Introduction

Towards Video Game Aesthetics

Since the 1970s, much research on video games has recapitulated parental and governmental concerns about the effects of print, radio, film and television on youth. Case studies, observational/participant analyses, controlled experiments, questionnaire-based surveys, and correlative analyses have addressed concerns that video games are a “mindless addiction” that promote aggression and sexism, impede social skills, reduce school attendance and performance, undermine family life, and generally take time away from other social activities or responsibilities (Creasey & Myers, 1986; Egli & Meyers, 1984; Funk, 1993; Greenfield, 1984; Kubey & Larson, 1990; Mitchell, 1985; Provenzo, 1991). Indeed, Kevin Durkin's (1995) review of research on Australian video games was commissioned specifically by the Australian Office of Film and Literature Classification to address “cause and effect issues relating to the behaviour of children and young people” (1995, iii) (also see Durkin & Aisbett, 1999).

However, as Durkin (1995) makes clear, early researchers generally disproved the assumptions that they set out to test. Video games may be distinct from print, film, and other media, but they do not produce the negative psychological and social effects to the extent that critics claim. The recognition that video games are not necessarily antisocial is evident in popular histories of the medium and its culture (Demaria & Wilson, 2004; Herz, 1997; Kent, 2001; King & Bordland, 2003; Kushner, 2003; Sheff, 1999). In fact, research in Cultural Studies has turned to the positive psychological and cultural uses of video games, especially in relation to issues of identity and community (Baym, 1995; Beaubien, 1996; Curtis, 1992; Dibbell, 1993; Holmes, 1997; Ito, 1997; Kollock & Smith, 1999; McLaughlin, Osborne & Smith, 1995; Turkle, 1995; Reid, 1994; Rheingold, 1993; Stone, 1995; Wertheim, 1999). Nonetheless, Cultural Studies is a newcomer to video game theory and in 1995 Ted Friedman observed that while game designers and players had long discussed the aesthetics of video gameplay there was as yet no “software theory” (p. 73). Recently, greater attention has been directed to the aesthetics of video games (Aarseth, 1997; Murray, 1998; Poole, 2001; Wolf, 2001, 2004), and there has been a burgeoning of conferences, and print and online publications. However, researchers are still in the early stages of defining video game aesthetics compared with print and film.

This thesis is partly concerned with applying semiotics and structuralist narratology to a video game—an approach which requires some justification given the shift to poststructuralist and ethnographic methodologies in Cultural Studies (Jenkins, 1992; Morley, 1980, 1992; Morris, 2004; Tulloch & Jenkins, 1995). As Aarseth (1997) has argued, semiotics and narratology have already been applied to video games, but usually as a form of “theoretical imperialism” (p. 16) in which new media are evaluated in terms of theories developed for print and film. However, this thesis is not concerned with making claims about the exclusive relevance of semiotics or narrative theory; rather it is premised upon the assumption that it is preferable to interrogate the usefulness of older theories, such as semiotics and narratology, before arguing about the radical newness of video games. This will hopefully provide a better perspective when it comes to identifying the distinctiveness of video games as new theoretical objects. In the absence of such qualification, the privileging of “new” theories of “new media” may naturalise the increased turnover of academic capital, paralleling the ideologies of progress, capitalist innovation, and technological determinism which motivate and justify the development of “new media.”

As Frasca (2003) notes:

Certainly, formal approaches are limited . . . but they are probably the easiest way to uncover the structural differences between stories and games. I personally see this structural approach as a first, necessary step in video game studies, which we will definitely outgrow once it helps us to better grasp the characteristics of video games. (p. 222)

This thesis similarly accepts that there is a need for a consideration of structure. The early structuralists, of course, elaborated upon Saussure's (1915/1966) argument that meaning is produced through differences within a system by analysing the differential structure of texts (Barthes, 1966/1988, 1972, 1975a). Narratology was a dominant field in this approach, and was premised upon a distinction between what Russian formalists, such as Jakobson (1960), called “fabula,” the supposedly objective structure of the story in time and space, and “syuzhet,” the way in which this story was told. Following Barthes (1966/1988, 1975a), narratologists considered how the minimal units of narratives, “events,” were selected and combined in “sequences” according to logical/causal relations. Barthes (1966/1988) distinguished between “nuclei,” the indispensable units in a sequence, and “catalyzers,” units which amplify or fill in a sequence. Indeed, a fairly exhaustive category of terminology was developed, especially in regards to temporal and spatial anachronisms and points of view, or focalisation (Chatman, 1978; Culler, 1975; Genette, 1980; Lanser, 1981; Prince, 1987; Propp, 1968; Scholes & Kellog, 1966). It was argued that narratology would reveal a universal grammar of narrative organization across all types of media, though narratology was also used to distinguish narrative genres (Todorov, 1973) and media ( Metz , 1974) on the basis of how they limited, or facilitated, this grammar.

However, the following makes no claims about the structure of video games in and of itself, accepting Bordwell's (1986) argument that the fabula is not a structure in the text, but is constructed and reconstructed by the audience through ongoing inferences about events represented by the syuzhet. This thesis is more concerned with the way “codes” are activated by the syuzhet and how they structure (in the verb sense) players' ongoing inferential and interpretative activity. Within Cultural Studies, of course, a code is a system of signs governed by rules agreed upon by members of a culture, and defines conventional ways of making sense of texts (O'Sullivan, Hartley, Saunders, Montgomery & Fiske, 1994, p. 42). Without these codes there is no basis for communication or a common culture, yet at the same time their existence is no guarantee of shared meaning or experience. As Barthes (1975a) argues, each code

is one of the forces that can take over the text (of which the text is the network), one of the voices out of which the text is woven. Alongside each utterance, one might say that off-stage voices can be heard: they are the codes: in their interweaving, these voices (whose origin is “lost” in the vast perspective of the already-written ) de-originate the utterance: the convergence of the voices (of the codes) becomes writing: a stereographic space where the . . . codes . . . intersect. (pp. 20-21)

Codes, then, may have relative autonomy as rules that enable readers to decode a text, but they do not exist apart from the act of reading, they do not inhere in any medium, and they are subject to continual transformation. They also cannot be neatly separated except in an “ideal” text, as texts are constructed by their inter-relationship. Indeed, what this thesis takes from semiotics and structuralism is the attitude that the subtle permutations of signification—the “flickers of meaning” (Barthes, 1975a, p. 19) that result from the inter-play of codes—are worthy of consideration. If this thesis passes through a “structuralist moment,” then, it is only in the sense of attending to codes which can structure a reader's reception.

This thesis makes reference to Barthes' (1975a) distinction between proairetic, symbolic, hermeneutic, semic, reference, and diegetic codes. These are useful in considering how narratives and games organise space, time, characters, plots and settings, and how they employ various narrative strategies to produce vraisemblance and suspense. This thesis also makes reference to technical codes associated with media. Cinema, for example, has codes of focus, framing, angle and editing, which contribute to the meaning of a particular shot. Video games appropriate technical codes from both print and film texts, but also have their own codes which govern players' use of the interface and the aesthetic experience of gameplay. Drawing from Aarseth's (1997) account of cybertexts, codes relating to permissible actions at the interface are referred to as “ergodic.”

Codes are usually clustered in genres. In Hollywood cinema, for example, genres are usually distinguished in terms of iconography, structure, theme, character, setting and style (see Neale, 1980, 1990, 2000). Video games often draw from established narrative (print or film) genres, but they also draw from game genres, the most well-known of which are action, simulation, strategy, adventure and role-playing (Herz, 1997). However, video game genres were never defined through a rigorous analysis of formal differences, they were defined ad hoc and reinforced for marketing purposes. While the industry recognises between five or eight genres, with various sub-genres, these genres have always overlapped. Commercial and creative imperatives have also led games to draw from a variety of genres and other media, such that they have become increasingly hybrid (Darley, 2000; Herz, 1997; Wolf, 2001).

Having accounted for some of the common codes which govern gameplay, researchers will be in a better position to analyse how players negotiate with them. Of course, since any “interpretation” depends upon the person engaging in the act of “reading,” this thesis is less concerned with “ a reading,” than with probable and potential relationships between the text and its readers; that is, with preferred, negotiated and oppositional readings or reading formations (Eco, 1981; Hall, 1997; Jenkins, 1992; Morley, 1980; Tulloch & Jenkins, 1995). In the absence of ethnography, the role of the critic can be loosely identified with the writer of hyper-fiction who creates a text of forking paths, or, rather, several possible texts that are irreducible to a single text or world. Yet the point of such analysis is not to be as exhaustive as Borges' library, but merely to acknowledge the plenitude of meaning-making and, by implication, privilege a politics based upon difference, diversity, creativity, and negotiation. By analysing how codes inter-relate, we become more sensitive to the complexities of gameplay as a context for semiosis. Holding open the possibilities of a “text” in this way militates against simplistic evaluations about the “effects” of video games on players.

However, this thesis is concerned with codes not simply in terms of defining the “meaning” or “values” of games as “texts,” but also their role in producing the experience of gameplay. In this respect, the issue is one of aesthetics. There is, of course, no need to turn to idealists, such as Plato, Kant and Hegel, for whom aesthetics refers to supposedly “universal” criteria for value, beauty and taste (see Cothey, 1990; Dickie & Sclafani, 1977; Eagleton, 1990), nor to literary distinctions between “high” and “low” culture or art, such as those derived from Schiller, Arnold and Coleridge (see Bourdieu, 1984; Eagleton, 1983; Turner, 1984). In Cultural Studies all human activity is “artistic” in the broad sense of creative expression, and aesthetic value is seen as socially and historically specific. Within this framework, “aesthetics” can be seen as akin to “poetics” in the sense of analysing how certain codes may operate during one's engagement with a medium to produce aesthetic effects (Todorov, 1977, 1981). By extension, just as shared codes are likely to form the basis of discursive formations, it can be argued that shared aesthetic codes inform the trajectories of aesthetic experience. These may be relative to a medium (“film aesthetics”), narrative genres (the “aesthetics of heroic fantasy”), or game genres (the “aesthetics of first person shooters”). In the following, then, “aesthetics” refers to either: the codes which govern the relationship between formal qualities (of the medium, genres, or individual games); to some quality of aesthetic experience, including uses and gratifications (for example, aesthetics of “mastery” and “kinaesthesia”); or to the discursive representation of an aesthetic (for example, Friedman's (1995) theory of an aesthetic of “demystification”).

However, aesthetic codes are more connotative, subjective and polysemic than other types of communication (Fiske, 1982; Melrose, 1994; Turner, 1990), and their articulation is especially subject to the perceptual, cognitive and emotional activity of the individual. The effect of aesthetic codes is usually evident only in some quality of experience, and while Cultural Studies provides a useful theoretical basis for analysing the “textual” reception of aesthetic texts, its account of aesthetic experience tends to fall back on psychoanalytic theory and phenomenology. Cultural Studies tends to display a prejudice against the “natural” or “biological,” with the assumption that any reference to either involves a claim to “innateness” or “truth” (Gibbs, 2002; Grodal, 1997; Sedgwick and Frank, 1995). Yet much of a player's experience is only intelligible in terms of perceptual, motor, cognitive and emotional faculties, each of which has its history in the evolution of humans.

This thesis, then, draws from research on cognition and emotion to account for players' aesthetic experiences. Theories of cognition generally developed out of Jean Piaget's (1929/1973, 1932/1965, 1950/2001, 1951, 1978) analyses of the development of children, and have given way to the dominant Constructivist model that underlies most (non-psychoanalytic) models of psychology. This thesis is therefore in dialogue with the cognitive branch of film theory, characterised by the work of David Bordwell (1986) and Edward Branigan (1992), which examines the cognitive structures that govern film comprehension. While video games also have a sensorimotor component and require strategic thinking, Loftus and Loftus' (1983) account of the cognitive and sensorimotor work of players provides a precedent for the use of a cognitive model.

What needs greater qualification is that, despite the stated value of cognitive science, this thesis makes greater use of structuralist narratology and other methods of textual analysis when accounting for the ways that players represent and organise their experience. There are three main reasons for this. First, as Grodal argues:

Cognitive science is compatible with, or has roots in, different, earlier theoretical models in the humanities, and their special variants within film and media studies. . . . Many aspects of cognitive science are genetically closely related to structural linguistics . . . and freely share terms like ‘code' (pp. 13-14)

More specifically, terms like “code,” “type” and “genre” are analogous to terms like “schemata,” the theoretical structures which govern an individual's cognitive activity. Second, while cognitive science may seem to offer a more dynamic account of the individual's response to a media artefact (Grodal, 1997, p. 14), textual-based accounts, including structuralist narratology, have proven extremely sensitive to the subtle possibilities of signification. If we set aside the problematic argument that the computational structures of video games makes structuralist accounts more relevant, it can be argued that the long-standing methods of textual analysis are (at this early stage of video game research) more useful in suggesting the meaningfulness of video games than cognitive accounts. Third, this thesis employs methods of textual analysis as a way of dramatising the limitations of structuralist methodology, in the hope that later research will integrate textual and cognitive accounts in more balanced ways.

Since cognition has increasingly been seen as inseparable from affect (Tomkins, 1963; Nathanson, 1992) and emotions (Frijda, 1986; Lazarus, 1991), this thesis is also aligned with Tan's (1997) and Grodal's (1997) attempts to combine theories of cognition and emotion in their analyses of film. However, some more qualifications need to be made in this regard. First, Carrol (1990) and Grodal (1997) are hostile to the psychoanalytic tradition's emphasis on sexuality, gender, and the unconscious, arguing that many aspects of a viewer's relationship to a film are non-sexual, non-gendered, and conscious. This thesis generally accepts Carrol's and Grodal's claims, but it follows Bordwell's (1986, p. 30) suggestion that psychoanalytic theories are useful when cognitive accounts fail. It argues that psychoanalysis is valuable in conjunction with cognitive accounts, especially when analysing some games and some types of gameplay. Precedents in this regard can be found in Kinder's (1991) and Turkle's (1995) use of both cognitive and psychoanalytic theories in their studies of video gameplay.

Second, while this thesis remains open to other research on emotion it does not perform experimental research. Following Tan (1997) this thesis argues that there are some “homogenous” emotional responses that most viewers (or players) are likely to experience and/or are comparable (p. 154). In this respect, just as textual analysis often concerns itself with dominant codes and preferred readings, this thesis is concerned with dominant ways in which emotions are structured and/or are likely to be experienced. It presumes the existence of, not a “Model Reader,” but what, following Tan (1997), can be called a “natural” player, that is, a player who has a preference for the (type of) game under analysis and who freely chooses to enter into the experience of play (p. 10). While all players may dislike and resist aspects of a game, a “natural” player “generally makes no effort to escape the attraction of the fictional world” (p. 10), preferring to progressively invest in the diegesis. At the same time, this thesis regularly indicates potential variations in players' emotional responses, analogous to textual analyses of negotiated or oppositional readings. In so doing, this thesis makes the kinds of qualifications that will hopefully make future quantitative or ethnographic analyses more sensitive to video games and their players.

However, it must be noted that while the following emphasises textual analysis over experimentation, it avoids referring to its human subjects as “readers,” as is common in Cultural Studies, or “viewers,” as is common in the cognitive tradition of film theory, preferring the term “player.” This is not to deny that viewing, or playing, is mediated by textuality, nor to offer a rhetorical statement that “readers” are, and always have, been “players” in a textual game. The role of textuality is too well-known to need foregrounding, and other perceptual, cognitive and emotional processes may be of equal or greater import, so always calling video game players “readers” is misleading. That said, this study occasionally uses the terms “reader” and “viewer” to indicate the mode of engagement that is of theoretical emphasis, given that a video game player (or a theorist of video gameplay) may focus on reading more than viewing, viewing more than reading, or playing more than reading or viewing.

Lastly, the theoretical conjunction this study offers is developed both as an account, and through an account, of a single game: Squaresoft's single-player, graphical adventure role-playing game Final Fantasy X ( FFX ) (2001). The Final Fantasy series is notable because most early computer role-playing games (CRPGs), including series like Bard's Tale (1982), Ultima (1980), Might and Magic (1986), and Wizardry (1981), were developed for PCs. Since the CRPG was a relatively minor genre on game consoles, the Final Fantasy series helped to popularise the genre for a broader console market, or at least make it more recogniseable. Unfortunately, the Final Fantasy titles IV , V , and VI were not released outside Japan until recently, creating some discrepancies between the titles in Japan and other countries. If we follow the Japanese sequence (recapitulated in the versions recently released for the Playstation), FFI , FFII , FFIII , FFIV , FFV , and FFVI were developed for the Nintendo cartridge market. With FFVII (1997) Square moved to the Sony Playstation, spending three years and US$30 million to produce detailed backgrounds, animated sequences, and a complex story that took advantage of its three CD-ROMs. FFVIII (1999) and FFIX (2000) utilised the same graphical capabilities, though FFVIII , with its futuristic setting and more complicated magic system, deviated more from the existing formula than fans wanted. FFIX was a more popular return to the medieval flavour and simpler gameplay of the earlier titles.

FFX was created for the Playstation 2 and offers enormous graphical improvements over its predecessors, but similar gameplay. Players navigate panoramic landscapes, collect items, fight monsters, solve puzzles, and develop characters' physical and magical powers while advancing through a complex narrative. The major difference is that, in addition to the traditional Field Screen, Battle Screen, and various Menu Screens, FFX uses the Grid Sphere to manage skill development. This is comprised of a network of nodes, each representing a particular skill or skill bonus, which characters traverse and activate as they accrue experience points. The story, at its most basic, follows Tidus, Yuna, and Yuna's Guardians (Wakka, Lulu, Kimahri, Auron and Rikku) on their pilgrimage to defeat Sin, a monstrous being who has ravaged the world of Spira for a thousand years. Familiarity with FFX 's ninety or more hours of gaming and narration would certainly give substance to the following arguments, but such familiarity is not presumed. Appendix One lists key settings and events; Appendix Two is a breakdown of codes in the opening sequence; and Appendix Three is a breakdown of dominant clusters of hermeneutic codes. Other relevant aspects of the game are described where necessary.

The choice to apply several different theories to this single game has its analogy in Genette's (1980) analysis of Proust's (1934/1970) Remembrance of Things Past . Like Genette, it would be possible to “put the specific subject at the service of a general aim, and critical analysis at the service of theory,” such that FFX would be a “reservoir of examples, and a flow of illustration” (p. 22) for several theories about gameplay. Alternately, one could “turn the concepts, classifications and procedures . . . into so many ad hoc instruments exclusively intended to allow a more precise description” (p. 22) of FFX .

However, the complexity of FFX makes it more useful to take a dual approach: to use several theories to better appreciate the experience of playing a popular game, and to use a popular game as a means to test and refine those theories. Early research was largely directed towards games oriented around, on the one hand, motor skills or strategic thinking (arcade games, first person shooters, war games), or, on the other hand, narrative (adventure and role-playing games). While this is changing, the emotional distinctiveness of video games may be owed to their peculiar convergence of sensorimotor and cognitive activity, and this has not been adequately theorised. While FFX has few of the kinaesthetic sequences that characterise first person shooters (FPSs) like Doom (1994), it nonetheless combines many types of gameplay. Indeed, the Final Fantasy series is significant not just because of its consistent narrative depth or gaming innovation (a few other PC and console games may claim equivalent or superior depth and innovation), but its popular integration of both for an international market.

Chapter Overview

While the organization of the following chapters owes something to Tan's (1997) analysis of emotion in the traditional feature film, the distinctiveness of video games, and FFX , requires some modifications of his structure. Chapter One reviews the relevant theories and research on play, cognition, aesthetics, narrative, and emotion. Chapter Two draws from Grodal (1997) to consider the feeling-tones experienced as a consequence of the perceptual construction of reality-status at FFX 's interface. Chapter Three adapts Tan's (1997) account of “interest” as the basic emotion in the traditional feature film to video gameplay, but makes greater use of structuralist methodology, and emphasises problems with the regulation and transfer of interest across FFX 's narrative and game macrostructures. Chapter Four looks at how FFX 's hermeneutic coding of reality regulates interest, or, more specifically, how analysing the game in terms of the marvellous, the uncanny and the fantastic suggests that curiosity, surprise and suspense about (dis-) expected events may give way to wonder, fear, and/or anxiety about events.

Chapter Five reviews structuralist accounts of character before analysing how interest is reinforced through cognitive identification of characters in FFX 's narrative and game macrostructures. Chapter Six addresses some of the criticisms against psychoanalytic accounts of identification, then considers the relevance of Cowie's (1997) psychoanalytic account of three modes of identification—with the ideal ego, ego-ideals, and super-egos—to gameplay. Chapter Seven elaborates upon the empathetic emotions players are likely to feel towards characters as a result of innate releasers and cognitive appraisal of situational contexts. On this basis it discusses some of the dimensions of happiness, anger, fear and sadness, but argues that FFX 's tragic macrostructures may elicit a dominant mood of sadness. The chapter concludes by arguing that allegiance with a character's moral type reinforces or blocks empathetic emotions towards video game characters.

Chapter Eight argues that the emotional experience of gameplay is often characterised by players' empathy towards themselves in the play-situation. This is addressed in terms of how gameplay promotes and blocks what Piaget (1951) calls “preoperatory” cognition, with its accompanying “egocentrism.” Within a psychoanalytic model, the blocking of players' egocentrism might be presumed to give rise to fear or anger. However, a better account of how players experience, and self-regulate, such blocking is offered using Silvan Tomkins' (1963) and Nathanson's (1992) model of shame, defined as an auxiliary affective response to a sudden, unexpected loss of interest. The chapter concludes that the transfer of empathy players feel for themselves to characters, and vice versa, may distinguish the video game as a medium.

Chapter Nine argues that ambivalent reality-status, the dysregulation of interest, the blocking of empathy, and other aspects of gameplay, produce an experience of “meaninglessness” that is recognised in self-psychology (Erikson, 1950, 1968; Kohut, 1971, 1977), but which finds its analogy in Sartrean (1943/1993) existentialism. However, existential feelings of “meaninglessness” may be interpreted as part of FFX 's meaning and aesthetics and/or may be covered over when “affect grouping” (Tomkins, 1991) motivates players' (re-)evaluations of gameplay.

Footnotes

1. This thesis is, of course, not a manual on game design, but it provides accounts which parallel, or implicitly explain the significance of, some rules of game design, such as those pertaining to character motivation and emotionally complex situations (see Bates, 2001; Freeman, 2003; Kelly, 1998; Kim, 2000; Lewinski, 2000; Rouse, 2000).

2. Personally, I have only played FFVI , FFVII , FFVIII , FFIX , FFX , and FFX-2 . Both FFVII and FFX were replayed and recorded onto (eight three-hour) video tapes during the course of this study to facilitate analysis.

3.

All the pictures used in this thesis were taken from the Square Central (http://www.square-central.com/) and the Final Fantasy Shrine (http://www.ffshrine.org/). Both sites were last checked January 1, 2005, and all pictures taken from them are public domain.

4.

In addition to the two guides cited in the references and the Websites noted above, the following sites are useful sources of information about the game and/or gaming culture: Unofficial Final Fantasy Site (http://www.uff9.net/), Final Fantasy Insider (http://www.ffinsider.net/), Final Fantasy Fusion (http://www.fffusion.com/), Fantasy Square (http://www.fantasysquare.com/), NeoMidgar (http://neomidgar.com/), Final Fantasy Spirit (http://www.ffspirit.net/) and Square Nation (http://www.squarenation.com/).