Canterbury Law Review
In his science fiction masterpiece Neuromancer, published in 1984, author William Gibson outlines a compelling vision of what he termed 'cyberspace'. Users of this futuristic worldwide network 'jack in' using hardware that interfaces directly with the brain; they then find themselves in an environment that has similar spatial characteristics to the material world. Users navigate through cyberspace by moving their digital presence, or 'avatar', throughout three dimensions, with greatly increased freedom.
The average reader of Neuromancer today is probably struck by the dissimilarity with our current experience of the internet. The most obvious difference is, of course, that we do not plug our computers directly into our brains. At a more abstract level, however, perhaps the most interesting difference is that we do not navigate through cyberspace in the same manner. 'Surfing' the net is not usually done in three dimensions using a visual representation of the self; movement through cyberspace is more of a metaphor. However, a growing number of people might see a similarity between what William Gibson describes, and how they spend their leisure, or, indeed, work time.
Currently, most inhabitants of the 'synthetic worlds' are essentially playing games. Many, such as the immensely popular World of Warcraft are set in fantasy environments, that would be familiar to a reader of Tolkien. After buying a particular software package, and arranging the regular payment of a subscription fee, players create an avatar by selecting various options, and can then move through the synthetic world environment and socialise
with other players and 'mobs'. Subsequent activities commonly involve vanquishing monsters, performing quests, and, consequently, multiplying and enhancing their avatar's abilities and possessions.
If this was the summation of the phenomena, it would not be particularly interesting, but there are certain characteristics of synthetic worlds that have had significant follow-on effects. The first is that they are 'massively multi-player': they have large numbers of interacting participants in each game. The second is that they are 'persistent state worlds'; when a player turns off her computer, the world itself keeps on running, without her in it. The third is that the world of possessions is constructed similarly to ours; there are virtual items in most synthetic worlds that can be made, acquired and possessed for various purposes. The ability to socialise, the persistent state nature, and the ability to possess separates these games from the visually similar 'first-person shooter' or 'twitch' games, where the sole prerogative is to kill or be killed.
These distinctive characteristics have meant that synthetic worlds have had effects in the real world. The most obvious is the commercial element; this has attracted the attention of both academics and the mainstream media. The attention has mostly focussed on the amounts of real world currency that virtual items and avatars are demanding on auction websites such as eBay. While sales have ostensibly been banned by many synthetic worlds, some participants earn a full time living, with six-figure incomes, through buying and selling virtual property.
With large amounts of money changing hands, it is inevitable that there will be interest in the legal status and implications of the activities surrounding synthetic worlds. While on a legal time-scale the phenomena is still in its infancy, there has already been much academic interest in the status of virtual property In some Asian countries, where synthetic world participation has been at higher rates than in the West, interest has gone beyond the academic, with the courts upholding legal rights concerning virtual property. However, there has been little legal analysis of the issues pertaining specifically to avatars. While consideration of this area requires attention to be directed to the issue of virtual property, there are other issues that are also relevant, perhaps some of the most interesting concern the visual appearance of the avatar.
The subject of avatar appearance is not as trivial as it first appears. The limits, or lack thereof, that are placed on the visual appearance of avatars will affect the individual conception of identity, the way in which social interaction is conducted, as well as have potential economic effects. Imagine that someone appears in a synthetic world with the same name and appearance as: Mickey Mouse; Donald Trump; or even you, yourself. Alternatively, imagine the same situation, but this time the 'copycat' avatars take part in commercial activity within the synthetic world: 'you' and 'Mickey' both act as hawkers for child cyber-prostitution, and 'Donald' profits by buying, developing, and selling virtual property. Alternatively, imagine that you have spent hundreds of hours socialising in a synthetic world as a particular avatar that you have carefully chosen and developed, and one day you discover there is someone who looks exactly like you, and is named almost identically. It is obvious that Disney, the real Donald Trump, and the real you, might see something objectionable about some, or all, of the given situations. As the prevalence of synthetic worlds existing on the internet increases, these issues will, likewise, become more important.
The following account focuses on the visual appearance of avatars in synthetic worlds. In Part II, a brief account is given of the history and possible future of synthetic worlds. In Part III, a survey is given of the relevant areas of law. The questions considered in Part III are what legal remedies are of relevance, and how efficacious they might be, when an avatar adopts the same name and appearance as some other entity (be it avatar, fictional character or person). Part IV concludes the paper.
The origins of modern synthetic worlds can be traced back to pre-computer forms of entertainment such as games and literature. The concept of the avatar, and the notion of moving a projection of the self around a projection of the world, can be seen in the oldest known board game, the Royal Game of Ur. Imaginary worlds have featured prominently in literature, from the Epic of Gilgamesh, through The Odyssey, to Alice in Wonderland. Even though these stories exist as mere strings of symbols, readers commonly become immersed in the imaginative worlds created by literature. It is hardly surprising that the source material for many synthetic worlds comes from one of the most detailed feats of literary imagination, Tolkien's Middle-Earth.
An important intermediary between the aforementioned sources and the modern synthetic world are role-playing games, such as the eponymous Dungeons & Dragons. These games, which frequently take place in a Middle-Earth-like setting, involve groups of people assuming the roles of characters, who are subsequently led on quests by a 'game-master'. The games consist of descriptive dialogue between the game-master and the players (eg, game-master: 'While walking through the forest you encounter a vicious looking troll'; player: 'I run forward and attack!'), with detailed numerically based rules governing many activities, notably combat. These games introduced the theatrical idea of identification with a character, a recognisable precursor to the modern avatar concept.
As role-playing games, at their simplest, require only linguistic communication and mathematical calculation, they made a relatively rapid jump to computer technology, with the first single player game appearing in 1976. With the advent of computer networking, this concept quickly became social, with the first 'MUD' (Multi-User Dungeon) being developed in 1979. Players typed commands and received textual descriptions of the environment and other avatars, and could chat between themselves. MUD's, along with the variant MOO's and MUSH's, feature similar essential features to the modern synthetic world; they are massively multi-player, persistent worlds, often with the facility for the 'ownership' of'property'.
The natural development of the internet has also contributed to the concept of the avatar, in its wider sense of artificial identity. E-mail addresses, auction site names and chat-room identifiers all contribute to the individual's internet persona. Often, internet users will use the same identifier in many different contexts; in so doing, building up an online personality and reputation; one that may or may not be similar to their 'real' identity. Most internet users are, in this context, familiar with the concept of an artificial identity.
The modern synthetic worlds are similar to their previous textual incarnations, and internet chat spaces, but they add a graphically rich and involving three-dimensional environment, and an equally visually impressive, and customisable, identity. This additional stimulus is, perhaps, responsible for the degree of immersion, and emotional involvement felt by players, that is, in some cases, taken to pathological extremes. However, it is interesting that in the development of synthetic worlds, fully immersive 'virtual reality', in the sense of goggles and gloves, was not as necessary as emotionally involving players through wide-ranging social interaction and the ability to produce, accumulate and trade possessions. Most of the graphically outdated precursors to cutting edge synthetic worlds, even some text-based incarnations, are still operational. Initially successful synthetic worlds hardly ever 'die'. Part of the reason for this may be the level of identification that many feel with their avatars.
Synthetic worlds have evolved beyond their original game origins. Perhaps the most interesting development has been into a form of purely social space, with no systemic game element pervading all activity. Notable examples include Second Life, There, and Entropia Universe. This type of synthetic world is more closely modelled on the real world. While socialising is the chief activity in these worlds, consumption and consumerism are also present.
The worlds commonly have a direct method of exchange between in-world currency and real-world money. Usually the forms of entertainment and consumables are designed and distributed by the games publishers. Second Life, however, provides in-game mechanisms for content creation; modelling tools to form objects, and scripting tools to make them do something. It appears that, in these types of synthetic world, avatar customisation is a favourite way to pass time, and spend money.
Synthetic worlds have also been adapted, and even built, for specific serious applications. An island in Second Life was purchased, and a community established there, to aid in treatment for people with Asperger's Syndrome. The US bank Wells Fargo developed a synthetic world, Stagecoach Island, ostensibly for the purpose of teaching young people money management. Various projects exist to aid in language learning. The US military had the makers of There build a Virtual Baghdad for training purposes, that can be utilised by soldiers no matter where in the world they are stationed. Those in charge of the project were apparently so impressed with the results that they commissioned a similar project, this time encompassing the entire planet.
There have been various commercial enterprises that utilise synthetic world environments and concepts in order to enhance the online retail experience. Notably, some clothing retailers have developed systems whereby the shopper creates an avatar of him or herself, and then tries on virtual clothes before ordering their real equivalents.
While predicting the future is a perilous exercise, it does not stretch credulity to imagine a huge future growth in the creation, adaptation, and use of synthetic worlds for various purposes. Bearing in mind the commercial applications, and the astounding and rapid uptake of technologies such as Google Earth, it is not out of the question that these technologies may fuse, and produce some kind of hybrid; a place where one wanders the
globe through an avatar, walking into clothes stores in Paris or Rome, chatting to the assistant, and trying on virtual clothes (‘I’ll take two: a real one for me, and a virtual one for my avatar'). In the future, much of our participation in the internet may be through avatars and synthetic worlds, resulting in a cyberspace experience closer to what William Gibson imagined.
In order to aid the following survey of the relevant law, two concepts require further explanation. First, a description is given of what it means to create an avatar. Following this is a schema outlining the types of activity that the survey is concerned with. The subsequent survey is then given, divided into subject based subsections. These are: copyright, trade marks, passing off, appropriation of personality/right of publicity, defamation, property and contract. The focus is, where possible, on New Zealand law. Where necessary, material from other jurisdiction is examined. Legal issues concerning the internet are, by definition, multi-jurisdictional. However, in order to prove manageable, a paper of this length must either be general to the point of abstraction, or based upon a particular jurisdiction.
The method by which an avatar is formed depends entirely upon the synthetic world itself. In most synthetic worlds, users select from various options to create their avatar. They will commonly select a race, a profession, and a gender; all of which will affect their appearance. Then, perhaps, they will be able to select clothing, a hairstyle and body-shape. Other methods of avatar creation are also possible. Some games feature the ability to upload content, meaning one can have one's own (or someone else's) exact appearance. When the same degree of freedom is generally given to avatar creation in synthetic worlds as is presently given to object creation in Second Life, which can only be a matter of time, anything will be possible.
The following table presents a schema of possible situations. It assigns reference numbers, which will be used in the subsequent discussion, to various possibilities where an avatar's name or image might potentially impinge upon the interests of another:
'Copycat' is not involved in commercial activity using the avatar
'Copycat' is involved in commercial activity using the avatar
Avatar bears a similar name/likeness to a pre-existing avatar
Avatar bears same name/likeness as a pre-existing fictional character
An avatar bears a similar name, and/or likeness as a real person
The table is not a complete survey of the possibilities; other conflicts are possible. For example, a copycat might make an out-of-game use of an avatar. Also, a game might allow the creation of characters that are likely to appear as pre-existing fictional characters. These possibilities will also be considered in the discussion on copyright below; the issues they raise contain elements of the above situations, but they are not as illustrative of the distinctive problems presented by avatars in synthetic worlds.
Copyright exists as a statutory protection for the fruits of creative endeavour. Copyright is a statutory scheme. The following account will initially address the possibility of subsistence of copyright in an avatar; this is of relevance to situations (i) and (iv). It is assumed in (ii) and (v) that there is a pre-existing work in which copyright subsists. After that, the issue of infringement will be considered; this is of relevance to situations (i), (ii), (iv), and (v). In situations (iii) and (vi), there could only be an infringement if there were an intervening work upon which to base a copyright claim; there can be no copyright in one's appearance in and of itself. The copying of an intervening work would make the situation effectively indistinguishable from (ii) and (v). Finally, the distinctive situations above will be discussed.
Copyright persists in original works, as defined by the Act. The first relevant question, therefore, is whether an avatar, or a part thereof, might constitute a qualifying work. There are two relevant types of work:
1. Literary works. This includes any written work, as well as 'a table or compilation'; 'writing' includes 'any form of notation or code', regardless of the medium 'in or on which' it is recorded.
2. Artistic works. This includes a 'collage' or a 'graphic work' which, in turn, includes any 'painting, drawing, diagram', or 'plan'.
The visual form of an avatar might be considered at three different levels of abstraction. At one level, an avatar is a construction of polygons and texture-maps that appears on the two-dimensional computer screen as if it was in three-dimensional space. At a more abstract level, both the polygon construction and the texture-maps will be either separate computer data files, or part of a larger file. Even more abstractly, if the visual nature of the avatar was chosen through the selection of options, then those options will themselves form a separate data file, or will be part of a larger database of avatar appearances.
The polygon construction and texture maps that constitute an avatar might be considered to be artistic works. The polygon construction could be seen as a 'diagram', 'plan', or perhaps even a 'drawing'. The texture-map could be argued to be a 'painting', or perhaps a 'drawing'. At the same time, the underlying data files could be considered literary works, as they are written in a 'form of notation or code'. If the avatar was built through the selection of options, then the particular combination of options could be seen as a 'table or compilation', and therefore as a further literary work. It could also possibly be a 'collage' of the particular features, and therefore a further artistic work.
Deciding between these various options could prove difficult. However, it is submitted that the most natural option is to conceive of an avatar as an artistic work, or as a compilation or collage of artistic works. An analogy could be made with a drawing done in Photoshop, or a similar computer graphics program. This could also be seen as a literary work, comprising of a data file, but the natural interpretation would be to consider it an artistic work. When a designer for a game constructs a three-dimensional appearing object, it will be with specialised graphics programs. It will not, for the most part, be done by directly writing text. This conception also accords with the reality of infringement. If an avatar is created that looks exactly like Mickey Mouse, it is more straightforward to say that it is an artistic work that infringes another artistic work.
If an avatar was created by selecting options, then it is possible that the selection in its own right could be considered to be a compilation, or a collage, of the separate prefabricated, and interlocking, artistic works (eg, a hairstyle, particular facial features, a body shape). The argument used above could be used to discount an avatar as being a compilation, as that is a type of literary work. The term collage has not received any attention as of yet, and is not
defined in the Act. 'Collage' has both a narrow and wide meaning; it is either 'a composition made of pieces of paper ... etc fixed to a surface', or 'an assembly of diverse fragments'. It could be argued that 'collage' is to artistic works as 'compilation' is to literary works, and it should, therefore, receive the wider interpretation. The matter, however, requires judicial determination, and the result is highly uncertain.
The next major issue is that of originality; the Act only provides protection to 'original' works. The level of originality required is, however, relatively low; the skill and labour involved must be 'more than minimal'. The dominant question here concerns the tension caused by the interaction between the player and the holder of the copyrights in works constituting the synthetic world when an avatar is created. Essentially, the problem is that the player is merely choosing from a limited range of possibilities given to them by the software's creators.
There are noticeable similarities between this question, and the question of whether the player of a game can claim copyright in a 'screen-shot', or film, obtained during play. It is probable that in that situation, even where the user has a high degree of control over the final image, there will not be sufficient variation between one user and the next in order to substantiate a claim of originality. However, this must surely be a matter of degree. At the opposite end of the spectrum, a three-dimensional design program, such as Maya, produces stills and films which will certainly attract copyright. The differentiating factor would seem to be the level of control given to the user: in Maya there is immense control; in most games, there is far less control. Even in games, though, the gap may be closing; the rise of'machinima' surely presents a challenge to preconceived notions of game-playing and originality.
The question as to whether an avatar could be original, for the purposes of the Act, will, therefore, be a matter of degree. It is submitted that the options presented by the creation process used in most synthetic worlds, such as World of Warcraft, will probably be insufficient. The 'palette' from which creation occurs will not be sufficiently wide. However, this is not a necessary state of affairs; if the number of possibilities becomes sufficiently
wide, originality will exist. Where synthetic worlds allow the uploading of models and textures, the level of user input is more closely related to Maya, and the avatars would then be potentially original.
The next relevant issue is that of authorship. The Act allows for joint ownership of a copyright work, where the contributions of each party are not distinct. This would seem to suggest the possibility of joint ownership between the user and the synthetic world company. This issue will, however, likely be determined by the question of originality. If a user uploads an original model and texture, they will, all other things being equal, be the author of the work. In the unlikely situation where it was found that copyright subsisted, as a collage or compilation, in a particular selection of predetermined elements, then the user will likely be the sole author of that particular work, while the company retains copyright of the component parts.
However, there is also the possibility that an avatar is considered a 'computer-generated' work. A work is 'computer-generated' where it is 'generated by computer in circumstances such that there is no human author'. The author is then the 'person by whom the arrangements necessary for the creation of the work are undertaken'. The first of these concepts is difficult to pin down. In most cases, such as a Word file there is clearly a human author. The only possibility is a computer-generated work exists where the human contribution is extremely slight. It is possible that, presuming copyright exists as a collage or compilation in a choice-based avatar, the contribution is small enough for the work to be computer-generated. However, there is a problem with that analysis: if the user contribution is enough to constitute originality, it is difficult to see how it could be minimal for the purposes of attributing authorship. It seems that the issue of originality makes that of computer-generation irrelevant.
Infringement of copyright in avatars does not present as many distinctive problems as its existence. Of primary relevance is infringement through copying, which is a restricted act. In the context of copyright, it is the act itself that is restricted. This is as opposed to a requirement of commercial use of the material. There is, therefore, no material difference between situations (i) and (ii), and situations (iv) and (v), respectively. This is worth noting, as such distinctions become important when trade marks and passing off are considered later on.
'Copying' receives a definition in the Act. The definition is 'reproducing or recording the work in any material form', and includes 'storing the work in any medium by any means', as well as a 'copy in three dimensions of a two-dimensional work' and vice versa.
The definition certainly encompasses copies made in the computer context. It would also seem to cover the dimensional shift involved in creating a 'three-dimensional' avatar based on a two-dimensional character. This could occur in situations (ii) and (v). Although an avatar is not, strictly speaking, three-dimensional, it is submitted that the intention of the legislation would not be fulfilled by excluding such a situation from the definition of copying simply on that basis. An avatar would be capable of direct reproduction in three dimensions utilising no information other than that which exists as part of the data file(s). Notwithstanding the specific provision of the Act, it would be no more of a stretch to allow that an avatar could be a copy of a drawing than it is to hold a flying disc, or a chainsaw chain, is a copy of a technical drawing.
Establishing actual infringement requires that a substantial part of the work is reproduced, that there is sufficient objective similarity, and that there is a causal connection between the works. The first two parts of the test are contingent upon the particular situation, and nothing generally relevant to avatars can be said about either of them, other than to note that if the level of originality is seen as low, the protection given will be correspondingly low. The requirement for causal connection, however, makes slightly more palatable the possibility of copyright being found in an avatar created by selecting options. Presuming that copyright exists in such an avatar, situations (i) and (iv) would not be an infringement where the similarity is merely coincidental, because there is no causal connection between the two.
Above, when the table of situations of conflict was outlined, two further examples were given. The first was where someone makes an out of game use of an avatar. A possible example of this is where someone uses images of a famous avatar, a hero in the particular world, as a character in a comic book. If copyright subsists in the avatar, the issue will then be the familiar question of infringement; whether a substantial part was copied. This will, again, depend on the circumstances.
The second situation is more interesting, not least because it has, arguably, already occurred. The example concerned a dispute that occurred between Marvel Comics, the owner of the rights to such characters as 'Spider-Man', and NCsoft, the publisher of the City of Heroes synthetic world. City of Heroes is a MMORPG which involves the players assuming the role of a super-hero that is created by the player using an options based system. Marvel's objection was that players were using the software to create resultant characters that closely resembled its creations, such as Spider-Man. The dispute was settled by way of an undisclosed settlement, however NCsoft stated that the terms of the settlement do not limit the players' expression in creating original characters.
The Marvel case exemplifies the complicated relationship between the creative acts of the game designers, and those of the player. It is, effectively the reverse of the previous situation. It potentially involves issues of copyright, trade marks and passing off. In the copyright context, the design of City of Heroes could, arguably, constitute a criminal offence under the Act, as NCsoft '[made] an object specifically designed or adapted for making copies of a particular copyright work'. It might also be seen as 'authorising' the players to infringe copyright byway of copying. Authorisation is, however, a concept with uncertain boundaries.
In conclusion, copyright may be of use in situations (i), (ii), (iv), and (v). In (ii) and (v), establishing the subsistence of copyright will present no distinctive difficulty. However, in situations (i) and (v), establishing copyright will entail successfully showing that there is a relevant work. If that work is a collage or compilation, this will require additional argument, the outcome of which is somewhat doubtful. Originality will need to be shown, which may prove difficult, unless the avatar is comprised of an uploaded model and texture map. If these, rather challenging, requirements are met, authorship will likely reside with the user. Infringement, in all circumstances, will take the form of copying. A causal connection will also need to be shown. Otherwise, infringement will depend on the circumstances. The first of the other circumstances relies on similar considerations to (i) and (iv). The second, reverse, situation is quite distinctive, and may constitute a criminal offence, or authorisation of user infringement.
The trade mark system is a statutory scheme by which a trade mark may be registered in order to obtain protection from unauthorised use in trade.A 'trade mark' is a 'sign' that is capable of graphical representation, and of distinguishing the goods and services of one person from another. A 'sign' can include a 'device', or a 'name'. The sign must be registered in relation to a specific goods or services within a specified class or classes, and the registration must not be over more goods or services than is 'justified by the use ... of the sign'. A trade mark must also possess 'distinctive character'. Infringement may then occur when an 'identical or similar' sign is used in the course of trade.
Avatars comprise of both a visual and textual element; they possess both an appearance and a name. The name is perhaps the most accurate identifier of who a particular avatar is. In most synthetic worlds it is exclusive; each avatar must have a unique identifier, with a minimum and maximum length. The difference, however, might be extremely slight: Aragorn', 'aragorn', ‘AragorN' and ‘Arag0rn' might all be different users. An examination of the definition reveals that both the visual appearance and the name of an avatar could potentially be a 'sign'. The visual appearance could be portrayed as a 'device', while the identifier could be seen as a 'name'.
However, as promising as it may initially seem, there are several reasons to believe that the trade mark regime may be of marginal use in relation to avatars. Firstly, infringement will only occur where an unauthorised person uses an identical or similar mark 'in the course of trade'. This will automatically cut out the utility of registration in situations (i)-(iii); only situations (iv)-(vi) remain. Restrictions on the possibilities of the avatar's name, and/or appearance, may mean that, presuming registrability, a strict line would be taken on the interpretation of 'similarity' for the purposes of infringement. The notion that there must be use in trade will also limit the varieties of (iv)-(vi) that receive useful protection.
There is more than one way in which an avatar could be engaged in commercial activity. The user could be involved strictly in activity within the 'rules of the game'. For example, a blacksmith avatar could be producing virtual horseshoes, selling them to other avatars for in-world currency, and then using the currency to purchase a new virtual anvil, all within the synthetic world. On the other hand, an avatar might be combining in-world and real world activity. She might buy the synthetic world software, set up an account, purchase a piece of virtual land through eBay, divide it up, develop it with night-clubs/casinos, etc, and finally sell the land on eBay for real money again.
The first of the given examples is intuitively not so much 'in the course of trade', in the context of the Act, so much as 'play'. However, the second example seems as close to 'trade' as one can get. The activity is conducted entirely for profit, which is realised as a matter of fact, and partially conducted in a recognised marketplace. Even in the first example, there is the potential for selling the acquired in-world currency on eBay. Entrepreneurs have seized on the profitability of in-world labour by 'farming' out repetitive currency generating tasks, such as blacksmithing, to low wage third world residents, with the resulting in-world currency being sold on eBay. However, it could be argued that as long as the final step of a real world transaction does not occur, as a matter of fact, there is no 'trade'. There are synthetic worlds were the transition between real and in-world currency is seamless as a matter of design; in that situation, it is arguable that all in-world commercial activity is 'trade'. The result of all this is that only in certain types of case would the victims of situations (iv)-(vi) be able to prevent infringement.
Infringement may also only be found where there 'the sign is used in such a manner as to render the use of the sign as likely to be taken as being use as a trade mark'. The issues here may be similar to those discussed concerning trade. An argument could be made that the name or appearance of the avatar that delivers goods sold on an auction website does not constitute 'use as a trade mark'. The appearance of the avatar will probably not be known until after the sale of a virtual item is complete, and delivery occurs. The avatar operates more in the nature of a post-sale delivery truck.
There are also reasons why a trade mark may not be registered, registrable, or relevant to a particular use in trade. In situations (iv) and (vi), it is somewhat unlikely that the user will have registered her appearance, or name, as a trade mark. There is theoretically registrable material infringed upon in situation (vi); the Act appears to allow registration of the name or representation of a person insofar as it requires that person's written consent. Even if registration has occurred, the user must have 'put the trade mark to genuine use in the course of trade in New Zealand' within three years of the application, 'in relation to the goods or services in respect of which it is registered'. It must also not be suspended from use for a continuous period of three years. Effectively, the user will only be able to maintain registration if she intends to employ the trade mark commercially herself within the relevant class.
A trade mark is not registrable if its use would be contrary to New Zealand law. In situation (iv), if an avatar was constructed by way of choosing options, it may be a collection of copyright artistic works. Use of such a trade mark would likely be seen as contrary to New Zealand law in that it infringes copyright by way of copying.
Infringement will also only occur with relation to similar or identical goods and services, or where the trade mark is 'well known in New Zealand', and use takes 'unfair advantage of, or is detrimental to, the distinctive character or the repute of the mark'. This means that, not only must the aggrieved person have registered a trade mark, it must be used, and used in the same field. The 'well known' trade mark exception will almost certainly only be of potential use in situations (v) and (vi); it is highly unlikely that in situation (iv) the avatar name or image would be well known in its own right. Examples where this could be plausible are the previous ones of 'Mickey Mouse' and 'Donald Trump', presuming, of course, that these are registered trade marks. It will depend on the context as to whether it takes 'unfair advantage of, or is detrimental to, the distinctive character or the repute of the mark'. It could conceivably be argued that dilution of any form is detrimental to the mark.
In order for an avatar to receive protection under the Act, then, there must be a relevant registered trade mark or marks. It must be registered in relation to similar goods or services, or be well known. It must be in use in trade within a certain time frame. It must not be contrary to law, and therefore, will probably not cover option generated avatars. The situation
must be of the (iv)-(vi) type, and the commercial activity must be seen as occurring 'in the course of trade'. It can be clearly seen that any protection that might exist is greatly limited.
At root, the tort of passing off provides an action where one trader misappropriates the goodwill of another. The condensed test in Reckitt & Colman requires three elements: goodwill or reputation attached to the goods or services; a misrepresentation by the defendant to the public, whether or not intentional, that the goods or services are those of the plaintiff; and demonstration of damage by the defendant suffered by way of that misrepresentation. The initially restrained interpretation of this tort has widened over time, notably to include 'character merchandising', in the context of which it is arguable whether the traditional requirements of passing off are fulfilled.
Here, the traditional categories of goodwill, misrepresentation, and damage will be discussed, and problems identified, before a discussion of character merchandising occurs. Finally, the relevance of the Fair Trading Act 1986 will be briefly considered here, as it shares many similarities with passing off.
The requirement that there be goodwill may be fulfilled by way of commercial, professional or artistic activity, but may also be present in regard to charities and churches. However, in even the widest cases, there is still an analogy with commercial goodwill: in Day v Brownrigg it was found that the name of a house occupied by a private individual could not form the basis of a cause of action. The requirement that there be a commercial element, or analogy, means that in order to have the basis of a cause of action, this must be present in the plaintiff's activities in situations (i)-(vi).
In order for there to be something able to be protected, there must be some name, mark or get-up which represents the underlying goodwill. This could be constituted by the name or appearance concerned, whether it be an avatar, a fictional character, or a real person. Distinctiveness in the name or appearance will still be required. However, a problem that may exist in relation to situations (i) and (iv) is the 'delivery truck' one, noted above; the avatar, in and of itself, is only delivering something that is already sold to the purchaser. The avatar is not in the nature of 'get-up', in that the
purchaser will only see the avatar once the trade has occurred. Arguably, the avatar itself bears no material relation to the relevant goodwill in the activities of the trader. The 'post-sale' doctrine is not relevant here, as it is highly unlikely that anyone else will see the avatar deliver the goods. This may not be the case if the synthetic world possesses seamless in-world and real world commerce; then, arguably, there could be goodwill in the name or appearance of an in-game vendor (ie, where the sale and transfers occur completely in-world, with in-world currency that is seamlessly convertible into real currency).
Another possible problem exists in regard to situations (iii) and (vi), if it were the case that plaintiff and defendant shared the same name. It has been stated, obiter, in New Zealand that there is probably no 'own name' exception in the context of passing off. This means that there could still be an action even if the defendant is using her own name or appearance in situations (iii) and (vi), which happens to be the same as that of the plaintiff. This is, in the synthetic world context, somewhat unlikely, as there seems to be an intuitive fear of using one's own identity to represent oneself in the worlds.
The requirement that there be a misrepresentation would, it seems, be destructive to many claims concerning avatars. There needs to be a deception of consumers, causing them to be confused as to what they are getting.There does not need to be an intention to deceive; but showing this will certainly advance the claim.
Implicit in the requirement of misrepresentation is that there must be some kind of commercial activity on the part of the defendant; no-one can be mislead with regard to what she is getting if she is not getting anything. It follows that only in situations (iv)-(vi) will the plaintiff be capable of seeking a remedy by way of passing off. However, this is commercial activity, widely construed and, as previously noted, it can include charitable causes.
A significant problem in relation to situations (v), (vi), and even possibly (iv), is the context of the synthetic world. Evidence can be admissible in relation to the special features of relevant markets. Those who enter a synthetic world are cognisant of the method involved to construct an avatar; they have had to create one themselves. The implication here is that most will be very aware that if they run into 'Mickey Mouse' or 'Donald
Trump' in the synthetic world, they are probably not communicating with a representative of Disney or Trump. The test for confusion has been stated to be that there is a 'reasonable likelihood' of confusion among a 'substantial number' of people. This is unlikely to be the case in this particular market. However, situation (iv) is less subject to this effect. The visual appearance of a copycat avatar could be exactly the same, and the login name could be a slight variation, such as a modification of case; it would then seem that there was a reasonable likelihood of confusion for a substantial number of people.
In the past the existence of a common field of activity between plaintiff and defendant has been seen as a necessary requirement. However, it is now accepted that it merely gives an indication of the likelihood of confusion and damage. It will, therefore, be relevant to situations (v) and (vi) as regards the field of activity of the character or the person. In situation (iv), it is almost certain that the field of activity will be the same, regardless of what kind of virtual items or services they trade in. It seems unlikely that, even if a court was progressive enough to allow the claim in the first place, that it would subdivide the types of commerce possible in synthetic worlds (eg, into land development, consumer goods sales, sales of services, etc).
Several different types of damage have been recognised: diversion of custom; damage to reputation; and, dilution of the value of the name or get up. To a certain extent, damage may be inferred from proof of interference with goodwill and misrepresentation. In situation (iv), the likely type of damage would be diversion of custom; for example, one trader impersonates another who is known to give bargains on certain types of 'virtual items'. Another possibility is damage to reputation, which could also apply to situations (v) and (vi), if the copycat avatar was engaging in commerce of an unsavoury nature. Dilution, while probably the most controversial head of damage, could potentially apply to all three situations; although, it would be advisable, if possible, to argue it in the alternative.
As passing off is an action of traders, not consumers, it is irrelevant that there is no difference in quality of the goods or services. This is of particular relevance in relation to the sale of virtual items, as it is unlikely that there will be any difference between the 'goods' on offer.
The courts, particularly in Australia, have established that there may be an action in passing off where the goodwill in a character or a person has been misappropriated. The use of the label 'misappropriation' in the cases suggests that they are tending towards the recognition of a property right in the character, or person, rather than strictly applying the tort of passing off.
The Australian courts acknowledged, relatively early on, that passing off could be utilised where one's image was used without permission. In Henderson v Radio Corporation, a pair of dancers were successful in their action, where their image was used without authorisation on a record sleeve. Success was based on the finding that the people for whom the record was intended would believe that this indicated recommendation or approval. The applicability of this case to situation (vi) is suspect, though, as it is clear in the above discussion, that there will be no such belief.
In Hogan v Koala Dundee, the protagonist of the film Crocodile Dundee, and the producers, successfully proceeded against a company that was utilising distinctive elements of the fictional character created by the plaintiffs. Specifically, the name 'Dundee', a koala featuring the bush hat with teeth in the band and the vest and knife of the original character, were placed on signs and merchandise. The plaintiffs claimed that there was a misrepresentation that there was a connection between the enterprises. The Court answered the claim by stating that 'the inventor of a sufficiently famous fictional character having certain visual or other traits, may prevent others from using his character to sell their goods ... even where he has never carried on any business at all, other than the ... making of the work in which the character appears'. This appeared to significantly alter the cause of action, suggesting a proprietary right.
In the similar case of Pacific Dunlop v Hogan, a scene from the same film was emulated in a shoe advertisement. Here, there was something of a dampener put on the breadth of the action as stated in Koala Dundee. The need for showing misrepresentation and damage to goodwill was maintained, as opposed to simple misappropriation. However, the Court went on to find for the plaintiff, and the misrepresentation and damage that was identified revealed a rather expansive definition of those concepts.
It should be remembered that the Crocodile Dundee cases were in the context of the peculiarities of character merchandising. Although they involve the same subject matter, and therefore seem intuitively relevant to avatar situations, the issues are, in some ways, different. In character merchandising situations the character is applied to, often quite disparate, items, in order to increase the value of those items through association. In avatar situations, it is as if the salesperson is selling her own items, while dressed as, and answering to the name of, the character. A true analogy with the character merchandising situation would only occur where either: the defendant is selling virtual items in the game, such as clothing, with the character appearing on it; or in situations (v) and (vi), if the commercial activity is the sale of the avatar itself. This would, more straightforwardly, involve character merchandising issues.
However, in both the archetypal character merchandising situation, and where avatars are concerned, the underlying problem is the same: showing that there is a misrepresentation, and that there has been damage. The problem of misrepresentation can chiefly be overcome by showing that a connection is supposed by the public. The problem of damage is often addressed by arguing dilution.
An extreme example where a connection was seemingly not required is Twentieth Century Fox, where 'Duff', applied to beer, was successfully argued to make a misrepresentation that the beer was connected with the television program The Simpsons, which featured a fictitious beer of the same name. This was the case even where there was a clear disclaimer on the label. If this approach is applied to situations (v) and (vi), it would suggest that there should be a remedy despite the fact that it is clear to most people that there is no connection between the plaintiff and the defendant.
What is clear is that the closer that the protection granted by passing off approaches a property right, the more likely an avatar is likely to receive protection. The New Zealand case of Tot Toys v Mitchell, provided a simultaneous cold shower for the exuberance of the Australian cases for the usefulness of passing of, and an indication that a preferable option may be to recognise an actual property right in personality. The judgement of Fisher J questioned whether there was deception or damage in most character merchandising cases. A forceful argument was made that, in
many situations, the argument in support of the application of passing off is circular. However, comments were also made that suggested that if personality or characters were to be protected, it might be better done through recognising an action for appropriation of personality.
The Fair Trading Act 1986 provides statutory causes of action that operate alongside passing off. The Act can be used by rival traders, in order to prevent misleading practices, for their own benefit. The Act provides that 'no person shall, in trade, engage in conduct that is misleading or deceptive or likely to mislead or deceive'. It is generally thought that this provides an action that is wider than passing off.
An initial requirement is that the defendant be in 'trade'. This receives definition in the Act, and would seem to entail that, like passing off, only situations (iv)-(vi) are included within the ambit. There is some complexity to the notion of 'trade', and it may not be all commercial activities involving avatars that qualify.
The requirement that the conduct be 'misleading or deceptive' appears to raise some of the problems considered above. Centrally, in situations (v) and (vi), it is arguable whether anyone within a synthetic world would be mislead or deceived into thinking there was any connection between the avatar, and the person or character's creators. There is no obvious reason why there could not be something misleading or deceptive in situation (iv). It is also notable that damage is not a requirement of the action. This would prevent the uncertainty of having to argue dilution if there is no, more direct, way of showing damage.
Overall, passing off and the actions available under the Fair Trading Act 1986, may be of assistance in situations (iv)-(vi), where both parties are engaged in commercial activity. In the context of passing off, the plaintiff will have to show goodwill, misrepresentation, and damage. There may be difficulty in showing a relationship between the avatar and any goodwill, in situation (iv). Showing that there has been a misrepresentation may be difficult in situations (v) and (vi). The approaches taken in the character merchandising cases may be of some assistance. The ease of showing damage will depend largely on the facts; it will, however, not be required with regard to the Fair Trading Act.
There is, in some jurisdictions, a cause of action directly protecting one's image from misappropriation. The action is notably existent in Canada, by way of the tort of appropriation of personality, and in some states of the United States, by way of the right of publicity. This section briefly considers: the requirements of such an action; the potential for application to avatar situations; and, the situation in New Zealand.
The basis of the action is that there has been a misappropriation of a person's personality that causes that person damage. 'Misappropriation' entails that there has been an unauthorised use of some referent of that person's personality. The requirement of damage generally entails that there must be some commercial value, or potential value, in the relevant personality. An additional requirement is that there not be a countervailing public interest; this prevents such an action from impinging upon important aspects of free speech.
Such a cause of action could obviously be of potential benefit in situation (vi); It may also be applicable to situation (iii); but it might be more difficult to show damage because of a lack of commercial activity. Unlikely as it seems, it may also be possible that there is redress in situation (iv). The key problem here is identification between the avatar, and its 'owner'. This may seem insurmountable, but there have been cases where there is an intermediary object that seems to act as a 'vehicle' for the personality of its controller. In Motschenbacher v Reynolds, the plaintiff obtained redress for the unauthorised use in an advertisement of his actual vehicle: a distinctively marked car. The markings were found to cause the public to infer that the car was the plaintiff's, and that he was driving. If this is applied to situation (iv), it suggests that, provided that the player was of sufficient renown, and he was, in the eyes of the relevant public, connected with the player, such an action could potentially be made out.
Despite the existence of this cause of action in other common law jurisdictions, it is not yet recognised in New Zealand. In Tot Toys v Mitchell, the comment was made that the attempt to provide redress for such situations has led to a distortion of the tort of passing off, and that a better course would be to consider the justifications for providing such a remedy on its own merits. However, any development in this area seems unlikely, after the Court of Appeal made a unanimous, if obiter, statement rejecting the existence of such a cause of action in New Zealand. While this does not completely close the door to such an argument, it certainly decreases the chances of the action being recognised any time in the near future.
Defamation is a tort which protects the reputation of a person against unjustified attack. While the action is a creature of the common law, some aspects have been reformed in New Zealand by statute. Here, the legal requirements are briefly laid out, and a discussion of the applicability of each aspect of the tort to the copycat avatar situations is given. The concern is that someone will adopt an avatar that clearly refers to a certain person, and then engage in activity that brings discredit to that person. Defamation has three central requirements: a defamatory statement has been made; the statement is about the plaintiff; and, the statement has been published by the defendant.
There is no single definition of defamation, but one of several popular judicial pronouncements is that it is 'a statement which may tend to lower the plaintiff in the estimation of right-thinking members of society generally'. Potentially relevant types of defamatory statement are those which present the plaintiff as: displaying antisocial behaviour; acting dishonestly or criminally; being someone to 'shun and avoid'; and, being in financial difficulty. It is not necessary that the defendant intended to defame the plaintiff.
It is possible that statements could be made by the defendant that fit any of the above categories. Whether the actual statements made are defamatory will be a matter of fact in each case. There is also a particular problem here, with regard to situations (ii), (iii), (v), and (vi). Suppose that 'Donald Trump' was wandering around a synthetic world, making offensive statements. It would be artificial to suggest that this is analogous to the, clearly defamatory, situation where a column is published in a newspaper under the name Trump, that contains statements that, upon reading, reasonable people would take the writer to be antisocial, a criminal, etc. Instead, it is closer to the situation
where a person is standing in a Trump mask on a street corner and making the same such statements: the person is obviously not Trump. Even the second analogy is, however, not perfect. There is an inference made that the man on the street in the mask is pretending to be Trump. Arguably, there is not the same inference in the synthetic world. In the synthetic world the residents know that the person behind the avatar is not necessarily pretending to be Trump at the relevant times; they may totally disassociate the two stimuli. If this is accepted, it may be that making such comments in the synthetic world situation is not defamatory. This is, however, an overly metaphysical distinction, and may not be seen as relevant to the potential for defamation.
The above issue aside, the lack of a requirement of intention means that it will not matter if the purpose of the defendant in adopting the particular avatar and name, and making potentially defamatory statements, was to defame the plaintiff. She might have merely picked an appearance she preferred, and then acted in her typical manner.
The supposed statement must also be about the plaintiff; it must be the plaintiff who is defamed. The relevant test is whether reasonable people would reasonably believe that the statement refers to the plaintiff. The plaintiff need not be actually named in the statement. Where there are extrinsic facts necessary to link the statements to the plaintiff, it must be shown that a reasonable person possessed of the facts would make that connection. A company can be a plaintiff, where the statements affect its business or trading reputation.
In situations (ii), (iii), (v), and (vi), this will largely be a matter of fact. With regard to the fact that avatars possess a name and an appearance, the best approach would seem to be that the overall effect of both should be taken into account. The question will likely be whether the combination of image and name are such that a reasonable person could reasonably see who the avatar was 'supposed' to be. The fact that a company can be a plaintiff would potentially allow corporate owners of the rights concerned in situations (ii) and (v) to obtain relief.
In situations (i) and (iv), the issue is slightly different. There is potential for someone to create an identical appearance to the plaintiff's avatar, and then to run around the synthetic world making objectionable statements. Arguably, all that is necessary here is that users mistake the copycat for the original avatar. The fact that the original avatar is actually John Smith may be treated as an extrinsic fact; this, however, might depend on the general accessibility of that information. In many contexts, there might be no real way of knowing who is supposed to be behind the particular avatar; in other contexts, it may be common knowledge. The identification issue will then
remain: how similar are the two avatars? If the avatar creation system is option based, arguably they will need to be quite close indeed. It is also possible that the fact that all users must have a distinctive name would make identification more difficult. Most users would be aware that a slight difference in the case of the different characters in a name means a different person.
There must also be publication by the defendant, to some person other than the plaintiff. 'Publication', however, does not require that there be a large number of recipients. It is clear that publication can occur on the internet. It has also been held that the internet is not a more permissive environment with regard to defamation. Every actor in the 'chain of publication' is potentially liable to the plaintiff, but there is a defence of 'innocent dissemination'.
There are two points to note here. The first is that it will, almost certainly, not matter that the publication takes place in a synthetic world. While the number of people that may be in receipt of the defamatory material is potentially far fewer, it is still has the potential to be one or more. Communication in a synthetic world is not always a one on one affair; it is possible for conversations to be 'overheard'.
The other point concerns the potential liability of those running the synthetic world under the chain of publication rule. Chat within synthetic worlds happens in real time, and subsequently in a natural and unmoderated way. It seems probable that in most circumstances the defence of innocent dissemination will apply.
Potential liability for defamation may occur where copycat avatars engage in certain activities. In situations (ii), (iii), (v), and (vi), it may be difficult to show that the statement is actually defamatory of the relevant person, although that need not be the intended effect. Identification will depend greatly on the facts, with the name and appearance of the avatar likely to be relevant. In situations (i) and (iv), there may be additional problems linking the avatar and the user. Publication will almost certainly be made out where there is communication to others within a synthetic world.
Over the previous discussion, two spectres hang, either of which could affect a determination of possible rights in the appearance of avatars. The first is the previously mentioned idea that there might, or should, be property in virtual items and user accounts. The second is the effect of contract upon the relationships involved. These issues are briefly examined in turn.
The idea of virtual property has, almost exclusively, occupied most legal commentators on synthetic worlds. The issue is whether there is, or indeed should be, property in virtual items and user accounts. This issue is worthy of its own, extensive, treatment, and it has already received several insightful analyses. It is necessary, however, to make a passing comment in relation to the foregoing discussion of avatars.
The central question is whether the grant of property rights in virtual items and avatars would affect the positions outlined above. There is good reason to believe that it should not. Intellectual property rights often exist in relation to physical items in which separate property rights exist. The classical example is a novel, in relation to which copyright exists. Someone might buy a particular copy of the novel, but this purchase does not confer a right to, amongst other things, copy it.
The major difference between the given example and the virtual item scenario concerns tangibility. With a novel, the focus of the rights concerned is tangible. Virtual property is, like the rights themselves, intangible. There are other examples of intangibles in which property rights may exist, such as shares. While there might be more potential issues to deal with, there is no conceptual reason why intellectual property rights should not exist in intangible virtual property, which may be owned by another.
Unfortunately, for most of the actions considered above, and for the property rights just mentioned, there is a contractual coup de grâce applied by most synthetic world providers before the users even enter. Every time users enter a synthetic world they must reaffirm, by way of a click-wrap agreement, a contract with the provider. These are commonly referred to as 'End-User Licensing Agreements' or EULA's.
EULA's generally purport to extinguish any and all of the rights that their drafters can conceive of as being relevant to the synthetic world. It seems likely that the underlying reason for these, seemingly draconian, measures
is that providers wish to avoid entanglement in any litigation. A notable exception to the above is Second Life, which is premised upon allowing users to retain any intellectual property rights that they might generate within the world.
Consideration of the effectiveness of contractual exclusion on each of the above causes of action is both beyond the scope of this paper, and dependant to an extent on interpretation of the relevant terms. However, one thing that can be said is that the doctrine of privity may prevent any contract between a user and a provider from negatively affecting a right of action between a third party and the user. This is most obviously the case in situations (ii), and (v), with regard to a copyright claim. However, it will not always be an issue; for example, in situations (i) and (iv), if the plaintiff is seen to have assigned the copyright to the provider, supposing that copyright subsisted in a generated avatar, she would have no cause of action, by way of copyright, against the defendant.
We are now in a position, with regard to the individual situations, to outline possible causes of action for each:
'Copycat' is not involved in commercial activity using the avatar
'Copycat' is involved in commercial activity using the avatar
Avatar bears a similar name/likeness to a preexisting avatar
(i) Copyright, Defamation
(iv) Copyright, Trade Mark, Passing Off, Fair Trading Act, Defamation
Avatar bears same name/likeness as a pre-existing fictional character
(ii) Copyright, Defamation
(v) Copyright, Trade Mark, Passing Off, Fair Trading Act, Defamation
An avatar bears a similar name, and/or likeness as a real person
(vi) Trade Mark, Passing Off, Fair Trading Act, Defamation
It should be noted that no reference is made here to the probability of success of these actions, merely to their possibility. As seen throughout this paper, many of the causes of action have a slim chance of success in many situations. Causes of action have been omitted where, in the view of the author, there is no hope of success.
There are several points that can be made in relation to these results. The first is that there is far more likelihood of a cause of action being available where there is a relevant economic interest to be protected. Such a trait can, of course, be ascribed more generally to common law legal systems as a whole.
Economic interests, here, are as opposed to the dignitary interests that one might have in concepts such as one's identity. However, the interests that might be threatened in the avatar situation are mostly of the dignitary kind. The closest the actions considered above come to the protection of such an interest is the tort of defamation.
Even where there are commercial interests involved, as in situations (iv)-(vi), the likelihood of an action being made out is highly uncertain. One reason for this is the current uncertainty over the status of synthetic world commerce. There is a bleeding edge between the synthetic world and the real, made obvious by worlds such as Entropia Universe and Second Life. This has produced a tension between; the appearance, that the residents are merely playing a version of Monopoly, and the reality, that some are exchanging Monopoly money (not to mention little green houses, and tiny metal tokens) for real currency.
With the knowledge that there is a paucity of protection in the area, the question arises, does it really matter? The best answer is, probably, not yet. But it may soon matter a great deal. While synthetic worlds are still a relatively minor phenomenon, that phenomenon has experienced startlingly rapid growth, with no sign of abatement. The media periodically focuses on some new parallel between the synthetic worlds and the real: the continuing evolution of Second Life and Entropia Universe being a favoured story source. Serious applications of the concept are emerging with great rapidity.
The limits inherent in the synthetic worlds themselves, ie, the degree of creativity allowed in avatar design, is a limitation on potential litigation. As the degree of user customisation increases, as seems inevitable, there will probably be an increasingly proprietary attitude adopted by users in relation to their creations. EULA's provide another, rather large, obstacle to the success of a user taking legal action. It remains to be seen the kind of effect that experiments, such as Second Life's, in allowing retention of intellectual property rights will have.
In one sense the relative lack of rights might have a positive effect. Some inhabitants of synthetic worlds are concerned both with the current commercialisation of their environment, and the anticipated advent of widespread litigation. There is also the ever present fear that, in view of the value of the transactions occurring, and the ease of conversion to conventional currency, that the Revenue might start taking a real interest. Some see these incursions of the real world as leading to a decline in the 'fun' factor of what are still called, perhaps increasingly nominally, games.
The problem outlined in this paper is, for the most part, one for the future. It was a future that William Gibson, back in the early eighties, anticipated, and one that may soon be with us. The one thing that might be said with certainty about this future is that as the number of people forming and maintaining identities in synthetic worlds increases, the greater the attachment they will feel to their avatar, and consequently, the greater their sense of grievance when the appearance of that avatar is misappropriated.
Scott Holdaway BA/LLB(Hons). He currently works as a clerk for the Christchurch City Council. Scott's future plans are to complete his Professionals and obtain employment in property and commercial law.
 W Gibson, Neuromancer (1984).
 The term 'avatar' is derived from Sanskrit, where it means the incarnation of a Hindu deity on earth: R Allen (ed), The New Penguin English Dictionary (2000) 92.
 A notable exception is VRML, or Virtual Reality Mark-up Language, and its descendants: L Bullard, Extensible 3D: XML Meets VRML (2003) xml.com <http://www.xml.com/pub/a/2003/08/06/x3d.html> at 25 January 2007.
 The term 'synthetic world' is adopted from Castronova: E Castronova, Synthetic Worlds: The Business and Culture of Online Games (2005) 9-10. This word is chosen over alternatives, such as the popular term 'MMORPG' ('massively multi-player online role-playing game'), as it possesses wide coverage, but avoids problematic terms such as 'virtual'.
 See, eg, Blizzard Entertainment Inc, Introduction to World of Warcraft (2007) World of Warcraft <http://www.worldofwarcraft.com/info/beginners/index.html> at 25 January 2007
 A 'mobile object'. This term applies to just about anything that moves of its own accord, but is not controlled directly by a human: Castronova, above n 5, 38.
 For further description, see G Lastowka & D Hunter, 'The Laws of the Virtual Worlds' (2004) 92 California Law Review 1, 26-29 for a simple account, and Castronova, above n 5, 29-43 for a more detailed discussion.
 Castronova, above n 5, 53.
 The number of synthetic world users worldwide was estimated in 2005 as being at least 10 million, with the likely number sitting somewhere between 20 and 30 million: Castronova, above n 5, 52-56. This figure was calculated before the unprecedented explosion in popularity of World of Warcraft: see A Schwarz & R Bullis, 'Rivalrous Consumption and the Boundaries of Copyright Law: Intellectual Property Lessons from Online Games'  Intellectual Property Law Bulletin 13, 15.
 Castronova, above n 5, 80-81.
 Ibid 35-42.
 Ibid 103-104.
 See,eg,ECastronova,See, eg, E Castronova, Virtual Worlds: A First-Hand Account of Market and Society on the Cyberian Frontier (2001) Social Science Research Network <http://papers.ssrn.com/abstract=294828> at 25 January 2007; J Fairfield, Virtual Property (2006) Social Science Research Network <http://ssrn.com/abstract=807966> at 25 January 2007
 See, eg, M Ward, Buy Yourself a Hero (2000) BBC News <http://news.bbc.co.uk/2/hi/science/nature/755614.stm> at 25 January 2007; BBC, Gamer Buys $26,500 Virtual Land (2004) BBC News <http://news.bbc.co.Uk/go/pr/fr/-/2/hi/technology/4104731.stm> at 25 January 2007
 See, eg, ibid. See, also, eBay's Category 1654, Internet Games.
 Castronova, above n 5, 149, 301.
 J Dibbell, The Unreal Estate Boom (2003) Wired <http://www.wired.com/wired/archive/l 1.01/gaming.html> at 25 January 2007
 See, eg, Fairfield, above n 14; Lastowka & Hunter, above n 8; E Castronova, 'The Right to Play' (2004) 49 New York Law School Law Review 185; J Grimmelmann, 'Virtual World as Comparative Law' (2004) 49 New York Law School Law Review 147; Fairfield, above n 14.
 Castronova, above n 5, 52-53.
 Fairfield, above n 14, 133-138.
 See R Reynolds, Hands off MY Avatar! Issues with Claims of Virtual Property and Identity (2003) Ren-Reynolds.com <http://www.ren-reynolds.com/downloads/HandsOfFMYavatar.htm> at 25 January 2007; S Crawford, 'Who's in Charge of Who I Am?: Identity and Law Online' (2004) 49 New York Law School Law Review 211.
 See M Ward, The Dark Side of Digital Utopia (2003) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/3334923.stm> at 25 January 2007.
 See, eg, BBC, Virtual Property Market Booming (2005) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/4421496.stm> at 25 January 2007.
 Lastowka & Hunter, above n 8, 14-21.
 Ibid 14.
 Ibid 16.
 Ibid 16-17.
 Ibid 17-18.
 Ibid 18.
 Ibid 19.
 Ibid 21.
 Ibid 21-22.
 MUD Object-Oriented.
 Multi-User Shared Hallucination.
 Interestingly ,there is reason to believe that the greater the degree of available avatar customisation, the greater the tendency for avatars to look in some way similar to their creator: see J Twist, Virtual Gamers Reveal Themselves (2004) BBC News <http://news.bbc. co.uk/go/pr/fr/-/2/hi/technology/3683260.stm> at 25 January 2007.
 See BBC, 'Game Theft' Led to Fatal Attack (2005) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/4397159.stm> at 25 January 2007; BBC, S Korean Dies After Games Session (2005) BBC News http://news.bbc.co.Uk/go/pr/fr/-/2/hi/technology/4137782.stm at 25 January 2007; D Griffiths, Treating China's Online Addicts (2005) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/asia-pacific/4327258.stm> at 25 January 2007; BBC, Head to Head: Is Online Gaming Bad For You? (2005) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/4273125.stm> at 25 January 2007.
 Castronova, above n 5, 285-294.
 See Castronova, above n 5, 56; Lastowka & Hunter, above n 8, 20.
 Castronova, above n 5, 133.
 See Lastowka & Hunter, above n 8, 66.
 Linden Research Inc, What is Second Life? (2007) Second Life <http://secondlife.com/whatis/> at 25 January 2007.
 Makena Technologies Inc, Welcome to There!! (2005) There <http://www.there.com/companyInfo.html> at 25 January 2007
 Mindark, About (2007) Entropia Universe <http://www.entropiauniverse.com/en/rich/5035.html> at 25 January 2007.
 See, eg, BBC, Cash Card Taps Virtual Game Funds (2006) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/4953620.stm> at 25 January 2007. In most synthetic worlds buying in-world currency involves bidding on an auction site, making payment, and then the two parties arranging for their avatars to meet in-world in order to effect the transfer. This is the method by which most virtual property is transferred. Avatars themselves are an exception; when they are sold, the vendor simply sends the purchaser their login name and password. The purchaser then changes the password: Castronova, above n 5, 149.
 C Ondrejka, 'Escaping the Gilded Cage: User Created Content and Building the Metaverse' (2004) 49 New York Law School Law Review 81, 92-93.
 J Twist, Picturing Online Gaming's Value (2005) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/4360654.stm> at 25 January 2007.
 Fairfield, above n 14, 111-112.
 BBC, Money Matters in Cybercash Game (2005) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/business/4248074.stm> at 25 January 2007.
 Fairfield, above n 14, 112-113.
 Ibid 113.
 Ibid 114-115.
 Google, Explore, Search and Discover (2007) Google Earth http://earth.google.com/tour/ at 25 January 2007
 English and Australian cases are frequently cited in areas, such as passing off, where they have a significant influence on New Zealand law: S Frankel & G McLay, Intellectual Property in New Zealand (2002) 5.
 Castronova, above n 14, 11.
 G Lastowka & D Hunter, The Laws of the Virtual Worlds (2003) Social Science Research Network <http://papers.ssrn.com/abstract=402860> at 25 January 2007, 87.
 D Bainbridge, Intellectual Property (4th ed, 1999) 38.
 See Copyright Act 1994.
 See, eg, Lyngstad v Anabas Products Ltd  FSR 62, 65.
 Copyright Act 1994, s 14.
 Copyright Act 1994, s 14(l)(a).
 Copyright Act 1994, s2(1).
 Copyright Act 1994, s 14(l)(a).
 Copyright Act 1994, s2(1).
 See Frankel & McLay, above n 58, 171-172; Bainbridge, above n 63, 48.
 Frankel & McLay, above n 58, 177
 Allen, above n 3, 269.
 Copyright Act 1994, s 14(1).
 Land Transport Safety Authority of New Zealand v Glogau  1 NZLR 261, 358.
 D Bainbridge, Software Copyright Law (4th ed, 1999) 138.
 See Autodesk, Autodesk Maya — Product Information (2007) Autodesk <http://usa.autodesk.com/adsk/servlet/index?siteID = 123112 & id=7635018> at 25 January 2007.
 'Machinima' is hugely popular field, devoted to the creation of short (and long) films, utilising computer games. This activity arose as a result of the ability in recent games, such as Quake, for: large numbers of simultaneous players; great freedom of player movement; and a controllable 'camera', or, point of view. Films range greatly in quality and originality. After viewing some of the best, it would seem impossible to deny that they would be original enough to attract copyright protection as films. This is, of course, disregarding the issue of their likely infringement of copyright elements in the particular games. See: Ondrejka, above n 48, 85.
 See Copyright Act 1994, ss 6(1), 8.
 See Copyright Act 1994, ss 2(1): Definition of 'computer-generated', 5(2)(a).
 Copyright Act 1994, s 2(1). Definition of'computer-generated'.
 Copyright Act 1994, s 5(2)(a).
 Bainbridge, above n 76, 77-78.
 Ibid 78.
 See Copyright Act 1994, ss 16, 29, 30.
 Copyright Act 1994, s 2(1).
 Copyright Act 1994, s 2(1). Definition of 'copying' (a).
 Copyright Act 1994, s 2(1). Definition of 'copying' (b).
 Copyright Act 1994, s 2(1). Definition of 'copying' (c).
 See Wham-O MFG Co v Lincoln Industries  1 NZLR 641 (CA).
 See Husqvarna Forest & Garden Ltd v Bridon New Zealand Ltd  3 NZLR 215.
 Wham-O MFG Co v Lincoln Industries  1 NZLR 641, 666 (CA).
 Bonz Group (Pty) Ltd v Cooke  NZHC 1004;  3 NZLR 216, 220.
 See T Loftus, Virtual Worlds Wind up in Real Courts (2005) MSNBC.com <http://wmsnbc.msn.com/id/6870901/> at 25 January 2007.
 Loftus, above n 94; Thelen Reid Brown Raysman & Steiner LLP, Can Virtual Property Gain Legal Protection? (2006) ThelenReid.com <http://www.thelenreid.com/resources/documents/060209_IPL360.pdf> at 25 January 2007.
 NCsoft, Game Info: Game Synopsis (2007) City of Heroes Community Site <http://www.cityofheroes.com/gameinfo/synopsis.html> at 25 January 2007
 Loftus, above n 94.
 Thelen Reid Brown Raysman & Steiner LLP, above n 95.
 Loftus, above n 94.
 See Copyright Act 1994, s 131(2).
 See Copyright Act 1994, s 16(l)(i).
 Frankel & McLay, above n 58, 230-232.
 See Trade Marks Act 2002.
 Trade Marks Act 2002, s 5(1). Definition of 'trade mark'.
 Trade Marks Act 2002, s 5(1). Definition of 'sign'.
 Trade Marks Act 2002, s 32.
 Trade Marks Act 2002, s 18(1).
 Trade Marks Act 2002, s 89.
 Castronova, above, n 5, 32-34.
 It should be borne in mind that such registration is in regard to the particular image that constitutes the sign, and not the image as such: H Beverley-Smith, The Commercial Appropriation of Personality (2002) 44.
 Trade Marks Act 2002, s 89(1).
 See Dibbell, above n 18.
 See BBC, above n 24.
 In reality such tasks are generally a mind-numbing sequence of repetitive clicking and dragging with the mouse: Dibbell, above n 18.
 Loftus, above n 94.
 It has been argued that it is the presence of economic transfers between the synthetic world and the real world that should indicate whether the law should intrude into the synthetic world: see, generally, Castronova, 'The Right to Play' above n 19.
 Entropia Universe has made an Eftpos card available, by which in-world funds ('Project Entropia Dollars') are able to be converted directly into US dollars; deposits into the account are likewise accessible in the synthetic world. The conversion rate is fixed at $10PED to $1US. See BBC, Cash Card Taps Virtual Game Funds (2006) BBC News <http://news.bbc.co.uk/go/pr/fr/-/2/hi/technology/4953620.stm> at 25 January 2007.
 Trade Marks Act 2002, s 89(2).
 Trade Marks Act 2002, s 23.
 Trade Marks Act 2002, s66(1)(a).
 Trade Marks Act 2002, s66(1)(b).
 Trade Marks Act 2002, s 17(l)(b)(i).
 This may not be the case within the synthetic world (there would, arguably, be an implied licence operating with regard to copyright within the world), but it certainly would if the mark was utilised in an, otherwise legitimate, real world context.
 Trade Marks Act 2002, s 89(1).
 See Frankel & McLay, above n 58, 580-587
 See S Todd, The Law of Torts in New Zealand (4th ed, 2005) 577.
 As opposed to Lord Diplock's five point test: Warnink v Townend & Sons (Hull)  AC 731, 742 (HL).
 Reckitt & Colman Products Ltd v Borden Inc  UKHL 12;  1 All ER 873, 880 (Lord Oliver) (HL).
 See British Diabetic Association v Diabetic Society Ltd  4 All ER 812; Arabian Muslim Association v Canadian Islamic Centre (2005) 246 DLR (4th) 351.
 (1878) 10 Ch D 294 (CA).
 See, eg, Cadbury Schweppes Pty Ltd v Pub Squash Co Pty Ltd  1 All ER 213, 218 (PC).
 See, eg, Tot Toys Ltd v Mitchell  1 NZLR 325, 344.
 See above, text associated with n 122.
 Levi Strauss & Co v Kimbyr Investments  1 NZLR 332.
 See above, n 121, and associated text.
 Neumegen v Neumegen and Co  3 NZLR 310, 317 (CA). The case concerned the Fair Trading Act 1986, and it was held that there was no own name defence in the statutory context, barring necessity..
 See, eg, Castronova, above n 5, 33.
 See Tot Toys Ltd v Mitchell  1 NZLR 325, 354-355.
 Anheuser-Busch Inc v Budweiser Budvar National Corporation  3 NZLR 666, .
 Tot Toys Ltd v Mitchell  1 NZLR 325, 354.
 Charity collections have occurred inside synthetic worlds, notably in the wake of Hurricane Katrina: J Twist, Storm Help Flows from 3D Worlds (2005) BBC News <http://news.bbc. co.uk/go/pr/fr/-/2/hi/technology/4219030.stm> at 25 January 2007.
 Clark v Associated Newspapers Ltd  1 WLR 1558.
 Klissers Farmhouse Bakeries Ltd v Harvest Bakeries Ltd (No 2)  2 NZLR 143, 153 (Davison CJ).
 Wombles Ltd v Wombles Skips Ltd  FSR 488.
 Taylor Bros Ltd v Taylors Group Ltd  2 NZLR 1, 18-20.
 See Todd, above n 130, 595-597.
 Taylor Bros Ltd v Taylors Group Ltd  2 NZLR 1, 37.
 See, Ward, above n 23.
 See Frankel & McLay, above n 58, 575-580, 587-588.
 Wineworths Group Ltd v Comite Interprofessionel du Vin de Champagne  2 NZLR 327, 343 (CA).
 See, eg, Children's Television Workshop Inc v Woolworths (New South Wales) Ltd  RPC 187; Hutchence v South Seas Bubble Co Pty Ltd (1986) 64 ALR 330.
 Todd, above n 130, 607-611.
 Henderson v Radio Corporation Pty Ltd  RPC 218.
 Ibid 232.
 See text associated with n 146-147
 Hogan v Koala Dundee Pty Ltd  FCA 333; (1988) 83 ALR 187
 Ibid 192(PincusJ).
 Beverley-Smith, above n 114, 93.
 Pacific Dunlop Ltd v Hogan  FCA 185; (1989) 87 ALR 14.
 Ibid 38.
 Ibid 44-45.
 For example, selling an avatar that looks like 'Mickey Mouse'.
 Twentieth Century Fox Film Corp v South Australian Brewing Co Ltd (1996) 34 IPR 225.
 This is potentially explicable by the fact that, because of the humorous nature of The Simpsons, viewers would be likely to be unperturbed by this in their impression of an association: Frankel & McLay, above n 58, 594. It is, perhaps, more likely that the Court was veering more dangerously close to creating a property right.
 It should be noted that the Australian courts have refused to openly countenance the existence of a proprietary right: Sony Music Australia Ltd and Michael Jackson v Tansing (1994) 27 IPR 649, 653-654.
 Tot Toys Ltd v Mitchell  1 NZLR 325, 359-366.
 Ibid 363-364.
 Ibid. See below, subpartE.
 Todd, above n 130, 600.
 Fair TradingAct 1986, s 9.
 Bleiman v News Media (Auckland) Ltd  2 NZLR 673, 680 (CA).
 Fair TradingAct 1986, s 2(1). Definition of 'trade'.
 Todd, above n 130, 189-190.
 Taylor Bros Ltd v Taylors Group Ltd  2 NZLR 1, 39.
 See, generally, Beverley-Smith, above n 114; Todd, n 130, 609-611.
 See Krouse v Chrysler Canada Ltd (1973) 40 DLR (3d) 15; Athans v Canadian Adventure Camps Ltd (1977) 80 DLR (3d) 583.
 See Zacchini v Scripps-Howard Broadcasting Co,  USSC 148; 433 US 562 (1977); Beverley-Smith, above n 114, 174-189.
 Todd, n 130, 609.
 Athans v Canadian Adventure Camps Ltd (1977) 80 DLR (3d) 583, 595.
 Holdke v Calgary Convention Centre Authority  ACWS (3d) 1281.
 Gould Estate v Stoddart Publishing Co (1997) 30 OR (3d) 520, 526-527.
Motschenbacher v R J Reynolds Tobacco Co,  USCA9 255; 498 F 2d 821 (9th Cir, 1974).
 Tot Toys Ltd v Mitchell  1 NZLR 325, 363-364.
 Hoskingv Runting  NZCA 34;  1 NZLR 1,  (CA).
 Todd, above n 130, 651.
 See Defamation Act 1992.
 Todd, above n 130, 652.
 Sim v Stretch  2 All ER 1237, 1240 (Lord Atkin) (HL).
 For example, rudeness: Hoare v Silverlock  EngR 667; (1848) 12 QB 624.
 Lloyd v David Syme Ltd  AC 350 (PC).
 Gray v Jones  1 All ER798.
 Villiers v Monsley (1769) 2 Wils 403;  EngR 526; 95 ER 886.
 Hill v National Bank of New Zealand Ltd  NZHC 116;  1 NZLR 736.
 Slim v Daily Telegraph Ltd  2 QB 157.
 Substitute with 'Mickey Mouse' and the Disney Corporation to see the application to situations (ii) and (v).
 Hyams v Peterson  3 NZLR 648, 654 (CA).
 Mt Cook Group Ltd v Johnstone Motors Ltd  NZHC 137;  2 NZLR 488.
 Bruce v Odhams Press Ltd  1 KB 697.
 See, eg, Mt Cook Group Ltd v Johnstone Motors Ltd  NZHC 137;  2 NZLR 488.
 Collerton v Maclean  NZLR 1045.
 O'Brien v Brown  DCR 1065.
 Dingle v Associated Newspapers Ltd  AC 371 (HL).
 Defamation Act 1992, s 21.
 See, eg, Castronova, above n 5, 89.
 See above, n 19.
 See Fairfield, above n 14, 144-145.
 Bainbridge, above n 63, 12-14.
 Unless, of course it is conceived of as electromagnetic pulses. However, it is submitted that this is an unhelpful way to conceptualise such matters, as it frequently leads to an analysis that places too great an emphasis on who owns the hardware on which the pulses reside. It is contended that in a distributed computing environment, such as the internet, such a focus is meaningless. For example, the physical instantiation of the data that represents a particular avatar will often be on several computers simultaneously: the server, and all the relevant clients. See Fairfield, above n 14, 125-131.
 Bainbridge, above n 63, 10.
 See Frankel & McLay, above n 58, 15.
 Castronova, above n 5, 208.
 See, for example, the EULA for Star Wars Galaxies, reproduced in: Castronova, above n 5, 219-223.
 J Grimmelmann, Free as in Gaming? (2004) LawMeme <http://research.yale.edu/lawmeme/modules.php?name=News & nle=article & sid=1290> at 25 January 2007.
 See J Burrows, J Finn & S Todd, Law of Contract in New Zealand (2002) 499-501.
 Beverley-Smith, above n 114, 141-142.
 Ibid 8-12.
 See the frequent references in the above footnotes.
 See Grimmelmann, above n 225.
 See J Dibbell, Dragon Slayers or Tax Evaders? (2006) LegalAffairs <http://www.legalaffairs.org/issues/January-February-2006/feature_dibbell_janfeb06.msp> at 25 January 2007; R Paul, Are MMORPG Goods Theoretically Taxable? (2006) <http://arstechnica.com/news.ars/post/20060106-5923.html> at 25 January 2007.
 See, eg, Castronova, above n 5, 150-151.