Toward a Visualization of Video Game Cultural History: Grasping the French Touch

Let’s Get Digital

As Ricoeur (2004) explains in his sweeping overview of historiography, the integration of “mobility in the historian’s gaze constitutes an important conquest of history during the last third of the twentieth century” (p. 268). In the course of the 20th century, the practice of historians benefited from methods associated with other disciplines, from the integration of sociological concepts and statistical methods (starting with l’École des Annales) to the development of microhistorical accounts inspired by ethnography and archaeology (from the 1970s onward). Relocated at the confluent of humanities, history as a discipline was able to observe and analyze from a multiplicity of vantage points. More recently, humanities have benefited from methodologies and tools emerging from the computer medium: digitization of existing archives, proliferation of user data on the web, and novel visualization interfaces. Considering that such techniques are already used by media scholars to develop possibilities of “distant readings” (Moretti, 2005; see also Boni’s work on TV series, 2015), one would think that a field like video game studies already benefits from the additional mobility afforded by digital humanities. Such techniques are not widespread in the discipline at the moment, even less so in the context of video game history.

In “Art and Chartjunk: a Guide for NEUVis,” Gough, Dunn, Bednarz, and Ho (2015) highlight different ideologies in the design of visualizations for nonexpert users. A few research projects are currently being conducted on large, game-related data sets. For instance, Triclot studies the expansive player metrics collected from game controllers under the term “rythmanalyse” (2013). These approaches are mostly synchronic in nature: Digital tools are used to inspect the video game phenomenon at this point in time. But what if game studies had access to a mobile vantage point, better suited to study diachronic data in a consistent manner?

There are many difficulties preventing us from undertaking vast historical inspections of video game culture. The discipline is still refining its conceptual underpinnings; as Jeffrey Schnapp would put it: Scholars in the field are still hard at work on “knowledge design.” In a recent publication from L’internationale summing up his reflections about digital humanities, Schnapp highlights three challenges: quantity of data, articulation of micro/macro scales, and fragility of media (2016). Many efforts have emerged in the past 10 years to tackle the issue of video game fragility and preservation (see Lowood, 2009; Stuckey, Swalwell, Ndalianis, & de Vries, 2015; Suominen, Sivula, & Garda, 2018; Swalwell, 2007). The issue of quantity is becoming more salient lately with the “discovery” of international video game histories; scholars entering the field at this moment have to tackle several tens of thousands of objects emerging from all over the world. Efforts are being made to establish metadata for video games (see Kaltman et al., 2016), but such contributions cannot be integrated within historical video game databases, for very few institutions have undertaken such a sweeping initiative at this point. For the time being, our historical investigations have to rely on user-populated databases such as MobyGames, which documents more than 200,000 games through dozens of overlapping concepts.

The dedication of the gaming community populating these historical databases needs to be acknowledged in a context where no institution is able to provide similar efforts. Such resources make the proliferation issue even more tangible; “The challenge intensifies in the case of vast, interconnected, multimedial digital repositories. There was already far too much to sift through and make sense of; now there is exponentially more” (Schnapp, 2016, p. 20). Moreover, fan labor cannot be held accountable for lack of consistency in applying dozens of conceptual terms over these objects, even more so in an increasingly globalized culture. The infrastructure introduced in this article seeks to provide consistent knowledge design and encoding protocols in order to populate an extensive database of video game culture and thus provide a reliable mobile gaze that will allow us to navigate data proliferation in a way that can make sense.

Already, a strong underlying assumption emerges from the previous mission statement: A specific culture can be understood in a satisfactory manner through the study of related props created by human groups. On top of artifact inspection, a more holistic approach should also integrate ethnographic studies with player communities. However, very few ethnographers work within a diachronic mind-set in game studies, and one can imagine the difficulty of conducting a project that would seek to interview players from the last 60 years. In this context, inspecting the games only might appear like a convenient cop-out. We should stress, inspired by Graeme Kirkpatrick’s (2015) integration of a Bourdieusian perspective, that games and gamers are caught in a reflexive entwinement; certain game forms proliferate in response to player habitus, and this habitus is informed by the gaming experiences readily available at any given time (p. 75). Jørgensen’s (2013) definition echoes these observations: “[gameplay] is not something one can design explicitly, but designers can facilitate a specific kind of gameplay through designing specific kinds of game mechanisms and the context in which they occur” (p. 33). Inspecting video games in order to conduct large historical inspections is not only convenient; it provides a clear reflection of how players engage with the medium. This leads us back to the contentious issue of selection.

Polyphonic Canon

The goal of our project is to encode a large amount of games using a consistent analytical tool in order to trace the evolution of video game culture. In order to address the articulation of micro/macro scales put forth by Schnapp (2016), and to develop the mobility of scale lauded by Ricoeur, we need a search engine “that focuses upon the retrieval, not of individual items, but of the networks of interconnections that run through them” (p. 22). Ideally, hundreds of titles in every possible genre would be integrated in the database. This article reflects a more practical scenario where 140 adventure games were selected and analyzed by research assistants (RAs), in order to provide a “proof of concept” that would lead to an understanding of common gameplay configurations, their relative prevalence at specific moments of history, and the original nature of a specific subcorpus (the French Touch) within this expanding ensemble. In this section, methodological issues pertaining to the selection process and the procedure that emerged to alleviate these issues are presented.

The adventure genre proved to be a relevant and realistic testing ground for this project. As Arsenault (2009) points out, building on Alastair Fowler and Rick Altman, genre labels reflect a certain consensus reached within a given community . The proliferation and recombination of these labels strive to keep up with the evolution of cultural productions. Arsenault’s account highlights how “genre is being used in a similar manner across popular, industrial and academic taxonomies: as a way of breaking down the vast continent of videogames into more manageable provinces.” His overview of different taxonomies underscores “the fluidity and impreciseness of the concept of genre itself”; generic markers include thematic, iconographic, ergodic, and technological elements (2009). Some game experiences have crystalized to the extent that a specific assemblage of generic terms can express their core elements with great clarity for members of a given discourse community, while others obviously challenge easy categorization. Such communicative shortcuts have been essential in the development of video game marketing, journalism, and have also been appropriated by gamers. Thankfully for our project, many of these communities have embraced the discursive genre known as the “listicle” in order to express their conception of exemplary titles in the history of specific video game genres. This discursive proliferation allowed us to develop a polyphonic corpus selection method.

Journalists from the specialized video game press have proposed “top games” features for a long time; such features can already be seen in Computer Gaming World (CGW) in the early 1980s. The CGW editorial team expanded multiple times on their initial division (strategy and action/adventure); This initial segmentation mixed up two genres—action and adventure—that would become quite distinct a few years later. In our contemporary setting, both journalists and gamers produce their own listicles, and it is easy to find documents that singularize the adventure genre. Five “top adventure games” listings emerging from a variety of channels (traditional press, professional and amateur websites, and YouTube channels) have been selected, providing dozens of potential titles to integrate in our data set. In order to form a list of relevant titles encompassing the majority of video game history, we have privileged recent publications. This criterion came with the risk of a certain historical blindness for older titles.

In order to reflect the consensual aspect of genre edification, we have decided to integrate in our database every title that was mentioned 2 times or more across these listings. More than 50 titles emerged from this process and confirmed the expected invisibility issue: Most of these games were published in the 1990s. This might be a consequence of the format more than pure historical blindness from the authors: A majority of genre listicles are created with an overt elevating bias (pointing out the “best” elements of a phenomenon), and many would argue that adventure games reached their paradigmatic perfection in the 1990s. Yet some parts of history are clearly obscured in these sources. Symptomatically, Adventure Gamers’ “Top 100 All-Time Adventure Games” integrates a vague mention of “Infocom text adventure games” alongside 99 very specific titles. According to MobyGames, the famous company created at least 44 distinct text adventures; it is doubtful that the author meant to populate this top 100 with all these additional games.

Many mechanisms have been integrated to compensate the elevating bias and potential historical blindness found in our sources. First, we have added sources that purport to list underrated gems (“underdogs”) and even the worst examples in the genre (“lemons”) in order to account for a greater variety of experiences. These additional sources emerging from fans and journalists often include the same titles that were mentioned in the extensive “top of the genre” listicles; “worst games” features typically seek to make a counterargument about the overrated nature of popular Sierra and Lucasfilm titles. This highlights an overall visibility problem since the worst games in the genre’s history are apparently so invisible that they don’t even manage to break through in dedicated features. In order to circumvent the visibility issue, we have hunted down top game features emerging in the specialized press; two such features were found in CGW and Edge, providing a few additional titles in our canon (Edge Editorial Team, 2000; Sipe, 1992).

The second compensating measure brought another discourse community to the table: genre experts. Two members of the original team have been adventure game players since the late 1980s. Additionally, two academics who have studied the genre extensively—Jonathan Lessard and Sébastien Genvo—were contacted. These scholars were able to complement our corpus in many ways. They could choose to act as the “second voice” in order to integrate titles from the extensive “best of” features. Their main contribution however came in the “underdog” category; video games that were not mentioned anywhere could be added, emphasizing titles released by lesser known studios and hybrid experiments. This allowed us, most notably, to add action-adventure and text adventure games to the mix. Finally, we have decided to consult resources curated by academics, most notably the LGBTQ Videogame Archive, in order to find additional early underdogs that reflect marginal voices: Caper in the Castro (Ralph, 1989), Rex Nebular and the Cosmic Gender Bender (MicroProse, 1992), and The Orion Conspiracy (Divide by Zero, 1995).

While nonprobabilistic, this sampling method provides an intersubjective view into genre formation and history. Items could have been selected randomly (in MobyGames listings for instance), but such a random sampling would have likely excluded most well-known titles, offering little anchoring to potential readers from the gaming community. Building on the knowledge shared by various discourse communities, we believe this method circumvents the contentious issue of authority in corpus selection. Altogether, a list of 116 adventure titles emerged from this polyphonic selection process. Thanks to the combined expertise of fans, journalists, and academics, a plurality of appreciative criteria were accounted for in the creation of our game canon. One of the obvious shortcomings comes from the Anglocentric nature of the procedure and the accentuated visibility of American games in the globalized video game industry. Ideally, our game canon would be supplemented with titles that had relevance in a variety of regions, thus reflecting the growing sensitivity toward international game histories (notably with the publication of Videogames Around the World [Wolf, 2015], Videogames and Postcolonialism [Mukherjee, 2017], and Gaming The Iron Curtain [Svelch, 2018]). For this article, we were able to integrate the criterion of “regional / local relevance” with a French corpus since many of the involved scholars speak the language fluently. Two of the contributors work from France and have provided an additional selection of 24 games. With this corpus of 140 games, we set out to test the comparative affordances of the visualization infrastructure.

HACS Adventures

In order to provide a consistent analysis of gameplay configurations inscribed within all these objects, we have used the Historical-Analytical Comparative System (Therrien, 2017). HACS emerged from a literature review of formalist or ontological approaches to gameplay. Some systems can be very economical (10 gameplay bricks, Djaouti, Alvarez, Jessel, Methel, & Molinier, 2008), while other approaches are incredibly expansive in scope (hundreds of concepts in Björk and Holopainen’s [2004] Patterns in Game Design and dozens in Juul’s [2016] inspection of tile-matching games). As Ea Christina Willumsen (2018) points out in a recent contribution, many examples of game formalism seek to achieve an ideal level of abstraction, “related to the need for a formal language that accounts for—and helps categorize—the specific elements of a system” (p. 140) while keeping a focus on the essential property that games are meant to be played. Björk and Lankoski (2015) recommend a meticulous formalist description of game content, acknowledging that some modern games are too large to be described with such a technique. A landmark contribution by Elverdam and Aarseth (2007) offers abstract categories of time and space that can be taxing for RAs to master and apply systematically on a large corpus of games.

Even though on the surface HACS appears to be very similar to other formalist approaches, Therrien stresses that the system doesn’t seek to make strong ontological claims about core game components and highlights the arbitrary nature of some conceptual selections. In order to facilitate the encoding process, HACS has privileged concepts that are commonly employed within games, game manuals, and by gamers themselves. The system encompasses 28 main concepts within 5 categories, which interlock in order to transcode a great variety of gameplay experiences. HACS was built with an experiential bias: It seeks to highlight the types of virtual actions performed within the game (such as “navigation,” “neutralization,” or “resource management”), the modeling of these mechanics through the interface (from manipulation props to the type of feedback provided to players), and the modes of performance associated with these mechanics (such as puzzle resolution or strategy building). A great level of precision can be reached, thanks to the conjunction of these categories; for instance, some games model navigation with a screen-augmented interface, a synchronic action mapping (often expressed with the idea of “real time” by gamers) and with elliptical visual feedback (one appears to be “jumping” through space). Overall, only one concept that was not integrated in the original framework was added to better account for textual adventure games (the notion of “verbal transcoding,” to highlight the specificity of various types of text parsers seen in early adventure games). It would be inconvenient to sum up all the concepts in a single overview of the analytical framework; we have preferred to introduce relevant aspects and definition as we move forward in the presentation of our results.

In order to ensure stable coding, each category has been defined through broad continuums commonly found in academic discussions (such as “Natural interaction < – > Symbolism”) in which every concept is situated. Additional definition elements and examples were provided to RAs, along with encoding guidelines for in-between cases. The synthetic nature of this analytical system allowed us to design a 1-hr-per-game encoding procedure: RAs were instructed to consult paratextual elements (such as the game boxes and database entry on MobyGames), to experience the game firsthand, and/or to look at audiovisual archives uploaded by players on YouTube. The proliferation of Let’s Plays and Longplays represented an ideal way to look up for mechanical variations in later stages of a given game without going over the 1-hr encoding protocol.

Interactive figures, the first category specified through HACS, refers to domains of human agency modeled in the game system through mechanics and challenges. A figure such as navigation or “social interaction” can be modeled in a great variety of ways. The working hypothesis is that a genre can be defined by the prevalent interactive figures it integrates at any given time and by the potential variations of these figures through interface modeling/mode of performance. A synoptic visualization of our data set reveals that there is indeed a great commonality of interactive figures in the adventure genre; 5 of the 10 figures listed in HACS occur in more than 81% of the encoded games (Figure 1). Interestingly, the sixth most common figure is neutralization (34%), which is more typically integrated in fast paced action games. Hopefully, the rest of the HACS architecture will help us solve this mystery before the end of this article.

OPEN IN VIEWER

Figure 1.

Most prevalent interactive figures in adventure games.

The visualization infrastructure was developed in a way that would provide some assistance to researchers in order to find out the most common configurations for each interactive figures. Thanks to this algorithm, we can observe that navigation has been implemented to a large extent via a screen-augmented interface; through on-screen pointers, menus and icons, players indicate where they wish to go (Figure 2). This type of interface is situated in the “symbolical” region of HACS’s “Natural interaction < – > Symbolism” continuum. The only concept that would go further in that direction is verbal transcoding, which refers more specifically to textual parsers where players had to formulate multiple word sentences in the hope of exerting any kind of virtual action (such as “go west”). In the realm of adventure game navigation, this was prevalent throughout the 1980s; from 1978 to 1988, 32% of our canon uses this model. Interestingly, many early adventure games used a “generic” interface for navigation, meaning that the action was controlled through a handheld prop (joystick or keypad) without any on-screen addition; this was most notably the case in Sierra On-Line games built with the Adventure Game Interpreter engine, where every other interactive figure was conducted through a text parser. In this context, action mapping can be said to be “synchronic”; it is carefully bound to the actual manipulations of players on the interface. Similar designs vanished during the heyday of graphical adventure games, but a notable comeback is visible in our data set after 2010 (in Telltale games notably, and Night in the Woods, Infinite Fall, 2017).

OPEN IN VIEWER

Figure 2.

Most common configuration for the navigation figure.

The category of “mapping” is certainly the most abstract in HACS and needs to be explained more thoroughly in order to make sense of gameplay configurations. It is based on the same continuum and seeks to highlight five common ways to map player actions on the interface with virtual actions exerted by an avatar/player character, from the most natural (“symbiotic mapping”) to the dissolution of this action mapping altogether (“punctual mapping”; the action is mapped only at a specific point in time before the system takes over; Figure 3). The original paper introducing HACS highlights how most of these terms have already been defined in other academic contributions (Therrien, 2017); punctual mapping was inspired in part by James Newman's notion of “ergodic punctuations” (2002). In the context of adventure games, it is common for players to merely initiate an action and witness the protagonist perform without any additional input, most notably for the social interaction figure; once selected, a specific phrase and its impact on the conversation will be automated. The fascination for graphical adventure games developed in part because of this spectacle element, which could be associated with a form of “spectacle-reward” (Surman, 2007); popular characters such as Guybrush Threepwood gained personality and the corollary fan admiration through this partially automated performance.

OPEN IN VIEWER

Figure 3.

Typology of action mapping (Therrien, 2017).

The feedback provided to players for navigation in this context emerges from the world itself (we can see characters moving); more abstract visual or auditory signalization (such as guiding arrows or a GPS) is rarely seen in our corpus. Here we refer to the third interface category, “gameworld feedback,” building on a continuum explored thoroughly in Gameworld Interfaces (Jørgensen, 2013), with diegetic feedback on one end (departing somewhat from Jørgensen’s terminology) and signaletic feedback on the other. Many other action subgenres have integrated guiding arrows, flashing overlays or a full-fledged GPS. Interestingly, graphical adventure games still provide verbal feedback for most of the typical interactive figures, including navigation. This is clearly a leftover from purely textual modes of engagement; verbal feedback, either extradiegetic or expressed through the voice of the main character, appears to compensate for a certain lack of clarity in early graphical adventures, but can be found commonly up to the late 1990s (for instance in The Journeyman Project 3: Legacy of Time, Presto studios, 1998), and even in more recent titles.

Another major navigational model (52% of the titles in our data set) developed in graphical adventure games relies on a synthetic form of visual feedback—a rapid passage from State A to State B that evokes a traditional visual ellipsis, for instance when the player character walks off-screen to remerge in a completely remote location. This type of feedback can also take part in a form of synchronic mapping that is quite distinct from the one enacted in action games and in early Sierra games. In a few games such as Caper in the Castro (Ralph, 1989), players “jump” from one portion of space to the next, as seen through the eyes of the protagonist; things move quickly enough to be experienced as a synchronic mapping. Such a configuration became typical of many role-playing games at the turn of the 1990s; “space by space” navigation through a screen-augmented interface and synthetic visual feedback has been described as a “real-time” experience by marketers and gamers. This is easier to understand now that we have exposed the previously dominant model of punctual navigation (either with verbal or visual feedback).

The case of cumulative mapping is fascinating; adding up multiple primitive actions on the manipulation interface in order to perform one virtual action implies an obvious symbolic convention, yet the accumulation of efforts in itself seeks to evoke the complexity of real-world gestures. In the history of adventure games, verbal transcoding interfaces often imply such mapping design; multiple word sentences need to be formulated properly in order to progress. Lessard (2014) and Fernández-Vara (2008) point toward an “ergonomic turn” with the widespread integration of mouse-based, screen-augmented interfaces; clicking on icons and on the screen display in itself corresponds to a streamlined versions of agency in the genre and in such cases action mapping could be punctual (single or double clicks trigger autonomous visual or verbal feedback to confirm that the action was registered). For the figure of resource management, action design quickly developed in a much more convoluted manner that reintegrates a form of cumulative mapping. Many “point and click” titles rely on a three-step formula: select an action verb (such as “use”) or action icon, select an item in inventory, and finally identify an element in the visual display to be acted upon. Between 1985 and 1995, 24% of games integrate cumulative mapping for resource management.

In the history of video games, more natural/symbiotic action mappings have exerted a constant fascination. Not every type of manipulation interface can afford such mappings. Techno-mimetic interfaces, such as steering wheels or guns on arcade machines, offered symbiotic actions even in the electromechanical era for navigation and neutralization figures, respectively; the latter has even been modeled through a corporeal interface such as the Wiimote. This ideal of natural interaction—the dream of transparent mediation reenacted in video games (see Jørgensen, 2013, pp. 34–36)—has been explored in action games to a large extent, but an interesting twist occurs more specifically in the history of adventure games.

While this might not seem as obvious from our contemporary perspective, the corporeal potential of mouse interfaces was clearly put forward when it was first integrated in Xerox’s Alto, Apple’s Macintosh, Atari’s ST, and Commodore’s Amiga computers. The metaphor of the desktop involved a lot of screen-augmented tokens but also the ability to “drag and drop” these elements between different areas. This feature, in essence, captures natural gestures in a slightly abstracted manner. In the realm of adventure games, it was integrated as early as 1985 (Deja Vu: A Nightmare Comes True!!, ICOM Simulations) for resource management; players “pick up” their trench coat in the environment and drag it in their inventory window. Genvo’s research exposed an even earlier example occurring in France; Jean-Louis Le Breton and Fabrice Gille proposed one of the early attempts to integrate this configuration in La souris Golotte, released in 1982 for the Apple IIc (and integrated by Apple as a demo to showcase the potential of mouse control). Outside of the desktop metaphor, the abstracted nature of this 2-D gesture comes forward. In the context of HACS, such abstraction (which still encompasses a strong mimetic component) is defined as a miniature—or metonymic—gesture. As we can see in Figure 4, this model really took off in the following years and can be associated closely with the ergonomic and graphical paradigm in the genre’s history.

OPEN IN VIEWER

Figure 4.

The rise of the “Resource management>Corporeal manipulation interface>Metonymic action mapping” configuration. The first bar for each year represents the amount of games integrating the interactive figure, the second bar the ones that accentuate the “corporeal” nature of the mouse interface, and the third highlights the amount that corresponds to the full configuration.

Coming back to the neutralization peculiarity, we can observe that games which integrate this component rely on a generic or screen-augmented interface and synchronic action mapping. As such, they aren’t configured in a strikingly different design compared to action games, and for good reason: They were conceived as action mini-games within the grand adventure ensemble (typically for a climatic final confrontation). How can HACS account for these hybrid scenarios? The last category seeks to highlight the overall performance mode associated with each figure. RAs encoded these modes on a figure-by-figure basis, but the system makes it easy to present the challenge of any given game in a synthetic manner. Thanks to the work of different scholars, HACS defines modes of performance according to three main criteria: length of action planning, type of skill set (minimal, sensorimotor, or imaginative), and the nature of underlying structures (static or dynamic). Most adventure games epitomize the mode of resolution, understood as a mid-term to long-term action planning, imaginative activity directed toward static structures (such as predetermined locations, conversation pieces, items, and puzzle structures). HACS encoding can be useful to understand which modes of performance are integrated within a game, and which mechanics are integrated within each mode. For instance, Figure 5 sums up the gameplay configuration from Amazon: Guardians of Eden (Access, 1992).

OPEN IN VIEWER

Figure 5.

Amazon: Guardians of Eden’s gameplay configuration.

The “execution + resolution” combination in Figure 5 transcodes the climatic confrontation, where players need to take down a giant ant by hitting the monster at the right moment (they perform a static action sequence which is taxing on the sensorimotor level) and need to figure out which item to use at the right moment on top of this motor challenge. Lessard (2014) might question our inclusion of “Navigation” in the mode of “resolution”; “Movement in these adventure games is usually trivial, the main challenge lying in identifying places to go to and gaining access (unlocking doors and so on).” Indeed, few adventure titles integrate spaces that are so labyrinthine that the figure in itself would be set up as a resolution challenge (it was not rare, however, to integrate labyrinths as a subsection in many Sierra and Lucasfilm games, and Westwood’s Legend of Kyrandia, 1992, for instance). We have decided to encode these modes not as an object-based assessment but from an experiential perspective. In the course of playing, gamers need to elaborate complex action sequences in order to progress; while the navigation of space might not be the primary element in these elaborations, these mechanics are inevitably part of the action sequence. As adventure players, we are coming up with resolution hypotheses that we could artificially fix with the following phrase: “navigate to X space, examine X element, talk with X character, activate X device with X tool.” Through this association between modes of performance and interactive figures, HACS can provide one of the most extensive, gameplay-centric definitions of game genres; in the case of adventure games, the core configuration relies on resolution challenges embodied through navigation, apprehension, social interaction, environmental activation, and resource management mechanics.

The Touch

The French incarnation of the adventure phenomenon has been epitomized in many historical books. Donovan’s (2010) Replay dedicates a few pages to the touch, alongside other major movements such as “English surrealism” (pp. 127–131), in an explicit effort to include more regional/international elements to supplement U.S.-centric efforts. As Donovan (2010) relates, “A summer 1984 article in Tilt reported how French game designers, having cut their teeth on simple arcade games, now wanted to create something more personal, more rooted in reality, more French”; based on many accounts, he associated the touch with “strong narrative, real-life settings and visuals inspired by the art of France’s vibrant comic book industry” (p. 128). Guillaume de Montagnon and Alexis Blanchet (2019) have done extensive field work to unearth unknown aspects of French video game history; the title of their book—French touche—will undoubtedly contribute to further cement this generic expression. While scholars question the idea of a concerted creative approach (which is strongly implied by the notion of a movement), it is widely accepted that French game developers proposed bold offerings most notably through the adventure genre. As Montagnon and Blanchet explain, the American console crash led to a reorientation of the market in France toward personal computers, giving the technical means for many small studios to emerge. Game authors had more freedom in this context, “going from one editor to the next depending on the nature of their project and professional affinities” (in press; freely translated). In this section, we seek to test our infrastructure’s ability to highlight the French touch’s originality (or lack thereof).

Early French titles encoded in the database through HACS reveal how designers were inspired by gameplay formulas proposed by their American contemporaries; as Arsenault (2009) point out, this mimicry phenomenon is common when it comes to successful new generic formulas. Le Vampire Fou (Le Breton, 1983)—the first French text adventure game—adopts a configuration reminiscent of Sierra’s Mystery House (1980). Many images could be integrated in these early graphical adventure games since visual elements were drawn through program instructions (and not stored as bitmaps). On top of different locations that illustrate player navigation (with synthetic visual feedback; players jump through space), the game features close-ups (for instance, a candelabra and character portraits). These apprehension mechanics provide vital information in order to complete the main quest: piece together a formula that will be able to get rid of Count Dracula. Apprehension makes up most of the resolution challenge since looking in some places (at the mirror, in the wrong drawer, etc.) will trigger an unescapable attack from the vampire. Players need to reconstruct an ideal sequence of action by monitoring where these static dead ends occur in order to win; the game manages to evade frustration through quick replays that avoid repeating long segments. HACS also integrate concepts that aim to account for ludic conventions and signifiers such as “score” or “failure marker.” Le Vampire Fou is quite original in that it associates such a failure marker to the figure of apprehension, an odd occurrence even in early Sierra games (in which players could be punished for exerting just about any other actions).

The will to remediate comic book aesthetics is one of the most common elements associated with the movement; thanks to Amiga and ST computers’ increased storage abilities, many corporations saw the opportunity to license and adapt popular French and Belgian comics. Infogrames was clearly the leader on this front, adapting, among others, the Tintin, Smurfs, Iznogoud, and Asterix franchises into platformers. However, the most interesting remediation attempts were inspired by the tabularity of the comic form (the disposition of many boxes on a page) in order to develop novel forms of video game agency. Infogrames adapted François Bourgeon’s Les passagers du vent—a troubling narrative dealing with triangular trade in the 19th century—in 1986 on the Amstrad CPC (1987 for the Amiga and ST versions). The design is streamlined in order to mimic the progression of the original narrative extensively. Users mostly perform the figure of social interaction through a minimal screen-augmented interface; buttons are hidden near character close-ups and “phylactères” (the French word referring to text bubbles). Progression in the game can be summed up by constructing dialogues, which appears as a minimal transposition of the comic narrative form. While other figures—such as navigation and environmental activation—can be evoked in the course of conversation choices, this limited mechanical realm means that the resolution challenge is restricted to a “trial and error” mode of engagement; users have to reconstruct the correct course of events, but their action hypothesis is far removed from the complexity of puzzle resolution seen in adventure games. It comes close to mere “actualization,” the simplest mode of performance integrated in HACS. Interestingly, contemporary listings used for this project often integrate a few recent visual novels, a long-standing but neglected genre that also relies on actualization, and popular walking simulators such as Gone Home (Fullbright, 2013) and Firewatch (Campo Santo, 2016).

It is important to note that comic book remediation also occurred outside of France, for instance with Master Designer/Cinemaware’s King of Chicago (1987) and Mike Singletown’s Grimblood (1990). The most elaborated attempt, however, is still French based. Ubisoft’s B.A.T. (1989) forces players to navigate diegetic space through comic boxes; virtual navigation is sometimes performed synchronically through synthetic feedback (players “jump through” space in real time from box to box), and sometimes players have the option to make the space of a box “scroll” through synchronic mapping/full visual feedback. To our knowledge, these forms of spatial remediation have only appeared outside of France in Sega’s 1995 beat’em up game Comix Zone. The complexity of the social interaction figure in B.A.T. is also noteworthy; through a screen-augmented interface, topics accumulate and greatly complicate the resolution of various scenes.

In terms of social interaction, one of the most well-known experimentations comes from Ère informatique’s Captain Blood (1988). Michel Rho’s graphical rendition of whimsical space creatures, which was also clearly inspired by the encompassing comics landscape, is responsible in part for the fascination emerging from this surreal science-fiction exploration game. Much like B.A.T., a strong emphasis is put in Blood on intricate conversation mechanics; players have to learn how to use dozens of different symbols on the screen-augmented interfaces in order to communicate with aliens properly. We can point out the game’s uniqueness through HACS: Resolution mode for this figure is exacerbated, thanks to the proliferation of static options and through a novel form of verbal transcoding/cumulative mapping (many symbols must be aligned to form sentences). Our data set reveals the relative prevalence of social interaction in the French Touch to some extent: 94% of adventure games produced in France incorporate the figure (vs. 75% in the United States).

Captain Blood is noteworthy for its diegetic integration of the screen-augmented interface; through the mouse, players appear to control the titular character’s hand. But in order to insist on this metonymic—almost symbiotic—form of mapping, game designers have relied on another prop: Every interactive figure—from spatial navigation to social interaction with the whimsical creatures—relies on a diegetic technological device (the captain’s ship; the alien transcoder) that is presented in a way that appears to echo the position of users. In HACS parlance, we should say that Captain Blood relies on a “screen-augmented + techno-mimetic” interface; the mimetism doesn’t come from the manipulation interface in itself but rather from a virtual prop introduced in the diegesis.

Blood’s designer Philip Ulrich formed the famous studio Cryo in 1990. It went on to create some of the most original and popular games in France; its success meant that the studio could invest a lot of money to concretize novel ideas. Their adaptation of Dune (1992), for instance, is a striking example of genre hybridity, incorporating elements from adventure and strategy games. The CD version (released in 1993) highlights the attractive nature that long animated sequences had acquired at the time. Their integration in this game as analyzed through HACS appears paradoxical: While players—taking up the role of Paul Atreides—can navigate the desert planet by piloting an ornithopter through its techno-mimetic screen-augmented interface (much like the one seen in Captain Blood), pointing their destination on the vehicle’s fictional GPS, this selection leads to a lengthy animated sequence prerendered through computer-generated imagery. The mapping is almost symbiotic for an instant but dissolves as soon as the animation begins; it corresponds to a “metonymic + punctual” configuration. In Lost Eden (1995) and Versailles 1685 (1996), Cryo settled for a punctual navigation that appears to mimic Trilobye’s The 7th Guest (1993). In China: The Forbidden City (1998), Cryo adopted a navigational model—the elliptical synchronic design that we have discussed above—that could be found already in Deja Vu (1985) and became widespread following the popularity of Myst (Cyan Worlds) in 1993. With these educative collaborations, Cryo lost part of its aura as the enfant terrible of the French industry. HACS proves to be a useful tool to expose the mimicry of established gameplay configurations and help us map out this stagnation process.

Nowadays, Quantic Dream is one of the main studios that could be associated with the French Touch. Genre hybridity has been a central design element in David Cage’s team since its inception, quite ostensibly so in early titles such as Omikron (1999). Interactive configurations developed for Heavy Rain (2010) represents a challenge for game analysis, mixing gameplay configurations and interface design from many different genres; it is an ideal case to test HACS’s ability to describe a wide spectrum of games and to sum up our findings. Through the PlayStation 3’s generic controller, navigation is handled in a synchronic manner (as in Adventure [Atari, 1980] and Pitfall! [Activision, 1982] on the Atari VCS), and other figures are implemented much like classical graphical adventure games outside of action segments. For instance, social interaction still relies on predetermined virtual phrases mapped in a punctual manner to user selections. These manipulations play an integral part in the encompassing resolution challenge; the main objective is to find out clues about the origami killer in order to stop him. The game’s originality comes from the integration of action sequences in a much greater proportion than in the typical graphical adventure game. These “quick time events” execution challenges are integrated mostly for the figures of neutralization and self-preservation, building on an intricate interplay of interface design. On the screen, buttons from the PS3 controller appear in order to incite the corresponding button press; actions are mapped in a punctual manner. These visual cues may also refer to the controller’s ability to capture hand gestures to some extent (a feature that becomes much more prominent in the Move edition of the game). This is what we have called miniature gestures or metonymic mapping. Feedback occurs in a signaletic form on top of the diegesis: Players can see the variations of visual elements on the screen augmented interface while their player character perform static actions in the background; ultimately, the player’s success (or lack thereof) will make the diegesis switch from one pre-scripted variation to another. Even in the course of these metonymic gestures mapped to signaletic feedback, the static diegetic fragments will inevitably take over and play independently of user manipulation. Using HACS’s typology of action mapping, this would be described as “metonymic + cumulative + punctual.” Our infrastructure reveals that only one previous game in our historical canon used such a configuration before: The Dark Eye (Inscape, 1995), a visual novel-like exploration of Edgar Allan Poe stories.

Data-Assisted Discussion

In this article, we have presented methodological underpinnings that allowed us to perform a distant reading of video game history through the case study of the adventure genre. Thanks to a data set that was assembled from the collective wisdom of many discourse communities, some of the most prevalent gameplay configurations in the history of the genre have been revealed. While interactive designs using the most symbolic and abstract elements in computer interface designs—on-screen pointers and buttons, verbal retroaction, visual forms of elliptical feedback, and convoluted cumulative or punctual action mappings—will remain associated with adventure games in popular imagination, we have seen that some figures like navigation and resource management could be integrated in synchronic and even metonymic ways. At the same time, HACS has proven useful to provide close readings of specific games, situate these titles within a network of similar configurations, and redirect our attention to lesser known titles. In order to develop the mobility of the historian’s gaze, more data sets would need to be encoded in our database. This would also open up the possibility of pursuing comparative studies similar to the one we have performed in this article with the French Touch.

While HACS encoding is able to account for the Touch’s originality, some elements can only be analyzed through a close encounter with these peculiar cultural objects. The most obvious limitation in the context of this comparative study comes from the system’s inability to acknowledge thematic complexity; our analysis in the previous section already hinted that this complexity defined the Touch to a great extent. Les passagers du vent deals with a difficult subject matter (colonialist exactions) that was also a core element in Muriel Tramis’s award winning Méwilo (1987). Much like its Sierra inspiration, Le Vampire Fou draws on a classic generic repertoire (detective drama/horror), but it also incorporates many puns and makes fun of its own silliness and poor graphic quality; this mix of tones and autoreferential nature was quite original at the time. While interactive figures, much like themes, are based on semantic categories, they were built from semantic pairings that sought to achieve a great level of generality in order to facilitate encoding. By contrast, thematic repertoires can integrate concepts that are more or less general. It is impossible with the current system to highlight the fact that the figure of social interaction in Paranoiak (Froggy, 1984) dovetails with psychotherapy, for instance. Another obvious HACS blind spot comes from its inability to acknowledge the integration of submecanics in any given interactive figure; spatial navigation or activation might encompass a full range of specific subactions that really alter the nature of the ludic experience. Such fuzziness is almost inevitable if we set out to observe from a more distant point of view.

Following this inspection, we have decided to integrate any generic marker seen on the MobyGames page—which includes some purely thematic dimensions—and on the original paratext (such as game packages) for each of the titles integrated in our canon. Further research could try to compare both data sets to find out which HACS configuration corresponds with typical generic markers. In order to build a comprehensive history of video game genres, our data set would need to integrate games that better reflect ongoing hybridization. The documents we used to build our canon inevitably aim toward “pure” examples of the genre; hybrid experiments are typically relegated to the outskirts of elevating listicles. While many other aspects of the methodology could be improved depending on research objectives, the results presented in this article highlight the interest of building systematic ways of encoding video game culture. Through more expansive data sets and visualization tools, we can envision the development of a wandering viewpoint that could assist video game historians in understanding and accounting for the intricacies of this culture.