Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Video Games as Unnatural Narratives

Astrid Ensslin, Bangor University, UK

Introduction
The aim of this paper is to approach video games from a territory within postclassical
narratology hitherto largely untouched by ludologists: the study of unnatural
narratives is one of the most recent fields within contemporary narrative theory and
deeply intertwined with the endeavour to understand how human beings make sense
of narrative texts and artefacts across media that cannot be sufficiently accounted for
or analysed using traditional narratological theories (Alber et al. 2010: 113). So this
article seeks to open up a new field of ludonarratological enquiry targeted in
particular at games that push ludonarratological boundaries and call out for metaludic
debate and reflection, albeit not necessarily with an avant-garde agenda in mind. In
fact, the games that I’m going to look at in the analytical part of this study are
commercially traded and played by global audiences, possibly because rather than
despite the fact that they prima facie go against the grain and call out for metaludic
debate, specialised analytical tools and conceptual frameworks.

The idea for this project grew out of my previous monograph project on Literary
Gaming (Ensslin 2014), which is situated at the junction between indie games,
electronic literature and ludostylistic analysis. For this book I deliberately chose the
title literary ‘gam-ing’ rather than ‘games’ because literary games proper (in the sense
of games that embed literary structures and encourage literary reading interwoven
with gameplay) are only one subform of what I see as a continuum between ludic
digital literature (where literary reading is foregrounded and games and play are
integrated in digital literary structures) and literary games as I’ve just described them.
So in a nutshell the study of literary gaming looks at hybrid digital media that
combine different types of gameplay and literary reading and which cause clashes and
creative interplay between what Hayles (2007) calls hyperattention and deep attention
in reader-players. Ludic forms of digital literature correlate mostly with the deep

1
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

attentive side of the spectrum, while literary games are experienced in a mostly
hyperattentive state.

The term literary gaming spans a wide range of ludo-literary media including poetry
games (e.g. by Jason Nelson, Jim Andrews and Gregory Weir), literary/narrative
auteur games1 (e.g. by the likes of Mike Bithell, Jonathan Blow, Tale of Tales, and
Davey Wreden/Galactic Café), Interactive Fiction (e.g. works by Nick Montfort,
Emily Short and Aaron Reed), ludic and metaludic types of hypertext and hypermedia
(e.g. by geniwate, Deena Larsen, Robert Kendall and Richard Holeton), as well as
more linear ludoliterary digital narratives produced in Flash, Shockwave and other
interactive animation technologies (e.g. works by Serge Bouchardon, Kate Pullinger
and Christine Wilks), as well as navigable 3D literary environments (e.g. by Andy
Campbell and Judi Alston).

This paper forms a first step towards my new book project (co-authored with digital
stylistician Alice Bell) on Unnatural Narratives and Digital Fiction (Ensslin and Bell,
forthcoming), which applies theoretical and analytical concepts of unnatural
narratology to various types of digital fictions (including hypertext fiction, Flash
fiction, Interactive fiction and narrative games). I’m going to discuss some ways of
understanding unnaturalness in games, using two very different definitions of the
unnatural in comparison, and explore the extent to which they may be useful to close
analysis. I will begin by providing some theoretical background on videogames’
narrativity, on unnatural narratives and unnatural narratology. I’ll then move on to a
dual argument based on two divergent definitions of unnaturalness: given a broad
conceptual framework, I propose that – in many ways - videogames are unnatural
narratives par excellence; therefore the term can be seen as somewhat tautological
when it comes to ludic narrativity.

Taking a more narrow, aesthetically oriented definition as starting point, I contend in

a second move that some games are more ‘unnatural’ than others, and that the
idiosyncratic ludonarrative mechanics exhibited by them allow us to apply, adapt and
further develop existing concepts and tools developed by unnatural narratologists. In

1
For an explanation of this term, see Ensslin (2014).

2
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

my analytical part I’ll then have a look at three games in particular that showcase
some key aspects of unnatural narratology at work: Tale of Tales’ (2009) The Path
and its uses of unnatural spatiality, Jonathan Blow’s (2009) Braid and its uses of
unnatural temporality and, finally, Galactic Café’s (2013) The Stanley Parable and its
uses of unnatural narration – in particular the role of the would-be omniscient narrator
and his conflict with the player-character. In my closing remarks I’ll sum up some of
my initial conclusions about the relevance and feasibility of unnatural narratology for
the study of games, and I’ll make some suggestions as to how we may develop this
analytical approach further to accommodate the media-specificity of digital games
and gaming.

Ludonarratological assumptions
The ludo-narratological approach taken here is set against the background of what are
by now widely agreed assumptions about the narrativity of games. First, a game isn’t
a narrative in the sense of a prescripted sequence of events, or indeed “any semiotic
object produced with the intent of evoking a [pre-intended] narrative script in the
mind of the audience” (Ryan 2004a: 9). Instead, they possess narrativity, as Marie-
Laure Ryan puts it in Narrative Across Media (ibid: 9), which means that they have
the potential to evoke multiple, individualised narrative scripts through settings,
characters and other elements that players interact with through choice and with the
intention to solve problems and make progress. Thus, in gameplay users are turned
into characters, and as players we enact the destiny, or the trajectory, of the
gameworld autotelically (ibid: 349), i.e. through our own motivated actions rather
than being told about or shown events as we are in fiction, drama or film.

In a more detailed analysis, Henry Jenkins (2004) breaks the narrative properties of
games into three core concepts:

 environmental storytelling: games are designed as environments, as worlds

full of characters and props for players to interact with (much like Disney
World and other amusement parks). Players explore games spatially, in an
episodic way, and this nonlinear model is kept coherent by an overarching
goal and repetitive mechanic. Games also form part of a larger storytelling
ecology, which brings to mind Jenkins' (2006) theory of transmedia
storytelling, which assumes that stories develop and evolve across media,
3
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

rather than simply being remediated or adapted). Finally, games are evocative
spaces with large mnemonic potential in that they evoke the structures of
existing stories and the genre traditions of other media (e.g. Red Dead
Redemption vis-a-vis the Western genre, the Lego series, Star Wars or Indiana
Jones);
 emergent narratives refers to the ways in which players create their own
stories by exploring the gameworld (corresponding roughly to Ryan’s [2004a]
autotelic enactment). These stories become manifest in oral storytelling or
participatory media, such as gamer fora or on YouTube, where gamers post
their own playthroughs, walkthroughs, Let’s Plays etc.;
 embedded narratives, which are any non-interactive narrative sequences
integrated into or surrounding gameplay, such as cutscenes, backstory
descriptions or dialogues (written or voiced-over). They tend to be embedded
in such a way as not to impede the interactive flow of gameplay, and they may
function as rewards or level-up markers; they may help drive the story forward
or bridge loading time. Needless to say, their usefulness and aesthetic potential
are controversial topics amongst gamers, and there are significant cultural
differences with respect to accepted duration and player patience (Ensslin
2011: 166).

Arguably there’s a lot more to say about videogame narrativity more generally, but
due to spatial constraints I’ll now move straight on to the core theoretical interest of
this paper: the varying concepts of ‘unnatural’ narrativity and how they may or may
not contribute to understanding video games as ludo-narrative media.

Unnatural narratives – unnatural narratology

Before embarking on an examination of what unnatural narratives are, it needs to be
acknowledged that the term is highly evocative of numerous problematic meanings
and uses: it carries ideology-ridden connotations of hegemonic ‘normality’, of
discursively constructed social and cultural hierarchies and oppositions and binary
thought more generally. Furthermore, the School of ‘Unnatural Narratologists’ (much
like that of ‘Natural Narratologists’) is deeply rooted in Western, Anglo-American
scholarship, and the vast majority of texts studied under this theoretical umbrella were

4
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

authored by Anglo-American writers, which leaves a large part of global narrativity

unaccounted for. Finally, popular notions of ‘unnaturalness’ are negatively connoted,
as it is often used to “denounce certain types of behavior (as well as sexual
orientations or practices) which the speaker considers to be deviant or perverse”
(Alber and Heinze 2011: 2), and this of course adds to the controversy surrounding
terminological choices underlying this theoretical apparatus.

Having said all that, the term ‘unnatural’ as used by narratologists carries a highly
specialised set of meanings, and can only be comprehended in the context of its
derivation. It was borrowed from Monika Fludernik’s idea of a ‘natural narratology’
(1996), which is anchored in a cognitive approach to human experientiality and the
ways in which narratives and narrativity can be re-evaluated from the point of view of
‘natural’, or “naturally occurring” (1996: 13) storytelling in the Labovian sense. So
the derivate ‘unnatural’ and its surrounding theories form a response to Fludernik’s
concept, and – despite its somewhat misleading negative prefix - the term needs to be
understood in a distinctly positive, productive sense for purposes of cognitive
narratological analysis: “[t]he aim of an unnatural theoretical approach is to
approximate and conceptualize Otherness, rather than to stigmatize or reify it; such an
approach is interested in various kinds of narrative strangeness and in particular in
texts that deviate from the mimetic norms of most narratological models” (Alber and
Heinze 2011: 2).2

One of the most frequently quoted definitions of unnatural narrative is by Jan Alber
(2013, n.p.), who describes them as a “subset of fictional narratives” that “[v]iolate[]
physical laws, logical principles, or standard anthropomorphic limitations of
knowledge by representing storytelling scenarios, narrators, characters, temporalities,
or spaces that could not exist in the actual world.” In other words, unnaturalness is
defined ex negativo in opposition to the ‘natural’ (see above), which relates to the
cognitive frames and scripts we have derived from our actual experience of being in
the world. So according to Alber’s fairly broad and inclusive notion, unnaturalness
refers to both physically and logically impossible narrative structures, which includes

2
Of course, the boundary between ‘natural’ and ‘unnatural’ isn’t clear-cut; unnaturalness has to be
understood as a matter of degree rather than an absolute quality.

5
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

the supernatural in fairy tales as much as it does, for example, multiple contradictory
endings of a story, or two parallel timelines that unfold at different speeds.

Another, more narrowly defined and aesthetically oriented concept of unnaturalness is

put forward by Brian Richardson. To him, unnatural narratives “conspicuously
violate[] conventions of standard narrative forms, in particular the conventions of
nonfictional narratives, oral or written, and fictional modes like realism that model
themselves on nonfictional narratives. Unnatural narratives furthermore follow fluid,
changing conventions and create new narratological patterns in each work. In a
phrase, unnatural narratives produce a defamiliarization of the basic elements of
narrative” (2011: 34; cf. Shklovsy 1917/65). Again, he defines unnaturalness ex
negativo as narrative structures that are anti-mimetic, which means they are “clearly
and strikingly impossible in the real world” (Alber et al. 2013: 102) and defy the
principles of (a) mimetic, realistic fictional storytelling and (b) the conventions of
nonfictional narratives, oral and written” for purposes of aesthetic innovation, critical
pleasure and meta-level reflection (Richardson 2011: 34).

Hence, whereas Alber puts both physically and logically scenarios in one ‘unnatural’
basket, Richardson makes a crucial distinction between so-called non-mimetic (or
physically impossible, fantastic) and anti-mimetic narrative structures, which defy the
principles of reality and realistic storytelling, but also the conventions of existing
media genres we tend to be familiar with, not in a deliberately negative, or alienating
way, but rather in a creative-productive manner that engenders various types of
reflective thinking in audiences.3 Hence, Richardson’s concept is geared more
towards the audience’s (projected or likely rather than empirically tested) response
than a textual quality. What matters to him is “the degree of unexpectedness that the
text produces, whether surprise, shock, or the wry smile that acknowledges that a
different, playful kind of representation is at work” (2015: 5).

The study of unnatural narratives, called unnatural narratology, is a subdomain of

postclassical narratology (Herman 1999), which represents a departure from classical

3
According to Alber (personal correspondence, 30/05/2015), Richardson’s “distinction between non-
and anti-mimetic elements is identical with [his] distinction between conventionalized and not yet
conventionalized instances of the unnatural”.

6
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

structuralist narratology (associated e.g. with Genette, Chatman, Bal and Prince) in
that it is both transmedia and transdisciplinary: (1) it broadens the scope of analytical
objects from print-based, literary fiction to narrative media more widely, as well as
nonfictional forms of storytelling (in the Labovian tradition of oral storytelling); and
(2) it expands the narratologist’s analytical and conceptual toolkit by integrating non-
literary disciplines such as post-Saussurean linguistics (e.g. discourse analysis,
possible-worlds semantics etc.), gender theory, ethnography, cognitive science
(schema and frames and scripts theory), film and media studies.

As far as its cross-media remit is concerned, unnatural narratology has reached out to
drama, film, comics, nonfictional testimonies and hypertext fiction, but very little
work exists (to my knowledge) that deals with the idiosyncratic narrativity of
videogames. So this is where my theoretical and analytical contribution lies with this
project.

It’s not surprising that many, if not most, unnatural narratologists have looked at
postmodernist narratives (novels, short stories, films) when developing their theories,
so if I’m proposing in this paper that in many ways mainstream videogames are
unnatural narratives par excellence, I’m doing something quite unconventional, or
theoretically unnatural in its own right, because I’m arguing that ‘unnatural’ is
actually quite ‘natural’ (or rather conventionalised) when it comes to videogames
(according to Alber’s definition), and that the body of games that we can
meaningfully refer to and analyse as ‘unnatural’ (using Richardson’s definition) is
still fairly small (but growing, and an exciting development to follow within the indie
sector in particular).

When studying unnatural narratives, a core, classical narratological distinction is

usually made between unnaturalness at story level (which concerns the actual
underlying fabula, or that which is told), and at discourse level (which is the level of
the telling, i.e. of narrative organisation, sequentialisation or design). So for example
unnatural temporality at the story level happens in time travel narratives (where the
protagonist may criss-cross between different historical periods consecutively),
whereas at discourse level the story may remain unaffected by discourse-level
fragmentation, mixing or reversal, such as in Nolan’s Memento (Heinze 2013). For
7
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

games, this distinction has to be substituted for by a concept that allows for the
executability of the underlying code, and Jenkins’s (2004) idea of emergent
narrativity, where the player’s interaction with the coded interface produces as many
stories as there are players and play-throughs.

Reading strategies and conventionalisation

We’ve established earlier that unnaturalness can be understood in terms of
unconventional and defamiliarising structures and experiences. Surely, however, what
we’ve come to accept as ‘conventional’ hasn’t always been that: basic cognitive
frames develop over time and the more often we’re exposed to specific ‘impossible’
scenarios, the more readily we’ll integrate them into our repository of ‘the possible’ –
so we naturalise (Culler 1975) initially unfamiliar, or defamiliarising structures, by
embedding them into our cognitive frames of reference. And these mental repositories
of ours tend to be genre-bound. We’ve become used, for example, to speaking
animals from fables, fairy tales and other types of fantasy (which are non-mimetic but
not anti-mimetic according to Richardson), we’ve conventionalised the omnimentality
of the omniscient narrator (which is a humanly impossible quality – no-one can know
everything, least of all what other people are thinking exactly, but we’ve become used
to it especially from classical realist novels), and we’re perfectly well accustomed to
time travel narratives (as they often occur in sci-fi) and physically impossible
geographies such as flying islands (e.g. Laputa in Gulliver’s Travels, and fantasy
fiction more generally).

So why and how does conventionalisation happen? Essentially, it’s within our human
nature that, when we encounter anything unfamiliar, or strange, as we do in unnatural
narratives, we try to make sense of it in some way, by applying a range of reading
strategies. As Alber puts it, we are “ultimately bound by [our] cognitive architecture,
even when trying to make sense of the unnatural. Hence, the only way to respond to
narratives of all sorts (including unnatural ones) is through cognitive frames and
scripts” (2013b: 451-54), so on the basis of cognitive theory, Alber proposes the
following reading strategies employed by readers (in any combination and any order;
cf Alber 2013c: 49) to help them “come to terms with the unnatural” (2013b: 451):

8
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

1) Frame blending: here we blend pre-existing frames that we previously

considered to be incompatible (e.g. that the flow of time may be tied to the
direction in which you move, which is the case in World 4, ‘Time and Place’, of
Braid)
2) Generification: evoking genre conventions from literary and media history; so
here the blending has already happened and we’ve integrated it as a possible
convention in a given genre or medium (e.g. time travel in sci-fi narratives; or
superhuman jump heights in platformers)
3) Subjectification: here we attribute the unnatural to internal states, such as dreams,
nightmares, or hallucinations. We know it’s perfectly natural for our unconscious
mind to produce highly surreal scenarios, so this option is part of our explanatory
repository, especially when we’re dealing with an unreliable narrator or a
vulnerable, victimised protagonist (such as the six sisters in The Path, who all
have to meet ‘their’ wolf in the form of an age-specific traumatic experience).
4) Thematic foregrounding: here we identify specific thematic elements in a
narrative that recur, in various configurations, to form an idee fixe (e.g. the
relationship between time and human experience, in Braid, or the metaludic
conflicts in The Stanley Parable [TSP])
5) Allegorical reading: here we understand unnatural structures as part of an
extended metaphor about the human condition, or the world in general (so the
impossibility of meaningful play in TSP can be seen as an allegory of illusory
agency [McCallum-Stewart and Parsler 2007] in gameplay more generally)
6) Satirization and parody: this occurs when narratives try to mock either other
narratives or elements of the world in general; the zero-player game Progress
Quest for example is unnatural in that it doesn’t allow players to do anything other
than watch the game ‘play itself’, thereby parodying MMOGs such as Everquest
(and particularly their auto-attack function, which is extremely sit-back).
7) Positing a transcendental realm: here we attribute the unnatural to some kind of
supernatural setting, e.g. heaven or hell (think of the god game, Black and White,
where players have godlike powers over characters and are infiltrated by the
impersonated voices of good (an angelic character) and evil (a demonic guide)
8) Invitation to ‘free play’, where mutually contradictory storylines, or endings, are
seen as an invitation to create one’s own story, which is a common feature of
hypertext fiction, but we may well ask other players what their preferred ending of
9
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

TSP is, since we’re dealing with a game that thematises non-closure,
multilinearity, logical contradiction and cyclicality.
And yet we may just as well adopt an unnaturalizing reading strategy (Nielsen
2013), in an attempt to accept the impossible as it is without trying to make sense of
it. This approach goes against “domesticating the unnatural” (Alber 2013a, n.d), and
can be described in terms of a “Zen way of reading” (Alber 2013c: 83-84): so here we
adopt the stoic position of simply leaving things unexplained and accepting the
feeling of confusion, frustration, or discomfort that the narrative experience may
evoke in us.

Video games as unnatural narratives

Having talked a lot about theory, let’s now turn to videogames as unnatural narratives.
I was wondering initially whether I should put a question mark behind the title of this
paper, but I think it is fair to claim that in many ways games are unnatural narratives
par excellence. Having said that, this proposition only holds true if we adopt a broad
concept of unnaturalness as, for example, put forward by Alber (2013a), who defines
it as physically or logically impossible when measured against the foil of our real-
world cognitive frames. In actual fact, we may go as far as to say that under this
definition the unnaturalness of games is what makes them so attractive to vast
amounts of people around the world because it enables us to escape into realms of
what’s normally thought to be humanly impossible or unthinkable. Furthermore, we
have to bear in mind that games are perhaps the most readily ‘naturalising’ media of
all because they integrate in their procedural mechanics the very structures (unnatural
or not) and ‘ecological’ interactions (Linderoth 2011) players are meant to internalise
as effectively as possible for fast progress and high levels of satisfaction.

Clearly, mainstream videogames are full of physical impossibilities; to list just a few
examples:
 respawning and rebirth are crucial replayability factors (violating the truth
condition of singular mortality);
 games thrive on using fantasy traditions from other media such as talking
animals, monsters and other forms of non-human yet anthropomorphised
creatures;

10
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

 human or not, the anatomic dimensions of some hypersexualised characters

would be anatomically impossible in the real world (think of the early Lara
Croft’s athletic abilities vis-a-vis her hyperfeminised physique);
 warping, or teleporting, between geographic areas is a standard form of fast in-
game movement (violating the limitations of physical movement)
 similarly, there are highly dexterous types of movement in some games that
are more akin to those of animals than human beings (think of the wall runs in
Prince of Persia, the jumping art of SuperMario and other platform characters,
the superhumanly fast-paced balancing act of Mirror’s Edge and quite
generally the fact that falling or jumping off high edges often doesn’t result in
character death or even the slightest degree of harm).
 Multiple impersonations of one and the same player are a key attraction of
RPGs and MMOGs: either synchronously (with more than one avatar of the
same player in a game world simultaneously) or asynchronously (anatomically
shape-shifting avatars through customisation)
But games also exhibit a range of conventionalised logical impossibilities, e.g.:
 The fact that avatars are ‘us’ in the game world makes them the interactional
metaleptic tool par excellence (Ryan 2004b), yet metalepsis in the sense of
transgressing ontological boundaries, and especially those of fictional and
actual world is both physically and logically impossible (because we can’t all
of a sudden lose our anatomic materiality; nor can we be in two places at the
same time, especially if they’re in different time zones)
 The success of a lot of games is based on the fact that they offer multiple and
either contradictory or incompatible endings. Dragon Age: Origins for
example converges (despite seemingly countless choices throughout) down to
four endings (which is a comparably low number, incidentally), which see
different characters ruling the game world, and some characters either dead or
alive (which is a logical incompatibility).

So does all this mean that we should stop here and simply conclude that unnatural
narratology doesn’t work with games because ‘video games as unnatural narratives’ is
a tautology? To me this would be slightly myopic because, clearly, when we look at
Richardson’s anti-mimetically oriented conceptualisation and move beyond the

11
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

conventionalised ‘unnatural’ media-specificity of games, there’s actually quite a lot

we can do with a specific type of games: games that seek to defamiliarise and
innovate the gaming experience through highly idiosyncratic ludonarrative
mechanics.

So in a second move I would argue that some games are more ‘unnatural’ (in a
Richardsonian sense of aesthetically ‘more estranging’) than others because they
deliberately violate the ludonarrative conventions of their genre and the medium itself
in order to evoke metaludic and metafictional reflections in the player – as well as
other types of philosophical and critical processes. With this premise in mind, I shall
now move on to the analytical part of this lecture and demonstrate three aspects of
unnaturalness at play: unnatural spatiality in The Path, unnatural temporality in Braid
and unnatural narration in The Stanley Parable.

Unnatural spatiality in The Path

According to Alber (2013, n.p.), “impossible spaces undo our assumptions about
space and spatial organization in the real world”. So typical types of unnatural spaces
include:
 containers that are bigger on the outside than on the inside, or vice versa
 shape-shifting settings
 non-actualisable geographies
 visions of the infinite and unimaginable universe
 metaleptic jumps between different ontological spheres.
What interests me most here is how impossible in the sense of anti-mimetic spatial
design contributes to reflexivity, and one game where this can be shown quite nicely
is Tale of Tales’s The Path (2009).

The Path is a short horror game in which the adolescent female characters we can
choose to play are exposed to different types of trauma – tailored to their age. The
gameworld and our interaction with it is designed in such a way as to evoke notions of
horror and premonition of what may happen, and the contemplative effects are partly
created through slow movement through the game world (it’s impossible to run for
more than a few seconds, and the forest seems endless – augmented by a wraparound

12
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

structure which causes us to move in circles). Whenever a girl has met her wolf (and
experienced spiritual death as a result), she ends up lying in front of grandmother’s
house, which on the outside looks fairly small. As she enters the house, however, it
obtains gigantic dimensions and the semi-cut-scene after her ‘fall’ takes us through
seemingly endless corridors with countless doors and huge rooms displaying and
staging objects evocative of her nightmarish experience (fig.1). So the logically
impossible spatial dimensions of the house (incompatibility between outside and
inside) can be read in terms of Alber’s subjectification strategy (the attribution of
unnaturalness to internal states, or trauma in this case).4

Fig.1: Grandmother’s house in The Path as a logically impossible space

Another interesting element of unnatural spatiality is the treatment of paratext vis-à-

vis gameworld: the girls’ journals, retrievable from the game’s official website, read
very much like they’ve been written by the fictional characters. However, numerous
real-world comments have been posted by players of the game, which are interspersed
with pre-scripted posts by fictional characters, and this creates an ontological
hesitation between players’ actual world and the fictional world of the game. In fact,
this occurrence of interactional metalepsis (or transgression between logically distinct
ontological spheres) adds to the eerie but also the philosophical/reflective atmosphere

4
Another way of understanding this spatial incompatibility would be in terms of positing a
transcendental realm (Alber 2013c) – that of some kind of highly unsettling afterlife.

13
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

of the game, and we accept the unnaturalness of this design feature because we can
read the game as an allegory (one of Alber’s [2013b] strategies) of adolescent female
trials and tribulations, and this element converges our actual world with the game
world.

Unnatural temporality in Braid

My second analysis focuses on unnatural temporality. In an essay on temporal
paradoxes in narrative Ryan (2009: 142) proposes four core intuitive human beliefs
about time:
1. Time flows, and it does so in a fixed direction.
2. You cannot fight this flow and go back in time.
3. Causes always precede their effects.
4. The past is written once for all.
In fiction of course at least some of these dictums are regularly subverted, e.g. in time
travel narratives or postmodern, multilinear narratives such as Groundhog Day or
Run, Lola, Run. Hence, some elements of (fictional) unnatural temporality have
already been conventionalised – depending on individual levels of exposure and
culture-specific media ecologies.

According to Alber (and his broad definition of unnaturalness), “unnatural

temporalities [which revolve around Ryan’s principles] challenge our real-world ideas
about time and temporal progression” (2013a: n.p.). So for this study we need to add
the assumption of antimimetic defamiliarisation as part of the developer’s intent.

Typical unnatural temporalities (Richardson 2002) include:

 retrogressive/antinomic temporality, where the scripts of everyday life are
reversed (e.g. in Martin Amis’s 1991 novel Time’s Arrow);
 eternal temporal loops / circular temporalities: where the narrative, or a
character, seems to be going round in circles (a common feature of hypertext
fiction)
 conflated time lines (a.k.a. ‘chronomontages’; conjoining different temporal
zones – an example would be the time traveller landing in the historical past

14
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

(e.g. Kevin from Time Bandits (Gilliam 1981), catapulted into Ancient Greece,
in his own contemporary clothes, taking photos with his Polaroid camera);
 reversed causalities (where the present is caused by future like for example in
D.M. Thomas’s The White Hotel (1981), where the protagonist’s pain is
caused by an anticipatory projection of a future event);
 contradictory temporalities (mutually exclusive events or sequences, e.g. in
Coover’s short story ‘The Babysitter’, where Mr Tucker both did and did not
go home to have sex with the babysitter);
 differential time lines (such as different aging speeds between characters, e.g.
in Virginia Woolf’s Orlando)
 multiple time lines (plotlines that begin and end at the same time yet take
different periods of time to unfold, e.g. in Shakespeare’s A Summer Night’s
Dream).

A game that embeds a number of unnatural temporalities quite firmly and

unavoidably in its mechanics is Jonathan Blow’s Braid (2009). It’s a 2D platform
game where the protagonist Tim is on a quest to save the princess, beg her forgiveness
and live happily ever after (although we later learn that she doesn’t actually want
him). The game as a whole can be read as an allegory of the Trinity nuclear bomb test
of 1945, which directly preceded the destruction of Hiroshima and Nagasaki.
Hence, against this backdrop, the impossible temporal mechanics of the game can
meaningfully be read as an allegory (see Alber’s reading strategies) of the
irreversibility of human action and suffering, while the mechanics seem to suggest
that the rules of temporal logic can be lifted by Faustian ambitions.

Each world in Braid has its own impossible temporal mechanics, which have to be
internalised by players (against their real-world and genre-specific assumptions) in
order to be successful. I’ll only mention 3 examples here:
 Retrogressive temporality: You can go back in time in exact reversed order
by holding the shift key (and speed up the time as needed); hence Tim can
‘un-die’(rather than respawn) indefinitely, and indeed certain achievements
only become possible through this rewind function.

15
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

 Reversed causality (which goes against the principle that causes always
precede their effects): in World 3, ‘Time and Mystery’, the player’s actions
can be rewound whilst other elements in the game world remain unaffected by
reversal. For example, through selective rewind a key can be brought back to
the immediate past (without having to drop the key again) to open a door that
otherwise wouldn’t be possible to open because Tim would be stuck forever in
a pit that he’s jumped into to grab the key.
 Differential timelines: in World 6, ‘Hesitance’, the player can slow down
time to get certain things done – time moves slower in the proximity of the all-
important ring, marked with a halo, enabling Tim to for example escape
certain monsters while either he or they are in the time warp, or to manipulate
the velocity of moving objects, such as clouds, to facilitate forward movement.

Unnatural narration in The Stanley Parable

Finally, I’d like to examine an instance of unnatural narration. Unnatural narration can
again be quite simply any physically impossible narrator (according to Alber), such as
an animal, a baby, a human bodily organ, a plant or object. Or it can, and in my mind
more powerfully so, manifest itself in unconventionalised forms of extreme narration,
such as second person narration, multiperson narration, certain forms of unreliable
narration and denarration (a narrator’s negation of previously stated or assumed
truths; see Richardson [2006] for a comprehensive overview). Strangely or not,
omniscient, or authorial narration is also generally held to be unnatural, largely
because no-one can possibly know as much as a standard omniscient narrator tends to
do, and the fascination associated with this paradox is reflected in recent fictional
creativity beyond print, as my example will show.

What’s important to note here is that unnatural narration in games is in itself

tautological. A narrator, or narrative voice telling the story, is impossible in
videogames because it would subvert or hinder the player’s decision making process
in the game world, as well as their individualised emergent experiences.

16
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

The game I’m going to look at, The Stanley Parable by Davey Wreden/Galactic Café
(the remake version of 2013), experiments with this paradox by employing an
intriguing type of unnatural narration – a shape-shifting, intrusive narrator whose
would-be omniscience is deconstructed by the player’s subversive behaviour. The
game stages a combat between player and narrator (who, as it turns out, isn’t as
empowered and omniscient as he pretends to be, and ultimately is at the mercy of the
player and, of course, the essence of gameplay and its implications for narrative
design). By the same token, we as players are confronted with the limitations of our
own agency as even the choices we can make are prescripted.

At this juncture, let me say a few words about omniscient narration. It was the
standard form of realistic storytelling in 19th century realism, was then superseded by
reflector mode and internal narrative styles in literary modernism (e.g. James Joyce
and Virginia Wolf) and has more recently, in 21st century fiction, seen a revival, yet in
very different forms that reflect the impossibility of godlike or representative
knowledge or insight. Writers such as Zadie Smith, Salman Rushdie and Martin Amis
have been experimenting with more vulnerable, fragmented, confessional forms of
authorial narration that reflect a crisis of fiction writing yet at the same time the
ambition to create new forms of literary authority and thereby regain cultural capital
vis-à-vis popular culture (Dawson 2013). So if, in The Stanley Parable, we are
confronted with an experimental battle between would-be omniscient narrator and
player-character, we have to see this design as a reflection of two current trends: (a)
the media-ecological crisis and cultural ambitions of 21st century fiction writing, and
(b) the literary gaming movement that’s been evolving over the past 15 years or so.

According to its official site, TSP is “an exploration of story, games, and choice.
Except the story doesn’t matter, it might not even be a game, and if you ever actually
do have a choice, well let me know how you did it” (Mularcyzk, n.d.). So what the
game tries to get across is the question of how much agency and choice players
actually have in a game, and that agency is ultimately illusory (MacCallum-Stewart
and Parsler 2007) given that choices, paths and endings tend to be pre-coded. We also
learn through procedural rhetoric that subversiveness on the part of the player is a sine
qua non to escape from illusory agency, which makes cheating not simply a legitimate
but indeed a recommended form of player engagement.
17
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Fig. 2: screenshot from the initial cutscene of TSP, where the camera zooms into zero
(omniscient) focalization mode contrasting already explored with yet-to-explore
(greyed out, symbolically endless) territory.

The protagonist is Stanley, an office worker in a Kafkaesque corporate, bureaucratic

environment, who pushes buttons upon command, day-in, day-out (fig.2). Initially
Stanley, played in the first person, sets out in his quest to find out what’s happened to
his co-workers, who have all disappeared. The narrator accompanies him on his way
throughout, giving instructions as to where to go next yet not in a directive, command
form, but in the past tense indicative, thereby making propositions about what Stanley
did (before he gets a chance to do so) rather than suggestions about where he might
go. Whereas in a film or print narrative, past tense indicative narration is the accepted,
default standard of storytelling, in a game it completely goes against the grain and
seems patronising at best. This makes us as players suspicious about the reliability, or
trustworthiness of the narrator right from the outset, and the more we attempt to
deviate from the narrator’s propositions, the more standoffish and annoyed he
becomes, so much so that in some extremely deviant endings his comments,
behaviour and designs suggest frustration, despair, resignation, or even madness.

The narrator (or implied author) becomes our main enemy in the game because there
aren’t really any further major obstacles or enemies to overcome, and this battle
between implied author (manifested in the choices built into the game) and the player

18
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

who strives to undermine his/her ‘implied’ counterpart (personified by conformist

Stanley) is orchestrated in at least 18 different paths, or endings.

In what follows, I’m going to briefly look at three of them to demonstrate the
transformation of the narrator’s projected authority: (1) the Life ending, which
follows the path of maximum conformism/obedience; (2) the Choice, or Real Person
ending, where you unplug a phone that you were supposed to answer; and (3) the
Museum ending, which adds another ontological sphere, or diegetic level to the
gameworld, lifting the story experienced so far onto a symbolical, or allegorical level.

The main decision players have to make for Stanley is at a set of two doors, where the
‘correct’ path is left and the ‘deviant’ path is right (paradoxically or not), whereby
Stanley can still ‘go wrong’ after taking the ‘right’ door, and very drastically so, going
by the narrator’s perplexed reactions. If the player follows all the narrator’s
propositions precisely, in the Life Ending (1), they’ll be rewarded with a ‘win’.
Stanley switches off the controls in the Orwellian Mind Control Facility and steps out
into the open. However, this is not supposed to be the point, or target of this game,
which is indicated two-fold: by the ensuing cut-scene, which leaves the player
passively watching and listening to the narrator’s freedom monologue on Stanley’s
alleged ‘happiness’; and by the fact that, after the cut scene the player is sent straight
back to the beginning, i.e. to Stanley’s office, with the interspersed loading message
‘the end is never the end’, suggesting that they’re supposed to explore the different
paths of deviance, or defiance.

Fig.3 shows the diegetic levels of the gameworld as evoked by the Life Ending. The
big box contains the diegetic, or fictional space of the game, with the intradiegetic, or
character level embedded in the narrator’s diegesis. We as players are extradiegetic,
or outside the story, but since we implement the narrator’s story, which he tells to a
narratee on his level, by steering Stanley, we’re also inevitably part of the gameworld.
Therefore the membranes between the levels, or spheres, are shown as semi-
permeable, and we can see that there are a lot of metaleptic cross-overs happening
already, even though the narrator isn’t actually addressing us directly (which he does
in other endings).

19
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Fig.3: diegetic levels of TSP, ‘Life Ending’.

A striking example of the narrator’s reaction to player deviance is the Choice, or Real
Person Ending (2). Here the player opts to unplug a phone that Stanley was supposed
to answer, thereby undertaking an action not contained within the narrator’s script.
This act of transgression causes the narrator to sense that there’s someone else behind
Stanley’s incorrect behaviour, and after first addressing Stanley and realising Stanley
couldn’t possibly have devised such an act, he turns to the implied player thus:

Oh no, no, no! Did you just unplug the phone? That wasn’t supposed
to be a choice. How did you do that? You actually chose incorrectly,
but I didn’t even know that was possible. Let me double-check
[shuffling his papers around yet finding no evidence of this choice in
his script] … I don’t understand. How on earth are you making
meaningful choices? What did you…Wait a second …. How had I not
noticed it sooner? You’re not Stanley. You’re a real person. I can’t
believe I was so mistaken. This is why you’ve been able to make
correct and incorrect choices, and to think I’ve been letting you run
around in this game for so long! If you’d made any more wrong
choices you might have negated it entirely. It’s as though you’d
completely ignored even the most basic safety protocol for real-world
decision-making. I’m going to stop the game for a moment so we can
educate you properly on safe decision making.

20
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

This is followed by a satirical educational video about the life-threatening potential of

human decision-making. So in this instance the narrator breaks the fourth wall and
moves the metaleptic interaction onto an (implied) extradiegetic level. Another
observation that can be made here is that this particular ending shows how indirect
communication works between the developer and the player outside the fictional
world (extradiegetically; marked by the dotted arrow in fig.4), as we learn that the
narrator is not responsible for the unpluggable phone as a pre-programmed choice. So
here the actual author communicates implicitly with the actual player, who
implements the communicated option intradiegetically.

Fig.4: diegetic levels as suggested by the Choice Ending.

Interestingly, as the camera returns to the gameworld and the room with the
telephone, we find the space transformed into a postmodern pastiche (fig.5), and the
narrator says, “Ah welcome back! You may have noticed that this room has begun to
deteriorate as a result of narrative contradiction”…and “we just need to get you home
as soon as possible before the narrative contradiction gets any worse. Unfortunately it
seems this place is not well equipped to deal with reality” – so here the narrator
himself explicates the anti-mimetic, logical impossibilities embedded in the game’s
narrative design.

21
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Fig.5: postmodernised telephone room in the Choice Ending.

Finally, in the Museum Ending (3), Stanley meets his intradiegetic death after
choosing to take the escape route branching off the ‘correct’ track to the Mind Control
Facility. He’s killed by some kind of machine, yet again we find that the game goes
on…and the player (now playing their own alter ego) finds him/herself in a fictional
developer’s museum, which exhibits all sorts of in-game props and concept art.
Strikingly, here we encounter another narrator, a female voice, which seems to be
superordinate to the diegesis of the initial male narrator, who now seems to have
disappeared along with Stanley. The female narrator comments on the paradoxical
love-hate relationship between player-character and narrator and advises the player to
stop the game to put an end to the endless, meaningless cycle of “walking someone
else’s path”. And here, finally, is where the player’s alter ego in the game dies by
being crushed by the ‘metal jaws’ – and after this ending players have to physically
reload the game by starting again from the main menu.

22
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Fig.6: diegetic levels as suggested by the Museum Ending.

In fig. 6 we can see that another level of diegesis, call it metadiegesis, has been added
to the ontological universe of the game, and the female narrator speaks to us directly
as implied players. Interestingly, though, although the female super-narrator seems to
be more empowered than the male narrator in the other endings, she is equally subject
to the player’s choices (and of course the game design), and as she frantically tries to
prevent them from having their alter ego killed in the game world, and from endlessly
perpetuating the cycle of following pre-designed paths and subjecting themselves to
illusory agency.

To wrap up this analysis, The Stanley Parable’s procedural rhetoric reinforces the
decoding strategy suggested by its title (a parable is an educational allegory). As
players we are made to read it as an allegory of illusory agency built into games to
give players the illusion of choice, power and control. In fact, we as players are all
Stanley because again and again we willingly or even enthusiastically subject
ourselves to the constraints set by the games we play, except of course when we cheat
– and this is where we have the power to ‘defeat’ the implied author-programmer.

Concluding remarks
To sum up the main insights I’ve drawn from this research so far, if we want to move
toward an unnatural ludo-narratology, there are several things to be aware of and take
23
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

into account. First, not every definition of unnaturalness is useful for close game
analysis, but if we take anti-mimeticism and defamiliarisation – for the sake of
entertainment, flow, innovation and critical reflection – as a starting point, we can
begin to make sense of the kinds of ‘unnatural structures’ featuring for example in
metagames like TSP, or generally in games that push the boundaries of ludonarrative
design.

Second, the ‘naturalising’ and ‘unnaturalising’ reading strategies put forward by

Alber (2013b) and Nielsen (2013) are useful starting points, yet they have to be
augmented by game-specific ways of making sense of what Jesper Juul (2005: 132)
calls “incoherent worlds”, i.e. by explaining unnaturalness in terms of the rules of the
game. Furthermore, there’s still a lot of work to be done on studying players’
individual nuances in understanding unnatural ludonarrative structures and their
underlying and resulting play styles and strategies. Closer insights into these
processes can only be gained through empirical player research.

Finally, I’d propose an inductive approach to developing a medium-specific toolkit

for unnatural ludonarrative structures, taking into account the multiple ways in which
game mechanics allow us to execute procedural rhetoric to ‘read’ design features such
as illusory agency, slow gaming, action reversal and character duplication
functionally and with a view to gaining a deeper understanding of videogames as a
narrative art.

Acknowledgements
I would like to thank a number of people whose comments on earlier drafts of this
paper have been invaluable: Jan Alber, Anne Dippel, Sonia Fizek, Jesper Juul, Matti
Karhulati, Niklas Schrape, and Annika Waern. Big thanks also to Mathias Fuchs and
his team at Leuphana for allowing me to present this paper as a keynote lecture at
DiGRA2015, and for their copy-editing assistance.

References
24
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Alber, Jan (2013a) ‘Unnatural Narrative’, in Peter Hühn et al. (eds) The Living
Handbook of Narratology. Hamburg: Hamburg University, http://www.lhn.uni-
hamburg.de/article/unnatural-narrative (23/04/2015).
Alber, Jan (2013b). “Unnatural Narratology: The Systematic Study of Anti-
Mimeticism.”Literature Compass 10.5, 449–60.
Alber, Jan (2013c) “Unnatural Spaces and Narrative Worlds”, in J. Alber, H.S.
Nielsen, and B. Richardson (eds) (2013), A Poetics of Unnatural Narrative.
Columbus, OH: Ohio State University Press, pp. 45-66.
Alber, Jan, Stefan Iversen, Henrik Skov Nielsen and Brian Richardson (2010)
‘Unnatural Narratives, Unnatural Narratology: Beyond Mimetic Models’,
Narrative 18:2, 113-136.
Alber, Jan and Ruediger Heinze (2011) (eds) Unnatural Narratives – Unnatural
Narratology. Berlin: De Gruyter.
Blow, Jonathan (2009) Braid. USA: Number None, Inc.
Culler, Jonathan (1975) Structuralist Poetics: Structuralism, Linguistics and the Study
of Literature. London: Routledge.
Dawson, Paul (2013) The Return of the Omniscient Narrator: Authorship in Twenty-
First Century Fiction. Columbus, OH: Ohio State University Press.
Ensslin, Astrid (2011) The Language of Gaming. Basingstoke: Palgrave Macmillan.
Ensslin, Astrid (2014) Literary Gaming. Cambridge, MA: MIT Press.
Ensslin, Astrid and Alice Bell (forthcoming) Unnatural Narratives and Digital
Fiction. Invited for review by Ohio State University Press.
Fludernik, Monika (1996). Towards a ‘Natural’ Narratology. London: Routledge.
Hayles, N. Katherine (2007) "Hyper and deep attention: The generational divide in
cognitive modes." Profession 13: 187-99.
Herman, David (1999) Narratologies: New Perspectives on Narrative Analysis.
Columbus, OH: Ohio State University Press.
Jenkins, Henry (2004) "Game design as narrative architecture." In First person: new
media as story, performance, and game, ed. Noah Wardrip-Fruin and Pat
Harrigan, 118-30. Cambridge, MA: MIT Press.
Jenkins, Henry (2006) Convergence Culture: Where Old and New Media Collide.
New York: New York University Press.
Juul, Jesper (2005) Half-Real: Video Games Between Real Rules and Fictional
Worlds. Cambridge, MA: MIT Press.
25
 Pre-proof article forthcoming with Meson Press (Lueneburg, July/August 2015)

Linderoth, Jonas (2011) “Beyond the Digital Divide: An Ecological Approach to

Gameplay,” Proceedings of DiGRA 2011, http://www.digra.org/wp-
content/uploads/digital-library/11307.03263.pdf. Accessed 20 May 2015.
McCallum-Stewart, Esther and Parsler, Justin (2007) "Illusory agency in Vampire:
The Masquerade – Bloodlines." dichtung-digital 37, accessed January 11, 2013,
http://dichtung-digital.de/2007/maccallumstewart_parsler.htm.
Mularcyzk, Sam (n.d.) ‘The Stanley Parable: a Galactic Café game’,
http://www.stanleyparable.com, accessed 29 May 2015.
Nielsen, Henrik Sklov (2013) “Naturalizing and Unnaturalizing Reading Strategies”,
in J. Alber, H.S. Nielsen, and B. Richardson (eds) (2013), A Poetics of Unnatural
Narrative. Columbus, OH: Ohio State University Press, pp. 67-93.
Richardson, Brian (2002) "Beyond Story and Discourse: Narrative Time in
Postmodern and Nonmimetic Fiction." In B. Richardson (ed) Narrative
Dynamics: Essays on Time, Plot, Closure, and Frames. Columbus: Ohio State
University Press. 47-63.
Richardson, Brian (2006). Unnatural Voices: Extreme Narration in Modern and
Contemporary Fiction. Columbus: Ohio State UP.
Richardson, Brian (2011). “What is Unnatural Narrative Theory?” J. Alber & R.
Heinze (eds.).Unnatural Narratives, Unnatural Narratology. Berlin: de Gruyter,
23–40.
Richardson, Brian (2015). Unnatural Narrative: Theory, History, and Practice.
Columbus, OH: Ohio State University Press.
Ryan, Marie-Laure (2004a) Narrative Across Media. Lincoln, NE: University of
Nebraska Press.
Ryan, Marie-Laure (2004b) “Metaleptic Machines”. Semiotica 150.1: 439-469
Ryan, Marie-Laure (2009) “Temporal Paradoxes in Narrative.” Style 43:2, 142-164.
Šklovskij, Viktor (Shklovsky, Victor) ([1917] 1965). “Art as Technique.” L. T.
Lemon & M. J. Reis (eds.). Russian Formalist Criticism. Lincoln: U of Nebraska
P, 3–24.
Tale of Tales (2009) The Path. Ghent: Tale of Tales.
Wreden, Davey (2013) The Stanley Parable. USA: Galactic Café.

26