3.1 Introduction

The creative endeavours of fashion designers have long been celebrated by consumers, journalists and even museums. But despite the widespread recognition of fashion design’s artistry, it has always had an uncertain status within intellectual property law. However beautiful a garment may be, it is also functional. It serves to clothe the body, to keep it warm and to enable the wearer to perform particular activities while wearing it.

Copyright, trademark and design patent laws in the United States are deeply anxious about protecting functionality. They each use various doctrinal devices to block functional content, and they channel that content instead towards the utility patent regime. It should be noted that in Commonwealth common law jurisdictions, the terminology is different. ‘Design patents’, that protect ornamental designs are known as ‘registered designs’ or simply ‘designs’.Footnote ¹

This chapter explains how garments function and how different intellectual property regimes should deal with that functionality. Garments are not simply functional to the extent that they serve the mechanical or technological ends of keeping people properly clothed. They are also functional to the extent that they influence the way the wearer is perceived. When aspects of a design affect the perception of the wearer, for example, by making them look taller, slimmer, broader or curvier, those aspects of the design must be treated as functional.Footnote ²

The US Supreme Court’s opinion in Star Athletica, L.L.C. v Varsity Brands, Inc.Footnote ³ on the copyrightability of cheerleading uniform designs did not recognise this important aspect of the nature of garment design. Because the Supreme Court did not accurately assess the functionality of the cheerleading uniform designs before it, the Court risks conferring a substantial competitive advantage on the designer-plaintiff, one that Congress did not intend to conferFootnote ⁴ and that the designer did not earn. As we explain here, the contested features of these uniform designs – stripes, chevrons, zigzags and colour-blocking – like many features of fashion design, affect the perception of the wearer by drawing attention to certain aspects of the wearer’s body over others and by making the wearer seem to be curvier and have a longer torso.

The account of fashion’s functionality that we develop here also has important implications for trademark law and design patent law. In this chapter, we explain how trademark and design patent laws can more successfully respond to attempts by claimants to protect functional aspects of garment design without satisfying utility patent law’s more strenuous demands.

Section 3.1 describes our account of fashion’s function. We explain how, in addition to technical or mechanical functions, fashion design is also functional when it incorporates features that influence the way the wearer is perceived. Section 3.2 explains how copyright law should treat the functional aspects of fashion design in light of copyright law’s functionality screening for useful articles, like garment designs, and the law’s originality and merger doctrines. In Section 3.3, we illustrate our approach with a case study of the cheerleading uniforms at issue in Star Athletica and demonstrate how the Court went wrong. Sections 3.4 and 3.5, respectively, show how our account of fashion’s function should affect trademark and design patent laws.

3.2 How Fashion Design Is Functional

Copyright, trademark and design patent laws each limit protection for functional aspects of designs. They each do so in somewhat different ways, as we address. Yet these three legal regimes are unified in limiting functional matter as a way to ensure that those design aspects are protected through utility patent law, if at all. Because of their functionality screens, copyright, trademark and design patent laws each need to identify the functional aspects of otherwise protectable designs.Footnote ⁵

The features of product designs fall into three mutually exclusive categories. Features can be (1) purely functional, (2) purely non-functional or (3) hybrid. Consider how these features might appear in different parts of an automobile. The car’s spark plugs are rarely seen, and their design may be motivated entirely by utilitarian concerns about size and performance. They would fall into the first category. The sculpture of a leaping jaguar that serves as the car’s hood ornament has no functional contribution to the car’s operation; it exists merely to look good. It falls into the second, purely non-functional, category. Many other aspects of the car’s design are simultaneously functional and serve some additional non-functional purpose. The shape of the car’s hood both makes it look attractive and influences its aerodynamics. The car’s wheels may be designed to look speedy, but they must also be able to support the car’s weight. These and other features of the car’s design are hybrid.

Knowing what to treat as functional for purposes of garment design requires understanding how clothing can serve utilitarian ends. In some cases this is fairly simple, while in other cases it is more complex. Decisions about the physical or technological performance of a garment should clearly be treated as functional. For example, a designer’s choices to use fabric that provides warmth even in the coldest temperature or fabric that wicks sweat away from the wearer’s body are functional.Footnote ⁶

On the other hand, various aspects of garment design could be purely expressive, or aesthetic. A logo for a sports team emblazoned on a t-shirt is one such example. Similarly, a paisley pattern woven into a necktie is probably a purely expressive element of the tie’s design. Two-dimensional pictorial or graphic additions to garment designs are the most plausible candidates of designs for purely expressive elements.

Many of a garment’s design features will be at least partially functional and fall into the hybrid category. For example, the incorporation and design of pockets on a pair of trousers may be particularly stylish and visually appealing, but they are also useful because the pockets give the wearer a place for his or her wallet and keys. Thus, the design of pockets and other similar features should also be treated as dual-nature components of garment design. The same might be true for the cut of sleeve that provides a certain degree of warmth, modesty or ease of movement while simultaneously being beautiful to perceive. Many of the aspects of the fit and cut of a garment fall into the hybrid category, because although they may look attractive in their own right, they also affect the garment’s technical or physical performance as a piece of apparel.

Designs can importantly also function in ways that go beyond standard mechanical or technological utility. In particular, the designs of many garments are intended to and are purchased because they make their wearers look attractive. A designer’s choices about hemline, neckline, darting and, sometimes, fabric prints influence people’s perceptions of the wearer. A shopper does not simply ask of a garment, ‘Does it look good?’ In addition, and more importantly, he or she asks, ‘Does it look good on me?’ or ‘Does it make me look good?’Footnote ⁷ In this way, the garment’s design is valued for what it can do when it is being used. Thus, in addition to the mechanical functions of garment designs, another intrinsic – or essential – function garment designs tend to share is that many of their design features are incorporated to affect the perception of the wearer.Footnote ⁸

Many features of garment design – line, shape, texture, colour and print – exploit features of human visual perception and optical illusions to influence the way in which the wearer’s body is perceived.Footnote ⁹ Importantly, these visual effects can be created with both three-dimensional design techniques such as garment shape and cut as well as with two-dimensional design techniques such as patterns, stripes and colour. Thus, the frequent assertion that two-dimensional fabric design is non-functionalFootnote ¹⁰ is in many cases wrong when the design affects how the wearer is perceived. Below, we illustrate a number of these techniques, ranging from camouflage to stripe placement.

As all shoppers know, a certain style or cut of a garment may look good on one person but not on another.Footnote ¹¹ And different sorts of designs may affect how people’s bodies look. This is not accidental. Design choices create visual effects that can vary the size or shape of the wearer’s body or body parts. In this section, we illustrate a number of the most prominent design techniques. Our goal here is not to catalogue garment design techniques exhaustively but rather, by choosing common examples, to illustrate the nature of design functionality.

It is common knowledge that the placement of horizontal lines on a garment can make the wearer look shorter and broader, while vertical lines have the opposite effect of lengthening and slimming the wearer’s body.Footnote ¹² Garment design can visually minimise the size of body parts judged to be too large by either appearing to subdivide those body parts or by counterbalancing them to increase the size of surrounding parts.Footnote ¹³ For example, a V-shaped neckline can make the neck look longer and narrower while also making shoulder width narrower.Footnote ¹⁴ A bodice can be smocked, pleated, draped or gathered at the bust to make the bust appear larger.Footnote ¹⁵ Similarly, as one design textbook indicates, ‘a line continuing around the body, diagonal stripes that are not too dominant, or a reflecting surface smoothly contouring the body call our attention to the rounded contours’.Footnote ¹⁶

Designers can also affect the perception of the shape of body parts through design choices. Importantly for our purposes, the combination of colours, shapes and lines can have enormous influence on how the wearer’s body is perceived. This is known as colour-blocking, and it was recently made famous by Stella McCartney’s line of dresses, garment designs employing a colour-blocked hourglass appearance that made wearers’ bodies look both curvier and slimmer. Two examples of these dresses, as worn by actress Kate Winslet, are shown in Figure 3.1. The black colour along the sides, which changes in width at various points, produces a particularly striking hourglass shape.Footnote ¹⁷ And, in the dress on the right, the brighter colour on the top of the dress produces a larger and more defined bust.Footnote ¹⁸

Figure 3.1

Stella McCartney colour-blocked dresses.Footnote ¹⁹

(a) © ABACA/Press Association Images (b) © Action Press / Rex Features

Optical illusions can also be deployed in garment design to influence perceptions of the wearer’s body.Footnote ²⁰ Consider the Müller-Lyer illusion, as shown in Figure 3.2, which causes a line to look longer if it is bracketed on each end by arrow tails and shorter if it is bracketed on each end by arrow heads.Footnote ²¹ This illusion can be incorporated into garment designs to lengthen or contract the body of the wearer through placement of arrow tails or heads, respectively, as shown in Figure 3.3.Footnote ²²

Figure 3.2

Müller-Lyer illusion.Footnote ²³

© PsyPhi > SciFi

Figure 3.3

The Müller-Lyer illusion applied to garment design.Footnote ²⁴

Figure 3.3 Long description

The figure on the left has a less structured bodice shape, while the figure on the right has an almost corset-like, shape around the torso. The torso on the left has an arrow with V-shaped heads, while the torso on the right has an arrow with arrowheads. Both figures have elongated arms and hands on their sides.

© International Journal of Science, Culture and Sports

These design techniques are not used just because they look attractive on their own. They are used to make the wearer look more attractive. They are part of what make designs inherently work as fashion designs. By manipulating features of human vision, these design techniques alter the way we see things in the world. As two designers explain, ‘By using art components in this way, [designers] can alter the frame of reference in which we see the human form, and in so doing, [they] can create illusions or effects that would not be possible in any other way.’Footnote ²⁵ In this sense, they are the same as camouflage or high-visibility patterns that serve a particular function. Unlike a painting or a drawing of a cartoon character, they do not exist merely to portray their own appearance or convey information.

As we argue in Section 3.3, the Copyright Act treats as functional aspects of visual design that influence the perception of another object, including the human body. Accordingly, as a matter of copyright statute and policy, the incorporation of design techniques that produce dependent value are unprotectable. Similarly, trademark and design patent laws must grapple with the issue of whether such functional aspects of fashion designs are unprotectable. We now address these laws in turn, beginning with copyright law.

3.3 Fashion’s Function in Copyright Law

Congress and the US courts have long expressed concern about the misuse of copyright law to protect functional objects from market competition.Footnote ²⁶ The fashion industry, in particular, has proven a consistent source of consternation in this regard.Footnote ²⁷ Over a century ago, one court explained that ‘a manufacturer of unpatented articles cannot practically monopolise their sale by copyrighting a catalogue containing illustrations of them’.Footnote ²⁸ As the Star Athletica case decided by the US Supreme Court indicates, many in the fashion industry are still seeking haven for their designs within copyright law. In this section, we review the general structure of the copyright-patent divide in intellectual property law and the doctrinal techniques that copyright law has historically used to screen out functional content. We then set out how copyright law and policy lead to the conclusion that fashion design features that affect the perception of the wearer are not protectable.

3.3.3 The Functionality of Garment Design Features

Copyright law’s useful articles doctrine is intended to handle cases involving works that incorporate both expression and function. Applying this doctrine to fashion designs has proven challenging to courts, as they try to sort and separate the aspects of garments that are expressive from those that are functional. In this section, we offer a robust approach to understanding the nature of fashion’s functionality. Our approach is grounded in the current copyright statute and responds to the Supreme Court’s Star Athletica decision. Correctly applied, it appropriately limits the scope of copyright protection for garment design to only those features that are purely expressive. According to our approach, components of a garment design should be treated as functional not only if they are mechanically or technologically functional but also if they are valued for their ability to influence the way that the wearer is perceived. These components are prevalent in fashion design and, even if they are also expressive, must be screened out of any copyright protection granted to an author.

It is clear that fashion designs are both ‘pictorial, graphic, [or] sculptural works’ and ‘useful articles’, so copyright law’s separability criteria apply to them.Footnote ⁵⁸ As noted above, the most important step in this analysis involves the identification and categorisation of the components of a claimed design. A court must determine which components of the design are purely expressive, which are purely functional and which are hybrid – that is, that they simultaneously affect the aesthetics and the utility of the garment.

We think that the Copyright Act treats aspects of garment design that influence the perception of the wearer’s body as ‘utilitarian features’ of a useful article, although they may also be expressive ‘pictorial, graphic, or sculptural’ aspects of the article. The reason is that the copyright statute defines a useful article as one having ‘an intrinsic utilitarian function that is not merely to portray the appearance of the article or to convey information’.Footnote ⁵⁹ As discussed in Section 3.3.2, the Act distinguishes between useful article features that ‘merely portray the appearance of the article or … convey information’ and useful article features that have additional utility. By this distinction, an aspect of a useful article that ‘merely … portray[s] [its] appearance … or … convey[s] information’ is not considered a utilitarian feature. It is purely an expression of authorship. By contrast, an aspect of a useful article that does not merely portray its appearance or convey information is a utilitarian feature of that article.

In the context of the useful articles doctrine, the Copyright Act establishes a distinction between aspects of designs that are valued inherently for their depictive effect and those whose value is dependent on their effect on other objects. The former it treats as expressive ‘pictorial, graphic, or sculptural features’, and the latter it treats as ‘utilitarian aspects’.Footnote ⁶⁰ Inherently valuable design elements alone may be treated as purely expressive features of a work. Elements that are dependently valuable for their ability to affect the perception of another object are (at least partly) functional, because they do not exist merely to portray their own appearance or convey information. For garments, this dependent visual utility represents an additional way in which aspects of garment design can be utilitarian. In addition to the mechanical utility related to use, warmth and modesty, garments have additional utility under the copyright statute when they incorporate design features that influence the wearer’s appearance.

Consider how different sorts of visual figures might be incorporated into the design of a garment. On the one hand, a drawing of a cartoon character on a T-shirt is inherently valuable. It is produced and consumed for the representation it provides. The drawing of the character plays the same role and has the same effect whether it is on a piece of paper, a digital device screen or a garment. On the other hand are useful article designs that are primarily dependently valuable. They produce value when brought into appropriate relationships with other objects and in terms of their effects on those objects: making them look bigger, smaller, different or better. Camouflage offers an ideal example of designs at this end of the spectrum, of those having dependent value. The design of a camouflage pattern achieves its value when it is placed in an appropriate environment. When it is, the pattern works: it functions to cloak the person or object that it covers.Footnote ⁶¹ Consider, by comparison, the two camouflage patterns in Figure 3.4. The camouflage pattern on the left was worn by soldiers in the US Army deployed in Afghanistan to cloak them well against their surroundings, whereas the camouflage pattern on the right, worn by North Korean soldiers, does not cloak them, but if anything, makes them easier to see.

Figure 3.4

Effective and ineffective camouflage.Footnote ⁶⁴

(a) © SpaceDaily (b) Courtesy of kotaku.com

When a design element, like an effective camouflage pattern, alters the way another object appears, that is, when it works in relation to some other object, that design element should be treated as utilitarian as a matter of copyright law.Footnote ⁶² That is, the design element is ‘not merely … portray[ing] the appearance of the article or … convey[ing] information’.Footnote ⁶³ Only if a design or an aspect of design exists merely to portray its own appearance should it be treated by copyright law as purely expressive. One such example would be an image of a character, such as Mickey Mouse, printed on a T-shirt. The character exists merely to portray its own appearance. Similarly, Pharrell Williams’ so-called ‘damoflage’ (a pixelated version of Louis Vuitton’s damier check print resembling camouflage) – shown in Figure 3.5 – does not serve to camouflage or do anything other than portray its own appearance.

Figure 3.5

Pharrell Williams wearing the ‘damoflage’ print.Footnote ⁶⁵

Credit: Giovanni Giannoni / WWD / Getty Images

In light of this discussion, we wish to respond to an argument that might be raised to challenge our reading of the Copyright Act: the claim by some courts and scholars that two-dimensional design features are always non-functional, which the Supreme Court recently rejected. It is important to note that expressive techniques that affect the perception of the wearer – and are thus also functional – include both three-dimensional design choices involving the shape and cut of garments as well as two-dimensional design choices involving shapes, colours and patterns. This is essential because some courts and scholars have often reflexively treated two- dimensional fabric designs as non-functional.Footnote ⁶⁶ Yet both two- and three- dimensional design features will often be utilitarian in nature. As discussed above, camouflage offers an obvious example of a functional two-dimensional pattern. If the developer of a new camouflage pattern desires intellectual property protection, they must look to the patent system, as some camouflage designers (including the US military itself) have done.Footnote ⁶⁷ High-visibility fabrics or patterns are also functional, even though they are two-dimensional. And the techniques discussed above in Section 3.1 use two-dimensional patterns in similar ways, by altering how the viewer sees the clothed person.Footnote ⁶⁸

As the Supreme Court emphasised in Star Athletica, the Copyright Act clearly anticipates that two-dimensional designs can be functional when it refers to ‘pictorial, graphic, or sculptural features’ that are inseparable from utilitarian aspects of the design.Footnote ⁶⁹ The first two objects in this list, pictorial and graphic features, are, in fact, two-dimensional design techniques. The statutory scheme thus makes clear that two-dimensional works and features, and not only three-dimensional works and features, can be utilitarian.Footnote ⁷⁰ By stating that ‘pictorial, graphic, or sculptural work[s]’ can be ‘useful article[s]’, rather than limiting the statutory text to ‘sculptural works’, it is apparent that Congress recognised that both two- and three-dimensional works can be utilitarian. Similarly, by indicating that ‘pictorial, graphic, or sculptural features’ might be inseparable from ‘utilitarian aspects’ of the design, rather than just referring to ‘sculptural features’, the copyright statute signifies that both two- and three-dimensional features can be utilitarian.

3.4 The Conceptual Shortcomings of Star Athletica

The account presented in Section 3.3 best comports with the Copyright Act’s useful articles doctrine as Congress drafted it and the policy goals undergirding it. In 2017, the US Supreme Court had an opportunity to affirm these principles in Star Athletica, LLC v Varsity Brands, Inc.Footnote ⁷¹ Instead, the Court charted an entirely new path through copyright law, upsetting decades of lower court precedents and altering the nature of copyright’s functionality screening. Here we attempt to explain what the Court did and how it went so wrong. We also discuss how our account of fashion’s function should still matter for copyright validity and infringement even if its role in the useful articles doctrine is now different.

The case centres on the copyrightability of the cheerleading uniform designs shown in Figure 3.6.Footnote ⁷² The Supreme Court held the arrangement of stripes, chevrons, zigzags and colour-blocking in the garment designs to be separable features of useful articles and potentially copyrightable.Footnote ⁷³ In light of our analysis above, we think the Supreme Court came to the wrong conclusion, after failing to recognise that the arrangement of stripes, chevrons, zigzags and colour-blocking in the garment designs at issue – like many components of clothing designs – has both expressive and utilitarian aspects.

Figure 3.6

Varsity’s claimed design.

Courtesy of US Supreme Court opinion and Copyright Office registration image

Varsity submitted the above designs for registration by the Copyright Office, and Varsity sued Star Athletica for producing uniforms displaying similar designs. The district court granted the defendant’s summary judgment motion, ruling that the designs were not protectable because they served the function of identifying their wearers as cheerleaders.Footnote ⁷⁴ The Sixth Circuit Court of Appeals reversed.Footnote ⁷⁵ It is worth spending a moment on the Sixth Circuit’s opinion because, even though we think it reached the wrong conclusion, it did so via a logical methodology. The appellate court engaged in the correct set of steps, first determining that the uniform designs were useful articles and then identifying their expressive and functional features.Footnote ⁷⁶ It was only in applying this step that we think the court erred. It concluded that the only features of the uniforms that were functional were those that contributed to their mechanical use as garments, such as enabling wearers to jump and flip or cover their bodies.Footnote ⁷⁷ The Sixth Circuit determined that the stripes, chevrons and colour-blocking were purely expressive features and were separable from the rest of the uniforms.Footnote ⁷⁸

The Sixth Circuit understood the basics of the useful articles doctrine and its separability analysis. Its only problem was its failure to grasp that the plaintiff’s designs are not purely expressive. Instead, the stripes, chevrons, zigzags and colour-blocking in the claimed designs serve a hybrid purpose, similar to those documented in Section 3.1. They are simultaneously both expressive and functional, serving to accentuate and elongate the cheerleader’s body and make it appear curvier in particular areas.Footnote ⁷⁹ We detail some of these hybrid features in light of the discussion in Section 3.1.

In Design 078, the white patches in the colour-blocked pattern help create the visual effect of curviness by creating an hourglass shape with contrasting colours, as in Stella McCartney’s dresses.Footnote ⁸⁰ The V-shaped neckline together with the inverted-V-shaped slit on the skirt elongates the body by exploiting the Müller-Lyer illusion. The V-shaped neckline also serves to point to the bust and elongate the neck. The colour-blocking, using contrasting colours, creates contrast and draws attention to the wearer’s body. The repeating stripe pattern at the neckline, bust, waist and skirt hem draws viewers’ attention from the top of the wearer’s neckline all the way to the bottom of the wearer’s skirt to see that part of the wearer’s body as unified.Footnote ⁸¹

Again, in Design 0815, Varsity employs colour-blocking; this time a darker blue colour along the side of the uniform to make the wearer look both more slender in certain places and curvier in others, just as in the Stella McCartney dresses. In addition, the V-shaped striping on the front centre of the uniform serves to further accentuate the bust. The V-shaped neckline both points to the bust and elongates the neck. Designs 299A and 299B have some of the same features associated with the previous two designs, plus there is a diagonal striping, which calls attention to the body’s rounded contours. Additionally, the chevron at the bottom of the uniform top is cut to display some of the belly and draw attention to it. Design 074 serves to accentuate the curves and elongate the body for many of the reasons discussed with regard to the previous designs. In addition, the colour-blocking serves to highlight the bust by colouring it in white in contrast to the green above it and navy blue below it.

The stripes, chevrons, zigzags and colour-blocking that form the heart of Varsity’s uniform designs do not merely portray their own appearance or convey information. They also serve the purpose of influencing the appearance of the uniform’s wearer. These features are the reason that the designs succeed as appropriate designs for garments meant to emphasise the fitness, athleticism, and attractiveness of those who don them. It is with these objectives in mind that Varsity here chose to design its uniforms in this fashion, and it is no accident that many cheerleading uniforms share similar design features.Footnote ⁸² Granting copyright protection to these designs would enable Varsity to monopolise functional aspects of garments without satisfying the exacting demands of patent law. Although other designers might be free to produce cheerleading uniforms incorporating features that flatten the bust, widen the waist, and shorten the legs, we suspect that they are unlikely to find a vibrant market for such products. This is not the sort of competitive advantage that copyright law is intended to foster.

Unlike the Sixth Circuit, which misinterpreted what makes a design feature utilitarian or expressive, the Supreme Court thoroughly misunderstood the statute it claimed to be interpreting. Justice Thomas’ majority opinion treats the separability criteria as both intellectually simple and doctrinally weak when they are, in fact, neither of these things. He first asks whether the decision maker can ‘spot some two- or three-dimensional element that appears to have pictorial, graphic, or sculptural qualities’.Footnote ⁸³ He describes this requirement as ‘not onerous’,Footnote ⁸⁴ as his application of the test to the uniforms proves. According to the majority, an element of a design counts as an expressive feature if it looks like something.Footnote ⁸⁵ As we discuss elsewhere in a commentary on the decision, ‘considering that this is the area of copyright law that applies to pictorial, graphic, and sculptural works, it is impossible that the claimed works would not have some such features’.Footnote ⁸⁶

The Supreme Court next asks whether the identified expressive feature can exist apart from the utilitarian aspects of the article.Footnote ⁸⁷ The Court explains that ‘the feature must be able to exist as its own pictorial, graphic, or sculptural work … once it is imagined apart from the useful article’.Footnote ⁸⁸ Finally, the Court indicates, ‘If the feature is not capable of existing as a pictorial, graphic, or sculptural work once separated from the useful article, then it was not a pictorial, graphic, or sculptural feature of that article, but rather one of its utilitarian aspects.’Footnote ⁸⁹ Although we do not find these statements to be entirely clear, we think that they imply that given ‘some [expressive] qualities that were spotted in the first step, the second step of the analysis asks whether they could be fixed in some form other than the useful article itself’.Footnote ⁹⁰ Or as we put it more simply elsewhere: ‘In effect, the majority’s test for separability amounts to: Could you draw it on a sheet of paper?’Footnote ⁹¹ ‘Here, because the [uniform designs] could have been, and actually were, reproduced on other (presumably non-useful) media[ – in drawings – ]they were separable.”Footnote ⁹²

The Supreme Court utterly fails to understand the statute and its requirement that decision makers identify distinct expressive and functional features of the claimed work. In its interpretation of the ‘ordinary, contemporary, common meaning’Footnote ⁹³ of the statute, the Court asks whether the spotted expressive features can be ‘imagined apart from the useful article’.Footnote ⁹⁴ But this is not how the statute reads.Footnote ⁹⁵ Instead, the statute asks whether the work’s expressive features can be separated from ‘the utilitarian aspects of the article’.Footnote ⁹⁶ The difference is not trivial. The Court’s statutory misreading substantially lowered the bar for claimants: Rather than distinguishing between and separating the uniform design’s pictorial, graphic and sculptural features from its utilitarian aspects, the Court only asks whether the pictorial, graphic and sculptural features were distinct and separate from the uniform itself. Varsity could then claim the stripes, chevrons and colour-blocking because, once they were imaginatively removed from the uniform, a uniform could still exist. That is a different exercise from asking whether a design’s pictorial, graphic and sculptural features could be separated from the design’s utilitarian aspects.Footnote ⁹⁷

Under the correct formulation of the test, the Court would have had to determine which components of the design it should treat as utilitarian aspects, something it never asked under the test it conceived for itself.

To make matters worse, the Court failed to appreciate how rigorously Congress intended copyright law to treat functionality. The Court seems to have rejected the notion that copyright law should only be used to protect a work’s exclusively expressive features. The majority explains that the imaginatively removed feature of the work ‘may not be a useful article’, but the majority concludes that the feature is still protectable even if ‘the imagined remainder [is not] a fully functioning useful article at all, much less an equally useful one’.Footnote ⁹⁸ This reading of the Act flips the statute on its head. Instead of copyright law protecting only solely expressive features of a work and excluding completely functional and dual-nature aspects, the Court’s opinion will prevent copyright from attaching only to purely functional features of the work.

Given the Supreme Court’s statutory interpretation, it never had to think carefully about the arguments we have raised above. Because the Court lowered the bar for copyright claimants to protect functional aspects of their designs, it did not significantly interrogate whether the features of the uniform designs that influenced perceptions of wearers’ bodies were hybrid. The Court further neglected to rely on the part of the copyright statute that explains what makes a feature utilitarian and which we think requires the classification of many fashion features as utilitarian and therefore inseparable and unprotectable.

Although this is a shame, it is not a complete catastrophe. The useful articles doctrine is not the only mechanism by which copyright law polices functionality, and we are sanguine that subsequent courts will utilise these tools. For example, for the same reasons as we have articulated as to why these design features are not separable, they likely run afoul of the merger doctrine and ought to be denied protectability on that ground as well. When there are a limited number of ways to express something, the merger doctrine denies copyrightability to that expression, lest it protect the function too.Footnote ⁹⁹ Given the particularities of human vision and human torsos, we suspect that there are only a limited number of ways to express the functions of accentuating the cheerleader’s curves, lengthening the cheerleader’s torso and so forth. There is thus good basis to deny copyrightability to these features for having merged function and expression.

In addition, when decision makers compare a plaintiff’s work with an allegedly infringing work, copyright law demands that they filter out from their analysis aspects of the plaintiff’s work that are functional.Footnote ¹⁰⁰ Accordingly, when decision makers are asked to compare garments like Varsity’s to allegedly infringing copies like Star Athletica’s, they should be instructed to ignore aspects of the plaintiff’s work that contribute to its function, including dual-nature features like the ones discussed above. Doing so will help achieve Congress’ goal of limiting copyright protection to functional designs and will minimise costs to consumers and subsequent designers.

3.5 Fashion’s Function in Trademark Law

Copyright law is not the only field of intellectual property law that is concerned with screening out functional aspects of product design. Both trademark law and design patent law also cast a wary eye toward functionality.Footnote ¹⁰¹ Although the doctrines these regimes use to screen functionality differ from those employed by copyright law, their concerns are similar. Accordingly, we analyse how these two fields should respond to the particular issue of fashion design’s dependent functionality, considering trademark law in this section and design patent law in Section 3.6.

Unlike copyright law, trademark law is not concerned with encouraging the creation of new aesthetic works. Instead, to promote fair competition and protect consumers, trademark law guards consumers from marketplace confusion by ensuring that certain symbols accurately reflect the source of goods.Footnote ¹⁰² If a pair of shoes has the distinctive Nike swoosh on it, consumers should be able to trust that the shoes are made by Nike. Trademark law covers more than just words or images that designate source, however. It can also be used to protect features of a product’s design or packaging that identify its source.Footnote ¹⁰³ This is known as trade dress.

Trademark law is similar to copyright in at least one important respect: it also needs to be concerned about functionality.Footnote ¹⁰⁴ Just as features of a pictorial, graphic or sculptural work can have utilitarian features, so too can a product’s design or packaging. Moreover, allowing mark owners to protect the functional aspects of their designs can have anticompetitive effects as it does in copyright law.Footnote ¹⁰⁵ Like copyrights, trademarks do not undergo the substantial ex ante evaluation that patents do to ensure that they are new and nonobvious. In many cases, allowing trademark protection for aspects of garment design that affect the perception of the wearer’s body will produce problematic anticompetitive effects.

To ensure that trademark law is not used to provide claimants with an undue competitive advantage, the law excludes functional trademarks and trade dress from protection.Footnote ¹⁰⁶ Trademark law enables owners to take advantage of the value of their reputations for producing high-quality products, but it seeks to prevent them from using trademarks to gain advantages unrelated to their reputation.Footnote ¹⁰⁷ This unfair advantage could result if the mark extended to cover useful product features that competitors were then prevented from producing.Footnote ¹⁰⁸ It could also occur if the mark extended to product features that consumers desired to purchase irrespective of their source-signifying characteristics.

Features of clothing design that influence the way the wearer’s body is perceived, such that they make him or her look thinner, bigger, taller or curvier should be treated as bearing on the functionality inquiry in trade dress law. Such features should be treated as functional to the extent that they confer an unfair competitive advantage. Thus, while Burberry’s plaid is purely ornamental, a design that used stripes to make the wearer’s legs look longer or one that used a pattern to make the wearer’s feet look smaller should be treated as functional and unprotectable.

3.6 Fashion’s Function in Design Patent Law

Creators of ‘any new, original and ornamental design for an article of manufacture’ may obtain a design patent.Footnote ¹⁰⁹ Design patents provide protection from infringement for 15 years from the date of patent grant.Footnote ¹¹⁰ To get a design patent, a design’s creator must apply to the Patent and Trademark Office, where the application will be examined for patentability.Footnote ¹¹¹ Design patent law does not treat the ‘ornamental[ity]’ requirement as an affirmative condition that designs be aesthetically pleasing.Footnote ¹¹² Rather, the law has long seen this threshold requirement to mean that designs simply be ‘non-functional’.Footnote ¹¹³ The courts have furthermore not understood this requirement to mean that entire designs are necessarily rendered unprotectable for functionality; after all, the ‘article[s] of manufacture’ that design patent law anticipates will be protected will typically have conventionally functional uses.Footnote ¹¹⁴ Instead, courts understand the ornamentality/non-functionality requirement to signify that a ‘design patent only protects the novel, ornamental features of the patented design’.Footnote ¹¹⁵

What ‘functional’ means in the context of design patent law seems less stringent than in copyright and trademark law. A design feature is functional, and therefore unprotectable, only if the feature is ‘“dictated by” the use or purpose of the article’.Footnote ¹¹⁶ By contrast, as the Federal Circuit has explained, ‘when there are several ways to achieve the function of an article of manufacture, the design of the article is more likely to serve a primarily ornamental purpose’.Footnote ¹¹⁷

Design patent law ought to take account of hybrid design features in fashion. That is, the law must sort out how to characterise these features: as ornamental or not ornamental. Design patent law must decide whether these features that function to change the appearance of the wearer rather than merely to look attractive in their own right are well accounted for in the calibrations of design patent law or are better addressed by utility patent law. To the extent that design patent law – or the other non-utility patent IP regimes – ignores fashion’s function, it risks imposing substantial limits on competition, stifling innovation and increasing costs to consumers.

3.7 Conclusion

The account of fashion’s function that we develop in this chapter clarifies how intellectual property law should treat fashion design. It recognises an important way in which designs are functional – when they affect how wearers’ bodies are perceived. This is not simply an issue with fashion design, however. As fashion designer and businesswoman Coco Chanel famously said, ‘Fashion is not something that exists in dresses only. Fashion is in the sky, in the street. Fashion has to do with the ideas, the way we live, what is happening.’Footnote ¹¹⁸ Looking beyond fashion, many visual media can affect the ways that we perceive objects. Trompe l’oeil painting and camouflage techniques have this effect. The same is true of chemical or physical processes for developing pigments, such as Vantablack, a carbon nanotube surface coating that produces the sensation of perfect blackness.Footnote ¹¹⁹ Attempts to gain intellectual property protection for any of these techniques, unless through the patent system, should be met with intense scepticism.

4.1 Introduction

4.1.2 The Major Players in the Sneaker Market

In Australia, traditional sneaker brands such as Nike and Adidas still dominate the performance and lifestyle shoe markets. However, since the sneaker revolution, sneakers are a now a regular offering from many high street and luxury brands, enlarging the lifestyle market with additional (and often more expensive) options. For example, sneakers are now a critical product line from designers such as Balenciaga, Gucci and Alexander McQueen.Footnote ¹⁰ Conversely, traditional sneaker brands have expanded into more fashion-conscious markets. For example, Nike founded NikeLabFootnote ¹¹ to create high end and occasionally avant-garde shoes and apparel. Likewise, Adidas maintains the brand Y-3, which is a long-standing collaboration with renowned avant-garde Japanese designer Yohji Yamamoto (Figure 4.1).Footnote ¹² Most of the traditional sneaker brands also create capsule collections with successful athletes, for example, Naomi Osaka with Nike.Footnote ¹³ However, many collaborations are now with celebrity designers such as Victoria Beckham (with Reebok),Footnote ¹⁴ and Stella McCartney (with Adidas).Footnote ¹⁵ Occasionally, these high-profile relationships experience well-publicised implosions, for example, Adidas’ partnership with rapper Ye (formerly known as Kanye West).Footnote ¹⁶

Figure 4.1

Example of cross-market entrants – Adidas’ Y-3 Fall/Winter 2022 Collection.Footnote ¹⁷

© Thue Norgaard

4.1.2.1 Goliath v Goliath

As fashion designers have entered the lucrative sneaker market, traditional sneaker brands have launched a myriad of lawsuits to protect the distinctive aspects of their shoes including their shape. For example, Adidas are well known for suing any designer using angled stripes on sneakers.Footnote ¹⁸ Likewise, Converse (now owned by Nike) has filed dozens of trademark suits to protect its iconic ‘Chuck Taylor’ sneaker, which have sold over a billion units since 1917.Footnote ¹⁹ In the last few years alone, sneaker infringement cases against fashion designers have become almost commonplace. Recently in 2023, Ralph Lauren was sued by NBA hall-of-famer (and Nike collaborator) George Gervin in relation to its high-top basketball sneaker named the ‘Gervin Mid’. Gervin alleged violations of his right of publicity, trademark infringement and false designation of origin.Footnote ²⁰ The case settled.Footnote ²¹ Also in 2023, Nike sued and settled a trade dress infringement lawsuit against designer John Geiger alleging his GF-01 shoes infringed Nike’s Air Force 1 trade dress (see Figure 4.2).Footnote ²² In another well-publicised case in 2022, Vans successfully restrained the sale of guerrilla art studio MSCHF’s ‘Wavy Baby’ shoes (a collaboration with singer Tyga), which Vans claimed were a copy of Vans’ iconic Sk8 shoes.Footnote ²³

Figure 4.2

Nike Air Force 1 trade dress vs Geiger’s alleged infringing products.Footnote ²⁴

Figure 4.2 Long description

The table demonstrates the designs the differences in design and branding between the genuine and infringing products. Two trade dress shoes are outlined in the first column. The genuine air force 1 shoes with a swoosh Nike logo is in the second column. The infringing or Geiger products illustrates six shoes that are arranged vertically. The shoes either have lace or no lace at all.

Courtesy of US District Court, Central District of California, opinion

4.1.2.2 Approaches Vary across Jurisdictions

These examples demonstrate that it has never been more important for sneaker designers to protect their distinctive trademarks or innovative designs. This competitiveness, coupled with the globalisation of brands, has heightened interest in cross-jurisdictional protection for footwear designs and revealed important differences between legal systems. Of relevance to shoes, this includes the treatment of 3D shape trademarks, the role of design laws, overlap rules and the treatment of functional signs and designs. For example, in the EU, UK, Singapore and Australia, fashion designers have access to both trademark rights and some form of registered or unregistered design rights. In contrast, shoe litigation in the United States typically proceeds as infringement of ‘trade dress’ – a trademark doctrine originally limited to packaging but that has expanded since the 1980s to ‘the shape and design of the product itself’.Footnote ²⁵ The limits of the latter doctrine continue to lead commentators to demand some form of sui generis fashion or design right.Footnote ²⁶

This chapter uses Australia as a case study to illustrate the empirical impact of the sneaker revolution on registered trademark and design rights. This case study also serves as a cautionary tale for those advocating sui generis design rights for fashion designers. This chapter reveals the problems with the Australian system, which implemented a registered design right (without a grace period for self-disclosure) and also allows for concurrent 3D shape trademarks (where there is no statutory bar against functional shapes). In addition to these technical lacunae, this chapter will argue that aspects of culture are relevant in designing an optimal IP framework. This chapter argues that there are unusual aspects of Australian culture as well as IP law that shepherd designers towards registering the overall appearance of a product (such as a shoe) as a 3D trademark but. not as a registered design. This is a problem because in Australia (as with many jurisdictions), a preference for trademark rights over other time-limited rights such as patents or designs can lead to perpetual trademark registrations and anti-competitive monopolies.

4.2 A Culture of ‘Casual Piracy’

In comparison to the jurisdictions discussed in this book, Australia is unusual in the sense of being a relatively wealthy countryFootnote ²⁷ but one that has a notorious attitude of casual piracy towards designer goods.Footnote ²⁸ By ‘casual piracy’, I am not referring to ‘frequency of downloads’, as that phrase is used in a copyright piracy context.Footnote ²⁹ Nor am I referring to so-called ‘benign copying’ done by design-intensive countries, which begets fresh innovation (typically discussed as the piracy paradox).Footnote ³⁰ ‘Casual piracy’ in this context refers to an observation that in Australia, buying and using copies of designer goods is not a hidden shame. Surveys have shown that from a cultural perspective, Australians are more accepting of counterfeit or replica designer products than other developed nations.Footnote ³¹ For example, YouTube videos can be found of Australians instructing others where to buy the best fake bags in Bali (a popular holiday destination for Australians).Footnote ³² This kind of ‘casual piracy’ is often open and frequently normalised. This casual attitude to piracy has been discussed in a copyright context,Footnote ³³ but I argue that the normalisation of piracy exists in the design context as well.

In a recent IP Australia survey,Footnote ³⁴ industry respondents reported ‘that consumers and other businesses in Australia do not sufficiently value original design or appreciate the impacts (financial and otherwise) of copying designers, particularly in fashion and furniture’.Footnote ³⁵ In reviewing public sentiment, IP Australia found that although designers were concerned about unauthorised copying, this concern was not reflected in the broader public conversation.Footnote ³⁶ In other words, Australians seem to have a strong appetite for replica goods particularly where the price for the original is much higher than the price for the copy.

In a classical paradigm, a designer might protect their goods from a copier by applying a trademark to the face of the article. However, under common law concepts of trademark law, the equity protected is in the trademark itself and not in the underlying product. Therefore, a counterfeiter could avoid trademark infringement by replicating the underlying article – for example, shoe, bag or dress – but omitting the trademark.Footnote ³⁷ In Australia and elsewhere, there would arguably still be a strong demand for this unbranded replica product. This has led some fashion houses to embrace the ‘logo-fication’ of their goods. For example, some designers with strong trademarks may incorporate or weave their trademarks into the surface of the articles. For example, Burberry’s check trademark and Louis Vuitton’s monogram trademarks have been incorporated into the entire surface of clothing, handbags and shoes (Figure 4.3).Footnote ³⁸ This is a sub-optimal choice for weaker brands and also for well-branded fashion houses who may refuse to confine their creativity to logo-fied goods.

Figure 4.3

Logo-fication of designs: (left) Burberry’s check trademark used upon a sneaker;Footnote ³⁹ (right) Louis Vuitton quatrefoils and flowers trademarks used upon a sneaker.Footnote ⁴⁰

(a) © Burberry Limited (b) © Louis Vuitton Australia Pty Ltd.

In Australia, for those designers who cannot or will not ‘logo-fy’ their products but want to protect the overall appearance of their goods, the most attractive option is to pursue protection for the overall appearance of the underlying article, for example, to pursue shoe protection as a 3D shape or as a registered design. In other words, the legal impetus to protect fashion items as 3D IP is important in Australia due to its culture of casual piracy (which happily consumes un-branded ‘knock-offs’) and where logo-fication is an unavailable or unappealing option.

4.3 Legal Options for Fashion Designers in Australia

If a fashion designer seeks to prevent the copying of an object itself – apparel, shoes, bags, etc – there are many options available to a designer under Australian IP law. For example, a designer can access formal registered IP rights under the Designs Act 2003 (Cth), Trade Marks Act 1995 (Cth) or Patents Act 1990 (Cth) or unregistered rights under the Copyright Act 1968 (Cth). Designers may also access allied rights under statutory consumer law or the common law tort of passing off.Footnote ⁴¹ However, this chapter argues that Australian law maintains peculiar lacunae within, and between, copyright, designs and trademark law that shepherd designers towards relying upon 3D trademark rights to protect their designs rather than other types of rights. This is an outcome most jurisdictions try to avoid due to anti-competitive issues specific to trademark law. The concern is that because registered trademarks can be renewed indefinitely, a perpetual monopoly over a shape (particularly a functional shape) might be granted to a designer/trademark owner. To understand how this sub-optimal situation arose and why designers nevertheless may default to, or strategically use, shape trademarks, a brief overview of relevant aspects of Australian copyright, designs and trademark law is required.

4.3.1 Copyright and Design Overlap Laws

In Australia, as with many jurisdictions, copyright subsists automatically once a work assumes material form; registration is not required, and copyright terms are long, typically the life of the author plus 70 years after death of the author.Footnote ⁴² This can make copyright an attractive ‘default’ if a fashion designer has not, or cannot, obtain other forms of protection, for example, a registered design right. In line with other Commonwealth jurisdictions, the Australian Copyright Act 1968 (Cth) prescribes a closed list of categories for copyrightable works.Footnote ⁴³ However, as with the United Kingdom, in Australia the exclusion of utilitarian works is via case lawFootnote ⁴⁴ rather than expressly proscribed by statute as in the United States.Footnote ⁴⁵

In contrast to the United States, Australia also maintains sui generis design laws that were originally modelled on UK legislation.Footnote ⁴⁶ However, unlike the United Kingdom and many EU member states,Footnote ⁴⁷ Australia pursues a strict approach to demarcation and limits the cumulation of copyrights.Footnote ⁴⁸ The legislative overlap provisions are complex, but essentially designers lose copyright protection if they apply their design industrially, that is, where the design is reproduced 50 times through embodiment into a 3D product.Footnote ⁴⁹ So a sketch of a shoe retains copyright as an artistic work (e.g., as a painting) if it is printed on a T-shirt (a 2D reproduction of an artistic work as surface ornamentation) but loses copyright if the reproduction is as a 3D product, for example, a shoe.Footnote ⁵⁰ In effect, these provisions tacitly encourage designers to use shorter-lived registered design rights rather than rely on lengthy unregistered and unsearchable copyrights for 2D sketched designs.Footnote ⁵¹ For reasons discussed later, an increased interest in registered designs failed to materialise.

An important exception to the overlap provisions exists for products categorised as works of artistic craftsmanship (WOACs).Footnote ⁵² WOACs may receive copyright protection even if registered or registrable under the Designs Act 2003 (Cth).Footnote ⁵³ There is no definition of a WOAC in the Copyright Act 1968 (Cth), but UK and Australian courts have cited as examples the works of ‘silversmiths, potters, woodworkers, hand embroiderers and many others’Footnote ⁵⁴ whose works are seen as deserving of longer protection due to ‘special training, skill and knowledge’.Footnote ⁵⁵ The absence of a strict definition has meant designers of more modern products, such as boat hulls,Footnote ⁵⁶ corkscrewsFootnote ⁵⁷ and handbagsFootnote ⁵⁸ have tried to shoehorn themselves into copyright in the absence of registered design right protection.

In relation to fashion, in a 1998 case,Footnote ⁵⁹ the knitwear brand Coogi successfully argued that a length of its highly textured knitted fabric (which featured three-dimensional elements) was a work of artistic craftsmanship.Footnote ⁶⁰ However, it failed to make out that copyright in this WOAC had been infringed.Footnote ⁶¹ Unlike Coogi, most other fashion plaintiffs have failed in subsistence arguments, that is, that their products rose to being works of artistic craftsmanship. For example in 2003, a court held in Muscat v LeFootnote ⁶² that mundane black pants lacked the requisite artistry and artistic intent to be classed as a WOAC.Footnote ⁶³ In a more recent case, the handbag designer of Escape’s popularFootnote ⁶⁴ neoprene tote bag failed to convince the Federal Court of Australia that the prototype for the bag was a WOAC (Figure 4.4).Footnote ⁶⁵ In that case, the court held that ‘whether the bag is a work of artistic craftsmanship turns on the extent to which the Escape Bag’s artistic expression, in its form, was unconstrained by functional considerations’.Footnote ⁶⁶ The court could not equate the designer’s choices, for example, of neoprene fabric or using sailing rope for handles, to that of an ‘artist-craftsperson’;Footnote ⁶⁷ instead, they were considered ‘an evolution in styling’Footnote ⁶⁸ rather than an act of artistic craftsmanship.

Figure 4.4

State of Escape neoprene bag.Footnote ⁷⁰

© Federal Court of Australia

These cases indicate that copyright protection for a 3D object is theoretically possible if the work expresses the necessary artistry and artistic intent to be considered a work of artistic craftsmanship. However, the bar for such classification is extremely high. In Coogi, there was significant evidence of artistic decision-making in Coogi’s unusual stitch structures and colours.Footnote ⁶⁹ There was much evidence of conscious design choices incorporated into the 3D forms of the knitted textiles. Designers of less unusual forms will likely fail to qualify their products as WOACs and (as the legislature intended) should consider alternative rights, such as a registered design right. Unfortunately, as discussed below, Australian design law has had significant limitations.

4.3.3 Trademarks

The advent of the Trade Marks Act 1995 (Cth) introduced major changes to comply with Australia’s obligations under the TRIPS Agreement.Footnote ⁹¹ The definition of a ‘sign’Footnote ⁹² was broadened to expressly include shapes, sounds, colours and scents.Footnote ⁹³ The protection of 3D shapes therefore has a different foundation to US trade dress law, which was influenced by state-based unfair competition law.Footnote ⁹⁴ In Australia, unlike the United States, there is no requirement to show actual use or an express intention to use at the time of application; instead marks can be subject to removal for non-use after registration.Footnote ⁹⁵ Indefinite renewals are permitted (without re-examination) so long as the renewal fees are paid.Footnote ⁹⁶

As with other TRIPS member countries, there were concerns that the express codificationFootnote ⁹⁷ of shape trademarks could have an anti-competitive effect if mark owners were to have monopolies over functional shapes.Footnote ⁹⁸ A classic example is a basic bottle shape, or in a fashion context, the shape of a basic T-shirt. Both have an opening and a shape designed to contain a substance, for example, a liquid or a human torso. These basic shapes serve a critical utilitarian function and to grant a trademark monopoly over such shapes might inhibit competition.

Most jurisdictions proscribe against registration of functional shapes though express statutory bars.Footnote ⁹⁹ However, for reasons that remain ‘obscure’,Footnote ¹⁰⁰ there are no express legislative provisions in Australia to refuse functional shapes despite inclusion of such provisions in earlier drafts of the current ActFootnote ¹⁰¹ and in similar Commonwealth jurisdictions.Footnote ¹⁰² Instead, functionality is dealt with as an aspect of distinctiveness at registration or when assessing an infringer’s impugned use (ie, where use of functional aspects of the mark are considered non-infringing).Footnote ¹⁰³ In his comparison of different jurisdictions, Michael Handler has argued that this is a ‘suboptimal way of managing the problems raised by functional marks’.Footnote ¹⁰⁴

The Australian Trade Marks Office (ATMO) tries to factor in functionality into the assessment of distinctiveness, that is, whether ‘other persons, trading in goods of the relevant kind and being actuated only by proper motives … will think of the sign and want to use it in connexion with similar goods’.Footnote ¹⁰⁵ If a shape mark has ‘significant functional features there is a strong likelihood that other traders will need to use that shape for their similar goods’ in the context of fair competition.Footnote ¹⁰⁶ However, evidence of use can be presented to show that a functional shape has become in fact distinctiveFootnote ¹⁰⁷ or has acquired ‘distinctiveness through use’.Footnote ¹⁰⁸ Nevertheless, it appears that using distinctiveness as a filter against functional marks is ‘imperfect’.Footnote ¹⁰⁹ In his analysis of the Australian Register, Handler identified a ‘significant number’ of shape and colour marks that were prima facie functional. Conceding that it was impossible to know if functionality was considered by the ATMO as such records are not publicly available, Handler identified several shape marks that were rejected in other jurisdictions on functionality grounds but were accepted for registration in Australia by the ATMO.Footnote ¹¹⁰

4.3.3.1 Australian Shoe Shape Litigation

In Global Brand Marketing Inc v YD Pty LtdFootnote ¹¹¹ (Global Brand) – the only reported shoe shape infringement case in Australia – details of the prosecution history were discussed in the judgment. In that case, Diesel S.p.A (Diesel) the licensee of Global Brand Marketing, was able to register the ‘Shape’ trademark (#986713) for the overall appearance of its popular square toed sneaker and the ‘Sole’ trademark (#986709) for the embossed under-sole of the shoe (Figure 4.5).

Figure 4.5

Registered shape marks of Diesel:Footnote ¹¹⁴ (left) shape mark (#986713); (right) sole mark (#986709).

(a) © Federal Court of Australia (b) © Federal Court of Australia

The Federal Court of Australia noted that the ATMO examiner originally objected to the shape mark on the grounds that: ‘[The] trade mark consists of the shape of a shoe. The shape is an ordinary shoe shape. Other traders should be able to use an ordinary shoe shape in connection with goods or services similar to yours.’Footnote ¹¹² In response, Diesel limited the registration with an endorsement emphasising distinctive aspects of the registration such as the ‘stylised letter D’, and the registration was accepted.Footnote ¹¹³

In evaluating the validity of the shape mark, the court reviewed the authorities and gave its support to the principle that non-functional shapes can operate as Australian trademarks. Moreover, the court in construing the infringement claim, declared that YD were also using its own shoe shape and sole as shape trademarks of YD. The court held that YD’s trademarks as used by YD were not deceptively similar to Diesel’s Shape and Sole marks and therefore not infringing. Although the plaintiffs were not successful, it was significant that Justice Sundberg found that both parties had valid trademark rights in the shape of their shoes as well as the soles of those shoes. In relation to the plaintiff’s registered marks, his Honour found that the non-functional and non-descriptive features overwhelmed the functional and common features of the marks.Footnote ¹¹⁵ This decision likely gave traders some confidence to apply for trademark registrations for distinctive shapes related to footwear.

Given all the factors discussed above – increased demand for sneakers, a culture of casual privacy, limitations of copyright and designs law and favourable trademark courts – it is hypothesised that shoe designers are increasingly reliant on protecting footwear by registering them as 3D shape trademarks. Although causation is not argued, it is suggested that 3D shoe registrations may be on the increase due to these various factors.

4.4 Method

To examine whether and how fashion designers have employed Australian trademark and design registrations in relation to shoes, publicly available trademark and design registration data from IP Australia was extracted. In relation to trademarks, this study uses two sources of data. First, trademark application data from IPGOD2021Footnote ¹¹⁶ was extracted by a research assistant for the period January 1, 1996 (commencement of the Trade Marks Act 1995 (Cth)) until December 31, 2021 (25 years).Footnote ¹¹⁷ This data was used to identify trends over time. Second, in order to examine specific trademark applications in detail, the author located these via the online Australian Trade Marks Register (the Register).Footnote ¹¹⁸ Additional data was retrieved from the online Australian Designs Register.Footnote ¹¹⁹ All coding was done by the author.

A key limitation of the study is that the number of applications in the dataset is small. Descriptive statistics based on a small number of applications are sensitive to change if a small number of new applications are added to the dataset. Nevertheless, the applications cover the entire population of available applications in the period of inquiry, so are in this sense representative. Still, this study does not make any predictive statements and inferences are only drawn with caution.

4.5 Results and Discussion

4.5.1 Overview of Trademark Application and Registration Data

4.5.1.1 Shape Relative to All Applications

Over the 25-year period (January 1, 1996–December 31, 2021) there were 1,486,174 trademark applications to IP Australia. Of these, only 3,631 were for shape trademarks, that is, less than 1% (on average only 0.24%). The bars of Figure 4.7 show that while there was some enthusiasm for shape mark applications in the first 10 years of the Trade Marks Act 1995 (Cth), the number of shape applications slightly declined after 2015. This decline is significant in light of the fact that trademark applications in general (ie, for non-shape marks) increased dramatically over the same period.Footnote ¹²⁰ The relative proportion of shape to total marks (line) can therefore be seen to decline over time. In other words, Figure 4.6 shows that both the nominal and relative popularity of shape trademarks fell after their initial popularity. This is consistent with previous findings by Amanda Scardamaglia and Mitchell Adams.Footnote ¹²¹

Figure 4.6

Shape trademark applications (actual) and relative to annual trademark applications per year (Janaury 1, 1996–December 31, 2021).

Figure 4.6 Long description

The vertical axis on the left ranges from 0 through 250, and the axis on the right ranges from 0.00 through 0.90 percent. The horizontal axis presents a time period. The bars represents the number of applications over the years which are as follows. (1996, 200), (2005, 170), (1997, 160), (1998, 175), (1999, 160), (2003, 160), (2004, 155), (2014, 154), (2015, 150), (2008, 150), (2010, 150), (2012, 140), (2014, 140), (2021, 135), and so on. The line represents the percentage of annual applications, which originates at (1996, 0.80), drops at (2000, 0.30), fluctuates, and terminates at (2021, 0.18). The values are estimated.

Credit: Vicki Huang

Figure 4.7

Class 25 shape trademark applications (actual) and relative to annual shape trademark applications per year (January 1, 1996–December 31, 2021).

Figure 4.7 Long description

The vertical axis on the left ranges from 0 through 250, and the axis on the right ranges from 0.0 through 18.0 percent. The horizontal axis presents a time period. The bars represent the number of shape applications over the years, which are as follows: (1996, 200), (2005, 170), (1997, 160), (1998, 175), (1999, 160), (2003, 160), (2004, 155), (2014, 154), (2015, 150), (2008, 150), (2010, 150), (2012, 140), (2014, 140), (2021, 135), and so on. A dashed line representing the linear number of shape applications also originates at (1996, 160), intersects the bars and terminates at (2021, 120). A solid line representing class 25 as a percentage of annual share applications originates at (1996, 6.0), fluctuates, and terminates at (2021, 16.0). The values are estimated. Another dashed line representing the linear class 25 as a percentage of annual share applications originates at (1996, 6.0), fluctuates, and terminates at (1996, 8.0). The values are estimated.

Credit: Vicki Huang

4.5.1.2 Applications Related to Fashion

Fashion designers can apply for marks across a wide range of goods and services. For the purpose of this chapter, the focus was on shape applications under Nice Class 25 (clothing, headgear and footwear)Footnote ¹²² between January 1, 1996, and December 31, 2021. Fashion designers also applied for shapes under other classes but in much smaller numbers. For example, under Class 18 (bags and wallets) there were 92 shape applications, and under Class 14 (jewellery) there were 105 shape applications for the same period.

Of the 3,631 shape mark applications, 226 (6.2% or 226/3,631) were for Class 25 (clothing, headgear and footwear). Figure 4.7 plots the total number of shape applications per year (bars) and the percentage of shape applications that were in Class 25 (line). Figure 4.7 shows that while the nominal number of shape applications declined over time (descending dotted trend line), the proportion that were shape applications in fashion (Class 25) grew over time (ascending dotted trend line).

Bearing in mind the small population size of 3,631 shape mark applications with 226 in Class 25, from Figures 4.7 and 4.8 we can infer that while shape applications lost their initial popularity, the interest in shape trademarks for Class 25 (ie, the fashion industry) grew in both nominal and relative terms.

Figure 4.8

Whole shoe trademark applications and status ($n=55$, January 1, 1996– December 31, 2021).

Figure 4.8 Long description

The vertical axis represents the number of applications that range from 0 to 10. The horizontal axis represents the shoe style, including flip-flops or thongs, sneakers, sandals, Derby lace-up, Crocs, boots, and Birkenstock Sandals. The maximum and minimum number of accepted applications are for sneakers, boots, crocs and flip-flops or thongs, which are 9, 5, 3, and 1, respectively. The maximum and minimum number of pending applications are for Birkenstock Sandal and Sneaker, which are 7 and 5, respectively. The maximum and minimum number of accepted applications are for boots, sandals, sneakers, flip flop or thongs, Derby lace up and Croc, which are 7, 6, 6, 4, 1, and 1, respectively.

Credit: Vicki Huang

4.6 Discussion

The empirical data reveals that while overall demand for Australian shape trademarks is low and declining, the demand for such marks for fashion goods has increased in real and relative terms. A significant portion of this demand has come from the fashion industry, particularly from shoe applicants seeking a 3D shape trademark. This would suggest that shoe designers (particularly some sneaker designers) view shape trademarks favourably. As discussed earlier, this demand could be a response to Australia’s culture of casual piracy and/or the perceived limitations of other IP rights such as design rights.

A more general explanation for the increased demand for shoe trademarks is an overall increase in the aggregate demand for shoes. In relation to Australia, the footwear market has grown significantly. For example, in 2004, the retail value of the market was US$1.3 billion.Footnote ¹²⁷ By 2017, the Australian footwear market reached a value of US$3.6 billion.Footnote ¹²⁸ This mirrors global trends, where shoe sales have grown exponentially, becoming a critical product category for fashion designers and sportswear brands.Footnote ¹²⁹ A key driver of this growth has been sneakers. MarketLine Australia reported in 2017Footnote ¹³⁰ and 2021Footnote ¹³¹ that the ‘growing popularity of sports footwear is driving growth in the market, thanks to the “athleisure” trend that has swept across the world’. Significantly in 2021 they noted, ‘The cultural shift of athletic footwear being repurposed as everyday wear has changed the dynamics of the footwear market.’Footnote ¹³² This is not just an Australian phenomenon. McKinsey & Company have reported that globally, fashion leaders predict that casual wear, sportswear and sneakers are the categories with the greatest growth potential.Footnote ¹³³ This trend began before the global pandemic but no doubt accelerated during stay-at-home work mandates in Australia as well as other countries.

There are also fascinating, systematic cultural changes that are impacting the growth in demand for sneakers. In particular, the growth of ‘sneakerhead culture’ and the increased interest by men for fashion (typically expressed through an interest in sneakers). Yuniya Kawamura argues that ‘men are reclaiming adornment and fashion’, expression men had abandoned due to the adoption of the business suit as ‘uniform since the Industrial Revolution’.Footnote ¹³⁴ However, ‘to avoid being “effeminate,” girls and women are shut out from the community’ leading to a male dominant sub-culture of ‘sneaker-heads’ or ‘sneaker-holics’.Footnote ¹³⁵ This gendered demand is reflected in economic studies that show the footwear market is also pivoting male. For example, in Australia in 2017, the ratio of women’s to men’s footwear was 53.4% to 34% of the dollar value of the overall market.Footnote ¹³⁶ In 2020, this gender gap closed with the ratio of women’s to men’s footwear becoming 42% to 41.4% of the market.Footnote ¹³⁷

Admittedly, it is difficult to parse out how much of this gender swing relates to the pandemic. Australia has an uneven labour force participation between genders,Footnote ¹³⁸ and the confinement of both male and female office workers may have increased demand for sneakers unevenly. Nevertheless, even pre-pandemic, the popularity and marketing of sneakers towards men was evident. Kawamura, points out that ‘boys and young men are the biggest consumers of sneakers’ and this is reflected in the supply of sneakers where ‘there are far more varieties and choices for male sneakers than female sneakers’.Footnote ¹³⁹ As sneakerhead culture becomes more visible,Footnote ¹⁴⁰ and profitable, how designers respond to the increased influence of the male customer will be fascinating.

The increase in competition in the global sneaker market shows no sign of slowing down. With that there will likely be increased demand for IP protection in all jurisdictions. This study has shown that in Australia shoe designers have sought out 3D shape trademarks as an addition to (or instead of) registered designs. Other jurisdictions are dealing with shoe designers pressing other types of rights. For example, difficulties with enforcing trade dress rights in the United States has led to interesting strategies from some designers. The brand Off-White has successfully been granted a US registration of the ironic trademark ‘FOR WALKING’, which is used on its sneakers and other shoes.Footnote ¹⁴¹ Nike, has amassed over 34,000 patents globallyFootnote ¹⁴² and recently sued Lululemon for infringement of textile patents used in Nike’s flyknit sneaker range.Footnote ¹⁴³

Nevertheless, for many designers, it is the overall appearance of the product, that is, the shoe design, that is seen as warranting protection. The increase in shoe demand and alleged copying of designs will undoubtedly lead to further calls for the United States to adopt a form of sui generis design right for fashion. Such advocates should view Australia as a cautionary tale. This chapter has argued that the success of fashion design rights requires a holistic consideration of the culture, commerce and the matrix of related rights. A problem for jurisdictions such as Australia where the matrix allows for and tacitly encourages the registration of shoes as shape trademarks, may be the anti-competitive effects of this demand. As with other jurisdictions, trademarks in Australia can be registered indefinitely without substantive re-examination. Although the dataset in this study is fairly small, it is clear that many significant brands are using the Australian trademark system to protect the overall appearance of their goods. In Australia, the majority of footwear trademark applicants are foreign companies. Given a concern of the government is the stimulation of the domestic design industry, an increase in foreign-owned trademarks over shapes is an issue that requires particular vigilance. And it will be interesting to observe whether demand for more time limited design rights increases now that changes (including the inclusion of a grace period) have been implemented.

5.1 Introduction

In its simplest form, the fashion industry may be described as the business of making apparel. Yet most will agree that the term ‘fashion’ goes beyond mere functionality in apparel to incorporate other attributes such as identities in styles, personal tastes and cultural biases. As an industry, it spans a wide range of offerings from haute couture, luxury fashion, bridge brands and diffusion lines to high street, fast fashion and economy brands,Footnote ¹ providing the market with a plethora of fashion products at varying price points and other attributes. To survive and thrive in the fashion industry, designers and artisanal craftsmen must constantly innovate and create iconic and memorable designs and fashion concepts to differentiate their products from those of others, towards the ultimate objective of establishing a loyal customer base. Originality in designs and fashion concepts is thus the lifeblood of fashion designers. The mass market covets celebrated designs, many of which have originated from talented designers in haute couture and luxury fashion as well as lesser-known freelancers, because of their aesthetic appeal and exclusivity. The prevalent use of social media has contributed significantly to the extent and speed that unique and distinctive designs are able to reach vast consumer markets. A few notable design concepts have transcended mere commercial merchandise, successfully permeating popular culture. As such, fashion designers are driven to claim ownership in their designs and design concepts with the objective of preventing others from copying and exploiting them without attribution or due compensation.

Unfortunately, such protection is not always easy to achieve. While registered design rights may provide the means to protect designs, which are novel (not common to the trade) and not functional in nature, for a limited number of years (on average between 15 and 25 years depending on the country), many designs in the fashion industry, in particular designs of apparels, shoes, bags and other fashion accessories, often fall short of the requisite standard of novelty because they are common to the trade. Yet it has been noted that certain aspects or features of these designs are able to capture consumers’ attention and create lasting impressions. Examples of such memorable designs often cited include the eye-catching red on the bottom of a Christian Louboutin heel,Footnote ² the intrecciato weave pattern and knot adornment on a Bottega Veneta product, the specific placement of buckles and zippers on a Balenciaga handbag, the red-green-red racing stripes of Gucci, the Adidas three-stripes design and the protruding ‘pocket tab’ in the form of a folded ribbon sewn with its ends captured in the seam of a rear patch pocket on a pair of Levi’s jeans, just to mention a few. The lack of adequate legal protection of fashion designs and design concepts continues to be a major source of dissatisfaction for fashion designers and is deeply rooted in a palpable sense of inequity.

At the same time, the modern fashion industry has grown into a huge lucrative global business and is a far cry from the bespoke artisanal craft business model of centuries past. At the outset of the twenty-first century, the modus operandi of fast-fashion conglomerates is the pursuit of profitable styles that can be translated, with the help of competitive technology, into timely in-season merchandise for the global mass apparel market. Aesthetically appealing designs and original design concepts are often the targets of cheap knock-offs almost immediately after they are presented to the public at fashion shows and on the catwalks. This has further fuelled the need for fashion designers and fashion houses to find effective ways to obtain property rights in their fashion designs and concepts in a bid to stop unauthorised exploitations.

There is a discernible trend in the fashion industry to obtain property rights by filing trademark applications for iconic design features applied on products, such as single colours, monograms and patterns, commonly referred to as non-traditional trademarks (NTTMs), and by defending vigorously the validities of these registered trademarks in courts. Recent examples of fashion houses going to court to lay claims over NTTMs that they consider distinctive of their goods and services include Bottega Veneta’s intrecciato weave and knot marks, Christian Louboutin’s red sole mark, the Adidas three-stripes design and the Levi’s protruding ‘pocket tab’.

The surge in registrations of NTTMs globally is the result of several factors. Following the 1994 Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), the world has seen the growth of the trademark universe. The broader definition of a trademark in TRIPS and the willingness of policymakers across all continents to facilitate the registration of traditional marks and NTTMs alike have contributed to the expansionist trend observed. In particular, modern branding and marketing strategies (which have had success in educating consumers to make a connection between a specific colour, design feature or smell and a particular origin of goods and services) have convinced businesses to seek exclusive rights over NTTMs. Yet it could be said that the protection of NTTMs have not been harmonised at the multilateral level despite their widespread acceptance and use. Indeed, courts in different countries have not always been consistent with the treatments of NTTMs, often resulting in differing judgments and creating uncertainties for trademark proprietors when protecting their NTTMs.

This chapter examines the use of NTTMs in the fashion industry. In particular, the authors study the arguments for and against the protection of NTTMs with reference to selected cases in the United States and the European Union. Although some fashion designers and enterprises have celebrated victories in court, it has been widely noted that the registration and protection of these NTTMs as trademarks can be problematic and potentially detrimental to continued innovations in the fashion industry. Closely connected to the conferral of trademark rights in NTTMs is the countervailing policy to ensure market competition and public accessibility to common design features and shapes in the fashion industry. Because of this, central to the discussion of where the appropriate balance ought to be struck is the element of distinctiveness of NTTMs, that is, their ability to distinguish goods and services, and consequently whether they should be conferred trademark rights. As it is generally accepted that NTTMs are inherently devoid of distinctiveness, the criteria in the assessment of whether they have acquired distinctiveness through use (in the alternative) becomes critical for proprietors seeking trademark protection of their NTTMs. To this end, the authors critically examine the different tests of ‘association’ and ‘reliance’ as discussed in recent court decisions on acquired distinctiveness of NTTMs and conclude with a recommendation that the ‘reliance’ test is preferred and ought to be adopted.

5.2 NTTMs in the United States and the European Union

As a starting point, there is no internationally agreed-upon and accepted definition of NTTMs.

Under Article 15.1 of TRIPS, any sign, or any combination of signs, capable of distinguishing a business’s goods or services from those of others is a trademark. This definition is expansive in scope and identifies the most important and universally accepted function of a trademark, that is, its ability to distinguish and to serve as a badge of origin for goods or services. In fact, an all-encompassing reference to ‘any sign’ in the definition of a trademark without qualification in Article 15.1 paves the way for the registrability of unconventional signs, including those that are not visually perceptible. As such, any signs outside of the traditional categories of words and images may be considered NTTMs and they possibly include visually perceptible signs like single colours, shapes, video clips, holograms, gestures and position marks, as well as non-visually perceptible ones like sounds and smells.

The broad definition of a trademark provided in TRIPS has opened up a whole new universe of possible signs, that of NTTMs, which may be protected through trademark registrations. Furthermore, registrations of NTTMs received a significant booster from post-TRIPS treaties such as the Singapore Treaty on the Law of Trade Mark 2006, which has provided the international trademark community with some degree of clarity and certainty in the registration procedure for novel types of trademarks, including non-visible signs. This has in turn facilitated the increase in registrations of NTTMs globally, including by prominent fashion houses and fashion designers. Indeed, over the years, fashion house and designers have not only taken advantage of the increasingly liberal and expansive legal norms in trademark protection to secure protection for NTTMs with regard to their fashion products but have also resulted in a marked increase in litigation in the fashion industry all over the world.

5.3 The Protection of NTTMs – Problems and Objections

As the above cases demonstrate, the protection of NTTMs is not without difficulties. In certain jurisdictions, the range of registrable NTTMs is limited by the requirement that, in order to be protected as a mark, a sign must be ‘capable of graphic representation’ or be ‘visually perceptible’. Based on this requirement, it becomes practically impossible, or considerably more difficult, to protect non-visual signs such as smells and sounds unless they could be visually represented in some way.

Moreover, from a public policy perspective, some, including one of the authors,Footnote ⁵⁴ have argued that granting registered trademark rights to NTTMs is problematic and ought to be discouraged or disallowed based on competition and freedom of expression concerns.

For one, NTTMs are signs that often consist of features in product designs such as shapes, patterns or colours that may be common or functional in nature and thus ought to remain in the public domain and be available for use by others in the course of trade. In trademark law lingo, NTTMs are either ‘devoid of distinctive character’ or descriptive of ‘the kind, quality, intended purpose … or other characteristics of good or services’ and as such it is in the public’s interest (in particular, from the perspective of fair competition) to keep them free for all to use. To use the examples referred to above, granting trademark protection to the Bottega Veneta Intrecciato pattern, LV Damier Checkerboard pattern or Louboutin red-sole shoes means that, in practice, no one can copy these designs or produce similar-looking ones. One key reason is because registered trademark protection is potentially indefinite, since marks can be renewed after a 10 year period. As a result, consumers are denied the opportunity to purchase other products with the same features (often with a much lower price tag). Some may argue that this may not be of major concern as consumers are still offered a large number of alternative choices in a market economy – several types of bags or shoes, even without these designs. Still, competitors are foreclosed from producing those product features and consumers cannot be offered less costly, or merely other, alternatives. This runs fundamentally against the principles of the trademark right, whose ultimate objective is to promote market competition, not to protect a monopoly on aesthetic and appealing features.

An additional concern is that the protection of NTTMs ultimately encourages prolonged repetitive use of iconic or popular design features of a product so as to secure and maintain trademark registrations under the concept of distinctiveness acquired through use, forgoing alternative novel and innovative design features. In particular, once a feature receives trademark protection, companies need to continue to replicate it ‘as is’, lest they lose protection because they would be using the ‘marks’ inconsistently. In turn, this may have a chilling effect on creativity and innovation of the enterprises that are proprietors of the NTTMs, with the result that society is worse off overall. For example, every pair of Louboutin shoes, be they for women or men, use the famed red soles today. Likewise, the LV Damier checkerboard pattern is omnipresent across most LV products, particularly its bags. The same applies to the intrecciato weave – albeit in different widths – at Bottega Veneta. Of course, these companies are eager to repeatedly use their marks because doing so drives more consumers to purchase their products. Yet, should companies desire to change their products, consumers would struggle to recognise them because they have become used to the same features in all products. Precisely because of this, LV’s recent attempts to innovate its products has been largely based on reproducing its signature monogram and checkerboard patterns on different materials such as transparent plastic, or as the new menswear designer Pharrell Williams had done, rendering the Damier pattern to resemble pixels and jungle camouflage. Likewise, Bottega Veneta has introduced products with the intrecciato weave in different larger sizes, but this innovation arguably does not depart significantly from the brand’s strategy of using the same iconic pattern on all its products.

In other words, granting excessive protection to NTTMs can also result in limiting investments in product innovation and quality. In particular, the possibility to continue to extract value from the same shapes, colours and patterns protected on an exclusive basis – without competitors being able to copy these features – may lead producers to avoid investing in new products as the above examples demonstrate. This is a serious concern. IP laws are supposed to promote innovation, yet NTTMs could become vehicles to promote standardisation and commercialisation of non-innovative products. This may lead to a loss of creativity, innovation and investment in product quality in the long term with negative consequences for everyone.

5.4 NTTMs and the Legal Requirement of Distinctiveness

The preceding sections discussed the approaches in the United States and European Union on assessing distinctiveness. In Singapore, the highest appellate court, the Court of Appeal, has observed that ‘the criteria for assessing distinctiveness character are the same for all categories of marks. Nevertheless, the perception of the relevant public may not always be the same for all categories of marks, and it may therefore be more difficult to establish distinctive character in relation to some categories (such as shapes, colours, personal names, advertising slogans and surface treatments) than others’.Footnote ⁵⁵ Regarding colour, for instance, one has witnessed protracted litigation by Cadbury to protect the use of the colour purple for its chocolates, with courts being cautious about granting too wide a monopoly should the trademark registration be approved.Footnote ⁵⁶

A mark’s inherent distinctiveness must be assessed by reference to the goods or services in respect of which registration is sought as well as the perception of relevant persons, namely the consumers of those goods or services, who is deemed to be reasonably well-informed, observant and circumspect.Footnote ⁵⁷ The critical question to ask is whether the average consumer would appreciate the trademark significance of the mark in question without being educated that it is used for that purpose.

It is surmised that the most significant objection to NTTMs as registered trademarks relates to the subsistence (or absence) of the element of distinctiveness, which is the cornerstone of a trademark registration. After all, the primary function of a trademark is that of an indicator of origin and distinctiveness, that is, the attribute in a trademark to distinguish goods or services having one business source from goods or services having a different business source. Conventional signs such as words or figurative devices, in particular coined or invented words and unique, one-of-a-kind logos, are examples of trademarks that possess the innate ability to distinguish goods or services. The law refers to such trademarks as possessing inherent distinctiveness.

In contrast, except in rare circumstances, NTTMs are generally found not to be inherently capable of distinguishing the relevant goods and services because it has been widely acknowledged by the courts that the average consumer is not normally accustomed to perceive these marks, such as shapes and single colours, as badges of trade origins before the products have been sold in the marketplace for some time. In other words, the average consumer does not, by reason alone of the mark (a shape, sound or a colour), perceive the goods in question as originating from a particular trader. To qualify for trademark protection, proprietors of NTTMs should thus adduce evidence to show that these signs, though lacking in inherent distinctiveness, have nonetheless acquired distinctiveness through extensive use and substantial market investments.

Indeed, Article 15.1 of TRIPS acknowledges the important distinction between signs that are inherently capable of distinguishing the relevant goods and services and those that are not. In the case of the latter, Article 15.1 provides that member states may make registrability dependent on distinctiveness acquired through use. The concept of ‘distinctiveness acquired through use’ and the applicable test for factual distinctiveness are thus the linchpin of the registrability of NTTMs and constitutes the basis of the validity of the registrations.

Based on this, this chapter now turns to the critical question of distinctiveness of NTTMs and the exposition of the test for acquired distinctiveness by the courts in Singapore and the United Kingdom in recent NTTM decisions with particular attention to shapes.

5.5 Conclusion

Because of their nature, there are unique and serious competition-related and expression-associated problems, which may result from affording overbroad protection to NTTMs. With the various safeguards identified and mentioned, courts around the world have tried, on the one hand, to meet the needs of business for novel branding and marketing strategies in the form of NTTMs (at least in shape marks), while, on the other hand, minimising the ills that NTTMs bring about so as to strike a balance in order to ensure market competition, minimise the overlaps of IP rights and prevent abuses of the trademark system. In our view, the sensible solution adopted by the Singapore Court of Appeal with the ‘reliance’ test strikes the correct balance in protecting trademarks while also safeguarding the general trademark principles and the objectives of trademark policy.

6.1 Introduction

Fashion is not the first industry that comes to mind as depending on patent law. Other, higher-tech industries like semiconductors and pharmaceuticals occupy much more of the time of patent lawyers and judges. Yet every year, large apparel manufacturers apply for and obtain patents, and every year patent offices issue thousands of patents in ‘wearing apparel’, ‘weaving’ and other categories relevant to the fashion industry.Footnote ¹ Many of these patents are directed to the sorts of fashion that seem obviously technological or routinely subjected to technological innovation, like athletic wear or watches. Others are directed towards new materials and construction techniques that can be used in a variety of garments. But still more are directed to garments that employ traditional methods of design and construction.

This chapter examines the fashion industry’s uses of utility patents. It first provides an overview of the fashion industry’s patenting activities, looking at how many patents are granted for fashion-related inventions and how large apparel companies obtain those patents (Section 6.2). It then examines companies’ use of patents, looking at how patent litigation in the fashion industry compares to other industries (Section 6.3). Finally, it turns to the uneasy tension in fashion between function and aesthetics, examining the specific technologies and inventions being claimed in fashion patents to see what can be inferred about the relationship between fashion and intellectual property (Section 6.4).

Before we dive in, a quick note on scope. This chapter focuses on utility patents, the type of patent most people think of when they think ‘patent’. Utility patents cover inventions – ‘any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof’, as it is put in US law.Footnote ² Utility patents are not, however, the only kind of patent; in addition to plant patents (which cover, well, plants), in the United States there are also design patents, which cover ‘any new, original and ornamental design for an article of manufacture’.Footnote ³ Similar rights are protected in other countries under other names, like ‘registered designs’. Whatever the name, design protections are a more conventional fit for fashion, since they can apply to purely aesthetic creations without any showing of utility. Those rights, however, are covered elsewhere in this volume; our focus here will be utility patents only.Footnote ⁴

6.2 Obtaining Patents: Patenting Fashion Inventions

6.2.1 Patents by Classification

Fashion may not be the most obvious industry to rely on patents, but numerous technologies are critical to the industry. Apparel itself has for centuries reflected technological change; otherwise, today clothing would be made without synthetic fibres, modern dyes or even the sewing machine. Some of these technological developments are relevant to almost every fashion house; consider new fabrics, new fabric dyes and treatments or new methods of construction. Others are specific to market segments like athletic wear or jewellery.

For an initial sense of how important patents are to fashion, we can look at how many patents are granted that appear to relate to fashion. This is not a trivial task, for a few reasons. Patents can at times be written in impenetrable language only occasionally resembling English. Even when read by a patent practitioner, though, it is not always obvious what are the different applications an invention might have. For instance, the original US patent for hook-and-loop fasteners (more commonly known as Velcro) mentioned use as ‘closing devices for clothing, blinds or the like’ but didn’t mention the countless other uses to which Velcro has been applied.Footnote ⁵ So it is impossible to get a definitive list of fashion-related patents.

Helpfully, the patent offices of different nations have agreed on standard classification schemes that make it possible to filter patents for those related to particular technologies. These classification schemes have numerous limitations: many of the categories are broad, the schemes have changed over the years and they rely on the patent applicant or patent examiner to identify a particular use as a possibility. Perhaps most critically, most patents fall within multiple technology classifications, and the fashion-related classifications may be minor, niche or hypothetical uses of a technology. For instance, US Patent No. 7,627,451, titled ‘movement and event systems and associated methods’, was granted to Apple Inc. in 2009. The patent explains that the invention ‘relates to sensing systems monitoring applications in sports, shipping, training, medicine, fitness, wellness and industrial production’ and includes 76 separate technology classifications, one of which – A43B3/34, for ‘footwear characterised by the shape or the use with electrical or electronic arrangements’ – relates to fashion. If Apple ever made use of the invention covered by this patent, it apparently did not do so in footwear. Yet the patent comes up in a search for patents falling within fashion-related classifications. Still, despite these limitations in using technology classifications and their potential to produce overly broad results, they can provide a helpful overview of fashion-related patenting activity.

To examine fashion-related patenting activity, the most comprehensive classification scheme – the Cooperative Patent Classification, which is maintained jointly by the US Patent and Trademark Office and the European Patent Office – was used to identify technology classifications related to fashion. The CPC contains approximately 250,000 classification entries, which are organised into nine sections (like ‘human necessities’ and ‘physics’) that are divided into classes, subclasses, groups and so forth. Ten classes were identified as being most likely to cover most fashion-related inventions:Footnote ⁶

• A41: Wearing Apparel

• A42: Headwear

• A43: Footwear

• A44: Haberdashery; Jewellery

• A45: Hand or Travelling Articles [which includes umbrellas, purses, luggage, handbags, and walking sticks, among other things]

• D01: Natural or Man-Made Threads or Fibres; Spinning

• D02: Yarns; Mechanical Finishing of Yarns or Ropes; Warping or Beaming

• D03: Weaving

• D04: Braiding; Lace-Making; Knitting; Trimmings; Non-Woven Fabrics

• D05: Sewing; Embroidering; Tufting

Each of these classes contains numerous specific classification entries. For instance, A41, the group covering ‘wearing apparel’, contains six subclasses, including ‘shirts; underwear; baby linen; handkerchiefs’; ‘corsets; brassieres’; ‘outerwear; protective garments; accessories’ and so forth. Each subclass contains dozens of specific classifications like ‘collars with supports for neckties or cravats’, ‘reversible collars’, ‘chemically stiffened collars’ and so forth.

The number of US patents granted each year containing any classification entry falling within one of these ten classes was determined from the US Patent and Trademark Office’s PatentsView database.Footnote ⁷ The results show a general upward trend in the number of patents granted each year, with recent years showing more than 5,000 patents a year (Figure 6.1):

Figure 6.1

Upward trend in fashion-related patents.Footnote ⁸

Figure 6.1 Long description

The lines represent the A 41, A 42, A 43, A 44, A 45, D 01, D 02, D 03, D 04, D 05, and any fashion-related classification. A 41 through A 45 originates at (1976, 500), (1976, 400), (1976, 300), (1976, 200), and (1976, 100), fluctuates and terminates at (1976, 1700), (1976, 1000), (1976, 800), (1976, 700), and (1976, 600). Similarly, D 01 through D 05 originates at (1976, 300), (1976, 200), (1976, 150), (1976, 100), and (1976, 50), fluctuates and terminates at (1976, 350), (1976, 300), (1976, 250), (1976, 200), and (1976, 100). Any fashion-related classification originates at (1976, 2000), fluctuates in an increasing trend, peaks at (2021, 5400) and drops at (2021, 5900). The values are estimated on the y-axis.

Credit: Roger Allan Ford

This trend, though, does not appear to be due to any growing importance of patents to the fashion industry. Instead, it appears to track the general growth in the number of patents issued each year, which increased from around 70,000 patent grants (in the United States) in 1976 to around 350,000 in 2020.Footnote ⁹ Indeed, the proportion of all patents with any classification entry from the ten fashion-related classes has fallen from just over 3% in 1976 to around 1.5% in 2020, suggesting that fashion has relied comparatively less on patents than have other industries over the last four decades.

Still, thousands of fashion-related patents a year is not a trivial number; it represents around 100 fashion-related inventions a week. It may seem like a lot for a field that isn’t especially technological. But compare it to the number of patent grants in 2022 in some other technology classes (Table 6.1).

Table 6.1Patent grants within selected technology classes (2022)

Table 6.1Patent grants within selected technology classes (2022)

CPC Class	Grants
A61: Medical or Veterinary Science; Hygiene	40,327
B33: Additive Manufacturing Technology [e.g., 3D printing]	2,658
B60: Vehicles In General	18,428
C01: Inorganic Chemistry	2,427
C02: Treatment of Water, Waste Water, Sewage, or Sludge	1,544
C07: Organic Chemistry	11,046
C08: Organic Macromolecular Compounds; Their Preparation or Chemical Working-Up; Compositions Based Thereon	6,717
G06: Computing; Calculating or Counting	83,820
G11: Information Storage	6,206
G21: Nuclear Physics; Nuclear Engineering	731
H03: Electronic Circuitry	7,122
H04: Electronic Communication Technique	69,774
H10: Semiconductor Devices; Electric Solid-State Devices Not Otherwise Provided For	9,077

Many of these classes cover far more patents than the fashion-related classifications. That is not surprising, since some of these classes are quite broad (A61, for instance, covers inventions as varied as surgical techniques, dental implements, veterinary tools, cardiac stents, devices to accommodate the disabled and radiation treatments, among others). But as Table 6.1 shows, even narrowly drawn classes in niche areas like wastewater treatment (C02) and nuclear physics (G21) have similar numbers of patents granted as the fashion-related classes.

Patents in the fashion-related classes are obtained by a wide array of applicants, ranging from expected firms like fashion houses and manufacturers to individual inventors and firms with no obvious link to fashion. Although patents are issued to inventors – individual humans – rather than companies, most patents are assigned to a company when issued, and that information is listed on the face of the patent. Between 2000 and 2022, 78,212 patents were issued that included one or more fashion-related classification entry, with 36 different firms accounting for 100 or more patents (Table 6.2).

Table 6.2Companies with more than 100 patents in fashion-related technology classes (2000–2022)¹⁰

Table 6.2Companies with more than 100 patents in fashion-related technology classes (2000–2022)

Company	Grants
Nike, Inc.	3,273
The Procter & Gamble Co.	707
L’Oreal	684
YKK Corporation	631
Kimberly-Clark Worldwide, Inc.	575
3M Innovative Properties Co.	509
Brother Kogyo Kabushiki Kaisha	505
E. I. du Pont de Nemours and Co.	458
Apple Inc.	331
Toray Industries, Inc.	251
adidas AG	227
Under Armour, Inc.	209
Honeywell International Inc.	196
Milliken & Company	194
Groz-Beckert KG	187
Illinois Tool Works Inc.	182
Toyota Jidosha Kabushiki Kaisha	181
Maschinenfabrik Rieter AG	181
Shima Seiki Mfg., Ltd.	177
Salomon S.A.S.	166
Reebok International Limited	146
Trutzschler GmbH & Co. KG	143
Janome Sewing Machine Co., Ltd.	134
Unicharm Corporation	134
Velcro Industries B.V.	133
Samsung Electronics Co., Ltd.	121
Johns Manville	120
Sony Corporation	120
Otter Products, LLC	112
Invista North America S.a.r.l.	109
Kao Corporation	108
BASF SE	108
The Boeing Company	104
Teijin Limited	103
Braun Gmbh	102
Dow Global Technologies LLC	101

In addition to apparel manufacturers, these firms include electronics makers like Apple, Samsung and Sony; consumer-goods makers like Procter & Gamble and Kimberly-Clark; materials and chemical companies like 3M, du Pont and Dow; even large industrial manufacturers like Honeywell and Boeing. Some of these non-fashion firms are patenting inventions with fashion-related applications, like inventions related to the Apple Watch, while others are patenting materials that happen to overlap in the fashion-related technology classes, like novel fabrics for aerospace purposes. Still, the firms listed in Table 6.2 accounted for less than 15% of the fashion-related patents, suggesting that the vast majority of fashion-related patents are issued to small firms, individual inventors and non-fashion firms with a small number of inventions that may peripherally touch on fashion.

Even though other technologies account for far more patents than those related to fashion, there are still thousands of fashion-related patents granted each year, and many of those patents are nevertheless granted to large fashion companies, as discussed in Section 6.2.2.

6.2.2 Large Fashion Firms

Companies in the fashion industry have a variety of approaches to utility patents. Some obtain many, while others obtain very few or none. Of those patents, some fall within the fashion-related classifications discussed in Section 6.2.1, while others relate to other aspects of their business. This diversity is seen even in the largest firms, as seen in this table of the worldwide number of patents obtained between 2000 and 2022 by apparel firms with market capitalisations of US$10 billion or more (Table 6.3).

Table 6.3Patents obtained by fashion-sector firms with market cap of more than $10bn¹¹

Table 6.3Patents obtained by fashion-sector firms with market cap of more than $10bn

Company	Country	Market Cap	Grants
LVMH	France	$386 billion	814
Nike	United States	$186 billion	5,694
Hermès	France	$171 billion	81
Christian Dior	France	$138 billion	198
TJX Companies	United States	$91 billion	0
Inditex	Spain	$84 billion	2
EssilorLuxottica	France	$81 billion	5,031
Kering	France	$65 billion	35
Fast Retailing	Japan	$63 billion	65
Lululemon	Canada	$41 billion	26
Ross Stores	United States	$40 billion	1
Adidas	Germany	$26 billion	1,772
H&M	Sweden	$19 billion	3
Prada	Italy	$15 billion	7
Burlington Stores	United States	$13 billion	0
VF Corp.	United States	$11 billion	7
Zalando	Germany	$10 billion	0

Several observations can be made about this data. First, there is a vast disparity in different firms’ patenting activities, which corresponds to differences in their underlying businesses. Just 4 of the 17 companies – LVMH, Nike, EssilorLuxottica and Adidas – account for almost 97% of the patent grants. What these four companies have in common is that their businesses are not limited to traditional apparel but include other categories of goods and services, both within the fashion industry (such as leather goods and jewellery) and in other markets. These non-apparel businesses account for most of their patents. Two of the companies, Nike and Adidas, are sportswear companies that make sneakers and a variety of sports equipment. EssilorLuxottica makes glasses, lenses and other optical goods. And LVMH is a conglomerate with dozens of luxury brands across numerous categories including wine and spirits (Dom Pérignon, Hennessy, Krug, Moët & Chandon, among others), watches and jewellery (Bulgari, Hublot, TAG Heuer, Tiffany & Co.), makeup and perfume (Acqua di Parma, Fenty Beauty, Sephora), even yachts and luxury trains. Moreover, the trend continues beyond the top four patent recipients: Christian Dior, the company with the fifth-most patent grants (and which has a complicated ownership structure that overlaps substantially with LVMH) sells a lot of perfume, which accounts for most of its patents.

Second, the number of patents obtained by firms is larger than the number of fashion-related patents obtained by the same firms, as discussed in Section 6.2.1.Footnote ¹² This may seem obvious given the first observation, but the two points are distinct, because there may or may not be a correspondence between a firm’s patenting activity and its business activity. Indeed, firms patent many technologies that come out of research without ever intending to commercialise them.Footnote ¹³ Still, large fashion firms obtain many non-fashion patents in addition to patents falling within the fashion-related technology classes. Sometimes, they obtain these patents because they are also engaged in non-fashion businesses, like LVMH’s wine and spirits business. Sometimes, these patents cover underlying technologies that may be used in fashion goods without being listed within those classes. And sometimes, these patents relate to business operations or infrastructure, even if they don’t relate directly to the firm’s products. Fast Retailing, for instance, has several patents that cover retail equipment, like one on a ‘reading apparatus and commodity sales data processing apparatus’.Footnote ¹⁴ That patent’s classification entries fall within classes G06 (‘computing; calculating or counting’) and G07 (‘checking-devices’), not any of the fashion-related classes, even though the invention is closely related to the company’s operations. Similarly, Ross Stores has one patent, which describes and claims a ‘real estate management system and method’ that has nothing to do with fashion and everything to do with running a large retail chain.Footnote ¹⁵

Third, among the long tail of firms that do not have hundreds or thousands of patents, there are still differences between different types of firms. For instance, those that sell their own goods have more patents than those who are retailers principally selling third-party goods. And this is true whether or not those brand sellers manufacture their own goods in house or rely on contract manufacturers. The pure retailers on the list – TJX (TJ Maxx and Marshalls, among others), Ross Stores, Burlington Stores and Zalando (a European e-commerce firm modelled on Zappos) – have only one patent between them, and it’s the Ross Stores real-estate patent mentioned above. Compare that to Hermès, Inditex (Zara), Kering (Gucci, Saint Laurent and Balenciaga, among others), Fast Retailing (Uniqlo), Lululemon, H&M, Prada and VF Corp., which have 226 patents combined. Some of these patents claim glasses, shoes or their components, like the companies with the most patents, and some claim business processes or technologies like retail point-of-sale terminals. But others claim traditional items of sewn apparel, like the jacket sleeves with built-in ventilation flaps covered by one Lululemon patent (Figure 6.2).

Figure 6.2

Lululemon sleeve patent.Footnote ¹⁶

Courtesy of United States Patent and Trademark Office

Fourth, other than being largely based in Europe and the United States, this collection of companies shows no particular correlation between geography and patent strategy. The same is true when breaking down patent grants by issuing nation or patent office. Across all the companies listed, 54% of patents were obtained from the United States; 25% were issued by the European Patent Office or one of the national patent offices in Europe; 13% were issued by China; and the rest came from Canada (3%), Japan (3%), Australia (2%) and a smattering of other nations (1% combined).

Individual companies also acted globally when obtaining patents: other than Inditex’s two Spanish patents and Ross Stores’ one US patent, every company that obtained patents obtained them from multiple nations. Hermès’ 81 patents, for instance, were obtained from China, Denmark, France, Germany, Spain, Taiwan, the United States and the European Patent Office. And though a few companies have obtained a disproportionate number of patents from their home nations, that home-nation bias was small compared to the bias in favour of large markets like the United States and China.

Finally, just as the number of fashion-related patents issued has increased over time, so has the number obtained by these large firms. Different firms, though, reflect different patterns, as seen in a graph of patents issued to the large fashion firms each year between 2000 and 2022 (Figure 6.3).

Figure 6.3

Patents issued to large fashion firms by year (2000–2022).

Figure 6.3 Long description

The graph plots multiple lines of which, Essilor Luxottica originates at (50, 2000), fluctuates and terminates at (2022, 380). Nike originates at (2000, 30), fluctuates, rises and terminates at (2000, 390). Addidas originates at (25, 2000), fluctuates over the horizontal axis and terminates at (2022, 120). L V M H originates at (2000, 20), fluctuates over the horizontal axis and terminates at (2022, 50). Christian Dior originates at (2000, 10), fluctuates over the horizontal axis and terminates at (2022, 0). The values are estimated on the vertical axis.

Credit: Roger Allan FordPatent counts are from Google Patents, following the same methods as Table 6.3.

While every company except Christian Dior has increased its patenting activity over time, some have increased far more quickly than others; compare, for instance, Nike and EssilorLuxottica.Footnote ¹⁷ This suggests that patenting growth is more likely to reflect individual companies’ growth, strategies and other circumstances more than broader trends in the fashion industry.

There are limitations to how much we can infer from patenting activity, since the act of obtaining a patent doesn’t say anything about how that patent is used. But patent activity does provide tentative indications that the fashion industry does make use of patents, albeit to a lesser degree than some other fields. That use is concentrated among firms that make athletic wear, watches, eyewear and other accessories more than traditional apparel manufacturers. It is also clear that not all fashion-related patents are obtained by fashion firms; plenty of other companies obtain patents that are classified, in part, in classes that relate to fashion. Likewise, not all patents obtained by fashion companies cover fashion-related inventions; others cover the full array of things these companies do, like Ross Stores’ real-estate patent. But to move beyond these inferences, it is helpful to widen the scope beyond just the act of acquiring patent rights to the use of those patents once acquired.

6.3 Using Patents: Patent Litigation

Patent scholars generally think about three broad ways that a patent can be valuable to a company: as an asset to be monetised (whether through infringement litigation, licensing or the rents that come from a legal monopoly), as a signalling tool (whether to investors, competitors, potential clients or employees or others) or as a defensive tool (in the event that a competitor brings a patent lawsuit).Footnote ¹⁸ These different sources of value can make it difficult to understand a firm’s patent strategy, since some, like litigation, are fairly transparent, while others, like licensing, may be entirely opaque to outsiders.

Still, there are ways to obtain insights about a firm’s patent strategy from its visible patent activities. Looking at litigation is maybe the simplest, since the possibility of infringement litigation is the implicit threat even in a friendly license agreement, so that a firm with significant licensing activity will generally also have significant litigation activity.

A surface look at patent-litigation activity by the large apparel companies discussed above shows a wide array of litigation activities, with some firms frequently litigating their patents and others never doing so, as seen in Table 6.4, showing the number of patents with related litigation activity.

On the whole, compared to other industries traditionally seen as depending on patents, the fashion industry obtains fewer patents and is less likely to litigate the patents it does get. For instance, the three largest pharmaceutical companies – Eli Lilly, Johnson & Johnson and Novo Nordisk – have collectively obtained more than 5,000 patents that had litigation activity, amounting to 14% of their patent portfolios. Similarly, the tech companies Apple, Microsoft and Samsung have collectively obtained more than 6,500 patents that had litigation activity, making up 11% of their patent portfolios. By these standards, Adidas is the only large apparel company that compares.

Even though these statistics suggest that patent litigation is not driving fashion firms’ patenting activities, they nevertheless likely overstate the degree to which large fashion firms assert their patents in litigation, for a few reasons. For one, the related litigation activity reflected in Table 6.4 refers to any litigation over any patent in a patent’s global family, which can be scores of patents from numerous countries. At the same time, a single patent-infringement lawsuit may assert many patents, even if only one or two are core to the dispute between the parties. Consequently, a large number of patents with related litigation may reflect only a small number of live patent disputes.

To obtain a better understanding of how fashion firms use patents in litigation, the Stanford NPE Litigation DatabaseFootnote ²⁰ and PACERFootnote ²¹ were used to identify US lawsuits in which these firms asserted patents. Although the count is not exhaustive due to older cases with incomplete files, remarkably, only twelve cases were identified in which these companies asserted utility patents (Table 6.5).Footnote ²²

These cases fall into two broad categories. The main category is lawsuits against large competitors, in which patented technologies are the core of the dispute; Adidas, for instance, has filed patent lawsuits against competing shoemakers Asics, Nike, Skechers and Under Armour,Footnote ²³ while Nike, in turn, has filed patent lawsuits against Adidas, Hummel and Puma.Footnote ²⁴ (Indeed, a significant chunk of these cases are part of a long-running dispute between Nike and Adidas, which extended beyond federal litigation to proceedings before the Patent and Trademark Office and the International Trade Commission, before it settled in 2022.)Footnote ²⁵ But considering the size of the fashion industry, there are not very many of these cases; they all appear to involve shoes or glasses. Beyond these cases, there are a few lawsuits against companies selling knockoffs, in which the asserted patents are sideshows to claims asserting trademarks, design patents, and other causes of action like unfair competition.Footnote ²⁶

The small number of patent lawsuits brought by fashion companies doesn’t necessarily mean that the fashion industry is failing to get value out of its patent portfolios. Not all the industries that are traditionally seen as dependent on patents litigate large numbers of them. Semiconductor manufacturing, for instance, is an intensely competitive industry that obtains many patents and considers them strategically important yet litigates very few of them. Intel, Nvidia and TSMC, for instance, have collectively obtained 995 patents that had related litigation, accounting for just 2% of their collective portfolios. One possible explanation is that these firms are more likely to rely on licensing – especially cross-licensing agreements with competitors – because the industry is fairly concentrated and has high barriers to entry. Since there are fewer players to worry about, it is easier to focus on resolving disputes without litigation. (It helps that semiconductor manufacturing is a classic patent thicket – a field with many overlapping patents that make it almost impossible to operate without practicing some patented invention – which gives firms a strong incentive to reach cross-licensing agreements and avoid becoming defendants in litigation.)Footnote ²⁷

The technology-centric portions of the fashion industry, like athletic wear and glasses, resemble the semiconductor-manufacturing industry in a key way: both have a small number of firms that account for most of the market share. But there are also critical differences. Fashion firms mostly sell to consumers, not businesses like semiconductor manufacturers, and they depend more on marketing and style to sell their products, compared to the continual technological innovation of semiconductor manufacturing. So, we should expect these fashion companies to rely less heavily on patents, whether through infringement litigation or cross-licensing. And that is consistent with the relatively few patent lawsuits brought by the fashion industry.

6.4 Patents and the Nature of Utility in Fashion

As discussed before, most of the patents obtained by large fashion companies don’t actually claim apparel; instead, they often claim sporting equipment or accessories or retail strategies and tools or constituent technologies for these things. But there are patents, like the Lululemon sleeve patent mentioned above (Figure 6.1), that claim traditional forms of apparel.

In addition to revealing information about the fashion industry and its strategic decision-making, these patents tell us something about the nature of utility in fashion and the relationship of fashion to different forms of intellectual-property protections. Intellectual property traditionally draws a distinction between utilitarian or functional creations – which are protected by utility patents – and creations that are creative or aesthetic or artistic rather than functional – which are protected by copyrights and design patents. Fashion, though, makes a mess of this traditional dichotomy, since even basic questions like what counts as utility versus aesthetics are debated and disputed. Fashion firms, policymakers, courts and others have responded to this uncertainty by attempting to use a wide array of protections – traditional ones like patents and copyright and more novel ones like design rights and trade dress claims within trademark law – and by using these protections in new ways, as when a designer plasters a logo all over a product to gain stronger trademark protections.

The US Supreme Court confronted these difficulties in 2017, concluding in Star Athletica, LLC v. Varsity Brands, Inc. that designs of cheerleader uniforms could be protected by copyright law notwithstanding that the uniforms were ‘useful articles’.Footnote ²⁸ At issue was whether elements of clothing designs, such as stripes, chevrons or zigzags, were protectable as aesthetic elements of an otherwise useful article of clothing. Under US copyright law, to be protectable, the uniforms’ ‘pictorial, graphic, or sculptural features’ must be capable of being ‘identified separately from’ and ‘existing independently of the utilitarian aspects of the article’.Footnote ²⁹ By concluding that the design of the uniforms could be copyrighted, the court necessarily concluded that the graphic features of the uniform could exist ‘independently of the utilitarian aspects’ of the uniform.

The court’s resolution of Star Athletica suggests that patent law would not cover those graphic features because they are separate and independent of the uniform’s utility. Indeed, in an important trademark case in 2001, the court adopted this reasoning, concluding that a company could not obtain trademark protections for a product feature – which, like copyright law, requires that the feature not be functional – precisely because that feature had been protected by a utility patent:

So, if a utility patent is strong evidence of functionality, and the graphic features of a cheerleader uniform can exist separately from the utilitarian aspects of the uniform, then those features should not be patentable (assuming functionality and utility are synonyms).

Remarkably, however, a review of patents obtained by fashion companies shows that graphic features of apparel are frequently patented, precisely because of the aesthetic characteristics of those features. These patents generally claim methods of using two-dimensional clothing design to make clothing that creates a favourable three-dimensional shape or a favourable perception of a person’s body shape. For instance, one patent obtained by VF Corp., for a ‘body-enhancing garment and garment design’,Footnote ³¹ describes the invention like this:

The patent goes on to describe and claim various methods for using shading, curved lines and other optical tricks on clothing to tweak how the shape of a person’s body will be perceived by others. For instance, if a garment has a pattern of dots, then the position of those dots could be shifted in a way that reflects the desired 3D body shape of the wearer. A viewer’s eye will assume that the dots reflect a fixed pattern and will see the tweaked pattern as reflecting the intended shape. The patent shows this concept in Figures 6.4 and 6.5.

Figure 6.4

Body-enhancing garment patent.Footnote ³²

Courtesy of United States Patent and Trademark Office

Figure 6.5

Shading patent for aesthetic effects via optical illusions.Footnote ³³

Figure 6.5 Long description

Three broad patterned stripes of cloth are illustrated vertically. The original pattern exhibits a consistent grid-like pattern. The anatomy warping shows a distorted pattern. Anatomy Warping + Shading has a pattern which is further distorted and has shading applied. The numerical values of 802, 804, and 806 are indicated with arrows in the original pattern, Anatomy warping and Anatomy Warping + Shading, respectively.

Courtesy of United States Patent and Trademark Office

And other patents describe and claim similar illusions. A European patent jointly obtained by VF Corp. and the University of California, for instance, describes and claims a ‘method for anatomy shading for garments’ that uses shading instead of pattern distortion to accomplish the same sort of optical tricks (Figure 6.6).

Figure 6.6

Anatomy-enhancing shading on jeans.Footnote ³⁴

Figure 6.6 Long description

The process shows a basic silhouette of jeans labeled original jeans, followed by a plus symbol, a grayscale template with buttocks in focus, an equal symbol, the original jean silhouette with the template's shading with bare buttocks.

Courtesy of the European Patent Organisation

The use of optical tricks to affect how clothes look on a person is, of course, not a new concept in fashion; designers have been doing similar things for centuries. Indeed, the cheerleader uniforms at issue in Star Athletica have the designs they do precisely because they create certain impressions when worn by cheerleaders, as discussed elsewhere in this volume (Figure 6.7).Footnote ³⁵

Figure 6.7

Cheerleader uniforms at issue in Star Athletica v. Varsity Brands.

Courtesy of US Supreme Court opinion and Copyright Office registration image

The combinations of colour blocking, chevrons, diagonal lines, inward-pointing angles and other features aren’t just artistic choices by the designer; they’re choices designed to make the wearer – a cheerleader – look better. As Christopher Buccafusco and Jeanne C Fromer have explained elsewhere, clothing is not generally designed to look good in isolation; it’s designed to look good when worn by someone.Footnote ³⁶

The fact, then, that companies have sought, and obtained, utility-patent protection for inventions designed to change the visual perception of a person wearing a particular garment provides additional evidence, if more were needed, that the aesthetic design of fashion has functionality. And in turn it bolsters the case, contra Star Athletica, against protecting fashion designs using copyright law or novel forms of intellectual-property protections. Instead, it suggests that utility patents may be a better fit, despite the small amounts of patent litigation engaged in by fashion companies.

6.5 Conclusion

Fashion is not an obvious fit for utility patents, since the things that make a fashion firm successful are generally not the sorts of things protectable by patent law. Yet despite that poor fit, thousands of patents are granted each year for fashion-related technologies, and large fashion companies have obtained thousands of patents between them. At the same time, evidence of how the industry uses these patents is hard to come by since few are litigated and the industry lacks the sort of cross-licensing arrangements that some more patent-dependent industries use.

The kinds of inventions patented by the fashion industry range from traditional apparel to athletic wear to fashion-adjacent technologies like business methods and retail equipment. The most interesting ones, however, provide insights into a core issue in fashion law: the line between functionality and aesthetic or artistic expression. These patents claim and protect ways of tweaking the aesthetic designs of clothing to change how they look on their wearers. The availability of utility patents for these inventions suggests that aesthetic choices in fashion do serve functional purposes, and therefore that that the law should look more sceptically upon efforts to extend copyright and other non-patent protections to aesthetic choices made by fashion designers.

7.1 Introduction

This chapter asks whether the United Kingdom’s intellectual property regime is adequate to meet the needs of fashion designers. As such, it rests on two assumptions: first, that fashion designers ought to enjoy some IP protection over their creations and, second, that fashion designers benefit from having rights to enforce against third party interlopers. Neither of these assumptions is unproblematic. In relation to the first, it is notable that design protection is generally justified by reference to a standard market failure account.Footnote ¹ However, if one engages with this justification seriously, one might query whether there is any meaningful danger of undersupply in the fashion industry that the law needs to guard against. This chimes with observations that fashion is characterised by a strong culture of design innovation that has not obviously been held back by widespread imitation and copying. Indeed, on at least one prominent account, the ‘low-IP equilibrium’ of the fashion industry may help spur innovation.Footnote ² An alternative justification for IP protection might derive from appeals to distributive justice. For instance, those concerned about the under-protection of fashion often point to the position of emerging and lesser-known designers whose works are copied by larger, more powerful, industry players.Footnote ³ But this puts into focus the second assumption, in relation to enforcement: even if the content of the law was entirely fit-for-purpose, designers would gain little if they lack the resources or knowledge to seek legal redress. Indeed, attempts to strengthen rights may backfire if the practical effect of any new or ‘improved’ regime is to provide large industry players with even more capacity to pursue smaller players, while themselves continuing to copy with impunity.

If this chapter rests on questionable assumptions, why write it at all? Part of the answer is that there is an active debate about IP protection for fashion.Footnote ⁴ There is therefore value in doctrinal work that engages closely with the existing legal infrastructure so that we properly understand fashion’s ‘law in books’. Although some scholarship – in particular, that emanating from the United States – focuses on the lack of IP protection for fashion,Footnote ⁵ in the United Kingdom, there are a number of rights that are relevant to the sector. In fact, far from being an IP wasteland, the legally astute UK designer might wonder why there are so many overlapping and, by and large, cumulative regimes that might protect their creations:Footnote ⁶ the UK registered design system;Footnote ⁷ the (new, EU-derived) supplementary unregistered design right;Footnote ⁸ the (old, UK-created) sui generis unregistered design right;Footnote ⁹ copyright; and (although outside the scope of this chapter) registered trade marks and passing off. This preponderance of regimes is part of the second reason why this topic is worth exploring: that design law in the United Kingdom is in need of a wide-ranging review.Footnote ¹⁰ One way of feeding into this process is by asking how the current system might operate for one community of designers. As we note at various points, the sort of doctrinal analysis in this chapter might be helpfully supplemented by empirical work in relation to creative norms and enforcement practices in the fashion sector.

The thesis of this chapter is as follows. UK designs law is widely regarded as being highly unsatisfactory. It has been described as ‘multi-layered, complex and lacking in logic’.Footnote ¹¹ The problems with the system have become more pronounced following the UK’s departure from the European Union. We start in Section 7.2 by summarising registered and unregistered design protection in the United Kingdom. We observe that prior to Brexit, it was often assumed that the EU’s unregistered design right offered the most appropriate form of protection for the fashion industry.Footnote ¹² Protection is short, lasting a maximum of three years, but it was said that any downsides in relation to duration were of lower concern in a fast-moving sector and were offset by the benefits of avoiding the need for registration. However, there are a number of matters that may limit the relevance of the UK’s new supplementary unregistered design right (the SUD), all of which may cause designers to consider other IP options. First, a three-year term of protection may not be suitable for designs that are intended or turn out to have longer market presence. Second, infringement of the SUD requires proof of copying, in contrast with registered designs, for which liability is strict. Finally, the lack of reciprocity between the SUD and the equivalent EU regime, now known as the UEUD, means that UK designers who would previously have enjoyed unregistered design rights across the EU no longer do so, unless the rules on first disclosure are interpreted in a way that permits simultaneous publication.

There are a number of ways that the UK government could address the issues identified in relation to the SUD. However, there are political, legal and philosophical impediments to each of these options, raising the question of what UK fashion designers will do in the absence of any (meaningful) reform. One option is to pro-actively register their designs before novelty-destroying disclosures are made. However, another option – and the focus of Section 7.3 – is to rely more heavily on copyright as part of their protection strategy. The viability of this approach increased following case law from the Court of Justice of the European Union (CJEU), which suggests that the EU understanding of a ‘work’ is incredibly broad. We observe that UK judges have sought to interpret the CDPA in a way that respects this case law, while also recognising that the CJEU’s approach maps awkwardly onto (1) the closed list of subject matter in the CDPA and (2) the policy imperative, expressed in much existing UK case law, to limit the copyright protection of designs. Our concern is that if UK judges expand the ambit of an ‘original artistic work’ too far, it may not be possible to correct for the overshoot at other points in the copyright system. That is, if we agree that certain designs (especially functional ones) are not the sort of works that copyright ought to protect, the more principled and effective way to achieve that state of affairs is for such designs to be excluded by a subject matter limitation, not for courts to seek to manipulate other rules (for instance, in relation to infringement or exceptions) to bring protection within reasonable bounds.

7.2 Protected Designs in the United Kingdom

7.2.2 Appropriateness of the UK Regime for Fashion Designers

A question for anyone considering design protection is whether to register their design, bearing in mind that registration does not preclude concurrent reliance on unregistered rights.Footnote ³⁶ As mentioned in the introduction, prior to Brexit, it was often assumed that the EU’s unregistered design right (then known as the UCD) offered the most appropriate form of protection for the fashion industry. Protection is short, lasting a maximum of three years, but any downsides in relation to duration were thought to be of relatively little concern in a fast-moving sector and were offset by the benefits of avoiding the financial and other costs of registration.

However, it is worth pausing to reflect on the operation of the (now) UEUD and its post-Brexit ‘mirror’ right, the SUD. This analysis discusses three features that might limit the relevance of these rights to fashion designers, these being features that also differentiate the UEUD and SUD from the registered system:Footnote ³⁷ first, their duration; second, the need to prove copying in order to establish infringement; and third, the rules for creation and the resulting post-Brexit challenges for designers who wish UK and EU protection.

Starting with duration, there is the question of whether a three-year term is too short. As mentioned above, a shorter right is said to be appropriate for fashion due to the rapid turnover of designs and collections, which suggests that consumer demand in fashion markets tends to be time-bound – bearing in mind there are exceptions, as exemplified by certain collectable luxury brand items such as the Hermès Birkin and Chanel 2.55 handbags. It is also uncontroversial that, within the fast fashion sector, there are copyists who can turn around knock-off versions within weeks of the public exposure of the original item on the fashion runways in Paris or Milan or on social media feeds. The business model of many such merchants is to satisfy the immediate public demand for a garment just like the one [X] wore or [Y] designed.Footnote ³⁸ Thus, there is good evidence that commercial-level copying can coincide with the market for source designs and certainly occur well within the three-year term of the unregistered right.

However, when thinking about duration, the question is not just when copying occurs – after all, the fashion cycle is also characterised by the old coming back (often to the horror of those who survived acid wash jeans or ultra low-slung trousers to begin with).Footnote ³⁹ Rather the question is the length of time for which the designer should enjoy exclusivity in their design. There are a number of matters that might inform this, including the typical commercial life of the products in question, but also the desirability of copying in order to satisfy consumer demand and support design innovation. Of course, it might be asked whether smaller and lesser-known designers – apparently key beneficiaries of an automatic right – have similar experiences to those described above. For instance, one can imagine that unlike established luxury fashion houses with celebrity designers who want to set a new trend with every runway show, emerging designers might wish to maintain consistency in their range so as to develop a clear voice or point of view as a designer. However, taken in the round, the three-year term suggests that for the UEUD and SUD, it has been decided that problematic copying – that is, that for which a legal response is appropriate – is that which occurs relatively quickly. For designers who want longer rights, there is the registered system. We can see the logic in this, although as discussed in Section 7.3, there is a fear that, on current trends, UK designers may in fact be channelled towards the turbo-charged protection offered by copyright.

Second, for both the UEUD and SUD, it is necessary to prove copying in order to establish infringement.Footnote ⁴⁰ The significance of this is illustrated by Rothy’s v Giesswein Walkwaren,Footnote ⁴¹ in which the UCD claim failed because copying could not be established. In legal circles, this case has been said to illustrate not only the benefits of registrationFootnote ⁴² but that ‘rights holders who rely mainly on unregistered rights [should] think twice about their protection strategy’.Footnote ⁴³

Rothy’s concerned shoe designs. The claimant was a Californian company that had registered a Community design for its ‘Pointed Loafer’ shoe. The Pointed Loafer was a ballerina shoe with a pointed toe and a protruding loafer-style tongue. Its upper components were made with a knitted yarn derived from recycled plastic. The claimant alleged that the defendant had infringed its registered Community design (RCD) and its UCD through the latter’s ‘Pointy Flat’ shoe.Footnote ⁴⁴ In addition to similarities in the shapes of the shoes, the upper parts of the defendant’s shoe were made from a knitted meshwork fabric produced from recycled plastic. For images of each shoe, see Figure 7.1.

Figure 7.1

(left) The claimant’s Pointed Loafer and (right) the defendant’s Pointy Flat.

Credit: Open Justice License v1.0

Rothy’s is important for a number of reasons. A key aspect of the claimant’s case was that the protected design included the knitted material on the upper surface of the shoe. To establish this for the RCD, it was necessary to show that the fabric was apparent in the images included in the registrationFootnote ⁴⁵ (it being impermissible to look to the Pointed Loafer itself to determine what the registration means).Footnote ⁴⁶ Here, HHJ Stone (sitting as a Deputy High Court Judge) was willing to zoom in on the images of the RCD in finding that the designer claimed a knitted fabric rather than a shape or textured appearance more generally.Footnote ⁴⁷ Another significant aspect of the case was that HHJ Stone had little difficulty in rejecting the attack on the validity of the RCD or in finding that the RCD had been infringed. The knitted upper stood out to the ‘informed user’ (a strange individual even when judged against English law’s other fictional legal characters). It conferred both the necessary ‘individual character’ for validity and ensured that the defendant’s shoe produced the same ‘overall impression’.Footnote ⁴⁸

For present purposes, however, the key thing to note is that although the registered design case succeeded, the claimant was unable to make out a case for infringement of its UCD because it could not demonstrate that the defendant had copied its Pointed Loafer. It was clear from documents disclosed by the defendant that its designers had used shoes of the claimant as inspiration for its Pointy Flat. However, amongst all those materials, there was no reference to the Pointed Loafer – and two of the designers gave evidence that they had never seen that design. The similarity in the material of the upper was explained as being the result of one of the designers viewing an article that described the claimant’s use of recycled plastic material in its shoes.

It would be going too far to claim that this case illustrates that, for the UCD (and, by implication, the UEUD and SUD), it is more difficult to prove copying or derivation than in copyright. As a general matter, the CJEU has clarified that Member States are entitled to apply national rules relating to the reversal of the burden of proof such that copying can be inferred from objective similarity alone in appropriate circumstances.Footnote ⁴⁹ In essence, the defendant in Rothy’s was able to present a plausible narrative that the Pointy Flat was ‘an independent work of creation’ vis-à-vis the Pointed Loafer. The plausibility of the defendant’s account – that is, that no one had seen the Pointed Loafer – was strengthened by the fact that the Pointed Loafer was a limited-edition product that was on the claimant’s website for a short period of time.

There are, moreover, copyright cases in a similar vein. Most famously, in Mitchell v BBC,Footnote ⁵⁰ the BBC was able to defend a copyright infringement action in relation to the animated characters in a children’s television programme. The claimant had pitched a show to the BBC. Subsequently, the BBC broadcast a programme featuring characters that displayed some striking similarities with the claimant’s drawings. However, the BBC prevailed as it was able to provide a convincing explanation to account for these similarities (including that characters aimed at young children tend to have certain common – and, given the idea/expression dichotomy, unprotectable – features).Footnote ⁵¹ Also important to the outcome in that case, much like in Rothy’s, was that the claimant was not able to make out a convincing case of access by the BBC animators, despite his pitch.

The point is therefore not that there is a different legal or evidential standard to be applied between copyright and designs cases but more that some combination of functional constraints, design norms and the possibility of inspiration from a common third party source in a sector where referencing and imitation are the norm mean that the requirement to prove copying acts as a much more formidable hurdle in designs law than in does in the typical copyright action.Footnote ⁵² As HHJ Stone observed in a designs case involving an allegation of copying of dress designs, ‘absent striking closeness between the designs in relation to features that were not well known’, a tribunal may not be willing to draw an inference of copying from other factors (such as the defendant’s ‘design process’).Footnote ⁵³

Rothy’s therefore helps highlight the significance of the requirement for copying and raises questions about the inferences that can be drawn in the absence of unambiguous evidence of access. Lawyers responded to Rothy’s by emphasising the desirability of designers registering their designs. We might, however, ask whether the rules on infringement undermine the stated aims of the UEUD and SUD. To put it another way, if there is consensus that a short but strong unregistered design right is best for fashion designers such that we don’t want them to have to bother with registration, should the UEUD and SUD be strengthened by removing the requirement that infringement requires copying?

This would, of course, involve abandoning the view that ‘absolute’ or ‘true’ monopolies are only justified for registered forms of IP. This view is often said to rest on the proposition that liability for independent creation can be justified because the act of registration provides public notice of the existence of a right. Whether this proposition withstands close scrutiny is open to question. For instance, it ignores the fact that registered rights can often be difficult to locate, sometimes deliberately so. Moreover, it is an idea that already has no application in the trademark sphere, where liability for passing off is not conditional on proof of copying or derivation.

However, even if agreement could be reached regarding the intellectual case for removing the copying requirement in the United Kingdom, it would be necessary to contend with post-Brexit arrangements. In particular, the ability of the United Kingdom to amend the SUD is constrained by Article 247(1) of the Trade and Cooperation Agreement between the EU and UK, which provides:

Each Party shall confer on holders of an unregistered design the right to prevent the use of the unregistered design by any third party not having the holder’s consent only if the contested use results from copying the unregistered design in their respective territory.Footnote ⁵⁴

Given the Brexit goal of ‘taking back control’,Footnote ⁵⁵ it might have been thought that the UK government would maximise its freedom to strike out in a different direction with its domestic laws, especially vis-à-vis the EU. However, when the Trade and Cooperation Agreement (which has relatively few provisions in relation to IP) is read alongside other new trade agreements (such as the Free Trade Agreement with Ukraine),Footnote ⁵⁶ we see a pattern of the United Kingdom limiting – at times significantly – its legislative freedom in the IP space. In relation to unregistered designs, while it is still open for the United Kingdom to extend the duration of the SUD,Footnote ⁵⁷ changing the rules around infringement is off the table.Footnote ⁵⁸

One explanation for the loss of regulatory freedom in Article 247 might have been to pave the way for reciprocal recognition of unregistered design rights. However, no such reciprocity has been established. And herein lies the issue for the UK designer who wishes to rely on unregistered rights in the UK and EU. Under Art 11(1) of the EUDR, the unregistered right comes into existence on the date the design ‘was first made available to the public within the Union’. However, the EUDR has not provided a clear resolution to the question of whether disclosures in other territories can nevertheless satisfy this requirement.Footnote ⁵⁹ The position in the United Kingdom is likewise unclear: although guidance issued by the UKIPO states that a SUD ‘is established by first disclosure in the UK’,Footnote ⁶⁰ the United Kingdom did not retain all the relevant provisions of the CDR on this matter.Footnote ⁶¹ Thus, it is open to question whether disclosure in the United Kingdom can establish the UEUD, and likewise whether disclosure in the EU (and indeed elsewhere) might establish the SUD.Footnote ⁶² Against this backdrop, well-advised parties are exploring ways to publish simultaneously in both jurisdictions. The success of such efforts is difficult to predict. Simultaneous here means truly simultaneous – there is no equivalent to the 30-day rule under the Berne Convention.Footnote ⁶³ The obvious step of placing a design on a website accessible in both jurisdictions may not result in simultaneous publication, given the emphasis that has been placed on the need for websites to ‘target’ a jurisdiction in order for there to be a relevant connection.Footnote ⁶⁴ In any event, for smaller market actors – said to be important recipients of automatic rights – it is almost inevitable that they will at times inadvertently lose their ability to access the unregistered design right system.

One practical response to all three issues is for designers to register their designs: this will give them protection for up to 25 years, means they do not have to prove copying to establish infringement and allows them to rely on the six-month priority period to register their designs in the UK and the EU. But as already mentioned, one of the main rationales for an unregistered right is that the knowledge and resources required for registration are outside the reach of large swathes of the fashion industry.

In sum, although the UK designs system is in need of reform, there are a number of matters that stand in the way of revising the SUD. In addition to the usual issues for law reform – limits in legislative bandwidth, challenges in reaching consensus amongst diverse stakeholders, etc – there is the fact that the United Kingdom is operating within a network of trade agreements in which it has agreed to certain standards for IP protections, including agreements that preserve numerous standards from EU law. For those who believe that fashion is under-protected by the SUD, one response might therefore be to bolster calls for fashion designers to rely on the registered system. However, another response might be to encourage designers to seek redress by reference to another unregistered right: copyright. We believe that this is far from obviously desirable, as reliance on copyright risks massively over-correcting for problems associated with the SUD and UEUD. We explore this in Section 7.3.

7.3 Copyright

In this section, we discuss our concerns about copyright playing a more prominent role in designs protection in the United Kingdom. We start with a brief examination of why copyright might be attractive to designers and why, in our view, copyright is an inappropriate IP ‘box’ for many designs. This includes discussion of case law from the CJEU that seems to have taken a very expansive approach to the definition of an original ‘work’ in European copyright law, such that many designs will have dual protection in copyright and designs law. We then discuss the reception of that case law in the United Kingdom and argue that UK judges should be slow to follow the approach of the CJEU, including due to the difficulties in sufficiently reining in rights, for instance, via infringement or exceptions, once subject matter has been expanded.

We start with the reasons why UK designers might be attracted to copyright, which are straightforward and can be dealt with relatively swiftly. In short, reliance on copyright not only avoids issues in relation to the SUD but also the registered design right. Copyright protection is automatic and, given the principle of national treatment that underpins the international copyright system, makes the time and place of creation (or first public exposure) less significant. To the extent that there are concerns about the short duration of the SUD, such concerns clearly do not apply to copyright, as the default rule for artistic works is that copyright subsists for the life of the author plus 70 years.Footnote ⁶⁵ This is also far longer than the maximum term of the registered right. Finally, although infringement of copyright requires that the defendant copied the claimant’s work, UK judges are willing to draw inferences about this based on the level of objective similarity between and the character of the two works and the defendant’s familiarity with the claimant’s work.Footnote ⁶⁶

Our concern with recourse to copyright is that copyright provides an unduly long and inappropriate form of protection over designs in a range of fields, including fashion. For avoidance of doubt, we are not saying that fashion should be excluded from copyright protection. We can imagine haute couture and avant garde creation in which the craftmanship and visual appeal are such that copyright protection is appropriate (or at least defensible). We can also imagine instances in which individual features might justifiably be protected by copyright, for instance, where the image on a T-shirt is an original drawing.Footnote ⁶⁷ However, we have concerns about clothing and footwear being protected by copyright, for reasons of doctrine and policy including that (1) the creative choices in fashion design are limited by the nature of the item and the function it serves; (2) clothing and footwear are basic human needs, which only reinforces the need for restraint in the exclusive rights thrown around such creations; (3) to the extent clothing goes beyond basic needs and is a vehicle for personal expression (of the designer, wearer, or both), imitation is baked-in to an industry known for being cyclical, referential and trend-driven; and (4) the designs of younger designers may enjoy a copyright term that is 5 times as long as the maximum protection of a registered design and over 40 times as long as the unregistered design right.

Some of these concerns can be illustrated by reference to the ballerina shoe, the subject of the Rothy’s case, discussed in Section 7.2. Expert evidence was given in relation to the ‘basic anatomy’ of ballerina shoes, their design heritage, their ‘important characteristics’ (e.g., toe shape, heel height, outsole profile, vamp and trim) and the types of materials they comprise (e.g., leather, suede and canvas).Footnote ⁶⁸ On the one hand, one can see the potential for creative decision-making in relation to each element. On the other hand, the product has to function as a shoe. Although the designer has choices, many relate to how the characteristics of the shoe manifest, not whether those characteristics are present to begin with. Furthermore, when thinking about each such manifestation, the designer may have to work within fairly limited parameters. For instance, the toe shape of a ballerina shoe can, according to the expert evidence in Rothy’s, be pointed, square, round or almond shaped.Footnote ⁶⁹ It may be that there are infinite, minute gradations between and within these options – for instance, a pointed toe can be sharper or more rounded.Footnote ⁷⁰ But ultimately, designers are using their creativity to make choices that are bounded to greater and lesser degrees by practical and technical considerations.

In Rothy’s, the elements of the claimant’s Pointed Loafer admitted by the defendant to be different from the relevant design corpus were the heel height, the material of the upper and the presence of a counterline.Footnote ⁷¹ These were held by HHJ Stone to be enough for the shoe to have a different overall impression to the informed user, meaning that the RCD (and, under the same reasoning, UCD) was valid. What might have happened if the claimant argued that the Pointed Loafer was an original work? To explore this in the European context, we consider two design-oriented judgments from the CJEU: Cofemel v G-Star RawFootnote ⁷² and Brompton Bicycle.Footnote ⁷³ Those cases concerned the definition of a work under the Information Society Directive (ISD). To understand the significance of those cases, it is useful to understand the legal backdrop when they were handed down.

Prior to Cofemel, there had been a long line of CJEU case law stating that, to be protected, a work must be original in terms of being its ‘author’s own intellectual creation’.Footnote ⁷⁴ In applying this standard, relevant questions included whether the author was able to make ‘free and creative choices’Footnote ⁷⁵ or had, through their choices, put their ‘personal touch’ on the work.Footnote ⁷⁶ In Levola Hengelo,Footnote ⁷⁷ handed down ten months prior to Cofemel, it was held that ‘two cumulative conditions must be satisfied for subject matter to be classified as a “work”’:Footnote ⁷⁸ first, that it passes the European originality standard and, second, that it is ‘the expression of the author’s own intellectual creation’,Footnote ⁷⁹ being an expression that is ‘identifiable with sufficient precision and objectivity.’Footnote ⁸⁰ Applying this test, it was held that the taste of a food product – in that case, a spreadable cheese dip – was not a work as it ‘cannot … be pinned down with precision and objectivity’.Footnote ⁸¹

Levola Hengelo generated numerous questions about the European approach to the meaning of ‘work’.Footnote ⁸² On the one hand, the CJEU recognised that, as a party to the WCT, the European Community must comply with Articles 1 to 21 of Berne.Footnote ⁸³ It referred to the definition of ‘literary and artistic works’ in BerneFootnote ⁸⁴ and statements in the WCT and TRIPS that copyright may be granted to expression but not to ideas.Footnote ⁸⁵ However, in considering the significance of the international copyright framework, the CJEU focused on the need for the claimed expression to be objectivity identifiable, not the nature of the expression to begin with. That is, the problem with the taste of food was not that it was outside the subject matter in Article 2 but that it was not sufficiently certain: people experience taste very subjectively, and there is currently no technology by which a machine can make that assessment objectively.Footnote ⁸⁶

Our concern with this approach is that, if taken seriously, it suggests the only thing standing in the way of tastes being ‘works’ is the lack of technology to objectively measure them – that there was no separate reason, coming from doctrine or policy, for them to be excluded from copyright protection. In his Opinion in Levola Hengelo, AG Wathelet discussed three additional reasons why tastes of foodstuffs are not works: (1) all the illustrative subject matter in Berne Article 2 are works that are perceived visually or aurally;Footnote ⁸⁷ (2) no multilateral treaties have been made or revised to confer copyright protection on tastes (as has occurred for other contentious subject matter, such as computer programs);Footnote ⁸⁸ and (3) the recipe as such of a foodstuff is not protectable, being an idea.Footnote ⁸⁹ However, the Grand Chamber focused mainly on another proposition discussed by AG Wathelet: (4) that the expression needs to be sufficiently precise.Footnote ⁹⁰ It also seemed to depart from AG Wathelet’s observation that originality should be considered in addition to whether the subject matter is a work, that is, that they are distinct concepts and should not be conflated.Footnote ⁹¹

This brings us to Cofemel, where the relevant legal uncertainty related to whether Member States could include additional requirements for subject matter to be a protected ‘work’ or otherwise have their own gloss on that question. G-Star had instituted proceedings in Portugal alleging that, inter alia, Cofemel had infringed its copyright in its ARC design for jeans and ROWDY design for sweatshirts and T-shirts. The Portuguese copyright statute only protected works of applied art and design ‘which constitute an artistic creation’.Footnote ⁹² The Supreme Court of Portugal referred to the CJEU two questions in relation to the permitted content of local subsistence rules: was it compatible with the ISD for a Member State to protect designs that ‘create their own visual and distinctive effect from an aesthetic point of view’ or to require that designs be ‘artistic creation[s]’ or ‘work[s] of art’.

The CJEU repeated the two-limb definition of work set out in Levola Hengelo.Footnote ⁹³ In doing so, it observed that there will be no originality where ‘the realisation of a subject matter has been dictated by technical considerations, rules or other constraints, which have left no room for creative freedom’.Footnote ⁹⁴ We pause here to note that if this is intended as exhaustive of the situations in which technical imperatives oust originality, and is read literally, it would seem to operate in very narrow circumstances. To return to the ballerina shoe in Rothy’s, it would be going too far to suggest that matters such as the shape of the toe or the height of the heel are dictated solely by technical considerations: there may be a range of choices that all result in functional shoes, and aesthetic matters might help mediate the choice between them. The CJEU also observed that if the two limbs in Levola Hengelo are met, the work ‘must, as such, qualify for copyright protection’.Footnote ⁹⁵ This suggests that additional requirements are not permitted, for instance, that a work be an ‘artistic creation’. Finally, the case suggests that Member States should take care in adding their own gloss by rewriting or reimagining European requirements. In the latter stages of the judgment, the CJEU observed that the Berne-compatible approach in Europe was that copyright and designs give separate and cumulative protection, albeit that concurrent protection ‘can be envisaged only in certain situations’.Footnote ⁹⁶ However, it said that copyright protection may not be linked to the ‘aesthetic effects’ produced by a design as (1) such effects produce intrinsically subjective internal experiences and (2) just because a design generates an aesthetic effect does not mean that it constitutes its author’s own intellectual creation.Footnote ⁹⁷

Taken to its logical conclusion, the approach in Cofemel suggests that any requirement that a work be ‘artistic’ is straight-out incompatible with European copyright law, unless a convincing way can be found to distinguish artistic nature from an aesthetic effect. Furthermore, the CJEU’s two-limb approach suggests that, under European copyright law, a work is an objectively certain expression that is the result of an intellectual process in which free and creative choices are made. In such a world, it is not obvious that the design of a Pointed Loafer is outside the scope of copyright protection as being the wrong sort of expression. Furthermore, the CJEU stated that the degree of protection afforded a work ‘does not depend on the degree of creative freedom exercised by its author’ – that is, works involving fewer creative choices do not get ‘inferior’ protection.Footnote ⁹⁸

As noted above, originality may provide some handbrake to the protection of functional and utilitarian items. However, this will depend on the circumstances in which courts find that technical considerations oust creative choices. This was considered by the CJEU in the Brompton Bicycle case. Brompton bicycles are marketed by an English company and contain a mechanism that allows them to exist in three different positions: folded, unfolded and on stand-by. Because of their compact size, they are attractive for those who need to take their bicycles onto England’s excellent public transport system as part of their commute. The question was whether Brompton could bring an action in copyright against a competitor who produced its own folding bike that resembled Brompton bicycles. The CJEU clarified that a work may be original despite its realisation being ‘dictated by technical considerations’, so long as these considerations have not ‘prevented the author from reflecting his personality in that subject matter, as an expression of free and creative choices’.Footnote ⁹⁹ It stated that the existence of other shapes that achieve the same technical result is not decisive.Footnote ¹⁰⁰ The CJEU left it to the national court to apply this guidance to the facts. The clear lesson is that copyright can protect functional shapes.

The approach in Cofemel and Brompton Bicycle conflicts with UK law in a number of ways. First, it sits awkwardly – although not necessarily irreconcilably – with the UK’s closed list of protected subject matter.Footnote ¹⁰¹ We say ‘not necessarily’ because a closed list might nevertheless permit expansive definitions that end up including all the subject matter that meets the CJEU two-limb test. However, when we look at the drafting of the UK closed list and interpretations developed by judges, we struggle to see a principled and intellectually compelling way that the UK approach can be completely aligned with that in Europe without the CDPA being amended.

In the CDPA, artistic work is defined exhaustively – it means ‘(a) a graphic work, photograph, sculpture or collage, irrespective of artistic quality, (b) a work of architecture being a building or a model for a building, or (c) a work of artistic craftsmanship’.Footnote ¹⁰² A conflict can immediately be seen between the European approach and the category works of artistic craftsmanship, unless the word ‘artistic’ – which in the UK has traditionally required some visual or aesthetic appealFootnote ¹⁰³ – is interpreted to refer to the nature of the work, for instance, that the relevant expression has a visual rather than a semiotic function. A further conflict arises when one considers interpretations in the case law. As mentioned earlier, the UK struggled for many decades with the boundary between copyright and designs. One cause was case law in which plaintiffs were able to successfully claim copyright in utilitarian objects such as the moulds for the heating plates of a sandwich toaster.Footnote ¹⁰⁴ These cases seemed to take very seriously the qualification that things like sculptures and engravings are protected ‘irrespective of artistic quality’, such that the focus was almost exclusively on the method of creation.

There was pushback against this trend, including because it was allowing the development of a definition of ‘sculpture’ that was entirely divorced from the ordinary meaning of the term and causing a significant expansion in the subject matter protected by copyright.Footnote ¹⁰⁵ Judges therefore sought to read some limiting principles into words like ‘sculpture’, for instance, by agreeing that a sculpture is not any work made through sculpting but one that is ‘made by an artist’s hand’.Footnote ¹⁰⁶ The meaning of sculpture was considered by the UK Supreme Court in 2011 in Lucasfilm v Ainsworth, which concerned the copyright status of Stormtrooper helmets and other artefacts made for the Star Wars movies.Footnote ¹⁰⁷ A key principle affirmed by the Supreme Court was that while judges should not assess artistic quality (this being precluded by the text of the CDPA), it was entirely acceptable to consider artistic purpose.Footnote ¹⁰⁸ The Supreme Court therefore endorsed the legal analysis of the trial judge, Mann J, who developed a list of factors relevant to the meaning of sculpture, the key elements of which were that (1) the ordinary usage of sculpture is relevant but not determinative, (2) artistic worth is not relevant and (3) the essence of a sculpture is that it is intended to have visual appeal (although it may also have other purposes).Footnote ¹⁰⁹ The Supreme Court observed that this definition was supported as a matter of policy:

The Court stated that it would ‘not accord with the normal use of language’ to describe a helmet used in the making of a film as a ‘sculpture’ and that the Stormtrooper helmet ‘was utilitarian in the sense that it was an element in the process of production of the film’.Footnote ¹¹¹ As such, the Court was not prepared to disturb the conclusions of the trial judge that the artefacts were not sculptures, as that term is used in the CDPA.Footnote ¹¹²

UK courts have now begun to explore the question of how the two-limb test in Cofemel maps onto the closed-list approach in the CDPA and the meaning of categories such as sculpture and works of artistic craftsmanship. For instance, in Response Clothing v Edinburgh Woollen Mill, which was handed down during the Brexit transition period, it was held that a wave fabric pattern constituted a work of artistic craftsmanship irrespective of whether the relevant test was the British one or the European test in Cofemel.Footnote ¹¹³ HHJ Hacon did not, therefore, need to determine what would happen if the fabric was a work under Cofemel but not a work of artistic craftsmanship as traditionally understood in the United Kingdom. That said, he expressed the view that the task of the British judge is not to determine whether the ISD has removed all gaps in protection; rather, the judge must ask whether it is possible to interpret the CDPA in conformity with the ISD. This would seem to reflect the view that there are limits to the Marleasing principle, under which national courts must seek to interpret national law in conformity with EU directives.Footnote ¹¹⁴ If it is simply not possible to interpret a local statute in accordance with EU law (bearing in mind the looser rules for statutory construction that govern that process), it is for the legislature to fix the situation.Footnote ¹¹⁵ Thinking about this in the context of copyright subsistence, it is one thing for UK judges to define the protected categories of work as broadly as possible but another for them to abandon the closed list entirely or push the boundaries of the statutory language in such a dramatic way that it amounts to much the same thing.

The position has now been further clarified by the decision in WaterRower v Liking,Footnote ¹¹⁶ which also considered the meaning of work of artistic craftsmanship. It was held that a series of water resistance rowing machines (an example of which appears in Figure 7.2) were not protected as original artistic works under the CDPA, meaning that the claimant’s action for copyright infringement failed. However, in the course of the reasoning, HHJ Forsyth concluded that each of the WaterRower iterations was a work under the European test.

Figure 7.2

A WaterRower machine.

Credit: Open Justice License v1.0

This meant that unlike Response Clothing, it was necessary for HHJ Forsyth to reach a view on the potential conflict between European and UK copyright law. On the European approach – which remained relevant as cases such as Cofemel and Brompton Bicycle form part of assimilated CJEU case law – HHJ Forsyth was not prepared to find that the technical constraints inherent in the device (for instance, that it needs runners, a seat with padding, footrests, and so forth) left no room for creative freedom, applying statements from Brompton.Footnote ¹¹⁷ In contrast, on the UK approach, he indicated that the device failed to qualify as a work of artistic craftsmanship within the meaning of s.4(1)(c) of the CDPA. HHJ Forsyth considered indications from a number of authorities, including Hensher and Response Clothing, in reaching this conclusion. The problem for the claimant was that even though Mr Duke, the creator of the WaterRower, had constructed the prototype with the requisite craftsmanship and with care and consideration as to its appearance, it would be going too far to suggest he was ever acting an artist-craftsman. To qualify as a work of artistic craftsmanship, the work has to have an ‘an artistic justification’ and not merely a commercial rationale.Footnote ¹¹⁸ With the prototopye held not to be protected, the modified WaterRowers fared no better as none of the changes were driven by any artistic rationale.

At first sight, the decision in WaterRower might seem to lessen the danger of copyright coming to play a major role in protecting fashion design. It now seems clear that, despite CJEU case law to the contrary, it will still be necessary for the work to be of an artist’s hand or have some artistic quality or appeal.

On further reflection, however, the position may not be as clear cut. One question, on which we shall not dwell, is whether a higher court might take a different approach to HHJ Forsyth in how to deal with case law such as Cofemel, being one which does much more to read away the language of the CDPA’s closed list. HHJ Forsyth attempted to find a way forward that both recognised the case law from the CJEU and paid respect to the UK drafting environment. If the goal was to achieve genuine reconciliation, his solution clearly fell short. However, we see merit in his choice to respect the words of the UK statute and to present a decision-making framework that centres the EU test.

However, even if future judges are likewise disinclined to ignore the language and interpretations of the CDPA, the decisions in Response Clothing and WaterRower are consistent with other cases that have applied a relatively lax definition of works of artistic craftsmanship. For instance, in Bonz Group, Tipping J held that handmade woollen sweaters and cardigans were works of artistic craftsmanship.Footnote ¹¹⁹ Mrs Rodwell, the principal designer of the garments, was said to have ‘brought sufficient artistry to the task’, such that the designs have ‘more than sufficient aesthetic appeal’ and Mrs Rodwell could ‘fairly be described as an artist’.Footnote ¹²⁰ In the Australian case, Coogi Australia v Hysport International,Footnote ¹²¹ it was held that brightly-coloured knitted fabric was also a work of artistic craftsmanship. The outcomes in these cases can be contrasted with that in Equisafety and Woof Wear,Footnote ¹²² in which it was held that high-visibility equestrian products, including a waistcoat, were not protected works of artistic craftsmanship. Focusing on the waistcoast, although the designer had choices in where to apply new elements to an existing waistcoat design, those elements did not ‘reflect the author’s personality as required [by Cofemel].’Footnote ¹²³ This meant that the new features did not themselves comprise a work in the EU sense, and it was further held that they were not protected under the ‘older’ UK law either, as they were not the product of a craftsman.Footnote ¹²⁴ Given the particular facts of this case, and the brevity of the reasoning, it would be unwise to suggest it has moved the dial on the copyright status of clothing in general.Footnote ¹²⁵

The upshot of the foregoing analysis is that anyone who can demonstrate that aesthetic considerations were a significant consideration throughout the creative process will be in a relatively strong position to claim copyright in their creations. In the context of fashion, the fact that a garment has some functional elements will not preclude it being protected, although the simpler the design and/or the more it is dictated by function, the less likely this will be the outcome. In that regard, we agree with the result in Equisafety, and indeed believe that judges should take great care before concluding that everyday items of clothing are protected by copyright.

Of course, even if a designer can establish copyright subsistence, this is not the end of the matter in relation to enforcement, as the claimant still needs to establish infringement and the defendant may point to an exception. However, we have concerns about relying on infringement and exceptions to deal with the inevitable overshoots in protection that will arise from a broad, European-inspired approach. One set of concerns relate to the suitability of infringement principles for functional works. For instance, in a world of ubiquitous online availability, there may be plenty of evidence to suggest access, meaning that much will turn on similarity and substantial part. We wonder whether differences that may be highly significant as a technical matter might result in works that remain highly similar as a visual matter. Another set of concerns relate to whether the current suite of exceptions will provide adequate breathing space for design innovation. For instance, the CJEU’s approach to the quotation exception suggests that it may be inapplicable in most designs cases, as the Court has defined quotation to relate to uses to illustrate a point or enter into a dialogue with the source work.Footnote ¹²⁶ This sort of intellectual engagement is likely to be lacking in most fashion design.

Finally, the more fundamental point is that if we agree that certain designs (especially functional ones) are not the sort of works that copyright ought to protect, the more principled and effective way to achieve that state of affairs is for such designs to be excluded by a subject matter limitation, not for courts to seek to manipulate other rules to bring protection within reasonable bounds. If we were to suggest any reform to the CDPA, it would be to re-instate the schema that was set up in 1988, for instance, by reviewing the decision to repeal section 52.

7.4 Conclusion

We wish to conclude our chapter with some brief remarks on post-Brexit lawmaking in the United Kingdom. The analysis presented in this chapter is far from uplifting. Designs law in the United Kingdom is in need of reform and fashion designers are one community that might benefit from a root-and-branch review of the current legislative framework (at least if one accepts the two premises set out in the introduction). However, as with much IP policymaking, we fear that positive reform is a distant prospect. The process of withdrawing from the EU saw the UK agree to limit its policy freedom, such that radical reform is no longer on the cards – but at the same time, Brexit has produced genuine friction and difficulties for creative communities. There is scope for good faith disagreement about the relative importance of policy freedom versus frictionless trade, but in the IP field at least, this trade-off has proven to be largely illusory. The terms on which the UK left the EU are such that the space for meaningful reform is limited, but the costs of the exclusion from EU-wide arrangements are real.

The irony is that in the design field, the costs of divergence are likely to provide a further reason for pushing designers towards reliance on copyright. This creates a conundrum for judges, who will have to decide whether to diverge from CJEU case law on copyright subsistence. Considered in isolation, divergence might seem an attractive option as the CJEU has taken EU copyright law in a strange and difficult to justify direction. At the same time, it is hard to imagine that a judge would not be conscious of the fact that the UK’s relationship with the EU remains far from settled and that divergence now would be undesirable as long as closer trade relations remains a real possibility.