Leo the Lion has been the most regular star of MGM Pictures since it was founded on this day in 1924, and his roar is probably the sound most commonly associated with the studio.

–Kat Eschner, The Story of Hollywood’s Most Famous Lion (2017)


There’s no roar quite like a Nittany Lion’s!

–Penn State Football @PennStateFball on 12/7/2020

So, there you go— Metro Goldwyn Mayer and Pennsylvania State University are two different entities each associated with a distinctive roar connected to their institution.  They are marketed through different, yet related, channels of trade (sports and entertainment, which were melded together as ESPN’s original name). The MGM roar and the Nittany Lion roar are certainly not identical, but they might be considered confusingly similar, should one ever need to distinguish or compare them. (I am unaware of any reported comparison, but I did find the latter compared to a toilet flushing and have seen the former’s trademark suffer indignities at times too).  I start with MGM and Penn State, not because MGM’s registered trademark and Penn State’s common law one are at odds, but because they are two well-known acoustic, or sound (or auditory), marks that can help us understand those sounding off recently about such non-traditional trademarks.  That noise includes this one, which is “the sound made by a drinks can being opened, followed by a silence of approximately one second and a fizzing sound lasting approximately nine seconds,” according to Ardagh Metal Beverage Holdings Gmbh & Co. Kg v. European Union Intellectual Property Office, T-668/19.  The uproar also includes numerous articles by diverse scholars and various bloggers expressing a renewed interest in sound marks in many countries.

The object then is to hear out all sides, and offer some thoughts on the recent clamor around acoustic marks and the questions that they present. Every sound trademark seems not only to signify something but to raise a question or two as well.  Hence, the title of this piece and the discussion below.

Looking first at the Ardagh case, we note that the applicant sought to register that sound file as a trademark. Here in a nutshell is the query the Court addressed:

16      Under Article 7(1)(b) of Regulation 2017/1001, trade marks which are devoid of any distinctive character shall not be registered.

17      For a trademark to possess distinctive character for the purposes of Article 7(1)(b) of Regulation 2017/1001, it must serve to identify the product or service in respect of which registration is applied for as originating from a particular undertaking, and thus to distinguish that product or service from those of other undertakings (see judgments of 21 January 2010, Audi OHIM, C‑398/08 P, EU:C:2010:29, paragraph 33 and the case-law cited, and of 20 October 2011, Freixenet v OHIM, C‑344/10 P and C‑345/10 P, EU:C:2011:680, paragraph 42 and the case-law cited).

18      Signs devoid of any distinctive character, referred to in Article 7(1)(b) of Regulation 2017/1001, are deemed incapable of fulfilling the essential function of the trade mark, namely to identify the origin of the product or service, so as to enable the consumer who purchases the product or service which the trade mark designates, when making a subsequent purchase, to make the same choice if the experience has been positive or to make another choice if it has been negative (see, to that effect, judgment of 16 September 2004, SAT.1 v OHIM, C‑329/02 P, EU:C:2004:532, paragraph 23).

19      The distinctive character of a trade mark must be assessed by reference, first, to the goods or services in respect of which registration is sought and, second, to the perception of the relevant public (see judgment of 12 February 2004, Henkel, C‑218/01, EU:C:2004:88, paragraph 50 and the case-law cited). The level of attention of the average consumer, deemed to be reasonably well-informed and reasonably observant and circumspect, is likely to vary according to the category of goods or services in question (judgment of 10 October 2007, Bang & Olufsen v OHIM (Shape of a loudspeaker), T‑460/05, EU:T:2007:304, paragraph 32).

20      In the present case, the trade mark applied for is the sound sign corresponding to the sequence of the sound of a drinks can being opened, a silence of approximately one second and a fizzing sound of approximately nine seconds.

Applying those standards, the Ardargh Court rejected that registration, holding that such a sound patter “cannot be registered as a trade mark in respect of various drinks and for metal containers for storage or transport, in so far as it is not distinctive.” General Court of the European Union PRESS RELEASE No 120/21 Luxembourg, 7 July 2021 entitled “The Court gives a ruling for the first time on the registration of a sound mark submitted in audio format.”; see also ILN Today, July 15, 2021.  The Court also noted that sound “to be a purely technical and functional element,” PRESS RELEASE No 120/21 Luxembourg, 7 July 2021, see also Ardagh at par. 36-41.

But the court did say that “it should be noted that the criteria for assessing distinctiveness are the same for all categories of trade mark, and that [applicable law] draws no distinction between the various categories. The criteria for assessing the distinctiveness of sound marks are therefore no different from those applicable to the other categories of mark.” Ardagh at par. 23.  Though reemphasizing that “case-law does not establish new criteria for assessing the distinctive character of a trade mark,” the Court did concede “that, in the context of the application of those criteria, the perception of the relevant public is likely to be influenced by the nature of the sign for which registration is sought.” Ardagh at par. 31.  Some wonder what that all means.

Ardagh itself seems to suggest that it means sound can be a trademark under the same rules as other trademarks, but that the unique issues endemic to sound trademarks will nonetheless influence any final decisions on registration and enforceability. The Ardagh analysis echoes the analysis found in American courts and registration controversies, such as in the case of Ride The Ducks, LLC v. Duck Boat Tours, Inc., where a federal district court noted that “a sound mark’s level of protection will depend on whether it is inherently distinctive or more common-place and nondistinctive.”  As the court noted, in order to register a commonplace sound, “a proponent must ‘show that purchasers, prospective purchasers and listeners do recognize and associate the sound with services offered . . . exclusively with a single, albeit anonymous, source.’” (quoting In re General Electric Broadcasting Company, Inc., 199 U.S.P.Q. 560, 563 (Tr. Tr. & App. Bd., Apr. 12, 1978)).  That, essentially, sounds like the Ardagh approach.

Certainly the reaction to Harley-Davidson’s efforts to register a sound mark prompted a reaction like that seen in Ardagh:

Harley-Davidson was famously unsuccessful in attempting to trademark the sound produced by its V-Twin engines. In 1994 Harley-Davidson sought to register the “exhaust sound of applicant’s motorcycles, produced by V-Twin, common crankpin motorcycle engines when the goods are in use” (Trademark Serial No 74485223). Several competitors opposed the application, arguing that the sound was not unique (eg, compared to other engines, see Michael Saperstein ‘The Trademark Registrability of the Harley-Davidson Roar: A Multimedia Analysis’, Boston College Intellectual Property & Technology Forum, (1998) at page 3). Harley-Davidson also faced potential challenges that the V-Twin engine sound is functional and that, unlike many other trademarked sounds, the V-Twin sound is not always the same. Rather, it varies based on which motorcycle model it is used in, and also how much the throttle is revved (Id pages 4 and 5). Eventually, Harley-Davidson abandoned the application six years later (see Trademark Serial No 74485223).

[G Warren Bleeker and Mike Koplow, Can you trademark sounds, smells, colours, motions and flavours? (2019)

The sound of the engine, like that of the pull-top, just did not impress the examiners and others as distinctive.

Though the Duck court thought that “the trademarking of a particular noise or sound appears to be a relatively rare phenomenon,” it really is not as rare, at least in the United States, as one might think.  As one commentator noted, “[t]he USPTO’s regulatory guidelines for the examination and registration of sound marks are set forth in Section 1202.15 of the Trademark Manual of Examining Procedure,” and:

Specifically, the TMEP states that, in order to function as a source indicator, a sound mark must ‘assume a definitive shape or arrangement’ and ‘create in the hearer’s mind an association of the sound’ with the underlying goods or services. The TMEP lists ‘a series of tones or musical notes’ and ‘wording accompanied by music’ as two examples of potentially registrable forms of sound marks. This same TMEP provision establishes that, like other types of trademarks and service marks, sound marks must be ‘arbitrary, unique or distinctive’ in connection with the goods or services with which they are associated. The TMEP then elaborates that ‘commonplace’ sounds, such as sounds which are frequently heard in the ‘normal course of operation’ of products or to which listeners have been exposed under different circumstances, can only be registered as sound marks upon a showing of ‘acquired distinctiveness’.

[John S. Miranda, Behind the Jingles: Legal Strategies to Protect Sound Trademarks (2021)

Because a number of registrants have met those criteria and more are interested in doing so, the USPTO itself publishes a list of trademark sound mark examples that is over 100 marks long, most with linked audio files. (Other lists exist as well, for example here, here, and here).

In addition to MGM’s Roaring Lion, the USPTO’s list includes Homer Simpson’s short D’Oh! (an exclamation I did not find on this literary inventory), the earlier mentioned ESPN’s well-known, slightly longer  DaDaDa, DaDaDa, the Green Giant’s Ho-Ho-Ho, the minute-long Lone Ranger intro featuring a piece of Rossini’s Willian Tell Overture, and the five-minute-thirty-nine-second Sweet Georgia Brown song as a mark for the Harlem Globe Trotters.  (While the lion’s roar, Homer’s exclamation, and ESPN’s six simple notes sound like trademarks to me, I am dreading the Christmas dinner infringement suit between the Green Giant and Santa Claus, and wondering whether Lone Ranger and Globetrotter length trademarks will have us on a Stairway To Heaven type descent into a fair-use type analysis).  In the US, other sound marks include Law & Order’s ca-chung chung (as Reg. No. 3137680), the lightsaber sound from Star Wars (as Reg. No. 3618321), and “[t]he NBC chimes sequence[, which] actually became the first ‘purely audible’ service mark to register with the USPTO in 1950.” as that early quoted commentator also noted. (As noted (at 498, fn. 44) of Iyanu Lipede’s  “Let’s Hear it For Sound Marks: Trademark Protection in Copyrighted Songs,” 16 Liberty University Law Review 493 (2022), “[t]he mark comprises the musical notes G, E, C played on chimes, Registration No. 523,616. NBC did not renew this mark, and the mark expired in 1992. NBC filed a new trademark application for the same sound in 1970, and the new mark was registered the following year. The mark comprises a sequence of chime-like musical notes which are in the key of C and sound the noted G, E, C, the ‘G’ being the one just below middle C, the ‘E’ the one just above middle C, and the ‘C’ being middle C, thereby to identify applicant’s broadcasting service, Registration No. 916,522”)  Thus, sound marks are not as new a phenomenon in the US as they are in Europe, though they remain somewhat rare even in the States.

Turning to matters outside the US, we observe that Simon Geiregat recently noted, in Trade Marks in Sounds and Gestures: A Critical Analysis of Two Non-Traditional Signs in the EU (2022), that sound marks “seem to be entirely accepted in the European trademark system.” But that is more in the sense of philosophical acknowledgment of the possibility than practical acceptance of very large numbers of sound marks.  This may be, as Geiregat himself notes, because hurdles for sound remarks remain, from case-by-case challenges to distinctiveness to the difficulty of searching sound marks on various registers to concerns of “overprotection, and even in depletion of sounds available for registration.” Others share these concerns as to sound and other non-conventional marks, noting that “[n]on-conventional trademarks, despite performing the basic purpose of a trademark, have proven to be a challenging procedure in the nation [of India], owing to the uniqueness requirement and the inability to be graphically depicted. There is also the possibility that these marks will cause consumer confusion, negating the purpose of trademarks.” Shaista Kahkeshani and Mohd Juned Ansari, Non-Conventional Trademarks: The Spectrum of Distinctiveness in the Era of Globalization, 5 International Journal of Law Management & Humanities 1117, 1122 (2022)

There are also concerns that reverberate to some of the concerns expressed in copyright and related cases like those involving the Blurred Lines controversies over use of limited musical notes.  Unsurprisingly, one recent article goes directly to this question, and proposes  layering trademark protections atop copyrighted song rights and calling for both the clarifying of Supreme Court precedent and the amending of the Lanham Act so that, in the author’s words (at 527), “plaintiffs may bring concurrent claims arising under the Lanham Act and Copyright Act when overlapping protection of the subject matter is available” and that “[m]usician-sound-mark-owners would be able to bring trademark claims against copiers regardless of whether an adequate remedy lies in the Copyright Act.”

In fact, sound marks have been the focus of many in the EU.

For instance, though  Giereget observed that sound marks (as well as non-traditional gesture marks) “have rarely been comprehensively assessed in recent literature,” a simple Google Scholar search for sound mark articles in 2022 suggests that is rapidly changing (Though the title of my post here is a paraphrase to the famous lines concerning “sound and fury, signifying nothing” in the Scottish play, I’ll leave for another day an examination of gesture marks.  Just as Shakespeare noted in Winter’s Tale (Act 5, Scene 2, lines 14-15), there is such a full body of “language in their very gestures” that doing that topic justice goes beyond the space available in this piece).

Likewise, sound marks have become more popular in the EU since the European Union Trade Mark Implementing Regulation (EUTMIR) was amended in October 2017 to allow the submission of audio files rather than requiring musical notation descriptions.  Though there are only 333 EU registered sound marks, according to EUIPO January 1996 to July 2022 statistics (section 5.2 at page 36 of 53), sixty-four of them have registered in the last three years, thus roughly 20+ per year (That compares to about 11.6 per year from 1996 to 2019).   The EUTMIR builds on earlier European recognitions of sound marks, such as in the case of Shield Mark v. Joost Kist Memex (2003), par. 27, where the Court held, in a Netherlands case, that “all signs capable of serving to distinguish an undertaking’s products or services from those of other undertakings may, in principle, serve as trade marks. It follows, in its submission, that since sound signs are clearly capable of doing so they may fulfil the role of a trade mark.” Id.; see also par. 34-41

Outside of the EU and the US, sound marks have also been recognized—and debated.

For example, the first sound mark in India was registered for Yahoo in 2008, and an Indian bank later received a sound even though Indian law at the time did not expressly specify that sound marks were allowed.  As one Indian lawyer noted, “At a later stage, India adopted the provisions of the EUIPO (European Union Intellectual Property Office) and came up with the Trademark Rules of 2017. Section 26 of those Rules stipulates how a sound mark should be registered as a trademark. Although sounds were never prohibited by law from obtaining trademark protection before the introduction of the Trademark Rules, 2017, there was no prescribed formula in which a sound mark could be registered.”  Now, “a sound that can be graphically represented by a succession of musical notes with or without words” can be registered, so long as it meets registration and is “submitted in an MP3 format no longer than 30 seconds in length.” Id.  (That time limitation does not exist in the EU and the US by contrast as another Indian lawyer noted, and as reflected in the comments above Sweet Georgia Brown, at almost six minutes in length, was almost 12 times longer than would be allowed in India)  As of April 2022, there have still been no contested sound mark cases in India, suggesting that “the trend for sound marks is still in its infancy in India.”

Likewise, sound mark applications have been acceptable in Thailand since late 2017, as noted here.  This was made possible through amendments to the Trademark Act B.E. 2534 (1991) and the issuance of Ministerial Regulation No. 5 B.E. 2560 (2017).  As Thai counsel has noted, a sound mark seeking Thai registration must be one having “distinctiveness, not being prohibited by law, and not being identical or confusingly similar to a registered or pending-for-examination mark.”  But it also must not be (i)  “a sound that is directly descriptive of the characteristics of the relevant goods or services,” such as mooing cows for milk; (ii) “ natural sound associated with the relevant goods or services,” such the sound of cutting something with a metal tool when used in relation to knives, or (iii) “a sound produced by the operation of specific goods or services,” such as a can opening in relation to cans or the drinks in them. Id.  The registration application should also provide “the closest Thai syllabic expression of the melody, if possible,” along with “the sound’s description as a musical notation, a sonograph, a spectrogram, or other forms of transcription to illustrate the proposed sound,” an mp3 audio file and statement of the type of sound mark (whether it is a human sound, animal sound, music/ melodic sound, or other type of sound) and the context in which the mark will be used.

Other countries also allow for registration of sound marks or have those interested in the issue promoting such allowance.  For instance, one commentator noted that, in addition to the US, Thailand, the EU, and India, the following allow sound marks:

Australia—As the Australian Trade Marks Manual of Practice and Procedure itself states in Section 21.6.2, ‘The 1995 Act allows for the registration of a sound as a trade mark.  As with any other kind of trade mark, a sound must first meet the basic requirements for registrability set out in the Act.’

China—’As early as May 1 2014, China’s revised trademark law cancelled the visibility requirements for trademark registration, and for the first time, evolved to incorporate sound into the scope of trademark registration’ and ‘[a]ccording to statistics, by the end of January 2021, 928 sound trademarks have been formally filed before the China Trademark Office,’ according to attorney Gang Hu.

UK—’In the UK, sounds that are protected by registered trade marks include the jingle used by Direct Line Insurance, Bach’s Air on a G String, trade marked by Hamlet Cigars, and Delibes’ Flower Duet from Lakmé, trade marked by British Airways,’ according to one source.

Canada—Since the 2012 Federal Court decision in Metro-Goldwyn-Mayer v Attorney General Of Canada ( T-1650-10),  sound marks have been registrable in Canada, as noted here; and

Japan—As Masaki Mikami noted, ‘On September 26, 2017, the Japan Patent Office (JPO) announced to grant protection to three sound trademarks solely consisting of a sound element for the very first time.’ They issued to Intel, BMW and Taiko Pharmaceutical Co., Ltd. Id.

Interestingly, Elnur Karimov, in The Protection of Non-Traditional Signs in the Republic of Azerbaijan: International Instruments as a Road to Registration? (2002)(at 78), has written that, while current Azerbaijani domestic trademark law and processes may not support sound and other non-traditional trademark applications, the Patent and Trademark Examination Center of the Republic of Azerbaijan should cite paragraph A (1) of Article 6 of the Paris Convention in an “attempt to protect these signs” under Azerbaijani law as ones submitted through the international registration route under the Madrid Protocol.  (Article 6 states in relevant part that “every trademark duly registered in the country of origin shall be accepted for filing and protected as is in the other countries of the Union, subject to the reservations indicated in this Article.”).  Thus, in addition to the domestic law of various countries, treaties may provide a path to the registration of sound marks.

This article, then, is really just a recognition that the crack of the starter’s pistol has been heard, and sound marks are off to the races.  Of course, sounds, like symbols and words, can be confused for one another, so challenges remain even if we sort through the legal issues.  But, as Nick March has said,  there is no turning back as “the power of sound to build emotional connections between consumers to brands has become evident… The equity of an instantly recognisable – ideally catchy – sound that creates subconscious connections with a brand is immense. An aural identity can be used in myriad ways across TV, radio and internet media… While sound marks are still quite rare, we are likely to see – and hear – much more of them as the future unfolds.”

Listen up, as they say.