Real Ethics in a Virtual World

Real Ethics in a Virtual World Victoria McArthur Abstract Centre for Digital Humanities This paper investigates the ethics of the appearance and behavior of avatars in massively multi-user online communities, in particular, avatars created for virtual business interactions in Second Life. The ethics of research conducted with avatars in 3D online environments is also discussed. Brock University St. Catharines, ON, Canada L2S 3A1 [email protected] Keywords Avatar, Second Life, ethics, virtual reality experiences ACM Classification Keywords H5.m. Information interfaces and presentation (e.g., HCI): Miscellaneous. Introduction Copyright is held by the author/owner(s). Second Life is a user-driven online virtual world, originally conceived by Philip Rosedale, CEO of Linden Labs [7]. It is a “metaverse”, a term coined in Neal Stephenson’s 1992 novel Snow Crash [10], in which people in real physical spaces create and maintain avatars for the purpose of interacting with one another in 3D virtual environments. Avatars, in the case of Second Life, are three-dimensional representations of the user. The appearance of users’ avatars can be manipulated with a powerful avatar appearance editor. CHI 2008, April 5 – April 10, 2008, Florence, Italy ACM 978-1-60558-012-8/08/04. Similar to Stephenson’s metaverse, Second Life goes beyond mere social interaction. Citizens of Second Life can purchase $100 of its currency, Linden Dollars, for approximately $0.68 USD1. Linden Dollars can be exchanged within Second Life for virtual goods and services and can be converted back into USD. The ability to conduct monetary business within Second Life is perhaps why real companies, such as IBM and Sun Microsystems, are moving into Second Life [11]. Representatives deployed within Second Life adopt an avatar to represent their presence in the virtual space. In this environment, interaction between avatars both represents and replaces face-to-face interaction. The importance of cues present in face-to-face interaction is known. Goffman stresses the importance of both verbal and nonverbal cues in our ability to generate impressions of ourselves in social situations [5]. In Second Life, these forms of communication may occur either by text entry or by voice chat and by avatar gestures and appearance respectively. In his chapter on communication out of character, Goffman suggests that we endeavor to suppress cues that detract from the impressions we are trying to make [5]. This can be seen, for example, in business interactions where a representative might suppress the use of slang for the duration of the interaction. Extensive work has been done to address the limitations currently present in avatar-based communication (see e.g., [2, 3]). The present work does not address this problem but instead focuses on some of the ethical concerns of such communication in business situations. 1 Exchange rate retrieved from the Linden Dollar Exchange website on January 7, 2008. Avatars Avatars are the visual representation of online personae and are often constructed solely for use in online social interaction. However, for many, the avatar is more than a visual representation of self; it is a vessel through which one projects oneself and develops an online identity. The essence of the avatar may bear striking resemblance to that of the creator, or the avatar may be a means for the creator to experiment with aspects of her/his identity within an online social context. This fragmentation of self may seem abnormal, but Turkle suggests that we are already accustomed to managing multiple selves in our daily lives and that this sense of multiplicity easily extends into virtual worlds [12]. Similarly, within real life, it is not uncommon to present limited versions of one’s self at any given time or to reserve certain idiosyncrasies for specific social situations [5]. This behaviour easily extends into virtual environments. As stated by Rehak, avatars are “ambassadors of agency”; they are vessels for action, but do not directly represent the personality of the individual at the helm [8]. Thus, the behaviour of the avatar may not be typical of the mannerisms of the person in control. Turkle goes on to suggest that the degree to which adopted personalities can differ from our own may vary. Some people enter virtual communities for the explicit purpose of role playing. Others may choose to create an accurate virtual representation of themselves [12]. This virtual representation can exist as an inworld duplicate of the creator, whose actions may be governed by a set of social guidelines identical to the pilot. The anonymity that is unique to online social interaction presents the user with the ability to create a mask through which (s)he may express an alternate, or perhaps even ideal self – either physically or socially. Bardzell and Bardzell argue that virtual existence provides a level of safety that extends the limits “to which people are willing to extend themselves” [1]. in Second Life further adds to the myriad of possibilities. Gee takes this complex relationship between player and avatar further by breaking it down into three identities: virtual, real and projective. The virtual identity belongs to the avatar; it represents the personality of the character being played. The real identity is that of the person piloting the avatar. The projective identity is the interface between the two and the “feedback loop” through which values are projected upon the virtual identity by the real identity. Essentially, projective identity is the space in which the player evaluates and reevaluates the morality that guides the actions of her/his avatar [4]. Avatars in Second Life In Second Life, avatars are three-dimensional representations of the user, created by the user. When signing up for a Second Life account, users must decide upon a user name for their avatar which, unlike the avatar’s appearance, is permanent. So, while one can continually manipulate the avatar’s appearance, any reputation attributed to the user name is not so easily altered. Second Life provides a powerful avatar appearance editor (see Figure 1) that grants the user the ability to manipulate a variety of physical features including body thickness, height and muscularity. Users can also modify the textures and colours on their bodies and clothing. Many users decide to remain human in appearance, but some decide to adopt anthropomorphic animal avatars. The ability to generate custom content Figure 1. The avatar editing tool in Second Life allows users to freely manipulate the many traits that contribute to their appearance. Company avatars in Second Life Companies such as IBM, Sun Microsystems and Dell are among the biggest corporations moving into Second Life [11]. IBM has been using Second Life as a recruitment tool and to meet with clients, for virtual teleconferencing and as a place to advertise some of its services. With their employees deployed in-world, companies may wish to develop dress codes that take advantage of the malleability of employee avatar appearance. Professional appearance in real-world situations is important in business relations. For this reason it is not uncommon for corporations to employ dress codes, which can extend beyond attire to deal with hair colour, body piercing and personal hygiene. These are the properties of appearance that are most easily manipulated in the real world. The degree to which one can manipulate the appearance of one’s avatar in Second Life is almost infinite by comparison. One can easily modify one’s avatar’s appearance down to the most minute details. Not only is this nearly instantaneous, but can be controlled with great precision. When translated into a 3D virtual world such as Second Life, the dress code has the potential to evolve into a more general “appearance code”. For example, in addition to enforcing rules upon attire and hair colour, the employer may require that the representative avatar match the employees’ real-world physical appearance as closely as possible, within a reasonable degree of accuracy. This raises an important question: how much control should an employer have over the appearance of employees’ avatars, given that they are representing the company in a 3D virtual environment? The employer has an interest in ensuring that they are represented in a manner consistent with company image. The employer’s desire for an avatar appearance code may simply be grounded in the same logic by which uniforms are utilized in real life; the company may simply want its employees to be easily recognized. People controlling avatars, on the other hand, may expect a certain degree of freedom of expression. These two factors may be at odds with one another. The former also assumes that the same conventions which guide successful interaction in real life can be directly mapped into virtual life; this may not always the case. Another issue raise is that of intellectual property. In general, one would expect that ownership of an avatar would reside with its creator. However, if the avatar is created by company request, ownership of the avatar is less clear. Figure 2. Visiting IBM in Second Life. IBM is one of many companies using Second Life as a recruitment tool. Moreover, does the appearance of the avatar, either in its professionalism or in its resemblance to the pilot personality, have any effect on the success of business conducted in these virtual environments? If established negotiations move from the virtual world into real life, how are the relationships that were forged in Second Life affected? Previous research has addressed issues regarding trust in non face-to-face communication [6]. If the appearance of the avatar differs greatly from that of the real identity, it is conceivable that there could be a loss of trust following the transition from Second Life to real life. Thus accuracy of the avatar’s representation of the pilot may have an impact on business. Furthermore, what should the company do about behaviours outside of office hours? As previously mentioned, account names in second life are static and appear floating above the user’s head in-world. This means that it may be possible to recognize the employee by name, outside of working hours. As certain areas of Second Life contain questionable content, this could be problematic since the actions of the employee (or the employee avatar) reflect on the company as well. One solution to this would be for employees to maintain at least two accounts: one for company-related interactions and one for use “off the clock.” Many of these issues could be easily alleviated by the implementation of a contractual clause. Such a clause could address the issues of the pilot’s conduct in Second Life and ownership of any virtual artifacts created therein. Although this is one of many potential solutions, it may not be immediately obvious to employers who are eager to explore virtual business. The aim of this research is to raise awareness of these issues. Virtual ethics In order to conduct research into the ethics surrounding these issues, the ethics approval process may also need to be revisited. Current human research ethics review practices are based on research conducted in real life with human participants. This may not directly apply to behavioural and interview-based research involving online personae. The intent of ethical review is often to protect the identity, personal information and physical safety of the research participants. Of course, avatars are immune to physical harm, so that aspect of the ethical review is irrelevant. However, it is less clear whose information we are protecting if we are interviewing an avatar – that of the human piloting it, or the avatar’s persona? The ethics we apply to research conducted on the Internet have sometimes had to be modified as necessary [9]. Those ethics review processes which are not prepared to handle research involving avatars may also need to be updated. Where do we go from here? The long term goal of this research is to explore the various practices adopted by companies with strong Second Life presence regarding the appearance of employee avatars. Another goal is to survey employees of such companies to determine how important it is for them to express themselves in this medium – if they even perceive a difference between appearance demands in either environment. The question posed is: are the demands of the employer considered reasonable by the employees? Current guidelines for research conducted involving avatars may need to be re-examined and perhaps updated, to address the differences that set avatars apart from human participants. Applying these standards to individuals who do not have physical bodies or the same types of privacy concerns is cumbersome and could possibly be missing some ethical concerns unique to their existence. Companies will need to begin to consider ethics and “appearance codes”, and research ethics boards will have to start thinking virtually. The majority of the requirements currently in place for research are wellsuited to research conducted in real life, but perhaps not in virtual worlds. Acknowledgements This research has been made possible by the assistance of Dr. Barry Joe, Dr. John Bonnett, and the Centre for Digital Humanities at Brock University. Special thanks to Robert J. Teather for assistance with editing. References [1] Bardzell, S., and Bardzell, J. 2006. Sex-interfaceaesthetics: The docile avatars and embodied pixels of Second Life BDSM. Workshop on Sexual Interactions: Why We Should Talk About Sex in HCI held at CHI ’06, 2006. [2] Colburn, A., Cohen, M. F., and Drucker, S., The role of eye gaze in avatar mediated conversational interfaces, Technical Report MSR-TR-2000-81. Microsoft Research, 2000. [3] Freier, N. G. Children distinguish conventional from moral violations in interactions with a personified agent. ACM student research competition: Ext. Abstracts CHI ’07. ACM Press (2007), 2195-2200. [4] Gee, J.P., What video games have to teach us about learning and literacy. Palgrave MacMillan, New York, NY, USA, 2003. [5] Goffman, E. The presentation of self in everyday life. Doubleday, Garden City, NY, USA, 1959. [6] Nguyen, D. and Canny, J. MultiView: Improving trust in group video conferencing through spatial faithfulness. In Proc. of CHI’07, ACM Press (2007), 1465-1474. [7] Rymaszewski, M., Au, W.J., Wallace, M., Winters, C., Ondrejka, C. and Batstone-Cunningham, B. Second Life: The official guide. Wiley, 2007. [8] Rehak, B. Playing at Being: Psychoanalysis and the avatar. The video game theory reader. Routledge, NY, USA, (2003), 103-127. [9] Smith, M.A. and Leigh, B. Virtual subjects: Using the Internet as an alternative source of subjects and research environment. Behavior Research Methods, Instruments, & Computers (1997) 29, 496-505. [10] Stephenson, N. Snowcrash. Bantam Books, NY, USA, 1992. [11] The top 8 Second Life virtual businesses. http://www.pcadvisor.co.uk/news/index.cfm?newsid=9 279. Accessed February 4, 2008. [12] Turkle, S. Identity Crisis. CyberReader. Allyn and Bacon, Boston, MA, USA, (1998), 78-91.