interactivecinema.org

Modern cinematic language has many beginnings that range from the development of the apparatus to the arrangement of shots to produce dynamic meaning. Whichever inception one chooses to look at, the bottom line is that the language itself emerged from a creative use of several disciplines such as literature and photography. One capability of this hybrid product that is cinema--its ease in capturing a slice of life in motion- is often mistaken for the goal, which begs the question: what were the collective efforts of such diverse disciplines aiming towards? The goal suggested by André Bazin is the "Myth of Total Cinema."

"Total Cinema" is the platonic ideal that Bazin argued has informed the birth and development of cinema; it is an ideal founded upon the desire for "a recreation of the world in its own image." (Bazin 1967, 21). Photography was a step in this direction but even such a perfect reproduction of reality lacked a defining quality of life, the quality that could only be achieved by the technique of rapidly sequencing the pictures together and setting them in apparent motion.

Bazin had likened Total Cinema to the flight of Icarus. Both represent ideals that existed long before the means to accomplish them. "Thus, the myth of Icarus had to wait on the internal combustion engine before descending from the platonic heavens." (Bazin 1967, 22). Similarly, a world reproduced with verisimilitude had long been an ideal-- literature, painting and photography have all approached this ideal but there are dimensions of reality which they cannot reproduce. Because it added the vital
component of motion in the passage of time, cinema was a major advance. For Bazin, it was "the myth of Total Cinema"--the belief that a seamless representation of reality was possible--that spurned technological inventions which gave birth to cinema; thus he proclaimed that "every new development added to the cinema must, paradoxically, take it nearer and nearer to its origins. In short, cinema has not yet been invented!" (Bazin 1967, 21).

Computing technology exploded at the end of the twentieth century. For example, processor speeds increased dramatically, allowing the complexity of applications to grow unabated. With the technological limits pushed back, the creative aspirations of designers have been freed. As designers use their tools in pioneering ways, Bazin's concept of Total Cinema seems to be increasingly relevant in the many growing forms of media. However, while many look to the future for the beginnings of Interactive Cinema, I suggest that we look, instead, to the past because, in its simplest form, Interactive Cinema has existed in popular culture since 1980 even though no one has properly examined the medium. Because there has not been an adequate description nor a proper definition of the medium, it is my aim to not only establish a working definition of "Interactive Cinema" but also provide the groundwork with which one can begin to properly analyze it as the valid medium that it is and has been for over twenty years.

Interactive Cinema

I think that the common understanding of these two words put together--"interactive" and "cinema"--is an approximation of Total Cinema; that is to say, we have in "interactive cinema" a narrative environment in which the viewer-player is positioned at the control center. Thus, the artist-creator of the narrative environment constructs a defined world rather than a defined work since this world contains many possible works and it is from the viewer-player that a singular work emerges. In other words, the total reproduction of reality--Bazin's "Total Cinema"--necessarily includes the subjectivity and activity of the viewer-player.

When the player becomes the main character and is immersed in a total world, everything happens from his or her eyes. Therefore the player becomes united with the camera for a continuous period of time. In fact, in having complete control of one's self, one also simultaneously has complete control of the camera (where it goes, from where it looks, etc.). Not only has the artist lost control of the apparatus, the apparatus has been subsumed into the player! With this new construction, the traditional relationships in cinema break down. To add further paradox to Bazin's reverse development of Total Cinema, this means that when Bazin's "Total Cinema" is finally achieved, not only may the apparatus of modern cinema have vanished, but its linguistic aesthetics may disappear as well. Current computing technology has already begun to push beyond the cinematic apparatus in an effort to reach this mythical ideal.

Although certain aspects of the "recreated world" must still be compromised the result is an experience that deviates from traditional cinema but is not able to abandon it entirely. While traditional cinema relies on a montage structure of a single medium, this new experience exists as a trans-media montage, in which the segments are quite different from one another. This difference forces a change in the relationship between the audience and the text so that a cycle is established in which the audience moves back-and-forth between the roles of viewer (who watches) and player (who commands). To find instances of this, one need look no further than the world of videogames that is growing larger all the time. While distinctions have been made between the "on-line" and "off-line" portions of these experiences-- which have even been identified as "interactive" and "cinematic"-- the relationship between the two has never been properly analyzed (Newman, 2002). However, I will show that it makes sense to understand the (a) on-line and (b) off-line segments as (a) the Interactive and (b) the Cinematic contributions to Interactive Cinema, which I define as the following:

An experience involving one or more alternations between juxtaposed

(a) non-trivial interactive sequences, and

(b) cinematic sequences.

It is a simple definition that accurately describes the components required for something to constitute Interactive Cinema. The challenge is that, before we can be comfortable with this definition, the terms "interactive" and "cinematic" must be clearly delineated.

The word "Interactive" falls victim to many false assumptions surrounding the dominant assumption that the ultimate goal of interactivity is the ability to navigate a narrative world at the player's own discretion. This has led to a number of theorists aiming for a specific concept of interactive best summarized in the study of ergodics.

Ergodicity, as proposed by Espen Aarseth, is the quality of a text that determines the amount of effort the reader must invest in order to traverse the text. The main question is: does the text exist apart from the reader or does it only truly exist when the reader makes certain choices that determine the way in which the text is sequenced (Aarseth, 1)? The goal of successful ergodicity is to create within a defined world probable works that are realized at the discretion of the user. The immediate correlation between ergodics and the misconception of Interactive Cinema is quite apparent, but to best understand how ergodics actually relate to Interactive Cinema, we need look no further than the realm of what I call "qualified interactivities," or interactivities with different features. For example, "branching-type interactivity" (figure 1.1) is a physical map of an interactive experience. The model exists as a web structure composed of many nodes. The user navigates by progressing from node to node and selecting a path at each node (provided a choice is presented). This is very much an ergodic experience, but what does it reveal about interactivity? The "branching" nature of the experience is made clear by the branching paths, but there is not any exploration into what there is about it that makes it interactive. As such, "branching-type interactivity" and indeed all qualified interactivities say nothing of the underlying process that takes place during such an experience. Unfortunately, because such terms describe instances of interactivity so well they have been accepted as descriptions of interactivity itself, leaving the actual interactive process undefined.

When the most acknowledged efforts to define interactivity do not even come close to defining it, the term is vague at best. This confusion devalues the term as an analytical tool. Aarseth himself proposed the term "ergodic" because of what he perceives to be a "dubiousness of the term." (Aarseth, 1) Following this lead, other theorists have made bold statements claiming that its proliferation of use "has rendered [the term interactivity] meaningless and of use only to the marketer." (Newman 2002) The result of such accusations is an academic shunning of the word interactive. Lev Manovich states it quite simply. "I find the concept too broad to be truly useful." (Manovich, 55)

I agree that the marketing departments have ravaged the term. I agree that qualified interactivities present good descriptions of instances of interactivity. I even agree that the term as it is currently used is far too general. I do not agree that these are sufficient reasons for dismissing the term entirely. "Interactive" can be a valuable tool if properly wielded, and to abandon the term because of those that lack the skill to accurately use it is unacceptable. The key to overcoming this classical conception of interactivity exists in a redefinition of the term.

There have been attempts at formal definitions of interactivity, but those definitions usually worked in theoretical instances alone, making any practical applications impossible. Aarseth criticizes one such definition proposed by Peter Bogh Andersen in which Andersen presents a semiotic definition of interactivity. While it can apply to an interactive process, Aarseth points out that it eliminates experiences such as hypertexts because it does not "describe discursive systems where the user's activity is limited to metasemiotic exploration" (Aarseth, 49).

The inadequacies that inform such definitions are usually the result of mistaking the potential of interactivity with the actual process of interactivity. Of course a definition that focuses only on the actual process would also result in problems. A proper definition of interactivity exists somewhere between the process and the potential, while being exclusive to neither.

The definition must accurately describe the interactive process with specific details while at the same time remaining general enough so as not to exclude the vast number of forms and structures that might facilitate interaction. Chris Crawford, a veteran game designer, has managed to do this. By observing the interplay between user and system, Crawford has proposed the following definition of interactivity:

a cyclic process in which two actors alternately listen, think, and speak

(Crawford 2000, 6)

The fact that Crawford has modeled his definition on the interplay of a conversation is quite clear. While his vocabulary can be interpreted in a number of ways, when the two actors are a person and a computer, the terms "listen, think, speak" become synonymous with "input, process, output." Crawford goes to great lengths to show that each step of the cycle is necessary to the interactive process but that the three steps do not need to be equal. Variations in the balance between the three not only produce different experiences, but they also revolutionize the classic concept of interactivity.

Previous conceptions of interactivity have always viewed it as a binary constant: an object was perceived as either interactive or non-interactive. When interactivity began to enter into popular culture, it was easy to accept this duality. Clearly the videogame PONG was interactive while a rubber ball was not. Sure, the ball would react to the forces we applied to it, but PONG would evaluate and respond to our input. Thirty years of continuous technological development since the birth of PONG have added complexity to the previously simple distinction between "interactive" and "non-interactive," creating a new depth.

To begin with, the interactive side of the scale is no longer constant. As Crawford points out, an interactive product that listens and speaks well but thinks poorly has an entirely different level of interactivity from a product that thinks and listens well but speaks poorly. These levels form the scale of interactivity in the redefinition of the interactive attribute.

Figure 1.2 illustrates the classic conception of the interactive attribute. It is an either/or construct with no room for variation. The redefinition of the interactive attribute, figure 1.3, is different. Objects still exist within the either/or framework of the classic model, but now there is a further distinction that needs to be made in how well an object performs each of the steps in the interactive process with another actor. The obvious concern that will arise over this is how to rate the overall interactivity of a project on this new scale.

At first glance, it may seem that the new conception of interactivity would simply replace the old ambiguity with a scale that appears to be overly subjective and equally nebulous. It certainly is the case that there is no objective scale that one can use to determine the level of interactivity, but is there a way to determine objectively how "cinematic" a film is or how literary a book is? No, but each medium does have complex linguistic structures that allow a differentiation of quality within them. The three steps of the interactive process are the foundation of the corresponding linguistics for interactivity. Though there is not yet a true codex on how these steps will be gauged, it is helpful to examine how they operate in regards to the traditional conceptions of interactivity.

It is believed that "there is the danger that we will interpret 'interaction' literally, equating it with physical interaction between a user and a media object... at the expense of psychological interaction." (Manovich, 57). That would mean that the physical process of interaction is completely divorced from the mental process of interaction. This concern seems to be informed by the classic concept of interactivity: the either/or mentality considers user input the only requirement for interactivity. "User input" is any information transmitted from the user to the other object; this could be as simple as striking a random key on a keyboard. This act of classical interaction does not require a connection between physical interaction and a higher mental process; all that is required is a single instance of random input and the binary standard is achieved. Under the new concept of interactivity, however, such a "single-click experience" would be interactive, but at varying levels depending on the context.

As Crawford himself explains, even a refrigerator is, according to his definition, interactive as it responds to the user input of opening the door with a light turning on (Crawford 2000, 7). What is missing, however, is any significant effort on the part of either actor when it comes to the thinking step. The user pulls a handle and opens a door. The machine responds to either open or closed states and acts accordingly by turning a light on or off. As a procedural step, thought exists; yet with only two possible evaluations, the thought process functions at only a minimal level; the refrigerator offers no real thought for either actor. Imagine, instead, a more thought-intensive process, such as a web narrative.

The collection of hypertexts that have emerged is a perfect example of emphasis being placed on thought. As a physical process of interaction, hypertexts are best described as a series of single-click experiences. Accordingly, in order to select the path along which the player/viewer wants to proceed, a single click of the mouse is all that is required from the user. This simple action certainly requires less physical effort than opening a refrigerator door. However, hypertexts have slightly more complex output than a refrigerator. A refrigerator simply turns a light on, but while hypertexts are also composed of light displays, the light of a hypertext is highly organized and imbued with meaning-- that is, the light of a hypertext represents words and, in turn, thoughts. Instead of choosing either "on" or "off," as the refrigerator does, a hypertext program must find a single file from (in the case of the World Wide Web) millions of possible files. This "thought" exists in machine language-- the link is an address stored in the computer -- but both actors do not need to speak the same language. This may seem obvious since we speak in words and images while computers speak in bits and bytes, but it raises the question of artificial intelligence with regards to interactivity.

Artificially intelligent creations may indeed be integral aspects to the development of interactive experiences, but they are not part of the interactive process. Aarseth claims that interaction requires a "functional equality," which implies the machine must therefore be intelligent, but Crawford's definition overcomes this problem (Aarseth, 49-50). The key is that the steps of interactivity for the two agents are not functionally equal: one actor speaks via light and sound while the other speaks via letters and a mouse. Do we need to know how the computer processes information in order to browse the Internet? No, because the speaking and listening between two heterogeneous actors becomes possible because of the interface. One actor may think about data and the other may think about ideas. This reveals itself most fully in a more thorough examination of a typical hypertext such as Judy Malloy's The Roar of Destiny Emanated from the Refrigerator (Malloy).

When first "opening" the text, the computer speaks to the user via a number of words on a screen positioned to produce a certain effect and to present a range of choices to the reader. This effect will clearly change from viewer to viewer. On the first page, the only selection for user input is a "RESET" link. This first page is indicative of a low-level interaction. Like the refrigerator door, the user can only "speak" with a binary voice (either to, or not to, reset). Once this initial speech has been made, the level of interactivity begins to increase in complexity.

The next page is filled with words, each with its own unique meanings. This time single words and sometimes groups of words are active links. The thought process of the user-- which was limited in speech previously-- has been given more freedom and the thoughts of the user will dictate more accurately what will be "spoken" (what selection will be made). Importantly, even when the range and depth of speech is limited, interaction still occurs. This is a perfect example of thought and physical interaction working together.

With Crawford's definition, evaluation of the other's speech is mandatory for both actors. Without the thought process, the interactive loop would be interrupted and the result is that there is no interactivity. This holds true for the absence of each of the three steps of the interactive process. This means that the "psychological interaction" that was feared to be absent is not only present but is essential in Crawford's definition of interactivity. With this understanding, let us examine more closely the cinematic medium.

Cinema is based on the concept of motion. This motion is created through the animation of sequential still photographs at the standard rate of twenty-four still images per second. Yet, this motion is very much an illusion, permitted by a persistence of vision which lets the eye see images in rapid succession and perceive them as a unified record of motion. While this trick of the eye was sufficient to entertain audiences in the early years of cinema, it soon became apparent that the objective recording of events limited what could be done with the medium. This led to the exploration of shot arrangements not dictated by the linearity of real time; concern was focused, instead, on the meaning created by the juxtaposition of non-linear shots in the cinematic montage.

In this way, far from producing a seamless continuity that effaces the work of the editor, montage editing provokes reaction precisely in the seams of the juxtaposed shots. As Manovich states, "All classical, and even more so modern, art is 'interactive' in a number of ways. Ellipses in literary narration, missing details of objects in visual art, and other representational 'shortcuts' require the user to fill in the missing information." (Manovich, 56)

Does the filling in of missing information constitute all three steps in the interactive process? Crawford engages this question in a number of ways and manages to sum up the argument perfectly when he differentiates between reaction and interaction. Reactive media, such as cinema, are deaf to the viewers. Since a film cannot accept input, it has nothing to think about, and without all three steps of the interactive process being performed by both actors, there is no interaction. "Movies, like books, confine themselves to speaking to their audiences, and they do that very, very well. Let's appreciate them for what they are good at instead of press-ganging them into something they're terrible at." (Crawford 2002, 10). To say that they are terrible at anything other than speech is an overstatement, since it is the only step in the interactive process that reactive media are capable of. Yet, thanks to the power of montage, cinema is capable of incredibly persuasive speech and so, while cinematic montage may be excluded from discussions of interactivity, this exclusion does not lessen its relevance to Interactive Cinema.

In his prolific writing on the topic, Sergei Eisenstein characterized different types of montage, ranging from being guided by the movement within shots to the intellectual associations the viewer has with each shot. What is important in each case is the conceptual foundation of cinematic montage.

Simply put, montage is a synergistic process. "An episode is not simply formed as the sum of its montage sequences but emerges in the process of the interaction between sequences as the generalizing image of them." (Glenny, 200). Despite Eisenstein's critique of Lev Kuleshov's analysis of montage as a "linkage" rather than a "collision," Lev Kuleshov's famous montage experiment still proves useful (Eisenstein, 36-38). Kuleshov screened three montage sequences. An object and then a close-up of an actor. The object changed each time, but the shot of the actor remained identical. The experiment showed that despite the reaction of the actor remaining completely unchanged, the audience perceived his reaction quite differently each time. "The public raved about the acting of the artist. They pointed out the heavy pensiveness of his mood over the forgotten soup, were touched and moved by the deep sorrow with which he looked on the dead woman, and admired the light happy smile with which he surveyed the girl at play. But we knew that in all three cases the face was exactly the same." (Cook, 137). This experiment led Kuleshov to theorize the duplicity of the shot and to conclude that "the shot has two distinct values: (1) that which it possesses in itself as a photographic image of reality and (2) that which it acquires when placed in relationship to other shots." (Cook, 137).

The cinematic theories of montage certainly indicate the importance of montage editing (in which separate elements are juxtaposed between separate shots) and montage within a shot (in which separate elements are juxtaposed within a single frame) imply that the essence of montage is the invisible product that results from a juxtaposition of two media objects. In other words, a wholly new experience is composed from its component parts.

Mixed-media montage is what lies at the center of Interactive Cinema. Let us return again to our definition of Interactive Cinema:

An experience involving one or more alternations between juxtaposed

(a) non-trivial interactive sequences, and

(b) cinematic sequences.

Despite the differing vocabulary, the presence of a montage construction in Interactive Cinema is unmistakable. The best way to show this is by modifying the concept of "branching-type interactivity."

In its original version, the entire concept represents interactivity; that is to say, each node and each connection combines to form an instance of interactivity, but there is no clearly predefined meaning of what each portion of the diagram represents. It is the assumption that the diagram as a whole represents "interactivity" that has caused the turmoil regarding the term "interactive." It is common for studies to assume such a diagrammatic representation of a specific instance of interactivity represents interactivity as a concept. Fracturing of the diagram and analyzing the component parts of the structure will give us a clear sense of how montage works in Interactive Cinema. Instead of meaningless nodes and connections, one has a diagram composed of interactive segments (nodes) and cinematic segments (connections). The nodes function according to interactivity as defined by Crawford, and the connections work according the established language of cinema.

In its simplest form, Interactive Cinema is best represented as a single node and a single connection; together they constitute a single alternation (figure 1.4). In other words, it is the juxtaposition of an interactive segment and a cinematic segment that forms the basic montage of Interactive Cinema. While each interactive node could have any number of connecting cinematic paths to follow (and vice versa), this single alternation is the basic component-- the atom-- of Interactive Cinema. From this construction, all types of Interactive Cinema are formed.

Traditional studies of videogames have looked at them as texts involving a single medium of interactivity rather than a composite of interactivity and cinema. Thirty years ago, this may have been a valid approach to study the medium because all that existed was PONG and its many clones, but interactivity took multiple paths and grew more diverse with each technological advancement. The terrain of interactive entertainment today is not composed of homogenous, uniform entities, but much discourse on the medium remains single-visioned and limited to the classic binary that focuses only on the either/or input from the user. This limitation in thinking about interactivity undermines the role of the cinematic sequence in interactive entertainment but it is pervasive in the thinking in both popular and academic cultures. Perhaps because of the incredible ease with which Interactive Cinema is now able to transition between interactive and cinematic segments, this montage between two media--which is precisely the structure that produces Interactive Cinema--has been overlooked in both popular and academic studies. What has been the cost of this oversight? To answer that question, let's take a look at the genesis of Interactive Cinema.