Animated game pieces. Avatars as roles, tools and props [1]

Jonas Linderoth (LearnIT/Göteborg University)


Since Turkle’s seminal work Life on the screen (1995), avatars have been described as a form of alter egos, being means for exploring and playing with identity. Following this line of reasoning, computer gaming is seen as an activity where we become immersed in a fictitious world, pretending to be the character we play. Drawing upon empirical observations of children’s game-play I argue that the relation between the avatar and the player is a more multifaceted affair. The meanings of avatars depend upon how they are framed by the player, thus they can have at least three different functions. Avatars can become roles for socio-dramatic interaction. As extensions of the player’s agency, avatars can become tools for handling the game state. Finally when choosing and using avatars in the presence of others, avatars can become a part of our identity, not as alter egos but as props for our presentation of self on the social arena surrounding the game.

Gaming and identity
Compared to film and literature, the interactive nature of computer games stands out as a unique property. The ability of games to let the viewer/reader/listener control fictive characters in a frame set by the designer has been subject for assumptions about the nature of the gaming experience and the potential impact games might have on players. The relation between the player and the game character is thought of as a special, mystic, and yet unexplored bond, a bond that has to do with identity. Game characters are for instance sometimes labelled as avatars[2] a term which originates from Hindu mythology and means the incarnation a God takes when descending to earth. This suggests that game characters are seen as a form of alter egos, as means for playing with identity. For instance Murray (1997) stated that: “Even when avatars are crudely drawn or offer a very limited choice of personalization, they can still provide alternate identities that can be energetically employed” (p. 113). The idea that identity is at stake when we enter digital worlds is even more developed in the writings of Sherry Turkle (1995). Even though her studies mainly concern virtual online communities like MUDs and chat rooms her works have influenced the way we understand the player – character relation in computer games.

Children are sometimes thought to be especially vulnerable to the influence that computer games might have on identity, still as Walkerdine (1998) puts it, there is a need for more research on this topic:

What for example, is the relationship between the screen world, the inner world and the outside world? While children talk clearly about the relation of fantasy to reality, they use the personal pronoun ´I´ to describe the actions that they make the protagonist do, as in ´I beat the boss´. This suggests that the learning of moves also means learning to move like that character, and therefore a complex relation of identification, which has not begun to be explored. (p. 244 f.)

This paper can be seen as a contribution to the study of this unexplored relation. The findings suggest that the assertion that children’s game play is a matter of “a complex relation of identification” based on the fact that children use the personal pronoun ´I´ can be questioned. Though, as Walkerdines quotation indicates, the usage of ‘I’ in game contexts are bye no means trivial. Avatars can have a roll in identity formation, not in the mystified sense of being “alternate” personalities, but rather as a potential resource for the child’s presentation of self in the social context at hand.

Against the idea of immersion
The idea that the game experience is based on immersion, that the player identifies herself with the avatar, has been criticized in the literature about computer games. Some writers describe this idea as an untrue fabrication which is not based upon systematic knowledge. Ryan (2001) calls Murray’s (1997) claims a holodeck myth. Newman (2002) talks about the myth of the ergodic and means that for the player the character is better understood as a piece of equipment that the player makes use of. Salen and Zimmerman (2004) label the idea of a player – character relation based on identification as the immersive fallacy (p. 450). This immersive fallacy is a misleading notion about how games work, something which according to the writers is problematic since it might effect game design. Instead Salen and Zimmerman relies upon an ethnographic study on paper and pencil role-playing games made by Fine (1983) and claim that the conclusions in this study is applicable to digital games as well. According to Fine’s study, playing role playing games is an activity where meaning is generated on different levels. The fictive ‘story’, the level where the game characters exist, is thus just one dimension of the game experience. Fine uses the concept frame (Goffman, 1961, 1974/1986) in order to describe how role-playing works. A frame is a definition of a situation which the participants in the situation more or less share with each other. This frame can be seen as their mutual answer to the question “what is going on here” (Goffman, 1974/1986, p. 8). Participants relate events, actions and utterances to the framework at hand and use the frame as a resource for giving meaning to their experiences. Frames can, according to Goffman, be laminated, meaning that participants can agree to have a number of different frames at work in the same situation. According to Fine (1983) role-playing is a laminated activity, constituted by three different frames. The outer rim for role-playing is the primary framework, the commonsense understandings we have of ‘reality’. Then there is a second frame which Fine calls the game context, here the participant’s actions are governed by the game rules so that they constrain their actions in order to uphold the game, i.e. take their turns, follow the game’s structure for interaction etc. Finally there is the inner, socio-dramatic frame where actions are governed by the participant’s mutual agreement to pretend that they are fantasy characters. Salen and Zimmerman (2004) comment upon this conclusion:

This three-fold framing of player consciousness – as character in a simulated world, as a player in a game, and as a person in a larger social setting – elegantly sketches out the experience of play./…/ In digital games the same multi-layered phenomena occurs. (p. 454)

One could argue that Salen and Zimmermans leap from the early pen and paper role-playing games to the digital games of today is a problematic generalization since they have no empirical support (other than their own game experiences) to back up their statement. Yet there have been some studies on computer games which give a similar picture.

In an ethnographic study of how children position themselves in computer settings, Johansson (2000), like Walkerdine (1998) observed how children used the personal pronoun ‘I’ when talking about the actions of their avatars. In her conclusions Johansson is critical to psychological interpretations of this phenomenon. Johansson claims that expressions like ‘I died’ are something that children say in play roles which are clearly separated from their own self. According to Johansson this is the same thing as taking on different perspectives in socio-dramatic play were children can switch between the present and the past tense, between talking as characters and talking about the characters. Even though Johansson does not use Goffmans frame concept her findings suggests that at least Fine’s distinction between a primary framework and a socio-dramatic frame can be applicable for digital games as well.

In previous studies I have explicitly focused children’s interaction patterns during computer game play (Linderoth, 2004). The results shows that children establish their interaction when playing computer games by shifting between different frameworks for handling the things they see on the screen and by relating, transforming and/or dissociating aspects of the world outside the gaming situation. There where three patterns where children related features in the game to different frames:

1. The aesthetics of the game, here the children made short comments about how they liked the game material (sounds, images, plot etc.). Sometimes the aesthetic agenda also became the driving force for their game play. Simulations then became design activities were the beautiful, cool and attractive were given priority over functionality.
2. The theme of the game, here children saw the game as a representation and utilized their previous experiences of the game in two ways. Either they establish short moments of socio-dramatic play, or they made assumptions about how the rules of the game worked based on their knowledge of the represented phenomenon.
3. The rules of the game, here children treated game features in accordance to the game’s goal and focused on how things could be used, not what they represented.

There were also dynamic patterns where things from the wider social setting had to be handled in the game situation. For example children who were close friends sometimes had to negotiate the situation in order to reach an agreement which allowed them to compete with each other. These dynamic patterns as well as the aesthetic focus correspond Fine’s observation about a primary, everyday framework as a part of game experiences (it is the child, not the player or the character who finds some parts of a game to be appealing). The socio-dramatic interaction pattern is consistent with Fine’s inner frame where players become characters and the rule focused pattern is similar to what Fine calls the game context. Thus the identified interaction patterns in the study support the idea that the experience of playing computer games is a “three-fold framing of player consciousness” (Salen & Zimmerman, 2004, p. 454).

Analysis of the player – avatar relation
As mentioned, Johanssons critique of psychological interpretations only makes the distinction between fiction and reality, between ‘I’ as a character and ‘I’ as my own self. In this paper I will draw upon empirical material in order to argue that while playing computer games participants can produce a third meaning of ‘I’.

The analytical method employed was Interaction Analysis as it is described by Jordan and Henderson (1995). The aim of Interaction Analysis, according to these writers, is to identify regularities and depict mechanisms in how people interact and conduct their affairs. Interaction Analysts share the assumption that knowledge and action are social phenomena, situated in social and material ecologies. Knowledge is not localized in each individual’s head, but seen as situated in the interaction between participants of specific communities. Interaction Analysis is done with the use of video data which the researcher transforms into detailed transcripts. Theories of learning and communication are then used as analytical tools in order to examine what the meaning of participants actions are in the analyzed session.

Design and collection of data
The excerpts in this paper come from a data collection of 23,5 hours of video data from 24 gaming session with totally 36 children in the age of 6 to 11. As a whole the material has been previously analyzed (see Linderoth, 2004; Linderoth, Lindström & Alexandersson, 2004). For this study two sessions which were especially interesting in accordance to the question about identity were chosen to be analyzed.

In the first session two girls, who are sisters, Bea (six years old) and Elin (eight years old) played Perfect Dark on a Nintendo 64 console for 60 minutes. The children were playing the multiplayer option and were sometimes on the same team fighting against bots, sometimes in opposing teams. The session took place in the children’s room at home. Recording was done with one stationary camera. A parent was called upon at two occasions. The researcher overheard the session from next door. The game is an agent style action adventure with the possibility to play a multiplayer scenario on a split screen. In the multiplayer option 1 - 4 players can be either with or against each other when playing. The players can also have ‘bots’ (character which are controlled by the games artificial intelligence) in their teams. The basic way of gaining points in order to win is to defeat members of the other team without getting your own avatar killed.

In the second session three boys who are friends, Felix (eight years old) Anders (eight years old) and Simon (eight years old) played Super Smash Bros. Melee on a Gamecube console for 60 minutes. The session took place in the living room at Anders home. Super Smash Bros. Melee is a humorous fighting game. The goal of the game is to gain points by knocking out your opponents. Before each round the players choose from a broad variation of different avatars. The avatars represent characters from other games. Recording was done with one stationary camera. The researcher overheard the session from next door.

The excerpts below are structured in five columns. The first and second columns are the number of the turn and the name of the participant. Column three tells what the participants are saying and column four tells about actions and events in the physical room. In column five events in the game are described.

‘I’ the player
The first two excerpts bellow illustrate how children switch between frames when playing computer games and thus alter the meaning of the words they use. In these excerpts Elin and Bea are playing with bots in their teams. Sometimes they enter short strips of socio-dramatic interaction where they pretend to talk with their bots. In these instances they call their bots ‘pal’. Since they are not familiar with the concept ‘bot’ they also use the word ‘pal’ outside the socio-dramatic frame.

In this excerpt Elin makes a frame shift within the same turn. She first talks to her fictive ‘pal’ and then makes a statement to her sister that she has seen her ‘pal’. Even though she uses the same word the meaning is different. Hi pal is a representation of a greeting; it is supposed to be treated in accordance to the tacit understanding that follows with socio-dramatic play. A special kind of awareness where we agree to treat the world as if it had other properties than those we perceive. The listener is supposed to accept the idea that we treat the bot as if he could hear what we say. This only goes for the first part of the turn. In the second part Elin addresses Bea and makes a statement which is not supposed to be treated from an as if approach. Here the word ‘pal’ does not signify a fictive friend but the actual bot within the game. In a second excerpt the same phenomenon occurs.

Turns 1 and 2 are utterances about the game state, what the children are doing and how they experience the things that happen. Still they use the term ‘pal’ which is illusory since it gives the impression that this is socio-dramatic interaction. In turn 3 Elin makes one utterance where she for a moment talks to the fictive characters.

The two excerpts above suggest that entering an alternate reality only is one potential way of experiencing computer games. An assumption which challenges the opinion that the relation between, screen world, inner world and outside world has a mystic, unexplored impact on children’s identity.

In the third excerpt Bea and Elin are playing in the same team against a team of bots. The girls’ team has yellow avatars, their opponents are red. In this excerpt it is possible to see the multilayered character of the gaming activity. At the beginning of the excerpt the children have established a socio-dramatic frame for their interaction. In turn 1 – 5 they thus pretend to be two agents which have radio contact with each other. In turn 6 - 7 they comment upon their own interaction. In turn 8 – 15 they try to re-establish the socio-dramatic frame. At the end of the excerpt in turn 16, Elin’s interaction becomes rule focused.

In accordance with the analytical focus on identity, turns 7 and 16 is interesting. When Elin in turn 7 says The red one killed me.. she is repeating what she said in turn 5. In turn 5, the statement was said ‘in character’ as a part of her pretence play to be an agent. In turn 7 this frame is broken and the utterance is best understood as a comment about the impossibility of making sense of what she has said within the socio-dramatic frame. Something that she explicitly comments in the rest of turn 7 the red one killed me then you can not say that the red one killed me. What happens here can be described as an unintentional frame break. Without thinking about it, Elin makes the statement that she is dead and when she realize that it is not possible to make a comment about your own death when you are dead, the situation becomes comical. Thus Elin is perceptive about the difference between her identity in a primary framework as the producer of the socio-dramatic frame and her identity within the socio-dramatic frame as an agent. It is also worth mentioning that the information Elin communicates to Bea in turns 2 – 5 tells the reality about her game state. When she says that she sees a red bot, she gives Bea the opportunity to use this information strategically. Thus in turn 16 the utterance I died is not said in a socio-dramatic frame. Something we can tell from the fact that the utterance is said in an everyday voice and the last part of the turn we will probably lose this time. Here we have a third frame at work, the competitive, rule focused level of meaning. The utterance I died is here said to a team mate in order to give her the information that the team has lost a point. While ‘I’ in the primary framework denotes the speakers own physical body and ‘I’ in the socio-dramatic frame signify the fictive character, ‘I’ in turn 16 means the speakers agency within the game activity. I claim that this way of using ‘I’ is an everyday phenomenon and not something unique for computer games. When our agency in a certain activity system is extended outside our own body we talk about this extension as a part of ourself. For instance a horse and a rider tend to become a unit, and, while only the horse is exhausted after a ride, we still say ‘I trotted’. Likewise we talk about our game pieces in board games as a part of ourself, and can in the game of Monopoly claim that ‘I’ stand on chance. [3]

A fourth excerpt strengthens this line of reasoning. Now Bea and Elin has changed game mode so that they compete against each other. Each child also has a bot in their team. Bea is the yellow team, Elin is the red. When we enter the excerpt the children are discussing their difficulties to keep up with their bots.

Turn 4 at the end of the excerpt shows us how fragile the meaning of utterances can be in gaming activities. Here Elin wants to help Bea so that she perceives that her bot is close to her avatar. In order to help Bea, Elin says I saw that you saw a yellow. Then like in the excerpt above she repeats her own words, this time with a smile and a giggling voice. The humor in the utterance comes from the unusual phrasing I saw that you saw. In order for this utterance to make sense the listener must be aware of the specific context and be receptive for the different frames that are at work here. The word ‘I’ signifies Elin’s identity in the everyday sense, a girl sitting on the floor with a control pad in her hands. On the other hand the word ‘you’ which follows closely here denotes Bea’s avatar since it is the avatars point of view that is represented on the screen. Thus the statement can only be understood if we have parallel frames at work. Again we see that children can be aware of the multilayered nature of the game experience and there is no need for speculations about the impact computer games have on children’s identity.

Avatars as part of the player’s setting
Does the line of reasoning above mean that there never are instances when it makes sense to talk about avatars in relation to identity? If we follow the theoretical tradition from Goffman, then another picture reveals it self. For Goffman (1956) the study of identity is not at matter of revealing the ‘core’ of someone’s personality. According to Goffman ‘the self’ is something that we produce and re-produce in our everyday life. When we enter the presence of others we try to manage the image we give of ourselves. We give performances were we, in order to control the information we give to our observers, use “expressive equipment” (p. 22) like clothing, facial expressions, body langue etc. One sort of expressive equipment Goffman refers to as the setting “involving furniture, décor, physical layout, and other background items which supply the scenery and stage props for the spate of human action played out before, within or upon it.” (1956, p. 22)

Applied to the gaming activity, avatars can be seen as a part of the setting, as possible means for the players to present themselves. In two excerpts below this is illustrated. In these instances of interaction one of the children shifts frame so that he becomes detached from the values and connotations of a female avatar. When we come in to the excerpt the three eight year old boys are just about to start a fight and are choosing avatars. The avatar Zelda is a blond princess with a pink dress.

In turn 2, Anders giggles while his cursor is on Zelda. A piece of interaction which signals that something out of the ordinary is happening. In turn 3 Felix confirms that choosing Zelda is not something that boys are supposed to do. After this excerpt Anders chooses to play with Zelda anyway. While the fight loads he sits quiet.

In turn 3, Anders claims that Zelda is awesome, and he does this immediately when the game starts, he has only made one unsuccessful attack. Since Simon in excerpt 5 did not think that it was appropriate to play with the pink princess, Anders risked being harassed for his choice. When choosing an avatar with so strong female connotations he risked having this meaning attached to his identity. Then in excerpt 6, he points out that the chosen avatar is awesome to play with and thus changes the frame. Zelda stops being a princess in a pink dress and becomes equipment in the game context.

The multilayered nature of gaming is probably what makes it possible for children to play with cultural material that they probably would have disregarded in another context (this phenomenon has been observed elsewhere see Newman, 2002, for a discussion). The fact that Anders has to handle the situation with a frame shift suggests that the cultural meaning associated with an avatar can blend into the gaming activity and become attached with the player. Avatars are part of the ‘setting’ and thus provide the players with means for presenting themselves. However this is not a simple process of causality. We can not simply make connections between the player and the avatar he or she plays. The relation is, as we have seen, enacted. Thus it is possible to play an avatar with detachment in an ironical manner. The player would then probably establish a self presentation which is contradictory to the avatar.

Conclusions: Rethinking the player - avatar relation
Fine suggested that pen and paper role-playing games is best understood as a three layered activity where the everyday reality is the outer frame, within this frame there is a game context governed by the rules, finally there is an inner frame where the players pretend to be in a fictive world. This paper suggests, in accordance to Salen & Zimmerman (2004), that the metaphor of ‘three layers’ is a reasonable way to understand digital games as well. This has specific implications when it comes to understanding the player – avatar relation.

1. A fictive character that you can pretend to be, a role
2. A piece of equipment, a tool which extends the player’s agency in the game activity
3. A part of the players setting, props which can be used as a part of the players presentation of self

When the avatar becomes a tool for the player, an extension of her or his agency, the term ‘I’ refers to the player – avatar unit. This is not a phenomenon which is unique for the gaming activity, it occurs in other cases when our ability to act in a certain activity systems is mediated by a tool.

This way of reasoning about the player – avatar relation gives a down-to-earth understanding about the nature of the game experience. The psychological/cognitive way of understanding the gaming activity depicts an image were we are to understand the relation between the screen world, the inner world and the outside world. A notion which wrongly takes for granted that the game experience always is a matter of representation.

Notes [ back ]

[1] The research presented here was funded by the Swedish Knowledge Foundation's (KK-stiftelsens) research programme Learning and IT (LearnIT)

[2] Sometimes this term is retained for incarnations in online communities. I will here use it for all player controlled characters in both on and off-line games.

[3] This line of reasoning resembles of how Wilhelmsson (2001) talks about a tactile motor/kinaesthetic link which makes it possible for us to constitute a “game ego.”



Fine, Gary Alan. 1983. Shared Fantasy. Chicago: University of Chicago.

Goffman, Erving. 1956. The presentation of self in everyday life. Edingburgh: University of Edingburgh social science research centre.

Goffman, Erving. 1961. Encounters: Two studies in the sociology of interaction. Indianapolis: Bobbs-Merill.

Goffman, Erving. 1974/1986. Frame analysis: an essay on the organization of experience. Boston: Northeastern Univ. Press.

Johansson, Barbro. 2000. Kom och ät - Jag ska bara dö först, Datorn i barns vardag. Göteborg: Etnologiska föreningen i Västsverige.

Jordan, B, and A Henderson. 1995. Interaction Analysis: Foundations and Practice. The Journal of the Learning Sciences 4 (1):39-103.

Linderoth, Jonas. 2004. Datorspelandets mening: Bortom idén om den interaktiva illusionen. Göteborg: Acta Universitatis Gothoburgensis.

Linderoth, Jonas, Berner Lindström, and Mikael Alexandersson. 2004. Learning with computer games. In Toys, games and media, edited by J. Goldstein, D. Buckingham and G. Brougere. London: Lawrence Earlbaum.

Newman, James. 2002. The myth of the ergodic videogame. Some thoughts on player-character relationships in videogames. In Game studies.

Murray, Janet H. 1997. Hamlet on the holodeck: the future of narrative in cyberspace. New York: Free Press.

Ryan, Marie-Laure. 2001. Beyond myth and metaphor - The case of narrative in digital media. In Game studies.

Salen, Katie, and Eric Zimmerman. 2004. Rules of play. Game design fundamentals. Cambridige: The MIT Press.

Turkle, Sherry. 1995. Life on the screen: identity in the age of the Internet. New York: Simon & Schuster.

Walkerdine, Valerie. 1998. Children in Cyberspace: A new frointer. In Children in culture. Approaches to childhood, edited by K. Lesnik-Oberstein. Basingstoke: Macmillian press LTD.

Wilhelmsson, Ulf. 2001. Enacting the point of being - Computer games, interaction and film theory. Ph.D, Department of Film and media studies, Copenhagen university, Copenhagen.