Multi-user online environments involve millions of participants world-wide. In these online communities participants can use their online personas --- avatars --- to chat, fight, make friends, have sex, kill monsters and even get married. Unfortunately participants can also use their avatars to stalk, kill, sexually assault, steal from and torture each other. Despite attempts to minimise the likelihood of interpersonal virtual harm, programmers cannot remove all possibility of online deviant behaviour. Participants are often greatly distressed when their avatars are harmed by other participants' malicious actions, yet there is a tendency in the literature on this topic to dismiss such distress as evidence of too great an involvement in and identification with the online character. In this paper I argue that this dismissal of virtual harm is based on a set of false assumptions about the nature of avatar attachment and its relation to genuine moral harm. I argue that we cannot dismiss avatar attachment as morally insignificant without being forced to also dismiss other, more acceptable, forms of attachment such as attachment to possessions, people and cultural objects and communities. Arguments against according moral significance to virtual harm fail because they do not reflect participants' and programmers' experiences and expectations of virtual communities and they have the unintended consequence of failing to grant significance to attachments that we take for granted, morally speaking. Avatar attachment is expressive of identity and self-conception and should therefore be accorded the moral significance we give to real-life attachments that play a similar role.
Read full abstract