Just Like Me Only Better

Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. JUST LIKE ME ONLY BETTER Suely Fragoso & Nísia Martinsdo Rosário (Unisinos, Brazil) Abstract. This paper presents the results of an investigation of Second Life avatars which represent subjects from different ethnic origins and various cultural backgrounds. Initial considerations about the representation of self in avatars, specially in non-thematic graphical MUVEs (Multi-User Virtual Environments) such as Second Life, contextualise and explain the research question. The methodological strategies adopted for the construction of a corpus comprised of 68 avatars and 18 nationalities collected from 5 different types of sources are detailed and discussed. Research findings include, but are not restricted to: (a) a strong preference for the human form; (b) a tendency towards the adoption of features similar to the physical body of the subject represented by the avatar; (c) a general similarity amongst all avatars, due to the adoption of a contemporary Caucasian pattern of beauty composed by youth, slenderness, fair skin, straight hair and other likely features (not contradictory to (b), as the personal characteristics of the ‘owners’ are often attenuated or emphasised according to the general standard of beauty) and (d) in the process reported in (c), a tendency to resort to stereotypical exaggerations. After the discussion of the findings, final considerations set the ground for further investigations. 1. Introduction The term avatar comes from Sanskrit and originally referred to the form that the gods assumed when they visited the world. Thus it does not denote a material body like ours, but a representation bearing divine characteristics, independent of the limitations of time and space. This peculiarity makes this Hindu concept particularly appropriate for denominating elements of the digital space that function as representations of the human users. In this usage, the word avatar includes a range from the simplest representations, from a pointer on the screen or usernames through to the most complex, such as the graphically refined images of characters in computer games. The continuity, interactivity and graphic detail involved are important factors in the complexity of the links that are forged between the subjects and their avatars. Nothing confers more reality on the avatar, however, that its history as a mediator in interactions between the subject and other people in shared digital environments. In this function, the avatar is a form of social representation of the self. It is, however, a different form, due to its immaterial nature, which frees the subject from his physical body, time and space, and due to the need for it to be explicitly created. This paper presents the results of an investigation of Second Life avatars of subjects from different ethnic origins and various cultural backgrounds. We start 1 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. with the theoretical background from the perspective of which we construct the research problem. Then we present a brief discussion of the history of the modes of representing the self online, which culminates with the description of the possibilities of avatars in Second Life. There follows the detailing of the research, with emphasis on the ethnicity and aesthetic values of users of Second Life from different regions of the world. Methodological strategies are presented and discussed in Section 3. Section 4 presents and discusses the results and the final section (5) carries the discussion of the findings further in the direction of their cultural and psychological implications as well as indicating possible further work to be done. 2. Online Representations of Self An unavoidable reference when discussing presentations of the self are the seminal works of Erwin Goffman, notably The Representation of Self in Everyday Life (first published in 1959). Underlying Goffman’s work is the presupposition that in all social interactions the desire of each participant is to regulate the impression that is transmitted to the others, through roles. In this environment the self is no more than a character being played, whose interest is being believed. The process of role construction established here is one that requires the use of a sign dimension – despite this not being mentioned by Goffman – that is capable of organizing these representations into a complex discourse in which the belief in the representation predominates. In this same way, the avatar is also a set of signs through which the subject seeks to construct a believable representation. Goffman postulates the existence of a radical asymmetry in the communication process, which would be divided between a high level of control over verbal expressions and the autonomy of the “ungovernable aspects of the expressive behaviour of the individual1” (Goffman, 1985, p. 16), which basically correspond to non-verbal manifestations. However, given that verbal expression is not under absolute control, and non-verbal manifestations are not totally spontaneous, this dichotomy seems better formulated in the words used by Miller (1995, online), who distinguishes between ‘information given’, that is, intentionally produced and controlled, and ‘information given off’, which 'leaks through' without intention. Physical co-presence implies the maximum number of channels for this ‘leaking’. The fact that face-to-face interactions have the greatest communication bandwidth explain why they occupy a central position in most of Goffman’s work. Without denying the potentialisation of the channels of meaning by the presence of the body, it is important to observe that technical or technologically-mediated interactions are also not restricted to a single flow of meaning. The image of the author of a letter, for example, is not only 1 All quotations from Goffman have been retranslated to English from Brazilian editions. 2 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. constructed by the content that it is chosen to be discussed (or avoided) and by the words that are selected to do express that content, but also by the peculiarities of the handwriting, the type of paper, the cleanness, etc. In e-mail, as with IRC, the interaction is effectively restricted to the text. Due to this, despite the fact that stylistic choices and other similar elements are still in play, Miller (1995) considers that, in e-mails, “aspects of the embodied self can only be apparent if described by the sender”. As early as 1991 an empirical observation of IRC had indicated to Reid that “the idea that as the communication bandwidth narrows interaction should become increasingly impersonal does not hold true for IRC”. Rheingold’s widely known book The Virtual Community (1998) supports this position. In 1994, studying interaction in MUDs, Reid identified more clearly the existence of compensations for the loss of many of the cultural tools that form the basis of conventional systems of interaction in text-only virtual environments through the creation of replacements and substitutes for physical cues. This process is highlighted by one of the principal differences between MUDs and other text-based social interaction systems, which is the invitation to explicitly detail the representations of the environment and of the self. Vasalou, Joinson and Pitt (2007) stress the importance of the processes of representation and identification in online social interactions. Along the same lines, Kafai, Fields and Cook (2007, p. 1) propose that “avatars designed by players might offer a particularly promising window into understanding how participants in the worlds (as designers and players) think about their representations of self in virtual life”. Beyond choosing a screen name for themselves, MUDders are invited to attach descriptions of their virtual bodies, clothing, possessions and other symbols of identity. These descriptions are often carefully detailed, and result in a consolidation of the content of various layers of meaning, which confers reality to the representations and intensifies the relationship between the avatar and the subject represented. 2.1. GRAPHIC ONLINE REPRESENTATIONS OF SELF The move towards graphical representations of users and environment in the end of the 1980s did not mean an immediate improvement in the possibilities for the creation of the avatars. Compared to the richness of the descriptions found in MUDs, the 2D representations of early 2D graphical MUVEs such as LucasFilm’s Habitat (1986) or The Palace (Time Warner and Partners, 1995) were a step backwards. Mostly driven by the game industry, the visual sophistication of such environments greatly improved over time. The thematic nature of a good part of the MUVEs implied restrictions on the characterization of users’ representations. In the majority of those cases, users’ graphical representations are built from sets of images for specific features (species, gender, hair colour, etc). Given that they are constructed from pre- 3 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. defined options, it is not surprising that the avatars that inhabit thematic MUVEs tend to be highly similar in appearance. Contrary to this, creative freedom is offered as an attraction by the non-thematic environments. The motto of Second Life, ‘your world, your imagination’, synthesizes the call to participate in the construction of the MUVE itself, suggesting freedom of creation at all scales, including in the customization of the avatars. It is true that in Second Life the new user has to choose their appearance from a small range of options, which are nearly all human and that are clearly divided into male and female. Whilst still in the basic tutorial, the users (‘residents’) learn how to change a set of characteristics of their avatars, including height, body proportions, hair, skin and eye colour, the size, format and distribution of facial features etc. With a little more expertise it is possible to bring into the MUVE totally new images and substitute the textures that wrap the 3D model of the avatar, radically altering its appearance. Skins and shapes are sold or freely distributed in various places. Thus, the residents that are less graphically talented are also able to find ways to change their avatars. For the expert users, customization of Second Life avatars have practically no limits: it is possible to look like an animal, a geometric form or even to become invisible. Despite this freedom to characterise the representation of one’s own body in Second Life, in-world experience and image searches on the web 2 indicated that the vast majority of residents choose to maintain a human form. Also, the great majority of skins traded in Second Life look toward realism, with attention given to details such as highlights and shadows for protrusions and hollows, colour and shape of nails and nipples and the inclusion of pubic and chest hair. Particular attention is given to the representation of the genitalia, which can also be obtained as complementary objects which, when provided with the appropriate scripts, ‘react’ to stimuli. These observations converge with the contention of Smith, Farnhman and Drucker, whose research suggests that “people might play with gesture and move around the 3D spaces without incorporating gestures and movements into their social interactions” (2000, p. 464). In both cases, the characterization of the avatar seems to be less directed toward deepening the interaction processes than it is toward the identity. Taylor (2002) calls attention to the importance of the avatar to increase the realism of users’ engagement with the virtual world. Avatars, in fact, come to provide access points in the creation of identity and social life. The bodies people use in these spaces provide a means to live digitally – to fully inhabit the world. It is not simply that users exist as just “mind”, but instead construct their identities through avatars (Taylor, 2002, p. 40). The tendency to replicate that which exists in the physical world is not a privilege of avatars: many, if not the majority, of objects in Second Life are 2 The searches referred to were carried out in Google Images (http://images.google.com.br/) and Yahoo! Images (http://br.images.search.yahoo.com/) in December 2007 and January, 2008 and used the phrase “second life”. 4 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. copies of things from the off-line world. Places are also recreated, in such a manner that it is possible to find the Second Life equivalent of various cities, national monuments and slums from around the world, all boasting of the fidelity of the copy. Attention to the residents that frequent these virtual places make it noticeable that they don’t transcribe into the MUVE only the appearance of environments known, but also their social and cultural marks. By doing so, these duplications inform of the current regulations and indicate the identities and othernesses expected there (Fragoso, 2007), and therefore they facilitate adaptation to and interaction within the on-line world: avatars use swimming costumes in environments that appear like beaches but not in shops, they know that they are more likely to encounter Portuguese speakers on Ilha Brasil than in Second Japan, that it may be acceptable to have sex in private spaces but not in public, etc. Taking this tendency to reproduce reality even further, there are indications that residents in Second Life seek to approximate the characteristics of their avatars to their physical bodies. A similar situation was observed by Vasalou, Joinson and Piit, who suggest that “users who customize avatars to reflect their own appearance experience increased private self-awareness in comparison to users who create avatars unlike themselves” (2007, p. 446). From our point of view, the similarity between the bodies of the user and the avatar can reduce the tension of representing a single self simultaneously in two worlds: the physical/biological and the virtual. This double projection sets up a peculiar exchange between the subjectivity of the individual, the experience of his fleshand-bone body, and the avatar. There are many reports of emotions and concerns that are manifest in the physical body despite having been motivated by events that exclusively involved the avatar (on-line sexual relations are the most widely known of this type of event). Overall, the inescapable connection between the self and its physical body is reframed by the existence of a representation that, despite having a corporeal appearance, belongs to the realm of the symbolic. 2.2. TRANS-CULTURAL GRAPHIC REPRESENTATIONS OF SELF Following the example of what occurs in the off-line world, the construction of a social identity in a MUVE is through the adoption and highlighting of some characteristics to the detriment of others. Each choice that relates to the appearance of the avatar emphasises certain attributes – physical, intellectual, aesthetic, economic – and generates signs that inform who that subject is, and to which group he belongs. The cultural outline of the subject that is being represented is the first reference point sought to find references to decide which elements are desirable to have associated (or not) with the representation of the self. Following the example of what happens with the representations of the self in the everyday off-line life, it is to be expected that people build their avatars with the features that they consider to be capable of expressing more clearly 5 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. what they want to state about themselves. In addition to the purpose of making the avatar body similar to their physical self, teenager users observed by Kafai, Fields and Cook (2007) listed six, non-mutually exclusive reasons for creating their avatars the way they did. Five of them focus on the looks: the desire to make the avatar look like their real selves (as formerly discussed here); the intention to exercise (or demonstrate) their artistical skills; to affiliate with something or someone they liked “such as a video game character, relative, hobby, or nationality” (p. 6); to have a look or belonging they can’t have in real life (a specific hair style, or pretty lips); to align oneself for or against a popular trend. The sixth reason was a functional one: as a disguise, to deceive or to surprise. In all cases, the information about what is common and what is rare is available in the physical reality that surrounds each individual, whilst the value filters of the culture are charged with distinguishing the normal from the mediocre and the exotic from the strange. If avatars are (a) modelled to be similar to the physical bodies of their ‘owners’ and (b) shaped by the values inherent to the cultural context of the individuals they represent, it would be reasonable to expecte that the avatars in a MUVE with global scope, such as is the case with Second Life, would comprise a highly varied range of different ethnicities and aesthetic values. However, the in-world experience of the authors suggests the contrary: that the avatars of Second Life are all very similar to each other. Due to its spontaneous character, this impression can derive from various factors, above all the natural tendency to habituate a restricted range of places (those associated with known or admired countries for example), which results in a high probability of meeting people of the same ethnic and cultural profile. To obtain more consistent information about the way that variations in the ethnicity and aesthetic values of users of Second Life that come from different regions of the world are manifested in the configuration of their avatars, a systematic observation was undertaken, following the procedures described below. 3. Metodological Strategies Linden Research Inc. periodically publishes data about the nationalities of the active residents in Second Life3. The list of the Top 100 Countries by Active Users Counts for November of 20074 included countries from five continents. There is a clear predominance of residents from North America and Western Europe, but it ought to be possible to encounter avatars whose ‘owners’ are 3 According to the SecondLife Wiki (http://wiki.secondlife.com/wiki/Metrics_Glossary#Active_Users), only residents who have logged into the system for longer than one hour for the reported time period are computed as Active Users. 4 Part of the document entitled Second Life Virtual Economy - Key Metrics (BETA) Through November2007. Despite October figures being included in November for most of this document, the breakdown of active users by country is specific to November (Linden, 2007) 6 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. located in different regions of the world, comprising a considerable variety of ethic groups and cultures. 3.1. CONSTRUCTION OF THE CORPUS Taking as a guide the table of nationalities present in Second Life in November 2007, a collection of images of avatars identified with users from different regions of the world5 was made. Wherever it was possible, images of the ‘owners’ of these avatars were also recorded. In the cases where the truthfulness of the data relating to the subjects represented by the avatars was extremely questionable (use of grotesque descriptions or images, identification with a media figure, etc.) they were not included in the sample. Despite the attempts to minimize the doubts it is impossible to guarantee the validity of a corpus constituted of online records that are deliberately created by users. Additionally the limits of reasonable generalization for the data found relative to this corpus are very clear. Three determining factors applied to the attempts to widen these limits: (1) Seeking breadth, that is, the inclusion of avatars constructed by residents coming from different nationalities, ethnic groups and cultures; (2) the restriction of consideration to those items with sufficient representativity, understood as the minimum recurrence of any characteristic in one subset and (3) the expressive intensity of the images; clarity of the visualization of the characteristics of the avatars and of the subjects that they represent. Images were collected from five source types: 1. Communities and profiles on the SNS Orkut (http://www.orkut.com ) related to 2. 3. 4. 5. Second Life; SLProfiles.com (http://slprofiles.com); Flickr groups related to Second Life (http://www.flickr.com); Thematic groups in Second Life ; Websites, forums and blogs relating to Second Life and that are explicitly linked to specific nationalities. In many of these environments it would have been foolish to take the nationality indicated by the residents at face value. In Orkut, for example, it is known to be common to declare oneself a resident of exotic countries or of those which you admire (Fragoso, 2006). The same is true for the membership of groups in Second Life and is likely to happen in the other sources as well. For the purposes of this work, it was considered that to identify the nationality of the avatars it was sufficient to have coherence of two of three indicators: declaration 5 All of the images used are publicly available (some sources such as Orkut and Second Life itself require registration, but this imposes no practical restrictions). Even so, for both ethical and legal reasons, the question of the reproduction of these images raises issues that drive from their multicultural origin. In some cultures, personal photographs published in one context are generally considered as available to be copied and published in other places. In Brazil even their comic use is relatively common (the site Kibe Loco (http://www.kibeloco.com.br/), for example, frequently ridicules images of celebrities and ordinary people without this generating moral protests or legal actions (Tabet, 2007, online)), which certainly would not be true in many other countries. To prevent from generating embarrassment as well as to avoid violating laws which are highly varied, we have chosen not to include example images, names or other data that could individually identify the avatars that make up part of the corpus in this paper. All of the data is available by request from <suely@unisinos.br> or <nisia@unisinos.br> , for consultation by other researchers. 7 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. (“i am Japanese”); demonstration of the ability to express oneself in the official language of the country (in descriptions, forums and blogs) and belonging to the majority ethnic profile of the country or geographical region in question (as the intention is to identify the representation of traces of the hegemonic ethnicity and of the majority cultural values in each region, immigrants and other minority ethnic groups were not included in the corpus) 3.1.1. Communities and Profiles in Orkut A search of communities in Orkut, in all languages for the expression ‘second life’ identified 616 results. An initial sweep allowed the identification of a large number of communities which included discussion topics such as ‘what is your name in Second Life?’ or ‘who is who’, whose replies sometimes permitted the identification of the subjects represented by the avatars and to locate their profiles on Orkut6. The collection of images was concentrated on topics from five communities (Table 1): Table 1. Avatars selected from Orkut Communities. COMMUNITY (nº of members 04/01/08) Second Life (474) TOPIC (number of posts 04/01/08) what is your name in Second Life?” (130) Second Life (236) who is who (55) Orkut Life- Second Life – 1728 apresente-se7 (93) Second Life Índia (89 What is your Avatar (The Virtual Reality Name)?(19) Second Life – Brasil – 1015 qual é o seu nome no SL8 (102) NUMBER OF AVATARS CHOSEN (by country) 2 Brazil; 5 India; 1 Germany; 1 Russia; 1 UK TOTAL = 10 2 USA; 2 Brazil; 2 India; 1 Finland; 1 Holland TOTAL = 8 3 Brazil TOTAL = 3 3 India TOTAL = 3 4 Brazil TOTAL = 4 TOTAL = 28 In total, 28 avatars were selected from Orkut: 11 Brazilian, 10 Indian, 2 USA, 1 German, 1 Russian, 1 British, 1 Dutch, 1 Fin. Three characteristics determined that Orkut was one of the best sources of avatars for the research: (1) the explicit connection between the user of the SNS and their avatar in Second Life; (2) the presence of many indicators of nationality in the information sent and received in the communities and scrapbooks, which allowed for greater certainty in establishing the nationality of the person behind the avatar., and (3) the availability of images of avatars and of their owners, normally in the photo albums of the users. As the majority of Orkut users are Brazilian and Indian, this implied an inequality from the start in the nationalities present in the sample 6 It is worth noting that many profiles of avatars were found on Orkut, which contained detailed information, photo albums, biographies and a collection of friends (also avatars) that left messages and comments. 7 Present yourself. 8 What is your name in SL 8 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. obtained from this source. The collection of avatars from other sources helped to ameliorate this imbalance. 3.1.2. SLProfiles.com In SLProfiles.com the avatars were selected on the basis of the photographs of the subjects represented, published in the area First Life Pics. In addition to observing the nationality, it was also possible to identify ethnic indicators and differences with regard to the more common physical types amongst those selected in the Orkut communities9. Thus it was possible to increase the range of not only nationalities but also of ethnicities present in the sample. Images were collected for 14 avatars (and of the subjects associated with them) coming from 8 countries: 7 from the US; 1 from the UK, 1 from France, 1 from Canada, 1 from Spain, 1 from Germany, 1 from Japan, 1 from South Africa. 3.1.3. Flickr Groups The users were identified on the basis of photographs published in the groups Second Life (4,949 members on 11/Jan/2008) and Second Life - Avatars (1,472 members on 11/Jan/2008). Working from the indication of the geographic origin of the photographs (according to the map for the associated Flickr group) and further indications found in the profile of the user or in the comments on the photos, 3 avatars were selected: 1 from Norway, 1 Australian, 1 Mexican. 3.1.4. Second Life Groups In Second Life, a sweep was made of the profiles of the avatars for the groups: Angola (30 members10), Japan e-kuni (302 members), Jeju korea (16 members), Jiang hu (208 members), Lobito – Angola (10 members), People’s Republic of China (32 members), Second Japan (469 members) and Vietnam (28 members). This was intended to increase the presence of African and East Asian users in the sample. Only 9 profiles (4 Japanese, 3 Chinese, 2 Korean) were considered to have sufficient identification of their nationality and a sufficiently clear image of the avatar. Just two of these also had clear photographs on their 1st Life tab. All attempts to contact members of the groups that were on-line during the process were unsuccessful. 3.1.5. Other sources: websites, forums and blogs Complementary images and information were also collected from 7 other sources, with 3 being associated with China (Second Life 中 国 同 盟 (第 二 人 生 ) 这 不 是 一 场 游 戏 (http://www.cnsecondlife.com/bbs/),飞 刀 塑 梦 的 secondlife - 飞 刀 塑 梦 - 新 浪 (http://blog.sina.com.cn/secondlife100), and estellecloud (http://www.photoblog.com/estellecloud); 1 with Germany (Slinside.com (http:// www.slinside.de/), 1 with Turkey (Second Life Türkiye Video Dizini (http://video.slturkiye.com/), 1 with Russia (Русское сообщество Second Life 9 The majority of Orkut users linked to Second Life avatars were young and in good physical shape. It is probable that this is a result of a combination between the profile of users for that SNS (in the majority young) and the prevalence of Brazilians, amongst which photos of those that are older and or overweight are less common. 10 The data for the number of members in the Second Life groups are for 11/Jan/2008. 9 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. (Вторая Жизнь) (http://world2.ru/), (http://slbeat.mytimesdispatch.com/). and 1 with the USA InRich 3.1.6. General nature of the Corpus The process of constructing the corpus implied the observation of hundreds of avatars.The final sample was comprised of images relating to 61 avatars and 19 nationalities, as presented in Table 2. The question of the relevance of the sample was addressed on the basis of quality, that is, we chose on the basis of the guarantee provided by the cross-referencing of various indicators for the nationality of each avatar and the discarding of all those cases where the information was either insufficient or inconsistent. This implied a sacrifice of the representativity provided by quantity, and resulted in the final corpus being comprised of a low number of avatars, but of avatars for which we could be more certain of the nationality of their creators.. During the process of constructing the corpus, very few non-human avatars were found, with the majority of those having human bodies with altered ears (the most common being feline ears, but with one Chinese avatar had elven ears.), tails (also feline) and/or with skins with patterns that suggest fur. Table 2. Composition of the Corpus Source Nationality Total Source Nationality Total 1 Orkut (11) Brazilian 11 11 Orkut (1) Fin 1 2 Orkut (10) Indian 10 12 SLProfiles (1) French 1 3 Orkut (2) USA 10 13 SLProfiles (1) Canadian 1 Chinese 6 14 SLProfiles (1) Spanish 1 Japanese 5 15 SLProfiles (1) South African 1 German 3 16 Flickr (1) Norwegian 1 Russian 2 17 Flickr (1) Australian 1 British 2 18 Flickr (1) Mexican 1 SLProfiles (7) Website (1) 4 Second Life Groups (3) Forum & Blog (3) 5 SLProfiles (1) Second Life Groups (4) 6 Orkut (1) SLProfiles (1) Website (1) 7 Orkut (1) 8 Orkut (1) Website (1) SLProfiles (1) 10 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. 9 Second Life Groups (2) Korean 2 10 Orkut (1) Dutch 1 9 19 Website (1) TOTAL Turkish 1 61 3.2. ANALYSIS The images were analysed to identify general patterns and the exceptions in the features selected to compose the body of the avatars. Initially all of the avatars were observed together. Then they were separated by nationality and analysed again. The results by nationality were compared to those for the general set of images. Finally the avatars were compared to the images of the subjects they represent. The patterns, repetitions, similarities and exceptions identified at each stage were recorded and discussed. 4. Results Despite the variation in nationalities that constituted the corpus, there was a clear predominance of certain facial features and physical shapes amongst the avatars analysed, principally: straight hair, fair skin, large eyes, strong cheekbones, small mouth, fine nose, tall and slender bodies with Vitruvian proportions. The majority of women had a slim waist, full breasts, delicate faces, small noses, full lips and long hair. Amongst the men predominated large chests, defined biceps, narrow hips, and short hair. These characteristics follow a contemporary Caucasian pattern of beauty that is ever present in Western media, and are those that are principally responsible for the similarity in the general set of Second Life avatars. Some variations by nationality are noticeable, despite the great majority of avatars being white, some have tanned skin. This is a preference amongst German avatars and is accompanied by straight hair (in general blond) and thin lips and nose. Amongst the Germans the majority of the tanned avatars are women, whilst for other nationalities it is more common that the male avatars are tanned. Amongst the Asians predominated straight dark hair and eyes that were slightly more almond shaped than the westerners (but did not look asiatic). The Vietnamese and Chinese seemed to prefer darker skins than the rest and the Japanese had larger eyes (but not so large that they could be associated with the aesthetics of manga). When compared to the photographs of their ‘owners’ the majority of avatars revealed to bear similarities, principally with respect to the hair. While this could be indicative of a general tendency that the avatars resemble the person that they represent, it could also be that the users of Second Life that choose to reveal their photographs are those that construct their avatars according to their own appearance. In any case, despite including the peculiarities of the subjects, the 11 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. avatars remain very similar to one another, given that the personal characteristics of the ‘owners’ are attenuated or emphasised according to the general standard of beauty. For example, practically all of the female avatars have longer hair and slimmer bodies than their ‘owners’. The aesthetic reference for Second Life does not appear to be, however, exactly the same as that that prevails in the media, in particular, that that rules in fashion. In place of the straight shapes of the models, what predominates are the curvaceous forms that recall the pin-ups of the 1940s and 1950s: firm generous breasts, extremely narrow waist , rounded hips and well turned legs. The only black and overweight avatar located belonged to a woman who was equally black and overweight. At the same time that it was differentiated from the others by the inclusion of these two characteristics, the avatar obeyed the predominant logic in emphasising the properties of its ‘owner’ that were in consonance with the recurrent standard of beauty, such as long hair and full breasts. This same tendency to ‘perfect’ reality was found in the avatars of both women and men, whose bodies were in line with the current aesthetic standard. The results are rarely viable combinations of stunningly large breasts with impossibly thin waists and, amongst men, biceps and pectorals excessively increased and hips drastically reduced. Caricatures or perfect, female bodies are shown of in tight clothes, small blouses, exposed shoulders, whilst the male avatars are presented shirtless. Youth was also a constant in the avatars analysed. This does not mean a preference for smooth, wrinkle free skin (which could be a result of technical limitations): independent of the age of the subjects represented in their photographs, the firmness and distribution of body mass always suggests young avatars. Within the predominant traits in the construction of avatars it appears that there are also some rules implicit in the possible and desirable combinations. There is the case already mentioned of tanned skin and blonde hair. The (few) black avatars encountered had straight hair, thin lips and noses11 whilst the avatars with dark or curly hair tended to have light eyes and fair skin. These associations also contribute to the convergence of the appearance of avatars with a Caucasian standard of beauty. Finally, the generalized adoption of elements valued by a beauty standard compatible with that found in the Western media points to a prevalence of the desire to create a positive impression on other residents (which, as Goffman has been careful to indicate throughout his work, is not always the case in the representations of the self in the everyday off-line world). During the entire process of collecting the images for the corpus only two images were found of avatars whose appearance did not suggest the intention to please: an animal that appeared aggressive and an avatar without skin (or, more precisely, whose skin simulated a dissected body). The first was not included in the corpus due to a 11 It is not difficult to alter these characteristics, given that one of the standard male avatars of Second Life hás black curly hair, a wide nose and thicker lips. 12 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. lack of information concerning the nationality, and the second, in addition to this, was presented as an experiment that did not correspond to the usual appearance of the avatar. In both the cases, the images were accompanied by comments that apologised for or justified their unusual appearance. The similarity observed between all the avatars examined in Second Life, independent of the ethnic origin and geography of the subjects concerned suggests that the standard of beauty obtained from the value of youth, of physical fitness and of Caucasian ethnicity has a global hegemony. Interestingly, the set of avatars of Second Life evoke the classical period, in referring to a world inhabited by youths with perfect bodies that go about half naked, as in the example of the human figures in the Parthenon friezes (Sennett, 1997, p.37). 5. Further Considerations Taken at face value, the hegemony identified of the caucasian standard of beauty is not in the least suprising. The variations in the reading and appropriation of this standard of beauty in the avatars of diferent nationalities , however, makes the results significantly more interesting. The images of avatars in Second Life were here considered under their communicative ends – the characterization of an avatar seeks to inform the other about the self. For this to happen it is necessary that the meanings associated with the avatar are readable to the other residents of Second Life, about whom, in principle, nothing is known. For this reason, stereotypical indicators (for example, of gender) are abundant, whilst subtle, exotic, ambiguous elements, or those whose interpretation requires a refined repertoire, are rare. As all stereotypes are, by definition, exaggerated representations, the tendency to emphasise characteristics considered positive often results in accentuating that which was already at the limit. The result is a recurrence of the bizarre in certain elements, particularly the size of breasts and, in men, the musculature of the pectorals and biceps. The best proportioned figures of Second Life, on the other hand, are examples of a perfection that is impossible for flesh-and-bone bodies. It is interesting to wonder about this with respect to the fact that the self which is being represented by the perfect virtual body is inescapably bound to his physical body. In his discussion of how the representation of the body in virtual environments affect the mind, Biocca (1997) emphasizes the mismapping and intersensory conflict between the objective and the phenomenal bodies. Taylor (2002) claims that the users are always working though a reconcicliation of their online and offline bodies. A further situation can be described in which there are not just two bodies in play at once, but three – the third one being the ‘imagined body’, that is, the mental construct that the individual makes of his own body (as defined by Freud, 1980a, 1980b, 1980c and Lacan 1949, 1956) is taken into account. The fact that both the digital body of the avatar and the imagined body 13 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. are essentially symbolic and that both have the physical body as a primary source helps explain why avatars function so well as identity projections (in many senses even better than physical bodies)12. Differently from our daily life experience, the subjects that interact in a MUVE are not dealing only with differences between their imagined bodies and the reality of their flesh-and-bone ones. The shapes commonly chosen for the avatars imply a situation in which one’s imagined body is projected onto a symbol which has the appearance of a human body - and often has been made to resemble the body the subject is used to identify as himself (or, more popularly, as his own). His body of flesh-andbone and his imagined body both remain present, and active. In this context, the similarities between the appearance of the physical body and of the digital body potentialise the linkage between the three simultaneous embodiments of the self: the imagined, the physical and the digital. In the specific case of residents in Second Life, it is possible that the adaptation of the characteristics of the ‘real’ body to the hegemonic standards of beauty is an attempt to alleviate the pressure generated by the difference between the flesh-and-bone body and the unrealisable parameters that the cultural setting insists on reiterating (that is, how one’s physical body should be: young, healthy, slender, fair skinned, etc). The experience of establishing social relations through representations of the self based on a body that looks like one’s own, but has been changed to fit the standards of beauty is bound to have psychological as well as social implications. This investigation shows that, in Second Life, the reference for those alterations is the same for users distributed in five continents and that its appropriations are often stereotypical, a first step towards the identification and understanding of those implications. Further studies are necessary to verify and extend these findings. The experience of the work described here suggests that the most beneficial path to follow in extending this work could well be to deepen the qualitative sample. This would require ethnographic techniques, in particular the cross-referencing of records, participant observation, interviews and surveys. Despite this path sacrificing the representativity of quantity in the research, it permits particular attention to be given to the validation of the nationalities of the avatars included in the corpus, which is of the highest importance given the research problem being addressed. Following the same direction, ethnography seems to be the most appropriate area within which to find the information needed to deepen the discussion on the motivations and cultural frameworks that are in play when the users characterize their avatars. Acknowledgements The investigation reported is part of a research project sponsored by Conselho Nacional de Desenvolvimento Científico e Tecnológico, CNPq, Brazil. 12 This has strong connections to the therapeutic potential that Turkle (1997) sees in on-line experiences. 14 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. References Biocca, F. (1997). The Cyborg’s Dilemma: Progressive Embodiment in Virtual Environments. Journal of Computer-Mediated Communication, v3(2). Available at http://jcmc.indiana.edu/ vol3/issue2/biocca2.html#Body [10/01/2008] Fragoso, S. (2006). WTF a Crazy Brazilian Invasion. In: C. Ess, F. Sudweeks and H. Hrachovec (Eds.), Proceedings of the Fifht International Conference on Cultural Attitudes Towards Technology and Communication (CATaC), June 2006, Tartu, Estonia. Murdoch: School of Information Technology, pp. 255-274. Fragoso, S. (2007). Games Online como Terceiros Lugares. Proceedings of Intercom 2007 - XXX Congresso Brasileiro de Ciências da Comunicação, September 2007, Santos, Brazil. CDROM. Freud, S. (1980a). Sugestão e libido. Obras completas, v. XVIII, 1920-1922. Rio de Janeiro: Imago. (pp.113-118). Freud, S. (1980b). A dissecação da personalidade psíquica. Conferência XXXI. Obras completas, v. XXII, 1932-1936. Rio de Janeiro: Imago, (pp. 63-84). Freud, S. (1980c). Identificação. In:_____ Obras completas. v.XVIII 1920-1922. Rio de Janeiro: Imago. (pp.133-139). Goffman, E. (1985). A Representação do Eu na Vida Cotidiana. Petrópolis: Vozes. Kafai, Y.B; Fields, D. A.; Cook, M. (2007). Your second selves: avatar designs and identity play in a teen virtual world. Proceedings of Situated Play - DiGRA 2007 Conference. Tokyo, Japan, 24-28/09/2007. Lacan, J., (1949). “El Estadio del Espejo como Formador de la Función del yo (je)”. Presented in XVI Congreso Internacional de Psicoanálisis, Zurich, 17/07/1949. Online at El Oriba, http:// www.elortiba.org/lacan5.html [14/01/2008] Lacan, J., (1956). La Pregunta Histérica. Seminario 3, Las Psicosis, clase 12. Online at El Oriba, http://www.elortiba.org/lacan6.html [14/01/2008] Linden Research Inc. (2007). Second Life Virtual Economy, Key Metrics (BETA) Through November 2007. Published 18 December 2007. Online at http://static-secondlifecom.s3.amazonaws.com/economy/stats_200712.xls [03/01/2008] Linden, M. (2007) Economic Key Metrics for October and November 2007, posted in Second Life Blog, 18/12/2007. Online at http://blog.secondlife.com/2007/12/18/economic-key-metricsfor-october-and-november-2007/ [03/01/2008] Miller, H. (1995). The Presentation of Self in Electronic Life: Goffman on the Internet. Embodied Knowledge and Virtual Space Conference, Goldsmiths' College, University of London. Online at http://www.psicopolis.com/psicopedia/selfweb.htm [03/01/2008] Reid, E. M. (1991). Electropolis: Communication and Community On Internet Relay Chat. Department of History, University of Melbourne, Australia, 1991. Online at http://www.irchelp.org/irchelp/misc/electropolis.html [03/01/2008] Reid, E. M. (1994). Cultural Formations in Text-Based Virtual Realities. Thesis submitted in fulfillment of the requirements for the degree of Master of Arts. Cultural Studies Program, Department of English, University of Melbourne, Australia. Online at http://www.aluluei.com/cult-form.htm [03/01/2008] Rheingold, H. (1998). The Virtual community. Online at http://www.rheingold.com/vc/book/ [12/01/2008] 15 Fragoso, S. & Rosario, N. M. (2008) Just like me only better. In: Fay Sudweeks, Herbert Hrachovec, Charles Ess. (Org.). Cultural Attitudes Towards Technology and Communications. Vol 4 Murdoch Australia: School of Information Technology - Murdoch University, p. 314-327. SecondLife Wiki (n.d.). Metrics Glossary. Online at http://wiki.secondlife.com/wiki/ [03/01/2008] Sennett, R. (1997) Carne e pedra. Rio de Janeiro: Record. Smith, M.A, S.D.Farnham & S.M. Drucker (2000). The social life of small graphical chat spaces. Proceedings of the SIGCHI conference on Human factors in computing systems, CHI Letters, 2(1), pp. 462-469. Tabet, A. (2007). O principal desafio de fazer humor é fazer o que é engraçado para você e para os outros, Interview to Cronicamente Viável, 9/10/2007. Online at http://www2.uol.com.br/cronicamenteviavel [12/01/2008] Taylor, T.L. (2002). Living Digitally: Embodiment in Virtual Worlds. In R. Schroeder (Ed.) The Social Life of Avatars: presence and interaction in shared virtual environments. London: Springer-Verlag. Turkle, S. (1997). A Vida no Ecrã: a identidade na era da internet. Lisboa: Relógio D'Àgua. Vasalou, A., Joinson, A. N.; Pitt J. (2007) Constructing my online self: avatars that increase selffocused attention. Proceedings of the SIGCHI conference on Human factors in computing systems. San Jose, California, USA, pp. 445-448. 16