Cornell University Law School Social Science and Law

How to Prove or Disprove Trade Dress Infringement

Compiled by Michael Baak and Eric Torres

Part 1. Legal Question

Trade dress refers to the characteristics of the visual appearance of a product that signify to the public the source of the product. The governing statute for trade dress violations is the Lanham Act, which is the primary federal trademark statute. Section 43(a) of the Trademark Act of 1946 (Lanham Act) provides:

(1) Any person who, on or in connection with any goods or services, or any container for goods, uses in commerce any word, term, name, symbol, or device, or any combination thereof, or any false designation of origin, false or misleading description of fact, or false or misleading representation of fact, which

1. is likely to cause confusion, or to cause mistake, or to deceive [...] as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person, or
2. in commercial advertising or promotion, misrepresents the nature, characteristics, qualities, or geographic origin of his or her or another person’s goods, services, or commercial activities,

shall be liable in a civil action by any person who believes that he or she is likely to be damaged by such an act. 15 U.S.C. §1125.

As the court goes on to say in Two Pesos v. Taco Cabana, the Lanham Act "was intended to make ‘actionable the deceptive and misleading use of marks’ and ‘to protect persons engaged in . . . commerce against unfair competition.’" Two Pesos v. Taco Cabana, 505 U.S. 763, 768 (1992).

Distinctiveness

Although nothing explicitly in § 43 of the Lanham Act requires a showing of distinctiveness, courts have universally imposed that requirement. To establish a claim for trade dress infringement, a plaintiff must demonstrate the distinctiveness of his or her product’s appearance. Thus, the plaintiff must show that consumers perceive the particular trade dress as identifying the source of a product. Marks (or products) are often classified into 5 different categories of increasing distinctiveness; including (1) generic, (2) descriptive, (3) suggestive, (4) arbitrary, and (5) fanciful. Two Pesos, 505 U.S. at 768. The latter three categories are regarded as being inherently distinctive and worthy of protection because "their intrinsic nature serves to identify a particular source of a product." Id. In other words, a mark is inherently distinctive if "its intrinsic nature serves to identify a particular source."

Wal-Mart Stores, Inc. v. Samara Bros., Inc., 529 U.S. 205, 209 (2000). If a mark is not found to be inherently distinctive, it still can be found to be a distinctive mark if it has developed a secondary meaning. A secondary meaning occurs when "in the minds of the public, the primary significance of a [mark] is to identify the source of the product rather than the product itself." Wal-Mart, 529 U.S. at 209. A mark that has a secondary meaning is still distinct because the consumer has created an association between the trade dress and the source of the trade dress (the company).

Functionality

To receive protection under the Lanham Act, a trade dress must not be functional. "The functionality doctrine prevents trademark law, which seeks to promote competition by protecting a firm’s reputation, from instead inhibiting legitimate competition by allowing a producer to control a useful product feature." Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. 159, 164 (1995). If trademarks were allowed for functional features of a product, then the company could achieve a monopoly over these features regardless of whether they qualify as patents, and the trademarks for these functional features could theoretically be renewed forever. Trademark law, and specifically trade dress law, can be contrasted with patent law, which clearly allows one to have a monopoly for a product design or function for some definite period of time.

This difference between trademark and patent law can be explained by their respective underlying rationales: trademark law seeks to "protect[] competitors against a disadvantage (unrelated to recognition or reputation) that trademark protection might otherwise impose," while patent law seeks to encourage invention. The court in Qualitex presents an example to illustrate this difference. They imagine a new, patented light bulb with a special illumination-enhancing shape. The court says that a manufacturer could not use the shape as a trademark because, after the patent expired, this would be an obstacle to competition, "not by protecting the reputation of the original bulb maker, but by frustrating competitors’ legitimate efforts to produce an equivalent illumination-enhancing bulb." Generally, a feature is functional if "it is essential to the use or purpose of the article or if it affects the cost or quality of the article." Qualitex, 514 U.S. at 165.

What Can Be Trademarked?

Here is a quick overview of some of the prominent trade dress cases and what objects were the focus of the trade dress infringement allegations:

• Restaurant décor – Two Pesos v. Taco Cabana, 505 U.S. 763 (1992).
  • Taco Cabana had established a décor consisting of a "festive eating atmosphere having interior dining and patio areas decorated with artifacts, bright colors, paintings, and murals."
• Color – Qualitex Co. v. Jacobson Products Co., Inc., 514 U.S. 159 (1995).
  • Qualitex used dry cleaning press pads that were colored a special shade of green gold and sought to have this deemed a trade dress.
• Children’s Clothing Design – Wal-Mart Stores, Inc. v. Samara Bros., Inc., 529 U.S. 205, 209 (2000).
  • Children's Clothing Design alleged that Wal-Mart infringed on their trade dress for children's clothes.
• Road/Outdoor Signs – TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001).
  • Petitioner argued that the "dual spring design" on their road side signs had established a secondary meaning that customers recognized.

Part 2. Case Studies

     Using two very well-known trade dress cases, Two Pesos v. Taco Cabana and Apple v. Samsung, this Part will examine in detail which type of trade dress has been found protectable under Lanham Act and what evidence was used to prove thepresence or absence of trade dress.  Particularly, we will look at how expert testimony, surveys and visual presentations were presented to the jury and how persuasive the evidence looked to courts.  Following our discussion, please keep in mind two questions: Should our feelings for a dining hall be protected by law?  Should a particular angle of a curve be exclusive to one type of cell-phone?

1) Two Pesos v. Taco Cabana

2) Apple v. Samsung

Part 3. fMRI as alternative methodology: cognitive science or cognitive silence

1) Inherent Weakness of Surveys and Interviews for Lanham Act purposes

     As shown above, the most frequently used methods for trademark research - surveys and interviews - are fundamentally dependent on self-reports. As modern research found that self-reports are vulnerable to implicit biases, some judges and commentators argue that survey results may be much more malleable than we customarily assume. The Seventh Circuit once criticized that survey results in trademark cases may veil "other tricks of the survey researcher’s black arts.²¹ Chief Judge Posner suggested that the confusion between "Baltimore CFL Colts" and "Indiana Colts" could have been caused by respondents’ unfamiliarity with the "CFL", abbreviation for the Canadian Football League.²² IP scholar Tushnet warned that survey results are subject to a complex interaction between the wording of survey questions and a survey respondent’s expectation.²³ Very often, humans are hooked on "a plausible-sounding reason for why we might like or dislike something, and then we adjust our true preference" to make us look good.²⁴ In the context of Lanham Act lawsuits, Posner and Tushnet raise important concerns about the use or over-use of survey evidence. Survey must be used with caution because, even though researchers may successfully narrow the scope of their research variables by statistically controlling factors that are not their interests, the data could have been – intentionally or unintentionally? – tainted from the beginning.

     Then, would it be possible to directly show what is going on in people’s minds without relying upon subjective responses? In other words, as lawyers, can we establish consumer confusion without ever asking actual consumers? According to neuroscientists, probably we can.

2) Promise of Neurological Evidence: Lessons from Neuromarketing

      Through observation of brain activities rather than consumer responses, neurological evidence examines how the brain processes external stimuli. Of all the neurological methodologies, fMRI (functional Magnetic Resonance Imaging) looks most promising.²⁵ In an fMRI research setting, the live subject remains still inside the fMRI scanner and performs a particular task.²⁶ The scanner then records changes, presumably caused by the task, in the subject’s Blood Oxygenation Level Dependent (BOLD) in the brain.²⁷ This possibility of observing live subjects performing specific tasks makes fMRI attractive. Through fMRI, we can measure psychological processes at the brain-region level, can avoid cognitive bias, and can track simultaneous activation of antithetical conditions.²⁸

A. Threshold Questions

     There are several threshold questions regarding the admissibility of fMRI evidence. First, fMRI evidence may not pass the Daubert test. In 2012, the Sixth Circuit affirmed a district court’s ruling that fMRI lie detection evidence is inadmissible under the Federal Rule of Evidence 702.²⁹ The district court for the Western District of Tennessee found that some of the Daubert factors, error rates, established standards, and general acceptance, were not met. Considering the novelty of the technology, the Sixth Circuit agreed with the district court and concluded that the trier of fact would not be assisted in determining the facts at issue. Still, if the nascency of the technology is the main concern, fMRI could be found admissible in the near future.³⁰ Notably, lie detection evidence is mostly inadmissible because "credibility is a matter solely for the jury and is clearly within the ken of the jury."³¹ Second, unless fMRI research findings have general implications, fMRI research designed for particular purposes may not be admissible in for a case with another set of facts. When fMRI research is performed on a limited number of individuals in a tightly controlled setting, how can we extract generality? One answer may be that fMRI evidence is a type of “social framework” evidence. This issue was addressed by an fMRI research on anti-smoking ads campaign: "Can neural responses of a small group of individuals predict the behavior of large-scale populations?"³² In this research, three series of TV ads (call A, B, and C) promoting an anti-smoking hotline (1-800-QUIT-NOW) were shown to subjects (smokers who want to quit), and their neural activities were measured to rank the ads in the order of effectiveness (the order was C>B>A).³³ The researchers also obtained the rankings from the subjects’ self-reports and from the industry experts (both were B>C>A). As the order of the call volumes turned out to be C>B>A, the research found that the fMRI data in the medial prefrontal cortex (MPFC) at the individual level predicted the hotline number call volume at the population level, while the subjects’ self-reports and the industry expert opinions failed (Fig.12).³⁴ MPFC is implicated in self-related processing for behavior change (whether one can relate to ads and modify behaviors).³⁵ Then, fMRI research of individuals can be used to "track the responses of large groups of people at the population level."

[Fig.12] Brain scan image of a person who watched anti-smoking ads campaigns; comparison of effectiveness rankings from survey, MPFC activation, and call volume (Falk et al., 2012)

B. Coke vs. Pepsi: So, I’m Drinking a Brand?

     Hoping, in the near future, that courtrooms can fully benefit from the potential of fMRI evidence, this Part examines neuromarketing research on consumer confusion.³⁶ fMRI research may provide more accurate understanding of mechanisms behind confusion. Though we commonly assume confusion is a linear match-up with the presence or absence of certain visual stimuli, confusion seems to be multi-dimensional. For example, an fMRI study about confusion over Coke and Pepsi (alphabetically ordered) shows that in blinded and non-blinded taste tests, the brain uses different areas.³⁷ Researchers tracked which area of the brain is activated when respondents showed taste preferences, with or without a "brand cue."³⁸ However, Coke and Pepsi are chemically identical, and in blinded tests without any brand cue, Pepsi was more favored; in non-blinded tests, where each brand visually exudes distinct cultural connotations, preference shifted to Coke (Fig.13).

[Fig.13] Results of Coke/Pepsi preference tests (McClure et al., 2004)³⁹

     Researchers also reported different neural responses for each test. Without a brand cue, taste as sole sensory information consistently had neural responses in the ventromedial prefrontal cortex (VMPFC).  VMPFC is "strongly implicated in signaling basic appetitive aspects of reward." When presented with a brand cue, the activated brain areas were the hippocampus, Dorso-Lateral Pre Frontal Cortex (DLPFC) and midbrain. The DLPFC and hippocampus are both implicated in emotional and affective behavior modification (Fig.14).⁴⁰ In other words, culturally-loaded messages shape perception; with proper visual stimuli, humans modify behavioral preferences only to associate with liquefied sugar. The brain has two separate neural pathways for food preference: one is governed by sensory information; the other by cultural information.⁴¹ When the incoming information is loaded with cultural aspects, the two pathways in the brain interact and end up with different preference.

[Fig.14] Comparison of brain images with or without brand cues (McClure et al., 2004)⁴²

     For purposes of trade dress, this study’s ramifications are two-fold. First, the current, broad definition of trade dress could be an accurate understanding of how the brain works. Confusion is an integral experience, a combination of the sensory information and the cultural connotations. Consumer confusion research that solely looks at sensory information is not a correct reflection of how the brain perceives the world full of multiple stimuli, even cultural ones. We noted above that courts in general define trade dress very broadly. That being said, second, we may need to avoid lumping all the stimuli into an umbrella concept of "trade dress" and examining them under the same standards. The law may separately treat sources of confusion in accordance with their types. As we have seen above, when the incoming stimuli are associated with food, the brain has two separate pathways.

C. But The Picture Seems Too Rosy: Limitations of Neurological Evidence

     Even if the picture looks very enticing, we should not get carried away. Neurological evidence has certain limitations, including admissibility issues. One should not forget that the lab is a decontextualized space.  By focusing on specific sets of variables and their interactions, a lab research can decrease noise and produce meaningful results. Yet, by the same token, lab research is deprived of real-life cues that will lead to actual purchase or non-purchase. The legal disclaimers and goals of the study given to the interviewee may lead to a response that the interviewee would not take in a purchasing context. At least one study showed that interviewees’ description of recent experience may change if interviewers specify the goal of the interview. "The different reasons for asking [the interviewee] to share this experience constituted different goals as well as different stimuli."⁴³

     Moreover, fMRI tests cannot measure reaffirmation effects.  In real life, the point of confusion is hard to determine.  Confusion may arise from lack of familiarity, and repeated exposure to the similar color or design of a newly introduced product can "reaffirm" the identity of the pre-existing product.  The advertising campaign for a tattoo parlor named "Hyatt" was found to increase the value of the Hyatt hotel chain.⁴⁴ Since an fMRI test observes a single moment in time, the neurological evidence may not capture the complexity of confusion, denial and reaffirmation of a trade dress.

     Stronger challenges come from neuroscientists themselves. One common caution is the use of reverse inference. The proper inference from an fMRI research data set is that, based on the regional activation in the brain during a particular task, an area appears to have a role in the brain function associated with the task. That being said, researchers often reverse-infer that a specific brain function arises from or occurs in a pre-identified area. For instance, a group of researchers claimed that a specific candidate may signal anxiety to viewers because watching the candidate activated the amygdala, the area known to have a role in anxiety.⁴⁵ Such a reverse-inference is not scientifically grounded because the amygdal is also known to be implicated in learning. Thus, interpretation of a specific fMRI research must be balanced with prior findings and theories with regard to the brain areas in question. Lastly, balancing out fMRI’s powerful spatial resolution, fMRI has many weaknesses in constructing a research design. fMRI research must have a lower number of stimuli and manipulations. For each fMRI research subject, stimuli must be repeated multiple times to obtain enough data. fMRI is not suitable for complex research with multiple stimuli and tasks. fMRI research is also temporally restricted. fMRI can only detect stimuli responses separated by one second. If the client desires to have research results down to milliseconds, which are the norm in marketing, fMRI may be complemented with another neuroimaging method such as EEG.⁴⁶

Footnote Sources:

²¹ Indianapolis Colts, Inc. v. Metropolitan Baltimore Football Club Ltd. Partnership, 34 F.3d 410, 416 (7th Cir. 2009).

²² The Seventh Circuit held that the district court did not commit clear error in finding that the use of "Baltimore CFL Colts" by the Canadian Football League's Baltimore team was likely to cause substantial consumer confusion withregard to "the Indiana Colts" of the National Football League. The plaintiff claimed that 64% of "self-identified football fans . . . thought that the ‘Baltimore CFL Colts’ was either the old Baltimore Colts or the Indianapolis Colts." The plaintiff’s survey data also showed that there was a higher confusion rate of "Baltimore CFL Colts" than "Baltimore Horses."

²³ Tushnet, Gone in Sixty Milliseconds: Trademark Law and Cognitive Science, 86 Tex. L. Rev. 507-68, 507 (2008).

²⁴ Tushnet, supra at 545 (quoting Gladwell, Blink: The Power of Thinking without Thinking, 181 (2005)).

²⁵ For methodological advantages, see Reimann et al., Functional Magnetic Resonance Imaging in Consumer Research: A Review and Application. Psychology and Marketing, Vol. 28(6): 608-637 (2011).

²⁶ U.S. v. Semrau, 693 F.3d 510, 517 (6th Cir. 2012).

²⁷ When a particular brain area is activated, more blood with oxygen is driven to the area, and the blood oxygenation level increases accordingly.  Hence, the BOLD signal is strongly correlated with brain activity. Reimann et al., supra at 614-15.

²⁸ Reimann et al., supra at 611-12.

²⁹ Semrau, 693 F.3d at 521-22.

³⁰ Kittay, Admissibility of fMRI Lie Detection, 72 Brook. L. Rev. 1351, 1376 (2007).

³¹ Wilson v. Corestaff Services L.P., 28 Misc.3d 425, 429 (N.Y. Sup. Ct. 2010).

³² Falk et al., From Neural Responses to Population Behavior, Psychological Science (April 2012).

³³ The sample size was 31 (16 males and 15 females, all right-handed, but ethnically and socioeconomically diverse).

³⁴ The call volume was gathered before and after the ads aired.

³⁵ It was also suggested that such behavioral change may not need consciousness awareness because similar regions of MPFC are implicated in valuation and affective judgment, independent of consciousness awareness.

³⁶ Concededly, there is not much research data on the effect of consumer brain activity on subsequent behavior. Tushnet suspects that one reason for such dearth of data is because neuromaketing projects are sponsored by large companies and performed in private. Tushnet, supra at 509.

³⁷ McClure et al., Neural Correlates of Behavioral Preference for Culturally Familiar Drinks, Neuron, Vol. 44(2), 379-387 (2004).  In the research, non-carbonated Coke and Pepsi were used for technical reasons (carbonation build-up in the test tubes).

³⁸ The total number of subjects was 67, separated into four groups.  Each group had a taste test outside the MRI scanner and a drink-delivery-paradigm test inside the MRI scanner. Group 1 and 2 had anonymous taste tests and anonymou drink delivery.  Group 3 had a semi-anonymous taste test and drink delivery for Coke. Group 4 had a semi-anonymous taste test and drink delivery for Pepsi. Two types of preferences were used: "stated preferences" (answers to “Which drink do you prefer to consume: Coke, Pepsi, or no preference?”) and "behavioral preferences" (choices of labeled or unlabeled cups containing Coke/Pepsi).

³⁹"Semi-Anonymous" tests in B and C are when brand information is provided to respondents. E is another test exclusively done with respondents in C. When participants in the C test did another anonymous test without brand information, the distribution of preference shifted back to the anonymous test in A.  McClure et al., supra at 383.

⁴⁰ Notably, the DLPFC and the hippocampus belong to a group of the brain areas implicated in "biasing behavior based on affect."  Other areas are the amygdala, ventral striatum, anterior cingulate cortex, and posterior cingulate cortex. These areas are reported in experiments when research subjects make decisions and/or judgments.

⁴¹ This point was raised by a blogger who self-identified as “third year grad student in Neuroscience.” http://corticalconfusion.blogspot.com/2006/06/coke-or-pepsi.html.

⁴² The caption to the “Pepsi Delivery” image means that the active brain areas were the same as the “Coke Delivery” image.

⁴³ Tushnet, supra at 545 (citing Zaltman, How Customers Think: Essential Insights into the Mind of the Market, 171 (2003)).

⁴⁴ Tushnet, supra at 538 (citing Steckel et al., Dilution Through the Looking Glass: A Marketing Look at the Trademark Dilution Revision Act of 2005, 96 Trademark Rep. 616, 635 (2006)).

⁴⁵ Reimann et al., supra at 612.

⁴⁶ EEG (Electroencephalography) can detect millisecond differences.  Reimann et al., supra at 613.