Even though the public could likely recognize that the camera manufacturer and the refrigeration company were separate entities, the court found that the uniqueness of the "Polaroid" mark was likely to be lessened. [145] If the public were to continue being exposed to the refrigeration company's use of "Polaraid," the camera manufacturer's legitimate uses of "Polaroid" would trigger associations, in the public's collective mind, not only with cameras but also with the defendant's refrigeration services. [146] Thus, "[t]he mental image would be blurred," [147] even though most consumers would not be confused concerning what each company provided.
Dilution claims based on the loss of a mark's distinctive quality depend heavily on the mental associations made by consumers when they see a mark in use. [148] The mental associations relevant to dilution are subtly yet significantly different from those relevant in trademark infringement cases. [149] In infringement cases, the impression resulting from the defendant's unconsented use of the plaintiff's mark must trigger a likelihood of public confusion concerning whether the plaintiff is providing the defendant's goods or services or is in some sense connected with the defendant. [150] The success of a dilution claim, however, depends upon a defendant's use that calls to mind both the defendant and the plaintiff without producing a mental association so strong as to cause a probability of confusion over source. [151]
The subtle distinction between the consumer associations relevant to dilution and those relevant to infringement cases makes the dilution theory conceptually difficult. There is no question, however, that by eliminating the need for the likelihood of confusion element required in trademark infringement law, the dilution doctrine "lays a heavy hand upon one who adopts the trade name or mark of another." [152] In trademark parody cases, plaintiffs have sometimes relied on this first type of dilution claim--dilution by loss of distinctiveness--as a basis for relief against parodists. The results in these cases have been mixed, and judicial analysis of the presence or absence of dilution has been generally superficial. [153]
Although generally not enunciated by the courts, two considerations help explain the decisions on parody-based dilution claims. First, dilution by loss of distinctiveness is more likely to be found if the defendant's use was in a trademark-like sense (as in the name of the defendant's product, service, or business) rather than in the nature of commentary. Second, if the defendant's use occurred only once, it is less likely to give rise to a successful loss of distinctiveness claim than if the use has been repeated numerous times or would be so repeated unless enjoined. [154] A one-time use is much less likely to effect the underlying concern of this sort of dilution claim--the "whittling away" of the distinctiveness of the trademark. [155]
In McDonald's Corp. v. McBagel's, Inc., [156] for example, both considerations weighed in favor of the plaintiff. The defendant's use of "McBagel's" as the name of a bakery and restaurant involved the name of the plaintiff's product and would presumably have continued indefinitely if not enjoined. [157] Concluding that a loss of the distinctiveness of McDonald's "Mc"-prefixed trademarks would be likely if the McBagel's name remained in use by the defendant, the court granted injunctive relief to McDonald's. [158] Conversely, in L.L. Bean, Inc. v. Drake Publishers, Inc., [159] the two considerations weighed in the defendant's favor. The First Circuit refused to recognize L.L. Bean's dilution claim concerning an adult magazine's parody of the L.L. Bean name and catalog. [160] The defendant's use was not in a trademark-like sense, [161] and was presumably a one- time occurrence. [162] Other decisions involving comparable uses by defendants are generally consistent with L.L. Bean. [163]

2. Loss of Business Reputation Associated with Mark
Under most anti-dilution statutes, the owner can also bring a successful dilution claim by demonstrating that the defendant's use of the mark is likely to damage the previously developed business reputation associated with the mark. [164] This alternative is broad enough to include two differing dilution actions. The first is a claim of probable damage to a trademark's business reputation as a result of the defendant's use of the mark, or a substantially similar version, in connection with goods that are inferior in quality to, or less prestigious than, the trademark owner's goods. The plaintiff's argument is that the mark's reputation and value are likely to be diminished because the public will assume that the lack of quality or lack of prestige reflected by the defendant's goods will also be present in the plaintiff's unrelated goods. [165]
The second sort of dilution claim pursued under the loss of business reputation is seen more frequently. The claim is based on another's use of the mark in a context the mark owner considers unwholesome. Such a use arguably casts a negative light on the mark, reducing its reputation and standing in the eyes of consumers as a wholesome identifier of the mark owner's products or services. This sort of claim is sometimes referred to as dilution by "tarnishment" because it deals with a mark whose previously unsullied public image has been tarnished by the defendant's offensive actions. [166]
Most of the dilution by tarnishment cases involve parodies. Although the trademark parody cases applying anti-dilution statutes do not establish ground rules for determining what is an unwholesome context, they tend to indicate that parodies with sexual overtones or dealing with illegal drug use are likely to tarnish the reputation of a plaintiff's mark. [167] In Pillsbury Co. v. Milky Way Productions, Inc., [168] the court concluded that a sexually- oriented parody was an unwholesome context and allowed the plaintiff to prevail on a dilution by tarnishment theory even though the absence of a likelihood of confusion caused the plaintiff to lose its trademark infringement claim. [169] In Coca-Cola Co. v. Gemini Rising, Inc., [170] the court relied on dilution by tarnishment as an alternative basis for its issuance of a preliminary injunction against the defendant's "Enjoy Cocaine" poster. [171] Although one court concluded that the Garbage Pail Kids stickers' parody of the Cabbage Patch Kids was unwholesome, [172] most courts have been hesitant to find unwholesome contexts when neither sexual content nor references to illegal drug use can be found in the defendant's parody. [173]
In the L.L. Bean case, the First Circuit departed from the prior trend of finding dilution by tarnishment almost as a matter of course when the plaintiff's trademark was used in a parody having sexual overtones. The court refused to grant relief to the trademark owner despite the appearance of the defendant's parody in an adult magazine. [174] In so ruling, the court rejected the view that a finding of tarnishment may be based on the defendant's single unauthorized use of the plaintiff's trademark in an arguably unwholesome setting. [175] It asserted that the dilution by tarnishment theory should be restricted in application to clearly commercial contexts where the defendant uses the plaintiff's trademark in connection with goods that are inferior to, or otherwise incompatible with, the plaintiff's goods or business. [176] The court called the defendant's parody "editorial or artistic" and held that as a noncommercial parody it was not subject to injunction under the anti-dilution statute. [177]
The foregoing cases leave considerable uncertainty as to what is an unwholesome context and whether it is appropriate to apply dilution by tarnishment to noncommercial parodies. [178] Because dilution by tarnishment claims necessarily require courts to make judgments on what is offensive, unwholesome, or tarnishing, trademark parody cases pursued on dilution grounds implicate important first amendment concerns. These first amendment issues are often ignored entirely by courts or, if not ignored, are dealt with unsatisfactorily. [179]
No discussion of the dilution doctrine would be complete without mention of a phenomenon dubbed "pseudo-dilution." [180] This phenomenon is the tendency of some courts to grant a trademark owner relief on dilution-like grounds when the dilution theory has not been raised by the plaintiff or is not even available under applicable law. [181] Pseudo-dilution is sometimes seen in cases where the court regards the defendant's actions as particularly objectionable. [182] When courts resort to pseudo-dilution, they do not openly acknowledge having done so. Instead, they inappropriately employ dilution-oriented language to bolster a finding of infringement or conclude that infringement was established even though the decision reveals an underlying dilution basis. [183] Plaintiffs who either failed to bring a dilution claim or were not entitled to do so under applicable law obviously are happy to acquiesce in a court's use of pseudo-dilution.
Whether consisting of its statutory forms or of the unofficial variety just described, the dilution doctrine jas joined the infringement theory as a major player in the trademark rights arena. When the case involves a parody, trademark owners have not always been content to restrict themselves to claims of infringement or dilution. The following subsection examines the alternative causes of action sometimes employed in these cases.


E. Other Theories Used by Plaintiffs in Trademark Parody Cases
In apparent attempts to cover all the possible bases, trademark parody plaintiffs have raised common law tort claims as alternative theories of recovery. These claims include passing off, misappropriation, defamation and disparagement.

1. Passing Off and Misappropriation
A passing off claim involves the allegation that the defendant competed unfairly by engaging in a scheme to have his goods or services "pass" in the marketplace as those of the plaintiff. [184] An alternative claim for passing off normally adds little to a trademark infringement case, because passing off involves the same sort of likelihood of confusion element essential to a finding of infringement. [185]
Other times, the plaintiff's unfair competition claim alleges that the defendant engaged in misappropriation. The misappropriation doctrine has had a checkered history. [186] It has been successfully employed in a variety of nontrademark contexts where the gravamen of the plaintiff's complaint is that the defendant seized for his own benefit something of value that the plaintiff had built up through expenditures of time, money, and effort. [187] The underlying rationale of this nebulous tort is to prevent defendants from "reap[ing] . . . the harvest of those who have sown." [188]
The application of the misappropriation doctrine to the trademark rights setting is subject to question. A trademark may properly be characterized as an item of value developed through its owner's expenditures of time, money and effort. [189] As such, trademark owners' claims could be actionable under the misappropriation rubric. [190] It must be remembered, however, that the misappropriation doctrine, which developed as part of the common law outside the trademark setting, contains no likelihood of confusion requirement in the usual trademark law sense. [191] Therefore, to recognize misappropriation as a separate common law basis for trademark-related litigation would be to alter substantially the common law's longstanding formulation of trademark rights. [192]
When faced with a trademark owner's misappropriation claim, courts generally seem inclined to avoid extended analysis of whether the doctrine fits in the trademark setting. Some courts have summarily disregarded the misappropriation claim, asserting that trademark law simply does not recognize such a claim. [193] Other courts have indicated that there is no need to consider the plaintiff's misappropriation claim, because such a claim would not provide the plaintiff with any advantages beyond those available under conventional trademark law. [194] In preferring to avoid careful analysis of whether the misappropriation theory may be used to enforce trademark rights, courts have perpetuated uncertainty concerning such claims.
Notwithstanding this apparent judicial reluctance to decide whether the misappropriation doctrine should be granted official recognition in the trademark setting, misappropriation influences have significantly infiltrated trademark law. Sometimes this infiltration is subtle. Although a misappropriation claim is not necessarily brought, some courts apply what amounts to a misappropriation analysis as a justification for a finding of trademark infringement. [195] Other times, the misappropriation influence is more obvious, and the court may even cite misappropriation precedents as a basis for granting the plaintiff relief on conventional trademark grounds. [196] Regardless of whether misappropriation claims are formally recognized or even pleaded by plaintiffs, the apparent receptivity of courts to misappropriation-oriented arguments assures the misappropriation doctrine a continued, albeit unofficial, role in trademark law. [197]

2. Defamation and Disparagement
Plaintiffs in trademark parody cases occasionally plead an alternative claim of defamation. There seems little reason to bother doing so, however, given the other theories potentially available and the meager prospects of success on a defamation claim. Defamation suits are based on actual or presumed harm to the reputation of a person, whether natural or corporate. [198] This requirement of harm to reputation makes it difficult for the trademark owner to prevail in a parody-based defamation action.
When a defamation claim is brought by a corporation or other business entity (the sort of plaintiff often seen in trademark parody litigation), a critical question is whether the defendant's statements would have a tendency to harm the plaintiff's overall business reputation for integrity, competence and solvency. [199] Arguable harm to property such as a trademark or to other economic interests of the plaintiff does not become relevant in a defamation case unless the requisite tendency to harm the plaintiff's reputation is shown. [200] It stretches plausibility almost to the breaking point to assert that a business plaintiff's overall reputation is harmed by a parody of its trademark--even a parody that is arguably in bad taste. As a result, trademark parody-based defamation claims tend to falter on the basic harm to reputation element, [201] even before the potentially more difficult questions of falsity [202] and constitutional fault [203] are encountered.
The tort of disparagement is potentially better suited to the trademark parody setting. It is classified under the broader heading of injurious falsehood, which concerns itself with false statements that cause harm to the plaintiff's economic interests but do not go so far as to harm the plaintiff's reputation. [204] The name "disparagement" tends to be used when the injurious falsehood case arises in a business context. [205] A disparagement claim is more plausible in the trademark parody setting than a defamation claim, because a trademark provides the sort of economic interest that may be adversely affected for purposes of a disparagement claim. [206] It is possible that the defendant's parody could diminish the value of the plaintiff's trademark, especially where the parody's context is arguably distasteful. [207]
The preceding discussion of harm to the value of a trademark should sound familiar, because similar notions are relevant to dilution claims. Indeed, trademark disparagement and trademark dilution, especially that of the tarnishment variety, bear strong conceptual resemblances to each other. [208] Nevertheless, there are differences in elements of proof that make disparagement more difficult to prove and hence less desirable to trademark parody plaintiffs. [209] In view of its stern proof requirements, it is likely that disparagement will maintain a secondary position behind infringement and dilution claims [210] in trademark parody litigation. [211]

III. THE FIRST AMENDMENT SPECTRUM
By its nature, parody involves expression of some sort. It is therefore potentially entitled to first amendment protection against undue governmental interference. Although some courts apparently prefer to avoid the first amendment thicket whenever possible in trademark parody litigation, [212] the constitutional guarantees of freedom of expression cannot readily be dismissed as irrelevant to such cases. In order to determine the extent and nature of the first amendment's proper role in trademark parody cases--an inquiry the U.S. Supreme Court has not directly undertaken--one must begin by examining relevant first amendment doctrines. [213]
The first amendment flatly states that there shall be "no law" restricting freedom of speech or press. [214] Nevertheless, the government's need to function effectively has led to the consistent judicial determination that the first amendment prohibits only government action that unduly restricts freedom of expression. [215] Incidental government restrictions on expression therefore do not violate the first amendment. [216] Courts balance two competing interests in deciding whether government action unduly restricts freedom of speech: the speaker's first amendment interest and the government's interest in regulating the affected expression. [217] Different sorts of speech may carry different degrees of constitutional protection. [218] The degree of protection, if any, given to speech and the corresponding strength of the speaker's first amendment interest depend upon where the speech falls on the first amendment spectrum. [219]

A. Full First Amendment Protection for Political Speech and Noncommercial
Equivalents
Supreme Court decisions establish that even though the first amendment spectrum is broad enough to include various types of speech, not all speech is protected. [220] Obscene expression, for example, receives no first amendment protection. [221] The same is true of speech that is both designed to incite or likely to incite imminent lawless activity. [222] At the opposite end of the spectrum is political speech. Although the full range of the political speech classification is uncertain, it may safely be said that it includes statements and certain expressive actions regarding the workings of government, major social and public policy issues and persons connected with such matters. [223] Political speech is given full first amendment protection [224] because such expression is considered central to the meaning of the first amendment. [225]
Other types of expression are also entitled to first amendment protection. [226] As the Supreme Court has observed, "[O]ur cases have never suggested that expression about philosophical, social, artistic, economic, literary, or ethical matters--to take a nonexhaustive list of labels--is not entitled to full First Amendment protection." [227] The Court has also emphasized that expression on matters of "public concern" rests comfortably under the first amendment umbrella. [228]
Thus, the novelist, composer, essayist, or commentator is entitled to rely on the freedom of speech or press clauses regardless of whether her work is of a political nature. [229] This necessarily means that the satirist or parodist carries a substantial constitutional entitlement to engage in humorous or caustic commentary without the chilling prospect that legal liability will be readily imposed upon her because of what she has stated. [230] The party who has served as the subject of the satire or parody may find the parody offensive, but offensiveness, without more, cannot serve as a constitutional basis for imposing liability on the satirist or parodist. [231] The Supreme Court recently made this proposition abundantly clear in Hustler Magazine Inc. v. Falwell. [232] Even though Falwell did not involve trademark rights, it is relevant to the issues dealt with in this article because it involved an attempt to impose liability for the harm allegedly caused by a parodist's expression of ideas. [233]
In Falwell, the Court struck down, as violative of the first amendment, an award of compensatory and punitive damages to the Reverend Jerry Falwell. [234] Falwell had brought a claim for intentional infliction of emotional distress against Hustler magazine and its publisher, Larry Flynt, because of a Hustler advertising parody that portrayed Falwell in an unwholesome and unflattering light. [235] The Court reasoned that the offensive character of speech--even when it is calculated to offend, as satire or parody often is-- does not strip the speech of the constitutional protection it would otherwise carry. [236] Therefore, the Court imposed stern first amendment requirements on public plaintiffs who seek to hold parodists liable for intentional infliction of emotional distress. [237]
The Supreme Court again underscored the substantial protection given potentially offensive political speech and its equivalents in Texas v. Johnson, [238] the notorious flag-burning decision. In Johnson, the Court held that a protestor who burned the American flag could not be convicted of a criminal offense for violating a Texas statute. [239] The Court classified the burning of the flag as expressive conduct entitled to first amendment protection, and concluded that the criminal prosecution amounted to an unconstitutional attempt to punish the protestor because of the content of his political expression. [240] The Court emphasized that "[i]f there is a bedrock principle underlying the First Amendment, it is that the Government may not prohibit the expression of an idea simply because society finds the idea itself offensive or disagreeable." [241]
Because the criminal prosecution in Johnson was regarded by the Court as an attempt to restrict the content of political expression, the Court applied "the most exacting scrutiny" to the government action. [242] In other words, the Court ruled that the appropriate test to be applied was the one amounting to full first amendment protection. [243] Under this test, in order for the government to justify a content-based restriction on political speech, it must demonstrate that the restriction was necessary to fulfill a compelling government interest. [244]
In applying this demanding test, the Johnson Court examined the interests asserted by the state in support of the prosecution of the defendant. According to the Court, the first asserted interest--that of preventing breaches of the peace which could result from offensive acts such as flag-burning--was not implicated on the facts of the case. [245] The Court further concluded that even though Texas had a "legitimate" interest in preserving the flag as a symbol of nationhood and national unity, there was no likelihood that the defendant's politically motivated burning of the flag would dilute the flag's symbolic strength in the eyes of the public. [246] The Court therefore held that the prosecution of the defendant was not necessary to preserve the flag as a symbol of national unity. [247] Accordingly, the prosecution of the defendant did not withstand first amendment scrutiny.
Johnson illustrates the difficult burden of justification placed on the government when affected expression is entitled to full first amendment protection. Not all constitutionally shielded speech receives such extensive protection, however. Current constitutional doctrine mandates a distinction between noncommercial speech, which is potentially entitled to full protection, and commercial speech, which receives some, though not full, protection. The following paragraphs examine the constitutional significance of the distinction between commercial and noncommercial speech--a distinction critical to the determination of the proper role of the first amendment in trademark parody cases. [248]

B. Lesser (and Lessening) Protection for Commercial Speech
The distinction between commercial and noncommercial speech is a mainstay of first amendment jurisprudence. Speech classified as commercial was long regarded as being outside the scope of the freedom of speech and press clauses. [249] That view was repudiated in 1976. In Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, Inc., [250] the Supreme Court brought commercial speech under the first amendment umbrella but declined to give it the full protection accorded political speech and its equivalents. [251] The commercial/noncommercial distinction thus assumed its present role. Instead of determining whether the subject speech fell within the first amendment, the distinction now establishes the degree of constitutional protection for the subject speech. [252] The less-than-full protection given commercial speech will be explored in more depth shortly. First, however, it is useful to consider what is and is not commercial speech.
"Commercial speech" is usually defined by the Supreme Court as expression that does "no more than propose a commercial transaction." [253] Advertising intended to promote the sale of a product or service is a typical example of commercial speech. Indeed, the Supreme Court's major decisions involving commercial speech have been advertising cases. [254] Sometimes, however, the Court has broadened the definition of commercial speech to include expression that is solely in the economic interest of the speaker and his audience. [255] This latter definition may allow the commercial speech label to be assigned to certain speech that does more than merely propose a commercial transaction. [256] The Court has not offered definitive guidance along these lines, however
Notwithstanding the lack of clarity concerning the full reach of the commercial speech classification, it may safely be said that a speaker's expectation of profit from his statements does not by itself make the expression commercial speech. Both older and recent Supreme Court decisions emphasize that the mere existence of a profit motive for speech does not reduce the degree of constitutional protection to which the expression is entitled. [257] If speech is to be classified as partially protected commercial speech or as wholly unprotected expression, factors other than the economic motivation of the speaker must justify the classification. [258] This approach to the economic gain issue is constitutionally compelled. Without it, the first amendment guarantees of freedom of speech and press would be little more than "empty vessels" [259] providing insufficient protection for the organized press, book publishers and other similarly situated parties whose expression has profitmaking aspects. [260]
Assuming, however, that speech restricted by government action is properly classified as commercial, a court charged with determining the validity of the government action must make suitable allowances for the partially-protected status of commercial expression. Doing so may be challenging, because the Supreme Court's notion of what constitutes an intermediate degree of first amendment protection has not remained constant. The following discussion focuses on the development and current status of first amendment protection for commercial speech.
The holding in Virginia Board of Pharmacy that commercial speech merits constitutional status was justified primarily by the Supreme Court's recognition of the public interest in the free flow of commercial information. [261] The Court made it clear that any constitutional protection extended to commercial speech was conditioned on the expression's being truthful and about a lawful activity. [262] Except for specifying these conditions, however, Virginia Board of Pharmacy left undermined the specific means of implementing commercial speech's partial first amendment protection. That task was undertaken in Central Hudson Gas & Electric Corp. v. Public Service Commission. [263] The Central Hudson Court developed a four-part test for determining the constitutionality of government action that restricts commercial speech. [264]
The first element of the Central Hudson test asks whether the affected commercial speech pertains to a lawful activity and is nonmisleading. [265] If this question is answered negatively, there is no need to apply the remaining elements of the test because the government action will not violate the first amendment. [266] If the question posed in the first element of the Central Hudson test is answered affirmatively, the affected commercial speech is entitled to first amendment protection. Nevertheless, the government action may be upheld if it clears the hurdles posed by the remaining three elements of the four-part test. [267] Central Hudson established the remaining elements as: (1) whether the government had a "substantial" underlying interest to further in taking the action; (2) whether the government action directly advanced the underlying interest; and (3) whether the government action was no more extensive than necessary to serve that interest. [268]
The partial firs amendment protection crafted in Virginia Board of Pharmacy and Central Hudson contemplated that even truthful commercial speech about lawful activities could constitutionally be subjected to government regulation under appropriate circumstances. Nevertheless, the Supreme Court's decisions prior to 1986 reflected an inclination to view skeptically government attempts to restrict commercial speech that was neither false nor promoted illegal activities. [269] For instance, in Bolger v. Youngs Drug Products Corp., [270] the Court struck down as unconstitutional a federal statute that prohibited unsolicited mailings of contraceptive advertisements. [271] The affected advertisements were presumably truthful, and the sale of contraceptives was unquestionably a lawful activity. Thus, the commercial speech at issue [272] merited first amendment protection under the first element of the Central Hudson test. According to the Court, the government was unable to override this protection because an insufficient nexus between the statute and the government's underlying interests caused the statute to fail the final two elements of the four-part test. [273]
The first decade of decisions after Virginia Board of Pharmacy collectively established the high-water mark for first amendment protection of commercial speech. A 1986 decision, Posadas de Puerto Rico Associates v. Tourism Co., [274] began an apparent lessening of the protection accorded commercial speech. In Posadas, the Court rejected a first amendment challenge to a Puerto Rico regulation that prohibited advertisements promoting casino gambling if the advertisements were directed at residents or citizens of Puerto Rico. [275] Casino gambling was legal in Puerto Rico. [276] Posadas thus stands in marked contrast to the Court's previous decisions which disapproved of government regulation of accurate speech concerning lawful commercial activities.
In upholding the advertising ban, the majority opinion by Justice Rehnquist professed allegiance to Central Hudson's four-part test. [277] The Court, however, altered the previous application of the third and fourth elements of the test. Instead of conducting a careful analysis of the relationship between the advertising ban and the government's underlying interest in protecting residents from the harmful effects of casino gambling, the Court seemed to defer to the legislature's judgment about whether the regulation directly advanced the government interest through a means no more extensive than necessary. [278] Observing that the Puerto Rico legislature would not have enacted the advertising ban if it did not believe that the restriction would be an appropriate means of advancing the government interest, the majority indicated that judicial interference with this "reasonable" judgment by the legislature would be inappropriate. [279]
With its deferential approach, Posadas effectively weakened the Central Hudson test while purporting to adhere to it. More government regulations of commercial speech may logically be expected to pass constitutional muster under Posada's deferential approach to the last two elements of the Central Hudson test than under prior judicial analysis. [280] Even though Posadas did not specifically designate the extent to which first amendment protection for commercial speech was being lessened, the decision nonetheless signalled that a lessening of protection was underway. [281] This signal took a more concrete form three years later in Board of Trustees of the State University of New York v. Fox. [282]
In Fox, the Court focused on the final element of Central Hudson's four-part test for determining whether government action that restricts commercial speech violates the first amendment. [283] As formulated in Central Hudson, the final element inquired whether the government action was no more extensive than necessary to serve the underlying government interest. [284] Some of the Court's previous commercial speech decisions seemed to indicate that this element effectively required a "least restrictive means" analysis. [285] Writing for the Fox majority, Justice Scalia stated that no such analysis was contemplated or required by Central Hudson. [286] The majority reasoned that the references in prior decisions to a least restrictive means approach were dicta because the Court had never actually established that a government restriction on commercial speech must be absolutely the narrowest means of furthering the underlying government interest. [287]
In rejecting the least-restrictive means analysis, the Fox Court held that the final element of the four-part test merely requires that when the government regulates commercial speech, its means must be "narrowly tailored to achieve [its] desired objective." [288] Justice Scalia observed that the Constitution mandated a "reasonable" fit, "not necessarily [a] perfect" fit, between the regulation and the underlying government interest. [289] After Fox's reformulation of the Central Hudson test, a regulation that is more extensive than necessary to serve an underlying government interest may now pass first amendment muster. Under Fox's "narrowly tailored" approach, a restriction that sweeps more broadly than the narrowest possible regulation may still be "reasonably" suited to the advancement of the government interest. [290]
Fox's definite lessening of the protection afforded by the commercial speech test signifies that government restrictions on commercial speech will be less likely to be struck down on first amendment grounds than they would have been under Central Hudson. Although the recent interpretation of the commercial speech test translates into a lessening of the first amendment protection afforded commercial speech, the Fox Court failed to acknowledge that it was effecting such a shift. Instead, the Court purported merely to clarify the degree of scrutiny given to government action that limits commercial speech. [291] Furthermore, the Court expressly denied that Fox amounted to an adoption of a lenient "rational basis" test for determining the constitutionality of restrictions on commercial speech. [292]
Although Fox should not be regarded as having eliminated first amendment protection for commercial speech, it is fairly read as having lowered commercial speech protection to a rung that is still "intermediate," yet farther removed from the highest rung of the first amendment ladder of protection. The highest rung, of course, is reserved for political speech and its noncommercial equivalents. [293] Fox thus signals a widening of the gulf between the respective levels of first amendment protection for noncommercial and commercial speech. [294]

IV. JUDICIAL TREATMENT OF FIRST AMENDMENT INTERESTS IN TRADEMARK PARODY CASES
The relevance of first amendment considerations in trademark parody cases has not always been recognized by courts. Although some of the more recent decisions reflect attempts to deal seriously with parodists' constitutional arguments, most courts deciding trademark parody cases have either ignored the possible first amendment issues or dealt with them in an abbreviated, unprincipled fashion. As a prelude to section V's proposals concerning a proper accomodation of first amendment interests in trademark parody cases, this section of the article reviews how courts have dealt with freedom of expression interests in these cases.
There is no question that litigation over trademark rights is much simpler for courts to resolve if the first amendment thicket is avoided. Of course, it is not unreasonable for a court to avoid this thicket by finding a defendant not liable on other "nonconstitutional" grounds. [295] Where the defendant would otherwise be liable under a substantive theory of trademark rights, however, the possible first amendment aspects of the case should not be dismissed as irrelevant or unnecessary to the decision. Some courts have simply omitted discussion of freedom of expression issues when ruling in favor of the trademark owner. [296] In other trademark parody cases in which the trademark owner prevailed, courts have grudgingly acknowledged that the parodist made a first amendment argument but have rejected it with virtually no analysis or explanation. [297]

A. The "Adequate Alternative Avenues of Communication" Approach
Other courts have rejected the parodists' first amendment argument because of an erroneous application of the Supreme Court's decision in Lloyd Corp. v. Tanner. [298] This decision should not be considered controlling in the trademark parody setting. Lloyd presented the question of whether the first amendment was violated when persons were prevented from distributing handbills concerning the military draft and the Vietnam War at a privately owned shopping center. [299] The Supreme Court concluded that there was no first amendment violation because there was no government action, only the action of a private property owner exercising property rights. [300] The Court also noted that the would-be distributors of handbills had various "adequate alternative avenues of communication" because they could have distributed their materials at other locations. [301]
In Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema, Ltd., [302] the defendants argued that the first amendment should protect them from liability for their use of a likeness of the plaintiff's cheerleading uniform in their sexually explicit movie. [303] The second Circuit, relying on Lloyd, rejected the first amendment argument. The court noted that the plaintiff's trademark was "in the nature of a property right" and therefore need not "yield to the exercise of First Amendment rights under circumstances where adequate alternative avenues of communication exist." [304] The court pointed out that if the defendant's movie was a commentary on sexuality in athletics, there were numerous ways in which the defendant could have made the comment without employing the plaintiff's trademark. [305]
The reliance in Dallas Cowboys on Lloyd involved an erroneous legal analysis that linked two considerably different forms of property: real estate and trademarks. Besides the obvious difference that real estate is tangible property while a trademark is intangible, the two forms of property function quite differently. Real estate is a location at which speech may or may not take place. A trademark, on the other hand, is hardly a location. It is a potential subject of speech or is otherwise incorporated into the content of speech. The trademark parodist pokes fun at a trademark or engages in commentary concerning the trademark, its owner, or some aspect of life. [306] It overextends the trademark owner's rights and devalues the parodist's interests in freedom of expression to conclude that the trademark owner's property rights in the mark necessarily must prevail over another party's ability to make an expressive use of the mark. [307] A proper accommodation of these two competing interests is essential. [308]
Whatever a proper accommodation of these competing interests may be, it is not effected by blindly applying the ill-fitting Lloyd approach to the trademark rights setting. It is one thing to recognize, as Lloyd did, that the real estate owner has a right to restrict the location of another party's speech by saying that the expression cannot take place on the real estate owner's private property. It is quite another to assert, as in Dallas Cowboys, that the trademark owner may automatically exclude her trademark from the list of permissible subjects about which another may speak. [309] The latter approach is constitutionally infirm because it becomes a judicially sanctioned content regulation, rather than a restriction on where speech may take place. [310]
The questionable first amendment framework adopted in Dallas Cowboys has been erroneously adopted in other trademark parody cases. [311] For instance, in Mutual of Omaha Insurance Co. v. Novak, [312] the Eighth Circuit affirmed the issuance of a permanent injunction in favor of Mutual of Omaha against an anti-nuclear weapons activist who sold t-shirts and other items on which "Mutant of Omaha" was printed. [313] The court, citing Dallas Cowboys, rejected the defendant's first amendment argument because there were "adequate alternative avenues of communication" available to the defendant. [314]

B. First Amendment Considerations as an Additional Rationale
Other courts have shown more of a willingness to recognize that freedom of expression considerations play a serious role in trademark parody cases. For instance, in Girl Scouts of the United States v. Personality Posters Manufacturing Co., [315] the court noted that enjoining the defendant's "Be Prepared" poster of a pregnant "Girl Scout" could run afoul of the substantial first amendment protection given to those exercising the "right of satirical expression." [316] Similarly, in University of Notre Dame v. Twentieth Century-Fox Film Corp., [317] the court refused to enjoin a satirical movie that contained a scene in which a team represented to be Notre Dame lost a "wild burlesque of a football game" to a Middle Eastern country's team. [318] The court noted the first amendment protection given to books, motion pictures and the like, and emphasized that whether the movie at issue was "good burlesque or bad, penetrating satire or blundering buffoonery, is not for us to decide. It is fundamental that courts may not muffle expression by passing judgment on its skill or clumsiness, its sensitivity or coarseness; nor on whether it pains or pleases." [319]
In both Notre Dame and Girl Scouts, as well as in other cases examined previously, the courts first concluded that the defendant would not be liable on the substantive trademark rights theory at issue, and then cited first amendment concerns as a further reason for denying relief to the trademark owner. [320] Therefore, because the first amendment issue was not crucial to the outcome, these decisions did not contain enunciations of principled mechanisms for addressing the first amendment aspects of trademark parody cases.

C. The First Amendment as Determinative of the Outcome
Unlike the cases just described, recent decisions of the First and Second Circuits have made first amendment limitations on trademark rights the centerpiece of consideration. These decisions have provided more insight than any of their predecessors into the appropriate accommodation of freedom of expression considerations in trademark parody cases.
In L.L. Bean, Inc. v. Drake Publishers, Inc., [321] the First Circuit overturned an injunction granted against the publisher of High Society magazine following the adult magazine's parody of the L.L. Bean name and catalog style. [322] The trial court had held that the parody violated Maine's anti- dilution statute. [323] The First Circuit rejected the argument accepted by various other courts that dilution by tarnishment could occur on the basis of a trademark's appearing on only one occasion in an arguably unwholesome context. [324] It thus adopted a narrow reading of the typical antidilution statute--a reading that was justified primarily by resort to the first amendment.
Before discussing the constitutional aspects of the case, the L.L. Bean court observed that the rights granted to trademark owners by the substantive theories infringement and dilution are properly restricted to commercial settings. [325] The decision's recognition of the need for a commercial/noncommercial distinction reflects a limitation on trademark rights that is both logical and consistent with relevant constitutional considerations. [326] The court concluded that serious first amendment problems would arise if anti-dilution statutes were interpreted to allow dilution by tarnishment claims to be established against noncommercial parodists just as easily as against commercial parodists. [327] The first amendment protects a broad range of statements that may offend some persons. If anti-dilution statutes created liability for all uses of a plaintiff's trademark in an unwholesome setting, protected expression would be at risk. This risk is particularly serious when the offensive statements are made in noncommercial settings. [328]
Although the L.L. Bean court did not set forth extensive guidelines for determining what constitutes a commercial use of a trademark, the court observed that a defendant who "unauthorizedly merchandis[es] his products with another's trademark" has engaged in a commercial use. The enjoining of such a use under an anti-dilution statute would likely be "a legitimate regulation of commercial speech." [329] Such a use, however, did not occur in L.L. Bean. Labeling High Society's parody of the L.L. Bean name and catalog style "editorial or artistic" in nature, the court determined that the defendant's use was noncommercial. [330] The defendant used the plaintiff's trademarks "solely to identify Bean as the object of its parody." [331] The court held that to allow the plaintiff to enjoin such a parody would be to "improperly expand the scope of the anti-dilution statute far beyond the frontiers of commerce and deep into the realm of expression." [332] As recognized by the L.L. Bean court, the property rights inuring to trademark owners cannot be enforced to the extent of suppressing all speech adverse or offensive to the trademark owners. [333] The court therefore concluded that the anti-dilution statute's reach was limited by the first amendment. [334]
Another significant aspect of the First Circuit's constitutional analysis in L.L. Bean is that it repudiated the "adequate alternative avenues of communication" approach taken by the trial court and other courts in cases such as Dallas Cowboys. [335] In holding that the defendant could have conveyed largely the same content without using the plaintiff's trademarks, the trial court in L.L. Bean disregarded the fact that L.L. Bean was a significant object of the parody. The First Circuit's reversal involved a recognition that granting L.L. Bean relief on the dilution theory would be an impermissible content restriction on speech. [336] As if to answer those who would question whether there was any meaningful content in the parody at issue in L.L. Bean, the court concluded its opinion with language that sheds light on the functions of trademark parodies: The central role which trademarks occupy in public discourse (a role eagerly encouraged by trademark owners), makes them a natural target of parodists. Trademark parodies, even when offensive, do convey a message. The message may be simply that business and product images need not always be taken too seriously; a trademark parody reminds us that we are free to laugh at the images and associations linked with the mark. The message also may be a simple form of entertainment conveyed by juxtaposing the irreverent representation of the trademark with the idealized image created by the mark's owner. While such a message lacks explicit political content, that is no reason to afford it less protection under the first amendment. Denying parodists the opportunity to poke fun at symbols and names which have become woven into the fabric of our daily life, would constitute a serious curtailment of a protected form of expression. [337]
L.L. Bean is not the only recent decision to make freedom of expression considerations the determining factor in a case involving a noncommercial use of another's trademark. In two recent decisions, the Second Circuit applied a first amendment analysis to resolve cases in which the plaintiffs claimed that an "artistic" work by the defendants violated the plaintiffs' trademark rights or other intangible property rights.
The first of these decisions, Rogers v. Grimaldi, [338] did not involve a trademark parody but is significant in that it recognized first amendment limitations on the application of section 43(a) of the Lanham Act, a major legal theory applied in trademark parody cases. [339] In Rogers, dancer-actress Ginger Rogers alleged that the defendants' use of the title "Ginger and Fred" for a movie directed by noted film-maker Federico Fellini violated section 43(a) and her common law right of publicity. [340] The movie told the story of two struggling dancers in Italy during the 1940s. These dancers sometimes imitated the international stars Ginger Rogers and Fred Astaire and were known to their audience as "Ginger and Fred." According to Rogers, the title falsely implied that she endorsed the film in some sense and that the movie was about her. She further alleged that the public was likely to be confused by the film's title in violation of section 43(a) of the Lanham Act. [341] Her right of publicity claim rested on the notion that the title impermissibly appropriated her well-known name and public identity without her consent. [342]
The trial court in Rogers granted summary judgment to the defendants, concluding that their use of the title was artistic in nature and thus protected expression falling outside the permissible scope of the Lanham Act. [343] The Second Circuit affirmed, despite its discomfort with the trial court's first amendment approach of creating a "nearly absolute privilege" for movie titles that are relevant to the content of the movie. [344] Reluctant to go that far, however, the Second Circuit forged a different framework for first amendment considerations applicable to section 43(a) and right of publicity claims concerning titles of artistic works.
In affirming the grant of summary judgment to the defendants, the court observed that books, movies and the like "are all indisputably works of artistic expression and deserve protection." [345] The court also noted, however, that titles of artistic works "are of a hybrid nature, combining artistic expression and commercial promotion," because they are at once intertwined with the protected work and useful in the marketing of it. [346] In addition, the Second Circuit observed that artistic works such as the one at issue in the case are sold in the marketplace along with purely commercial goods and services and may sometimes create a risk of consumer confusion. [347] The court conceded that Rogers had produced some evidence of consumer confusion concerning whether she had endorsed the movie and whether the movie was about her. [348]
According to the Rogers court, any system recognizing first amendment protection for artistic works and their titles must also make suitable allowance for the rights of consumers not to be misled when the works are sold in the marketplace. [349] The court recognized, however, that the "expressive element" of titles of artistic works causes such titles to merit "more protection than the labeling of ordinary commercial products." [350]
The Rogers court cautioned that "[b]ecause overextension of Lanham Act restrictions in the area of titles might intrude on First Amendment values, we must construe the Act narrowly to avoid such a conflict." [351] The court rejected the plaintiff's argument that holding the defendants liable under section 43(a) would not violate the first amendment because the defendants could have conveyed the messages of their film in many other ways without employing the plaintiff's name and without engendering possible consumer confusion. [352] This argument was based on the "adequate alternative avenues of communication" approach adopted by the Second Circuit a decade earlier in the trademark parody case of Dallas Cowboys. [353] In spurning the Dallas Cowboys approach, the court noted that such an approach made insufficient allowance for freedom of expression interests. [354] Instead, the court concluded that a balancing approach should be adopted, with section 43(a) "be[ing] construed to apply to artistic works only where the public interest in avoiding consumer confusion outweighs the public interest in free expression." [355] The Second Circuit also added substance to this balancing approach: In the context of allegedly misleading titles using a celebrity's name, that balance will normally not support application of [section 43(a)] unless the title has no artistic relevance to the underlying work whatsoever, or, if it has some artistic relevance, unless the title explicitly misleads as to the source or the content of the work. [356]
In applying this standard, the Rogers court concluded that the title "Ginger and Fred" was artistically relevant and not explicitly misleading. [357] Therefore, the court reasoned that the prospect of some public confusion about whether the film was about or endorsed by Rogers was outweighed by the first amendment interest in free expression. [358]
As for the right of publicity claim asserted by Rogers, the Second Circuit observed that consideration of first amendment interests was just as important as in the section 43(a) claim, if not more so, because the right of publicity cause of action does not require likelihood of confusion. [359] The court held that in recognition of freedom of expression concerns the right of publicity claim of a plaintiff would not "bar the use of a celebrity's name in a movie title unless the title was 'wholly unrelated' to the movie or was 'simply a disguised commercial advertisement for the sale of goods or services."' [360]
Although Rogers was not a trademark parody case, its first amendment analysis is highly relevant to the trademark parody setting for three major reasons. First, as in Rogers, many trademark parodies occur in the context of works that are more "artistic" than they are commercial. [361] Second, section 43(a) is among the major theories relied upon by plaintiffs in trademark parody cases because of the breadth of its language prohibiting false or misleading representations. Many section 43(a) cases, including the plaintiff's claim in Rogers as well as the typical trademark parody case brought under section 43(a), are subject to a likelihood of confusion standard for the imposition of liability. [362] The section 43(a) claim of Ginger Rogers that her name was used in a confusing fashion is quite similar to trademark owners' claims about the arguably confusing use of their marks in parodies. Third, the right of publicity claim asserted in Rogers is analogous to claims often raised by parodists. The right of publicity, which focuses on the unconsented use of a party's name, likeness, or image, does not require proof of likelihood of confusion. Neither does the dilution theory often employed in trademark parody cases by plaintiffs complaining about unconsented uses of their trademarks. [363]
Three months after Rogers v. Grimaldi was decided, the Second Circuit recognized the decision's applicability to the trademark parody setting in Cliffs Notes, Inc. v. Bantam Doubleday Dell Publishing Group, Inc. [364] In Cliffs Notes, the Second Circuit overturned a preliminary injunction against the defendant's sale of Spy Notes, which parodied the cover and style of the familiar Cliffs Notes summaries of novels. [365] The Second Circuit disagreed with the trial court's conclusion that the defendant's parody violated section 43(a). [366] According to the Second Circuit, the speech at issue in the case was sufficiently similar to that in Rogers to warrant the application of the Rogers balancing test. [367]
In "weigh[ing] the public interest in free expression against the public interest in avoiding consumer confusion," [368] the Cliffs Notes court asserted that the issue of likelihood of consumer confusion must be given close scrutiny. [369] Such an approach is sensible in the context of noncommercial parodies. A cavalier assumption that likelihood of confusion exists may result in giving too much weight to the rights of the trademark owner and insufficient weight to the first amendment rights of the parodist. [370] The Second Circuit concluded that the district court had erred as a matter of law in finding "a strong likelihood of confusion." [371] Instead, there was at most only a "slight risk of consumer confusion" as a result of the defendant's "literary parody." [372] This risk was "outweighed by the public interest in free expression, especially in a form of expression that must to some extent resemble the original." [373]

V. A PROPER ACCOMMODATION OF FIRST AMENDMENT INTERESTS IN TRADEMARK PARODY
CASES
With the discussion and analysis in earlier sections providing necessary background and support, this section considers the role that first amendment considerations should play in trademark parody litigation. It also proposes a framework for accommodating those considerations in different varieties of trademark parody cases. The proposed approach sometimes involves restricting the scope of substantive trademark rights theories in order to minimize their reach into the realm of protected expression. The suggested framework also depends significantly upon a proper balancing of harm to trademark owners on the one hand and the strength of first amendment interests of parodists on the other.

A. Judicial Enforcement of Trademark Rights as Government Action
It is fundamental, as has already been noted, that there can be no first amendment violation unless government action accounts for the alleged restriction on expression. [374] Some courts in trademark parody cases have mistakenly concluded that first amendment arguments raised by defendants should be rejected because of an absence of government action. Those courts have reasoned that granting relief to a trademark owner constitutes merely an enforcement of a private party's property rights rather than the government action necessary for purposes of first amendment analysis. [375]
This approach to the government action question must be abandoned in future trademark parody cases because it is not consistent with Supreme Court precedent in other relevant first amendment contexts. The Court has made clear that judicial enforcement of rights and remedies granted to private parties under state or federal law may constitute government action for first amendment purposes if such enforcement would chill interests in free expression. In the landmark case of New York Times Co. v. Sullivan, [376] the government action invoking the first amendment was judicial enforcement of state defamation law which granted a right to relief for reputational harm. [377]
After New York Times, the numerous Supreme Court decisions setting forth the first amendment aspects of defamation law have all been grounded on the basic notion that judicial enforcement of state defamation law is sufficient government action to implicate first amendment concerns. [378] Similarly, the Court recently found government action for first amendment purposes when a lower court held a magazine liable for the tort of intentional infliction of emotional distress. [379] The Court has also recognized that judicial enforcement of the state law-based right of publicity is government action that may raise first amendment questions. [380]
The respective rights claimed by the holder of the right of publicity, a defamation plaintiff and a trademark owner are quite similar. [381] Therefore, if judicial enforcement of the right of publicity or a rule of defamation law constitutes government action for first amendment purposes, the same conclusion should be reached concerning judicial enforcement of trademark rights. The Supreme Court has recognized that first amendment concerns may be implicated by state and federal statutes dealing with trademarks or trademark- related matters. [382] In light of this recognition, and the first amendment analyses employed by the Court in defamation, emotional distress, and right of publicity cases, it is unreasonable to conclude that judicial enforcement of trademark rights--whether those rights come from federal law or from state statutes or common law--would not constitute government action.
As discussed above, many of the more recent and better-reasoned trademark parody cases have proceeded on the basis that government action exists, for first amendment purposes, when courts enforce trademark rights. [383] These cases properly focused on whether the government action at issue violated the first amendment. Any vestiges of the misguided "no government action" rationale of earlier trademark parody cases should not be perpetuated by courts deciding parody cases in the future. [384]

B. The Need to Distinguish Between Commercial and Noncommercial Parodies
The treatment given by the Supreme Court to first amendment considerations in other relevant contexts compels a conclusion that a commercial/noncommercial distinction is essential to a useful first amendment framework in trademark parody litigation. This subsection examines the reasons for such a distinction in the trademark parody setting and deals with how the commercial/noncommercial distinction should be made. Later subsections will propose a framework that addresses and accommodates the differing first amendment considerations in commercial and noncommercial parody cases.

1. Reasons for Drawing the Distinction
As earlier discussion revealed, the Supreme Court has long distinguished between commercial and noncommercial expression. It initially did so by giving the former no first amendment protection whatsoever. During the past fifteen years, the Court has distinguished between the two by giving commercial speech partial first amendment protection and its noncommercial counterpart full protection. [385] The distinction between commercial and noncommercial expression will not disappear from first amendment jurisprudence for two major reasons. First, the Supreme Court is not inclined to give commercial speech the full protection accorded noncommercial expression. Second, even if the Court continues its apparent course of lessening the partial protection given to commercial speech [386] and ultimately goes so far as to abolish that protection, the commercial/noncommercial distinction would still exist. If the court were to classify commercial speech as wholly unprotected, the commercial/noncommercial distinction would again assume its pre-1970s role of determining whether speech is protected at all. In any event, therefore, the commercial/noncommercial distinction would still be a first amendment mainstay. Accordingly, tying the freedom of expression of trademark parody cases to a distinction between commercial and noncommercial uses is hardly hitching one's wagon to a fading star.
Although the Supreme Court has not decided a trademark parody case, it has decided two significant trademark-oriented cases and a noteworthy parody case outside of the trademark setting. [387] Read together, these cases further underscore the need for a commercial/noncommercial distinction in the first amendment aspects of trademark parody litigation and shed considerable light on how that distinction should be made.
In Friedman v. Rogers, [388] the Court rejected a first amendment challenge to a Texas statute that prohibited optometrists from practicing their profession under a trade name. [389] The Court observed that the use of a trade name is a form of commercial speech. [390] According to the Court, the statute was a legitimate restriction on potentially misleading commercial speech and therefore did not violate the first amendment. [391]
If, as Friedman indicates, using a trade name is a form of commercial speech, the Court presumably would say the same about a trademark because of the similar functions of trademarks and trade names. Moreover, if the use of a trademark is a form of commercial speech, certain trademark parodies must also be commercial speech. In some instances, the parodist employs a version of the parodied trademark to identify the parodist's goods or services. [392] Friedman, as extended in this fashion, would thus seem to indicate that when the trademark parodist makes a trademark-like use of the imitated trademark, the parodist's expression is effectively commercial speech that merits partial first amendment protection if it is not misleading, but no protection if it is misleading.
San Francisco Arts & Athletics, Inc. [SFAA] v. United States Olympic Committee [393] bolsters the conclusion that distinctions must be drawn, for first amendment purposes, between commercial and noncommercial uses of another party's trademark. In SFAA, the Supreme Court considered whether section 110 of the Amateur Sports Act of 1978 violated the first amendment by granting the Olympic Committee an exclusive right to the use of the "Olympic" name and by not requiring the Olympic Committee to prove that an alleged violator's use created a likelihood of confusion. [394] The Court concluded that the exclusive right granted by the statute was basically restricted to policing the marketplace for commercial uses of the "Olympic" name, and that few noncommercial uses would be prohibited by the statute. [395] Therefore, the Court reasoned, the statute was a permissible regulation of commercial speech and a tolerable time, place and manner restriction on the minimal amount of noncommercial expression subject to its provisions. [396]
Although it focused on the peculiarities of the Amateur Sports Act's grant of special privileges to the Olympic Committee, SFAA was essentially a trademark rights case. Echoing a message from Friedman, the Court noted in SFAA that, if other parties used the Olympic name in connection with the sale of goods or services, their uses would amount to commercial speech. [397] Much of the Court's discussion of the first amendment issue in the case focused upon whether other parties' uses of the Olympic name and symbol would be commercial or noncommercial. [398] SFAA thus leads to the conclusion that when other parties use some version of an owner's trademark--whether in a parody or otherwise--those uses may merit different first amendment treatment depending on whether they are commercial or noncommercial.
The Falwell decision also offers insight on the proper handling of trademark parody cases, even though the plaintiff in Falwell was not asserting trademark rights. [399] In Falwell, the Supreme Court struck down an award of damages to Reverend Jerry Falwell and imposed stern first amendment proof requirements on public figure plaintiffs who claim that a defendant's speech constituted intentional infliction of emotional distress. [400] The decision demonstrates the substantial first amendment latitude given to parodists when their parodies are noncommercial in nature. Admittedly, the parody in Hustler magazine had a political connection because of Falwell's role as a political figure in addition to his status as a religious leader. Noncommercial trademark parodies will not always have that connection. Nonetheless, first amendment jurisprudence establishes that a wide variety of noncommercial expression is considered to be the equivalent of political speech and therefore entitled to full first amendment protection. [401]
The Court's opinion in Falwell omitted any discussion of a commercial/noncommercial distinction. Without question, however, the Court considered the parody at issue to be noncommercial expression. [402] The extent of protection given by the Court to the defendant's speech was clearly the full first amendment protection accorded noncommercial speech, not the watered-down variety extended to commercial expression. [403] Falwell would thus seem to stand for the proposition that an arguably offensive parody in a magazine--even a sexually explicit magazine--is likely to be considered noncommercial expression despite the desire of the magazine's publisher to profit economically from the sale of the magazine. Falwell's treatment of noncommercial speech and the Court's recent commercial speech decisions demonstrate the widening gulf between the respective levels of first amendment protection given the two varieties of speech, and the corresponding increase in importance of the commercial/noncommercial distinction. [404]
In some of the more recent parody cases, most notably L.L. Bean, Cliffs Notes and even Mutual of Omaha, courts have recognized that not all trademark parodies are alike: sometimes parodies are engaged in for the purpose of marketing a product or service, whereas other times they are more artistic, literary, or editorial in nature. These courts seem to have concluded that drawing commercial/noncommercial distinctions among trademark parodies would be consistent with general first amendment jurisprudence and sensibly accommodating of the differing natures of parodies. [405] There are two critical questions: 1) How are courts to determine whether a parody should be classified as commercial or noncommercial?; and 2) what should be done with a trademark parody once it is classified as either commercial or noncommercial? Although these recent decisions have given some helpful guidance, they have provided only partial answers to the two questions. [406]

2. Making the Distinction Between Commercial and Noncommercial Parodies
In its first amendment decisions, the Supreme Court has provided three guiding principles that should aid the decisionmaker in applying the sometimes elusive commercial/noncommercial distinction to a trademark parody. The first principle suggested by the Friedman and SFAA decisions is that when the trademark parodist has used his version of a trademark to help market a good or service other than the parody itself, the parody is almost certainly commercial in nature. [407] The same would be true if the parodist used the parody as the name of his business. [408] Conversely, if the parody was not used in one of these commercial ways, the parody should usually be considered noncommercial. For purposes of this principle, a newspaper, magazine, book, movie, or similar item containing a parody would not be considered a "good" or "service" even though it is sold in the marketplace.
The second principle, that speech otherwise appearing to be noncommercial is not transformed into less protected commercial speech simply because the speaker has an economic motive, is a longstanding notion recently reaffirmed by the Supreme Court. [409] For purposes of this second principle, a parody appearing in a newspaper, magazine, book, movie, or similar item will usually be speech that otherwise appears to be noncommercial. Falwell's implicit, yet obvious, classification of the parody in Hustler magazine as noncommercial reinforces this notion. [410] Indeed, if the presence of a profit motive were by itself sufficient to make speech commercial in nature, there would be little meaningful first amendment protection for the press, book publishers and similar parties such as film-makers. [411] Out of necessity, then, speech of this sort should normally be labeled noncommercial because its functions of informing, educating and entertaining outweigh any profit-motivated characteristics.
Under the third guiding principle, if the trademark parody possesses both commercial and noncommercial characteristics, it should be treated as a commercial parody when the commercial aspects predominate and as a noncommercial parody when the noncommercial aspects weight more heavily. This principle comports with the Supreme Court's approach in classifying speech that seems at once both commercial and noncommercial. [412]
A few examples drawn from cases discussed earlier in this article will demonstrate the operation of these three principles. As the examples will reveal, some trademark parodies are relatively easy to classify as either commercial or noncommercial. Others, however, fall quite close to the line and are more difficult to classify.
The parodies of the Anheuser-Busch slogans, "This Bud's for you" and "where there's life there's Bud," are readily classifiable as commercial parodies under the above principles. In the cases involving these slogans, the florists' association and floor wax-insecticide manufacturer used the parodies to market their own service or product. [413] Similarly, the "Lardashe" name assigned by a manufacturer to its large-size designer jeans, [414] the "Gucchi Goo" name affixed to diaper bags by their maker and seller [415] and the "McSleep" and "McBagel's" trade names [416] would all be classified as commercial parodies. Although these parodies may have some entertaining features (depending upon one's sense of humor), their overwhelmingly commercial character compels their classification as commercial under principles one and three.
What parodies are readily classifiable as noncommercial under the foregoing principles? The poster in Stop the Olympic Prison (to the extent that the poster was a parody) would be noncommercial. [417] So would the environmental group's parody of the Reddy Kilowatt character in its literature criticizing the electric utility industry, [418] and the political advertisements employing the "star wars" designation to describe the Reagan administration's strategic defense initiative. [419] All of these parodies were nonprofit disseminations of viewpoints on matters of public concern. But, as has been shown, the presence of a profit motive is not necessarily fatal to a parodist's assertion that her parody was noncommercial. There were also clear noncommercial parodies involved in University of Notre Dame (satirical movie employing the Notre Dame name and "football team"), [420] L.L. Bean and Pillsbury (parodies in adult magazines), [421] Cliffs Notes (parody in a book-like form) [422] and Fisher v. Dees (song that parodied a well-known song). [423] The defendants in these latter cases all hoped and expected to profit economically from their parodies, but well-established first amendment doctrines dictate that these parodies are predominantly characterized by their noncommercial functions of entertaining and providing editorial comment. [424]
In other cases, however, line-drawing is more difficult. The t-shirt, sticker and poster cases are representative of those cases in which the commercial/noncommercial line is blurred. [425] When the parody is on a shirt, sticker, or poster, the parody (not so much the object on which it appears) is the item sold. [426] When the parody itself is the item sold, as opposed to being an aid in the marketing of another good or service, the first principle--relating to whether the parody is used to market a good or service-- does not apply. [427] Yet the producer of a shirt, sticker, or poster should not automatically be equated with the press, the moviemaker and the songwriter for purposes of principle number two, which provides that the mere presence of a profit motive does not necessarily render speech commercial. Principle number three (the balancing of commercial versus noncommercial characteristics and some supplementary considerations) must therefore govern the classification decision in these close-to-the-line cases. These supplementary considerations are: the sort of "speaker" responsible for the parody; the apparent motivations underlying the speaker's use of parody; and the presence and degree of expression of ideas or viewpoints through the parody.
Of the cases involving parodies on t-shirts or stickers, the most difficult is Mutual of Omaha Insurance Co. v. Novak. [428] There, the defendant produced and sold "Mutant of Omaha" shirts that parodied the Mutual of Omaha name and ostensibly offered commentary on issues related to nuclear weapons. The defendant claimed that the shirts he sold were expressions of his anti- nuclear weapons stance. [429] His parody went beyond being humorous and seemingly reflected the expression of a viewpoint on an important public issue. Further, the defendant apparently was motivated to some extent by a desire to communicate this viewpoint.
These arguable indicators of noncommercial use in Mutual of Omaha must be considered, however, alongside other relevant factors that seem to cut the other way. The defendant did not sell the shirts as a one-time endeavor. Instead, he sold reasonably large quantities of them at various locations that were typical outlets for miscellaneous goods. He also sold various other items bearing his Mutant of Omaha designs. [430] The fact that the defendant had several different designs which he used on the shirts and other items is also significant: it gives the impression that, in a business sense, he had launched a whole Mutant of Omaha "line." [431] Consideration of these facts leads to a conclusion that the parody in Mutual of Omaha was predominantly commercial despite having a significant noncommercial component. Therefore, though the call was close, the Mutual of Omaha court was probably correct in finding a commercial use. [432]
Other t-shirt and sticker cases have involved parodies that reflect some blurring of commercial and noncommercial features, but are not so problematic as the parody in Mutual of Omaha. For instance, the "Miami Mice" shirts, which parodied the "Miami Vice" name and very loosely parodied the television show, arguably expressed ideas. Nonetheless, the relatively minimal expressive content of the parodies cannot outweigh the inescapable commercial flavor of the defendants' sales operations. Thus, the "Miami Mice" parodies were commercial. [433] The same is true of Topps Chewing Gum, Inc.'s "Petley Flea Bags" and "Garbage Pail Kids" stickers, which parodied the Tetley Tea trademarks and the "Cabbage Patch Kids" dolls. [434] Whatever expressive content there may have been beyond the humor of the stickers pales by comparison with the magnitude of the commercial operations in which Topps sold the stickers. The mere labeling of a trademark parody as commercial, however, does not necessarily mean that the parodist is or should be liable to the trademark owner. [435]
The previously noted and now settled litigation instituted by Miller Brewing Co. against the Doctors Ought to Care (DOC) organization [436] provides a useful illustration of a parody that is, on balance, noncommercial, despite possessing some arguable commercial characteristics. The "Killer Lite" t- shirts sold by DOC contained arguably humorous parodies of Miller slogans and trademarks. These parodies were designed to communicate-- and did in fact effectively communicate--DOC's critical attitude toward the marketing practices of Miller and other producers of alcoholic beverages. The parodies, when coupled with a listing on the t-shirts of health risks associated with alcohol consumption, helped to express DOC's view that alcohol abuse is a significant public health problem. Although DOC sold the shirts, it used the proceeds to fund its mission of advocacy on public health issues, most notably the dangers posed by alcohol and tobacco. [437] DOC itself was a nonprofit organization. [438] The strong expressive content of the parodies, the forces motivating DOC, and the nature of the DOC organization are factors which lead to a conclusion that the parodies were noncommercial, even though the shirts were actually sold rather than given away.
When parodies appear on posters, commercial/ noncommercial line- drawing is especially difficult. Consistent line-drawing is even more difficult. Posters are troublesome in this context because they are both ornamental products sold in the marketplace and potential vehicles for literary, artistic, or editorial expression. Poster manufacturers are not customarily considered part of the organized press in the same sense as newspapers and magazines, which we necessarily tend to label noncommercial. Yet on the spectrum of potential communicative tools, posters lie closer to newspapers and magazines than t-shirts, for example. [439] Girl Scouts of United States v. Personality Posters Manufacturing Co. and Coca-Cola Co. v. Gemini Rising, Inc. are the leading cases involving parodies on posters. [440] The parodies in these cases possessed commercial and noncommercial traits: both were sold in the marketplace and both contained an expressive and arguably humorous content.
Although the distinction between the two parodies is admittedly narrow, the parody in Girl Scouts should fall on the noncommercial side while that in Coca-Cola should be treated as commercial. [441] The "Be Prepared" poster, with its picture and accompanying inscription, uses humor as a vehicle for implicit commentary on larger social issues such as the sexual "revolution," teenage pregnancy, and availability of contraceptives to minors. [442] The expressive and artistic components of the "Be Prepared" poster exceed those of the "Enjoy Cocaine" poster. Although the use of illegal drugs is a matter of obvious public concern, it is difficult to construe the "Enjoy Cocaine" language as any sort of intended or meaningful commentary on the appropriateness or inappropriateness of drug use. Instead, the poster is merely a witty juxtaposition of a familiar product name and the name of an unwholesome, illegal substance. [443] This juxtaposition may have some expressive value, [444] but not enough to outweigh the otherwise commercial character of the poster. [445]
As has been shown, the making of the commercial/noncommercial distinction in trademark parody cases is necessary, but sometimes difficult. Once the parody at issue has been properly categorized, the next consideration is how to properly accommodate the respective first amendment interests of commercial and noncommercial parodists.
Some of the trademark parody decisions discussed earlier have begun the process of developing a first amendment framework for trademark parody cases. None of them, however, has completed the job. This is primarily because a court must confine its rulings to the issues in the case before it. For instance, in L.L. Bean, the court's first amendment discussion was, for the most part, restricted to the context of a noncommercial parody and a dilution claim, under which likelihood of confusion is not an issue. [446] In Cliffs Notes, the court's first amendment analysis was in the context of a noncommercial parody and a section 43(a) claim under which likelihood of confusion is necessary. [447] Additionally, no court has meaningfully explored the proper first amendment analysis of a case involving a commercial parody and a claim brought under an anti-dilution statute.
The next two subsections set out a comprehensive first amendment framework that accounts for the differences in types of parodies and the differences in the elements of trademark rights theories. The proper accommodation of first amendment concerns in trademark parody cases will be considered first with regard to commercial parodies and then with regard to noncommercial parodies.

C. Accounting for First Amendment Concerns in Commercial Parody Cases

1. When Trademark Owner's Claim Is Brought on Infringement or Section 43(a) Theory
Assuming that the parody before the court has been classified as commercial and that the plaintiff's claim is for trademark infringement or for an alleged violation of section 43(a), there is obviously no need to reach the freedom of expression issue if the court concludes that the parody did not create a likelihood of confusion. In that event, the defendant would not be liable under the trademark rights theory at issue. [448]
If the commercial parody is attacked on infringement or section 43(a) grounds and the court properly concludes that a likelihood of confusion was shown, the parodist's first amendment defense must fail. A commercial parody is a form of commercial speech, which receives no first amendment protection whatsoever if it is misleading. Logically, a commercial parody that creates a likelihood of confusion as to source, endorsement, sponsorship, or affiliation must be treated as misleading speech. [449]
Note that the analysis in the preceding paragraph was based on a court properly concluding that the commercial parody created a likelihood of confusion. As noted earlier, some courts in trademark parody cases have adopted a rather loose view of what constitutes likelihood of confusion, often because of the courts' distaste for the defendant's parody. [450] Such a hasty and unreasoned finding of likelihood of confusion creates constitutional difficulties, because an incorrect finding on that issue necessarily leads to an improper denial of the partial first amendment protection a commercial parody would otherwise receive.
In order to minimize the potential first amendment difficulties created by a cavalier approach to the establishment of likelihood of confusion, courts must give careful attention to the critical likelihood of confusion element. Courts must give special attention to this element in cases involving parodies such as those in Mutual of Omaha and Coca-Cola which are predominantly commercial, but also contain a significant noncommercial component. [451] The careful attention urged here should translate into a disinclination to find likelihood of confusion unless the evidence demonstrates a realistic probability of confusion, not merely some possibility of confusion. [452]
2. When Trademark Owner's Claim Is Brought on Dilution Theory
A common strategy of trademark owners is to sue parodists on a dilution theory. Unlike the trademark infringement and section 43(a) theories, the dilution theory does not require proof of likelihood of confusion as to source, endorsement or affiliation. [453] This is precisely why trademark owners prefer the dilution doctrine. It is also why the dilution doctrine poses a potentially greater threat to the first amendment freedoms of parodists than do the infringement and section 43(a) theories. [454]
The absence of a likelihood of confusion "check" in the dilution doctrine creates a strong prospect of conflict with protected speech. [455] This potential threat to expression must be kept in proper perspective, however. In noncommercial parody cases in which the trademark owner seeks relief under a dilution theory, the threat to first amendment rights may be quite serious. [456] In contrast, with respect to commercial parodies, the danger posed to the parodist's first amendment rights by a dilution claim is generally less serious. The difference in the degree of threat to first amendment rights posed by the dilution doctrine is tied to the different levels of protection given to commercial and noncommercial expression.
It must be remembered that a commercial parody does not receive the full first amendment protection accorded noncommercial speech. When a commercial parody is attacked on dilution grounds, any affected speech associated with the parody will normally be entitled to the partial first amendment protection given commercial speech because the parody probably will be unlikely to cause confusion. [457] This partial protection is made plain by the Supreme Court's test for governmental restrictions on commercial speech. Under this test, the partial protection given to commercial speech means that government action restricting commercial speech does not violate the first amendment if the government possesses a substantial underlying interest and the government action directly advances the underlying interest by narrowly tailored means. [458] If the government passes this test, its action restricting commercial speech automatically passes first amendment muster even though the affected speech was nonmisleading.
In a commercial parody case involving a dilution claim by a trademark owner, the application of the Supreme Court's commercial speech test should proceed in the following manner. The relevant government action would be the imposition of liability on the parodist for the alleged violation of an anti-dilution statute. There would logically be a substantial government interest underlying the enforcement of the anti-dilution statute. For many years, states and the federal government have passed legislation on intellectual property matters in order to prevent unfair business competition. [459] State anti-dilution laws thus have a substantial purpose. Enforcement of an anti-dilution statute by injunctive relief--the form of relief typically allowed by such a statute-- logically advances the state's interest underlying the statute and is reasonably tailored to the state's achievement of that interest. [460]
Therefore, when the governing test for the constitutionality of commercial speech restrictions is applied to trademark parody cases involving a commercial parody and a dilution claim, there is unlikely to be a first amendment bar to the entry of judgment in favor of the trademark owner. This assumes, however, that the court will properly apply the dilution doctrine. In order for the dilution doctrine to avoid running afoul of the partial protection given the commercial parodist, the dilution doctrine must be applied with caution. If it is not cautiously applied, judicial enforcement of the anti-dilution statute will not be sufficiently narrowly tailored to further the proper objectives of the statute and, as such, may prohibit commercial speech that should otherwise be protected. [461]
How, then, should the dilution doctrine be limited in application so not to create first amendment problems by restricting too much commercial speech? Courts should begin by interpreting less expansively the typical anti-dilution statute's alternative routes to a finding of dilution. For cases in which the trademark owner claims dilution by loss of distinctiveness, courts should be skeptical of any such claim if the defendant's use was a one-time use rather than a repeated use or a use that would be repeated with frequency if not enjoined. A defendant's isolated use, particularly parody, is considerably less likely to "whittle away" at the distinctiveness of the plaintiff's mark than is a recurring use. [462] For example, assume that a Kentucky Fried Chicken advertisement parodies, in an obvious fashion, the familiar Ronald McDonald character belonging to fast food rival McDonald's. [463] The single advertisement would not be sufficient to support a dilution by loss of distinctiveness claim unless McDonald's could demonstrate an ongoing use of that sort by Kentucky Fried Chicken or could produce clear evidence that Kentucky Fried Chicken would engage in a long-term use if not enjoined. Ronald McDonald would be no less clearly identified with McDonald's after the rival's single advertising "jab" than before the advertisement. [464] To uncritically assume that isolated parodies in the commercial sense necessarily diminish the distinctiveness of the parodied trademark is to overextend the permissible reach of the dilution doctrine at the expense of legitimate commercial expression.
As a second appropriate limit on dilution claims, courts should be reluctant to find dilution by tarnishment simply because the trademark owner--and perhaps the court as well--does not like the defendant's parody. Simply because the plaintiff may find the defendant's humorous commentary unwelcome does not necessarily make the context unwholesome and therefore a possible candidate for a dilution by tarnishment claim. [465] An example of a court's falling into this trap is the previously discussed litigation over the "Garbage Pail Kids" stickers, which parodied the "Cabbage Patch Kids" dolls. In that case, the court rushed to find dilution by tarnishment, seemingly because neither the plaintiff nor the court was pleased with the connotations suggested by the Garbage Pail Kids. [466]
By nature, however, a parody is not pleasing to the object thereof. Even though commercial parodies are not fully protected under the first amendment, their partially protected character must allow them reasonable latitude to offend, as long as any offensiveness stems from ideas expressed in the parody rather than from a truly unwholesome context. [467] The parody in the Garbage Pail Kids litigation was hardly unwholesome when compared to the contexts more commonly considered to be unwholesome: those displaying sexual explicitness or referring to illegal drug use. [468]

D. Accounting for First Amendment Concerns in Noncommercial Parody Cases
The question of how to accommodate first amendment interests in trademark rights cases involving noncommercial parodies has an easy and tempting answer: Courts need not worry about first amendment issues in such cases because trademark owners do not have valid claims against noncommercial users of their marks. Thus, the trademark owner's claim against the noncommercial user should be succinctly dismissed without the court's having to address potentially complicated constitutional questions.
This answer was given by the court in Lucasfilm Ltd. v. High Frontier. [469] In Lucasfilm, the owner of the "Star Wars" trademark failed to obtain relief on a variety of theories against a public interest group that used the term "star wars" in political advertisements protesting the Reagan administration's strategic defense initiative. [470] The answer was an appropriate one in the context of that case, where the defendant's use of the mark was as unambiguously noncommercial as possible. Not all uses that are classified as noncommercial for purposes of first amendment analysis are exclusively noncommercial, however. For instance, books, magazines, newspapers, movies and similar items are properly given the noncommercial label even though they have the arguably commercial feature of being sold in the marketplace. [471]
Courts have concluded that trademark rights theories have the potential to reach predominantly noncommercial uses when a secondary commercial component of the sort just noted is present. The Second Circuit has recently recognized that a movie title and a parody in book form were assailable under section 43(a) of the Lanham Act. [472] First amendment concerns, however, were sufficient on the facts of each case to keep liability from being imposed. Some noncommercial parodies--though not always recognized as noncommercial by courts--have been the subject of trademark rights litigation won by trademark owners. [473] Trademark owners have prevailed in these cases despite the full first amendment protection that noncommercial parodies should merit. These decisions demonstrate the need for a suitable framework to accommodate freedom of expression concerns.

1. When Trademark Owner's Claim Is Brought on Infringement or Section 43(a) Theory
When a parody is properly classified as noncommercial for first amendment purposes and the trademark owner proceeds on an infringement or section 43(a) theory, the familiar likelihood of confusion element of course becomes critical. To properly account for the full first amendment protection given to noncommercial parodies, a shift in focus on the likelihood of confusion element is needed. Whereas a commercial parody that is likely to cause confusion obtains no first amendment protection, [474] a noncommercial parody that may cause confusion does not necessarily forfeit first amendment protection. In the realm of noncommercial speech, there is considerably greater tolerance for misleading statements and even falsehoods.
In the unusual event that a noncommercial parody is found to create a likelihood of confusion, [475] the first amendment should nevertheless insulate the parodist from liability. In the Rogers case, for instance, there was reasonable evidence tending to indicate that the movie title used by the defendants created consumer confusion concerning whether the movie was about or endorsed by Ginger Rogers. Despite this evidence, the Rogers court held that the first amendment protected the defendants from liability. [476] In doing so, the court called for an approach that balanced the extent of the harm to the plaintiff, who claimed a violation of section 43(a), against the strength of the expressive interests of the defendants. [477]
Although Rogers was not a trademark parody case, its similarities make it useful in the trademark parody context. The balancing approach contemplated by Rogers provides a useful means of giving analytical substance to the full first amendment protection the noncommercial parodist receives. In balancing the harm to trademark owners against the expressive interests of the noncommercial parodist, it is also necessary to commence this balancing process in an uneven fashion by placing heavier weight on the parodist's side of the scale. [478] This balance can be achieved by establishing a rule that the trademark owner cannot prevail on an infringement or section 43(a) claim against a noncommercial parodist unless he or she proves a strong likelihood of confusion. This proposed standard of strong likelihood of confusion [479] is consistent with the constitutionally valid notion that it should be quite difficult for a trademark owner to win an infringement or section 43(a) claim against a noncommercial parodist. If full first amendment protection means that government action restricting the content of such speech cannot pass constitutional scrutiny unless it is necessary to the fulfillment of a compelling government interest, it would seem that any government interest that is not intended to prevent a strong likelihood of confusion should not suffice. [480]

2. When Trademark Owner's Claim Is Brought on Dilution Theory
The absence of a likelihood of confusion element--let alone an element of strong likelihood of confusion--makes the dilution theory a potentially broad- ranging device that may extend far into the realm of protected speech. Because first amendment concerns are so strong with respect to noncommercial speech, the threat posed by the dilution doctrine to freedom of expression is even greater in the noncommercial context. Therefore, it is necessary to restrict the dilution doctrine within appropriate limits and to again employ a version of the Rogers balancing approach.
In restricting application of the dilution doctrine in the noncommercial setting, it would be appropriate to begin by limiting the use of the dilution by loss of distinctiveness theory to cases against commercial users of the plaintiff's trademark. The dilution by loss of distinctiveness theory is ill- fitting in the noncommercial parody context because it contemplates commercial trademark-like uses by a defendant rather than what is customarily seen in the noncommercial setting. In addition, the noncommercial parody is unlikely to reflect the repeated use factor that is often important to loss of distinctiveness claims. Thus, results such as that in Dallas Cowboys, where dilution by loss of distinctiveness was found, [481] would be eliminated by this approach.
The dilution by tarnishment theory is more problematic. The L.L. Bean court apparently would eliminate this possible basis of liability in the noncommercial parody context. [482] It is not necessary, however, to ban all such claims from the noncommercial context as long as courts restrict tarnishment claims to instances where the alleged tarnishment of the trademark's reputation stems from a truly unwholesome context in which the mark was used by the parodist, and not from the unwelcome impact of the ideas being expressed. [483] For example, if the Miller Brewing Company suit against Doctors Ought to Care (DOC) had gone to trial instead of being settled, [484] dilution by tarnishment would not have been found in that case under this proposed restriction. Any tarnishment of the Miller name and slogans came from the striking content of the ideas and viewpoints expressed by DOC. To impose dilution liability in such an instance would penalize the noncommercial parodist for exercising legitimate and substantial expressive rights and would hardly be consistent with the first amendment prohibition against content restrictions on fully protected speech. [485]
Cases resembling Pillsbury and Dallas Cowboys make analysis of a dilution by tarnishment claim more difficult. In these cases, the offensive parodies appeared, respectively, in a sexually explicit magazine and a sexually explicit movie. [486] It is more difficult for parodists in such factual settings to argue credibly that the powerful force of their ideas caused the trademark owner to complain. Although the parodies in Pillsbury and Dallas Cowboys presumably were not devoid of ideas, they were not brimming with profound content. [487] Any arguable tarnishment of the parodied trademarks in such cases would seem to stem from the unwholesome context of the parodies, not from the expression of the parodists' ideas or viewpoints.
The analysis should not stop at this point. Even if it appears that the unwholesome context of the noncommercial parody (rather than the unwelcome content of the ideas expressed therein) is the apparent "culprit," the court should not immediately find dilution by tarnishment. Instead, the Rogers balancing approach must be applied. In applying this approach, the court must take a serious, realistic look at the likelihood of tarnishment and whether there is meaningful evidence tending to show this likelihood. Without such evidence, a successful dilution by tarnishment claim would seem to rest primarily on a knee-jerk reaction to the parody's unwholesome context. Regrettably, such a poorly grounded ruling is not uncommon in trademark parody litigation, [488] but is constitutionally inadequate because it ascribes too much weight in the balancing process to the supposed harm to the trademark owner and too little weight to the first amendment interests of the noncommercial parodist. [489]
As a further means of keeping dilution by tarnishment claims against noncommercial parodies within permissible first amendment bounds, the court should also consider the audience. For instance, if the parody appeared in an adult magazine, as in Pillsbury, the court should consider the likelihood of tarnishment issue in terms of whether the magazine audience's attitude toward the mark would be adversely affected. If the answer to this question in "no"-- the answer one would expect in a Pillsbury-like case--the court should be reluctant to find a genuine likelihood that the mark's reputation would be tarnished. [490] This careful scrutiny of dilution by tarnishment claims is warranted by the first amendment. It demands the conclusion that the mere placement of a trademark parody in an unwholesome but noncommercial context should not automatically give rise to a meritorious dilution by tarnishment claim by the trademark owner.

VI. CONCLUSION
Trademark parody cases present a significant conflict between the property rights asserted by trademark owners and the freedom of expression concerns advanced by parodists. The entertaining facts of these cases sometimes belie the seriousness of the constitutional issues at stake. It is noteworthy that some of the very recent trademark parody decisions display a judicial willingness to address the first amendment aspects of these cases on a more principled basis than had often been displayed previously. Nevertheless, the cases reveal a need for a coherent analytical framework that accommodates the different first amendment considerations present in different varieties of trademark parody cases. This article has offered proposals to that end.
The analytical framework explained herein recognizes and accounts for the practical and constitutional differences between parodies that are commercial (meriting at best partial first amendment protection) and parodies that are noncommercial (meriting full first amendment protection). The article has also addressed the proper implementation of these respective levels of protection in the differing contexts of the infringement, dilution and federal statutory claims brought by trademark owners against parodists. It has urged the adoption of appropriate modifications of or restrictions on these trademark rights theories in order to minimize their potential for reaching too far into the realm of protected expression. If courts adopt this article's analytical framework, trademark parody decisions will reflect the drawing of a far better balance between trademark owners' property rights and parodists' first amendment interests than has generally been the case until now.