The Problems With "Immersion"

A lot of appeal resides within the word “immersion”. Game journalists use it to describe good, qualitative, elements in a game. Usually whenever the word is used, it is within contexts of positive quality. As far as journalism is concerned, the word is closely tied with narrative and graphical criticisms about a game. If the journalist is critiquing the product, it wouldn’t be short of ordinary to find “...this game’s immersive cut-scenes...” or ...“the way in which characters talk and interact with you helps the immersion...”. Both examples have dense clouds of ambiguity surrounding them, and the nature of the word remains somewhat vague within the context of the literature. Likewise, the journalist might be publicising a post-mortem on behalf of a developer or leisurely researching aspects of the game where one may find such sentences like “...immersion is a difficult aspect of design climax to attain these days in the industry...” or “...the sound was designed so it would immerse the player, but executed in such a way that it breaks the immersion...”; again, the word is used in positive context but unlike the previous two examples, may infer completely different things.
 
Other instances where the word is used include the video-gaming community. Gamers would use the word much in the same way journalists would; to describe positive aspects of design, and attempt to communicate a “golden climax” of this design. Usually the term would imply the player’s complete disembodiment of the physical world surrounding them and an “immersion” into the virtual space that the game presents. However because of the many different models surrounding the word put forth by academics, it is hard to exactly pinpoint what the word clearly defines or attempts to describe. It is clear, however, that the word has positive implications both inside and outside the fields of academia, and that players constantly use it in attempts to communicate desirable achievements through the marriage of development and production.
 
Before refining the word’s definitions into a more understandable and approachable concept, we must first address its problems. These problems reside in the obvious ambiguity surrounding the term and the plethora of different models put forth by researchers and academics around the world. Unlike Csikszentmihalyi’s Flow model (1990), “immersion” has no solid platform of research on which to build off. Csikszentmihalyi created a model to describe seamless play in which the duality of challenge and skill resulted in the revelation of “flow” - a state of play that allows the user to be involved within the activity without the non-diegetic and negative experiences of frustration (too much challenge for the player) or boredom (too much skill of the player). This model is used to this day and referenced heavily in psychology and video game theory.
 
With the evident need for the construction of a solid “immersion” model and the obvious ambiguity around the term, we should first look at the current models and how they differ and resonate with and from each other.
 
Firstly we should look at Laurie N. Taylor’s (2002) approach to immersion. Taylor claims there are two levels or “stages” when we experience immersion: first, the “diegetic” stage of immersion; an experience characterised not by the total embodiment of the text / media, but the dissipation of the exterior space (real world). This first stage of immersion relies heavily on the player being distracted by the game. It is less reliant on the player becoming totally absorbed by the narrative and other immersive elements in the game and rather uses the generic nature of the real world to deter the player from that dimension. Essentially, it manipulates the player’s inattention to the zero-frame spatial field (a psychology-term used to describe where the player / user currently perceives him or her to exist), not because the game world is more “real”. The second “stage” argued by Taylor is the “intra-diegetic” stage of immersion which implies that the “player is not acting upon the game, but within the game”. This concept differs from the first stage in that the player is unaware of their diegetic actions upon the gaming interface as opposed to the first stage’s non-diegetic interaction and ignorance to the real world. Furthermore it is argued that in order for intra-diegetic immersion to occur, the first stage of immersion must first be reached and that the gameworld must be consistent; inconsistency breaks immersion and is therefore a key concept to understand.
 
Interestingly, Taylor uses Diablo (1996) and Diablo II (2000) as chief case studies for inconsistency and how they break immersion. While I agree that inconsistencies do break “immersion” (a term we still have yet to define), I disagree that good examples of diegetic inconsistencies can be found within the Diablo series. For example, Taylor argues that “the level randomisation and regenerating enemies are inconsistent with the overall game structure when one plays the game as single-player”. While Taylor fails to define what “overall game structure” she is exactly referring to, we can clearly make out that she is referring to the procedurally generated levels in the game, and their apparent diegetically nonsensical purpose. While this may typically be a good argument to hold against a game that “[randomises levels] when a player leaves and returns to an area”, Diablo II remembers the procedurally generated layout of the level when you exit and enter the gamespace so it doesn’t have to generate the world more than once, and so it makes sense diegetically.

Conversely, McMahan (2003) admits to the ambiguity of the term in that “immersion has become an excessively vague, all-inclusive concept... [in need of] specific terminology”. Similar to Taylor’s model, McMahan describes two distinct levels of immersion: diegetic and non-diegetic and her main argument rests in that immersion is not purely dependent on technology or game mechanics, but rather more dependent on user perceptions and how the player perceives the game world and experience. The ineffability of “immersion” for the individual emerges from the apparent realisation that the concept is largely experiential. There are three conditions that create a sense of immersion / presence according to McMahan: The user’s expectation must be met (no graphic walls / diegetic limitations / inconsistencies); the user’s actions must have a non-trivial impact upon the virtual space / gameworld; and the world conventions must be consistent with regards to player and non-player actions and cause / effect scripts.

McMahan then presents another more refined model: perceptual and psychological immersion. “The first is accomplished by blocking as many senses as possible to the outside world... the second results from the user’s mental absorption of the world.” Not only is perceptual and psychological immersion proven to be effective in the realm of video games, it is proven that “immersive virtual reality environments have been shown to be effective in the treatment of fear of heights, fear of flying, arachnophobia, claustrophobia, and agoraphobia.” This realisation proves the effectiveness of perceptual and psychological variables in helping to create “immersion”. While McMahan tends to explore the causes / effects of immersion more-so than Taylor, her studies do hold resemblance with Taylor’s findings in that there are two distinct levels or stages when defining “immersion”. It appears Taylor has more of an urge to propose a model while McMahan wishes to explore the variables in attaining the experience.

Like Taylor and McMahan, Ermi & Mayra (2005) argue there are two elements of immersion (or, as they put it, “dimensions”): participation, and connection. “The dimension of participation varies from active to passive... and the dimension of connection varies from absorption to immersion.” These “sub-domains” of the two dimensions initially complicate the concept, but once we reveal the context of these “four realms of experience” from the two initial dimensions, Ermi & Mayra’s model becomes somewhat clearer. First we have the experience of “entertainment (absorption and passive participation), educational (absorption and active participation), aesthetic (passive participation and immersion), and escapist (immersiona nd active participation)”. While Ermi & Mayra do touch on the vagaries of the term in certain communities (“...players, designers, and researchers use it as well, but often in an unspecified and vague way without clearly stating to what kind of experiences or phenomena it actually refers to”), I can’t help but disagree with the very strict and concise categorisation of experiences in their “two dimension / four experience” model. While they urge that “absorption and active participation” leads to an “educational” experience, surely it could also lead to a different realm of immersion such as “entertainment”. For instance, a player reacting to the sudden change of colour in a level due to the game’s rule-sets (whatever they may be) could be an absorbed, learned, and actively participated reaction. The game in question does not have to be educational. Osmos (2009) might be dependent on the player learning from the environmental shifts and changes and reacting according to those variables in an actively participated way - but this is not to say Osmos is an educational experience. There are severe faults in the experiential model put forth by Ermi & Mayra that dampen the categorisation of more complex game systems that we see today. The mind could well be processing educational activities while playing non-educational games, but the player could be thinking “I am not learning. This is not educational. This is entertainment.”

Then there is the argument that games originate from the need and desire to learn - the human race plays games to learn about our environment and we are learning about that environment constantly, every day. These arguments among others deflate the legitimacy of an “experiential” model because players above all else govern those experiences. One player might feel an active-participation in Tetris, while another might simply feel it on a more passive level. The main problem with the experiential model is that it is dependent on player variables.