University of Notre Dame

Against Secondary Meaning

November 16, 2022

View PDF



Jeanne C. Fromer*

Trademark law premises protection and scope of marks on secondary meaning, which is established when a mark develops sufficient association to consumers with a business as a source of goods or services in addition to the mark’s linguistic primary meaning.  In recent years, scholars have proposed that secondary meaning play an even more central role in trademark law than it already does.  Yet enshrining secondary meaning in the law undermines the ultimate goals of trademark law: promoting fair competition and protecting consumers.  The dangers of enshrining secondary meaning include the problematic doctrine that has built up to assess it or presume it, including the ease of establishing secondary meaning and inaccurate categorizations for assessing protectability and scope; the competitive inequalities secondary meaning creates which hurt smaller and newer businesses; and neglect of competitive harms caused by protecting marks whose primary meaning is too conceptually related to the associated goods or services.  Alongside these dangers, businesses intrinsically have the incentive to establish secondary meaning regardless of whether trademark law requires it, making its enshrinement less necessary than the law and scholars suppose.  Trademark law therefore ought to dethrone secondary meaning from its central role in establishing protectability and scope.  Instead, primary meaning ought to be used to gauge protectability by assessing how conceptually related a mark is to its associated goods or services.  Secondary meaning should also serve a more carefully tailored role in assessing scope, with a sliding scale of secondary meaning.  These adjustments would address the dangers caused by the current role of secondary meaning and would help restore trademark law’s promotion of fair competition and consumer protection.


BOOKING.COM for online travel booking services.  “FOR WALKING” (with quotation marks) used on Off-White footwear.  VAGISIL for vaginal-health products.  All of these terms have been involved in recent trademark disputes, a core issue being that protection for these terms arguably runs contrary to trademark law’s goals of fair competition and consumer protection because of how closely associated each term is conceptually with the category of goods or services for which it is being used.  (In fact, BOOKING.COM touted this close conceptual connection in its widely seen Super Bowl ad this year in which it refers to its name as “lit” for literal.)  In particular, giving one business exclusive rights in these marks can prevent competing businesses from using these or similar terms, granting the markholder an outsized competitive advantage.

Even so, each instance also underscores how trademark protection turns on consumers’ perception that a word or symbol is a trademark signifying goods or services, known in trademark law as “secondary meaning”—that is, meaning in addition to its linguistic primary meaning.  With regard to BOOKING.COM, the Supreme Court held that even though “[a] generic name—the name of a class of products or services—is ineligible for federal trademark registration,” BOOKING.COM is protectable because “[c]onsumers . . . do not perceive the term . . . to signify online hotel-reservation services as a class” but rather as a term with secondary meaning.  That is, a “” term is not always generic and might be protectable based on the presence of secondary meaning.  For “FOR WALKING,” the Patent and Trademark Office (PTO) refused to register the mark to Off-White on the ground that it is descriptive of footwear despite the fashion business’s argument that “the use of the quotation marks changes the commercial impression and therefore creates a non-descriptive mark.”  Off-White can nonetheless overcome this refusal and register the mark if it responds by submitting sufficient evidence that the term has secondary meaning.  As to VAGISIL, the Fourth Circuit recently allowed the markholder to oppose the trademark registration of VAGISAN for competing products even though the mark was closely associated conceptually with vaginal-health products because of the strong secondary meaning of VAGISIL.

More generally, trademark law premises protection for a term or symbol on either a showing of secondary meaning or an irrebuttable presumption that the term or symbol has secondary meaning.  It does so through its rules of distinctiveness, allowing marks that are descriptive of the goods or services they signify to be protectable upon a showing of secondary meaning and protecting marks that are suggestive of the goods or services they signify, arbitrary in relation to them, or coined terms without any further showing on the reasoning that these marks intrinsically possess secondary meaning.  The law also awards broader scope to marks that have well-established secondary meaning.10   Atop that, trademark law makes it easy to establish secondary meaning, especially for deep-pocketed businesses.11 

Yet these rules that base protection and trademark scope on secondary meaning undermine trademark law’s goals when there is a close conceptual connection between the mark and the associated goods or services, as in the examples just noted.  As discussed herein, this close conceptual connection can exist regardless of whether the law classifies the mark as descriptive, suggestive, arbitrary, or fanciful.12   When this close conceptual connection exists, a business securing trademark rights in such a mark will inherently possess a tremendous commercial advantage over its competitors.  All else being equal, the business can lure in customers more easily than competitors using a less conceptually related mark and it can assert its trademark rights to prevent competitors from choosing a mark that is similar to its advantageous mark.13   That is precisely antithetical to trademark’s goals of fair competition and consumer protection in allowing businesses to distinguish themselves via their marks as source indicator and repository of goodwill.14   Indeed, trademark protection in this scenario does the contrary by creating unfair competition and consumer harm.15 

Even so, recent scholarship has sought to make secondary meaning yet more important in trademark law and enshrine it as the basis for protecting just about any term as a mark, even otherwise generic terms like COMPUTER COMPANY for a computer company.16   The Supreme Court, in its ruling on BOOKING.COM, seems headed in the same direction.17   The move to enshrine secondary meaning seems to be about grounding trademark protection in what consumers actually perceive as terms indicating product source.  Yet it is too easy—especially for well-resourced firms—to show that consumers perceive terms as source-indicating, including for marks whose protection undermines trademark law’s goals.  Making it harder to establish secondary meaning as a way to promote the law’s goals in turn creates undesirable competitive consequences, making it too tough for new entrants or smaller businesses to establish secondary meaning.18   Additionally, the law conclusively presumes secondary meaning for certain marks where that frustrates trademark’s goals.19   Moreover, even when marks have secondary meaning, they retain their primary meaning, which might be closely related enough conceptually to the associated goods or services that protection impairs trademark law’s goals.20   To the extent that establishing secondary meaning is an independent goal trademark law deems worth encouraging, legitimate businesses have strong incentives external to trademark law to develop secondary meaning anyhow.21   For these reasons, trademark law should not be centered on secondary meaning because the consecration of secondary meaning counters trademark law’s fundamental goals of fair competition and consumer protection.

This Article thus advocates against enshrining secondary meaning as a basis of protectability in trademark law.  In its place, the Article proposes recentering distinctiveness doctrine on the primary meaning of terms as the gauge of protectability.  That is, a word, image, or other symbol should be protectable as a trademark only when there is enough conceptual distance between the primary meaning of that symbol and the goods or services for which it is being used.  Recentering distinctiveness on primary meaning and moving away from secondary meaning would better promote trademark law’s goals by encouraging businesses to choose certain marks and avoid others at the outset.  Furthermore, because it is difficult to eradicate considerations of secondary meaning from assessment of the scope of trademark rights, the Article recommends that a sliding scale of secondary meaning be established for that assessment.  A sliding scale would make the requisite showing of secondary meaning directly proportional to the size of a business’s footprint as a way to navigate between the relevance of assessing secondary meaning for scope and the problematic consequences of doing so.

Part I explores the role of secondary meaning in trademark law with regard to both protectability and scope of protection.  Part II turns to the dangers of enshrining secondary meaning based on current doctrine, consumer understandings of source indications, and incentives external to trademark law to develop secondary meaning.  Part III proposes instead recentering trademark distinctiveness doctrine on primary meaning as the gauge of protectability.  Part IV presents a sliding scale for secondary meaning in assessing the scope of trademark rights.


Words, symbols, logos, and sometimes a product’s design or packaging may be protected under trademark law.22   According to the Lanham Act, these are protectable under federal law so long as they are “used by a person” in commerce “to identify and distinguish his or her goods . . . from those manufactured or sold by others and to indicate the source of the goods, even if that source is unknown.”23   Federal law similarly protects marks that designate services.24   Distinctive marks used in commerce are protectable, either via registration and enforcement,25  or through a provision allowing enforcement of unregistered marks.26   Federal law protects trademark registrants against another’s “use in commerce [of] any reproduction . . . of [their] registered mark in connection with the sale, offering for sale, distribution, or advertising of any goods or services on or in connection with which such use is likely to cause confusion, or to cause mistake, or to deceive.”27   Unregistered marks are similarly protected.28 

Trademark law originated to protect indications of an article’s source by guarding producers from competitors’ illegitimate interference with their trade.29   The law has broadened beyond these narrower purposes to also emphasize protecting consumers from confusion to foster fair competition.30   The increased production capacity for goods ushered in by the Industrial Revolution boosted instances of advertising to the public to distinguish goods, which popularized trademarks as identifiers of the source of goods.31   As time marched on, consumers began to know less and less about the specific source of particular goods.32   Nonetheless, a trademark affixed to goods would help consumers know that those goods originated from the same source, whatever that source was.33   Relatedly, a trademark would help distinguish a product from others with different marks.34 

Trademark protection strives to bolster trade, as Frank Schechter explains, by “identify[ing] a product as satisfactory and thereby . . . stimulat[ing] further purchases by the consuming public.”35   Scholars theorize that producers of trademarked goods will have the incentive to invest in the goods’ quality.36   They suggest that this investment will occur because consumers will use the trademark as a way to identify a desirable good only if their past experiences reliably forecast the good’s worth.37   Protecting against trademark infringement, from this vantage point, thus prevents others from trading on the goodwill represented by the trademark.38   Businesses might then leverage the goodwill engendered by consumers’ association of a particular mark with positive value and meaning by using the mark for an ever-wider range of goods and services.

In these ways, trademarks also aim to reduce consumers’ search costs—the expenditures they must make to discern important qualities of goods or services, which are frequently hard to measure.39   Therefore, trademark protection guards against use of a too-similar mark that is likely to cause consumer confusion as to goods’ or services’ origin.40   In these ways, trademark law seeks to promote fair competition and protect consumers.

With that background, this Part delves into the role secondary meaning plays in establishing the protectability of trademarks and the scope of trademark rights.  Section A sets out how secondary meaning informs the protectability of all trademarks.  Section B shows how the scope of trademark rights also turns on secondary meaning.

A.   Protectability

Trademark law’s distinctiveness requirement is situated in the Lanham Act’s rule that a mark “identify and distinguish [a business’s] goods . . . from those manufactured or sold by others and . . . indicate the source of the goods.”41   As Barton Beebe explains, “a trademark is distinctive of source if it is recognized by consumers as a designation of the source of the product to which it is affixed rather than as, say, a decoration on or a description of that product.”42   This source distinctiveness is also known as “secondary meaning,” as this association between source and product supplements the mark’s linguistic primary meaning as a word, image, or the like.43   Protectability turns on distinctiveness for a cluster of related reasons.  First, consumers would never be likely to be confused as to source—trademark’s central inquiry for infringement—unless they recognize a particular designation as source-indicating in the first place.44   Relatedly, consumers can reduce their search costs of associating goods or services with a particular source—something trademark law seeks to encourage—only if they know that the term or symbol associated with them is a source designator.45   Moreover, from a business’s perspective, if consumers know to associate a term or symbol with the business as the source of goods or services, it will be encouraged to invest in the quality of its goods or services, an important goal of trademark law.46   In addition to the benefits of establishing distinctiveness, there is cost to granting trademark rights for marks that are not distinctive.  The principal worry is that trademark rights in such a case would inefficiently prevent other businesses from using terms or symbols as they compete that fail to distinguish source.47 

A framework to assess distinctiveness and thus protectability in trademark law is set out most famously in Abercrombie & Fitch Co. v. Hunting World, Inc., a 1976 Second Circuit decision authored by Judge Friendly,48  and on which other courts—including the Supreme Court—rely.49   Abercrombie laid out five categories of marks—generic, descriptive, suggestive, arbitrary, and fanciful—to sort terms and assess their protectability.50   As explained by the court, a term is categorized based on its primary meaning in relation to the category of goods or services for which the term is being used.51 

As per Abercrombie, a generic term “refers, or has come to be understood as referring, to the genus of which the particular product is a species.”52   The court provided as an example the of-its-era reference of IVORY, which would be generic when used to indicate the product made from elephant tusks (now principally illegal to import into or sell in the United States53 ) but not soap (a less iconic brand now than in the 1970s54 ).55   Other examples include URGENT CARE for urgent care medical centers56  and SUMMER JAM for a musical concert occurring during the summer months.57   Generic terms are never protectable under trademark law, even if they accrue secondary meaning.58   The reason for this ban on protection is for the sake of fair competition and consumer protection: competitors ought to have the absolute right to call their goods or services by their category name rather than be deprived of it because one business has exclusive rights in it as a trademark; if they could not, consumers would concomitantly be disserved too through confusion and unfair competition.59 

Abercrombie situated descriptive terms on better footing for protection.  As explained by the Second Circuit, a descriptive mark “describe[s] a product or its attributes,”60  with the Seventh Circuit giving as examples HOLIDAY INN for inns in which people stay while on holiday, ALL BRAN for all-bran cereal, and AMERICAN GIRL for American girl dolls.61   As the Seventh Circuit elaborated, “[a] descriptive mark is not a complete description, . . . but it picks out a product characteristic that figures prominently in the consumer’s decision whether to buy the product or service in question.”62   Similar to generic terms, the fear with protecting descriptive marks is that competitors might want to use a term because it describes their product too and also be unfairly disadvantaged if one business in the competitive landscape has exclusive rights in such a term.63   Moreover, descriptive terms are understood to lack inherent distinctiveness because they describe the products to which they are affixed and thereby do not readily signify source.64   Trademark law nonetheless allows descriptive marks to be protected if they acquire secondary meaning.65   The implication of this rule is that descriptive marks cannot be protected immediately upon use until this secondary meaning has developed.66   As Judge Friendly explained in Abercrombie, in allowing protection for descriptive marks that have acquired secondary meaning, trademark law “strikes the balance . . . between the hardships to a competitor in hampering the use of an appropriate word and those to the owner who, having invested money and energy to endow a word with the good will adhering to his enterprise, would be deprived of the fruits of his efforts.”67 

Secondary meaning can be shown through direct or circumstantial evidence, and courts weigh this evidence using a multifactor test to assess whether there is a “mental association by a substantial segment of consumers and potential consumers ‘between the alleged mark and a single source of the product.’”68   Though the circuits’ tests are somewhat different, the Fifth Circuit’s seven-factor test is representative, looking to “(1) length and manner of use of the mark or trade dress, (2) volume of sales, (3) amount and manner of advertising, (4) nature of use of the mark or trade dress in newspapers and magazines, (5) consumer-survey evidence, (6) direct consumer testimony, and (7) the defendant’s intent in copying the [mark].”69   Despite variations, surveys of secondary meaning tend to ask whether respondents associate a claimed trademark with one or more than one company.70   If enough respondents associate it with one company, secondary meaning is established.71   While these surveys are generally well accepted, there are concerns with them, including that respondents might associate a claimed trademark with only one company merely because it offers a unique product72  and respondents might not associate a claimed trademark with any company at all.73 

Despite generic marks not being protectable, the Supreme Court recently held that an assessment of secondary meaning can be as relevant for the protectability of marks that might potentially be classified as generic.  In ruling that BOOKING.COM for online travel-reservation services was not generic, the Court relied in large part on a survey credited by the district court in the case that 74.8% of respondents thought BOOKING.COM was a brand name rather than a generic name.74   More broadly, the Court established that terms of the form “” are not necessarily generic and are protectable if “consumers in fact perceive that term as . . . [one] capable of distinguishing among members of the class” of goods or services.75   This ruling underscores how evidence of secondary meaning (and its close cousin, capability of having secondary meaning) can move a term that might plausibly be generic into the descriptive category, thus rendering it protectable.76 

All other marks—suggestive, arbitrary, and fanciful terms—are protectable without proof of secondary meaning as inherently distinctive.77   As explained by the Fifth Circuit, a suggestive mark “suggests, rather than describes, some particular characteristic of the goods or services to which it applies and requires the consumer to exercise the imagination in order to draw a conclusion as to the nature of the goods and services.”78   Courts have found to be suggestive SWAP for a watch with interchangeable parts,79  5 HR ENERGY for an energy drink,80  and GLASS DOCTOR for glass installation and repair services.81   Arbitrary marks are preexisting words that are used in a way that is conceptually unrelated to the category of goods or services at hand.82   Examples of marks courts have classified as arbitrary are STARBUCKS for coffee,83  VEUVE (meaning “widow” in French) for champagne,84  and KIRBY for vacuum cleaners.85   Fanciful marks, as per Abercrombie, are terms (typically words) “invented solely for their use as trademarks”86  (or an “obscure or archaic term not familiar to buyers”87 ).  Courts have deemed to be fanciful CARSONITE for highway markers88  and LUMAR for fabric softener.89   The law understands these three categories of terms to be easily protectable because, as the Second Circuit put it, they do not “depriv[e] others of a means of describing their products to the market.”90 

Even though there is no requirement that secondary meaning be shown to protect inherently distinctive marks, secondary meaning remains central to protection of inherently distinctive marks, as with the other categories.  That is, inherently distinctive marks are given an irrebuttable presumption that they have automatic secondary meaning—that consumers will understand these marks as source indicators as soon as they start being used—because of their categorization as suggestive, arbitrary, or fanciful.91   In that sense, secondary meaning is at the core of protection of suggestive, arbitrary, and fanciful marks too.  The notion is that anyone who sees such terms or symbols affixed to a product will immediately understand that the only possible reason they are so affixed is to serve as a trademark because they have so little conceptual relationship to the underlying goods or services.92   Therefore, suggestive, arbitrary, and fanciful marks can be protected upon first use, unlike descriptive marks.93 

In all, this overview of protectability reveals that secondary meaning is relevant across the Abercrombie distinctiveness spectrum, whether to show that a mark has developed secondary meaning, to presume that a mark inherently has secondary meaning, or to show whether a mark is understood to be generic or have (or be capable of having) secondary meaning.

Courts understand the distinctiveness spectrum as striking a balance, as the Second Circuit put it, between conferring trademark rights on businesses using marks to promote competition and help consumers and not “creat[ing] a ‘linguistic monopoly’ which would stifle competitors’ efforts to market similar goods to consumers.”94   In this framework, denying rights in generic terms permits all competitors in the category to use the term to refer to their goods or services (such as COMPUTER for computers) without any single business having rights in the term as a mark.95   As to descriptive terms, Tom McCarthy relatedly explains that “all sellers should be free to truthfully use [such] terms to describe their [merchandise]” and “[n]o one seller should be given the exclusive right to describe a product by an important characteristic.”96   At the other end of the spectrum, fanciful marks are thought to present little concern to competitive goals.  As Jake Linford explains, “[l]ack of connection between mark and product means there is likely no competitive harm if the mark is granted broad protection from first use.”97 

As discussed in previous work, this spectrum would seem to encourage businesses to adopt fanciful or arbitrary marks more than other types because they can be easily adopted without a need to establish secondary meaning.98   Moreover, even though suggestive marks can be protected without secondary meaning, it would also seem that they are less likely to be chosen than fanciful or arbitrary marks due to the risk that a mark thought to be suggestive might be classified as descriptive—given the gray border between the two99 —and thereby require evidence of secondary meaning for protectability.100   It would furthermore appear that businesses might avoid choosing a descriptive mark because of the need to show secondary meaning, as well as the fear that it gets categorized as generic and is thus unprotectable.101   Nonetheless, these structural incentives do not fully account for the realities of showing secondary meaning as discussed in Part II, which warp this picture.  In light of these realities, this Article returns to a broader view of incentives to choose certain trademarks over others in Part III.

B.   Trademark Scope

Secondary meaning is not pertinent only to the protectability of trademarks.  The breadth of trademark protection also turns on secondary meaning.  Specifically, courts have developed doctrine that the likelihood of confusion with a plaintiff’s (valid) trademark—the lynchpin of whether a defendant’s use of a mark infringes the plaintiff’s rights102 —turns on multiple factors, including the degree of secondary meaning of the mark.  Different circuits vary somewhat in assessing the likelihood of confusion, yet the Second Circuit’s test is characteristic.  Its multifactor test looks to

(1) the strength of the mark; (2) the degree of similarity between the two marks; (3) the proximity of the products; (4) the likelihood that the prior owner will “bridge the gap” . . . ; (5) actual confusion; (6) the defendant’s good faith in adopting its mark; (7) the quality of the defendant’s product; and (8) the sophistication of the buyers.103 

The first factor, the mark’s strength, is grounded in two different forms of strength: conceptual and commercial strength.104   Conceptual strength is based on where a valid mark falls on the Abercrombie spectrum: arbitrary and fanciful marks are the strongest, followed by suggestive marks, with descriptive marks as the weakest.105   Commercial strength is about how well consumers actually recognize the mark and is treated akin to an assessment of the degree of secondary meaning.106   The overall strength of the mark is determined based on the union of conceptual and commercial strength.107   The stronger a plaintiff’s mark is, the broader the scope of the plaintiff’s trademark rights, as it makes it easier to conclude that there is a likelihood of confusion between the plaintiff’s and defendant’s marks.108 

As Barton Beebe and Scott Hemphill explain, “[t]he most common rationale [for assessing strength in evaluating likelihood of confusion] asserts that the stronger a mark is, the more likely a consumer is to think of it when exposed to a similar or identical mark and thus the more likely it is that the consumer will conclude that the two marks originate in the same source.”109   Courts sometimes also explain that the stronger a mark is conceptually, the likelier it is that a consumer would assume that a business using a mark similar to a preexisting mark is the same source for both marks simply because the mark bears little conceptual relation to the goods or services with which they are associated.110   On the flip side, as the Supreme Court observes, “[w]hen a mark incorporates generic or highly descriptive components, consumers are less likely to think that other uses of the common element emanate from the mark’s owner,” meaning its use by another is less likely to confuse consumers.111   For these reasons, the law grants broader scope to stronger marks, while hemming in that scope for weaker marks and granting more leeway to competitors to use such terms without worry of confusing consumers.112 

Secondary meaning is materially relevant to assessing both conceptual and commercial strength.  In empirical work of trademark infringement cases, Barton Beebe shows that while a mark’s conceptual strength correlates with a finding of likelihood of confusion, its commercial strength trumps its conceptual strength when the two conflict, making commercial strength more meaningful.113   That is, secondary meaning is more pertinent to a mark’s strength than conceptual strength is.  Moreover, conceptual strength also is grounded in secondary meaning, as just shown with regard to all categories in the Abercrombie spectrum.114 

In all, secondary meaning is a central concept in trademark law, forming a crucial basis for a mark’s protectability—regardless of its categorization along the Abercrombie spectrum—and its scope.


Given how central secondary meaning is to trademark law, it is important to probe how and whether it serves trademark’s goals.  Secondary meaning is supposed to ensure that a trademark is distinctive or strong, suggesting that a business is using it in a way that protects consumers by lowering their search costs and in a way that promotes competition.115   However, as this Part shows, secondary meaning is easy to establish under current doctrine, which undermines trademark law’s effectuation of its goals.  Additionally, making secondary meaning harder to establish to right trademark’s balance would aggravate trademark law’s competitive equity, by creating barriers to entry and competition for less deep-pocketed businesses.  Moreover, trademark law wrongly presumes secondary meaning for many marks that are categorized as suggestive, arbitrary, or fanciful when those marks might not have much, if any, secondary meaning and therefore can be as competitively dangerous as descriptive marks without secondary meaning.  Trademark law’s focus on secondary meaning also obfuscates that it coexists with a mark’s primary meaning—which can remain front and center.  As such, protecting marks like descriptive ones upon a showing of secondary meaning does not obviate the concerns that exist for fair competition and consumer protection.  That all said, establishing secondary meaning—at least as understood at a conceptual level—can advance trademark’s purposes.  Yet most businesses generally have incentives external to the trademark system to establish secondary meaning anyway, so there is no need for trademark law to dwell on requiring it for such businesses.  If the law can screen out protection for marks that are undermining trademark law’s goals as discussed in this Part in other ways, that can obviate the need for trademark law to require secondary meaning.

A.   Ease of Establishing Secondary Meaning

As set out in the previous Part, secondary meaning is presumed for inherently distinctive marks, rendering them protectable immediately upon use, whereas businesses using descriptive marks must acquire secondary meaning after use before becoming protectable.116   The inquiries involved for marks with inherent distinctiveness as compared with acquired distinctiveness are thus different: the former focuses on the primary meaning of the mark in relation to the associated category of goods or services, whereas the latter focuses on whether consumers actually associate a mark with a single source of goods or services.117 

Because secondary meaning is conclusively presumed for inherently distinctive marks, it is thus trivially easy to prove once a mark is classified as either suggestive, arbitrary, or fanciful.  Yet under current doctrine, it tends to be not that much more complicated for marks that must acquire distinctiveness with secondary meaning, particularly for businesses with sufficient financial resources.

As discussed above, secondary meaning can be shown directly or circumstantially.118   Direct evidence—particularly in the form of consumer surveys—is expensive to generate, so many businesses show secondary meaning instead using circumstantial evidence: often the breadth, reach, and cost of advertising ventures.119   Recall that other circumstantial evidence includes length and manner of a mark’s use, volume of sales, nature of use in newspapers and magazines, and a defendant’s intent in copying the mark.120   The asserted relevance of most of this circumstantial evidence is that consumers’ likely wide exposure to the mark in these various ways leads to secondary meaning for the mark.121   Yet the logical inference from this evidence to secondary meaning can be weak.  For example, just because a business spends a lot of money on advertising does not mean it has succeeded in prompting consumers to associate the mark with a single source.122   Consider the case of Coca-Cola—a company with very deep pockets for advertising—which spent millions of dollars advertising the little-remembered Vault Soda, ultimately discontinued in 2011 after six years on the market.123   Similar critiques can be made about relying on length of a mark’s use or sales volume.124 

While some courts are careful to make a distinction between a business’s efforts and success in establishing secondary meaning,125  others are not as fastidious.  In particular, courts tend to rely heavily on mere advertising expenditures as a proxy for secondary meaning.  For example, the Third Circuit has stated that “[s]econdary meaning is generally established through extensive advertising which creates in the mind of consumers an association between different products bearing the same mark.”126   Other courts emphasize advertising expenditures plus sales as evidence that advertising has succeeded in creating secondary meaning.127   In one representative case, the Second Circuit found secondary meaning for T-50 used in connection with staple guns after the plaintiff submitted evidence of “substantial advertising expenditures, high sales volumes, use of the mark for more than 30 years, and recognition of the mark by competitors.”128   In another, the Fifth Circuit found BEEFEATER to have secondary meaning for use in connection with restaurants based on advertising and other promotion.129 

All in all, this framework and the resulting outcomes show that businesses that have the resources to spend enough on advertising can manufacture a finding of secondary meaning for their otherwise unprotectable marks.

There are other contexts in which courts easily conclude there is secondary meaning.  When a defendant intentionally copies a plaintiff’s mark, many courts presume the plaintiff’s mark has secondary meaning as the explanation for why the defendant copied.  For example, the Fourth Circuit concluded there was secondary meaning for LARSEN as used on decorative compact disc holders after finding that the defendant had intentionally copied that from the plaintiff.130   Yet as Tom McCarthy observes, the logic that there must be secondary meaning if a defendant copied the plaintiff’s mark is flawed, as there are many legitimate and procompetitive reasons for using the same term (including its descriptiveness).131 

Perhaps the easiest way of all to establish secondary meaning for a mark is through the mere passage of time, combined with knowledge to use trademark law to one’s advantage.  Marks that are registered in the PTO can become incontestable after five years if the registrant complies with certain formalities.132   Incontestable marks cannot be invalidated on many grounds that ordinarily disqualify trademark protection, including a lack of acquired or inherent distinctiveness.133   In that vein, the Supreme Court held that PARK ’N FLY for airport parking services cannot be challenged for being merely descriptive and lacking secondary meaning because the mark had become incontestable.134 

Even beyond the doctrine, it can be simple to manufacture secondary meaning in fact if a business uses the right cues to consumers.  In an empirical study, Tom Lee, Eric DeRosia, and Glenn Christensen show the ease of establishing source distinctiveness in the eyes of consumers, even with descriptive marks.135   They demonstrate that “when descriptive words are used in a typical trademark-use context, they are statistically just as likely to be perceived as source indicators as are their suggestive, arbitrary, and fanciful counterparts.”136   They show this across a range of fake products—packaged cookies, snack chips, vitamin food supplements, and laundry stain removers—with packaging design on which these words were placed with typical location and other trademark cues, where the words did not possess any particular secondary meaning ex ante to consumers.137   In effect, they find that “the non-lexical cues of typical trademark use (e.g., the location and size of the mark) . . . cause consumers to find descriptive marks to be source-indicating.”138 

In sum, secondary meaning is relatively easy to establish under current doctrine, particularly for deep-pocketed businesses.

B.   Perils of Raising the Bar for Secondary Meaning

If it is easy to establish secondary meaning under current doctrine—especially for businesses with deep pockets—and it is also simple to manufacture with contextual “tricks,” one might worry that marks that should not be readily protectable—particularly descriptive marks139 —can be protected by trademark law too easily.  Similarly, marks found to have secondary meaning would also have broader scope of protection than they might have were the showing harder to make.140   As discussed above, trademark strikes a balance to protect descriptive marks only upon a showing of secondary meaning as a way to calibrate concerns about fair and unfair competition.141   In essence, trademark law makes the determination that these marks can promote fair competition and help consumers only if such marks have truly developed secondary meaning.142   Yet allowing descriptive marks to be protected without sufficient secondary meaning and conferring on all marks broader scope without sufficient secondary meaning runs counter to trademark law’s goals.  How so?  In applying the current rules for establishing secondary meaning, trademark law blocks off descriptive marks—among those most necessary for fair competition—readily and allows businesses to use trademark law to stave off competitors with a finding of infringement too readily.  These outcomes undermine fair competition and hurt consumers.

In response, one might suggest raising the bar for finding secondary meaning as a way to restore trademark law’s promotion of fair competition and consumer protection.143   That is, if secondary meaning serves values trademark law wants to promote, perhaps the law ought to make it harder for businesses to qualify as having established secondary meaning.  Indeed, there are many ways the law could do that.  It could raise the threshold of the percentage of consumers that must associate a mark with the business at hand when relying on survey evidence.  When resting on circumstantial evidence, the law could require a tighter showing of success in establishing secondary meaning from the proffered evidence.  One way it could do so is by raising the evidentiary requirements for showing a link between circumstantial evidence—such as advertising—and success on that metric in establishing secondary meaning.  Another way it could do so is by raising the bar on what counts as qualifying circumstantial evidence, such as greater advertising expenditures than are currently sufficient.

Although these sorts of increased showings might do a better job of recognizing only true, extensive secondary meaning, they would nonetheless have significantly deleterious equitable and competitive consequences.  Raising the bar for establishing secondary meaning would almost always require greater expenditures by businesses to establish it.  To show that (more) consumers readily recognize a business’s mark, the business would likely have to accomplish more sales, spend more on advertising, and appear more prominently in media, not to mention spending significant sums on any surveys of secondary meaning.144   Businesses with deeper pockets could then more readily establish secondary meaning, thereby claiming descriptive marks and establishing broader scope for their marks.  By contrast, new entrants and other businesses with fewer resources could not as easily establish secondary meaning for their marks.145   To be sure, these distributive consequences146  are already inherent in the current laxer doctrines of secondary meaning, though less severely, but raising the bar would exacerbate these consequences.  In previous work, Barton Beebe and I establish that businesses with less choice among competitively effective marks have significant barriers to entry that persist in disadvantaging them over the course of their commercial life.147   That is, firms that were able to choose better marks have had persistently higher stock prices, greater access to funds, higher valuation ratios, and other benefits.148   The same can be true here: businesses with deeper pockets would likely have enduring advantages from having a wider range of marks from which to choose, including descriptive marks, and broader scope accorded to them, whereas lesser-resourced businesses would choose such marks at their own peril and be afforded lesser scope.  Because of these equitable and competitive consequences, raising the bar to establish secondary meaning is ill-advised.

C.   Improperly Presuming Secondary Meaning

As easy as it is to establish secondary meaning currently and as detrimental as it would be to raise the bar on secondary meaning, trademark law also improperly and irrefutably presumes there is secondary meaning in the case of some suggestive, arbitrary, and fanciful marks.  Recall that a mark categorized as suggestive, arbitrary, or fanciful is presumed to possess secondary meaning inherently because its primary meaning is distant enough from the goods or services with which it is associated that the law presumes it can serve no other purpose than be a trademark.149   For that reason, the law concludes that all consumers would understand it to be a trademark, hence the immediate and irrefutable secondary meaning for it.  Yet many of the marks placed into these categories by the doctrinal rules for sorting them in fact have close conceptual relations to their associated goods or services, rendering faulty the logic for conclusively presuming they have secondary meaning.

Consider first fanciful marks.  Recall that doctrinally, as per the leading trademark treatise, fanciful marks “consist of invented words that have been created for the sole purpose of serving as a trademark.”150   Accordingly, many courts have applied this rule to properly find that certain coined terms are fanciful: POTENZA and TURANZA for tires,151  M2 for software for the film and music
industries,152  PETERBILT for trucks,153  CARSONITE for highway markers,154  and so forth.  Yet courts’ conclusions that other marks are fanciful frequently miss the mark with formalistic rulings that marks are fanciful merely because they do not appear in the dictionary even though they convey meaning closely related to the goods or services at hand.  One court determined that SEGWAY for motorized personal mobility devices is fanciful because it was “coined ‘for the sole purpose of functioning as a trademark’”155  even though it is pronounced exactly like “segue,”156  an existing word which means “any smooth, uninterrupted transition from one thing to another.”157   Another found that PERCHVILLE for apparel associated with a town’s festival with fishing contest is fanciful because it was coined to be a trademark even though its two root words (“perch” and “-ville”) connote a town of fish.158   Yet another found MIRACLESUIT for slimming swimsuits to be fanciful.159   Moreover, the PTO’s Trademark Trial and Appeal Board found TELMEX for a Mexican telecommunications company to be fanciful160  even though it is composed of discernable parts that make it seem to be better categorized as generic or descriptive.  Furthermore, a court found that JAYCEES as a civic group for young men is fanciful even though its pronunciation is the same as J.C.s, which stands for Junior Chamber.161   Additionally, a court held that PSYCHO-CALISTHENICS for a series of physical exercises designed to achieve personal growth is fanciful.162   That said, more rarely, some courts—especially the Ninth Circuit—are careful to go beyond whether a term is coined in determining that it is fanciful, as with the Ninth Circuit’s conclusion that SURFVIVOR for Hawaiian-themed beach products is not fanciful but suggestive.163   As the Ninth Circuit elaborated in analyzing the categorization of EPIX for electronic pictures, “‘EPIX’ is not listed in the Oxford English Dictionary . . . .  However, . . . a coined term can still be
suggestive or descriptive.”164   The Ninth Circuit seems utterly correct: If a coined term describes the associated goods or services, why treat it as inherently distinctive rather than be as wary of protecting it as a run-of-the-mill descriptive mark?165 

Jake Linford, Barton Beebe, and I all relatedly show that many marks classified as fanciful in fact have associated built-in meaning in other ways.166   Sometimes, that is because of sound symbolism, in that particular sounds convey properties, like size, speed, and friendliness.167   For instance, Barton Beebe and I compare “the aggressiveness of VIAGRA with the calmer, more sensual CIALIS, both for drugs treating erectile dysfunction using very different marketing approaches coinciding with their respective names.”168   Jake Linford worries that “[f]or an empty vessel like a fanciful mark, sound symbolism may actually have an outsized effect because the fanciful mark has no lexical meaning when first coined.”169 

As with fanciful marks, courts sometimes conclude that marks are arbitrary even though they seem far from it, perhaps even descriptive.  The leading treatise explains arbitrary marks as “words in common linguistic use but which, when used with the goods or services in issue, neither suggest nor describe any ingredient, quality or characteristic of those goods or services.”170   Courts apply this rule seemingly correctly in, say, concluding that HARD CANDY for cosmetics is arbitrary.171   While they do not err in classification as frequently as with fanciful marks, they sometimes neglect that the mark in question is not quite arbitrary, as with RADIO SHACK for a store selling radio equipment,172  K2 for skis when the term is a reference to the second highest mountain in the world,173  OXYDOL for soap products including oxygen bleach,174  FOOTHRILLS for shoes,175  MATERNALLY YOURS for a maternity clothing shop,176  and SEVENTEEN for a magazine aimed at teenage girls, including seventeen-year-old ones.177 

Finally, consider the last category of inherently distinctive marks, those that are suggestive.  This category is the one where the line blurs most easily with descriptive marks, something explored in previous work.178   As discussed there,

[o]ne court determined that ‘L.A.’ as a mark for low alcohol beer was suggestive of it, while another held that the same mark was descriptive of the beer’s quality. And while a court held that ‘Family Market’ as a mark for a retail store was suggestive, other courts found ‘Shoppers Fair’ and ‘Food Fair’ to be descriptive marks for retail stores.179 

Consider similarly the following marks that courts or administrative agencies have found to be suggestive, and thus inherently distinctive and protectable as trademarks: CITIBANK for urban banks,180  DIAL-A-MATTRESS for telephonic mattress sales,181  WET/DRY BROOM for electric vacuum cleaners,182  SPRAY ‘N VAC for aerosol rug cleaners,183  and POM for pomegranate juice beverages.184   They all seem potentially mischaracterized, as they are arguably descriptive of the goods with which they are associated.

All of this is to suggest that the Abercrombie spectrum’s conclusive presumption of secondary meaning for suggestive, arbitrary, and fanciful marks might be a poor idea if the courts frequently enough categorize marks as inherently distinctive even though they are so closely related conceptually to the goods or services with which they are associated.  If such marks are more like descriptive marks—where there is strong worry that granting protection to them will undermine competition and hurt consumers—granting these marks automatic protection upon use is troublesome.185 

D.   Coexistence of Secondary Meaning and Primary Meaning

A final concern with enshrining secondary meaning in trademark law is that it obfuscates that a mark’s primary meaning is still extant even when there is secondary meaning.  Just as we might understand that a “bank” can be a place where money is deposited or the land along a river, among other things,186  we can understand AMERICAN AIRLINES to refer both to a particular airline company as well as an American airline more generally.187   More generally, even though acquisition of secondary meaning can activate new cognitive links between a mark and the source of goods or services using the mark,188  the mark still retains its cognitive link to its primary meaning(s).189   It is for this reason that some businesses like to choose as a mark a term with existing meaning, so that they “more readily impart a feeling of familiarity and authenticity, and have proven themselves to be relatively easy to pronounce, hear, read, and remember,” something I explore with Barton Beebe in previous work.190   Moreover, as we observe, it “generally requires less effort to instill common words with brand meaning, especially when the words’ meanings and connotations sync with the brand’s message.”191 

The presence of both primary and secondary meaning for marks poses concerns for fair competition and consumer protection in trademark law when the primary meaning of the term is closely related conceptually to the associated goods or services.192   As the marketing literature attests, a mark that describes or connotes important aspects of its associated good or service gives the business using it an advantage in convincing consumers of those aspects and thus their purchasing decisions.193   Marketing expert Kevin Keller observes that “[c]onsumers will find it easier to believe that a laundry detergent ‘adds fresh scent’ to clothes if it has a name like ‘Blossom’ than if it’s called something neutral like ‘Circle.’”194   It also follows from the previous Section’s discussion that this can be as true of marks categorized as suggestive, arbitrary, and fanciful as those classified as descriptive.  For example, CLOROX for bleach has been classified as fanciful, despite describing or suggesting the major ingredients of bleach (sodium hypochlorite, sodium chloride, sodium chlorate, and sodium hydroxide among other things), connoting both chloride and oxide.195 

A recent Second Circuit decision authored by Judge Leval lays bare this concern: that even a mark properly categorized as suggestive can have a primary meaning so closely linked to the associated goods or services that the mark is weak, presumably because it is unlikely to have strong—if any—secondary meaning.196   In that case, the court concluded that the mark RISE for nitro-brewed coffee is both suggestive and weak.197   It reasons that, “Rising is generally associated with the morning, a time when many crave a cup of coffee, relying on its caffeine to jumpstart their energy for the day.”198   The court elaborated that “[t]he word ‘Rise’ may also refer directly to energy itself; after consuming caffeine, one’s energy levels can be expected to ‘rise.’”199   The court then concluded that “[t]he close associations between the word ‘Rise’ and coffee constituted a weakness of the mark under the trademark law.”200   Judge Leval laid out more general reasoning undermining the conclusivity of the Abercrombie spectrum’s categorizations: “If the suggestion conveyed by a suggestive mark conjures up an essential or important aspect of the product, while the description conveyed by a descriptive mark refers to a relatively trivial or insignificant aspect of the product, the particular suggestive mark could be deemed weaker than the descriptive.”201 

The recognition that a mark’s primary meaning continues to coexist with secondary meaning underscores a serious problem with recent proposals to prioritize secondary meaning even further in trademark law as a basis to protect even generic marks or fanciful marks that are not truly fanciful.  In particular, Jake Linford advocates in favor of protecting generic marks when they have acquired secondary meaning such as with regard to HOG as a mark for Harley-Davidson’s motorcycles or HOTELS.COM as a mark for online hotel booking services.202   This is worrisome unless their generic meaning has been wholly lost to the world because of the disproportionate competitive advantage to the markholder and the disservice done to competitors and consumers sorting between the primary and secondary meaning.  Similarly, Linford suggests that marks currently categorized as fanciful but which have sound symbolism should be assessed for secondary meaning instead of being accorded inherent distinctiveness.203   Moreover, in light of the ease of miscategorizing suggestive and descriptive marks as one another given the fuzziness of the boundary between the two categories,204  Linford suggests that marks that might be suggestive should be protectable only upon evidence of secondary meaning.205   These proposals to elevate secondary meaning neglect that the primary meaning that remains extant too for a mark might swamp the benefits of trademark protection with unfair competitive advantages that disserve competitors and consumers and underscore the concerns of protecting descriptive terms too upon a showing of secondary meaning, including the competitive inequities inherent in the inquiry.206 

E.   Businesses’ Existing Incentives to Develop Secondary Meaning

The previous Sections expose multiple dangers with enshrining secondary meaning in trademark law: the ease of establishing it, the negative competitive consequences of raising the bar to establish secondary meaning, the wrongly presumed secondary meaning for many marks classified as inherently distinctive, and the obfuscation that primary meaning still coexists with marks that have secondary meaning.

Even so, it is important not to lose sight of the fact that establishing secondary meaning can serve important values for trademark law.  Trademark law is premised on encouraging a business to choose a mark that consumers will actually associate with that business’s goods or services, whether readily or with efforts put in by a business to foster secondary meaning.207   As discussed at the outset, it is only when consumers actually associate a mark with a business as a source of goods or services that a business can use this mark as a repository for goodwill for its goods or services as a way to compete in the marketplace and that consumers can use such marks to lower their search costs.208   That is precisely why trademark law requires (or presumes) secondary meaning for all marks.209 

What to make, then, of the quandary of secondary meaning the foregoing analysis has revealed?  Secondary meaning is valuable and reasonably something trademark law seeks to promote, yet trademark law currently makes it too easy to establish or presume secondary meaning when it might not exist, and yet making it harder to establish secondary meaning would have undesirable competitive consequences.  Moreover, trademark law tends to forget that secondary meaning coexists with primary meaning and that the negative consequences of protecting marks that are closely related conceptually to their associated goods or services therefore persist even when there is secondary meaning.  In light of the foregoing analysis, should the law maintain the requirement of secondary meaning?  Make it harder to satisfy?

In the next Part, I propose that this quandary—perhaps surprisingly—can be resolved by abolishing the requirement of secondary meaning in trademark law altogether (along with refashioning the inquiry for distinctiveness).  Yet before getting to that proposal, it is worth setting out that legitimate businesses generally want to develop secondary meaning for their marks, regardless of whether trademark law requires it.

Whether a business coins a mark, chooses one that would be considered arbitrary, or selects the most obviously descriptive mark, it generally wants consumers to associate that mark with itself as the source of the relevant goods or services.  The American Marketing Association explains that a brand is a “name, term, sign, symbol, or design, or a combination of them, intended to identify the goods and services of one seller or group of sellers and to differentiate them from those of competition.”210   As a leading marketing textbook explains, “many practicing managers refer to a brand as more than that—as something that has actually created a certain amount of awareness, reputation, prominence, and so on in the marketplace.”211   Businesses strive mightily to make consumers aware of their mark, develop positive associations with it, and differentiate it from competitors’ marks.212   They do so because it helps their business grow, not merely because trademark law sometimes requires it.  High degrees of brand awareness (which is brand recognition plus brand recall) translate to advantages a business has in terms of consumers learning about the business, as well as consumers considering and choosing goods or services from the business.213   Businesses know that to get to a point of helpful brand awareness, they typically need to advertise their products memorably, create strong brand imagery, and offer quality products for consumers to experience.214 

Strikingly, the extreme sums spent on advertising arbitrary and fanciful marks even though their secondary meaning is conclusively presumed by trademark law underscores how businesses very much labor to create actual secondary meaning for all marks regardless of trademark law.215   A look at the most expensive commercials aired illustrates as much, as many businesses with inherently distinctive marks and no need to establish secondary meaning are among the heavy spenders.  For example, the car company Kia spent $8 million on one advertisement, and the insurance company Aviva spent $13.4 million on an advertisement after it changed its name from Norwich Union.216 

To be sure, not all businesses want to develop secondary meaning for their marks, but those businesses that do not tend to be choosing marks and avoiding developing secondary meaning for illegitimate reasons.  One large class of this behavior involves businesses that choose a mark that resembles the mark of another business to piggyback on the preexisting mark’s goodwill and secondary meaning.  In fact, a defendant’s intentional copying of a plaintiff’s mark in a trademark infringement case weighs in favor of finding infringement.217   Courts weighing this factor in favor of a likelihood of confusion infer that a defendant that has copied a plaintiff’s mark is attempting to freeride on the trademark’s goodwill and confuse consumers that the plaintiff is the source of the defendant’s products rather than establish its own association with the chosen mark.218   Indeed, this evidence is considered so strong that Barton Beebe shows in an empirical study of courts’ applications of the multifactor infringement tests that “a finding of bad faith intent creates, if not in doctrine, then at least in practice, a nearly un-rebuttable presumption of a likelihood of confusion.”219 

The other large swath of this behavior involves those businesses that choose a descriptive (or other) term as a mark that they think consumers will rely on to seek out their products even though consumers do not associate that term with them as a source.220   The most poignant illustration of this in today’s day and age occurs when a business chooses to use as a mark a descriptive term that it thinks consumers will enter into Google or Amazon’s search engine to locate goods or services the business offers, prompting increased sales, regardless of the mark’s secondary meaning or lack thereof.221   Courts have recently pointed this out in the context of establishing that consumers might find certain marks easily at the top of search results because they are descriptive of what consumers have been searching rather than because they have secondary meaning.222   Examples include 24 HOUR FITNESS for a gym that is always open,223  1-800 CONTACTS for contact lenses,224  HOME-MARKET.COM for referral services for homeowners including to real estate agents and insurance companies,225  and this Article’s opening example of BOOKING.COM for travel booking services.226   In fact, with regard to the last example, the Supreme Court observed that “[a]ll descriptive marks are intuitively linked to the[ir associated] product or service and thus might be easy for consumers to find using a search engine or telephone directory.”227   Of course, some of these businesses might try to establish or even succeed in establishing secondary meaning for these terms.  But other businesses might not even be trying or accruing secondary meaning for their marks, all the while benefiting from sales that flow merely from consumers’ associations with and searches for those terms as descriptors.228 

In all, businesses using trademarks for the purposes set out by trademark law are generally trying to establish or succeeding in establishing secondary meaning independent of trademark law’s requirement of it.  As such, even if achieving secondary meaning is a worthy goal for trademarks, trademark law does not need to require it for businesses trying to develop marks and brands legitimately.  By contrast, others that are using trademarks for less legitimate reasons—such as to confuse consumers or to freeride on descriptive associations with a term—are not seeking to establish secondary meaning (and are often avoiding doing so).  Those trying to confuse consumers will tend to be liable for infringement if pursued as just discussed, thereby deterring this practice.  And those choosing descriptive terms to freeride on their meaning could be deterred from doing so in other ways, as discussed in the next Part.  With this view, requiring, presuming, and enshrining secondary meaning might not be a task with which trademark law needs to occupy itself.


The previous Part sets out the myriad problems with enshrining secondary meaning in trademark law, including the ease of establishing it, the perils of raising the bar on establishing it, the improper presumptions of it for marks categorized as inherently distinctive, and the obfuscation that primary meaning still sits aside any secondary meaning that is established.  In addition, the fact that most legitimate businesses are intrinsically motivated, well beyond any requirement in trademark law, to establish secondary meaning obviates much of the law’s need to step into the morass of secondary meaning.

To fix these problems with secondary meaning and ensure that trademark law promotes its goals of fair competition and consumer protection, I propose instead to use primary meaning as the gauge of protectability in trademark law.  And in the next Part, I propose how to think about a more careful role for secondary meaning in fashioning the scope of trademark protection.

The primary meaning of a term can be used to gauge protectability by assessing its conceptual relatedness to the goods or services for which it is used.  When that relatedness is too high, the term should not be protectable as a trademark.  Conceptual relatedness can be measured based on standard linguistic evidence—such as dictionary meaning and common word usage in media or corpora229 —or software tools like WordNet, which group words into cognitive synonyms and interlink them using conceptual-semantic and lexical relations, such as hyponyms (as “spoon” is a subtype of “cutlery”), hypernyms (as “color” is a hypernym of “red,” “blue,” and so forth), and meronyms (as “nose” is a part of a “face” or “engine” is a part of a “car”).230 

This analysis should be functional rather than formalistic.  In particular, coined terms should not be adjudged to lack conceptual relation to their associated goods or services merely because they are new words.  Rather, a term should be analyzed to see whether it communicates one or more existing words within it and conceptual distance should be measured in turn based on any such words.231 

Moreover, it should be emphasized that in assessing conceptual relatedness, primary meaning is not just about the most widely understood sense of a term.  The primary meaning of a term can include more than one sense, such as how “bank” means both the rising ground bordering a body of water and a place that transacts over and holds money, among other senses.232   “Primary meaning” is merely to be contrasted with “secondary meaning,” which is developed by associating the term with a source of goods or services.

Consider some illustrations of how a measure of conceptual relatedness would operate.  For example, COMPUTER for computers would be unprotectable because there is no conceptual distance between the term and the goods or services with which it is associated.  Relatedly, COOL for air conditioners, AMERICAN AIRLINES for an airline providing flights in the United States, and DRIVE for cars should also be similarly unprotectable because of the high conceptual relatedness between the term and the respective goods or services at issue.

Return now also to the examples at this Article’s outset: BOOKING.COM for online travel services, “FOR WALKING” for footwear, and VAGISIL for vaginal-health products.  They are each so close conceptually to their associated goods or services that they should not be protectable.  Say BOOKING.COM to anyone these past few decades, and regardless of whether they might also understand the term to refer to a particular business, they surely would think it means online booking services.  Rendering secondary meaning irrelevant and dwelling on the term’s primary meaning in relation to the associated services—the opposite of what the Supreme Court did for that mark—expose how harmful it is to competition and consumers to provide trademark protection for BOOKING.COM.  “FOR WALKING” is also closely conceptually related to footwear, despite the possibility that fashion aficionados might see the quotation marks surrounding the term and understand that they signal that the footwear comes from OFF-WHITE.  Consider now the third example of VAGISIL.  Although VAGISIL is a coined term, it readily conveys the concept of “vaginal”: its first four letters are the same as those in “vaginal” and there are almost no other English words that start with those four letters unrelated to the vagina.233   It is thus too closely related conceptually to the vaginal-health products with which it is associated.  This is true even if many consumers also associate VAGISIL with a single source in light of advertising expenditures exceeding $400 million since 1993, sales revenue greater than $1 billion since 1991, media coverage including on Saturday Night Live, and market dominance in the relevant market.234   Given that a word starting with VAGI- conveys “vaginal,” it is perhaps no surprise that at least sixty-six other vaginal-health products use the same first four letters to begin the mark for their products, as they know these first four letters will convey that they are vaginal-relevant products.235 

By contrast to these examples, STARBUCKS for coffee would be protectable because there is little conceptual relatedness between the term and the goods, even if the term might call to mind the Moby Dick character Starbuck or the words “star” and “bucks,” of which it is comprised, as none of these have a sufficient conceptual association with coffee.236   NIKE for sportswear would also be protectable because sportswear is not sufficiently conceptually related to the Greek goddess of victory with the same name,237  even if sportswear and victory are not entirely unconnected.  They are merely insufficiently directly related.  To take another example, KODAK, a coined term, conveyed no primary meaning in English when it was adopted for photography equipment.238   Founder George Eastman merely wanted a mark that started with a “K,” was short, easy to pronounce, and did not resemble other marks.239   It would thus be protectable as a trademark as it is sufficiently conceptually distant from photography.  Consider as a final illustration VIAGRA, another coined term, for an erectile-dysfunction drug.  As noted above, its sound connotes aggression240  and as Barton Beebe and I explain elsewhere, “VIAGRA calls to mind, all at once, ‘vigor,’ ‘vitality,’ ‘aggression,’ and ‘Niagara’ (suggesting both water and honeymoons).”241   Even with these associations, this coined term is sufficiently conceptually distant from erectile-dysfunction drugs that it would be protectable.

As these examples illustrate what should and should not be protectable under a test of primary meaning, trademark law would not protect marks that are conceptually necessary or important for competitors to use and that would grant a single competitor a persistent competitive advantage by allowing that business to have rights in that term for its sector.242   Consumers would be benefited by a more competitive, level playing field and not be drawn to businesses merely because of a mark that is conceptually close to the goods or services they are seeking.243   To be sure, consumers might benefit from locating a business providing a particular good or service easily, but consumers ultimately are worse off if they are led to a particular business merely because the mark conceptually matches the good or service.  In particular instances, consumers might be better off locating goods and services that they like, but categorically, they are likely to be hurt because conceptual matches between mark and the associated good or service bear no relation to the quality of the associated good or service.244 

In fact, the overarching goal of the Abercrombie spectrum in the first instance is to prevent rights from being granted—whether at all or too easily—in terms too conceptually related to their associated goods or services.245   Yet as discussed in the previous Part, the categories and the doctrinal buildup around them have undermined this goal, as seen from so many marks being improperly categorized.246   Moreover, the ease of finding secondary meaning, the distributive consequences of relying on secondary meaning for protection, and the neglect of the primary meaning still associated with marks that have secondary meaning247  counsel against protecting marks that might be classified as descriptive even upon a showing of secondary meaning.  Their current protection advantages the (often well-resourced) businesses that claimed these terms first for their goods or services in ways that can undermine trademark’s goals.

Scholars who defend the Abercrombie spectrum do so on the understanding that it is approximately helpful enough.  For example, Jake Linford generally supports the spectrum “[a]t a rough cut” for its “different treatment of fanciful, arbitrary, suggestive, and descriptive marks . . . in light of how listeners and readers process confusing ambiguity” about the signification of a mark in light of its various possible meanings.248   Some further argue that it would be costly to replace this spectrum—which conserves on detailed examination for suggestive, arbitrary, and fanciful marks—with a more in-depth inquiry about protectability.249   Yet these thinkers also presume that a more detailed investigation would be about providing evidence of secondary meaning,250  something that is problematic to competition and consumer protection for the reasons excavated in the previous Part.  Others take an opposite approach and would allow protection for all nongeneric marks upon use,251  a plan that is similarly detrimental to competition and consumer protection.252 

Restricting trademark protection to terms whose primary meaning is conceptually distant enough from the goods or services with which they are associated would instill in businesses the incentive to choose marks that do not undermine trademark’s goals, unlike under the current trademark regime.253   This would be a welcome refinement of the incentives created by the Abercrombie spectrum discussed above, to choose marks that are arbitrary or fanciful—so as to get immediate protection—over suggestive marks—which get immediate protection but might be (mis)categorized as descriptive—over descriptive marks—which require secondary meaning and might be (mis)categorized as generic.254   Moreover, it would also nudge trademark law in a somewhat more normative direction—by encouraging the adoption and use of marks that advance trademark law’s goals—instead of the more empirical grounding it currently has, of assessing trademark protection (and scope) in large part based on secondary meaning—on what consumers actually think—even when that undermines the law’s aims.255 

When a business chooses a mark whose primary meaning has sufficient conceptual distance from the goods or services with which it is associated, the mark will readily signal to consumers that it is a trademark—that it is signifying source in the way trademark law seeks to encourage.  As discussed above, marks can readily signify that they are source indicators through context and placement.256   Moreover, a business engaging in legitimate trademark use will go further and seek to establish broad consumer awareness of its mark and the mark’s source indication through advertising and the like.257   Therefore, even absent a screen for secondary meaning, a test of primary meaning to gauge protectability would allow protection for marks that will likely readily establish secondary meaning.  It would also exclude protection for the most worrisome categories of trademarks noted above: those trying to copy a competitor’s trademark and those trying to freeride on the primary meaning of a mark that is closely related to the associated goods or services.  For the former, if the competitor’s trademark is protectable in the first instance, there is a good chance the business would be deemed to be an infringer.258   And for the latter, the primary-meaning test would screen out protection for these marks in the first instance.

Moreover, making primary meaning the touchstone of protectability rather than secondary meaning would even out trademark law’s current distributive effects as to protectability.  A small business is on a relatively even playing field with a better-resourced one with regard to choosing a mark whose primary meaning is conceptually distant enough from the goods or services offered by the business.259   This result would leave businesses more or less at an equal starting point with regard to choosing a trademark, giving them all a fair shot independent of their marks.  Contrast that with the inequalities faced by a smaller business needing to establish secondary meaning to protect a mark compared with an established firm with significantly more money.260   In that context, the rich get richer with them able to secure protection for more descriptive and other trademarks that others might not easily protect.261   For this reason, it subverts competitive equity to raise the bar on secondary meaning, as Christopher Buccafusco, Jonathan Masur, and Mark McKenna propose be done to address many of the foregoing competitive concerns discussed herein.262   While raising this bar would indeed go beyond the current requirement—which to them “imposes only small marginal costs on putative mark holders,”263  as similarly discussed above264 —it would allow deep-pocketed businesses to choose these marks while preventing other businesses, including many new entrants, from doing so.265 

In response to this Article’s proposal to center protectability on primary meaning, one might suggest that it is difficult to excavate a term’s primary meaning and predictably assess its conceptual relatedness to the goods or services at hand.  Yet trademark law already concerns itself with assessing primary meaning in the context of evaluating the categorization of a mark along the Abercrombie spectrum.266   More broadly, interpretation and construction of terms is a central feature of legal decisionmaking.267   An analysis that the Abercrombie spectrum is unpredictable and unreliable in its classifications—including this Article’s—relates to the existence of numerous closely related buckets into which terms must be sorted.268   It is therefore hard to sort terms reliably (particularly, but not only, between the descriptive and suggestive categories).  By contrast, collapsing the sorting into a single yes-no inquiry into primary meaning would be more predictable.269   Of course, there would be some lack of clarity at the margins in hard cases, but that is to be expected of most inquires.  In addition, the current analysis is unpredictable in that courts are inconsistent in applying formalistic or functional approaches to categorize terms, especially with regard to fanciful marks.270   That source of unreliability would be eliminated by dictating that the primary-meaning inquiry be functional in nature.271 

To sharpen the clarity of the inquiry into primary meaning, it is important to differentiate between the conceptual relatedness of a term and its associated goods or services based on their respective primary meanings and the relatedness that might be drawn out by a marketing campaign for those goods or services using that term as a mark.  For example, consider APPLE for computers, frequently listed by courts and scholars as a prototypical arbitrary mark on the Abercrombie spectrum.272   APPLE has little conceptual relation with computers in terms of their respective primary meanings of a fruit and a technological device.  It would therefore be a protectable mark under a test of primary meaning.  Yet it surely was not chosen randomly by the business as a mark.  Perhaps the business thought it was a catchy term or perhaps the business’s founders had particular associations with apples, some of which have been told throughout the years, including that cofounder Steve Jobs liked to eat apples273  and that he had previously planted and harvested apples on an Oregon farm.274   Whatever the history, once the term APPLE is associated with computers, the business can take advantage of the conceptual associations with apples to help sell the company’s computers and develop a coherent and memorable brand.  The business can signify the biblical story in the Garden of Eden, which came to associate apples with knowledge, something that might help develop a positive brand association for its computers.275   Or it can convey the tale of Isaac Newton discovering the scientific principle of gravity when he witnessed an apple fall from a tree.276   (Indeed, Apple’s first logo featured a complex drawing of an apple hanging on a tree above Newton’s head.277 )  These and other possible brand associations that a business might develop are different than the conceptual relatedness of a mark and its associated goods and services.278   One way to think about this distinction is that between the mark and the brand.  As Deven Desai puts it, a trademark provides information about the source of goods or services, whereas a brand not only does that but also is related to “convey[ing] image components regarding power, value, and personality.”279 

As with any proposal to change which trademarks are protectable, there is the concern about how to address the many trademarks that are already protected under the current regime that would not pass muster under the new regime.  It is hard to remove protection en masse from so many businesses using these marks on an ongoing basis.  Yet it is also doubly unfair to businesses going forward to be disallowed from using marks that are protectable for those that got in under the wire.  One way to navigate this issue is with attention to narrow scope—possibly more so than under current doctrine280 —for marks whose protectability hinges on secondary meaning.  That would steer between the foregoing concerns to some extent.  Another possibility is to do a better job of clearing unused marks that are still nonetheless protected from the trademark register and to make it much more expensive to retain rights in marks that would no longer pass muster under the doctrine of primary meaning.281 

In sum, a doctrine of primary meaning to replace the doctrines requiring or presuming secondary meaning for trademarks would better promote fair competition and protect consumers while also promoting competitive equity among businesses, whether small or large.  It would also instill helpful incentives in businesses to choose marks that promote trademark’s goals and steer them away from those more anticompetitive marks that undermine them.


If secondary meaning is removed as a gauge of or presumption for protectability and replaced with a doctrine of primary meaning, many anticompetitive marks would no longer generally be protectable.282   Businesses would need to choose and use a mark from the set of terms conceptually unrelated enough to the associated goods and services and would then be able to stop third parties’ use of a mark likely to confuse consumers as infringing.283 

Recall that a mark’s conceptual strength and commercial strength (secondary meaning) factor into whether a third-party use of a mark is likely to confuse consumers.284   If a doctrine of primary meaning were to gauge protectability, a mark’s conceptual strength should no longer be measured by where it falls on the Abercrombie spectrum for the same reasons that spectrum should no longer sit at the center of protectability.  Rather, the doctrine of primary meaning—with its similar theoretical grounding as the Abercrombie spectrum285 —can be used to assess conceptual strength.  Conceptual strength would correlate with the conceptual distance between the primary meaning of the mark and the associated goods or services: the greater the distance, the greater the conceptual strength.

By contrast, commercial strength is more complicated to address if trademark law demotes secondary meaning.  Given that this Article critiques the role of secondary meaning in trademark law and argues it should be excised from protectability assessments, should it be eliminated as a relevant factor in assessing likelihood of confusion?  One might argue that it should for many of the same reasons discussed above: it is currently too easy to establish, especially for well-resourced businesses, yet raising the bar to establish it would create inequalities favoring these businesses yet further at the expense of those with fewer resources.286   However, the concern with eliminating a consideration of secondary meaning from how likely a defendant’s use of a mark is to confuse consumers is that the question logically turns on whether consumers are aware of the plaintiff’s mark in the first instance.  Consumers are more likely to be confused by the third-party use of a mark that is similar to the plaintiff’s when they actually are aware of the plaintiff’s mark.287   Thus, it makes sense for infringement and thus the scope of trademark rights to turn in part on secondary meaning.288   Yet the problematic ease of showing secondary meaning and the competitive inequalities in proving it persist alongside that logic.

To navigate between these considerations in assessing trademark scope, this Article proposes a sliding scale for secondary meaning for scope as a way to recognize the relevance of secondary meaning to scope, to raise the bar for secondary meaning in appropriate circumstances, and to mitigate secondary meaning’s competitive inequality.  That is, smaller and less well-resourced businesses—such as a local restaurant—should not need to establish secondary meaning to the same extent as a large established business—such as Coca-Cola or L’Oréal.  On this thinking, smaller businesses might need about the same or less evidence to show secondary meaning as currently, whereas larger businesses might require much more than currently.

Having a sliding scale in this context is sensible because it recognizes the relevance of secondary meaning to the inquiry into likelihood of consumer confusion while also remaining cognizant that smaller businesses have a smaller footprint, in that they tend to operate over a smaller geographic zone or with fewer potential consumers.  They therefore can establish consumer awareness of their mark with less advertising, less expenditure, fewer overall consumers being aware of them, awareness in a smaller geographic area, and so forth.289   At the other end of the spectrum, larger businesses have a correspondingly larger footprint and would need greater degrees of evidence of expenditures, sales, and awareness across a broader geographic area to establish secondary meaning.  As one court analyzing secondary meaning puts it, “the question is not the extent of [a business’s] promotional efforts, but [its] effectiveness in altering the meaning of [a term] to the consuming public.”290   This effectiveness varies based on a business’s footprint.  Not only would a sliding scale reflect the commercial realities of different businesses, but it would also ease the competitive inequities of a uniform secondary-meaning inquiry,291  regardless of where the bar to establish secondary meaning is set but especially if the bar is raised.292 

In fact, a sliding scale of secondary meaning tends to reflect how courts adjust their inquiries into secondary meaning based on the extent of a business’s footprint, even if they do not expressly say so.  As illustration, compare two cases.  In one, the Fifth Circuit found secondary meaning for Zatarain’s use of FISH-FRI for a coating to fry fish by looking only to the New Orleans area and by crediting the following evidence as establishing secondary meaning: $400,000 in advertising expenditures over five years, an increase in sales of just over 20,000 cases of the coating over a decade, total sales of over 900,000 cases over fifteen years, and evidence from a telephone survey of 100 women in the New Orleans area who fry fish three or more times per month.293   This corresponds to the business being a relatively local one based in the New Orleans area at the time.294   Contrast that with a court’s finding from about the same era that Purina—a pet food business with a bigger, national footprint—had not established secondary meaning for its TENDER VITTLES semimoist cat food despite having spent more than $4 million on advertising in the course of a year, having sales of approximately $6 million during that same time, and being the biggest seller in the category of semimoist cat food.295 

Some courts similarly acknowledge a lower burden for showing secondary meaning for smaller businesses in the context of a claim of reverse confusion.  Unlike the typical scenario of consumer confusion in trademark law, reverse confusion happens when—as one court explains it—“a larger, more powerful company uses the trademark of a smaller, less powerful senior owner and thereby causes likely confusion as to the source of the senior user’s goods or services.”296   In this scenario, because the junior user of the mark is so much larger, it will spend so much more on advertising and promotion that it drowns out the senior user’s use of the mark so that consumers associate the mark with the junior user and think the senior user’s goods are the junior user’s.297   Some courts, like the Third Circuit, hold that “in the context of a reverse confusion case, the evidentiary burden upon a smaller, senior user to establish the existence of secondary meaning is placed somewhat lower.”298   The Third Circuit explained this rule as ensuring that a larger business cannot “with impunity infringe the senior mark of a smaller one.”299   The court elaborated that “in a reverse confusion claim, a plaintiff with a commercially weak mark is more likely to prevail than a plaintiff with a stronger mark, and this is particularly true when the plaintiff’s weaker mark is pitted against a defendant with a far stronger mark.”300   Other courts have refused to lower this bar, explaining that the smaller business has to establish the same secondary meaning that a larger one would need to before making out a claim.301   Jeremy Sheff analyzes those court decisions that lower the threshold for secondary meaning in reverse confusion cases as showing concern for “a large disparity in the size and economic power of the parties.”302   As with the proposed sliding scale of secondary meaning, this regard for trademark law’s distributive consequences prompts some courts to lower the bar for establishing secondary meaning for smaller businesses in reverse confusion cases.

In essence, using a sliding scale for secondary meaning with regard to scope is a way to take account of the dangers of enshrining secondary meaning in trademark law—including the distributional effects of doing so—while at the same time appreciating the logical relevance of secondary meaning to assessing consumers’ likelihood of confusion and thus trademark scope.  Trademark law already adjusts its inquiry into secondary meaning in ways proportional to the size of a business’s footprint, so introducing a sliding scale would be consistent with that reality.


While secondary meaning has long been considered to be at the heart of trademark law, an analysis of it from multiple angles reveals that it actually undermines what trademark law is trying to promote: fair competition and consumer protection.  The hazards with enshrining secondary meaning include the problematic doctrine that has built up to assess it or presume it, the competitive inequalities secondary meaning creates which hurt smaller and newer businesses, and the neglect of the competitive harms caused by protecting marks whose primary meaning is too close conceptually to the associated goods or services.  Alongside these dangers, businesses intrinsically have the incentive to establish secondary meaning regardless of whether trademark law requires it, making its enshrinement less necessary than the law and scholars suppose.

This Article exposes that trademark law has it backward in protecting marks like BOOKING.COM, “FOR WALKING,” and VAGISIL as long as they have (and are capable of) secondary meaning.  Rather, trademark law should investigate whether it promotes or undermines trademark law’s goals to protect such marks.  If counterproductive to the law’s aims—and this Article’s investigation reveals that such protection is—trademark law should not protect such marks regardless of whether there is secondary meaning.  Instead, trademark law should encourage businesses to choose and use trademarks in the first instance that promote the law’s goals.  Secondary meaning should serve no role in the protectability of trademarks.  Rather, primary meaning should be used to gauge protectability.  And secondary meaning should serve a more carefully tailored role in assessing likelihood of consumer confusion and thus scope, with a sliding scale of secondary meaning.  These adjustments would help restore trademark law’s promotion of fair competition and consumer protection.

© 2022 Jeanne C. Fromer.  Individuals and nonprofit institutions may reproduce and distribute copies of this Article in any format at or below cost, for educational purposes, so long as each copy identifies the author, provides a citation to the Notre Dame Law Review, and includes this provision in the copyright notice.

* Walter J. Derenberg Professor of Intellectual Property Law, New York University School of Law; Faculty Co-Director, Engelberg Center on Innovation Law & Policy.  Thanks to Amy Adler, Arnaud Ajdler, Stefan Bechtold, Barton Beebe, Ryan Bubb, Mala Chatterjee, Graeme Dinwoodie, Rochelle Dreyfuss, Clay Gillette, Leah Chan Grinvald, Daniel Hemel, Scott Hemphill, Cynthia Ho, Annette Kur, Edward Lee, Matthias Leistner, Jake Linford, Glynn Lunney, Gerard Magliocca, Jonathan Masur, Bill McGeveran, Mark McKenna, Michael Meurer, Joseph Scott Miller, Ansgar Ohly, Alexandra Roberts, Rachel Sachs, Jason Schultz, Jessica Silbey, Christopher Sprigman, Jennifer Sturiale, Haochen Sun, Rebecca Tushnet, and Jeremy Waldron, and participants in colloquia at the Frauenchiemsee LMU Munich Faculty of Law-New York University School of Law IP and Fairness conference, Indiana University Robert H. McKinney School of Law, New York University School of Law, University of Hong Kong Faculty of Law, Chicago Intellectual Property Colloquium, Trademark Scholars Roundtable, and Works in Progress in Intellectual Property Colloquium for their first-rate comments.  Thanks to Amanda Adian, Mala Chatterjee, and Tim Keegan for extraordinary research assistance.  I gratefully acknowledge support from the Filomen D’Agostino and Max E. Greenberg Research Fund.

Infra Part I., | Idris Elba Says Things | 2022 Big Game Ad, YOUTUBE (Feb. 8, 2022),

Infra Part I.

U.S. Pat. & Trademark Off. v. B.V., 140 S. Ct. 2298, 2301 (2020).

See id.

Final Office Action, U.S. Trademark Application Serial No. 90/277,796 (filed Oct. 26, 2020) (Aug. 2, 2021).

See 15 U.S.C. § 1052(f) (2018); U.S. PAT. & TRADEMARK OFF., TRADEMARK MANUAL OF EXAMINING PROCEDURE § 1212 (July 2022).  Indeed, Off-White recently filed a response in the PTO that it has established the requisite secondary meaning and the mark is thus protectable.  See Request for Reconsideration After Final Action, U.S. Trademark Application Serial No. 90/277,796 (filed Oct. 26, 2020) (Jan. 31, 2022); Response to Office Action, U.S. Trademark Application Serial No. 90/277,796 (filed Oct. 26, 2020) (Aug. 5, 2022).  Nonetheless, the PTO refused the application on the ground that the applied-for mark fails to function as a trademark in the sense that it “would be perceived by consumers as merely conveying information about applicant’s or similar goods and/or services.”  Nonfinal Office Action, U.S. Trademark Application Serial No. 90/277,796 (filed Oct. 26, 2020) (Sept. 21, 2022).  This refusal therein purports to refute the possibility of secondary meaning.  See generally Alexandra J. Roberts, Trademark Failure to Function, 104 IOWA L. REV. 1977 (2019) (unpacking this trademark doctrine).

Combe Inc. v. Dr. August Wolff GMBH & Co. KG Arzneimittel, 851 F. App’x 357, 360–63 (4th Cir. 2021).

See infra Section I.A.

10 See infra Section I.B.

11 See infra Section II.A.

12 See infra Section II.C.

13 See infra Section II.B.

14 See infra Part I.

15 See infra Part II.

16 See Timothy Denny Greene & Jeff Wilkerson, Understanding Trademark Strength, 16 STAN. TECH. L. REV. 535, 582 (2013); Jake Linford, A Linguistic Justification for Protecting “Generic” Trademarks, 17 YALE J.L. & TECH. 110 (2015) [hereinafter Linford, Generic Marks]; Jake Linford, Are Trademarks Ever Fanciful?, 105 GEO. L.J. 731, 738 (2017) [hereinafter Linford, Fanciful Marks].

17 See infra Section I.A.

18 See infra Section II.B.

19 See infra Section II.C.

20 See infra Section II.D.

21 See infra Section II.E.

22 15 U.S.C. § 1127 (2018) (defining trademarks to include certain “word[s], name[s], symbol[s], or device[s], or any combination thereof”); Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205, 209–16 (2000) (holding that product design or packaging might constitute a protectable trademark).  Also potentially protectable are sounds, scents, and colors.  Qualitex Co. v. Jacobson Prods. Co., 514 U.S. 159, 162, 174 (1995).  Although much of what I discuss in this Article might also apply to images and trade dress—a product’s design or packaging—I focus explicitly on words.

23 15 U.S.C. § 1127 (2018).  Likewise, they are protectable if a person has a “bona fide intention to use [them] in commerce and applies to register [them] on the principal register established by [federal law].”  Id.

24 Id.

25 See id. §§ 1052, 1114, 1127.

26 See id. §§ 1125(a)(1), 1127.

27 Id. §§ 1114(1)(a), 1127.

28 Id. §§ 1125(a)(1), 1127.  Even though protection against infringement is provided whether or not a mark is registered, there are significant advantages to registration.  If an entity succeeds in demonstrating to the PTO that a mark it uses is entitled to trademark registration (including that an identical or sufficiently similar mark is not already in use in a way that would confuse consumers), id. §§ 1052(d), 1127, that entity benefits in multiple ways.  Among the advantages conferred are nationwide constructive use conferring priority over most users of the same or similar marks, see id. §§ 1057(c), 1072, 1127, enhanced remedies in federal court, id. §§ 1117(b), 1127, and the possibility of the mark becoming incontestable in many ways after five years, id. §§ 1065, 1127.  Conversely, there are disadvantages for the owner of an unregistered mark.  For that owner, an infringement action might lie only if the unregistered user has priority, typically due to an allegedly infringing use in a geographic area where the marked products or services were already being sold or advertised or where the mark’s reputation has been established.  See RESTATEMENT (THIRD) OF UNFAIR COMPETITION § 30 cmt. a (AM. L. INST. 1995).

29 See Sidney A. Diamond, The Historical Development of Trademarks, 65 TRADEMARK REP. 265, 270 (1975); Mark P. McKenna, The Normative Foundations of Trademark Law, 82 NOTRE DAME L. REV. 1839, 1841 (2007); Beverly W. Pattishall, Two Hundred Years of American Trademark Law, 68 TRADEMARK REP. 121, 128 (1978); Frank I. Schechter, The Rational Basis of Trademark Protection, 40 HARV. L. REV. 813, 814–19 (1927).

30 See Park ’N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 198 (1985); McKenna, supra note 29, at 1843; Schechter, supra note 29, at 814–19.

31 See Diamond, supra note 29, at 280–81.

32 See Schechter, supra note 29, at 814–15.

33 See id. at 817.

34 See Barton Beebe, The Semiotic Analysis of Trademark Law, 51 UCLA L. REV. 621, 677–82 (2004).

35 Schechter, supra note 29, at 818.

36 See, e.g., William M. Landes & Richard A. Posner, Trademark Law: An Economic Perspective, 30 J.L. & ECON. 265, 269–70 (1987).

37 William M. Landes & Richard A. Posner, The Economics of Trademark Law, 78 TRADEMARK REP. 267, 271 (1988).  There has been a good deal of debate among trademark scholars about whether the law ought to view the consumer as a free individual making a choice to pay more for intangible values conveyed through advertising or as a person to be protected from the irrational encouragement that advertising and trademarks provide, beyond the underlying goods’ quality.  See generally Ralph S. Brown, Jr., Advertising and the Public Interest: Legal Protection of Trade Symbols, 57 YALE L.J. 1165 (1948) (describing this debate).

38 See Robert G. Bone, Hunting Goodwill: A History of the Concept of Goodwill in Trademark Law, 86 B.U. L. REV. 547, 549 (2006).

39 Beebe, supra note 34, at 623.

40 Graeme B. Dinwoodie, The Death of Ontology: A Teleological Approach to Trademark Law, 84 IOWA L. REV. 611, 614 (1999).  Rochelle Dreyfuss observes that, in recent years, trademarks have begun to serve a communicative purpose, of “becom[ing] products in their own right, valued as indicators of the status, preferences, and aspirations of those who use them.”  Rochelle Cooper Dreyfuss, Expressive Genericity: Trademarks as Language in the Pepsi Generation, 65 NOTRE DAME L. REV. 397, 397 (1990); accord Beebe, supra note 34, at 624, 656.  To take one litigated instance, a band sings a song entitled Barbie Girl to offer commentary on the values a Barbie doll represents: “I’m a blond bimbo girl, in a fantasy world/Dress me up, make it tight, I’m your dolly.”  Mattel, Inc. v. MCA Recs., Inc., 296 F.3d 894, 901 (9th Cir. 2002) (quoting AQUA, BARBIE GIRL (MCA Records 1997)).  As the ensuing Ninth Circuit decision puts it:

Some trademarks enter our public discourse and become an integral part of our vocabulary.  How else do you say that something’s ‘the Rolls Royce of its class’?  What else is a quick fix, but a Band-Aid?  Does the average consumer know to ask for aspirin as ‘acetyl salicylic acid’? . . . Trademarks often fill in gaps in our vocabulary and add a contemporary flavor to our expressions.  Once imbued with such expressive value, the trademark becomes a word in our language and assumes a role outside the bounds of trademark law.

Id. at 900 (citations omitted).  Because these communicative uses enrich society’s conversations, scholars recommend that trademark protection not encompass such uses.  See, e.g., Dreyfuss, supra, at 418.

41 15 U.S.C. § 1127 (2018).

42 Barton Beebe, Search and Persuasion in Trademark Law, 103 MICH. L. REV. 2020, 2028, 2028–29 (2005); accord Mark P. McKenna, Teaching Trademark Theory Through the Lens of Distinctiveness, 52 ST. LOUIS U. L.J. 843, 847 (2008) (“Distinctiveness . . . refers to the extent to which a claimed designation conveys to consumers information about the source of products or services as opposed to merely conveying product-related information.”).  There are yet more refined understandings of distinctiveness than trademark law incorporates.  See Beebe, supra, at 2028 (distinguishing “source distinctiveness, which a trademark must possess to fall within the subject matter of trademark protection,” and “differential distinctiveness, the extent of which prescribes the scope of trademark protection when protection is given”).

43 See, e.g., RJR Foods, Inc. v. White Rock Corp., 603 F.2d 1058, 1059 (2d Cir. 1979) (“To be entitled to the benefit of the secondary meaning doctrine, a plaintiff must establish that the purchasing public has come to associate certain words, symbols, collocations of colors and designs, or other advertising materials or techniques, with goods from a single source.”).  As Christopher Buccafusco, Jonathan Masur, and Mark McKenna emphasize, the reference to this meaning being “secondary” is “not in the sense of being of secondary importance, but in the sense of being second in time to the primary, descriptive meaning.”  Christopher Buccafusco, Jonathan S. Masur & Mark P. McKenna, Screening Meaning, 15 (Univ. of Chi. L. Sch., Pub. L. & Legal Theory Working Paper No. 784, 2022), [].

44 Graeme B. Dinwoodie, Reconceptualizing the Inherent Distinctiveness of Product Design Trade Dress, 75 N.C. L. REV. 471, 483 (1997).

45 McKenna, supra note 42, at 850.

46 Id. at 851.

47 Landes & Posner, supra note 36, at 288.

48 537 F.2d 4, 7, 9–11 (2d Cir. 1976).

49 Wal-Mart Stores, Inc. v. Samara Bros., 529 U.S. 205, 210–12 (2000) (citing Abercrombie, 537 F.2d at 10–11); Qualitex Co. v. Jacobson Prods. Co., 514 U.S. 159, 162–63 (1995) (citing Abercrombie, 537 F.2d at 9–10).

50 Abercrombie, 537 F.2d at 9.  These categories are often referred to as the Abercrombie spectrum for reflecting a continuum of protectability ranging from generic terms that get no protection to arbitrary and fanciful marks which get protection most easily, see, e.g., Lisa Larrimore Ouellette, The Google Shortcut to Trademark Law, 102 CALIF. L. REV. 351, 359 (2014),even though the term “spectrum” does not appear in the court’s opinion.  The first published federal court opinion that appears to refer to these categories as a “spectrum” in the context of Abercrombie is Miller Brewing Co. v. G. Heileman Brewing Co., 561 F.2d 75, 79 (7th Cir. 1977).  That said, there is sometimes reference to a spectrum with regard to these categories that predates Abercrombie.  See Redken Lab’ys, Inc. v. Clairol Inc., 350 F. Supp. 1301, 1306 n.12 (C.D. Cal. 1972).  Characterization of a term in one of these categories is a factual issue to be reviewed for clear error.  2 J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 11:3 (5th ed. 2017).

51 See Abercrombie, 537 F.2d at 9.

52 Id.

53 See Sofía G. de la Rocha, Note, Tusk Tusk: A Comparative Analysis into the Effects of Ivory Trade Regulation and the International Art Market, 49 CAL. W. INT’L L.J. 425, 432–40 (2019) (surveying the state of U.S. law on ivory trade).

54 See Judy Kirpich, Ivory Soap Dope: A Brand Evolution, GRAFIK, []; see Jane L. Levere, Ivory Soap Refreshes Its Ads and Its Look, but Is Resolutely Simple, N.Y. TIMES (Nov. 7, 2011), [].

55 Abercrombie, 537 F.2d at 9 n.6.

56 Urgent Care Inc. v. S. Miss. Urgent Care Inc., 289 F. App’x 741, 744 (5th Cir. 2008).

57 Small Bus. Assistance Corp. v. Clear Channel Broad., Inc., 210 F.3d 278, 279–80 (5th Cir. 2000).

58 Abercrombie, 537 F.2d at 9.

59 See id.

60 TCPIP Holding Co. v. Haar Commc’ns Inc., 244 F.3d 88, 93 (2d Cir. 2001); accord 2 MCCARTHY, supra note 50, § 11:16 (discussing what makes a mark descriptive, including to whom it must be descriptive and as to which aspects of a product, including ingredients, size, function, and provider).

61 Custom Vehicles, Inc. v. Forest River, Inc., 476 F.3d 481, 483 (7th Cir. 2007).

62 Id.

63 Id.

64 TCPIP Holding Co., 244 F.3d at 93.

65 15 U.S.C. §§ 1052(e)(1), (f) (2018); see also Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4, 10 (2d Cir. 1976); Custom Vehicles, 476 F.3d at 483; TCPIP Holding Co., 244 F.3d at 94.  However, trademark law recognizes that competitors might need to use such terms not as marks but to describe their goods and sometimes allows that as descriptive fair use.  The descriptive fair use defense permits a business to use a competitor’s protected descriptive mark so long as this use is “in good faith only to describe the [business’s] goods or services,” rather than as a mark.  15 U.S.C. § 1115(b)(4) (2018); see also KP Permanent Make-Up, Inc. v. Lasting Impression I, Inc., 543 U.S. 111, 124 (2004) (holding that a defendant raising such a defense need not demonstrate that consumers will not be confused by the use).  For example, if a mark is used to describe another’s goods or services rather than indicate source, such as if Delta Airlines described itself patriotically as “an American airline,” that might be permissible as a descriptive fair use of the AMERICAN AIRLINES mark.  Although defendants can prevail in an infringement claim with this defense, see Deborah R. Gerhardt, A Masterclass in Trademark’s Descriptive Fair Use Defense, 52 AKRON L. REV. 739 (2018), the defense is narrow because it does not allow the defendant to use the term as a mark—including a domain name or slogan—even if it describes the defendant’s goods or services, Lisa P. Ramsey, Descriptive Trademarks and the First Amendment, 70 TENN. L. REV. 1095, 1167–68 (2003) (citing cases); see also Stephen L. Carter, The Trouble with Trademark, 99 YALE L.J. 759, 771 (1990); Rebecca Tushnet, Fixing Incontestability: The Next Frontier?, 23 B.U. J. SCI. & TECH. L. 434, 436 (2017).  Lisa Ramsey documents that the defense is murky, and “[r]elevant factors for determining whether a use is a trademark or descriptive use include the size, style, location, and prominence of the descriptive term in comparison to the defendant’s use of its own trademark or other descriptive matter in advertising or product packaging.”  Ramsey, supra, at 1168.  She concludes that “[w]ith such a limited and uncertain fair use defense, competitors will likely self-censor their commercial expression rather than risk the cost and inconvenience of having to defend a trademark infringement action.”  Id. at 1169.  Rebecca Tushnet additionally worries that the defense “can be burdensome to prove and often requires at least extensive discovery.”  Tushnet, supra, at 436.

66 Beebe, supra note 42, at 2029; Dinwoodie, supra note 44, at 487–88.

67 Abercrombie, 537 F.2d at 10.

68 Levi Strauss & Co. v. Blue Bell, Inc., 778 F.2d 1352, 1354 (9th Cir. 1985) (quoting 1 J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 15:2, 15:11 (2d ed. 1984)).  A single source of consumer association suffices even if buyers do not know the source’s corporate identity.  Tone Bros. v. Sysco Corp., 28 F.3d 1192, 1203–04 (Fed. Cir. 1994); see supra text accompanying notes 32–33 (describing this evolution in trademark history).

69 Test Masters Educ. Servs., Inc. v. Robin Singh Educ. Servs., Inc., 799 F.3d 437, 445 (5th Cir. 2015) (alteration in original) (quoting Bd. of Supervisors for La. State Univ. Agric. & Mech. Coll. v. Smack Apparel Co., 550 F.3d 465, 476 (5th Cir. 2008)).  For more on what survey evidence of consumer association of a term with a single source of goods or services might entail, see Dominic A. Azzopardi, Note, Disarray Among the Circuits: When Are Consumer Surveys Persuasive?, 104 IOWA L. REV. 829 (2019).  Courts are not uniform on the threshold percentage of consumers that make this association to suffice to establish secondary meaning, but it is generally more than 30% and 50% is clearly sufficient.  Sarah Butler & Healey Whitsett, Commentary, A Tale of Two Cups: Acquired Distinctiveness and Survey Evidence Before the TTAB, 110 TRADEMARK REP. 698, 699 (2020).  As Alexandra Roberts has recently proposed, trademark law might also rely on the way consumers refer to marks online, including on social media, to assess whether marks have acquired secondary meaning.  Alexandra J. Roberts, Mark Talk, 39 CARDOZO ARTS & ENT. L.J. 1001 (2021).

70 See Vincent N. Palladino, Surveying Secondary Meaning, 84 TRADEMARK REP. 155, 165 (1994).

71 Id.

72 Vincent N. Palladino, Assessing Trademark Significance: Genericness, Secondary Meaning and Surveys, 92 TRADEMARK REP. 857, 876–89 (2002).

73 Vincent N. Palladino, Secondary Meaning Surveys in Light of Lund, 91 TRADEMARK REP. 573, 618–21 (2001).

74 U.S. Pat. & Trademark Off. v. B.V., 140 S. Ct. 2298, 2305 (2020); id. at 2313–14 (Breyer, J., dissenting).  This form of survey is known as a Teflon survey after a 1973 survey used as evidence in court to determine that TEFLON is a protectable mark for being assessed to be a brand name rather than generic.  2 MCCARTHY, supra note 50, § 12:16.  After explaining the distinction between generic terms and brand names to respondents, such a survey asks them whether a term is a brand name or a generic name.  Id.  There are potential flaws in these surveys, including that a term can concurrently be understood as a generic name and a brand name., 140 S. Ct. at 2307 n.6.

75, 140 S. Ct. at 2307.

76 Cf. Laura A. Heymann, Trademarks in Conversation: Assessing Genericism After, 39 CARDOZO ARTS & ENT. L.J. 955, 957 (2021) (positing that “[t]he Court was correct [in] when it reaffirmed that a trademark’s status as [generic or not] depends on consumer understanding”).

77 Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4, 11 (2d Cir. 1976).

78 Zatarains, Inc. v. Oak Grove Smokehouse, Inc., 698 F.2d 786, 791 (5th Cir. 1983).  Suggestive marks were created as a protectable category in response to a bar in the common law on protecting descriptive marks.  2 MCCARTHY, supra note 50, § 11:63.  This new category was made to grant protection to marks that “were neither exactly descriptive on the one hand nor truly fanciful on the other.”  Abercrombie, 537 F.2d at 10.

79 Swatch AG v. Beehive Wholesale, LLC, 739 F.3d 150, 157–58 (4th Cir. 2014).

80 Innovation Ventures, LLC v. N.V.E., Inc., 694 F.3d 723, 730 (6th Cir. 2012).

81 Synergistic Int’l, LLC v. Korman, 470 F.3d 162, 172 (4th Cir. 2006).

82 See Abercrombie, 537 F.2d at 11.

83 Starbucks Corp. v. Wolfe’s Borough Coffee, Inc., 736 F.3d 198, 212 (2d Cir. 2013).

84 Palm Bay Imps., Inc. v. Veuve Clicquot Ponsardin Maison Fondee en 1772, 396 F.3d 1369, 1372 (Fed. Cir. 2005).

85 Scott Fetzer Co. v. House of Vacuums Inc., 381 F.3d 477, 485 (5th Cir. 2004).

86 Abercrombie, 537 F.2d at 11 n.12.

87 2 MCCARTHY, supra note 50, § 11:4.  A danger for a product associated with a fanciful mark for which there is no familiar term is that consumers might start using the mark as a term for the product, at which point the mark would become generic, as has happened with once-fanciful marks ASPIRIN and CELLOPHANE.  Id. § 11:9.

88 Carson Mfg. Co. v. Carsonite Int’l Corp., 686 F.2d 665, 670 (9th Cir. 1981).

89 Luzier Inc. v. Marlyn Chem. Co., 442 F.2d 973, 974 (C.C.P.A. 1971).  Abercrombie characterized suggestive marks as less distinctive than arbitrary or fanciful marks, which it placed on roughly equal footing.  See Abercrombie, 537 F.2d at 9.  By contrast, some other courts—including the Supreme Court—have characterized fanciful marks as more distinctive than arbitrary ones.  Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 768 (1992).

90 Landscape Forms, Inc. v. Columbia Cascade Co., 113 F.3d 373, 380 (2d Cir. 1997).

91 2 MCCARTHY, supra note 50, § 11:2; Dinwoodie, supra note 44, at 486.

92 Beebe, supra note 42, at 2029; Dinwoodie, supra note 44, at 486; Linford, Fanciful Marks, supra note 16, at 743–45.  Graeme Dinwoodie argues that notions of inherent distinctiveness principally developed in the context of word marks and are less well-suited to trade dress and product design.  Dinwoodie, supra note 44, at 475.  Rebecca Tushnet relatedly argues that the Abercrombie spectrum does not suit images either.  Rebecca Tushnet, Looking at the Lanham Act: Images in Trademark and Advertising Law, 48 HOUS. L. REV. 861, 866–67 (2011).  Dustin Marlan provides a framework for analyzing distinctiveness of images that is an extension of the Abercrombie spectrum.  Dustin Marlan, Visual Metaphor and Trademark Distinctiveness, 93 WASH. L. REV. 767 (2018).  While this Article’s analysis might have broader applicability beyond words to images and trade dress, the treatment of these topics is beyond this Article’s scope.

93 Dinwoodie, supra note 44, at 487–88.  Bob Bone objects to this explanation on the ground that “[s]ource identification is not something intrinsic or natural to a symbol; it is an empirical property that results from the way a symbol actually operates in the marketplace.”  Robert G. Bone, Enforcement Costs and Trademark Puzzles, 90 VA. L. REV. 2099, 2130 (2004).  Rather, he thinks that the doctrine of inherent distinctiveness for these categories is premised on “minimiz[ing] the sum of error and administrative costs” related to proving secondary meaning for these categories of terms.  Id.  More specifically, he reasons that “[b]ecause of the high probability that these marks already have or will soon acquire secondary meaning, the conclusive presumption is very likely to be correct most of the time.  Moreover, the conclusive presumption saves the litigation costs of actually proving secondary meaning in individual cases.”  Id. at 2131–32 (footnote omitted).

94 Landscape Forms, Inc., 113 F.3d at 380; accord Thomas R. Lee, Eric D. DeRosia & Glenn L. Christensen, An Empirical and Consumer Psychology Analysis of Trademark Distinctiveness, 41 ARIZ. ST. L.J. 1033, 1041 (2009).

95 Supra text accompanying note 59.

96 2 MCCARTHY, supra note 50, § 11:18.

97 Linford, Fanciful Marks, supra note 16, at 746.

98 Jeanne C. Fromer, The Role of Creativity in Trademark Law, 86 NOTRE DAME L. REV. 1885, 1910–12 (2011); cf. Greene & Wilkerson, supra note 16, at 565 (theorizing that the Abercrombie spectrum “properly creates incentives for trademark adoption and brand-building”).

99 Infra SectionII.C.

100 Fromer, supra note 98, at 1911–13.

101 Id. at 1913–14.  Incentives for businesses to adopt marks in one category or the other can look different from the perspective of marketing research.  This body of research suggests that businesses ought to pick an existing word as a mark (be it descriptive, suggestive, or arbitrary) because it will be easier for consumers to remember and for businesses to draw on the existing associations with that word.  See KEVIN LANE KELLER, STRATEGIC BRAND MANAGEMENT: BUILDING, MEASURING, AND MANAGING BRAND EQUITY 149–50 (3d ed. 2008).  Descriptive and suggestive marks can help consumers yet further and help in brand recall.  Id. at 150; see also Barton Beebe & Jeanne C. Fromer, Are We Running out of Trademarks? An Empirical Study of Trademark Depletion and Congestion, 131 HARV. L. REV. 945, 964–70 (2018); cf. Justin Hughes, Cognitive and Aesthetic Functionality in Trademark Law, 36 CARDOZO L. REV. 1227, 1230 (2015) (“[A]esthetic functionality should only be found by courts when the product feature at issue triggers a positive cognitive, psychological, or aesthetic response among a substantial composite of the relevant consumers and that response predates the trademark owner’s activities.” (emphasis omitted)).

102 Savin Corp. v. Savin Grp., 391 F.3d 439, 456 (2d Cir. 2004).

103 Time, Inc. v. Petersen Publ’g Co. L.L.C., 173 F.3d 113, 117 (2d Cir. 1999) (quoting Arrow Fastener Co. v. Stanley Works, 59 F.3d 384, 391 (2d Cir. 1995)).

104 Hornady Mfg. Co. v. Doubletap, Inc., 746 F.3d 995, 1007 (10th Cir. 2014).

105 Time, Inc., 173 F.3d at 118.  Some courts treat all inherently distinctive marks—whether suggestive, arbitrary, or fanciful—as conceptually strong.  E.g., Hornady, 746 F.3d at 1007.

106 Hornady, 746 F.3d at 1007–08.

107 There is no necessary correlation between conceptual and commercial strength.  For example, descriptive marks, such as COCA-COLA, can have enormous consumer awareness and thus commercial strength.  Greene & Wilkerson, supra note 16, at 552–53.  Conversely, the fanciful mark of a relatively unknown or new business might have limited commercial strength.

108 Id. at 535.  Timothy Greene and Jeff Wilkerson would define trademark strength as “the scope of protection afforded a trademark by courts based on: (1) the mark’s tendency to signify to consumers the source of the products to which the mark is affixed; and (2) the mark’s ability to influence a consumer’s purchasing decisions.”  Id. at 546.  Barton Beebe and Scott Hemphill argue that trademark law should not always afford stronger marks broader scope because the actual likelihood of confusion with strong marks is actually lower than with weaker marks.  Barton Beebe & C. Scott Hemphill, The Scope of Strong Marks: Should Trademark Law Protect the Strong More than the Weak?, 92 N.Y.U. L. REV. 1339, 1345–47 (2017).

109 Beebe & Hemphill, supra note 108, at 1361.

110 TCPIP Holding Co. v. Haar Commc’ns Inc., 244 F.3d 88, 100–01 (2d Cir. 2001).

111 E.g., U.S. Pat. & Trademark Off. v. B.V., 140 S. Ct. 2298, 2307 (2020) (citing 2 MCCARTHY, supra note 50, § 11:76).

112 Id. at 2307–08.

113 Barton Beebe, An Empirical Study of the Multifactor Tests for Trademark Infringement, 94 CALIF. L. REV. 1581,1633–39 (2006).  But cf. Pierre N. Leval, Trademark: Champion of Free Speech, 27 COLUM. J.L. & ARTS 187, 192 (2004) (“[A] descriptive mark, once it has achieved secondary meaning designating its owner, commands some area of exclusivity, but very little.  It cannot prevent use of the same mark in slightly different areas of commerce or use of slightly different marks in the same area of commerce.  If the owner tries to enforce the mark beyond the narrow area of exclusivity, the weak nature of the descriptive mark should defeat the claim.”).

114 Supra Section I.A.

115 Supra Part I.

116 Supra Section I.A.  As Graeme Dinwoodie relatedly explains, “[t]he inherent distinctiveness analysis is prospective, even speculative.  It is a predictive inquiry.  In contrast, the secondary meaning inquiry focuses on evidence of actual consumer association.”  Dinwoodie, supra note 44, at 487.

117 Dinwoodie, supra note 44, at 487; supra Section I.A; cf. Laura A. Heymann, The Grammar of Trademarks, 14 LEWIS & CLARK L. REV. 1313, 1317–18 (2010) (“On the one hand, the ultimate answer to th[e] question [of protectability] is whether a word will have or continues to have trademark significance to consumers in the relevant market, a question that is typically answered by reference to empirical evidence.  On the other hand, this inquiry is in service of a more prescriptive goal of trademark holders: to shape the language usage not only of competitors in the marketplace but of the public more generally.”).

118 Supra text accompanying notes 69–72.

119 2 MCCARTHY, supra note 50, §§ 15:42, 15:48; Bone, supra note 93, at 2131.

120 Supra Section I.A.

121 2 MCCARTHY, supra note 50, § 15:48 (citing cases).

122 Greene & Wilkerson, supra note 16, at 556–57.

123 Id. at 557.

124 2 MCCARTHY, supra note 50, §§ 15:47, 15:53.

125 See, e.g., Co-Rect Prods., Inc. v. Marvy! Advert. Photography, Inc., 780 F.2d 1324, 1332 (8th Cir. 1985) (“Although it is true that advertising is a relevant factor in determining whether a mark has acquired a secondary meaning, it is the effect of such advertising that is important, not its extent. . . . To be effective in this respect, the advertising must cause the public to equate the mark with the source of the product.” (quoting Sec. Ctr., Ltd. v. First Nat’l Sec. Ctrs., 750 F.2d 1295, 1301 (5th Cir. 1985)); Aloe Creme Lab’ys, Inc. v. Milsan, Inc., 423 F.2d 845, 849–50 (5th Cir. 1970) (finding no secondary meaning for ALO as used on ointment even though the business had been using the mark for almost twenty years, had developed a nationwide market, had the product written up in newspapers and magazines, had spent over three million dollars in advertising, and had witness testimony on secondary meaning because “the question is not the extent of the promotional efforts, but their effectiveness in altering the meaning of ‘Alo’ to the consuming public.”).

126 Scott Paper Co. v. Scott’s Liquid Gold, Inc., 589 F.2d 1225, 1228 (3d Cir. 1978).

127 E.g., Beneficial Corp. v. Beneficial Cap. Corp., 529 F. Supp. 445, 448 (S.D.N.Y. 1982).

128 Arrow Fastener Co. v. Stanley Works, 59 F.3d 384, 393–94 (2d Cir. 1995) (presuming secondary meaning also based on the PTO’s registration of the mark).

129 Beef/Eater Rests., Inc. v. James Burrough Ltd., 398 F.2d 637, 639–40 (5th Cir. 1968).  For some other cases relying heavily on circumstantial evidence in this way, see Zatarains, Inc. v. Oak Grove Smokehouse, Inc., 698 F.2d 786, 795 (5th Cir. 1983) (upholding a district court finding of secondary meaning for FISH-FRI for fish-fry batter when the court relied heavily on advertising expenditures of over $400,000 over five years and an increase in annual sales by just over 20,000 cases over a decade); Beneficial, 529 F. Supp. at 447–48 (finding secondary meaning for BENEFICIAL as applied to money lending services based on evidence of extensive advertising, billions of dollars of loan agreements using this name, and survey evidence).

130 Larsen v. Terk Techs. Corp., 151 F.3d 140, 148–49 (4th Cir. 1998); accord Marker Int’l v. DeBruler, 844 F.2d 763, 764 (10th Cir. 1988) (“Defendant deBruler stated that he continued to use the Marker name and sloping ‘M’ logo because Marker had a reputation for quality products and he believed people might associate that reputation with the Marker Surf America products. . . . These statements by the defendant constitute an admission that the Marker International trademark has secondary meaning.” (citing Marker Int’l v. deBruler, 635 F.Supp. 986, 1000 (D. Utah 1986)).

131 2 MCCARTHY, supra note 50, § 15:38.

132 15 U.S.C. § 1065 (2018).

133 Id. § 1115(b).  See generally Tushnet, supra note 65.

134 Park ’N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189, 205 (1985).  Of course, to take this pathway to establish secondary meaning, one must first secure a trademark registration, which in theory does not allow registration of descriptive trademarks without the applicant demonstrating secondary meaning.  See 15 U.S.C. § 1052(f) (2018).  Yet the PARK ’N FLY example itself demonstrates that plausibly descriptive marks can be registered without showing secondary meaning.  Moreover, trademark law makes registration of descriptive marks yet easier by presuming that a mark in use for five years before registration has acquired secondary meaning.  Id.

135 Lee DeRosia & Christensen, supra note 94, at 1038.

136 Id. at 1078.  The generic words used in the experiment were not, however, perceived as source-indicating.  Id. at 1090.

137 See id. at 1083–86.  They also find that consumers do not recognize descriptive words as source indicators only when such contextual markers of trademark use are weaker.  Id. at 1096–99.

138 Id. at 1038.

139 Supra Section I.A.

140 Supra Section I.B.

141 Supra Section I.A.

142 Supra Part I.

143 For a proposal along these lines, see Buccafusco, Masur & McKenna, supra note 43.

144 In an age of social media, it is possible that a mark could go viral without as much expenditure as classically.  That could be a path by which small businesses can establish secondary meaning without as much expenditure, as likely happened to apparel company Roman Originals when its dress—“Is it blue and black, or is it white and gold?”—went viral and led to exponentially increased sales and 10,000 tweets per minute.  Meg Prater, Help! My Brand Went Viral: 16 Small Brands That Made It Big, HUBSPOT (Jan. 11, 2022), [].  Even in today’s age, however, that is the exception rather than the rule.  See id.

145 Cf. Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 775 (1992) (“[A]dding a secondary meaning requirement [for inherently distinctive trade dress] could have anticompetitive effects, creating particular burdens on the startup of small companies.”).

146 In discussing the consequences as “distributive,” I merely mean that new entrants and businesses with fewer resources tend to suffer as compared with businesses with more resources.  Of course, a billionaire can sit behind a new or small company and shareholders in the largest companies can be teachers investing through pension funds, so there is no necessary relationship between company age or size and the wealth of its owners.

147 Beebe & Fromer, supra note 101, at 1021–29 (discussing these consequences in the face of increasing levels of trademark depletion and congestion).

148 Id. at 964 & n.85 (first citing Michael J. Cooper, Orlin Dimitrov & P. Raghavendra Rau, A by Any Other Name, 56 J. FIN. 2371 (2001); then citing Michael J. Cooper, Huseyin Gulen & P. Raghavendra Rau, Changing Names with Style: Mutual Fund Name Changes and Their Effects on Fund Flows, 60 J. FIN. 2825 (2005); and then citing T. Clifton Green & Russell Jame, Company Name Fluency, Investor Recognition, and Firm Value, 109 J. FIN. ECONS. 813 (2013)).

149 Supra Section I.A.

150 2 MCCARTHY, supra note 50, § 11:5; accord supra Section I.A.

151 Bridgestone Ams. Tire Operations, LLC v. Fed. Corp., 673 F.3d 1330, 1334 (Fed. Cir. 2012).

152 M2 Software, Inc. v. M2 Commc’ns, Inc., 450 F.3d 1378, 1384 (Fed. Cir. 2006).

153 PACCAR Inc. v. TeleScan Techs., L.L.C., 319 F.3d 243, 251 (6th Cir. 2003).

154 Carson Mfg. Co. v. Carsonite Int’l Corp., 686 F.2d 665, 670 (9th Cir. 1981).

155 Swagway, LLC v. Int’l Trade Comm’n, 934 F.3d 1332, 1337 (Fed. Cir. 2019) (quoting Nonconfidential Joint Appendix at 235, Swagway, 934 F.3d 1332 (No. 2018-1672)).

156 Segue or Segway?, MERRIAM-WEBSTER, [].

157 Segue, DICTIONARY.COM, [].

158 Ausable River Trading Post, LLC v. Dovetail Sols., Inc., 902 F.3d 567, 571 (6th Cir. 2018) (acknowledging, however, that the mark might thus be suggestive).

159 A & H Sportswear Inc. v. Victoria’s Secret Stores, Inc., 166 F.3d 191, 195 (3d Cir. 1999).

160 Estrada v. Telefonos de Mexico, S.A.B. de C.V., 447 Fed. App’x 197, 202 (Fed. Cir. 2011).

161 U.S. Jaycees v. Phila. Jaycees, 639 F.2d 134, 143 (3d Cir. 1981).

162 W. & Co. v. Arica Inst., Inc., 557 F.2d 338, 340 (2d Cir. 1977) (per curiam).

163 Surfvivor Media, Inc. v. Survivor Prods., 406 F.3d 625, 632 (9th Cir. 2005).

164 Interstellar Starship Servs., Ltd. v. Epix Inc., 184 F.3d 1107, 1111 (9th Cir. 1999) (internal citation omitted) (first citing THE COMPACT OXFORD ENGLISH DICTIONARY (J.A. Simpson, E.S.C. Weiner & Donna Lee Berg eds., 2d ed. 1991); and then citing AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 349 (9th Cir. 1979)).

165 Cf. Beverly W. Pattishall, The Lanham Trademark Act—Its Impact over Four Decades, 76 TRADEMARK REP. 193, 220 (1986) (criticizing the Abercrombie spectrum as “artificial and regrettable”).

166 Cf. KELLER, supra note 101, at 151–52 (discussing this feature from a marketing perspective).

167 Beebe & Fromer, supra note 101, at 967–68; Linford, Fanciful Marks, supra note 16, at 750–54.

168 Beebe & Fromer, supra note 101, at 967.

169 Linford, Fanciful Marks, supra note 16, at 756.  Another more recent issue related to fanciful marks is the rise of nonsense marks—such as LJXOAIEU for hair clips, QIANDLEE for apparel, JANRSTIC for headphones, and AEZLHJYA for jewelry—typically used on Amazon or other platforms.  Note, Fanciful Failures: Keeping Nonsense Marks off the Trademark Register, 134 HARV. L. REV. 1804, 1804–05 (2021).  While these marks are coined, they are conceptually unrelated to the goods with which they are associated because they are nonsense, yet they raise the different legal issue of failing to function as trademarks.  Id. at 1817–25.

170 2 MCCARTHY, supra note 50, § 11:11; accord supra Section I.A.

171 Hard Candy, LLC v. Anastasia Beverly Hills, Inc., 921 F.3d 1343, 1361 (11th Cir. 2019).

172 Radio Shack Corp. v. Radio Shack, Inc., 180 F.2d 200, 205–06 (7th Cir. 1950).

173 Philip Morris Inc. v. K2 Corp., 555 F.2d 815, 816 (C.C.P.A. 1977).

174 See Procter & Gamble Co. v. A. E. Staley Mfg. Co., 342 F.2d 476, 478 (C.C.P.A. 1965); Greg Johnson, New Owners Hope Oxydol Brand Isn’t All Washed up, L.A. TIMES (July 7, 2000, 12:00 AM), [] (noting that OXYDOL received its name from the oxygen in bleach).

175 Rex Shoe Co. v. Juv. Shoe Corp. of Am., 273 F.2d 179, 180 (C.C.P.A. 1959).

176 Maternally Yours, Inc. v. Your Maternity Shop, Inc., 234 F.2d 538, 544 (2d Cir. 1956).

177 Triangle Publ’ns, Inc. v. Rohrlich, 167 F.2d 969, 972 (2d Cir. 1948).

178 Fromer, supra note 98, at 1908, 1912–13.

179 Id. at 1912–13 (first citing Anheuser-Busch Inc. v. Stroh Brewery Co., 750 F.2d 631, 633 (8th Cir. 1984); then citing G. Heileman Brewing Co. v. Anheuser-Busch, Inc., 873 F.2d 985, 1000 (7th Cir. 1989); then citing Tradewell Stores, Inc. v. T.B. & M., Inc., 500 P.2d 1290, 1294 (Wash. Ct. App. 1972); then citing Shoppers Fair of Ark., Inc. v. Sanders Co., 328 F.2d 496, 499 (8th Cir. 1964); and then citing Food Fair Stores, Inc. v. Lakeland Grocery Corp., 301 F.2d 156, 160–61 (4th Cir. 1962)).

180 Citibank, N.A. v. Citibanc Grp., Inc., 724 F.2d 1540, 1545 (11th Cir. 1984).

181 Dial-A-Mattress Operating Corp. v. Mattress Madness, Inc., 841 F. Supp. 1339, 1347–48 (E.D.N.Y. 1994).

182 In re Shop-Vac Corp., 219 U.S.P.Q. (BNA) 470, 471–72 (T.T.A.B. 1983).

183 Glamorene Prods. Corp. v. Boyle-Midway, Inc., 188 U.S.P.Q. (BNA) 145, 164–65 (S.D.N.Y. 1975).

184 Pom Wonderful LLC v. Hubbard, 775 F.3d 1118, 1126–27 (9th Cir. 2014).

185 This concern is distinct from, though related to, Bob Bone’s recognition that a conclusive presumption of secondary meaning for suggestive, arbitrary, and fanciful marks minimizes administrative and error costs in trademark law because these categories are likely to have secondary meaning.  Bone, supra note 93, at 2127–34.  His reasoning depends on proper categorization of marks into these categories, which he then uses to suggest that marks in these categories more often than not have secondary meaning and the law can thus avoid the expensive task of assessing secondary meaning.  Id.  By contrast, what I suggest here is that the law regularly enough miscategorizes marks as suggestive, arbitrary, or fanciful when they should perhaps be placed instead in the descriptive category and asked to show secondary meaning before being protectable.

186 Cf. Daniel J. Hemel, Polysemy and the Law, 76 VAND. L. REV. 1 (forthcoming 2023) (manuscript at 1), (analyzing the costs and benefits of polysemy—”the existence of multiple related meanings for the same word or phrase”—as well as a framework for it, in legal language).

187 See Jake Linford, The False Dichotomy Between Suggestive and Descriptive Trademarks,76 OHIO ST. L.J. 1367, 1402–20 (2015) (discussing polysemy with regard to descriptive and suggestive marks).

188 Jacob Jacoby, The Psychological Foundations of Trademark Law: Secondary Meaning, Genericism, Fame, Confusion and Dilution, 91 TRADEMARK REP. 1013, 1028–29 (2001) (“By repeated exposure to and apprehension of the outside world, consumers develop cognitive networks to mentally represent the companies, products, brands and stores that they encounter.  Once ensconced in memory, these networks assist in the interpretation . . . of what is experienced.” (footnote omitted)).  According to marketing expert Kevin Keller, “[t]he associative network memory model views memory as a network of nodes and connecting links.  According to this model, recall or retrieval of information occurs through a concept called spreading activation.”  KELLER, supra note 101, at 358.  He continues, “When the activation of a particular node exceeds a threshold level, the person recalls the contents of that node.  The spread of activation depends on the number and strength of the links connected to the activated node: Concepts whose linkages have the greatest strength will receive the most activation.”  Id.

189 Cf. Ramsey, supra note 65, at 1099 (maintaining that protecting descriptive marks is inconsistent with the First Amendment because even when such marks indicate source—as with the example of FAIR & BALANCED for Fox News—they still “retain their original descriptive meaning”).  Even if consumers can differentiate between the multiple associations a word has, Heymann, supra note 117, at 1338–39 (providing the example of “Shirley Temple,” referring both to the person and to the nonalcoholic cocktail), that does not mean that one business should be able to monopolize such terms when doing so would confer an unfair competitive advantage.

190 Beebe & Fromer, supra note 101, at 965 (giving as examples APPLE for computers and KIND for snack bars).

191 Id. at 966 (footnote omitted); see also supra note 101 (discussing marketing research on choosing different types of words as marks).

192 Cf. Dinwoodie, supra note 44, at 503–04 (“Distinctiveness is a measure of consumer association, and analysis of competitive impact adds little to our understanding of whether word marks identify and distinguish a particular product.  Competitive concerns may underlie trademark protection, but they do not provide appropriate guidance on the concrete question of distinctiveness.” (footnote omitted)).  Lisa Ramsey worries about how protection for descriptive marks raises worries not necessarily for competition but for the First Amendment.  Ramsey, supra note 65.

193 See KELLER, supra note 101, at 151; Kevin Lane Keller, Susan E. Heckler & Michael J. Houston, The Effects of Brand Name Suggestiveness on Advertising Recall, 62 J. MKTG. 48, 56 (1998).

194 KELLER, supra note 101, at 151.

195 Clorox Chem. Co. v. Chlorit Mfg. Corp., 25 F. Supp. 702, 705 (E.D.N.Y. 1938) (“‘Clorox’ is a fanciful word, arbitrarily selected and in no wise describes its ingredients.”).  The concerns of enshrining secondary meaning in trademark law generally likely also apply to deceptively misdescriptive marks like THCTEA (for tea that does not contain THC), In re Hinton, 116 U.S.P.Q.2d (BNA) 1051 (T.T.A.B. 2015), and FURNITURE MAKERS (for a furniture store that sells, but does not make, furniture), In re Berman Bros. Harlem Furniture, 26 U.S.P.Q.2d (BNA) 1514 (T.T.A.B. 1993).  Deceptively misdescriptive marks are protectable upon a showing of secondary meaning, just as descriptive marks are.  15 U.S.C. §§ 1052(e)–(f) (2018).  (By contrast, marks are categorized as “deceptive” when the deceptiveness materially affects a consumer’s decision to purchase and are never protectable regardless of the presence of secondary meaning.  Id. § 1052(a); In re Budge Mfg. Co., 857 F.2d 773, 774–75 (Fed. Cir. 1988).  Deceptively misdescriptive marks lack this materiality.  Glendale Int’l Corp. v. U.S. Pat. & Trademark Off., 374 F. Supp. 2d 479, 485–86, 485 n.10 (E.D. Va. 2005).  As with run-of-the-mill descriptive marks, deceptively misdescriptive marks can be protectable too easily when the bar to establish secondary meaning is set too low, and there is competitive inequity to setting the bar higher.  See supra Sections II.A–B.  Similarly, there can be a coexistence of secondary and primary meaning for deceptively misdescriptive marks, yet here, the worry is less about other businesses having a strong competitive need to use the mark with a certain primary meaning than about consumers being misled by the use of the mark.

196 See RiseandShine Corp. v. PepsiCo, Inc., 41 F.4th 112, 121–23 (2d Cir. 2022).

197 Id.

198 Id. at 122.

199 Id.

200 Id.

201 Id.

202 Linford, Generic Marks, supra note 16, at 115–16, 121–22 (first citing Harley-Davidson, Inc. v. Grottanelli, 164 F.3d 806 (2d Cir. 1999); and then citing In re, L.P., 573 F.3d 1300 (Fed. Cir. 2009)).  To be fair, Linford seems to be advocating this allowance mostly for terms that have undergone semantic shift and were once but are no longer perceived as generic.  Cf. Deven R. Desai & Sandra L. Rierson, Confronting the Genericism Conundrum, 28 CARDOZO L. REV. 1789, 1832–33 (2007) (arguing that secondary meaning should take precedence for once-but-no-longer generic terms and they should be accorded protection to reduce consumer search costs).  That said, the examples and categories of words for which he would allow protection seem also to encompass still-generic terms where enough consumers also perceive them as having secondary meaning.  Cf. Timothy Greene, Trademark Hybridity and Brand Protection, 46 LOY. U. CHI. L.J. 75, 79 (2014) (“A given term need not be understood only as a generic term or only as a source identifier. . . . Rather, all terms—and perhaps especially words used as trademarks—embody multiple concepts . . . .” (footnote omitted)).

203 Linford, Fanciful Marks, supra note 16, at 760–61.  Linford also suggests that Congress and courts might consider requiring secondary meaning of all marks before protecting them.  Id. at 764.

204 Fromer, supra note 98, at 1912–13.

205 Linford, supra note 187, at 1367, 1402–03, 1416.

206 As discussed above, it is little comfort for competitors to rely on the limited and murky doctrine of descriptive fair use for third-party use of a descriptive mark to describe goods and services.  See supra note 65.  Beyond the narrowness of this defense, it applies only to marks categorized as descriptive, not to those classified as suggestive, arbitrary, or fanciful, though there might be just as much reason for a third party to use such marks to describe their goods or services, particularly given the odd categorizations of many marks by courts.  See supra Section II.C; cf. Joseph Scott Miller, Abercrombie 2.0—Can We Get There from Here? Thoughts on “Suggestive Fair Use”, 77 OHIO ST. L.J. FURTHERMORE 1, 9–14 (2016) (proposing a doctrine of suggestive fair use to manage the scope of suggestive marks given that they—like descriptive marks—describe aspects of their associated goods or services).

207 See supra Part I.

208 Supra text accompanying notes 44–46.

209 See supra Part I.

210 KELLER, supra note 101, at 2.

211 Id.

212 See id. at 10, 39, 49.

213 Id. at 54–55.

214 See id. at 55–56, 64–66.  Of course, the mark itself adopted might make the establishment of brand awareness easier or harder.  As the marketing textbook explains, “consumers can form brand associations . . . by assumptions or inferences consumers make about the brand itself, its name, logo, or identification with a company, country, channel of distribution, or person, place, or event.”  Id. at 56.

215 Relatedly, the sums spent underscore that when there is a great conceptual distance between a mark and its associated good or service, as there should be with arbitrary and fanciful marks, advertising is important to establish with what type of good or service the mark is to be associated.  That is, without any learning, it is hard for consumers to know whether EXXON is a mark for gasoline or coffee and whether APPLE is a mark for computers or socks.

216 Alex Huntsberger, Here Are the 10 Most Expensive Commercials Ever Made, OPPU (Dec. 22, 2021), [].  To be sure, such businesses might also be striving to establish secondary meaning even if not necessary for trademark protection as a way to achieve broader scope for their marks by establishing greater commercial strength.  See supra SectionI.B.

217 See, e.g., Mobil Oil Corp. v. Pegasus Petroleum Corp., 818 F.2d 254, 258 (2d Cir. 1987).  But see, e.g., Lois Sportswear, U.S.A., Inc. v. Levi Strauss & Co., 799 F.2d 867, 875 (2d Cir. 1986) (“[I]ntent is largely irrelevant in determining if consumers likely will be confused as to source.”).

218 TCPIP Holding Co. v. Haar Commc’ns Inc., 244 F.3d 88, 96, 102 (2d Cir. 2001).

219 Beebe, supra note 113, at 1628.

220 Cf. Graeme B. Dinwoodie, What Linguistics Can Do for Trademark Law, in TRADE MARKS AND BRANDS: AN INTERDISCIPLINARY CRITIQUE 140, 152 (Lionel Bently, Jennifer Davis & Jane C. Ginsburg eds., 2008) (suggesting that businesses that adopt “marks at the margin of descriptiveness” do so “because they would like to appropriate not only the source-identifying capacity of the term, but any evocative power”).

221 This practice has likely occurred for some time, such as when a business would choose a descriptive mark so that it might be provided by a telephone operator when a consumer would call and ask for the category of goods or services at hand.  Yet the practice has likely increased sharply in an age of the internet and search engines.

222 For an exploration of how search engine results might help resolve trademark law issues more broadly, see Ouellette, supra note 50.

223 24 Hour Fitness USA, Inc. v. 24/7 Tribeca Fitness, LLC, 277 F. Supp. 2d 356, 366 (S.D.N.Y. 2003).

224 1-800 Contacts, Inc. v., Inc., 755 F. Supp. 2d 1151, 1179 (D. Utah 2010).

225 Shade’s Landing, Inc. v. Williams, 76 F. Supp. 2d 983, 989–90 (D. Minn. 1999).

226 U.S. Pat. & Trademark Off. v. B.V., 140 S. Ct. 2298, 2308 (2020).

227 Id.

228 The situation might also fall in between these two possibilities, where some consumers purchase a business’s goods or services with awareness of secondary meaning for its descriptive mark and others do so after locating the business’s products based on its descriptive mark as descriptor.

229 On dictionaries to interpret terms in other legal contexts—like statutes and patents—see Joseph Scott Miller & James A. Hilsenteger, The Proven Key: Roles and Rules for Dictionaries at the Patent Office and the Courts, 54 AM. U. L. REV. 829 (2005);A. Raymond Randolph, Dictionaries, Plain Meaning, and Context in Statutory Interpretation, 17 HARV. J.L. & PUB. POL’Y 71 (1994).  On the use of corpora for legal construction of terms, see Thomas R. Lee & Stephen C. Mouritsen, Judging Ordinary Meaning, 127 YALE L.J. 788 (2018); Thomas R. Lee & Stephen C. Mouritsen, The Corpus and the Critics, 88 U. CHI. L. REV. 275 (2021).  Using dictionaries and corpora might be more fraught in these other contexts than in trademark law because of the more complex normative questions about interpreting and construing laws and technical legal documents.


231 Cf. Dinwoodie, supra note 220 (asserting that trademark law should assess the meaning of marks in context of their use).

232 Bank, MERRIAM-WEBSTER, [].


234 Combe Inc. v. Dr. August Wolff GMBH & Co. KG Arzneimittel, 851 F. App’x 357, 360 (4th Cir. 2021).

235 Id. at 361.  It also cannot do consumers much good to have at least sixty-seven different vaginal-health product names begin with VAGI-, but that is another topic.

236 This is true even though Starbucks in part chose its mark because it calls to mind “the romance of the high seas and the seafaring tradition of the early coffee traders,” which connects to the Moby Dick character.  Emma Kumer, How Did Starbucks Gets Its Name? The Real Story Might Surprise You, TASTE OF HOME (Mar. 14, 2018), [].

237 Philip Levinson, How Nike Almost Ended up with a Very Different Name, INSIDER (Jan. 19, 2016, 2:35 PM), [].

238 See James, The Meaning of the Word ‘Kodak’, DARK LANE CREATIVE (Jan. 11, 2014), [].

239 Id.

240 Supra text accompanying note 168.

241 Beebe & Fromer, supra note 101, at 966 (citing Sharon Begley, StrawBerry Is No BlackBerry: Building Brands Using Sound, WALL ST. J. (Aug. 26, 2002, 3:47 PM), [].

242 See supra Part I.

243 See supra Part I.

244 But cf. Buccafusco, Masur & McKenna, supra note 43, at 28 (analyzing the competitive worries of protecting descriptive and suggestive marks, but rejecting a ban on these categories because “those marks can have some value to consumers,” providing as examples LYFT for ride-sharing services and ICYHOT for topical pain relief products).  Moreover, this analysis presumes that there is not an unlimited number of equally attractive descriptive marks for competitors to use, an assumption suggested by empirical research showing that competitively effective trademarks are being readily depleted over time.  See generally Beebe & Fromer, supra note 101.

245 See supra Section I.A.  Moreover, this proposal bears some relationship to a return to nineteenth-century-era protection only for what were known as “technical trademarks,” terms that were arbitrary or fanciful, those categories whose terms are thought to be the most distant conceptually from their associated goods or services.  See Mark P. McKenna, Trademark Use and the Problem of Source, 2009 U. ILL. L. REV. 773, 781–82.  All other terms—descriptive ones, suggestive ones (which were not yet a legal category of their own), and surnames—could not be technical trademarks.  See id. at 782.  Yet they might be protectable as trade names and enforceable under unfair competition law.  See id.  Unfair competition law was narrower than trademark protection, affording protection only against deceptive practices.  See id. at 782–85.  Yet this Article’s proposal differs by eliminating the Abercrombie spectrum, including its categories for arbitrary and fanciful marks, as it often categorizes terms closely related conceptually to their associated goods or services as inherently distinctive and thus protectable.  See supra Section II.C.  Moreover, this Article’s proposal would not allow protection for marks that are closely conceptually related to their associated goods or services, something that the older system would allow in some measure under the law of unfair competition.  This proposal is also similar to a ban on descriptive and suggestive marks, but because the Abercrombie spectrum has been so convoluted in ways described herein, it is preferable to strip down the elaborate spectrum to its core.

246 See supra Section II.C.  It would be possible to sharpen up the inquiries in each Abercrombie category to make the categorization of terms more meaningful.  For example, one might ask courts to be more functional in analyzing whether a term is fanciful rather than just asking whether it is coined.  See supra Section II.C.  At this juncture, however, that feels like a fool’s errand because it would impose more complexity on the law than is necessary—as compared with a direct inquiry into primary meaning—as it would be to refine a Ptolemaic model of the solar system instead of simplifying it to account for scientific knowledge.  Cf. Carol M. Rose, Crystals and Mud in Property Law, 40 STAN. L. REV. 577, 580 (1988) (analyzing “the blurring of clear and distinct property rules with the muddy doctrines of ‘maybe or maybe not,’ and about the reverse tendency to try to clear up the blur with new crystalline rules”).  Moreover, a refined Abercrombie analysis does not address the concerns with protecting descriptive marks upon a showing of secondary meaning.  See supra Section II.A–B.

247 See supra Sections II.A–B, D.

248 Linford, Generic Marks, supra note 16, at 139.  Rebecca Tushnet criticizes the Abercrombie spectrum as being dissociated from how consumers actually attach meaning to source indicators.  See Tushnet, supra note 92, at 873–74.  She proposes instead that the law “divide words into generic terms, terms that are protectable with secondary meaning, and terms that are inherently distinctive, with the presumption being that secondary meaning ought to be required in cases of doubt.”  Id. at 874.

249 See Bone, supra note 93, at 2130–34; Robert C. Denicola, Freedom to Copy, 108 YALE L.J. 1661, 1673 (1999); Michael Grynberg, The Judicial Role in Trademark Law, 52 B.C. L. REV. 1283, 1290 n.41 (2011).

250 See, e.g., Linford, Fanciful Marks, supra note 16, at 759–60.

251 Lee, DeRosia & Christensen, supra note 94, at 1038.

252 See supra text accompanying notes 246–47.

253 See supra Part II.

254 See supra text accompanying notes 98–101.

255 Cf. Graeme B. Dinwoodie & Dev S. Gangjee, The Image of the Consumer in EU Trade Mark Law, in THE IMAGES OF THE CONSUMER IN EU LAW: LEGISLATION, FREE MOVEMENT AND COMPETITION LAW 339, 339 (Dorota Leczykiewicz & Stephen Weatherill eds., 2016) (exploring “the role played by the consumer in EU trade mark law,” and maintaining that “most efforts by [EU] courts to identify or construct consumers and the marketplace they inhabit are blended exercises that are part-empirical and part-normative”).  In many ways, a shift toward primary meaning and away from secondary meaning would emphasize (expert and other) linguistic evidence over consumer evidence.

256 See supra text accompanying notes 135–38.

257 See supra Section II.E.

258 See supra Section II.E.

259 To be sure, the smaller business might not be able to afford a branding consultant that the deeper-pocketed business could more easily pay to help find an optimal mark.  See Pia Silva, How Much Does Branding Cost?, FORBES (July 13, 2017, 2:12 PM), [] (discussing the various costs of branding firms).  Yet both firms are decently placed to come up with possible marks in consultation with a reasonably priced trademark lawyer to clear the marks.

260 See supra Section II.B.

261 See supra Section II.B.  With regard to distributive consequences, it is worth recognizing that this Article’s proposal to gauge protectability via primary meaning would restrict the number of potential competitively effective marks available to businesses from which to choose, an issue of growing concern in light of the ever-increasing rates of trademark depletion and congestion.  See generally Beebe & Fromer, supra note 101 (empirically studying trademark depletion and congestion in the United States).  Though a shrinking pool is likely, it is nonetheless ultimately preferable to leaving protectability rules as they are because the distributive and competitive consequences of the current trademark system are sufficiently severe.  Yet this effect might mean that earlier firms have an advantage over later firms in terms of choosing marks.  See id. at 1021–29 (discussing this consequence generally with regard to depletion and congestion).  That said, trademark law might be reformed to decrease these temporal consequences, as explored in previous work.  See id. at 1029–41.

262 See Buccafusco, Masur & McKenna, supra note 43, at 51–60.

263 Id. at 44.

264 See supra Section II.A.

265 In response to this concern, Buccafusco, Masur, and McKenna note that “wealth effects are unavoidable consequences of using prices as a sorting mechanism,” yet they think that their proposal would ultimately help small businesses by deterring businesses from choosing descriptive and suggestive marks in the first place.  See Buccafusco, Masur & McKenna, supra note 43, at 59–60.  That seems unlikely unless the price for establishing secondary meaning is set incredibly higher than it is currently, which seems improbable.

266 See Dinwoodie, supra note 220, at 142, 153; supra Section I.A.

267 See, e.g., Markman v. Westview Instruments, Inc., 517 U.S. 370, 388 (1996) (stating in the context of patent claim construction that “[t]he construction of written instruments is one of those things that judges often do and are likely to do better than jurors unburdened by training in exegesis”); supra note 229.

268 See Fromer, supra note 98, at 1910–14; supra Section II.C.

269 Courts would have to take care not to elaborate the inquiry back into an Abercrombie-like inquiry to avoid the current concerns.

270 See supra Section II.C.

271 See supra text accompanying notes 230–31.

272 See, e.g., Custom Vehicles, Inc. v. Forest River, Inc., 476 F.3d 481, 486 (7th Cir. 2007); Flynn v. AK Peters, Ltd., 377 F.3d 13, 19 (1st Cir. 2004); 2 MCCARTHY, supra note 50, § 11:11; Fromer, supra note 98, at 1888; Michael Grynberg, A Trademark Defense of the Disparagement Bar, 126 YALE L.J.F. 178, 185 (2016); Jake Linford, Trademark Owner as Adverse Possessor: Productive Use and Property Acquisition, 63 CASE W. RSRV. L. REV. 703, 710 (2013).

273 See Apple Logo and the History Behind the Company, LOGOMYWAY, [] [hereinafter Apple Logo].

274 See Karl Klooster, Mac’s Link to Mac, OR. WINE PRESS (Dec. 1, 2011), [].

275 See Apple Logo, supra note 273; Holden Frith, Unraveling the Tale Behind the Apple Logo, CNN (Oct. 7, 2011, 3:07 PM), [].

276 See Frith, supra note 275.

277 Apple Logo Evolution—It All Started with a Fruit, LOGO CREATIVE (Aug. 10, 2018), [].

278 Laura Heymann conceptualizes a similar differentiation in terms of metaphor.  See Heymann, supra note 117, at 1330.  She gives the examples of BRAWNY for paper towels “not only to serve as the name of the product but also to encourage consumers to think of the product in terms of human strength,” and APPLE for computers rather than SPINACH because of the difference in connotation to the youth market.  Id. at 1331.  To her, those marks are more metaphorical in nature and thus would be readily protectable.  See id. at 1334.

279 Deven R. Desai, From Trademarks to Brands, 64 FLA. L. REV. 981, 988 (2012); accord KELLER, supra note 101, at 2–5 (distinguishing between the “small b” brand, which seems akin to the mark, and the “big B” brand, which relates to the “awareness, reputation, [and] prominence” of a mark, as well as the “rational and tangible,” and “symbolic, emotional, and intangible” aspects that the mark represents).

280 See supra Section I.B; infra Part IV.

281 Cf. Beebe & Fromer, supra note 101, at 1030–35 (navigating similar concerns in advocating reform to address trademark depletion and congestion).

282 See supra Part III.

283 See supra Section I.B.

284 Supra Section I.B.

285 See supra text accompanying notes 245–47.

286 See supra Part II.

287 See supra Section I.B.

288 As noted above, Barton Beebe and Scott Hemphill maintain that trademark law should not always afford commercially stronger marks broader scope because the actual likelihood of confusion with strong marks is actually lower than with weaker marks.  See Beebe & Hemphill, supra note 108, at 1366–70, 1393.  Even if that is right and trademark law should often draw an inference against likelihood of confusion the stronger the mark, secondary meaning still remains relevant to the determination.

289 That is not to say that a small business looking to break into the national marketplace and compete with the Coca-Colas, Amazons, and Toyotas of the world is not going to need to provide similar evidence as those business giants would to establish secondary meaning.

290 Aloe Creme Lab’ys, Inc. v. Milsan, Inc., 423 F.2d 845, 850 (5th Cir. 1970).

291 See supra Section II.B.

292 To be clear, this proposal for a sliding scale is not a proposal to recognize what is known as “secondary meaning in the making.”  As Tom McCarthy explains it, according to this theory, “a firm that is making efforts to create secondary meaning in a designation, but has not yet succeeded, should be protected against a competitor who knowingly rushes in to market a product under a similar mark.”  2 MCCARTHY, supra note 50, § 15:58.  Courts have rejected this theory on the ground that trademark law does not protect terms that need secondary meaning to be protected before secondary meaning is established.  For example, the Eighth Circuit has reasoned that this “theory focuses solely upon the intent and actions of the seller of the product to the exclusion of the consuming public; but the very essence of secondary meaning is the association in the mind of the public of particular aspects of trade dress with a particular product and producer.”  Black & Decker Mfg. Co. v. Ever-Ready Appliance Mfg. Co., 684 F.2d 546, 550 (8th Cir. 1982) (citing Truck Equip. Serv. Co. v. Fruehauf Corp., 536 F.2d 1210, 1215 (8th Cir. 1976)); accord Laureyssens v. Idea Grp., Inc., 964 F.2d 131, 137 (2d Cir. 1992); Cicena Ltd. v. Columbia Telecomms. Grp., 900 F.2d 1546, 1550 (Fed. Cir. 1990); A.J. Canfield Co. v. Concord Beverage Co., 629 F. Supp. 200, 212 (E.D. Pa. 1985).  Rather than give protection to a business that is on its way to establishing secondary meaning but has not yet done so, a sliding scale would recognize that secondary meaning is actually established in relation to the size of a business’s footprint: smaller for smaller businesses, and larger for larger ones.

293 Zatarains, Inc. v. Oak Grove Smokehouse, Inc., 698 F.2d 786, 793–96 (5th Cir. 1983).

294 See Zatarain’s, WIKIPEDIA,’s []; Keeping It Real Since 1889, ZATARAIN’S, [].

295 Ralston Purina Co. v. Thomas J. Lipton, Inc., 341 F. Supp. 129, 131–32, 134 (S.D.N.Y. 1972).  Another context in which some courts arguably meter how much secondary meaning must be shown, albeit in order to protect a mark, is with regard to the well-known marks doctrine.  According to Tom McCarthy, this doctrine provides that “even if a mark is used only on products or services sold abroad, if its reputation is well known in the United States, then that mark should be legally recognized in the United States.”  5 MCCARTHY, supra note 50, § 29:4. Courts are divided on whether to recognize this doctrine.  Compare Grupo Gigante SA De CV v. Dallo & Co., 391 F.3d 1088, 1094 (9th Cir. 2004) (approving the doctrine), with ITC Ltd. v. Punchgini, Inc., 482 F.3d 135, 163 (2d Cir. 2007) (rejecting it).  In recognizing the doctrine, the Ninth Circuit emphasized that “a substantial percentage of consumers in the relevant American market [must be] familiar with the foreign mark.”  Grupo Gigante, 391 F.3d at 1098.  Judge Graber, concurring in the decision, made clear that he understood the majority decision to require “a foreign owner of a supposedly famous or well-known foreign trademark [to] show a higher level of ‘fame’ or recognition than that required to establish secondary meaning.”  Id. at 1106 (Graber, J., concurring).

296 Fisons Horticulture, Inc. v. Vigoro Indus., Inc., 30 F.3d 466, 474 (3d Cir. 1994).  On reverse confusion generally, see Anthony L. Fletcher, The Curious Doctrine of Reverse Confusion—Getting It Right in Reverse, 95 TRADEMARK REP. 1273 (2005);Thad G. Long & Alfred M. Marks, Reverse Confusion: Fundamentals and Limits, 84 TRADEMARK REP. 1 (1994); Jeremy N. Sheff, Reverse Confusion and the Justification of Trademark Protection (Sept. 13, 2022) (unpublished manuscript) (on file with author).

297 4 MCCARTHY, supra note 50, § 23:10.

298 Com. Nat’l Ins. Servs., Inc. v. Com. Ins. Agency, Inc., 214 F.3d 432, 444 (3d Cir. 2000).

299 Id. (quoting Banff, Ltd. v. Federated Dep’t Stores, Inc., 841 F.2d 486, 491 (2d Cir. 1988)).

300 A & H Sportswear, Inc. v. Victoria’s Secret Stores, Inc., 237 F.3d 198, 231 (3d Cir. 2000).

301 E.g., Dick’s Sporting Goods, Inc. v. Dick’s Clothing & Sporting Goods, Inc., 188 F.3d 501 (4th Cir. 1999).

302 See Sheff, supra note 296 (manuscript at 21–22).